diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000000000000000000000000000000000000..b073a2ccc58540b26e260de2d727a797bb3030d4 --- /dev/null +++ b/.gitignore @@ -0,0 +1 @@ +**_pycache** \ No newline at end of file diff --git a/app.py b/app.py new file mode 100644 index 0000000000000000000000000000000000000000..936053250ebcea9a74377542f4ac44b89bb12a03 --- /dev/null +++ b/app.py @@ -0,0 +1,766 @@ +import os +import signal +import time +import csv +import sys +import warnings +import random +import gradio as gr +import torch +import torch.nn as nn +import torch.nn.functional as F +import torch.distributed as dist +from torch.nn.parallel import DistributedDataParallel as DDP +import torch.multiprocessing as mp +import numpy as np +import time +import pprint +from loguru import logger +import smplx +from torch.utils.tensorboard import SummaryWriter +import wandb +import matplotlib.pyplot as plt +from utils import config, logger_tools, other_tools_hf, metric, data_transfer, other_tools +from dataloaders import data_tools +from dataloaders.build_vocab import Vocab +from optimizers.optim_factory import create_optimizer +from optimizers.scheduler_factory import create_scheduler +from optimizers.loss_factory import get_loss_func +from dataloaders.data_tools import joints_list +from utils import rotation_conversions as rc +import soundfile as sf +import librosa +import subprocess +from transformers import pipeline +from diffusion.model_util import create_gaussian_diffusion +from diffusion.resample import create_named_schedule_sampler +from models.vq.model import RVQVAE +import train +import spaces + +command = ["bash","./demo/install_mfs.sh"] +result = subprocess.run(command, capture_output=True, text=True) + +device = "cuda" if torch.cuda.is_available() else "cpu" + +pipe = pipeline( + "automatic-speech-recognition", + model="openai/whisper-tiny.en", + chunk_length_s=30, + device=device, +) + +debug = False + +class BaseTrainer(object): + def __init__(self, args,ap): + args.use_ddim=True + hf_dir = "hf" + time_local = time.localtime() + time_name_expend = "%02d%02d_%02d%02d%02d_"%(time_local[1], time_local[2],time_local[3], time_local[4], time_local[5]) + self.time_name_expend = time_name_expend + tmp_dir = args.out_path + "custom/"+ time_name_expend + hf_dir + if not os.path.exists(tmp_dir + "/"): + os.makedirs(tmp_dir + "/") + self.audio_path = tmp_dir + "/tmp.wav" + sf.write(self.audio_path, ap[1], ap[0]) + + + audio, ssr = librosa.load(self.audio_path,sr=args.audio_sr) + + # use asr model to get corresponding text transcripts + file_path = tmp_dir+"/tmp.lab" + self.textgrid_path = tmp_dir + "/tmp.TextGrid" + if not debug: + text = pipe(audio, batch_size=8)["text"] + with open(file_path, "w", encoding="utf-8") as file: + file.write(text) + + # use montreal forced aligner to get textgrid + + command = ["mfa", "align", tmp_dir, "english_us_arpa", "english_us_arpa", tmp_dir] + result = subprocess.run(command, capture_output=True, text=True) + + + ap = (ssr, audio) + self.args = args + self.rank = 0 # dist.get_rank() + + args.textgrid_file_path = self.textgrid_path + args.audio_file_path = self.audio_path + + + self.rank = 0 # dist.get_rank() + + self.checkpoint_path = tmp_dir + args.tmp_dir = tmp_dir + if self.rank == 0: + self.test_data = __import__(f"dataloaders.{args.dataset}", fromlist=["something"]).CustomDataset(args, "test") + self.test_loader = torch.utils.data.DataLoader( + self.test_data, + batch_size=1, + shuffle=False, + num_workers=args.loader_workers, + drop_last=False, + ) + logger.info(f"Init test dataloader success") + model_module = __import__(f"models.{args.model}", fromlist=["something"]) + + self.model = torch.nn.DataParallel(getattr(model_module, args.g_name)(args), args.gpus).cuda() + + if self.rank == 0: + logger.info(self.model) + logger.info(f"init {args.g_name} success") + + self.smplx = smplx.create( + self.args.data_path_1+"smplx_models/", + model_type='smplx', + gender='NEUTRAL_2020', + use_face_contour=False, + num_betas=300, + num_expression_coeffs=100, + ext='npz', + use_pca=False, + ).to(self.rank).eval() + + + + + + self.args = args + self.ori_joint_list = joints_list[self.args.ori_joints] + self.tar_joint_list_face = joints_list["beat_smplx_face"] + self.tar_joint_list_upper = joints_list["beat_smplx_upper"] + self.tar_joint_list_hands = joints_list["beat_smplx_hands"] + self.tar_joint_list_lower = joints_list["beat_smplx_lower"] + + self.joint_mask_face = np.zeros(len(list(self.ori_joint_list.keys()))*3) + self.joints = 55 + for joint_name in self.tar_joint_list_face: + self.joint_mask_face[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + self.joint_mask_upper = np.zeros(len(list(self.ori_joint_list.keys()))*3) + for joint_name in self.tar_joint_list_upper: + self.joint_mask_upper[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + self.joint_mask_hands = np.zeros(len(list(self.ori_joint_list.keys()))*3) + for joint_name in self.tar_joint_list_hands: + self.joint_mask_hands[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + self.joint_mask_lower = np.zeros(len(list(self.ori_joint_list.keys()))*3) + for joint_name in self.tar_joint_list_lower: + self.joint_mask_lower[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + + self.tracker = other_tools.EpochTracker(["fid", "l1div", "bc", "rec", "trans", "vel", "transv", 'dis', 'gen', 'acc', 'transa', 'exp', 'lvd', 'mse', "cls", "rec_face", "latent", "cls_full", "cls_self", "cls_word", "latent_word","latent_self","predict_x0_loss"], [False,True,True, False, False, False, False, False, False, False, False, False, False, False, False, False, False,False, False, False,False,False,False]) + + vq_model_module = __import__(f"models.motion_representation", fromlist=["something"]) + self.args.vae_layer = 2 + self.args.vae_length = 256 + self.args.vae_test_dim = 106 + self.vq_model_face = getattr(vq_model_module, "VQVAEConvZero")(self.args).to(self.rank) + other_tools.load_checkpoints(self.vq_model_face, "./datasets/hub/pretrained_vq/face_vertex_1layer_790.bin", args.e_name) + + + vq_type = self.args.vqvae_type + if vq_type=="vqvae": + + self.args.vae_layer = 4 + self.args.vae_test_dim = 78 + self.vq_model_upper = getattr(vq_model_module, "VQVAEConvZero")(self.args).to(self.rank) + other_tools.load_checkpoints(self.vq_model_upper, args.vqvae_upper_path, args.e_name) + self.args.vae_test_dim = 180 + self.vq_model_hands = getattr(vq_model_module, "VQVAEConvZero")(self.args).to(self.rank) + other_tools.load_checkpoints(self.vq_model_hands, args.vqvae_hands_path, args.e_name) + self.args.vae_test_dim = 54 + self.args.vae_layer = 4 + self.vq_model_lower = getattr(vq_model_module, "VQVAEConvZero")(self.args).to(self.rank) + other_tools.load_checkpoints(self.vq_model_lower, args.vqvae_lower_path, args.e_name) + + elif vq_type=="rvqvae": + + args.num_quantizers = 6 + args.shared_codebook = False + args.quantize_dropout_prob = 0.2 + args.mu = 0.99 + + args.nb_code = 512 + args.code_dim = 512 + args.code_dim = 512 + args.down_t = 2 + args.stride_t = 2 + args.width = 512 + args.depth = 3 + args.dilation_growth_rate = 3 + args.vq_act = "relu" + args.vq_norm = None + + dim_pose = 78 + args.body_part = "upper" + self.vq_model_upper = RVQVAE(args, + dim_pose, + args.nb_code, + args.code_dim, + args.code_dim, + args.down_t, + args.stride_t, + args.width, + args.depth, + args.dilation_growth_rate, + args.vq_act, + args.vq_norm) + + dim_pose = 180 + args.body_part = "hands" + self.vq_model_hands = RVQVAE(args, + dim_pose, + args.nb_code, + args.code_dim, + args.code_dim, + args.down_t, + args.stride_t, + args.width, + args.depth, + args.dilation_growth_rate, + args.vq_act, + args.vq_norm) + + dim_pose = 54 + if args.use_trans: + dim_pose = 57 + self.args.vqvae_lower_path = self.args.vqvae_lower_trans_path + args.body_part = "lower" + self.vq_model_lower = RVQVAE(args, + dim_pose, + args.nb_code, + args.code_dim, + args.code_dim, + args.down_t, + args.stride_t, + args.width, + args.depth, + args.dilation_growth_rate, + args.vq_act, + args.vq_norm) + + self.vq_model_upper.load_state_dict(torch.load(self.args.vqvae_upper_path)['net']) + self.vq_model_hands.load_state_dict(torch.load(self.args.vqvae_hands_path)['net']) + self.vq_model_lower.load_state_dict(torch.load(self.args.vqvae_lower_path)['net']) + + self.vqvae_latent_scale = self.args.vqvae_latent_scale + + self.vq_model_upper.eval().to(self.rank) + self.vq_model_hands.eval().to(self.rank) + self.vq_model_lower.eval().to(self.rank) + + + + + + self.args.vae_test_dim = 61 + self.args.vae_layer = 4 + self.args.vae_test_dim = 330 + self.args.vae_layer = 4 + self.args.vae_length = 240 + + + self.vq_model_face.eval() + self.vq_model_upper.eval() + self.vq_model_hands.eval() + self.vq_model_lower.eval() + + self.cls_loss = nn.NLLLoss().to(self.rank) + self.reclatent_loss = nn.MSELoss().to(self.rank) + self.vel_loss = torch.nn.L1Loss(reduction='mean').to(self.rank) + self.rec_loss = get_loss_func("GeodesicLoss").to(self.rank) + self.log_softmax = nn.LogSoftmax(dim=2).to(self.rank) + + self.diffusion = create_gaussian_diffusion(use_ddim=args.use_ddim) + self.schedule_sampler_type = 'uniform' + self.schedule_sampler = create_named_schedule_sampler(self.schedule_sampler_type, self.diffusion) + self.mean = np.load(args.mean_pose_path) + self.std = np.load(args.std_pose_path) + + self.use_trans = args.use_trans + if self.use_trans: + self.trans_mean = np.load(args.mean_trans_path) + self.trans_std = np.load(args.std_trans_path) + self.trans_mean = torch.from_numpy(self.trans_mean).cuda() + self.trans_std = torch.from_numpy(self.trans_std).cuda() + + + joints = [3,6,9,12,13,14,15,16,17,18,19,20,21] + upper_body_mask = [] + for i in joints: + upper_body_mask.extend([i*6, i*6+1, i*6+2, i*6+3, i*6+4, i*6+5]) + + joints = list(range(25,55)) + hands_body_mask = [] + for i in joints: + hands_body_mask.extend([i*6, i*6+1, i*6+2, i*6+3, i*6+4, i*6+5]) + + joints = [0,1,2,4,5,7,8,10,11] + lower_body_mask = [] + for i in joints: + lower_body_mask.extend([i*6, i*6+1, i*6+2, i*6+3, i*6+4, i*6+5]) + + self.mean_upper = self.mean[upper_body_mask] + self.mean_hands = self.mean[hands_body_mask] + self.mean_lower = self.mean[lower_body_mask] + self.std_upper = self.std[upper_body_mask] + self.std_hands = self.std[hands_body_mask] + self.std_lower = self.std[lower_body_mask] + + self.mean_upper = torch.from_numpy(self.mean_upper).cuda() + self.mean_hands = torch.from_numpy(self.mean_hands).cuda() + self.mean_lower = torch.from_numpy(self.mean_lower).cuda() + self.std_upper = torch.from_numpy(self.std_upper).cuda() + self.std_hands = torch.from_numpy(self.std_hands).cuda() + self.std_lower = torch.from_numpy(self.std_lower).cuda() + + + def inverse_selection(self, filtered_t, selection_array, n): + original_shape_t = np.zeros((n, selection_array.size)) + selected_indices = np.where(selection_array == 1)[0] + for i in range(n): + original_shape_t[i, selected_indices] = filtered_t[i] + return original_shape_t + + def inverse_selection_tensor(self, filtered_t, selection_array, n): + selection_array = torch.from_numpy(selection_array).cuda() + original_shape_t = torch.zeros((n, 165)).cuda() + selected_indices = torch.where(selection_array == 1)[0] + for i in range(n): + original_shape_t[i, selected_indices] = filtered_t[i] + return original_shape_t + + def _load_data(self, dict_data): + tar_pose_raw = dict_data["pose"] + tar_pose = tar_pose_raw[:, :, :165].to(self.rank) + tar_contact = tar_pose_raw[:, :, 165:169].to(self.rank) + tar_trans = dict_data["trans"].to(self.rank) + tar_trans_v = dict_data["trans_v"].to(self.rank) + tar_exps = dict_data["facial"].to(self.rank) + in_audio = dict_data["audio"].to(self.rank) + in_word = dict_data["word"].to(self.rank) + tar_beta = dict_data["beta"].to(self.rank) + tar_id = dict_data["id"].to(self.rank).long() + bs, n, j = tar_pose.shape[0], tar_pose.shape[1], self.joints + + tar_pose_jaw = tar_pose[:, :, 66:69] + tar_pose_jaw = rc.axis_angle_to_matrix(tar_pose_jaw.reshape(bs, n, 1, 3)) + tar_pose_jaw = rc.matrix_to_rotation_6d(tar_pose_jaw).reshape(bs, n, 1*6) + tar_pose_face = torch.cat([tar_pose_jaw, tar_exps], dim=2) + + tar_pose_hands = tar_pose[:, :, 25*3:55*3] + tar_pose_hands = rc.axis_angle_to_matrix(tar_pose_hands.reshape(bs, n, 30, 3)) + tar_pose_hands = rc.matrix_to_rotation_6d(tar_pose_hands).reshape(bs, n, 30*6) + + tar_pose_upper = tar_pose[:, :, self.joint_mask_upper.astype(bool)] + tar_pose_upper = rc.axis_angle_to_matrix(tar_pose_upper.reshape(bs, n, 13, 3)) + tar_pose_upper = rc.matrix_to_rotation_6d(tar_pose_upper).reshape(bs, n, 13*6) + + tar_pose_leg = tar_pose[:, :, self.joint_mask_lower.astype(bool)] + tar_pose_leg = rc.axis_angle_to_matrix(tar_pose_leg.reshape(bs, n, 9, 3)) + tar_pose_leg = rc.matrix_to_rotation_6d(tar_pose_leg).reshape(bs, n, 9*6) + + tar_pose_lower = tar_pose_leg + + + tar4dis = torch.cat([tar_pose_jaw, tar_pose_upper, tar_pose_hands, tar_pose_leg], dim=2) + + + if self.args.pose_norm: + tar_pose_upper = (tar_pose_upper - self.mean_upper) / self.std_upper + tar_pose_hands = (tar_pose_hands - self.mean_hands) / self.std_hands + tar_pose_lower = (tar_pose_lower - self.mean_lower) / self.std_lower + + if self.use_trans: + tar_trans_v = (tar_trans_v - self.trans_mean)/self.trans_std + tar_pose_lower = torch.cat([tar_pose_lower,tar_trans_v], dim=-1) + + latent_face_top = self.vq_model_face.map2latent(tar_pose_face) # bs*n/4 + latent_upper_top = self.vq_model_upper.map2latent(tar_pose_upper) + latent_hands_top = self.vq_model_hands.map2latent(tar_pose_hands) + latent_lower_top = self.vq_model_lower.map2latent(tar_pose_lower) + + latent_in = torch.cat([latent_upper_top, latent_hands_top, latent_lower_top], dim=2)/self.args.vqvae_latent_scale + + + tar_pose_6d = rc.axis_angle_to_matrix(tar_pose.reshape(bs, n, 55, 3)) + tar_pose_6d = rc.matrix_to_rotation_6d(tar_pose_6d).reshape(bs, n, 55*6) + latent_all = torch.cat([tar_pose_6d, tar_trans, tar_contact], dim=-1) + style_feature = None + if self.args.use_motionclip: + motionclip_feat = tar_pose_6d[...,:22*6] + batch = {} + bs,seq,feat = motionclip_feat.shape + batch['x']=motionclip_feat.permute(0,2,1).contiguous() + batch['y']=torch.zeros(bs).int().cuda() + batch['mask']=torch.ones([bs,seq]).bool().cuda() + style_feature = self.motionclip.encoder(batch)['mu'].detach().float() + + + + # print(tar_index_value_upper_top.shape, index_in.shape) + return { + "tar_pose_jaw": tar_pose_jaw, + "tar_pose_face": tar_pose_face, + "tar_pose_upper": tar_pose_upper, + "tar_pose_lower": tar_pose_lower, + "tar_pose_hands": tar_pose_hands, + 'tar_pose_leg': tar_pose_leg, + "in_audio": in_audio, + "in_word": in_word, + "tar_trans": tar_trans, + "tar_exps": tar_exps, + "tar_beta": tar_beta, + "tar_pose": tar_pose, + "tar4dis": tar4dis, + "latent_face_top": latent_face_top, + "latent_upper_top": latent_upper_top, + "latent_hands_top": latent_hands_top, + "latent_lower_top": latent_lower_top, + "latent_in": latent_in, + "tar_id": tar_id, + "latent_all": latent_all, + "tar_pose_6d": tar_pose_6d, + "tar_contact": tar_contact, + "style_feature":style_feature, + } + + def _g_test(self, loaded_data): + sample_fn = self.diffusion.p_sample_loop + if self.args.use_ddim: + sample_fn = self.diffusion.ddim_sample_loop + mode = 'test' + bs, n, j = loaded_data["tar_pose"].shape[0], loaded_data["tar_pose"].shape[1], self.joints + tar_pose = loaded_data["tar_pose"] + tar_beta = loaded_data["tar_beta"] + tar_exps = loaded_data["tar_exps"] + tar_contact = loaded_data["tar_contact"] + tar_trans = loaded_data["tar_trans"] + in_word = loaded_data["in_word"] + in_audio = loaded_data["in_audio"] + in_x0 = loaded_data['latent_in'] + in_seed = loaded_data['latent_in'] + + remain = n%8 + if remain != 0: + tar_pose = tar_pose[:, :-remain, :] + tar_beta = tar_beta[:, :-remain, :] + tar_trans = tar_trans[:, :-remain, :] + in_word = in_word[:, :-remain] + tar_exps = tar_exps[:, :-remain, :] + tar_contact = tar_contact[:, :-remain, :] + in_x0 = in_x0[:, :in_x0.shape[1]-(remain//self.args.vqvae_squeeze_scale), :] + in_seed = in_seed[:, :in_x0.shape[1]-(remain//self.args.vqvae_squeeze_scale), :] + n = n - remain + + tar_pose_jaw = tar_pose[:, :, 66:69] + tar_pose_jaw = rc.axis_angle_to_matrix(tar_pose_jaw.reshape(bs, n, 1, 3)) + tar_pose_jaw = rc.matrix_to_rotation_6d(tar_pose_jaw).reshape(bs, n, 1*6) + tar_pose_face = torch.cat([tar_pose_jaw, tar_exps], dim=2) + + tar_pose_hands = tar_pose[:, :, 25*3:55*3] + tar_pose_hands = rc.axis_angle_to_matrix(tar_pose_hands.reshape(bs, n, 30, 3)) + tar_pose_hands = rc.matrix_to_rotation_6d(tar_pose_hands).reshape(bs, n, 30*6) + + tar_pose_upper = tar_pose[:, :, self.joint_mask_upper.astype(bool)] + tar_pose_upper = rc.axis_angle_to_matrix(tar_pose_upper.reshape(bs, n, 13, 3)) + tar_pose_upper = rc.matrix_to_rotation_6d(tar_pose_upper).reshape(bs, n, 13*6) + + tar_pose_leg = tar_pose[:, :, self.joint_mask_lower.astype(bool)] + tar_pose_leg = rc.axis_angle_to_matrix(tar_pose_leg.reshape(bs, n, 9, 3)) + tar_pose_leg = rc.matrix_to_rotation_6d(tar_pose_leg).reshape(bs, n, 9*6) + tar_pose_lower = torch.cat([tar_pose_leg, tar_trans, tar_contact], dim=2) + + tar_pose_6d = rc.axis_angle_to_matrix(tar_pose.reshape(bs, n, 55, 3)) + tar_pose_6d = rc.matrix_to_rotation_6d(tar_pose_6d).reshape(bs, n, 55*6) + latent_all = torch.cat([tar_pose_6d, tar_trans, tar_contact], dim=-1) + + rec_all_face = [] + rec_all_upper = [] + rec_all_lower = [] + rec_all_hands = [] + vqvae_squeeze_scale = self.args.vqvae_squeeze_scale + roundt = (n - self.args.pre_frames * vqvae_squeeze_scale) // (self.args.pose_length - self.args.pre_frames * vqvae_squeeze_scale) + remain = (n - self.args.pre_frames * vqvae_squeeze_scale) % (self.args.pose_length - self.args.pre_frames * vqvae_squeeze_scale) + round_l = self.args.pose_length - self.args.pre_frames * vqvae_squeeze_scale + + + for i in range(0, roundt): + in_word_tmp = in_word[:, i*(round_l):(i+1)*(round_l)+self.args.pre_frames * vqvae_squeeze_scale] + + in_audio_tmp = in_audio[:, i*(16000//30*round_l):(i+1)*(16000//30*round_l)+16000//30*self.args.pre_frames * vqvae_squeeze_scale] + in_id_tmp = loaded_data['tar_id'][:, i*(round_l):(i+1)*(round_l)+self.args.pre_frames] + in_seed_tmp = in_seed[:, i*(round_l)//vqvae_squeeze_scale:(i+1)*(round_l)//vqvae_squeeze_scale+self.args.pre_frames] + in_x0_tmp = in_x0[:, i*(round_l)//vqvae_squeeze_scale:(i+1)*(round_l)//vqvae_squeeze_scale+self.args.pre_frames] + mask_val = torch.ones(bs, self.args.pose_length, self.args.pose_dims+3+4).float().cuda() + mask_val[:, :self.args.pre_frames, :] = 0.0 + if i == 0: + in_seed_tmp = in_seed_tmp[:, :self.args.pre_frames, :] + else: + in_seed_tmp = last_sample[:, -self.args.pre_frames:, :] + + cond_ = {'y':{}} + cond_['y']['audio'] = in_audio_tmp + cond_['y']['word'] = in_word_tmp + cond_['y']['id'] = in_id_tmp + cond_['y']['seed'] =in_seed_tmp + cond_['y']['mask'] = (torch.zeros([self.args.batch_size, 1, 1, self.args.pose_length]) < 1).cuda() + + + + cond_['y']['style_feature'] = torch.zeros([bs, 512]).cuda() + + shape_ = (bs, 1536, 1, 32) + sample = sample_fn( + self.model, + shape_, + clip_denoised=False, + model_kwargs=cond_, + skip_timesteps=0, + init_image=None, + progress=True, + dump_steps=None, + noise=None, + const_noise=False, + ) + sample = sample.squeeze().permute(1,0).unsqueeze(0) + + last_sample = sample.clone() + + rec_latent_upper = sample[...,:512] + rec_latent_hands = sample[...,512:1024] + rec_latent_lower = sample[...,1024:1536] + + + + if i == 0: + rec_all_upper.append(rec_latent_upper) + rec_all_hands.append(rec_latent_hands) + rec_all_lower.append(rec_latent_lower) + else: + rec_all_upper.append(rec_latent_upper[:, self.args.pre_frames:]) + rec_all_hands.append(rec_latent_hands[:, self.args.pre_frames:]) + rec_all_lower.append(rec_latent_lower[:, self.args.pre_frames:]) + + rec_all_upper = torch.cat(rec_all_upper, dim=1) * self.vqvae_latent_scale + rec_all_hands = torch.cat(rec_all_hands, dim=1) * self.vqvae_latent_scale + rec_all_lower = torch.cat(rec_all_lower, dim=1) * self.vqvae_latent_scale + + rec_upper = self.vq_model_upper.latent2origin(rec_all_upper)[0] + rec_hands = self.vq_model_hands.latent2origin(rec_all_hands)[0] + rec_lower = self.vq_model_lower.latent2origin(rec_all_lower)[0] + + + if self.use_trans: + rec_trans_v = rec_lower[...,-3:] + rec_trans_v = rec_trans_v * self.trans_std + self.trans_mean + rec_trans = torch.zeros_like(rec_trans_v) + rec_trans = torch.cumsum(rec_trans_v, dim=-2) + rec_trans[...,1]=rec_trans_v[...,1] + rec_lower = rec_lower[...,:-3] + + if self.args.pose_norm: + rec_upper = rec_upper * self.std_upper + self.mean_upper + rec_hands = rec_hands * self.std_hands + self.mean_hands + rec_lower = rec_lower * self.std_lower + self.mean_lower + + + + + n = n - remain + tar_pose = tar_pose[:, :n, :] + tar_exps = tar_exps[:, :n, :] + tar_trans = tar_trans[:, :n, :] + tar_beta = tar_beta[:, :n, :] + + + rec_exps = tar_exps + #rec_pose_jaw = rec_face[:, :, :6] + rec_pose_legs = rec_lower[:, :, :54] + bs, n = rec_pose_legs.shape[0], rec_pose_legs.shape[1] + rec_pose_upper = rec_upper.reshape(bs, n, 13, 6) + rec_pose_upper = rc.rotation_6d_to_matrix(rec_pose_upper)# + rec_pose_upper = rc.matrix_to_axis_angle(rec_pose_upper).reshape(bs*n, 13*3) + rec_pose_upper_recover = self.inverse_selection_tensor(rec_pose_upper, self.joint_mask_upper, bs*n) + rec_pose_lower = rec_pose_legs.reshape(bs, n, 9, 6) + rec_pose_lower = rc.rotation_6d_to_matrix(rec_pose_lower) + rec_lower2global = rc.matrix_to_rotation_6d(rec_pose_lower.clone()).reshape(bs, n, 9*6) + rec_pose_lower = rc.matrix_to_axis_angle(rec_pose_lower).reshape(bs*n, 9*3) + rec_pose_lower_recover = self.inverse_selection_tensor(rec_pose_lower, self.joint_mask_lower, bs*n) + rec_pose_hands = rec_hands.reshape(bs, n, 30, 6) + rec_pose_hands = rc.rotation_6d_to_matrix(rec_pose_hands) + rec_pose_hands = rc.matrix_to_axis_angle(rec_pose_hands).reshape(bs*n, 30*3) + rec_pose_hands_recover = self.inverse_selection_tensor(rec_pose_hands, self.joint_mask_hands, bs*n) + rec_pose = rec_pose_upper_recover + rec_pose_lower_recover + rec_pose_hands_recover + rec_pose[:, 66:69] = tar_pose.reshape(bs*n, 55*3)[:, 66:69] + + rec_pose = rc.axis_angle_to_matrix(rec_pose.reshape(bs*n, j, 3)) + rec_pose = rc.matrix_to_rotation_6d(rec_pose).reshape(bs, n, j*6) + tar_pose = rc.axis_angle_to_matrix(tar_pose.reshape(bs*n, j, 3)) + tar_pose = rc.matrix_to_rotation_6d(tar_pose).reshape(bs, n, j*6) + + return { + 'rec_pose': rec_pose, + 'rec_trans': rec_trans, + 'tar_pose': tar_pose, + 'tar_exps': tar_exps, + 'tar_beta': tar_beta, + 'tar_trans': tar_trans, + 'rec_exps': rec_exps, + } + + + def test_demo(self, epoch): + ''' + input audio and text, output motion + do not calculate loss and metric + save video + ''' + results_save_path = self.checkpoint_path + f"/{epoch}/" + if os.path.exists(results_save_path): + import shutil + shutil.rmtree(results_save_path) + os.makedirs(results_save_path) + start_time = time.time() + total_length = 0 + test_seq_list = self.test_data.selected_file + align = 0 + latent_out = [] + latent_ori = [] + l2_all = 0 + lvel = 0 + self.model.eval() + self.smplx.eval() + # self.eval_copy.eval() + with torch.no_grad(): + for its, batch_data in enumerate(self.test_loader): + loaded_data = self._load_data(batch_data) + net_out = self._g_test(loaded_data) + tar_pose = net_out['tar_pose'] + rec_pose = net_out['rec_pose'] + tar_exps = net_out['tar_exps'] + tar_beta = net_out['tar_beta'] + rec_trans = net_out['rec_trans'] + tar_trans = net_out['tar_trans'] + rec_exps = net_out['rec_exps'] + bs, n, j = tar_pose.shape[0], tar_pose.shape[1], self.joints + if (30/self.args.pose_fps) != 1: + assert 30%self.args.pose_fps == 0 + n *= int(30/self.args.pose_fps) + tar_pose = torch.nn.functional.interpolate(tar_pose.permute(0, 2, 1), scale_factor=30/self.args.pose_fps, mode='linear').permute(0,2,1) + rec_pose = torch.nn.functional.interpolate(rec_pose.permute(0, 2, 1), scale_factor=30/self.args.pose_fps, mode='linear').permute(0,2,1) + + + rec_pose = rc.rotation_6d_to_matrix(rec_pose.reshape(bs*n, j, 6)) + rec_pose = rc.matrix_to_rotation_6d(rec_pose).reshape(bs, n, j*6) + tar_pose = rc.rotation_6d_to_matrix(tar_pose.reshape(bs*n, j, 6)) + tar_pose = rc.matrix_to_rotation_6d(tar_pose).reshape(bs, n, j*6) + + rec_pose = rc.rotation_6d_to_matrix(rec_pose.reshape(bs*n, j, 6)) + rec_pose = rc.matrix_to_axis_angle(rec_pose).reshape(bs*n, j*3) + tar_pose = rc.rotation_6d_to_matrix(tar_pose.reshape(bs*n, j, 6)) + tar_pose = rc.matrix_to_axis_angle(tar_pose).reshape(bs*n, j*3) + + + tar_pose_np = tar_pose.detach().cpu().numpy() + rec_pose_np = rec_pose.detach().cpu().numpy() + rec_trans_np = rec_trans.detach().cpu().numpy().reshape(bs*n, 3) + rec_exp_np = rec_exps.detach().cpu().numpy().reshape(bs*n, 100) + tar_exp_np = tar_exps.detach().cpu().numpy().reshape(bs*n, 100) + tar_trans_np = tar_trans.detach().cpu().numpy().reshape(bs*n, 3) + gt_npz = np.load("./demo/examples/2_scott_0_1_1.npz", allow_pickle=True) + + results_npz_file_save_path = results_save_path+f"result_{self.time_name_expend[:-1]}"+'.npz' + np.savez(results_npz_file_save_path, + betas=gt_npz["betas"], + poses=rec_pose_np, + expressions=rec_exp_np, + trans=rec_trans_np, + model='smplx2020', + gender='neutral', + mocap_frame_rate = 30, + ) + total_length += n + render_vid_path = other_tools_hf.render_one_sequence_no_gt( + results_npz_file_save_path, + # results_save_path+"gt_"+test_seq_list.iloc[its]['id']+'.npz', + results_save_path, + self.audio_path, + self.args.data_path_1+"smplx_models/", + use_matplotlib = False, + args = self.args, + ) + + result = [ + gr.Video(value=render_vid_path, visible=True), + gr.File(value=results_npz_file_save_path, label="download motion and visualize in blender"), + ] + + end_time = time.time() - start_time + logger.info(f"total inference time: {int(end_time)} s for {int(total_length/self.args.pose_fps)} s motion") + return result + +@logger.catch +@spaces.GPU +def syntalker(audio_path,sample_stratege): + args = config.parse_args() + if sample_stratege==0: + args.use_ddim=True + elif sample_stratege==1: + args.use_ddim=False + print(sample_stratege) + print(args.use_ddim) + #os.environ['TRANSFORMERS_CACHE'] = args.data_path_1 + "hub/" + if not sys.warnoptions: + warnings.simplefilter("ignore") + # dist.init_process_group(backend="gloo", rank=rank, world_size=world_size) + + #logger_tools.set_args_and_logger(args, rank) + other_tools_hf.set_random_seed(args) + other_tools_hf.print_exp_info(args) + + # return one intance of trainer + trainer = BaseTrainer(args, ap = audio_path) + other_tools.load_checkpoints(trainer.model, args.test_ckpt, args.g_name) + + result = trainer.test_demo(999) + return result + +examples = [ + ["demo/examples/2_scott_0_1_1.wav"], + ["demo/examples/2_scott_0_2_2.wav"], + ["demo/examples/2_scott_0_3_3.wav"], + ["demo/examples/2_scott_0_4_4.wav"], + ["demo/examples/2_scott_0_5_5.wav"], +] + +demo = gr.Interface( + syntalker, # function + inputs=[ + # gr.File(label="Please upload SMPL-X file with npz format here.", file_types=["npz", "NPZ"]), + gr.Audio(), + gr.Radio(choices=["DDIM", "DDPM"], label="Please select a sample strategy", type="index", value="DDIM"), # 0 for DDIM, 1 for DDPM + # gr.File(label="Please upload textgrid format file here.", file_types=["TextGrid", "Textgrid", "textgrid"]) + ], # input type + outputs=[ + gr.Video(format="mp4", visible=True), + gr.File(label="download motion and visualize in blender") + ], + title='SynTalker: Enabling Synergistic Full-Body Control in Prompt-Based Co-Speech Motion Generation', + description="1. Upload your audio.
\ + 2. Then, sit back and wait for the rendering to happen! This may take a while (e.g. 2 minutes)
\ + 3. After, you can view the videos.
\ + 4. Notice that we use a fix face animation, our method only produce body motion.
\ + 5. Use DDPM sample strategy will generate a better result, while it will take more inference time. \ + ", + article="Project links: [SynTalker](https://robinwitch.github.io/SynTalker-Page).
\ + Reference links: [EMAGE](https://pantomatrix.github.io/EMAGE/). ", + examples=examples, +) + + +if __name__ == "__main__": + os.environ["MASTER_ADDR"]='127.0.0.1' + os.environ["MASTER_PORT"]='8675' + #os.environ["TORCH_DISTRIBUTED_DEBUG"] = "DETAIL" + demo.launch(share=True) diff --git a/bash_raw_cospeech_download.sh b/bash_raw_cospeech_download.sh new file mode 100644 index 0000000000000000000000000000000000000000..2f3752f2d12cfa0ea2bd1ad2b0e9091632ef26aa --- /dev/null +++ b/bash_raw_cospeech_download.sh @@ -0,0 +1,4 @@ +mkdir -p datasets/BEAT_SMPL +cd datasets/BEAT_SMPL +gdown https://drive.google.com/uc?id=1_iXr0XiT_EdslXe4b0HwDr2OoOCrtlrB +unzip beat_v2.0.0.zip \ No newline at end of file diff --git a/ckpt/beatx2_cospeech_diffusion/0403_212319_diffusion_rvqvae_128.txt b/ckpt/beatx2_cospeech_diffusion/0403_212319_diffusion_rvqvae_128.txt new file mode 100644 index 0000000000000000000000000000000000000000..fee18c4b1cc22e5c623fd6d25739ef9964e1235c --- /dev/null +++ b/ckpt/beatx2_cospeech_diffusion/0403_212319_diffusion_rvqvae_128.txt @@ -0,0 +1,19476 @@ + 04-03 21:23:19 | {'a_encoder': None, + 'a_fix_pre': False, + 'a_pre_encoder': None, + 'acc': 1, + 'acc_weight': 0.0, + 'additional_data': False, + 'adv_weight': 20.0, + 'ali_weight': 0.0, + 'amsgrad': False, + 'apex': False, + 'asmr': 0.0, + 'atcont': 0.0, + 'atmr': 0.0, + 'aud_prob': 1.0, + 'audio_dims': 1, + 'audio_f': 256, + 'audio_fps': 16000, + 'audio_norm': False, + 'audio_rep': 'onset+amplitude', + 'audio_sr': 16000, + 'batch_size': 40, + 'beat_align': True, + 'benchmark': True, + 'cache_only': False, + 'cache_path': 'datasets/beat_cache/beat_smplx_en_emage_2_128/', + 'cf': 0.0, + 'ch': 1.0, + 'cl': 1.0, + 'clean_final_seconds': 0, + 'clean_first_seconds': 0, + 'commit': 0.02, + 'config': 'configs/diffusion_rvqvae_128.yaml', + 'csv_name': 'a2g_0', + 'cu': 1.0, + 'cudnn_enabled': True, + 'd_lr_weight': 0.2, + 'd_name': None, + 'data_path': '/mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/', + 'data_path_1': '/mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/datasets/hub/', + 'dataset': 'beat_sep_lower', + 'ddp': False, + 'debug': False, + 'decay_epochs': 200, + 'decay_rate': 0.1, + 'decode_fusion': None, + 'depth': 3, + 'deterministic': True, + 'dilation_growth_rate': 3, + 'disable_filtering': False, + 'div_reg_weight': 0.0, + 'downs_t': [3], + 'dropout_prob': 0.3, + 'e_name': 'VAESKConv', + 'e_path': 'weights/AESKConv_240_100.bin', + 'emb_width': 512, + 'emo_rep': None, + 'emotion_dims': 8, + 'emotion_f': 0, + 'epoch_stage': 0, + 'epochs': 1000, + 'eval_model': 'motion_representation', + 'f_encoder': 'null', + 'f_fix_pre': False, + 'f_pre_encoder': 'null', + 'fac_prob': 1.0, + 'facial_dims': 100, + 'facial_f': 0, + 'facial_fps': 15, + 'facial_norm': False, + 'facial_rep': 'smplxflame_30', + 'fid_weight': 0.0, + 'finger_net': 'original', + 'freeze_wordembed': False, + 'fsmr': 0.0, + 'ftmr': 0.0, + 'fusion_mode': 'sum', + 'g_name': 'MDM', + 'gap_weight': 0.0, + 'gpus': [0], + 'grad_norm': 0.99, + 'hidden_size': 768, + 'hvqvae_multipliers': [1], + 'id_rep': 'onehot', + 'input_context': 'both', + 'is_train': True, + 'ita_weight': 0.0, + 'iwa_weight': 0.0, + 'joint_channel': 3, + 'kld_aud_weight': 0.0, + 'kld_fac_weight': 0.0, + 'kld_weight': 0.0, + 'l': 4, + 'l_bins': 512, + 'l_mu': 0.99, + 'levels': 1, + 'lf': 3.0, + 'lh': 3.0, + 'll': 3.0, + 'loader_workers': 0, + 'log_period': 10, + 'loss_contrastive_neg_weight': 0.005, + 'loss_contrastive_pos_weight': 0.2, + 'loss_gan_weight': 5.0, + 'loss_kld_weight': 0.1, + 'loss_physical_weight': 0.0, + 'loss_reg_weight': 0.05, + 'loss_regression_weight': 70.0, + 'lr_base': 5e-05, + 'lr_min': 1e-07, + 'lr_policy': 'step', + 'lu': 3.0, + 'm_conv': 1.0, + 'm_decoder': None, + 'm_encoder': 'null', + 'm_fix_pre': False, + 'm_pre_encoder': 'null', + 'mean_pose_path': '/mnt/fu09a/chenbohong/PantoMatrix/beatx_2_330_mean.npy', + 'mean_trans_path': '/mnt/fu09a/chenbohong/PantoMatrix/beatx_2_trans_mean.npy', + 'model': 'denoiser', + 'momentum': 0.8, + 'motion_f': 256, + 'msmr': 0.0, + 'mtmr': 0.0, + 'multi_length_training': [1.0], + 'n_layer': 1, + 'n_poses': 34, + 'n_pre_poses': 4, + 'name': '0403_212319_diffusion_rvqvae_128', + 'nesterov': True, + 'new_cache': False, + 'no_adv_epoch': 999, + 'notes': '', + 'opt': 'adam', + 'opt_betas': [0.5, 0.999], + 'ori_joints': 'beat_smplx_joints', + 'out_path': '/mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/outputs/audio2pose/', + 'pos_encoding_type': 'sin', + 'pos_prob': 1.0, + 'pose_dims': 330, + 'pose_fps': 30, + 'pose_length': 128, + 'pose_norm': True, + 'pose_rep': 'smplxflame_30', + 'pre_frames': 4, + 'pre_type': 'zero', + 'pretrain': False, + 'project': 's2g', + 'queue_size': 1024, + 'random_seed': 2021, + 'rec_aud_weight': 0.0, + 'rec_fac_weight': 0.0, + 'rec_pos_weight': 0.0, + 'rec_txt_weight': 0.0, + 'rec_ver_weight': 0.0, + 'rec_weight': 1.0, + 'root_path': '/mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/', + 'root_weight': 1.0, + 'rot6d': True, + 'sample_length': 34, + 'sem_rep': None, + 'sparse': 1, + 'speaker_dims': 4, + 'speaker_f': 0, + 'speaker_id': 'onehot', + 'stat': 'ts', + 'std_pose_path': '/mnt/fu09a/chenbohong/PantoMatrix/beatx_2_330_std.npy', + 'std_trans_path': '/mnt/fu09a/chenbohong/PantoMatrix/beatx_2_trans_std.npy', + 'stride': 20, + 'strides_t': [2], + 't_encoder': 'null', + 't_fix_pre': False, + 't_pre_encoder': 'fasttext', + 'tar_joints': 'beat_smplx_full', + 'test_ckpt': '/mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/outputs/audio2pose/custom/0330_140056_diffusion_rvqvae/last_300.bin', + 'test_data_path': '/datasets/trinity/test/', + 'test_length': 128, + 'test_period': 20, + 'train_data_path': '/datasets/trinity/train/', + 'train_trans': True, + 'trainer': 'diffusion_rvqvae', + 'training_speakers': [2], + 'tsmr': 0.0, + 'ttmr': 0.0, + 'txt_prob': 1.0, + 'use_amass': False, + 'use_aug': False, + 'use_bottleneck': True, + 'use_trans': True, + 'vae_codebook_size': 256, + 'vae_grow': [1, 1, 2, 1], + 'vae_layer': 4, + 'vae_length': 240, + 'vae_quantizer_lambda': 1.0, + 'vae_test_dim': 330, + 'vae_test_len': 32, + 'vae_test_stride': 20, + 'val_data_path': '/datasets/trinity/val/', + 'variational': False, + 'vel': 1, + 'vel_weight': 0.0, + 'vqvae_ckpt': None, + 'vqvae_hands_path': '/mnt/fu09a/chenbohong/gdc/T2M-GPT/output_beatx2/RVQVAE_hands/net_300000.pth', + 'vqvae_latent_scale': 5.0, + 'vqvae_lower_path': '/mnt/fu09a/chenbohong/gdc/T2M-GPT/output_beatx2/RVQVAE_lower/net_300000.pth', + 'vqvae_lower_trans_path': '/mnt/fu09a/chenbohong/gdc/T2M-GPT/output_beatx2/RVQVAE_lower_trans/net_300000.pth', + 'vqvae_reverse_decoder_dilation': True, + 'vqvae_squeeze_scale': 4, + 'vqvae_type': 'rvqvae', + 'vqvae_upper_path': '/mnt/fu09a/chenbohong/gdc/T2M-GPT/output_beatx2/RVQVAE_upper/net_300000.pth', + 'warmup_epochs': 0, + 'warmup_lr': 0.0005, + 'wei_weight': 0.0, + 'weight_decay': 0.0, + 'width': 512, + 'word_cache': False, + 'word_dims': 300, + 'word_f': 256, + 'word_index_num': 11195, + 'word_rep': 'textgrid', + 'z_type': 'speaker'} + 04-03 21:23:19 | # ------------ 0403_212319_diffusion_rvqvae_128 ----------- # + 04-03 21:23:19 | PyTorch version: 2.0.1+cu117 + 04-03 21:23:19 | CUDA version: 11.7 + 04-03 21:23:19 | 1 GPUs + 04-03 21:23:19 | Random Seed: 2021 + 04-03 21:23:20 | Audio bit rate: 16000 + 04-03 21:23:20 | Reading data '/mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/'... + 04-03 21:23:20 | Creating the dataset cache... + 04-03 21:23:20 | Found the cache /mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/datasets/beat_cache/beat_smplx_en_emage_2_128/train/smplxflame_30_cache + 04-03 21:23:20 | Init train dataloader success + 04-03 21:23:21 | Audio bit rate: 16000 + 04-03 21:23:21 | Reading data '/mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/'... + 04-03 21:23:21 | Creating the dataset cache... + 04-03 21:23:21 | Found the cache /mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/datasets/beat_cache/beat_smplx_en_emage_2_128/val/smplxflame_30_cache + 04-03 21:23:21 | Init val dataloader success + 04-03 21:23:21 | Audio bit rate: 16000 + 04-03 21:23:21 | Reading data '/mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/'... + 04-03 21:23:21 | Creating the dataset cache... + 04-03 21:23:21 | Found the cache /mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/datasets/beat_cache/beat_smplx_en_emage_2_128/test/smplxflame_30_cache + 04-03 21:23:21 | Init test dataloader success + 04-03 21:23:21 | DataParallel( + (module): MDM( + (WavEncoder): WavEncoder( + (feat_extractor): Sequential( + (0): BasicBlock( + (conv1): Conv1d(2, 64, kernel_size=(15,), stride=(5,), padding=(1700,)) + (bn1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act1): LeakyReLU(negative_slope=0.01, inplace=True) + (conv2): Conv1d(64, 64, kernel_size=(15,), stride=(1,), padding=(7,)) + (bn2): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act2): LeakyReLU(negative_slope=0.01, inplace=True) + (downsample): Sequential( + (0): Conv1d(2, 64, kernel_size=(15,), stride=(5,), padding=(1700,)) + (1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (1): BasicBlock( + (conv1): Conv1d(64, 64, kernel_size=(15,), stride=(6,)) + (bn1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act1): LeakyReLU(negative_slope=0.01, inplace=True) + (conv2): Conv1d(64, 64, kernel_size=(15,), stride=(1,), padding=(7,)) + (bn2): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act2): LeakyReLU(negative_slope=0.01, inplace=True) + (downsample): Sequential( + (0): Conv1d(64, 64, kernel_size=(15,), stride=(6,)) + (1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (2): BasicBlock( + (conv1): Conv1d(64, 64, kernel_size=(15,), stride=(1,), padding=(7,)) + (bn1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act1): LeakyReLU(negative_slope=0.01, inplace=True) + (conv2): Conv1d(64, 64, kernel_size=(15,), stride=(1,), padding=(7,)) + (bn2): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act2): LeakyReLU(negative_slope=0.01, inplace=True) + ) + (3): BasicBlock( + (conv1): Conv1d(64, 128, kernel_size=(15,), stride=(6,)) + (bn1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act1): LeakyReLU(negative_slope=0.01, inplace=True) + (conv2): Conv1d(128, 128, kernel_size=(15,), stride=(1,), padding=(7,)) + (bn2): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act2): LeakyReLU(negative_slope=0.01, inplace=True) + (downsample): Sequential( + (0): Conv1d(64, 128, kernel_size=(15,), stride=(6,)) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (4): BasicBlock( + (conv1): Conv1d(128, 128, kernel_size=(15,), stride=(1,), padding=(7,)) + (bn1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act1): LeakyReLU(negative_slope=0.01, inplace=True) + (conv2): Conv1d(128, 128, kernel_size=(15,), stride=(1,), padding=(7,)) + (bn2): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act2): LeakyReLU(negative_slope=0.01, inplace=True) + ) + (5): BasicBlock( + (conv1): Conv1d(128, 256, kernel_size=(15,), stride=(3,)) + (bn1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act1): LeakyReLU(negative_slope=0.01, inplace=True) + (conv2): Conv1d(256, 256, kernel_size=(15,), stride=(1,), padding=(7,)) + (bn2): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act2): LeakyReLU(negative_slope=0.01, inplace=True) + (downsample): Sequential( + (0): Conv1d(128, 256, kernel_size=(15,), stride=(3,)) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + ) + ) + (text_encoder_body): Linear(in_features=300, out_features=256, bias=True) + (text_pre_encoder_body): Embedding(11195, 300) + (sequence_pos_encoder): PositionalEncoding( + (dropout): Dropout(p=0.1, inplace=False) + ) + (mytimmblocks): ModuleList( + (0-7): 8 x Block( + (norm1): LayerNorm((512,), eps=1e-05, elementwise_affine=True) + (attn): Attention( + (qkv): Linear(in_features=512, out_features=1536, bias=False) + (q_norm): Identity() + (k_norm): Identity() + (attn_drop): Dropout(p=0.0, inplace=False) + (proj): Linear(in_features=512, out_features=512, bias=True) + (proj_drop): Dropout(p=0.0, inplace=False) + ) + (ls1): Identity() + (drop_path1): DropPath(drop_prob=0.100) + (norm2): LayerNorm((512,), eps=1e-05, elementwise_affine=True) + (mlp): Mlp( + (fc1): Linear(in_features=512, out_features=1024, bias=True) + (act): GELU(approximate='none') + (drop1): Dropout(p=0.0, inplace=False) + (norm): Identity() + (fc2): Linear(in_features=1024, out_features=512, bias=True) + (drop2): Dropout(p=0.0, inplace=False) + ) + (ls2): Identity() + (drop_path2): DropPath(drop_prob=0.100) + ) + ) + (embed_timestep): TimestepEmbedder( + (sequence_pos_encoder): PositionalEncoding( + (dropout): Dropout(p=0.1, inplace=False) + ) + (time_embed): Sequential( + (0): Linear(in_features=512, out_features=512, bias=True) + (1): SiLU() + (2): Linear(in_features=512, out_features=512, bias=True) + ) + ) + (embed_style): Linear(in_features=6, out_features=64, bias=True) + (embed_text): Linear(in_features=6144, out_features=512, bias=True) + (output_process): OutputProcess( + (poseFinal): Linear(in_features=512, out_features=1536, bias=True) + ) + (rel_pos): SinusoidalEmbeddings() + (input_process): InputProcess( + (poseEmbedding): Linear(in_features=1536, out_features=512, bias=True) + ) + (input_process2): Linear(in_features=1280, out_features=512, bias=True) + (mix_audio_text): Linear(in_features=512, out_features=256, bias=True) + ) +) + 04-03 21:23:21 | init MDM success + 04-03 21:23:21 | load self-pretrained checkpoints for VAESKConv + 04-03 21:23:21 | load self-pretrained checkpoints for VAESKConv + 04-03 21:23:21 | VAESKConv( + (encoder): LocalEncoder( + (layers): ModuleList( + (0): Sequential( + (0): SkeletonResidual( + (residual): Sequential( + (0): SkeletonConv() + (1): GroupNorm(10, 330, eps=1e-05, affine=True) + ) + (shortcut): SkeletonConv() + (common): Sequential( + (0): SkeletonPool() + (1): Tanh() + ) + ) + ) + (1): Sequential( + (0): SkeletonResidual( + (residual): Sequential( + (0): SkeletonConv() + (1): GroupNorm(10, 210, eps=1e-05, affine=True) + ) + (shortcut): SkeletonConv() + (common): Sequential( + (0): SkeletonPool() + (1): Tanh() + ) + ) + ) + (2-3): 2 x Sequential( + (0): SkeletonResidual( + (residual): Sequential( + (0): SkeletonConv() + (1): GroupNorm(10, 240, eps=1e-05, affine=True) + ) + (shortcut): SkeletonConv() + (common): Sequential( + (0): Tanh() + ) + ) + ) + ) + ) + (decoder): VQDecoderV3( + (main): Sequential( + (0): ResBlock( + (model): Sequential( + (0): Conv1d(240, 240, kernel_size=(3,), stride=(1,), padding=(1,)) + (1): LeakyReLU(negative_slope=0.2, inplace=True) + (2): Conv1d(240, 240, kernel_size=(3,), stride=(1,), padding=(1,)) + ) + ) + (1): ResBlock( + (model): Sequential( + (0): Conv1d(240, 240, kernel_size=(3,), stride=(1,), padding=(1,)) + (1): LeakyReLU(negative_slope=0.2, inplace=True) + (2): Conv1d(240, 240, kernel_size=(3,), stride=(1,), padding=(1,)) + ) + ) + (2): Upsample(scale_factor=2.0, mode='nearest') + (3): Conv1d(240, 240, kernel_size=(3,), stride=(1,), padding=(1,)) + (4): LeakyReLU(negative_slope=0.2, inplace=True) + (5): Upsample(scale_factor=2.0, mode='nearest') + (6): Conv1d(240, 240, kernel_size=(3,), stride=(1,), padding=(1,)) + (7): LeakyReLU(negative_slope=0.2, inplace=True) + (8): Upsample(scale_factor=2.0, mode='nearest') + (9): Conv1d(240, 240, kernel_size=(3,), stride=(1,), padding=(1,)) + (10): LeakyReLU(negative_slope=0.2, inplace=True) + (11): Upsample(scale_factor=2.0, mode='nearest') + (12): Conv1d(240, 330, kernel_size=(3,), stride=(1,), padding=(1,)) + (13): LeakyReLU(negative_slope=0.2, inplace=True) + (14): Conv1d(330, 330, kernel_size=(3,), stride=(1,), padding=(1,)) + ) + ) + (fc_mu): Linear(in_features=240, out_features=240, bias=True) + (fc_logvar): Linear(in_features=240, out_features=240, bias=True) +) + 04-03 21:23:21 | init VAESKConv success + 04-03 21:23:22 | load self-pretrained checkpoints for VAESKConv + 04-03 21:23:22 | load self-pretrained checkpoints for VAESKConv + 04-03 21:23:22 | Training from scratch ... + 04-03 21:23:22 | Time info >>>> elapsed: 0.00 mins remain: 476.84 mins + 04-03 21:23:25 | [000][000/179] predict_x0_loss: 0.257 glr: 5.0e-05 dtime: 1527 ntime: 1605 mem: 1.75 + 04-03 21:23:27 | [000][010/179] predict_x0_loss: 0.191 glr: 5.0e-05 dtime: 0056 ntime: 0082 mem: 3.36 + 04-03 21:23:28 | [000][020/179] predict_x0_loss: 0.161 glr: 5.0e-05 dtime: 0086 ntime: 0081 mem: 3.36 + 04-03 21:23:30 | [000][030/179] predict_x0_loss: 0.143 glr: 5.0e-05 dtime: 0069 ntime: 0083 mem: 3.36 + 04-03 21:23:31 | [000][040/179] predict_x0_loss: 0.130 glr: 5.0e-05 dtime: 0067 ntime: 0082 mem: 3.36 + 04-03 21:23:33 | [000][050/179] predict_x0_loss: 0.121 glr: 5.0e-05 dtime: 0058 ntime: 0083 mem: 3.36 + 04-03 21:23:34 | [000][060/179] predict_x0_loss: 0.115 glr: 5.0e-05 dtime: 0058 ntime: 0088 mem: 3.36 + 04-03 21:23:36 | [000][070/179] predict_x0_loss: 0.110 glr: 5.0e-05 dtime: 0064 ntime: 0077 mem: 3.36 + 04-03 21:23:37 | [000][080/179] predict_x0_loss: 0.106 glr: 5.0e-05 dtime: 0076 ntime: 0084 mem: 3.36 + 04-03 21:23:39 | [000][090/179] predict_x0_loss: 0.103 glr: 5.0e-05 dtime: 0067 ntime: 0084 mem: 3.36 + 04-03 21:23:40 | [000][100/179] predict_x0_loss: 0.100 glr: 5.0e-05 dtime: 0061 ntime: 0080 mem: 3.36 + 04-03 21:23:42 | [000][110/179] predict_x0_loss: 0.098 glr: 5.0e-05 dtime: 0059 ntime: 0075 mem: 3.36 + 04-03 21:23:43 | [000][120/179] predict_x0_loss: 0.096 glr: 5.0e-05 dtime: 0066 ntime: 0082 mem: 3.36 + 04-03 21:23:44 | [000][130/179] predict_x0_loss: 0.094 glr: 5.0e-05 dtime: 0061 ntime: 0079 mem: 3.36 + 04-03 21:23:46 | [000][140/179] predict_x0_loss: 0.092 glr: 5.0e-05 dtime: 0054 ntime: 0073 mem: 3.36 + 04-03 21:23:47 | [000][150/179] predict_x0_loss: 0.091 glr: 5.0e-05 dtime: 0067 ntime: 0082 mem: 3.36 + 04-03 21:23:49 | [000][160/179] predict_x0_loss: 0.090 glr: 5.0e-05 dtime: 0065 ntime: 0079 mem: 3.36 + 04-03 21:23:50 | [000][170/179] predict_x0_loss: 0.089 glr: 5.0e-05 dtime: 0071 ntime: 0081 mem: 3.36 + 04-03 21:23:51 | Time info >>>> elapsed: 0.48 mins remain: 478.24 mins + 04-03 21:23:51 | [001][000/179] predict_x0_loss: 0.066 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:23:52 | [001][010/179] predict_x0_loss: 0.068 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:23:54 | [001][020/179] predict_x0_loss: 0.067 glr: 5.0e-05 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 21:23:55 | [001][030/179] predict_x0_loss: 0.067 glr: 5.0e-05 dtime: 0045 ntime: 0082 mem: 3.36 + 04-03 21:23:56 | [001][040/179] predict_x0_loss: 0.066 glr: 5.0e-05 dtime: 0052 ntime: 0073 mem: 3.36 + 04-03 21:23:58 | [001][050/179] predict_x0_loss: 0.066 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:23:59 | [001][060/179] predict_x0_loss: 0.065 glr: 5.0e-05 dtime: 0043 ntime: 0081 mem: 3.36 + 04-03 21:24:00 | [001][070/179] predict_x0_loss: 0.065 glr: 5.0e-05 dtime: 0045 ntime: 0074 mem: 3.36 + 04-03 21:24:02 | [001][080/179] predict_x0_loss: 0.064 glr: 5.0e-05 dtime: 0044 ntime: 0079 mem: 3.36 + 04-03 21:24:03 | [001][090/179] predict_x0_loss: 0.064 glr: 5.0e-05 dtime: 0043 ntime: 0075 mem: 3.36 + 04-03 21:24:04 | [001][100/179] predict_x0_loss: 0.063 glr: 5.0e-05 dtime: 0044 ntime: 0075 mem: 3.36 + 04-03 21:24:05 | [001][110/179] predict_x0_loss: 0.063 glr: 5.0e-05 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 21:24:07 | [001][120/179] predict_x0_loss: 0.062 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 21:24:08 | [001][130/179] predict_x0_loss: 0.062 glr: 5.0e-05 dtime: 0044 ntime: 0081 mem: 3.36 + 04-03 21:24:09 | [001][140/179] predict_x0_loss: 0.062 glr: 5.0e-05 dtime: 0054 ntime: 0075 mem: 3.36 + 04-03 21:24:11 | [001][150/179] predict_x0_loss: 0.061 glr: 5.0e-05 dtime: 0042 ntime: 0078 mem: 3.36 + 04-03 21:24:12 | [001][160/179] predict_x0_loss: 0.061 glr: 5.0e-05 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 21:24:13 | [001][170/179] predict_x0_loss: 0.060 glr: 5.0e-05 dtime: 0043 ntime: 0078 mem: 3.36 + 04-03 21:24:14 | Time info >>>> elapsed: 0.86 mins remain: 430.79 mins + 04-03 21:24:14 | [002][000/179] predict_x0_loss: 0.055 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:24:16 | [002][010/179] predict_x0_loss: 0.053 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 21:24:17 | [002][020/179] predict_x0_loss: 0.052 glr: 5.0e-05 dtime: 0046 ntime: 0082 mem: 3.36 + 04-03 21:24:18 | [002][030/179] predict_x0_loss: 0.052 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:24:20 | [002][040/179] predict_x0_loss: 0.051 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:24:21 | [002][050/179] predict_x0_loss: 0.051 glr: 5.0e-05 dtime: 0045 ntime: 0074 mem: 3.36 + 04-03 21:24:22 | [002][060/179] predict_x0_loss: 0.051 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:24:23 | [002][070/179] predict_x0_loss: 0.051 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 21:24:25 | [002][080/179] predict_x0_loss: 0.051 glr: 5.0e-05 dtime: 0053 ntime: 0079 mem: 3.36 + 04-03 21:24:26 | [002][090/179] predict_x0_loss: 0.050 glr: 5.0e-05 dtime: 0046 ntime: 0079 mem: 3.36 + 04-03 21:24:27 | [002][100/179] predict_x0_loss: 0.050 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 21:24:28 | [002][110/179] predict_x0_loss: 0.050 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 21:24:30 | [002][120/179] predict_x0_loss: 0.050 glr: 5.0e-05 dtime: 0056 ntime: 0079 mem: 3.36 + 04-03 21:24:31 | [002][130/179] predict_x0_loss: 0.050 glr: 5.0e-05 dtime: 0045 ntime: 0083 mem: 3.36 + 04-03 21:24:32 | [002][140/179] predict_x0_loss: 0.049 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:24:34 | [002][150/179] predict_x0_loss: 0.049 glr: 5.0e-05 dtime: 0046 ntime: 0075 mem: 3.36 + 04-03 21:24:35 | [002][160/179] predict_x0_loss: 0.049 glr: 5.0e-05 dtime: 0045 ntime: 0081 mem: 3.36 + 04-03 21:24:36 | [002][170/179] predict_x0_loss: 0.049 glr: 5.0e-05 dtime: 0044 ntime: 0077 mem: 3.36 + 04-03 21:24:37 | Time info >>>> elapsed: 1.25 mins remain: 414.42 mins + 04-03 21:24:37 | [003][000/179] predict_x0_loss: 0.047 glr: 5.0e-05 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 21:24:39 | [003][010/179] predict_x0_loss: 0.045 glr: 5.0e-05 dtime: 0055 ntime: 0086 mem: 3.36 + 04-03 21:24:40 | [003][020/179] predict_x0_loss: 0.045 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 21:24:41 | [003][030/179] predict_x0_loss: 0.045 glr: 5.0e-05 dtime: 0050 ntime: 0085 mem: 3.36 + 04-03 21:24:43 | [003][040/179] predict_x0_loss: 0.045 glr: 5.0e-05 dtime: 0044 ntime: 0080 mem: 3.36 + 04-03 21:24:44 | [003][050/179] predict_x0_loss: 0.045 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:24:45 | [003][060/179] predict_x0_loss: 0.045 glr: 5.0e-05 dtime: 0045 ntime: 0078 mem: 3.36 + 04-03 21:24:46 | [003][070/179] predict_x0_loss: 0.044 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:24:48 | [003][080/179] predict_x0_loss: 0.044 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 21:24:49 | [003][090/179] predict_x0_loss: 0.044 glr: 5.0e-05 dtime: 0053 ntime: 0079 mem: 3.36 + 04-03 21:24:50 | [003][100/179] predict_x0_loss: 0.044 glr: 5.0e-05 dtime: 0045 ntime: 0078 mem: 3.36 + 04-03 21:24:52 | [003][110/179] predict_x0_loss: 0.044 glr: 5.0e-05 dtime: 0055 ntime: 0082 mem: 3.36 + 04-03 21:24:53 | [003][120/179] predict_x0_loss: 0.044 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:24:55 | [003][130/179] predict_x0_loss: 0.044 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:24:56 | [003][140/179] predict_x0_loss: 0.043 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 21:24:57 | [003][150/179] predict_x0_loss: 0.043 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 21:24:58 | [003][160/179] predict_x0_loss: 0.043 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 21:25:00 | [003][170/179] predict_x0_loss: 0.043 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:25:01 | Time info >>>> elapsed: 1.64 mins remain: 408.97 mins + 04-03 21:25:01 | [004][000/179] predict_x0_loss: 0.039 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 21:25:02 | [004][010/179] predict_x0_loss: 0.040 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:25:04 | [004][020/179] predict_x0_loss: 0.040 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:25:05 | [004][030/179] predict_x0_loss: 0.040 glr: 5.0e-05 dtime: 0060 ntime: 0088 mem: 3.36 + 04-03 21:25:06 | [004][040/179] predict_x0_loss: 0.040 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 21:25:08 | [004][050/179] predict_x0_loss: 0.040 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 21:25:09 | [004][060/179] predict_x0_loss: 0.040 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:25:11 | [004][070/179] predict_x0_loss: 0.040 glr: 5.0e-05 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 21:25:12 | [004][080/179] predict_x0_loss: 0.040 glr: 5.0e-05 dtime: 0065 ntime: 0085 mem: 3.36 + 04-03 21:25:13 | [004][090/179] predict_x0_loss: 0.040 glr: 5.0e-05 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 21:25:15 | [004][100/179] predict_x0_loss: 0.040 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 21:25:16 | [004][110/179] predict_x0_loss: 0.040 glr: 5.0e-05 dtime: 0058 ntime: 0079 mem: 3.36 + 04-03 21:25:17 | [004][120/179] predict_x0_loss: 0.040 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 21:25:19 | [004][130/179] predict_x0_loss: 0.040 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 21:25:20 | [004][140/179] predict_x0_loss: 0.039 glr: 5.0e-05 dtime: 0057 ntime: 0073 mem: 3.36 + 04-03 21:25:21 | [004][150/179] predict_x0_loss: 0.039 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:25:23 | [004][160/179] predict_x0_loss: 0.039 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:25:24 | [004][170/179] predict_x0_loss: 0.039 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:25:25 | Time info >>>> elapsed: 2.04 mins remain: 406.53 mins + 04-03 21:25:25 | [005][000/179] predict_x0_loss: 0.040 glr: 5.0e-05 dtime: 0058 ntime: 0073 mem: 3.36 + 04-03 21:25:26 | [005][010/179] predict_x0_loss: 0.039 glr: 5.0e-05 dtime: 0045 ntime: 0070 mem: 3.36 + 04-03 21:25:28 | [005][020/179] predict_x0_loss: 0.038 glr: 5.0e-05 dtime: 0051 ntime: 0074 mem: 3.36 + 04-03 21:25:29 | [005][030/179] predict_x0_loss: 0.038 glr: 5.0e-05 dtime: 0055 ntime: 0077 mem: 3.36 + 04-03 21:25:30 | [005][040/179] predict_x0_loss: 0.037 glr: 5.0e-05 dtime: 0057 ntime: 0082 mem: 3.36 + 04-03 21:25:32 | [005][050/179] predict_x0_loss: 0.037 glr: 5.0e-05 dtime: 0055 ntime: 0085 mem: 3.36 + 04-03 21:25:33 | [005][060/179] predict_x0_loss: 0.037 glr: 5.0e-05 dtime: 0055 ntime: 0080 mem: 3.36 + 04-03 21:25:34 | [005][070/179] predict_x0_loss: 0.037 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:25:36 | [005][080/179] predict_x0_loss: 0.037 glr: 5.0e-05 dtime: 0048 ntime: 0086 mem: 3.36 + 04-03 21:25:37 | [005][090/179] predict_x0_loss: 0.037 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 21:25:38 | [005][100/179] predict_x0_loss: 0.037 glr: 5.0e-05 dtime: 0057 ntime: 0078 mem: 3.36 + 04-03 21:25:40 | [005][110/179] predict_x0_loss: 0.037 glr: 5.0e-05 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 21:25:41 | [005][120/179] predict_x0_loss: 0.037 glr: 5.0e-05 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 21:25:43 | [005][130/179] predict_x0_loss: 0.037 glr: 5.0e-05 dtime: 0071 ntime: 0087 mem: 3.36 + 04-03 21:25:44 | [005][140/179] predict_x0_loss: 0.037 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 21:25:45 | [005][150/179] predict_x0_loss: 0.037 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 21:25:47 | [005][160/179] predict_x0_loss: 0.036 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 21:25:48 | [005][170/179] predict_x0_loss: 0.036 glr: 5.0e-05 dtime: 0053 ntime: 0075 mem: 3.36 + 04-03 21:25:49 | Time info >>>> elapsed: 2.45 mins remain: 405.32 mins + 04-03 21:25:49 | [006][000/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 21:25:51 | [006][010/179] predict_x0_loss: 0.036 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 21:25:52 | [006][020/179] predict_x0_loss: 0.035 glr: 5.0e-05 dtime: 0046 ntime: 0076 mem: 3.36 + 04-03 21:25:53 | [006][030/179] predict_x0_loss: 0.035 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:25:55 | [006][040/179] predict_x0_loss: 0.035 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 21:25:56 | [006][050/179] predict_x0_loss: 0.035 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 21:25:57 | [006][060/179] predict_x0_loss: 0.035 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:25:59 | [006][070/179] predict_x0_loss: 0.035 glr: 5.0e-05 dtime: 0049 ntime: 0087 mem: 3.36 + 04-03 21:26:00 | [006][080/179] predict_x0_loss: 0.035 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:26:01 | [006][090/179] predict_x0_loss: 0.035 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 21:26:02 | [006][100/179] predict_x0_loss: 0.035 glr: 5.0e-05 dtime: 0057 ntime: 0085 mem: 3.36 + 04-03 21:26:04 | [006][110/179] predict_x0_loss: 0.035 glr: 5.0e-05 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 21:26:05 | [006][120/179] predict_x0_loss: 0.034 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 21:26:07 | [006][130/179] predict_x0_loss: 0.034 glr: 5.0e-05 dtime: 0045 ntime: 0075 mem: 3.36 + 04-03 21:26:08 | [006][140/179] predict_x0_loss: 0.034 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 21:26:09 | [006][150/179] predict_x0_loss: 0.034 glr: 5.0e-05 dtime: 0054 ntime: 0074 mem: 3.36 + 04-03 21:26:10 | [006][160/179] predict_x0_loss: 0.034 glr: 5.0e-05 dtime: 0047 ntime: 0084 mem: 3.36 + 04-03 21:26:12 | [006][170/179] predict_x0_loss: 0.034 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:26:13 | Time info >>>> elapsed: 2.84 mins remain: 403.17 mins + 04-03 21:26:13 | [007][000/179] predict_x0_loss: 0.036 glr: 5.0e-05 dtime: 0053 ntime: 0086 mem: 3.36 + 04-03 21:26:14 | [007][010/179] predict_x0_loss: 0.032 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:26:16 | [007][020/179] predict_x0_loss: 0.032 glr: 5.0e-05 dtime: 0045 ntime: 0083 mem: 3.36 + 04-03 21:26:17 | [007][030/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0052 ntime: 0074 mem: 3.36 + 04-03 21:26:18 | [007][040/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:26:20 | [007][050/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0045 ntime: 0075 mem: 3.36 + 04-03 21:26:21 | [007][060/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 21:26:22 | [007][070/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:26:23 | [007][080/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 21:26:25 | [007][090/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 21:26:26 | [007][100/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 21:26:27 | [007][110/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 21:26:29 | [007][120/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 21:26:30 | [007][130/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0057 ntime: 0088 mem: 3.36 + 04-03 21:26:31 | [007][140/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 21:26:33 | [007][150/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0053 ntime: 0083 mem: 3.36 + 04-03 21:26:34 | [007][160/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0041 ntime: 0062 mem: 3.36 + 04-03 21:26:35 | [007][170/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 21:26:36 | Time info >>>> elapsed: 3.23 mins remain: 400.60 mins + 04-03 21:26:36 | [008][000/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0079 ntime: 0085 mem: 3.36 + 04-03 21:26:38 | [008][010/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0046 ntime: 0080 mem: 3.36 + 04-03 21:26:39 | [008][020/179] predict_x0_loss: 0.032 glr: 5.0e-05 dtime: 0055 ntime: 0084 mem: 3.36 + 04-03 21:26:40 | [008][030/179] predict_x0_loss: 0.032 glr: 5.0e-05 dtime: 0054 ntime: 0083 mem: 3.36 + 04-03 21:26:42 | [008][040/179] predict_x0_loss: 0.032 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 21:26:43 | [008][050/179] predict_x0_loss: 0.032 glr: 5.0e-05 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 21:26:44 | [008][060/179] predict_x0_loss: 0.032 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 21:26:46 | [008][070/179] predict_x0_loss: 0.032 glr: 5.0e-05 dtime: 0060 ntime: 0083 mem: 3.36 + 04-03 21:26:47 | [008][080/179] predict_x0_loss: 0.032 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:26:48 | [008][090/179] predict_x0_loss: 0.032 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 21:26:50 | [008][100/179] predict_x0_loss: 0.032 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:26:51 | [008][110/179] predict_x0_loss: 0.032 glr: 5.0e-05 dtime: 0049 ntime: 0071 mem: 3.36 + 04-03 21:26:52 | [008][120/179] predict_x0_loss: 0.032 glr: 5.0e-05 dtime: 0045 ntime: 0082 mem: 3.36 + 04-03 21:26:54 | [008][130/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0056 ntime: 0079 mem: 3.36 + 04-03 21:26:55 | [008][140/179] predict_x0_loss: 0.032 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 21:26:56 | [008][150/179] predict_x0_loss: 0.032 glr: 5.0e-05 dtime: 0053 ntime: 0083 mem: 3.36 + 04-03 21:26:58 | [008][160/179] predict_x0_loss: 0.032 glr: 5.0e-05 dtime: 0056 ntime: 0075 mem: 3.36 + 04-03 21:26:59 | [008][170/179] predict_x0_loss: 0.032 glr: 5.0e-05 dtime: 0071 ntime: 0087 mem: 3.36 + 04-03 21:27:00 | Time info >>>> elapsed: 3.63 mins remain: 399.82 mins + 04-03 21:27:00 | [009][000/179] predict_x0_loss: 0.033 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:27:02 | [009][010/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 21:27:03 | [009][020/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 21:27:04 | [009][030/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 21:27:06 | [009][040/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0047 ntime: 0086 mem: 3.36 + 04-03 21:27:07 | [009][050/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 21:27:08 | [009][060/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0046 ntime: 0082 mem: 3.36 + 04-03 21:27:10 | [009][070/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0046 ntime: 0082 mem: 3.36 + 04-03 21:27:11 | [009][080/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:27:12 | [009][090/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0077 ntime: 0081 mem: 3.36 + 04-03 21:27:14 | [009][100/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0059 ntime: 0084 mem: 3.36 + 04-03 21:27:15 | [009][110/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0064 ntime: 0083 mem: 3.36 + 04-03 21:27:17 | [009][120/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:27:18 | [009][130/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0067 ntime: 0082 mem: 3.36 + 04-03 21:27:19 | [009][140/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 21:27:21 | [009][150/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 21:27:22 | [009][160/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 21:27:23 | [009][170/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 21:27:24 | Time info >>>> elapsed: 4.03 mins remain: 399.44 mins + 04-03 21:27:25 | [010][000/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0046 ntime: 0081 mem: 3.36 + 04-03 21:27:26 | [010][010/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:27:27 | [010][020/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0043 ntime: 0078 mem: 3.36 + 04-03 21:27:29 | [010][030/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0063 ntime: 0075 mem: 3.36 + 04-03 21:27:30 | [010][040/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0054 ntime: 0086 mem: 3.36 + 04-03 21:27:31 | [010][050/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0051 ntime: 0087 mem: 3.36 + 04-03 21:27:33 | [010][060/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0053 ntime: 0085 mem: 3.36 + 04-03 21:27:34 | [010][070/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 21:27:35 | [010][080/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 21:27:37 | [010][090/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0046 ntime: 0083 mem: 3.36 + 04-03 21:27:38 | [010][100/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:27:39 | [010][110/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0061 ntime: 0080 mem: 3.36 + 04-03 21:27:41 | [010][120/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0047 ntime: 0083 mem: 3.36 + 04-03 21:27:42 | [010][130/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 21:27:43 | [010][140/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 21:27:45 | [010][150/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 21:27:46 | [010][160/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0053 ntime: 0084 mem: 3.36 + 04-03 21:27:47 | [010][170/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 21:27:48 | Time info >>>> elapsed: 4.43 mins remain: 398.59 mins + 04-03 21:27:48 | [011][000/179] predict_x0_loss: 0.031 glr: 5.0e-05 dtime: 0046 ntime: 0080 mem: 3.36 + 04-03 21:27:50 | [011][010/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:27:51 | [011][020/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:27:52 | [011][030/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 21:27:54 | [011][040/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:27:55 | [011][050/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 21:27:56 | [011][060/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:27:58 | [011][070/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 21:27:59 | [011][080/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0046 ntime: 0083 mem: 3.36 + 04-03 21:28:00 | [011][090/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 21:28:02 | [011][100/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0049 ntime: 0086 mem: 3.36 + 04-03 21:28:03 | [011][110/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 21:28:04 | [011][120/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 21:28:06 | [011][130/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:28:07 | [011][140/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0052 ntime: 0085 mem: 3.36 + 04-03 21:28:08 | [011][150/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0050 ntime: 0070 mem: 3.36 + 04-03 21:28:10 | [011][160/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0054 ntime: 0074 mem: 3.36 + 04-03 21:28:11 | [011][170/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 21:28:12 | Time info >>>> elapsed: 4.83 mins remain: 397.63 mins + 04-03 21:28:12 | [012][000/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 21:28:14 | [012][010/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0052 ntime: 0072 mem: 3.36 + 04-03 21:28:15 | [012][020/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 21:28:16 | [012][030/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 21:28:18 | [012][040/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0049 ntime: 0088 mem: 3.36 + 04-03 21:28:19 | [012][050/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0058 ntime: 0081 mem: 3.36 + 04-03 21:28:20 | [012][060/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0067 ntime: 0083 mem: 3.36 + 04-03 21:28:22 | [012][070/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0046 ntime: 0081 mem: 3.36 + 04-03 21:28:23 | [012][080/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0070 ntime: 0084 mem: 3.36 + 04-03 21:28:24 | [012][090/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 21:28:26 | [012][100/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 21:28:27 | [012][110/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 21:28:28 | [012][120/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:28:30 | [012][130/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 21:28:31 | [012][140/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 21:28:32 | [012][150/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:28:34 | [012][160/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 21:28:35 | [012][170/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 21:28:36 | Time info >>>> elapsed: 5.23 mins remain: 396.89 mins + 04-03 21:28:36 | [013][000/179] predict_x0_loss: 0.034 glr: 5.0e-05 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 21:28:37 | [013][010/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0054 ntime: 0076 mem: 3.36 + 04-03 21:28:39 | [013][020/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0045 ntime: 0077 mem: 3.36 + 04-03 21:28:40 | [013][030/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 21:28:41 | [013][040/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 21:28:43 | [013][050/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0047 ntime: 0071 mem: 3.36 + 04-03 21:28:44 | [013][060/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 21:28:45 | [013][070/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0051 ntime: 0088 mem: 3.36 + 04-03 21:28:47 | [013][080/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 21:28:48 | [013][090/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 21:28:49 | [013][100/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 21:28:50 | [013][110/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 21:28:52 | [013][120/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0057 ntime: 0077 mem: 3.36 + 04-03 21:28:53 | [013][130/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0052 ntime: 0094 mem: 3.36 + 04-03 21:28:54 | [013][140/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0048 ntime: 0085 mem: 3.36 + 04-03 21:28:56 | [013][150/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:28:57 | [013][160/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 21:28:58 | [013][170/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0057 ntime: 0076 mem: 3.36 + 04-03 21:28:59 | Time info >>>> elapsed: 5.62 mins remain: 395.65 mins + 04-03 21:29:00 | [014][000/179] predict_x0_loss: 0.030 glr: 5.0e-05 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 21:29:01 | [014][010/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0046 ntime: 0084 mem: 3.36 + 04-03 21:29:02 | [014][020/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:29:04 | [014][030/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 21:29:05 | [014][040/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0073 ntime: 0086 mem: 3.36 + 04-03 21:29:06 | [014][050/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:29:08 | [014][060/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 21:29:09 | [014][070/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 21:29:10 | [014][080/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 21:29:12 | [014][090/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:29:13 | [014][100/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0055 ntime: 0082 mem: 3.36 + 04-03 21:29:14 | [014][110/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 21:29:16 | [014][120/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0054 ntime: 0086 mem: 3.36 + 04-03 21:29:17 | [014][130/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 21:29:19 | [014][140/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0056 ntime: 0077 mem: 3.36 + 04-03 21:29:20 | [014][150/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0075 ntime: 0076 mem: 3.36 + 04-03 21:29:21 | [014][160/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0057 ntime: 0080 mem: 3.36 + 04-03 21:29:23 | [014][170/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0059 ntime: 0072 mem: 3.36 + 04-03 21:29:24 | Time info >>>> elapsed: 6.02 mins remain: 395.64 mins + 04-03 21:29:24 | [015][000/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 21:29:25 | [015][010/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 21:29:27 | [015][020/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0059 ntime: 0077 mem: 3.36 + 04-03 21:29:28 | [015][030/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0056 ntime: 0076 mem: 3.36 + 04-03 21:29:29 | [015][040/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0047 ntime: 0085 mem: 3.36 + 04-03 21:29:31 | [015][050/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 21:29:32 | [015][060/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0053 ntime: 0085 mem: 3.36 + 04-03 21:29:33 | [015][070/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 21:29:35 | [015][080/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 21:29:36 | [015][090/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:29:37 | [015][100/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 21:29:38 | [015][110/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0054 ntime: 0083 mem: 3.36 + 04-03 21:29:40 | [015][120/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 21:29:41 | [015][130/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0050 ntime: 0086 mem: 3.36 + 04-03 21:29:42 | [015][140/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 21:29:44 | [015][150/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 21:29:45 | [015][160/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0056 ntime: 0078 mem: 3.36 + 04-03 21:29:46 | [015][170/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 21:29:48 | Time info >>>> elapsed: 6.42 mins remain: 394.81 mins + 04-03 21:29:48 | [016][000/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0046 ntime: 0075 mem: 3.36 + 04-03 21:29:49 | [016][010/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0048 ntime: 0071 mem: 3.36 + 04-03 21:29:50 | [016][020/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0050 ntime: 0061 mem: 3.36 + 04-03 21:29:51 | [016][030/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 21:29:53 | [016][040/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0050 ntime: 0073 mem: 3.36 + 04-03 21:29:54 | [016][050/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0053 ntime: 0072 mem: 3.36 + 04-03 21:29:55 | [016][060/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:29:57 | [016][070/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0053 ntime: 0084 mem: 3.36 + 04-03 21:29:58 | [016][080/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 21:29:59 | [016][090/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0049 ntime: 0091 mem: 3.36 + 04-03 21:30:01 | [016][100/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0052 ntime: 0093 mem: 3.36 + 04-03 21:30:02 | [016][110/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 21:30:03 | [016][120/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 21:30:05 | [016][130/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:30:06 | [016][140/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0056 ntime: 0082 mem: 3.36 + 04-03 21:30:07 | [016][150/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:30:09 | [016][160/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:30:10 | [016][170/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0053 ntime: 0083 mem: 3.36 + 04-03 21:30:11 | Time info >>>> elapsed: 6.81 mins remain: 393.75 mins + 04-03 21:30:11 | [017][000/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 21:30:12 | [017][010/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0049 ntime: 0092 mem: 3.36 + 04-03 21:30:14 | [017][020/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 21:30:15 | [017][030/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:30:16 | [017][040/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0052 ntime: 0085 mem: 3.36 + 04-03 21:30:18 | [017][050/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 21:30:19 | [017][060/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 21:30:20 | [017][070/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 21:30:22 | [017][080/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:30:23 | [017][090/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0059 ntime: 0076 mem: 3.36 + 04-03 21:30:24 | [017][100/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:30:26 | [017][110/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 21:30:27 | [017][120/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0047 ntime: 0073 mem: 3.36 + 04-03 21:30:28 | [017][130/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0049 ntime: 0089 mem: 3.36 + 04-03 21:30:30 | [017][140/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 21:30:31 | [017][150/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:30:32 | [017][160/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 21:30:34 | [017][170/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0052 ntime: 0086 mem: 3.36 + 04-03 21:30:35 | Time info >>>> elapsed: 7.21 mins remain: 393.10 mins + 04-03 21:30:35 | [018][000/179] predict_x0_loss: 0.029 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:30:36 | [018][010/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 21:30:40 | [018][020/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:30:42 | [018][030/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0068 ntime: 0084 mem: 3.36 + 04-03 21:30:43 | [018][040/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:30:44 | [018][050/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0046 ntime: 0081 mem: 3.36 + 04-03 21:30:45 | [018][060/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0053 ntime: 0076 mem: 3.36 + 04-03 21:30:47 | [018][070/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0053 ntime: 0079 mem: 3.36 + 04-03 21:30:48 | [018][080/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:30:49 | [018][090/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 21:30:51 | [018][100/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0071 ntime: 0079 mem: 3.36 + 04-03 21:30:52 | [018][110/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0061 ntime: 0080 mem: 3.36 + 04-03 21:30:53 | [018][120/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 21:30:55 | [018][130/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0056 ntime: 0084 mem: 3.36 + 04-03 21:30:56 | [018][140/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0051 ntime: 0088 mem: 3.36 + 04-03 21:30:58 | [018][150/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0063 ntime: 0077 mem: 3.36 + 04-03 21:30:59 | [018][160/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 21:31:00 | [018][170/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0070 ntime: 0083 mem: 3.36 + 04-03 21:31:02 | Time info >>>> elapsed: 7.65 mins remain: 395.12 mins + 04-03 21:31:02 | [019][000/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0051 ntime: 0095 mem: 3.36 + 04-03 21:31:03 | [019][010/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0055 ntime: 0082 mem: 3.36 + 04-03 21:31:04 | [019][020/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0059 ntime: 0081 mem: 3.36 + 04-03 21:31:06 | [019][030/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0054 ntime: 0086 mem: 3.36 + 04-03 21:31:07 | [019][040/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 21:31:09 | [019][050/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 21:31:10 | [019][060/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 21:31:11 | [019][070/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0054 ntime: 0082 mem: 3.36 + 04-03 21:31:13 | [019][080/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0085 ntime: 0086 mem: 3.36 + 04-03 21:31:14 | [019][090/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0056 ntime: 0077 mem: 3.36 + 04-03 21:31:16 | [019][100/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0056 ntime: 0081 mem: 3.36 + 04-03 21:31:17 | [019][110/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 21:31:18 | [019][120/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0057 ntime: 0085 mem: 3.36 + 04-03 21:31:20 | [019][130/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 21:31:21 | [019][140/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0052 ntime: 0088 mem: 3.36 + 04-03 21:31:22 | [019][150/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 21:31:24 | [019][160/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:31:25 | [019][170/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0055 ntime: 0075 mem: 3.36 + 04-03 21:31:26 | Time info >>>> elapsed: 8.06 mins remain: 394.93 mins + 04-03 21:31:26 | [020][000/179] predict_x0_loss: 0.028 glr: 5.0e-05 dtime: 0063 ntime: 0093 mem: 3.36 + 04-03 21:31:27 | [020][010/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0045 ntime: 0082 mem: 3.36 + 04-03 21:31:29 | [020][020/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:31:30 | [020][030/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0055 ntime: 0077 mem: 3.36 + 04-03 21:31:31 | [020][040/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 21:31:33 | [020][050/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 21:31:34 | [020][060/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0058 ntime: 0078 mem: 3.36 + 04-03 21:31:35 | [020][070/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:31:37 | [020][080/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 21:31:38 | [020][090/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 21:31:39 | [020][100/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:31:41 | [020][110/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 21:31:42 | [020][120/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 21:31:43 | [020][130/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 21:31:45 | [020][140/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0053 ntime: 0085 mem: 3.36 + 04-03 21:31:46 | [020][150/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0048 ntime: 0069 mem: 3.36 + 04-03 21:31:47 | [020][160/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 21:31:49 | [020][170/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0053 ntime: 0086 mem: 3.36 + 04-03 21:31:50 | Time info >>>> elapsed: 8.46 mins remain: 394.25 mins + 04-03 21:31:50 | [021][000/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 21:31:51 | [021][010/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 21:31:53 | [021][020/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 21:31:54 | [021][030/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0054 ntime: 0074 mem: 3.36 + 04-03 21:31:55 | [021][040/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 21:31:57 | [021][050/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 21:31:58 | [021][060/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0052 ntime: 0073 mem: 3.36 + 04-03 21:31:59 | [021][070/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0056 ntime: 0081 mem: 3.36 + 04-03 21:32:01 | [021][080/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 21:32:02 | [021][090/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0048 ntime: 0086 mem: 3.36 + 04-03 21:32:03 | [021][100/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0055 ntime: 0084 mem: 3.36 + 04-03 21:32:05 | [021][110/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0062 ntime: 0086 mem: 3.36 + 04-03 21:32:06 | [021][120/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0051 ntime: 0087 mem: 3.36 + 04-03 21:32:07 | [021][130/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0049 ntime: 0086 mem: 3.36 + 04-03 21:32:09 | [021][140/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 21:32:10 | [021][150/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0047 ntime: 0085 mem: 3.36 + 04-03 21:32:11 | [021][160/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 21:32:13 | [021][170/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:32:14 | Time info >>>> elapsed: 8.85 mins remain: 393.59 mins + 04-03 21:32:14 | [022][000/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0051 ntime: 0088 mem: 3.36 + 04-03 21:32:15 | [022][010/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:32:17 | [022][020/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:32:18 | [022][030/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0045 ntime: 0075 mem: 3.36 + 04-03 21:32:19 | [022][040/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:32:20 | [022][050/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0043 ntime: 0080 mem: 3.36 + 04-03 21:32:22 | [022][060/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 21:32:23 | [022][070/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0046 ntime: 0078 mem: 3.36 + 04-03 21:32:24 | [022][080/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0046 ntime: 0082 mem: 3.36 + 04-03 21:32:26 | [022][090/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0046 ntime: 0084 mem: 3.36 + 04-03 21:32:27 | [022][100/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 21:32:28 | [022][110/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0045 ntime: 0076 mem: 3.36 + 04-03 21:32:29 | [022][120/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 21:32:31 | [022][130/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:32:32 | [022][140/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0052 ntime: 0073 mem: 3.36 + 04-03 21:32:33 | [022][150/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 21:32:35 | [022][160/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 21:32:36 | [022][170/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0047 ntime: 0084 mem: 3.36 + 04-03 21:32:37 | Time info >>>> elapsed: 9.25 mins remain: 392.73 mins + 04-03 21:32:37 | [023][000/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0052 ntime: 0074 mem: 3.36 + 04-03 21:32:39 | [023][010/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 21:32:40 | [023][020/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 21:32:41 | [023][030/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0057 ntime: 0087 mem: 3.36 + 04-03 21:32:43 | [023][040/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 21:32:44 | [023][050/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 21:32:45 | [023][060/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0054 ntime: 0087 mem: 3.36 + 04-03 21:32:47 | [023][070/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0053 ntime: 0074 mem: 3.36 + 04-03 21:32:48 | [023][080/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0053 ntime: 0073 mem: 3.36 + 04-03 21:32:49 | [023][090/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 21:32:51 | [023][100/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 21:32:52 | [023][110/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 21:32:53 | [023][120/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0056 ntime: 0085 mem: 3.36 + 04-03 21:32:55 | [023][130/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:32:56 | [023][140/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0046 ntime: 0080 mem: 3.36 + 04-03 21:32:57 | [023][150/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 21:32:59 | [023][160/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0054 ntime: 0093 mem: 3.36 + 04-03 21:33:00 | [023][170/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0057 ntime: 0083 mem: 3.36 + 04-03 21:33:01 | Time info >>>> elapsed: 9.65 mins remain: 392.36 mins + 04-03 21:33:01 | [024][000/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0051 ntime: 0088 mem: 3.36 + 04-03 21:33:03 | [024][010/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 21:33:04 | [024][020/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0056 ntime: 0083 mem: 3.36 + 04-03 21:33:05 | [024][030/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0056 ntime: 0079 mem: 3.36 + 04-03 21:33:07 | [024][040/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0059 ntime: 0094 mem: 3.36 + 04-03 21:33:08 | [024][050/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 21:33:09 | [024][060/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 21:33:11 | [024][070/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0084 ntime: 0082 mem: 3.36 + 04-03 21:33:12 | [024][080/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:33:14 | [024][090/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 21:33:15 | [024][100/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0052 ntime: 0092 mem: 3.36 + 04-03 21:33:16 | [024][110/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 21:33:17 | [024][120/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0053 ntime: 0079 mem: 3.36 + 04-03 21:33:19 | [024][130/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:33:20 | [024][140/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 21:33:21 | [024][150/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 21:33:23 | [024][160/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 21:33:24 | [024][170/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 21:33:25 | Time info >>>> elapsed: 10.05 mins remain: 391.79 mins + 04-03 21:33:25 | [025][000/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 21:33:27 | [025][010/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:33:28 | [025][020/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 21:33:29 | [025][030/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 21:33:31 | [025][040/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 21:33:32 | [025][050/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:33:33 | [025][060/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 21:33:34 | [025][070/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 21:33:36 | [025][080/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:33:37 | [025][090/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0055 ntime: 0077 mem: 3.36 + 04-03 21:33:38 | [025][100/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0046 ntime: 0081 mem: 3.36 + 04-03 21:33:40 | [025][110/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0054 ntime: 0087 mem: 3.36 + 04-03 21:33:41 | [025][120/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 21:33:42 | [025][130/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0046 ntime: 0078 mem: 3.36 + 04-03 21:33:44 | [025][140/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 21:33:45 | [025][150/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0048 ntime: 0085 mem: 3.36 + 04-03 21:33:46 | [025][160/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:33:48 | [025][170/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0045 ntime: 0084 mem: 3.36 + 04-03 21:33:49 | Time info >>>> elapsed: 10.44 mins remain: 391.08 mins + 04-03 21:33:49 | [026][000/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:33:50 | [026][010/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:33:51 | [026][020/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0045 ntime: 0079 mem: 3.36 + 04-03 21:33:53 | [026][030/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:33:54 | [026][040/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:33:55 | [026][050/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 21:33:57 | [026][060/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 21:33:58 | [026][070/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 21:33:59 | [026][080/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0047 ntime: 0074 mem: 3.36 + 04-03 21:34:00 | [026][090/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0055 ntime: 0072 mem: 3.36 + 04-03 21:34:02 | [026][100/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 21:34:03 | [026][110/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 21:34:04 | [026][120/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 21:34:05 | [026][130/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:34:07 | [026][140/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 21:34:08 | [026][150/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:34:09 | [026][160/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 21:34:11 | [026][170/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 21:34:12 | Time info >>>> elapsed: 10.82 mins remain: 390.09 mins + 04-03 21:34:12 | [027][000/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0053 ntime: 0084 mem: 3.36 + 04-03 21:34:13 | [027][010/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:34:15 | [027][020/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:34:16 | [027][030/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0050 ntime: 0089 mem: 3.36 + 04-03 21:34:17 | [027][040/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0049 ntime: 0071 mem: 3.36 + 04-03 21:34:19 | [027][050/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0046 ntime: 0079 mem: 3.36 + 04-03 21:34:20 | [027][060/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0045 ntime: 0075 mem: 3.36 + 04-03 21:34:21 | [027][070/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:34:22 | [027][080/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 21:34:24 | [027][090/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 21:34:25 | [027][100/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 21:34:26 | [027][110/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:34:27 | [027][120/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 21:34:29 | [027][130/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 21:34:30 | [027][140/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0045 ntime: 0074 mem: 3.36 + 04-03 21:34:31 | [027][150/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 21:34:32 | [027][160/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0047 ntime: 0074 mem: 3.36 + 04-03 21:34:34 | [027][170/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0046 ntime: 0079 mem: 3.36 + 04-03 21:34:35 | Time info >>>> elapsed: 11.21 mins remain: 389.00 mins + 04-03 21:34:35 | [028][000/179] predict_x0_loss: 0.026 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:34:36 | [028][010/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0094 ntime: 0077 mem: 3.36 + 04-03 21:34:37 | [028][020/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 21:34:39 | [028][030/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0052 ntime: 0088 mem: 3.36 + 04-03 21:34:40 | [028][040/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 21:34:42 | [028][050/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 21:34:43 | [028][060/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 21:34:44 | [028][070/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0061 ntime: 0084 mem: 3.36 + 04-03 21:34:46 | [028][080/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 21:34:47 | [028][090/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 21:34:48 | [028][100/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:34:50 | [028][110/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 21:34:51 | [028][120/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0047 ntime: 0083 mem: 3.36 + 04-03 21:34:52 | [028][130/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 21:34:54 | [028][140/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 21:34:55 | [028][150/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:34:56 | [028][160/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 21:34:58 | [028][170/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0048 ntime: 0088 mem: 3.36 + 04-03 21:34:59 | Time info >>>> elapsed: 11.61 mins remain: 388.60 mins + 04-03 21:34:59 | [029][000/179] predict_x0_loss: 0.027 glr: 5.0e-05 dtime: 0047 ntime: 0085 mem: 3.36 + 04-03 21:35:00 | [029][010/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0047 ntime: 0088 mem: 3.36 + 04-03 21:35:02 | [029][020/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:35:03 | [029][030/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 21:35:04 | [029][040/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:35:06 | [029][050/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:35:07 | [029][060/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 21:35:08 | [029][070/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0046 ntime: 0084 mem: 3.36 + 04-03 21:35:09 | [029][080/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:35:11 | [029][090/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0053 ntime: 0085 mem: 3.36 + 04-03 21:35:12 | [029][100/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:35:14 | [029][110/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:35:15 | [029][120/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 21:35:16 | [029][130/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0045 ntime: 0076 mem: 3.36 + 04-03 21:35:17 | [029][140/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0044 ntime: 0072 mem: 3.36 + 04-03 21:35:19 | [029][150/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:35:20 | [029][160/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0062 ntime: 0086 mem: 3.36 + 04-03 21:35:21 | [029][170/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 21:35:23 | Time info >>>> elapsed: 12.00 mins remain: 388.10 mins + 04-03 21:35:23 | [030][000/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0063 ntime: 0080 mem: 3.36 + 04-03 21:35:24 | [030][010/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:35:25 | [030][020/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0050 ntime: 0073 mem: 3.36 + 04-03 21:35:26 | [030][030/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 21:35:28 | [030][040/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 21:35:29 | [030][050/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 21:35:30 | [030][060/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:35:32 | [030][070/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:35:33 | [030][080/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0049 ntime: 0070 mem: 3.36 + 04-03 21:35:34 | [030][090/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0053 ntime: 0075 mem: 3.36 + 04-03 21:35:36 | [030][100/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0051 ntime: 0087 mem: 3.36 + 04-03 21:35:37 | [030][110/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 21:35:38 | [030][120/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0050 ntime: 0086 mem: 3.36 + 04-03 21:35:40 | [030][130/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 21:35:41 | [030][140/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0048 ntime: 0073 mem: 3.36 + 04-03 21:35:42 | [030][150/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 21:35:43 | [030][160/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 21:35:45 | [030][170/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 21:35:46 | Time info >>>> elapsed: 12.39 mins remain: 387.25 mins + 04-03 21:35:46 | [031][000/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0045 ntime: 0083 mem: 3.36 + 04-03 21:35:47 | [031][010/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 21:35:48 | [031][020/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 21:35:50 | [031][030/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 21:35:51 | [031][040/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 21:35:52 | [031][050/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:35:54 | [031][060/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:35:55 | [031][070/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 21:35:57 | [031][080/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0046 ntime: 0075 mem: 3.36 + 04-03 21:35:58 | [031][090/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 21:35:59 | [031][100/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 21:36:00 | [031][110/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 21:36:02 | [031][120/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:36:03 | [031][130/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 21:36:04 | [031][140/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:36:06 | [031][150/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 21:36:07 | [031][160/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0054 ntime: 0083 mem: 3.36 + 04-03 21:36:08 | [031][170/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0054 ntime: 0082 mem: 3.36 + 04-03 21:36:09 | Time info >>>> elapsed: 12.78 mins remain: 386.72 mins + 04-03 21:36:10 | [032][000/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 21:36:11 | [032][010/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 21:36:12 | [032][020/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 21:36:13 | [032][030/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0063 ntime: 0080 mem: 3.36 + 04-03 21:36:15 | [032][040/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 21:36:16 | [032][050/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 21:36:17 | [032][060/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 21:36:19 | [032][070/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 21:36:20 | [032][080/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 21:36:21 | [032][090/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 21:36:22 | [032][100/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:36:24 | [032][110/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 21:36:25 | [032][120/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0053 ntime: 0076 mem: 3.36 + 04-03 21:36:26 | [032][130/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0054 ntime: 0075 mem: 3.36 + 04-03 21:36:28 | [032][140/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0057 ntime: 0079 mem: 3.36 + 04-03 21:36:29 | [032][150/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0068 ntime: 0082 mem: 3.36 + 04-03 21:36:31 | [032][160/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 21:36:32 | [032][170/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0070 ntime: 0084 mem: 3.36 + 04-03 21:36:33 | Time info >>>> elapsed: 13.18 mins remain: 386.23 mins + 04-03 21:36:33 | [033][000/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0053 ntime: 0071 mem: 3.36 + 04-03 21:36:35 | [033][010/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0060 ntime: 0081 mem: 3.36 + 04-03 21:36:36 | [033][020/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0066 ntime: 0081 mem: 3.36 + 04-03 21:36:38 | [033][030/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0070 ntime: 0080 mem: 3.36 + 04-03 21:36:39 | [033][040/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0056 ntime: 0071 mem: 3.36 + 04-03 21:36:41 | [033][050/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0059 ntime: 0078 mem: 3.36 + 04-03 21:36:42 | [033][060/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0080 ntime: 0077 mem: 3.36 + 04-03 21:36:44 | [033][070/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0067 ntime: 0078 mem: 3.36 + 04-03 21:36:45 | [033][080/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 21:36:46 | [033][090/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 21:36:48 | [033][100/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0067 ntime: 0085 mem: 3.36 + 04-03 21:36:49 | [033][110/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0073 ntime: 0078 mem: 3.36 + 04-03 21:36:51 | [033][120/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0063 ntime: 0084 mem: 3.36 + 04-03 21:36:52 | [033][130/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0073 ntime: 0076 mem: 3.36 + 04-03 21:36:53 | [033][140/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0055 ntime: 0082 mem: 3.36 + 04-03 21:36:55 | [033][150/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0065 ntime: 0066 mem: 3.36 + 04-03 21:36:56 | [033][160/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0052 ntime: 0074 mem: 3.36 + 04-03 21:36:57 | [033][170/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 21:36:58 | Time info >>>> elapsed: 13.60 mins remain: 386.47 mins + 04-03 21:36:59 | [034][000/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0046 ntime: 0081 mem: 3.36 + 04-03 21:37:00 | [034][010/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0047 ntime: 0073 mem: 3.36 + 04-03 21:37:01 | [034][020/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 21:37:03 | [034][030/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 21:37:04 | [034][040/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0052 ntime: 0090 mem: 3.36 + 04-03 21:37:05 | [034][050/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:37:07 | [034][060/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 21:37:08 | [034][070/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0055 ntime: 0077 mem: 3.36 + 04-03 21:37:09 | [034][080/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 21:37:11 | [034][090/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0065 ntime: 0081 mem: 3.36 + 04-03 21:37:12 | [034][100/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0058 ntime: 0080 mem: 3.36 + 04-03 21:37:14 | [034][110/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0054 ntime: 0087 mem: 3.36 + 04-03 21:37:15 | [034][120/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 21:37:16 | [034][130/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:37:18 | [034][140/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:37:19 | [034][150/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:37:20 | [034][160/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:37:22 | [034][170/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 21:37:23 | Time info >>>> elapsed: 14.01 mins remain: 386.16 mins + 04-03 21:37:23 | [035][000/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:37:24 | [035][010/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0048 ntime: 0073 mem: 3.36 + 04-03 21:37:25 | [035][020/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0047 ntime: 0083 mem: 3.36 + 04-03 21:37:27 | [035][030/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 21:37:28 | [035][040/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0045 ntime: 0080 mem: 3.36 + 04-03 21:37:29 | [035][050/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 21:37:30 | [035][060/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0056 ntime: 0083 mem: 3.36 + 04-03 21:37:32 | [035][070/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0062 ntime: 0080 mem: 3.36 + 04-03 21:37:33 | [035][080/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0043 ntime: 0072 mem: 3.36 + 04-03 21:37:34 | [035][090/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 21:37:36 | [035][100/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:37:37 | [035][110/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0070 ntime: 0082 mem: 3.36 + 04-03 21:37:39 | [035][120/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0062 ntime: 0085 mem: 3.36 + 04-03 21:37:40 | [035][130/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:37:41 | [035][140/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 21:37:43 | [035][150/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 21:37:44 | [035][160/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0086 mem: 3.36 + 04-03 21:37:45 | [035][170/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:37:46 | Time info >>>> elapsed: 14.40 mins remain: 385.57 mins + 04-03 21:37:46 | [036][000/179] predict_x0_loss: 0.024 glr: 5.0e-05 dtime: 0054 ntime: 0076 mem: 3.36 + 04-03 21:37:48 | [036][010/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 21:37:49 | [036][020/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 21:37:50 | [036][030/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:37:52 | [036][040/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0059 ntime: 0084 mem: 3.36 + 04-03 21:37:53 | [036][050/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0058 ntime: 0082 mem: 3.36 + 04-03 21:37:54 | [036][060/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0055 ntime: 0072 mem: 3.36 + 04-03 21:37:56 | [036][070/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0050 ntime: 0070 mem: 3.36 + 04-03 21:37:57 | [036][080/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0047 ntime: 0070 mem: 3.36 + 04-03 21:37:58 | [036][090/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 21:37:59 | [036][100/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 21:38:01 | [036][110/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 21:38:02 | [036][120/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0047 ntime: 0083 mem: 3.36 + 04-03 21:38:03 | [036][130/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:38:05 | [036][140/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0048 ntime: 0073 mem: 3.36 + 04-03 21:38:06 | [036][150/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 21:38:07 | [036][160/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 21:38:09 | [036][170/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:38:10 | Time info >>>> elapsed: 14.79 mins remain: 384.90 mins + 04-03 21:38:10 | [037][000/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0045 ntime: 0078 mem: 3.36 + 04-03 21:38:11 | [037][010/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:38:12 | [037][020/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 21:38:14 | [037][030/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 21:38:15 | [037][040/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:38:16 | [037][050/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0049 ntime: 0071 mem: 3.36 + 04-03 21:38:18 | [037][060/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0056 ntime: 0079 mem: 3.36 + 04-03 21:38:19 | [037][070/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:38:20 | [037][080/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:38:21 | [037][090/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0047 ntime: 0084 mem: 3.36 + 04-03 21:38:23 | [037][100/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 21:38:24 | [037][110/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0057 ntime: 0091 mem: 3.36 + 04-03 21:38:26 | [037][120/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0046 ntime: 0078 mem: 3.36 + 04-03 21:38:27 | [037][130/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0053 ntime: 0074 mem: 3.36 + 04-03 21:38:28 | [037][140/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 21:38:29 | [037][150/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 21:38:31 | [037][160/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 21:38:32 | [037][170/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 21:38:33 | Time info >>>> elapsed: 15.18 mins remain: 384.21 mins + 04-03 21:38:33 | [038][000/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 21:38:34 | [038][010/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0046 ntime: 0078 mem: 3.36 + 04-03 21:38:36 | [038][020/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:38:37 | [038][030/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 21:38:38 | [038][040/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 21:38:40 | [038][050/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0053 ntime: 0092 mem: 3.36 + 04-03 21:38:41 | [038][060/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:38:42 | [038][070/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:38:43 | [038][080/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0047 ntime: 0071 mem: 3.36 + 04-03 21:38:45 | [038][090/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 21:38:46 | [038][100/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 21:38:47 | [038][110/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 21:38:49 | [038][120/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 21:38:50 | [038][130/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 21:38:51 | [038][140/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 21:38:53 | [038][150/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 21:38:54 | [038][160/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:38:55 | [038][170/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0052 ntime: 0092 mem: 3.36 + 04-03 21:38:56 | Time info >>>> elapsed: 15.57 mins remain: 383.57 mins + 04-03 21:38:56 | [039][000/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0046 ntime: 0081 mem: 3.36 + 04-03 21:38:58 | [039][010/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 21:38:59 | [039][020/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0051 ntime: 0086 mem: 3.36 + 04-03 21:39:01 | [039][030/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0050 ntime: 0086 mem: 3.36 + 04-03 21:39:02 | [039][040/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0047 ntime: 0085 mem: 3.36 + 04-03 21:39:03 | [039][050/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 21:39:05 | [039][060/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0091 ntime: 0088 mem: 3.36 + 04-03 21:39:06 | [039][070/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0076 ntime: 0085 mem: 3.36 + 04-03 21:39:08 | [039][080/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0058 ntime: 0080 mem: 3.36 + 04-03 21:39:09 | [039][090/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:39:10 | [039][100/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0071 ntime: 0080 mem: 3.36 + 04-03 21:39:12 | [039][110/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:39:13 | [039][120/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0058 ntime: 0080 mem: 3.36 + 04-03 21:39:14 | [039][130/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0056 ntime: 0084 mem: 3.36 + 04-03 21:39:16 | [039][140/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0058 ntime: 0084 mem: 3.36 + 04-03 21:39:17 | [039][150/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:39:18 | [039][160/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0065 ntime: 0077 mem: 3.36 + 04-03 21:39:20 | [039][170/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0104 ntime: 0079 mem: 3.36 + 04-03 21:39:21 | Time info >>>> elapsed: 15.98 mins remain: 383.41 mins + 04-03 21:39:21 | [040][000/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0054 ntime: 0073 mem: 3.36 + 04-03 21:39:22 | [040][010/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 21:39:24 | [040][020/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0053 ntime: 0076 mem: 3.36 + 04-03 21:39:25 | [040][030/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0057 ntime: 0081 mem: 3.36 + 04-03 21:39:26 | [040][040/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0067 ntime: 0084 mem: 3.36 + 04-03 21:39:28 | [040][050/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 21:39:29 | [040][060/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:39:30 | [040][070/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 21:39:32 | [040][080/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 21:39:33 | [040][090/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 21:39:35 | [040][100/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0065 ntime: 0086 mem: 3.36 + 04-03 21:39:36 | [040][110/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 21:39:37 | [040][120/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:39:39 | [040][130/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 21:39:40 | [040][140/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0059 ntime: 0078 mem: 3.36 + 04-03 21:39:41 | [040][150/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 21:39:43 | [040][160/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 21:39:44 | [040][170/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 21:39:45 | Time info >>>> elapsed: 16.38 mins remain: 383.17 mins + 04-03 21:39:45 | [041][000/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 21:39:47 | [041][010/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 21:39:48 | [041][020/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 21:39:49 | [041][030/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 21:39:51 | [041][040/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0055 ntime: 0084 mem: 3.36 + 04-03 21:39:52 | [041][050/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 21:39:53 | [041][060/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0065 ntime: 0082 mem: 3.36 + 04-03 21:39:55 | [041][070/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0061 ntime: 0076 mem: 3.36 + 04-03 21:39:56 | [041][080/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0052 ntime: 0073 mem: 3.36 + 04-03 21:39:57 | [041][090/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 21:39:59 | [041][100/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0052 ntime: 0069 mem: 3.36 + 04-03 21:40:00 | [041][110/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0054 ntime: 0089 mem: 3.36 + 04-03 21:40:01 | [041][120/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0075 ntime: 0079 mem: 3.36 + 04-03 21:40:03 | [041][130/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 21:40:04 | [041][140/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 21:40:05 | [041][150/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 21:40:07 | [041][160/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0070 ntime: 0082 mem: 3.36 + 04-03 21:40:08 | [041][170/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0057 ntime: 0078 mem: 3.36 + 04-03 21:40:09 | Time info >>>> elapsed: 16.78 mins remain: 382.73 mins + 04-03 21:40:09 | [042][000/179] predict_x0_loss: 0.025 glr: 5.0e-05 dtime: 0060 ntime: 0076 mem: 3.36 + 04-03 21:40:11 | [042][010/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 21:40:12 | [042][020/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 21:40:13 | [042][030/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:40:14 | [042][040/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 21:40:16 | [042][050/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 21:40:17 | [042][060/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0062 ntime: 0086 mem: 3.36 + 04-03 21:40:19 | [042][070/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0068 ntime: 0080 mem: 3.36 + 04-03 21:40:20 | [042][080/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0070 ntime: 0083 mem: 3.36 + 04-03 21:40:22 | [042][090/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0071 ntime: 0088 mem: 3.36 + 04-03 21:40:23 | [042][100/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 21:40:24 | [042][110/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0056 ntime: 0082 mem: 3.36 + 04-03 21:40:26 | [042][120/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0048 ntime: 0086 mem: 3.36 + 04-03 21:40:27 | [042][130/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0058 ntime: 0060 mem: 3.36 + 04-03 21:40:28 | [042][140/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 21:40:29 | [042][150/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0057 ntime: 0077 mem: 3.36 + 04-03 21:40:31 | [042][160/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 21:40:32 | [042][170/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0046 ntime: 0075 mem: 3.36 + 04-03 21:40:33 | Time info >>>> elapsed: 17.18 mins remain: 382.31 mins + 04-03 21:40:33 | [043][000/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0046 ntime: 0075 mem: 3.36 + 04-03 21:40:34 | [043][010/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 21:40:36 | [043][020/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0046 ntime: 0080 mem: 3.36 + 04-03 21:40:37 | [043][030/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0044 ntime: 0072 mem: 3.36 + 04-03 21:40:38 | [043][040/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0046 ntime: 0086 mem: 3.36 + 04-03 21:40:39 | [043][050/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0044 ntime: 0079 mem: 3.36 + 04-03 21:40:41 | [043][060/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:40:42 | [043][070/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 21:40:43 | [043][080/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0088 mem: 3.36 + 04-03 21:40:45 | [043][090/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0046 ntime: 0083 mem: 3.36 + 04-03 21:40:46 | [043][100/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:40:47 | [043][110/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0046 ntime: 0072 mem: 3.36 + 04-03 21:40:48 | [043][120/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 21:40:50 | [043][130/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0046 ntime: 0075 mem: 3.36 + 04-03 21:40:51 | [043][140/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:40:52 | [043][150/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 21:40:53 | [043][160/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0052 ntime: 0073 mem: 3.36 + 04-03 21:40:55 | [043][170/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 21:40:56 | Time info >>>> elapsed: 17.56 mins remain: 381.46 mins + 04-03 21:40:56 | [044][000/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0048 ntime: 0085 mem: 3.36 + 04-03 21:40:57 | [044][010/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:40:58 | [044][020/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 21:41:00 | [044][030/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 21:41:01 | [044][040/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 21:41:02 | [044][050/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0080 ntime: 0089 mem: 3.36 + 04-03 21:41:04 | [044][060/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0046 ntime: 0087 mem: 3.36 + 04-03 21:41:05 | [044][070/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0054 ntime: 0086 mem: 3.36 + 04-03 21:41:06 | [044][080/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0057 ntime: 0082 mem: 3.36 + 04-03 21:41:08 | [044][090/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0053 ntime: 0071 mem: 3.36 + 04-03 21:41:09 | [044][100/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 21:41:10 | [044][110/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 21:41:12 | [044][120/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0047 ntime: 0073 mem: 3.36 + 04-03 21:41:13 | [044][130/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 21:41:14 | [044][140/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:41:16 | [044][150/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0054 ntime: 0085 mem: 3.36 + 04-03 21:41:17 | [044][160/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0087 mem: 3.36 + 04-03 21:41:18 | [044][170/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0046 ntime: 0079 mem: 3.36 + 04-03 21:41:19 | Time info >>>> elapsed: 17.95 mins remain: 380.93 mins + 04-03 21:41:19 | [045][000/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 21:41:21 | [045][010/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0047 ntime: 0085 mem: 3.36 + 04-03 21:41:22 | [045][020/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 21:41:23 | [045][030/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 21:41:25 | [045][040/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0063 ntime: 0082 mem: 3.36 + 04-03 21:41:26 | [045][050/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0056 ntime: 0078 mem: 3.36 + 04-03 21:41:28 | [045][060/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0070 ntime: 0079 mem: 3.36 + 04-03 21:41:29 | [045][070/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0075 ntime: 0078 mem: 3.36 + 04-03 21:41:31 | [045][080/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0057 ntime: 0079 mem: 3.36 + 04-03 21:41:32 | [045][090/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0074 ntime: 0083 mem: 3.36 + 04-03 21:41:33 | [045][100/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:41:35 | [045][110/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0047 ntime: 0085 mem: 3.36 + 04-03 21:41:36 | [045][120/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:41:37 | [045][130/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 21:41:38 | [045][140/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:41:40 | [045][150/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 21:41:41 | [045][160/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:41:42 | [045][170/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 21:41:43 | Time info >>>> elapsed: 18.35 mins remain: 380.57 mins + 04-03 21:41:43 | [046][000/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0044 ntime: 0076 mem: 3.36 + 04-03 21:41:45 | [046][010/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0048 ntime: 0072 mem: 3.36 + 04-03 21:41:46 | [046][020/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0069 mem: 3.36 + 04-03 21:41:47 | [046][030/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 21:41:49 | [046][040/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 21:41:50 | [046][050/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 21:41:51 | [046][060/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0054 ntime: 0073 mem: 3.36 + 04-03 21:41:52 | [046][070/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 21:41:54 | [046][080/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 21:41:55 | [046][090/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 21:41:56 | [046][100/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:41:58 | [046][110/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0047 ntime: 0072 mem: 3.36 + 04-03 21:41:59 | [046][120/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:42:00 | [046][130/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:42:02 | [046][140/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:42:03 | [046][150/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0044 ntime: 0083 mem: 3.36 + 04-03 21:42:04 | [046][160/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:42:05 | [046][170/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:42:06 | Time info >>>> elapsed: 18.73 mins remain: 379.88 mins + 04-03 21:42:07 | [047][000/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:42:08 | [047][010/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 21:42:09 | [047][020/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:42:11 | [047][030/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0062 ntime: 0080 mem: 3.36 + 04-03 21:42:12 | [047][040/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 21:42:13 | [047][050/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:42:15 | [047][060/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 21:42:16 | [047][070/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0045 ntime: 0088 mem: 3.36 + 04-03 21:42:17 | [047][080/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0053 ntime: 0072 mem: 3.36 + 04-03 21:42:19 | [047][090/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0045 ntime: 0081 mem: 3.36 + 04-03 21:42:20 | [047][100/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:42:21 | [047][110/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0046 ntime: 0081 mem: 3.36 + 04-03 21:42:22 | [047][120/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0046 ntime: 0074 mem: 3.36 + 04-03 21:42:24 | [047][130/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 21:42:25 | [047][140/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 21:42:26 | [047][150/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:42:28 | [047][160/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:42:29 | [047][170/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:42:30 | Time info >>>> elapsed: 19.12 mins remain: 379.31 mins + 04-03 21:42:30 | [048][000/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:42:31 | [048][010/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 21:42:33 | [048][020/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0057 ntime: 0078 mem: 3.36 + 04-03 21:42:34 | [048][030/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:42:35 | [048][040/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:42:36 | [048][050/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 21:42:38 | [048][060/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 21:42:39 | [048][070/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0072 mem: 3.36 + 04-03 21:42:40 | [048][080/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 21:42:42 | [048][090/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 21:42:43 | [048][100/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:42:44 | [048][110/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 21:42:46 | [048][120/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0054 ntime: 0083 mem: 3.36 + 04-03 21:42:47 | [048][130/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0048 ntime: 0090 mem: 3.36 + 04-03 21:42:48 | [048][140/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:42:50 | [048][150/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 21:42:51 | [048][160/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 21:42:52 | [048][170/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:42:53 | Time info >>>> elapsed: 19.52 mins remain: 378.75 mins + 04-03 21:42:53 | [049][000/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0045 ntime: 0076 mem: 3.36 + 04-03 21:42:55 | [049][010/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0052 ntime: 0089 mem: 3.36 + 04-03 21:42:56 | [049][020/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 21:42:57 | [049][030/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0051 ntime: 0086 mem: 3.36 + 04-03 21:42:59 | [049][040/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0089 mem: 3.36 + 04-03 21:43:00 | [049][050/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0052 ntime: 0092 mem: 3.36 + 04-03 21:43:01 | [049][060/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:43:03 | [049][070/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 21:43:04 | [049][080/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 21:43:05 | [049][090/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0053 ntime: 0076 mem: 3.36 + 04-03 21:43:07 | [049][100/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0048 ntime: 0085 mem: 3.36 + 04-03 21:43:08 | [049][110/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0050 ntime: 0087 mem: 3.36 + 04-03 21:43:09 | [049][120/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0053 ntime: 0089 mem: 3.36 + 04-03 21:43:11 | [049][130/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0054 ntime: 0074 mem: 3.36 + 04-03 21:43:12 | [049][140/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 21:43:14 | [049][150/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0048 ntime: 0090 mem: 3.36 + 04-03 21:43:15 | [049][160/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0052 ntime: 0087 mem: 3.36 + 04-03 21:43:16 | [049][170/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0050 ntime: 0088 mem: 3.36 + 04-03 21:43:17 | Time info >>>> elapsed: 19.92 mins remain: 378.44 mins + 04-03 21:43:18 | [050][000/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0057 ntime: 0073 mem: 3.36 + 04-03 21:43:19 | [050][010/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 21:43:20 | [050][020/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0048 ntime: 0073 mem: 3.36 + 04-03 21:43:21 | [050][030/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 21:43:23 | [050][040/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:43:24 | [050][050/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 21:43:25 | [050][060/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 21:43:27 | [050][070/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0071 mem: 3.36 + 04-03 21:43:28 | [050][080/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:43:29 | [050][090/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:43:31 | [050][100/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 21:43:32 | [050][110/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:43:33 | [050][120/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 21:43:34 | [050][130/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:43:36 | [050][140/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:43:37 | [050][150/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:43:38 | [050][160/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 21:43:40 | [050][170/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:43:41 | Time info >>>> elapsed: 20.30 mins remain: 377.80 mins + 04-03 21:43:41 | [051][000/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 21:43:42 | [051][010/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0053 ntime: 0076 mem: 3.36 + 04-03 21:43:43 | [051][020/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 21:43:45 | [051][030/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:43:46 | [051][040/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 21:43:47 | [051][050/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:43:48 | [051][060/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 21:43:50 | [051][070/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:43:51 | [051][080/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0050 ntime: 0086 mem: 3.36 + 04-03 21:43:52 | [051][090/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 21:43:54 | [051][100/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 21:43:55 | [051][110/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0058 ntime: 0084 mem: 3.36 + 04-03 21:43:56 | [051][120/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:43:58 | [051][130/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0046 ntime: 0060 mem: 3.36 + 04-03 21:43:59 | [051][140/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 21:44:00 | [051][150/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 21:44:02 | [051][160/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0046 ntime: 0074 mem: 3.36 + 04-03 21:44:03 | [051][170/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 21:44:04 | Time info >>>> elapsed: 20.69 mins remain: 377.23 mins + 04-03 21:44:04 | [052][000/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0055 ntime: 0075 mem: 3.36 + 04-03 21:44:05 | [052][010/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0054 ntime: 0077 mem: 3.36 + 04-03 21:44:07 | [052][020/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0057 ntime: 0077 mem: 3.36 + 04-03 21:44:08 | [052][030/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0059 ntime: 0079 mem: 3.36 + 04-03 21:44:09 | [052][040/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0067 ntime: 0085 mem: 3.36 + 04-03 21:44:11 | [052][050/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:44:12 | [052][060/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0052 ntime: 0074 mem: 3.36 + 04-03 21:44:13 | [052][070/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0053 ntime: 0089 mem: 3.36 + 04-03 21:44:15 | [052][080/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0045 ntime: 0074 mem: 3.36 + 04-03 21:44:16 | [052][090/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 21:44:17 | [052][100/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0045 ntime: 0076 mem: 3.36 + 04-03 21:44:18 | [052][110/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0046 ntime: 0078 mem: 3.36 + 04-03 21:44:20 | [052][120/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0062 ntime: 0080 mem: 3.36 + 04-03 21:44:21 | [052][130/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 21:44:22 | [052][140/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 21:44:24 | [052][150/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:44:25 | [052][160/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 21:44:26 | [052][170/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0072 ntime: 0084 mem: 3.36 + 04-03 21:44:28 | Time info >>>> elapsed: 21.09 mins remain: 376.76 mins + 04-03 21:44:28 | [053][000/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:44:29 | [053][010/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 21:44:30 | [053][020/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0046 ntime: 0080 mem: 3.36 + 04-03 21:44:31 | [053][030/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:44:33 | [053][040/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0045 ntime: 0079 mem: 3.36 + 04-03 21:44:34 | [053][050/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:44:35 | [053][060/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 21:44:37 | [053][070/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0072 mem: 3.36 + 04-03 21:44:38 | [053][080/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:44:39 | [053][090/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 21:44:40 | [053][100/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 21:44:42 | [053][110/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0063 ntime: 0084 mem: 3.36 + 04-03 21:44:43 | [053][120/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0051 ntime: 0072 mem: 3.36 + 04-03 21:44:45 | [053][130/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 21:44:46 | [053][140/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:44:47 | [053][150/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:44:48 | [053][160/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 21:44:50 | [053][170/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0055 ntime: 0075 mem: 3.36 + 04-03 21:44:51 | Time info >>>> elapsed: 21.47 mins remain: 376.20 mins + 04-03 21:44:51 | [054][000/179] predict_x0_loss: 0.023 glr: 5.0e-05 dtime: 0052 ntime: 0087 mem: 3.36 + 04-03 21:44:52 | [054][010/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0054 ntime: 0085 mem: 3.36 + 04-03 21:44:54 | [054][020/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0055 ntime: 0083 mem: 3.36 + 04-03 21:44:55 | [054][030/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 21:44:56 | [054][040/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:44:58 | [054][050/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0088 mem: 3.36 + 04-03 21:44:59 | [054][060/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0050 ntime: 0070 mem: 3.36 + 04-03 21:45:00 | [054][070/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0050 ntime: 0086 mem: 3.36 + 04-03 21:45:02 | [054][080/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 21:45:03 | [054][090/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 21:45:04 | [054][100/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0046 ntime: 0082 mem: 3.36 + 04-03 21:45:06 | [054][110/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:45:07 | [054][120/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 21:45:08 | [054][130/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 21:45:10 | [054][140/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:45:11 | [054][150/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0047 ntime: 0087 mem: 3.36 + 04-03 21:45:12 | [054][160/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:45:14 | [054][170/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0048 ntime: 0085 mem: 3.36 + 04-03 21:45:15 | Time info >>>> elapsed: 21.87 mins remain: 375.81 mins + 04-03 21:45:15 | [055][000/179] predict_x0_loss: 0.022 glr: 5.0e-05 dtime: 0046 ntime: 0078 mem: 3.36 + 04-03 21:45:16 | [055][010/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 21:45:18 | [055][020/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0089 mem: 3.36 + 04-03 21:45:19 | [055][030/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0048 ntime: 0086 mem: 3.36 + 04-03 21:45:20 | [055][040/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0048 ntime: 0071 mem: 3.36 + 04-03 21:45:21 | [055][050/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 21:45:23 | [055][060/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:45:24 | [055][070/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 21:45:25 | [055][080/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 21:45:26 | [055][090/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 21:45:28 | [055][100/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 21:45:29 | [055][110/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 21:45:30 | [055][120/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0047 ntime: 0074 mem: 3.36 + 04-03 21:45:32 | [055][130/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0044 ntime: 0075 mem: 3.36 + 04-03 21:45:33 | [055][140/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0083 mem: 3.36 + 04-03 21:45:34 | [055][150/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0067 ntime: 0071 mem: 3.36 + 04-03 21:45:35 | [055][160/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:45:37 | [055][170/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 21:45:38 | Time info >>>> elapsed: 22.26 mins remain: 375.20 mins + 04-03 21:45:38 | [056][000/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 21:45:39 | [056][010/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0075 ntime: 0095 mem: 3.36 + 04-03 21:45:41 | [056][020/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0061 ntime: 0071 mem: 3.36 + 04-03 21:45:42 | [056][030/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:45:44 | [056][040/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:45:45 | [056][050/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0056 ntime: 0080 mem: 3.36 + 04-03 21:45:46 | [056][060/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0083 ntime: 0089 mem: 3.36 + 04-03 21:45:48 | [056][070/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0065 ntime: 0080 mem: 3.36 + 04-03 21:45:49 | [056][080/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 21:45:50 | [056][090/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0052 ntime: 0086 mem: 3.36 + 04-03 21:45:52 | [056][100/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:45:53 | [056][110/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:45:54 | [056][120/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 21:45:56 | [056][130/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 21:45:57 | [056][140/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 21:45:58 | [056][150/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 21:46:00 | [056][160/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 21:46:01 | [056][170/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 21:46:02 | Time info >>>> elapsed: 22.66 mins remain: 374.92 mins + 04-03 21:46:02 | [057][000/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:46:04 | [057][010/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:46:05 | [057][020/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 21:46:06 | [057][030/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0050 ntime: 0086 mem: 3.36 + 04-03 21:46:08 | [057][040/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:46:09 | [057][050/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 21:46:10 | [057][060/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0078 ntime: 0084 mem: 3.36 + 04-03 21:46:12 | [057][070/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0058 ntime: 0081 mem: 3.36 + 04-03 21:46:13 | [057][080/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 21:46:14 | [057][090/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0053 ntime: 0083 mem: 3.36 + 04-03 21:46:16 | [057][100/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0057 ntime: 0078 mem: 3.36 + 04-03 21:46:17 | [057][110/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:46:18 | [057][120/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0043 ntime: 0070 mem: 3.36 + 04-03 21:46:20 | [057][130/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 21:46:21 | [057][140/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 21:46:22 | [057][150/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:46:24 | [057][160/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 21:46:25 | [057][170/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 21:46:26 | Time info >>>> elapsed: 23.06 mins remain: 374.56 mins + 04-03 21:46:26 | [058][000/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0065 ntime: 0085 mem: 3.36 + 04-03 21:46:28 | [058][010/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:46:29 | [058][020/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 21:46:30 | [058][030/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 21:46:31 | [058][040/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 21:46:33 | [058][050/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 21:46:34 | [058][060/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:46:35 | [058][070/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0053 ntime: 0075 mem: 3.36 + 04-03 21:46:37 | [058][080/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:46:38 | [058][090/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 21:46:39 | [058][100/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0071 mem: 3.36 + 04-03 21:46:41 | [058][110/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0054 ntime: 0083 mem: 3.36 + 04-03 21:46:42 | [058][120/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 21:46:43 | [058][130/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 21:46:44 | [058][140/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0050 ntime: 0072 mem: 3.36 + 04-03 21:46:46 | [058][150/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:46:47 | [058][160/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0045 ntime: 0077 mem: 3.36 + 04-03 21:46:48 | [058][170/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0046 ntime: 0075 mem: 3.36 + 04-03 21:46:49 | Time info >>>> elapsed: 23.45 mins remain: 373.99 mins + 04-03 21:46:49 | [059][000/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 21:46:51 | [059][010/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0052 ntime: 0088 mem: 3.36 + 04-03 21:46:52 | [059][020/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:46:54 | [059][030/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0057 ntime: 0086 mem: 3.36 + 04-03 21:46:55 | [059][040/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 21:46:56 | [059][050/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 21:46:58 | [059][060/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0055 ntime: 0089 mem: 3.36 + 04-03 21:46:59 | [059][070/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 21:47:00 | [059][080/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 21:47:02 | [059][090/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 21:47:03 | [059][100/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0056 ntime: 0083 mem: 3.36 + 04-03 21:47:04 | [059][110/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0074 ntime: 0084 mem: 3.36 + 04-03 21:47:06 | [059][120/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 21:47:07 | [059][130/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0055 ntime: 0082 mem: 3.36 + 04-03 21:47:09 | [059][140/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0050 ntime: 0086 mem: 3.36 + 04-03 21:47:10 | [059][150/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 21:47:11 | [059][160/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0050 ntime: 0073 mem: 3.36 + 04-03 21:47:12 | [059][170/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 21:47:14 | Time info >>>> elapsed: 23.85 mins remain: 373.69 mins + 04-03 21:47:14 | [060][000/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:47:15 | [060][010/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0048 ntime: 0085 mem: 3.36 + 04-03 21:47:16 | [060][020/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:47:18 | [060][030/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0046 ntime: 0073 mem: 3.36 + 04-03 21:47:19 | [060][040/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:47:20 | [060][050/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0090 mem: 3.36 + 04-03 21:47:21 | [060][060/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0073 mem: 3.36 + 04-03 21:47:23 | [060][070/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0066 ntime: 0081 mem: 3.36 + 04-03 21:47:24 | [060][080/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 21:47:26 | [060][090/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0046 ntime: 0088 mem: 3.36 + 04-03 21:47:27 | [060][100/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 21:47:28 | [060][110/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 21:47:29 | [060][120/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:47:31 | [060][130/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 21:47:32 | [060][140/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0042 ntime: 0064 mem: 3.36 + 04-03 21:47:33 | [060][150/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0054 ntime: 0077 mem: 3.36 + 04-03 21:47:35 | [060][160/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0051 ntime: 0072 mem: 3.36 + 04-03 21:47:36 | [060][170/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 21:47:37 | Time info >>>> elapsed: 24.25 mins remain: 373.29 mins + 04-03 21:47:37 | [061][000/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 21:47:39 | [061][010/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0055 ntime: 0082 mem: 3.36 + 04-03 21:47:40 | [061][020/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0048 ntime: 0071 mem: 3.36 + 04-03 21:47:42 | [061][030/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0060 ntime: 0081 mem: 3.36 + 04-03 21:47:43 | [061][040/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 21:47:44 | [061][050/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0055 ntime: 0083 mem: 3.36 + 04-03 21:47:46 | [061][060/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 21:47:47 | [061][070/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0083 mem: 3.36 + 04-03 21:47:48 | [061][080/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0084 mem: 3.36 + 04-03 21:47:50 | [061][090/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0045 ntime: 0077 mem: 3.36 + 04-03 21:47:51 | [061][100/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0046 ntime: 0069 mem: 3.36 + 04-03 21:47:52 | [061][110/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 21:47:53 | [061][120/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:47:55 | [061][130/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 21:47:56 | [061][140/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0087 mem: 3.36 + 04-03 21:47:57 | [061][150/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 21:47:59 | [061][160/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 21:48:00 | [061][170/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:48:01 | Time info >>>> elapsed: 24.64 mins remain: 372.85 mins + 04-03 21:48:01 | [062][000/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:48:02 | [062][010/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0048 ntime: 0088 mem: 3.36 + 04-03 21:48:04 | [062][020/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0095 mem: 3.36 + 04-03 21:48:05 | [062][030/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 21:48:07 | [062][040/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0053 ntime: 0084 mem: 3.36 + 04-03 21:48:08 | [062][050/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0089 mem: 3.36 + 04-03 21:48:09 | [062][060/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0046 ntime: 0082 mem: 3.36 + 04-03 21:48:10 | [062][070/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:48:12 | [062][080/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 21:48:13 | [062][090/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:48:14 | [062][100/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:48:16 | [062][110/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 21:48:17 | [062][120/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:48:18 | [062][130/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0044 ntime: 0074 mem: 3.36 + 04-03 21:48:19 | [062][140/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 21:48:21 | [062][150/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0047 ntime: 0071 mem: 3.36 + 04-03 21:48:22 | [062][160/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 21:48:23 | [062][170/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0045 ntime: 0081 mem: 3.36 + 04-03 21:48:24 | Time info >>>> elapsed: 25.03 mins remain: 372.27 mins + 04-03 21:48:24 | [063][000/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 21:48:25 | [063][010/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 21:48:27 | [063][020/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0045 ntime: 0077 mem: 3.36 + 04-03 21:48:28 | [063][030/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0046 ntime: 0079 mem: 3.36 + 04-03 21:48:29 | [063][040/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0045 ntime: 0073 mem: 3.36 + 04-03 21:48:30 | [063][050/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 21:48:32 | [063][060/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:48:33 | [063][070/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0046 ntime: 0082 mem: 3.36 + 04-03 21:48:34 | [063][080/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 21:48:36 | [063][090/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:48:37 | [063][100/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:48:38 | [063][110/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 21:48:39 | [063][120/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0053 ntime: 0083 mem: 3.36 + 04-03 21:48:41 | [063][130/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 21:48:42 | [063][140/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 21:48:43 | [063][150/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:48:45 | [063][160/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:48:46 | [063][170/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0055 ntime: 0084 mem: 3.36 + 04-03 21:48:47 | Time info >>>> elapsed: 25.41 mins remain: 371.63 mins + 04-03 21:48:47 | [064][000/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:48:49 | [064][010/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 21:48:50 | [064][020/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 21:48:51 | [064][030/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:48:53 | [064][040/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0058 ntime: 0091 mem: 3.36 + 04-03 21:48:54 | [064][050/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:48:55 | [064][060/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0080 ntime: 0083 mem: 3.36 + 04-03 21:48:57 | [064][070/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0048 ntime: 0085 mem: 3.36 + 04-03 21:48:58 | [064][080/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 21:48:59 | [064][090/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0086 mem: 3.36 + 04-03 21:49:01 | [064][100/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0055 ntime: 0084 mem: 3.36 + 04-03 21:49:02 | [064][110/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0049 ntime: 0072 mem: 3.36 + 04-03 21:49:03 | [064][120/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0052 ntime: 0085 mem: 3.36 + 04-03 21:49:05 | [064][130/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0047 ntime: 0073 mem: 3.36 + 04-03 21:49:06 | [064][140/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0052 ntime: 0090 mem: 3.36 + 04-03 21:49:07 | [064][150/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 21:49:09 | [064][160/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:49:10 | [064][170/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 21:49:11 | Time info >>>> elapsed: 25.81 mins remain: 371.29 mins + 04-03 21:49:11 | [065][000/179] predict_x0_loss: 0.020 glr: 5.0e-05 dtime: 0047 ntime: 0074 mem: 3.36 + 04-03 21:49:12 | [065][010/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 21:49:14 | [065][020/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0053 ntime: 0069 mem: 3.36 + 04-03 21:49:15 | [065][030/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 21:49:16 | [065][040/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 21:49:17 | [065][050/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0044 ntime: 0075 mem: 3.36 + 04-03 21:49:19 | [065][060/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0047 ntime: 0084 mem: 3.36 + 04-03 21:49:20 | [065][070/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0045 ntime: 0075 mem: 3.36 + 04-03 21:49:21 | [065][080/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0066 ntime: 0078 mem: 3.36 + 04-03 21:49:23 | [065][090/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0047 ntime: 0084 mem: 3.36 + 04-03 21:49:24 | [065][100/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:49:26 | [065][110/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0057 ntime: 0083 mem: 3.36 + 04-03 21:49:27 | [065][120/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0056 ntime: 0079 mem: 3.36 + 04-03 21:49:28 | [065][130/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0046 ntime: 0082 mem: 3.36 + 04-03 21:49:30 | [065][140/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0054 ntime: 0076 mem: 3.36 + 04-03 21:49:31 | [065][150/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 21:49:32 | [065][160/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0045 ntime: 0077 mem: 3.36 + 04-03 21:49:33 | [065][170/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 21:49:34 | Time info >>>> elapsed: 26.20 mins remain: 370.80 mins + 04-03 21:49:35 | [066][000/179] predict_x0_loss: 0.021 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:49:36 | [066][010/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0043 ntime: 0078 mem: 3.36 + 04-03 21:49:37 | [066][020/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0046 ntime: 0076 mem: 3.36 + 04-03 21:49:38 | [066][030/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 21:49:40 | [066][040/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:49:41 | [066][050/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0046 ntime: 0076 mem: 3.36 + 04-03 21:49:42 | [066][060/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0046 ntime: 0079 mem: 3.36 + 04-03 21:49:44 | [066][070/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 21:49:45 | [066][080/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 21:49:46 | [066][090/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:49:47 | [066][100/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 21:49:49 | [066][110/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:49:50 | [066][120/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:49:51 | [066][130/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0046 ntime: 0075 mem: 3.36 + 04-03 21:49:53 | [066][140/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 21:49:54 | [066][150/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 21:49:55 | [066][160/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 21:49:57 | [066][170/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 21:49:58 | Time info >>>> elapsed: 26.59 mins remain: 370.24 mins + 04-03 21:49:58 | [067][000/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0051 ntime: 0074 mem: 3.36 + 04-03 21:49:59 | [067][010/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:50:00 | [067][020/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0051 ntime: 0073 mem: 3.36 + 04-03 21:50:02 | [067][030/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 21:50:03 | [067][040/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:50:04 | [067][050/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:50:06 | [067][060/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0060 ntime: 0086 mem: 3.36 + 04-03 21:50:07 | [067][070/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 21:50:09 | [067][080/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0066 ntime: 0078 mem: 3.36 + 04-03 21:50:10 | [067][090/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0048 ntime: 0072 mem: 3.36 + 04-03 21:50:11 | [067][100/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0046 ntime: 0073 mem: 3.36 + 04-03 21:50:12 | [067][110/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:50:14 | [067][120/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0065 ntime: 0081 mem: 3.36 + 04-03 21:50:15 | [067][130/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 21:50:16 | [067][140/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0047 ntime: 0072 mem: 3.36 + 04-03 21:50:17 | [067][150/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0053 ntime: 0079 mem: 3.36 + 04-03 21:50:19 | [067][160/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 21:50:20 | [067][170/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 21:50:21 | Time info >>>> elapsed: 26.98 mins remain: 369.75 mins + 04-03 21:50:21 | [068][000/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 21:50:22 | [068][010/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 21:50:24 | [068][020/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0068 ntime: 0072 mem: 3.36 + 04-03 21:50:25 | [068][030/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:50:26 | [068][040/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 21:50:28 | [068][050/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0086 mem: 3.36 + 04-03 21:50:29 | [068][060/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 21:50:30 | [068][070/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0087 mem: 3.36 + 04-03 21:50:32 | [068][080/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 21:50:33 | [068][090/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0071 mem: 3.36 + 04-03 21:50:34 | [068][100/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 21:50:36 | [068][110/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 21:50:37 | [068][120/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 21:50:38 | [068][130/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0045 ntime: 0082 mem: 3.36 + 04-03 21:50:40 | [068][140/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0047 ntime: 0083 mem: 3.36 + 04-03 21:50:41 | [068][150/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0046 ntime: 0084 mem: 3.36 + 04-03 21:50:42 | [068][160/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 21:50:43 | [068][170/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:50:45 | Time info >>>> elapsed: 27.37 mins remain: 369.29 mins + 04-03 21:50:45 | [069][000/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0073 mem: 3.36 + 04-03 21:50:46 | [069][010/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0044 ntime: 0070 mem: 3.36 + 04-03 21:50:47 | [069][020/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0072 mem: 3.36 + 04-03 21:50:49 | [069][030/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 21:50:50 | [069][040/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0054 ntime: 0076 mem: 3.36 + 04-03 21:50:52 | [069][050/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0057 ntime: 0083 mem: 3.36 + 04-03 21:50:53 | [069][060/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 21:50:54 | [069][070/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0085 mem: 3.36 + 04-03 21:50:56 | [069][080/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0049 ntime: 0087 mem: 3.36 + 04-03 21:50:57 | [069][090/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 21:50:58 | [069][100/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0072 mem: 3.36 + 04-03 21:51:00 | [069][110/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0086 mem: 3.36 + 04-03 21:51:01 | [069][120/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0069 ntime: 0080 mem: 3.36 + 04-03 21:51:02 | [069][130/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 21:51:04 | [069][140/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 21:51:05 | [069][150/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0045 ntime: 0057 mem: 3.36 + 04-03 21:51:06 | [069][160/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0073 ntime: 0084 mem: 3.36 + 04-03 21:51:08 | [069][170/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0055 ntime: 0088 mem: 3.36 + 04-03 21:51:09 | Time info >>>> elapsed: 27.77 mins remain: 368.99 mins + 04-03 21:51:09 | [070][000/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 21:51:10 | [070][010/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 21:51:12 | [070][020/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0073 ntime: 0082 mem: 3.36 + 04-03 21:51:13 | [070][030/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0053 ntime: 0083 mem: 3.36 + 04-03 21:51:14 | [070][040/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0055 ntime: 0082 mem: 3.36 + 04-03 21:51:16 | [070][050/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 21:51:17 | [070][060/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:51:18 | [070][070/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:51:20 | [070][080/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 21:51:21 | [070][090/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:51:22 | [070][100/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:51:24 | [070][110/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:51:25 | [070][120/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0070 ntime: 0086 mem: 3.36 + 04-03 21:51:26 | [070][130/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 21:51:28 | [070][140/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 21:51:29 | [070][150/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 21:51:30 | [070][160/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 21:51:32 | [070][170/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0054 ntime: 0082 mem: 3.36 + 04-03 21:51:33 | Time info >>>> elapsed: 28.17 mins remain: 368.60 mins + 04-03 21:51:33 | [071][000/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 21:51:34 | [071][010/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 21:51:35 | [071][020/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:51:37 | [071][030/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0047 ntime: 0071 mem: 3.36 + 04-03 21:51:38 | [071][040/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0056 ntime: 0079 mem: 3.36 + 04-03 21:51:39 | [071][050/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0085 mem: 3.36 + 04-03 21:51:41 | [071][060/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 21:51:42 | [071][070/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 21:51:43 | [071][080/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:51:45 | [071][090/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 21:51:46 | [071][100/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 21:51:47 | [071][110/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 21:51:49 | [071][120/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 21:51:50 | [071][130/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:51:51 | [071][140/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 21:51:52 | [071][150/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 21:51:54 | [071][160/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:51:55 | [071][170/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0088 mem: 3.36 + 04-03 21:51:56 | Time info >>>> elapsed: 28.56 mins remain: 368.14 mins + 04-03 21:51:56 | [072][000/179] predict_x0_loss: 0.019 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 21:51:58 | [072][010/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 21:51:59 | [072][020/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 21:52:00 | [072][030/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 21:52:01 | [072][040/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:52:03 | [072][050/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 21:52:04 | [072][060/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0045 ntime: 0078 mem: 3.36 + 04-03 21:52:05 | [072][070/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 21:52:06 | [072][080/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0060 ntime: 0079 mem: 3.36 + 04-03 21:52:08 | [072][090/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0060 ntime: 0077 mem: 3.36 + 04-03 21:52:09 | [072][100/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 21:52:10 | [072][110/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:52:12 | [072][120/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0071 mem: 3.36 + 04-03 21:52:13 | [072][130/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 21:52:14 | [072][140/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:52:16 | [072][150/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 21:52:17 | [072][160/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 21:52:18 | [072][170/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 21:52:19 | Time info >>>> elapsed: 28.95 mins remain: 367.61 mins + 04-03 21:52:19 | [073][000/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 21:52:21 | [073][010/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 21:52:22 | [073][020/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0055 ntime: 0077 mem: 3.36 + 04-03 21:52:23 | [073][030/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0046 ntime: 0084 mem: 3.36 + 04-03 21:52:25 | [073][040/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0047 ntime: 0073 mem: 3.36 + 04-03 21:52:26 | [073][050/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 21:52:27 | [073][060/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 21:52:28 | [073][070/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 21:52:30 | [073][080/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:52:31 | [073][090/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:52:32 | [073][100/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 21:52:34 | [073][110/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 21:52:35 | [073][120/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:52:36 | [073][130/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:52:37 | [073][140/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 21:52:39 | [073][150/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 21:52:40 | [073][160/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 21:52:41 | [073][170/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 21:52:42 | Time info >>>> elapsed: 29.33 mins remain: 367.07 mins + 04-03 21:52:42 | [074][000/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0047 ntime: 0074 mem: 3.36 + 04-03 21:52:44 | [074][010/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:52:45 | [074][020/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 21:52:47 | [074][030/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 21:52:48 | [074][040/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0054 ntime: 0083 mem: 3.36 + 04-03 21:52:49 | [074][050/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 21:52:51 | [074][060/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0081 ntime: 0095 mem: 3.36 + 04-03 21:52:52 | [074][070/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0056 ntime: 0088 mem: 3.36 + 04-03 21:52:53 | [074][080/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0088 mem: 3.36 + 04-03 21:52:55 | [074][090/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0087 mem: 3.36 + 04-03 21:52:56 | [074][100/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 21:52:57 | [074][110/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:52:59 | [074][120/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0053 ntime: 0076 mem: 3.36 + 04-03 21:53:00 | [074][130/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0087 mem: 3.36 + 04-03 21:53:01 | [074][140/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0046 ntime: 0070 mem: 3.36 + 04-03 21:53:03 | [074][150/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:53:04 | [074][160/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:53:05 | [074][170/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 21:53:06 | Time info >>>> elapsed: 29.73 mins remain: 366.73 mins + 04-03 21:53:07 | [075][000/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:53:08 | [075][010/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:53:09 | [075][020/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:53:10 | [075][030/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0045 ntime: 0071 mem: 3.36 + 04-03 21:53:12 | [075][040/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0046 ntime: 0074 mem: 3.36 + 04-03 21:53:13 | [075][050/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0047 ntime: 0084 mem: 3.36 + 04-03 21:53:14 | [075][060/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 21:53:16 | [075][070/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 21:53:17 | [075][080/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:53:18 | [075][090/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 21:53:20 | [075][100/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0073 mem: 3.36 + 04-03 21:53:21 | [075][110/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 21:53:22 | [075][120/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 21:53:24 | [075][130/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 21:53:25 | [075][140/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:53:26 | [075][150/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 21:53:27 | [075][160/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:53:29 | [075][170/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0087 mem: 3.36 + 04-03 21:53:30 | Time info >>>> elapsed: 30.13 mins remain: 366.26 mins + 04-03 21:53:30 | [076][000/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0053 ntime: 0073 mem: 3.36 + 04-03 21:53:31 | [076][010/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 21:53:33 | [076][020/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:53:34 | [076][030/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:53:35 | [076][040/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 21:53:36 | [076][050/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 21:53:38 | [076][060/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:53:39 | [076][070/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 21:53:40 | [076][080/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0055 ntime: 0075 mem: 3.36 + 04-03 21:53:42 | [076][090/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 21:53:43 | [076][100/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 21:53:44 | [076][110/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:53:46 | [076][120/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0073 mem: 3.36 + 04-03 21:53:47 | [076][130/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0053 ntime: 0086 mem: 3.36 + 04-03 21:53:48 | [076][140/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:53:50 | [076][150/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:53:51 | [076][160/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 21:53:52 | [076][170/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:53:53 | Time info >>>> elapsed: 30.52 mins remain: 365.81 mins + 04-03 21:53:54 | [077][000/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:53:55 | [077][010/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 21:53:56 | [077][020/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 21:53:57 | [077][030/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:53:59 | [077][040/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 21:54:00 | [077][050/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 21:54:01 | [077][060/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0055 ntime: 0084 mem: 3.36 + 04-03 21:54:03 | [077][070/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:54:04 | [077][080/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 21:54:06 | [077][090/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 21:54:07 | [077][100/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0087 mem: 3.36 + 04-03 21:54:08 | [077][110/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:54:10 | [077][120/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0046 ntime: 0078 mem: 3.36 + 04-03 21:54:11 | [077][130/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0054 ntime: 0076 mem: 3.36 + 04-03 21:54:12 | [077][140/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 21:54:13 | [077][150/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:54:15 | [077][160/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0046 ntime: 0075 mem: 3.36 + 04-03 21:54:16 | [077][170/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 21:54:17 | Time info >>>> elapsed: 30.91 mins remain: 365.36 mins + 04-03 21:54:17 | [078][000/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0044 ntime: 0079 mem: 3.36 + 04-03 21:54:18 | [078][010/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:54:20 | [078][020/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 21:54:21 | [078][030/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:54:22 | [078][040/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:54:24 | [078][050/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:54:25 | [078][060/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0052 ntime: 0075 mem: 3.36 + 04-03 21:54:26 | [078][070/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 21:54:28 | [078][080/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 21:54:29 | [078][090/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:54:30 | [078][100/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 21:54:31 | [078][110/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:54:33 | [078][120/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0056 ntime: 0083 mem: 3.36 + 04-03 21:54:34 | [078][130/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0055 ntime: 0086 mem: 3.36 + 04-03 21:54:35 | [078][140/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0056 ntime: 0085 mem: 3.36 + 04-03 21:54:37 | [078][150/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 21:54:38 | [078][160/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0072 mem: 3.36 + 04-03 21:54:39 | [078][170/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 21:54:40 | Time info >>>> elapsed: 31.30 mins remain: 364.92 mins + 04-03 21:54:41 | [079][000/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0051 ntime: 0073 mem: 3.36 + 04-03 21:54:42 | [079][010/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 21:54:43 | [079][020/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 21:54:44 | [079][030/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0074 mem: 3.36 + 04-03 21:54:46 | [079][040/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0057 ntime: 0078 mem: 3.36 + 04-03 21:54:47 | [079][050/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 21:54:48 | [079][060/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0054 ntime: 0071 mem: 3.36 + 04-03 21:54:50 | [079][070/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0053 ntime: 0087 mem: 3.36 + 04-03 21:54:51 | [079][080/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 21:54:52 | [079][090/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0074 ntime: 0082 mem: 3.36 + 04-03 21:54:54 | [079][100/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0047 ntime: 0084 mem: 3.36 + 04-03 21:54:55 | [079][110/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 21:54:56 | [079][120/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 21:54:58 | [079][130/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0087 mem: 3.36 + 04-03 21:54:59 | [079][140/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:55:00 | [079][150/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0046 ntime: 0076 mem: 3.36 + 04-03 21:55:02 | [079][160/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 21:55:03 | [079][170/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0053 ntime: 0089 mem: 3.36 + 04-03 21:55:04 | Time info >>>> elapsed: 31.70 mins remain: 364.53 mins + 04-03 21:55:04 | [080][000/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:55:06 | [080][010/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 21:55:07 | [080][020/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0054 ntime: 0077 mem: 3.36 + 04-03 21:55:08 | [080][030/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0070 mem: 3.36 + 04-03 21:55:10 | [080][040/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 21:55:11 | [080][050/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0072 mem: 3.36 + 04-03 21:55:12 | [080][060/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 21:55:14 | [080][070/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 21:55:15 | [080][080/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0052 ntime: 0085 mem: 3.36 + 04-03 21:55:16 | [080][090/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0046 ntime: 0079 mem: 3.36 + 04-03 21:55:18 | [080][100/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 21:55:19 | [080][110/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:55:20 | [080][120/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:55:21 | [080][130/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0046 ntime: 0081 mem: 3.36 + 04-03 21:55:23 | [080][140/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 21:55:24 | [080][150/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:55:25 | [080][160/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 21:55:27 | [080][170/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 21:55:28 | Time info >>>> elapsed: 32.09 mins remain: 364.07 mins + 04-03 21:55:28 | [081][000/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:55:29 | [081][010/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 21:55:30 | [081][020/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0054 ntime: 0086 mem: 3.36 + 04-03 21:55:32 | [081][030/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0086 mem: 3.36 + 04-03 21:55:33 | [081][040/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0056 ntime: 0081 mem: 3.36 + 04-03 21:55:35 | [081][050/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0053 ntime: 0074 mem: 3.36 + 04-03 21:55:36 | [081][060/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0073 mem: 3.36 + 04-03 21:55:37 | [081][070/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:55:39 | [081][080/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 21:55:40 | [081][090/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0047 ntime: 0073 mem: 3.36 + 04-03 21:55:41 | [081][100/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 21:55:42 | [081][110/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0054 ntime: 0088 mem: 3.36 + 04-03 21:55:44 | [081][120/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 21:55:45 | [081][130/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:55:46 | [081][140/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 21:55:48 | [081][150/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0053 ntime: 0083 mem: 3.36 + 04-03 21:55:49 | [081][160/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 21:55:50 | [081][170/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 21:55:51 | Time info >>>> elapsed: 32.48 mins remain: 363.65 mins + 04-03 21:55:51 | [082][000/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0055 ntime: 0087 mem: 3.36 + 04-03 21:55:53 | [082][010/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0072 mem: 3.36 + 04-03 21:55:54 | [082][020/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:55:55 | [082][030/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0053 ntime: 0083 mem: 3.36 + 04-03 21:55:57 | [082][040/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:55:58 | [082][050/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0046 ntime: 0079 mem: 3.36 + 04-03 21:55:59 | [082][060/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 21:56:01 | [082][070/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:56:02 | [082][080/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:56:03 | [082][090/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0072 mem: 3.36 + 04-03 21:56:05 | [082][100/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0085 mem: 3.36 + 04-03 21:56:06 | [082][110/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0085 mem: 3.36 + 04-03 21:56:07 | [082][120/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:56:09 | [082][130/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0090 mem: 3.36 + 04-03 21:56:10 | [082][140/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 21:56:12 | [082][150/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:56:13 | [082][160/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:56:14 | [082][170/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 21:56:15 | Time info >>>> elapsed: 32.88 mins remain: 363.26 mins + 04-03 21:56:15 | [083][000/179] predict_x0_loss: 0.018 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 21:56:17 | [083][010/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 21:56:18 | [083][020/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 21:56:19 | [083][030/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0060 ntime: 0080 mem: 3.36 + 04-03 21:56:21 | [083][040/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:56:22 | [083][050/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 21:56:23 | [083][060/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 21:56:25 | [083][070/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 21:56:26 | [083][080/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0046 ntime: 0073 mem: 3.36 + 04-03 21:56:27 | [083][090/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 21:56:29 | [083][100/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0061 ntime: 0083 mem: 3.36 + 04-03 21:56:30 | [083][110/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0062 ntime: 0078 mem: 3.36 + 04-03 21:56:31 | [083][120/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 21:56:33 | [083][130/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0046 ntime: 0073 mem: 3.36 + 04-03 21:56:34 | [083][140/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:56:35 | [083][150/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 21:56:36 | [083][160/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 21:56:38 | [083][170/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0046 ntime: 0075 mem: 3.36 + 04-03 21:56:39 | Time info >>>> elapsed: 33.27 mins remain: 362.83 mins + 04-03 21:56:39 | [084][000/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:56:40 | [084][010/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0054 ntime: 0084 mem: 3.36 + 04-03 21:56:41 | [084][020/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0085 mem: 3.36 + 04-03 21:56:43 | [084][030/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0057 ntime: 0088 mem: 3.36 + 04-03 21:56:44 | [084][040/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 21:56:46 | [084][050/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0221 ntime: 0075 mem: 3.36 + 04-03 21:56:47 | [084][060/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 21:56:49 | [084][070/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0058 ntime: 0070 mem: 3.36 + 04-03 21:56:50 | [084][080/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 21:56:52 | [084][090/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:56:53 | [084][100/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 21:56:54 | [084][110/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0053 ntime: 0084 mem: 3.36 + 04-03 21:56:56 | [084][120/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0360 ntime: 0088 mem: 3.36 + 04-03 21:56:57 | [084][130/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0086 mem: 3.36 + 04-03 21:56:59 | [084][140/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0056 ntime: 0080 mem: 3.36 + 04-03 21:57:00 | [084][150/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0059 ntime: 0082 mem: 3.36 + 04-03 21:57:01 | [084][160/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:57:03 | [084][170/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 21:57:04 | Time info >>>> elapsed: 33.69 mins remain: 362.69 mins + 04-03 21:57:04 | [085][000/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0045 ntime: 0077 mem: 3.36 + 04-03 21:57:05 | [085][010/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 21:57:07 | [085][020/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 21:57:08 | [085][030/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0052 ntime: 0075 mem: 3.36 + 04-03 21:57:09 | [085][040/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0053 ntime: 0074 mem: 3.36 + 04-03 21:57:11 | [085][050/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 21:57:12 | [085][060/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 21:57:13 | [085][070/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 21:57:14 | [085][080/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0053 ntime: 0079 mem: 3.36 + 04-03 21:57:16 | [085][090/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:57:17 | [085][100/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0072 mem: 3.36 + 04-03 21:57:18 | [085][110/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:57:20 | [085][120/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 21:57:21 | [085][130/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0069 ntime: 0083 mem: 3.36 + 04-03 21:57:22 | [085][140/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 21:57:24 | [085][150/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:57:25 | [085][160/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 21:57:26 | [085][170/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 21:57:27 | Time info >>>> elapsed: 34.08 mins remain: 362.25 mins + 04-03 21:57:28 | [086][000/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0053 ntime: 0076 mem: 3.36 + 04-03 21:57:29 | [086][010/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 21:57:30 | [086][020/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0085 mem: 3.36 + 04-03 21:57:32 | [086][030/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 21:57:33 | [086][040/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0061 ntime: 0083 mem: 3.36 + 04-03 21:57:34 | [086][050/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0072 mem: 3.36 + 04-03 21:57:36 | [086][060/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0058 ntime: 0079 mem: 3.36 + 04-03 21:57:37 | [086][070/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:57:38 | [086][080/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0085 mem: 3.36 + 04-03 21:57:40 | [086][090/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 21:57:41 | [086][100/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0062 ntime: 0076 mem: 3.36 + 04-03 21:57:42 | [086][110/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0062 ntime: 0079 mem: 3.36 + 04-03 21:57:44 | [086][120/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0061 ntime: 0074 mem: 3.36 + 04-03 21:57:45 | [086][130/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 21:57:46 | [086][140/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 21:57:48 | [086][150/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 21:57:49 | [086][160/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 21:57:50 | [086][170/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 21:57:51 | Time info >>>> elapsed: 34.48 mins remain: 361.86 mins + 04-03 21:57:51 | [087][000/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:57:53 | [087][010/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0072 mem: 3.36 + 04-03 21:57:54 | [087][020/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:57:55 | [087][030/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 21:57:57 | [087][040/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 21:57:58 | [087][050/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0073 mem: 3.36 + 04-03 21:57:59 | [087][060/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0072 mem: 3.36 + 04-03 21:58:00 | [087][070/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 21:58:02 | [087][080/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 21:58:03 | [087][090/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:58:04 | [087][100/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0056 ntime: 0082 mem: 3.36 + 04-03 21:58:06 | [087][110/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 21:58:07 | [087][120/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 21:58:09 | [087][130/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0065 ntime: 0079 mem: 3.36 + 04-03 21:58:10 | [087][140/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0059 ntime: 0074 mem: 3.36 + 04-03 21:58:11 | [087][150/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0067 ntime: 0077 mem: 3.36 + 04-03 21:58:13 | [087][160/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0055 ntime: 0076 mem: 3.36 + 04-03 21:58:14 | [087][170/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 21:58:15 | Time info >>>> elapsed: 34.88 mins remain: 361.45 mins + 04-03 21:58:15 | [088][000/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0046 ntime: 0083 mem: 3.36 + 04-03 21:58:16 | [088][010/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 21:58:18 | [088][020/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 21:58:19 | [088][030/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 21:58:20 | [088][040/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 21:58:22 | [088][050/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0059 ntime: 0074 mem: 3.36 + 04-03 21:58:23 | [088][060/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0056 ntime: 0079 mem: 3.36 + 04-03 21:58:24 | [088][070/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0046 ntime: 0076 mem: 3.36 + 04-03 21:58:26 | [088][080/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0047 ntime: 0073 mem: 3.36 + 04-03 21:58:27 | [088][090/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0069 mem: 3.36 + 04-03 21:58:28 | [088][100/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 21:58:29 | [088][110/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:58:31 | [088][120/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0053 ntime: 0076 mem: 3.36 + 04-03 21:58:32 | [088][130/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0047 ntime: 0086 mem: 3.36 + 04-03 21:58:34 | [088][140/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0054 ntime: 0087 mem: 3.36 + 04-03 21:58:35 | [088][150/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 21:58:36 | [088][160/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 21:58:38 | [088][170/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0056 ntime: 0076 mem: 3.36 + 04-03 21:58:39 | Time info >>>> elapsed: 35.27 mins remain: 361.07 mins + 04-03 21:58:39 | [089][000/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0068 ntime: 0069 mem: 3.36 + 04-03 21:58:40 | [089][010/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:58:42 | [089][020/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 21:58:43 | [089][030/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0055 ntime: 0074 mem: 3.36 + 04-03 21:58:44 | [089][040/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0054 ntime: 0075 mem: 3.36 + 04-03 21:58:47 | [089][050/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0052 ntime: 0091 mem: 3.36 + 04-03 21:58:48 | [089][060/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:58:49 | [089][070/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 21:58:51 | [089][080/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 21:58:52 | [089][090/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0056 ntime: 0075 mem: 3.36 + 04-03 21:58:53 | [089][100/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0045 ntime: 0080 mem: 3.36 + 04-03 21:58:55 | [089][110/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 21:58:56 | [089][120/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0054 ntime: 0073 mem: 3.36 + 04-03 21:58:57 | [089][130/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 21:58:59 | [089][140/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 21:59:00 | [089][150/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 21:59:01 | [089][160/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 21:59:03 | [089][170/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 21:59:04 | Time info >>>> elapsed: 35.69 mins remain: 360.86 mins + 04-03 21:59:04 | [090][000/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 21:59:05 | [090][010/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 21:59:06 | [090][020/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:59:08 | [090][030/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 21:59:09 | [090][040/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 21:59:10 | [090][050/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 21:59:12 | [090][060/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0070 mem: 3.36 + 04-03 21:59:13 | [090][070/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0047 ntime: 0091 mem: 3.36 + 04-03 21:59:14 | [090][080/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0052 ntime: 0074 mem: 3.36 + 04-03 21:59:16 | [090][090/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 21:59:17 | [090][100/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0060 ntime: 0083 mem: 3.36 + 04-03 21:59:18 | [090][110/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0057 ntime: 0077 mem: 3.36 + 04-03 21:59:20 | [090][120/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:59:21 | [090][130/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 21:59:22 | [090][140/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0076 ntime: 0074 mem: 3.36 + 04-03 21:59:24 | [090][150/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 21:59:25 | [090][160/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 21:59:26 | [090][170/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 21:59:27 | Time info >>>> elapsed: 36.08 mins remain: 360.44 mins + 04-03 21:59:27 | [091][000/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 21:59:29 | [091][010/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 21:59:30 | [091][020/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 21:59:32 | [091][030/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0048 ntime: 0087 mem: 3.36 + 04-03 21:59:33 | [091][040/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 21:59:34 | [091][050/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0051 ntime: 0087 mem: 3.36 + 04-03 21:59:35 | [091][060/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0045 ntime: 0069 mem: 3.36 + 04-03 21:59:37 | [091][070/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 21:59:38 | [091][080/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 21:59:39 | [091][090/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 21:59:41 | [091][100/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 21:59:42 | [091][110/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0067 ntime: 0071 mem: 3.36 + 04-03 21:59:43 | [091][120/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0050 ntime: 0073 mem: 3.36 + 04-03 21:59:45 | [091][130/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0086 mem: 3.36 + 04-03 21:59:46 | [091][140/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 21:59:47 | [091][150/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 21:59:49 | [091][160/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 21:59:50 | [091][170/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0056 ntime: 0085 mem: 3.36 + 04-03 21:59:51 | Time info >>>> elapsed: 36.48 mins remain: 360.04 mins + 04-03 21:59:51 | [092][000/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 21:59:53 | [092][010/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0087 mem: 3.36 + 04-03 21:59:54 | [092][020/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 21:59:55 | [092][030/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0051 ntime: 0073 mem: 3.36 + 04-03 21:59:57 | [092][040/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0052 ntime: 0086 mem: 3.36 + 04-03 21:59:58 | [092][050/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0059 ntime: 0091 mem: 3.36 + 04-03 21:59:59 | [092][060/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 22:00:01 | [092][070/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:00:02 | [092][080/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:00:03 | [092][090/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0054 ntime: 0074 mem: 3.36 + 04-03 22:00:04 | [092][100/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 22:00:06 | [092][110/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0084 ntime: 0093 mem: 3.36 + 04-03 22:00:07 | [092][120/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 22:00:09 | [092][130/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0055 ntime: 0073 mem: 3.36 + 04-03 22:00:10 | [092][140/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0082 ntime: 0077 mem: 3.36 + 04-03 22:00:11 | [092][150/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0073 mem: 3.36 + 04-03 22:00:13 | [092][160/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0085 mem: 3.36 + 04-03 22:00:14 | [092][170/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0046 ntime: 0081 mem: 3.36 + 04-03 22:00:15 | Time info >>>> elapsed: 36.88 mins remain: 359.65 mins + 04-03 22:00:15 | [093][000/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0046 ntime: 0076 mem: 3.36 + 04-03 22:00:16 | [093][010/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 22:00:18 | [093][020/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0053 ntime: 0076 mem: 3.36 + 04-03 22:00:19 | [093][030/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0072 ntime: 0081 mem: 3.36 + 04-03 22:00:21 | [093][040/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0056 ntime: 0085 mem: 3.36 + 04-03 22:00:22 | [093][050/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0069 ntime: 0074 mem: 3.36 + 04-03 22:00:24 | [093][060/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0065 ntime: 0076 mem: 3.36 + 04-03 22:00:25 | [093][070/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0105 ntime: 0080 mem: 3.36 + 04-03 22:00:26 | [093][080/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0051 ntime: 0089 mem: 3.36 + 04-03 22:00:28 | [093][090/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0063 ntime: 0081 mem: 3.36 + 04-03 22:00:29 | [093][100/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0057 ntime: 0076 mem: 3.36 + 04-03 22:00:30 | [093][110/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 22:00:32 | [093][120/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 22:00:33 | [093][130/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 22:00:34 | [093][140/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 22:00:36 | [093][150/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0054 ntime: 0074 mem: 3.36 + 04-03 22:00:37 | [093][160/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 22:00:38 | [093][170/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 22:00:40 | Time info >>>> elapsed: 37.29 mins remain: 359.37 mins + 04-03 22:00:40 | [094][000/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0059 ntime: 0079 mem: 3.36 + 04-03 22:00:42 | [094][010/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0140 ntime: 0075 mem: 3.36 + 04-03 22:00:44 | [094][020/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0054 ntime: 0084 mem: 3.36 + 04-03 22:00:45 | [094][030/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0054 ntime: 0082 mem: 3.36 + 04-03 22:00:46 | [094][040/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 22:00:48 | [094][050/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 22:00:49 | [094][060/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0057 ntime: 0081 mem: 3.36 + 04-03 22:00:50 | [094][070/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 22:00:52 | [094][080/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0074 ntime: 0074 mem: 3.36 + 04-03 22:00:53 | [094][090/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 22:00:55 | [094][100/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0054 ntime: 0090 mem: 3.36 + 04-03 22:00:56 | [094][110/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 22:00:57 | [094][120/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 22:00:59 | [094][130/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0056 ntime: 0077 mem: 3.36 + 04-03 22:01:00 | [094][140/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0069 ntime: 0096 mem: 3.36 + 04-03 22:01:01 | [094][150/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:01:03 | [094][160/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 22:01:04 | [094][170/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:01:05 | Time info >>>> elapsed: 37.71 mins remain: 359.26 mins + 04-03 22:01:05 | [095][000/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0086 mem: 3.36 + 04-03 22:01:07 | [095][010/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 22:01:08 | [095][020/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 22:01:09 | [095][030/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 22:01:10 | [095][040/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 22:01:12 | [095][050/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0055 ntime: 0083 mem: 3.36 + 04-03 22:01:13 | [095][060/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0057 ntime: 0079 mem: 3.36 + 04-03 22:01:15 | [095][070/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 22:01:16 | [095][080/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0057 ntime: 0083 mem: 3.36 + 04-03 22:01:17 | [095][090/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0055 ntime: 0085 mem: 3.36 + 04-03 22:01:19 | [095][100/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0064 ntime: 0080 mem: 3.36 + 04-03 22:01:20 | [095][110/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0064 ntime: 0083 mem: 3.36 + 04-03 22:01:22 | [095][120/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0073 ntime: 0078 mem: 3.36 + 04-03 22:01:23 | [095][130/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 22:01:24 | [095][140/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0075 mem: 3.36 + 04-03 22:01:26 | [095][150/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0057 ntime: 0082 mem: 3.36 + 04-03 22:01:27 | [095][160/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0087 mem: 3.36 + 04-03 22:01:28 | [095][170/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0058 ntime: 0083 mem: 3.36 + 04-03 22:01:30 | Time info >>>> elapsed: 38.12 mins remain: 359.01 mins + 04-03 22:01:30 | [096][000/179] predict_x0_loss: 0.016 glr: 5.0e-05 dtime: 0073 ntime: 0078 mem: 3.36 + 04-03 22:01:31 | [096][010/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 22:01:33 | [096][020/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0086 mem: 3.36 + 04-03 22:01:34 | [096][030/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0072 mem: 3.36 + 04-03 22:01:35 | [096][040/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 22:01:36 | [096][050/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 22:01:38 | [096][060/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0047 ntime: 0063 mem: 3.36 + 04-03 22:01:39 | [096][070/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:01:40 | [096][080/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 22:01:42 | [096][090/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0070 ntime: 0075 mem: 3.36 + 04-03 22:01:43 | [096][100/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 22:01:44 | [096][110/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0078 ntime: 0082 mem: 3.36 + 04-03 22:01:46 | [096][120/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:01:47 | [096][130/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0065 ntime: 0076 mem: 3.36 + 04-03 22:01:48 | [096][140/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:01:50 | [096][150/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0045 ntime: 0075 mem: 3.36 + 04-03 22:01:51 | [096][160/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0045 ntime: 0076 mem: 3.36 + 04-03 22:01:52 | [096][170/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0043 ntime: 0072 mem: 3.36 + 04-03 22:01:53 | Time info >>>> elapsed: 38.52 mins remain: 358.55 mins + 04-03 22:01:53 | [097][000/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 22:01:55 | [097][010/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 22:01:56 | [097][020/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:01:57 | [097][030/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0067 ntime: 0076 mem: 3.36 + 04-03 22:01:59 | [097][040/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0045 ntime: 0080 mem: 3.36 + 04-03 22:02:00 | [097][050/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:02:01 | [097][060/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:02:03 | [097][070/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 22:02:04 | [097][080/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 22:02:05 | [097][090/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 22:02:07 | [097][100/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0057 ntime: 0081 mem: 3.36 + 04-03 22:02:08 | [097][110/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0080 ntime: 0087 mem: 3.36 + 04-03 22:02:10 | [097][120/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0067 ntime: 0081 mem: 3.36 + 04-03 22:02:11 | [097][130/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 22:02:13 | [097][140/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 22:02:14 | [097][150/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 22:02:16 | [097][160/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0055 ntime: 0074 mem: 3.36 + 04-03 22:02:17 | [097][170/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 22:02:18 | Time info >>>> elapsed: 38.93 mins remain: 358.29 mins + 04-03 22:02:18 | [098][000/179] predict_x0_loss: 0.017 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:02:19 | [098][010/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0075 mem: 3.36 + 04-03 22:02:21 | [098][020/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0071 mem: 3.36 + 04-03 22:02:22 | [098][030/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0071 ntime: 0076 mem: 3.36 + 04-03 22:02:24 | [098][040/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 22:02:25 | [098][050/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0068 ntime: 0075 mem: 3.36 + 04-03 22:02:26 | [098][060/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:02:28 | [098][070/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:02:29 | [098][080/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 22:02:30 | [098][090/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:02:31 | [098][100/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0073 mem: 3.36 + 04-03 22:02:33 | [098][110/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:02:34 | [098][120/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:02:35 | [098][130/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 22:02:37 | [098][140/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:02:38 | [098][150/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0054 ntime: 0084 mem: 3.36 + 04-03 22:02:40 | [098][160/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:02:41 | [098][170/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0085 mem: 3.36 + 04-03 22:02:42 | Time info >>>> elapsed: 39.33 mins remain: 357.96 mins + 04-03 22:02:42 | [099][000/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 22:02:44 | [099][010/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0055 ntime: 0085 mem: 3.36 + 04-03 22:02:45 | [099][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0065 ntime: 0082 mem: 3.36 + 04-03 22:02:47 | [099][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 22:02:48 | [099][040/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 22:02:49 | [099][050/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:02:51 | [099][060/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0055 ntime: 0082 mem: 3.36 + 04-03 22:02:52 | [099][070/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0086 mem: 3.36 + 04-03 22:02:53 | [099][080/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 22:02:55 | [099][090/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 22:02:56 | [099][100/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0047 ntime: 0089 mem: 3.36 + 04-03 22:02:57 | [099][110/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 22:02:59 | [099][120/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 22:03:00 | [099][130/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0087 mem: 3.36 + 04-03 22:03:01 | [099][140/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 22:03:03 | [099][150/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0075 mem: 3.36 + 04-03 22:03:04 | [099][160/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 22:03:05 | [099][170/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 22:03:06 | Time info >>>> elapsed: 39.73 mins remain: 357.60 mins + 04-03 22:03:06 | [100][000/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 22:03:08 | [100][010/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 22:03:09 | [100][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 22:03:10 | [100][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 22:03:12 | [100][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0070 ntime: 0081 mem: 3.36 + 04-03 22:03:13 | [100][050/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0058 ntime: 0083 mem: 3.36 + 04-03 22:03:15 | [100][060/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0077 ntime: 0076 mem: 3.36 + 04-03 22:03:16 | [100][070/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 22:03:18 | [100][080/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 22:03:19 | [100][090/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0046 ntime: 0082 mem: 3.36 + 04-03 22:03:20 | [100][100/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 22:03:22 | [100][110/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 22:03:23 | [100][120/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 22:03:24 | [100][130/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 22:03:26 | [100][140/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:03:27 | [100][150/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 22:03:28 | [100][160/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0058 ntime: 0082 mem: 3.36 + 04-03 22:03:29 | [100][170/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0046 ntime: 0086 mem: 3.36 + 04-03 22:03:31 | Time info >>>> elapsed: 40.14 mins remain: 357.28 mins + 04-03 22:03:31 | [101][000/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0053 ntime: 0068 mem: 3.36 + 04-03 22:03:32 | [101][010/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 22:03:33 | [101][020/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0061 ntime: 0079 mem: 3.36 + 04-03 22:03:35 | [101][030/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0051 ntime: 0087 mem: 3.36 + 04-03 22:03:36 | [101][040/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0071 ntime: 0081 mem: 3.36 + 04-03 22:03:38 | [101][050/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:03:39 | [101][060/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0072 mem: 3.36 + 04-03 22:03:40 | [101][070/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0087 mem: 3.36 + 04-03 22:03:42 | [101][080/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 22:03:43 | [101][090/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:03:44 | [101][100/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 22:03:46 | [101][110/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 22:03:47 | [101][120/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0046 ntime: 0080 mem: 3.36 + 04-03 22:03:48 | [101][130/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 22:03:49 | [101][140/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0052 ntime: 0086 mem: 3.36 + 04-03 22:03:51 | [101][150/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 22:03:52 | [101][160/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 22:03:54 | [101][170/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0062 ntime: 0079 mem: 3.36 + 04-03 22:03:55 | Time info >>>> elapsed: 40.54 mins remain: 356.89 mins + 04-03 22:03:55 | [102][000/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 22:03:56 | [102][010/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 22:03:57 | [102][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 22:03:59 | [102][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 22:04:00 | [102][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 22:04:01 | [102][050/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:04:02 | [102][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 22:04:04 | [102][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0057 ntime: 0079 mem: 3.36 + 04-03 22:04:05 | [102][080/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 22:04:06 | [102][090/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 22:04:08 | [102][100/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0062 ntime: 0078 mem: 3.36 + 04-03 22:04:09 | [102][110/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 22:04:11 | [102][120/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0068 ntime: 0082 mem: 3.36 + 04-03 22:04:12 | [102][130/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0068 ntime: 0077 mem: 3.36 + 04-03 22:04:14 | [102][140/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 22:04:15 | [102][150/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0073 mem: 3.36 + 04-03 22:04:16 | [102][160/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 22:04:17 | [102][170/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 22:04:19 | Time info >>>> elapsed: 40.94 mins remain: 356.51 mins + 04-03 22:04:19 | [103][000/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 22:04:20 | [103][010/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0057 ntime: 0077 mem: 3.36 + 04-03 22:04:21 | [103][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:04:23 | [103][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0045 ntime: 0074 mem: 3.36 + 04-03 22:04:24 | [103][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 22:04:25 | [103][050/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0071 mem: 3.36 + 04-03 22:04:26 | [103][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 22:04:28 | [103][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 22:04:29 | [103][080/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 22:04:30 | [103][090/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 22:04:32 | [103][100/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0083 mem: 3.36 + 04-03 22:04:33 | [103][110/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0072 ntime: 0093 mem: 3.36 + 04-03 22:04:34 | [103][120/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 22:04:36 | [103][130/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 22:04:37 | [103][140/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:04:38 | [103][150/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 22:04:40 | [103][160/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 22:04:41 | [103][170/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 22:04:42 | Time info >>>> elapsed: 41.33 mins remain: 356.09 mins + 04-03 22:04:42 | [104][000/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0067 ntime: 0083 mem: 3.36 + 04-03 22:04:44 | [104][010/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0063 ntime: 0097 mem: 3.36 + 04-03 22:04:45 | [104][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0062 ntime: 0085 mem: 3.36 + 04-03 22:04:47 | [104][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0057 ntime: 0079 mem: 3.36 + 04-03 22:04:49 | [104][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0061 ntime: 0086 mem: 3.36 + 04-03 22:04:50 | [104][050/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 22:04:51 | [104][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0068 ntime: 0076 mem: 3.36 + 04-03 22:04:53 | [104][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0055 ntime: 0085 mem: 3.36 + 04-03 22:04:54 | [104][080/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0066 ntime: 0087 mem: 3.36 + 04-03 22:04:56 | [104][090/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0058 ntime: 0082 mem: 3.36 + 04-03 22:04:57 | [104][100/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0053 ntime: 0085 mem: 3.36 + 04-03 22:04:58 | [104][110/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0086 mem: 3.36 + 04-03 22:05:00 | [104][120/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 22:05:01 | [104][130/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0092 mem: 3.36 + 04-03 22:05:02 | [104][140/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0057 ntime: 0093 mem: 3.36 + 04-03 22:05:04 | [104][150/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0063 mem: 3.36 + 04-03 22:05:05 | [104][160/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0070 ntime: 0081 mem: 3.36 + 04-03 22:05:06 | [104][170/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0074 mem: 3.36 + 04-03 22:05:07 | Time info >>>> elapsed: 41.75 mins remain: 355.87 mins + 04-03 22:05:07 | [105][000/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0079 mem: 3.36 + 04-03 22:05:09 | [105][010/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 22:05:10 | [105][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:05:11 | [105][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0059 ntime: 0086 mem: 3.36 + 04-03 22:05:13 | [105][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 22:05:14 | [105][050/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0072 mem: 3.36 + 04-03 22:05:15 | [105][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0055 ntime: 0074 mem: 3.36 + 04-03 22:05:17 | [105][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:05:18 | [105][080/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 22:05:19 | [105][090/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 22:05:21 | [105][100/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0075 mem: 3.36 + 04-03 22:05:22 | [105][110/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 22:05:23 | [105][120/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0073 mem: 3.36 + 04-03 22:05:24 | [105][130/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:05:26 | [105][140/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 22:05:27 | [105][150/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:05:28 | [105][160/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 22:05:30 | [105][170/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 22:05:31 | Time info >>>> elapsed: 42.14 mins remain: 355.39 mins + 04-03 22:05:31 | [106][000/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:05:32 | [106][010/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0058 ntime: 0072 mem: 3.36 + 04-03 22:05:33 | [106][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0057 ntime: 0082 mem: 3.36 + 04-03 22:05:35 | [106][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 22:05:36 | [106][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 22:05:37 | [106][050/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 22:05:39 | [106][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0066 ntime: 0081 mem: 3.36 + 04-03 22:05:40 | [106][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0074 mem: 3.36 + 04-03 22:05:41 | [106][080/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 22:05:43 | [106][090/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0071 mem: 3.36 + 04-03 22:05:44 | [106][100/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 22:05:45 | [106][110/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 22:05:47 | [106][120/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0086 mem: 3.36 + 04-03 22:05:48 | [106][130/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0069 ntime: 0082 mem: 3.36 + 04-03 22:05:49 | [106][140/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0062 ntime: 0079 mem: 3.36 + 04-03 22:05:51 | [106][150/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 22:05:52 | [106][160/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0054 ntime: 0082 mem: 3.36 + 04-03 22:05:54 | [106][170/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0086 mem: 3.36 + 04-03 22:05:55 | Time info >>>> elapsed: 42.54 mins remain: 355.04 mins + 04-03 22:05:55 | [107][000/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0085 mem: 3.36 + 04-03 22:05:56 | [107][010/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0086 mem: 3.36 + 04-03 22:05:58 | [107][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 22:05:59 | [107][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0086 mem: 3.36 + 04-03 22:06:00 | [107][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:06:02 | [107][050/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 22:06:03 | [107][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 22:06:04 | [107][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 22:06:06 | [107][080/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0057 ntime: 0086 mem: 3.36 + 04-03 22:06:07 | [107][090/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0061 ntime: 0078 mem: 3.36 + 04-03 22:06:08 | [107][100/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 22:06:10 | [107][110/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0072 mem: 3.36 + 04-03 22:06:11 | [107][120/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0086 mem: 3.36 + 04-03 22:06:12 | [107][130/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:06:14 | [107][140/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 22:06:15 | [107][150/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:06:16 | [107][160/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0085 mem: 3.36 + 04-03 22:06:18 | [107][170/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0070 mem: 3.36 + 04-03 22:06:19 | Time info >>>> elapsed: 42.94 mins remain: 354.66 mins + 04-03 22:06:19 | [108][000/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0054 ntime: 0083 mem: 3.36 + 04-03 22:06:20 | [108][010/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0062 ntime: 0074 mem: 3.36 + 04-03 22:06:22 | [108][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0054 ntime: 0077 mem: 3.36 + 04-03 22:06:23 | [108][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 22:06:24 | [108][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0060 ntime: 0082 mem: 3.36 + 04-03 22:06:26 | [108][050/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 22:06:27 | [108][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 22:06:28 | [108][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 22:06:30 | [108][080/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 22:06:31 | [108][090/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 22:06:32 | [108][100/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0045 ntime: 0071 mem: 3.36 + 04-03 22:06:34 | [108][110/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0088 mem: 3.36 + 04-03 22:06:35 | [108][120/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 22:06:36 | [108][130/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 22:06:37 | [108][140/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 22:06:39 | [108][150/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:06:40 | [108][160/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 22:06:41 | [108][170/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 22:06:42 | Time info >>>> elapsed: 43.33 mins remain: 354.21 mins + 04-03 22:06:42 | [109][000/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 22:06:44 | [109][010/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:06:45 | [109][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 22:06:46 | [109][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:06:48 | [109][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 22:06:49 | [109][050/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0092 mem: 3.36 + 04-03 22:06:51 | [109][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0055 ntime: 0080 mem: 3.36 + 04-03 22:06:52 | [109][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 22:06:53 | [109][080/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 22:06:55 | [109][090/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0060 ntime: 0080 mem: 3.36 + 04-03 22:06:56 | [109][100/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0089 mem: 3.36 + 04-03 22:06:57 | [109][110/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 22:06:59 | [109][120/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:07:00 | [109][130/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 22:07:01 | [109][140/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0066 ntime: 0080 mem: 3.36 + 04-03 22:07:03 | [109][150/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 22:07:04 | [109][160/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0073 mem: 3.36 + 04-03 22:07:05 | [109][170/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0072 mem: 3.36 + 04-03 22:07:06 | Time info >>>> elapsed: 43.73 mins remain: 353.83 mins + 04-03 22:07:06 | [110][000/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:07:08 | [110][010/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 22:07:09 | [110][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0045 ntime: 0076 mem: 3.36 + 04-03 22:07:10 | [110][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0045 ntime: 0076 mem: 3.36 + 04-03 22:07:11 | [110][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0078 mem: 3.36 + 04-03 22:07:13 | [110][050/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0074 mem: 3.36 + 04-03 22:07:14 | [110][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 22:07:15 | [110][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:07:17 | [110][080/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:07:18 | [110][090/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 22:07:19 | [110][100/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0078 mem: 3.36 + 04-03 22:07:20 | [110][110/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0057 ntime: 0081 mem: 3.36 + 04-03 22:07:22 | [110][120/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0068 ntime: 0085 mem: 3.36 + 04-03 22:07:23 | [110][130/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0045 ntime: 0080 mem: 3.36 + 04-03 22:07:24 | [110][140/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0044 ntime: 0082 mem: 3.36 + 04-03 22:07:26 | [110][150/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 22:07:27 | [110][160/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0078 mem: 3.36 + 04-03 22:07:28 | [110][170/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0056 ntime: 0083 mem: 3.36 + 04-03 22:07:29 | Time info >>>> elapsed: 44.12 mins remain: 353.35 mins + 04-03 22:07:30 | [111][000/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0046 ntime: 0075 mem: 3.36 + 04-03 22:07:31 | [111][010/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0066 ntime: 0083 mem: 3.36 + 04-03 22:07:32 | [111][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0073 mem: 3.36 + 04-03 22:07:33 | [111][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 22:07:35 | [111][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:07:36 | [111][050/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 22:07:37 | [111][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0078 mem: 3.36 + 04-03 22:07:38 | [111][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:07:40 | [111][080/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 22:07:41 | [111][090/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 22:07:42 | [111][100/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 22:07:44 | [111][110/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0055 ntime: 0087 mem: 3.36 + 04-03 22:07:45 | [111][120/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0057 ntime: 0083 mem: 3.36 + 04-03 22:07:47 | [111][130/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0056 ntime: 0079 mem: 3.36 + 04-03 22:07:48 | [111][140/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0054 ntime: 0073 mem: 3.36 + 04-03 22:07:49 | [111][150/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0054 ntime: 0073 mem: 3.36 + 04-03 22:07:51 | [111][160/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 22:07:52 | [111][170/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 22:07:53 | Time info >>>> elapsed: 44.51 mins remain: 352.90 mins + 04-03 22:07:53 | [112][000/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 22:07:55 | [112][010/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0065 ntime: 0089 mem: 3.36 + 04-03 22:07:56 | [112][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 22:07:57 | [112][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0055 ntime: 0075 mem: 3.36 + 04-03 22:07:59 | [112][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:08:00 | [112][050/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 22:08:01 | [112][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 22:08:02 | [112][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 22:08:04 | [112][080/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0056 ntime: 0072 mem: 3.36 + 04-03 22:08:05 | [112][090/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0086 mem: 3.36 + 04-03 22:08:06 | [112][100/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0055 ntime: 0082 mem: 3.36 + 04-03 22:08:08 | [112][110/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 22:08:09 | [112][120/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 22:08:10 | [112][130/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:08:11 | [112][140/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 22:08:13 | [112][150/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 22:08:14 | [112][160/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 22:08:15 | [112][170/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0078 mem: 3.36 + 04-03 22:08:16 | Time info >>>> elapsed: 44.90 mins remain: 352.44 mins + 04-03 22:08:16 | [113][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0045 ntime: 0073 mem: 3.36 + 04-03 22:08:18 | [113][010/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0075 mem: 3.36 + 04-03 22:08:19 | [113][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0069 mem: 3.36 + 04-03 22:08:20 | [113][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0045 ntime: 0074 mem: 3.36 + 04-03 22:08:21 | [113][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0073 mem: 3.36 + 04-03 22:08:23 | [113][050/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0091 mem: 3.36 + 04-03 22:08:24 | [113][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 22:08:25 | [113][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0080 mem: 3.36 + 04-03 22:08:26 | [113][080/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 22:08:28 | [113][090/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 22:08:29 | [113][100/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 22:08:30 | [113][110/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0045 ntime: 0075 mem: 3.36 + 04-03 22:08:31 | [113][120/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 22:08:33 | [113][130/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:08:34 | [113][140/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0071 mem: 3.36 + 04-03 22:08:35 | [113][150/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0053 ntime: 0083 mem: 3.36 + 04-03 22:08:37 | [113][160/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0056 ntime: 0077 mem: 3.36 + 04-03 22:08:38 | [113][170/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0056 ntime: 0083 mem: 3.36 + 04-03 22:08:39 | Time info >>>> elapsed: 45.27 mins remain: 351.85 mins + 04-03 22:08:39 | [114][000/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 22:08:40 | [114][010/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 22:08:41 | [114][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:08:43 | [114][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 22:08:44 | [114][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 22:08:45 | [114][050/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:08:47 | [114][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0085 mem: 3.36 + 04-03 22:08:48 | [114][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 22:08:50 | [114][080/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 22:08:51 | [114][090/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 22:08:52 | [114][100/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 22:08:54 | [114][110/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 22:08:55 | [114][120/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0072 mem: 3.36 + 04-03 22:08:56 | [114][130/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0053 ntime: 0088 mem: 3.36 + 04-03 22:08:58 | [114][140/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0075 mem: 3.36 + 04-03 22:08:59 | [114][150/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0053 ntime: 0084 mem: 3.36 + 04-03 22:09:01 | [114][160/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 22:09:02 | [114][170/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0056 ntime: 0081 mem: 3.36 + 04-03 22:09:03 | Time info >>>> elapsed: 45.68 mins remain: 351.51 mins + 04-03 22:09:03 | [115][000/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 22:09:04 | [115][010/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:09:06 | [115][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0072 mem: 3.36 + 04-03 22:09:07 | [115][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:09:08 | [115][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 22:09:10 | [115][050/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0054 ntime: 0083 mem: 3.36 + 04-03 22:09:11 | [115][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0080 mem: 3.36 + 04-03 22:09:12 | [115][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0073 mem: 3.36 + 04-03 22:09:14 | [115][080/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 22:09:15 | [115][090/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0089 mem: 3.36 + 04-03 22:09:16 | [115][100/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 22:09:17 | [115][110/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:09:19 | [115][120/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 22:09:20 | [115][130/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:09:22 | [115][140/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0055 ntime: 0082 mem: 3.36 + 04-03 22:09:23 | [115][150/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 22:09:24 | [115][160/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 22:09:26 | [115][170/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0075 mem: 3.36 + 04-03 22:09:27 | Time info >>>> elapsed: 46.07 mins remain: 351.10 mins + 04-03 22:09:27 | [116][000/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0046 ntime: 0081 mem: 3.36 + 04-03 22:09:28 | [116][010/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 22:09:29 | [116][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0086 mem: 3.36 + 04-03 22:09:31 | [116][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 22:09:32 | [116][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 22:09:33 | [116][050/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0055 ntime: 0084 mem: 3.36 + 04-03 22:09:35 | [116][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0084 ntime: 0080 mem: 3.36 + 04-03 22:09:36 | [116][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0055 ntime: 0080 mem: 3.36 + 04-03 22:09:38 | [116][080/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 22:09:39 | [116][090/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0054 ntime: 0077 mem: 3.36 + 04-03 22:09:40 | [116][100/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0045 ntime: 0079 mem: 3.36 + 04-03 22:09:41 | [116][110/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:09:43 | [116][120/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 22:09:44 | [116][130/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 22:09:45 | [116][140/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:09:47 | [116][150/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0053 ntime: 0075 mem: 3.36 + 04-03 22:09:48 | [116][160/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 22:09:49 | [116][170/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0056 ntime: 0088 mem: 3.36 + 04-03 22:09:50 | Time info >>>> elapsed: 46.47 mins remain: 350.70 mins + 04-03 22:09:51 | [117][000/179] predict_x0_loss: 0.015 glr: 5.0e-05 dtime: 0056 ntime: 0081 mem: 3.36 + 04-03 22:09:52 | [117][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 22:09:53 | [117][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0085 mem: 3.36 + 04-03 22:09:55 | [117][030/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0074 mem: 3.36 + 04-03 22:09:56 | [117][040/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 22:09:57 | [117][050/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:09:58 | [117][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0062 ntime: 0101 mem: 3.36 + 04-03 22:10:00 | [117][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0053 ntime: 0079 mem: 3.36 + 04-03 22:10:01 | [117][080/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0053 ntime: 0073 mem: 3.36 + 04-03 22:10:02 | [117][090/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:10:04 | [117][100/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:10:05 | [117][110/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0054 ntime: 0076 mem: 3.36 + 04-03 22:10:06 | [117][120/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 22:10:07 | [117][130/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0055 ntime: 0087 mem: 3.36 + 04-03 22:10:09 | [117][140/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:10:10 | [117][150/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 22:10:11 | [117][160/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 22:10:13 | [117][170/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 22:10:14 | Time info >>>> elapsed: 46.85 mins remain: 350.22 mins + 04-03 22:10:14 | [118][000/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 22:10:15 | [118][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 22:10:16 | [118][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 22:10:18 | [118][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 22:10:19 | [118][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:10:20 | [118][050/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 22:10:21 | [118][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 22:10:23 | [118][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0074 mem: 3.36 + 04-03 22:10:24 | [118][080/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 22:10:25 | [118][090/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 22:10:27 | [118][100/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:10:28 | [118][110/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 22:10:29 | [118][120/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 22:10:31 | [118][130/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0047 ntime: 0083 mem: 3.36 + 04-03 22:10:32 | [118][140/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0073 mem: 3.36 + 04-03 22:10:33 | [118][150/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0073 mem: 3.36 + 04-03 22:10:34 | [118][160/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 22:10:36 | [118][170/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 22:10:37 | Time info >>>> elapsed: 47.24 mins remain: 349.74 mins + 04-03 22:10:37 | [119][000/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 22:10:38 | [119][010/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 22:10:40 | [119][020/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 22:10:41 | [119][030/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0054 ntime: 0075 mem: 3.36 + 04-03 22:10:42 | [119][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:10:44 | [119][050/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0055 ntime: 0085 mem: 3.36 + 04-03 22:10:45 | [119][060/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 22:10:46 | [119][070/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 22:10:48 | [119][080/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0085 ntime: 0079 mem: 3.36 + 04-03 22:10:49 | [119][090/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 22:10:50 | [119][100/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 22:10:52 | [119][110/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:10:53 | [119][120/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 22:10:54 | [119][130/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0082 ntime: 0086 mem: 3.36 + 04-03 22:10:56 | [119][140/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 22:10:57 | [119][150/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0090 mem: 3.36 + 04-03 22:10:58 | [119][160/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0090 mem: 3.36 + 04-03 22:11:00 | [119][170/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 22:11:01 | Time info >>>> elapsed: 47.64 mins remain: 349.37 mins + 04-03 22:11:01 | [120][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0071 mem: 3.36 + 04-03 22:11:02 | [120][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:11:04 | [120][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 22:11:05 | [120][030/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:11:06 | [120][040/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0046 ntime: 0080 mem: 3.36 + 04-03 22:11:07 | [120][050/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0074 mem: 3.36 + 04-03 22:11:09 | [120][060/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 22:11:10 | [120][070/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 22:11:11 | [120][080/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0083 mem: 3.36 + 04-03 22:11:13 | [120][090/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0069 mem: 3.36 + 04-03 22:11:14 | [120][100/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 22:11:15 | [120][110/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0071 mem: 3.36 + 04-03 22:11:16 | [120][120/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0087 mem: 3.36 + 04-03 22:11:18 | [120][130/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:11:19 | [120][140/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 22:11:21 | [120][150/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 22:11:22 | [120][160/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0089 mem: 3.36 + 04-03 22:11:23 | [120][170/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0088 mem: 3.36 + 04-03 22:11:24 | Time info >>>> elapsed: 48.03 mins remain: 348.94 mins + 04-03 22:11:25 | [121][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0045 ntime: 0075 mem: 3.36 + 04-03 22:11:26 | [121][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 22:11:27 | [121][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 22:11:28 | [121][030/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0056 ntime: 0085 mem: 3.36 + 04-03 22:11:30 | [121][040/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 22:11:31 | [121][050/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:11:33 | [121][060/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 22:11:34 | [121][070/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 22:11:35 | [121][080/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 22:11:36 | [121][090/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:11:38 | [121][100/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 22:11:39 | [121][110/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0045 ntime: 0084 mem: 3.36 + 04-03 22:11:40 | [121][120/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0086 mem: 3.36 + 04-03 22:11:42 | [121][130/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:11:43 | [121][140/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0070 mem: 3.36 + 04-03 22:11:44 | [121][150/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 22:11:46 | [121][160/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 22:11:47 | [121][170/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0063 ntime: 0071 mem: 3.36 + 04-03 22:11:48 | Time info >>>> elapsed: 48.43 mins remain: 348.53 mins + 04-03 22:11:48 | [122][000/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0057 ntime: 0078 mem: 3.36 + 04-03 22:11:50 | [122][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0056 ntime: 0074 mem: 3.36 + 04-03 22:11:51 | [122][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 22:11:52 | [122][030/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0082 mem: 3.36 + 04-03 22:11:54 | [122][040/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 22:11:55 | [122][050/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0067 ntime: 0079 mem: 3.36 + 04-03 22:11:56 | [122][060/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0076 mem: 3.36 + 04-03 22:11:58 | [122][070/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 22:11:59 | [122][080/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 22:12:00 | [122][090/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0084 mem: 3.36 + 04-03 22:12:02 | [122][100/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0060 ntime: 0086 mem: 3.36 + 04-03 22:12:03 | [122][110/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 22:12:04 | [122][120/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 22:12:06 | [122][130/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0043 ntime: 0065 mem: 3.36 + 04-03 22:12:07 | [122][140/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 22:12:08 | [122][150/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0089 mem: 3.36 + 04-03 22:12:09 | [122][160/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0045 ntime: 0083 mem: 3.36 + 04-03 22:12:10 | [122][170/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0046 ntime: 0075 mem: 3.36 + 04-03 22:12:12 | Time info >>>> elapsed: 48.82 mins remain: 348.09 mins + 04-03 22:12:12 | [123][000/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 22:12:13 | [123][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 22:12:14 | [123][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0045 ntime: 0075 mem: 3.36 + 04-03 22:12:15 | [123][030/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0046 ntime: 0078 mem: 3.36 + 04-03 22:12:17 | [123][040/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 22:12:18 | [123][050/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 22:12:19 | [123][060/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 22:12:21 | [123][070/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0045 ntime: 0083 mem: 3.36 + 04-03 22:12:22 | [123][080/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0054 ntime: 0082 mem: 3.36 + 04-03 22:12:23 | [123][090/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0062 ntime: 0083 mem: 3.36 + 04-03 22:12:25 | [123][100/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 22:12:26 | [123][110/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0072 mem: 3.36 + 04-03 22:12:28 | [123][120/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 22:12:29 | [123][130/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 22:12:30 | [123][140/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 22:12:31 | [123][150/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 22:12:33 | [123][160/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0069 mem: 3.36 + 04-03 22:12:34 | [123][170/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 22:12:35 | Time info >>>> elapsed: 49.21 mins remain: 347.65 mins + 04-03 22:12:35 | [124][000/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 22:12:36 | [124][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 22:12:38 | [124][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0072 mem: 3.36 + 04-03 22:12:39 | [124][030/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0056 ntime: 0079 mem: 3.36 + 04-03 22:12:40 | [124][040/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0085 mem: 3.36 + 04-03 22:12:42 | [124][050/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0054 ntime: 0082 mem: 3.36 + 04-03 22:12:43 | [124][060/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0072 mem: 3.36 + 04-03 22:12:44 | [124][070/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0072 mem: 3.36 + 04-03 22:12:46 | [124][080/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 22:12:47 | [124][090/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:12:48 | [124][100/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 22:12:50 | [124][110/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0083 mem: 3.36 + 04-03 22:12:51 | [124][120/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 22:12:52 | [124][130/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:12:54 | [124][140/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0046 ntime: 0076 mem: 3.36 + 04-03 22:12:55 | [124][150/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0093 mem: 3.36 + 04-03 22:12:56 | [124][160/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0086 mem: 3.36 + 04-03 22:12:58 | [124][170/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 22:12:59 | Time info >>>> elapsed: 49.61 mins remain: 347.25 mins + 04-03 22:12:59 | [125][000/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 22:13:00 | [125][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0071 mem: 3.36 + 04-03 22:13:02 | [125][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 22:13:03 | [125][030/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0084 mem: 3.36 + 04-03 22:13:04 | [125][040/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 22:13:06 | [125][050/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0066 ntime: 0077 mem: 3.36 + 04-03 22:13:07 | [125][060/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 22:13:08 | [125][070/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0057 ntime: 0080 mem: 3.36 + 04-03 22:13:10 | [125][080/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 22:13:11 | [125][090/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 22:13:12 | [125][100/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0059 ntime: 0086 mem: 3.36 + 04-03 22:13:14 | [125][110/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:13:15 | [125][120/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0059 ntime: 0090 mem: 3.36 + 04-03 22:13:17 | [125][130/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0054 ntime: 0083 mem: 3.36 + 04-03 22:13:18 | [125][140/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 22:13:20 | [125][150/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 22:13:21 | [125][160/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 22:13:22 | [125][170/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0054 ntime: 0085 mem: 3.36 + 04-03 22:13:23 | Time info >>>> elapsed: 50.02 mins remain: 346.94 mins + 04-03 22:13:23 | [126][000/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0056 ntime: 0078 mem: 3.36 + 04-03 22:13:25 | [126][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 22:13:26 | [126][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 22:13:27 | [126][030/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:13:29 | [126][040/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 22:13:30 | [126][050/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0054 ntime: 0083 mem: 3.36 + 04-03 22:13:31 | [126][060/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0044 ntime: 0071 mem: 3.36 + 04-03 22:13:33 | [126][070/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:13:34 | [126][080/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 22:13:35 | [126][090/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0075 ntime: 0086 mem: 3.36 + 04-03 22:13:37 | [126][100/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 22:13:38 | [126][110/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 22:13:39 | [126][120/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0058 ntime: 0082 mem: 3.36 + 04-03 22:13:41 | [126][130/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0060 ntime: 0081 mem: 3.36 + 04-03 22:13:42 | [126][140/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 22:13:44 | [126][150/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 22:13:45 | [126][160/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0056 ntime: 0082 mem: 3.36 + 04-03 22:13:46 | [126][170/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:13:47 | Time info >>>> elapsed: 50.41 mins remain: 346.55 mins + 04-03 22:13:47 | [127][000/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0045 ntime: 0079 mem: 3.36 + 04-03 22:13:49 | [127][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 22:13:50 | [127][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0073 mem: 3.36 + 04-03 22:13:51 | [127][030/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0061 ntime: 0081 mem: 3.36 + 04-03 22:13:53 | [127][040/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 22:13:54 | [127][050/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:13:55 | [127][060/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:13:56 | [127][070/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 22:13:58 | [127][080/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0054 ntime: 0083 mem: 3.36 + 04-03 22:13:59 | [127][090/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 22:14:01 | [127][100/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 22:14:02 | [127][110/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 22:14:03 | [127][120/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0056 ntime: 0081 mem: 3.36 + 04-03 22:14:05 | [127][130/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:14:06 | [127][140/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:14:07 | [127][150/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:14:09 | [127][160/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 22:14:10 | [127][170/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 22:14:11 | Time info >>>> elapsed: 50.81 mins remain: 346.14 mins + 04-03 22:14:11 | [128][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:14:12 | [128][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0080 mem: 3.36 + 04-03 22:14:14 | [128][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0062 ntime: 0080 mem: 3.36 + 04-03 22:14:15 | [128][030/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0045 ntime: 0082 mem: 3.36 + 04-03 22:14:16 | [128][040/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 22:14:18 | [128][050/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 22:14:19 | [128][060/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 22:14:20 | [128][070/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:14:22 | [128][080/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0074 mem: 3.36 + 04-03 22:14:23 | [128][090/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0045 ntime: 0074 mem: 3.36 + 04-03 22:14:24 | [128][100/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0087 mem: 3.36 + 04-03 22:14:25 | [128][110/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 22:14:27 | [128][120/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 22:14:28 | [128][130/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0046 ntime: 0074 mem: 3.36 + 04-03 22:14:29 | [128][140/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 22:14:31 | [128][150/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0059 ntime: 0088 mem: 3.36 + 04-03 22:14:32 | [128][160/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 22:14:33 | [128][170/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:14:35 | Time info >>>> elapsed: 51.20 mins remain: 345.72 mins + 04-03 22:14:35 | [129][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 22:14:36 | [129][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0074 mem: 3.36 + 04-03 22:14:37 | [129][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0045 ntime: 0076 mem: 3.36 + 04-03 22:14:38 | [129][030/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:14:40 | [129][040/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 22:14:41 | [129][050/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0079 mem: 3.36 + 04-03 22:14:42 | [129][060/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 22:14:44 | [129][070/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 22:14:45 | [129][080/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0058 ntime: 0083 mem: 3.36 + 04-03 22:14:46 | [129][090/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 22:14:48 | [129][100/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 22:14:49 | [129][110/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 22:14:50 | [129][120/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0072 mem: 3.36 + 04-03 22:14:52 | [129][130/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 22:14:53 | [129][140/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 22:14:55 | [129][150/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:14:56 | [129][160/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0056 ntime: 0089 mem: 3.36 + 04-03 22:14:57 | [129][170/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 22:14:58 | Time info >>>> elapsed: 51.60 mins remain: 345.33 mins + 04-03 22:14:59 | [130][000/179] predict_x0_loss: 0.014 glr: 5.0e-05 dtime: 0066 ntime: 0085 mem: 3.36 + 04-03 22:15:00 | [130][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 22:15:01 | [130][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 22:15:03 | [130][030/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0080 mem: 3.36 + 04-03 22:15:04 | [130][040/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0060 ntime: 0072 mem: 3.36 + 04-03 22:15:05 | [130][050/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0068 ntime: 0081 mem: 3.36 + 04-03 22:15:07 | [130][060/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0071 ntime: 0081 mem: 3.36 + 04-03 22:15:08 | [130][070/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0072 ntime: 0076 mem: 3.36 + 04-03 22:15:10 | [130][080/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0066 ntime: 0079 mem: 3.36 + 04-03 22:15:11 | [130][090/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0090 mem: 3.36 + 04-03 22:15:13 | [130][100/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0087 mem: 3.36 + 04-03 22:15:14 | [130][110/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0090 mem: 3.36 + 04-03 22:15:15 | [130][120/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 22:15:17 | [130][130/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:15:18 | [130][140/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0070 mem: 3.36 + 04-03 22:15:19 | [130][150/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0071 mem: 3.36 + 04-03 22:15:21 | [130][160/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0057 ntime: 0081 mem: 3.36 + 04-03 22:15:22 | [130][170/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 22:15:23 | Time info >>>> elapsed: 52.01 mins remain: 345.02 mins + 04-03 22:15:23 | [131][000/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 22:15:25 | [131][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 22:15:26 | [131][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 22:15:27 | [131][030/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0054 ntime: 0076 mem: 3.36 + 04-03 22:15:28 | [131][040/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 22:15:30 | [131][050/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0087 mem: 3.36 + 04-03 22:15:31 | [131][060/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 22:15:32 | [131][070/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0058 ntime: 0078 mem: 3.36 + 04-03 22:15:34 | [131][080/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 22:15:35 | [131][090/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0065 ntime: 0078 mem: 3.36 + 04-03 22:15:37 | [131][100/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0069 mem: 3.36 + 04-03 22:15:38 | [131][110/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 22:15:39 | [131][120/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0072 ntime: 0073 mem: 3.36 + 04-03 22:15:41 | [131][130/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0062 ntime: 0072 mem: 3.36 + 04-03 22:15:42 | [131][140/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0058 ntime: 0074 mem: 3.36 + 04-03 22:15:44 | [131][150/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0073 ntime: 0082 mem: 3.36 + 04-03 22:15:45 | [131][160/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 22:15:46 | [131][170/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0057 ntime: 0078 mem: 3.36 + 04-03 22:15:47 | Time info >>>> elapsed: 52.42 mins remain: 344.68 mins + 04-03 22:15:48 | [132][000/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0063 ntime: 0087 mem: 3.36 + 04-03 22:15:49 | [132][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 22:15:50 | [132][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:15:51 | [132][030/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 22:15:53 | [132][040/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0058 ntime: 0083 mem: 3.36 + 04-03 22:15:54 | [132][050/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 22:15:56 | [132][060/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 22:15:57 | [132][070/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 22:15:58 | [132][080/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0074 mem: 3.36 + 04-03 22:16:00 | [132][090/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 22:16:01 | [132][100/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 22:16:02 | [132][110/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 22:16:04 | [132][120/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0078 ntime: 0083 mem: 3.36 + 04-03 22:16:05 | [132][130/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0046 ntime: 0078 mem: 3.36 + 04-03 22:16:06 | [132][140/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0073 mem: 3.36 + 04-03 22:16:08 | [132][150/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:16:09 | [132][160/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 22:16:10 | [132][170/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 22:16:12 | Time info >>>> elapsed: 52.82 mins remain: 344.32 mins + 04-03 22:16:12 | [133][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 22:16:13 | [133][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0089 mem: 3.36 + 04-03 22:16:14 | [133][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0079 mem: 3.36 + 04-03 22:16:16 | [133][030/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 22:16:17 | [133][040/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 22:16:18 | [133][050/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 22:16:20 | [133][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 22:16:21 | [133][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0070 mem: 3.36 + 04-03 22:16:22 | [133][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 22:16:23 | [133][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0055 ntime: 0083 mem: 3.36 + 04-03 22:16:25 | [133][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0074 mem: 3.36 + 04-03 22:16:26 | [133][110/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0057 ntime: 0079 mem: 3.36 + 04-03 22:16:27 | [133][120/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0073 mem: 3.36 + 04-03 22:16:29 | [133][130/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0071 ntime: 0074 mem: 3.36 + 04-03 22:16:30 | [133][140/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0058 ntime: 0080 mem: 3.36 + 04-03 22:16:31 | [133][150/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 22:16:33 | [133][160/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0053 ntime: 0073 mem: 3.36 + 04-03 22:16:34 | [133][170/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0059 ntime: 0076 mem: 3.36 + 04-03 22:16:35 | Time info >>>> elapsed: 53.22 mins remain: 343.91 mins + 04-03 22:16:35 | [134][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0067 ntime: 0080 mem: 3.36 + 04-03 22:16:37 | [134][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0072 mem: 3.36 + 04-03 22:16:38 | [134][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 22:16:39 | [134][030/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:16:41 | [134][040/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:16:42 | [134][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0087 mem: 3.36 + 04-03 22:16:43 | [134][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 22:16:45 | [134][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0091 mem: 3.36 + 04-03 22:16:46 | [134][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0088 mem: 3.36 + 04-03 22:16:47 | [134][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0069 ntime: 0087 mem: 3.36 + 04-03 22:16:49 | [134][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 22:16:50 | [134][110/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0056 ntime: 0084 mem: 3.36 + 04-03 22:16:51 | [134][120/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 22:16:53 | [134][130/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0074 mem: 3.36 + 04-03 22:16:54 | [134][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0072 ntime: 0082 mem: 3.36 + 04-03 22:16:56 | [134][150/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0086 mem: 3.36 + 04-03 22:16:57 | [134][160/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 22:16:59 | [134][170/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0055 ntime: 0090 mem: 3.36 + 04-03 22:17:00 | Time info >>>> elapsed: 53.62 mins remain: 343.59 mins + 04-03 22:17:00 | [135][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0054 ntime: 0077 mem: 3.36 + 04-03 22:17:01 | [135][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 22:17:03 | [135][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 22:17:04 | [135][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 22:17:05 | [135][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0073 mem: 3.36 + 04-03 22:17:06 | [135][050/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 22:17:08 | [135][060/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 22:17:09 | [135][070/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0073 mem: 3.36 + 04-03 22:17:10 | [135][080/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 22:17:12 | [135][090/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0091 mem: 3.36 + 04-03 22:17:13 | [135][100/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 22:17:14 | [135][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 22:17:15 | [135][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0064 ntime: 0080 mem: 3.36 + 04-03 22:17:17 | [135][130/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0054 ntime: 0074 mem: 3.36 + 04-03 22:17:18 | [135][140/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 22:17:19 | [135][150/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 22:17:21 | [135][160/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:17:22 | [135][170/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 22:17:23 | Time info >>>> elapsed: 54.01 mins remain: 343.14 mins + 04-03 22:17:23 | [136][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 22:17:25 | [136][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0054 ntime: 0077 mem: 3.36 + 04-03 22:17:26 | [136][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 22:17:28 | [136][030/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0065 ntime: 0081 mem: 3.36 + 04-03 22:17:29 | [136][040/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:17:30 | [136][050/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 22:17:31 | [136][060/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 22:17:33 | [136][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:17:34 | [136][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 22:17:35 | [136][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 22:17:37 | [136][100/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 22:17:38 | [136][110/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0075 mem: 3.36 + 04-03 22:17:39 | [136][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 22:17:41 | [136][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0046 ntime: 0076 mem: 3.36 + 04-03 22:17:42 | [136][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 22:17:43 | [136][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0060 ntime: 0079 mem: 3.36 + 04-03 22:17:45 | [136][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 22:17:46 | [136][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0070 ntime: 0084 mem: 3.36 + 04-03 22:17:47 | Time info >>>> elapsed: 54.41 mins remain: 342.77 mins + 04-03 22:17:47 | [137][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0063 ntime: 0081 mem: 3.36 + 04-03 22:17:49 | [137][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0065 ntime: 0083 mem: 3.36 + 04-03 22:17:50 | [137][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0073 mem: 3.36 + 04-03 22:17:52 | [137][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0065 ntime: 0084 mem: 3.36 + 04-03 22:17:53 | [137][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 22:17:54 | [137][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 22:17:56 | [137][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 22:17:57 | [137][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0046 ntime: 0076 mem: 3.36 + 04-03 22:17:58 | [137][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 22:18:00 | [137][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 22:18:01 | [137][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 22:18:02 | [137][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 22:18:03 | [137][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0072 mem: 3.36 + 04-03 22:18:05 | [137][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 22:18:06 | [137][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 22:18:07 | [137][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0074 mem: 3.36 + 04-03 22:18:09 | [137][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0054 ntime: 0082 mem: 3.36 + 04-03 22:18:10 | [137][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 22:18:11 | Time info >>>> elapsed: 54.81 mins remain: 342.37 mins + 04-03 22:18:11 | [138][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0053 ntime: 0074 mem: 3.36 + 04-03 22:18:12 | [138][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 22:18:14 | [138][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0046 ntime: 0080 mem: 3.36 + 04-03 22:18:15 | [138][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 22:18:16 | [138][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 22:18:18 | [138][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0089 mem: 3.36 + 04-03 22:18:19 | [138][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 22:18:20 | [138][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 22:18:22 | [138][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 22:18:23 | [138][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:18:24 | [138][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0083 mem: 3.36 + 04-03 22:18:26 | [138][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:18:27 | [138][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 22:18:28 | [138][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 22:18:30 | [138][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:18:31 | [138][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0045 ntime: 0074 mem: 3.36 + 04-03 22:18:32 | [138][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 22:18:33 | [138][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0057 ntime: 0087 mem: 3.36 + 04-03 22:18:35 | Time info >>>> elapsed: 55.20 mins remain: 341.94 mins + 04-03 22:18:35 | [139][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0084 mem: 3.36 + 04-03 22:18:36 | [139][010/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 22:18:37 | [139][020/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 22:18:39 | [139][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 22:18:40 | [139][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0085 mem: 3.36 + 04-03 22:18:41 | [139][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 22:18:43 | [139][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 22:18:44 | [139][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0090 mem: 3.36 + 04-03 22:18:45 | [139][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0116 ntime: 0086 mem: 3.36 + 04-03 22:18:47 | [139][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 22:18:48 | [139][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 22:18:50 | [139][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 22:18:51 | [139][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:18:52 | [139][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0067 ntime: 0078 mem: 3.36 + 04-03 22:18:54 | [139][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:18:55 | [139][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 22:18:56 | [139][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 22:18:58 | [139][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0055 ntime: 0093 mem: 3.36 + 04-03 22:18:59 | Time info >>>> elapsed: 55.61 mins remain: 341.61 mins + 04-03 22:18:59 | [140][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 22:19:01 | [140][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0086 mem: 3.36 + 04-03 22:19:02 | [140][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0057 ntime: 0089 mem: 3.36 + 04-03 22:19:03 | [140][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 22:19:05 | [140][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:19:06 | [140][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 22:19:07 | [140][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0063 mem: 3.36 + 04-03 22:19:09 | [140][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0056 ntime: 0084 mem: 3.36 + 04-03 22:19:10 | [140][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:19:11 | [140][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0054 ntime: 0077 mem: 3.36 + 04-03 22:19:12 | [140][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0074 mem: 3.36 + 04-03 22:19:14 | [140][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 22:19:15 | [140][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 22:19:16 | [140][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0074 mem: 3.36 + 04-03 22:19:18 | [140][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 22:19:19 | [140][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0086 mem: 3.36 + 04-03 22:19:20 | [140][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0073 mem: 3.36 + 04-03 22:19:22 | [140][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 22:19:23 | Time info >>>> elapsed: 56.01 mins remain: 341.21 mins + 04-03 22:19:23 | [141][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 22:19:24 | [141][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0071 mem: 3.36 + 04-03 22:19:26 | [141][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:19:27 | [141][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:19:28 | [141][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 22:19:30 | [141][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 22:19:31 | [141][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 22:19:32 | [141][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 22:19:34 | [141][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0055 ntime: 0089 mem: 3.36 + 04-03 22:19:35 | [141][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 22:19:36 | [141][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 22:19:38 | [141][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0054 ntime: 0083 mem: 3.36 + 04-03 22:19:39 | [141][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 22:19:40 | [141][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 22:19:42 | [141][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 22:19:43 | [141][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0073 mem: 3.36 + 04-03 22:19:44 | [141][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0046 ntime: 0071 mem: 3.36 + 04-03 22:19:45 | [141][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 22:19:47 | Time info >>>> elapsed: 56.40 mins remain: 340.80 mins + 04-03 22:19:47 | [142][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0070 mem: 3.36 + 04-03 22:19:48 | [142][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 22:19:49 | [142][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 22:19:51 | [142][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 22:19:52 | [142][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 22:19:53 | [142][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 22:19:55 | [142][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:19:56 | [142][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 22:19:57 | [142][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 22:19:59 | [142][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:20:00 | [142][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 22:20:01 | [142][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:20:03 | [142][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0087 mem: 3.36 + 04-03 22:20:04 | [142][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 22:20:05 | [142][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:20:06 | [142][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0073 mem: 3.36 + 04-03 22:20:08 | [142][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0086 mem: 3.36 + 04-03 22:20:09 | [142][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 22:20:10 | Time info >>>> elapsed: 56.80 mins remain: 340.39 mins + 04-03 22:20:10 | [143][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 22:20:12 | [143][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 22:20:13 | [143][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0085 mem: 3.36 + 04-03 22:20:14 | [143][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 22:20:16 | [143][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:20:17 | [143][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 22:20:18 | [143][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0084 mem: 3.36 + 04-03 22:20:20 | [143][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0056 ntime: 0080 mem: 3.36 + 04-03 22:20:21 | [143][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0061 ntime: 0081 mem: 3.36 + 04-03 22:20:22 | [143][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 22:20:24 | [143][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 22:20:25 | [143][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0071 mem: 3.36 + 04-03 22:20:26 | [143][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:20:27 | [143][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0062 ntime: 0085 mem: 3.36 + 04-03 22:20:29 | [143][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 22:20:30 | [143][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 22:20:31 | [143][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 22:20:33 | [143][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0088 mem: 3.36 + 04-03 22:20:34 | Time info >>>> elapsed: 57.19 mins remain: 339.97 mins + 04-03 22:20:34 | [144][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 22:20:35 | [144][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:20:37 | [144][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 22:20:38 | [144][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 22:20:39 | [144][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 22:20:41 | [144][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:20:42 | [144][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0061 ntime: 0080 mem: 3.36 + 04-03 22:20:43 | [144][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0087 mem: 3.36 + 04-03 22:20:45 | [144][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0086 ntime: 0080 mem: 3.36 + 04-03 22:20:46 | [144][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0096 mem: 3.36 + 04-03 22:20:48 | [144][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0055 ntime: 0084 mem: 3.36 + 04-03 22:20:49 | [144][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 22:20:50 | [144][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 22:20:52 | [144][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 22:20:53 | [144][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0056 ntime: 0078 mem: 3.36 + 04-03 22:20:54 | [144][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 22:20:56 | [144][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0090 mem: 3.36 + 04-03 22:20:57 | [144][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 22:20:58 | Time info >>>> elapsed: 57.60 mins remain: 339.62 mins + 04-03 22:20:58 | [145][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 22:21:00 | [145][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0094 mem: 3.36 + 04-03 22:21:01 | [145][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0073 ntime: 0092 mem: 3.36 + 04-03 22:21:02 | [145][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0060 ntime: 0083 mem: 3.36 + 04-03 22:21:04 | [145][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0087 mem: 3.36 + 04-03 22:21:05 | [145][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0068 ntime: 0075 mem: 3.36 + 04-03 22:21:07 | [145][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0066 ntime: 0079 mem: 3.36 + 04-03 22:21:08 | [145][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0091 ntime: 0088 mem: 3.36 + 04-03 22:21:10 | [145][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0063 ntime: 0080 mem: 3.36 + 04-03 22:21:11 | [145][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 22:21:12 | [145][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 22:21:14 | [145][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 22:21:15 | [145][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0056 ntime: 0077 mem: 3.36 + 04-03 22:21:16 | [145][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0071 mem: 3.36 + 04-03 22:21:17 | [145][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 22:21:19 | [145][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 22:21:20 | [145][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 22:21:21 | [145][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 22:21:22 | Time info >>>> elapsed: 58.00 mins remain: 339.27 mins + 04-03 22:21:23 | [146][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 22:21:24 | [146][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0089 mem: 3.36 + 04-03 22:21:25 | [146][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 22:21:27 | [146][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 22:21:28 | [146][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0056 ntime: 0072 mem: 3.36 + 04-03 22:21:29 | [146][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 22:21:31 | [146][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 22:21:32 | [146][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:21:33 | [146][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0089 mem: 3.36 + 04-03 22:21:35 | [146][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 22:21:36 | [146][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 22:21:37 | [146][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:21:39 | [146][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 22:21:40 | [146][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0057 ntime: 0080 mem: 3.36 + 04-03 22:21:41 | [146][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:21:42 | [146][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 22:21:44 | [146][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 22:21:45 | [146][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 22:21:46 | Time info >>>> elapsed: 58.40 mins remain: 338.86 mins + 04-03 22:21:46 | [147][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0045 ntime: 0075 mem: 3.36 + 04-03 22:21:48 | [147][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 22:21:49 | [147][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 22:21:50 | [147][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 22:21:52 | [147][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 22:21:53 | [147][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0086 ntime: 0075 mem: 3.36 + 04-03 22:21:54 | [147][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 22:21:55 | [147][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 22:21:57 | [147][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0074 mem: 3.36 + 04-03 22:21:58 | [147][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 22:21:59 | [147][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0072 mem: 3.36 + 04-03 22:22:00 | [147][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 22:22:02 | [147][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 22:22:03 | [147][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 22:22:04 | [147][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 22:22:06 | [147][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 22:22:07 | [147][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 22:22:08 | [147][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 22:22:09 | Time info >>>> elapsed: 58.78 mins remain: 338.40 mins + 04-03 22:22:09 | [148][000/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 22:22:11 | [148][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:22:12 | [148][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 22:22:13 | [148][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:22:15 | [148][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 22:22:16 | [148][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:22:17 | [148][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0054 ntime: 0085 mem: 3.36 + 04-03 22:22:19 | [148][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0046 ntime: 0082 mem: 3.36 + 04-03 22:22:20 | [148][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 22:22:21 | [148][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 22:22:22 | [148][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 22:22:24 | [148][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0064 ntime: 0084 mem: 3.36 + 04-03 22:22:25 | [148][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 22:22:27 | [148][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 22:22:28 | [148][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0070 mem: 3.36 + 04-03 22:22:29 | [148][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 22:22:30 | [148][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:22:32 | [148][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 22:22:33 | Time info >>>> elapsed: 59.17 mins remain: 337.97 mins + 04-03 22:22:33 | [149][000/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 22:22:34 | [149][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 22:22:36 | [149][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0078 ntime: 0084 mem: 3.36 + 04-03 22:22:37 | [149][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 22:22:38 | [149][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 22:22:40 | [149][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 22:22:41 | [149][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 22:22:42 | [149][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:22:44 | [149][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 22:22:45 | [149][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 22:22:46 | [149][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:22:48 | [149][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 22:22:49 | [149][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 22:22:50 | [149][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 22:22:52 | [149][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 22:22:53 | [149][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 22:22:54 | [149][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0079 mem: 3.36 + 04-03 22:22:56 | [149][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 22:22:57 | Time info >>>> elapsed: 59.58 mins remain: 337.61 mins + 04-03 22:22:57 | [150][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0057 ntime: 0080 mem: 3.36 + 04-03 22:22:59 | [150][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 22:23:00 | [150][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 22:23:01 | [150][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 22:23:02 | [150][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0086 mem: 3.36 + 04-03 22:23:04 | [150][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 22:23:05 | [150][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:23:06 | [150][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:23:08 | [150][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 22:23:09 | [150][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 22:23:10 | [150][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:23:12 | [150][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:23:13 | [150][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0046 ntime: 0076 mem: 3.36 + 04-03 22:23:14 | [150][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0058 ntime: 0085 mem: 3.36 + 04-03 22:23:16 | [150][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0056 ntime: 0083 mem: 3.36 + 04-03 22:23:17 | [150][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 22:23:18 | [150][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 22:23:20 | [150][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 22:23:21 | Time info >>>> elapsed: 59.97 mins remain: 337.19 mins + 04-03 22:23:21 | [151][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0046 ntime: 0075 mem: 3.36 + 04-03 22:23:22 | [151][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0072 mem: 3.36 + 04-03 22:23:23 | [151][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 22:23:25 | [151][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 22:23:26 | [151][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 22:23:27 | [151][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 22:23:29 | [151][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 22:23:30 | [151][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0058 ntime: 0079 mem: 3.36 + 04-03 22:23:31 | [151][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 22:23:33 | [151][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0058 ntime: 0073 mem: 3.36 + 04-03 22:23:34 | [151][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0055 ntime: 0074 mem: 3.36 + 04-03 22:23:35 | [151][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 22:23:37 | [151][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0083 mem: 3.36 + 04-03 22:23:38 | [151][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 22:23:39 | [151][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:23:41 | [151][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:23:42 | [151][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0056 ntime: 0082 mem: 3.36 + 04-03 22:23:43 | [151][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 22:23:44 | Time info >>>> elapsed: 60.37 mins remain: 336.77 mins + 04-03 22:23:44 | [152][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0058 ntime: 0077 mem: 3.36 + 04-03 22:23:46 | [152][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 22:23:47 | [152][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0074 mem: 3.36 + 04-03 22:23:48 | [152][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 22:23:50 | [152][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 22:23:51 | [152][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0073 mem: 3.36 + 04-03 22:23:52 | [152][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0058 ntime: 0081 mem: 3.36 + 04-03 22:23:54 | [152][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0073 mem: 3.36 + 04-03 22:23:55 | [152][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 22:23:56 | [152][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0054 ntime: 0084 mem: 3.36 + 04-03 22:23:58 | [152][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 22:23:59 | [152][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 22:24:00 | [152][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:24:01 | [152][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0072 mem: 3.36 + 04-03 22:24:03 | [152][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 22:24:04 | [152][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0071 mem: 3.36 + 04-03 22:24:05 | [152][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:24:07 | [152][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 22:24:08 | Time info >>>> elapsed: 60.76 mins remain: 336.35 mins + 04-03 22:24:08 | [153][000/179] predict_x0_loss: 0.013 glr: 5.0e-05 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 22:24:09 | [153][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 22:24:10 | [153][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0068 ntime: 0080 mem: 3.36 + 04-03 22:24:12 | [153][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 22:24:14 | [153][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 22:24:15 | [153][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0046 ntime: 0078 mem: 3.36 + 04-03 22:24:16 | [153][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 22:24:18 | [153][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:24:19 | [153][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0046 ntime: 0082 mem: 3.36 + 04-03 22:24:20 | [153][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0045 ntime: 0079 mem: 3.36 + 04-03 22:24:21 | [153][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:24:23 | [153][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0072 mem: 3.36 + 04-03 22:24:24 | [153][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0056 ntime: 0079 mem: 3.36 + 04-03 22:24:25 | [153][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0073 mem: 3.36 + 04-03 22:24:27 | [153][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 22:24:28 | [153][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 22:24:29 | [153][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:24:31 | [153][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0060 ntime: 0085 mem: 3.36 + 04-03 22:24:32 | Time info >>>> elapsed: 61.15 mins remain: 335.95 mins + 04-03 22:24:32 | [154][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0072 mem: 3.36 + 04-03 22:24:33 | [154][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 22:24:34 | [154][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 22:24:36 | [154][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 22:24:37 | [154][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:24:38 | [154][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0045 ntime: 0086 mem: 3.36 + 04-03 22:24:39 | [154][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 22:24:41 | [154][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0091 mem: 3.36 + 04-03 22:24:42 | [154][080/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0085 mem: 3.36 + 04-03 22:24:44 | [154][090/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0059 ntime: 0084 mem: 3.36 + 04-03 22:24:45 | [154][100/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0056 ntime: 0080 mem: 3.36 + 04-03 22:24:46 | [154][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:24:48 | [154][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0060 ntime: 0080 mem: 3.36 + 04-03 22:24:49 | [154][130/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0054 ntime: 0084 mem: 3.36 + 04-03 22:24:50 | [154][140/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0090 mem: 3.36 + 04-03 22:24:52 | [154][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 22:24:53 | [154][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 22:24:55 | [154][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 22:24:56 | Time info >>>> elapsed: 61.56 mins remain: 335.59 mins + 04-03 22:24:56 | [155][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0085 mem: 3.36 + 04-03 22:24:58 | [155][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 22:24:59 | [155][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 22:25:00 | [155][030/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 22:25:02 | [155][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 22:25:03 | [155][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 22:25:05 | [155][060/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0396 ntime: 0077 mem: 3.36 + 04-03 22:25:07 | [155][070/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0515 ntime: 0083 mem: 3.36 + 04-03 22:25:10 | [155][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0245 ntime: 0085 mem: 3.36 + 04-03 22:25:13 | [155][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0424 ntime: 0079 mem: 3.36 + 04-03 22:25:16 | [155][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0193 ntime: 0083 mem: 3.36 + 04-03 22:25:18 | [155][110/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:25:21 | [155][120/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0439 ntime: 0080 mem: 3.36 + 04-03 22:25:25 | [155][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0161 ntime: 0081 mem: 3.36 + 04-03 22:25:27 | [155][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 22:25:30 | [155][150/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0066 ntime: 0077 mem: 3.36 + 04-03 22:25:33 | [155][160/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0125 ntime: 0084 mem: 3.36 + 04-03 22:25:35 | [155][170/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 22:25:38 | Time info >>>> elapsed: 62.25 mins remain: 336.80 mins + 04-03 22:25:38 | [156][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 22:25:40 | [156][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0228 ntime: 0089 mem: 3.36 + 04-03 22:25:42 | [156][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0291 ntime: 0086 mem: 3.36 + 04-03 22:25:45 | [156][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0148 ntime: 0077 mem: 3.36 + 04-03 22:25:47 | [156][040/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0316 ntime: 0083 mem: 3.36 + 04-03 22:25:51 | [156][050/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0275 ntime: 0079 mem: 3.36 + 04-03 22:25:53 | [156][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 22:25:55 | [156][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0148 ntime: 0080 mem: 3.36 + 04-03 22:25:58 | [156][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0055 ntime: 0086 mem: 3.36 + 04-03 22:26:01 | [156][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 22:26:04 | [156][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0714 ntime: 0080 mem: 3.36 + 04-03 22:26:06 | [156][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0174 ntime: 0080 mem: 3.36 + 04-03 22:26:08 | [156][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0168 ntime: 0078 mem: 3.36 + 04-03 22:26:10 | [156][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0190 ntime: 0081 mem: 3.36 + 04-03 22:26:12 | [156][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:26:16 | [156][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0064 ntime: 0093 mem: 3.36 + 04-03 22:26:18 | [156][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0292 ntime: 0086 mem: 3.36 + 04-03 22:26:21 | [156][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 22:26:23 | Time info >>>> elapsed: 63.02 mins remain: 338.36 mins + 04-03 22:26:24 | [157][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0422 ntime: 0079 mem: 3.36 + 04-03 22:26:26 | [157][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0047 ntime: 0073 mem: 3.36 + 04-03 22:26:28 | [157][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 22:26:31 | [157][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0651 ntime: 0084 mem: 3.36 + 04-03 22:26:34 | [157][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0118 ntime: 0082 mem: 3.36 + 04-03 22:26:36 | [157][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0062 ntime: 0081 mem: 3.36 + 04-03 22:26:38 | [157][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0123 ntime: 0084 mem: 3.36 + 04-03 22:26:41 | [157][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0059 ntime: 0085 mem: 3.36 + 04-03 22:26:43 | [157][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 22:26:46 | [157][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 22:26:48 | [157][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 22:26:50 | [157][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0241 ntime: 0079 mem: 3.36 + 04-03 22:26:52 | [157][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 22:26:54 | [157][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 22:26:56 | [157][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 22:26:57 | [157][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0054 ntime: 0086 mem: 3.36 + 04-03 22:27:00 | [157][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0065 ntime: 0079 mem: 3.36 + 04-03 22:27:02 | [157][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0247 ntime: 0080 mem: 3.36 + 04-03 22:27:03 | Time info >>>> elapsed: 63.68 mins remain: 339.34 mins + 04-03 22:27:03 | [158][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0181 ntime: 0077 mem: 3.36 + 04-03 22:27:05 | [158][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0167 ntime: 0083 mem: 3.36 + 04-03 22:27:07 | [158][020/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 22:27:09 | [158][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0163 ntime: 0083 mem: 3.36 + 04-03 22:27:11 | [158][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0157 ntime: 0090 mem: 3.36 + 04-03 22:27:13 | [158][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0052 ntime: 0086 mem: 3.36 + 04-03 22:27:15 | [158][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0085 ntime: 0080 mem: 3.36 + 04-03 22:27:17 | [158][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0063 ntime: 0074 mem: 3.36 + 04-03 22:27:19 | [158][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0119 ntime: 0084 mem: 3.36 + 04-03 22:27:22 | [158][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0066 ntime: 0077 mem: 3.36 + 04-03 22:27:24 | [158][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 22:27:26 | [158][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 22:27:28 | [158][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0289 ntime: 0081 mem: 3.36 + 04-03 22:27:30 | [158][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 22:27:32 | [158][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0351 ntime: 0078 mem: 3.36 + 04-03 22:27:34 | [158][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 22:27:36 | [158][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0254 ntime: 0079 mem: 3.36 + 04-03 22:27:38 | [158][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0150 ntime: 0080 mem: 3.36 + 04-03 22:27:39 | Time info >>>> elapsed: 64.29 mins remain: 340.03 mins + 04-03 22:27:40 | [159][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 22:27:42 | [159][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0155 ntime: 0075 mem: 3.36 + 04-03 22:27:44 | [159][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0053 ntime: 0084 mem: 3.36 + 04-03 22:27:46 | [159][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0330 ntime: 0081 mem: 3.36 + 04-03 22:27:48 | [159][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0061 ntime: 0077 mem: 3.36 + 04-03 22:27:50 | [159][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0151 ntime: 0081 mem: 3.36 + 04-03 22:27:53 | [159][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0058 ntime: 0091 mem: 3.36 + 04-03 22:27:55 | [159][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0073 ntime: 0081 mem: 3.36 + 04-03 22:27:58 | [159][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0403 ntime: 0084 mem: 3.36 + 04-03 22:27:59 | [159][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 22:28:01 | [159][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0223 ntime: 0085 mem: 3.36 + 04-03 22:28:03 | [159][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 22:28:05 | [159][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0055 ntime: 0087 mem: 3.36 + 04-03 22:28:08 | [159][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0057 ntime: 0084 mem: 3.36 + 04-03 22:28:10 | [159][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0326 ntime: 0082 mem: 3.36 + 04-03 22:28:12 | [159][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0053 ntime: 0086 mem: 3.36 + 04-03 22:28:14 | [159][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0183 ntime: 0079 mem: 3.36 + 04-03 22:28:16 | [159][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0226 ntime: 0085 mem: 3.36 + 04-03 22:28:18 | Time info >>>> elapsed: 64.92 mins remain: 340.84 mins + 04-03 22:28:18 | [160][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0069 ntime: 0082 mem: 3.36 + 04-03 22:28:20 | [160][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0062 ntime: 0084 mem: 3.36 + 04-03 22:28:22 | [160][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0072 ntime: 0080 mem: 3.36 + 04-03 22:28:25 | [160][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0055 ntime: 0085 mem: 3.36 + 04-03 22:28:26 | [160][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0137 ntime: 0081 mem: 3.36 + 04-03 22:28:28 | [160][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0087 ntime: 0068 mem: 3.36 + 04-03 22:28:30 | [160][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 22:28:32 | [160][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0052 ntime: 0087 mem: 3.36 + 04-03 22:28:34 | [160][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0078 ntime: 0079 mem: 3.36 + 04-03 22:28:36 | [160][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:28:39 | [160][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0052 ntime: 0088 mem: 3.36 + 04-03 22:28:41 | [160][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0345 ntime: 0077 mem: 3.36 + 04-03 22:28:43 | [160][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0246 ntime: 0071 mem: 3.36 + 04-03 22:28:45 | [160][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0233 ntime: 0077 mem: 3.36 + 04-03 22:28:47 | [160][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0058 ntime: 0076 mem: 3.36 + 04-03 22:28:49 | [160][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 22:28:52 | [160][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 22:28:54 | [160][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0204 ntime: 0083 mem: 3.36 + 04-03 22:28:56 | Time info >>>> elapsed: 65.56 mins remain: 341.63 mins + 04-03 22:28:56 | [161][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0285 ntime: 0080 mem: 3.36 + 04-03 22:28:58 | [161][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0060 ntime: 0079 mem: 3.36 + 04-03 22:29:00 | [161][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0046 ntime: 0075 mem: 3.36 + 04-03 22:29:02 | [161][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0186 ntime: 0086 mem: 3.36 + 04-03 22:29:05 | [161][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 22:29:07 | [161][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0105 ntime: 0078 mem: 3.36 + 04-03 22:29:10 | [161][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 22:29:12 | [161][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0346 ntime: 0077 mem: 3.36 + 04-03 22:29:14 | [161][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0292 ntime: 0083 mem: 3.36 + 04-03 22:29:16 | [161][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0054 ntime: 0086 mem: 3.36 + 04-03 22:29:18 | [161][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0074 ntime: 0073 mem: 3.36 + 04-03 22:29:20 | [161][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 22:29:22 | [161][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0154 ntime: 0080 mem: 3.36 + 04-03 22:29:24 | [161][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0157 ntime: 0080 mem: 3.36 + 04-03 22:29:26 | [161][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:29:28 | [161][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0085 mem: 3.36 + 04-03 22:29:30 | [161][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0056 ntime: 0075 mem: 3.36 + 04-03 22:29:32 | [161][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0086 mem: 3.36 + 04-03 22:29:34 | Time info >>>> elapsed: 66.19 mins remain: 342.39 mins + 04-03 22:29:34 | [162][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0301 ntime: 0079 mem: 3.36 + 04-03 22:29:37 | [162][010/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0072 ntime: 0080 mem: 3.36 + 04-03 22:29:38 | [162][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0071 ntime: 0074 mem: 3.36 + 04-03 22:29:41 | [162][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 22:29:43 | [162][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0069 ntime: 0079 mem: 3.36 + 04-03 22:29:46 | [162][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0322 ntime: 0078 mem: 3.36 + 04-03 22:29:47 | [162][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0072 ntime: 0074 mem: 3.36 + 04-03 22:29:50 | [162][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0184 ntime: 0080 mem: 3.36 + 04-03 22:29:52 | [162][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0173 ntime: 0080 mem: 3.36 + 04-03 22:29:54 | [162][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0046 ntime: 0079 mem: 3.36 + 04-03 22:29:57 | [162][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0198 ntime: 0086 mem: 3.36 + 04-03 22:29:59 | [162][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0156 ntime: 0080 mem: 3.36 + 04-03 22:30:01 | [162][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0228 ntime: 0082 mem: 3.36 + 04-03 22:30:04 | [162][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0384 ntime: 0074 mem: 3.36 + 04-03 22:30:06 | [162][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0121 ntime: 0076 mem: 3.36 + 04-03 22:30:08 | [162][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0062 ntime: 0082 mem: 3.36 + 04-03 22:30:10 | [162][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 22:30:13 | [162][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0067 ntime: 0079 mem: 3.36 + 04-03 22:30:16 | Time info >>>> elapsed: 66.89 mins remain: 343.48 mins + 04-03 22:30:16 | [163][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:30:21 | [163][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 22:30:24 | [163][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0157 ntime: 0087 mem: 3.36 + 04-03 22:30:26 | [163][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0085 ntime: 0076 mem: 3.36 + 04-03 22:30:28 | [163][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0157 ntime: 0077 mem: 3.36 + 04-03 22:30:31 | [163][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0059 ntime: 0084 mem: 3.36 + 04-03 22:30:34 | [163][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0320 ntime: 0087 mem: 3.36 + 04-03 22:30:36 | [163][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0195 ntime: 0081 mem: 3.36 + 04-03 22:30:39 | [163][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0053 ntime: 0079 mem: 3.36 + 04-03 22:30:41 | [163][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0118 ntime: 0077 mem: 3.36 + 04-03 22:30:44 | [163][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 22:30:46 | [163][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0190 ntime: 0083 mem: 3.36 + 04-03 22:30:48 | [163][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0272 ntime: 0082 mem: 3.36 + 04-03 22:30:51 | [163][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0209 ntime: 0086 mem: 3.36 + 04-03 22:30:53 | [163][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0167 ntime: 0078 mem: 3.36 + 04-03 22:30:55 | [163][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0055 ntime: 0077 mem: 3.36 + 04-03 22:30:57 | [163][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0276 ntime: 0074 mem: 3.36 + 04-03 22:31:00 | [163][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0066 ntime: 0085 mem: 3.36 + 04-03 22:31:02 | Time info >>>> elapsed: 67.66 mins remain: 344.92 mins + 04-03 22:31:02 | [164][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 22:31:05 | [164][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0176 ntime: 0082 mem: 3.36 + 04-03 22:31:07 | [164][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0087 mem: 3.36 + 04-03 22:31:10 | [164][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0053 ntime: 0090 mem: 3.36 + 04-03 22:31:13 | [164][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:31:15 | [164][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 22:31:18 | [164][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0271 ntime: 0082 mem: 3.36 + 04-03 22:31:20 | [164][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0095 ntime: 0082 mem: 3.36 + 04-03 22:31:23 | [164][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0232 ntime: 0079 mem: 3.36 + 04-03 22:31:25 | [164][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0220 ntime: 0079 mem: 3.36 + 04-03 22:31:27 | [164][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0328 ntime: 0089 mem: 3.36 + 04-03 22:31:31 | [164][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0200 ntime: 0083 mem: 3.36 + 04-03 22:31:35 | [164][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0054 ntime: 0085 mem: 3.36 + 04-03 22:31:36 | [164][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 22:31:39 | [164][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0046 ntime: 0080 mem: 3.36 + 04-03 22:31:42 | [164][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0086 mem: 3.36 + 04-03 22:31:45 | [164][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0217 ntime: 0081 mem: 3.36 + 04-03 22:31:47 | [164][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0055 ntime: 0082 mem: 3.36 + 04-03 22:31:49 | Time info >>>> elapsed: 68.45 mins remain: 346.37 mins + 04-03 22:31:49 | [165][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0137 ntime: 0081 mem: 3.36 + 04-03 22:31:51 | [165][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0059 ntime: 0081 mem: 3.36 + 04-03 22:31:53 | [165][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0137 ntime: 0078 mem: 3.36 + 04-03 22:31:55 | [165][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 22:31:58 | [165][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0059 ntime: 0084 mem: 3.36 + 04-03 22:32:00 | [165][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:32:02 | [165][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 22:32:04 | [165][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0253 ntime: 0085 mem: 3.36 + 04-03 22:32:07 | [165][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:32:09 | [165][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 22:32:12 | [165][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0256 ntime: 0085 mem: 3.36 + 04-03 22:32:14 | [165][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0309 ntime: 0079 mem: 3.36 + 04-03 22:32:17 | [165][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0191 ntime: 0079 mem: 3.36 + 04-03 22:32:20 | [165][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0056 ntime: 0079 mem: 3.36 + 04-03 22:32:23 | [165][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0056 ntime: 0084 mem: 3.36 + 04-03 22:32:25 | [165][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0133 ntime: 0077 mem: 3.36 + 04-03 22:32:27 | [165][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0154 ntime: 0074 mem: 3.36 + 04-03 22:32:29 | [165][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0188 ntime: 0076 mem: 3.36 + 04-03 22:32:31 | Time info >>>> elapsed: 69.14 mins remain: 347.39 mins + 04-03 22:32:31 | [166][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 22:32:34 | [166][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0403 ntime: 0088 mem: 3.36 + 04-03 22:32:36 | [166][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0215 ntime: 0085 mem: 3.36 + 04-03 22:32:38 | [166][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0111 ntime: 0080 mem: 3.36 + 04-03 22:32:40 | [166][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0208 ntime: 0072 mem: 3.36 + 04-03 22:32:42 | [166][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0186 ntime: 0080 mem: 3.36 + 04-03 22:32:45 | [166][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0236 ntime: 0082 mem: 3.36 + 04-03 22:32:47 | [166][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0234 ntime: 0079 mem: 3.36 + 04-03 22:32:49 | [166][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0508 ntime: 0081 mem: 3.36 + 04-03 22:32:51 | [166][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:32:53 | [166][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0060 ntime: 0083 mem: 3.36 + 04-03 22:32:55 | [166][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 22:32:57 | [166][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 22:32:59 | [166][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0171 ntime: 0087 mem: 3.36 + 04-03 22:33:01 | [166][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:33:04 | [166][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0585 ntime: 0080 mem: 3.36 + 04-03 22:33:07 | [166][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0321 ntime: 0081 mem: 3.36 + 04-03 22:33:09 | [166][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0219 ntime: 0080 mem: 3.36 + 04-03 22:33:10 | Time info >>>> elapsed: 69.80 mins remain: 348.16 mins + 04-03 22:33:10 | [167][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0062 ntime: 0074 mem: 3.36 + 04-03 22:33:13 | [167][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0046 ntime: 0061 mem: 3.36 + 04-03 22:33:15 | [167][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0257 ntime: 0090 mem: 3.36 + 04-03 22:33:17 | [167][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0148 ntime: 0082 mem: 3.36 + 04-03 22:33:19 | [167][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0211 ntime: 0082 mem: 3.36 + 04-03 22:33:21 | [167][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0057 ntime: 0083 mem: 3.36 + 04-03 22:33:24 | [167][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 22:33:27 | [167][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0076 ntime: 0075 mem: 3.36 + 04-03 22:33:29 | [167][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0087 mem: 3.36 + 04-03 22:33:31 | [167][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0071 ntime: 0078 mem: 3.36 + 04-03 22:33:33 | [167][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0074 ntime: 0079 mem: 3.36 + 04-03 22:33:36 | [167][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 22:33:38 | [167][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0188 ntime: 0084 mem: 3.36 + 04-03 22:33:39 | [167][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0046 ntime: 0080 mem: 3.36 + 04-03 22:33:42 | [167][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0104 ntime: 0081 mem: 3.36 + 04-03 22:33:44 | [167][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0077 ntime: 0080 mem: 3.36 + 04-03 22:33:47 | [167][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0398 ntime: 0080 mem: 3.36 + 04-03 22:33:49 | [167][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0054 ntime: 0082 mem: 3.36 + 04-03 22:33:50 | Time info >>>> elapsed: 70.47 mins remain: 348.98 mins + 04-03 22:33:51 | [168][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0225 ntime: 0085 mem: 3.36 + 04-03 22:33:54 | [168][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:33:57 | [168][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0128 ntime: 0084 mem: 3.36 + 04-03 22:33:59 | [168][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0080 ntime: 0074 mem: 3.36 + 04-03 22:34:02 | [168][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 22:34:04 | [168][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0271 ntime: 0081 mem: 3.36 + 04-03 22:34:06 | [168][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0135 ntime: 0079 mem: 3.36 + 04-03 22:34:08 | [168][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0189 ntime: 0081 mem: 3.36 + 04-03 22:34:10 | [168][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0059 ntime: 0088 mem: 3.36 + 04-03 22:34:12 | [168][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0495 ntime: 0077 mem: 3.36 + 04-03 22:34:14 | [168][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 22:34:16 | [168][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0210 ntime: 0086 mem: 3.36 + 04-03 22:34:19 | [168][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0373 ntime: 0082 mem: 3.36 + 04-03 22:34:20 | [168][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 22:34:23 | [168][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0054 ntime: 0075 mem: 3.36 + 04-03 22:34:25 | [168][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0141 ntime: 0080 mem: 3.36 + 04-03 22:34:27 | [168][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0350 ntime: 0057 mem: 3.36 + 04-03 22:34:30 | [168][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0139 ntime: 0076 mem: 3.36 + 04-03 22:34:31 | Time info >>>> elapsed: 71.15 mins remain: 349.86 mins + 04-03 22:34:32 | [169][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 22:34:33 | [169][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0046 ntime: 0071 mem: 3.36 + 04-03 22:34:36 | [169][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0218 ntime: 0083 mem: 3.36 + 04-03 22:34:38 | [169][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0329 ntime: 0083 mem: 3.36 + 04-03 22:34:40 | [169][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 22:34:42 | [169][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:34:44 | [169][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 22:34:47 | [169][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0193 ntime: 0086 mem: 3.36 + 04-03 22:34:49 | [169][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0234 ntime: 0083 mem: 3.36 + 04-03 22:34:51 | [169][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0207 ntime: 0079 mem: 3.36 + 04-03 22:34:53 | [169][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0063 ntime: 0084 mem: 3.36 + 04-03 22:34:56 | [169][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0254 ntime: 0079 mem: 3.36 + 04-03 22:34:58 | [169][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0206 ntime: 0077 mem: 3.36 + 04-03 22:35:00 | [169][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 22:35:03 | [169][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0052 ntime: 0075 mem: 3.36 + 04-03 22:35:05 | [169][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0426 ntime: 0072 mem: 3.36 + 04-03 22:35:07 | [169][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0240 ntime: 0077 mem: 3.36 + 04-03 22:35:09 | [169][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0054 ntime: 0077 mem: 3.36 + 04-03 22:35:11 | Time info >>>> elapsed: 71.81 mins remain: 350.61 mins + 04-03 22:35:11 | [170][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0196 ntime: 0078 mem: 3.36 + 04-03 22:35:13 | [170][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0065 ntime: 0089 mem: 3.36 + 04-03 22:35:15 | [170][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0061 ntime: 0083 mem: 3.36 + 04-03 22:35:17 | [170][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0052 ntime: 0087 mem: 3.36 + 04-03 22:35:19 | [170][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0082 ntime: 0079 mem: 3.36 + 04-03 22:35:22 | [170][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0119 ntime: 0081 mem: 3.36 + 04-03 22:35:24 | [170][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0147 ntime: 0078 mem: 3.36 + 04-03 22:35:26 | [170][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0089 mem: 3.36 + 04-03 22:35:29 | [170][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 22:35:31 | [170][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0165 ntime: 0085 mem: 3.36 + 04-03 22:35:32 | [170][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0066 ntime: 0078 mem: 3.36 + 04-03 22:35:34 | [170][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0053 ntime: 0091 mem: 3.36 + 04-03 22:35:36 | [170][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 22:35:38 | [170][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0091 ntime: 0078 mem: 3.36 + 04-03 22:35:41 | [170][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0333 ntime: 0091 mem: 3.36 + 04-03 22:35:42 | [170][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:35:45 | [170][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0170 ntime: 0087 mem: 3.36 + 04-03 22:35:47 | [170][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0055 ntime: 0080 mem: 3.36 + 04-03 22:35:49 | Time info >>>> elapsed: 72.44 mins remain: 351.18 mins + 04-03 22:35:49 | [171][000/179] predict_x0_loss: 0.012 glr: 5.0e-05 dtime: 0098 ntime: 0077 mem: 3.36 + 04-03 22:35:52 | [171][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0077 ntime: 0080 mem: 3.36 + 04-03 22:35:54 | [171][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0053 ntime: 0072 mem: 3.36 + 04-03 22:35:56 | [171][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0052 ntime: 0085 mem: 3.36 + 04-03 22:35:58 | [171][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0176 ntime: 0084 mem: 3.36 + 04-03 22:36:00 | [171][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0053 ntime: 0073 mem: 3.36 + 04-03 22:36:02 | [171][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 22:36:05 | [171][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 22:36:08 | [171][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 22:36:10 | [171][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0061 ntime: 0079 mem: 3.36 + 04-03 22:36:12 | [171][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0058 ntime: 0077 mem: 3.36 + 04-03 22:36:14 | [171][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0055 ntime: 0075 mem: 3.36 + 04-03 22:36:17 | [171][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0402 ntime: 0083 mem: 3.36 + 04-03 22:36:19 | [171][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 22:36:21 | [171][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0358 ntime: 0081 mem: 3.36 + 04-03 22:36:23 | [171][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0259 ntime: 0078 mem: 3.36 + 04-03 22:36:26 | [171][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0345 ntime: 0084 mem: 3.36 + 04-03 22:36:27 | [171][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 22:36:29 | Time info >>>> elapsed: 73.11 mins remain: 351.96 mins + 04-03 22:36:29 | [172][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 22:36:31 | [172][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 22:36:33 | [172][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 22:36:36 | [172][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0246 ntime: 0080 mem: 3.36 + 04-03 22:36:38 | [172][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 22:36:40 | [172][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0152 ntime: 0086 mem: 3.36 + 04-03 22:36:42 | [172][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0064 ntime: 0068 mem: 3.36 + 04-03 22:36:44 | [172][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 22:36:46 | [172][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0096 ntime: 0077 mem: 3.36 + 04-03 22:36:48 | [172][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0061 ntime: 0087 mem: 3.36 + 04-03 22:36:50 | [172][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0053 ntime: 0085 mem: 3.36 + 04-03 22:36:52 | [172][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0139 ntime: 0079 mem: 3.36 + 04-03 22:36:54 | [172][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:36:56 | [172][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0315 ntime: 0085 mem: 3.36 + 04-03 22:36:58 | [172][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0052 ntime: 0087 mem: 3.36 + 04-03 22:37:00 | [172][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0117 ntime: 0085 mem: 3.36 + 04-03 22:37:02 | [172][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0057 ntime: 0083 mem: 3.36 + 04-03 22:37:04 | [172][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0275 ntime: 0076 mem: 3.36 + 04-03 22:37:06 | Time info >>>> elapsed: 73.72 mins remain: 352.41 mins + 04-03 22:37:06 | [173][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0178 ntime: 0083 mem: 3.36 + 04-03 22:37:08 | [173][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0171 ntime: 0084 mem: 3.36 + 04-03 22:37:10 | [173][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0259 ntime: 0083 mem: 3.36 + 04-03 22:37:11 | [173][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 22:37:14 | [173][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0085 mem: 3.36 + 04-03 22:37:15 | [173][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0053 ntime: 0079 mem: 3.36 + 04-03 22:37:18 | [173][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0288 ntime: 0084 mem: 3.36 + 04-03 22:37:19 | [173][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0044 ntime: 0054 mem: 3.36 + 04-03 22:37:21 | [173][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 22:37:23 | [173][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0208 ntime: 0078 mem: 3.36 + 04-03 22:37:25 | [173][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 22:37:27 | [173][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0158 ntime: 0085 mem: 3.36 + 04-03 22:37:29 | [173][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0095 ntime: 0073 mem: 3.36 + 04-03 22:37:31 | [173][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 22:37:33 | [173][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:37:36 | [173][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0399 ntime: 0076 mem: 3.36 + 04-03 22:37:38 | [173][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0057 ntime: 0087 mem: 3.36 + 04-03 22:37:40 | [173][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 22:37:43 | Time info >>>> elapsed: 74.35 mins remain: 352.95 mins + 04-03 22:37:43 | [174][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 22:37:46 | [174][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0070 ntime: 0072 mem: 3.36 + 04-03 22:37:49 | [174][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0088 mem: 3.36 + 04-03 22:37:52 | [174][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0456 ntime: 0088 mem: 3.36 + 04-03 22:37:54 | [174][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0141 ntime: 0076 mem: 3.36 + 04-03 22:37:58 | [174][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0087 mem: 3.36 + 04-03 22:38:00 | [174][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0275 ntime: 0060 mem: 3.36 + 04-03 22:38:02 | [174][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0052 ntime: 0085 mem: 3.36 + 04-03 22:38:04 | [174][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0310 ntime: 0083 mem: 3.36 + 04-03 22:38:06 | [174][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:38:09 | [174][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:38:11 | [174][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0224 ntime: 0078 mem: 3.36 + 04-03 22:38:13 | [174][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0181 ntime: 0082 mem: 3.36 + 04-03 22:38:15 | [174][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 22:38:17 | [174][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0062 ntime: 0079 mem: 3.36 + 04-03 22:38:19 | [174][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0492 ntime: 0085 mem: 3.36 + 04-03 22:38:21 | [174][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0085 ntime: 0078 mem: 3.36 + 04-03 22:38:24 | [174][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0265 ntime: 0080 mem: 3.36 + 04-03 22:38:26 | Time info >>>> elapsed: 75.05 mins remain: 353.82 mins + 04-03 22:38:26 | [175][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0325 ntime: 0077 mem: 3.36 + 04-03 22:38:28 | [175][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0091 ntime: 0077 mem: 3.36 + 04-03 22:38:30 | [175][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0065 ntime: 0075 mem: 3.36 + 04-03 22:38:32 | [175][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0124 ntime: 0083 mem: 3.36 + 04-03 22:38:35 | [175][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0077 ntime: 0081 mem: 3.36 + 04-03 22:38:37 | [175][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0058 ntime: 0081 mem: 3.36 + 04-03 22:38:39 | [175][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 22:38:42 | [175][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0143 ntime: 0086 mem: 3.36 + 04-03 22:38:44 | [175][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0057 ntime: 0082 mem: 3.36 + 04-03 22:38:48 | [175][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0060 ntime: 0083 mem: 3.36 + 04-03 22:38:50 | [175][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 22:38:52 | [175][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0081 ntime: 0077 mem: 3.36 + 04-03 22:38:54 | [175][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 22:38:57 | [175][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0171 ntime: 0079 mem: 3.36 + 04-03 22:38:59 | [175][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 22:39:01 | [175][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0054 ntime: 0088 mem: 3.36 + 04-03 22:39:03 | [175][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0163 ntime: 0080 mem: 3.36 + 04-03 22:39:06 | [175][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:39:08 | Time info >>>> elapsed: 75.76 mins remain: 354.68 mins + 04-03 22:39:08 | [176][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0103 ntime: 0085 mem: 3.36 + 04-03 22:39:10 | [176][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 22:39:12 | [176][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0059 ntime: 0071 mem: 3.36 + 04-03 22:39:15 | [176][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0268 ntime: 0077 mem: 3.36 + 04-03 22:39:17 | [176][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 22:39:19 | [176][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0058 ntime: 0080 mem: 3.36 + 04-03 22:39:20 | [176][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0052 ntime: 0072 mem: 3.36 + 04-03 22:39:23 | [176][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0362 ntime: 0080 mem: 3.36 + 04-03 22:39:25 | [176][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 22:39:27 | [176][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0224 ntime: 0079 mem: 3.36 + 04-03 22:39:29 | [176][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0144 ntime: 0079 mem: 3.36 + 04-03 22:39:31 | [176][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 22:39:34 | [176][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0906 ntime: 0080 mem: 3.36 + 04-03 22:39:37 | [176][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0257 ntime: 0079 mem: 3.36 + 04-03 22:39:39 | [176][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0331 ntime: 0085 mem: 3.36 + 04-03 22:39:41 | [176][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0062 ntime: 0078 mem: 3.36 + 04-03 22:39:43 | [176][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 22:39:46 | [176][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 22:39:48 | Time info >>>> elapsed: 76.43 mins remain: 355.36 mins + 04-03 22:39:48 | [177][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0137 ntime: 0085 mem: 3.36 + 04-03 22:39:50 | [177][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0072 ntime: 0074 mem: 3.36 + 04-03 22:39:53 | [177][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0080 ntime: 0078 mem: 3.36 + 04-03 22:39:54 | [177][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0166 ntime: 0083 mem: 3.36 + 04-03 22:39:57 | [177][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0166 ntime: 0081 mem: 3.36 + 04-03 22:39:59 | [177][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0199 ntime: 0080 mem: 3.36 + 04-03 22:40:01 | [177][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 22:40:04 | [177][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0477 ntime: 0076 mem: 3.36 + 04-03 22:40:06 | [177][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0141 ntime: 0084 mem: 3.36 + 04-03 22:40:09 | [177][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0159 ntime: 0085 mem: 3.36 + 04-03 22:40:11 | [177][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0177 ntime: 0079 mem: 3.36 + 04-03 22:40:13 | [177][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0057 ntime: 0077 mem: 3.36 + 04-03 22:40:15 | [177][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0158 ntime: 0087 mem: 3.36 + 04-03 22:40:17 | [177][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0069 ntime: 0087 mem: 3.36 + 04-03 22:40:19 | [177][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0064 ntime: 0074 mem: 3.36 + 04-03 22:40:22 | [177][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0085 ntime: 0073 mem: 3.36 + 04-03 22:40:24 | [177][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 22:40:26 | [177][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0191 ntime: 0082 mem: 3.36 + 04-03 22:40:28 | Time info >>>> elapsed: 77.09 mins remain: 356.02 mins + 04-03 22:40:28 | [178][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:40:30 | [178][010/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0055 ntime: 0088 mem: 3.36 + 04-03 22:40:33 | [178][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0130 ntime: 0085 mem: 3.36 + 04-03 22:40:36 | [178][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:40:38 | [178][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0146 ntime: 0082 mem: 3.36 + 04-03 22:40:40 | [178][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0480 ntime: 0078 mem: 3.36 + 04-03 22:40:43 | [178][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0063 ntime: 0084 mem: 3.36 + 04-03 22:40:45 | [178][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0175 ntime: 0077 mem: 3.36 + 04-03 22:40:47 | [178][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0066 ntime: 0078 mem: 3.36 + 04-03 22:40:49 | [178][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0141 ntime: 0079 mem: 3.36 + 04-03 22:40:51 | [178][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 22:40:54 | [178][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 22:40:57 | [178][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 22:40:59 | [178][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0096 ntime: 0077 mem: 3.36 + 04-03 22:41:01 | [178][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 22:41:04 | [178][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0194 ntime: 0073 mem: 3.36 + 04-03 22:41:06 | [178][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 22:41:08 | [178][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 22:41:09 | Time info >>>> elapsed: 77.78 mins remain: 356.74 mins + 04-03 22:41:09 | [179][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0273 ntime: 0082 mem: 3.36 + 04-03 22:41:11 | [179][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0206 ntime: 0077 mem: 3.36 + 04-03 22:41:14 | [179][020/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0717 ntime: 0081 mem: 3.36 + 04-03 22:41:16 | [179][030/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0064 ntime: 0078 mem: 3.36 + 04-03 22:41:19 | [179][040/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0261 ntime: 0086 mem: 3.36 + 04-03 22:41:21 | [179][050/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 22:41:23 | [179][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0180 ntime: 0080 mem: 3.36 + 04-03 22:41:26 | [179][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0216 ntime: 0080 mem: 3.36 + 04-03 22:41:28 | [179][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0316 ntime: 0080 mem: 3.36 + 04-03 22:41:31 | [179][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0056 ntime: 0083 mem: 3.36 + 04-03 22:41:33 | [179][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0295 ntime: 0081 mem: 3.36 + 04-03 22:41:35 | [179][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 22:41:39 | [179][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 1168 ntime: 0080 mem: 3.36 + 04-03 22:41:41 | [179][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0293 ntime: 0083 mem: 3.36 + 04-03 22:41:43 | [179][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 22:41:46 | [179][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0085 ntime: 0078 mem: 3.36 + 04-03 22:41:48 | [179][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0204 ntime: 0080 mem: 3.36 + 04-03 22:41:50 | [179][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0049 ntime: 0087 mem: 3.36 + 04-03 22:41:51 | Time info >>>> elapsed: 78.48 mins remain: 357.54 mins + 04-03 22:41:51 | [180][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 22:41:54 | [180][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0151 ntime: 0086 mem: 3.36 + 04-03 22:41:56 | [180][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 22:41:59 | [180][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0362 ntime: 0081 mem: 3.36 + 04-03 22:42:01 | [180][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0116 ntime: 0075 mem: 3.36 + 04-03 22:42:03 | [180][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0056 ntime: 0084 mem: 3.36 + 04-03 22:42:05 | [180][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0076 ntime: 0087 mem: 3.36 + 04-03 22:42:08 | [180][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 22:42:10 | [180][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0509 ntime: 0081 mem: 3.36 + 04-03 22:42:13 | [180][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0058 ntime: 0084 mem: 3.36 + 04-03 22:42:15 | [180][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0056 ntime: 0082 mem: 3.36 + 04-03 22:42:17 | [180][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0059 ntime: 0081 mem: 3.36 + 04-03 22:42:19 | [180][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0180 ntime: 0087 mem: 3.36 + 04-03 22:42:21 | [180][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0136 ntime: 0074 mem: 3.36 + 04-03 22:42:24 | [180][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0279 ntime: 0087 mem: 3.36 + 04-03 22:42:26 | [180][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0120 ntime: 0079 mem: 3.36 + 04-03 22:42:28 | [180][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 22:42:30 | [180][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0088 ntime: 0073 mem: 3.36 + 04-03 22:42:32 | Time info >>>> elapsed: 79.17 mins remain: 358.21 mins + 04-03 22:42:33 | [181][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0189 ntime: 0082 mem: 3.36 + 04-03 22:42:35 | [181][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0364 ntime: 0085 mem: 3.36 + 04-03 22:42:37 | [181][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 22:42:39 | [181][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0059 ntime: 0084 mem: 3.36 + 04-03 22:42:41 | [181][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:42:43 | [181][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:42:46 | [181][060/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:42:48 | [181][070/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0179 ntime: 0083 mem: 3.36 + 04-03 22:42:51 | [181][080/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0716 ntime: 0073 mem: 3.36 + 04-03 22:42:53 | [181][090/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0054 ntime: 0075 mem: 3.36 + 04-03 22:42:55 | [181][100/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0064 ntime: 0080 mem: 3.36 + 04-03 22:42:58 | [181][110/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0056 ntime: 0079 mem: 3.36 + 04-03 22:43:01 | [181][120/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0046 ntime: 0057 mem: 3.36 + 04-03 22:43:03 | [181][130/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0068 ntime: 0074 mem: 3.36 + 04-03 22:43:08 | [181][140/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0093 ntime: 0078 mem: 3.36 + 04-03 22:43:10 | [181][150/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0186 ntime: 0077 mem: 3.36 + 04-03 22:43:15 | [181][160/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0058 ntime: 0079 mem: 3.36 + 04-03 22:43:18 | [181][170/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0056 ntime: 0080 mem: 3.36 + 04-03 22:43:20 | Time info >>>> elapsed: 79.96 mins remain: 359.38 mins + 04-03 22:43:20 | [182][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0046 ntime: 0083 mem: 3.36 + 04-03 22:43:23 | [182][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0337 ntime: 0081 mem: 3.36 + 04-03 22:43:25 | [182][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0400 ntime: 0077 mem: 3.36 + 04-03 22:43:28 | [182][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:43:30 | [182][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0046 ntime: 0084 mem: 3.36 + 04-03 22:43:33 | [182][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0050 ntime: 0085 mem: 3.36 + 04-03 22:43:36 | [182][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0361 ntime: 0080 mem: 3.36 + 04-03 22:43:38 | [182][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0163 ntime: 0077 mem: 3.36 + 04-03 22:43:40 | [182][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0313 ntime: 0076 mem: 3.36 + 04-03 22:43:42 | [182][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0234 ntime: 0087 mem: 3.36 + 04-03 22:43:44 | [182][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0131 ntime: 0077 mem: 3.36 + 04-03 22:43:46 | [182][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 22:43:49 | [182][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 22:43:51 | [182][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0166 ntime: 0075 mem: 3.36 + 04-03 22:43:53 | [182][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0049 ntime: 0073 mem: 3.36 + 04-03 22:43:56 | [182][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0062 ntime: 0079 mem: 3.36 + 04-03 22:43:58 | [182][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 22:44:01 | [182][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0162 ntime: 0077 mem: 3.36 + 04-03 22:44:02 | Time info >>>> elapsed: 80.66 mins remain: 360.10 mins + 04-03 22:44:02 | [183][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0288 ntime: 0074 mem: 3.36 + 04-03 22:44:04 | [183][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 22:44:07 | [183][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0225 ntime: 0079 mem: 3.36 + 04-03 22:44:09 | [183][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0065 ntime: 0078 mem: 3.36 + 04-03 22:44:11 | [183][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0062 ntime: 0083 mem: 3.36 + 04-03 22:44:15 | [183][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0065 ntime: 0079 mem: 3.36 + 04-03 22:44:19 | [183][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0065 ntime: 0084 mem: 3.36 + 04-03 22:44:21 | [183][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0056 ntime: 0080 mem: 3.36 + 04-03 22:44:23 | [183][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0074 ntime: 0085 mem: 3.36 + 04-03 22:44:25 | [183][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0073 ntime: 0079 mem: 3.36 + 04-03 22:44:27 | [183][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0167 ntime: 0081 mem: 3.36 + 04-03 22:44:30 | [183][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0056 ntime: 0080 mem: 3.36 + 04-03 22:44:32 | [183][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:44:34 | [183][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0207 ntime: 0081 mem: 3.36 + 04-03 22:44:36 | [183][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0231 ntime: 0088 mem: 3.36 + 04-03 22:44:39 | [183][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:44:42 | [183][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0172 ntime: 0077 mem: 3.36 + 04-03 22:44:44 | [183][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0428 ntime: 0075 mem: 3.36 + 04-03 22:44:46 | Time info >>>> elapsed: 81.39 mins remain: 360.96 mins + 04-03 22:44:46 | [184][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 22:44:48 | [184][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0126 ntime: 0083 mem: 3.36 + 04-03 22:44:51 | [184][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0058 ntime: 0087 mem: 3.36 + 04-03 22:44:53 | [184][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:44:55 | [184][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0236 ntime: 0078 mem: 3.36 + 04-03 22:44:58 | [184][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 22:45:00 | [184][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0236 ntime: 0079 mem: 3.36 + 04-03 22:45:02 | [184][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0069 ntime: 0076 mem: 3.36 + 04-03 22:45:05 | [184][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0049 ntime: 0075 mem: 3.36 + 04-03 22:45:07 | [184][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 22:45:09 | [184][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0258 ntime: 0085 mem: 3.36 + 04-03 22:45:11 | [184][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0219 ntime: 0081 mem: 3.36 + 04-03 22:45:14 | [184][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0348 ntime: 0090 mem: 3.36 + 04-03 22:45:17 | [184][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0160 ntime: 0078 mem: 3.36 + 04-03 22:45:20 | [184][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0260 ntime: 0078 mem: 3.36 + 04-03 22:45:21 | [184][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 22:45:25 | [184][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0248 ntime: 0080 mem: 3.36 + 04-03 22:45:27 | [184][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0061 ntime: 0076 mem: 3.36 + 04-03 22:45:28 | Time info >>>> elapsed: 82.10 mins remain: 361.69 mins + 04-03 22:45:29 | [185][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0149 ntime: 0091 mem: 3.36 + 04-03 22:45:31 | [185][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0222 ntime: 0081 mem: 3.36 + 04-03 22:45:33 | [185][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0378 ntime: 0082 mem: 3.36 + 04-03 22:45:36 | [185][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0061 ntime: 0085 mem: 3.36 + 04-03 22:45:38 | [185][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0086 ntime: 0079 mem: 3.36 + 04-03 22:45:40 | [185][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 22:45:42 | [185][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0082 ntime: 0078 mem: 3.36 + 04-03 22:45:44 | [185][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0078 ntime: 0078 mem: 3.36 + 04-03 22:45:47 | [185][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0084 ntime: 0081 mem: 3.36 + 04-03 22:45:50 | [185][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0053 ntime: 0085 mem: 3.36 + 04-03 22:45:52 | [185][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0203 ntime: 0078 mem: 3.36 + 04-03 22:45:54 | [185][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0067 ntime: 0076 mem: 3.36 + 04-03 22:46:03 | [185][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 1191 ntime: 0082 mem: 3.36 + 04-03 22:46:09 | [185][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0076 ntime: 0079 mem: 3.36 + 04-03 22:46:11 | [185][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0046 ntime: 0080 mem: 3.36 + 04-03 22:46:14 | [185][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:46:17 | [185][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 22:46:19 | [185][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0056 ntime: 0083 mem: 3.36 + 04-03 22:46:21 | Time info >>>> elapsed: 82.98 mins remain: 363.16 mins + 04-03 22:46:21 | [186][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0117 ntime: 0075 mem: 3.36 + 04-03 22:46:25 | [186][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0069 ntime: 0078 mem: 3.36 + 04-03 22:46:27 | [186][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0083 ntime: 0077 mem: 3.36 + 04-03 22:46:29 | [186][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0136 ntime: 0078 mem: 3.36 + 04-03 22:46:31 | [186][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 22:46:34 | [186][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0053 ntime: 0085 mem: 3.36 + 04-03 22:46:37 | [186][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0152 ntime: 0081 mem: 3.36 + 04-03 22:46:41 | [186][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 1286 ntime: 0082 mem: 3.36 + 04-03 22:46:46 | [186][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0045 ntime: 0086 mem: 3.36 + 04-03 22:46:48 | [186][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0327 ntime: 0080 mem: 3.36 + 04-03 22:46:51 | [186][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 22:46:54 | [186][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0181 ntime: 0078 mem: 3.36 + 04-03 22:46:56 | [186][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0063 ntime: 0078 mem: 3.36 + 04-03 22:46:58 | [186][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0057 ntime: 0084 mem: 3.36 + 04-03 22:47:01 | [186][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0152 ntime: 0083 mem: 3.36 + 04-03 22:47:03 | [186][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 22:47:06 | [186][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0188 ntime: 0079 mem: 3.36 + 04-03 22:47:08 | [186][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0268 ntime: 0080 mem: 3.36 + 04-03 22:47:10 | Time info >>>> elapsed: 83.79 mins remain: 364.28 mins + 04-03 22:47:10 | [187][000/179] predict_x0_loss: 0.011 glr: 5.0e-05 dtime: 0305 ntime: 0078 mem: 3.36 + 04-03 22:47:14 | [187][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0166 ntime: 0078 mem: 3.36 + 04-03 22:47:16 | [187][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 22:47:20 | [187][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 22:47:22 | [187][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0150 ntime: 0079 mem: 3.36 + 04-03 22:47:25 | [187][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 22:47:27 | [187][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0156 ntime: 0073 mem: 3.36 + 04-03 22:47:30 | [187][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0210 ntime: 0082 mem: 3.36 + 04-03 22:47:32 | [187][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0211 ntime: 0061 mem: 3.36 + 04-03 22:47:34 | [187][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0314 ntime: 0080 mem: 3.36 + 04-03 22:47:36 | [187][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0064 ntime: 0077 mem: 3.36 + 04-03 22:47:39 | [187][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0057 ntime: 0083 mem: 3.36 + 04-03 22:47:41 | [187][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0057 ntime: 0077 mem: 3.36 + 04-03 22:47:44 | [187][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0081 ntime: 0076 mem: 3.36 + 04-03 22:47:46 | [187][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0249 ntime: 0086 mem: 3.36 + 04-03 22:47:49 | [187][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0394 ntime: 0079 mem: 3.36 + 04-03 22:47:50 | [187][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0074 ntime: 0080 mem: 3.36 + 04-03 22:47:53 | [187][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 22:47:54 | Time info >>>> elapsed: 84.54 mins remain: 365.12 mins + 04-03 22:47:55 | [188][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0273 ntime: 0078 mem: 3.36 + 04-03 22:47:57 | [188][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 22:47:59 | [188][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0062 ntime: 0085 mem: 3.36 + 04-03 22:48:02 | [188][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 22:48:04 | [188][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 22:48:06 | [188][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0185 ntime: 0076 mem: 3.36 + 04-03 22:48:08 | [188][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0148 ntime: 0082 mem: 3.36 + 04-03 22:48:10 | [188][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 22:48:13 | [188][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0057 ntime: 0081 mem: 3.36 + 04-03 22:48:15 | [188][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 22:48:17 | [188][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0280 ntime: 0084 mem: 3.36 + 04-03 22:48:19 | [188][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0069 ntime: 0084 mem: 3.36 + 04-03 22:48:21 | [188][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0148 ntime: 0079 mem: 3.36 + 04-03 22:48:24 | [188][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0272 ntime: 0087 mem: 3.36 + 04-03 22:48:26 | [188][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0137 ntime: 0078 mem: 3.36 + 04-03 22:48:28 | [188][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0151 ntime: 0082 mem: 3.36 + 04-03 22:48:30 | [188][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0045 ntime: 0079 mem: 3.36 + 04-03 22:48:33 | [188][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0161 ntime: 0079 mem: 3.36 + 04-03 22:48:35 | Time info >>>> elapsed: 85.20 mins remain: 365.61 mins + 04-03 22:48:35 | [189][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 22:48:37 | [189][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0366 ntime: 0085 mem: 3.36 + 04-03 22:48:39 | [189][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0307 ntime: 0082 mem: 3.36 + 04-03 22:48:42 | [189][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0309 ntime: 0085 mem: 3.36 + 04-03 22:48:44 | [189][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0196 ntime: 0087 mem: 3.36 + 04-03 22:48:46 | [189][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0044 ntime: 0059 mem: 3.36 + 04-03 22:48:48 | [189][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0257 ntime: 0083 mem: 3.36 + 04-03 22:48:51 | [189][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0543 ntime: 0088 mem: 3.36 + 04-03 22:48:53 | [189][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0186 ntime: 0080 mem: 3.36 + 04-03 22:48:55 | [189][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0356 ntime: 0077 mem: 3.36 + 04-03 22:48:59 | [189][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0694 ntime: 0077 mem: 3.36 + 04-03 22:49:02 | [189][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0177 ntime: 0078 mem: 3.36 + 04-03 22:49:05 | [189][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0059 ntime: 0084 mem: 3.36 + 04-03 22:49:07 | [189][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 22:49:10 | [189][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0335 ntime: 0081 mem: 3.36 + 04-03 22:49:12 | [189][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0069 ntime: 0081 mem: 3.36 + 04-03 22:49:14 | [189][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0290 ntime: 0079 mem: 3.36 + 04-03 22:49:16 | [189][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0233 ntime: 0086 mem: 3.36 + 04-03 22:49:18 | Time info >>>> elapsed: 85.92 mins remain: 366.31 mins + 04-03 22:49:18 | [190][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0331 ntime: 0084 mem: 3.36 + 04-03 22:49:21 | [190][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:49:23 | [190][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0294 ntime: 0081 mem: 3.36 + 04-03 22:49:26 | [190][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0197 ntime: 0084 mem: 3.36 + 04-03 22:49:29 | [190][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0281 ntime: 0081 mem: 3.36 + 04-03 22:49:30 | [190][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0174 ntime: 0082 mem: 3.36 + 04-03 22:49:32 | [190][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0066 ntime: 0084 mem: 3.36 + 04-03 22:49:35 | [190][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0072 ntime: 0081 mem: 3.36 + 04-03 22:49:37 | [190][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0193 ntime: 0082 mem: 3.36 + 04-03 22:49:40 | [190][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0225 ntime: 0076 mem: 3.36 + 04-03 22:49:41 | [190][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0079 ntime: 0075 mem: 3.36 + 04-03 22:49:44 | [190][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0135 ntime: 0083 mem: 3.36 + 04-03 22:49:46 | [190][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0188 ntime: 0082 mem: 3.36 + 04-03 22:49:48 | [190][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0060 ntime: 0078 mem: 3.36 + 04-03 22:49:51 | [190][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0099 ntime: 0085 mem: 3.36 + 04-03 22:49:53 | [190][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0054 ntime: 0074 mem: 3.36 + 04-03 22:49:55 | [190][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0079 ntime: 0078 mem: 3.36 + 04-03 22:49:58 | [190][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 22:49:59 | Time info >>>> elapsed: 86.62 mins remain: 366.87 mins + 04-03 22:49:59 | [191][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0056 ntime: 0080 mem: 3.36 + 04-03 22:50:02 | [191][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0067 ntime: 0076 mem: 3.36 + 04-03 22:50:04 | [191][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0091 ntime: 0079 mem: 3.36 + 04-03 22:50:06 | [191][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0165 ntime: 0079 mem: 3.36 + 04-03 22:50:09 | [191][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0050 ntime: 0075 mem: 3.36 + 04-03 22:50:11 | [191][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:50:13 | [191][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0218 ntime: 0082 mem: 3.36 + 04-03 22:50:15 | [191][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0290 ntime: 0084 mem: 3.36 + 04-03 22:50:17 | [191][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 22:50:19 | [191][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0061 ntime: 0085 mem: 3.36 + 04-03 22:50:22 | [191][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0154 ntime: 0083 mem: 3.36 + 04-03 22:50:25 | [191][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 22:50:28 | [191][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0169 ntime: 0076 mem: 3.36 + 04-03 22:50:30 | [191][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0147 ntime: 0078 mem: 3.36 + 04-03 22:50:32 | [191][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0292 ntime: 0078 mem: 3.36 + 04-03 22:50:34 | [191][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0186 ntime: 0081 mem: 3.36 + 04-03 22:50:36 | [191][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0220 ntime: 0082 mem: 3.36 + 04-03 22:50:38 | [191][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0178 ntime: 0090 mem: 3.36 + 04-03 22:50:40 | Time info >>>> elapsed: 87.29 mins remain: 367.36 mins + 04-03 22:50:40 | [192][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:50:43 | [192][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0069 ntime: 0077 mem: 3.36 + 04-03 22:50:45 | [192][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0465 ntime: 0082 mem: 3.36 + 04-03 22:50:47 | [192][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0076 ntime: 0078 mem: 3.36 + 04-03 22:50:50 | [192][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 22:50:52 | [192][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0252 ntime: 0089 mem: 3.36 + 04-03 22:50:55 | [192][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0158 ntime: 0077 mem: 3.36 + 04-03 22:50:57 | [192][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0304 ntime: 0087 mem: 3.36 + 04-03 22:50:59 | [192][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0259 ntime: 0076 mem: 3.36 + 04-03 22:51:02 | [192][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0067 ntime: 0074 mem: 3.36 + 04-03 22:51:05 | [192][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0260 ntime: 0078 mem: 3.36 + 04-03 22:51:07 | [192][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0074 ntime: 0082 mem: 3.36 + 04-03 22:51:09 | [192][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0055 ntime: 0080 mem: 3.36 + 04-03 22:51:11 | [192][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0337 ntime: 0088 mem: 3.36 + 04-03 22:51:14 | [192][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0164 ntime: 0081 mem: 3.36 + 04-03 22:51:16 | [192][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0232 ntime: 0077 mem: 3.36 + 04-03 22:51:18 | [192][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0104 ntime: 0084 mem: 3.36 + 04-03 22:51:20 | [192][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0057 ntime: 0080 mem: 3.36 + 04-03 22:51:22 | Time info >>>> elapsed: 87.99 mins remain: 367.90 mins + 04-03 22:51:22 | [193][000/179] predict_x0_loss: 0.009 glr: 5.0e-05 dtime: 0153 ntime: 0077 mem: 3.36 + 04-03 22:51:24 | [193][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0164 ntime: 0078 mem: 3.36 + 04-03 22:51:27 | [193][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0247 ntime: 0089 mem: 3.36 + 04-03 22:51:28 | [193][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 22:51:30 | [193][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0071 ntime: 0079 mem: 3.36 + 04-03 22:51:32 | [193][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0095 ntime: 0078 mem: 3.36 + 04-03 22:51:34 | [193][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0048 ntime: 0087 mem: 3.36 + 04-03 22:51:36 | [193][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0190 ntime: 0057 mem: 3.36 + 04-03 22:51:38 | [193][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0045 ntime: 0071 mem: 3.36 + 04-03 22:51:40 | [193][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 22:51:42 | [193][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0065 ntime: 0078 mem: 3.36 + 04-03 22:51:44 | [193][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0116 ntime: 0082 mem: 3.36 + 04-03 22:51:46 | [193][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0133 ntime: 0075 mem: 3.36 + 04-03 22:51:48 | [193][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0163 ntime: 0076 mem: 3.36 + 04-03 22:51:50 | [193][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0266 ntime: 0081 mem: 3.36 + 04-03 22:51:52 | [193][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 22:51:54 | [193][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0062 ntime: 0076 mem: 3.36 + 04-03 22:51:56 | [193][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0145 ntime: 0080 mem: 3.36 + 04-03 22:51:57 | Time info >>>> elapsed: 88.58 mins remain: 368.04 mins + 04-03 22:51:58 | [194][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 22:51:59 | [194][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0051 ntime: 0073 mem: 3.36 + 04-03 22:52:01 | [194][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 22:52:03 | [194][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 22:52:05 | [194][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0193 ntime: 0079 mem: 3.36 + 04-03 22:52:07 | [194][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0092 ntime: 0080 mem: 3.36 + 04-03 22:52:09 | [194][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0095 ntime: 0078 mem: 3.36 + 04-03 22:52:11 | [194][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0271 ntime: 0086 mem: 3.36 + 04-03 22:52:13 | [194][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0046 ntime: 0073 mem: 3.36 + 04-03 22:52:15 | [194][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0296 ntime: 0081 mem: 3.36 + 04-03 22:52:18 | [194][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:52:20 | [194][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0180 ntime: 0085 mem: 3.36 + 04-03 22:52:22 | [194][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0168 ntime: 0090 mem: 3.36 + 04-03 22:52:24 | [194][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0059 ntime: 0069 mem: 3.36 + 04-03 22:52:26 | [194][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0142 ntime: 0085 mem: 3.36 + 04-03 22:52:29 | [194][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0195 ntime: 0079 mem: 3.36 + 04-03 22:52:32 | [194][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0205 ntime: 0088 mem: 3.36 + 04-03 22:52:34 | [194][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0056 ntime: 0078 mem: 3.36 + 04-03 22:52:35 | Time info >>>> elapsed: 89.22 mins remain: 368.30 mins + 04-03 22:52:35 | [195][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0076 ntime: 0079 mem: 3.36 + 04-03 22:52:38 | [195][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0244 ntime: 0076 mem: 3.36 + 04-03 22:52:41 | [195][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 22:52:43 | [195][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0055 ntime: 0086 mem: 3.36 + 04-03 22:52:45 | [195][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0312 ntime: 0079 mem: 3.36 + 04-03 22:52:47 | [195][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0056 ntime: 0083 mem: 3.36 + 04-03 22:52:51 | [195][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0072 ntime: 0086 mem: 3.36 + 04-03 22:52:54 | [195][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0048 ntime: 0086 mem: 3.36 + 04-03 22:52:56 | [195][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0104 ntime: 0079 mem: 3.36 + 04-03 22:52:58 | [195][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:53:01 | [195][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0163 ntime: 0088 mem: 3.36 + 04-03 22:53:03 | [195][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0276 ntime: 0080 mem: 3.36 + 04-03 22:53:05 | [195][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0091 ntime: 0089 mem: 3.36 + 04-03 22:53:07 | [195][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0075 ntime: 0075 mem: 3.36 + 04-03 22:53:09 | [195][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0104 ntime: 0082 mem: 3.36 + 04-03 22:53:11 | [195][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 22:53:13 | [195][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 22:53:16 | [195][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0217 ntime: 0082 mem: 3.36 + 04-03 22:53:17 | Time info >>>> elapsed: 89.92 mins remain: 368.85 mins + 04-03 22:53:18 | [196][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0207 ntime: 0075 mem: 3.36 + 04-03 22:53:20 | [196][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 22:53:22 | [196][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 22:53:24 | [196][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0219 ntime: 0083 mem: 3.36 + 04-03 22:53:26 | [196][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 22:53:27 | [196][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0069 ntime: 0081 mem: 3.36 + 04-03 22:53:29 | [196][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0143 ntime: 0086 mem: 3.36 + 04-03 22:53:31 | [196][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0080 ntime: 0090 mem: 3.36 + 04-03 22:53:33 | [196][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 22:53:35 | [196][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0083 ntime: 0075 mem: 3.36 + 04-03 22:53:37 | [196][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0278 ntime: 0076 mem: 3.36 + 04-03 22:53:39 | [196][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0255 ntime: 0079 mem: 3.36 + 04-03 22:53:42 | [196][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0198 ntime: 0080 mem: 3.36 + 04-03 22:53:44 | [196][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 22:53:46 | [196][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 22:53:49 | [196][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0171 ntime: 0083 mem: 3.36 + 04-03 22:53:51 | [196][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0322 ntime: 0078 mem: 3.36 + 04-03 22:53:53 | [196][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0056 ntime: 0081 mem: 3.36 + 04-03 22:53:55 | Time info >>>> elapsed: 90.55 mins remain: 369.09 mins + 04-03 22:53:56 | [197][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0323 ntime: 0082 mem: 3.36 + 04-03 22:53:58 | [197][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0209 ntime: 0074 mem: 3.36 + 04-03 22:54:00 | [197][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0229 ntime: 0081 mem: 3.36 + 04-03 22:54:02 | [197][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 22:54:05 | [197][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0211 ntime: 0078 mem: 3.36 + 04-03 22:54:06 | [197][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0062 ntime: 0076 mem: 3.36 + 04-03 22:54:09 | [197][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0376 ntime: 0089 mem: 3.36 + 04-03 22:54:12 | [197][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0077 ntime: 0077 mem: 3.36 + 04-03 22:54:15 | [197][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0271 ntime: 0083 mem: 3.36 + 04-03 22:54:17 | [197][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0166 ntime: 0083 mem: 3.36 + 04-03 22:54:19 | [197][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0046 ntime: 0074 mem: 3.36 + 04-03 22:54:21 | [197][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0161 ntime: 0076 mem: 3.36 + 04-03 22:54:23 | [197][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0047 ntime: 0085 mem: 3.36 + 04-03 22:54:25 | [197][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 22:54:27 | [197][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0280 ntime: 0083 mem: 3.36 + 04-03 22:54:30 | [197][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0905 ntime: 0085 mem: 3.36 + 04-03 22:54:32 | [197][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0052 ntime: 0091 mem: 3.36 + 04-03 22:54:34 | [197][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0105 ntime: 0084 mem: 3.36 + 04-03 22:54:37 | Time info >>>> elapsed: 91.24 mins remain: 369.55 mins + 04-03 22:54:37 | [198][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0178 ntime: 0078 mem: 3.36 + 04-03 22:54:39 | [198][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0071 ntime: 0079 mem: 3.36 + 04-03 22:54:41 | [198][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 22:54:44 | [198][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0059 ntime: 0080 mem: 3.36 + 04-03 22:54:47 | [198][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0264 ntime: 0075 mem: 3.36 + 04-03 22:54:49 | [198][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0444 ntime: 0077 mem: 3.36 + 04-03 22:54:51 | [198][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0174 ntime: 0077 mem: 3.36 + 04-03 22:54:53 | [198][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 22:54:55 | [198][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0076 ntime: 0072 mem: 3.36 + 04-03 22:54:58 | [198][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0316 ntime: 0078 mem: 3.36 + 04-03 22:55:00 | [198][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0158 ntime: 0086 mem: 3.36 + 04-03 22:55:02 | [198][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0176 ntime: 0079 mem: 3.36 + 04-03 22:55:04 | [198][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0059 ntime: 0079 mem: 3.36 + 04-03 22:55:06 | [198][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0389 ntime: 0074 mem: 3.36 + 04-03 22:55:09 | [198][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0065 ntime: 0081 mem: 3.36 + 04-03 22:55:11 | [198][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 22:55:13 | [198][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0190 ntime: 0078 mem: 3.36 + 04-03 22:55:16 | [198][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0126 ntime: 0081 mem: 3.36 + 04-03 22:55:17 | Time info >>>> elapsed: 91.92 mins remain: 369.98 mins + 04-03 22:55:18 | [199][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0343 ntime: 0079 mem: 3.36 + 04-03 22:55:20 | [199][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0063 ntime: 0081 mem: 3.36 + 04-03 22:55:23 | [199][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 22:55:25 | [199][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0051 ntime: 0072 mem: 3.36 + 04-03 22:55:27 | [199][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0109 ntime: 0078 mem: 3.36 + 04-03 22:55:30 | [199][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 22:55:32 | [199][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0230 ntime: 0077 mem: 3.36 + 04-03 22:55:35 | [199][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0056 ntime: 0085 mem: 3.36 + 04-03 22:55:37 | [199][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 22:55:39 | [199][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0267 ntime: 0081 mem: 3.36 + 04-03 22:55:41 | [199][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0059 ntime: 0079 mem: 3.36 + 04-03 22:55:43 | [199][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 22:55:46 | [199][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0258 ntime: 0076 mem: 3.36 + 04-03 22:55:47 | [199][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0045 ntime: 0080 mem: 3.36 + 04-03 22:55:49 | [199][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0061 ntime: 0078 mem: 3.36 + 04-03 22:55:51 | [199][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0113 ntime: 0077 mem: 3.36 + 04-03 22:55:54 | [199][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0062 ntime: 0077 mem: 3.36 + 04-03 22:55:56 | [199][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0045 ntime: 0081 mem: 3.36 + 04-03 22:55:57 | Time info >>>> elapsed: 92.58 mins remain: 370.34 mins + 04-03 22:55:58 | [200][000/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 22:56:00 | [200][010/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0134 ntime: 0077 mem: 3.36 + 04-03 22:56:02 | [200][020/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0311 ntime: 0083 mem: 3.36 + 04-03 22:56:03 | [200][030/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0050 ntime: 0085 mem: 3.36 + 04-03 22:56:06 | [200][040/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0245 ntime: 0080 mem: 3.36 + 04-03 22:56:08 | [200][050/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0402 ntime: 0081 mem: 3.36 + 04-03 22:56:10 | [200][060/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0093 ntime: 0078 mem: 3.36 + 04-03 22:56:12 | [200][070/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0307 ntime: 0083 mem: 3.36 + 04-03 22:56:15 | [200][080/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 22:56:17 | [200][090/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 22:56:19 | [200][100/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0228 ntime: 0074 mem: 3.36 + 04-03 22:56:21 | [200][110/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 22:56:24 | [200][120/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0182 ntime: 0078 mem: 3.36 + 04-03 22:56:26 | [200][130/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0054 ntime: 0089 mem: 3.36 + 04-03 22:56:28 | [200][140/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0053 ntime: 0073 mem: 3.36 + 04-03 22:56:31 | [200][150/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0058 ntime: 0079 mem: 3.36 + 04-03 22:56:33 | [200][160/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0367 ntime: 0077 mem: 3.36 + 04-03 22:56:36 | [200][170/179] predict_x0_loss: 0.010 glr: 5.0e-05 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 22:56:38 | Time info >>>> elapsed: 93.26 mins remain: 370.74 mins + 04-03 22:56:39 | [201][000/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0413 ntime: 0085 mem: 3.36 + 04-03 22:56:42 | [201][010/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0051 ntime: 0086 mem: 3.36 + 04-03 22:56:44 | [201][020/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0063 ntime: 0081 mem: 3.36 + 04-03 22:56:47 | [201][030/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0183 ntime: 0077 mem: 3.36 + 04-03 22:56:49 | [201][040/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0150 ntime: 0083 mem: 3.36 + 04-03 22:56:50 | [201][050/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0060 ntime: 0076 mem: 3.36 + 04-03 22:56:53 | [201][060/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0235 ntime: 0078 mem: 3.36 + 04-03 22:56:55 | [201][070/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0438 ntime: 0081 mem: 3.36 + 04-03 22:56:57 | [201][080/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 22:57:00 | [201][090/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0062 ntime: 0079 mem: 3.36 + 04-03 22:57:02 | [201][100/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0294 ntime: 0071 mem: 3.36 + 04-03 22:57:05 | [201][110/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0058 ntime: 0082 mem: 3.36 + 04-03 22:57:07 | [201][120/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0445 ntime: 0081 mem: 3.36 + 04-03 22:57:09 | [201][130/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 22:57:13 | [201][140/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0594 ntime: 0079 mem: 3.36 + 04-03 22:57:15 | [201][150/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 22:57:17 | [201][160/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0059 ntime: 0075 mem: 3.36 + 04-03 22:57:19 | [201][170/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0098 ntime: 0075 mem: 3.36 + 04-03 22:57:21 | Time info >>>> elapsed: 93.98 mins remain: 371.28 mins + 04-03 22:57:22 | [202][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 22:57:24 | [202][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 22:57:26 | [202][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 22:57:28 | [202][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 22:57:31 | [202][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0207 ntime: 0078 mem: 3.36 + 04-03 22:57:32 | [202][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 22:57:35 | [202][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 22:57:37 | [202][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 22:57:39 | [202][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0297 ntime: 0087 mem: 3.36 + 04-03 22:57:41 | [202][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0135 ntime: 0074 mem: 3.36 + 04-03 22:57:43 | [202][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0150 ntime: 0079 mem: 3.36 + 04-03 22:57:45 | [202][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:57:47 | [202][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0078 mem: 3.36 + 04-03 22:57:49 | [202][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0076 mem: 3.36 + 04-03 22:57:52 | [202][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0345 ntime: 0084 mem: 3.36 + 04-03 22:57:55 | [202][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0082 mem: 3.36 + 04-03 22:57:57 | [202][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0212 ntime: 0083 mem: 3.36 + 04-03 22:58:00 | [202][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0421 ntime: 0078 mem: 3.36 + 04-03 22:58:02 | Time info >>>> elapsed: 94.66 mins remain: 371.63 mins + 04-03 22:58:02 | [203][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 22:58:04 | [203][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0084 mem: 3.36 + 04-03 22:58:06 | [203][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0077 mem: 3.36 + 04-03 22:58:08 | [203][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0303 ntime: 0081 mem: 3.36 + 04-03 22:58:10 | [203][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0090 mem: 3.36 + 04-03 22:58:13 | [203][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0078 mem: 3.36 + 04-03 22:58:16 | [203][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0414 ntime: 0077 mem: 3.36 + 04-03 22:58:18 | [203][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0089 ntime: 0080 mem: 3.36 + 04-03 22:58:20 | [203][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0084 mem: 3.36 + 04-03 22:58:22 | [203][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0103 ntime: 0076 mem: 3.36 + 04-03 22:58:24 | [203][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0336 ntime: 0081 mem: 3.36 + 04-03 22:58:27 | [203][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0192 ntime: 0082 mem: 3.36 + 04-03 22:58:30 | [203][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:58:32 | [203][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 22:58:34 | [203][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 22:58:36 | [203][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 22:58:38 | [203][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0098 ntime: 0075 mem: 3.36 + 04-03 22:58:41 | [203][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0152 ntime: 0081 mem: 3.36 + 04-03 22:58:43 | Time info >>>> elapsed: 95.34 mins remain: 372.01 mins + 04-03 22:58:43 | [204][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0146 ntime: 0070 mem: 3.36 + 04-03 22:58:46 | [204][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0091 ntime: 0084 mem: 3.36 + 04-03 22:58:48 | [204][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0477 ntime: 0081 mem: 3.36 + 04-03 22:58:51 | [204][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0393 ntime: 0081 mem: 3.36 + 04-03 22:58:53 | [204][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 22:58:56 | [204][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0196 ntime: 0086 mem: 3.36 + 04-03 22:58:59 | [204][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0845 ntime: 0083 mem: 3.36 + 04-03 22:59:01 | [204][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0184 ntime: 0083 mem: 3.36 + 04-03 22:59:03 | [204][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0074 mem: 3.36 + 04-03 22:59:05 | [204][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0075 mem: 3.36 + 04-03 22:59:08 | [204][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0408 ntime: 0078 mem: 3.36 + 04-03 22:59:10 | [204][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0080 mem: 3.36 + 04-03 22:59:12 | [204][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0079 mem: 3.36 + 04-03 22:59:15 | [204][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0190 ntime: 0084 mem: 3.36 + 04-03 22:59:17 | [204][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0376 ntime: 0078 mem: 3.36 + 04-03 22:59:20 | [204][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0077 mem: 3.36 + 04-03 22:59:22 | [204][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0201 ntime: 0084 mem: 3.36 + 04-03 22:59:25 | [204][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0430 ntime: 0076 mem: 3.36 + 04-03 22:59:26 | Time info >>>> elapsed: 96.07 mins remain: 372.56 mins + 04-03 22:59:27 | [205][000/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 22:59:29 | [205][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 22:59:32 | [205][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0268 ntime: 0084 mem: 3.36 + 04-03 22:59:34 | [205][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0430 ntime: 0079 mem: 3.36 + 04-03 22:59:37 | [205][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 22:59:39 | [205][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0082 ntime: 0079 mem: 3.36 + 04-03 22:59:41 | [205][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0081 mem: 3.36 + 04-03 22:59:44 | [205][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0176 ntime: 0076 mem: 3.36 + 04-03 22:59:46 | [205][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0268 ntime: 0080 mem: 3.36 + 04-03 22:59:48 | [205][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0081 mem: 3.36 + 04-03 22:59:50 | [205][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0085 mem: 3.36 + 04-03 22:59:53 | [205][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0321 ntime: 0085 mem: 3.36 + 04-03 22:59:55 | [205][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0145 ntime: 0083 mem: 3.36 + 04-03 22:59:58 | [205][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0270 ntime: 0085 mem: 3.36 + 04-03 22:59:59 | [205][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 23:00:02 | [205][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0103 ntime: 0081 mem: 3.36 + 04-03 23:00:04 | [205][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0397 ntime: 0081 mem: 3.36 + 04-03 23:00:08 | [205][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0081 mem: 3.36 + 04-03 23:00:09 | Time info >>>> elapsed: 96.78 mins remain: 373.04 mins + 04-03 23:00:10 | [206][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0306 ntime: 0079 mem: 3.36 + 04-03 23:00:12 | [206][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0236 ntime: 0088 mem: 3.36 + 04-03 23:00:14 | [206][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0179 ntime: 0079 mem: 3.36 + 04-03 23:00:17 | [206][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0079 mem: 3.36 + 04-03 23:00:19 | [206][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0394 ntime: 0079 mem: 3.36 + 04-03 23:00:22 | [206][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0090 mem: 3.36 + 04-03 23:00:24 | [206][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 23:00:27 | [206][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0078 mem: 3.36 + 04-03 23:00:31 | [206][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0910 ntime: 0081 mem: 3.36 + 04-03 23:00:33 | [206][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0467 ntime: 0082 mem: 3.36 + 04-03 23:00:35 | [206][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0073 mem: 3.36 + 04-03 23:00:37 | [206][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 23:00:40 | [206][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0172 ntime: 0074 mem: 3.36 + 04-03 23:00:42 | [206][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0086 mem: 3.36 + 04-03 23:00:46 | [206][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0939 ntime: 0079 mem: 3.36 + 04-03 23:00:48 | [206][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0278 ntime: 0080 mem: 3.36 + 04-03 23:00:50 | [206][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 23:00:53 | [206][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 23:00:55 | Time info >>>> elapsed: 97.54 mins remain: 373.68 mins + 04-03 23:00:55 | [207][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0078 ntime: 0079 mem: 3.36 + 04-03 23:00:57 | [207][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0268 ntime: 0076 mem: 3.36 + 04-03 23:00:59 | [207][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 23:01:02 | [207][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 23:01:04 | [207][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0077 mem: 3.36 + 04-03 23:01:07 | [207][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0320 ntime: 0085 mem: 3.36 + 04-03 23:01:09 | [207][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0354 ntime: 0087 mem: 3.36 + 04-03 23:01:11 | [207][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0082 mem: 3.36 + 04-03 23:01:13 | [207][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0083 mem: 3.36 + 04-03 23:01:15 | [207][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0077 ntime: 0081 mem: 3.36 + 04-03 23:01:17 | [207][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0296 ntime: 0079 mem: 3.36 + 04-03 23:01:20 | [207][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0240 ntime: 0078 mem: 3.36 + 04-03 23:01:22 | [207][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0105 ntime: 0083 mem: 3.36 + 04-03 23:01:24 | [207][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0070 mem: 3.36 + 04-03 23:01:26 | [207][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0290 ntime: 0080 mem: 3.36 + 04-03 23:01:28 | [207][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0078 mem: 3.36 + 04-03 23:01:30 | [207][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0080 mem: 3.36 + 04-03 23:01:33 | [207][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0273 ntime: 0087 mem: 3.36 + 04-03 23:01:34 | Time info >>>> elapsed: 98.20 mins remain: 373.92 mins + 04-03 23:01:35 | [208][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0093 ntime: 0083 mem: 3.36 + 04-03 23:01:37 | [208][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0077 mem: 3.36 + 04-03 23:01:39 | [208][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 23:01:42 | [208][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0265 ntime: 0081 mem: 3.36 + 04-03 23:01:46 | [208][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0465 ntime: 0077 mem: 3.36 + 04-03 23:01:48 | [208][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0081 mem: 3.36 + 04-03 23:01:51 | [208][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 23:01:53 | [208][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 23:01:56 | [208][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0266 ntime: 0081 mem: 3.36 + 04-03 23:01:59 | [208][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0203 ntime: 0081 mem: 3.36 + 04-03 23:02:01 | [208][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0083 ntime: 0085 mem: 3.36 + 04-03 23:02:03 | [208][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0085 mem: 3.36 + 04-03 23:02:06 | [208][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0482 ntime: 0080 mem: 3.36 + 04-03 23:02:08 | [208][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 23:02:10 | [208][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0077 mem: 3.36 + 04-03 23:02:14 | [208][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0079 mem: 3.36 + 04-03 23:02:16 | [208][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0519 ntime: 0083 mem: 3.36 + 04-03 23:02:19 | [208][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0270 ntime: 0078 mem: 3.36 + 04-03 23:02:21 | Time info >>>> elapsed: 98.98 mins remain: 374.59 mins + 04-03 23:02:21 | [209][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0077 mem: 3.36 + 04-03 23:02:23 | [209][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0085 mem: 3.36 + 04-03 23:02:26 | [209][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0078 mem: 3.36 + 04-03 23:02:28 | [209][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 23:02:30 | [209][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0084 mem: 3.36 + 04-03 23:02:32 | [209][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0070 ntime: 0074 mem: 3.36 + 04-03 23:02:35 | [209][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0074 mem: 3.36 + 04-03 23:02:37 | [209][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 23:02:40 | [209][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0045 ntime: 0085 mem: 3.36 + 04-03 23:02:42 | [209][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0097 ntime: 0085 mem: 3.36 + 04-03 23:02:44 | [209][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0075 mem: 3.36 + 04-03 23:02:46 | [209][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 23:02:49 | [209][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0194 ntime: 0071 mem: 3.36 + 04-03 23:02:50 | [209][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0076 mem: 3.36 + 04-03 23:02:53 | [209][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0086 mem: 3.36 + 04-03 23:02:55 | [209][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0350 ntime: 0086 mem: 3.36 + 04-03 23:02:57 | [209][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0085 mem: 3.36 + 04-03 23:02:59 | [209][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0270 ntime: 0079 mem: 3.36 + 04-03 23:03:01 | Time info >>>> elapsed: 99.64 mins remain: 374.83 mins + 04-03 23:03:01 | [210][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0198 ntime: 0078 mem: 3.36 + 04-03 23:03:03 | [210][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0040 ntime: 0058 mem: 3.36 + 04-03 23:03:05 | [210][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0085 mem: 3.36 + 04-03 23:03:07 | [210][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0076 mem: 3.36 + 04-03 23:03:10 | [210][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0072 mem: 3.36 + 04-03 23:03:12 | [210][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0248 ntime: 0077 mem: 3.36 + 04-03 23:03:14 | [210][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0264 ntime: 0074 mem: 3.36 + 04-03 23:03:17 | [210][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0128 ntime: 0075 mem: 3.36 + 04-03 23:03:20 | [210][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0067 ntime: 0079 mem: 3.36 + 04-03 23:03:22 | [210][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0088 mem: 3.36 + 04-03 23:03:24 | [210][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0078 mem: 3.36 + 04-03 23:03:26 | [210][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0085 mem: 3.36 + 04-03 23:03:29 | [210][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0091 mem: 3.36 + 04-03 23:03:31 | [210][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0100 ntime: 0082 mem: 3.36 + 04-03 23:03:33 | [210][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0082 mem: 3.36 + 04-03 23:03:35 | [210][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 23:03:37 | [210][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0180 ntime: 0087 mem: 3.36 + 04-03 23:03:40 | [210][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0342 ntime: 0079 mem: 3.36 + 04-03 23:03:42 | Time info >>>> elapsed: 100.33 mins remain: 375.15 mins + 04-03 23:03:42 | [211][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0081 mem: 3.36 + 04-03 23:03:44 | [211][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0084 mem: 3.36 + 04-03 23:03:47 | [211][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0086 mem: 3.36 + 04-03 23:03:49 | [211][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0385 ntime: 0075 mem: 3.36 + 04-03 23:03:52 | [211][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0071 mem: 3.36 + 04-03 23:03:54 | [211][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 23:03:56 | [211][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0088 mem: 3.36 + 04-03 23:03:59 | [211][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 23:04:01 | [211][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 23:04:04 | [211][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1131 ntime: 0080 mem: 3.36 + 04-03 23:04:07 | [211][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0686 ntime: 0071 mem: 3.36 + 04-03 23:04:09 | [211][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0120 ntime: 0085 mem: 3.36 + 04-03 23:04:11 | [211][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0150 ntime: 0086 mem: 3.36 + 04-03 23:04:13 | [211][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0272 ntime: 0087 mem: 3.36 + 04-03 23:04:15 | [211][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 23:04:18 | [211][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0077 ntime: 0081 mem: 3.36 + 04-03 23:04:20 | [211][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0075 mem: 3.36 + 04-03 23:04:23 | [211][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0082 mem: 3.36 + 04-03 23:04:25 | Time info >>>> elapsed: 101.05 mins remain: 375.59 mins + 04-03 23:04:25 | [212][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0077 mem: 3.36 + 04-03 23:04:28 | [212][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0230 ntime: 0079 mem: 3.36 + 04-03 23:04:31 | [212][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 23:04:33 | [212][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0085 mem: 3.36 + 04-03 23:04:36 | [212][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0082 mem: 3.36 + 04-03 23:04:38 | [212][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0207 ntime: 0076 mem: 3.36 + 04-03 23:04:40 | [212][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0077 mem: 3.36 + 04-03 23:04:42 | [212][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0082 mem: 3.36 + 04-03 23:04:45 | [212][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0081 mem: 3.36 + 04-03 23:04:48 | [212][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0561 ntime: 0080 mem: 3.36 + 04-03 23:04:51 | [212][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0180 ntime: 0085 mem: 3.36 + 04-03 23:04:54 | [212][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0217 ntime: 0085 mem: 3.36 + 04-03 23:04:56 | [212][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0080 mem: 3.36 + 04-03 23:04:59 | [212][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0340 ntime: 0078 mem: 3.36 + 04-03 23:05:01 | [212][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0206 ntime: 0079 mem: 3.36 + 04-03 23:05:03 | [212][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0389 ntime: 0083 mem: 3.36 + 04-03 23:05:06 | [212][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 23:05:09 | [212][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0075 mem: 3.36 + 04-03 23:05:10 | Time info >>>> elapsed: 101.80 mins remain: 376.13 mins + 04-03 23:05:11 | [213][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0162 ntime: 0076 mem: 3.36 + 04-03 23:05:13 | [213][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0276 ntime: 0084 mem: 3.36 + 04-03 23:05:15 | [213][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0238 ntime: 0080 mem: 3.36 + 04-03 23:05:17 | [213][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0189 ntime: 0078 mem: 3.36 + 04-03 23:05:19 | [213][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0077 mem: 3.36 + 04-03 23:05:22 | [213][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0076 mem: 3.36 + 04-03 23:05:24 | [213][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0083 ntime: 0079 mem: 3.36 + 04-03 23:05:27 | [213][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0077 mem: 3.36 + 04-03 23:05:30 | [213][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0147 ntime: 0079 mem: 3.36 + 04-03 23:05:33 | [213][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0181 ntime: 0084 mem: 3.36 + 04-03 23:05:35 | [213][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0151 ntime: 0080 mem: 3.36 + 04-03 23:05:38 | [213][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0336 ntime: 0084 mem: 3.36 + 04-03 23:05:40 | [213][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0171 ntime: 0082 mem: 3.36 + 04-03 23:05:42 | [213][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0120 ntime: 0080 mem: 3.36 + 04-03 23:05:46 | [213][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 23:05:48 | [213][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0078 mem: 3.36 + 04-03 23:05:51 | [213][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 23:05:53 | [213][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0218 ntime: 0078 mem: 3.36 + 04-03 23:05:55 | Time info >>>> elapsed: 102.55 mins remain: 376.66 mins + 04-03 23:05:55 | [214][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0057 ntime: 0076 mem: 3.36 + 04-03 23:05:59 | [214][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0345 ntime: 0082 mem: 3.36 + 04-03 23:06:01 | [214][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0297 ntime: 0077 mem: 3.36 + 04-03 23:06:03 | [214][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0211 ntime: 0081 mem: 3.36 + 04-03 23:06:06 | [214][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0326 ntime: 0081 mem: 3.36 + 04-03 23:06:08 | [214][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0221 ntime: 0082 mem: 3.36 + 04-03 23:06:10 | [214][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0189 ntime: 0075 mem: 3.36 + 04-03 23:06:12 | [214][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0080 mem: 3.36 + 04-03 23:06:15 | [214][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0362 ntime: 0080 mem: 3.36 + 04-03 23:06:17 | [214][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0076 mem: 3.36 + 04-03 23:06:19 | [214][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0079 mem: 3.36 + 04-03 23:06:21 | [214][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0078 mem: 3.36 + 04-03 23:06:23 | [214][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 23:06:25 | [214][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0077 mem: 3.36 + 04-03 23:06:27 | [214][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0080 mem: 3.36 + 04-03 23:06:29 | [214][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0089 mem: 3.36 + 04-03 23:06:31 | [214][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0078 mem: 3.36 + 04-03 23:06:33 | [214][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0067 ntime: 0070 mem: 3.36 + 04-03 23:06:35 | Time info >>>> elapsed: 103.20 mins remain: 376.81 mins + 04-03 23:06:35 | [215][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0213 ntime: 0080 mem: 3.36 + 04-03 23:06:37 | [215][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0080 mem: 3.36 + 04-03 23:06:40 | [215][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0045 ntime: 0078 mem: 3.36 + 04-03 23:06:42 | [215][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0328 ntime: 0080 mem: 3.36 + 04-03 23:06:45 | [215][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 23:06:47 | [215][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0083 mem: 3.36 + 04-03 23:06:49 | [215][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0182 ntime: 0079 mem: 3.36 + 04-03 23:06:51 | [215][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0119 ntime: 0087 mem: 3.36 + 04-03 23:06:53 | [215][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0286 ntime: 0077 mem: 3.36 + 04-03 23:06:55 | [215][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 23:06:57 | [215][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0080 mem: 3.36 + 04-03 23:06:59 | [215][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 23:07:01 | [215][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 23:07:03 | [215][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 23:07:05 | [215][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 23:07:07 | [215][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0085 mem: 3.36 + 04-03 23:07:09 | [215][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0193 ntime: 0085 mem: 3.36 + 04-03 23:07:12 | [215][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0156 ntime: 0084 mem: 3.36 + 04-03 23:07:14 | Time info >>>> elapsed: 103.86 mins remain: 376.96 mins + 04-03 23:07:14 | [216][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0085 mem: 3.36 + 04-03 23:07:16 | [216][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0187 ntime: 0086 mem: 3.36 + 04-03 23:07:18 | [216][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0083 mem: 3.36 + 04-03 23:07:21 | [216][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 23:07:23 | [216][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0087 mem: 3.36 + 04-03 23:07:25 | [216][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 23:07:28 | [216][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0195 ntime: 0082 mem: 3.36 + 04-03 23:07:30 | [216][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0309 ntime: 0091 mem: 3.36 + 04-03 23:07:32 | [216][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0140 ntime: 0078 mem: 3.36 + 04-03 23:07:36 | [216][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0399 ntime: 0084 mem: 3.36 + 04-03 23:07:37 | [216][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0111 ntime: 0075 mem: 3.36 + 04-03 23:07:40 | [216][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0214 ntime: 0085 mem: 3.36 + 04-03 23:07:42 | [216][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0085 mem: 3.36 + 04-03 23:07:44 | [216][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0083 mem: 3.36 + 04-03 23:07:47 | [216][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1134 ntime: 0083 mem: 3.36 + 04-03 23:07:50 | [216][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0077 mem: 3.36 + 04-03 23:07:53 | [216][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0378 ntime: 0080 mem: 3.36 + 04-03 23:07:55 | [216][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0076 mem: 3.36 + 04-03 23:07:56 | Time info >>>> elapsed: 104.57 mins remain: 377.31 mins + 04-03 23:07:57 | [217][000/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0246 ntime: 0092 mem: 3.36 + 04-03 23:08:00 | [217][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0075 mem: 3.36 + 04-03 23:08:02 | [217][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1139 ntime: 0080 mem: 3.36 + 04-03 23:08:05 | [217][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0082 mem: 3.36 + 04-03 23:08:07 | [217][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0152 ntime: 0081 mem: 3.36 + 04-03 23:08:09 | [217][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0226 ntime: 0085 mem: 3.36 + 04-03 23:08:11 | [217][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0085 mem: 3.36 + 04-03 23:08:14 | [217][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0395 ntime: 0078 mem: 3.36 + 04-03 23:08:16 | [217][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0083 mem: 3.36 + 04-03 23:08:19 | [217][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0179 ntime: 0089 mem: 3.36 + 04-03 23:08:21 | [217][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 23:08:23 | [217][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0088 ntime: 0078 mem: 3.36 + 04-03 23:08:26 | [217][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 23:08:28 | [217][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 23:08:30 | [217][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0082 mem: 3.36 + 04-03 23:08:32 | [217][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0086 mem: 3.36 + 04-03 23:08:34 | [217][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0084 mem: 3.36 + 04-03 23:08:37 | [217][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0233 ntime: 0079 mem: 3.36 + 04-03 23:08:39 | Time info >>>> elapsed: 105.28 mins remain: 377.65 mins + 04-03 23:08:39 | [218][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0359 ntime: 0082 mem: 3.36 + 04-03 23:08:41 | [218][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0190 ntime: 0085 mem: 3.36 + 04-03 23:08:43 | [218][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0088 mem: 3.36 + 04-03 23:08:46 | [218][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0193 ntime: 0055 mem: 3.36 + 04-03 23:08:48 | [218][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 23:08:50 | [218][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0234 ntime: 0078 mem: 3.36 + 04-03 23:08:52 | [218][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0085 mem: 3.36 + 04-03 23:08:54 | [218][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0125 ntime: 0077 mem: 3.36 + 04-03 23:08:58 | [218][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0411 ntime: 0075 mem: 3.36 + 04-03 23:09:00 | [218][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 23:09:04 | [218][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0480 ntime: 0079 mem: 3.36 + 04-03 23:09:06 | [218][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0217 ntime: 0084 mem: 3.36 + 04-03 23:09:08 | [218][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0101 ntime: 0071 mem: 3.36 + 04-03 23:09:11 | [218][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0091 ntime: 0080 mem: 3.36 + 04-03 23:09:14 | [218][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0072 mem: 3.36 + 04-03 23:09:16 | [218][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0080 mem: 3.36 + 04-03 23:09:18 | [218][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0332 ntime: 0080 mem: 3.36 + 04-03 23:09:21 | [218][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0067 ntime: 0082 mem: 3.36 + 04-03 23:09:22 | Time info >>>> elapsed: 106.00 mins remain: 378.02 mins + 04-03 23:09:23 | [219][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0085 mem: 3.36 + 04-03 23:09:25 | [219][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 23:09:28 | [219][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0250 ntime: 0080 mem: 3.36 + 04-03 23:09:31 | [219][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0087 mem: 3.36 + 04-03 23:09:33 | [219][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0082 mem: 3.36 + 04-03 23:09:35 | [219][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0211 ntime: 0086 mem: 3.36 + 04-03 23:09:37 | [219][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0086 mem: 3.36 + 04-03 23:09:40 | [219][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 23:09:42 | [219][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0130 ntime: 0077 mem: 3.36 + 04-03 23:09:44 | [219][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0067 ntime: 0077 mem: 3.36 + 04-03 23:09:46 | [219][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0124 ntime: 0081 mem: 3.36 + 04-03 23:09:48 | [219][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0078 mem: 3.36 + 04-03 23:09:51 | [219][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 23:09:53 | [219][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 23:09:56 | [219][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 23:09:58 | [219][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0265 ntime: 0082 mem: 3.36 + 04-03 23:10:01 | [219][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0184 ntime: 0088 mem: 3.36 + 04-03 23:10:03 | [219][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0186 ntime: 0076 mem: 3.36 + 04-03 23:10:05 | Time info >>>> elapsed: 106.70 mins remain: 378.32 mins + 04-03 23:10:05 | [220][000/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0045 ntime: 0078 mem: 3.36 + 04-03 23:10:07 | [220][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0219 ntime: 0084 mem: 3.36 + 04-03 23:10:10 | [220][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0262 ntime: 0083 mem: 3.36 + 04-03 23:10:11 | [220][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0074 mem: 3.36 + 04-03 23:10:14 | [220][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 23:10:16 | [220][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0359 ntime: 0077 mem: 3.36 + 04-03 23:10:19 | [220][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0082 mem: 3.36 + 04-03 23:10:21 | [220][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0317 ntime: 0085 mem: 3.36 + 04-03 23:10:23 | [220][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0349 ntime: 0083 mem: 3.36 + 04-03 23:10:25 | [220][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0153 ntime: 0078 mem: 3.36 + 04-03 23:10:27 | [220][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0231 ntime: 0081 mem: 3.36 + 04-03 23:10:29 | [220][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0350 ntime: 0080 mem: 3.36 + 04-03 23:10:32 | [220][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0078 mem: 3.36 + 04-03 23:10:34 | [220][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0218 ntime: 0082 mem: 3.36 + 04-03 23:10:36 | [220][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0285 ntime: 0080 mem: 3.36 + 04-03 23:10:39 | [220][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0102 ntime: 0074 mem: 3.36 + 04-03 23:10:41 | [220][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0079 mem: 3.36 + 04-03 23:10:43 | [220][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 23:10:45 | Time info >>>> elapsed: 107.37 mins remain: 378.48 mins + 04-03 23:10:45 | [221][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 23:10:47 | [221][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 23:10:49 | [221][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0090 mem: 3.36 + 04-03 23:10:52 | [221][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0288 ntime: 0078 mem: 3.36 + 04-03 23:10:54 | [221][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0087 mem: 3.36 + 04-03 23:10:56 | [221][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 23:10:59 | [221][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0149 ntime: 0088 mem: 3.36 + 04-03 23:11:00 | [221][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 23:11:02 | [221][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0199 ntime: 0078 mem: 3.36 + 04-03 23:11:05 | [221][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0079 mem: 3.36 + 04-03 23:11:08 | [221][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 23:11:10 | [221][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0130 ntime: 0073 mem: 3.36 + 04-03 23:11:12 | [221][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0208 ntime: 0086 mem: 3.36 + 04-03 23:11:14 | [221][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0500 ntime: 0079 mem: 3.36 + 04-03 23:11:16 | [221][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0239 ntime: 0086 mem: 3.36 + 04-03 23:11:19 | [221][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0273 ntime: 0090 mem: 3.36 + 04-03 23:11:21 | [221][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0195 ntime: 0086 mem: 3.36 + 04-03 23:11:23 | [221][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 23:11:25 | Time info >>>> elapsed: 108.04 mins remain: 378.61 mins + 04-03 23:11:25 | [222][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0086 mem: 3.36 + 04-03 23:11:27 | [222][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0083 mem: 3.36 + 04-03 23:11:29 | [222][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0282 ntime: 0076 mem: 3.36 + 04-03 23:11:32 | [222][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0322 ntime: 0080 mem: 3.36 + 04-03 23:11:35 | [222][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0236 ntime: 0081 mem: 3.36 + 04-03 23:11:37 | [222][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0061 mem: 3.36 + 04-03 23:11:38 | [222][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0071 mem: 3.36 + 04-03 23:11:41 | [222][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0259 ntime: 0079 mem: 3.36 + 04-03 23:11:44 | [222][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 23:11:46 | [222][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0145 ntime: 0076 mem: 3.36 + 04-03 23:11:49 | [222][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 23:11:51 | [222][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0222 ntime: 0075 mem: 3.36 + 04-03 23:11:54 | [222][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 23:11:56 | [222][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0365 ntime: 0081 mem: 3.36 + 04-03 23:11:58 | [222][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 23:12:01 | [222][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0159 ntime: 0079 mem: 3.36 + 04-03 23:12:03 | [222][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0244 ntime: 0076 mem: 3.36 + 04-03 23:12:05 | [222][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0136 ntime: 0084 mem: 3.36 + 04-03 23:12:07 | Time info >>>> elapsed: 108.74 mins remain: 378.89 mins + 04-03 23:12:07 | [223][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0121 ntime: 0073 mem: 3.36 + 04-03 23:12:10 | [223][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0362 ntime: 0081 mem: 3.36 + 04-03 23:12:12 | [223][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0265 ntime: 0074 mem: 3.36 + 04-03 23:12:14 | [223][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0198 ntime: 0085 mem: 3.36 + 04-03 23:12:17 | [223][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0079 mem: 3.36 + 04-03 23:12:19 | [223][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 23:12:21 | [223][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0342 ntime: 0079 mem: 3.36 + 04-03 23:12:23 | [223][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0302 ntime: 0085 mem: 3.36 + 04-03 23:12:25 | [223][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 23:12:28 | [223][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 23:12:31 | [223][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0076 mem: 3.36 + 04-03 23:12:33 | [223][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0073 ntime: 0086 mem: 3.36 + 04-03 23:12:35 | [223][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0079 mem: 3.36 + 04-03 23:12:37 | [223][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0224 ntime: 0086 mem: 3.36 + 04-03 23:12:39 | [223][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 23:12:42 | [223][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 23:12:44 | [223][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 23:12:46 | [223][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 23:12:48 | Time info >>>> elapsed: 109.42 mins remain: 379.07 mins + 04-03 23:12:48 | [224][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0228 ntime: 0073 mem: 3.36 + 04-03 23:12:50 | [224][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0085 mem: 3.36 + 04-03 23:12:52 | [224][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0076 mem: 3.36 + 04-03 23:12:54 | [224][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0070 ntime: 0079 mem: 3.36 + 04-03 23:12:56 | [224][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0077 mem: 3.36 + 04-03 23:12:59 | [224][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0077 mem: 3.36 + 04-03 23:13:02 | [224][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0218 ntime: 0081 mem: 3.36 + 04-03 23:13:04 | [224][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0278 ntime: 0074 mem: 3.36 + 04-03 23:13:06 | [224][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0077 mem: 3.36 + 04-03 23:13:10 | [224][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0076 mem: 3.36 + 04-03 23:13:12 | [224][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0076 mem: 3.36 + 04-03 23:13:14 | [224][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 23:13:16 | [224][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0141 ntime: 0079 mem: 3.36 + 04-03 23:13:18 | [224][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0143 ntime: 0078 mem: 3.36 + 04-03 23:13:20 | [224][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0086 mem: 3.36 + 04-03 23:13:22 | [224][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 23:13:24 | [224][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0078 mem: 3.36 + 04-03 23:13:26 | [224][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 23:13:28 | Time info >>>> elapsed: 110.10 mins remain: 379.23 mins + 04-03 23:13:29 | [225][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0320 ntime: 0079 mem: 3.36 + 04-03 23:13:31 | [225][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0280 ntime: 0080 mem: 3.36 + 04-03 23:13:33 | [225][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0433 ntime: 0080 mem: 3.36 + 04-03 23:13:35 | [225][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0225 ntime: 0089 mem: 3.36 + 04-03 23:13:37 | [225][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0075 mem: 3.36 + 04-03 23:13:40 | [225][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 23:13:42 | [225][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0381 ntime: 0079 mem: 3.36 + 04-03 23:13:45 | [225][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 23:13:47 | [225][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0154 ntime: 0076 mem: 3.36 + 04-03 23:13:49 | [225][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0076 mem: 3.36 + 04-03 23:13:52 | [225][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0077 mem: 3.36 + 04-03 23:13:54 | [225][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0085 ntime: 0081 mem: 3.36 + 04-03 23:13:56 | [225][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0253 ntime: 0076 mem: 3.36 + 04-03 23:13:58 | [225][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0113 ntime: 0081 mem: 3.36 + 04-03 23:14:01 | [225][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0307 ntime: 0082 mem: 3.36 + 04-03 23:14:04 | [225][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0151 ntime: 0081 mem: 3.36 + 04-03 23:14:06 | [225][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0148 ntime: 0076 mem: 3.36 + 04-03 23:14:08 | [225][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0090 mem: 3.36 + 04-03 23:14:09 | Time info >>>> elapsed: 110.78 mins remain: 379.39 mins + 04-03 23:14:09 | [226][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0086 mem: 3.36 + 04-03 23:14:11 | [226][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0070 ntime: 0079 mem: 3.36 + 04-03 23:14:14 | [226][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0194 ntime: 0082 mem: 3.36 + 04-03 23:14:16 | [226][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0045 ntime: 0077 mem: 3.36 + 04-03 23:14:18 | [226][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0075 mem: 3.36 + 04-03 23:14:21 | [226][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0073 mem: 3.36 + 04-03 23:14:23 | [226][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0225 ntime: 0084 mem: 3.36 + 04-03 23:14:25 | [226][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 23:14:27 | [226][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0319 ntime: 0079 mem: 3.36 + 04-03 23:14:29 | [226][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0310 ntime: 0083 mem: 3.36 + 04-03 23:14:31 | [226][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0054 mem: 3.36 + 04-03 23:14:33 | [226][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0077 mem: 3.36 + 04-03 23:14:35 | [226][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0090 ntime: 0075 mem: 3.36 + 04-03 23:14:37 | [226][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 23:14:39 | [226][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0084 ntime: 0081 mem: 3.36 + 04-03 23:14:41 | [226][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0120 ntime: 0079 mem: 3.36 + 04-03 23:14:44 | [226][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0254 ntime: 0087 mem: 3.36 + 04-03 23:14:46 | [226][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0085 mem: 3.36 + 04-03 23:14:47 | Time info >>>> elapsed: 111.42 mins remain: 379.41 mins + 04-03 23:14:48 | [227][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0185 ntime: 0076 mem: 3.36 + 04-03 23:14:50 | [227][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 23:14:52 | [227][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0239 ntime: 0063 mem: 3.36 + 04-03 23:14:55 | [227][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0240 ntime: 0081 mem: 3.36 + 04-03 23:14:57 | [227][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0293 ntime: 0083 mem: 3.36 + 04-03 23:14:59 | [227][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0081 mem: 3.36 + 04-03 23:15:02 | [227][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0079 ntime: 0079 mem: 3.36 + 04-03 23:15:04 | [227][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 23:15:06 | [227][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0093 ntime: 0081 mem: 3.36 + 04-03 23:15:08 | [227][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0082 ntime: 0083 mem: 3.36 + 04-03 23:15:10 | [227][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0078 mem: 3.36 + 04-03 23:15:13 | [227][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0431 ntime: 0079 mem: 3.36 + 04-03 23:15:15 | [227][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0216 ntime: 0079 mem: 3.36 + 04-03 23:15:17 | [227][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 23:15:19 | [227][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0078 mem: 3.36 + 04-03 23:15:22 | [227][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0119 ntime: 0082 mem: 3.36 + 04-03 23:15:24 | [227][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0075 mem: 3.36 + 04-03 23:15:26 | [227][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0362 ntime: 0077 mem: 3.36 + 04-03 23:15:28 | Time info >>>> elapsed: 112.09 mins remain: 379.52 mins + 04-03 23:15:28 | [228][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0108 ntime: 0080 mem: 3.36 + 04-03 23:15:29 | [228][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0101 ntime: 0091 mem: 3.36 + 04-03 23:15:32 | [228][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0084 mem: 3.36 + 04-03 23:15:34 | [228][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0250 ntime: 0084 mem: 3.36 + 04-03 23:15:36 | [228][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0171 ntime: 0081 mem: 3.36 + 04-03 23:15:38 | [228][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0114 ntime: 0079 mem: 3.36 + 04-03 23:15:40 | [228][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 23:15:42 | [228][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0086 mem: 3.36 + 04-03 23:15:43 | [228][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0081 ntime: 0080 mem: 3.36 + 04-03 23:15:46 | [228][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0082 ntime: 0078 mem: 3.36 + 04-03 23:15:47 | [228][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 23:15:49 | [228][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 23:15:51 | [228][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0225 ntime: 0078 mem: 3.36 + 04-03 23:15:53 | [228][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0113 ntime: 0075 mem: 3.36 + 04-03 23:15:56 | [228][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0215 ntime: 0081 mem: 3.36 + 04-03 23:15:57 | [228][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0088 mem: 3.36 + 04-03 23:15:59 | [228][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0080 mem: 3.36 + 04-03 23:16:02 | [228][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0093 ntime: 0084 mem: 3.36 + 04-03 23:16:03 | Time info >>>> elapsed: 112.68 mins remain: 379.38 mins + 04-03 23:16:04 | [229][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0209 ntime: 0081 mem: 3.36 + 04-03 23:16:05 | [229][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 23:16:07 | [229][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0082 mem: 3.36 + 04-03 23:16:09 | [229][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0133 ntime: 0078 mem: 3.36 + 04-03 23:16:11 | [229][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 23:16:12 | [229][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 23:16:14 | [229][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0196 ntime: 0081 mem: 3.36 + 04-03 23:16:16 | [229][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0082 mem: 3.36 + 04-03 23:16:18 | [229][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 23:16:20 | [229][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0212 ntime: 0081 mem: 3.36 + 04-03 23:16:23 | [229][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0070 mem: 3.36 + 04-03 23:16:25 | [229][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0084 mem: 3.36 + 04-03 23:16:27 | [229][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0082 mem: 3.36 + 04-03 23:16:29 | [229][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0247 ntime: 0081 mem: 3.36 + 04-03 23:16:31 | [229][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0079 mem: 3.36 + 04-03 23:16:33 | [229][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0095 ntime: 0080 mem: 3.36 + 04-03 23:16:35 | [229][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0080 mem: 3.36 + 04-03 23:16:37 | [229][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0080 mem: 3.36 + 04-03 23:16:39 | Time info >>>> elapsed: 113.27 mins remain: 379.21 mins + 04-03 23:16:39 | [230][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0265 ntime: 0077 mem: 3.36 + 04-03 23:16:41 | [230][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 23:16:44 | [230][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0204 ntime: 0080 mem: 3.36 + 04-03 23:16:47 | [230][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0085 ntime: 0081 mem: 3.36 + 04-03 23:16:50 | [230][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0081 mem: 3.36 + 04-03 23:16:52 | [230][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 23:16:54 | [230][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 23:16:57 | [230][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0086 mem: 3.36 + 04-03 23:16:59 | [230][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 23:17:01 | [230][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0110 ntime: 0084 mem: 3.36 + 04-03 23:17:03 | [230][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0081 mem: 3.36 + 04-03 23:17:06 | [230][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0129 ntime: 0084 mem: 3.36 + 04-03 23:17:08 | [230][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0187 ntime: 0084 mem: 3.36 + 04-03 23:17:11 | [230][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0225 ntime: 0082 mem: 3.36 + 04-03 23:17:13 | [230][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0148 ntime: 0086 mem: 3.36 + 04-03 23:17:15 | [230][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 23:17:17 | [230][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0084 mem: 3.36 + 04-03 23:17:19 | [230][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0081 mem: 3.36 + 04-03 23:17:21 | Time info >>>> elapsed: 113.97 mins remain: 379.42 mins + 04-03 23:17:21 | [231][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0254 ntime: 0082 mem: 3.36 + 04-03 23:17:23 | [231][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0073 mem: 3.36 + 04-03 23:17:26 | [231][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0258 ntime: 0082 mem: 3.36 + 04-03 23:17:28 | [231][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0206 ntime: 0057 mem: 3.36 + 04-03 23:17:30 | [231][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 23:17:32 | [231][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 23:17:34 | [231][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0078 ntime: 0074 mem: 3.36 + 04-03 23:17:37 | [231][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0408 ntime: 0089 mem: 3.36 + 04-03 23:17:39 | [231][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 23:17:41 | [231][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0158 ntime: 0078 mem: 3.36 + 04-03 23:17:43 | [231][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0080 mem: 3.36 + 04-03 23:17:45 | [231][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0112 ntime: 0075 mem: 3.36 + 04-03 23:17:48 | [231][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0082 ntime: 0079 mem: 3.36 + 04-03 23:17:49 | [231][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0080 mem: 3.36 + 04-03 23:17:51 | [231][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0263 ntime: 0084 mem: 3.36 + 04-03 23:17:54 | [231][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0290 ntime: 0076 mem: 3.36 + 04-03 23:17:56 | [231][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0080 mem: 3.36 + 04-03 23:17:58 | [231][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 23:18:00 | Time info >>>> elapsed: 114.62 mins remain: 379.44 mins + 04-03 23:18:00 | [232][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0403 ntime: 0079 mem: 3.36 + 04-03 23:18:02 | [232][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0274 ntime: 0079 mem: 3.36 + 04-03 23:18:05 | [232][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0083 mem: 3.36 + 04-03 23:18:07 | [232][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 23:18:09 | [232][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0083 mem: 3.36 + 04-03 23:18:11 | [232][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 23:18:13 | [232][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0085 mem: 3.36 + 04-03 23:18:16 | [232][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0077 mem: 3.36 + 04-03 23:18:18 | [232][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0139 ntime: 0089 mem: 3.36 + 04-03 23:18:20 | [232][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0072 mem: 3.36 + 04-03 23:18:22 | [232][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 23:18:24 | [232][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0083 ntime: 0079 mem: 3.36 + 04-03 23:18:26 | [232][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0151 ntime: 0076 mem: 3.36 + 04-03 23:18:28 | [232][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0081 mem: 3.36 + 04-03 23:18:31 | [232][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0408 ntime: 0084 mem: 3.36 + 04-03 23:18:33 | [232][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0091 mem: 3.36 + 04-03 23:18:36 | [232][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0044 ntime: 0078 mem: 3.36 + 04-03 23:18:37 | [232][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 23:18:39 | Time info >>>> elapsed: 115.28 mins remain: 379.49 mins + 04-03 23:18:40 | [233][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0363 ntime: 0075 mem: 3.36 + 04-03 23:18:41 | [233][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 23:18:44 | [233][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0076 mem: 3.36 + 04-03 23:18:47 | [233][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0081 mem: 3.36 + 04-03 23:18:49 | [233][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0090 mem: 3.36 + 04-03 23:18:52 | [233][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0197 ntime: 0078 mem: 3.36 + 04-03 23:18:54 | [233][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0094 ntime: 0080 mem: 3.36 + 04-03 23:18:56 | [233][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0082 mem: 3.36 + 04-03 23:18:59 | [233][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0230 ntime: 0085 mem: 3.36 + 04-03 23:19:01 | [233][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0079 mem: 3.36 + 04-03 23:19:04 | [233][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0249 ntime: 0085 mem: 3.36 + 04-03 23:19:06 | [233][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0074 mem: 3.36 + 04-03 23:19:08 | [233][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0075 mem: 3.36 + 04-03 23:19:11 | [233][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0148 ntime: 0084 mem: 3.36 + 04-03 23:19:14 | [233][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0257 ntime: 0084 mem: 3.36 + 04-03 23:19:16 | [233][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0195 ntime: 0079 mem: 3.36 + 04-03 23:19:18 | [233][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0073 mem: 3.36 + 04-03 23:19:21 | [233][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0107 ntime: 0081 mem: 3.36 + 04-03 23:19:23 | Time info >>>> elapsed: 116.01 mins remain: 379.76 mins + 04-03 23:19:23 | [234][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0204 ntime: 0080 mem: 3.36 + 04-03 23:19:26 | [234][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0081 mem: 3.36 + 04-03 23:19:28 | [234][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0125 ntime: 0075 mem: 3.36 + 04-03 23:19:30 | [234][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 23:19:32 | [234][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0131 ntime: 0085 mem: 3.36 + 04-03 23:19:34 | [234][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 23:19:37 | [234][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0074 mem: 3.36 + 04-03 23:19:38 | [234][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0098 ntime: 0072 mem: 3.36 + 04-03 23:19:40 | [234][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 23:19:43 | [234][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0167 ntime: 0086 mem: 3.36 + 04-03 23:19:46 | [234][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0312 ntime: 0079 mem: 3.36 + 04-03 23:19:48 | [234][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0079 mem: 3.36 + 04-03 23:19:50 | [234][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0110 ntime: 0078 mem: 3.36 + 04-03 23:19:52 | [234][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 23:19:54 | [234][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0076 mem: 3.36 + 04-03 23:19:57 | [234][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0071 mem: 3.36 + 04-03 23:19:59 | [234][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0290 ntime: 0086 mem: 3.36 + 04-03 23:20:02 | [234][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0288 ntime: 0083 mem: 3.36 + 04-03 23:20:04 | Time info >>>> elapsed: 116.69 mins remain: 379.86 mins + 04-03 23:20:04 | [235][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0241 ntime: 0081 mem: 3.36 + 04-03 23:20:06 | [235][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0208 ntime: 0085 mem: 3.36 + 04-03 23:20:08 | [235][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0227 ntime: 0082 mem: 3.36 + 04-03 23:20:10 | [235][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0077 mem: 3.36 + 04-03 23:20:12 | [235][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0086 mem: 3.36 + 04-03 23:20:14 | [235][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 23:20:17 | [235][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0297 ntime: 0055 mem: 3.36 + 04-03 23:20:19 | [235][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0270 ntime: 0077 mem: 3.36 + 04-03 23:20:21 | [235][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0101 ntime: 0077 mem: 3.36 + 04-03 23:20:23 | [235][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 23:20:25 | [235][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0132 ntime: 0081 mem: 3.36 + 04-03 23:20:27 | [235][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0086 mem: 3.36 + 04-03 23:20:29 | [235][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0070 ntime: 0077 mem: 3.36 + 04-03 23:20:32 | [235][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0244 ntime: 0080 mem: 3.36 + 04-03 23:20:37 | [235][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 23:20:39 | [235][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0083 mem: 3.36 + 04-03 23:20:41 | [235][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 23:20:43 | [235][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0199 ntime: 0081 mem: 3.36 + 04-03 23:20:44 | Time info >>>> elapsed: 117.37 mins remain: 379.95 mins + 04-03 23:20:44 | [236][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0079 mem: 3.36 + 04-03 23:20:47 | [236][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0433 ntime: 0086 mem: 3.36 + 04-03 23:20:49 | [236][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0263 ntime: 0077 mem: 3.36 + 04-03 23:20:52 | [236][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0366 ntime: 0078 mem: 3.36 + 04-03 23:20:54 | [236][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0181 ntime: 0083 mem: 3.36 + 04-03 23:20:55 | [236][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0090 mem: 3.36 + 04-03 23:20:57 | [236][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0080 mem: 3.36 + 04-03 23:20:59 | [236][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 23:21:01 | [236][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0202 ntime: 0087 mem: 3.36 + 04-03 23:21:03 | [236][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0075 mem: 3.36 + 04-03 23:21:05 | [236][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 23:21:08 | [236][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0110 ntime: 0075 mem: 3.36 + 04-03 23:21:10 | [236][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0083 mem: 3.36 + 04-03 23:21:12 | [236][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 23:21:14 | [236][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0402 ntime: 0087 mem: 3.36 + 04-03 23:21:16 | [236][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0293 ntime: 0080 mem: 3.36 + 04-03 23:21:18 | [236][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0075 mem: 3.36 + 04-03 23:21:20 | [236][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0166 ntime: 0076 mem: 3.36 + 04-03 23:21:22 | Time info >>>> elapsed: 117.99 mins remain: 379.86 mins + 04-03 23:21:22 | [237][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0059 ntime: 0082 mem: 3.36 + 04-03 23:21:24 | [237][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0205 ntime: 0082 mem: 3.36 + 04-03 23:21:26 | [237][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0236 ntime: 0074 mem: 3.36 + 04-03 23:21:28 | [237][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 23:21:30 | [237][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0076 mem: 3.36 + 04-03 23:21:32 | [237][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0234 ntime: 0079 mem: 3.36 + 04-03 23:21:34 | [237][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0168 ntime: 0079 mem: 3.36 + 04-03 23:21:36 | [237][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0082 mem: 3.36 + 04-03 23:21:38 | [237][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0076 ntime: 0079 mem: 3.36 + 04-03 23:21:40 | [237][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 23:21:42 | [237][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0251 ntime: 0082 mem: 3.36 + 04-03 23:21:45 | [237][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 23:21:47 | [237][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0126 ntime: 0084 mem: 3.36 + 04-03 23:21:49 | [237][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0073 ntime: 0078 mem: 3.36 + 04-03 23:21:51 | [237][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0081 mem: 3.36 + 04-03 23:21:52 | [237][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 23:21:54 | [237][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 23:21:56 | [237][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0076 mem: 3.36 + 04-03 23:21:58 | Time info >>>> elapsed: 118.60 mins remain: 379.72 mins + 04-03 23:21:59 | [238][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0072 mem: 3.36 + 04-03 23:22:01 | [238][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0194 ntime: 0078 mem: 3.36 + 04-03 23:22:03 | [238][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0151 ntime: 0086 mem: 3.36 + 04-03 23:22:05 | [238][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0183 ntime: 0074 mem: 3.36 + 04-03 23:22:07 | [238][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0407 ntime: 0085 mem: 3.36 + 04-03 23:22:09 | [238][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 23:22:11 | [238][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0181 ntime: 0085 mem: 3.36 + 04-03 23:22:13 | [238][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0293 ntime: 0074 mem: 3.36 + 04-03 23:22:15 | [238][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0081 ntime: 0080 mem: 3.36 + 04-03 23:22:18 | [238][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0530 ntime: 0076 mem: 3.36 + 04-03 23:22:20 | [238][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0076 mem: 3.36 + 04-03 23:22:22 | [238][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 23:22:24 | [238][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0648 ntime: 0085 mem: 3.36 + 04-03 23:22:27 | [238][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 23:22:29 | [238][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0075 mem: 3.36 + 04-03 23:22:31 | [238][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0114 ntime: 0075 mem: 3.36 + 04-03 23:22:33 | [238][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 23:22:36 | [238][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0355 ntime: 0081 mem: 3.36 + 04-03 23:22:38 | Time info >>>> elapsed: 119.27 mins remain: 379.76 mins + 04-03 23:22:38 | [239][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0077 mem: 3.36 + 04-03 23:22:41 | [239][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0072 mem: 3.36 + 04-03 23:22:43 | [239][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0103 ntime: 0082 mem: 3.36 + 04-03 23:22:45 | [239][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 23:22:47 | [239][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0083 mem: 3.36 + 04-03 23:22:49 | [239][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0197 ntime: 0080 mem: 3.36 + 04-03 23:22:51 | [239][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 23:22:53 | [239][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0086 mem: 3.36 + 04-03 23:22:56 | [239][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0077 mem: 3.36 + 04-03 23:22:58 | [239][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0081 ntime: 0079 mem: 3.36 + 04-03 23:23:01 | [239][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0081 mem: 3.36 + 04-03 23:23:03 | [239][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0221 ntime: 0077 mem: 3.36 + 04-03 23:23:05 | [239][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0197 ntime: 0081 mem: 3.36 + 04-03 23:23:08 | [239][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0131 ntime: 0080 mem: 3.36 + 04-03 23:23:09 | [239][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 23:23:12 | [239][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0144 ntime: 0063 mem: 3.36 + 04-03 23:23:14 | [239][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0356 ntime: 0074 mem: 3.36 + 04-03 23:23:16 | [239][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0393 ntime: 0076 mem: 3.36 + 04-03 23:23:18 | Time info >>>> elapsed: 119.92 mins remain: 379.76 mins + 04-03 23:23:18 | [240][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0083 mem: 3.36 + 04-03 23:23:20 | [240][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0121 ntime: 0074 mem: 3.36 + 04-03 23:23:22 | [240][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0195 ntime: 0076 mem: 3.36 + 04-03 23:23:24 | [240][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0085 mem: 3.36 + 04-03 23:23:26 | [240][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 23:23:28 | [240][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0076 mem: 3.36 + 04-03 23:23:30 | [240][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0085 mem: 3.36 + 04-03 23:23:33 | [240][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0084 mem: 3.36 + 04-03 23:23:35 | [240][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0223 ntime: 0080 mem: 3.36 + 04-03 23:23:38 | [240][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0311 ntime: 0076 mem: 3.36 + 04-03 23:23:41 | [240][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0081 ntime: 0079 mem: 3.36 + 04-03 23:23:43 | [240][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0181 ntime: 0079 mem: 3.36 + 04-03 23:23:45 | [240][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0080 mem: 3.36 + 04-03 23:23:47 | [240][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0074 mem: 3.36 + 04-03 23:23:49 | [240][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0070 ntime: 0080 mem: 3.36 + 04-03 23:23:51 | [240][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0249 ntime: 0078 mem: 3.36 + 04-03 23:23:53 | [240][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0302 ntime: 0077 mem: 3.36 + 04-03 23:23:55 | [240][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0078 mem: 3.36 + 04-03 23:23:57 | Time info >>>> elapsed: 120.58 mins remain: 379.75 mins + 04-03 23:23:57 | [241][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0281 ntime: 0078 mem: 3.36 + 04-03 23:24:00 | [241][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0215 ntime: 0083 mem: 3.36 + 04-03 23:24:02 | [241][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0243 ntime: 0076 mem: 3.36 + 04-03 23:24:04 | [241][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 23:24:07 | [241][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0438 ntime: 0081 mem: 3.36 + 04-03 23:24:09 | [241][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0310 ntime: 0086 mem: 3.36 + 04-03 23:24:12 | [241][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0067 ntime: 0076 mem: 3.36 + 04-03 23:24:15 | [241][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0121 ntime: 0084 mem: 3.36 + 04-03 23:24:18 | [241][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0074 mem: 3.36 + 04-03 23:24:20 | [241][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0241 ntime: 0081 mem: 3.36 + 04-03 23:24:22 | [241][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 23:24:24 | [241][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0089 mem: 3.36 + 04-03 23:24:25 | [241][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0144 ntime: 0078 mem: 3.36 + 04-03 23:24:28 | [241][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 23:24:31 | [241][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0791 ntime: 0074 mem: 3.36 + 04-03 23:24:33 | [241][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0088 mem: 3.36 + 04-03 23:24:35 | [241][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0123 ntime: 0079 mem: 3.36 + 04-03 23:24:37 | [241][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0072 mem: 3.36 + 04-03 23:24:39 | Time info >>>> elapsed: 121.28 mins remain: 379.86 mins + 04-03 23:24:39 | [242][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0282 ntime: 0086 mem: 3.36 + 04-03 23:24:41 | [242][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0091 ntime: 0084 mem: 3.36 + 04-03 23:24:43 | [242][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 23:24:45 | [242][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0079 mem: 3.36 + 04-03 23:24:47 | [242][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 23:24:49 | [242][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0139 ntime: 0080 mem: 3.36 + 04-03 23:24:51 | [242][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0174 ntime: 0080 mem: 3.36 + 04-03 23:24:53 | [242][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0078 mem: 3.36 + 04-03 23:24:55 | [242][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-03 23:24:57 | [242][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0435 ntime: 0087 mem: 3.36 + 04-03 23:24:59 | [242][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 23:25:01 | [242][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0115 ntime: 0077 mem: 3.36 + 04-03 23:25:03 | [242][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0078 mem: 3.36 + 04-03 23:25:05 | [242][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0195 ntime: 0087 mem: 3.36 + 04-03 23:25:07 | [242][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0123 ntime: 0077 mem: 3.36 + 04-03 23:25:09 | [242][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0176 ntime: 0076 mem: 3.36 + 04-03 23:25:11 | [242][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0081 mem: 3.36 + 04-03 23:25:13 | [242][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0095 ntime: 0083 mem: 3.36 + 04-03 23:25:14 | Time info >>>> elapsed: 121.87 mins remain: 379.64 mins + 04-03 23:25:15 | [243][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0193 ntime: 0075 mem: 3.36 + 04-03 23:25:17 | [243][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0078 mem: 3.36 + 04-03 23:25:19 | [243][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0185 ntime: 0082 mem: 3.36 + 04-03 23:25:21 | [243][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 23:25:23 | [243][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0299 ntime: 0081 mem: 3.36 + 04-03 23:25:25 | [243][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0112 ntime: 0080 mem: 3.36 + 04-03 23:25:28 | [243][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0203 ntime: 0086 mem: 3.36 + 04-03 23:25:30 | [243][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0086 mem: 3.36 + 04-03 23:25:32 | [243][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 23:25:34 | [243][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0307 ntime: 0084 mem: 3.36 + 04-03 23:25:36 | [243][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 23:25:38 | [243][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0083 mem: 3.36 + 04-03 23:25:40 | [243][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0103 ntime: 0080 mem: 3.36 + 04-03 23:25:43 | [243][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0535 ntime: 0080 mem: 3.36 + 04-03 23:25:45 | [243][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0081 mem: 3.36 + 04-03 23:25:48 | [243][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0090 mem: 3.36 + 04-03 23:25:51 | [243][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 23:25:53 | [243][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0204 ntime: 0080 mem: 3.36 + 04-03 23:25:55 | Time info >>>> elapsed: 122.54 mins remain: 379.67 mins + 04-03 23:25:55 | [244][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0303 ntime: 0083 mem: 3.36 + 04-03 23:25:57 | [244][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0130 ntime: 0085 mem: 3.36 + 04-03 23:25:59 | [244][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0145 ntime: 0077 mem: 3.36 + 04-03 23:26:02 | [244][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0284 ntime: 0081 mem: 3.36 + 04-03 23:26:03 | [244][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0080 mem: 3.36 + 04-03 23:26:05 | [244][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0110 ntime: 0055 mem: 3.36 + 04-03 23:26:07 | [244][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 23:26:11 | [244][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0283 ntime: 0079 mem: 3.36 + 04-03 23:26:13 | [244][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 23:26:16 | [244][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0235 ntime: 0086 mem: 3.36 + 04-03 23:26:18 | [244][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 23:26:20 | [244][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0107 ntime: 0078 mem: 3.36 + 04-03 23:26:23 | [244][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0387 ntime: 0079 mem: 3.36 + 04-03 23:26:25 | [244][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0083 mem: 3.36 + 04-03 23:26:28 | [244][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0082 mem: 3.36 + 04-03 23:26:30 | [244][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0087 mem: 3.36 + 04-03 23:26:32 | [244][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0161 ntime: 0075 mem: 3.36 + 04-03 23:26:34 | [244][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0070 ntime: 0084 mem: 3.36 + 04-03 23:26:36 | Time info >>>> elapsed: 123.22 mins remain: 379.72 mins + 04-03 23:26:36 | [245][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0077 mem: 3.36 + 04-03 23:26:38 | [245][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0077 mem: 3.36 + 04-03 23:26:40 | [245][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0087 mem: 3.36 + 04-03 23:26:42 | [245][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0070 ntime: 0080 mem: 3.36 + 04-03 23:26:45 | [245][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0127 ntime: 0083 mem: 3.36 + 04-03 23:26:48 | [245][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0219 ntime: 0080 mem: 3.36 + 04-03 23:26:51 | [245][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0846 ntime: 0088 mem: 3.36 + 04-03 23:26:53 | [245][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0242 ntime: 0080 mem: 3.36 + 04-03 23:26:55 | [245][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0083 mem: 3.36 + 04-03 23:26:57 | [245][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0044 ntime: 0082 mem: 3.36 + 04-03 23:26:59 | [245][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 23:27:01 | [245][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0078 mem: 3.36 + 04-03 23:27:04 | [245][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0081 mem: 3.36 + 04-03 23:27:06 | [245][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0080 mem: 3.36 + 04-03 23:27:08 | [245][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0092 ntime: 0078 mem: 3.36 + 04-03 23:27:11 | [245][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0084 mem: 3.36 + 04-03 23:27:14 | [245][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0297 ntime: 0080 mem: 3.36 + 04-03 23:27:16 | [245][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0079 mem: 3.36 + 04-03 23:27:18 | Time info >>>> elapsed: 123.92 mins remain: 379.83 mins + 04-03 23:27:18 | [246][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 23:27:20 | [246][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0088 mem: 3.36 + 04-03 23:27:23 | [246][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0238 ntime: 0080 mem: 3.36 + 04-03 23:27:25 | [246][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0315 ntime: 0076 mem: 3.36 + 04-03 23:27:27 | [246][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0123 ntime: 0078 mem: 3.36 + 04-03 23:27:29 | [246][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 23:27:31 | [246][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 23:27:33 | [246][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0203 ntime: 0080 mem: 3.36 + 04-03 23:27:36 | [246][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0239 ntime: 0078 mem: 3.36 + 04-03 23:27:38 | [246][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0149 ntime: 0076 mem: 3.36 + 04-03 23:27:41 | [246][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0381 ntime: 0076 mem: 3.36 + 04-03 23:27:43 | [246][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 23:27:45 | [246][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0075 mem: 3.36 + 04-03 23:27:48 | [246][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0078 mem: 3.36 + 04-03 23:27:50 | [246][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0081 mem: 3.36 + 04-03 23:27:52 | [246][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0222 ntime: 0082 mem: 3.36 + 04-03 23:27:54 | [246][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0073 ntime: 0076 mem: 3.36 + 04-03 23:27:57 | [246][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0390 ntime: 0077 mem: 3.36 + 04-03 23:28:01 | Time info >>>> elapsed: 124.64 mins remain: 379.98 mins + 04-03 23:28:01 | [247][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0320 ntime: 0083 mem: 3.36 + 04-03 23:28:03 | [247][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 23:28:06 | [247][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0297 ntime: 0081 mem: 3.36 + 04-03 23:28:08 | [247][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0086 mem: 3.36 + 04-03 23:28:10 | [247][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0432 ntime: 0075 mem: 3.36 + 04-03 23:28:12 | [247][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0131 ntime: 0077 mem: 3.36 + 04-03 23:28:15 | [247][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0086 ntime: 0079 mem: 3.36 + 04-03 23:28:18 | [247][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 23:28:20 | [247][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0156 ntime: 0087 mem: 3.36 + 04-03 23:28:22 | [247][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0232 ntime: 0071 mem: 3.36 + 04-03 23:28:24 | [247][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 23:28:27 | [247][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0316 ntime: 0078 mem: 3.36 + 04-03 23:28:30 | [247][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0082 mem: 3.36 + 04-03 23:28:32 | [247][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0079 mem: 3.36 + 04-03 23:28:34 | [247][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0344 ntime: 0089 mem: 3.36 + 04-03 23:28:36 | [247][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 23:28:39 | [247][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0248 ntime: 0084 mem: 3.36 + 04-03 23:28:41 | [247][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0077 mem: 3.36 + 04-03 23:28:43 | Time info >>>> elapsed: 125.35 mins remain: 380.08 mins + 04-03 23:28:43 | [248][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 23:28:46 | [248][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0077 ntime: 0083 mem: 3.36 + 04-03 23:28:48 | [248][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0081 mem: 3.36 + 04-03 23:28:50 | [248][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0203 ntime: 0090 mem: 3.36 + 04-03 23:28:54 | [248][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0328 ntime: 0085 mem: 3.36 + 04-03 23:28:56 | [248][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 23:28:58 | [248][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0361 ntime: 0083 mem: 3.36 + 04-03 23:29:01 | [248][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0291 ntime: 0079 mem: 3.36 + 04-03 23:29:03 | [248][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0071 mem: 3.36 + 04-03 23:29:06 | [248][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0276 ntime: 0069 mem: 3.36 + 04-03 23:29:08 | [248][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 23:29:10 | [248][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0080 mem: 3.36 + 04-03 23:29:12 | [248][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0133 ntime: 0081 mem: 3.36 + 04-03 23:29:14 | [248][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 23:29:17 | [248][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0076 mem: 3.36 + 04-03 23:29:19 | [248][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 23:29:21 | [248][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0076 mem: 3.36 + 04-03 23:29:23 | [248][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 23:29:24 | Time info >>>> elapsed: 126.03 mins remain: 380.13 mins + 04-03 23:29:25 | [249][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0076 mem: 3.36 + 04-03 23:29:26 | [249][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0070 ntime: 0074 mem: 3.36 + 04-03 23:29:28 | [249][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0326 ntime: 0086 mem: 3.36 + 04-03 23:29:31 | [249][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0085 mem: 3.36 + 04-03 23:29:33 | [249][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0434 ntime: 0081 mem: 3.36 + 04-03 23:29:35 | [249][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0147 ntime: 0078 mem: 3.36 + 04-03 23:29:37 | [249][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0125 ntime: 0082 mem: 3.36 + 04-03 23:29:39 | [249][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 23:29:41 | [249][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0078 mem: 3.36 + 04-03 23:29:44 | [249][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0079 mem: 3.36 + 04-03 23:29:45 | [249][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0078 mem: 3.36 + 04-03 23:29:47 | [249][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0076 mem: 3.36 + 04-03 23:29:50 | [249][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0184 ntime: 0089 mem: 3.36 + 04-03 23:29:51 | [249][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0079 mem: 3.36 + 04-03 23:29:55 | [249][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0172 ntime: 0074 mem: 3.36 + 04-03 23:29:57 | [249][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0091 mem: 3.36 + 04-03 23:29:59 | [249][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0135 ntime: 0088 mem: 3.36 + 04-03 23:30:00 | [249][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0168 ntime: 0084 mem: 3.36 + 04-03 23:30:02 | Time info >>>> elapsed: 126.66 mins remain: 379.98 mins + 04-03 23:30:02 | [250][000/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0281 ntime: 0080 mem: 3.36 + 04-03 23:30:04 | [250][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0076 mem: 3.36 + 04-03 23:30:07 | [250][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0072 mem: 3.36 + 04-03 23:30:09 | [250][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0222 ntime: 0086 mem: 3.36 + 04-03 23:30:11 | [250][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0191 ntime: 0077 mem: 3.36 + 04-03 23:30:13 | [250][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0093 ntime: 0076 mem: 3.36 + 04-03 23:30:15 | [250][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0212 ntime: 0072 mem: 3.36 + 04-03 23:30:17 | [250][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0083 mem: 3.36 + 04-03 23:30:19 | [250][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0088 mem: 3.36 + 04-03 23:30:21 | [250][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0214 ntime: 0087 mem: 3.36 + 04-03 23:30:23 | [250][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0189 ntime: 0075 mem: 3.36 + 04-03 23:30:26 | [250][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0258 ntime: 0078 mem: 3.36 + 04-03 23:30:28 | [250][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0259 ntime: 0079 mem: 3.36 + 04-03 23:30:31 | [250][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0076 mem: 3.36 + 04-03 23:30:33 | [250][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0361 ntime: 0087 mem: 3.36 + 04-03 23:30:36 | [250][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0106 ntime: 0077 mem: 3.36 + 04-03 23:30:39 | [250][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 23:30:41 | [250][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0240 ntime: 0084 mem: 3.36 + 04-03 23:30:43 | Time info >>>> elapsed: 127.35 mins remain: 380.01 mins + 04-03 23:30:43 | [251][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0088 mem: 3.36 + 04-03 23:30:46 | [251][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 23:30:49 | [251][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 23:30:51 | [251][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0083 mem: 3.36 + 04-03 23:30:54 | [251][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0209 ntime: 0086 mem: 3.36 + 04-03 23:30:56 | [251][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0125 ntime: 0077 mem: 3.36 + 04-03 23:30:59 | [251][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0080 mem: 3.36 + 04-03 23:31:01 | [251][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0074 mem: 3.36 + 04-03 23:31:04 | [251][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0356 ntime: 0087 mem: 3.36 + 04-03 23:31:06 | [251][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0081 mem: 3.36 + 04-03 23:31:09 | [251][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0206 ntime: 0077 mem: 3.36 + 04-03 23:31:11 | [251][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0078 mem: 3.36 + 04-03 23:31:14 | [251][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 23:31:16 | [251][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0222 ntime: 0086 mem: 3.36 + 04-03 23:31:18 | [251][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 23:31:21 | [251][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 23:31:23 | [251][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0142 ntime: 0087 mem: 3.36 + 04-03 23:31:26 | [251][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0232 ntime: 0087 mem: 3.36 + 04-03 23:31:27 | Time info >>>> elapsed: 128.08 mins remain: 380.18 mins + 04-03 23:31:28 | [252][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0078 mem: 3.36 + 04-03 23:31:30 | [252][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 23:31:32 | [252][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0168 ntime: 0080 mem: 3.36 + 04-03 23:31:34 | [252][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0070 mem: 3.36 + 04-03 23:31:36 | [252][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0311 ntime: 0084 mem: 3.36 + 04-03 23:31:38 | [252][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0143 ntime: 0076 mem: 3.36 + 04-03 23:31:41 | [252][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0081 mem: 3.36 + 04-03 23:31:43 | [252][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0289 ntime: 0079 mem: 3.36 + 04-03 23:31:47 | [252][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0058 mem: 3.36 + 04-03 23:31:50 | [252][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0076 mem: 3.36 + 04-03 23:31:52 | [252][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0075 mem: 3.36 + 04-03 23:31:54 | [252][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0075 mem: 3.36 + 04-03 23:31:56 | [252][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 23:31:59 | [252][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 23:32:02 | [252][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0210 ntime: 0074 mem: 3.36 + 04-03 23:32:04 | [252][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0178 ntime: 0082 mem: 3.36 + 04-03 23:32:06 | [252][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0171 ntime: 0080 mem: 3.36 + 04-03 23:32:08 | [252][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0116 ntime: 0083 mem: 3.36 + 04-03 23:32:10 | Time info >>>> elapsed: 128.80 mins remain: 380.29 mins + 04-03 23:32:10 | [253][000/179] predict_x0_loss: 0.010 glr: 5.0e-06 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 23:32:13 | [253][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0231 ntime: 0083 mem: 3.36 + 04-03 23:32:15 | [253][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0077 mem: 3.36 + 04-03 23:32:18 | [253][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0086 mem: 3.36 + 04-03 23:32:20 | [253][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0292 ntime: 0082 mem: 3.36 + 04-03 23:32:22 | [253][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 23:32:24 | [253][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0221 ntime: 0078 mem: 3.36 + 04-03 23:32:26 | [253][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0214 ntime: 0080 mem: 3.36 + 04-03 23:32:28 | [253][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0074 mem: 3.36 + 04-03 23:32:30 | [253][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0178 ntime: 0081 mem: 3.36 + 04-03 23:32:33 | [253][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 23:32:36 | [253][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 23:32:39 | [253][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0205 ntime: 0084 mem: 3.36 + 04-03 23:32:41 | [253][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0318 ntime: 0079 mem: 3.36 + 04-03 23:32:43 | [253][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0251 ntime: 0076 mem: 3.36 + 04-03 23:32:45 | [253][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 23:32:47 | [253][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0082 mem: 3.36 + 04-03 23:32:50 | [253][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0076 ntime: 0074 mem: 3.36 + 04-03 23:32:51 | Time info >>>> elapsed: 129.49 mins remain: 380.30 mins + 04-03 23:32:52 | [254][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0084 ntime: 0078 mem: 3.36 + 04-03 23:32:54 | [254][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0403 ntime: 0079 mem: 3.36 + 04-03 23:32:56 | [254][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0074 mem: 3.36 + 04-03 23:32:58 | [254][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0243 ntime: 0079 mem: 3.36 + 04-03 23:33:00 | [254][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0182 ntime: 0086 mem: 3.36 + 04-03 23:33:02 | [254][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0079 mem: 3.36 + 04-03 23:33:04 | [254][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0079 mem: 3.36 + 04-03 23:33:07 | [254][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0352 ntime: 0084 mem: 3.36 + 04-03 23:33:09 | [254][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0082 mem: 3.36 + 04-03 23:33:12 | [254][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0080 mem: 3.36 + 04-03 23:33:14 | [254][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0071 mem: 3.36 + 04-03 23:33:16 | [254][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0141 ntime: 0085 mem: 3.36 + 04-03 23:33:18 | [254][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0081 mem: 3.36 + 04-03 23:33:21 | [254][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0045 ntime: 0078 mem: 3.36 + 04-03 23:33:23 | [254][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0084 mem: 3.36 + 04-03 23:33:25 | [254][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 23:33:27 | [254][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0161 ntime: 0083 mem: 3.36 + 04-03 23:33:30 | [254][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0218 ntime: 0080 mem: 3.36 + 04-03 23:33:31 | Time info >>>> elapsed: 130.15 mins remain: 380.23 mins + 04-03 23:33:31 | [255][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0270 ntime: 0079 mem: 3.36 + 04-03 23:33:34 | [255][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0239 ntime: 0087 mem: 3.36 + 04-03 23:33:37 | [255][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0134 ntime: 0088 mem: 3.36 + 04-03 23:33:39 | [255][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0113 ntime: 0073 mem: 3.36 + 04-03 23:33:41 | [255][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0148 ntime: 0075 mem: 3.36 + 04-03 23:33:43 | [255][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 23:33:46 | [255][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0096 ntime: 0087 mem: 3.36 + 04-03 23:33:48 | [255][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0319 ntime: 0080 mem: 3.36 + 04-03 23:33:51 | [255][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0082 mem: 3.36 + 04-03 23:33:53 | [255][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0211 ntime: 0087 mem: 3.36 + 04-03 23:33:55 | [255][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0088 mem: 3.36 + 04-03 23:33:58 | [255][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0106 ntime: 0072 mem: 3.36 + 04-03 23:34:00 | [255][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0208 ntime: 0080 mem: 3.36 + 04-03 23:34:02 | [255][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0202 ntime: 0078 mem: 3.36 + 04-03 23:34:05 | [255][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0299 ntime: 0083 mem: 3.36 + 04-03 23:34:06 | [255][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0077 ntime: 0078 mem: 3.36 + 04-03 23:34:09 | [255][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0087 mem: 3.36 + 04-03 23:34:11 | [255][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0227 ntime: 0078 mem: 3.36 + 04-03 23:34:13 | Time info >>>> elapsed: 130.84 mins remain: 380.27 mins + 04-03 23:34:13 | [256][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 23:34:15 | [256][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 23:34:18 | [256][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0130 ntime: 0081 mem: 3.36 + 04-03 23:34:20 | [256][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0358 ntime: 0079 mem: 3.36 + 04-03 23:34:23 | [256][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0168 ntime: 0077 mem: 3.36 + 04-03 23:34:25 | [256][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0378 ntime: 0081 mem: 3.36 + 04-03 23:34:28 | [256][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0159 ntime: 0079 mem: 3.36 + 04-03 23:34:30 | [256][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 23:34:33 | [256][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0092 ntime: 0087 mem: 3.36 + 04-03 23:34:35 | [256][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0076 mem: 3.36 + 04-03 23:34:38 | [256][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0077 mem: 3.36 + 04-03 23:34:39 | [256][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 23:34:41 | [256][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0247 ntime: 0081 mem: 3.36 + 04-03 23:34:44 | [256][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0306 ntime: 0080 mem: 3.36 + 04-03 23:34:46 | [256][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0329 ntime: 0088 mem: 3.36 + 04-03 23:34:49 | [256][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0135 ntime: 0076 mem: 3.36 + 04-03 23:34:51 | [256][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 23:34:54 | [256][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0340 ntime: 0078 mem: 3.36 + 04-03 23:34:55 | Time info >>>> elapsed: 131.55 mins remain: 380.32 mins + 04-03 23:34:56 | [257][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0172 ntime: 0081 mem: 3.36 + 04-03 23:34:58 | [257][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0317 ntime: 0078 mem: 3.36 + 04-03 23:35:01 | [257][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0075 mem: 3.36 + 04-03 23:35:03 | [257][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0070 mem: 3.36 + 04-03 23:35:05 | [257][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0202 ntime: 0081 mem: 3.36 + 04-03 23:35:08 | [257][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 23:35:10 | [257][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0079 ntime: 0076 mem: 3.36 + 04-03 23:35:13 | [257][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0267 ntime: 0074 mem: 3.36 + 04-03 23:35:15 | [257][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0087 mem: 3.36 + 04-03 23:35:17 | [257][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0221 ntime: 0078 mem: 3.36 + 04-03 23:35:21 | [257][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0600 ntime: 0076 mem: 3.36 + 04-03 23:35:23 | [257][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0075 mem: 3.36 + 04-03 23:35:25 | [257][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0163 ntime: 0082 mem: 3.36 + 04-03 23:35:28 | [257][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0264 ntime: 0078 mem: 3.36 + 04-03 23:35:31 | [257][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0150 ntime: 0082 mem: 3.36 + 04-03 23:35:33 | [257][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0291 ntime: 0088 mem: 3.36 + 04-03 23:35:36 | [257][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0296 ntime: 0076 mem: 3.36 + 04-03 23:35:39 | [257][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0658 ntime: 0078 mem: 3.36 + 04-03 23:35:42 | Time info >>>> elapsed: 132.33 mins remain: 380.57 mins + 04-03 23:35:42 | [258][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0103 ntime: 0081 mem: 3.36 + 04-03 23:35:44 | [258][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0262 ntime: 0086 mem: 3.36 + 04-03 23:35:46 | [258][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0076 ntime: 0080 mem: 3.36 + 04-03 23:35:49 | [258][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0271 ntime: 0084 mem: 3.36 + 04-03 23:35:52 | [258][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0077 mem: 3.36 + 04-03 23:35:55 | [258][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0237 ntime: 0077 mem: 3.36 + 04-03 23:35:58 | [258][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0356 ntime: 0074 mem: 3.36 + 04-03 23:36:01 | [258][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 23:36:03 | [258][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0087 mem: 3.36 + 04-03 23:36:05 | [258][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0079 ntime: 0081 mem: 3.36 + 04-03 23:36:07 | [258][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0208 ntime: 0079 mem: 3.36 + 04-03 23:36:09 | [258][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0180 ntime: 0078 mem: 3.36 + 04-03 23:36:12 | [258][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0219 ntime: 0083 mem: 3.36 + 04-03 23:36:14 | [258][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0078 ntime: 0086 mem: 3.36 + 04-03 23:36:16 | [258][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0324 ntime: 0084 mem: 3.36 + 04-03 23:36:18 | [258][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0079 mem: 3.36 + 04-03 23:36:20 | [258][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0078 mem: 3.36 + 04-03 23:36:22 | [258][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0085 mem: 3.36 + 04-03 23:36:24 | Time info >>>> elapsed: 133.03 mins remain: 380.59 mins + 04-03 23:36:24 | [259][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 23:36:26 | [259][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 23:36:28 | [259][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0076 mem: 3.36 + 04-03 23:36:31 | [259][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0205 ntime: 0084 mem: 3.36 + 04-03 23:36:32 | [259][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0081 mem: 3.36 + 04-03 23:36:35 | [259][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0283 ntime: 0095 mem: 3.36 + 04-03 23:36:37 | [259][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0194 ntime: 0076 mem: 3.36 + 04-03 23:36:39 | [259][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0079 mem: 3.36 + 04-03 23:36:41 | [259][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 23:36:43 | [259][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0087 mem: 3.36 + 04-03 23:36:45 | [259][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0085 mem: 3.36 + 04-03 23:36:47 | [259][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0088 ntime: 0083 mem: 3.36 + 04-03 23:36:49 | [259][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0246 ntime: 0083 mem: 3.36 + 04-03 23:36:51 | [259][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0168 ntime: 0086 mem: 3.36 + 04-03 23:36:53 | [259][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 23:36:55 | [259][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0083 mem: 3.36 + 04-03 23:36:57 | [259][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0212 ntime: 0087 mem: 3.36 + 04-03 23:36:59 | [259][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0183 ntime: 0087 mem: 3.36 + 04-03 23:37:00 | Time info >>>> elapsed: 133.63 mins remain: 380.34 mins + 04-03 23:37:01 | [260][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0085 ntime: 0078 mem: 3.36 + 04-03 23:37:02 | [260][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0154 ntime: 0081 mem: 3.36 + 04-03 23:37:05 | [260][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0317 ntime: 0084 mem: 3.36 + 04-03 23:37:07 | [260][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0087 mem: 3.36 + 04-03 23:37:09 | [260][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0087 mem: 3.36 + 04-03 23:37:11 | [260][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0316 ntime: 0080 mem: 3.36 + 04-03 23:37:13 | [260][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0164 ntime: 0088 mem: 3.36 + 04-03 23:37:15 | [260][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0255 ntime: 0092 mem: 3.36 + 04-03 23:37:18 | [260][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0082 mem: 3.36 + 04-03 23:37:20 | [260][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0077 mem: 3.36 + 04-03 23:37:23 | [260][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0448 ntime: 0081 mem: 3.36 + 04-03 23:37:25 | [260][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0653 ntime: 0075 mem: 3.36 + 04-03 23:37:27 | [260][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0176 ntime: 0057 mem: 3.36 + 04-03 23:37:28 | [260][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0071 mem: 3.36 + 04-03 23:37:31 | [260][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0165 ntime: 0078 mem: 3.36 + 04-03 23:37:33 | [260][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0501 ntime: 0076 mem: 3.36 + 04-03 23:37:36 | [260][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0078 mem: 3.36 + 04-03 23:37:38 | [260][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 23:37:39 | Time info >>>> elapsed: 134.29 mins remain: 380.22 mins + 04-03 23:37:40 | [261][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0242 ntime: 0082 mem: 3.36 + 04-03 23:37:42 | [261][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0272 ntime: 0080 mem: 3.36 + 04-03 23:37:44 | [261][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0083 mem: 3.36 + 04-03 23:37:46 | [261][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0285 ntime: 0083 mem: 3.36 + 04-03 23:37:48 | [261][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0080 mem: 3.36 + 04-03 23:37:50 | [261][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 23:37:52 | [261][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0119 ntime: 0075 mem: 3.36 + 04-03 23:37:54 | [261][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 23:37:56 | [261][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0075 mem: 3.36 + 04-03 23:37:58 | [261][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0083 mem: 3.36 + 04-03 23:38:00 | [261][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0073 mem: 3.36 + 04-03 23:38:02 | [261][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0132 ntime: 0081 mem: 3.36 + 04-03 23:38:04 | [261][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0084 ntime: 0078 mem: 3.36 + 04-03 23:38:06 | [261][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0281 ntime: 0075 mem: 3.36 + 04-03 23:38:09 | [261][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 23:38:11 | [261][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0148 ntime: 0088 mem: 3.36 + 04-03 23:38:13 | [261][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0190 ntime: 0079 mem: 3.36 + 04-03 23:38:15 | [261][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0116 ntime: 0075 mem: 3.36 + 04-03 23:38:16 | Time info >>>> elapsed: 134.90 mins remain: 379.99 mins + 04-03 23:38:17 | [262][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0072 mem: 3.36 + 04-03 23:38:19 | [262][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0118 ntime: 0084 mem: 3.36 + 04-03 23:38:21 | [262][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 23:38:23 | [262][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0075 mem: 3.36 + 04-03 23:38:25 | [262][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0079 mem: 3.36 + 04-03 23:38:27 | [262][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0157 ntime: 0082 mem: 3.36 + 04-03 23:38:29 | [262][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0088 mem: 3.36 + 04-03 23:38:31 | [262][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0148 ntime: 0082 mem: 3.36 + 04-03 23:38:33 | [262][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0075 mem: 3.36 + 04-03 23:38:35 | [262][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0211 ntime: 0078 mem: 3.36 + 04-03 23:38:37 | [262][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0081 mem: 3.36 + 04-03 23:38:39 | [262][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 23:38:41 | [262][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0076 ntime: 0075 mem: 3.36 + 04-03 23:38:43 | [262][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0161 ntime: 0083 mem: 3.36 + 04-03 23:38:45 | [262][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0076 ntime: 0085 mem: 3.36 + 04-03 23:38:46 | [262][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0080 mem: 3.36 + 04-03 23:38:48 | [262][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0076 mem: 3.36 + 04-03 23:38:51 | [262][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0081 mem: 3.36 + 04-03 23:38:52 | Time info >>>> elapsed: 135.49 mins remain: 379.69 mins + 04-03 23:38:52 | [263][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0067 ntime: 0082 mem: 3.36 + 04-03 23:38:54 | [263][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0186 ntime: 0088 mem: 3.36 + 04-03 23:38:56 | [263][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 23:38:59 | [263][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0540 ntime: 0083 mem: 3.36 + 04-03 23:39:01 | [263][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0258 ntime: 0078 mem: 3.36 + 04-03 23:39:03 | [263][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0078 mem: 3.36 + 04-03 23:39:05 | [263][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 23:39:07 | [263][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0086 mem: 3.36 + 04-03 23:39:09 | [263][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0165 ntime: 0086 mem: 3.36 + 04-03 23:39:11 | [263][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0083 mem: 3.36 + 04-03 23:39:14 | [263][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0268 ntime: 0090 mem: 3.36 + 04-03 23:39:16 | [263][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0077 mem: 3.36 + 04-03 23:39:18 | [263][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0073 mem: 3.36 + 04-03 23:39:22 | [263][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0078 mem: 3.36 + 04-03 23:39:24 | [263][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0466 ntime: 0082 mem: 3.36 + 04-03 23:39:28 | [263][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0143 ntime: 0082 mem: 3.36 + 04-03 23:39:30 | [263][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 23:39:32 | [263][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 23:39:34 | Time info >>>> elapsed: 136.20 mins remain: 379.71 mins + 04-03 23:39:35 | [264][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0262 ntime: 0086 mem: 3.36 + 04-03 23:39:37 | [264][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0077 mem: 3.36 + 04-03 23:39:39 | [264][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0078 mem: 3.36 + 04-03 23:39:41 | [264][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 23:39:44 | [264][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0080 mem: 3.36 + 04-03 23:39:46 | [264][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0076 mem: 3.36 + 04-03 23:39:48 | [264][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0264 ntime: 0073 mem: 3.36 + 04-03 23:39:50 | [264][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0191 ntime: 0081 mem: 3.36 + 04-03 23:39:52 | [264][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0173 ntime: 0072 mem: 3.36 + 04-03 23:39:54 | [264][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0082 mem: 3.36 + 04-03 23:39:57 | [264][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0263 ntime: 0079 mem: 3.36 + 04-03 23:39:59 | [264][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0198 ntime: 0081 mem: 3.36 + 04-03 23:40:01 | [264][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0266 ntime: 0081 mem: 3.36 + 04-03 23:40:04 | [264][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0087 mem: 3.36 + 04-03 23:40:06 | [264][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0202 ntime: 0079 mem: 3.36 + 04-03 23:40:08 | [264][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 23:40:10 | [264][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0083 mem: 3.36 + 04-03 23:40:12 | [264][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 23:40:15 | Time info >>>> elapsed: 136.87 mins remain: 379.63 mins + 04-03 23:40:15 | [265][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0235 ntime: 0077 mem: 3.36 + 04-03 23:40:17 | [265][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 23:40:20 | [265][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0043 ntime: 0058 mem: 3.36 + 04-03 23:40:21 | [265][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0044 ntime: 0070 mem: 3.36 + 04-03 23:40:24 | [265][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0168 ntime: 0074 mem: 3.36 + 04-03 23:40:26 | [265][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0373 ntime: 0079 mem: 3.36 + 04-03 23:40:28 | [265][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0207 ntime: 0073 mem: 3.36 + 04-03 23:40:31 | [265][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0217 ntime: 0076 mem: 3.36 + 04-03 23:40:33 | [265][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0077 mem: 3.36 + 04-03 23:40:35 | [265][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 23:40:37 | [265][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0084 mem: 3.36 + 04-03 23:40:39 | [265][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0078 mem: 3.36 + 04-03 23:40:41 | [265][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0083 mem: 3.36 + 04-03 23:40:43 | [265][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 23:40:45 | [265][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0248 ntime: 0081 mem: 3.36 + 04-03 23:40:48 | [265][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0167 ntime: 0081 mem: 3.36 + 04-03 23:40:51 | [265][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0433 ntime: 0082 mem: 3.36 + 04-03 23:40:53 | [265][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 23:40:57 | Time info >>>> elapsed: 137.57 mins remain: 379.61 mins + 04-03 23:40:57 | [266][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0077 mem: 3.36 + 04-03 23:41:00 | [266][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0073 mem: 3.36 + 04-03 23:41:02 | [266][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0164 ntime: 0083 mem: 3.36 + 04-03 23:41:05 | [266][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0272 ntime: 0083 mem: 3.36 + 04-03 23:41:07 | [266][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0315 ntime: 0088 mem: 3.36 + 04-03 23:41:09 | [266][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 23:41:12 | [266][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0082 mem: 3.36 + 04-03 23:41:14 | [266][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0147 ntime: 0079 mem: 3.36 + 04-03 23:41:17 | [266][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0081 mem: 3.36 + 04-03 23:41:20 | [266][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0250 ntime: 0084 mem: 3.36 + 04-03 23:41:22 | [266][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0074 mem: 3.36 + 04-03 23:41:24 | [266][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0236 ntime: 0084 mem: 3.36 + 04-03 23:41:26 | [266][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0074 mem: 3.36 + 04-03 23:41:29 | [266][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0087 mem: 3.36 + 04-03 23:41:31 | [266][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0080 mem: 3.36 + 04-03 23:41:34 | [266][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0276 ntime: 0088 mem: 3.36 + 04-03 23:41:36 | [266][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0085 mem: 3.36 + 04-03 23:41:38 | [266][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0076 mem: 3.36 + 04-03 23:41:40 | Time info >>>> elapsed: 138.29 mins remain: 379.65 mins + 04-03 23:41:40 | [267][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0385 ntime: 0078 mem: 3.36 + 04-03 23:41:43 | [267][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0184 ntime: 0079 mem: 3.36 + 04-03 23:41:45 | [267][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0079 mem: 3.36 + 04-03 23:41:47 | [267][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0085 mem: 3.36 + 04-03 23:41:49 | [267][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0267 ntime: 0079 mem: 3.36 + 04-03 23:41:52 | [267][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0159 ntime: 0085 mem: 3.36 + 04-03 23:41:54 | [267][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0920 ntime: 0077 mem: 3.36 + 04-03 23:41:56 | [267][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0083 mem: 3.36 + 04-03 23:41:59 | [267][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0093 ntime: 0098 mem: 3.36 + 04-03 23:42:01 | [267][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0259 ntime: 0083 mem: 3.36 + 04-03 23:42:04 | [267][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0534 ntime: 0083 mem: 3.36 + 04-03 23:42:05 | [267][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0085 mem: 3.36 + 04-03 23:42:08 | [267][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0386 ntime: 0071 mem: 3.36 + 04-03 23:42:12 | [267][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0157 ntime: 0080 mem: 3.36 + 04-03 23:42:14 | [267][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0203 ntime: 0080 mem: 3.36 + 04-03 23:42:16 | [267][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0540 ntime: 0078 mem: 3.36 + 04-03 23:42:18 | [267][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0206 ntime: 0084 mem: 3.36 + 04-03 23:42:21 | [267][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0176 ntime: 0086 mem: 3.36 + 04-03 23:42:23 | Time info >>>> elapsed: 139.00 mins remain: 379.66 mins + 04-03 23:42:23 | [268][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0135 ntime: 0077 mem: 3.36 + 04-03 23:42:25 | [268][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0114 ntime: 0079 mem: 3.36 + 04-03 23:42:27 | [268][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0401 ntime: 0081 mem: 3.36 + 04-03 23:42:29 | [268][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0073 ntime: 0082 mem: 3.36 + 04-03 23:42:31 | [268][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0074 mem: 3.36 + 04-03 23:42:34 | [268][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0156 ntime: 0081 mem: 3.36 + 04-03 23:42:36 | [268][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0131 ntime: 0077 mem: 3.36 + 04-03 23:42:39 | [268][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0112 ntime: 0077 mem: 3.36 + 04-03 23:42:41 | [268][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0077 mem: 3.36 + 04-03 23:42:43 | [268][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0089 ntime: 0084 mem: 3.36 + 04-03 23:42:45 | [268][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 23:42:47 | [268][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0083 mem: 3.36 + 04-03 23:42:49 | [268][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0224 ntime: 0084 mem: 3.36 + 04-03 23:42:51 | [268][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0169 ntime: 0083 mem: 3.36 + 04-03 23:42:54 | [268][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0076 mem: 3.36 + 04-03 23:42:57 | [268][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 23:42:58 | [268][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0078 mem: 3.36 + 04-03 23:43:00 | [268][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0086 mem: 3.36 + 04-03 23:43:03 | Time info >>>> elapsed: 139.67 mins remain: 379.56 mins + 04-03 23:43:03 | [269][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 23:43:06 | [269][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1306 ntime: 0081 mem: 3.36 + 04-03 23:43:08 | [269][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0446 ntime: 0077 mem: 3.36 + 04-03 23:43:11 | [269][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0057 mem: 3.36 + 04-03 23:43:13 | [269][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0322 ntime: 0078 mem: 3.36 + 04-03 23:43:15 | [269][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0219 ntime: 0072 mem: 3.36 + 04-03 23:43:17 | [269][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0150 ntime: 0076 mem: 3.36 + 04-03 23:43:19 | [269][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0082 mem: 3.36 + 04-03 23:43:20 | [269][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 23:43:23 | [269][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0283 ntime: 0074 mem: 3.36 + 04-03 23:43:25 | [269][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0175 ntime: 0086 mem: 3.36 + 04-03 23:43:27 | [269][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0137 ntime: 0082 mem: 3.36 + 04-03 23:43:30 | [269][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 23:43:32 | [269][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0085 mem: 3.36 + 04-03 23:43:34 | [269][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 23:43:36 | [269][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 23:43:39 | [269][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0265 ntime: 0085 mem: 3.36 + 04-03 23:43:41 | [269][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0308 ntime: 0084 mem: 3.36 + 04-03 23:43:43 | Time info >>>> elapsed: 140.34 mins remain: 379.43 mins + 04-03 23:43:43 | [270][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0156 ntime: 0080 mem: 3.36 + 04-03 23:43:45 | [270][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-03 23:43:48 | [270][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0457 ntime: 0088 mem: 3.36 + 04-03 23:43:50 | [270][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0134 ntime: 0080 mem: 3.36 + 04-03 23:43:53 | [270][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0086 mem: 3.36 + 04-03 23:43:55 | [270][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0077 mem: 3.36 + 04-03 23:43:57 | [270][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0087 mem: 3.36 + 04-03 23:43:59 | [270][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0363 ntime: 0081 mem: 3.36 + 04-03 23:44:02 | [270][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0078 mem: 3.36 + 04-03 23:44:04 | [270][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0075 mem: 3.36 + 04-03 23:44:07 | [270][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0084 mem: 3.36 + 04-03 23:44:09 | [270][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0276 ntime: 0082 mem: 3.36 + 04-03 23:44:12 | [270][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0083 ntime: 0075 mem: 3.36 + 04-03 23:44:14 | [270][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0200 ntime: 0088 mem: 3.36 + 04-03 23:44:17 | [270][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0073 mem: 3.36 + 04-03 23:44:20 | [270][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0200 ntime: 0079 mem: 3.36 + 04-03 23:44:23 | [270][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0512 ntime: 0080 mem: 3.36 + 04-03 23:44:25 | [270][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0238 ntime: 0086 mem: 3.36 + 04-03 23:44:27 | Time info >>>> elapsed: 141.09 mins remain: 379.52 mins + 04-03 23:44:28 | [271][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 23:44:31 | [271][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0314 ntime: 0084 mem: 3.36 + 04-03 23:44:33 | [271][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0326 ntime: 0081 mem: 3.36 + 04-03 23:44:35 | [271][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0323 ntime: 0077 mem: 3.36 + 04-03 23:44:37 | [271][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0160 ntime: 0077 mem: 3.36 + 04-03 23:44:39 | [271][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0336 ntime: 0085 mem: 3.36 + 04-03 23:44:42 | [271][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 23:44:44 | [271][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0270 ntime: 0090 mem: 3.36 + 04-03 23:44:46 | [271][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0122 ntime: 0081 mem: 3.36 + 04-03 23:44:50 | [271][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0088 mem: 3.36 + 04-03 23:44:51 | [271][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0092 mem: 3.36 + 04-03 23:44:53 | [271][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 23:44:55 | [271][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0385 ntime: 0086 mem: 3.36 + 04-03 23:44:57 | [271][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0274 ntime: 0080 mem: 3.36 + 04-03 23:45:00 | [271][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0883 ntime: 0084 mem: 3.36 + 04-03 23:45:02 | [271][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0090 mem: 3.36 + 04-03 23:45:05 | [271][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0202 ntime: 0082 mem: 3.36 + 04-03 23:45:07 | [271][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0075 mem: 3.36 + 04-03 23:45:09 | Time info >>>> elapsed: 141.78 mins remain: 379.46 mins + 04-03 23:45:09 | [272][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0174 ntime: 0089 mem: 3.36 + 04-03 23:45:11 | [272][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0240 ntime: 0083 mem: 3.36 + 04-03 23:45:14 | [272][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0194 ntime: 0091 mem: 3.36 + 04-03 23:45:16 | [272][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 23:45:18 | [272][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0184 ntime: 0081 mem: 3.36 + 04-03 23:45:20 | [272][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0174 ntime: 0080 mem: 3.36 + 04-03 23:45:22 | [272][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0316 ntime: 0084 mem: 3.36 + 04-03 23:45:24 | [272][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0359 ntime: 0085 mem: 3.36 + 04-03 23:45:27 | [272][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 23:45:29 | [272][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 23:45:31 | [272][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0080 mem: 3.36 + 04-03 23:45:34 | [272][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0076 ntime: 0080 mem: 3.36 + 04-03 23:45:35 | [272][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 23:45:37 | [272][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0131 ntime: 0078 mem: 3.36 + 04-03 23:45:40 | [272][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0330 ntime: 0080 mem: 3.36 + 04-03 23:45:42 | [272][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 23:45:44 | [272][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0110 ntime: 0081 mem: 3.36 + 04-03 23:45:46 | [272][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0077 mem: 3.36 + 04-03 23:45:48 | Time info >>>> elapsed: 142.43 mins remain: 379.29 mins + 04-03 23:45:48 | [273][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0079 mem: 3.36 + 04-03 23:45:51 | [273][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 23:45:53 | [273][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0176 ntime: 0086 mem: 3.36 + 04-03 23:45:55 | [273][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0078 mem: 3.36 + 04-03 23:45:57 | [273][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 23:46:00 | [273][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0086 mem: 3.36 + 04-03 23:46:02 | [273][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0171 ntime: 0079 mem: 3.36 + 04-03 23:46:04 | [273][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0085 mem: 3.36 + 04-03 23:46:06 | [273][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0084 mem: 3.36 + 04-03 23:46:08 | [273][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0144 ntime: 0075 mem: 3.36 + 04-03 23:46:10 | [273][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0074 mem: 3.36 + 04-03 23:46:12 | [273][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0079 mem: 3.36 + 04-03 23:46:14 | [273][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0085 mem: 3.36 + 04-03 23:46:16 | [273][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0273 ntime: 0079 mem: 3.36 + 04-03 23:46:18 | [273][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0082 mem: 3.36 + 04-03 23:46:20 | [273][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0093 ntime: 0079 mem: 3.36 + 04-03 23:46:22 | [273][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0083 mem: 3.36 + 04-03 23:46:25 | [273][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0364 ntime: 0078 mem: 3.36 + 04-03 23:46:26 | Time info >>>> elapsed: 143.07 mins remain: 379.08 mins + 04-03 23:46:27 | [274][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0259 ntime: 0077 mem: 3.36 + 04-03 23:46:29 | [274][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0080 mem: 3.36 + 04-03 23:46:31 | [274][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0155 ntime: 0078 mem: 3.36 + 04-03 23:46:33 | [274][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0193 ntime: 0085 mem: 3.36 + 04-03 23:46:36 | [274][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0362 ntime: 0084 mem: 3.36 + 04-03 23:46:39 | [274][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0810 ntime: 0083 mem: 3.36 + 04-03 23:46:41 | [274][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0077 ntime: 0077 mem: 3.36 + 04-03 23:46:43 | [274][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0204 ntime: 0088 mem: 3.36 + 04-03 23:46:46 | [274][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0297 ntime: 0087 mem: 3.36 + 04-03 23:46:49 | [274][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0110 ntime: 0077 mem: 3.36 + 04-03 23:46:51 | [274][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0086 mem: 3.36 + 04-03 23:46:54 | [274][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0113 ntime: 0080 mem: 3.36 + 04-03 23:46:56 | [274][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0095 ntime: 0089 mem: 3.36 + 04-03 23:46:58 | [274][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0079 mem: 3.36 + 04-03 23:47:01 | [274][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0076 mem: 3.36 + 04-03 23:47:03 | [274][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0430 ntime: 0087 mem: 3.36 + 04-03 23:47:06 | [274][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0161 ntime: 0087 mem: 3.36 + 04-03 23:47:08 | [274][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0174 ntime: 0075 mem: 3.36 + 04-03 23:47:10 | Time info >>>> elapsed: 143.80 mins remain: 379.11 mins + 04-03 23:47:11 | [275][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0081 ntime: 0080 mem: 3.36 + 04-03 23:47:13 | [275][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 23:47:15 | [275][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0087 mem: 3.36 + 04-03 23:47:17 | [275][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0098 ntime: 0082 mem: 3.36 + 04-03 23:47:20 | [275][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0082 mem: 3.36 + 04-03 23:47:22 | [275][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0171 ntime: 0080 mem: 3.36 + 04-03 23:47:24 | [275][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0076 ntime: 0083 mem: 3.36 + 04-03 23:47:27 | [275][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0299 ntime: 0073 mem: 3.36 + 04-03 23:47:29 | [275][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 23:47:31 | [275][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0229 ntime: 0086 mem: 3.36 + 04-03 23:47:33 | [275][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0082 mem: 3.36 + 04-03 23:47:35 | [275][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0077 mem: 3.36 + 04-03 23:47:38 | [275][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0167 ntime: 0079 mem: 3.36 + 04-03 23:47:40 | [275][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0198 ntime: 0085 mem: 3.36 + 04-03 23:47:42 | [275][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0159 ntime: 0077 mem: 3.36 + 04-03 23:47:44 | [275][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0073 mem: 3.36 + 04-03 23:47:46 | [275][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0200 ntime: 0089 mem: 3.36 + 04-03 23:47:48 | [275][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 23:47:50 | Time info >>>> elapsed: 144.46 mins remain: 378.93 mins + 04-03 23:47:50 | [276][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0086 mem: 3.36 + 04-03 23:47:53 | [276][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0512 ntime: 0079 mem: 3.36 + 04-03 23:47:55 | [276][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0081 mem: 3.36 + 04-03 23:47:57 | [276][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0084 mem: 3.36 + 04-03 23:47:59 | [276][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0080 mem: 3.36 + 04-03 23:48:01 | [276][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0325 ntime: 0086 mem: 3.36 + 04-03 23:48:03 | [276][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0175 ntime: 0079 mem: 3.36 + 04-03 23:48:05 | [276][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0080 mem: 3.36 + 04-03 23:48:07 | [276][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0178 ntime: 0087 mem: 3.36 + 04-03 23:48:09 | [276][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 23:48:11 | [276][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0184 ntime: 0083 mem: 3.36 + 04-03 23:48:14 | [276][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0080 mem: 3.36 + 04-03 23:48:16 | [276][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0086 ntime: 0080 mem: 3.36 + 04-03 23:48:18 | [276][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0077 ntime: 0082 mem: 3.36 + 04-03 23:48:20 | [276][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0196 ntime: 0083 mem: 3.36 + 04-03 23:48:22 | [276][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0080 mem: 3.36 + 04-03 23:48:24 | [276][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 23:48:26 | [276][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0330 ntime: 0079 mem: 3.36 + 04-03 23:48:28 | Time info >>>> elapsed: 145.10 mins remain: 378.73 mins + 04-03 23:48:29 | [277][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0166 ntime: 0077 mem: 3.36 + 04-03 23:48:31 | [277][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0170 ntime: 0075 mem: 3.36 + 04-03 23:48:34 | [277][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0361 ntime: 0081 mem: 3.36 + 04-03 23:48:36 | [277][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0224 ntime: 0078 mem: 3.36 + 04-03 23:48:38 | [277][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0074 mem: 3.36 + 04-03 23:48:40 | [277][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0084 ntime: 0082 mem: 3.36 + 04-03 23:48:43 | [277][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 23:48:46 | [277][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0082 mem: 3.36 + 04-03 23:48:49 | [277][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0210 ntime: 0085 mem: 3.36 + 04-03 23:48:52 | [277][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0085 mem: 3.36 + 04-03 23:48:54 | [277][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 23:48:56 | [277][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 23:48:59 | [277][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0224 ntime: 0080 mem: 3.36 + 04-03 23:49:01 | [277][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0288 ntime: 0078 mem: 3.36 + 04-03 23:49:03 | [277][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0081 mem: 3.36 + 04-03 23:49:05 | [277][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0187 ntime: 0079 mem: 3.36 + 04-03 23:49:07 | [277][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0306 ntime: 0084 mem: 3.36 + 04-03 23:49:09 | [277][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 23:49:11 | Time info >>>> elapsed: 145.80 mins remain: 378.67 mins + 04-03 23:49:11 | [278][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 23:49:13 | [278][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0083 mem: 3.36 + 04-03 23:49:15 | [278][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0339 ntime: 0088 mem: 3.36 + 04-03 23:49:18 | [278][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0160 ntime: 0077 mem: 3.36 + 04-03 23:49:20 | [278][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0220 ntime: 0085 mem: 3.36 + 04-03 23:49:22 | [278][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0179 ntime: 0084 mem: 3.36 + 04-03 23:49:25 | [278][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0072 mem: 3.36 + 04-03 23:49:28 | [278][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0499 ntime: 0084 mem: 3.36 + 04-03 23:49:30 | [278][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0073 mem: 3.36 + 04-03 23:49:33 | [278][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0264 ntime: 0082 mem: 3.36 + 04-03 23:49:35 | [278][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0082 mem: 3.36 + 04-03 23:49:38 | [278][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0263 ntime: 0076 mem: 3.36 + 04-03 23:49:40 | [278][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0076 mem: 3.36 + 04-03 23:49:42 | [278][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 23:49:44 | [278][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0080 mem: 3.36 + 04-03 23:49:47 | [278][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0176 ntime: 0084 mem: 3.36 + 04-03 23:49:49 | [278][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0105 ntime: 0076 mem: 3.36 + 04-03 23:49:51 | [278][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0080 mem: 3.36 + 04-03 23:49:53 | Time info >>>> elapsed: 146.52 mins remain: 378.64 mins + 04-03 23:49:54 | [279][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0053 ntime: 0084 mem: 3.36 + 04-03 23:49:56 | [279][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0168 ntime: 0084 mem: 3.36 + 04-03 23:49:58 | [279][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0403 ntime: 0079 mem: 3.36 + 04-03 23:50:00 | [279][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 23:50:02 | [279][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0272 ntime: 0078 mem: 3.36 + 04-03 23:50:05 | [279][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0164 ntime: 0076 mem: 3.36 + 04-03 23:50:07 | [279][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0177 ntime: 0081 mem: 3.36 + 04-03 23:50:09 | [279][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0214 ntime: 0078 mem: 3.36 + 04-03 23:50:12 | [279][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0076 mem: 3.36 + 04-03 23:50:14 | [279][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0111 ntime: 0074 mem: 3.36 + 04-03 23:50:16 | [279][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0199 ntime: 0081 mem: 3.36 + 04-03 23:50:18 | [279][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0118 ntime: 0079 mem: 3.36 + 04-03 23:50:20 | [279][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0227 ntime: 0073 mem: 3.36 + 04-03 23:50:23 | [279][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0076 mem: 3.36 + 04-03 23:50:26 | [279][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0078 ntime: 0080 mem: 3.36 + 04-03 23:50:28 | [279][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0087 mem: 3.36 + 04-03 23:50:30 | [279][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0085 mem: 3.36 + 04-03 23:50:32 | [279][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0288 ntime: 0088 mem: 3.36 + 04-03 23:50:35 | Time info >>>> elapsed: 147.20 mins remain: 378.52 mins + 04-03 23:50:35 | [280][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0080 mem: 3.36 + 04-03 23:50:39 | [280][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0984 ntime: 0084 mem: 3.36 + 04-03 23:50:41 | [280][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0339 ntime: 0079 mem: 3.36 + 04-03 23:50:43 | [280][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0076 ntime: 0082 mem: 3.36 + 04-03 23:50:45 | [280][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0105 ntime: 0079 mem: 3.36 + 04-03 23:50:48 | [280][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0080 mem: 3.36 + 04-03 23:50:50 | [280][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0079 mem: 3.36 + 04-03 23:50:52 | [280][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0077 mem: 3.36 + 04-03 23:50:54 | [280][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0084 mem: 3.36 + 04-03 23:50:57 | [280][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0211 ntime: 0080 mem: 3.36 + 04-03 23:50:59 | [280][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0070 ntime: 0080 mem: 3.36 + 04-03 23:51:01 | [280][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0322 ntime: 0078 mem: 3.36 + 04-03 23:51:04 | [280][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0077 mem: 3.36 + 04-03 23:51:07 | [280][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0237 ntime: 0078 mem: 3.36 + 04-03 23:51:09 | [280][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0197 ntime: 0086 mem: 3.36 + 04-03 23:51:11 | [280][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0083 mem: 3.36 + 04-03 23:51:14 | [280][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0309 ntime: 0086 mem: 3.36 + 04-03 23:51:16 | [280][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0209 ntime: 0079 mem: 3.36 + 04-03 23:51:18 | Time info >>>> elapsed: 147.93 mins remain: 378.51 mins + 04-03 23:51:18 | [281][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0156 ntime: 0078 mem: 3.36 + 04-03 23:51:21 | [281][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0538 ntime: 0081 mem: 3.36 + 04-03 23:51:23 | [281][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0080 mem: 3.36 + 04-03 23:51:25 | [281][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0090 mem: 3.36 + 04-03 23:51:27 | [281][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0074 mem: 3.36 + 04-03 23:51:30 | [281][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0527 ntime: 0082 mem: 3.36 + 04-03 23:51:32 | [281][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0077 mem: 3.36 + 04-03 23:51:34 | [281][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0081 mem: 3.36 + 04-03 23:51:36 | [281][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 23:51:38 | [281][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0080 mem: 3.36 + 04-03 23:51:40 | [281][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0200 ntime: 0080 mem: 3.36 + 04-03 23:51:42 | [281][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0229 ntime: 0081 mem: 3.36 + 04-03 23:51:44 | [281][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0382 ntime: 0083 mem: 3.36 + 04-03 23:51:46 | [281][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0164 ntime: 0070 mem: 3.36 + 04-03 23:51:48 | [281][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0185 ntime: 0089 mem: 3.36 + 04-03 23:51:50 | [281][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0076 mem: 3.36 + 04-03 23:51:53 | [281][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0081 mem: 3.36 + 04-03 23:51:55 | [281][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0366 ntime: 0079 mem: 3.36 + 04-03 23:51:57 | Time info >>>> elapsed: 148.57 mins remain: 378.27 mins + 04-03 23:51:57 | [282][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 23:51:59 | [282][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0075 mem: 3.36 + 04-03 23:52:01 | [282][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0086 mem: 3.36 + 04-03 23:52:03 | [282][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0159 ntime: 0076 mem: 3.36 + 04-03 23:52:05 | [282][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0253 ntime: 0081 mem: 3.36 + 04-03 23:52:07 | [282][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0074 mem: 3.36 + 04-03 23:52:10 | [282][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0079 mem: 3.36 + 04-03 23:52:12 | [282][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0187 ntime: 0079 mem: 3.36 + 04-03 23:52:14 | [282][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0288 ntime: 0086 mem: 3.36 + 04-03 23:52:16 | [282][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0268 ntime: 0078 mem: 3.36 + 04-03 23:52:18 | [282][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0081 mem: 3.36 + 04-03 23:52:20 | [282][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 23:52:22 | [282][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0090 ntime: 0077 mem: 3.36 + 04-03 23:52:24 | [282][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0462 ntime: 0081 mem: 3.36 + 04-03 23:52:26 | [282][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0133 ntime: 0082 mem: 3.36 + 04-03 23:52:29 | [282][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0101 ntime: 0082 mem: 3.36 + 04-03 23:52:31 | [282][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0073 mem: 3.36 + 04-03 23:52:33 | [282][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 23:52:36 | Time info >>>> elapsed: 149.22 mins remain: 378.07 mins + 04-03 23:52:36 | [283][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0044 ntime: 0078 mem: 3.36 + 04-03 23:52:38 | [283][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0080 ntime: 0082 mem: 3.36 + 04-03 23:52:41 | [283][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0451 ntime: 0088 mem: 3.36 + 04-03 23:52:43 | [283][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0082 mem: 3.36 + 04-03 23:52:45 | [283][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0083 mem: 3.36 + 04-03 23:52:47 | [283][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0167 ntime: 0080 mem: 3.36 + 04-03 23:52:50 | [283][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0114 ntime: 0075 mem: 3.36 + 04-03 23:52:52 | [283][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0163 ntime: 0086 mem: 3.36 + 04-03 23:52:54 | [283][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0174 ntime: 0086 mem: 3.36 + 04-03 23:52:57 | [283][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0043 ntime: 0081 mem: 3.36 + 04-03 23:52:59 | [283][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0181 ntime: 0083 mem: 3.36 + 04-03 23:53:01 | [283][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 23:53:03 | [283][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0248 ntime: 0079 mem: 3.36 + 04-03 23:53:06 | [283][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0077 mem: 3.36 + 04-03 23:53:08 | [283][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0147 ntime: 0077 mem: 3.36 + 04-03 23:53:09 | [283][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 23:53:12 | [283][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0161 ntime: 0080 mem: 3.36 + 04-03 23:53:15 | [283][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0968 ntime: 0080 mem: 3.36 + 04-03 23:53:17 | Time info >>>> elapsed: 149.90 mins remain: 377.93 mins + 04-03 23:53:17 | [284][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0044 ntime: 0078 mem: 3.36 + 04-03 23:53:19 | [284][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0474 ntime: 0089 mem: 3.36 + 04-03 23:53:21 | [284][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0081 mem: 3.36 + 04-03 23:53:23 | [284][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0079 mem: 3.36 + 04-03 23:53:26 | [284][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0085 mem: 3.36 + 04-03 23:53:28 | [284][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0220 ntime: 0086 mem: 3.36 + 04-03 23:53:30 | [284][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0461 ntime: 0084 mem: 3.36 + 04-03 23:53:33 | [284][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0088 ntime: 0072 mem: 3.36 + 04-03 23:53:36 | [284][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0713 ntime: 0078 mem: 3.36 + 04-03 23:53:38 | [284][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 23:53:40 | [284][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0077 mem: 3.36 + 04-03 23:53:42 | [284][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0081 ntime: 0079 mem: 3.36 + 04-03 23:53:44 | [284][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 23:53:46 | [284][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0075 mem: 3.36 + 04-03 23:53:48 | [284][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0098 ntime: 0084 mem: 3.36 + 04-03 23:53:50 | [284][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0084 mem: 3.36 + 04-03 23:53:53 | [284][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0312 ntime: 0079 mem: 3.36 + 04-03 23:53:56 | [284][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 23:53:58 | Time info >>>> elapsed: 150.59 mins remain: 377.80 mins + 04-03 23:53:58 | [285][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 23:54:00 | [285][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0280 ntime: 0083 mem: 3.36 + 04-03 23:54:02 | [285][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0253 ntime: 0080 mem: 3.36 + 04-03 23:54:04 | [285][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0213 ntime: 0076 mem: 3.36 + 04-03 23:54:07 | [285][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0079 mem: 3.36 + 04-03 23:54:09 | [285][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0076 ntime: 0071 mem: 3.36 + 04-03 23:54:11 | [285][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 23:54:13 | [285][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0070 ntime: 0083 mem: 3.36 + 04-03 23:54:15 | [285][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0081 mem: 3.36 + 04-03 23:54:18 | [285][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0139 ntime: 0086 mem: 3.36 + 04-03 23:54:20 | [285][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0085 mem: 3.36 + 04-03 23:54:23 | [285][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1204 ntime: 0083 mem: 3.36 + 04-03 23:54:27 | [285][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0076 mem: 3.36 + 04-03 23:54:29 | [285][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0088 mem: 3.36 + 04-03 23:54:31 | [285][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0079 mem: 3.36 + 04-03 23:54:33 | [285][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0072 mem: 3.36 + 04-03 23:54:35 | [285][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0417 ntime: 0070 mem: 3.36 + 04-03 23:54:37 | [285][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0177 ntime: 0082 mem: 3.36 + 04-03 23:54:39 | Time info >>>> elapsed: 151.28 mins remain: 377.67 mins + 04-03 23:54:39 | [286][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0066 ntime: 0075 mem: 3.36 + 04-03 23:54:41 | [286][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0068 mem: 3.36 + 04-03 23:54:44 | [286][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0473 ntime: 0078 mem: 3.36 + 04-03 23:54:46 | [286][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0209 ntime: 0081 mem: 3.36 + 04-03 23:54:48 | [286][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0327 ntime: 0077 mem: 3.36 + 04-03 23:54:50 | [286][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0083 mem: 3.36 + 04-03 23:54:52 | [286][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0080 mem: 3.36 + 04-03 23:54:55 | [286][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0794 ntime: 0079 mem: 3.36 + 04-03 23:54:57 | [286][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0083 mem: 3.36 + 04-03 23:54:59 | [286][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0137 ntime: 0086 mem: 3.36 + 04-03 23:55:01 | [286][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 23:55:04 | [286][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0387 ntime: 0076 mem: 3.36 + 04-03 23:55:06 | [286][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0067 ntime: 0084 mem: 3.36 + 04-03 23:55:09 | [286][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0329 ntime: 0080 mem: 3.36 + 04-03 23:55:11 | [286][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0186 ntime: 0084 mem: 3.36 + 04-03 23:55:13 | [286][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0315 ntime: 0086 mem: 3.36 + 04-03 23:55:15 | [286][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0329 ntime: 0082 mem: 3.36 + 04-03 23:55:17 | [286][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0157 ntime: 0078 mem: 3.36 + 04-03 23:55:19 | Time info >>>> elapsed: 151.94 mins remain: 377.46 mins + 04-03 23:55:19 | [287][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0078 mem: 3.36 + 04-03 23:55:21 | [287][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0073 mem: 3.36 + 04-03 23:55:23 | [287][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0309 ntime: 0082 mem: 3.36 + 04-03 23:55:25 | [287][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0088 mem: 3.36 + 04-03 23:55:27 | [287][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0190 ntime: 0086 mem: 3.36 + 04-03 23:55:29 | [287][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0240 ntime: 0073 mem: 3.36 + 04-03 23:55:32 | [287][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0222 ntime: 0082 mem: 3.36 + 04-03 23:55:34 | [287][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0083 mem: 3.36 + 04-03 23:55:36 | [287][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0080 mem: 3.36 + 04-03 23:55:38 | [287][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0074 mem: 3.36 + 04-03 23:55:40 | [287][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0081 mem: 3.36 + 04-03 23:55:42 | [287][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0272 ntime: 0079 mem: 3.36 + 04-03 23:55:44 | [287][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0265 ntime: 0083 mem: 3.36 + 04-03 23:55:47 | [287][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0159 ntime: 0085 mem: 3.36 + 04-03 23:55:49 | [287][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 23:55:51 | [287][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0210 ntime: 0084 mem: 3.36 + 04-03 23:55:53 | [287][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0223 ntime: 0081 mem: 3.36 + 04-03 23:55:55 | [287][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0085 mem: 3.36 + 04-03 23:55:57 | Time info >>>> elapsed: 152.58 mins remain: 377.21 mins + 04-03 23:55:57 | [288][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0079 mem: 3.36 + 04-03 23:56:01 | [288][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0260 ntime: 0078 mem: 3.36 + 04-03 23:56:03 | [288][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0078 mem: 3.36 + 04-03 23:56:05 | [288][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0272 ntime: 0086 mem: 3.36 + 04-03 23:56:07 | [288][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0286 ntime: 0084 mem: 3.36 + 04-03 23:56:10 | [288][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0091 mem: 3.36 + 04-03 23:56:13 | [288][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0145 ntime: 0080 mem: 3.36 + 04-03 23:56:15 | [288][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0082 mem: 3.36 + 04-03 23:56:17 | [288][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0081 ntime: 0087 mem: 3.36 + 04-03 23:56:19 | [288][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0078 mem: 3.36 + 04-03 23:56:22 | [288][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0078 mem: 3.36 + 04-03 23:56:24 | [288][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0177 ntime: 0073 mem: 3.36 + 04-03 23:56:27 | [288][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0082 mem: 3.36 + 04-03 23:56:29 | [288][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0084 ntime: 0082 mem: 3.36 + 04-03 23:56:31 | [288][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 23:56:34 | [288][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0080 mem: 3.36 + 04-03 23:56:36 | [288][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0233 ntime: 0081 mem: 3.36 + 04-03 23:56:38 | [288][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0071 mem: 3.36 + 04-03 23:56:40 | Time info >>>> elapsed: 153.30 mins remain: 377.15 mins + 04-03 23:56:41 | [289][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0384 ntime: 0074 mem: 3.36 + 04-03 23:56:43 | [289][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0078 mem: 3.36 + 04-03 23:56:46 | [289][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0188 ntime: 0081 mem: 3.36 + 04-03 23:56:48 | [289][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0086 mem: 3.36 + 04-03 23:56:50 | [289][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0163 ntime: 0086 mem: 3.36 + 04-03 23:56:52 | [289][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0088 mem: 3.36 + 04-03 23:56:55 | [289][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0347 ntime: 0079 mem: 3.36 + 04-03 23:56:57 | [289][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0084 mem: 3.36 + 04-03 23:57:00 | [289][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0153 ntime: 0086 mem: 3.36 + 04-03 23:57:02 | [289][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0157 ntime: 0085 mem: 3.36 + 04-03 23:57:04 | [289][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0262 ntime: 0080 mem: 3.36 + 04-03 23:57:06 | [289][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0128 ntime: 0079 mem: 3.36 + 04-03 23:57:09 | [289][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0086 mem: 3.36 + 04-03 23:57:11 | [289][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0074 mem: 3.36 + 04-03 23:57:13 | [289][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0116 ntime: 0080 mem: 3.36 + 04-03 23:57:15 | [289][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0119 ntime: 0075 mem: 3.36 + 04-03 23:57:17 | [289][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0080 mem: 3.36 + 04-03 23:57:20 | [289][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0234 ntime: 0073 mem: 3.36 + 04-03 23:57:21 | Time info >>>> elapsed: 153.98 mins remain: 376.99 mins + 04-03 23:57:22 | [290][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0208 ntime: 0080 mem: 3.36 + 04-03 23:57:24 | [290][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0252 ntime: 0080 mem: 3.36 + 04-03 23:57:26 | [290][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0355 ntime: 0080 mem: 3.36 + 04-03 23:57:28 | [290][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0340 ntime: 0080 mem: 3.36 + 04-03 23:57:30 | [290][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-03 23:57:32 | [290][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0184 ntime: 0057 mem: 3.36 + 04-03 23:57:34 | [290][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0251 ntime: 0081 mem: 3.36 + 04-03 23:57:36 | [290][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0224 ntime: 0078 mem: 3.36 + 04-03 23:57:38 | [290][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0084 mem: 3.36 + 04-03 23:57:41 | [290][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0079 mem: 3.36 + 04-03 23:57:42 | [290][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-03 23:57:44 | [290][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0075 mem: 3.36 + 04-03 23:57:47 | [290][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0407 ntime: 0084 mem: 3.36 + 04-03 23:57:49 | [290][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0074 mem: 3.36 + 04-03 23:57:52 | [290][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0082 mem: 3.36 + 04-03 23:57:54 | [290][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0280 ntime: 0083 mem: 3.36 + 04-03 23:57:56 | [290][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0075 mem: 3.36 + 04-03 23:57:59 | [290][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0340 ntime: 0082 mem: 3.36 + 04-03 23:58:00 | Time info >>>> elapsed: 154.63 mins remain: 376.75 mins + 04-03 23:58:01 | [291][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0059 ntime: 0077 mem: 3.36 + 04-03 23:58:03 | [291][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0083 mem: 3.36 + 04-03 23:58:06 | [291][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0295 ntime: 0080 mem: 3.36 + 04-03 23:58:08 | [291][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0176 ntime: 0085 mem: 3.36 + 04-03 23:58:11 | [291][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0083 ntime: 0077 mem: 3.36 + 04-03 23:58:13 | [291][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0365 ntime: 0079 mem: 3.36 + 04-03 23:58:15 | [291][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0196 ntime: 0080 mem: 3.36 + 04-03 23:58:17 | [291][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0077 mem: 3.36 + 04-03 23:58:20 | [291][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 23:58:22 | [291][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0218 ntime: 0080 mem: 3.36 + 04-03 23:58:24 | [291][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0081 mem: 3.36 + 04-03 23:58:27 | [291][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0081 mem: 3.36 + 04-03 23:58:29 | [291][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0159 ntime: 0082 mem: 3.36 + 04-03 23:58:32 | [291][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0169 ntime: 0080 mem: 3.36 + 04-03 23:58:34 | [291][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0100 ntime: 0078 mem: 3.36 + 04-03 23:58:36 | [291][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-03 23:58:39 | [291][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0194 ntime: 0079 mem: 3.36 + 04-03 23:58:41 | [291][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0076 mem: 3.36 + 04-03 23:58:43 | Time info >>>> elapsed: 155.35 mins remain: 376.67 mins + 04-03 23:58:43 | [292][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0082 mem: 3.36 + 04-03 23:58:46 | [292][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0070 mem: 3.36 + 04-03 23:58:48 | [292][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0143 ntime: 0084 mem: 3.36 + 04-03 23:58:50 | [292][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0163 ntime: 0078 mem: 3.36 + 04-03 23:58:52 | [292][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0071 mem: 3.36 + 04-03 23:58:54 | [292][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0079 mem: 3.36 + 04-03 23:58:56 | [292][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0327 ntime: 0081 mem: 3.36 + 04-03 23:58:58 | [292][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0081 mem: 3.36 + 04-03 23:59:00 | [292][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0084 mem: 3.36 + 04-03 23:59:03 | [292][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0077 mem: 3.36 + 04-03 23:59:05 | [292][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0227 ntime: 0079 mem: 3.36 + 04-03 23:59:07 | [292][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0175 ntime: 0081 mem: 3.36 + 04-03 23:59:10 | [292][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0260 ntime: 0083 mem: 3.36 + 04-03 23:59:12 | [292][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0081 mem: 3.36 + 04-03 23:59:14 | [292][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0291 ntime: 0088 mem: 3.36 + 04-03 23:59:16 | [292][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-03 23:59:19 | [292][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0083 mem: 3.36 + 04-03 23:59:21 | [292][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0098 ntime: 0079 mem: 3.36 + 04-03 23:59:23 | Time info >>>> elapsed: 156.01 mins remain: 376.45 mins + 04-03 23:59:23 | [293][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0084 ntime: 0079 mem: 3.36 + 04-03 23:59:25 | [293][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0212 ntime: 0076 mem: 3.36 + 04-03 23:59:27 | [293][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0083 mem: 3.36 + 04-03 23:59:30 | [293][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-03 23:59:32 | [293][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0316 ntime: 0089 mem: 3.36 + 04-03 23:59:34 | [293][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0083 mem: 3.36 + 04-03 23:59:37 | [293][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0081 mem: 3.36 + 04-03 23:59:39 | [293][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0086 mem: 3.36 + 04-03 23:59:42 | [293][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0080 mem: 3.36 + 04-03 23:59:45 | [293][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0232 ntime: 0082 mem: 3.36 + 04-03 23:59:48 | [293][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0085 mem: 3.36 + 04-03 23:59:51 | [293][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0212 ntime: 0089 mem: 3.36 + 04-03 23:59:53 | [293][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0045 ntime: 0077 mem: 3.36 + 04-03 23:59:55 | [293][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0294 ntime: 0079 mem: 3.36 + 04-03 23:59:58 | [293][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0087 mem: 3.36 + 04-04 00:00:00 | [293][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 00:00:03 | [293][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0198 ntime: 0087 mem: 3.36 + 04-04 00:00:05 | [293][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0061 mem: 3.36 + 04-04 00:00:06 | Time info >>>> elapsed: 156.73 mins remain: 376.37 mins + 04-04 00:00:07 | [294][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0309 ntime: 0087 mem: 3.36 + 04-04 00:00:09 | [294][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0077 mem: 3.36 + 04-04 00:00:11 | [294][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0086 mem: 3.36 + 04-04 00:00:13 | [294][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0405 ntime: 0086 mem: 3.36 + 04-04 00:00:15 | [294][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 00:00:17 | [294][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0185 ntime: 0078 mem: 3.36 + 04-04 00:00:19 | [294][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0080 mem: 3.36 + 04-04 00:00:21 | [294][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0073 mem: 3.36 + 04-04 00:00:23 | [294][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0190 ntime: 0077 mem: 3.36 + 04-04 00:00:26 | [294][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0266 ntime: 0081 mem: 3.36 + 04-04 00:00:28 | [294][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0440 ntime: 0089 mem: 3.36 + 04-04 00:00:30 | [294][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0078 mem: 3.36 + 04-04 00:00:33 | [294][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0275 ntime: 0083 mem: 3.36 + 04-04 00:00:35 | [294][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0246 ntime: 0083 mem: 3.36 + 04-04 00:00:37 | [294][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0082 mem: 3.36 + 04-04 00:00:40 | [294][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0075 mem: 3.36 + 04-04 00:00:42 | [294][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0075 mem: 3.36 + 04-04 00:00:44 | [294][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0178 ntime: 0083 mem: 3.36 + 04-04 00:00:46 | Time info >>>> elapsed: 157.40 mins remain: 376.15 mins + 04-04 00:00:46 | [295][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0096 ntime: 0088 mem: 3.36 + 04-04 00:00:48 | [295][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0356 ntime: 0079 mem: 3.36 + 04-04 00:00:52 | [295][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 00:00:54 | [295][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0077 mem: 3.36 + 04-04 00:00:57 | [295][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0204 ntime: 0079 mem: 3.36 + 04-04 00:00:59 | [295][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0081 mem: 3.36 + 04-04 00:01:01 | [295][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0081 mem: 3.36 + 04-04 00:01:03 | [295][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0088 mem: 3.36 + 04-04 00:01:05 | [295][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0087 mem: 3.36 + 04-04 00:01:08 | [295][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 00:01:10 | [295][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0330 ntime: 0093 mem: 3.36 + 04-04 00:01:12 | [295][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0182 ntime: 0078 mem: 3.36 + 04-04 00:01:15 | [295][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0085 mem: 3.36 + 04-04 00:01:17 | [295][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0194 ntime: 0087 mem: 3.36 + 04-04 00:01:19 | [295][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0075 mem: 3.36 + 04-04 00:01:22 | [295][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0183 ntime: 0078 mem: 3.36 + 04-04 00:01:23 | [295][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0088 mem: 3.36 + 04-04 00:01:26 | [295][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0076 mem: 3.36 + 04-04 00:01:27 | Time info >>>> elapsed: 158.08 mins remain: 375.98 mins + 04-04 00:01:28 | [296][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0161 ntime: 0085 mem: 3.36 + 04-04 00:01:30 | [296][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0112 ntime: 0082 mem: 3.36 + 04-04 00:01:32 | [296][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0089 ntime: 0078 mem: 3.36 + 04-04 00:01:34 | [296][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0244 ntime: 0078 mem: 3.36 + 04-04 00:01:37 | [296][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0076 mem: 3.36 + 04-04 00:01:39 | [296][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0212 ntime: 0090 mem: 3.36 + 04-04 00:01:41 | [296][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0089 mem: 3.36 + 04-04 00:01:43 | [296][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0088 mem: 3.36 + 04-04 00:01:45 | [296][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0123 ntime: 0080 mem: 3.36 + 04-04 00:01:47 | [296][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0075 mem: 3.36 + 04-04 00:01:50 | [296][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0789 ntime: 0075 mem: 3.36 + 04-04 00:01:52 | [296][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0081 ntime: 0073 mem: 3.36 + 04-04 00:01:54 | [296][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0127 ntime: 0082 mem: 3.36 + 04-04 00:01:56 | [296][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0216 ntime: 0083 mem: 3.36 + 04-04 00:01:59 | [296][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0079 mem: 3.36 + 04-04 00:02:02 | [296][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0325 ntime: 0090 mem: 3.36 + 04-04 00:02:04 | [296][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0209 ntime: 0084 mem: 3.36 + 04-04 00:02:06 | [296][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0075 mem: 3.36 + 04-04 00:02:08 | Time info >>>> elapsed: 158.76 mins remain: 375.79 mins + 04-04 00:02:09 | [297][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0653 ntime: 0086 mem: 3.36 + 04-04 00:02:11 | [297][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0082 mem: 3.36 + 04-04 00:02:13 | [297][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0083 mem: 3.36 + 04-04 00:02:15 | [297][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0076 mem: 3.36 + 04-04 00:02:17 | [297][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 00:02:20 | [297][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0081 mem: 3.36 + 04-04 00:02:22 | [297][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0080 mem: 3.36 + 04-04 00:02:25 | [297][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0079 mem: 3.36 + 04-04 00:02:27 | [297][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0076 mem: 3.36 + 04-04 00:02:29 | [297][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0084 mem: 3.36 + 04-04 00:02:31 | [297][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0045 ntime: 0085 mem: 3.36 + 04-04 00:02:33 | [297][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0224 ntime: 0080 mem: 3.36 + 04-04 00:02:36 | [297][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0078 mem: 3.36 + 04-04 00:02:38 | [297][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0194 ntime: 0083 mem: 3.36 + 04-04 00:02:40 | [297][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0131 ntime: 0087 mem: 3.36 + 04-04 00:02:42 | [297][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0085 ntime: 0079 mem: 3.36 + 04-04 00:02:45 | [297][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 00:02:47 | [297][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0079 mem: 3.36 + 04-04 00:02:49 | Time info >>>> elapsed: 159.44 mins remain: 375.60 mins + 04-04 00:02:49 | [298][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0075 mem: 3.36 + 04-04 00:02:52 | [298][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0179 ntime: 0083 mem: 3.36 + 04-04 00:02:54 | [298][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0086 mem: 3.36 + 04-04 00:02:56 | [298][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0086 ntime: 0072 mem: 3.36 + 04-04 00:02:59 | [298][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0082 mem: 3.36 + 04-04 00:03:01 | [298][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0086 mem: 3.36 + 04-04 00:03:03 | [298][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0114 ntime: 0080 mem: 3.36 + 04-04 00:03:05 | [298][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0149 ntime: 0081 mem: 3.36 + 04-04 00:03:07 | [298][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0084 ntime: 0078 mem: 3.36 + 04-04 00:03:10 | [298][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0075 mem: 3.36 + 04-04 00:03:12 | [298][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0170 ntime: 0080 mem: 3.36 + 04-04 00:03:14 | [298][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0078 mem: 3.36 + 04-04 00:03:16 | [298][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0083 mem: 3.36 + 04-04 00:03:19 | [298][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0382 ntime: 0079 mem: 3.36 + 04-04 00:03:21 | [298][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0080 mem: 3.36 + 04-04 00:03:24 | [298][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0076 mem: 3.36 + 04-04 00:03:26 | [298][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0080 mem: 3.36 + 04-04 00:03:28 | [298][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0233 ntime: 0086 mem: 3.36 + 04-04 00:03:30 | Time info >>>> elapsed: 160.12 mins remain: 375.41 mins + 04-04 00:03:30 | [299][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0083 mem: 3.36 + 04-04 00:03:32 | [299][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0080 mem: 3.36 + 04-04 00:03:34 | [299][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0082 mem: 3.36 + 04-04 00:03:37 | [299][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 00:03:39 | [299][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0080 mem: 3.36 + 04-04 00:03:42 | [299][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1002 ntime: 0082 mem: 3.36 + 04-04 00:03:44 | [299][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0077 mem: 3.36 + 04-04 00:03:46 | [299][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0079 mem: 3.36 + 04-04 00:03:47 | [299][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0170 ntime: 0084 mem: 3.36 + 04-04 00:03:49 | [299][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0192 ntime: 0079 mem: 3.36 + 04-04 00:03:52 | [299][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0252 ntime: 0088 mem: 3.36 + 04-04 00:03:54 | [299][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0504 ntime: 0075 mem: 3.36 + 04-04 00:03:56 | [299][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0084 mem: 3.36 + 04-04 00:03:58 | [299][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0074 mem: 3.36 + 04-04 00:04:00 | [299][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0354 ntime: 0083 mem: 3.36 + 04-04 00:04:02 | [299][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0083 mem: 3.36 + 04-04 00:04:04 | [299][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0092 ntime: 0081 mem: 3.36 + 04-04 00:04:07 | [299][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0170 ntime: 0082 mem: 3.36 + 04-04 00:04:09 | Time info >>>> elapsed: 160.78 mins remain: 375.16 mins + 04-04 00:04:09 | [300][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 00:04:12 | [300][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0234 ntime: 0081 mem: 3.36 + 04-04 00:04:14 | [300][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 00:04:15 | [300][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0075 mem: 3.36 + 04-04 00:04:18 | [300][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0239 ntime: 0084 mem: 3.36 + 04-04 00:04:20 | [300][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0076 mem: 3.36 + 04-04 00:04:22 | [300][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0077 mem: 3.36 + 04-04 00:04:24 | [300][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0147 ntime: 0085 mem: 3.36 + 04-04 00:04:26 | [300][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0131 ntime: 0083 mem: 3.36 + 04-04 00:04:28 | [300][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0080 mem: 3.36 + 04-04 00:04:30 | [300][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0116 ntime: 0083 mem: 3.36 + 04-04 00:04:32 | [300][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0079 mem: 3.36 + 04-04 00:04:36 | [300][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0083 mem: 3.36 + 04-04 00:04:39 | [300][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0085 ntime: 0081 mem: 3.36 + 04-04 00:04:41 | [300][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0334 ntime: 0080 mem: 3.36 + 04-04 00:04:43 | [300][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 00:04:45 | [300][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0143 ntime: 0076 mem: 3.36 + 04-04 00:04:47 | [300][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0185 ntime: 0076 mem: 3.36 + 04-04 00:04:49 | Time info >>>> elapsed: 161.44 mins remain: 374.90 mins + 04-04 00:04:49 | [301][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0085 mem: 3.36 + 04-04 00:04:50 | [301][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0076 mem: 3.36 + 04-04 00:04:52 | [301][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0323 ntime: 0083 mem: 3.36 + 04-04 00:04:54 | [301][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0083 mem: 3.36 + 04-04 00:04:56 | [301][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 00:04:58 | [301][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0108 ntime: 0079 mem: 3.36 + 04-04 00:05:00 | [301][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0129 ntime: 0080 mem: 3.36 + 04-04 00:05:02 | [301][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0079 mem: 3.36 + 04-04 00:05:04 | [301][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0077 mem: 3.36 + 04-04 00:05:06 | [301][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0084 mem: 3.36 + 04-04 00:05:08 | [301][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0079 mem: 3.36 + 04-04 00:05:10 | [301][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0135 ntime: 0081 mem: 3.36 + 04-04 00:05:12 | [301][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 00:05:13 | [301][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0111 ntime: 0080 mem: 3.36 + 04-04 00:05:15 | [301][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-04 00:05:17 | [301][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0079 mem: 3.36 + 04-04 00:05:20 | [301][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0378 ntime: 0077 mem: 3.36 + 04-04 00:05:22 | [301][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0161 ntime: 0075 mem: 3.36 + 04-04 00:05:23 | Time info >>>> elapsed: 162.01 mins remain: 374.45 mins + 04-04 00:05:23 | [302][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0072 mem: 3.36 + 04-04 00:05:25 | [302][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0100 ntime: 0084 mem: 3.36 + 04-04 00:05:27 | [302][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0083 mem: 3.36 + 04-04 00:05:29 | [302][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0175 ntime: 0077 mem: 3.36 + 04-04 00:05:31 | [302][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0178 ntime: 0084 mem: 3.36 + 04-04 00:05:33 | [302][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0073 mem: 3.36 + 04-04 00:05:35 | [302][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0084 mem: 3.36 + 04-04 00:05:37 | [302][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0096 ntime: 0081 mem: 3.36 + 04-04 00:05:39 | [302][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0205 ntime: 0085 mem: 3.36 + 04-04 00:05:41 | [302][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0227 ntime: 0079 mem: 3.36 + 04-04 00:05:43 | [302][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0077 mem: 3.36 + 04-04 00:05:45 | [302][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0081 mem: 3.36 + 04-04 00:05:48 | [302][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0162 ntime: 0085 mem: 3.36 + 04-04 00:05:50 | [302][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0082 mem: 3.36 + 04-04 00:05:53 | [302][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0178 ntime: 0090 mem: 3.36 + 04-04 00:05:55 | [302][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0138 ntime: 0079 mem: 3.36 + 04-04 00:05:57 | [302][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0106 ntime: 0090 mem: 3.36 + 04-04 00:05:59 | [302][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0084 mem: 3.36 + 04-04 00:06:01 | Time info >>>> elapsed: 162.65 mins remain: 374.15 mins + 04-04 00:06:02 | [303][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0054 ntime: 0073 mem: 3.36 + 04-04 00:06:04 | [303][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0041 ntime: 0058 mem: 3.36 + 04-04 00:06:06 | [303][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0165 ntime: 0077 mem: 3.36 + 04-04 00:06:08 | [303][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0214 ntime: 0075 mem: 3.36 + 04-04 00:06:10 | [303][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 00:06:12 | [303][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0077 mem: 3.36 + 04-04 00:06:14 | [303][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0074 mem: 3.36 + 04-04 00:06:17 | [303][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0314 ntime: 0081 mem: 3.36 + 04-04 00:06:19 | [303][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0231 ntime: 0082 mem: 3.36 + 04-04 00:06:21 | [303][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0169 ntime: 0079 mem: 3.36 + 04-04 00:06:23 | [303][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0087 ntime: 0077 mem: 3.36 + 04-04 00:06:26 | [303][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0086 mem: 3.36 + 04-04 00:06:28 | [303][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0227 ntime: 0077 mem: 3.36 + 04-04 00:06:30 | [303][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0086 ntime: 0079 mem: 3.36 + 04-04 00:06:32 | [303][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0070 mem: 3.36 + 04-04 00:06:35 | [303][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0077 mem: 3.36 + 04-04 00:06:37 | [303][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0133 ntime: 0082 mem: 3.36 + 04-04 00:06:38 | [303][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0080 mem: 3.36 + 04-04 00:06:40 | Time info >>>> elapsed: 163.30 mins remain: 373.88 mins + 04-04 00:06:41 | [304][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0107 ntime: 0076 mem: 3.36 + 04-04 00:06:43 | [304][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0077 mem: 3.36 + 04-04 00:06:46 | [304][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0986 ntime: 0083 mem: 3.36 + 04-04 00:06:48 | [304][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0077 mem: 3.36 + 04-04 00:06:50 | [304][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0079 mem: 3.36 + 04-04 00:06:52 | [304][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0120 ntime: 0083 mem: 3.36 + 04-04 00:06:54 | [304][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 00:06:55 | [304][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0179 ntime: 0079 mem: 3.36 + 04-04 00:06:57 | [304][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0143 ntime: 0080 mem: 3.36 + 04-04 00:06:59 | [304][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0260 ntime: 0072 mem: 3.36 + 04-04 00:07:01 | [304][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0083 mem: 3.36 + 04-04 00:07:03 | [304][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0083 mem: 3.36 + 04-04 00:07:05 | [304][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0090 mem: 3.36 + 04-04 00:07:07 | [304][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0074 mem: 3.36 + 04-04 00:07:10 | [304][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0965 ntime: 0074 mem: 3.36 + 04-04 00:07:12 | [304][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 00:07:14 | [304][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0175 ntime: 0080 mem: 3.36 + 04-04 00:07:15 | [304][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0086 mem: 3.36 + 04-04 00:07:17 | Time info >>>> elapsed: 163.91 mins remain: 373.51 mins + 04-04 00:07:18 | [305][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0599 ntime: 0080 mem: 3.36 + 04-04 00:07:20 | [305][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0086 mem: 3.36 + 04-04 00:07:22 | [305][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0083 ntime: 0082 mem: 3.36 + 04-04 00:07:24 | [305][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0078 mem: 3.36 + 04-04 00:07:26 | [305][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0090 mem: 3.36 + 04-04 00:07:27 | [305][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0151 ntime: 0076 mem: 3.36 + 04-04 00:07:29 | [305][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0081 mem: 3.36 + 04-04 00:07:31 | [305][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0160 ntime: 0083 mem: 3.36 + 04-04 00:07:33 | [305][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0080 mem: 3.36 + 04-04 00:07:35 | [305][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0213 ntime: 0081 mem: 3.36 + 04-04 00:07:37 | [305][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0080 mem: 3.36 + 04-04 00:07:39 | [305][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0263 ntime: 0075 mem: 3.36 + 04-04 00:07:41 | [305][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0178 ntime: 0081 mem: 3.36 + 04-04 00:07:43 | [305][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0207 ntime: 0081 mem: 3.36 + 04-04 00:07:45 | [305][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0146 ntime: 0080 mem: 3.36 + 04-04 00:07:47 | [305][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0081 mem: 3.36 + 04-04 00:07:49 | [305][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 00:07:51 | [305][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0108 ntime: 0078 mem: 3.36 + 04-04 00:07:53 | Time info >>>> elapsed: 164.50 mins remain: 373.09 mins + 04-04 00:07:53 | [306][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-04 00:07:55 | [306][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0234 ntime: 0077 mem: 3.36 + 04-04 00:07:57 | [306][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 00:07:59 | [306][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0088 mem: 3.36 + 04-04 00:08:01 | [306][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0150 ntime: 0088 mem: 3.36 + 04-04 00:08:03 | [306][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0079 mem: 3.36 + 04-04 00:08:05 | [306][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0167 ntime: 0073 mem: 3.36 + 04-04 00:08:07 | [306][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0251 ntime: 0086 mem: 3.36 + 04-04 00:08:08 | [306][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0076 mem: 3.36 + 04-04 00:08:11 | [306][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0516 ntime: 0076 mem: 3.36 + 04-04 00:08:13 | [306][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0097 ntime: 0075 mem: 3.36 + 04-04 00:08:15 | [306][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0083 mem: 3.36 + 04-04 00:08:16 | [306][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0216 ntime: 0082 mem: 3.36 + 04-04 00:08:18 | [306][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0137 ntime: 0080 mem: 3.36 + 04-04 00:08:20 | [306][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0085 mem: 3.36 + 04-04 00:08:22 | [306][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0170 ntime: 0075 mem: 3.36 + 04-04 00:08:24 | [306][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 00:08:27 | [306][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0162 ntime: 0077 mem: 3.36 + 04-04 00:08:28 | Time info >>>> elapsed: 165.10 mins remain: 372.69 mins + 04-04 00:08:29 | [307][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 00:08:31 | [307][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0078 mem: 3.36 + 04-04 00:08:33 | [307][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0281 ntime: 0084 mem: 3.36 + 04-04 00:08:35 | [307][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0085 mem: 3.36 + 04-04 00:08:37 | [307][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0201 ntime: 0082 mem: 3.36 + 04-04 00:08:38 | [307][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0075 mem: 3.36 + 04-04 00:08:40 | [307][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0263 ntime: 0082 mem: 3.36 + 04-04 00:08:42 | [307][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0296 ntime: 0088 mem: 3.36 + 04-04 00:08:45 | [307][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0156 ntime: 0089 mem: 3.36 + 04-04 00:08:47 | [307][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 00:08:50 | [307][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0227 ntime: 0080 mem: 3.36 + 04-04 00:08:52 | [307][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0400 ntime: 0082 mem: 3.36 + 04-04 00:08:54 | [307][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0086 mem: 3.36 + 04-04 00:08:56 | [307][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0096 mem: 3.36 + 04-04 00:08:58 | [307][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0078 mem: 3.36 + 04-04 00:09:01 | [307][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0146 ntime: 0077 mem: 3.36 + 04-04 00:09:03 | [307][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0077 mem: 3.36 + 04-04 00:09:05 | [307][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0078 mem: 3.36 + 04-04 00:09:06 | Time info >>>> elapsed: 165.73 mins remain: 372.36 mins + 04-04 00:09:07 | [308][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0532 ntime: 0079 mem: 3.36 + 04-04 00:09:09 | [308][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0533 ntime: 0074 mem: 3.36 + 04-04 00:09:12 | [308][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0070 mem: 3.36 + 04-04 00:09:15 | [308][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0139 ntime: 0085 mem: 3.36 + 04-04 00:09:17 | [308][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0155 ntime: 0082 mem: 3.36 + 04-04 00:09:19 | [308][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0142 ntime: 0078 mem: 3.36 + 04-04 00:09:21 | [308][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0150 ntime: 0077 mem: 3.36 + 04-04 00:09:23 | [308][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0076 mem: 3.36 + 04-04 00:09:26 | [308][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0266 ntime: 0083 mem: 3.36 + 04-04 00:09:28 | [308][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 00:09:31 | [308][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0253 ntime: 0077 mem: 3.36 + 04-04 00:09:34 | [308][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0197 ntime: 0080 mem: 3.36 + 04-04 00:09:36 | [308][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 00:09:38 | [308][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 00:09:41 | [308][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0238 ntime: 0073 mem: 3.36 + 04-04 00:09:43 | [308][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0094 ntime: 0074 mem: 3.36 + 04-04 00:09:45 | [308][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0163 ntime: 0082 mem: 3.36 + 04-04 00:09:47 | [308][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0272 ntime: 0080 mem: 3.36 + 04-04 00:09:49 | Time info >>>> elapsed: 166.44 mins remain: 372.21 mins + 04-04 00:09:49 | [309][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0080 mem: 3.36 + 04-04 00:09:51 | [309][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0076 mem: 3.36 + 04-04 00:09:53 | [309][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0258 ntime: 0075 mem: 3.36 + 04-04 00:09:56 | [309][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0082 ntime: 0081 mem: 3.36 + 04-04 00:09:58 | [309][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0271 ntime: 0077 mem: 3.36 + 04-04 00:10:00 | [309][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0107 ntime: 0081 mem: 3.36 + 04-04 00:10:02 | [309][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0136 ntime: 0079 mem: 3.36 + 04-04 00:10:05 | [309][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0425 ntime: 0081 mem: 3.36 + 04-04 00:10:07 | [309][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0081 mem: 3.36 + 04-04 00:10:09 | [309][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0226 ntime: 0083 mem: 3.36 + 04-04 00:10:12 | [309][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0081 mem: 3.36 + 04-04 00:10:14 | [309][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0080 mem: 3.36 + 04-04 00:10:16 | [309][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0080 mem: 3.36 + 04-04 00:10:18 | [309][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0083 mem: 3.36 + 04-04 00:10:21 | [309][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0445 ntime: 0075 mem: 3.36 + 04-04 00:10:23 | [309][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0083 mem: 3.36 + 04-04 00:10:26 | [309][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0303 ntime: 0084 mem: 3.36 + 04-04 00:10:28 | [309][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0163 ntime: 0082 mem: 3.36 + 04-04 00:10:30 | Time info >>>> elapsed: 167.13 mins remain: 371.99 mins + 04-04 00:10:30 | [310][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0271 ntime: 0082 mem: 3.36 + 04-04 00:10:32 | [310][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0189 ntime: 0079 mem: 3.36 + 04-04 00:10:34 | [310][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0284 ntime: 0078 mem: 3.36 + 04-04 00:10:37 | [310][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0362 ntime: 0071 mem: 3.36 + 04-04 00:10:39 | [310][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 00:10:42 | [310][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0076 mem: 3.36 + 04-04 00:10:44 | [310][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 00:10:46 | [310][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0155 ntime: 0082 mem: 3.36 + 04-04 00:10:49 | [310][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0133 ntime: 0082 mem: 3.36 + 04-04 00:10:52 | [310][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0091 ntime: 0073 mem: 3.36 + 04-04 00:10:54 | [310][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0080 mem: 3.36 + 04-04 00:10:56 | [310][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0087 mem: 3.36 + 04-04 00:10:58 | [310][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0221 ntime: 0089 mem: 3.36 + 04-04 00:11:00 | [310][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0275 ntime: 0088 mem: 3.36 + 04-04 00:11:03 | [310][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0241 ntime: 0086 mem: 3.36 + 04-04 00:11:06 | [310][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0113 ntime: 0077 mem: 3.36 + 04-04 00:11:08 | [310][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0073 ntime: 0077 mem: 3.36 + 04-04 00:11:10 | [310][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 00:11:11 | Time info >>>> elapsed: 167.82 mins remain: 371.78 mins + 04-04 00:11:12 | [311][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0193 ntime: 0078 mem: 3.36 + 04-04 00:11:13 | [311][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 00:11:16 | [311][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0088 mem: 3.36 + 04-04 00:11:18 | [311][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0080 mem: 3.36 + 04-04 00:11:20 | [311][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0087 mem: 3.36 + 04-04 00:11:22 | [311][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0082 mem: 3.36 + 04-04 00:11:25 | [311][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0339 ntime: 0088 mem: 3.36 + 04-04 00:11:27 | [311][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0170 ntime: 0087 mem: 3.36 + 04-04 00:11:29 | [311][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0081 mem: 3.36 + 04-04 00:11:31 | [311][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0246 ntime: 0082 mem: 3.36 + 04-04 00:11:34 | [311][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0082 ntime: 0085 mem: 3.36 + 04-04 00:11:36 | [311][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0118 ntime: 0085 mem: 3.36 + 04-04 00:11:38 | [311][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0162 ntime: 0085 mem: 3.36 + 04-04 00:11:40 | [311][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0078 ntime: 0079 mem: 3.36 + 04-04 00:11:42 | [311][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0083 mem: 3.36 + 04-04 00:11:45 | [311][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0116 ntime: 0079 mem: 3.36 + 04-04 00:11:47 | [311][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0080 mem: 3.36 + 04-04 00:11:49 | [311][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 00:11:50 | Time info >>>> elapsed: 168.46 mins remain: 371.49 mins + 04-04 00:11:51 | [312][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0224 ntime: 0077 mem: 3.36 + 04-04 00:11:53 | [312][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0265 ntime: 0055 mem: 3.36 + 04-04 00:11:54 | [312][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0145 ntime: 0080 mem: 3.36 + 04-04 00:11:56 | [312][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0278 ntime: 0079 mem: 3.36 + 04-04 00:11:58 | [312][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0179 ntime: 0080 mem: 3.36 + 04-04 00:12:00 | [312][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0077 mem: 3.36 + 04-04 00:12:03 | [312][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0287 ntime: 0083 mem: 3.36 + 04-04 00:12:06 | [312][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0071 mem: 3.36 + 04-04 00:12:09 | [312][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0207 ntime: 0081 mem: 3.36 + 04-04 00:12:11 | [312][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0074 mem: 3.36 + 04-04 00:12:13 | [312][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0084 ntime: 0074 mem: 3.36 + 04-04 00:12:15 | [312][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0084 mem: 3.36 + 04-04 00:12:17 | [312][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 00:12:19 | [312][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0147 ntime: 0078 mem: 3.36 + 04-04 00:12:22 | [312][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0200 ntime: 0084 mem: 3.36 + 04-04 00:12:24 | [312][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0317 ntime: 0079 mem: 3.36 + 04-04 00:12:26 | [312][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0384 ntime: 0082 mem: 3.36 + 04-04 00:12:28 | [312][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0082 mem: 3.36 + 04-04 00:12:31 | Time info >>>> elapsed: 169.14 mins remain: 371.24 mins + 04-04 00:12:32 | [313][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1230 ntime: 0074 mem: 3.36 + 04-04 00:12:35 | [313][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0368 ntime: 0078 mem: 3.36 + 04-04 00:12:37 | [313][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0210 ntime: 0086 mem: 3.36 + 04-04 00:12:39 | [313][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0298 ntime: 0079 mem: 3.36 + 04-04 00:12:42 | [313][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0473 ntime: 0079 mem: 3.36 + 04-04 00:12:45 | [313][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0079 ntime: 0074 mem: 3.36 + 04-04 00:12:47 | [313][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0080 mem: 3.36 + 04-04 00:12:49 | [313][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0380 ntime: 0089 mem: 3.36 + 04-04 00:12:51 | [313][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0074 mem: 3.36 + 04-04 00:12:53 | [313][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0244 ntime: 0078 mem: 3.36 + 04-04 00:12:56 | [313][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0169 ntime: 0079 mem: 3.36 + 04-04 00:12:57 | [313][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0253 ntime: 0070 mem: 3.36 + 04-04 00:13:00 | [313][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0168 ntime: 0081 mem: 3.36 + 04-04 00:13:02 | [313][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0105 ntime: 0078 mem: 3.36 + 04-04 00:13:04 | [313][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0087 mem: 3.36 + 04-04 00:13:05 | [313][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0079 mem: 3.36 + 04-04 00:13:07 | [313][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0134 ntime: 0079 mem: 3.36 + 04-04 00:13:09 | [313][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0080 mem: 3.36 + 04-04 00:13:11 | Time info >>>> elapsed: 169.82 mins remain: 371.00 mins + 04-04 00:13:12 | [314][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0087 ntime: 0086 mem: 3.36 + 04-04 00:13:13 | [314][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0067 ntime: 0081 mem: 3.36 + 04-04 00:13:16 | [314][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0078 mem: 3.36 + 04-04 00:13:17 | [314][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 00:13:19 | [314][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0080 mem: 3.36 + 04-04 00:13:21 | [314][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 00:13:23 | [314][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0078 mem: 3.36 + 04-04 00:13:25 | [314][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0077 mem: 3.36 + 04-04 00:13:27 | [314][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0080 mem: 3.36 + 04-04 00:13:29 | [314][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0081 mem: 3.36 + 04-04 00:13:32 | [314][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0095 ntime: 0083 mem: 3.36 + 04-04 00:13:34 | [314][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0212 ntime: 0086 mem: 3.36 + 04-04 00:13:36 | [314][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0080 ntime: 0075 mem: 3.36 + 04-04 00:13:38 | [314][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0272 ntime: 0084 mem: 3.36 + 04-04 00:13:40 | [314][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0085 mem: 3.36 + 04-04 00:13:42 | [314][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0083 mem: 3.36 + 04-04 00:13:44 | [314][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0231 ntime: 0081 mem: 3.36 + 04-04 00:13:46 | [314][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 00:13:48 | Time info >>>> elapsed: 170.42 mins remain: 370.60 mins + 04-04 00:13:48 | [315][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0195 ntime: 0077 mem: 3.36 + 04-04 00:13:51 | [315][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0461 ntime: 0077 mem: 3.36 + 04-04 00:13:53 | [315][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0081 ntime: 0076 mem: 3.36 + 04-04 00:13:55 | [315][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0169 ntime: 0087 mem: 3.36 + 04-04 00:13:58 | [315][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0228 ntime: 0077 mem: 3.36 + 04-04 00:13:59 | [315][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0139 ntime: 0077 mem: 3.36 + 04-04 00:14:01 | [315][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0165 ntime: 0081 mem: 3.36 + 04-04 00:14:03 | [315][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0079 mem: 3.36 + 04-04 00:14:05 | [315][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0123 ntime: 0081 mem: 3.36 + 04-04 00:14:07 | [315][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0090 mem: 3.36 + 04-04 00:14:09 | [315][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 00:14:11 | [315][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0194 ntime: 0083 mem: 3.36 + 04-04 00:14:13 | [315][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0082 mem: 3.36 + 04-04 00:14:16 | [315][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0106 ntime: 0082 mem: 3.36 + 04-04 00:14:17 | [315][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0077 mem: 3.36 + 04-04 00:14:19 | [315][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0417 ntime: 0086 mem: 3.36 + 04-04 00:14:22 | [315][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0217 ntime: 0087 mem: 3.36 + 04-04 00:14:23 | [315][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0080 mem: 3.36 + 04-04 00:14:25 | Time info >>>> elapsed: 171.05 mins remain: 370.24 mins + 04-04 00:14:26 | [316][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0257 ntime: 0073 mem: 3.36 + 04-04 00:14:28 | [316][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0220 ntime: 0080 mem: 3.36 + 04-04 00:14:29 | [316][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0087 mem: 3.36 + 04-04 00:14:32 | [316][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0964 ntime: 0080 mem: 3.36 + 04-04 00:14:34 | [316][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0147 ntime: 0080 mem: 3.36 + 04-04 00:14:36 | [316][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0080 mem: 3.36 + 04-04 00:14:38 | [316][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0126 ntime: 0079 mem: 3.36 + 04-04 00:14:40 | [316][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0304 ntime: 0077 mem: 3.36 + 04-04 00:14:42 | [316][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0075 mem: 3.36 + 04-04 00:14:44 | [316][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0080 ntime: 0078 mem: 3.36 + 04-04 00:14:47 | [316][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0044 ntime: 0058 mem: 3.36 + 04-04 00:14:49 | [316][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0085 mem: 3.36 + 04-04 00:14:50 | [316][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0084 mem: 3.36 + 04-04 00:14:53 | [316][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0081 mem: 3.36 + 04-04 00:14:55 | [316][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0144 ntime: 0081 mem: 3.36 + 04-04 00:14:58 | [316][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0077 mem: 3.36 + 04-04 00:15:00 | [316][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0132 ntime: 0078 mem: 3.36 + 04-04 00:15:02 | [316][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0152 ntime: 0077 mem: 3.36 + 04-04 00:15:03 | Time info >>>> elapsed: 171.69 mins remain: 369.91 mins + 04-04 00:15:04 | [317][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0365 ntime: 0078 mem: 3.36 + 04-04 00:15:06 | [317][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0079 mem: 3.36 + 04-04 00:15:08 | [317][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0083 mem: 3.36 + 04-04 00:15:10 | [317][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 00:15:13 | [317][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0191 ntime: 0075 mem: 3.36 + 04-04 00:15:15 | [317][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0076 mem: 3.36 + 04-04 00:15:17 | [317][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 00:15:19 | [317][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0362 ntime: 0090 mem: 3.36 + 04-04 00:15:22 | [317][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0083 mem: 3.36 + 04-04 00:15:24 | [317][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0080 mem: 3.36 + 04-04 00:15:26 | [317][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0175 ntime: 0080 mem: 3.36 + 04-04 00:15:28 | [317][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0220 ntime: 0076 mem: 3.36 + 04-04 00:15:31 | [317][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0338 ntime: 0083 mem: 3.36 + 04-04 00:15:33 | [317][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0072 mem: 3.36 + 04-04 00:15:36 | [317][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0142 ntime: 0073 mem: 3.36 + 04-04 00:15:38 | [317][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0084 mem: 3.36 + 04-04 00:15:40 | [317][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0202 ntime: 0075 mem: 3.36 + 04-04 00:15:43 | [317][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0343 ntime: 0088 mem: 3.36 + 04-04 00:15:44 | Time info >>>> elapsed: 172.36 mins remain: 369.66 mins + 04-04 00:15:45 | [318][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0462 ntime: 0078 mem: 3.36 + 04-04 00:15:46 | [318][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0177 ntime: 0083 mem: 3.36 + 04-04 00:15:48 | [318][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0223 ntime: 0080 mem: 3.36 + 04-04 00:15:51 | [318][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0123 ntime: 0087 mem: 3.36 + 04-04 00:15:53 | [318][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0081 mem: 3.36 + 04-04 00:15:55 | [318][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0164 ntime: 0076 mem: 3.36 + 04-04 00:15:57 | [318][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0140 ntime: 0086 mem: 3.36 + 04-04 00:15:59 | [318][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0082 mem: 3.36 + 04-04 00:16:01 | [318][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0120 ntime: 0078 mem: 3.36 + 04-04 00:16:03 | [318][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0076 mem: 3.36 + 04-04 00:16:04 | [318][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0120 ntime: 0082 mem: 3.36 + 04-04 00:16:07 | [318][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0090 ntime: 0082 mem: 3.36 + 04-04 00:16:09 | [318][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0436 ntime: 0080 mem: 3.36 + 04-04 00:16:11 | [318][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0172 ntime: 0080 mem: 3.36 + 04-04 00:16:13 | [318][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0082 mem: 3.36 + 04-04 00:16:15 | [318][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0078 mem: 3.36 + 04-04 00:16:17 | [318][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0075 mem: 3.36 + 04-04 00:16:19 | [318][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0188 ntime: 0080 mem: 3.36 + 04-04 00:16:21 | Time info >>>> elapsed: 172.97 mins remain: 369.26 mins + 04-04 00:16:21 | [319][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0259 ntime: 0076 mem: 3.36 + 04-04 00:16:24 | [319][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1300 ntime: 0081 mem: 3.36 + 04-04 00:16:26 | [319][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0304 ntime: 0088 mem: 3.36 + 04-04 00:16:28 | [319][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0091 ntime: 0086 mem: 3.36 + 04-04 00:16:30 | [319][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0215 ntime: 0080 mem: 3.36 + 04-04 00:16:32 | [319][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0079 mem: 3.36 + 04-04 00:16:34 | [319][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0081 mem: 3.36 + 04-04 00:16:37 | [319][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0086 mem: 3.36 + 04-04 00:16:39 | [319][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0076 mem: 3.36 + 04-04 00:16:41 | [319][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0081 mem: 3.36 + 04-04 00:16:43 | [319][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0086 mem: 3.36 + 04-04 00:16:46 | [319][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0218 ntime: 0080 mem: 3.36 + 04-04 00:16:49 | [319][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0416 ntime: 0084 mem: 3.36 + 04-04 00:16:50 | [319][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0076 mem: 3.36 + 04-04 00:16:52 | [319][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0067 ntime: 0077 mem: 3.36 + 04-04 00:16:54 | [319][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0230 ntime: 0080 mem: 3.36 + 04-04 00:16:57 | [319][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0080 mem: 3.36 + 04-04 00:16:59 | [319][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0111 ntime: 0076 mem: 3.36 + 04-04 00:17:01 | Time info >>>> elapsed: 173.65 mins remain: 369.00 mins + 04-04 00:17:01 | [320][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0090 mem: 3.36 + 04-04 00:17:04 | [320][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0086 ntime: 0079 mem: 3.36 + 04-04 00:17:06 | [320][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0168 ntime: 0078 mem: 3.36 + 04-04 00:17:09 | [320][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0081 mem: 3.36 + 04-04 00:17:11 | [320][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0285 ntime: 0081 mem: 3.36 + 04-04 00:17:13 | [320][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0078 mem: 3.36 + 04-04 00:17:17 | [320][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0525 ntime: 0077 mem: 3.36 + 04-04 00:17:20 | [320][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0081 mem: 3.36 + 04-04 00:17:22 | [320][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0290 ntime: 0081 mem: 3.36 + 04-04 00:17:24 | [320][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0152 ntime: 0087 mem: 3.36 + 04-04 00:17:26 | [320][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0075 mem: 3.36 + 04-04 00:17:28 | [320][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0085 mem: 3.36 + 04-04 00:17:30 | [320][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0079 ntime: 0076 mem: 3.36 + 04-04 00:17:32 | [320][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0081 mem: 3.36 + 04-04 00:17:35 | [320][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0079 mem: 3.36 + 04-04 00:17:37 | [320][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0041 ntime: 0056 mem: 3.36 + 04-04 00:17:39 | [320][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0079 mem: 3.36 + 04-04 00:17:42 | [320][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 00:17:43 | Time info >>>> elapsed: 174.35 mins remain: 368.80 mins + 04-04 00:17:44 | [321][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0088 ntime: 0080 mem: 3.36 + 04-04 00:17:46 | [321][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0258 ntime: 0075 mem: 3.36 + 04-04 00:17:49 | [321][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0080 mem: 3.36 + 04-04 00:17:51 | [321][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 00:17:54 | [321][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0077 mem: 3.36 + 04-04 00:17:56 | [321][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0374 ntime: 0088 mem: 3.36 + 04-04 00:17:58 | [321][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0214 ntime: 0078 mem: 3.36 + 04-04 00:18:00 | [321][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0075 mem: 3.36 + 04-04 00:18:02 | [321][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0159 ntime: 0075 mem: 3.36 + 04-04 00:18:04 | [321][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0087 mem: 3.36 + 04-04 00:18:07 | [321][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0082 mem: 3.36 + 04-04 00:18:09 | [321][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0074 mem: 3.36 + 04-04 00:18:11 | [321][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 00:18:13 | [321][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0085 mem: 3.36 + 04-04 00:18:16 | [321][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0213 ntime: 0081 mem: 3.36 + 04-04 00:18:19 | [321][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0202 ntime: 0083 mem: 3.36 + 04-04 00:18:21 | [321][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 00:18:23 | [321][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0079 ntime: 0075 mem: 3.36 + 04-04 00:18:25 | Time info >>>> elapsed: 175.04 mins remain: 368.57 mins + 04-04 00:18:25 | [322][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0145 ntime: 0079 mem: 3.36 + 04-04 00:18:29 | [322][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0535 ntime: 0080 mem: 3.36 + 04-04 00:18:31 | [322][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0085 mem: 3.36 + 04-04 00:18:33 | [322][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0114 ntime: 0074 mem: 3.36 + 04-04 00:18:34 | [322][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0073 mem: 3.36 + 04-04 00:18:36 | [322][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0074 mem: 3.36 + 04-04 00:18:39 | [322][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0245 ntime: 0085 mem: 3.36 + 04-04 00:18:41 | [322][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0082 mem: 3.36 + 04-04 00:18:43 | [322][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0070 ntime: 0079 mem: 3.36 + 04-04 00:18:45 | [322][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0173 ntime: 0085 mem: 3.36 + 04-04 00:18:48 | [322][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0074 mem: 3.36 + 04-04 00:18:50 | [322][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0085 mem: 3.36 + 04-04 00:18:52 | [322][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 00:18:54 | [322][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0158 ntime: 0078 mem: 3.36 + 04-04 00:18:57 | [322][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0189 ntime: 0078 mem: 3.36 + 04-04 00:18:59 | [322][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 00:19:01 | [322][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0153 ntime: 0085 mem: 3.36 + 04-04 00:19:03 | [322][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0080 mem: 3.36 + 04-04 00:19:05 | Time info >>>> elapsed: 175.71 mins remain: 368.29 mins + 04-04 00:19:05 | [323][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0053 ntime: 0077 mem: 3.36 + 04-04 00:19:08 | [323][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0085 mem: 3.36 + 04-04 00:19:10 | [323][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 00:19:12 | [323][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0078 ntime: 0080 mem: 3.36 + 04-04 00:19:14 | [323][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0392 ntime: 0078 mem: 3.36 + 04-04 00:19:16 | [323][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0324 ntime: 0076 mem: 3.36 + 04-04 00:19:18 | [323][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0086 ntime: 0078 mem: 3.36 + 04-04 00:19:20 | [323][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 00:19:22 | [323][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0203 ntime: 0081 mem: 3.36 + 04-04 00:19:24 | [323][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0074 mem: 3.36 + 04-04 00:19:26 | [323][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0134 ntime: 0081 mem: 3.36 + 04-04 00:19:28 | [323][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0116 ntime: 0081 mem: 3.36 + 04-04 00:19:30 | [323][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0083 mem: 3.36 + 04-04 00:19:32 | [323][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0151 ntime: 0082 mem: 3.36 + 04-04 00:19:34 | [323][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0077 mem: 3.36 + 04-04 00:19:35 | [323][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0079 mem: 3.36 + 04-04 00:19:37 | [323][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 00:19:40 | [323][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0086 mem: 3.36 + 04-04 00:19:41 | Time info >>>> elapsed: 176.31 mins remain: 367.86 mins + 04-04 00:19:41 | [324][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 00:19:44 | [324][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0299 ntime: 0085 mem: 3.36 + 04-04 00:19:46 | [324][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0083 mem: 3.36 + 04-04 00:19:48 | [324][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0084 mem: 3.36 + 04-04 00:19:50 | [324][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0229 ntime: 0082 mem: 3.36 + 04-04 00:19:52 | [324][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0346 ntime: 0083 mem: 3.36 + 04-04 00:19:54 | [324][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0083 ntime: 0081 mem: 3.36 + 04-04 00:19:57 | [324][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 00:19:59 | [324][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0223 ntime: 0083 mem: 3.36 + 04-04 00:20:02 | [324][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0084 mem: 3.36 + 04-04 00:20:04 | [324][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0080 mem: 3.36 + 04-04 00:20:07 | [324][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0077 mem: 3.36 + 04-04 00:20:09 | [324][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0233 ntime: 0086 mem: 3.36 + 04-04 00:20:11 | [324][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 00:20:13 | [324][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0078 mem: 3.36 + 04-04 00:20:17 | [324][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0067 ntime: 0076 mem: 3.36 + 04-04 00:20:19 | [324][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0082 mem: 3.36 + 04-04 00:20:21 | [324][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0087 mem: 3.36 + 04-04 00:20:24 | Time info >>>> elapsed: 177.02 mins remain: 367.66 mins + 04-04 00:20:24 | [325][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 00:20:27 | [325][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0045 ntime: 0057 mem: 3.36 + 04-04 00:20:29 | [325][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 00:20:31 | [325][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0297 ntime: 0079 mem: 3.36 + 04-04 00:20:34 | [325][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0284 ntime: 0083 mem: 3.36 + 04-04 00:20:36 | [325][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0096 ntime: 0078 mem: 3.36 + 04-04 00:20:38 | [325][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0085 mem: 3.36 + 04-04 00:20:40 | [325][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0082 mem: 3.36 + 04-04 00:20:44 | [325][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0659 ntime: 0073 mem: 3.36 + 04-04 00:20:46 | [325][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0199 ntime: 0083 mem: 3.36 + 04-04 00:20:49 | [325][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0082 mem: 3.36 + 04-04 00:20:51 | [325][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0152 ntime: 0082 mem: 3.36 + 04-04 00:20:53 | [325][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0228 ntime: 0080 mem: 3.36 + 04-04 00:20:56 | [325][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0325 ntime: 0087 mem: 3.36 + 04-04 00:20:58 | [325][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0169 ntime: 0084 mem: 3.36 + 04-04 00:21:00 | [325][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0079 mem: 3.36 + 04-04 00:21:03 | [325][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0248 ntime: 0077 mem: 3.36 + 04-04 00:21:05 | [325][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0085 mem: 3.36 + 04-04 00:21:07 | Time info >>>> elapsed: 177.75 mins remain: 367.49 mins + 04-04 00:21:07 | [326][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0142 ntime: 0079 mem: 3.36 + 04-04 00:21:09 | [326][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0076 mem: 3.36 + 04-04 00:21:11 | [326][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0222 ntime: 0080 mem: 3.36 + 04-04 00:21:14 | [326][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0261 ntime: 0090 mem: 3.36 + 04-04 00:21:16 | [326][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 00:21:18 | [326][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0330 ntime: 0087 mem: 3.36 + 04-04 00:21:21 | [326][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0078 mem: 3.36 + 04-04 00:21:23 | [326][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0086 mem: 3.36 + 04-04 00:21:25 | [326][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 00:21:27 | [326][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0102 ntime: 0077 mem: 3.36 + 04-04 00:21:29 | [326][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0503 ntime: 0082 mem: 3.36 + 04-04 00:21:31 | [326][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0234 ntime: 0084 mem: 3.36 + 04-04 00:21:33 | [326][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0082 mem: 3.36 + 04-04 00:21:35 | [326][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0133 ntime: 0079 mem: 3.36 + 04-04 00:21:37 | [326][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0070 ntime: 0080 mem: 3.36 + 04-04 00:21:39 | [326][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0162 ntime: 0081 mem: 3.36 + 04-04 00:21:41 | [326][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0307 ntime: 0083 mem: 3.36 + 04-04 00:21:43 | [326][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0077 mem: 3.36 + 04-04 00:21:45 | Time info >>>> elapsed: 178.38 mins remain: 367.12 mins + 04-04 00:21:45 | [327][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0054 ntime: 0075 mem: 3.36 + 04-04 00:21:48 | [327][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0118 ntime: 0081 mem: 3.36 + 04-04 00:21:50 | [327][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 00:21:52 | [327][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 00:21:54 | [327][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0233 ntime: 0080 mem: 3.36 + 04-04 00:21:56 | [327][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0224 ntime: 0089 mem: 3.36 + 04-04 00:21:59 | [327][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0945 ntime: 0077 mem: 3.36 + 04-04 00:22:01 | [327][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0125 ntime: 0083 mem: 3.36 + 04-04 00:22:03 | [327][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0118 ntime: 0081 mem: 3.36 + 04-04 00:22:05 | [327][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0137 ntime: 0074 mem: 3.36 + 04-04 00:22:07 | [327][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0086 ntime: 0077 mem: 3.36 + 04-04 00:22:09 | [327][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0082 mem: 3.36 + 04-04 00:22:13 | [327][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0402 ntime: 0077 mem: 3.36 + 04-04 00:22:15 | [327][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0205 ntime: 0085 mem: 3.36 + 04-04 00:22:17 | [327][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0175 ntime: 0079 mem: 3.36 + 04-04 00:22:19 | [327][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0239 ntime: 0087 mem: 3.36 + 04-04 00:22:21 | [327][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0144 ntime: 0078 mem: 3.36 + 04-04 00:22:24 | [327][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0303 ntime: 0075 mem: 3.36 + 04-04 00:22:26 | Time info >>>> elapsed: 179.06 mins remain: 366.86 mins + 04-04 00:22:26 | [328][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0076 mem: 3.36 + 04-04 00:22:28 | [328][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0079 mem: 3.36 + 04-04 00:22:31 | [328][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0078 mem: 3.36 + 04-04 00:22:33 | [328][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0075 mem: 3.36 + 04-04 00:22:35 | [328][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0074 mem: 3.36 + 04-04 00:22:37 | [328][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0308 ntime: 0081 mem: 3.36 + 04-04 00:22:39 | [328][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0071 mem: 3.36 + 04-04 00:22:42 | [328][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0140 ntime: 0084 mem: 3.36 + 04-04 00:22:44 | [328][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0314 ntime: 0080 mem: 3.36 + 04-04 00:22:47 | [328][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0209 ntime: 0081 mem: 3.36 + 04-04 00:22:48 | [328][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0081 mem: 3.36 + 04-04 00:22:51 | [328][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0519 ntime: 0088 mem: 3.36 + 04-04 00:22:53 | [328][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0084 mem: 3.36 + 04-04 00:22:56 | [328][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0079 mem: 3.36 + 04-04 00:22:58 | [328][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0080 mem: 3.36 + 04-04 00:23:00 | [328][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0079 mem: 3.36 + 04-04 00:23:03 | [328][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0194 ntime: 0087 mem: 3.36 + 04-04 00:23:06 | [328][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0067 ntime: 0079 mem: 3.36 + 04-04 00:23:08 | Time info >>>> elapsed: 179.75 mins remain: 366.61 mins + 04-04 00:23:08 | [329][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0079 mem: 3.36 + 04-04 00:23:10 | [329][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0082 mem: 3.36 + 04-04 00:23:12 | [329][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0076 mem: 3.36 + 04-04 00:23:14 | [329][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0058 mem: 3.36 + 04-04 00:23:16 | [329][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0081 ntime: 0074 mem: 3.36 + 04-04 00:23:19 | [329][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0079 mem: 3.36 + 04-04 00:23:20 | [329][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0082 mem: 3.36 + 04-04 00:23:22 | [329][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0080 mem: 3.36 + 04-04 00:23:25 | [329][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 00:23:27 | [329][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0155 ntime: 0079 mem: 3.36 + 04-04 00:23:29 | [329][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0099 ntime: 0077 mem: 3.36 + 04-04 00:23:32 | [329][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0073 mem: 3.36 + 04-04 00:23:34 | [329][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 00:23:37 | [329][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0080 mem: 3.36 + 04-04 00:23:39 | [329][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0076 mem: 3.36 + 04-04 00:23:42 | [329][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0128 ntime: 0078 mem: 3.36 + 04-04 00:23:44 | [329][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0248 ntime: 0079 mem: 3.36 + 04-04 00:23:47 | [329][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0082 mem: 3.36 + 04-04 00:23:48 | Time info >>>> elapsed: 180.43 mins remain: 366.33 mins + 04-04 00:23:49 | [330][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0076 mem: 3.36 + 04-04 00:23:52 | [330][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0352 ntime: 0077 mem: 3.36 + 04-04 00:23:54 | [330][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0369 ntime: 0083 mem: 3.36 + 04-04 00:23:57 | [330][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0069 mem: 3.36 + 04-04 00:23:59 | [330][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0112 ntime: 0080 mem: 3.36 + 04-04 00:24:01 | [330][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0081 mem: 3.36 + 04-04 00:24:05 | [330][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0172 ntime: 0072 mem: 3.36 + 04-04 00:24:07 | [330][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 00:24:09 | [330][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0075 mem: 3.36 + 04-04 00:24:11 | [330][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0227 ntime: 0081 mem: 3.36 + 04-04 00:24:13 | [330][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0291 ntime: 0075 mem: 3.36 + 04-04 00:24:15 | [330][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0080 mem: 3.36 + 04-04 00:24:17 | [330][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0085 mem: 3.36 + 04-04 00:24:20 | [330][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0245 ntime: 0079 mem: 3.36 + 04-04 00:24:23 | [330][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0094 ntime: 0074 mem: 3.36 + 04-04 00:24:26 | [330][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0276 ntime: 0087 mem: 3.36 + 04-04 00:24:28 | [330][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0080 mem: 3.36 + 04-04 00:24:30 | [330][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0155 ntime: 0082 mem: 3.36 + 04-04 00:24:33 | Time info >>>> elapsed: 181.17 mins remain: 366.18 mins + 04-04 00:24:33 | [331][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0087 mem: 3.36 + 04-04 00:24:35 | [331][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 00:24:38 | [331][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0146 ntime: 0079 mem: 3.36 + 04-04 00:24:41 | [331][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0264 ntime: 0081 mem: 3.36 + 04-04 00:24:43 | [331][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0186 ntime: 0081 mem: 3.36 + 04-04 00:24:45 | [331][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0248 ntime: 0085 mem: 3.36 + 04-04 00:24:48 | [331][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0422 ntime: 0077 mem: 3.36 + 04-04 00:24:51 | [331][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0081 mem: 3.36 + 04-04 00:24:53 | [331][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0124 ntime: 0074 mem: 3.36 + 04-04 00:24:56 | [331][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 00:24:59 | [331][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0402 ntime: 0086 mem: 3.36 + 04-04 00:25:01 | [331][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0085 ntime: 0079 mem: 3.36 + 04-04 00:25:04 | [331][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 00:25:06 | [331][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0083 mem: 3.36 + 04-04 00:25:08 | [331][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0067 ntime: 0080 mem: 3.36 + 04-04 00:25:10 | [331][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0279 ntime: 0085 mem: 3.36 + 04-04 00:25:12 | [331][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0077 mem: 3.36 + 04-04 00:25:15 | [331][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0342 ntime: 0080 mem: 3.36 + 04-04 00:25:16 | Time info >>>> elapsed: 181.90 mins remain: 365.98 mins + 04-04 00:25:16 | [332][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0082 mem: 3.36 + 04-04 00:25:19 | [332][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0515 ntime: 0081 mem: 3.36 + 04-04 00:25:21 | [332][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0238 ntime: 0077 mem: 3.36 + 04-04 00:25:24 | [332][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0260 ntime: 0086 mem: 3.36 + 04-04 00:25:25 | [332][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0081 mem: 3.36 + 04-04 00:25:27 | [332][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0142 ntime: 0079 mem: 3.36 + 04-04 00:25:30 | [332][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0084 mem: 3.36 + 04-04 00:25:32 | [332][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0073 ntime: 0077 mem: 3.36 + 04-04 00:25:33 | [332][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0078 ntime: 0079 mem: 3.36 + 04-04 00:25:36 | [332][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 00:25:38 | [332][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0186 ntime: 0076 mem: 3.36 + 04-04 00:25:40 | [332][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0082 mem: 3.36 + 04-04 00:25:43 | [332][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0080 mem: 3.36 + 04-04 00:25:45 | [332][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 00:25:47 | [332][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0078 ntime: 0085 mem: 3.36 + 04-04 00:25:49 | [332][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0085 mem: 3.36 + 04-04 00:25:52 | [332][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0362 ntime: 0082 mem: 3.36 + 04-04 00:25:56 | [332][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0182 ntime: 0082 mem: 3.36 + 04-04 00:25:58 | Time info >>>> elapsed: 182.59 mins remain: 365.72 mins + 04-04 00:25:58 | [333][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0335 ntime: 0079 mem: 3.36 + 04-04 00:26:00 | [333][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0080 mem: 3.36 + 04-04 00:26:02 | [333][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0272 ntime: 0078 mem: 3.36 + 04-04 00:26:04 | [333][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0088 ntime: 0078 mem: 3.36 + 04-04 00:26:06 | [333][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0072 mem: 3.36 + 04-04 00:26:08 | [333][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0079 mem: 3.36 + 04-04 00:26:10 | [333][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 00:26:13 | [333][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 00:26:15 | [333][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0085 mem: 3.36 + 04-04 00:26:17 | [333][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 00:26:19 | [333][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0079 mem: 3.36 + 04-04 00:26:22 | [333][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0192 ntime: 0085 mem: 3.36 + 04-04 00:26:24 | [333][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0118 ntime: 0080 mem: 3.36 + 04-04 00:26:25 | [333][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0205 ntime: 0083 mem: 3.36 + 04-04 00:26:28 | [333][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0498 ntime: 0078 mem: 3.36 + 04-04 00:26:30 | [333][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0089 ntime: 0080 mem: 3.36 + 04-04 00:26:33 | [333][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0290 ntime: 0074 mem: 3.36 + 04-04 00:26:35 | [333][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0167 ntime: 0080 mem: 3.36 + 04-04 00:26:36 | Time info >>>> elapsed: 183.23 mins remain: 365.37 mins + 04-04 00:26:36 | [334][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0079 mem: 3.36 + 04-04 00:26:39 | [334][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0165 ntime: 0079 mem: 3.36 + 04-04 00:26:41 | [334][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0238 ntime: 0077 mem: 3.36 + 04-04 00:26:43 | [334][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0078 mem: 3.36 + 04-04 00:26:46 | [334][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0085 mem: 3.36 + 04-04 00:26:48 | [334][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0086 mem: 3.36 + 04-04 00:26:50 | [334][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0124 ntime: 0079 mem: 3.36 + 04-04 00:26:52 | [334][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0145 ntime: 0080 mem: 3.36 + 04-04 00:26:54 | [334][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0074 mem: 3.36 + 04-04 00:26:56 | [334][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0083 mem: 3.36 + 04-04 00:26:58 | [334][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0234 ntime: 0082 mem: 3.36 + 04-04 00:27:01 | [334][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0337 ntime: 0079 mem: 3.36 + 04-04 00:27:03 | [334][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 00:27:05 | [334][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0200 ntime: 0086 mem: 3.36 + 04-04 00:27:07 | [334][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0284 ntime: 0087 mem: 3.36 + 04-04 00:27:09 | [334][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0080 mem: 3.36 + 04-04 00:27:11 | [334][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0083 mem: 3.36 + 04-04 00:27:14 | [334][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0090 mem: 3.36 + 04-04 00:27:16 | Time info >>>> elapsed: 183.89 mins remain: 365.04 mins + 04-04 00:27:16 | [335][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 00:27:19 | [335][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0044 ntime: 0083 mem: 3.36 + 04-04 00:27:21 | [335][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0045 ntime: 0076 mem: 3.36 + 04-04 00:27:23 | [335][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 00:27:26 | [335][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0180 ntime: 0088 mem: 3.36 + 04-04 00:27:29 | [335][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0076 mem: 3.36 + 04-04 00:27:31 | [335][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0077 mem: 3.36 + 04-04 00:27:33 | [335][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0085 mem: 3.36 + 04-04 00:27:36 | [335][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0081 mem: 3.36 + 04-04 00:27:38 | [335][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0244 ntime: 0084 mem: 3.36 + 04-04 00:27:40 | [335][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0155 ntime: 0081 mem: 3.36 + 04-04 00:27:43 | [335][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0083 mem: 3.36 + 04-04 00:27:45 | [335][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0080 mem: 3.36 + 04-04 00:27:48 | [335][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0088 mem: 3.36 + 04-04 00:27:50 | [335][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 00:27:52 | [335][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0258 ntime: 0082 mem: 3.36 + 04-04 00:27:54 | [335][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0079 mem: 3.36 + 04-04 00:27:57 | [335][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0242 ntime: 0082 mem: 3.36 + 04-04 00:27:58 | Time info >>>> elapsed: 184.60 mins remain: 364.80 mins + 04-04 00:27:58 | [336][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0074 mem: 3.36 + 04-04 00:28:01 | [336][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0199 ntime: 0079 mem: 3.36 + 04-04 00:28:03 | [336][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0201 ntime: 0082 mem: 3.36 + 04-04 00:28:06 | [336][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0486 ntime: 0086 mem: 3.36 + 04-04 00:28:08 | [336][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0076 mem: 3.36 + 04-04 00:28:10 | [336][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0224 ntime: 0083 mem: 3.36 + 04-04 00:28:12 | [336][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0084 mem: 3.36 + 04-04 00:28:14 | [336][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0235 ntime: 0088 mem: 3.36 + 04-04 00:28:16 | [336][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0081 mem: 3.36 + 04-04 00:28:19 | [336][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0085 mem: 3.36 + 04-04 00:28:21 | [336][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0084 mem: 3.36 + 04-04 00:28:23 | [336][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0067 ntime: 0085 mem: 3.36 + 04-04 00:28:25 | [336][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0097 ntime: 0081 mem: 3.36 + 04-04 00:28:28 | [336][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0082 mem: 3.36 + 04-04 00:28:30 | [336][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0227 ntime: 0081 mem: 3.36 + 04-04 00:28:32 | [336][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0081 mem: 3.36 + 04-04 00:28:34 | [336][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0212 ntime: 0087 mem: 3.36 + 04-04 00:28:37 | [336][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0211 ntime: 0084 mem: 3.36 + 04-04 00:28:38 | Time info >>>> elapsed: 185.27 mins remain: 364.49 mins + 04-04 00:28:39 | [337][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0094 ntime: 0081 mem: 3.36 + 04-04 00:28:41 | [337][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0100 ntime: 0081 mem: 3.36 + 04-04 00:28:43 | [337][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0089 mem: 3.36 + 04-04 00:28:46 | [337][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0070 mem: 3.36 + 04-04 00:28:48 | [337][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0055 mem: 3.36 + 04-04 00:28:49 | [337][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0118 ntime: 0075 mem: 3.36 + 04-04 00:28:52 | [337][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0076 mem: 3.36 + 04-04 00:28:55 | [337][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0252 ntime: 0078 mem: 3.36 + 04-04 00:28:57 | [337][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 00:28:59 | [337][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0150 ntime: 0081 mem: 3.36 + 04-04 00:29:01 | [337][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0045 ntime: 0078 mem: 3.36 + 04-04 00:29:03 | [337][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0080 ntime: 0074 mem: 3.36 + 04-04 00:29:06 | [337][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0078 mem: 3.36 + 04-04 00:29:08 | [337][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0272 ntime: 0080 mem: 3.36 + 04-04 00:29:11 | [337][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0256 ntime: 0087 mem: 3.36 + 04-04 00:29:13 | [337][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0125 ntime: 0080 mem: 3.36 + 04-04 00:29:15 | [337][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 00:29:17 | [337][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 00:29:19 | Time info >>>> elapsed: 185.95 mins remain: 364.19 mins + 04-04 00:29:19 | [338][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0079 mem: 3.36 + 04-04 00:29:21 | [338][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0071 mem: 3.36 + 04-04 00:29:23 | [338][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0218 ntime: 0078 mem: 3.36 + 04-04 00:29:27 | [338][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0283 ntime: 0079 mem: 3.36 + 04-04 00:29:29 | [338][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0083 mem: 3.36 + 04-04 00:29:31 | [338][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0089 mem: 3.36 + 04-04 00:29:33 | [338][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0332 ntime: 0086 mem: 3.36 + 04-04 00:29:35 | [338][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0241 ntime: 0090 mem: 3.36 + 04-04 00:29:37 | [338][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0076 mem: 3.36 + 04-04 00:29:41 | [338][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 00:29:43 | [338][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0219 ntime: 0086 mem: 3.36 + 04-04 00:29:45 | [338][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0189 ntime: 0081 mem: 3.36 + 04-04 00:29:47 | [338][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0308 ntime: 0081 mem: 3.36 + 04-04 00:29:50 | [338][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0296 ntime: 0079 mem: 3.36 + 04-04 00:29:52 | [338][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0090 mem: 3.36 + 04-04 00:29:54 | [338][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 00:29:56 | [338][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0373 ntime: 0072 mem: 3.36 + 04-04 00:29:58 | [338][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0290 ntime: 0078 mem: 3.36 + 04-04 00:30:00 | Time info >>>> elapsed: 186.63 mins remain: 363.90 mins + 04-04 00:30:01 | [339][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0351 ntime: 0080 mem: 3.36 + 04-04 00:30:03 | [339][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0088 mem: 3.36 + 04-04 00:30:05 | [339][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0082 ntime: 0078 mem: 3.36 + 04-04 00:30:07 | [339][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0076 mem: 3.36 + 04-04 00:30:09 | [339][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0373 ntime: 0078 mem: 3.36 + 04-04 00:30:11 | [339][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0084 mem: 3.36 + 04-04 00:30:14 | [339][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0229 ntime: 0089 mem: 3.36 + 04-04 00:30:16 | [339][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0080 mem: 3.36 + 04-04 00:30:18 | [339][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0087 ntime: 0079 mem: 3.36 + 04-04 00:30:20 | [339][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0254 ntime: 0083 mem: 3.36 + 04-04 00:30:22 | [339][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0356 ntime: 0082 mem: 3.36 + 04-04 00:30:25 | [339][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0397 ntime: 0076 mem: 3.36 + 04-04 00:30:27 | [339][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0239 ntime: 0078 mem: 3.36 + 04-04 00:30:29 | [339][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0398 ntime: 0081 mem: 3.36 + 04-04 00:30:31 | [339][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 00:30:34 | [339][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0354 ntime: 0080 mem: 3.36 + 04-04 00:30:36 | [339][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0079 mem: 3.36 + 04-04 00:30:38 | [339][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0077 mem: 3.36 + 04-04 00:30:40 | Time info >>>> elapsed: 187.29 mins remain: 363.56 mins + 04-04 00:30:40 | [340][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0183 ntime: 0090 mem: 3.36 + 04-04 00:30:42 | [340][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0098 ntime: 0080 mem: 3.36 + 04-04 00:30:44 | [340][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0077 mem: 3.36 + 04-04 00:30:46 | [340][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 00:30:48 | [340][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0272 ntime: 0079 mem: 3.36 + 04-04 00:30:50 | [340][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0231 ntime: 0075 mem: 3.36 + 04-04 00:30:52 | [340][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0206 ntime: 0084 mem: 3.36 + 04-04 00:30:55 | [340][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 00:30:57 | [340][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0313 ntime: 0082 mem: 3.36 + 04-04 00:30:59 | [340][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0076 mem: 3.36 + 04-04 00:31:01 | [340][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0088 ntime: 0082 mem: 3.36 + 04-04 00:31:04 | [340][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0075 mem: 3.36 + 04-04 00:31:06 | [340][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0201 ntime: 0070 mem: 3.36 + 04-04 00:31:08 | [340][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0111 ntime: 0073 mem: 3.36 + 04-04 00:31:10 | [340][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0084 mem: 3.36 + 04-04 00:31:13 | [340][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0077 mem: 3.36 + 04-04 00:31:15 | [340][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0067 ntime: 0079 mem: 3.36 + 04-04 00:31:18 | [340][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0208 ntime: 0078 mem: 3.36 + 04-04 00:31:20 | Time info >>>> elapsed: 187.96 mins remain: 363.24 mins + 04-04 00:31:20 | [341][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0107 ntime: 0084 mem: 3.36 + 04-04 00:31:23 | [341][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0083 mem: 3.36 + 04-04 00:31:25 | [341][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0236 ntime: 0077 mem: 3.36 + 04-04 00:31:28 | [341][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0083 mem: 3.36 + 04-04 00:31:31 | [341][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0076 mem: 3.36 + 04-04 00:31:33 | [341][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0147 ntime: 0083 mem: 3.36 + 04-04 00:31:35 | [341][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0229 ntime: 0062 mem: 3.36 + 04-04 00:31:38 | [341][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 00:31:40 | [341][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0078 mem: 3.36 + 04-04 00:31:42 | [341][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0138 ntime: 0085 mem: 3.36 + 04-04 00:31:44 | [341][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0078 mem: 3.36 + 04-04 00:31:47 | [341][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0270 ntime: 0080 mem: 3.36 + 04-04 00:31:49 | [341][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0109 ntime: 0078 mem: 3.36 + 04-04 00:31:51 | [341][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 00:31:54 | [341][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0465 ntime: 0078 mem: 3.36 + 04-04 00:31:56 | [341][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0224 ntime: 0078 mem: 3.36 + 04-04 00:31:58 | [341][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0244 ntime: 0080 mem: 3.36 + 04-04 00:32:00 | [341][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0293 ntime: 0081 mem: 3.36 + 04-04 00:32:02 | Time info >>>> elapsed: 188.67 mins remain: 362.99 mins + 04-04 00:32:03 | [342][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0218 ntime: 0077 mem: 3.36 + 04-04 00:32:05 | [342][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0089 mem: 3.36 + 04-04 00:32:08 | [342][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0086 mem: 3.36 + 04-04 00:32:10 | [342][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 00:32:13 | [342][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0078 mem: 3.36 + 04-04 00:32:15 | [342][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0220 ntime: 0081 mem: 3.36 + 04-04 00:32:17 | [342][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0094 ntime: 0081 mem: 3.36 + 04-04 00:32:19 | [342][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0089 mem: 3.36 + 04-04 00:32:21 | [342][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0076 mem: 3.36 + 04-04 00:32:24 | [342][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0088 ntime: 0079 mem: 3.36 + 04-04 00:32:26 | [342][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0087 mem: 3.36 + 04-04 00:32:28 | [342][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 00:32:31 | [342][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0366 ntime: 0087 mem: 3.36 + 04-04 00:32:33 | [342][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0252 ntime: 0082 mem: 3.36 + 04-04 00:32:35 | [342][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 00:32:37 | [342][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0173 ntime: 0085 mem: 3.36 + 04-04 00:32:39 | [342][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0152 ntime: 0083 mem: 3.36 + 04-04 00:32:41 | [342][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0081 mem: 3.36 + 04-04 00:32:43 | Time info >>>> elapsed: 189.35 mins remain: 362.68 mins + 04-04 00:32:43 | [343][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0247 ntime: 0079 mem: 3.36 + 04-04 00:32:46 | [343][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0747 ntime: 0084 mem: 3.36 + 04-04 00:32:48 | [343][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0081 mem: 3.36 + 04-04 00:32:51 | [343][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0317 ntime: 0081 mem: 3.36 + 04-04 00:32:52 | [343][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0085 mem: 3.36 + 04-04 00:32:55 | [343][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0079 mem: 3.36 + 04-04 00:32:57 | [343][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 00:33:00 | [343][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0084 mem: 3.36 + 04-04 00:33:02 | [343][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0080 mem: 3.36 + 04-04 00:33:05 | [343][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0119 ntime: 0080 mem: 3.36 + 04-04 00:33:07 | [343][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0082 mem: 3.36 + 04-04 00:33:10 | [343][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0373 ntime: 0084 mem: 3.36 + 04-04 00:33:13 | [343][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0370 ntime: 0075 mem: 3.36 + 04-04 00:33:15 | [343][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0206 ntime: 0082 mem: 3.36 + 04-04 00:33:17 | [343][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0083 mem: 3.36 + 04-04 00:33:20 | [343][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0182 ntime: 0090 mem: 3.36 + 04-04 00:33:23 | [343][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0142 ntime: 0089 mem: 3.36 + 04-04 00:33:25 | [343][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0082 mem: 3.36 + 04-04 00:33:27 | Time info >>>> elapsed: 190.08 mins remain: 362.48 mins + 04-04 00:33:27 | [344][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0053 ntime: 0088 mem: 3.36 + 04-04 00:33:31 | [344][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 00:33:34 | [344][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0072 mem: 3.36 + 04-04 00:33:36 | [344][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0079 mem: 3.36 + 04-04 00:33:38 | [344][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0098 ntime: 0080 mem: 3.36 + 04-04 00:33:40 | [344][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0083 mem: 3.36 + 04-04 00:33:42 | [344][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0157 ntime: 0084 mem: 3.36 + 04-04 00:33:44 | [344][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0074 mem: 3.36 + 04-04 00:33:47 | [344][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0489 ntime: 0076 mem: 3.36 + 04-04 00:33:49 | [344][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0362 ntime: 0085 mem: 3.36 + 04-04 00:33:52 | [344][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0079 mem: 3.36 + 04-04 00:33:54 | [344][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0082 mem: 3.36 + 04-04 00:33:56 | [344][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0179 ntime: 0089 mem: 3.36 + 04-04 00:33:58 | [344][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 00:34:00 | [344][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0226 ntime: 0082 mem: 3.36 + 04-04 00:34:02 | [344][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 00:34:05 | [344][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0107 ntime: 0083 mem: 3.36 + 04-04 00:34:07 | [344][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0085 ntime: 0080 mem: 3.36 + 04-04 00:34:08 | Time info >>>> elapsed: 190.76 mins remain: 362.17 mins + 04-04 00:34:08 | [345][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0249 ntime: 0081 mem: 3.36 + 04-04 00:34:10 | [345][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0080 mem: 3.36 + 04-04 00:34:12 | [345][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0162 ntime: 0080 mem: 3.36 + 04-04 00:34:15 | [345][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0168 ntime: 0085 mem: 3.36 + 04-04 00:34:17 | [345][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0204 ntime: 0081 mem: 3.36 + 04-04 00:34:19 | [345][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0132 ntime: 0085 mem: 3.36 + 04-04 00:34:21 | [345][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0081 mem: 3.36 + 04-04 00:34:23 | [345][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0201 ntime: 0070 mem: 3.36 + 04-04 00:34:25 | [345][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0092 ntime: 0081 mem: 3.36 + 04-04 00:34:27 | [345][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0078 mem: 3.36 + 04-04 00:34:29 | [345][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0075 mem: 3.36 + 04-04 00:34:32 | [345][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0280 ntime: 0079 mem: 3.36 + 04-04 00:34:34 | [345][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0081 mem: 3.36 + 04-04 00:34:37 | [345][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0295 ntime: 0077 mem: 3.36 + 04-04 00:34:40 | [345][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0540 ntime: 0080 mem: 3.36 + 04-04 00:34:41 | [345][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0045 ntime: 0079 mem: 3.36 + 04-04 00:34:44 | [345][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 00:34:47 | [345][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0071 mem: 3.36 + 04-04 00:34:48 | Time info >>>> elapsed: 191.43 mins remain: 361.84 mins + 04-04 00:34:49 | [346][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0060 ntime: 0075 mem: 3.36 + 04-04 00:34:51 | [346][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0122 ntime: 0078 mem: 3.36 + 04-04 00:34:53 | [346][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0092 ntime: 0079 mem: 3.36 + 04-04 00:34:55 | [346][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0182 ntime: 0082 mem: 3.36 + 04-04 00:34:57 | [346][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0105 ntime: 0078 mem: 3.36 + 04-04 00:34:59 | [346][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0076 mem: 3.36 + 04-04 00:35:01 | [346][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0515 ntime: 0076 mem: 3.36 + 04-04 00:35:04 | [346][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0081 mem: 3.36 + 04-04 00:35:06 | [346][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0202 ntime: 0086 mem: 3.36 + 04-04 00:35:08 | [346][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0173 ntime: 0085 mem: 3.36 + 04-04 00:35:10 | [346][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0081 ntime: 0077 mem: 3.36 + 04-04 00:35:12 | [346][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0077 mem: 3.36 + 04-04 00:35:16 | [346][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0135 ntime: 0083 mem: 3.36 + 04-04 00:35:17 | [346][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0074 mem: 3.36 + 04-04 00:35:20 | [346][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0083 mem: 3.36 + 04-04 00:35:22 | [346][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0175 ntime: 0090 mem: 3.36 + 04-04 00:35:24 | [346][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0178 ntime: 0078 mem: 3.36 + 04-04 00:35:27 | [346][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0076 mem: 3.36 + 04-04 00:35:28 | Time info >>>> elapsed: 192.10 mins remain: 361.50 mins + 04-04 00:35:29 | [347][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0310 ntime: 0083 mem: 3.36 + 04-04 00:35:31 | [347][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0474 ntime: 0073 mem: 3.36 + 04-04 00:35:33 | [347][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0118 ntime: 0082 mem: 3.36 + 04-04 00:35:35 | [347][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0076 mem: 3.36 + 04-04 00:35:37 | [347][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0300 ntime: 0081 mem: 3.36 + 04-04 00:35:40 | [347][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0311 ntime: 0077 mem: 3.36 + 04-04 00:35:42 | [347][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0089 mem: 3.36 + 04-04 00:35:44 | [347][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 00:35:46 | [347][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0243 ntime: 0075 mem: 3.36 + 04-04 00:35:48 | [347][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0085 ntime: 0079 mem: 3.36 + 04-04 00:35:50 | [347][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0082 mem: 3.36 + 04-04 00:35:53 | [347][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0278 ntime: 0089 mem: 3.36 + 04-04 00:35:56 | [347][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0578 ntime: 0081 mem: 3.36 + 04-04 00:35:58 | [347][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0079 mem: 3.36 + 04-04 00:36:01 | [347][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0082 mem: 3.36 + 04-04 00:36:03 | [347][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0089 mem: 3.36 + 04-04 00:36:05 | [347][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0089 ntime: 0078 mem: 3.36 + 04-04 00:36:07 | [347][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0075 mem: 3.36 + 04-04 00:36:09 | Time info >>>> elapsed: 192.77 mins remain: 361.17 mins + 04-04 00:36:09 | [348][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0110 ntime: 0087 mem: 3.36 + 04-04 00:36:11 | [348][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0292 ntime: 0079 mem: 3.36 + 04-04 00:36:14 | [348][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0077 ntime: 0083 mem: 3.36 + 04-04 00:36:16 | [348][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0082 mem: 3.36 + 04-04 00:36:18 | [348][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0245 ntime: 0087 mem: 3.36 + 04-04 00:36:21 | [348][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0308 ntime: 0080 mem: 3.36 + 04-04 00:36:23 | [348][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0067 ntime: 0081 mem: 3.36 + 04-04 00:36:25 | [348][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0082 ntime: 0078 mem: 3.36 + 04-04 00:36:29 | [348][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0094 ntime: 0081 mem: 3.36 + 04-04 00:36:32 | [348][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0331 ntime: 0081 mem: 3.36 + 04-04 00:36:34 | [348][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0082 mem: 3.36 + 04-04 00:36:37 | [348][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0081 mem: 3.36 + 04-04 00:36:40 | [348][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0078 ntime: 0077 mem: 3.36 + 04-04 00:36:42 | [348][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0084 mem: 3.36 + 04-04 00:36:44 | [348][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0078 ntime: 0080 mem: 3.36 + 04-04 00:36:47 | [348][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0178 ntime: 0083 mem: 3.36 + 04-04 00:36:50 | [348][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0136 ntime: 0083 mem: 3.36 + 04-04 00:36:52 | [348][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0077 mem: 3.36 + 04-04 00:36:54 | Time info >>>> elapsed: 193.54 mins remain: 361.01 mins + 04-04 00:36:55 | [349][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0083 mem: 3.36 + 04-04 00:36:57 | [349][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0072 mem: 3.36 + 04-04 00:36:58 | [349][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 00:37:01 | [349][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0086 mem: 3.36 + 04-04 00:37:03 | [349][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0371 ntime: 0081 mem: 3.36 + 04-04 00:37:05 | [349][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0081 mem: 3.36 + 04-04 00:37:09 | [349][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0401 ntime: 0080 mem: 3.36 + 04-04 00:37:12 | [349][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0060 mem: 3.36 + 04-04 00:37:15 | [349][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0247 ntime: 0083 mem: 3.36 + 04-04 00:37:18 | [349][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0217 ntime: 0080 mem: 3.36 + 04-04 00:37:20 | [349][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0265 ntime: 0077 mem: 3.36 + 04-04 00:37:22 | [349][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0096 ntime: 0076 mem: 3.36 + 04-04 00:37:24 | [349][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0168 ntime: 0083 mem: 3.36 + 04-04 00:37:26 | [349][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0114 ntime: 0080 mem: 3.36 + 04-04 00:37:28 | [349][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0086 mem: 3.36 + 04-04 00:37:31 | [349][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0078 mem: 3.36 + 04-04 00:37:34 | [349][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0257 ntime: 0078 mem: 3.36 + 04-04 00:37:36 | [349][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0073 mem: 3.36 + 04-04 00:37:38 | Time info >>>> elapsed: 194.26 mins remain: 360.77 mins + 04-04 00:37:38 | [350][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0077 mem: 3.36 + 04-04 00:37:40 | [350][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0171 ntime: 0088 mem: 3.36 + 04-04 00:37:42 | [350][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0080 mem: 3.36 + 04-04 00:37:44 | [350][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0202 ntime: 0080 mem: 3.36 + 04-04 00:37:46 | [350][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-04 00:37:49 | [350][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-04 00:37:51 | [350][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0086 ntime: 0076 mem: 3.36 + 04-04 00:37:53 | [350][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0087 mem: 3.36 + 04-04 00:37:55 | [350][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0216 ntime: 0080 mem: 3.36 + 04-04 00:37:57 | [350][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0073 ntime: 0084 mem: 3.36 + 04-04 00:38:00 | [350][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0763 ntime: 0075 mem: 3.36 + 04-04 00:38:02 | [350][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 00:38:05 | [350][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0252 ntime: 0080 mem: 3.36 + 04-04 00:38:07 | [350][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0262 ntime: 0082 mem: 3.36 + 04-04 00:38:09 | [350][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0088 mem: 3.36 + 04-04 00:38:11 | [350][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0387 ntime: 0086 mem: 3.36 + 04-04 00:38:14 | [350][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0178 ntime: 0082 mem: 3.36 + 04-04 00:38:16 | [350][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 00:38:17 | Time info >>>> elapsed: 194.92 mins remain: 360.40 mins + 04-04 00:38:18 | [351][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0055 ntime: 0074 mem: 3.36 + 04-04 00:38:20 | [351][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0045 ntime: 0078 mem: 3.36 + 04-04 00:38:22 | [351][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0077 mem: 3.36 + 04-04 00:38:24 | [351][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0078 mem: 3.36 + 04-04 00:38:26 | [351][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0073 mem: 3.36 + 04-04 00:38:29 | [351][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0137 ntime: 0079 mem: 3.36 + 04-04 00:38:31 | [351][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0236 ntime: 0080 mem: 3.36 + 04-04 00:38:33 | [351][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 00:38:36 | [351][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0087 mem: 3.36 + 04-04 00:38:38 | [351][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 00:38:40 | [351][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0073 ntime: 0076 mem: 3.36 + 04-04 00:38:42 | [351][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0077 mem: 3.36 + 04-04 00:38:45 | [351][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 00:38:48 | [351][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0073 mem: 3.36 + 04-04 00:38:50 | [351][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0238 ntime: 0089 mem: 3.36 + 04-04 00:38:52 | [351][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0336 ntime: 0077 mem: 3.36 + 04-04 00:38:54 | [351][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0081 mem: 3.36 + 04-04 00:38:57 | [351][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0145 ntime: 0090 mem: 3.36 + 04-04 00:38:58 | Time info >>>> elapsed: 195.60 mins remain: 360.09 mins + 04-04 00:38:59 | [352][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0055 ntime: 0082 mem: 3.36 + 04-04 00:39:02 | [352][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0081 mem: 3.36 + 04-04 00:39:04 | [352][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0254 ntime: 0077 mem: 3.36 + 04-04 00:39:06 | [352][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0083 mem: 3.36 + 04-04 00:39:08 | [352][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0091 mem: 3.36 + 04-04 00:39:10 | [352][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 00:39:12 | [352][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 00:39:14 | [352][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0086 mem: 3.36 + 04-04 00:39:16 | [352][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0078 mem: 3.36 + 04-04 00:39:18 | [352][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0083 mem: 3.36 + 04-04 00:39:21 | [352][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0079 mem: 3.36 + 04-04 00:39:22 | [352][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0076 mem: 3.36 + 04-04 00:39:24 | [352][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 00:39:26 | [352][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0086 mem: 3.36 + 04-04 00:39:28 | [352][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0067 ntime: 0084 mem: 3.36 + 04-04 00:39:30 | [352][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0136 ntime: 0087 mem: 3.36 + 04-04 00:39:32 | [352][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0139 ntime: 0079 mem: 3.36 + 04-04 00:39:34 | [352][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0327 ntime: 0077 mem: 3.36 + 04-04 00:39:36 | Time info >>>> elapsed: 196.23 mins remain: 359.66 mins + 04-04 00:39:36 | [353][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0070 ntime: 0087 mem: 3.36 + 04-04 00:39:38 | [353][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0129 ntime: 0086 mem: 3.36 + 04-04 00:39:40 | [353][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0081 ntime: 0082 mem: 3.36 + 04-04 00:39:44 | [353][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0236 ntime: 0078 mem: 3.36 + 04-04 00:39:46 | [353][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0089 mem: 3.36 + 04-04 00:39:49 | [353][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0472 ntime: 0092 mem: 3.36 + 04-04 00:39:51 | [353][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0079 mem: 3.36 + 04-04 00:39:53 | [353][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0074 mem: 3.36 + 04-04 00:39:56 | [353][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 00:39:58 | [353][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0173 ntime: 0085 mem: 3.36 + 04-04 00:40:00 | [353][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0175 ntime: 0060 mem: 3.36 + 04-04 00:40:02 | [353][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0349 ntime: 0069 mem: 3.36 + 04-04 00:40:04 | [353][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0254 ntime: 0074 mem: 3.36 + 04-04 00:40:06 | [353][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0077 mem: 3.36 + 04-04 00:40:08 | [353][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0077 mem: 3.36 + 04-04 00:40:11 | [353][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0077 mem: 3.36 + 04-04 00:40:13 | [353][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 00:40:15 | [353][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0085 mem: 3.36 + 04-04 00:40:17 | Time info >>>> elapsed: 196.91 mins remain: 359.33 mins + 04-04 00:40:17 | [354][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0076 mem: 3.36 + 04-04 00:40:19 | [354][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0109 ntime: 0076 mem: 3.36 + 04-04 00:40:21 | [354][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0078 mem: 3.36 + 04-04 00:40:24 | [354][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0312 ntime: 0080 mem: 3.36 + 04-04 00:40:26 | [354][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0074 mem: 3.36 + 04-04 00:40:28 | [354][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0082 mem: 3.36 + 04-04 00:40:30 | [354][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0141 ntime: 0079 mem: 3.36 + 04-04 00:40:31 | [354][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0045 ntime: 0078 mem: 3.36 + 04-04 00:40:34 | [354][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0236 ntime: 0080 mem: 3.36 + 04-04 00:40:36 | [354][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0080 mem: 3.36 + 04-04 00:40:39 | [354][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0237 ntime: 0085 mem: 3.36 + 04-04 00:40:40 | [354][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0086 mem: 3.36 + 04-04 00:40:43 | [354][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0074 mem: 3.36 + 04-04 00:40:45 | [354][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0224 ntime: 0084 mem: 3.36 + 04-04 00:40:47 | [354][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0368 ntime: 0082 mem: 3.36 + 04-04 00:40:49 | [354][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 00:40:52 | [354][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0088 ntime: 0073 mem: 3.36 + 04-04 00:40:53 | [354][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0079 mem: 3.36 + 04-04 00:40:55 | Time info >>>> elapsed: 197.55 mins remain: 358.93 mins + 04-04 00:40:55 | [355][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0047 ntime: 0081 mem: 3.36 + 04-04 00:40:57 | [355][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0282 ntime: 0097 mem: 3.36 + 04-04 00:40:59 | [355][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0082 mem: 3.36 + 04-04 00:41:02 | [355][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 00:41:03 | [355][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0078 mem: 3.36 + 04-04 00:41:05 | [355][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0082 mem: 3.36 + 04-04 00:41:08 | [355][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0081 mem: 3.36 + 04-04 00:41:10 | [355][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0152 ntime: 0082 mem: 3.36 + 04-04 00:41:12 | [355][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0074 mem: 3.36 + 04-04 00:41:14 | [355][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0279 ntime: 0081 mem: 3.36 + 04-04 00:41:17 | [355][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0173 ntime: 0078 mem: 3.36 + 04-04 00:41:19 | [355][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 00:41:21 | [355][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0243 ntime: 0076 mem: 3.36 + 04-04 00:41:24 | [355][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0085 mem: 3.36 + 04-04 00:41:26 | [355][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0087 mem: 3.36 + 04-04 00:41:29 | [355][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0085 mem: 3.36 + 04-04 00:41:31 | [355][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 00:41:33 | [355][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0085 ntime: 0081 mem: 3.36 + 04-04 00:41:35 | Time info >>>> elapsed: 198.21 mins remain: 358.56 mins + 04-04 00:41:35 | [356][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0083 mem: 3.36 + 04-04 00:41:37 | [356][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0184 ntime: 0080 mem: 3.36 + 04-04 00:41:39 | [356][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0078 ntime: 0077 mem: 3.36 + 04-04 00:41:42 | [356][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0374 ntime: 0090 mem: 3.36 + 04-04 00:41:43 | [356][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0089 mem: 3.36 + 04-04 00:41:46 | [356][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0089 ntime: 0082 mem: 3.36 + 04-04 00:41:48 | [356][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0513 ntime: 0086 mem: 3.36 + 04-04 00:41:50 | [356][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0078 mem: 3.36 + 04-04 00:41:52 | [356][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0078 mem: 3.36 + 04-04 00:41:55 | [356][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 00:41:57 | [356][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0078 mem: 3.36 + 04-04 00:41:59 | [356][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 00:42:02 | [356][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0352 ntime: 0084 mem: 3.36 + 04-04 00:42:04 | [356][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0077 mem: 3.36 + 04-04 00:42:07 | [356][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0076 mem: 3.36 + 04-04 00:42:09 | [356][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0454 ntime: 0079 mem: 3.36 + 04-04 00:42:12 | [356][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 00:42:14 | [356][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0202 ntime: 0080 mem: 3.36 + 04-04 00:42:15 | Time info >>>> elapsed: 198.88 mins remain: 358.21 mins + 04-04 00:42:16 | [357][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0116 ntime: 0080 mem: 3.36 + 04-04 00:42:18 | [357][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0206 ntime: 0072 mem: 3.36 + 04-04 00:42:21 | [357][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0241 ntime: 0087 mem: 3.36 + 04-04 00:42:23 | [357][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0351 ntime: 0080 mem: 3.36 + 04-04 00:42:25 | [357][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0180 ntime: 0076 mem: 3.36 + 04-04 00:42:28 | [357][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0241 ntime: 0083 mem: 3.36 + 04-04 00:42:31 | [357][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0090 mem: 3.36 + 04-04 00:42:34 | [357][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 00:42:36 | [357][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0330 ntime: 0077 mem: 3.36 + 04-04 00:42:38 | [357][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0094 ntime: 0080 mem: 3.36 + 04-04 00:42:40 | [357][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0088 mem: 3.36 + 04-04 00:42:42 | [357][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0083 mem: 3.36 + 04-04 00:42:44 | [357][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0194 ntime: 0088 mem: 3.36 + 04-04 00:42:46 | [357][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 00:42:48 | [357][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0187 ntime: 0084 mem: 3.36 + 04-04 00:42:51 | [357][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0386 ntime: 0058 mem: 3.36 + 04-04 00:42:52 | [357][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0285 ntime: 0074 mem: 3.36 + 04-04 00:42:55 | [357][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-04 00:42:56 | Time info >>>> elapsed: 199.57 mins remain: 357.88 mins + 04-04 00:42:57 | [358][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0248 ntime: 0089 mem: 3.36 + 04-04 00:42:59 | [358][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0097 ntime: 0073 mem: 3.36 + 04-04 00:43:01 | [358][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0081 mem: 3.36 + 04-04 00:43:03 | [358][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0351 ntime: 0081 mem: 3.36 + 04-04 00:43:05 | [358][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 00:43:07 | [358][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 00:43:09 | [358][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0085 ntime: 0076 mem: 3.36 + 04-04 00:43:12 | [358][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0369 ntime: 0083 mem: 3.36 + 04-04 00:43:14 | [358][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0070 ntime: 0083 mem: 3.36 + 04-04 00:43:16 | [358][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0085 mem: 3.36 + 04-04 00:43:18 | [358][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0330 ntime: 0080 mem: 3.36 + 04-04 00:43:20 | [358][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0167 ntime: 0080 mem: 3.36 + 04-04 00:43:22 | [358][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0190 ntime: 0087 mem: 3.36 + 04-04 00:43:24 | [358][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0091 mem: 3.36 + 04-04 00:43:27 | [358][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0185 ntime: 0086 mem: 3.36 + 04-04 00:43:30 | [358][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0074 mem: 3.36 + 04-04 00:43:32 | [358][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0237 ntime: 0080 mem: 3.36 + 04-04 00:43:35 | [358][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0222 ntime: 0084 mem: 3.36 + 04-04 00:43:37 | Time info >>>> elapsed: 200.24 mins remain: 357.53 mins + 04-04 00:43:37 | [359][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0048 ntime: 0081 mem: 3.36 + 04-04 00:43:40 | [359][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0054 ntime: 0080 mem: 3.36 + 04-04 00:43:42 | [359][020/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0233 ntime: 0081 mem: 3.36 + 04-04 00:43:44 | [359][030/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0055 ntime: 0083 mem: 3.36 + 04-04 00:43:47 | [359][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0082 mem: 3.36 + 04-04 00:43:49 | [359][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0083 mem: 3.36 + 04-04 00:43:52 | [359][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0077 mem: 3.36 + 04-04 00:43:54 | [359][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0086 mem: 3.36 + 04-04 00:43:56 | [359][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0146 ntime: 0078 mem: 3.36 + 04-04 00:43:58 | [359][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0188 ntime: 0077 mem: 3.36 + 04-04 00:44:00 | [359][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 00:44:02 | [359][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0045 ntime: 0080 mem: 3.36 + 04-04 00:44:05 | [359][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0120 ntime: 0079 mem: 3.36 + 04-04 00:44:07 | [359][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0076 mem: 3.36 + 04-04 00:44:09 | [359][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0076 mem: 3.36 + 04-04 00:44:13 | [359][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0347 ntime: 0077 mem: 3.36 + 04-04 00:44:15 | [359][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0124 ntime: 0081 mem: 3.36 + 04-04 00:44:17 | [359][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0212 ntime: 0080 mem: 3.36 + 04-04 00:44:18 | Time info >>>> elapsed: 200.93 mins remain: 357.22 mins + 04-04 00:44:19 | [360][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0418 ntime: 0078 mem: 3.36 + 04-04 00:44:22 | [360][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0131 ntime: 0084 mem: 3.36 + 04-04 00:44:24 | [360][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0090 ntime: 0084 mem: 3.36 + 04-04 00:44:26 | [360][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0213 ntime: 0087 mem: 3.36 + 04-04 00:44:28 | [360][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0703 ntime: 0079 mem: 3.36 + 04-04 00:44:31 | [360][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0075 mem: 3.36 + 04-04 00:44:33 | [360][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0080 mem: 3.36 + 04-04 00:44:36 | [360][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0253 ntime: 0081 mem: 3.36 + 04-04 00:44:39 | [360][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0180 ntime: 0083 mem: 3.36 + 04-04 00:44:41 | [360][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0118 ntime: 0085 mem: 3.36 + 04-04 00:44:43 | [360][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0220 ntime: 0079 mem: 3.36 + 04-04 00:44:46 | [360][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0231 ntime: 0078 mem: 3.36 + 04-04 00:44:48 | [360][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0075 mem: 3.36 + 04-04 00:44:50 | [360][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0079 mem: 3.36 + 04-04 00:44:52 | [360][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0080 mem: 3.36 + 04-04 00:44:55 | [360][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0290 ntime: 0087 mem: 3.36 + 04-04 00:44:57 | [360][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0298 ntime: 0085 mem: 3.36 + 04-04 00:44:59 | [360][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0070 ntime: 0080 mem: 3.36 + 04-04 00:45:01 | Time info >>>> elapsed: 201.64 mins remain: 356.93 mins + 04-04 00:45:01 | [361][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0169 ntime: 0087 mem: 3.36 + 04-04 00:45:04 | [361][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0083 ntime: 0083 mem: 3.36 + 04-04 00:45:06 | [361][020/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 00:45:08 | [361][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0083 mem: 3.36 + 04-04 00:45:10 | [361][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0073 mem: 3.36 + 04-04 00:45:12 | [361][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 00:45:14 | [361][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0426 ntime: 0090 mem: 3.36 + 04-04 00:45:16 | [361][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0108 ntime: 0079 mem: 3.36 + 04-04 00:45:18 | [361][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0075 mem: 3.36 + 04-04 00:45:20 | [361][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0095 ntime: 0074 mem: 3.36 + 04-04 00:45:23 | [361][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0283 ntime: 0087 mem: 3.36 + 04-04 00:45:25 | [361][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0264 ntime: 0080 mem: 3.36 + 04-04 00:45:27 | [361][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0084 mem: 3.36 + 04-04 00:45:30 | [361][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0082 mem: 3.36 + 04-04 00:45:32 | [361][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0202 ntime: 0080 mem: 3.36 + 04-04 00:45:34 | [361][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0075 mem: 3.36 + 04-04 00:45:37 | [361][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0283 ntime: 0055 mem: 3.36 + 04-04 00:45:39 | [361][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 00:45:41 | Time info >>>> elapsed: 202.30 mins remain: 356.55 mins + 04-04 00:45:41 | [362][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0079 mem: 3.36 + 04-04 00:45:43 | [362][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0185 ntime: 0079 mem: 3.36 + 04-04 00:45:45 | [362][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0079 ntime: 0080 mem: 3.36 + 04-04 00:45:47 | [362][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0404 ntime: 0081 mem: 3.36 + 04-04 00:45:49 | [362][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0139 ntime: 0074 mem: 3.36 + 04-04 00:45:51 | [362][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0070 ntime: 0073 mem: 3.36 + 04-04 00:45:53 | [362][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0089 mem: 3.36 + 04-04 00:45:55 | [362][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0175 ntime: 0081 mem: 3.36 + 04-04 00:45:57 | [362][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0073 mem: 3.36 + 04-04 00:45:59 | [362][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0088 mem: 3.36 + 04-04 00:46:01 | [362][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0073 ntime: 0078 mem: 3.36 + 04-04 00:46:04 | [362][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0406 ntime: 0082 mem: 3.36 + 04-04 00:46:06 | [362][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0081 mem: 3.36 + 04-04 00:46:08 | [362][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0077 ntime: 0088 mem: 3.36 + 04-04 00:46:11 | [362][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0500 ntime: 0090 mem: 3.36 + 04-04 00:46:14 | [362][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0079 ntime: 0076 mem: 3.36 + 04-04 00:46:16 | [362][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0085 mem: 3.36 + 04-04 00:46:18 | [362][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0215 ntime: 0086 mem: 3.36 + 04-04 00:46:20 | Time info >>>> elapsed: 202.96 mins remain: 356.17 mins + 04-04 00:46:20 | [363][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0056 ntime: 0071 mem: 3.36 + 04-04 00:46:22 | [363][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0081 mem: 3.36 + 04-04 00:46:25 | [363][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0225 ntime: 0085 mem: 3.36 + 04-04 00:46:27 | [363][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0089 mem: 3.36 + 04-04 00:46:29 | [363][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0090 mem: 3.36 + 04-04 00:46:31 | [363][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0097 ntime: 0083 mem: 3.36 + 04-04 00:46:34 | [363][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0104 ntime: 0080 mem: 3.36 + 04-04 00:46:36 | [363][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0146 ntime: 0080 mem: 3.36 + 04-04 00:46:38 | [363][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0120 ntime: 0081 mem: 3.36 + 04-04 00:46:41 | [363][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0074 mem: 3.36 + 04-04 00:46:43 | [363][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0213 ntime: 0084 mem: 3.36 + 04-04 00:46:45 | [363][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0079 mem: 3.36 + 04-04 00:46:47 | [363][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0230 ntime: 0084 mem: 3.36 + 04-04 00:46:49 | [363][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0079 mem: 3.36 + 04-04 00:46:51 | [363][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0072 ntime: 0077 mem: 3.36 + 04-04 00:46:53 | [363][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0107 ntime: 0083 mem: 3.36 + 04-04 00:46:55 | [363][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0156 ntime: 0084 mem: 3.36 + 04-04 00:46:57 | [363][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0082 mem: 3.36 + 04-04 00:46:59 | Time info >>>> elapsed: 203.62 mins remain: 355.77 mins + 04-04 00:47:00 | [364][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0394 ntime: 0078 mem: 3.36 + 04-04 00:47:02 | [364][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0085 mem: 3.36 + 04-04 00:47:04 | [364][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0077 mem: 3.36 + 04-04 00:47:06 | [364][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0075 ntime: 0079 mem: 3.36 + 04-04 00:47:09 | [364][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0077 mem: 3.36 + 04-04 00:47:12 | [364][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0079 mem: 3.36 + 04-04 00:47:14 | [364][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0251 ntime: 0082 mem: 3.36 + 04-04 00:47:16 | [364][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0083 mem: 3.36 + 04-04 00:47:18 | [364][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0145 ntime: 0086 mem: 3.36 + 04-04 00:47:20 | [364][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0084 ntime: 0077 mem: 3.36 + 04-04 00:47:23 | [364][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0141 ntime: 0080 mem: 3.36 + 04-04 00:47:26 | [364][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0278 ntime: 0086 mem: 3.36 + 04-04 00:47:28 | [364][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0215 ntime: 0084 mem: 3.36 + 04-04 00:47:30 | [364][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0078 mem: 3.36 + 04-04 00:47:33 | [364][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0285 ntime: 0078 mem: 3.36 + 04-04 00:47:36 | [364][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0266 ntime: 0085 mem: 3.36 + 04-04 00:47:38 | [364][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0241 ntime: 0074 mem: 3.36 + 04-04 00:47:41 | [364][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0496 ntime: 0087 mem: 3.36 + 04-04 00:47:43 | Time info >>>> elapsed: 204.34 mins remain: 355.49 mins + 04-04 00:47:43 | [365][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 00:47:45 | [365][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0081 ntime: 0087 mem: 3.36 + 04-04 00:47:48 | [365][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0325 ntime: 0082 mem: 3.36 + 04-04 00:47:50 | [365][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0090 mem: 3.36 + 04-04 00:47:52 | [365][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0080 mem: 3.36 + 04-04 00:47:54 | [365][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0082 mem: 3.36 + 04-04 00:47:56 | [365][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0086 mem: 3.36 + 04-04 00:47:59 | [365][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0246 ntime: 0079 mem: 3.36 + 04-04 00:48:01 | [365][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0136 ntime: 0079 mem: 3.36 + 04-04 00:48:03 | [365][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 00:48:05 | [365][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0082 mem: 3.36 + 04-04 00:48:07 | [365][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0167 ntime: 0088 mem: 3.36 + 04-04 00:48:09 | [365][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0089 mem: 3.36 + 04-04 00:48:11 | [365][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0084 mem: 3.36 + 04-04 00:48:13 | [365][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0449 ntime: 0076 mem: 3.36 + 04-04 00:48:16 | [365][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0225 ntime: 0086 mem: 3.36 + 04-04 00:48:19 | [365][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0087 mem: 3.36 + 04-04 00:48:21 | [365][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0088 mem: 3.36 + 04-04 00:48:23 | Time info >>>> elapsed: 205.01 mins remain: 355.12 mins + 04-04 00:48:23 | [366][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0178 ntime: 0081 mem: 3.36 + 04-04 00:48:25 | [366][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0315 ntime: 0079 mem: 3.36 + 04-04 00:48:27 | [366][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0203 ntime: 0081 mem: 3.36 + 04-04 00:48:29 | [366][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 00:48:31 | [366][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0076 mem: 3.36 + 04-04 00:48:33 | [366][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0286 ntime: 0088 mem: 3.36 + 04-04 00:48:35 | [366][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0078 mem: 3.36 + 04-04 00:48:37 | [366][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0145 ntime: 0079 mem: 3.36 + 04-04 00:48:40 | [366][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0253 ntime: 0079 mem: 3.36 + 04-04 00:48:42 | [366][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0074 mem: 3.36 + 04-04 00:48:44 | [366][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0084 ntime: 0077 mem: 3.36 + 04-04 00:48:46 | [366][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0298 ntime: 0080 mem: 3.36 + 04-04 00:48:49 | [366][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0080 ntime: 0084 mem: 3.36 + 04-04 00:48:51 | [366][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0160 ntime: 0082 mem: 3.36 + 04-04 00:48:53 | [366][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0181 ntime: 0082 mem: 3.36 + 04-04 00:48:55 | [366][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0497 ntime: 0088 mem: 3.36 + 04-04 00:48:57 | [366][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0203 ntime: 0077 mem: 3.36 + 04-04 00:48:59 | [366][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0237 ntime: 0083 mem: 3.36 + 04-04 00:49:01 | Time info >>>> elapsed: 205.65 mins remain: 354.70 mins + 04-04 00:49:01 | [367][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 00:49:04 | [367][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0259 ntime: 0081 mem: 3.36 + 04-04 00:49:07 | [367][020/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0053 ntime: 0074 mem: 3.36 + 04-04 00:49:09 | [367][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0178 ntime: 0077 mem: 3.36 + 04-04 00:49:11 | [367][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0231 ntime: 0083 mem: 3.36 + 04-04 00:49:13 | [367][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0087 ntime: 0078 mem: 3.36 + 04-04 00:49:16 | [367][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0084 mem: 3.36 + 04-04 00:49:18 | [367][070/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0193 ntime: 0076 mem: 3.36 + 04-04 00:49:20 | [367][080/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 00:49:22 | [367][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0076 mem: 3.36 + 04-04 00:49:24 | [367][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0077 mem: 3.36 + 04-04 00:49:26 | [367][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 00:49:28 | [367][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0083 mem: 3.36 + 04-04 00:49:30 | [367][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 00:49:33 | [367][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0268 ntime: 0086 mem: 3.36 + 04-04 00:49:36 | [367][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0205 ntime: 0075 mem: 3.36 + 04-04 00:49:38 | [367][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0082 mem: 3.36 + 04-04 00:49:40 | [367][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0082 ntime: 0078 mem: 3.36 + 04-04 00:49:42 | Time info >>>> elapsed: 206.33 mins remain: 354.34 mins + 04-04 00:49:42 | [368][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0299 ntime: 0074 mem: 3.36 + 04-04 00:49:44 | [368][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0099 ntime: 0084 mem: 3.36 + 04-04 00:49:47 | [368][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 00:49:49 | [368][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 00:49:51 | [368][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0090 mem: 3.36 + 04-04 00:49:54 | [368][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0083 mem: 3.36 + 04-04 00:49:57 | [368][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0085 mem: 3.36 + 04-04 00:49:59 | [368][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 00:50:02 | [368][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0153 ntime: 0086 mem: 3.36 + 04-04 00:50:04 | [368][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0081 mem: 3.36 + 04-04 00:50:07 | [368][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0146 ntime: 0076 mem: 3.36 + 04-04 00:50:10 | [368][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1271 ntime: 0086 mem: 3.36 + 04-04 00:50:12 | [368][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0080 mem: 3.36 + 04-04 00:50:14 | [368][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 00:50:16 | [368][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0146 ntime: 0083 mem: 3.36 + 04-04 00:50:18 | [368][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0154 ntime: 0079 mem: 3.36 + 04-04 00:50:19 | [368][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0084 mem: 3.36 + 04-04 00:50:21 | [368][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0045 ntime: 0089 mem: 3.36 + 04-04 00:50:23 | Time info >>>> elapsed: 207.01 mins remain: 354.00 mins + 04-04 00:50:24 | [369][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0227 ntime: 0075 mem: 3.36 + 04-04 00:50:26 | [369][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0369 ntime: 0086 mem: 3.36 + 04-04 00:50:28 | [369][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0079 mem: 3.36 + 04-04 00:50:30 | [369][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0191 ntime: 0081 mem: 3.36 + 04-04 00:50:32 | [369][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0078 mem: 3.36 + 04-04 00:50:34 | [369][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0081 mem: 3.36 + 04-04 00:50:36 | [369][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0084 mem: 3.36 + 04-04 00:50:38 | [369][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0079 mem: 3.36 + 04-04 00:50:40 | [369][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0085 mem: 3.36 + 04-04 00:50:43 | [369][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 00:50:45 | [369][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0081 mem: 3.36 + 04-04 00:50:47 | [369][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0076 mem: 3.36 + 04-04 00:50:49 | [369][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0109 ntime: 0078 mem: 3.36 + 04-04 00:50:51 | [369][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0156 ntime: 0084 mem: 3.36 + 04-04 00:50:53 | [369][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0078 mem: 3.36 + 04-04 00:50:55 | [369][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0076 mem: 3.36 + 04-04 00:50:57 | [369][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0308 ntime: 0074 mem: 3.36 + 04-04 00:51:00 | [369][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0076 mem: 3.36 + 04-04 00:51:01 | Time info >>>> elapsed: 207.65 mins remain: 353.56 mins + 04-04 00:51:01 | [370][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0110 ntime: 0078 mem: 3.36 + 04-04 00:51:04 | [370][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 00:51:06 | [370][020/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0052 ntime: 0083 mem: 3.36 + 04-04 00:51:08 | [370][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0075 mem: 3.36 + 04-04 00:51:10 | [370][040/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0054 ntime: 0077 mem: 3.36 + 04-04 00:51:12 | [370][050/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0044 ntime: 0056 mem: 3.36 + 04-04 00:51:14 | [370][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0076 ntime: 0084 mem: 3.36 + 04-04 00:51:16 | [370][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0080 mem: 3.36 + 04-04 00:51:18 | [370][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0092 ntime: 0082 mem: 3.36 + 04-04 00:51:20 | [370][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0079 mem: 3.36 + 04-04 00:51:22 | [370][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0300 ntime: 0081 mem: 3.36 + 04-04 00:51:24 | [370][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0078 mem: 3.36 + 04-04 00:51:27 | [370][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0121 ntime: 0078 mem: 3.36 + 04-04 00:51:29 | [370][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0196 ntime: 0078 mem: 3.36 + 04-04 00:51:31 | [370][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 00:51:34 | [370][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 00:51:36 | [370][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0140 ntime: 0081 mem: 3.36 + 04-04 00:51:38 | [370][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0065 ntime: 0080 mem: 3.36 + 04-04 00:51:40 | Time info >>>> elapsed: 208.29 mins remain: 353.14 mins + 04-04 00:51:40 | [371][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0239 ntime: 0083 mem: 3.36 + 04-04 00:51:43 | [371][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0083 mem: 3.36 + 04-04 00:51:45 | [371][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 00:51:47 | [371][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0229 ntime: 0076 mem: 3.36 + 04-04 00:51:50 | [371][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0081 mem: 3.36 + 04-04 00:51:52 | [371][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0073 mem: 3.36 + 04-04 00:51:55 | [371][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0083 mem: 3.36 + 04-04 00:51:57 | [371][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0187 ntime: 0079 mem: 3.36 + 04-04 00:51:59 | [371][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0124 ntime: 0079 mem: 3.36 + 04-04 00:52:02 | [371][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0044 ntime: 0077 mem: 3.36 + 04-04 00:52:04 | [371][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0067 ntime: 0078 mem: 3.36 + 04-04 00:52:06 | [371][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0279 ntime: 0085 mem: 3.36 + 04-04 00:52:08 | [371][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0185 ntime: 0079 mem: 3.36 + 04-04 00:52:10 | [371][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0082 mem: 3.36 + 04-04 00:52:12 | [371][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0178 ntime: 0078 mem: 3.36 + 04-04 00:52:14 | [371][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0090 mem: 3.36 + 04-04 00:52:17 | [371][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0081 mem: 3.36 + 04-04 00:52:18 | [371][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0072 mem: 3.36 + 04-04 00:52:20 | Time info >>>> elapsed: 208.96 mins remain: 352.76 mins + 04-04 00:52:20 | [372][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0153 ntime: 0081 mem: 3.36 + 04-04 00:52:22 | [372][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0273 ntime: 0080 mem: 3.36 + 04-04 00:52:24 | [372][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0088 ntime: 0079 mem: 3.36 + 04-04 00:52:26 | [372][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 00:52:28 | [372][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0080 mem: 3.36 + 04-04 00:52:30 | [372][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0361 ntime: 0081 mem: 3.36 + 04-04 00:52:32 | [372][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0085 mem: 3.36 + 04-04 00:52:34 | [372][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0073 mem: 3.36 + 04-04 00:52:36 | [372][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0158 ntime: 0078 mem: 3.36 + 04-04 00:52:38 | [372][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0086 mem: 3.36 + 04-04 00:52:40 | [372][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0081 mem: 3.36 + 04-04 00:52:42 | [372][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0090 mem: 3.36 + 04-04 00:52:44 | [372][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0214 ntime: 0081 mem: 3.36 + 04-04 00:52:46 | [372][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 00:52:48 | [372][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0083 mem: 3.36 + 04-04 00:52:50 | [372][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0174 ntime: 0078 mem: 3.36 + 04-04 00:52:52 | [372][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0084 mem: 3.36 + 04-04 00:52:55 | [372][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0082 mem: 3.36 + 04-04 00:52:57 | Time info >>>> elapsed: 209.57 mins remain: 352.28 mins + 04-04 00:52:57 | [373][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0254 ntime: 0081 mem: 3.36 + 04-04 00:52:59 | [373][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0072 mem: 3.36 + 04-04 00:53:00 | [373][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0075 mem: 3.36 + 04-04 00:53:02 | [373][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0075 mem: 3.36 + 04-04 00:53:03 | [373][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0076 mem: 3.36 + 04-04 00:53:04 | [373][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0073 ntime: 0076 mem: 3.36 + 04-04 00:53:06 | [373][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0082 mem: 3.36 + 04-04 00:53:07 | [373][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0083 mem: 3.36 + 04-04 00:53:08 | [373][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0077 mem: 3.36 + 04-04 00:53:09 | [373][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0080 mem: 3.36 + 04-04 00:53:11 | [373][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0085 mem: 3.36 + 04-04 00:53:12 | [373][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 00:53:13 | [373][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0077 mem: 3.36 + 04-04 00:53:15 | [373][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0089 mem: 3.36 + 04-04 00:53:16 | [373][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0077 mem: 3.36 + 04-04 00:53:17 | [373][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0082 mem: 3.36 + 04-04 00:53:19 | [373][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0081 mem: 3.36 + 04-04 00:53:20 | [373][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 00:53:21 | Time info >>>> elapsed: 209.98 mins remain: 351.46 mins + 04-04 00:53:21 | [374][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0051 ntime: 0074 mem: 3.36 + 04-04 00:53:22 | [374][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0049 ntime: 0084 mem: 3.36 + 04-04 00:53:24 | [374][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0085 mem: 3.36 + 04-04 00:53:25 | [374][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0082 mem: 3.36 + 04-04 00:53:26 | [374][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0043 ntime: 0083 mem: 3.36 + 04-04 00:53:28 | [374][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 00:53:29 | [374][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0082 mem: 3.36 + 04-04 00:53:30 | [374][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0084 mem: 3.36 + 04-04 00:53:32 | [374][080/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0047 ntime: 0082 mem: 3.36 + 04-04 00:53:33 | [374][090/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0048 ntime: 0082 mem: 3.36 + 04-04 00:53:34 | [374][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0081 mem: 3.36 + 04-04 00:53:36 | [374][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0080 mem: 3.36 + 04-04 00:53:37 | [374][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0078 ntime: 0086 mem: 3.36 + 04-04 00:53:39 | [374][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0096 ntime: 0084 mem: 3.36 + 04-04 00:53:40 | [374][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0092 mem: 3.36 + 04-04 00:53:41 | [374][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 00:53:43 | [374][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0081 mem: 3.36 + 04-04 00:53:44 | [374][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0064 ntime: 0078 mem: 3.36 + 04-04 00:53:45 | Time info >>>> elapsed: 210.38 mins remain: 350.63 mins + 04-04 00:53:45 | [375][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0087 mem: 3.36 + 04-04 00:53:47 | [375][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0091 mem: 3.36 + 04-04 00:53:48 | [375][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0073 mem: 3.36 + 04-04 00:53:49 | [375][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0074 ntime: 0079 mem: 3.36 + 04-04 00:53:51 | [375][040/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0049 ntime: 0080 mem: 3.36 + 04-04 00:53:52 | [375][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0080 mem: 3.36 + 04-04 00:53:53 | [375][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0088 mem: 3.36 + 04-04 00:53:55 | [375][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0080 mem: 3.36 + 04-04 00:53:56 | [375][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0081 mem: 3.36 + 04-04 00:53:57 | [375][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0080 mem: 3.36 + 04-04 00:53:59 | [375][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0086 mem: 3.36 + 04-04 00:54:00 | [375][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0087 mem: 3.36 + 04-04 00:54:02 | [375][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0076 mem: 3.36 + 04-04 00:54:03 | [375][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 00:54:04 | [375][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0081 mem: 3.36 + 04-04 00:54:06 | [375][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0079 mem: 3.36 + 04-04 00:54:07 | [375][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0076 mem: 3.36 + 04-04 00:54:08 | [375][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0079 mem: 3.36 + 04-04 00:54:09 | Time info >>>> elapsed: 210.78 mins remain: 349.81 mins + 04-04 00:54:09 | [376][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0045 ntime: 0079 mem: 3.36 + 04-04 00:54:11 | [376][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0067 ntime: 0086 mem: 3.36 + 04-04 00:54:12 | [376][020/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0047 ntime: 0072 mem: 3.36 + 04-04 00:54:13 | [376][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0074 mem: 3.36 + 04-04 00:54:14 | [376][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0080 mem: 3.36 + 04-04 00:54:16 | [376][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0081 mem: 3.36 + 04-04 00:54:17 | [376][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0043 ntime: 0058 mem: 3.36 + 04-04 00:54:18 | [376][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 00:54:20 | [376][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0079 mem: 3.36 + 04-04 00:54:21 | [376][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0075 mem: 3.36 + 04-04 00:54:22 | [376][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0070 mem: 3.36 + 04-04 00:54:23 | [376][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0072 mem: 3.36 + 04-04 00:54:25 | [376][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0076 mem: 3.36 + 04-04 00:54:26 | [376][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0074 mem: 3.36 + 04-04 00:54:27 | [376][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0080 mem: 3.36 + 04-04 00:54:29 | [376][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0084 mem: 3.36 + 04-04 00:54:30 | [376][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 00:54:31 | [376][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0077 mem: 3.36 + 04-04 00:54:32 | Time info >>>> elapsed: 211.16 mins remain: 348.95 mins + 04-04 00:54:32 | [377][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0069 ntime: 0078 mem: 3.36 + 04-04 00:54:34 | [377][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0082 mem: 3.36 + 04-04 00:54:35 | [377][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0079 mem: 3.36 + 04-04 00:54:36 | [377][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0077 mem: 3.36 + 04-04 00:54:38 | [377][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0077 mem: 3.36 + 04-04 00:54:39 | [377][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0081 mem: 3.36 + 04-04 00:54:40 | [377][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0085 mem: 3.36 + 04-04 00:54:41 | [377][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0045 ntime: 0076 mem: 3.36 + 04-04 00:54:43 | [377][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0072 mem: 3.36 + 04-04 00:54:44 | [377][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 00:54:45 | [377][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 00:54:47 | [377][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0080 mem: 3.36 + 04-04 00:54:48 | [377][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0077 mem: 3.36 + 04-04 00:54:49 | [377][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0076 mem: 3.36 + 04-04 00:54:51 | [377][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0072 mem: 3.36 + 04-04 00:54:52 | [377][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0083 mem: 3.36 + 04-04 00:54:53 | [377][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 00:54:55 | [377][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0072 mem: 3.36 + 04-04 00:54:56 | Time info >>>> elapsed: 211.56 mins remain: 348.12 mins + 04-04 00:54:56 | [378][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 00:54:57 | [378][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0045 ntime: 0075 mem: 3.36 + 04-04 00:54:59 | [378][020/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0045 ntime: 0083 mem: 3.36 + 04-04 00:55:00 | [378][030/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0047 ntime: 0074 mem: 3.36 + 04-04 00:55:01 | [378][040/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0048 ntime: 0079 mem: 3.36 + 04-04 00:55:02 | [378][050/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 00:55:04 | [378][060/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 00:55:05 | [378][070/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0052 ntime: 0080 mem: 3.36 + 04-04 00:55:06 | [378][080/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0052 ntime: 0080 mem: 3.36 + 04-04 00:55:08 | [378][090/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0047 ntime: 0081 mem: 3.36 + 04-04 00:55:09 | [378][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0075 mem: 3.36 + 04-04 00:55:10 | [378][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0097 ntime: 0079 mem: 3.36 + 04-04 00:55:12 | [378][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0059 ntime: 0087 mem: 3.36 + 04-04 00:55:13 | [378][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0078 mem: 3.36 + 04-04 00:55:15 | [378][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0074 mem: 3.36 + 04-04 00:55:16 | [378][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0081 mem: 3.36 + 04-04 00:55:17 | [378][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 00:55:18 | [378][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0082 mem: 3.36 + 04-04 00:55:19 | Time info >>>> elapsed: 211.95 mins remain: 347.29 mins + 04-04 00:55:20 | [379][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 00:55:21 | [379][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0076 mem: 3.36 + 04-04 00:55:22 | [379][020/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 00:55:23 | [379][030/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 00:55:25 | [379][040/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0048 ntime: 0081 mem: 3.36 + 04-04 00:55:26 | [379][050/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0049 ntime: 0084 mem: 3.36 + 04-04 00:55:27 | [379][060/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0051 ntime: 0069 mem: 3.36 + 04-04 00:55:29 | [379][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0076 mem: 3.36 + 04-04 00:55:30 | [379][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0087 mem: 3.36 + 04-04 00:55:31 | [379][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0081 mem: 3.36 + 04-04 00:55:33 | [379][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 00:55:34 | [379][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0075 mem: 3.36 + 04-04 00:55:35 | [379][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 00:55:36 | [379][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 00:55:38 | [379][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0080 mem: 3.36 + 04-04 00:55:39 | [379][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0084 mem: 3.36 + 04-04 00:55:40 | [379][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0079 mem: 3.36 + 04-04 00:55:42 | [379][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0074 mem: 3.36 + 04-04 00:55:43 | Time info >>>> elapsed: 212.34 mins remain: 346.44 mins + 04-04 00:55:43 | [380][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 00:55:44 | [380][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0090 mem: 3.36 + 04-04 00:55:45 | [380][020/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0049 ntime: 0083 mem: 3.36 + 04-04 00:55:47 | [380][030/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0047 ntime: 0090 mem: 3.36 + 04-04 00:55:48 | [380][040/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0053 ntime: 0081 mem: 3.36 + 04-04 00:55:50 | [380][050/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 00:55:51 | [380][060/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 00:55:52 | [380][070/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 00:55:54 | [380][080/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0049 ntime: 0088 mem: 3.36 + 04-04 00:55:55 | [380][090/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0058 ntime: 0081 mem: 3.36 + 04-04 00:55:57 | [380][100/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0059 ntime: 0090 mem: 3.36 + 04-04 00:55:58 | [380][110/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0053 ntime: 0080 mem: 3.36 + 04-04 00:55:59 | [380][120/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0050 ntime: 0083 mem: 3.36 + 04-04 00:56:01 | [380][130/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0045 ntime: 0077 mem: 3.36 + 04-04 00:56:02 | [380][140/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0053 ntime: 0084 mem: 3.36 + 04-04 00:56:03 | [380][150/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 00:56:05 | [380][160/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0050 ntime: 0081 mem: 3.36 + 04-04 00:56:06 | [380][170/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 00:56:07 | Time info >>>> elapsed: 212.75 mins remain: 345.65 mins + 04-04 00:56:07 | [381][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0050 ntime: 0074 mem: 3.36 + 04-04 00:56:09 | [381][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 00:56:10 | [381][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 00:56:11 | [381][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0080 mem: 3.36 + 04-04 00:56:13 | [381][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0082 mem: 3.36 + 04-04 00:56:14 | [381][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0076 mem: 3.36 + 04-04 00:56:15 | [381][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0084 mem: 3.36 + 04-04 00:56:16 | [381][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0080 mem: 3.36 + 04-04 00:56:18 | [381][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0084 mem: 3.36 + 04-04 00:56:19 | [381][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0081 mem: 3.36 + 04-04 00:56:20 | [381][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0081 mem: 3.36 + 04-04 00:56:22 | [381][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0077 mem: 3.36 + 04-04 00:56:23 | [381][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 00:56:24 | [381][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-04 00:56:26 | [381][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0078 mem: 3.36 + 04-04 00:56:27 | [381][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0079 mem: 3.36 + 04-04 00:56:28 | [381][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 00:56:30 | [381][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0085 mem: 3.36 + 04-04 00:56:31 | Time info >>>> elapsed: 213.14 mins remain: 344.82 mins + 04-04 00:56:31 | [382][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0077 mem: 3.36 + 04-04 00:56:32 | [382][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 00:56:34 | [382][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0085 mem: 3.36 + 04-04 00:56:35 | [382][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0071 mem: 3.36 + 04-04 00:56:36 | [382][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0045 ntime: 0074 mem: 3.36 + 04-04 00:56:38 | [382][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0044 ntime: 0078 mem: 3.36 + 04-04 00:56:39 | [382][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 00:56:40 | [382][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0073 mem: 3.36 + 04-04 00:56:41 | [382][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0079 mem: 3.36 + 04-04 00:56:43 | [382][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 00:56:44 | [382][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0075 mem: 3.36 + 04-04 00:56:45 | [382][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0072 mem: 3.36 + 04-04 00:56:46 | [382][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 00:56:48 | [382][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0083 mem: 3.36 + 04-04 00:56:49 | [382][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0076 mem: 3.36 + 04-04 00:56:50 | [382][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-04 00:56:52 | [382][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 00:56:53 | [382][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0082 mem: 3.36 + 04-04 00:56:54 | Time info >>>> elapsed: 213.53 mins remain: 343.99 mins + 04-04 00:56:54 | [383][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0083 mem: 3.36 + 04-04 00:56:55 | [383][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 00:56:57 | [383][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 00:56:58 | [383][030/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0052 ntime: 0079 mem: 3.36 + 04-04 00:56:59 | [383][040/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0055 ntime: 0081 mem: 3.36 + 04-04 00:57:01 | [383][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0079 mem: 3.36 + 04-04 00:57:02 | [383][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0081 mem: 3.36 + 04-04 00:57:03 | [383][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0078 mem: 3.36 + 04-04 00:57:05 | [383][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0079 mem: 3.36 + 04-04 00:57:06 | [383][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0078 mem: 3.36 + 04-04 00:57:07 | [383][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0072 mem: 3.36 + 04-04 00:57:09 | [383][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0083 mem: 3.36 + 04-04 00:57:10 | [383][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 00:57:11 | [383][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0084 mem: 3.36 + 04-04 00:57:13 | [383][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0073 mem: 3.36 + 04-04 00:57:14 | [383][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0071 mem: 3.36 + 04-04 00:57:15 | [383][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0084 mem: 3.36 + 04-04 00:57:16 | [383][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0077 mem: 3.36 + 04-04 00:57:17 | Time info >>>> elapsed: 213.92 mins remain: 343.16 mins + 04-04 00:57:17 | [384][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 00:57:19 | [384][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0052 ntime: 0086 mem: 3.36 + 04-04 00:57:20 | [384][020/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0059 ntime: 0089 mem: 3.36 + 04-04 00:57:22 | [384][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0088 mem: 3.36 + 04-04 00:57:23 | [384][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0083 mem: 3.36 + 04-04 00:57:24 | [384][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0079 mem: 3.36 + 04-04 00:57:26 | [384][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0076 mem: 3.36 + 04-04 00:57:27 | [384][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-04 00:57:28 | [384][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0079 mem: 3.36 + 04-04 00:57:30 | [384][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0074 mem: 3.36 + 04-04 00:57:31 | [384][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0082 mem: 3.36 + 04-04 00:57:32 | [384][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0073 mem: 3.36 + 04-04 00:57:33 | [384][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0072 mem: 3.36 + 04-04 00:57:35 | [384][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0072 mem: 3.36 + 04-04 00:57:36 | [384][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0077 mem: 3.36 + 04-04 00:57:37 | [384][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0086 mem: 3.36 + 04-04 00:57:39 | [384][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0081 mem: 3.36 + 04-04 00:57:40 | [384][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0081 mem: 3.36 + 04-04 00:57:41 | Time info >>>> elapsed: 214.31 mins remain: 342.34 mins + 04-04 00:57:41 | [385][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0071 mem: 3.36 + 04-04 00:57:42 | [385][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0054 ntime: 0084 mem: 3.36 + 04-04 00:57:44 | [385][020/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 00:57:45 | [385][030/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0056 ntime: 0080 mem: 3.36 + 04-04 00:57:46 | [385][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0091 mem: 3.36 + 04-04 00:57:48 | [385][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0093 mem: 3.36 + 04-04 00:57:49 | [385][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0081 mem: 3.36 + 04-04 00:57:50 | [385][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 00:57:52 | [385][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0078 mem: 3.36 + 04-04 00:57:53 | [385][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0076 mem: 3.36 + 04-04 00:57:54 | [385][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0077 mem: 3.36 + 04-04 00:57:56 | [385][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0085 mem: 3.36 + 04-04 00:57:57 | [385][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0083 mem: 3.36 + 04-04 00:57:58 | [385][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 00:58:00 | [385][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0060 ntime: 0086 mem: 3.36 + 04-04 00:58:01 | [385][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 00:58:02 | [385][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0081 mem: 3.36 + 04-04 00:58:04 | [385][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 00:58:05 | Time info >>>> elapsed: 214.71 mins remain: 341.53 mins + 04-04 00:58:05 | [386][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0085 mem: 3.36 + 04-04 00:58:06 | [386][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0078 mem: 3.36 + 04-04 00:58:07 | [386][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0077 mem: 3.36 + 04-04 00:58:09 | [386][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0075 mem: 3.36 + 04-04 00:58:10 | [386][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0086 mem: 3.36 + 04-04 00:58:11 | [386][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0077 mem: 3.36 + 04-04 00:58:13 | [386][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0084 mem: 3.36 + 04-04 00:58:14 | [386][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0080 mem: 3.36 + 04-04 00:58:15 | [386][080/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0051 ntime: 0088 mem: 3.36 + 04-04 00:58:17 | [386][090/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0052 ntime: 0087 mem: 3.36 + 04-04 00:58:18 | [386][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0081 mem: 3.36 + 04-04 00:58:19 | [386][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0080 mem: 3.36 + 04-04 00:58:20 | [386][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0044 ntime: 0072 mem: 3.36 + 04-04 00:58:22 | [386][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 00:58:23 | [386][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 00:58:24 | [386][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 00:58:26 | [386][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0080 mem: 3.36 + 04-04 00:58:27 | [386][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0080 mem: 3.36 + 04-04 00:58:28 | Time info >>>> elapsed: 215.09 mins remain: 340.70 mins + 04-04 00:58:28 | [387][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0084 mem: 3.36 + 04-04 00:58:30 | [387][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0087 mem: 3.36 + 04-04 00:58:31 | [387][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0084 mem: 3.36 + 04-04 00:58:32 | [387][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0083 mem: 3.36 + 04-04 00:58:34 | [387][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0081 mem: 3.36 + 04-04 00:58:35 | [387][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0075 mem: 3.36 + 04-04 00:58:36 | [387][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0053 ntime: 0076 mem: 3.36 + 04-04 00:58:38 | [387][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 00:58:39 | [387][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0077 mem: 3.36 + 04-04 00:58:40 | [387][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 00:58:41 | [387][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 00:58:43 | [387][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0076 mem: 3.36 + 04-04 00:58:44 | [387][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-04 00:58:45 | [387][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0080 mem: 3.36 + 04-04 00:58:47 | [387][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0080 mem: 3.36 + 04-04 00:58:48 | [387][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0087 mem: 3.36 + 04-04 00:58:49 | [387][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0086 mem: 3.36 + 04-04 00:58:51 | [387][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 00:58:52 | Time info >>>> elapsed: 215.49 mins remain: 339.89 mins + 04-04 00:58:52 | [388][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 00:58:53 | [388][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 00:58:54 | [388][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0054 ntime: 0082 mem: 3.36 + 04-04 00:58:56 | [388][030/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0048 ntime: 0072 mem: 3.36 + 04-04 00:58:57 | [388][040/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0049 ntime: 0076 mem: 3.36 + 04-04 00:58:58 | [388][050/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0048 ntime: 0082 mem: 3.36 + 04-04 00:59:00 | [388][060/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0049 ntime: 0089 mem: 3.36 + 04-04 00:59:01 | [388][070/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0050 ntime: 0077 mem: 3.36 + 04-04 00:59:02 | [388][080/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0048 ntime: 0083 mem: 3.36 + 04-04 00:59:03 | [388][090/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0049 ntime: 0071 mem: 3.36 + 04-04 00:59:05 | [388][100/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0056 ntime: 0072 mem: 3.36 + 04-04 00:59:06 | [388][110/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0050 ntime: 0082 mem: 3.36 + 04-04 00:59:07 | [388][120/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0051 ntime: 0076 mem: 3.36 + 04-04 00:59:09 | [388][130/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0049 ntime: 0083 mem: 3.36 + 04-04 00:59:10 | [388][140/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0048 ntime: 0072 mem: 3.36 + 04-04 00:59:11 | [388][150/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 00:59:13 | [388][160/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 00:59:14 | [388][170/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0051 ntime: 0084 mem: 3.36 + 04-04 00:59:15 | Time info >>>> elapsed: 215.88 mins remain: 339.07 mins + 04-04 00:59:15 | [389][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0083 mem: 3.36 + 04-04 00:59:16 | [389][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0088 mem: 3.36 + 04-04 00:59:18 | [389][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 00:59:19 | [389][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0080 mem: 3.36 + 04-04 00:59:20 | [389][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0071 mem: 3.36 + 04-04 00:59:22 | [389][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0077 mem: 3.36 + 04-04 00:59:23 | [389][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 00:59:24 | [389][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 00:59:25 | [389][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 00:59:27 | [389][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 00:59:28 | [389][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0076 mem: 3.36 + 04-04 00:59:29 | [389][110/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0049 ntime: 0080 mem: 3.36 + 04-04 00:59:31 | [389][120/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 00:59:32 | [389][130/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0045 ntime: 0074 mem: 3.36 + 04-04 00:59:33 | [389][140/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0049 ntime: 0077 mem: 3.36 + 04-04 00:59:34 | [389][150/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 00:59:36 | [389][160/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0065 ntime: 0087 mem: 3.36 + 04-04 00:59:37 | [389][170/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 00:59:38 | Time info >>>> elapsed: 216.27 mins remain: 338.26 mins + 04-04 00:59:38 | [390][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0047 ntime: 0084 mem: 3.36 + 04-04 00:59:40 | [390][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0048 ntime: 0072 mem: 3.36 + 04-04 00:59:41 | [390][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 00:59:42 | [390][030/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0047 ntime: 0081 mem: 3.36 + 04-04 00:59:44 | [390][040/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 00:59:45 | [390][050/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0048 ntime: 0086 mem: 3.36 + 04-04 00:59:46 | [390][060/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0050 ntime: 0090 mem: 3.36 + 04-04 00:59:48 | [390][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0080 ntime: 0088 mem: 3.36 + 04-04 00:59:49 | [390][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0046 ntime: 0082 mem: 3.36 + 04-04 00:59:50 | [390][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0080 mem: 3.36 + 04-04 00:59:52 | [390][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0082 ntime: 0088 mem: 3.36 + 04-04 00:59:53 | [390][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0075 mem: 3.36 + 04-04 00:59:54 | [390][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0075 mem: 3.36 + 04-04 00:59:56 | [390][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 00:59:57 | [390][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0050 ntime: 0082 mem: 3.36 + 04-04 00:59:59 | [390][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0055 ntime: 0080 mem: 3.36 + 04-04 01:00:00 | [390][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0107 ntime: 0089 mem: 3.36 + 04-04 01:00:02 | [390][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 01:00:03 | Time info >>>> elapsed: 216.67 mins remain: 337.48 mins + 04-04 01:00:03 | [391][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0066 ntime: 0087 mem: 3.36 + 04-04 01:00:05 | [391][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0165 ntime: 0078 mem: 3.36 + 04-04 01:00:07 | [391][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0123 ntime: 0089 mem: 3.36 + 04-04 01:00:10 | [391][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0138 ntime: 0077 mem: 3.36 + 04-04 01:00:11 | [391][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0062 ntime: 0088 mem: 3.36 + 04-04 01:00:13 | [391][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0255 ntime: 0078 mem: 3.36 + 04-04 01:00:15 | [391][060/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0057 ntime: 0074 mem: 3.36 + 04-04 01:00:17 | [391][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0063 ntime: 0082 mem: 3.36 + 04-04 01:00:19 | [391][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0057 ntime: 0084 mem: 3.36 + 04-04 01:00:21 | [391][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0224 ntime: 0085 mem: 3.36 + 04-04 01:00:23 | [391][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0093 mem: 3.36 + 04-04 01:00:26 | [391][110/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0054 ntime: 0072 mem: 3.36 + 04-04 01:00:27 | [391][120/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0123 ntime: 0074 mem: 3.36 + 04-04 01:00:29 | [391][130/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0047 ntime: 0081 mem: 3.36 + 04-04 01:00:31 | [391][140/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0139 ntime: 0081 mem: 3.36 + 04-04 01:00:34 | [391][150/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0165 ntime: 0076 mem: 3.36 + 04-04 01:00:36 | [391][160/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0365 ntime: 0080 mem: 3.36 + 04-04 01:00:38 | [391][170/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0053 ntime: 0088 mem: 3.36 + 04-04 01:00:40 | Time info >>>> elapsed: 217.29 mins remain: 337.02 mins + 04-04 01:00:40 | [392][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0088 ntime: 0079 mem: 3.36 + 04-04 01:00:42 | [392][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0056 ntime: 0079 mem: 3.36 + 04-04 01:00:44 | [392][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0247 ntime: 0082 mem: 3.36 + 04-04 01:00:47 | [392][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0197 ntime: 0087 mem: 3.36 + 04-04 01:00:49 | [392][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0061 ntime: 0081 mem: 3.36 + 04-04 01:00:51 | [392][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0123 ntime: 0086 mem: 3.36 + 04-04 01:00:54 | [392][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0145 ntime: 0080 mem: 3.36 + 04-04 01:00:57 | [392][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0398 ntime: 0073 mem: 3.36 + 04-04 01:01:00 | [392][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0290 ntime: 0084 mem: 3.36 + 04-04 01:01:03 | [392][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0138 ntime: 0056 mem: 3.36 + 04-04 01:01:08 | [392][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0085 ntime: 0071 mem: 3.36 + 04-04 01:01:10 | [392][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0111 ntime: 0077 mem: 3.36 + 04-04 01:01:21 | [392][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1496 ntime: 0081 mem: 3.36 + 04-04 01:01:29 | [392][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0400 ntime: 0086 mem: 3.36 + 04-04 01:01:37 | [392][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0096 ntime: 0084 mem: 3.36 + 04-04 01:01:47 | [392][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1205 ntime: 0076 mem: 3.36 + 04-04 01:01:56 | [392][160/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0707 ntime: 0076 mem: 3.36 + 04-04 01:02:05 | [392][170/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1502 ntime: 0079 mem: 3.36 + 04-04 01:02:15 | Time info >>>> elapsed: 218.87 mins remain: 338.05 mins + 04-04 01:02:16 | [393][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0874 ntime: 0075 mem: 3.36 + 04-04 01:02:27 | [393][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1485 ntime: 0083 mem: 3.36 + 04-04 01:02:37 | [393][020/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0818 ntime: 0074 mem: 3.36 + 04-04 01:02:46 | [393][030/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1489 ntime: 0077 mem: 3.36 + 04-04 01:02:58 | [393][040/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1120 ntime: 0082 mem: 3.36 + 04-04 01:03:08 | [393][050/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1340 ntime: 0080 mem: 3.36 + 04-04 01:03:19 | [393][060/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1180 ntime: 0075 mem: 3.36 + 04-04 01:03:28 | [393][070/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0106 ntime: 0079 mem: 3.36 + 04-04 01:03:41 | [393][080/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1207 ntime: 0082 mem: 3.36 + 04-04 01:03:51 | [393][090/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1754 ntime: 0079 mem: 3.36 + 04-04 01:04:02 | [393][100/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0054 ntime: 0082 mem: 3.36 + 04-04 01:04:11 | [393][110/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0690 ntime: 0078 mem: 3.36 + 04-04 01:04:19 | [393][120/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0146 ntime: 0079 mem: 3.36 + 04-04 01:04:30 | [393][130/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1343 ntime: 0076 mem: 3.36 + 04-04 01:04:41 | [393][140/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1436 ntime: 0077 mem: 3.36 + 04-04 01:04:50 | [393][150/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0090 ntime: 0077 mem: 3.36 + 04-04 01:05:01 | [393][160/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1259 ntime: 0083 mem: 3.36 + 04-04 01:05:15 | [393][170/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1390 ntime: 0084 mem: 3.36 + 04-04 01:05:24 | Time info >>>> elapsed: 222.02 mins remain: 341.48 mins + 04-04 01:05:24 | [394][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0629 ntime: 0081 mem: 3.36 + 04-04 01:05:33 | [394][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1221 ntime: 0078 mem: 3.36 + 04-04 01:05:41 | [394][020/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0088 ntime: 0076 mem: 3.36 + 04-04 01:05:50 | [394][030/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1398 ntime: 0081 mem: 3.36 + 04-04 01:05:59 | [394][040/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1214 ntime: 0085 mem: 3.36 + 04-04 01:06:11 | [394][050/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1512 ntime: 0062 mem: 3.36 + 04-04 01:06:23 | [394][060/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1453 ntime: 0089 mem: 3.36 + 04-04 01:06:34 | [394][070/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0247 ntime: 0079 mem: 3.36 + 04-04 01:06:43 | [394][080/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1547 ntime: 0085 mem: 3.36 + 04-04 01:06:52 | [394][090/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1410 ntime: 0083 mem: 3.36 + 04-04 01:07:01 | [394][100/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0906 ntime: 0079 mem: 3.36 + 04-04 01:07:10 | [394][110/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0105 ntime: 0080 mem: 3.36 + 04-04 01:07:19 | [394][120/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 01:07:29 | [394][130/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0767 ntime: 0078 mem: 3.36 + 04-04 01:07:41 | [394][140/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1399 ntime: 0087 mem: 3.36 + 04-04 01:07:54 | [394][150/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1454 ntime: 0078 mem: 3.36 + 04-04 01:08:02 | [394][160/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0391 ntime: 0078 mem: 3.36 + 04-04 01:08:11 | [394][170/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1305 ntime: 0083 mem: 3.36 + 04-04 01:08:19 | Time info >>>> elapsed: 224.95 mins remain: 344.55 mins + 04-04 01:08:21 | [395][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1297 ntime: 0083 mem: 3.36 + 04-04 01:08:32 | [395][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1417 ntime: 0083 mem: 3.36 + 04-04 01:08:41 | [395][020/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0075 ntime: 0084 mem: 3.36 + 04-04 01:08:50 | [395][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1383 ntime: 0086 mem: 3.36 + 04-04 01:09:00 | [395][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0846 ntime: 0088 mem: 3.36 + 04-04 01:09:09 | [395][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1265 ntime: 0080 mem: 3.36 + 04-04 01:09:19 | [395][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1405 ntime: 0083 mem: 3.36 + 04-04 01:09:29 | [395][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1367 ntime: 0082 mem: 3.36 + 04-04 01:09:36 | [395][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0107 ntime: 0088 mem: 3.36 + 04-04 01:09:45 | [395][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1470 ntime: 0082 mem: 3.36 + 04-04 01:09:50 | [395][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0079 ntime: 0078 mem: 3.36 + 04-04 01:09:59 | [395][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1316 ntime: 0077 mem: 3.36 + 04-04 01:10:08 | [395][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0048 ntime: 0086 mem: 3.36 + 04-04 01:10:17 | [395][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1035 ntime: 0074 mem: 3.36 + 04-04 01:10:27 | [395][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1433 ntime: 0088 mem: 3.36 + 04-04 01:10:35 | [395][150/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0054 ntime: 0079 mem: 3.36 + 04-04 01:10:43 | [395][160/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1476 ntime: 0088 mem: 3.36 + 04-04 01:10:51 | [395][170/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1100 ntime: 0079 mem: 3.36 + 04-04 01:10:58 | Time info >>>> elapsed: 227.59 mins remain: 347.14 mins + 04-04 01:11:00 | [396][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1528 ntime: 0084 mem: 3.36 + 04-04 01:11:07 | [396][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0089 ntime: 0079 mem: 3.36 + 04-04 01:11:17 | [396][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1458 ntime: 0076 mem: 3.36 + 04-04 01:11:25 | [396][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1005 ntime: 0081 mem: 3.36 + 04-04 01:11:34 | [396][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0090 ntime: 0074 mem: 3.36 + 04-04 01:11:43 | [396][050/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1231 ntime: 0078 mem: 3.36 + 04-04 01:11:50 | [396][060/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0083 mem: 3.36 + 04-04 01:11:56 | [396][070/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0068 ntime: 0083 mem: 3.36 + 04-04 01:12:03 | [396][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0069 ntime: 0077 mem: 3.36 + 04-04 01:12:12 | [396][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0049 ntime: 0081 mem: 3.36 + 04-04 01:12:19 | [396][100/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0079 ntime: 0079 mem: 3.36 + 04-04 01:12:28 | [396][110/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0408 ntime: 0077 mem: 3.36 + 04-04 01:12:38 | [396][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0486 ntime: 0080 mem: 3.36 + 04-04 01:12:48 | [396][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1434 ntime: 0075 mem: 3.36 + 04-04 01:12:57 | [396][140/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1593 ntime: 0081 mem: 3.36 + 04-04 01:13:07 | [396][150/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1385 ntime: 0076 mem: 3.36 + 04-04 01:13:17 | [396][160/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1478 ntime: 0082 mem: 3.36 + 04-04 01:13:25 | [396][170/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0084 ntime: 0075 mem: 3.36 + 04-04 01:13:32 | Time info >>>> elapsed: 230.15 mins remain: 349.58 mins + 04-04 01:13:33 | [397][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1570 ntime: 0075 mem: 3.36 + 04-04 01:13:43 | [397][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1001 ntime: 0078 mem: 3.36 + 04-04 01:13:52 | [397][020/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0732 ntime: 0078 mem: 3.36 + 04-04 01:14:04 | [397][030/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1567 ntime: 0080 mem: 3.36 + 04-04 01:14:13 | [397][040/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1571 ntime: 0081 mem: 3.36 + 04-04 01:14:24 | [397][050/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1344 ntime: 0086 mem: 3.36 + 04-04 01:14:36 | [397][060/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1460 ntime: 0080 mem: 3.36 + 04-04 01:14:44 | [397][070/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1308 ntime: 0086 mem: 3.36 + 04-04 01:14:54 | [397][080/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1630 ntime: 0087 mem: 3.36 + 04-04 01:15:06 | [397][090/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1233 ntime: 0077 mem: 3.36 + 04-04 01:15:15 | [397][100/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0802 ntime: 0080 mem: 3.36 + 04-04 01:15:24 | [397][110/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0663 ntime: 0084 mem: 3.36 + 04-04 01:15:33 | [397][120/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0608 ntime: 0081 mem: 3.36 + 04-04 01:15:41 | [397][130/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1496 ntime: 0079 mem: 3.36 + 04-04 01:15:50 | [397][140/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1000 ntime: 0078 mem: 3.36 + 04-04 01:16:00 | [397][150/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1031 ntime: 0082 mem: 3.36 + 04-04 01:16:10 | [397][160/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0736 ntime: 0079 mem: 3.36 + 04-04 01:16:18 | [397][170/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0213 ntime: 0079 mem: 3.36 + 04-04 01:16:27 | Time info >>>> elapsed: 233.07 mins remain: 352.54 mins + 04-04 01:16:27 | [398][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0058 ntime: 0083 mem: 3.36 + 04-04 01:16:37 | [398][010/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0562 ntime: 0094 mem: 3.36 + 04-04 01:16:47 | [398][020/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0150 ntime: 0081 mem: 3.36 + 04-04 01:16:58 | [398][030/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0948 ntime: 0078 mem: 3.36 + 04-04 01:17:06 | [398][040/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0088 ntime: 0080 mem: 3.36 + 04-04 01:17:18 | [398][050/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1419 ntime: 0080 mem: 3.36 + 04-04 01:17:28 | [398][060/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1401 ntime: 0084 mem: 3.36 + 04-04 01:17:40 | [398][070/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1410 ntime: 0078 mem: 3.36 + 04-04 01:17:50 | [398][080/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1227 ntime: 0085 mem: 3.36 + 04-04 01:17:59 | [398][090/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0188 ntime: 0079 mem: 3.36 + 04-04 01:18:08 | [398][100/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0055 ntime: 0080 mem: 3.36 + 04-04 01:18:17 | [398][110/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0087 ntime: 0079 mem: 3.36 + 04-04 01:18:28 | [398][120/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1386 ntime: 0079 mem: 3.36 + 04-04 01:18:37 | [398][130/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1691 ntime: 0087 mem: 3.36 + 04-04 01:18:46 | [398][140/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1490 ntime: 0081 mem: 3.36 + 04-04 01:18:57 | [398][150/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1467 ntime: 0078 mem: 3.36 + 04-04 01:19:11 | [398][160/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1357 ntime: 0084 mem: 3.36 + 04-04 01:19:20 | [398][170/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1506 ntime: 0087 mem: 3.36 + 04-04 01:19:28 | Time info >>>> elapsed: 236.09 mins remain: 355.61 mins + 04-04 01:19:29 | [399][000/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1357 ntime: 0080 mem: 3.36 + 04-04 01:19:39 | [399][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0479 ntime: 0080 mem: 3.36 + 04-04 01:19:51 | [399][020/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1287 ntime: 0079 mem: 3.36 + 04-04 01:20:00 | [399][030/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1413 ntime: 0086 mem: 3.36 + 04-04 01:20:09 | [399][040/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0805 ntime: 0077 mem: 3.36 + 04-04 01:20:18 | [399][050/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0074 ntime: 0079 mem: 3.36 + 04-04 01:20:28 | [399][060/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1360 ntime: 0080 mem: 3.36 + 04-04 01:20:37 | [399][070/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1367 ntime: 0079 mem: 3.36 + 04-04 01:20:45 | [399][080/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1649 ntime: 0062 mem: 3.36 + 04-04 01:20:55 | [399][090/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1400 ntime: 0080 mem: 3.36 + 04-04 01:21:02 | [399][100/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0741 ntime: 0082 mem: 3.36 + 04-04 01:21:11 | [399][110/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1225 ntime: 0078 mem: 3.36 + 04-04 01:21:21 | [399][120/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0071 ntime: 0077 mem: 3.36 + 04-04 01:21:29 | [399][130/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0605 ntime: 0083 mem: 3.36 + 04-04 01:21:39 | [399][140/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1038 ntime: 0082 mem: 3.36 + 04-04 01:21:47 | [399][150/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1370 ntime: 0084 mem: 3.36 + 04-04 01:21:55 | [399][160/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0372 ntime: 0085 mem: 3.36 + 04-04 01:22:04 | [399][170/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0119 ntime: 0085 mem: 3.36 + 04-04 01:22:12 | Time info >>>> elapsed: 238.83 mins remain: 358.25 mins + 04-04 01:22:14 | [400][000/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1011 ntime: 0084 mem: 3.36 + 04-04 01:22:22 | [400][010/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1526 ntime: 0082 mem: 3.36 + 04-04 01:22:32 | [400][020/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0052 ntime: 0080 mem: 3.36 + 04-04 01:22:45 | [400][030/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1279 ntime: 0080 mem: 3.36 + 04-04 01:22:55 | [400][040/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 1539 ntime: 0079 mem: 3.36 + 04-04 01:23:05 | [400][050/179] predict_x0_loss: 0.009 glr: 5.0e-06 dtime: 0139 ntime: 0085 mem: 3.36 + 04-04 01:23:15 | [400][060/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0126 ntime: 0080 mem: 3.36 + 04-04 01:23:22 | [400][070/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0605 ntime: 0084 mem: 3.36 + 04-04 01:23:32 | [400][080/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1007 ntime: 0081 mem: 3.36 + 04-04 01:23:41 | [400][090/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1293 ntime: 0079 mem: 3.36 + 04-04 01:23:52 | [400][100/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0965 ntime: 0082 mem: 3.36 + 04-04 01:24:01 | [400][110/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1165 ntime: 0081 mem: 3.36 + 04-04 01:24:11 | [400][120/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1533 ntime: 0088 mem: 3.36 + 04-04 01:24:21 | [400][130/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1546 ntime: 0083 mem: 3.36 + 04-04 01:24:31 | [400][140/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1584 ntime: 0075 mem: 3.36 + 04-04 01:24:39 | [400][150/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1128 ntime: 0082 mem: 3.36 + 04-04 01:24:49 | [400][160/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 1466 ntime: 0071 mem: 3.36 + 04-04 01:24:59 | [400][170/179] predict_x0_loss: 0.008 glr: 5.0e-06 dtime: 0676 ntime: 0081 mem: 3.36 + 04-04 01:25:07 | Time info >>>> elapsed: 241.74 mins remain: 361.11 mins + 04-04 01:25:07 | [401][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 01:25:15 | [401][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1009 ntime: 0082 mem: 3.36 + 04-04 01:25:25 | [401][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1282 ntime: 0087 mem: 3.36 + 04-04 01:25:34 | [401][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1044 ntime: 0082 mem: 3.36 + 04-04 01:25:44 | [401][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0968 ntime: 0087 mem: 3.36 + 04-04 01:25:55 | [401][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1421 ntime: 0082 mem: 3.36 + 04-04 01:26:05 | [401][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0077 ntime: 0082 mem: 3.36 + 04-04 01:26:14 | [401][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0162 ntime: 0082 mem: 3.36 + 04-04 01:26:21 | [401][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1198 ntime: 0081 mem: 3.36 + 04-04 01:26:32 | [401][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1332 ntime: 0088 mem: 3.36 + 04-04 01:26:41 | [401][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1144 ntime: 0079 mem: 3.36 + 04-04 01:26:50 | [401][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0766 ntime: 0087 mem: 3.36 + 04-04 01:26:57 | [401][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0551 ntime: 0077 mem: 3.36 + 04-04 01:27:07 | [401][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1525 ntime: 0088 mem: 3.36 + 04-04 01:27:17 | [401][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0909 ntime: 0079 mem: 3.36 + 04-04 01:27:24 | [401][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0329 ntime: 0081 mem: 3.36 + 04-04 01:27:36 | [401][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0076 mem: 3.36 + 04-04 01:27:47 | [401][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0075 ntime: 0081 mem: 3.36 + 04-04 01:27:56 | Time info >>>> elapsed: 244.56 mins remain: 363.80 mins + 04-04 01:27:57 | [402][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1449 ntime: 0085 mem: 3.36 + 04-04 01:28:07 | [402][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1521 ntime: 0080 mem: 3.36 + 04-04 01:28:17 | [402][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1363 ntime: 0079 mem: 3.36 + 04-04 01:28:25 | [402][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0108 ntime: 0081 mem: 3.36 + 04-04 01:28:36 | [402][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1101 ntime: 0084 mem: 3.36 + 04-04 01:28:46 | [402][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1225 ntime: 0090 mem: 3.36 + 04-04 01:28:56 | [402][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1007 ntime: 0075 mem: 3.36 + 04-04 01:29:05 | [402][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0103 ntime: 0078 mem: 3.36 + 04-04 01:29:13 | [402][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0092 ntime: 0078 mem: 3.36 + 04-04 01:29:24 | [402][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1625 ntime: 0088 mem: 3.36 + 04-04 01:29:37 | [402][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1326 ntime: 0081 mem: 3.36 + 04-04 01:29:48 | [402][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0080 ntime: 0085 mem: 3.36 + 04-04 01:29:56 | [402][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0068 ntime: 0085 mem: 3.36 + 04-04 01:30:05 | [402][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0266 ntime: 0080 mem: 3.36 + 04-04 01:30:15 | [402][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1383 ntime: 0085 mem: 3.36 + 04-04 01:30:23 | [402][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0832 ntime: 0080 mem: 3.36 + 04-04 01:30:33 | [402][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1544 ntime: 0087 mem: 3.36 + 04-04 01:30:40 | [402][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0907 ntime: 0079 mem: 3.36 + 04-04 01:30:49 | Time info >>>> elapsed: 247.44 mins remain: 366.56 mins + 04-04 01:30:49 | [403][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0121 ntime: 0085 mem: 3.36 + 04-04 01:30:59 | [403][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1454 ntime: 0076 mem: 3.36 + 04-04 01:31:08 | [403][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0964 ntime: 0082 mem: 3.36 + 04-04 01:31:18 | [403][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0074 ntime: 0081 mem: 3.36 + 04-04 01:31:29 | [403][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1496 ntime: 0087 mem: 3.36 + 04-04 01:31:37 | [403][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1412 ntime: 0088 mem: 3.36 + 04-04 01:31:47 | [403][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1337 ntime: 0079 mem: 3.36 + 04-04 01:31:54 | [403][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0915 ntime: 0084 mem: 3.36 + 04-04 01:32:03 | [403][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0813 ntime: 0089 mem: 3.36 + 04-04 01:32:12 | [403][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1371 ntime: 0088 mem: 3.36 + 04-04 01:32:22 | [403][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0899 ntime: 0079 mem: 3.36 + 04-04 01:32:29 | [403][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0091 ntime: 0082 mem: 3.36 + 04-04 01:32:38 | [403][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0064 ntime: 0078 mem: 3.36 + 04-04 01:32:46 | [403][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0078 ntime: 0076 mem: 3.36 + 04-04 01:32:55 | [403][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1029 ntime: 0086 mem: 3.36 + 04-04 01:33:05 | [403][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1318 ntime: 0077 mem: 3.36 + 04-04 01:33:14 | [403][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1139 ntime: 0076 mem: 3.36 + 04-04 01:33:24 | [403][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0955 ntime: 0083 mem: 3.36 + 04-04 01:33:32 | Time info >>>> elapsed: 250.16 mins remain: 369.05 mins + 04-04 01:33:34 | [404][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1482 ntime: 0086 mem: 3.36 + 04-04 01:33:42 | [404][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0885 ntime: 0076 mem: 3.36 + 04-04 01:33:50 | [404][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0062 ntime: 0079 mem: 3.36 + 04-04 01:33:59 | [404][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0076 mem: 3.36 + 04-04 01:34:08 | [404][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0527 ntime: 0081 mem: 3.36 + 04-04 01:34:18 | [404][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0823 ntime: 0080 mem: 3.36 + 04-04 01:34:26 | [404][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0160 ntime: 0074 mem: 3.36 + 04-04 01:34:37 | [404][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1395 ntime: 0084 mem: 3.36 + 04-04 01:34:45 | [404][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1091 ntime: 0082 mem: 3.36 + 04-04 01:34:54 | [404][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1528 ntime: 0080 mem: 3.36 + 04-04 01:35:02 | [404][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1634 ntime: 0081 mem: 3.36 + 04-04 01:35:13 | [404][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1134 ntime: 0083 mem: 3.36 + 04-04 01:35:22 | [404][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0172 ntime: 0078 mem: 3.36 + 04-04 01:35:31 | [404][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0176 ntime: 0081 mem: 3.36 + 04-04 01:35:39 | [404][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1318 ntime: 0087 mem: 3.36 + 04-04 01:35:49 | [404][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0065 ntime: 0081 mem: 3.36 + 04-04 01:35:59 | [404][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0080 mem: 3.36 + 04-04 01:36:08 | [404][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0114 ntime: 0079 mem: 3.36 + 04-04 01:36:16 | Time info >>>> elapsed: 252.90 mins remain: 371.55 mins + 04-04 01:36:17 | [405][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0541 ntime: 0082 mem: 3.36 + 04-04 01:36:26 | [405][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0080 ntime: 0079 mem: 3.36 + 04-04 01:36:35 | [405][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0636 ntime: 0075 mem: 3.36 + 04-04 01:36:44 | [405][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1393 ntime: 0081 mem: 3.36 + 04-04 01:36:54 | [405][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0084 mem: 3.36 + 04-04 01:37:03 | [405][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1488 ntime: 0084 mem: 3.36 + 04-04 01:37:12 | [405][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1007 ntime: 0072 mem: 3.36 + 04-04 01:37:21 | [405][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1181 ntime: 0081 mem: 3.36 + 04-04 01:37:30 | [405][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1340 ntime: 0079 mem: 3.36 + 04-04 01:37:40 | [405][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0969 ntime: 0088 mem: 3.36 + 04-04 01:37:51 | [405][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1429 ntime: 0068 mem: 3.36 + 04-04 01:38:01 | [405][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1010 ntime: 0084 mem: 3.36 + 04-04 01:38:12 | [405][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1352 ntime: 0077 mem: 3.36 + 04-04 01:38:21 | [405][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0065 ntime: 0078 mem: 3.36 + 04-04 01:38:31 | [405][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0494 ntime: 0081 mem: 3.36 + 04-04 01:38:41 | [405][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1017 ntime: 0090 mem: 3.36 + 04-04 01:38:50 | [405][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1492 ntime: 0057 mem: 3.36 + 04-04 01:39:01 | [405][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1302 ntime: 0086 mem: 3.36 + 04-04 01:39:09 | Time info >>>> elapsed: 255.78 mins remain: 374.22 mins + 04-04 01:39:10 | [406][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1394 ntime: 0077 mem: 3.36 + 04-04 01:39:19 | [406][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1020 ntime: 0083 mem: 3.36 + 04-04 01:39:30 | [406][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0899 ntime: 0080 mem: 3.36 + 04-04 01:39:39 | [406][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1491 ntime: 0092 mem: 3.36 + 04-04 01:39:48 | [406][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1418 ntime: 0086 mem: 3.36 + 04-04 01:39:57 | [406][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0511 ntime: 0062 mem: 3.36 + 04-04 01:40:07 | [406][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1125 ntime: 0088 mem: 3.36 + 04-04 01:40:15 | [406][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1425 ntime: 0076 mem: 3.36 + 04-04 01:40:24 | [406][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0960 ntime: 0082 mem: 3.36 + 04-04 01:40:32 | [406][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0074 ntime: 0081 mem: 3.36 + 04-04 01:40:40 | [406][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1560 ntime: 0079 mem: 3.36 + 04-04 01:40:50 | [406][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1396 ntime: 0076 mem: 3.36 + 04-04 01:40:59 | [406][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0988 ntime: 0079 mem: 3.36 + 04-04 01:41:08 | [406][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1182 ntime: 0082 mem: 3.36 + 04-04 01:41:21 | [406][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1960 ntime: 0089 mem: 3.36 + 04-04 01:41:30 | [406][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0073 ntime: 0080 mem: 3.36 + 04-04 01:41:38 | [406][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1048 ntime: 0081 mem: 3.36 + 04-04 01:41:46 | [406][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1523 ntime: 0080 mem: 3.36 + 04-04 01:41:55 | Time info >>>> elapsed: 258.55 mins remain: 376.70 mins + 04-04 01:41:56 | [407][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0780 ntime: 0080 mem: 3.36 + 04-04 01:42:06 | [407][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1328 ntime: 0075 mem: 3.36 + 04-04 01:42:14 | [407][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0081 mem: 3.36 + 04-04 01:42:25 | [407][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1476 ntime: 0084 mem: 3.36 + 04-04 01:42:32 | [407][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0771 ntime: 0082 mem: 3.36 + 04-04 01:42:41 | [407][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1200 ntime: 0078 mem: 3.36 + 04-04 01:42:49 | [407][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0868 ntime: 0083 mem: 3.36 + 04-04 01:42:58 | [407][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0101 ntime: 0084 mem: 3.36 + 04-04 01:43:09 | [407][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1213 ntime: 0080 mem: 3.36 + 04-04 01:43:20 | [407][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1422 ntime: 0086 mem: 3.36 + 04-04 01:43:29 | [407][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0428 ntime: 0079 mem: 3.36 + 04-04 01:43:40 | [407][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0106 ntime: 0078 mem: 3.36 + 04-04 01:43:53 | [407][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1512 ntime: 0080 mem: 3.36 + 04-04 01:44:04 | [407][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0746 ntime: 0079 mem: 3.36 + 04-04 01:44:14 | [407][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1458 ntime: 0076 mem: 3.36 + 04-04 01:44:24 | [407][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1440 ntime: 0081 mem: 3.36 + 04-04 01:44:31 | [407][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0277 ntime: 0080 mem: 3.36 + 04-04 01:44:42 | [407][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1511 ntime: 0083 mem: 3.36 + 04-04 01:44:49 | Time info >>>> elapsed: 261.44 mins remain: 379.35 mins + 04-04 01:44:50 | [408][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0995 ntime: 0079 mem: 3.36 + 04-04 01:45:00 | [408][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0702 ntime: 0082 mem: 3.36 + 04-04 01:45:12 | [408][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1368 ntime: 0077 mem: 3.36 + 04-04 01:45:23 | [408][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1419 ntime: 0074 mem: 3.36 + 04-04 01:45:32 | [408][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1528 ntime: 0083 mem: 3.36 + 04-04 01:45:43 | [408][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1315 ntime: 0080 mem: 3.36 + 04-04 01:45:52 | [408][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0079 mem: 3.36 + 04-04 01:46:03 | [408][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1355 ntime: 0076 mem: 3.36 + 04-04 01:46:12 | [408][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0675 ntime: 0069 mem: 3.36 + 04-04 01:46:22 | [408][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0932 ntime: 0068 mem: 3.36 + 04-04 01:46:33 | [408][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0068 ntime: 0082 mem: 3.36 + 04-04 01:46:42 | [408][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1042 ntime: 0088 mem: 3.36 + 04-04 01:46:52 | [408][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1346 ntime: 0084 mem: 3.36 + 04-04 01:47:02 | [408][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1477 ntime: 0085 mem: 3.36 + 04-04 01:47:12 | [408][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1386 ntime: 0091 mem: 3.36 + 04-04 01:47:21 | [408][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1155 ntime: 0072 mem: 3.36 + 04-04 01:47:31 | [408][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0073 mem: 3.36 + 04-04 01:47:42 | [408][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1071 ntime: 0078 mem: 3.36 + 04-04 01:47:49 | Time info >>>> elapsed: 264.45 mins remain: 382.12 mins + 04-04 01:47:50 | [409][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1132 ntime: 0071 mem: 3.36 + 04-04 01:48:00 | [409][010/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1504 ntime: 0090 mem: 3.36 + 04-04 01:48:10 | [409][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1139 ntime: 0081 mem: 3.36 + 04-04 01:48:19 | [409][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0119 ntime: 0082 mem: 3.36 + 04-04 01:48:28 | [409][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1161 ntime: 0080 mem: 3.36 + 04-04 01:48:38 | [409][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1754 ntime: 0085 mem: 3.36 + 04-04 01:48:50 | [409][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1280 ntime: 0081 mem: 3.36 + 04-04 01:48:58 | [409][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0222 ntime: 0083 mem: 3.36 + 04-04 01:49:09 | [409][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0665 ntime: 0082 mem: 3.36 + 04-04 01:49:15 | [409][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0087 ntime: 0083 mem: 3.36 + 04-04 01:49:25 | [409][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0907 ntime: 0080 mem: 3.36 + 04-04 01:49:33 | [409][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0205 ntime: 0080 mem: 3.36 + 04-04 01:49:42 | [409][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0075 mem: 3.36 + 04-04 01:49:50 | [409][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0066 ntime: 0073 mem: 3.36 + 04-04 01:50:01 | [409][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1428 ntime: 0087 mem: 3.36 + 04-04 01:50:10 | [409][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0086 ntime: 0075 mem: 3.36 + 04-04 01:50:18 | [409][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1056 ntime: 0081 mem: 3.36 + 04-04 01:50:28 | [409][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 2367 ntime: 0084 mem: 3.36 + 04-04 01:50:38 | Time info >>>> elapsed: 267.26 mins remain: 384.59 mins + 04-04 01:50:39 | [410][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0800 ntime: 0072 mem: 3.36 + 04-04 01:50:48 | [410][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1778 ntime: 0088 mem: 3.36 + 04-04 01:50:58 | [410][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0083 ntime: 0080 mem: 3.36 + 04-04 01:51:08 | [410][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1190 ntime: 0085 mem: 3.36 + 04-04 01:51:17 | [410][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0787 ntime: 0079 mem: 3.36 + 04-04 01:51:29 | [410][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1338 ntime: 0078 mem: 3.36 + 04-04 01:51:41 | [410][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1515 ntime: 0083 mem: 3.36 + 04-04 01:51:52 | [410][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1433 ntime: 0085 mem: 3.36 + 04-04 01:52:00 | [410][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0059 ntime: 0076 mem: 3.36 + 04-04 01:52:11 | [410][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0578 ntime: 0076 mem: 3.36 + 04-04 01:52:19 | [410][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0965 ntime: 0088 mem: 3.36 + 04-04 01:52:28 | [410][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0059 ntime: 0082 mem: 3.36 + 04-04 01:52:37 | [410][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0774 ntime: 0080 mem: 3.36 + 04-04 01:52:46 | [410][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0069 ntime: 0084 mem: 3.36 + 04-04 01:52:57 | [410][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0668 ntime: 0074 mem: 3.36 + 04-04 01:53:07 | [410][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0076 ntime: 0074 mem: 3.36 + 04-04 01:53:18 | [410][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1086 ntime: 0078 mem: 3.36 + 04-04 01:53:27 | [410][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1041 ntime: 0085 mem: 3.36 + 04-04 01:53:34 | Time info >>>> elapsed: 270.19 mins remain: 387.21 mins + 04-04 01:53:34 | [411][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0125 ntime: 0080 mem: 3.36 + 04-04 01:53:45 | [411][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1465 ntime: 0078 mem: 3.36 + 04-04 01:53:55 | [411][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1056 ntime: 0083 mem: 3.36 + 04-04 01:54:08 | [411][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1320 ntime: 0080 mem: 3.36 + 04-04 01:54:19 | [411][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1421 ntime: 0079 mem: 3.36 + 04-04 01:54:27 | [411][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1011 ntime: 0079 mem: 3.36 + 04-04 01:54:39 | [411][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0092 ntime: 0075 mem: 3.36 + 04-04 01:54:49 | [411][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1466 ntime: 0081 mem: 3.36 + 04-04 01:55:00 | [411][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0302 ntime: 0094 mem: 3.36 + 04-04 01:55:08 | [411][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0851 ntime: 0082 mem: 3.36 + 04-04 01:55:20 | [411][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1543 ntime: 0081 mem: 3.36 + 04-04 01:55:29 | [411][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0123 ntime: 0080 mem: 3.36 + 04-04 01:55:38 | [411][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1440 ntime: 0080 mem: 3.36 + 04-04 01:55:48 | [411][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1202 ntime: 0079 mem: 3.36 + 04-04 01:55:58 | [411][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0078 mem: 3.36 + 04-04 01:56:06 | [411][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0673 ntime: 0078 mem: 3.36 + 04-04 01:56:16 | [411][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1581 ntime: 0078 mem: 3.36 + 04-04 01:56:25 | [411][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0683 ntime: 0078 mem: 3.36 + 04-04 01:56:33 | Time info >>>> elapsed: 273.18 mins remain: 389.88 mins + 04-04 01:56:34 | [412][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1183 ntime: 0086 mem: 3.36 + 04-04 01:56:43 | [412][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0598 ntime: 0077 mem: 3.36 + 04-04 01:56:56 | [412][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0756 ntime: 0079 mem: 3.36 + 04-04 01:57:05 | [412][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1180 ntime: 0081 mem: 3.36 + 04-04 01:57:13 | [412][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1374 ntime: 0086 mem: 3.36 + 04-04 01:57:22 | [412][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1492 ntime: 0079 mem: 3.36 + 04-04 01:57:31 | [412][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0076 mem: 3.36 + 04-04 01:57:40 | [412][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1160 ntime: 0086 mem: 3.36 + 04-04 01:57:48 | [412][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0553 ntime: 0080 mem: 3.36 + 04-04 01:57:57 | [412][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1670 ntime: 0084 mem: 3.36 + 04-04 01:58:07 | [412][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0122 ntime: 0072 mem: 3.36 + 04-04 01:58:17 | [412][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0102 ntime: 0078 mem: 3.36 + 04-04 01:58:25 | [412][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0095 ntime: 0079 mem: 3.36 + 04-04 01:58:37 | [412][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1068 ntime: 0084 mem: 3.36 + 04-04 01:58:45 | [412][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1442 ntime: 0084 mem: 3.36 + 04-04 01:58:54 | [412][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0968 ntime: 0085 mem: 3.36 + 04-04 01:59:03 | [412][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0073 mem: 3.36 + 04-04 01:59:14 | [412][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1442 ntime: 0072 mem: 3.36 + 04-04 01:59:20 | Time info >>>> elapsed: 275.97 mins remain: 392.23 mins + 04-04 01:59:21 | [413][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1024 ntime: 0092 mem: 3.36 + 04-04 01:59:31 | [413][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1110 ntime: 0083 mem: 3.36 + 04-04 01:59:39 | [413][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1133 ntime: 0081 mem: 3.36 + 04-04 01:59:48 | [413][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0093 ntime: 0078 mem: 3.36 + 04-04 01:59:59 | [413][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1414 ntime: 0080 mem: 3.36 + 04-04 02:00:09 | [413][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1265 ntime: 0084 mem: 3.36 + 04-04 02:00:22 | [413][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1372 ntime: 0083 mem: 3.36 + 04-04 02:00:30 | [413][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0125 ntime: 0079 mem: 3.36 + 04-04 02:00:40 | [413][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0706 ntime: 0080 mem: 3.36 + 04-04 02:00:51 | [413][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1576 ntime: 0084 mem: 3.36 + 04-04 02:01:01 | [413][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1443 ntime: 0076 mem: 3.36 + 04-04 02:01:11 | [413][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1523 ntime: 0090 mem: 3.36 + 04-04 02:01:20 | [413][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1021 ntime: 0072 mem: 3.36 + 04-04 02:01:30 | [413][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0581 ntime: 0080 mem: 3.36 + 04-04 02:01:38 | [413][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0080 mem: 3.36 + 04-04 02:01:48 | [413][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1405 ntime: 0085 mem: 3.36 + 04-04 02:01:56 | [413][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1120 ntime: 0073 mem: 3.36 + 04-04 02:02:05 | [413][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1425 ntime: 0080 mem: 3.36 + 04-04 02:02:12 | Time info >>>> elapsed: 278.83 mins remain: 394.68 mins + 04-04 02:02:14 | [414][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1398 ntime: 0078 mem: 3.36 + 04-04 02:02:23 | [414][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0076 mem: 3.36 + 04-04 02:02:32 | [414][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0067 ntime: 0081 mem: 3.36 + 04-04 02:02:40 | [414][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0941 ntime: 0085 mem: 3.36 + 04-04 02:02:47 | [414][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0085 mem: 3.36 + 04-04 02:02:57 | [414][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1392 ntime: 0083 mem: 3.36 + 04-04 02:03:06 | [414][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0588 ntime: 0078 mem: 3.36 + 04-04 02:03:19 | [414][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1089 ntime: 0089 mem: 3.36 + 04-04 02:03:28 | [414][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0885 ntime: 0074 mem: 3.36 + 04-04 02:03:38 | [414][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1424 ntime: 0080 mem: 3.36 + 04-04 02:03:46 | [414][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1433 ntime: 0076 mem: 3.36 + 04-04 02:03:58 | [414][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0816 ntime: 0090 mem: 3.36 + 04-04 02:04:08 | [414][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0984 ntime: 0086 mem: 3.36 + 04-04 02:04:15 | [414][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1401 ntime: 0077 mem: 3.36 + 04-04 02:04:23 | [414][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1231 ntime: 0077 mem: 3.36 + 04-04 02:04:31 | [414][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0408 ntime: 0077 mem: 3.36 + 04-04 02:04:37 | [414][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0064 ntime: 0083 mem: 3.36 + 04-04 02:04:47 | [414][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1325 ntime: 0073 mem: 3.36 + 04-04 02:04:55 | Time info >>>> elapsed: 281.54 mins remain: 396.87 mins + 04-04 02:04:55 | [415][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0067 ntime: 0079 mem: 3.36 + 04-04 02:05:04 | [415][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0151 ntime: 0086 mem: 3.36 + 04-04 02:05:12 | [415][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0862 ntime: 0078 mem: 3.36 + 04-04 02:05:19 | [415][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0864 ntime: 0077 mem: 3.36 + 04-04 02:05:28 | [415][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0793 ntime: 0077 mem: 3.36 + 04-04 02:05:38 | [415][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1235 ntime: 0078 mem: 3.36 + 04-04 02:05:49 | [415][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0958 ntime: 0077 mem: 3.36 + 04-04 02:05:58 | [415][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0071 mem: 3.36 + 04-04 02:06:06 | [415][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0113 ntime: 0080 mem: 3.36 + 04-04 02:06:15 | [415][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0152 ntime: 0074 mem: 3.36 + 04-04 02:06:23 | [415][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1345 ntime: 0089 mem: 3.36 + 04-04 02:06:33 | [415][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0069 ntime: 0079 mem: 3.36 + 04-04 02:06:42 | [415][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1379 ntime: 0084 mem: 3.36 + 04-04 02:06:50 | [415][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0768 ntime: 0080 mem: 3.36 + 04-04 02:06:56 | [415][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0258 ntime: 0078 mem: 3.36 + 04-04 02:07:03 | [415][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0055 ntime: 0091 mem: 3.36 + 04-04 02:07:13 | [415][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1078 ntime: 0079 mem: 3.36 + 04-04 02:07:24 | [415][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0084 ntime: 0078 mem: 3.36 + 04-04 02:07:31 | Time info >>>> elapsed: 284.14 mins remain: 398.89 mins + 04-04 02:07:31 | [416][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0087 ntime: 0067 mem: 3.36 + 04-04 02:07:40 | [416][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1366 ntime: 0078 mem: 3.36 + 04-04 02:07:51 | [416][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0130 ntime: 0080 mem: 3.36 + 04-04 02:07:58 | [416][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0140 ntime: 0083 mem: 3.36 + 04-04 02:08:08 | [416][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0138 ntime: 0079 mem: 3.36 + 04-04 02:08:16 | [416][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0680 ntime: 0080 mem: 3.36 + 04-04 02:08:24 | [416][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0905 ntime: 0079 mem: 3.36 + 04-04 02:08:34 | [416][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0476 ntime: 0079 mem: 3.36 + 04-04 02:08:43 | [416][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0190 ntime: 0078 mem: 3.36 + 04-04 02:08:55 | [416][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1374 ntime: 0078 mem: 3.36 + 04-04 02:09:03 | [416][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0081 ntime: 0081 mem: 3.36 + 04-04 02:09:13 | [416][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1084 ntime: 0071 mem: 3.36 + 04-04 02:09:22 | [416][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0155 ntime: 0074 mem: 3.36 + 04-04 02:09:32 | [416][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1008 ntime: 0080 mem: 3.36 + 04-04 02:09:40 | [416][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1379 ntime: 0080 mem: 3.36 + 04-04 02:09:49 | [416][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1499 ntime: 0092 mem: 3.36 + 04-04 02:09:57 | [416][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1416 ntime: 0084 mem: 3.36 + 04-04 02:10:05 | [416][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1091 ntime: 0093 mem: 3.36 + 04-04 02:10:10 | Time info >>>> elapsed: 286.80 mins remain: 400.97 mins + 04-04 02:10:11 | [417][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0060 ntime: 0083 mem: 3.36 + 04-04 02:10:22 | [417][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1313 ntime: 0079 mem: 3.36 + 04-04 02:10:33 | [417][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0903 ntime: 0075 mem: 3.36 + 04-04 02:10:42 | [417][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0080 mem: 3.36 + 04-04 02:10:50 | [417][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0700 ntime: 0086 mem: 3.36 + 04-04 02:11:01 | [417][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0108 ntime: 0078 mem: 3.36 + 04-04 02:11:10 | [417][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1464 ntime: 0077 mem: 3.36 + 04-04 02:11:18 | [417][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1547 ntime: 0071 mem: 3.36 + 04-04 02:11:27 | [417][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0794 ntime: 0079 mem: 3.36 + 04-04 02:11:38 | [417][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0502 ntime: 0081 mem: 3.36 + 04-04 02:11:47 | [417][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0945 ntime: 0075 mem: 3.36 + 04-04 02:11:56 | [417][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0593 ntime: 0080 mem: 3.36 + 04-04 02:12:04 | [417][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1384 ntime: 0079 mem: 3.36 + 04-04 02:12:12 | [417][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1329 ntime: 0074 mem: 3.36 + 04-04 02:12:19 | [417][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1122 ntime: 0083 mem: 3.36 + 04-04 02:12:26 | [417][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0952 ntime: 0087 mem: 3.36 + 04-04 02:12:35 | [417][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 02:12:44 | [417][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0068 ntime: 0077 mem: 3.36 + 04-04 02:12:52 | Time info >>>> elapsed: 289.49 mins remain: 403.07 mins + 04-04 02:12:52 | [418][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0481 ntime: 0079 mem: 3.36 + 04-04 02:13:01 | [418][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1456 ntime: 0083 mem: 3.36 + 04-04 02:13:10 | [418][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0691 ntime: 0071 mem: 3.36 + 04-04 02:13:18 | [418][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1269 ntime: 0089 mem: 3.36 + 04-04 02:13:28 | [418][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0936 ntime: 0077 mem: 3.36 + 04-04 02:13:36 | [418][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1506 ntime: 0078 mem: 3.36 + 04-04 02:13:45 | [418][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0798 ntime: 0078 mem: 3.36 + 04-04 02:13:53 | [418][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0089 ntime: 0078 mem: 3.36 + 04-04 02:14:02 | [418][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1013 ntime: 0075 mem: 3.36 + 04-04 02:14:10 | [418][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0093 ntime: 0079 mem: 3.36 + 04-04 02:14:18 | [418][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0740 ntime: 0080 mem: 3.36 + 04-04 02:14:27 | [418][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1661 ntime: 0081 mem: 3.36 + 04-04 02:14:37 | [418][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0080 mem: 3.36 + 04-04 02:14:45 | [418][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1368 ntime: 0074 mem: 3.36 + 04-04 02:14:55 | [418][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0379 ntime: 0079 mem: 3.36 + 04-04 02:15:06 | [418][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1528 ntime: 0074 mem: 3.36 + 04-04 02:15:14 | [418][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0153 ntime: 0081 mem: 3.36 + 04-04 02:15:24 | [418][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1179 ntime: 0082 mem: 3.36 + 04-04 02:15:32 | Time info >>>> elapsed: 292.16 mins remain: 405.12 mins + 04-04 02:15:33 | [419][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1089 ntime: 0083 mem: 3.36 + 04-04 02:15:45 | [419][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1403 ntime: 0082 mem: 3.36 + 04-04 02:15:54 | [419][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0136 ntime: 0078 mem: 3.36 + 04-04 02:16:03 | [419][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1216 ntime: 0078 mem: 3.36 + 04-04 02:16:14 | [419][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1080 ntime: 0078 mem: 3.36 + 04-04 02:16:24 | [419][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0728 ntime: 0081 mem: 3.36 + 04-04 02:16:33 | [419][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1294 ntime: 0077 mem: 3.36 + 04-04 02:16:43 | [419][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1532 ntime: 0084 mem: 3.36 + 04-04 02:16:51 | [419][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0182 ntime: 0057 mem: 3.36 + 04-04 02:17:01 | [419][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0870 ntime: 0078 mem: 3.36 + 04-04 02:17:11 | [419][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1235 ntime: 0080 mem: 3.36 + 04-04 02:17:22 | [419][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1474 ntime: 0083 mem: 3.36 + 04-04 02:17:33 | [419][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1491 ntime: 0078 mem: 3.36 + 04-04 02:17:42 | [419][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1357 ntime: 0092 mem: 3.36 + 04-04 02:17:52 | [419][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1757 ntime: 0079 mem: 3.36 + 04-04 02:18:05 | [419][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1583 ntime: 0086 mem: 3.36 + 04-04 02:18:15 | [419][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1396 ntime: 0082 mem: 3.36 + 04-04 02:18:23 | [419][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0110 ntime: 0081 mem: 3.36 + 04-04 02:18:29 | Time info >>>> elapsed: 295.11 mins remain: 407.54 mins + 04-04 02:18:29 | [420][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0062 ntime: 0074 mem: 3.36 + 04-04 02:18:42 | [420][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1380 ntime: 0078 mem: 3.36 + 04-04 02:18:51 | [420][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0134 ntime: 0079 mem: 3.36 + 04-04 02:19:01 | [420][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1147 ntime: 0079 mem: 3.36 + 04-04 02:19:11 | [420][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0998 ntime: 0078 mem: 3.36 + 04-04 02:19:21 | [420][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0756 ntime: 0088 mem: 3.36 + 04-04 02:19:32 | [420][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0688 ntime: 0078 mem: 3.36 + 04-04 02:19:40 | [420][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1466 ntime: 0090 mem: 3.36 + 04-04 02:19:49 | [420][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0169 ntime: 0073 mem: 3.36 + 04-04 02:19:57 | [420][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0888 ntime: 0080 mem: 3.36 + 04-04 02:20:06 | [420][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0111 ntime: 0078 mem: 3.36 + 04-04 02:20:15 | [420][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0981 ntime: 0070 mem: 3.36 + 04-04 02:20:23 | [420][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0076 ntime: 0072 mem: 3.36 + 04-04 02:20:33 | [420][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1033 ntime: 0078 mem: 3.36 + 04-04 02:20:40 | [420][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0537 ntime: 0079 mem: 3.36 + 04-04 02:20:48 | [420][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0550 ntime: 0081 mem: 3.36 + 04-04 02:20:56 | [420][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0078 mem: 3.36 + 04-04 02:21:07 | [420][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0787 ntime: 0073 mem: 3.36 + 04-04 02:21:13 | Time info >>>> elapsed: 297.84 mins remain: 409.61 mins + 04-04 02:21:14 | [421][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1481 ntime: 0083 mem: 3.36 + 04-04 02:21:23 | [421][010/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1184 ntime: 0076 mem: 3.36 + 04-04 02:21:29 | [421][020/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1104 ntime: 0086 mem: 3.36 + 04-04 02:21:37 | [421][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0083 mem: 3.36 + 04-04 02:21:48 | [421][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1222 ntime: 0074 mem: 3.36 + 04-04 02:21:56 | [421][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0069 ntime: 0077 mem: 3.36 + 04-04 02:22:09 | [421][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1452 ntime: 0078 mem: 3.36 + 04-04 02:22:19 | [421][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0193 ntime: 0080 mem: 3.36 + 04-04 02:22:29 | [421][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0734 ntime: 0080 mem: 3.36 + 04-04 02:22:43 | [421][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0991 ntime: 0078 mem: 3.36 + 04-04 02:22:52 | [421][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1228 ntime: 0077 mem: 3.36 + 04-04 02:23:03 | [421][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1427 ntime: 0077 mem: 3.36 + 04-04 02:23:10 | [421][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1382 ntime: 0084 mem: 3.36 + 04-04 02:23:20 | [421][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1082 ntime: 0074 mem: 3.36 + 04-04 02:23:31 | [421][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1382 ntime: 0078 mem: 3.36 + 04-04 02:23:40 | [421][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0600 ntime: 0082 mem: 3.36 + 04-04 02:23:48 | [421][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0747 ntime: 0077 mem: 3.36 + 04-04 02:23:58 | [421][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1302 ntime: 0079 mem: 3.36 + 04-04 02:24:05 | Time info >>>> elapsed: 300.71 mins remain: 411.87 mins + 04-04 02:24:07 | [422][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1460 ntime: 0079 mem: 3.36 + 04-04 02:24:16 | [422][010/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1499 ntime: 0087 mem: 3.36 + 04-04 02:24:27 | [422][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0629 ntime: 0079 mem: 3.36 + 04-04 02:24:38 | [422][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0946 ntime: 0084 mem: 3.36 + 04-04 02:24:46 | [422][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0092 ntime: 0082 mem: 3.36 + 04-04 02:24:56 | [422][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1126 ntime: 0083 mem: 3.36 + 04-04 02:25:04 | [422][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0843 ntime: 0081 mem: 3.36 + 04-04 02:25:12 | [422][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0120 ntime: 0077 mem: 3.36 + 04-04 02:25:19 | [422][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1107 ntime: 0081 mem: 3.36 + 04-04 02:25:29 | [422][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1345 ntime: 0076 mem: 3.36 + 04-04 02:25:38 | [422][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0117 ntime: 0078 mem: 3.36 + 04-04 02:25:47 | [422][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1368 ntime: 0081 mem: 3.36 + 04-04 02:25:57 | [422][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0605 ntime: 0082 mem: 3.36 + 04-04 02:26:08 | [422][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1126 ntime: 0090 mem: 3.36 + 04-04 02:26:18 | [422][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0911 ntime: 0078 mem: 3.36 + 04-04 02:26:30 | [422][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1550 ntime: 0080 mem: 3.36 + 04-04 02:26:39 | [422][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1388 ntime: 0081 mem: 3.36 + 04-04 02:26:48 | [422][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0107 ntime: 0090 mem: 3.36 + 04-04 02:26:55 | Time info >>>> elapsed: 303.54 mins remain: 414.05 mins + 04-04 02:26:55 | [423][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0167 ntime: 0075 mem: 3.36 + 04-04 02:27:04 | [423][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1017 ntime: 0078 mem: 3.36 + 04-04 02:27:14 | [423][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1383 ntime: 0074 mem: 3.36 + 04-04 02:27:23 | [423][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1514 ntime: 0082 mem: 3.36 + 04-04 02:27:31 | [423][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0229 ntime: 0084 mem: 3.36 + 04-04 02:27:41 | [423][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0971 ntime: 0083 mem: 3.36 + 04-04 02:27:51 | [423][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1279 ntime: 0075 mem: 3.36 + 04-04 02:28:02 | [423][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0108 ntime: 0091 mem: 3.36 + 04-04 02:28:11 | [423][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1098 ntime: 0087 mem: 3.36 + 04-04 02:28:19 | [423][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0990 ntime: 0079 mem: 3.36 + 04-04 02:28:30 | [423][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1202 ntime: 0079 mem: 3.36 + 04-04 02:28:39 | [423][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0082 ntime: 0077 mem: 3.36 + 04-04 02:28:49 | [423][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0095 ntime: 0077 mem: 3.36 + 04-04 02:28:58 | [423][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0929 ntime: 0078 mem: 3.36 + 04-04 02:29:04 | [423][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0079 ntime: 0078 mem: 3.36 + 04-04 02:29:14 | [423][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1464 ntime: 0082 mem: 3.36 + 04-04 02:29:22 | [423][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1101 ntime: 0078 mem: 3.36 + 04-04 02:29:29 | [423][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0077 ntime: 0081 mem: 3.36 + 04-04 02:29:35 | Time info >>>> elapsed: 306.22 mins remain: 415.99 mins + 04-04 02:29:37 | [424][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1308 ntime: 0084 mem: 3.36 + 04-04 02:29:45 | [424][010/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1308 ntime: 0093 mem: 3.36 + 04-04 02:29:55 | [424][020/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1166 ntime: 0081 mem: 3.36 + 04-04 02:30:05 | [424][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0967 ntime: 0087 mem: 3.36 + 04-04 02:30:15 | [424][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0124 ntime: 0077 mem: 3.36 + 04-04 02:30:24 | [424][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1390 ntime: 0078 mem: 3.36 + 04-04 02:30:36 | [424][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0775 ntime: 0088 mem: 3.36 + 04-04 02:30:46 | [424][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1334 ntime: 0081 mem: 3.36 + 04-04 02:30:54 | [424][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1294 ntime: 0078 mem: 3.36 + 04-04 02:31:02 | [424][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0099 ntime: 0078 mem: 3.36 + 04-04 02:31:13 | [424][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1493 ntime: 0082 mem: 3.36 + 04-04 02:31:18 | [424][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0770 ntime: 0078 mem: 3.36 + 04-04 02:31:27 | [424][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1203 ntime: 0085 mem: 3.36 + 04-04 02:31:37 | [424][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1117 ntime: 0086 mem: 3.36 + 04-04 02:31:45 | [424][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1019 ntime: 0081 mem: 3.36 + 04-04 02:31:55 | [424][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1129 ntime: 0083 mem: 3.36 + 04-04 02:32:04 | [424][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0099 ntime: 0080 mem: 3.36 + 04-04 02:32:12 | [424][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1193 ntime: 0076 mem: 3.36 + 04-04 02:32:20 | Time info >>>> elapsed: 308.96 mins remain: 418.01 mins + 04-04 02:32:21 | [425][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1315 ntime: 0088 mem: 3.36 + 04-04 02:32:31 | [425][010/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1273 ntime: 0082 mem: 3.36 + 04-04 02:32:39 | [425][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1468 ntime: 0076 mem: 3.36 + 04-04 02:32:48 | [425][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1426 ntime: 0081 mem: 3.36 + 04-04 02:32:57 | [425][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1287 ntime: 0084 mem: 3.36 + 04-04 02:33:05 | [425][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1432 ntime: 0089 mem: 3.36 + 04-04 02:33:14 | [425][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1286 ntime: 0080 mem: 3.36 + 04-04 02:33:21 | [425][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0531 ntime: 0077 mem: 3.36 + 04-04 02:33:33 | [425][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0943 ntime: 0083 mem: 3.36 + 04-04 02:33:42 | [425][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1520 ntime: 0079 mem: 3.36 + 04-04 02:33:53 | [425][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1388 ntime: 0086 mem: 3.36 + 04-04 02:34:03 | [425][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1223 ntime: 0081 mem: 3.36 + 04-04 02:34:11 | [425][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0936 ntime: 0081 mem: 3.36 + 04-04 02:34:22 | [425][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1177 ntime: 0088 mem: 3.36 + 04-04 02:34:32 | [425][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1470 ntime: 0081 mem: 3.36 + 04-04 02:34:42 | [425][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0976 ntime: 0077 mem: 3.36 + 04-04 02:34:50 | [425][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1331 ntime: 0071 mem: 3.36 + 04-04 02:34:59 | [425][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1288 ntime: 0082 mem: 3.36 + 04-04 02:35:07 | Time info >>>> elapsed: 311.74 mins remain: 420.05 mins + 04-04 02:35:09 | [426][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 2375 ntime: 0088 mem: 3.36 + 04-04 02:35:18 | [426][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1394 ntime: 0072 mem: 3.36 + 04-04 02:35:28 | [426][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0987 ntime: 0083 mem: 3.36 + 04-04 02:35:36 | [426][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0552 ntime: 0078 mem: 3.36 + 04-04 02:35:49 | [426][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1654 ntime: 0073 mem: 3.36 + 04-04 02:35:56 | [426][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0065 ntime: 0086 mem: 3.36 + 04-04 02:36:05 | [426][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0550 ntime: 0075 mem: 3.36 + 04-04 02:36:15 | [426][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1343 ntime: 0085 mem: 3.36 + 04-04 02:36:26 | [426][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0584 ntime: 0077 mem: 3.36 + 04-04 02:36:40 | [426][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1478 ntime: 0071 mem: 3.36 + 04-04 02:36:53 | [426][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0238 ntime: 0077 mem: 3.36 + 04-04 02:37:05 | [426][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1879 ntime: 0080 mem: 3.36 + 04-04 02:37:19 | [426][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1618 ntime: 0082 mem: 3.36 + 04-04 02:37:30 | [426][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0097 ntime: 0094 mem: 3.36 + 04-04 02:37:44 | [426][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0873 ntime: 0079 mem: 3.36 + 04-04 02:37:56 | [426][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1646 ntime: 0069 mem: 3.36 + 04-04 02:38:11 | [426][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1420 ntime: 0080 mem: 3.36 + 04-04 02:38:25 | [426][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0913 ntime: 0078 mem: 3.36 + 04-04 02:38:35 | Time info >>>> elapsed: 315.22 mins remain: 423.00 mins + 04-04 02:38:35 | [427][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 02:38:50 | [427][010/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0942 ntime: 0079 mem: 3.36 + 04-04 02:39:03 | [427][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1596 ntime: 0090 mem: 3.36 + 04-04 02:39:15 | [427][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0482 ntime: 0080 mem: 3.36 + 04-04 02:39:28 | [427][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0729 ntime: 0083 mem: 3.36 + 04-04 02:39:41 | [427][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0079 ntime: 0080 mem: 3.36 + 04-04 02:39:56 | [427][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1495 ntime: 0086 mem: 3.36 + 04-04 02:40:08 | [427][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1098 ntime: 0085 mem: 3.36 + 04-04 02:40:15 | [427][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1043 ntime: 0085 mem: 3.36 + 04-04 02:40:23 | [427][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0160 ntime: 0079 mem: 3.36 + 04-04 02:40:31 | [427][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0583 ntime: 0086 mem: 3.36 + 04-04 02:40:38 | [427][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1373 ntime: 0079 mem: 3.36 + 04-04 02:40:44 | [427][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0350 ntime: 0080 mem: 3.36 + 04-04 02:40:51 | [427][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0092 ntime: 0082 mem: 3.36 + 04-04 02:40:58 | [427][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0099 ntime: 0077 mem: 3.36 + 04-04 02:41:07 | [427][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1350 ntime: 0081 mem: 3.36 + 04-04 02:41:15 | [427][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0923 ntime: 0079 mem: 3.36 + 04-04 02:41:23 | [427][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0116 ntime: 0086 mem: 3.36 + 04-04 02:41:28 | Time info >>>> elapsed: 318.09 mins remain: 425.11 mins + 04-04 02:41:28 | [428][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0084 mem: 3.36 + 04-04 02:41:35 | [428][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0537 ntime: 0077 mem: 3.36 + 04-04 02:41:44 | [428][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1059 ntime: 0082 mem: 3.36 + 04-04 02:41:53 | [428][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1201 ntime: 0086 mem: 3.36 + 04-04 02:42:01 | [428][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1041 ntime: 0084 mem: 3.36 + 04-04 02:42:11 | [428][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1439 ntime: 0084 mem: 3.36 + 04-04 02:42:17 | [428][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0120 ntime: 0076 mem: 3.36 + 04-04 02:42:24 | [428][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0245 ntime: 0076 mem: 3.36 + 04-04 02:42:31 | [428][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1604 ntime: 0086 mem: 3.36 + 04-04 02:42:38 | [428][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0121 ntime: 0080 mem: 3.36 + 04-04 02:42:46 | [428][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1119 ntime: 0083 mem: 3.36 + 04-04 02:42:55 | [428][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0958 ntime: 0090 mem: 3.36 + 04-04 02:43:01 | [428][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0321 ntime: 0092 mem: 3.36 + 04-04 02:43:09 | [428][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0779 ntime: 0085 mem: 3.36 + 04-04 02:43:17 | [428][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0642 ntime: 0081 mem: 3.36 + 04-04 02:43:26 | [428][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0828 ntime: 0081 mem: 3.36 + 04-04 02:43:35 | [428][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0078 mem: 3.36 + 04-04 02:43:42 | [428][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1509 ntime: 0078 mem: 3.36 + 04-04 02:43:50 | Time info >>>> elapsed: 320.47 mins remain: 426.54 mins + 04-04 02:43:52 | [429][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1385 ntime: 0079 mem: 3.36 + 04-04 02:43:58 | [429][010/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1302 ntime: 0080 mem: 3.36 + 04-04 02:44:07 | [429][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0077 mem: 3.36 + 04-04 02:44:16 | [429][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0851 ntime: 0077 mem: 3.36 + 04-04 02:44:23 | [429][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0814 ntime: 0079 mem: 3.36 + 04-04 02:44:31 | [429][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0099 ntime: 0078 mem: 3.36 + 04-04 02:44:41 | [429][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0082 ntime: 0083 mem: 3.36 + 04-04 02:44:47 | [429][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1089 ntime: 0090 mem: 3.36 + 04-04 02:44:54 | [429][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1290 ntime: 0084 mem: 3.36 + 04-04 02:45:01 | [429][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0643 ntime: 0079 mem: 3.36 + 04-04 02:45:10 | [429][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0066 ntime: 0082 mem: 3.36 + 04-04 02:45:18 | [429][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1009 ntime: 0093 mem: 3.36 + 04-04 02:45:24 | [429][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0983 ntime: 0088 mem: 3.36 + 04-04 02:45:31 | [429][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0078 mem: 3.36 + 04-04 02:45:39 | [429][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0080 mem: 3.36 + 04-04 02:45:47 | [429][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1007 ntime: 0081 mem: 3.36 + 04-04 02:45:53 | [429][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0136 ntime: 0085 mem: 3.36 + 04-04 02:46:01 | [429][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0098 ntime: 0076 mem: 3.36 + 04-04 02:46:10 | Time info >>>> elapsed: 322.79 mins remain: 427.88 mins + 04-04 02:46:11 | [430][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1143 ntime: 0083 mem: 3.36 + 04-04 02:46:19 | [430][010/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0550 ntime: 0079 mem: 3.36 + 04-04 02:46:27 | [430][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1156 ntime: 0086 mem: 3.36 + 04-04 02:46:35 | [430][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0887 ntime: 0075 mem: 3.36 + 04-04 02:46:44 | [430][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0363 ntime: 0074 mem: 3.36 + 04-04 02:46:51 | [430][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0739 ntime: 0082 mem: 3.36 + 04-04 02:47:00 | [430][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1108 ntime: 0075 mem: 3.36 + 04-04 02:47:06 | [430][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0905 ntime: 0083 mem: 3.36 + 04-04 02:47:15 | [430][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1049 ntime: 0091 mem: 3.36 + 04-04 02:47:24 | [430][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1099 ntime: 0081 mem: 3.36 + 04-04 02:47:32 | [430][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1147 ntime: 0071 mem: 3.36 + 04-04 02:47:42 | [430][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1811 ntime: 0083 mem: 3.36 + 04-04 02:47:50 | [430][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1398 ntime: 0084 mem: 3.36 + 04-04 02:47:59 | [430][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0436 ntime: 0080 mem: 3.36 + 04-04 02:48:09 | [430][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1414 ntime: 0078 mem: 3.36 + 04-04 02:48:17 | [430][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0097 ntime: 0073 mem: 3.36 + 04-04 02:48:25 | [430][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0112 ntime: 0079 mem: 3.36 + 04-04 02:48:34 | [430][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1362 ntime: 0085 mem: 3.36 + 04-04 02:48:41 | Time info >>>> elapsed: 325.31 mins remain: 429.46 mins + 04-04 02:48:41 | [431][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0169 ntime: 0080 mem: 3.36 + 04-04 02:48:50 | [431][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1232 ntime: 0084 mem: 3.36 + 04-04 02:48:56 | [431][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0732 ntime: 0080 mem: 3.36 + 04-04 02:49:05 | [431][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0155 ntime: 0075 mem: 3.36 + 04-04 02:49:13 | [431][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1137 ntime: 0084 mem: 3.36 + 04-04 02:49:20 | [431][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1162 ntime: 0060 mem: 3.36 + 04-04 02:49:27 | [431][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0100 ntime: 0083 mem: 3.36 + 04-04 02:49:36 | [431][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0512 ntime: 0081 mem: 3.36 + 04-04 02:49:43 | [431][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1601 ntime: 0075 mem: 3.36 + 04-04 02:49:53 | [431][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1303 ntime: 0074 mem: 3.36 + 04-04 02:50:01 | [431][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1149 ntime: 0085 mem: 3.36 + 04-04 02:50:09 | [431][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1049 ntime: 0085 mem: 3.36 + 04-04 02:50:17 | [431][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0398 ntime: 0072 mem: 3.36 + 04-04 02:50:23 | [431][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0423 ntime: 0079 mem: 3.36 + 04-04 02:50:32 | [431][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0079 mem: 3.36 + 04-04 02:50:39 | [431][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0954 ntime: 0082 mem: 3.36 + 04-04 02:50:49 | [431][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0196 ntime: 0078 mem: 3.36 + 04-04 02:50:56 | [431][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0816 ntime: 0080 mem: 3.36 + 04-04 02:51:03 | Time info >>>> elapsed: 327.68 mins remain: 430.84 mins + 04-04 02:51:03 | [432][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0103 ntime: 0079 mem: 3.36 + 04-04 02:51:10 | [432][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0651 ntime: 0090 mem: 3.36 + 04-04 02:51:18 | [432][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1389 ntime: 0081 mem: 3.36 + 04-04 02:51:26 | [432][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0062 ntime: 0075 mem: 3.36 + 04-04 02:51:36 | [432][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1232 ntime: 0078 mem: 3.36 + 04-04 02:51:45 | [432][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0897 ntime: 0082 mem: 3.36 + 04-04 02:51:55 | [432][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1179 ntime: 0077 mem: 3.36 + 04-04 02:52:04 | [432][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1294 ntime: 0078 mem: 3.36 + 04-04 02:52:12 | [432][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0444 ntime: 0083 mem: 3.36 + 04-04 02:52:21 | [432][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1412 ntime: 0081 mem: 3.36 + 04-04 02:52:30 | [432][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0075 mem: 3.36 + 04-04 02:52:38 | [432][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0793 ntime: 0075 mem: 3.36 + 04-04 02:52:46 | [432][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0812 ntime: 0087 mem: 3.36 + 04-04 02:52:56 | [432][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1160 ntime: 0077 mem: 3.36 + 04-04 02:53:04 | [432][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0101 ntime: 0079 mem: 3.36 + 04-04 02:53:11 | [432][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0909 ntime: 0083 mem: 3.36 + 04-04 02:53:18 | [432][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0121 ntime: 0076 mem: 3.36 + 04-04 02:53:25 | [432][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0135 ntime: 0072 mem: 3.36 + 04-04 02:53:31 | Time info >>>> elapsed: 330.14 mins remain: 432.30 mins + 04-04 02:53:32 | [433][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1342 ntime: 0078 mem: 3.36 + 04-04 02:53:39 | [433][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0525 ntime: 0084 mem: 3.36 + 04-04 02:53:48 | [433][020/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0075 ntime: 0082 mem: 3.36 + 04-04 02:53:56 | [433][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0080 mem: 3.36 + 04-04 02:54:02 | [433][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0769 ntime: 0084 mem: 3.36 + 04-04 02:54:11 | [433][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0419 ntime: 0079 mem: 3.36 + 04-04 02:54:20 | [433][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1377 ntime: 0089 mem: 3.36 + 04-04 02:54:29 | [433][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0986 ntime: 0082 mem: 3.36 + 04-04 02:54:38 | [433][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1395 ntime: 0078 mem: 3.36 + 04-04 02:54:47 | [433][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0124 ntime: 0082 mem: 3.36 + 04-04 02:54:55 | [433][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0969 ntime: 0076 mem: 3.36 + 04-04 02:55:03 | [433][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1160 ntime: 0079 mem: 3.36 + 04-04 02:55:12 | [433][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1807 ntime: 0076 mem: 3.36 + 04-04 02:55:22 | [433][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0073 ntime: 0079 mem: 3.36 + 04-04 02:55:29 | [433][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1025 ntime: 0062 mem: 3.36 + 04-04 02:55:36 | [433][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0074 mem: 3.36 + 04-04 02:55:44 | [433][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0717 ntime: 0077 mem: 3.36 + 04-04 02:55:53 | [433][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1418 ntime: 0079 mem: 3.36 + 04-04 02:55:59 | Time info >>>> elapsed: 332.61 mins remain: 433.77 mins + 04-04 02:55:59 | [434][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0098 ntime: 0081 mem: 3.36 + 04-04 02:56:06 | [434][010/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0128 ntime: 0079 mem: 3.36 + 04-04 02:56:16 | [434][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1176 ntime: 0070 mem: 3.36 + 04-04 02:56:23 | [434][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1807 ntime: 0083 mem: 3.36 + 04-04 02:56:33 | [434][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1488 ntime: 0084 mem: 3.36 + 04-04 02:56:41 | [434][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1160 ntime: 0082 mem: 3.36 + 04-04 02:56:47 | [434][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0698 ntime: 0080 mem: 3.36 + 04-04 02:56:54 | [434][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1190 ntime: 0082 mem: 3.36 + 04-04 02:57:02 | [434][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0603 ntime: 0074 mem: 3.36 + 04-04 02:57:09 | [434][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1359 ntime: 0082 mem: 3.36 + 04-04 02:57:18 | [434][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0811 ntime: 0081 mem: 3.36 + 04-04 02:57:28 | [434][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1275 ntime: 0075 mem: 3.36 + 04-04 02:57:37 | [434][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0733 ntime: 0084 mem: 3.36 + 04-04 02:57:46 | [434][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0080 ntime: 0080 mem: 3.36 + 04-04 02:57:53 | [434][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1115 ntime: 0077 mem: 3.36 + 04-04 02:58:01 | [434][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0556 ntime: 0078 mem: 3.36 + 04-04 02:58:09 | [434][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0237 ntime: 0078 mem: 3.36 + 04-04 02:58:18 | [434][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0085 ntime: 0084 mem: 3.36 + 04-04 02:58:23 | Time info >>>> elapsed: 335.01 mins remain: 435.13 mins + 04-04 02:58:24 | [435][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1278 ntime: 0079 mem: 3.36 + 04-04 02:58:30 | [435][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0610 ntime: 0073 mem: 3.36 + 04-04 02:58:37 | [435][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0088 mem: 3.36 + 04-04 02:58:43 | [435][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1135 ntime: 0084 mem: 3.36 + 04-04 02:58:50 | [435][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0425 ntime: 0080 mem: 3.36 + 04-04 02:58:59 | [435][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0643 ntime: 0083 mem: 3.36 + 04-04 02:59:07 | [435][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0306 ntime: 0083 mem: 3.36 + 04-04 02:59:16 | [435][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0083 mem: 3.36 + 04-04 02:59:23 | [435][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0077 mem: 3.36 + 04-04 02:59:32 | [435][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0687 ntime: 0084 mem: 3.36 + 04-04 02:59:41 | [435][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0208 ntime: 0087 mem: 3.36 + 04-04 02:59:51 | [435][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1276 ntime: 0081 mem: 3.36 + 04-04 02:59:59 | [435][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1159 ntime: 0079 mem: 3.36 + 04-04 03:00:07 | [435][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0888 ntime: 0084 mem: 3.36 + 04-04 03:00:15 | [435][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0639 ntime: 0078 mem: 3.36 + 04-04 03:00:24 | [435][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1365 ntime: 0078 mem: 3.36 + 04-04 03:00:31 | [435][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0090 ntime: 0078 mem: 3.36 + 04-04 03:00:40 | [435][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1068 ntime: 0087 mem: 3.36 + 04-04 03:00:46 | Time info >>>> elapsed: 337.40 mins remain: 436.45 mins + 04-04 03:00:47 | [436][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0170 ntime: 0085 mem: 3.36 + 04-04 03:00:55 | [436][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0840 ntime: 0085 mem: 3.36 + 04-04 03:01:03 | [436][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0558 ntime: 0081 mem: 3.36 + 04-04 03:01:09 | [436][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0081 mem: 3.36 + 04-04 03:01:17 | [436][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0108 ntime: 0075 mem: 3.36 + 04-04 03:01:24 | [436][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0110 ntime: 0073 mem: 3.36 + 04-04 03:01:34 | [436][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0092 ntime: 0078 mem: 3.36 + 04-04 03:01:44 | [436][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0137 ntime: 0081 mem: 3.36 + 04-04 03:01:53 | [436][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0080 ntime: 0081 mem: 3.36 + 04-04 03:02:03 | [436][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0762 ntime: 0082 mem: 3.36 + 04-04 03:02:11 | [436][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1422 ntime: 0078 mem: 3.36 + 04-04 03:02:17 | [436][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1312 ntime: 0078 mem: 3.36 + 04-04 03:02:26 | [436][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0928 ntime: 0087 mem: 3.36 + 04-04 03:02:34 | [436][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0105 ntime: 0079 mem: 3.36 + 04-04 03:02:42 | [436][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0971 ntime: 0079 mem: 3.36 + 04-04 03:02:48 | [436][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0078 ntime: 0076 mem: 3.36 + 04-04 03:02:57 | [436][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0743 ntime: 0081 mem: 3.36 + 04-04 03:03:06 | [436][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1395 ntime: 0076 mem: 3.36 + 04-04 03:03:12 | Time info >>>> elapsed: 339.83 mins remain: 437.82 mins + 04-04 03:03:13 | [437][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0508 ntime: 0079 mem: 3.36 + 04-04 03:03:22 | [437][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1182 ntime: 0076 mem: 3.36 + 04-04 03:03:32 | [437][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1103 ntime: 0080 mem: 3.36 + 04-04 03:03:40 | [437][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0096 ntime: 0077 mem: 3.36 + 04-04 03:03:50 | [437][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1239 ntime: 0077 mem: 3.36 + 04-04 03:03:56 | [437][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0737 ntime: 0081 mem: 3.36 + 04-04 03:04:03 | [437][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1321 ntime: 0086 mem: 3.36 + 04-04 03:04:11 | [437][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0539 ntime: 0064 mem: 3.36 + 04-04 03:04:21 | [437][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0449 ntime: 0079 mem: 3.36 + 04-04 03:04:31 | [437][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0120 ntime: 0082 mem: 3.36 + 04-04 03:04:41 | [437][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1388 ntime: 0076 mem: 3.36 + 04-04 03:04:48 | [437][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0875 ntime: 0081 mem: 3.36 + 04-04 03:04:59 | [437][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1168 ntime: 0078 mem: 3.36 + 04-04 03:05:09 | [437][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0699 ntime: 0083 mem: 3.36 + 04-04 03:05:20 | [437][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1495 ntime: 0082 mem: 3.36 + 04-04 03:05:28 | [437][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1188 ntime: 0089 mem: 3.36 + 04-04 03:05:36 | [437][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0074 mem: 3.36 + 04-04 03:05:45 | [437][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1478 ntime: 0081 mem: 3.36 + 04-04 03:05:50 | Time info >>>> elapsed: 342.47 mins remain: 439.42 mins + 04-04 03:05:51 | [438][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0107 ntime: 0079 mem: 3.36 + 04-04 03:06:01 | [438][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0763 ntime: 0086 mem: 3.36 + 04-04 03:06:09 | [438][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0991 ntime: 0077 mem: 3.36 + 04-04 03:06:15 | [438][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0395 ntime: 0080 mem: 3.36 + 04-04 03:06:21 | [438][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0217 ntime: 0078 mem: 3.36 + 04-04 03:06:29 | [438][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1045 ntime: 0083 mem: 3.36 + 04-04 03:06:39 | [438][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0135 ntime: 0079 mem: 3.36 + 04-04 03:06:49 | [438][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1576 ntime: 0080 mem: 3.36 + 04-04 03:06:58 | [438][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1338 ntime: 0061 mem: 3.36 + 04-04 03:07:06 | [438][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1134 ntime: 0083 mem: 3.36 + 04-04 03:07:14 | [438][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0134 ntime: 0083 mem: 3.36 + 04-04 03:07:23 | [438][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1432 ntime: 0077 mem: 3.36 + 04-04 03:07:29 | [438][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0233 ntime: 0079 mem: 3.36 + 04-04 03:07:38 | [438][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0907 ntime: 0077 mem: 3.36 + 04-04 03:07:47 | [438][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0091 ntime: 0080 mem: 3.36 + 04-04 03:07:51 | [438][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0159 ntime: 0079 mem: 3.36 + 04-04 03:08:00 | [438][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0077 mem: 3.36 + 04-04 03:08:07 | [438][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1209 ntime: 0081 mem: 3.36 + 04-04 03:08:13 | Time info >>>> elapsed: 344.85 mins remain: 440.68 mins + 04-04 03:08:14 | [439][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0707 ntime: 0082 mem: 3.36 + 04-04 03:08:23 | [439][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1153 ntime: 0090 mem: 3.36 + 04-04 03:08:29 | [439][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0077 mem: 3.36 + 04-04 03:08:39 | [439][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1403 ntime: 0084 mem: 3.36 + 04-04 03:08:45 | [439][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0870 ntime: 0084 mem: 3.36 + 04-04 03:08:54 | [439][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1264 ntime: 0088 mem: 3.36 + 04-04 03:09:01 | [439][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 03:09:09 | [439][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0784 ntime: 0082 mem: 3.36 + 04-04 03:09:16 | [439][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1247 ntime: 0090 mem: 3.36 + 04-04 03:09:22 | [439][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1370 ntime: 0075 mem: 3.36 + 04-04 03:09:28 | [439][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0055 ntime: 0084 mem: 3.36 + 04-04 03:09:36 | [439][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0879 ntime: 0080 mem: 3.36 + 04-04 03:09:44 | [439][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1344 ntime: 0085 mem: 3.36 + 04-04 03:09:51 | [439][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 03:09:59 | [439][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1367 ntime: 0093 mem: 3.36 + 04-04 03:10:07 | [439][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0114 ntime: 0079 mem: 3.36 + 04-04 03:10:15 | [439][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1450 ntime: 0080 mem: 3.36 + 04-04 03:10:20 | [439][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0740 ntime: 0075 mem: 3.36 + 04-04 03:10:28 | Time info >>>> elapsed: 347.09 mins remain: 441.75 mins + 04-04 03:10:29 | [440][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0905 ntime: 0080 mem: 3.36 + 04-04 03:10:34 | [440][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0066 ntime: 0083 mem: 3.36 + 04-04 03:10:43 | [440][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1349 ntime: 0078 mem: 3.36 + 04-04 03:10:52 | [440][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1478 ntime: 0085 mem: 3.36 + 04-04 03:10:59 | [440][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0838 ntime: 0086 mem: 3.36 + 04-04 03:11:08 | [440][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1406 ntime: 0076 mem: 3.36 + 04-04 03:11:17 | [440][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0082 mem: 3.36 + 04-04 03:11:28 | [440][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1328 ntime: 0080 mem: 3.36 + 04-04 03:11:38 | [440][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1353 ntime: 0084 mem: 3.36 + 04-04 03:11:47 | [440][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0135 ntime: 0084 mem: 3.36 + 04-04 03:11:53 | [440][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1376 ntime: 0083 mem: 3.36 + 04-04 03:12:00 | [440][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0173 ntime: 0079 mem: 3.36 + 04-04 03:12:09 | [440][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0593 ntime: 0084 mem: 3.36 + 04-04 03:12:16 | [440][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1054 ntime: 0079 mem: 3.36 + 04-04 03:12:24 | [440][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0059 ntime: 0082 mem: 3.36 + 04-04 03:12:35 | [440][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0999 ntime: 0054 mem: 3.36 + 04-04 03:12:42 | [440][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0083 mem: 3.36 + 04-04 03:12:49 | [440][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0097 ntime: 0086 mem: 3.36 + 04-04 03:12:55 | Time info >>>> elapsed: 349.55 mins remain: 443.08 mins + 04-04 03:12:55 | [441][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0083 mem: 3.36 + 04-04 03:13:04 | [441][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1195 ntime: 0084 mem: 3.36 + 04-04 03:13:15 | [441][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0135 ntime: 0079 mem: 3.36 + 04-04 03:13:21 | [441][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0055 ntime: 0084 mem: 3.36 + 04-04 03:13:28 | [441][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0636 ntime: 0083 mem: 3.36 + 04-04 03:13:38 | [441][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0964 ntime: 0083 mem: 3.36 + 04-04 03:13:46 | [441][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1460 ntime: 0081 mem: 3.36 + 04-04 03:13:52 | [441][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1419 ntime: 0084 mem: 3.36 + 04-04 03:13:59 | [441][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0082 mem: 3.36 + 04-04 03:14:05 | [441][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0071 ntime: 0079 mem: 3.36 + 04-04 03:14:13 | [441][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0596 ntime: 0086 mem: 3.36 + 04-04 03:14:20 | [441][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1369 ntime: 0076 mem: 3.36 + 04-04 03:14:27 | [441][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0129 ntime: 0082 mem: 3.36 + 04-04 03:14:36 | [441][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0944 ntime: 0076 mem: 3.36 + 04-04 03:14:43 | [441][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0096 ntime: 0076 mem: 3.36 + 04-04 03:14:54 | [441][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0961 ntime: 0085 mem: 3.36 + 04-04 03:15:02 | [441][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1203 ntime: 0082 mem: 3.36 + 04-04 03:15:10 | [441][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1326 ntime: 0075 mem: 3.36 + 04-04 03:15:15 | Time info >>>> elapsed: 351.88 mins remain: 444.23 mins + 04-04 03:15:16 | [442][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1234 ntime: 0082 mem: 3.36 + 04-04 03:15:24 | [442][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0942 ntime: 0082 mem: 3.36 + 04-04 03:15:30 | [442][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1020 ntime: 0080 mem: 3.36 + 04-04 03:15:38 | [442][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0922 ntime: 0081 mem: 3.36 + 04-04 03:15:46 | [442][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0592 ntime: 0083 mem: 3.36 + 04-04 03:15:54 | [442][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0069 ntime: 0076 mem: 3.36 + 04-04 03:16:04 | [442][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1419 ntime: 0078 mem: 3.36 + 04-04 03:16:10 | [442][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0200 ntime: 0083 mem: 3.36 + 04-04 03:16:18 | [442][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1139 ntime: 0083 mem: 3.36 + 04-04 03:16:27 | [442][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0937 ntime: 0087 mem: 3.36 + 04-04 03:16:36 | [442][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1490 ntime: 0077 mem: 3.36 + 04-04 03:16:44 | [442][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0908 ntime: 0080 mem: 3.36 + 04-04 03:16:53 | [442][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1295 ntime: 0083 mem: 3.36 + 04-04 03:17:00 | [442][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0883 ntime: 0087 mem: 3.36 + 04-04 03:17:08 | [442][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0530 ntime: 0089 mem: 3.36 + 04-04 03:17:15 | [442][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0781 ntime: 0082 mem: 3.36 + 04-04 03:17:21 | [442][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0598 ntime: 0080 mem: 3.36 + 04-04 03:17:32 | [442][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1283 ntime: 0076 mem: 3.36 + 04-04 03:17:37 | Time info >>>> elapsed: 354.25 mins remain: 445.41 mins + 04-04 03:17:38 | [443][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0980 ntime: 0088 mem: 3.36 + 04-04 03:17:47 | [443][010/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0621 ntime: 0077 mem: 3.36 + 04-04 03:17:56 | [443][020/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1048 ntime: 0080 mem: 3.36 + 04-04 03:18:04 | [443][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0565 ntime: 0081 mem: 3.36 + 04-04 03:18:16 | [443][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0082 ntime: 0073 mem: 3.36 + 04-04 03:18:26 | [443][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0069 ntime: 0077 mem: 3.36 + 04-04 03:18:34 | [443][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1068 ntime: 0079 mem: 3.36 + 04-04 03:18:43 | [443][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1033 ntime: 0081 mem: 3.36 + 04-04 03:18:48 | [443][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0076 ntime: 0085 mem: 3.36 + 04-04 03:18:56 | [443][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0591 ntime: 0081 mem: 3.36 + 04-04 03:19:04 | [443][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0993 ntime: 0088 mem: 3.36 + 04-04 03:19:13 | [443][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1775 ntime: 0080 mem: 3.36 + 04-04 03:19:21 | [443][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1097 ntime: 0083 mem: 3.36 + 04-04 03:19:28 | [443][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0064 ntime: 0076 mem: 3.36 + 04-04 03:19:38 | [443][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1258 ntime: 0083 mem: 3.36 + 04-04 03:19:47 | [443][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0065 ntime: 0082 mem: 3.36 + 04-04 03:19:55 | [443][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1337 ntime: 0080 mem: 3.36 + 04-04 03:20:04 | [443][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1364 ntime: 0086 mem: 3.36 + 04-04 03:20:11 | Time info >>>> elapsed: 356.81 mins remain: 446.82 mins + 04-04 03:20:11 | [444][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0061 ntime: 0090 mem: 3.36 + 04-04 03:20:18 | [444][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0731 ntime: 0081 mem: 3.36 + 04-04 03:20:24 | [444][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0292 ntime: 0075 mem: 3.36 + 04-04 03:20:31 | [444][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0402 ntime: 0082 mem: 3.36 + 04-04 03:20:39 | [444][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1026 ntime: 0082 mem: 3.36 + 04-04 03:20:48 | [444][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0666 ntime: 0083 mem: 3.36 + 04-04 03:20:55 | [444][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0998 ntime: 0079 mem: 3.36 + 04-04 03:21:03 | [444][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1236 ntime: 0076 mem: 3.36 + 04-04 03:21:11 | [444][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0884 ntime: 0079 mem: 3.36 + 04-04 03:21:21 | [444][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0943 ntime: 0058 mem: 3.36 + 04-04 03:21:28 | [444][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0120 ntime: 0073 mem: 3.36 + 04-04 03:21:35 | [444][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0409 ntime: 0084 mem: 3.36 + 04-04 03:21:45 | [444][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1353 ntime: 0074 mem: 3.36 + 04-04 03:21:52 | [444][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0702 ntime: 0080 mem: 3.36 + 04-04 03:21:58 | [444][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1012 ntime: 0075 mem: 3.36 + 04-04 03:22:06 | [444][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1466 ntime: 0077 mem: 3.36 + 04-04 03:22:13 | [444][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0588 ntime: 0086 mem: 3.36 + 04-04 03:22:20 | [444][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0196 ntime: 0077 mem: 3.36 + 04-04 03:22:27 | Time info >>>> elapsed: 359.07 mins remain: 447.83 mins + 04-04 03:22:28 | [445][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1400 ntime: 0088 mem: 3.36 + 04-04 03:22:36 | [445][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1040 ntime: 0075 mem: 3.36 + 04-04 03:22:47 | [445][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1652 ntime: 0081 mem: 3.36 + 04-04 03:22:53 | [445][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0049 ntime: 0085 mem: 3.36 + 04-04 03:23:01 | [445][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1254 ntime: 0072 mem: 3.36 + 04-04 03:23:10 | [445][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0059 ntime: 0081 mem: 3.36 + 04-04 03:23:19 | [445][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 03:23:28 | [445][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1294 ntime: 0087 mem: 3.36 + 04-04 03:23:34 | [445][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0055 ntime: 0074 mem: 3.36 + 04-04 03:23:42 | [445][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1088 ntime: 0084 mem: 3.36 + 04-04 03:23:51 | [445][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0110 ntime: 0079 mem: 3.36 + 04-04 03:24:01 | [445][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1226 ntime: 0079 mem: 3.36 + 04-04 03:24:06 | [445][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0626 ntime: 0082 mem: 3.36 + 04-04 03:24:12 | [445][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0099 ntime: 0085 mem: 3.36 + 04-04 03:24:21 | [445][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0058 mem: 3.36 + 04-04 03:24:30 | [445][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0980 ntime: 0082 mem: 3.36 + 04-04 03:24:37 | [445][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1060 ntime: 0084 mem: 3.36 + 04-04 03:24:45 | [445][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1045 ntime: 0081 mem: 3.36 + 04-04 03:24:52 | Time info >>>> elapsed: 361.49 mins remain: 449.03 mins + 04-04 03:24:52 | [446][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0106 ntime: 0076 mem: 3.36 + 04-04 03:25:01 | [446][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0829 ntime: 0073 mem: 3.36 + 04-04 03:25:10 | [446][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1041 ntime: 0084 mem: 3.36 + 04-04 03:25:18 | [446][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1019 ntime: 0091 mem: 3.36 + 04-04 03:25:28 | [446][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1232 ntime: 0080 mem: 3.36 + 04-04 03:25:34 | [446][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0911 ntime: 0077 mem: 3.36 + 04-04 03:25:41 | [446][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0963 ntime: 0088 mem: 3.36 + 04-04 03:25:50 | [446][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0635 ntime: 0082 mem: 3.36 + 04-04 03:25:58 | [446][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0170 ntime: 0076 mem: 3.36 + 04-04 03:26:06 | [446][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0070 ntime: 0075 mem: 3.36 + 04-04 03:26:15 | [446][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0068 ntime: 0079 mem: 3.36 + 04-04 03:26:23 | [446][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0615 ntime: 0084 mem: 3.36 + 04-04 03:26:31 | [446][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0755 ntime: 0088 mem: 3.36 + 04-04 03:26:39 | [446][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1274 ntime: 0088 mem: 3.36 + 04-04 03:26:45 | [446][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0980 ntime: 0081 mem: 3.36 + 04-04 03:26:53 | [446][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1053 ntime: 0083 mem: 3.36 + 04-04 03:27:00 | [446][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0226 ntime: 0077 mem: 3.36 + 04-04 03:27:07 | [446][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0539 ntime: 0075 mem: 3.36 + 04-04 03:27:13 | Time info >>>> elapsed: 363.85 mins remain: 450.13 mins + 04-04 03:27:13 | [447][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0099 ntime: 0058 mem: 3.36 + 04-04 03:27:21 | [447][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0076 mem: 3.36 + 04-04 03:27:29 | [447][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1305 ntime: 0080 mem: 3.36 + 04-04 03:27:35 | [447][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0602 ntime: 0078 mem: 3.36 + 04-04 03:27:43 | [447][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 03:27:52 | [447][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0070 ntime: 0079 mem: 3.36 + 04-04 03:28:01 | [447][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1137 ntime: 0091 mem: 3.36 + 04-04 03:28:09 | [447][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1174 ntime: 0085 mem: 3.36 + 04-04 03:28:17 | [447][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0080 mem: 3.36 + 04-04 03:28:26 | [447][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0114 ntime: 0077 mem: 3.36 + 04-04 03:28:35 | [447][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1430 ntime: 0076 mem: 3.36 + 04-04 03:28:42 | [447][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1241 ntime: 0074 mem: 3.36 + 04-04 03:28:52 | [447][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1359 ntime: 0080 mem: 3.36 + 04-04 03:29:01 | [447][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0811 ntime: 0078 mem: 3.36 + 04-04 03:29:07 | [447][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0072 ntime: 0081 mem: 3.36 + 04-04 03:29:17 | [447][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0456 ntime: 0074 mem: 3.36 + 04-04 03:29:26 | [447][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1494 ntime: 0081 mem: 3.36 + 04-04 03:29:35 | [447][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1405 ntime: 0080 mem: 3.36 + 04-04 03:29:40 | Time info >>>> elapsed: 366.30 mins remain: 451.33 mins + 04-04 03:29:41 | [448][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0103 ntime: 0090 mem: 3.36 + 04-04 03:29:48 | [448][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0503 ntime: 0080 mem: 3.36 + 04-04 03:29:56 | [448][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0876 ntime: 0081 mem: 3.36 + 04-04 03:30:03 | [448][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0155 ntime: 0076 mem: 3.36 + 04-04 03:30:10 | [448][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1187 ntime: 0075 mem: 3.36 + 04-04 03:30:18 | [448][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0675 ntime: 0082 mem: 3.36 + 04-04 03:30:26 | [448][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1315 ntime: 0083 mem: 3.36 + 04-04 03:30:35 | [448][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0471 ntime: 0075 mem: 3.36 + 04-04 03:30:45 | [448][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 03:30:54 | [448][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1547 ntime: 0077 mem: 3.36 + 04-04 03:31:03 | [448][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1409 ntime: 0074 mem: 3.36 + 04-04 03:31:10 | [448][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1151 ntime: 0077 mem: 3.36 + 04-04 03:31:19 | [448][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0116 ntime: 0074 mem: 3.36 + 04-04 03:31:27 | [448][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0732 ntime: 0083 mem: 3.36 + 04-04 03:31:35 | [448][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0069 ntime: 0084 mem: 3.36 + 04-04 03:31:43 | [448][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1185 ntime: 0081 mem: 3.36 + 04-04 03:31:55 | [448][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1464 ntime: 0084 mem: 3.36 + 04-04 03:32:02 | [448][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1105 ntime: 0078 mem: 3.36 + 04-04 03:32:09 | Time info >>>> elapsed: 368.78 mins remain: 452.56 mins + 04-04 03:32:09 | [449][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0073 ntime: 0071 mem: 3.36 + 04-04 03:32:18 | [449][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0793 ntime: 0088 mem: 3.36 + 04-04 03:32:29 | [449][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1359 ntime: 0076 mem: 3.36 + 04-04 03:32:35 | [449][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0092 ntime: 0073 mem: 3.36 + 04-04 03:32:43 | [449][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1382 ntime: 0086 mem: 3.36 + 04-04 03:32:54 | [449][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1565 ntime: 0083 mem: 3.36 + 04-04 03:33:03 | [449][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1151 ntime: 0078 mem: 3.36 + 04-04 03:33:13 | [449][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1045 ntime: 0076 mem: 3.36 + 04-04 03:33:22 | [449][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1474 ntime: 0081 mem: 3.36 + 04-04 03:33:30 | [449][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1266 ntime: 0074 mem: 3.36 + 04-04 03:33:38 | [449][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0974 ntime: 0079 mem: 3.36 + 04-04 03:33:47 | [449][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1377 ntime: 0080 mem: 3.36 + 04-04 03:33:56 | [449][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1332 ntime: 0088 mem: 3.36 + 04-04 03:34:04 | [449][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1070 ntime: 0082 mem: 3.36 + 04-04 03:34:13 | [449][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1026 ntime: 0076 mem: 3.36 + 04-04 03:34:21 | [449][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0863 ntime: 0086 mem: 3.36 + 04-04 03:34:29 | [449][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0160 ntime: 0078 mem: 3.36 + 04-04 03:34:37 | [449][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0548 ntime: 0087 mem: 3.36 + 04-04 03:34:45 | Time info >>>> elapsed: 371.38 mins remain: 453.91 mins + 04-04 03:34:47 | [450][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1509 ntime: 0088 mem: 3.36 + 04-04 03:34:56 | [450][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0109 ntime: 0081 mem: 3.36 + 04-04 03:35:04 | [450][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1343 ntime: 0078 mem: 3.36 + 04-04 03:35:12 | [450][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1022 ntime: 0077 mem: 3.36 + 04-04 03:35:20 | [450][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1039 ntime: 0079 mem: 3.36 + 04-04 03:35:26 | [450][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1315 ntime: 0081 mem: 3.36 + 04-04 03:35:36 | [450][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0128 ntime: 0078 mem: 3.36 + 04-04 03:35:44 | [450][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1630 ntime: 0076 mem: 3.36 + 04-04 03:35:53 | [450][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0086 ntime: 0076 mem: 3.36 + 04-04 03:36:01 | [450][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1237 ntime: 0078 mem: 3.36 + 04-04 03:36:08 | [450][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1191 ntime: 0077 mem: 3.36 + 04-04 03:36:14 | [450][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0146 ntime: 0087 mem: 3.36 + 04-04 03:36:23 | [450][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0113 ntime: 0081 mem: 3.36 + 04-04 03:36:32 | [450][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1314 ntime: 0081 mem: 3.36 + 04-04 03:36:39 | [450][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0737 ntime: 0089 mem: 3.36 + 04-04 03:36:48 | [450][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1187 ntime: 0081 mem: 3.36 + 04-04 03:36:57 | [450][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1121 ntime: 0089 mem: 3.36 + 04-04 03:37:05 | [450][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0563 ntime: 0088 mem: 3.36 + 04-04 03:37:11 | Time info >>>> elapsed: 373.81 mins remain: 455.04 mins + 04-04 03:37:11 | [451][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0211 ntime: 0076 mem: 3.36 + 04-04 03:37:20 | [451][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0898 ntime: 0084 mem: 3.36 + 04-04 03:37:29 | [451][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0081 mem: 3.36 + 04-04 03:37:36 | [451][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0119 ntime: 0077 mem: 3.36 + 04-04 03:37:47 | [451][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0077 ntime: 0077 mem: 3.36 + 04-04 03:37:55 | [451][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1070 ntime: 0088 mem: 3.36 + 04-04 03:38:04 | [451][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0129 ntime: 0077 mem: 3.36 + 04-04 03:38:12 | [451][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0627 ntime: 0077 mem: 3.36 + 04-04 03:38:21 | [451][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0077 mem: 3.36 + 04-04 03:38:30 | [451][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0159 ntime: 0079 mem: 3.36 + 04-04 03:38:40 | [451][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0117 ntime: 0078 mem: 3.36 + 04-04 03:38:50 | [451][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1000 ntime: 0079 mem: 3.36 + 04-04 03:38:58 | [451][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 03:39:06 | [451][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0878 ntime: 0085 mem: 3.36 + 04-04 03:39:15 | [451][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1435 ntime: 0078 mem: 3.36 + 04-04 03:39:24 | [451][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0078 ntime: 0077 mem: 3.36 + 04-04 03:39:34 | [451][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0071 ntime: 0080 mem: 3.36 + 04-04 03:39:43 | [451][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0096 ntime: 0079 mem: 3.36 + 04-04 03:39:51 | Time info >>>> elapsed: 376.48 mins remain: 456.44 mins + 04-04 03:39:52 | [452][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1414 ntime: 0080 mem: 3.36 + 04-04 03:40:01 | [452][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0123 ntime: 0078 mem: 3.36 + 04-04 03:40:09 | [452][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1201 ntime: 0082 mem: 3.36 + 04-04 03:40:20 | [452][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1159 ntime: 0090 mem: 3.36 + 04-04 03:40:30 | [452][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1381 ntime: 0082 mem: 3.36 + 04-04 03:40:39 | [452][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0097 ntime: 0082 mem: 3.36 + 04-04 03:40:48 | [452][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1297 ntime: 0085 mem: 3.36 + 04-04 03:40:57 | [452][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0109 ntime: 0079 mem: 3.36 + 04-04 03:41:06 | [452][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0689 ntime: 0076 mem: 3.36 + 04-04 03:41:13 | [452][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0725 ntime: 0084 mem: 3.36 + 04-04 03:41:22 | [452][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1564 ntime: 0078 mem: 3.36 + 04-04 03:41:30 | [452][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0076 ntime: 0060 mem: 3.36 + 04-04 03:41:40 | [452][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1471 ntime: 0080 mem: 3.36 + 04-04 03:41:49 | [452][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0930 ntime: 0083 mem: 3.36 + 04-04 03:41:55 | [452][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0055 ntime: 0080 mem: 3.36 + 04-04 03:42:05 | [452][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0115 ntime: 0077 mem: 3.36 + 04-04 03:42:12 | [452][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0766 ntime: 0079 mem: 3.36 + 04-04 03:42:19 | [452][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0081 ntime: 0076 mem: 3.36 + 04-04 03:42:27 | Time info >>>> elapsed: 379.07 mins remain: 457.73 mins + 04-04 03:42:27 | [453][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0063 ntime: 0079 mem: 3.36 + 04-04 03:42:36 | [453][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0094 ntime: 0079 mem: 3.36 + 04-04 03:42:44 | [453][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0092 ntime: 0085 mem: 3.36 + 04-04 03:42:53 | [453][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0093 ntime: 0082 mem: 3.36 + 04-04 03:43:02 | [453][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0898 ntime: 0083 mem: 3.36 + 04-04 03:43:12 | [453][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0998 ntime: 0080 mem: 3.36 + 04-04 03:43:18 | [453][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0080 mem: 3.36 + 04-04 03:43:26 | [453][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1565 ntime: 0082 mem: 3.36 + 04-04 03:43:34 | [453][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0680 ntime: 0084 mem: 3.36 + 04-04 03:43:40 | [453][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0082 ntime: 0082 mem: 3.36 + 04-04 03:43:49 | [453][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1358 ntime: 0086 mem: 3.36 + 04-04 03:43:59 | [453][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0252 ntime: 0080 mem: 3.36 + 04-04 03:44:08 | [453][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0429 ntime: 0076 mem: 3.36 + 04-04 03:44:18 | [453][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1010 ntime: 0082 mem: 3.36 + 04-04 03:44:26 | [453][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1389 ntime: 0082 mem: 3.36 + 04-04 03:44:33 | [453][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 03:44:44 | [453][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0059 ntime: 0082 mem: 3.36 + 04-04 03:44:54 | [453][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0918 ntime: 0083 mem: 3.36 + 04-04 03:45:00 | Time info >>>> elapsed: 381.63 mins remain: 458.96 mins + 04-04 03:45:01 | [454][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1014 ntime: 0084 mem: 3.36 + 04-04 03:45:10 | [454][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0913 ntime: 0083 mem: 3.36 + 04-04 03:45:18 | [454][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0913 ntime: 0080 mem: 3.36 + 04-04 03:45:27 | [454][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0853 ntime: 0091 mem: 3.36 + 04-04 03:45:36 | [454][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1529 ntime: 0083 mem: 3.36 + 04-04 03:45:45 | [454][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1369 ntime: 0080 mem: 3.36 + 04-04 03:45:53 | [454][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0494 ntime: 0084 mem: 3.36 + 04-04 03:46:03 | [454][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1257 ntime: 0080 mem: 3.36 + 04-04 03:46:11 | [454][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0534 ntime: 0082 mem: 3.36 + 04-04 03:46:19 | [454][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1434 ntime: 0083 mem: 3.36 + 04-04 03:46:27 | [454][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0133 ntime: 0079 mem: 3.36 + 04-04 03:46:34 | [454][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0191 ntime: 0075 mem: 3.36 + 04-04 03:46:44 | [454][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1281 ntime: 0086 mem: 3.36 + 04-04 03:46:53 | [454][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0858 ntime: 0087 mem: 3.36 + 04-04 03:47:03 | [454][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1845 ntime: 0081 mem: 3.36 + 04-04 03:47:11 | [454][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0595 ntime: 0083 mem: 3.36 + 04-04 03:47:16 | [454][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0080 mem: 3.36 + 04-04 03:47:26 | [454][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1393 ntime: 0082 mem: 3.36 + 04-04 03:47:33 | Time info >>>> elapsed: 384.17 mins remain: 460.16 mins + 04-04 03:47:33 | [455][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0074 ntime: 0073 mem: 3.36 + 04-04 03:47:40 | [455][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1080 ntime: 0086 mem: 3.36 + 04-04 03:47:50 | [455][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1472 ntime: 0088 mem: 3.36 + 04-04 03:47:58 | [455][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0752 ntime: 0085 mem: 3.36 + 04-04 03:48:06 | [455][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0601 ntime: 0086 mem: 3.36 + 04-04 03:48:12 | [455][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0156 ntime: 0080 mem: 3.36 + 04-04 03:48:22 | [455][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1020 ntime: 0077 mem: 3.36 + 04-04 03:48:29 | [455][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0047 ntime: 0077 mem: 3.36 + 04-04 03:48:36 | [455][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0095 ntime: 0084 mem: 3.36 + 04-04 03:48:44 | [455][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0213 ntime: 0085 mem: 3.36 + 04-04 03:48:51 | [455][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0071 ntime: 0074 mem: 3.36 + 04-04 03:48:57 | [455][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0298 ntime: 0085 mem: 3.36 + 04-04 03:49:05 | [455][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1001 ntime: 0082 mem: 3.36 + 04-04 03:49:13 | [455][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0097 ntime: 0090 mem: 3.36 + 04-04 03:49:20 | [455][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1073 ntime: 0085 mem: 3.36 + 04-04 03:49:27 | [455][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0581 ntime: 0083 mem: 3.36 + 04-04 03:49:37 | [455][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1360 ntime: 0083 mem: 3.36 + 04-04 03:49:44 | [455][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0131 ntime: 0080 mem: 3.36 + 04-04 03:49:51 | Time info >>>> elapsed: 386.48 mins remain: 461.07 mins + 04-04 03:49:53 | [456][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1638 ntime: 0081 mem: 3.36 + 04-04 03:50:01 | [456][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0074 ntime: 0080 mem: 3.36 + 04-04 03:50:08 | [456][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0892 ntime: 0078 mem: 3.36 + 04-04 03:50:15 | [456][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0984 ntime: 0088 mem: 3.36 + 04-04 03:50:24 | [456][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1268 ntime: 0087 mem: 3.36 + 04-04 03:50:32 | [456][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0369 ntime: 0084 mem: 3.36 + 04-04 03:50:38 | [456][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0078 ntime: 0088 mem: 3.36 + 04-04 03:50:45 | [456][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 03:50:55 | [456][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1328 ntime: 0084 mem: 3.36 + 04-04 03:51:03 | [456][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0085 ntime: 0081 mem: 3.36 + 04-04 03:51:11 | [456][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0092 mem: 3.36 + 04-04 03:51:20 | [456][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1027 ntime: 0082 mem: 3.36 + 04-04 03:51:28 | [456][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1367 ntime: 0080 mem: 3.36 + 04-04 03:51:37 | [456][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1249 ntime: 0086 mem: 3.36 + 04-04 03:51:46 | [456][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1469 ntime: 0074 mem: 3.36 + 04-04 03:51:54 | [456][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1039 ntime: 0079 mem: 3.36 + 04-04 03:52:05 | [456][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1100 ntime: 0090 mem: 3.36 + 04-04 03:52:14 | [456][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0117 ntime: 0075 mem: 3.36 + 04-04 03:52:19 | Time info >>>> elapsed: 388.95 mins remain: 462.15 mins + 04-04 03:52:21 | [457][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1334 ntime: 0082 mem: 3.36 + 04-04 03:52:28 | [457][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0516 ntime: 0080 mem: 3.36 + 04-04 03:52:35 | [457][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1085 ntime: 0088 mem: 3.36 + 04-04 03:52:43 | [457][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0062 ntime: 0080 mem: 3.36 + 04-04 03:52:52 | [457][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1080 ntime: 0077 mem: 3.36 + 04-04 03:52:59 | [457][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0524 ntime: 0079 mem: 3.36 + 04-04 03:53:07 | [457][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0772 ntime: 0085 mem: 3.36 + 04-04 03:53:19 | [457][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1918 ntime: 0087 mem: 3.36 + 04-04 03:53:29 | [457][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1234 ntime: 0076 mem: 3.36 + 04-04 03:53:36 | [457][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0660 ntime: 0080 mem: 3.36 + 04-04 03:53:41 | [457][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0086 ntime: 0089 mem: 3.36 + 04-04 03:53:47 | [457][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1041 ntime: 0084 mem: 3.36 + 04-04 03:53:55 | [457][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0107 ntime: 0080 mem: 3.36 + 04-04 03:54:04 | [457][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1473 ntime: 0086 mem: 3.36 + 04-04 03:54:10 | [457][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0375 ntime: 0085 mem: 3.36 + 04-04 03:54:19 | [457][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0492 ntime: 0078 mem: 3.36 + 04-04 03:54:26 | [457][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0373 ntime: 0079 mem: 3.36 + 04-04 03:54:34 | [457][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1260 ntime: 0079 mem: 3.36 + 04-04 03:54:37 | Time info >>>> elapsed: 391.25 mins remain: 463.01 mins + 04-04 03:54:38 | [458][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0732 ntime: 0087 mem: 3.36 + 04-04 03:54:46 | [458][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0582 ntime: 0083 mem: 3.36 + 04-04 03:54:56 | [458][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0109 ntime: 0079 mem: 3.36 + 04-04 03:55:06 | [458][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1150 ntime: 0085 mem: 3.36 + 04-04 03:55:12 | [458][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0209 ntime: 0077 mem: 3.36 + 04-04 03:55:21 | [458][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0604 ntime: 0079 mem: 3.36 + 04-04 03:55:29 | [458][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0876 ntime: 0083 mem: 3.36 + 04-04 03:55:38 | [458][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1273 ntime: 0085 mem: 3.36 + 04-04 03:55:49 | [458][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1064 ntime: 0081 mem: 3.36 + 04-04 03:55:57 | [458][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0955 ntime: 0084 mem: 3.36 + 04-04 03:56:05 | [458][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0079 mem: 3.36 + 04-04 03:56:13 | [458][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0111 ntime: 0084 mem: 3.36 + 04-04 03:56:21 | [458][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0087 mem: 3.36 + 04-04 03:56:31 | [458][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0076 mem: 3.36 + 04-04 03:56:41 | [458][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0104 ntime: 0078 mem: 3.36 + 04-04 03:56:50 | [458][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0610 ntime: 0080 mem: 3.36 + 04-04 03:56:56 | [458][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0621 ntime: 0076 mem: 3.36 + 04-04 03:57:03 | [458][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0803 ntime: 0084 mem: 3.36 + 04-04 03:57:12 | Time info >>>> elapsed: 393.83 mins remain: 464.19 mins + 04-04 03:57:14 | [459][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1273 ntime: 0089 mem: 3.36 + 04-04 03:57:24 | [459][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0584 ntime: 0080 mem: 3.36 + 04-04 03:57:33 | [459][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0081 ntime: 0085 mem: 3.36 + 04-04 03:57:43 | [459][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1251 ntime: 0082 mem: 3.36 + 04-04 03:57:54 | [459][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0915 ntime: 0088 mem: 3.36 + 04-04 03:58:03 | [459][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1133 ntime: 0076 mem: 3.36 + 04-04 03:58:13 | [459][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0198 ntime: 0077 mem: 3.36 + 04-04 03:58:23 | [459][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1392 ntime: 0079 mem: 3.36 + 04-04 03:58:33 | [459][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0917 ntime: 0086 mem: 3.36 + 04-04 03:58:41 | [459][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0757 ntime: 0080 mem: 3.36 + 04-04 03:58:51 | [459][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0707 ntime: 0077 mem: 3.36 + 04-04 03:59:01 | [459][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1457 ntime: 0079 mem: 3.36 + 04-04 03:59:10 | [459][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0079 mem: 3.36 + 04-04 03:59:18 | [459][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1349 ntime: 0078 mem: 3.36 + 04-04 03:59:28 | [459][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1002 ntime: 0075 mem: 3.36 + 04-04 03:59:37 | [459][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0065 ntime: 0079 mem: 3.36 + 04-04 03:59:47 | [459][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0887 ntime: 0084 mem: 3.36 + 04-04 03:59:59 | [459][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1021 ntime: 0085 mem: 3.36 + 04-04 04:00:05 | Time info >>>> elapsed: 396.71 mins remain: 465.71 mins + 04-04 04:00:06 | [460][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0977 ntime: 0082 mem: 3.36 + 04-04 04:00:15 | [460][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0074 ntime: 0086 mem: 3.36 + 04-04 04:00:25 | [460][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1359 ntime: 0082 mem: 3.36 + 04-04 04:00:34 | [460][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0852 ntime: 0082 mem: 3.36 + 04-04 04:00:44 | [460][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1232 ntime: 0078 mem: 3.36 + 04-04 04:00:52 | [460][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0988 ntime: 0080 mem: 3.36 + 04-04 04:00:59 | [460][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0132 ntime: 0086 mem: 3.36 + 04-04 04:01:07 | [460][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0544 ntime: 0084 mem: 3.36 + 04-04 04:01:14 | [460][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0572 ntime: 0089 mem: 3.36 + 04-04 04:01:25 | [460][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1187 ntime: 0080 mem: 3.36 + 04-04 04:01:30 | [460][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0822 ntime: 0079 mem: 3.36 + 04-04 04:01:38 | [460][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0082 ntime: 0083 mem: 3.36 + 04-04 04:01:48 | [460][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1102 ntime: 0083 mem: 3.36 + 04-04 04:01:58 | [460][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1375 ntime: 0088 mem: 3.36 + 04-04 04:02:08 | [460][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0849 ntime: 0081 mem: 3.36 + 04-04 04:02:16 | [460][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1168 ntime: 0084 mem: 3.36 + 04-04 04:02:23 | [460][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0536 ntime: 0072 mem: 3.36 + 04-04 04:02:30 | [460][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0156 ntime: 0074 mem: 3.36 + 04-04 04:02:36 | Time info >>>> elapsed: 399.24 mins remain: 466.79 mins + 04-04 04:02:37 | [461][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0154 ntime: 0079 mem: 3.36 + 04-04 04:02:47 | [461][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0656 ntime: 0082 mem: 3.36 + 04-04 04:02:55 | [461][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1245 ntime: 0084 mem: 3.36 + 04-04 04:03:03 | [461][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0739 ntime: 0083 mem: 3.36 + 04-04 04:03:12 | [461][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0781 ntime: 0083 mem: 3.36 + 04-04 04:03:22 | [461][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0110 ntime: 0073 mem: 3.36 + 04-04 04:03:30 | [461][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1352 ntime: 0079 mem: 3.36 + 04-04 04:03:39 | [461][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0175 ntime: 0084 mem: 3.36 + 04-04 04:03:47 | [461][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0585 ntime: 0081 mem: 3.36 + 04-04 04:03:55 | [461][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 04:04:07 | [461][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1374 ntime: 0085 mem: 3.36 + 04-04 04:04:16 | [461][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1134 ntime: 0080 mem: 3.36 + 04-04 04:04:26 | [461][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0759 ntime: 0078 mem: 3.36 + 04-04 04:04:37 | [461][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1229 ntime: 0057 mem: 3.36 + 04-04 04:04:46 | [461][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0138 ntime: 0078 mem: 3.36 + 04-04 04:04:56 | [461][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0818 ntime: 0083 mem: 3.36 + 04-04 04:05:03 | [461][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0514 ntime: 0078 mem: 3.36 + 04-04 04:05:13 | [461][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1067 ntime: 0087 mem: 3.36 + 04-04 04:05:21 | Time info >>>> elapsed: 401.97 mins remain: 468.10 mins + 04-04 04:05:22 | [462][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1086 ntime: 0076 mem: 3.36 + 04-04 04:05:29 | [462][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0935 ntime: 0087 mem: 3.36 + 04-04 04:05:38 | [462][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0091 ntime: 0084 mem: 3.36 + 04-04 04:05:47 | [462][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1132 ntime: 0081 mem: 3.36 + 04-04 04:05:55 | [462][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1244 ntime: 0081 mem: 3.36 + 04-04 04:06:03 | [462][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1192 ntime: 0079 mem: 3.36 + 04-04 04:06:09 | [462][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0521 ntime: 0087 mem: 3.36 + 04-04 04:06:17 | [462][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0503 ntime: 0083 mem: 3.36 + 04-04 04:06:27 | [462][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0086 ntime: 0081 mem: 3.36 + 04-04 04:06:33 | [462][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0578 ntime: 0078 mem: 3.36 + 04-04 04:06:42 | [462][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0819 ntime: 0081 mem: 3.36 + 04-04 04:06:50 | [462][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0109 ntime: 0077 mem: 3.36 + 04-04 04:06:58 | [462][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0080 ntime: 0085 mem: 3.36 + 04-04 04:07:07 | [462][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1410 ntime: 0082 mem: 3.36 + 04-04 04:07:16 | [462][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1225 ntime: 0081 mem: 3.36 + 04-04 04:07:25 | [462][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1255 ntime: 0083 mem: 3.36 + 04-04 04:07:33 | [462][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0996 ntime: 0081 mem: 3.36 + 04-04 04:07:40 | [462][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0726 ntime: 0080 mem: 3.36 + 04-04 04:07:49 | Time info >>>> elapsed: 404.44 mins remain: 469.08 mins + 04-04 04:07:50 | [463][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0861 ntime: 0083 mem: 3.36 + 04-04 04:07:57 | [463][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0093 ntime: 0086 mem: 3.36 + 04-04 04:08:06 | [463][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0045 ntime: 0078 mem: 3.36 + 04-04 04:08:13 | [463][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0062 ntime: 0079 mem: 3.36 + 04-04 04:08:21 | [463][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0064 ntime: 0075 mem: 3.36 + 04-04 04:08:31 | [463][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0080 mem: 3.36 + 04-04 04:08:38 | [463][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1244 ntime: 0078 mem: 3.36 + 04-04 04:08:46 | [463][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0746 ntime: 0079 mem: 3.36 + 04-04 04:08:54 | [463][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0647 ntime: 0085 mem: 3.36 + 04-04 04:09:02 | [463][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0118 ntime: 0079 mem: 3.36 + 04-04 04:09:09 | [463][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0998 ntime: 0077 mem: 3.36 + 04-04 04:09:17 | [463][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0638 ntime: 0079 mem: 3.36 + 04-04 04:09:23 | [463][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0707 ntime: 0088 mem: 3.36 + 04-04 04:09:32 | [463][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0075 mem: 3.36 + 04-04 04:09:41 | [463][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0102 ntime: 0080 mem: 3.36 + 04-04 04:09:49 | [463][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0079 mem: 3.36 + 04-04 04:09:57 | [463][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0734 ntime: 0081 mem: 3.36 + 04-04 04:10:05 | [463][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0953 ntime: 0088 mem: 3.36 + 04-04 04:10:13 | Time info >>>> elapsed: 406.85 mins remain: 469.98 mins + 04-04 04:10:13 | [464][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0047 ntime: 0077 mem: 3.36 + 04-04 04:10:22 | [464][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0055 ntime: 0076 mem: 3.36 + 04-04 04:10:32 | [464][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1304 ntime: 0078 mem: 3.36 + 04-04 04:10:39 | [464][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0073 ntime: 0072 mem: 3.36 + 04-04 04:10:48 | [464][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1058 ntime: 0077 mem: 3.36 + 04-04 04:10:55 | [464][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0073 mem: 3.36 + 04-04 04:11:04 | [464][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1056 ntime: 0077 mem: 3.36 + 04-04 04:11:12 | [464][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1207 ntime: 0076 mem: 3.36 + 04-04 04:11:19 | [464][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1585 ntime: 0088 mem: 3.36 + 04-04 04:11:27 | [464][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0074 mem: 3.36 + 04-04 04:11:34 | [464][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0816 ntime: 0089 mem: 3.36 + 04-04 04:11:42 | [464][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0116 ntime: 0084 mem: 3.36 + 04-04 04:11:50 | [464][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0092 ntime: 0081 mem: 3.36 + 04-04 04:11:59 | [464][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0818 ntime: 0076 mem: 3.36 + 04-04 04:12:06 | [464][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0594 ntime: 0082 mem: 3.36 + 04-04 04:12:14 | [464][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0095 ntime: 0079 mem: 3.36 + 04-04 04:12:21 | [464][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0196 ntime: 0079 mem: 3.36 + 04-04 04:12:31 | [464][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0794 ntime: 0087 mem: 3.36 + 04-04 04:12:39 | Time info >>>> elapsed: 409.28 mins remain: 470.90 mins + 04-04 04:12:41 | [465][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1273 ntime: 0083 mem: 3.36 + 04-04 04:12:49 | [465][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1235 ntime: 0085 mem: 3.36 + 04-04 04:12:58 | [465][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1503 ntime: 0080 mem: 3.36 + 04-04 04:13:11 | [465][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1133 ntime: 0086 mem: 3.36 + 04-04 04:13:20 | [465][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1038 ntime: 0083 mem: 3.36 + 04-04 04:13:28 | [465][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1051 ntime: 0083 mem: 3.36 + 04-04 04:13:36 | [465][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1369 ntime: 0078 mem: 3.36 + 04-04 04:13:45 | [465][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0806 ntime: 0075 mem: 3.36 + 04-04 04:13:53 | [465][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1067 ntime: 0085 mem: 3.36 + 04-04 04:14:03 | [465][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1409 ntime: 0078 mem: 3.36 + 04-04 04:14:12 | [465][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1223 ntime: 0072 mem: 3.36 + 04-04 04:14:19 | [465][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 04:14:28 | [465][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1211 ntime: 0087 mem: 3.36 + 04-04 04:14:37 | [465][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0621 ntime: 0084 mem: 3.36 + 04-04 04:14:45 | [465][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1104 ntime: 0076 mem: 3.36 + 04-04 04:14:53 | [465][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0635 ntime: 0077 mem: 3.36 + 04-04 04:15:02 | [465][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1192 ntime: 0086 mem: 3.36 + 04-04 04:15:09 | [465][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0076 mem: 3.36 + 04-04 04:15:14 | Time info >>>> elapsed: 411.85 mins remain: 471.95 mins + 04-04 04:15:15 | [466][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1443 ntime: 0079 mem: 3.36 + 04-04 04:15:23 | [466][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0096 ntime: 0084 mem: 3.36 + 04-04 04:15:33 | [466][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1306 ntime: 0092 mem: 3.36 + 04-04 04:15:43 | [466][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1007 ntime: 0077 mem: 3.36 + 04-04 04:15:53 | [466][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1426 ntime: 0080 mem: 3.36 + 04-04 04:15:59 | [466][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0970 ntime: 0081 mem: 3.36 + 04-04 04:16:08 | [466][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0971 ntime: 0055 mem: 3.36 + 04-04 04:16:16 | [466][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1279 ntime: 0077 mem: 3.36 + 04-04 04:16:23 | [466][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0530 ntime: 0080 mem: 3.36 + 04-04 04:16:30 | [466][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0083 ntime: 0079 mem: 3.36 + 04-04 04:16:37 | [466][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0885 ntime: 0089 mem: 3.36 + 04-04 04:16:44 | [466][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0525 ntime: 0076 mem: 3.36 + 04-04 04:16:50 | [466][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0079 mem: 3.36 + 04-04 04:17:00 | [466][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1035 ntime: 0078 mem: 3.36 + 04-04 04:17:08 | [466][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0085 ntime: 0080 mem: 3.36 + 04-04 04:17:15 | [466][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0944 ntime: 0077 mem: 3.36 + 04-04 04:17:21 | [466][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0077 mem: 3.36 + 04-04 04:17:28 | [466][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0876 ntime: 0083 mem: 3.36 + 04-04 04:17:36 | Time info >>>> elapsed: 414.22 mins remain: 472.76 mins + 04-04 04:17:37 | [467][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1021 ntime: 0077 mem: 3.36 + 04-04 04:17:43 | [467][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0972 ntime: 0072 mem: 3.36 + 04-04 04:17:52 | [467][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1389 ntime: 0086 mem: 3.36 + 04-04 04:18:00 | [467][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0071 ntime: 0088 mem: 3.36 + 04-04 04:18:07 | [467][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0105 ntime: 0082 mem: 3.36 + 04-04 04:18:15 | [467][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1011 ntime: 0082 mem: 3.36 + 04-04 04:18:25 | [467][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1228 ntime: 0081 mem: 3.36 + 04-04 04:18:31 | [467][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0093 ntime: 0089 mem: 3.36 + 04-04 04:18:40 | [467][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0627 ntime: 0077 mem: 3.36 + 04-04 04:18:45 | [467][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0082 ntime: 0073 mem: 3.36 + 04-04 04:18:55 | [467][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1060 ntime: 0077 mem: 3.36 + 04-04 04:19:04 | [467][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0075 ntime: 0073 mem: 3.36 + 04-04 04:19:14 | [467][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0687 ntime: 0076 mem: 3.36 + 04-04 04:19:20 | [467][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0717 ntime: 0082 mem: 3.36 + 04-04 04:19:26 | [467][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 04:19:36 | [467][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0152 ntime: 0085 mem: 3.36 + 04-04 04:19:45 | [467][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0062 ntime: 0080 mem: 3.36 + 04-04 04:19:54 | [467][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1276 ntime: 0075 mem: 3.36 + 04-04 04:20:01 | Time info >>>> elapsed: 416.65 mins remain: 473.63 mins + 04-04 04:20:03 | [468][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1294 ntime: 0076 mem: 3.36 + 04-04 04:20:13 | [468][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1300 ntime: 0079 mem: 3.36 + 04-04 04:20:21 | [468][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0494 ntime: 0078 mem: 3.36 + 04-04 04:20:29 | [468][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1063 ntime: 0082 mem: 3.36 + 04-04 04:20:40 | [468][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1243 ntime: 0081 mem: 3.36 + 04-04 04:20:50 | [468][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1563 ntime: 0087 mem: 3.36 + 04-04 04:20:59 | [468][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0107 ntime: 0077 mem: 3.36 + 04-04 04:21:06 | [468][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0090 ntime: 0080 mem: 3.36 + 04-04 04:21:14 | [468][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0062 ntime: 0081 mem: 3.36 + 04-04 04:21:22 | [468][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0864 ntime: 0083 mem: 3.36 + 04-04 04:21:30 | [468][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1041 ntime: 0086 mem: 3.36 + 04-04 04:21:40 | [468][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1127 ntime: 0076 mem: 3.36 + 04-04 04:21:48 | [468][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0251 ntime: 0085 mem: 3.36 + 04-04 04:22:00 | [468][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1439 ntime: 0074 mem: 3.36 + 04-04 04:22:08 | [468][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1279 ntime: 0082 mem: 3.36 + 04-04 04:22:16 | [468][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0088 ntime: 0088 mem: 3.36 + 04-04 04:22:26 | [468][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1359 ntime: 0088 mem: 3.36 + 04-04 04:22:34 | [468][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1354 ntime: 0071 mem: 3.36 + 04-04 04:22:40 | Time info >>>> elapsed: 419.30 mins remain: 474.73 mins + 04-04 04:22:41 | [469][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0292 ntime: 0077 mem: 3.36 + 04-04 04:22:48 | [469][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0918 ntime: 0083 mem: 3.36 + 04-04 04:22:54 | [469][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0364 ntime: 0079 mem: 3.36 + 04-04 04:23:04 | [469][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0104 ntime: 0078 mem: 3.36 + 04-04 04:23:13 | [469][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0095 ntime: 0075 mem: 3.36 + 04-04 04:23:20 | [469][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0800 ntime: 0078 mem: 3.36 + 04-04 04:23:29 | [469][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0084 mem: 3.36 + 04-04 04:23:37 | [469][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0698 ntime: 0081 mem: 3.36 + 04-04 04:23:45 | [469][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1097 ntime: 0084 mem: 3.36 + 04-04 04:23:55 | [469][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0744 ntime: 0083 mem: 3.36 + 04-04 04:24:05 | [469][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1567 ntime: 0087 mem: 3.36 + 04-04 04:24:13 | [469][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1135 ntime: 0084 mem: 3.36 + 04-04 04:24:20 | [469][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0957 ntime: 0075 mem: 3.36 + 04-04 04:24:28 | [469][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1253 ntime: 0083 mem: 3.36 + 04-04 04:24:37 | [469][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1333 ntime: 0084 mem: 3.36 + 04-04 04:24:46 | [469][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0770 ntime: 0079 mem: 3.36 + 04-04 04:24:52 | [469][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0079 mem: 3.36 + 04-04 04:24:59 | [469][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0349 ntime: 0101 mem: 3.36 + 04-04 04:25:06 | Time info >>>> elapsed: 421.73 mins remain: 475.57 mins + 04-04 04:25:06 | [470][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0120 ntime: 0078 mem: 3.36 + 04-04 04:25:14 | [470][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0074 ntime: 0074 mem: 3.36 + 04-04 04:25:22 | [470][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0819 ntime: 0077 mem: 3.36 + 04-04 04:25:30 | [470][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0826 ntime: 0088 mem: 3.36 + 04-04 04:25:39 | [470][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0965 ntime: 0085 mem: 3.36 + 04-04 04:25:48 | [470][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1196 ntime: 0079 mem: 3.36 + 04-04 04:25:58 | [470][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1637 ntime: 0078 mem: 3.36 + 04-04 04:26:06 | [470][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1415 ntime: 0077 mem: 3.36 + 04-04 04:26:12 | [470][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0106 ntime: 0080 mem: 3.36 + 04-04 04:26:22 | [470][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1109 ntime: 0082 mem: 3.36 + 04-04 04:26:33 | [470][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1225 ntime: 0086 mem: 3.36 + 04-04 04:26:43 | [470][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1356 ntime: 0085 mem: 3.36 + 04-04 04:26:52 | [470][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0167 ntime: 0080 mem: 3.36 + 04-04 04:26:59 | [470][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1097 ntime: 0081 mem: 3.36 + 04-04 04:27:05 | [470][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0191 ntime: 0083 mem: 3.36 + 04-04 04:27:13 | [470][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0080 mem: 3.36 + 04-04 04:27:20 | [470][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0324 ntime: 0082 mem: 3.36 + 04-04 04:27:27 | [470][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0076 ntime: 0074 mem: 3.36 + 04-04 04:27:34 | Time info >>>> elapsed: 424.19 mins remain: 476.43 mins + 04-04 04:27:35 | [471][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1061 ntime: 0085 mem: 3.36 + 04-04 04:27:42 | [471][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1343 ntime: 0082 mem: 3.36 + 04-04 04:27:47 | [471][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0895 ntime: 0073 mem: 3.36 + 04-04 04:27:54 | [471][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0526 ntime: 0077 mem: 3.36 + 04-04 04:28:01 | [471][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0156 ntime: 0081 mem: 3.36 + 04-04 04:28:09 | [471][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1165 ntime: 0081 mem: 3.36 + 04-04 04:28:18 | [471][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0924 ntime: 0087 mem: 3.36 + 04-04 04:28:27 | [471][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1025 ntime: 0077 mem: 3.36 + 04-04 04:28:35 | [471][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0870 ntime: 0078 mem: 3.36 + 04-04 04:28:44 | [471][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1172 ntime: 0078 mem: 3.36 + 04-04 04:28:51 | [471][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0107 ntime: 0081 mem: 3.36 + 04-04 04:28:59 | [471][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0861 ntime: 0084 mem: 3.36 + 04-04 04:29:09 | [471][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1431 ntime: 0083 mem: 3.36 + 04-04 04:29:18 | [471][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1414 ntime: 0079 mem: 3.36 + 04-04 04:29:25 | [471][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1156 ntime: 0086 mem: 3.36 + 04-04 04:29:32 | [471][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0890 ntime: 0083 mem: 3.36 + 04-04 04:29:40 | [471][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0087 ntime: 0073 mem: 3.36 + 04-04 04:29:50 | [471][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0928 ntime: 0081 mem: 3.36 + 04-04 04:29:55 | Time info >>>> elapsed: 426.54 mins remain: 477.14 mins + 04-04 04:29:55 | [472][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0831 ntime: 0081 mem: 3.36 + 04-04 04:30:05 | [472][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0778 ntime: 0090 mem: 3.36 + 04-04 04:30:13 | [472][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0079 mem: 3.36 + 04-04 04:30:21 | [472][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1590 ntime: 0076 mem: 3.36 + 04-04 04:30:30 | [472][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0963 ntime: 0084 mem: 3.36 + 04-04 04:30:39 | [472][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1352 ntime: 0080 mem: 3.36 + 04-04 04:30:46 | [472][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1109 ntime: 0078 mem: 3.36 + 04-04 04:30:54 | [472][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0082 mem: 3.36 + 04-04 04:31:05 | [472][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0877 ntime: 0087 mem: 3.36 + 04-04 04:31:14 | [472][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1008 ntime: 0078 mem: 3.36 + 04-04 04:31:21 | [472][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1208 ntime: 0081 mem: 3.36 + 04-04 04:31:28 | [472][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1066 ntime: 0074 mem: 3.36 + 04-04 04:31:36 | [472][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0078 ntime: 0079 mem: 3.36 + 04-04 04:31:45 | [472][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1316 ntime: 0077 mem: 3.36 + 04-04 04:31:52 | [472][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1531 ntime: 0075 mem: 3.36 + 04-04 04:32:02 | [472][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1347 ntime: 0077 mem: 3.36 + 04-04 04:32:09 | [472][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0977 ntime: 0085 mem: 3.36 + 04-04 04:32:16 | [472][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1200 ntime: 0083 mem: 3.36 + 04-04 04:32:23 | Time info >>>> elapsed: 429.02 mins remain: 477.99 mins + 04-04 04:32:23 | [473][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0083 mem: 3.36 + 04-04 04:32:33 | [473][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1388 ntime: 0076 mem: 3.36 + 04-04 04:32:41 | [473][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1042 ntime: 0076 mem: 3.36 + 04-04 04:32:49 | [473][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0755 ntime: 0078 mem: 3.36 + 04-04 04:32:57 | [473][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0876 ntime: 0083 mem: 3.36 + 04-04 04:33:05 | [473][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1284 ntime: 0081 mem: 3.36 + 04-04 04:33:11 | [473][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0467 ntime: 0089 mem: 3.36 + 04-04 04:33:21 | [473][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0083 mem: 3.36 + 04-04 04:33:29 | [473][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0882 ntime: 0062 mem: 3.36 + 04-04 04:33:35 | [473][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0908 ntime: 0082 mem: 3.36 + 04-04 04:33:43 | [473][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0870 ntime: 0092 mem: 3.36 + 04-04 04:33:52 | [473][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0055 ntime: 0081 mem: 3.36 + 04-04 04:34:01 | [473][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0726 ntime: 0084 mem: 3.36 + 04-04 04:34:11 | [473][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1300 ntime: 0079 mem: 3.36 + 04-04 04:34:18 | [473][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0477 ntime: 0080 mem: 3.36 + 04-04 04:34:25 | [473][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0121 ntime: 0088 mem: 3.36 + 04-04 04:34:33 | [473][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0124 ntime: 0083 mem: 3.36 + 04-04 04:34:39 | [473][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0617 ntime: 0085 mem: 3.36 + 04-04 04:34:46 | Time info >>>> elapsed: 431.39 mins remain: 478.72 mins + 04-04 04:34:47 | [474][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0985 ntime: 0074 mem: 3.36 + 04-04 04:34:57 | [474][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1457 ntime: 0084 mem: 3.36 + 04-04 04:35:06 | [474][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0882 ntime: 0082 mem: 3.36 + 04-04 04:35:14 | [474][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0049 ntime: 0074 mem: 3.36 + 04-04 04:35:21 | [474][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0706 ntime: 0084 mem: 3.36 + 04-04 04:35:32 | [474][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1108 ntime: 0082 mem: 3.36 + 04-04 04:35:41 | [474][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0142 ntime: 0082 mem: 3.36 + 04-04 04:35:51 | [474][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1320 ntime: 0081 mem: 3.36 + 04-04 04:35:57 | [474][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0083 ntime: 0085 mem: 3.36 + 04-04 04:36:06 | [474][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0620 ntime: 0079 mem: 3.36 + 04-04 04:36:16 | [474][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0910 ntime: 0079 mem: 3.36 + 04-04 04:36:23 | [474][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0042 ntime: 0059 mem: 3.36 + 04-04 04:36:31 | [474][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0074 mem: 3.36 + 04-04 04:36:38 | [474][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0092 ntime: 0074 mem: 3.36 + 04-04 04:36:45 | [474][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0762 ntime: 0083 mem: 3.36 + 04-04 04:36:53 | [474][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1304 ntime: 0083 mem: 3.36 + 04-04 04:37:00 | [474][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0109 ntime: 0073 mem: 3.36 + 04-04 04:37:08 | [474][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1224 ntime: 0076 mem: 3.36 + 04-04 04:37:17 | Time info >>>> elapsed: 433.91 mins remain: 479.59 mins + 04-04 04:37:18 | [475][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0872 ntime: 0086 mem: 3.36 + 04-04 04:37:27 | [475][010/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0868 ntime: 0080 mem: 3.36 + 04-04 04:37:34 | [475][020/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0056 ntime: 0077 mem: 3.36 + 04-04 04:37:44 | [475][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0156 ntime: 0080 mem: 3.36 + 04-04 04:37:52 | [475][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0116 ntime: 0081 mem: 3.36 + 04-04 04:37:59 | [475][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0911 ntime: 0084 mem: 3.36 + 04-04 04:38:07 | [475][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1213 ntime: 0082 mem: 3.36 + 04-04 04:38:15 | [475][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1174 ntime: 0078 mem: 3.36 + 04-04 04:38:22 | [475][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0085 mem: 3.36 + 04-04 04:38:29 | [475][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0586 ntime: 0082 mem: 3.36 + 04-04 04:38:36 | [475][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0505 ntime: 0077 mem: 3.36 + 04-04 04:38:45 | [475][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0905 ntime: 0078 mem: 3.36 + 04-04 04:38:55 | [475][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1181 ntime: 0080 mem: 3.36 + 04-04 04:39:03 | [475][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1187 ntime: 0076 mem: 3.36 + 04-04 04:39:10 | [475][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1026 ntime: 0075 mem: 3.36 + 04-04 04:39:19 | [475][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1025 ntime: 0075 mem: 3.36 + 04-04 04:39:28 | [475][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1213 ntime: 0079 mem: 3.36 + 04-04 04:39:37 | [475][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1293 ntime: 0075 mem: 3.36 + 04-04 04:39:43 | Time info >>>> elapsed: 436.35 mins remain: 480.35 mins + 04-04 04:39:45 | [476][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1024 ntime: 0079 mem: 3.36 + 04-04 04:39:52 | [476][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0894 ntime: 0079 mem: 3.36 + 04-04 04:40:00 | [476][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0985 ntime: 0086 mem: 3.36 + 04-04 04:40:07 | [476][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0249 ntime: 0079 mem: 3.36 + 04-04 04:40:17 | [476][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0090 ntime: 0079 mem: 3.36 + 04-04 04:40:25 | [476][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1349 ntime: 0081 mem: 3.36 + 04-04 04:40:32 | [476][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0438 ntime: 0074 mem: 3.36 + 04-04 04:40:41 | [476][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1303 ntime: 0076 mem: 3.36 + 04-04 04:40:53 | [476][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1351 ntime: 0083 mem: 3.36 + 04-04 04:41:00 | [476][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1155 ntime: 0086 mem: 3.36 + 04-04 04:41:09 | [476][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0116 ntime: 0077 mem: 3.36 + 04-04 04:41:18 | [476][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1345 ntime: 0088 mem: 3.36 + 04-04 04:41:26 | [476][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0123 ntime: 0078 mem: 3.36 + 04-04 04:41:37 | [476][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0623 ntime: 0078 mem: 3.36 + 04-04 04:41:46 | [476][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1000 ntime: 0084 mem: 3.36 + 04-04 04:41:57 | [476][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1059 ntime: 0091 mem: 3.36 + 04-04 04:42:08 | [476][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0058 mem: 3.36 + 04-04 04:42:17 | [476][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1036 ntime: 0083 mem: 3.36 + 04-04 04:42:24 | Time info >>>> elapsed: 439.03 mins remain: 481.36 mins + 04-04 04:42:24 | [477][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0068 ntime: 0070 mem: 3.36 + 04-04 04:42:31 | [477][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0692 ntime: 0082 mem: 3.36 + 04-04 04:42:39 | [477][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1047 ntime: 0078 mem: 3.36 + 04-04 04:42:46 | [477][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0072 ntime: 0077 mem: 3.36 + 04-04 04:42:55 | [477][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1047 ntime: 0079 mem: 3.36 + 04-04 04:43:04 | [477][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1249 ntime: 0088 mem: 3.36 + 04-04 04:43:12 | [477][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0826 ntime: 0092 mem: 3.36 + 04-04 04:43:21 | [477][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1342 ntime: 0083 mem: 3.36 + 04-04 04:43:30 | [477][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1158 ntime: 0079 mem: 3.36 + 04-04 04:43:40 | [477][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1646 ntime: 0085 mem: 3.36 + 04-04 04:43:48 | [477][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1021 ntime: 0083 mem: 3.36 + 04-04 04:43:56 | [477][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0785 ntime: 0081 mem: 3.36 + 04-04 04:44:04 | [477][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0956 ntime: 0071 mem: 3.36 + 04-04 04:44:12 | [477][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1293 ntime: 0085 mem: 3.36 + 04-04 04:44:21 | [477][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1518 ntime: 0075 mem: 3.36 + 04-04 04:44:31 | [477][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1132 ntime: 0083 mem: 3.36 + 04-04 04:44:39 | [477][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0864 ntime: 0076 mem: 3.36 + 04-04 04:44:45 | [477][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0661 ntime: 0071 mem: 3.36 + 04-04 04:44:52 | Time info >>>> elapsed: 441.49 mins remain: 482.13 mins + 04-04 04:44:53 | [478][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0809 ntime: 0085 mem: 3.36 + 04-04 04:45:01 | [478][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0745 ntime: 0087 mem: 3.36 + 04-04 04:45:08 | [478][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0629 ntime: 0076 mem: 3.36 + 04-04 04:45:17 | [478][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0985 ntime: 0079 mem: 3.36 + 04-04 04:45:26 | [478][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1348 ntime: 0078 mem: 3.36 + 04-04 04:45:35 | [478][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1028 ntime: 0086 mem: 3.36 + 04-04 04:45:42 | [478][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0079 mem: 3.36 + 04-04 04:45:51 | [478][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0742 ntime: 0081 mem: 3.36 + 04-04 04:46:01 | [478][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0084 ntime: 0087 mem: 3.36 + 04-04 04:46:10 | [478][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0080 mem: 3.36 + 04-04 04:46:19 | [478][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1012 ntime: 0083 mem: 3.36 + 04-04 04:46:31 | [478][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1364 ntime: 0085 mem: 3.36 + 04-04 04:46:39 | [478][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0212 ntime: 0074 mem: 3.36 + 04-04 04:46:48 | [478][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1447 ntime: 0088 mem: 3.36 + 04-04 04:46:55 | [478][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0094 ntime: 0085 mem: 3.36 + 04-04 04:47:02 | [478][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0664 ntime: 0081 mem: 3.36 + 04-04 04:47:10 | [478][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0083 mem: 3.36 + 04-04 04:47:19 | [478][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0144 ntime: 0083 mem: 3.36 + 04-04 04:47:25 | Time info >>>> elapsed: 444.05 mins remain: 482.98 mins + 04-04 04:47:26 | [479][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0844 ntime: 0086 mem: 3.36 + 04-04 04:47:36 | [479][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1221 ntime: 0082 mem: 3.36 + 04-04 04:47:44 | [479][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0847 ntime: 0086 mem: 3.36 + 04-04 04:47:53 | [479][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0786 ntime: 0088 mem: 3.36 + 04-04 04:48:02 | [479][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1085 ntime: 0086 mem: 3.36 + 04-04 04:48:10 | [479][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1088 ntime: 0079 mem: 3.36 + 04-04 04:48:17 | [479][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0074 mem: 3.36 + 04-04 04:48:25 | [479][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0742 ntime: 0071 mem: 3.36 + 04-04 04:48:31 | [479][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0066 ntime: 0079 mem: 3.36 + 04-04 04:48:40 | [479][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0825 ntime: 0072 mem: 3.36 + 04-04 04:48:48 | [479][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1490 ntime: 0077 mem: 3.36 + 04-04 04:48:56 | [479][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0099 ntime: 0077 mem: 3.36 + 04-04 04:49:05 | [479][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1231 ntime: 0080 mem: 3.36 + 04-04 04:49:12 | [479][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0739 ntime: 0079 mem: 3.36 + 04-04 04:49:21 | [479][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1322 ntime: 0084 mem: 3.36 + 04-04 04:49:29 | [479][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0066 ntime: 0077 mem: 3.36 + 04-04 04:49:38 | [479][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0697 ntime: 0086 mem: 3.36 + 04-04 04:49:49 | [479][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1641 ntime: 0080 mem: 3.36 + 04-04 04:49:54 | Time info >>>> elapsed: 446.53 mins remain: 483.74 mins + 04-04 04:49:54 | [480][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0257 ntime: 0082 mem: 3.36 + 04-04 04:50:01 | [480][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1289 ntime: 0079 mem: 3.36 + 04-04 04:50:09 | [480][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0679 ntime: 0079 mem: 3.36 + 04-04 04:50:15 | [480][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1114 ntime: 0081 mem: 3.36 + 04-04 04:50:22 | [480][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0082 mem: 3.36 + 04-04 04:50:31 | [480][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1385 ntime: 0084 mem: 3.36 + 04-04 04:50:43 | [480][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1437 ntime: 0079 mem: 3.36 + 04-04 04:50:51 | [480][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1119 ntime: 0079 mem: 3.36 + 04-04 04:51:00 | [480][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0784 ntime: 0073 mem: 3.36 + 04-04 04:51:08 | [480][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0557 ntime: 0075 mem: 3.36 + 04-04 04:51:15 | [480][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0072 mem: 3.36 + 04-04 04:51:25 | [480][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1469 ntime: 0081 mem: 3.36 + 04-04 04:51:33 | [480][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0078 ntime: 0076 mem: 3.36 + 04-04 04:51:44 | [480][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0085 mem: 3.36 + 04-04 04:51:53 | [480][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1067 ntime: 0083 mem: 3.36 + 04-04 04:52:02 | [480][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1356 ntime: 0086 mem: 3.36 + 04-04 04:52:10 | [480][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1123 ntime: 0084 mem: 3.36 + 04-04 04:52:19 | [480][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0141 ntime: 0082 mem: 3.36 + 04-04 04:52:27 | Time info >>>> elapsed: 449.08 mins remain: 484.56 mins + 04-04 04:52:28 | [481][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0912 ntime: 0075 mem: 3.36 + 04-04 04:52:37 | [481][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1258 ntime: 0085 mem: 3.36 + 04-04 04:52:45 | [481][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0834 ntime: 0081 mem: 3.36 + 04-04 04:52:52 | [481][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0093 ntime: 0079 mem: 3.36 + 04-04 04:52:59 | [481][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1241 ntime: 0084 mem: 3.36 + 04-04 04:53:07 | [481][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1103 ntime: 0090 mem: 3.36 + 04-04 04:53:15 | [481][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0085 mem: 3.36 + 04-04 04:53:23 | [481][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0789 ntime: 0082 mem: 3.36 + 04-04 04:53:32 | [481][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0064 ntime: 0074 mem: 3.36 + 04-04 04:53:40 | [481][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0952 ntime: 0080 mem: 3.36 + 04-04 04:53:48 | [481][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1008 ntime: 0076 mem: 3.36 + 04-04 04:53:56 | [481][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0134 ntime: 0073 mem: 3.36 + 04-04 04:54:03 | [481][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0089 mem: 3.36 + 04-04 04:54:11 | [481][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0769 ntime: 0080 mem: 3.36 + 04-04 04:54:19 | [481][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1023 ntime: 0081 mem: 3.36 + 04-04 04:54:25 | [481][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 04:54:32 | [481][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1423 ntime: 0078 mem: 3.36 + 04-04 04:54:39 | [481][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1355 ntime: 0082 mem: 3.36 + 04-04 04:54:43 | Time info >>>> elapsed: 451.35 mins remain: 485.06 mins + 04-04 04:54:44 | [482][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0879 ntime: 0085 mem: 3.36 + 04-04 04:54:52 | [482][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0944 ntime: 0090 mem: 3.36 + 04-04 04:55:01 | [482][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1196 ntime: 0074 mem: 3.36 + 04-04 04:55:09 | [482][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0109 ntime: 0085 mem: 3.36 + 04-04 04:55:18 | [482][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1257 ntime: 0087 mem: 3.36 + 04-04 04:55:24 | [482][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1016 ntime: 0084 mem: 3.36 + 04-04 04:55:33 | [482][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1292 ntime: 0089 mem: 3.36 + 04-04 04:55:42 | [482][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0119 ntime: 0077 mem: 3.36 + 04-04 04:55:51 | [482][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1305 ntime: 0081 mem: 3.36 + 04-04 04:56:00 | [482][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1421 ntime: 0079 mem: 3.36 + 04-04 04:56:08 | [482][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1382 ntime: 0083 mem: 3.36 + 04-04 04:56:16 | [482][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0207 ntime: 0081 mem: 3.36 + 04-04 04:56:26 | [482][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0908 ntime: 0079 mem: 3.36 + 04-04 04:56:34 | [482][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1378 ntime: 0081 mem: 3.36 + 04-04 04:56:46 | [482][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1181 ntime: 0081 mem: 3.36 + 04-04 04:56:55 | [482][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1070 ntime: 0077 mem: 3.36 + 04-04 04:57:04 | [482][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1056 ntime: 0079 mem: 3.36 + 04-04 04:57:12 | [482][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1442 ntime: 0080 mem: 3.36 + 04-04 04:57:21 | Time info >>>> elapsed: 453.97 mins remain: 485.93 mins + 04-04 04:57:21 | [483][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0066 ntime: 0076 mem: 3.36 + 04-04 04:57:30 | [483][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0922 ntime: 0081 mem: 3.36 + 04-04 04:57:42 | [483][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1405 ntime: 0075 mem: 3.36 + 04-04 04:57:54 | [483][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0108 ntime: 0079 mem: 3.36 + 04-04 04:58:04 | [483][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1490 ntime: 0087 mem: 3.36 + 04-04 04:58:10 | [483][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1086 ntime: 0078 mem: 3.36 + 04-04 04:58:18 | [483][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0988 ntime: 0074 mem: 3.36 + 04-04 04:58:27 | [483][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0785 ntime: 0077 mem: 3.36 + 04-04 04:58:35 | [483][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0988 ntime: 0079 mem: 3.36 + 04-04 04:58:42 | [483][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1314 ntime: 0078 mem: 3.36 + 04-04 04:58:48 | [483][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1348 ntime: 0079 mem: 3.36 + 04-04 04:58:56 | [483][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0565 ntime: 0078 mem: 3.36 + 04-04 04:59:04 | [483][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0922 ntime: 0086 mem: 3.36 + 04-04 04:59:12 | [483][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1223 ntime: 0087 mem: 3.36 + 04-04 04:59:18 | [483][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0478 ntime: 0088 mem: 3.36 + 04-04 04:59:28 | [483][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0927 ntime: 0080 mem: 3.36 + 04-04 04:59:36 | [483][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0099 ntime: 0083 mem: 3.36 + 04-04 04:59:44 | [483][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1245 ntime: 0054 mem: 3.36 + 04-04 04:59:50 | Time info >>>> elapsed: 456.47 mins remain: 486.65 mins + 04-04 04:59:52 | [484][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1543 ntime: 0073 mem: 3.36 + 04-04 04:59:57 | [484][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0074 mem: 3.36 + 04-04 05:00:04 | [484][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1327 ntime: 0081 mem: 3.36 + 04-04 05:00:12 | [484][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1288 ntime: 0072 mem: 3.36 + 04-04 05:00:18 | [484][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0143 ntime: 0080 mem: 3.36 + 04-04 05:00:26 | [484][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1097 ntime: 0081 mem: 3.36 + 04-04 05:00:33 | [484][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0653 ntime: 0081 mem: 3.36 + 04-04 05:00:43 | [484][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1098 ntime: 0074 mem: 3.36 + 04-04 05:00:51 | [484][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1005 ntime: 0090 mem: 3.36 + 04-04 05:00:59 | [484][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0574 ntime: 0079 mem: 3.36 + 04-04 05:01:07 | [484][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1273 ntime: 0078 mem: 3.36 + 04-04 05:01:15 | [484][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1459 ntime: 0083 mem: 3.36 + 04-04 05:01:22 | [484][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0097 ntime: 0074 mem: 3.36 + 04-04 05:01:29 | [484][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0096 ntime: 0073 mem: 3.36 + 04-04 05:01:37 | [484][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0100 ntime: 0079 mem: 3.36 + 04-04 05:01:44 | [484][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0088 ntime: 0089 mem: 3.36 + 04-04 05:01:51 | [484][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0225 ntime: 0081 mem: 3.36 + 04-04 05:02:03 | [484][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1492 ntime: 0081 mem: 3.36 + 04-04 05:02:10 | Time info >>>> elapsed: 458.80 mins remain: 487.17 mins + 04-04 05:02:12 | [485][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1350 ntime: 0076 mem: 3.36 + 04-04 05:02:20 | [485][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0759 ntime: 0087 mem: 3.36 + 04-04 05:02:27 | [485][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0569 ntime: 0078 mem: 3.36 + 04-04 05:02:34 | [485][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1107 ntime: 0078 mem: 3.36 + 04-04 05:02:45 | [485][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0151 ntime: 0071 mem: 3.36 + 04-04 05:02:55 | [485][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1072 ntime: 0082 mem: 3.36 + 04-04 05:03:04 | [485][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1388 ntime: 0086 mem: 3.36 + 04-04 05:03:11 | [485][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0347 ntime: 0081 mem: 3.36 + 04-04 05:03:20 | [485][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0877 ntime: 0080 mem: 3.36 + 04-04 05:03:31 | [485][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1344 ntime: 0084 mem: 3.36 + 04-04 05:03:38 | [485][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0759 ntime: 0075 mem: 3.36 + 04-04 05:03:46 | [485][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0462 ntime: 0080 mem: 3.36 + 04-04 05:03:55 | [485][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0083 mem: 3.36 + 04-04 05:04:03 | [485][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1009 ntime: 0091 mem: 3.36 + 04-04 05:04:09 | [485][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0783 ntime: 0081 mem: 3.36 + 04-04 05:04:17 | [485][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1559 ntime: 0087 mem: 3.36 + 04-04 05:04:24 | [485][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0498 ntime: 0074 mem: 3.36 + 04-04 05:04:32 | [485][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0087 ntime: 0076 mem: 3.36 + 04-04 05:04:38 | Time info >>>> elapsed: 461.26 mins remain: 487.84 mins + 04-04 05:04:39 | [486][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0286 ntime: 0082 mem: 3.36 + 04-04 05:04:48 | [486][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1000 ntime: 0086 mem: 3.36 + 04-04 05:04:58 | [486][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1402 ntime: 0081 mem: 3.36 + 04-04 05:05:06 | [486][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1559 ntime: 0076 mem: 3.36 + 04-04 05:05:16 | [486][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0967 ntime: 0081 mem: 3.36 + 04-04 05:05:24 | [486][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0930 ntime: 0084 mem: 3.36 + 04-04 05:05:35 | [486][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1373 ntime: 0078 mem: 3.36 + 04-04 05:05:44 | [486][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1414 ntime: 0078 mem: 3.36 + 04-04 05:05:54 | [486][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1156 ntime: 0082 mem: 3.36 + 04-04 05:06:08 | [486][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1416 ntime: 0079 mem: 3.36 + 04-04 05:06:19 | [486][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0862 ntime: 0080 mem: 3.36 + 04-04 05:06:30 | [486][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0650 ntime: 0077 mem: 3.36 + 04-04 05:06:38 | [486][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0524 ntime: 0089 mem: 3.36 + 04-04 05:06:46 | [486][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0285 ntime: 0079 mem: 3.36 + 04-04 05:06:56 | [486][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0965 ntime: 0085 mem: 3.36 + 04-04 05:07:05 | [486][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0474 ntime: 0085 mem: 3.36 + 04-04 05:07:14 | [486][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0162 ntime: 0075 mem: 3.36 + 04-04 05:07:23 | [486][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0351 ntime: 0079 mem: 3.36 + 04-04 05:07:28 | Time info >>>> elapsed: 464.09 mins remain: 488.87 mins + 04-04 05:07:28 | [487][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0094 ntime: 0083 mem: 3.36 + 04-04 05:07:38 | [487][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0941 ntime: 0081 mem: 3.36 + 04-04 05:07:46 | [487][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0113 ntime: 0079 mem: 3.36 + 04-04 05:07:54 | [487][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0172 ntime: 0081 mem: 3.36 + 04-04 05:08:03 | [487][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1549 ntime: 0073 mem: 3.36 + 04-04 05:08:12 | [487][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0119 ntime: 0080 mem: 3.36 + 04-04 05:08:20 | [487][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0938 ntime: 0056 mem: 3.36 + 04-04 05:08:31 | [487][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1381 ntime: 0081 mem: 3.36 + 04-04 05:08:40 | [487][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1015 ntime: 0078 mem: 3.36 + 04-04 05:08:48 | [487][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1090 ntime: 0086 mem: 3.36 + 04-04 05:08:55 | [487][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0989 ntime: 0082 mem: 3.36 + 04-04 05:09:03 | [487][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1544 ntime: 0074 mem: 3.36 + 04-04 05:09:10 | [487][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0078 mem: 3.36 + 04-04 05:09:20 | [487][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1173 ntime: 0079 mem: 3.36 + 04-04 05:09:27 | [487][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0563 ntime: 0081 mem: 3.36 + 04-04 05:09:36 | [487][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0704 ntime: 0084 mem: 3.36 + 04-04 05:09:44 | [487][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0708 ntime: 0084 mem: 3.36 + 04-04 05:09:51 | [487][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0934 ntime: 0079 mem: 3.36 + 04-04 05:09:58 | Time info >>>> elapsed: 466.59 mins remain: 489.53 mins + 04-04 05:09:59 | [488][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1039 ntime: 0084 mem: 3.36 + 04-04 05:10:06 | [488][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0116 ntime: 0083 mem: 3.36 + 04-04 05:10:15 | [488][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0646 ntime: 0080 mem: 3.36 + 04-04 05:10:22 | [488][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0634 ntime: 0079 mem: 3.36 + 04-04 05:10:31 | [488][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0663 ntime: 0081 mem: 3.36 + 04-04 05:10:39 | [488][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0586 ntime: 0082 mem: 3.36 + 04-04 05:10:47 | [488][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0337 ntime: 0084 mem: 3.36 + 04-04 05:10:54 | [488][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0499 ntime: 0087 mem: 3.36 + 04-04 05:11:03 | [488][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1429 ntime: 0076 mem: 3.36 + 04-04 05:11:10 | [488][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0163 ntime: 0080 mem: 3.36 + 04-04 05:11:18 | [488][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0886 ntime: 0083 mem: 3.36 + 04-04 05:11:29 | [488][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0844 ntime: 0088 mem: 3.36 + 04-04 05:11:38 | [488][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0106 ntime: 0080 mem: 3.36 + 04-04 05:11:47 | [488][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1488 ntime: 0087 mem: 3.36 + 04-04 05:11:54 | [488][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0792 ntime: 0083 mem: 3.36 + 04-04 05:12:03 | [488][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1302 ntime: 0081 mem: 3.36 + 04-04 05:12:12 | [488][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1126 ntime: 0086 mem: 3.36 + 04-04 05:12:21 | [488][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0755 ntime: 0079 mem: 3.36 + 04-04 05:12:26 | Time info >>>> elapsed: 469.07 mins remain: 490.17 mins + 04-04 05:12:28 | [489][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1247 ntime: 0084 mem: 3.36 + 04-04 05:12:36 | [489][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0940 ntime: 0085 mem: 3.36 + 04-04 05:12:43 | [489][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0067 ntime: 0078 mem: 3.36 + 04-04 05:12:54 | [489][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1052 ntime: 0077 mem: 3.36 + 04-04 05:13:03 | [489][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0950 ntime: 0078 mem: 3.36 + 04-04 05:13:10 | [489][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0140 ntime: 0080 mem: 3.36 + 04-04 05:13:18 | [489][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0781 ntime: 0086 mem: 3.36 + 04-04 05:13:26 | [489][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1668 ntime: 0087 mem: 3.36 + 04-04 05:13:36 | [489][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0877 ntime: 0076 mem: 3.36 + 04-04 05:13:43 | [489][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0046 ntime: 0076 mem: 3.36 + 04-04 05:13:52 | [489][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0081 ntime: 0077 mem: 3.36 + 04-04 05:14:02 | [489][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1367 ntime: 0079 mem: 3.36 + 04-04 05:14:09 | [489][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0705 ntime: 0085 mem: 3.36 + 04-04 05:14:16 | [489][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0321 ntime: 0074 mem: 3.36 + 04-04 05:14:24 | [489][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1341 ntime: 0079 mem: 3.36 + 04-04 05:14:31 | [489][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0711 ntime: 0082 mem: 3.36 + 04-04 05:14:41 | [489][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1155 ntime: 0083 mem: 3.36 + 04-04 05:14:47 | [489][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0926 ntime: 0084 mem: 3.36 + 04-04 05:14:53 | Time info >>>> elapsed: 471.51 mins remain: 490.75 mins + 04-04 05:14:54 | [490][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1160 ntime: 0090 mem: 3.36 + 04-04 05:15:02 | [490][010/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0329 ntime: 0086 mem: 3.36 + 04-04 05:15:09 | [490][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0087 ntime: 0064 mem: 3.36 + 04-04 05:15:14 | [490][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0114 ntime: 0081 mem: 3.36 + 04-04 05:15:21 | [490][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0163 ntime: 0083 mem: 3.36 + 04-04 05:15:29 | [490][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1402 ntime: 0084 mem: 3.36 + 04-04 05:15:36 | [490][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0095 ntime: 0075 mem: 3.36 + 04-04 05:15:44 | [490][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0585 ntime: 0085 mem: 3.36 + 04-04 05:15:52 | [490][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0112 ntime: 0087 mem: 3.36 + 04-04 05:16:00 | [490][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1018 ntime: 0088 mem: 3.36 + 04-04 05:16:07 | [490][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1088 ntime: 0082 mem: 3.36 + 04-04 05:16:14 | [490][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0437 ntime: 0073 mem: 3.36 + 04-04 05:16:25 | [490][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1379 ntime: 0088 mem: 3.36 + 04-04 05:16:32 | [490][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0080 ntime: 0081 mem: 3.36 + 04-04 05:16:38 | [490][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0081 mem: 3.36 + 04-04 05:16:45 | [490][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0114 ntime: 0072 mem: 3.36 + 04-04 05:16:53 | [490][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0358 ntime: 0073 mem: 3.36 + 04-04 05:17:00 | [490][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0101 ntime: 0055 mem: 3.36 + 04-04 05:17:06 | Time info >>>> elapsed: 473.73 mins remain: 491.09 mins + 04-04 05:17:07 | [491][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1274 ntime: 0079 mem: 3.36 + 04-04 05:17:16 | [491][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1507 ntime: 0082 mem: 3.36 + 04-04 05:17:24 | [491][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0502 ntime: 0085 mem: 3.36 + 04-04 05:17:32 | [491][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0931 ntime: 0077 mem: 3.36 + 04-04 05:17:40 | [491][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0562 ntime: 0079 mem: 3.36 + 04-04 05:17:47 | [491][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0047 ntime: 0079 mem: 3.36 + 04-04 05:17:58 | [491][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1439 ntime: 0087 mem: 3.36 + 04-04 05:18:08 | [491][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1124 ntime: 0082 mem: 3.36 + 04-04 05:18:18 | [491][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1443 ntime: 0077 mem: 3.36 + 04-04 05:18:25 | [491][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0109 ntime: 0086 mem: 3.36 + 04-04 05:18:33 | [491][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0861 ntime: 0083 mem: 3.36 + 04-04 05:18:42 | [491][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1002 ntime: 0079 mem: 3.36 + 04-04 05:18:51 | [491][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1047 ntime: 0079 mem: 3.36 + 04-04 05:19:00 | [491][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0105 ntime: 0085 mem: 3.36 + 04-04 05:19:09 | [491][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0224 ntime: 0083 mem: 3.36 + 04-04 05:19:17 | [491][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0082 mem: 3.36 + 04-04 05:19:24 | [491][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0047 ntime: 0080 mem: 3.36 + 04-04 05:19:32 | [491][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0094 ntime: 0077 mem: 3.36 + 04-04 05:19:38 | Time info >>>> elapsed: 476.25 mins remain: 491.74 mins + 04-04 05:19:38 | [492][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 05:19:49 | [492][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1030 ntime: 0081 mem: 3.36 + 04-04 05:19:56 | [492][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1150 ntime: 0082 mem: 3.36 + 04-04 05:20:03 | [492][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1095 ntime: 0077 mem: 3.36 + 04-04 05:20:09 | [492][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0290 ntime: 0080 mem: 3.36 + 04-04 05:20:14 | [492][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1058 ntime: 0079 mem: 3.36 + 04-04 05:20:22 | [492][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1506 ntime: 0085 mem: 3.36 + 04-04 05:20:28 | [492][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0942 ntime: 0078 mem: 3.36 + 04-04 05:20:37 | [492][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0598 ntime: 0083 mem: 3.36 + 04-04 05:20:43 | [492][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1237 ntime: 0080 mem: 3.36 + 04-04 05:20:50 | [492][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0776 ntime: 0080 mem: 3.36 + 04-04 05:20:57 | [492][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0500 ntime: 0083 mem: 3.36 + 04-04 05:21:03 | [492][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1280 ntime: 0082 mem: 3.36 + 04-04 05:21:10 | [492][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1150 ntime: 0079 mem: 3.36 + 04-04 05:21:17 | [492][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 05:21:26 | [492][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1338 ntime: 0084 mem: 3.36 + 04-04 05:21:34 | [492][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1491 ntime: 0083 mem: 3.36 + 04-04 05:21:41 | [492][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0911 ntime: 0085 mem: 3.36 + 04-04 05:21:48 | Time info >>>> elapsed: 478.43 mins remain: 492.02 mins + 04-04 05:21:50 | [493][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1333 ntime: 0079 mem: 3.36 + 04-04 05:21:57 | [493][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0291 ntime: 0077 mem: 3.36 + 04-04 05:22:03 | [493][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1095 ntime: 0084 mem: 3.36 + 04-04 05:22:12 | [493][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0700 ntime: 0078 mem: 3.36 + 04-04 05:22:20 | [493][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0770 ntime: 0086 mem: 3.36 + 04-04 05:22:27 | [493][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0997 ntime: 0083 mem: 3.36 + 04-04 05:22:35 | [493][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1102 ntime: 0079 mem: 3.36 + 04-04 05:22:45 | [493][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1352 ntime: 0076 mem: 3.36 + 04-04 05:22:53 | [493][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0246 ntime: 0075 mem: 3.36 + 04-04 05:23:00 | [493][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1057 ntime: 0072 mem: 3.36 + 04-04 05:23:10 | [493][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0642 ntime: 0080 mem: 3.36 + 04-04 05:23:18 | [493][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0683 ntime: 0086 mem: 3.36 + 04-04 05:23:26 | [493][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0882 ntime: 0080 mem: 3.36 + 04-04 05:23:34 | [493][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1144 ntime: 0084 mem: 3.36 + 04-04 05:23:41 | [493][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0765 ntime: 0078 mem: 3.36 + 04-04 05:23:50 | [493][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0916 ntime: 0081 mem: 3.36 + 04-04 05:24:01 | [493][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0785 ntime: 0077 mem: 3.36 + 04-04 05:24:08 | [493][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1189 ntime: 0080 mem: 3.36 + 04-04 05:24:15 | Time info >>>> elapsed: 480.87 mins remain: 492.55 mins + 04-04 05:24:16 | [494][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1645 ntime: 0077 mem: 3.36 + 04-04 05:24:24 | [494][010/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0795 ntime: 0087 mem: 3.36 + 04-04 05:24:32 | [494][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1169 ntime: 0078 mem: 3.36 + 04-04 05:24:41 | [494][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0077 ntime: 0081 mem: 3.36 + 04-04 05:24:51 | [494][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0087 mem: 3.36 + 04-04 05:25:01 | [494][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0784 ntime: 0085 mem: 3.36 + 04-04 05:25:10 | [494][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0112 ntime: 0079 mem: 3.36 + 04-04 05:25:18 | [494][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1483 ntime: 0079 mem: 3.36 + 04-04 05:25:26 | [494][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0065 ntime: 0092 mem: 3.36 + 04-04 05:25:37 | [494][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0338 ntime: 0076 mem: 3.36 + 04-04 05:25:44 | [494][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0040 ntime: 0059 mem: 3.36 + 04-04 05:25:54 | [494][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0961 ntime: 0076 mem: 3.36 + 04-04 05:26:01 | [494][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0083 mem: 3.36 + 04-04 05:26:09 | [494][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0757 ntime: 0071 mem: 3.36 + 04-04 05:26:18 | [494][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1248 ntime: 0078 mem: 3.36 + 04-04 05:26:27 | [494][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0867 ntime: 0085 mem: 3.36 + 04-04 05:26:34 | [494][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0046 ntime: 0075 mem: 3.36 + 04-04 05:26:42 | [494][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0064 ntime: 0075 mem: 3.36 + 04-04 05:26:47 | Time info >>>> elapsed: 483.42 mins remain: 493.18 mins + 04-04 05:26:49 | [495][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1248 ntime: 0079 mem: 3.36 + 04-04 05:26:56 | [495][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0069 ntime: 0079 mem: 3.36 + 04-04 05:27:04 | [495][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0109 ntime: 0082 mem: 3.36 + 04-04 05:27:11 | [495][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0390 ntime: 0080 mem: 3.36 + 04-04 05:27:18 | [495][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1117 ntime: 0085 mem: 3.36 + 04-04 05:27:27 | [495][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1303 ntime: 0077 mem: 3.36 + 04-04 05:27:35 | [495][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0857 ntime: 0079 mem: 3.36 + 04-04 05:27:43 | [495][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1295 ntime: 0071 mem: 3.36 + 04-04 05:27:51 | [495][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1063 ntime: 0079 mem: 3.36 + 04-04 05:27:58 | [495][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0981 ntime: 0080 mem: 3.36 + 04-04 05:28:04 | [495][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0278 ntime: 0082 mem: 3.36 + 04-04 05:28:14 | [495][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0702 ntime: 0081 mem: 3.36 + 04-04 05:28:22 | [495][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1057 ntime: 0082 mem: 3.36 + 04-04 05:28:28 | [495][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0137 ntime: 0079 mem: 3.36 + 04-04 05:28:37 | [495][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0132 ntime: 0078 mem: 3.36 + 04-04 05:28:47 | [495][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1257 ntime: 0058 mem: 3.36 + 04-04 05:28:56 | [495][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0068 ntime: 0089 mem: 3.36 + 04-04 05:29:03 | [495][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0674 ntime: 0073 mem: 3.36 + 04-04 05:29:10 | Time info >>>> elapsed: 485.79 mins remain: 493.62 mins + 04-04 05:29:10 | [496][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0384 ntime: 0084 mem: 3.36 + 04-04 05:29:17 | [496][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0194 ntime: 0078 mem: 3.36 + 04-04 05:29:27 | [496][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0150 ntime: 0083 mem: 3.36 + 04-04 05:29:37 | [496][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0753 ntime: 0078 mem: 3.36 + 04-04 05:29:45 | [496][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1337 ntime: 0087 mem: 3.36 + 04-04 05:29:54 | [496][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1138 ntime: 0076 mem: 3.36 + 04-04 05:30:03 | [496][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0904 ntime: 0085 mem: 3.36 + 04-04 05:30:11 | [496][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0674 ntime: 0082 mem: 3.36 + 04-04 05:30:20 | [496][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1402 ntime: 0079 mem: 3.36 + 04-04 05:30:27 | [496][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0481 ntime: 0085 mem: 3.36 + 04-04 05:30:36 | [496][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0914 ntime: 0079 mem: 3.36 + 04-04 05:30:44 | [496][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0100 ntime: 0085 mem: 3.36 + 04-04 05:30:50 | [496][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0485 ntime: 0081 mem: 3.36 + 04-04 05:30:58 | [496][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0433 ntime: 0080 mem: 3.36 + 04-04 05:31:09 | [496][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0288 ntime: 0080 mem: 3.36 + 04-04 05:31:17 | [496][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1322 ntime: 0079 mem: 3.36 + 04-04 05:31:25 | [496][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1160 ntime: 0080 mem: 3.36 + 04-04 05:31:34 | [496][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1159 ntime: 0075 mem: 3.36 + 04-04 05:31:41 | Time info >>>> elapsed: 488.31 mins remain: 494.20 mins + 04-04 05:31:41 | [497][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0464 ntime: 0079 mem: 3.36 + 04-04 05:31:50 | [497][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1424 ntime: 0083 mem: 3.36 + 04-04 05:32:00 | [497][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0747 ntime: 0078 mem: 3.36 + 04-04 05:32:08 | [497][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1017 ntime: 0080 mem: 3.36 + 04-04 05:32:18 | [497][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0688 ntime: 0086 mem: 3.36 + 04-04 05:32:26 | [497][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0637 ntime: 0079 mem: 3.36 + 04-04 05:32:35 | [497][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1075 ntime: 0087 mem: 3.36 + 04-04 05:32:44 | [497][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1199 ntime: 0079 mem: 3.36 + 04-04 05:32:55 | [497][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1049 ntime: 0077 mem: 3.36 + 04-04 05:33:04 | [497][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0882 ntime: 0084 mem: 3.36 + 04-04 05:33:13 | [497][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1187 ntime: 0078 mem: 3.36 + 04-04 05:33:20 | [497][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0142 ntime: 0076 mem: 3.36 + 04-04 05:33:27 | [497][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0644 ntime: 0083 mem: 3.36 + 04-04 05:33:35 | [497][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1288 ntime: 0084 mem: 3.36 + 04-04 05:33:42 | [497][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0081 mem: 3.36 + 04-04 05:33:50 | [497][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1140 ntime: 0086 mem: 3.36 + 04-04 05:33:58 | [497][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0104 ntime: 0081 mem: 3.36 + 04-04 05:34:05 | [497][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1402 ntime: 0081 mem: 3.36 + 04-04 05:34:11 | Time info >>>> elapsed: 490.81 mins remain: 494.75 mins + 04-04 05:34:12 | [498][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1226 ntime: 0085 mem: 3.36 + 04-04 05:34:19 | [498][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0172 ntime: 0080 mem: 3.36 + 04-04 05:34:25 | [498][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0231 ntime: 0080 mem: 3.36 + 04-04 05:34:34 | [498][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1059 ntime: 0077 mem: 3.36 + 04-04 05:34:41 | [498][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1280 ntime: 0078 mem: 3.36 + 04-04 05:34:50 | [498][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1116 ntime: 0082 mem: 3.36 + 04-04 05:34:56 | [498][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0659 ntime: 0087 mem: 3.36 + 04-04 05:35:05 | [498][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1051 ntime: 0082 mem: 3.36 + 04-04 05:35:14 | [498][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1107 ntime: 0083 mem: 3.36 + 04-04 05:35:23 | [498][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0068 ntime: 0079 mem: 3.36 + 04-04 05:35:31 | [498][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1364 ntime: 0078 mem: 3.36 + 04-04 05:35:38 | [498][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0077 ntime: 0078 mem: 3.36 + 04-04 05:35:48 | [498][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0591 ntime: 0083 mem: 3.36 + 04-04 05:35:57 | [498][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1026 ntime: 0077 mem: 3.36 + 04-04 05:36:04 | [498][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0449 ntime: 0079 mem: 3.36 + 04-04 05:36:15 | [498][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1151 ntime: 0084 mem: 3.36 + 04-04 05:36:24 | [498][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1225 ntime: 0082 mem: 3.36 + 04-04 05:36:31 | [498][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0939 ntime: 0082 mem: 3.36 + 04-04 05:36:37 | Time info >>>> elapsed: 493.24 mins remain: 495.22 mins + 04-04 05:36:37 | [499][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0064 ntime: 0093 mem: 3.36 + 04-04 05:36:45 | [499][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0130 ntime: 0078 mem: 3.36 + 04-04 05:36:51 | [499][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0177 ntime: 0083 mem: 3.36 + 04-04 05:36:58 | [499][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1013 ntime: 0087 mem: 3.36 + 04-04 05:37:05 | [499][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0741 ntime: 0087 mem: 3.36 + 04-04 05:37:13 | [499][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0156 ntime: 0078 mem: 3.36 + 04-04 05:37:22 | [499][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1371 ntime: 0088 mem: 3.36 + 04-04 05:37:29 | [499][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0742 ntime: 0068 mem: 3.36 + 04-04 05:37:37 | [499][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0059 ntime: 0078 mem: 3.36 + 04-04 05:37:45 | [499][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1060 ntime: 0085 mem: 3.36 + 04-04 05:37:54 | [499][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 2108 ntime: 0089 mem: 3.36 + 04-04 05:38:01 | [499][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0080 mem: 3.36 + 04-04 05:38:10 | [499][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1338 ntime: 0078 mem: 3.36 + 04-04 05:38:17 | [499][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1015 ntime: 0083 mem: 3.36 + 04-04 05:38:27 | [499][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1809 ntime: 0079 mem: 3.36 + 04-04 05:38:37 | [499][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0121 ntime: 0084 mem: 3.36 + 04-04 05:38:45 | [499][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0635 ntime: 0082 mem: 3.36 + 04-04 05:38:55 | [499][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0105 ntime: 0080 mem: 3.36 + 04-04 05:39:01 | Time info >>>> elapsed: 495.64 mins remain: 495.64 mins + 04-04 05:39:02 | [500][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1270 ntime: 0084 mem: 3.36 + 04-04 05:39:11 | [500][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1318 ntime: 0077 mem: 3.36 + 04-04 05:39:18 | [500][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0149 ntime: 0076 mem: 3.36 + 04-04 05:39:28 | [500][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1465 ntime: 0086 mem: 3.36 + 04-04 05:39:37 | [500][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1315 ntime: 0085 mem: 3.36 + 04-04 05:39:43 | [500][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0088 ntime: 0081 mem: 3.36 + 04-04 05:39:50 | [500][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0629 ntime: 0081 mem: 3.36 + 04-04 05:39:58 | [500][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0639 ntime: 0092 mem: 3.36 + 04-04 05:40:08 | [500][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1279 ntime: 0081 mem: 3.36 + 04-04 05:40:18 | [500][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0077 ntime: 0081 mem: 3.36 + 04-04 05:40:27 | [500][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1357 ntime: 0083 mem: 3.36 + 04-04 05:40:36 | [500][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0083 mem: 3.36 + 04-04 05:40:44 | [500][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1342 ntime: 0080 mem: 3.36 + 04-04 05:40:54 | [500][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0842 ntime: 0089 mem: 3.36 + 04-04 05:40:59 | [500][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0055 ntime: 0074 mem: 3.36 + 04-04 05:41:08 | [500][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1035 ntime: 0081 mem: 3.36 + 04-04 05:41:17 | [500][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1207 ntime: 0084 mem: 3.36 + 04-04 05:41:27 | [500][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1434 ntime: 0084 mem: 3.36 + 04-04 05:41:32 | Time info >>>> elapsed: 498.16 mins remain: 496.17 mins + 04-04 05:41:32 | [501][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0094 ntime: 0079 mem: 3.36 + 04-04 05:41:41 | [501][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0403 ntime: 0085 mem: 3.36 + 04-04 05:41:51 | [501][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0159 ntime: 0075 mem: 3.36 + 04-04 05:42:02 | [501][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1462 ntime: 0087 mem: 3.36 + 04-04 05:42:13 | [501][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1384 ntime: 0074 mem: 3.36 + 04-04 05:42:22 | [501][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0918 ntime: 0074 mem: 3.36 + 04-04 05:42:32 | [501][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0977 ntime: 0070 mem: 3.36 + 04-04 05:42:41 | [501][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1237 ntime: 0082 mem: 3.36 + 04-04 05:42:50 | [501][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0807 ntime: 0083 mem: 3.36 + 04-04 05:42:56 | [501][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0526 ntime: 0081 mem: 3.36 + 04-04 05:43:04 | [501][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0208 ntime: 0076 mem: 3.36 + 04-04 05:43:13 | [501][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1675 ntime: 0058 mem: 3.36 + 04-04 05:43:20 | [501][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0073 ntime: 0077 mem: 3.36 + 04-04 05:43:27 | [501][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0637 ntime: 0089 mem: 3.36 + 04-04 05:43:36 | [501][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0760 ntime: 0086 mem: 3.36 + 04-04 05:43:46 | [501][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1428 ntime: 0079 mem: 3.36 + 04-04 05:43:54 | [501][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0094 ntime: 0080 mem: 3.36 + 04-04 05:44:01 | [501][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0079 ntime: 0078 mem: 3.36 + 04-04 05:44:08 | Time info >>>> elapsed: 500.75 mins remain: 496.76 mins + 04-04 05:44:08 | [502][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0687 ntime: 0078 mem: 3.36 + 04-04 05:44:15 | [502][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0498 ntime: 0080 mem: 3.36 + 04-04 05:44:24 | [502][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1243 ntime: 0084 mem: 3.36 + 04-04 05:44:31 | [502][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0992 ntime: 0096 mem: 3.36 + 04-04 05:44:40 | [502][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1039 ntime: 0079 mem: 3.36 + 04-04 05:44:49 | [502][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1361 ntime: 0092 mem: 3.36 + 04-04 05:44:57 | [502][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0619 ntime: 0076 mem: 3.36 + 04-04 05:45:05 | [502][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0994 ntime: 0077 mem: 3.36 + 04-04 05:45:12 | [502][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0774 ntime: 0092 mem: 3.36 + 04-04 05:45:21 | [502][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0531 ntime: 0084 mem: 3.36 + 04-04 05:45:29 | [502][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0950 ntime: 0083 mem: 3.36 + 04-04 05:45:37 | [502][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0083 ntime: 0076 mem: 3.36 + 04-04 05:45:46 | [502][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0875 ntime: 0085 mem: 3.36 + 04-04 05:45:55 | [502][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0069 ntime: 0078 mem: 3.36 + 04-04 05:46:05 | [502][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0235 ntime: 0086 mem: 3.36 + 04-04 05:46:13 | [502][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0823 ntime: 0085 mem: 3.36 + 04-04 05:46:20 | [502][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1105 ntime: 0079 mem: 3.36 + 04-04 05:46:27 | [502][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0744 ntime: 0077 mem: 3.36 + 04-04 05:46:33 | Time info >>>> elapsed: 503.18 mins remain: 497.17 mins + 04-04 05:46:34 | [503][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0499 ntime: 0080 mem: 3.36 + 04-04 05:46:41 | [503][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0446 ntime: 0080 mem: 3.36 + 04-04 05:46:48 | [503][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0666 ntime: 0081 mem: 3.36 + 04-04 05:46:58 | [503][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1309 ntime: 0084 mem: 3.36 + 04-04 05:47:07 | [503][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0083 mem: 3.36 + 04-04 05:47:14 | [503][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0662 ntime: 0083 mem: 3.36 + 04-04 05:47:22 | [503][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0075 ntime: 0084 mem: 3.36 + 04-04 05:47:30 | [503][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0928 ntime: 0079 mem: 3.36 + 04-04 05:47:40 | [503][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0081 mem: 3.36 + 04-04 05:47:47 | [503][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0453 ntime: 0084 mem: 3.36 + 04-04 05:47:56 | [503][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0423 ntime: 0088 mem: 3.36 + 04-04 05:48:05 | [503][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1062 ntime: 0082 mem: 3.36 + 04-04 05:48:13 | [503][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1777 ntime: 0081 mem: 3.36 + 04-04 05:48:20 | [503][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0984 ntime: 0078 mem: 3.36 + 04-04 05:48:29 | [503][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1104 ntime: 0083 mem: 3.36 + 04-04 05:48:36 | [503][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0069 ntime: 0079 mem: 3.36 + 04-04 05:48:44 | [503][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0805 ntime: 0090 mem: 3.36 + 04-04 05:48:52 | [503][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0085 mem: 3.36 + 04-04 05:48:59 | Time info >>>> elapsed: 505.61 mins remain: 497.58 mins + 04-04 05:49:00 | [504][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1001 ntime: 0083 mem: 3.36 + 04-04 05:49:06 | [504][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0527 ntime: 0087 mem: 3.36 + 04-04 05:49:14 | [504][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1289 ntime: 0085 mem: 3.36 + 04-04 05:49:21 | [504][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0105 ntime: 0079 mem: 3.36 + 04-04 05:49:30 | [504][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0065 ntime: 0077 mem: 3.36 + 04-04 05:49:39 | [504][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0590 ntime: 0076 mem: 3.36 + 04-04 05:49:48 | [504][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1340 ntime: 0072 mem: 3.36 + 04-04 05:49:56 | [504][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1636 ntime: 0080 mem: 3.36 + 04-04 05:50:03 | [504][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0333 ntime: 0081 mem: 3.36 + 04-04 05:50:12 | [504][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0066 ntime: 0084 mem: 3.36 + 04-04 05:50:23 | [504][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1155 ntime: 0079 mem: 3.36 + 04-04 05:50:32 | [504][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1444 ntime: 0080 mem: 3.36 + 04-04 05:50:38 | [504][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0090 ntime: 0081 mem: 3.36 + 04-04 05:50:45 | [504][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0958 ntime: 0076 mem: 3.36 + 04-04 05:50:53 | [504][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0084 mem: 3.36 + 04-04 05:51:01 | [504][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0257 ntime: 0088 mem: 3.36 + 04-04 05:51:08 | [504][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1039 ntime: 0079 mem: 3.36 + 04-04 05:51:16 | [504][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0086 ntime: 0080 mem: 3.36 + 04-04 05:51:22 | Time info >>>> elapsed: 507.99 mins remain: 497.93 mins + 04-04 05:51:22 | [505][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0149 ntime: 0084 mem: 3.36 + 04-04 05:51:32 | [505][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1209 ntime: 0078 mem: 3.36 + 04-04 05:51:39 | [505][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0083 ntime: 0072 mem: 3.36 + 04-04 05:51:49 | [505][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1155 ntime: 0079 mem: 3.36 + 04-04 05:51:57 | [505][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0563 ntime: 0082 mem: 3.36 + 04-04 05:52:04 | [505][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1270 ntime: 0078 mem: 3.36 + 04-04 05:52:12 | [505][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0506 ntime: 0079 mem: 3.36 + 04-04 05:52:19 | [505][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0099 ntime: 0080 mem: 3.36 + 04-04 05:52:26 | [505][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0116 ntime: 0077 mem: 3.36 + 04-04 05:52:35 | [505][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1224 ntime: 0080 mem: 3.36 + 04-04 05:52:44 | [505][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1411 ntime: 0079 mem: 3.36 + 04-04 05:52:51 | [505][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0136 ntime: 0083 mem: 3.36 + 04-04 05:53:00 | [505][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0074 ntime: 0088 mem: 3.36 + 04-04 05:53:08 | [505][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0055 ntime: 0077 mem: 3.36 + 04-04 05:53:18 | [505][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1356 ntime: 0075 mem: 3.36 + 04-04 05:53:26 | [505][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1047 ntime: 0082 mem: 3.36 + 04-04 05:53:37 | [505][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1206 ntime: 0094 mem: 3.36 + 04-04 05:53:45 | [505][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0062 ntime: 0082 mem: 3.36 + 04-04 05:53:51 | Time info >>>> elapsed: 510.48 mins remain: 498.38 mins + 04-04 05:53:52 | [506][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0158 ntime: 0076 mem: 3.36 + 04-04 05:53:59 | [506][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0080 ntime: 0077 mem: 3.36 + 04-04 05:54:08 | [506][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0064 ntime: 0083 mem: 3.36 + 04-04 05:54:16 | [506][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1104 ntime: 0082 mem: 3.36 + 04-04 05:54:24 | [506][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0192 ntime: 0080 mem: 3.36 + 04-04 05:54:33 | [506][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0871 ntime: 0083 mem: 3.36 + 04-04 05:54:40 | [506][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0079 ntime: 0080 mem: 3.36 + 04-04 05:54:48 | [506][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1193 ntime: 0079 mem: 3.36 + 04-04 05:54:58 | [506][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0998 ntime: 0081 mem: 3.36 + 04-04 05:55:06 | [506][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1414 ntime: 0078 mem: 3.36 + 04-04 05:55:16 | [506][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0590 ntime: 0086 mem: 3.36 + 04-04 05:55:26 | [506][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0988 ntime: 0073 mem: 3.36 + 04-04 05:55:36 | [506][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1386 ntime: 0082 mem: 3.36 + 04-04 05:55:44 | [506][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0122 ntime: 0078 mem: 3.36 + 04-04 05:55:54 | [506][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0274 ntime: 0079 mem: 3.36 + 04-04 05:56:05 | [506][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0681 ntime: 0077 mem: 3.36 + 04-04 05:56:14 | [506][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1115 ntime: 0084 mem: 3.36 + 04-04 05:56:21 | [506][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0790 ntime: 0086 mem: 3.36 + 04-04 05:56:29 | Time info >>>> elapsed: 513.12 mins remain: 498.95 mins + 04-04 05:56:30 | [507][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0212 ntime: 0080 mem: 3.36 + 04-04 05:56:39 | [507][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1428 ntime: 0078 mem: 3.36 + 04-04 05:56:47 | [507][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0988 ntime: 0076 mem: 3.36 + 04-04 05:56:54 | [507][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0092 ntime: 0070 mem: 3.36 + 04-04 05:57:02 | [507][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0781 ntime: 0083 mem: 3.36 + 04-04 05:57:09 | [507][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0609 ntime: 0085 mem: 3.36 + 04-04 05:57:19 | [507][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0334 ntime: 0084 mem: 3.36 + 04-04 05:57:27 | [507][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1483 ntime: 0083 mem: 3.36 + 04-04 05:57:40 | [507][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1579 ntime: 0088 mem: 3.36 + 04-04 05:57:48 | [507][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0045 ntime: 0058 mem: 3.36 + 04-04 05:57:59 | [507][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1456 ntime: 0080 mem: 3.36 + 04-04 05:58:08 | [507][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0139 ntime: 0078 mem: 3.36 + 04-04 05:58:15 | [507][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0115 ntime: 0083 mem: 3.36 + 04-04 05:58:23 | [507][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1369 ntime: 0084 mem: 3.36 + 04-04 05:58:30 | [507][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0383 ntime: 0084 mem: 3.36 + 04-04 05:58:38 | [507][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0568 ntime: 0079 mem: 3.36 + 04-04 05:58:48 | [507][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1085 ntime: 0080 mem: 3.36 + 04-04 05:58:55 | [507][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1275 ntime: 0081 mem: 3.36 + 04-04 05:59:03 | Time info >>>> elapsed: 515.69 mins remain: 499.44 mins + 04-04 05:59:04 | [508][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0104 ntime: 0084 mem: 3.36 + 04-04 05:59:12 | [508][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0988 ntime: 0074 mem: 3.36 + 04-04 05:59:22 | [508][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0963 ntime: 0083 mem: 3.36 + 04-04 05:59:32 | [508][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0077 mem: 3.36 + 04-04 05:59:43 | [508][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1602 ntime: 0078 mem: 3.36 + 04-04 05:59:51 | [508][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1026 ntime: 0086 mem: 3.36 + 04-04 05:59:59 | [508][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1057 ntime: 0080 mem: 3.36 + 04-04 06:00:08 | [508][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0604 ntime: 0078 mem: 3.36 + 04-04 06:00:16 | [508][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1049 ntime: 0076 mem: 3.36 + 04-04 06:00:24 | [508][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0078 mem: 3.36 + 04-04 06:00:32 | [508][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0698 ntime: 0078 mem: 3.36 + 04-04 06:00:39 | [508][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0179 ntime: 0081 mem: 3.36 + 04-04 06:00:48 | [508][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1088 ntime: 0083 mem: 3.36 + 04-04 06:00:54 | [508][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0049 ntime: 0076 mem: 3.36 + 04-04 06:01:03 | [508][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0075 ntime: 0079 mem: 3.36 + 04-04 06:01:12 | [508][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1243 ntime: 0061 mem: 3.36 + 04-04 06:01:21 | [508][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0068 ntime: 0072 mem: 3.36 + 04-04 06:01:28 | [508][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0072 ntime: 0078 mem: 3.36 + 04-04 06:01:35 | Time info >>>> elapsed: 518.21 mins remain: 499.89 mins + 04-04 06:01:35 | [509][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0107 ntime: 0087 mem: 3.36 + 04-04 06:01:43 | [509][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 06:01:53 | [509][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0862 ntime: 0080 mem: 3.36 + 04-04 06:02:02 | [509][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1375 ntime: 0076 mem: 3.36 + 04-04 06:02:10 | [509][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1178 ntime: 0084 mem: 3.36 + 04-04 06:02:19 | [509][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0088 ntime: 0082 mem: 3.36 + 04-04 06:02:27 | [509][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1143 ntime: 0073 mem: 3.36 + 04-04 06:02:34 | [509][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 06:02:42 | [509][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1032 ntime: 0081 mem: 3.36 + 04-04 06:02:47 | [509][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1150 ntime: 0081 mem: 3.36 + 04-04 06:02:56 | [509][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1591 ntime: 0089 mem: 3.36 + 04-04 06:03:04 | [509][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0067 ntime: 0082 mem: 3.36 + 04-04 06:03:13 | [509][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0507 ntime: 0081 mem: 3.36 + 04-04 06:03:21 | [509][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1247 ntime: 0092 mem: 3.36 + 04-04 06:03:27 | [509][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0612 ntime: 0095 mem: 3.36 + 04-04 06:03:38 | [509][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0694 ntime: 0083 mem: 3.36 + 04-04 06:03:47 | [509][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0833 ntime: 0089 mem: 3.36 + 04-04 06:03:54 | [509][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0072 ntime: 0089 mem: 3.36 + 04-04 06:04:00 | Time info >>>> elapsed: 520.62 mins remain: 500.20 mins + 04-04 06:04:01 | [510][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1128 ntime: 0077 mem: 3.36 + 04-04 06:04:09 | [510][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0718 ntime: 0086 mem: 3.36 + 04-04 06:04:18 | [510][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0174 ntime: 0077 mem: 3.36 + 04-04 06:04:28 | [510][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0055 ntime: 0088 mem: 3.36 + 04-04 06:04:36 | [510][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0960 ntime: 0077 mem: 3.36 + 04-04 06:04:45 | [510][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0079 mem: 3.36 + 04-04 06:04:52 | [510][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0252 ntime: 0088 mem: 3.36 + 04-04 06:04:59 | [510][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0081 ntime: 0080 mem: 3.36 + 04-04 06:05:08 | [510][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0661 ntime: 0084 mem: 3.36 + 04-04 06:05:16 | [510][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1454 ntime: 0080 mem: 3.36 + 04-04 06:05:26 | [510][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0083 mem: 3.36 + 04-04 06:05:36 | [510][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0880 ntime: 0079 mem: 3.36 + 04-04 06:05:44 | [510][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1275 ntime: 0084 mem: 3.36 + 04-04 06:05:52 | [510][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0081 mem: 3.36 + 04-04 06:06:02 | [510][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0446 ntime: 0077 mem: 3.36 + 04-04 06:06:10 | [510][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0875 ntime: 0091 mem: 3.36 + 04-04 06:06:18 | [510][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1139 ntime: 0086 mem: 3.36 + 04-04 06:06:27 | [510][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0090 ntime: 0076 mem: 3.36 + 04-04 06:06:33 | Time info >>>> elapsed: 523.18 mins remain: 500.65 mins + 04-04 06:06:35 | [511][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1482 ntime: 0074 mem: 3.36 + 04-04 06:06:40 | [511][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0079 mem: 3.36 + 04-04 06:06:47 | [511][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0102 ntime: 0073 mem: 3.36 + 04-04 06:06:55 | [511][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1194 ntime: 0084 mem: 3.36 + 04-04 06:07:02 | [511][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0670 ntime: 0078 mem: 3.36 + 04-04 06:07:13 | [511][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1055 ntime: 0085 mem: 3.36 + 04-04 06:07:23 | [511][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0922 ntime: 0082 mem: 3.36 + 04-04 06:07:32 | [511][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1097 ntime: 0084 mem: 3.36 + 04-04 06:07:38 | [511][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0520 ntime: 0083 mem: 3.36 + 04-04 06:07:46 | [511][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0086 ntime: 0080 mem: 3.36 + 04-04 06:07:54 | [511][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1158 ntime: 0087 mem: 3.36 + 04-04 06:08:02 | [511][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0072 ntime: 0074 mem: 3.36 + 04-04 06:08:09 | [511][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0307 ntime: 0082 mem: 3.36 + 04-04 06:08:17 | [511][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1405 ntime: 0074 mem: 3.36 + 04-04 06:08:25 | [511][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1051 ntime: 0079 mem: 3.36 + 04-04 06:08:32 | [511][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1360 ntime: 0078 mem: 3.36 + 04-04 06:08:39 | [511][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1007 ntime: 0085 mem: 3.36 + 04-04 06:08:46 | [511][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0333 ntime: 0080 mem: 3.36 + 04-04 06:08:52 | Time info >>>> elapsed: 525.49 mins remain: 500.86 mins + 04-04 06:08:53 | [512][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1453 ntime: 0080 mem: 3.36 + 04-04 06:09:02 | [512][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1123 ntime: 0084 mem: 3.36 + 04-04 06:09:11 | [512][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1412 ntime: 0075 mem: 3.36 + 04-04 06:09:20 | [512][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0713 ntime: 0059 mem: 3.36 + 04-04 06:09:28 | [512][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0704 ntime: 0078 mem: 3.36 + 04-04 06:09:36 | [512][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1028 ntime: 0077 mem: 3.36 + 04-04 06:09:44 | [512][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0167 ntime: 0079 mem: 3.36 + 04-04 06:09:53 | [512][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0110 ntime: 0078 mem: 3.36 + 04-04 06:10:02 | [512][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1410 ntime: 0079 mem: 3.36 + 04-04 06:10:10 | [512][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0977 ntime: 0083 mem: 3.36 + 04-04 06:10:17 | [512][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0953 ntime: 0077 mem: 3.36 + 04-04 06:10:25 | [512][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0729 ntime: 0085 mem: 3.36 + 04-04 06:10:32 | [512][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0091 mem: 3.36 + 04-04 06:10:40 | [512][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0845 ntime: 0082 mem: 3.36 + 04-04 06:10:51 | [512][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1591 ntime: 0085 mem: 3.36 + 04-04 06:10:57 | [512][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0751 ntime: 0088 mem: 3.36 + 04-04 06:11:06 | [512][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0079 ntime: 0091 mem: 3.36 + 04-04 06:11:13 | [512][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0843 ntime: 0074 mem: 3.36 + 04-04 06:11:19 | Time info >>>> elapsed: 527.94 mins remain: 501.18 mins + 04-04 06:11:20 | [513][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1269 ntime: 0080 mem: 3.36 + 04-04 06:11:29 | [513][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0711 ntime: 0077 mem: 3.36 + 04-04 06:11:37 | [513][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0933 ntime: 0085 mem: 3.36 + 04-04 06:11:45 | [513][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0905 ntime: 0076 mem: 3.36 + 04-04 06:11:53 | [513][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0101 ntime: 0080 mem: 3.36 + 04-04 06:12:01 | [513][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0905 ntime: 0081 mem: 3.36 + 04-04 06:12:08 | [513][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0912 ntime: 0081 mem: 3.36 + 04-04 06:12:16 | [513][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0118 ntime: 0081 mem: 3.36 + 04-04 06:12:23 | [513][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0529 ntime: 0080 mem: 3.36 + 04-04 06:12:29 | [513][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0686 ntime: 0078 mem: 3.36 + 04-04 06:12:39 | [513][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0726 ntime: 0080 mem: 3.36 + 04-04 06:12:48 | [513][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0612 ntime: 0075 mem: 3.36 + 04-04 06:12:57 | [513][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0644 ntime: 0071 mem: 3.36 + 04-04 06:13:06 | [513][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0474 ntime: 0077 mem: 3.36 + 04-04 06:13:13 | [513][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0688 ntime: 0081 mem: 3.36 + 04-04 06:13:19 | [513][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0975 ntime: 0077 mem: 3.36 + 04-04 06:13:27 | [513][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1170 ntime: 0074 mem: 3.36 + 04-04 06:13:34 | [513][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1119 ntime: 0075 mem: 3.36 + 04-04 06:13:40 | Time info >>>> elapsed: 530.30 mins remain: 501.41 mins + 04-04 06:13:42 | [514][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1294 ntime: 0079 mem: 3.36 + 04-04 06:13:48 | [514][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0171 ntime: 0081 mem: 3.36 + 04-04 06:13:56 | [514][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0084 mem: 3.36 + 04-04 06:14:06 | [514][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1314 ntime: 0084 mem: 3.36 + 04-04 06:14:15 | [514][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0106 ntime: 0083 mem: 3.36 + 04-04 06:14:20 | [514][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0170 ntime: 0075 mem: 3.36 + 04-04 06:14:30 | [514][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0514 ntime: 0080 mem: 3.36 + 04-04 06:14:41 | [514][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1481 ntime: 0079 mem: 3.36 + 04-04 06:14:50 | [514][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1456 ntime: 0088 mem: 3.36 + 04-04 06:14:58 | [514][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0545 ntime: 0085 mem: 3.36 + 04-04 06:15:05 | [514][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0104 ntime: 0077 mem: 3.36 + 04-04 06:15:13 | [514][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1104 ntime: 0082 mem: 3.36 + 04-04 06:15:20 | [514][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1145 ntime: 0058 mem: 3.36 + 04-04 06:15:26 | [514][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0078 mem: 3.36 + 04-04 06:15:35 | [514][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1250 ntime: 0078 mem: 3.36 + 04-04 06:15:44 | [514][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1145 ntime: 0080 mem: 3.36 + 04-04 06:15:53 | [514][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0156 ntime: 0080 mem: 3.36 + 04-04 06:16:02 | [514][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1339 ntime: 0078 mem: 3.36 + 04-04 06:16:11 | Time info >>>> elapsed: 532.81 mins remain: 501.77 mins + 04-04 06:16:11 | [515][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0073 ntime: 0076 mem: 3.36 + 04-04 06:16:19 | [515][010/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0869 ntime: 0078 mem: 3.36 + 04-04 06:16:27 | [515][020/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1119 ntime: 0078 mem: 3.36 + 04-04 06:16:36 | [515][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0724 ntime: 0083 mem: 3.36 + 04-04 06:16:43 | [515][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1439 ntime: 0084 mem: 3.36 + 04-04 06:16:51 | [515][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1426 ntime: 0076 mem: 3.36 + 04-04 06:16:59 | [515][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0171 ntime: 0083 mem: 3.36 + 04-04 06:17:08 | [515][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1307 ntime: 0083 mem: 3.36 + 04-04 06:17:16 | [515][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1250 ntime: 0081 mem: 3.36 + 04-04 06:17:25 | [515][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1123 ntime: 0081 mem: 3.36 + 04-04 06:17:32 | [515][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0978 ntime: 0084 mem: 3.36 + 04-04 06:17:42 | [515][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1274 ntime: 0084 mem: 3.36 + 04-04 06:17:51 | [515][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0468 ntime: 0077 mem: 3.36 + 04-04 06:17:59 | [515][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0086 ntime: 0079 mem: 3.36 + 04-04 06:18:07 | [515][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0322 ntime: 0080 mem: 3.36 + 04-04 06:18:14 | [515][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1225 ntime: 0086 mem: 3.36 + 04-04 06:18:25 | [515][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1067 ntime: 0082 mem: 3.36 + 04-04 06:18:30 | [515][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0961 ntime: 0081 mem: 3.36 + 04-04 06:18:34 | Time info >>>> elapsed: 535.19 mins remain: 502.00 mins + 04-04 06:18:34 | [516][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0075 ntime: 0078 mem: 3.36 + 04-04 06:18:41 | [516][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0689 ntime: 0090 mem: 3.36 + 04-04 06:18:47 | [516][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0638 ntime: 0079 mem: 3.36 + 04-04 06:18:56 | [516][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0321 ntime: 0078 mem: 3.36 + 04-04 06:19:05 | [516][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1084 ntime: 0081 mem: 3.36 + 04-04 06:19:14 | [516][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1288 ntime: 0079 mem: 3.36 + 04-04 06:19:20 | [516][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0059 ntime: 0077 mem: 3.36 + 04-04 06:19:28 | [516][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1418 ntime: 0089 mem: 3.36 + 04-04 06:19:38 | [516][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0816 ntime: 0083 mem: 3.36 + 04-04 06:19:46 | [516][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0062 ntime: 0084 mem: 3.36 + 04-04 06:19:54 | [516][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0188 ntime: 0076 mem: 3.36 + 04-04 06:20:02 | [516][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1136 ntime: 0092 mem: 3.36 + 04-04 06:20:10 | [516][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0430 ntime: 0077 mem: 3.36 + 04-04 06:20:18 | [516][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1122 ntime: 0076 mem: 3.36 + 04-04 06:20:24 | [516][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1253 ntime: 0073 mem: 3.36 + 04-04 06:20:32 | [516][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0993 ntime: 0077 mem: 3.36 + 04-04 06:20:41 | [516][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0098 ntime: 0075 mem: 3.36 + 04-04 06:20:50 | [516][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1329 ntime: 0084 mem: 3.36 + 04-04 06:20:58 | Time info >>>> elapsed: 537.59 mins remain: 502.24 mins + 04-04 06:20:59 | [517][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0689 ntime: 0080 mem: 3.36 + 04-04 06:21:07 | [517][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1212 ntime: 0079 mem: 3.36 + 04-04 06:21:17 | [517][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1540 ntime: 0081 mem: 3.36 + 04-04 06:21:24 | [517][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0080 ntime: 0073 mem: 3.36 + 04-04 06:21:31 | [517][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1426 ntime: 0083 mem: 3.36 + 04-04 06:21:38 | [517][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 06:21:45 | [517][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1143 ntime: 0085 mem: 3.36 + 04-04 06:21:51 | [517][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0132 ntime: 0079 mem: 3.36 + 04-04 06:21:58 | [517][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0114 ntime: 0075 mem: 3.36 + 04-04 06:22:05 | [517][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1103 ntime: 0075 mem: 3.36 + 04-04 06:22:13 | [517][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1352 ntime: 0086 mem: 3.36 + 04-04 06:22:22 | [517][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0072 mem: 3.36 + 04-04 06:22:30 | [517][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1003 ntime: 0080 mem: 3.36 + 04-04 06:22:39 | [517][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0100 ntime: 0079 mem: 3.36 + 04-04 06:22:47 | [517][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0619 ntime: 0082 mem: 3.36 + 04-04 06:22:54 | [517][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0622 ntime: 0077 mem: 3.36 + 04-04 06:23:01 | [517][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1206 ntime: 0079 mem: 3.36 + 04-04 06:23:10 | [517][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1410 ntime: 0079 mem: 3.36 + 04-04 06:23:16 | Time info >>>> elapsed: 539.89 mins remain: 502.37 mins + 04-04 06:23:17 | [518][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0955 ntime: 0076 mem: 3.36 + 04-04 06:23:25 | [518][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1152 ntime: 0078 mem: 3.36 + 04-04 06:23:32 | [518][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0595 ntime: 0085 mem: 3.36 + 04-04 06:23:41 | [518][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1123 ntime: 0081 mem: 3.36 + 04-04 06:23:49 | [518][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1156 ntime: 0080 mem: 3.36 + 04-04 06:23:55 | [518][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1371 ntime: 0090 mem: 3.36 + 04-04 06:24:03 | [518][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0081 mem: 3.36 + 04-04 06:24:09 | [518][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0796 ntime: 0089 mem: 3.36 + 04-04 06:24:15 | [518][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0659 ntime: 0079 mem: 3.36 + 04-04 06:24:24 | [518][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0087 mem: 3.36 + 04-04 06:24:31 | [518][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1387 ntime: 0080 mem: 3.36 + 04-04 06:24:42 | [518][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1212 ntime: 0081 mem: 3.36 + 04-04 06:24:50 | [518][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0136 ntime: 0078 mem: 3.36 + 04-04 06:24:57 | [518][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0083 mem: 3.36 + 04-04 06:25:05 | [518][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0066 ntime: 0077 mem: 3.36 + 04-04 06:25:13 | [518][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1436 ntime: 0073 mem: 3.36 + 04-04 06:25:20 | [518][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0124 ntime: 0083 mem: 3.36 + 04-04 06:25:28 | [518][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0753 ntime: 0081 mem: 3.36 + 04-04 06:25:33 | Time info >>>> elapsed: 542.18 mins remain: 502.49 mins + 04-04 06:25:34 | [519][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0890 ntime: 0082 mem: 3.36 + 04-04 06:25:42 | [519][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0926 ntime: 0080 mem: 3.36 + 04-04 06:25:52 | [519][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0577 ntime: 0079 mem: 3.36 + 04-04 06:26:00 | [519][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0478 ntime: 0078 mem: 3.36 + 04-04 06:26:09 | [519][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1148 ntime: 0086 mem: 3.36 + 04-04 06:26:17 | [519][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0986 ntime: 0084 mem: 3.36 + 04-04 06:26:27 | [519][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1231 ntime: 0077 mem: 3.36 + 04-04 06:26:36 | [519][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0125 ntime: 0078 mem: 3.36 + 04-04 06:26:45 | [519][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0494 ntime: 0076 mem: 3.36 + 04-04 06:26:55 | [519][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0203 ntime: 0080 mem: 3.36 + 04-04 06:27:03 | [519][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1098 ntime: 0083 mem: 3.36 + 04-04 06:27:10 | [519][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0627 ntime: 0068 mem: 3.36 + 04-04 06:27:20 | [519][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0264 ntime: 0073 mem: 3.36 + 04-04 06:27:29 | [519][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1788 ntime: 0083 mem: 3.36 + 04-04 06:27:36 | [519][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0069 ntime: 0074 mem: 3.36 + 04-04 06:27:44 | [519][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0076 mem: 3.36 + 04-04 06:27:51 | [519][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0865 ntime: 0079 mem: 3.36 + 04-04 06:28:00 | [519][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0092 ntime: 0088 mem: 3.36 + 04-04 06:28:07 | Time info >>>> elapsed: 544.75 mins remain: 502.85 mins + 04-04 06:28:09 | [520][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1248 ntime: 0076 mem: 3.36 + 04-04 06:28:16 | [520][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1056 ntime: 0079 mem: 3.36 + 04-04 06:28:25 | [520][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1046 ntime: 0087 mem: 3.36 + 04-04 06:28:32 | [520][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1324 ntime: 0087 mem: 3.36 + 04-04 06:28:41 | [520][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0666 ntime: 0079 mem: 3.36 + 04-04 06:28:49 | [520][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0066 ntime: 0083 mem: 3.36 + 04-04 06:28:57 | [520][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1166 ntime: 0082 mem: 3.36 + 04-04 06:29:05 | [520][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0265 ntime: 0080 mem: 3.36 + 04-04 06:29:13 | [520][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0080 ntime: 0085 mem: 3.36 + 04-04 06:29:20 | [520][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0078 mem: 3.36 + 04-04 06:29:28 | [520][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1199 ntime: 0084 mem: 3.36 + 04-04 06:29:36 | [520][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0078 mem: 3.36 + 04-04 06:29:47 | [520][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0623 ntime: 0082 mem: 3.36 + 04-04 06:29:57 | [520][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1366 ntime: 0082 mem: 3.36 + 04-04 06:30:04 | [520][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1327 ntime: 0080 mem: 3.36 + 04-04 06:30:15 | [520][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1072 ntime: 0088 mem: 3.36 + 04-04 06:30:24 | [520][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1069 ntime: 0086 mem: 3.36 + 04-04 06:30:33 | [520][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1493 ntime: 0083 mem: 3.36 + 04-04 06:30:37 | Time info >>>> elapsed: 547.25 mins remain: 503.13 mins + 04-04 06:30:37 | [521][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 06:30:47 | [521][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1544 ntime: 0086 mem: 3.36 + 04-04 06:30:56 | [521][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0714 ntime: 0080 mem: 3.36 + 04-04 06:31:06 | [521][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1904 ntime: 0084 mem: 3.36 + 04-04 06:31:14 | [521][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1364 ntime: 0086 mem: 3.36 + 04-04 06:31:23 | [521][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0923 ntime: 0085 mem: 3.36 + 04-04 06:31:30 | [521][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0372 ntime: 0080 mem: 3.36 + 04-04 06:31:38 | [521][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0444 ntime: 0078 mem: 3.36 + 04-04 06:31:47 | [521][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1196 ntime: 0085 mem: 3.36 + 04-04 06:31:55 | [521][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0135 ntime: 0080 mem: 3.36 + 04-04 06:32:03 | [521][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1690 ntime: 0072 mem: 3.36 + 04-04 06:32:12 | [521][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0122 ntime: 0078 mem: 3.36 + 04-04 06:32:21 | [521][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1140 ntime: 0085 mem: 3.36 + 04-04 06:32:29 | [521][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0097 ntime: 0087 mem: 3.36 + 04-04 06:32:36 | [521][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0074 mem: 3.36 + 04-04 06:32:42 | [521][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0059 ntime: 0077 mem: 3.36 + 04-04 06:32:50 | [521][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0471 ntime: 0073 mem: 3.36 + 04-04 06:32:58 | [521][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0078 ntime: 0083 mem: 3.36 + 04-04 06:33:05 | Time info >>>> elapsed: 549.71 mins remain: 503.37 mins + 04-04 06:33:06 | [522][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1515 ntime: 0075 mem: 3.36 + 04-04 06:33:13 | [522][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0078 mem: 3.36 + 04-04 06:33:21 | [522][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0137 ntime: 0081 mem: 3.36 + 04-04 06:33:29 | [522][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0918 ntime: 0078 mem: 3.36 + 04-04 06:33:35 | [522][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0480 ntime: 0076 mem: 3.36 + 04-04 06:33:43 | [522][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0077 mem: 3.36 + 04-04 06:33:51 | [522][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0815 ntime: 0087 mem: 3.36 + 04-04 06:33:59 | [522][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0169 ntime: 0099 mem: 3.36 + 04-04 06:34:07 | [522][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0803 ntime: 0079 mem: 3.36 + 04-04 06:34:13 | [522][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0055 ntime: 0082 mem: 3.36 + 04-04 06:34:23 | [522][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0084 mem: 3.36 + 04-04 06:34:31 | [522][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0993 ntime: 0085 mem: 3.36 + 04-04 06:34:38 | [522][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0916 ntime: 0080 mem: 3.36 + 04-04 06:34:46 | [522][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0483 ntime: 0080 mem: 3.36 + 04-04 06:34:55 | [522][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1372 ntime: 0077 mem: 3.36 + 04-04 06:35:02 | [522][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0077 mem: 3.36 + 04-04 06:35:09 | [522][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0067 ntime: 0080 mem: 3.36 + 04-04 06:35:19 | [522][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0652 ntime: 0087 mem: 3.36 + 04-04 06:35:26 | Time info >>>> elapsed: 552.07 mins remain: 503.51 mins + 04-04 06:35:28 | [523][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1424 ntime: 0085 mem: 3.36 + 04-04 06:35:38 | [523][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1494 ntime: 0080 mem: 3.36 + 04-04 06:35:46 | [523][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0072 ntime: 0081 mem: 3.36 + 04-04 06:35:54 | [523][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1420 ntime: 0076 mem: 3.36 + 04-04 06:36:03 | [523][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1227 ntime: 0075 mem: 3.36 + 04-04 06:36:12 | [523][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0197 ntime: 0076 mem: 3.36 + 04-04 06:36:20 | [523][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0756 ntime: 0084 mem: 3.36 + 04-04 06:36:28 | [523][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0093 ntime: 0080 mem: 3.36 + 04-04 06:36:38 | [523][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0866 ntime: 0078 mem: 3.36 + 04-04 06:36:46 | [523][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0082 mem: 3.36 + 04-04 06:36:54 | [523][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1459 ntime: 0078 mem: 3.36 + 04-04 06:37:04 | [523][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0073 mem: 3.36 + 04-04 06:37:11 | [523][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0074 ntime: 0078 mem: 3.36 + 04-04 06:37:20 | [523][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0084 mem: 3.36 + 04-04 06:37:27 | [523][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0982 ntime: 0080 mem: 3.36 + 04-04 06:37:34 | [523][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0129 ntime: 0074 mem: 3.36 + 04-04 06:37:43 | [523][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0630 ntime: 0086 mem: 3.36 + 04-04 06:37:53 | [523][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1153 ntime: 0074 mem: 3.36 + 04-04 06:38:00 | Time info >>>> elapsed: 554.62 mins remain: 503.82 mins + 04-04 06:38:00 | [524][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0342 ntime: 0079 mem: 3.36 + 04-04 06:38:10 | [524][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1112 ntime: 0082 mem: 3.36 + 04-04 06:38:18 | [524][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1303 ntime: 0078 mem: 3.36 + 04-04 06:38:27 | [524][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0298 ntime: 0081 mem: 3.36 + 04-04 06:38:37 | [524][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0342 ntime: 0089 mem: 3.36 + 04-04 06:38:45 | [524][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0071 ntime: 0073 mem: 3.36 + 04-04 06:38:52 | [524][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1248 ntime: 0089 mem: 3.36 + 04-04 06:39:00 | [524][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1304 ntime: 0077 mem: 3.36 + 04-04 06:39:07 | [524][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0556 ntime: 0074 mem: 3.36 + 04-04 06:39:17 | [524][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1305 ntime: 0081 mem: 3.36 + 04-04 06:39:27 | [524][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1270 ntime: 0080 mem: 3.36 + 04-04 06:39:39 | [524][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1407 ntime: 0084 mem: 3.36 + 04-04 06:39:48 | [524][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0643 ntime: 0087 mem: 3.36 + 04-04 06:39:57 | [524][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1504 ntime: 0061 mem: 3.36 + 04-04 06:40:04 | [524][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0989 ntime: 0088 mem: 3.36 + 04-04 06:40:14 | [524][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1185 ntime: 0082 mem: 3.36 + 04-04 06:40:23 | [524][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1442 ntime: 0078 mem: 3.36 + 04-04 06:40:33 | [524][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0076 mem: 3.36 + 04-04 06:40:39 | Time info >>>> elapsed: 557.28 mins remain: 504.21 mins + 04-04 06:40:39 | [525][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0059 ntime: 0076 mem: 3.36 + 04-04 06:40:48 | [525][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1060 ntime: 0087 mem: 3.36 + 04-04 06:40:55 | [525][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0341 ntime: 0081 mem: 3.36 + 04-04 06:41:02 | [525][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1335 ntime: 0089 mem: 3.36 + 04-04 06:41:11 | [525][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1464 ntime: 0075 mem: 3.36 + 04-04 06:41:17 | [525][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0796 ntime: 0084 mem: 3.36 + 04-04 06:41:27 | [525][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0102 ntime: 0083 mem: 3.36 + 04-04 06:41:34 | [525][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0114 ntime: 0059 mem: 3.36 + 04-04 06:41:43 | [525][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1387 ntime: 0074 mem: 3.36 + 04-04 06:41:52 | [525][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0079 mem: 3.36 + 04-04 06:42:01 | [525][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0055 ntime: 0074 mem: 3.36 + 04-04 06:42:09 | [525][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1045 ntime: 0084 mem: 3.36 + 04-04 06:42:16 | [525][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0630 ntime: 0075 mem: 3.36 + 04-04 06:42:23 | [525][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0074 ntime: 0060 mem: 3.36 + 04-04 06:42:33 | [525][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1272 ntime: 0077 mem: 3.36 + 04-04 06:42:41 | [525][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1330 ntime: 0078 mem: 3.36 + 04-04 06:42:50 | [525][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1407 ntime: 0079 mem: 3.36 + 04-04 06:43:00 | [525][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0571 ntime: 0076 mem: 3.36 + 04-04 06:43:06 | Time info >>>> elapsed: 559.73 mins remain: 504.40 mins + 04-04 06:43:07 | [526][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0902 ntime: 0076 mem: 3.36 + 04-04 06:43:13 | [526][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0080 mem: 3.36 + 04-04 06:43:20 | [526][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0070 ntime: 0082 mem: 3.36 + 04-04 06:43:27 | [526][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0081 mem: 3.36 + 04-04 06:43:36 | [526][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1492 ntime: 0078 mem: 3.36 + 04-04 06:43:43 | [526][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0923 ntime: 0076 mem: 3.36 + 04-04 06:43:53 | [526][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1140 ntime: 0087 mem: 3.36 + 04-04 06:44:00 | [526][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0687 ntime: 0081 mem: 3.36 + 04-04 06:44:08 | [526][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1580 ntime: 0076 mem: 3.36 + 04-04 06:44:15 | [526][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0078 mem: 3.36 + 04-04 06:44:22 | [526][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0110 ntime: 0079 mem: 3.36 + 04-04 06:44:29 | [526][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0904 ntime: 0085 mem: 3.36 + 04-04 06:44:37 | [526][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0085 ntime: 0057 mem: 3.36 + 04-04 06:44:45 | [526][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0590 ntime: 0077 mem: 3.36 + 04-04 06:44:53 | [526][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0049 ntime: 0076 mem: 3.36 + 04-04 06:45:01 | [526][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0082 ntime: 0084 mem: 3.36 + 04-04 06:45:09 | [526][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0711 ntime: 0081 mem: 3.36 + 04-04 06:45:16 | [526][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1185 ntime: 0080 mem: 3.36 + 04-04 06:45:24 | Time info >>>> elapsed: 562.03 mins remain: 504.44 mins + 04-04 06:45:26 | [527][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1333 ntime: 0069 mem: 3.36 + 04-04 06:45:34 | [527][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0088 ntime: 0076 mem: 3.36 + 04-04 06:45:42 | [527][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0399 ntime: 0078 mem: 3.36 + 04-04 06:45:51 | [527][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0734 ntime: 0079 mem: 3.36 + 04-04 06:45:59 | [527][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0398 ntime: 0083 mem: 3.36 + 04-04 06:46:08 | [527][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 06:46:19 | [527][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0747 ntime: 0083 mem: 3.36 + 04-04 06:46:26 | [527][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0134 ntime: 0084 mem: 3.36 + 04-04 06:46:36 | [527][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1156 ntime: 0088 mem: 3.36 + 04-04 06:46:44 | [527][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1045 ntime: 0087 mem: 3.36 + 04-04 06:46:53 | [527][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0089 mem: 3.36 + 04-04 06:47:01 | [527][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1034 ntime: 0079 mem: 3.36 + 04-04 06:47:11 | [527][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1118 ntime: 0091 mem: 3.36 + 04-04 06:47:18 | [527][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1370 ntime: 0081 mem: 3.36 + 04-04 06:47:28 | [527][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1216 ntime: 0056 mem: 3.36 + 04-04 06:47:35 | [527][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0613 ntime: 0085 mem: 3.36 + 04-04 06:47:45 | [527][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1566 ntime: 0081 mem: 3.36 + 04-04 06:47:56 | [527][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0775 ntime: 0081 mem: 3.36 + 04-04 06:48:04 | Time info >>>> elapsed: 564.69 mins remain: 504.80 mins + 04-04 06:48:04 | [528][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 06:48:13 | [528][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1227 ntime: 0070 mem: 3.36 + 04-04 06:48:19 | [528][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0075 mem: 3.36 + 04-04 06:48:29 | [528][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0104 ntime: 0078 mem: 3.36 + 04-04 06:48:40 | [528][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1224 ntime: 0081 mem: 3.36 + 04-04 06:48:48 | [528][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1311 ntime: 0086 mem: 3.36 + 04-04 06:48:56 | [528][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0121 ntime: 0079 mem: 3.36 + 04-04 06:49:04 | [528][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0803 ntime: 0076 mem: 3.36 + 04-04 06:49:13 | [528][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0094 ntime: 0079 mem: 3.36 + 04-04 06:49:19 | [528][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0064 ntime: 0080 mem: 3.36 + 04-04 06:49:27 | [528][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0537 ntime: 0080 mem: 3.36 + 04-04 06:49:34 | [528][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0955 ntime: 0083 mem: 3.36 + 04-04 06:49:43 | [528][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0957 ntime: 0078 mem: 3.36 + 04-04 06:49:54 | [528][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0374 ntime: 0079 mem: 3.36 + 04-04 06:50:02 | [528][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1262 ntime: 0080 mem: 3.36 + 04-04 06:50:11 | [528][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0515 ntime: 0080 mem: 3.36 + 04-04 06:50:18 | [528][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0047 ntime: 0076 mem: 3.36 + 04-04 06:50:23 | [528][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0076 mem: 3.36 + 04-04 06:50:30 | Time info >>>> elapsed: 567.12 mins remain: 504.94 mins + 04-04 06:50:30 | [529][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0105 ntime: 0080 mem: 3.36 + 04-04 06:50:37 | [529][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1205 ntime: 0083 mem: 3.36 + 04-04 06:50:43 | [529][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1176 ntime: 0077 mem: 3.36 + 04-04 06:50:49 | [529][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0763 ntime: 0082 mem: 3.36 + 04-04 06:50:57 | [529][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0074 mem: 3.36 + 04-04 06:51:05 | [529][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0923 ntime: 0079 mem: 3.36 + 04-04 06:51:11 | [529][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0806 ntime: 0083 mem: 3.36 + 04-04 06:51:19 | [529][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0979 ntime: 0076 mem: 3.36 + 04-04 06:51:24 | [529][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0180 ntime: 0079 mem: 3.36 + 04-04 06:51:32 | [529][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1127 ntime: 0074 mem: 3.36 + 04-04 06:51:43 | [529][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0888 ntime: 0075 mem: 3.36 + 04-04 06:51:50 | [529][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0067 ntime: 0081 mem: 3.36 + 04-04 06:51:57 | [529][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 06:52:04 | [529][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0343 ntime: 0076 mem: 3.36 + 04-04 06:52:11 | [529][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0079 mem: 3.36 + 04-04 06:52:19 | [529][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0080 ntime: 0077 mem: 3.36 + 04-04 06:52:25 | [529][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1094 ntime: 0085 mem: 3.36 + 04-04 06:52:34 | [529][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1310 ntime: 0078 mem: 3.36 + 04-04 06:52:39 | Time info >>>> elapsed: 569.28 mins remain: 504.84 mins + 04-04 06:52:40 | [530][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0215 ntime: 0076 mem: 3.36 + 04-04 06:52:46 | [530][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0047 ntime: 0076 mem: 3.36 + 04-04 06:52:53 | [530][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0085 mem: 3.36 + 04-04 06:53:01 | [530][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0669 ntime: 0086 mem: 3.36 + 04-04 06:53:08 | [530][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0805 ntime: 0089 mem: 3.36 + 04-04 06:53:15 | [530][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0598 ntime: 0083 mem: 3.36 + 04-04 06:53:22 | [530][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0088 ntime: 0082 mem: 3.36 + 04-04 06:53:32 | [530][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0080 mem: 3.36 + 04-04 06:53:39 | [530][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0658 ntime: 0074 mem: 3.36 + 04-04 06:53:47 | [530][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0632 ntime: 0084 mem: 3.36 + 04-04 06:53:56 | [530][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0112 ntime: 0078 mem: 3.36 + 04-04 06:54:04 | [530][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1155 ntime: 0077 mem: 3.36 + 04-04 06:54:09 | [530][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0081 mem: 3.36 + 04-04 06:54:15 | [530][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0072 mem: 3.36 + 04-04 06:54:23 | [530][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0129 ntime: 0078 mem: 3.36 + 04-04 06:54:33 | [530][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0085 mem: 3.36 + 04-04 06:54:42 | [530][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1144 ntime: 0081 mem: 3.36 + 04-04 06:54:49 | [530][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0509 ntime: 0079 mem: 3.36 + 04-04 06:54:58 | Time info >>>> elapsed: 571.59 mins remain: 504.85 mins + 04-04 06:54:59 | [531][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0690 ntime: 0077 mem: 3.36 + 04-04 06:55:07 | [531][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0317 ntime: 0082 mem: 3.36 + 04-04 06:55:14 | [531][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0082 ntime: 0078 mem: 3.36 + 04-04 06:55:20 | [531][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0069 ntime: 0084 mem: 3.36 + 04-04 06:55:25 | [531][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0895 ntime: 0080 mem: 3.36 + 04-04 06:55:34 | [531][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1235 ntime: 0079 mem: 3.36 + 04-04 06:55:41 | [531][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0609 ntime: 0082 mem: 3.36 + 04-04 06:55:50 | [531][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1302 ntime: 0077 mem: 3.36 + 04-04 06:55:58 | [531][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1224 ntime: 0082 mem: 3.36 + 04-04 06:56:05 | [531][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1258 ntime: 0086 mem: 3.36 + 04-04 06:56:12 | [531][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0848 ntime: 0079 mem: 3.36 + 04-04 06:56:18 | [531][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0075 ntime: 0080 mem: 3.36 + 04-04 06:56:26 | [531][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1302 ntime: 0081 mem: 3.36 + 04-04 06:56:31 | [531][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0109 ntime: 0079 mem: 3.36 + 04-04 06:56:39 | [531][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1100 ntime: 0077 mem: 3.36 + 04-04 06:56:46 | [531][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1329 ntime: 0079 mem: 3.36 + 04-04 06:56:53 | [531][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0623 ntime: 0085 mem: 3.36 + 04-04 06:57:00 | [531][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0072 mem: 3.36 + 04-04 06:57:07 | Time info >>>> elapsed: 573.75 mins remain: 504.72 mins + 04-04 06:57:08 | [532][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1306 ntime: 0082 mem: 3.36 + 04-04 06:57:16 | [532][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0945 ntime: 0088 mem: 3.36 + 04-04 06:57:22 | [532][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0957 ntime: 0082 mem: 3.36 + 04-04 06:57:30 | [532][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 06:57:37 | [532][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1047 ntime: 0085 mem: 3.36 + 04-04 06:57:44 | [532][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0885 ntime: 0089 mem: 3.36 + 04-04 06:57:51 | [532][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0081 mem: 3.36 + 04-04 06:57:57 | [532][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0082 mem: 3.36 + 04-04 06:58:07 | [532][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0610 ntime: 0081 mem: 3.36 + 04-04 06:58:13 | [532][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0963 ntime: 0078 mem: 3.36 + 04-04 06:58:21 | [532][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0913 ntime: 0085 mem: 3.36 + 04-04 06:58:29 | [532][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0715 ntime: 0080 mem: 3.36 + 04-04 06:58:35 | [532][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1046 ntime: 0082 mem: 3.36 + 04-04 06:58:41 | [532][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1213 ntime: 0079 mem: 3.36 + 04-04 06:58:49 | [532][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0913 ntime: 0082 mem: 3.36 + 04-04 06:58:54 | [532][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0124 ntime: 0078 mem: 3.36 + 04-04 06:59:01 | [532][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0112 ntime: 0082 mem: 3.36 + 04-04 06:59:10 | [532][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1302 ntime: 0075 mem: 3.36 + 04-04 06:59:14 | Time info >>>> elapsed: 575.86 mins remain: 504.56 mins + 04-04 06:59:15 | [533][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0546 ntime: 0072 mem: 3.36 + 04-04 06:59:22 | [533][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0842 ntime: 0081 mem: 3.36 + 04-04 06:59:29 | [533][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1107 ntime: 0082 mem: 3.36 + 04-04 06:59:38 | [533][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0644 ntime: 0082 mem: 3.36 + 04-04 06:59:45 | [533][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0881 ntime: 0079 mem: 3.36 + 04-04 06:59:53 | [533][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0872 ntime: 0094 mem: 3.36 + 04-04 07:00:01 | [533][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0957 ntime: 0073 mem: 3.36 + 04-04 07:00:10 | [533][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1425 ntime: 0085 mem: 3.36 + 04-04 07:00:18 | [533][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0535 ntime: 0078 mem: 3.36 + 04-04 07:00:26 | [533][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0089 ntime: 0078 mem: 3.36 + 04-04 07:00:33 | [533][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0195 ntime: 0077 mem: 3.36 + 04-04 07:00:40 | [533][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0415 ntime: 0081 mem: 3.36 + 04-04 07:00:49 | [533][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0674 ntime: 0076 mem: 3.36 + 04-04 07:00:53 | [533][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0138 ntime: 0080 mem: 3.36 + 04-04 07:01:01 | [533][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0105 ntime: 0080 mem: 3.36 + 04-04 07:01:07 | [533][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0124 ntime: 0079 mem: 3.36 + 04-04 07:01:12 | [533][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0185 ntime: 0083 mem: 3.36 + 04-04 07:01:19 | [533][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0080 mem: 3.36 + 04-04 07:01:23 | Time info >>>> elapsed: 578.01 mins remain: 504.41 mins + 04-04 07:01:24 | [534][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0860 ntime: 0077 mem: 3.36 + 04-04 07:01:32 | [534][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0088 ntime: 0081 mem: 3.36 + 04-04 07:01:38 | [534][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0343 ntime: 0082 mem: 3.36 + 04-04 07:01:47 | [534][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0755 ntime: 0083 mem: 3.36 + 04-04 07:01:53 | [534][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0625 ntime: 0082 mem: 3.36 + 04-04 07:01:58 | [534][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0082 mem: 3.36 + 04-04 07:02:05 | [534][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0058 mem: 3.36 + 04-04 07:02:12 | [534][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1027 ntime: 0081 mem: 3.36 + 04-04 07:02:18 | [534][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0074 ntime: 0083 mem: 3.36 + 04-04 07:02:25 | [534][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0463 ntime: 0073 mem: 3.36 + 04-04 07:02:32 | [534][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0083 mem: 3.36 + 04-04 07:02:40 | [534][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0970 ntime: 0087 mem: 3.36 + 04-04 07:02:45 | [534][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0077 ntime: 0073 mem: 3.36 + 04-04 07:02:51 | [534][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0094 ntime: 0080 mem: 3.36 + 04-04 07:02:58 | [534][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0873 ntime: 0087 mem: 3.36 + 04-04 07:03:05 | [534][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0128 ntime: 0077 mem: 3.36 + 04-04 07:03:12 | [534][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0105 ntime: 0084 mem: 3.36 + 04-04 07:03:19 | [534][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0062 ntime: 0079 mem: 3.36 + 04-04 07:03:24 | Time info >>>> elapsed: 580.03 mins remain: 504.14 mins + 04-04 07:03:25 | [535][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1000 ntime: 0079 mem: 3.36 + 04-04 07:03:33 | [535][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1105 ntime: 0075 mem: 3.36 + 04-04 07:03:42 | [535][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0574 ntime: 0085 mem: 3.36 + 04-04 07:03:49 | [535][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0107 ntime: 0079 mem: 3.36 + 04-04 07:03:55 | [535][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0214 ntime: 0081 mem: 3.36 + 04-04 07:04:01 | [535][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1132 ntime: 0087 mem: 3.36 + 04-04 07:04:08 | [535][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0059 ntime: 0075 mem: 3.36 + 04-04 07:04:17 | [535][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0682 ntime: 0087 mem: 3.36 + 04-04 07:04:24 | [535][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0978 ntime: 0081 mem: 3.36 + 04-04 07:04:30 | [535][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0283 ntime: 0081 mem: 3.36 + 04-04 07:04:35 | [535][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0764 ntime: 0077 mem: 3.36 + 04-04 07:04:41 | [535][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0791 ntime: 0082 mem: 3.36 + 04-04 07:04:47 | [535][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1331 ntime: 0080 mem: 3.36 + 04-04 07:04:55 | [535][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1044 ntime: 0082 mem: 3.36 + 04-04 07:05:01 | [535][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0137 ntime: 0073 mem: 3.36 + 04-04 07:05:08 | [535][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1032 ntime: 0079 mem: 3.36 + 04-04 07:05:14 | [535][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0083 ntime: 0076 mem: 3.36 + 04-04 07:05:21 | [535][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0084 mem: 3.36 + 04-04 07:05:26 | Time info >>>> elapsed: 582.07 mins remain: 503.88 mins + 04-04 07:05:27 | [536][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0084 ntime: 0076 mem: 3.36 + 04-04 07:05:32 | [536][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0066 ntime: 0079 mem: 3.36 + 04-04 07:05:39 | [536][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0758 ntime: 0078 mem: 3.36 + 04-04 07:05:46 | [536][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0803 ntime: 0078 mem: 3.36 + 04-04 07:05:54 | [536][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1134 ntime: 0073 mem: 3.36 + 04-04 07:06:00 | [536][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1184 ntime: 0074 mem: 3.36 + 04-04 07:06:08 | [536][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0325 ntime: 0081 mem: 3.36 + 04-04 07:06:15 | [536][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1305 ntime: 0084 mem: 3.36 + 04-04 07:06:23 | [536][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1004 ntime: 0075 mem: 3.36 + 04-04 07:06:31 | [536][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0668 ntime: 0089 mem: 3.36 + 04-04 07:06:40 | [536][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0071 mem: 3.36 + 04-04 07:06:47 | [536][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0175 ntime: 0078 mem: 3.36 + 04-04 07:06:56 | [536][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0570 ntime: 0079 mem: 3.36 + 04-04 07:07:04 | [536][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0282 ntime: 0077 mem: 3.36 + 04-04 07:07:12 | [536][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0403 ntime: 0088 mem: 3.36 + 04-04 07:07:19 | [536][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0150 ntime: 0079 mem: 3.36 + 04-04 07:07:29 | [536][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1192 ntime: 0079 mem: 3.36 + 04-04 07:07:37 | [536][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0115 ntime: 0087 mem: 3.36 + 04-04 07:07:43 | Time info >>>> elapsed: 584.35 mins remain: 503.83 mins + 04-04 07:07:44 | [537][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0119 ntime: 0084 mem: 3.36 + 04-04 07:07:50 | [537][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0081 mem: 3.36 + 04-04 07:07:59 | [537][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0694 ntime: 0075 mem: 3.36 + 04-04 07:08:04 | [537][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0073 mem: 3.36 + 04-04 07:08:11 | [537][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0490 ntime: 0085 mem: 3.36 + 04-04 07:08:18 | [537][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0972 ntime: 0079 mem: 3.36 + 04-04 07:08:24 | [537][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1098 ntime: 0084 mem: 3.36 + 04-04 07:08:29 | [537][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0069 ntime: 0081 mem: 3.36 + 04-04 07:08:36 | [537][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0348 ntime: 0083 mem: 3.36 + 04-04 07:08:44 | [537][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1216 ntime: 0076 mem: 3.36 + 04-04 07:08:49 | [537][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0629 ntime: 0089 mem: 3.36 + 04-04 07:08:58 | [537][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0763 ntime: 0086 mem: 3.36 + 04-04 07:09:03 | [537][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0994 ntime: 0080 mem: 3.36 + 04-04 07:09:09 | [537][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0750 ntime: 0079 mem: 3.36 + 04-04 07:09:18 | [537][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0970 ntime: 0088 mem: 3.36 + 04-04 07:09:24 | [537][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0511 ntime: 0084 mem: 3.36 + 04-04 07:09:31 | [537][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0144 ntime: 0087 mem: 3.36 + 04-04 07:09:38 | [537][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0139 ntime: 0080 mem: 3.36 + 04-04 07:09:43 | Time info >>>> elapsed: 586.35 mins remain: 503.52 mins + 04-04 07:09:45 | [538][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1308 ntime: 0083 mem: 3.36 + 04-04 07:09:50 | [538][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1333 ntime: 0082 mem: 3.36 + 04-04 07:09:55 | [538][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0918 ntime: 0081 mem: 3.36 + 04-04 07:10:02 | [538][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0566 ntime: 0081 mem: 3.36 + 04-04 07:10:09 | [538][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0570 ntime: 0083 mem: 3.36 + 04-04 07:10:17 | [538][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0920 ntime: 0087 mem: 3.36 + 04-04 07:10:23 | [538][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0106 ntime: 0081 mem: 3.36 + 04-04 07:10:31 | [538][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0082 ntime: 0080 mem: 3.36 + 04-04 07:10:38 | [538][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0060 mem: 3.36 + 04-04 07:10:42 | [538][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0142 ntime: 0057 mem: 3.36 + 04-04 07:10:49 | [538][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0076 mem: 3.36 + 04-04 07:10:57 | [538][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0500 ntime: 0070 mem: 3.36 + 04-04 07:11:04 | [538][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0623 ntime: 0080 mem: 3.36 + 04-04 07:11:09 | [538][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0071 ntime: 0077 mem: 3.36 + 04-04 07:11:18 | [538][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0279 ntime: 0081 mem: 3.36 + 04-04 07:11:27 | [538][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0708 ntime: 0076 mem: 3.36 + 04-04 07:11:36 | [538][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1236 ntime: 0080 mem: 3.36 + 04-04 07:11:42 | [538][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0758 ntime: 0084 mem: 3.36 + 04-04 07:11:48 | Time info >>>> elapsed: 588.43 mins remain: 503.28 mins + 04-04 07:11:48 | [539][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0050 ntime: 0071 mem: 3.36 + 04-04 07:11:55 | [539][010/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1322 ntime: 0082 mem: 3.36 + 04-04 07:12:01 | [539][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0085 ntime: 0084 mem: 3.36 + 04-04 07:12:09 | [539][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1100 ntime: 0088 mem: 3.36 + 04-04 07:12:15 | [539][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0085 mem: 3.36 + 04-04 07:12:23 | [539][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0667 ntime: 0088 mem: 3.36 + 04-04 07:12:31 | [539][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0142 ntime: 0077 mem: 3.36 + 04-04 07:12:37 | [539][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0116 ntime: 0087 mem: 3.36 + 04-04 07:12:45 | [539][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0896 ntime: 0078 mem: 3.36 + 04-04 07:12:54 | [539][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1350 ntime: 0090 mem: 3.36 + 04-04 07:12:59 | [539][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0702 ntime: 0079 mem: 3.36 + 04-04 07:13:08 | [539][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1459 ntime: 0079 mem: 3.36 + 04-04 07:13:14 | [539][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1091 ntime: 0075 mem: 3.36 + 04-04 07:13:22 | [539][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0810 ntime: 0083 mem: 3.36 + 04-04 07:13:27 | [539][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1345 ntime: 0074 mem: 3.36 + 04-04 07:13:34 | [539][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0067 ntime: 0079 mem: 3.36 + 04-04 07:13:41 | [539][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0087 ntime: 0078 mem: 3.36 + 04-04 07:13:48 | [539][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0419 ntime: 0069 mem: 3.36 + 04-04 07:13:54 | Time info >>>> elapsed: 590.53 mins remain: 503.05 mins + 04-04 07:13:55 | [540][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0175 ntime: 0079 mem: 3.36 + 04-04 07:14:03 | [540][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1349 ntime: 0079 mem: 3.36 + 04-04 07:14:09 | [540][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0183 ntime: 0078 mem: 3.36 + 04-04 07:14:17 | [540][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0446 ntime: 0092 mem: 3.36 + 04-04 07:14:25 | [540][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1279 ntime: 0082 mem: 3.36 + 04-04 07:14:32 | [540][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0307 ntime: 0076 mem: 3.36 + 04-04 07:14:39 | [540][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1347 ntime: 0082 mem: 3.36 + 04-04 07:14:46 | [540][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0536 ntime: 0082 mem: 3.36 + 04-04 07:14:53 | [540][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1076 ntime: 0083 mem: 3.36 + 04-04 07:14:59 | [540][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0586 ntime: 0081 mem: 3.36 + 04-04 07:15:06 | [540][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0350 ntime: 0079 mem: 3.36 + 04-04 07:15:14 | [540][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0185 ntime: 0085 mem: 3.36 + 04-04 07:15:21 | [540][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0046 ntime: 0076 mem: 3.36 + 04-04 07:15:28 | [540][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0166 ntime: 0076 mem: 3.36 + 04-04 07:15:35 | [540][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0697 ntime: 0077 mem: 3.36 + 04-04 07:15:43 | [540][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0069 ntime: 0072 mem: 3.36 + 04-04 07:15:51 | [540][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0120 ntime: 0083 mem: 3.36 + 04-04 07:15:58 | [540][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0935 ntime: 0075 mem: 3.36 + 04-04 07:16:03 | Time info >>>> elapsed: 592.68 mins remain: 502.85 mins + 04-04 07:16:05 | [541][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1282 ntime: 0083 mem: 3.36 + 04-04 07:16:11 | [541][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0560 ntime: 0088 mem: 3.36 + 04-04 07:16:17 | [541][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0151 ntime: 0080 mem: 3.36 + 04-04 07:16:29 | [541][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1378 ntime: 0058 mem: 3.36 + 04-04 07:16:37 | [541][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0229 ntime: 0081 mem: 3.36 + 04-04 07:16:43 | [541][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0072 mem: 3.36 + 04-04 07:16:48 | [541][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0090 ntime: 0078 mem: 3.36 + 04-04 07:16:55 | [541][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0862 ntime: 0076 mem: 3.36 + 04-04 07:17:02 | [541][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1108 ntime: 0076 mem: 3.36 + 04-04 07:17:08 | [541][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0068 ntime: 0080 mem: 3.36 + 04-04 07:17:15 | [541][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1019 ntime: 0084 mem: 3.36 + 04-04 07:17:22 | [541][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0333 ntime: 0084 mem: 3.36 + 04-04 07:17:28 | [541][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0044 ntime: 0081 mem: 3.36 + 04-04 07:17:36 | [541][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1045 ntime: 0077 mem: 3.36 + 04-04 07:17:43 | [541][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0966 ntime: 0083 mem: 3.36 + 04-04 07:17:50 | [541][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0109 ntime: 0080 mem: 3.36 + 04-04 07:17:57 | [541][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0930 ntime: 0077 mem: 3.36 + 04-04 07:18:03 | [541][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0075 mem: 3.36 + 04-04 07:18:09 | Time info >>>> elapsed: 594.78 mins remain: 502.60 mins + 04-04 07:18:10 | [542][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0833 ntime: 0084 mem: 3.36 + 04-04 07:18:17 | [542][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1276 ntime: 0080 mem: 3.36 + 04-04 07:18:24 | [542][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1215 ntime: 0086 mem: 3.36 + 04-04 07:18:31 | [542][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1028 ntime: 0081 mem: 3.36 + 04-04 07:18:41 | [542][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0858 ntime: 0077 mem: 3.36 + 04-04 07:18:50 | [542][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0501 ntime: 0086 mem: 3.36 + 04-04 07:18:58 | [542][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1089 ntime: 0091 mem: 3.36 + 04-04 07:19:05 | [542][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1156 ntime: 0078 mem: 3.36 + 04-04 07:19:11 | [542][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0494 ntime: 0079 mem: 3.36 + 04-04 07:19:19 | [542][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0705 ntime: 0082 mem: 3.36 + 04-04 07:19:26 | [542][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0987 ntime: 0056 mem: 3.36 + 04-04 07:19:32 | [542][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0856 ntime: 0082 mem: 3.36 + 04-04 07:19:39 | [542][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0077 mem: 3.36 + 04-04 07:19:46 | [542][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0055 ntime: 0080 mem: 3.36 + 04-04 07:19:51 | [542][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0333 ntime: 0084 mem: 3.36 + 04-04 07:19:59 | [542][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1079 ntime: 0084 mem: 3.36 + 04-04 07:20:07 | [542][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0260 ntime: 0078 mem: 3.36 + 04-04 07:20:16 | [542][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1533 ntime: 0084 mem: 3.36 + 04-04 07:20:20 | Time info >>>> elapsed: 596.96 mins remain: 502.42 mins + 04-04 07:20:21 | [543][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0604 ntime: 0084 mem: 3.36 + 04-04 07:20:26 | [543][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0732 ntime: 0085 mem: 3.36 + 04-04 07:20:34 | [543][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0629 ntime: 0083 mem: 3.36 + 04-04 07:20:39 | [543][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0320 ntime: 0074 mem: 3.36 + 04-04 07:20:47 | [543][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1343 ntime: 0080 mem: 3.36 + 04-04 07:20:53 | [543][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0075 mem: 3.36 + 04-04 07:21:01 | [543][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1354 ntime: 0084 mem: 3.36 + 04-04 07:21:07 | [543][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0351 ntime: 0084 mem: 3.36 + 04-04 07:21:14 | [543][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0623 ntime: 0089 mem: 3.36 + 04-04 07:21:21 | [543][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0886 ntime: 0081 mem: 3.36 + 04-04 07:21:28 | [543][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0047 ntime: 0079 mem: 3.36 + 04-04 07:21:36 | [543][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0904 ntime: 0079 mem: 3.36 + 04-04 07:21:43 | [543][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1301 ntime: 0082 mem: 3.36 + 04-04 07:21:47 | [543][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0139 ntime: 0081 mem: 3.36 + 04-04 07:21:53 | [543][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0159 ntime: 0078 mem: 3.36 + 04-04 07:21:59 | [543][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0070 ntime: 0071 mem: 3.36 + 04-04 07:22:07 | [543][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1070 ntime: 0067 mem: 3.36 + 04-04 07:22:13 | [543][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1031 ntime: 0081 mem: 3.36 + 04-04 07:22:18 | Time info >>>> elapsed: 598.93 mins remain: 502.05 mins + 04-04 07:22:19 | [544][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0074 mem: 3.36 + 04-04 07:22:25 | [544][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0073 mem: 3.36 + 04-04 07:22:31 | [544][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0458 ntime: 0094 mem: 3.36 + 04-04 07:22:37 | [544][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1013 ntime: 0082 mem: 3.36 + 04-04 07:22:43 | [544][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0819 ntime: 0085 mem: 3.36 + 04-04 07:22:50 | [544][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1065 ntime: 0079 mem: 3.36 + 04-04 07:22:58 | [544][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0074 ntime: 0091 mem: 3.36 + 04-04 07:23:06 | [544][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0832 ntime: 0080 mem: 3.36 + 04-04 07:23:12 | [544][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1122 ntime: 0082 mem: 3.36 + 04-04 07:23:18 | [544][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0540 ntime: 0084 mem: 3.36 + 04-04 07:23:23 | [544][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0170 ntime: 0080 mem: 3.36 + 04-04 07:23:28 | [544][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0085 mem: 3.36 + 04-04 07:23:34 | [544][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0085 ntime: 0082 mem: 3.36 + 04-04 07:23:40 | [544][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0150 ntime: 0082 mem: 3.36 + 04-04 07:23:48 | [544][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0989 ntime: 0076 mem: 3.36 + 04-04 07:23:54 | [544][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0659 ntime: 0087 mem: 3.36 + 04-04 07:24:01 | [544][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1235 ntime: 0077 mem: 3.36 + 04-04 07:24:09 | [544][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1337 ntime: 0076 mem: 3.36 + 04-04 07:24:14 | Time info >>>> elapsed: 600.86 mins remain: 501.63 mins + 04-04 07:24:15 | [545][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0805 ntime: 0082 mem: 3.36 + 04-04 07:24:23 | [545][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0775 ntime: 0088 mem: 3.36 + 04-04 07:24:31 | [545][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0092 ntime: 0078 mem: 3.36 + 04-04 07:24:38 | [545][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0751 ntime: 0080 mem: 3.36 + 04-04 07:24:43 | [545][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0079 mem: 3.36 + 04-04 07:24:51 | [545][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 07:24:58 | [545][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0753 ntime: 0084 mem: 3.36 + 04-04 07:25:06 | [545][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1242 ntime: 0079 mem: 3.36 + 04-04 07:25:14 | [545][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0990 ntime: 0057 mem: 3.36 + 04-04 07:25:22 | [545][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0551 ntime: 0073 mem: 3.36 + 04-04 07:25:29 | [545][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0081 ntime: 0075 mem: 3.36 + 04-04 07:25:35 | [545][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0096 ntime: 0079 mem: 3.36 + 04-04 07:25:43 | [545][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0729 ntime: 0082 mem: 3.36 + 04-04 07:25:50 | [545][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0808 ntime: 0081 mem: 3.36 + 04-04 07:25:58 | [545][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0077 ntime: 0076 mem: 3.36 + 04-04 07:26:04 | [545][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 07:26:12 | [545][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0910 ntime: 0082 mem: 3.36 + 04-04 07:26:19 | [545][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0666 ntime: 0082 mem: 3.36 + 04-04 07:26:25 | Time info >>>> elapsed: 603.05 mins remain: 501.44 mins + 04-04 07:26:26 | [546][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0699 ntime: 0077 mem: 3.36 + 04-04 07:26:34 | [546][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0265 ntime: 0080 mem: 3.36 + 04-04 07:26:42 | [546][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0589 ntime: 0078 mem: 3.36 + 04-04 07:26:48 | [546][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1330 ntime: 0079 mem: 3.36 + 04-04 07:26:55 | [546][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0065 ntime: 0077 mem: 3.36 + 04-04 07:27:03 | [546][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0126 ntime: 0086 mem: 3.36 + 04-04 07:27:11 | [546][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0634 ntime: 0084 mem: 3.36 + 04-04 07:27:15 | [546][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0344 ntime: 0088 mem: 3.36 + 04-04 07:27:24 | [546][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1067 ntime: 0090 mem: 3.36 + 04-04 07:27:31 | [546][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1412 ntime: 0085 mem: 3.36 + 04-04 07:27:38 | [546][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0572 ntime: 0087 mem: 3.36 + 04-04 07:27:46 | [546][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0677 ntime: 0081 mem: 3.36 + 04-04 07:27:55 | [546][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0075 ntime: 0079 mem: 3.36 + 04-04 07:28:04 | [546][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0055 mem: 3.36 + 04-04 07:28:09 | [546][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0199 ntime: 0080 mem: 3.36 + 04-04 07:28:15 | [546][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0155 ntime: 0089 mem: 3.36 + 04-04 07:28:21 | [546][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0064 ntime: 0077 mem: 3.36 + 04-04 07:28:30 | [546][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1220 ntime: 0081 mem: 3.36 + 04-04 07:28:34 | Time info >>>> elapsed: 605.19 mins remain: 501.19 mins + 04-04 07:28:34 | [547][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0072 mem: 3.36 + 04-04 07:28:41 | [547][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1330 ntime: 0077 mem: 3.36 + 04-04 07:28:47 | [547][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0916 ntime: 0080 mem: 3.36 + 04-04 07:28:53 | [547][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0084 ntime: 0077 mem: 3.36 + 04-04 07:29:01 | [547][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1317 ntime: 0085 mem: 3.36 + 04-04 07:29:07 | [547][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0049 ntime: 0073 mem: 3.36 + 04-04 07:29:13 | [547][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0906 ntime: 0087 mem: 3.36 + 04-04 07:29:19 | [547][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0064 ntime: 0070 mem: 3.36 + 04-04 07:29:26 | [547][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0207 ntime: 0082 mem: 3.36 + 04-04 07:29:34 | [547][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0083 ntime: 0082 mem: 3.36 + 04-04 07:29:42 | [547][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1011 ntime: 0080 mem: 3.36 + 04-04 07:29:49 | [547][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0381 ntime: 0075 mem: 3.36 + 04-04 07:29:55 | [547][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0047 ntime: 0080 mem: 3.36 + 04-04 07:30:04 | [547][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1184 ntime: 0076 mem: 3.36 + 04-04 07:30:14 | [547][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1051 ntime: 0081 mem: 3.36 + 04-04 07:30:20 | [547][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0610 ntime: 0077 mem: 3.36 + 04-04 07:30:28 | [547][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0912 ntime: 0084 mem: 3.36 + 04-04 07:30:38 | [547][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1196 ntime: 0076 mem: 3.36 + 04-04 07:30:46 | Time info >>>> elapsed: 607.39 mins remain: 500.98 mins + 04-04 07:30:47 | [548][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1425 ntime: 0079 mem: 3.36 + 04-04 07:30:56 | [548][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1318 ntime: 0084 mem: 3.36 + 04-04 07:31:02 | [548][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1238 ntime: 0078 mem: 3.36 + 04-04 07:31:10 | [548][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1364 ntime: 0077 mem: 3.36 + 04-04 07:31:18 | [548][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0992 ntime: 0082 mem: 3.36 + 04-04 07:31:25 | [548][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0751 ntime: 0084 mem: 3.36 + 04-04 07:31:33 | [548][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1297 ntime: 0086 mem: 3.36 + 04-04 07:31:40 | [548][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1073 ntime: 0084 mem: 3.36 + 04-04 07:31:49 | [548][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0920 ntime: 0085 mem: 3.36 + 04-04 07:31:55 | [548][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0076 mem: 3.36 + 04-04 07:32:02 | [548][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0845 ntime: 0077 mem: 3.36 + 04-04 07:32:11 | [548][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1208 ntime: 0080 mem: 3.36 + 04-04 07:32:18 | [548][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1061 ntime: 0090 mem: 3.36 + 04-04 07:32:23 | [548][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0636 ntime: 0086 mem: 3.36 + 04-04 07:32:30 | [548][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0083 mem: 3.36 + 04-04 07:32:34 | [548][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0140 ntime: 0076 mem: 3.36 + 04-04 07:32:41 | [548][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0576 ntime: 0076 mem: 3.36 + 04-04 07:32:47 | [548][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0247 ntime: 0078 mem: 3.36 + 04-04 07:32:54 | Time info >>>> elapsed: 609.52 mins remain: 500.72 mins + 04-04 07:32:55 | [549][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1320 ntime: 0074 mem: 3.36 + 04-04 07:33:02 | [549][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0282 ntime: 0082 mem: 3.36 + 04-04 07:33:09 | [549][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0916 ntime: 0083 mem: 3.36 + 04-04 07:33:16 | [549][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1066 ntime: 0086 mem: 3.36 + 04-04 07:33:23 | [549][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0520 ntime: 0082 mem: 3.36 + 04-04 07:33:30 | [549][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1135 ntime: 0076 mem: 3.36 + 04-04 07:33:39 | [549][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1090 ntime: 0080 mem: 3.36 + 04-04 07:33:46 | [549][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0059 ntime: 0081 mem: 3.36 + 04-04 07:33:54 | [549][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0143 ntime: 0084 mem: 3.36 + 04-04 07:34:02 | [549][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1320 ntime: 0079 mem: 3.36 + 04-04 07:34:11 | [549][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1125 ntime: 0083 mem: 3.36 + 04-04 07:34:19 | [549][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1354 ntime: 0084 mem: 3.36 + 04-04 07:34:26 | [549][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0742 ntime: 0082 mem: 3.36 + 04-04 07:34:35 | [549][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0855 ntime: 0085 mem: 3.36 + 04-04 07:34:41 | [549][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0692 ntime: 0079 mem: 3.36 + 04-04 07:34:47 | [549][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0587 ntime: 0083 mem: 3.36 + 04-04 07:34:54 | [549][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0076 ntime: 0082 mem: 3.36 + 04-04 07:35:02 | [549][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0715 ntime: 0079 mem: 3.36 + 04-04 07:35:07 | Time info >>>> elapsed: 611.75 mins remain: 500.52 mins + 04-04 07:35:08 | [550][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0499 ntime: 0079 mem: 3.36 + 04-04 07:35:14 | [550][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 07:35:23 | [550][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1403 ntime: 0077 mem: 3.36 + 04-04 07:35:28 | [550][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0928 ntime: 0082 mem: 3.36 + 04-04 07:35:35 | [550][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0093 ntime: 0079 mem: 3.36 + 04-04 07:35:44 | [550][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0846 ntime: 0079 mem: 3.36 + 04-04 07:35:51 | [550][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0229 ntime: 0089 mem: 3.36 + 04-04 07:36:00 | [550][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0656 ntime: 0084 mem: 3.36 + 04-04 07:36:08 | [550][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1167 ntime: 0082 mem: 3.36 + 04-04 07:36:15 | [550][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0080 mem: 3.36 + 04-04 07:36:24 | [550][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0935 ntime: 0080 mem: 3.36 + 04-04 07:36:29 | [550][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0066 ntime: 0081 mem: 3.36 + 04-04 07:36:37 | [550][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0093 ntime: 0076 mem: 3.36 + 04-04 07:36:45 | [550][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1108 ntime: 0079 mem: 3.36 + 04-04 07:36:53 | [550][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0662 ntime: 0077 mem: 3.36 + 04-04 07:37:01 | [550][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0841 ntime: 0080 mem: 3.36 + 04-04 07:37:07 | [550][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0080 mem: 3.36 + 04-04 07:37:14 | [550][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0591 ntime: 0089 mem: 3.36 + 04-04 07:37:19 | Time info >>>> elapsed: 613.95 mins remain: 500.29 mins + 04-04 07:37:20 | [551][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0806 ntime: 0075 mem: 3.36 + 04-04 07:37:27 | [551][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0974 ntime: 0079 mem: 3.36 + 04-04 07:37:33 | [551][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0066 ntime: 0076 mem: 3.36 + 04-04 07:37:41 | [551][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0864 ntime: 0083 mem: 3.36 + 04-04 07:37:48 | [551][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1294 ntime: 0081 mem: 3.36 + 04-04 07:37:53 | [551][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0889 ntime: 0077 mem: 3.36 + 04-04 07:38:00 | [551][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0089 ntime: 0077 mem: 3.36 + 04-04 07:38:06 | [551][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0078 ntime: 0077 mem: 3.36 + 04-04 07:38:12 | [551][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0832 ntime: 0078 mem: 3.36 + 04-04 07:38:20 | [551][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0073 mem: 3.36 + 04-04 07:38:27 | [551][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1399 ntime: 0085 mem: 3.36 + 04-04 07:38:32 | [551][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0069 ntime: 0077 mem: 3.36 + 04-04 07:38:39 | [551][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1011 ntime: 0092 mem: 3.36 + 04-04 07:38:47 | [551][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1183 ntime: 0083 mem: 3.36 + 04-04 07:38:54 | [551][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0221 ntime: 0079 mem: 3.36 + 04-04 07:39:02 | [551][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0562 ntime: 0075 mem: 3.36 + 04-04 07:39:09 | [551][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0591 ntime: 0081 mem: 3.36 + 04-04 07:39:14 | [551][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1055 ntime: 0077 mem: 3.36 + 04-04 07:39:18 | Time info >>>> elapsed: 615.93 mins remain: 499.89 mins + 04-04 07:39:19 | [552][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0512 ntime: 0071 mem: 3.36 + 04-04 07:39:25 | [552][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1296 ntime: 0079 mem: 3.36 + 04-04 07:39:33 | [552][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0099 ntime: 0078 mem: 3.36 + 04-04 07:39:42 | [552][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1065 ntime: 0074 mem: 3.36 + 04-04 07:39:49 | [552][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0910 ntime: 0066 mem: 3.36 + 04-04 07:39:55 | [552][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0688 ntime: 0075 mem: 3.36 + 04-04 07:40:02 | [552][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0888 ntime: 0087 mem: 3.36 + 04-04 07:40:09 | [552][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1039 ntime: 0078 mem: 3.36 + 04-04 07:40:15 | [552][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0095 ntime: 0079 mem: 3.36 + 04-04 07:40:21 | [552][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0080 mem: 3.36 + 04-04 07:40:28 | [552][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0875 ntime: 0083 mem: 3.36 + 04-04 07:40:35 | [552][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0505 ntime: 0078 mem: 3.36 + 04-04 07:40:40 | [552][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1287 ntime: 0082 mem: 3.36 + 04-04 07:40:48 | [552][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0659 ntime: 0079 mem: 3.36 + 04-04 07:40:54 | [552][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0996 ntime: 0085 mem: 3.36 + 04-04 07:41:01 | [552][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1243 ntime: 0073 mem: 3.36 + 04-04 07:41:09 | [552][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0070 ntime: 0076 mem: 3.36 + 04-04 07:41:14 | [552][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0704 ntime: 0082 mem: 3.36 + 04-04 07:41:19 | Time info >>>> elapsed: 617.95 mins remain: 499.50 mins + 04-04 07:41:21 | [553][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1293 ntime: 0079 mem: 3.36 + 04-04 07:41:26 | [553][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0666 ntime: 0083 mem: 3.36 + 04-04 07:41:34 | [553][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0481 ntime: 0074 mem: 3.36 + 04-04 07:41:42 | [553][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0869 ntime: 0081 mem: 3.36 + 04-04 07:41:48 | [553][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1059 ntime: 0090 mem: 3.36 + 04-04 07:41:58 | [553][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0945 ntime: 0081 mem: 3.36 + 04-04 07:42:05 | [553][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1110 ntime: 0073 mem: 3.36 + 04-04 07:42:12 | [553][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0140 ntime: 0076 mem: 3.36 + 04-04 07:42:19 | [553][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0067 ntime: 0084 mem: 3.36 + 04-04 07:42:27 | [553][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0912 ntime: 0078 mem: 3.36 + 04-04 07:42:33 | [553][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0637 ntime: 0082 mem: 3.36 + 04-04 07:42:40 | [553][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1125 ntime: 0077 mem: 3.36 + 04-04 07:42:45 | [553][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0466 ntime: 0067 mem: 3.36 + 04-04 07:42:51 | [553][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0604 ntime: 0071 mem: 3.36 + 04-04 07:42:57 | [553][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 07:43:05 | [553][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0480 ntime: 0087 mem: 3.36 + 04-04 07:43:12 | [553][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1292 ntime: 0086 mem: 3.36 + 04-04 07:43:19 | [553][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1051 ntime: 0082 mem: 3.36 + 04-04 07:43:25 | Time info >>>> elapsed: 620.05 mins remain: 499.18 mins + 04-04 07:43:27 | [554][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1332 ntime: 0078 mem: 3.36 + 04-04 07:43:34 | [554][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0915 ntime: 0075 mem: 3.36 + 04-04 07:43:41 | [554][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0068 ntime: 0081 mem: 3.36 + 04-04 07:43:49 | [554][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0081 mem: 3.36 + 04-04 07:43:57 | [554][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1068 ntime: 0089 mem: 3.36 + 04-04 07:44:03 | [554][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0878 ntime: 0086 mem: 3.36 + 04-04 07:44:09 | [554][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0077 mem: 3.36 + 04-04 07:44:15 | [554][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0090 ntime: 0085 mem: 3.36 + 04-04 07:44:23 | [554][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0673 ntime: 0083 mem: 3.36 + 04-04 07:44:30 | [554][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1351 ntime: 0079 mem: 3.36 + 04-04 07:44:38 | [554][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0545 ntime: 0079 mem: 3.36 + 04-04 07:44:44 | [554][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0340 ntime: 0082 mem: 3.36 + 04-04 07:44:50 | [554][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0084 ntime: 0084 mem: 3.36 + 04-04 07:44:58 | [554][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1114 ntime: 0085 mem: 3.36 + 04-04 07:45:03 | [554][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0324 ntime: 0077 mem: 3.36 + 04-04 07:45:10 | [554][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0402 ntime: 0079 mem: 3.36 + 04-04 07:45:17 | [554][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0700 ntime: 0076 mem: 3.36 + 04-04 07:45:25 | [554][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1280 ntime: 0084 mem: 3.36 + 04-04 07:45:29 | Time info >>>> elapsed: 622.11 mins remain: 498.81 mins + 04-04 07:45:30 | [555][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0841 ntime: 0085 mem: 3.36 + 04-04 07:45:36 | [555][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1165 ntime: 0086 mem: 3.36 + 04-04 07:45:42 | [555][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0830 ntime: 0082 mem: 3.36 + 04-04 07:45:49 | [555][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0718 ntime: 0077 mem: 3.36 + 04-04 07:45:56 | [555][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1141 ntime: 0080 mem: 3.36 + 04-04 07:46:03 | [555][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0873 ntime: 0076 mem: 3.36 + 04-04 07:46:11 | [555][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0277 ntime: 0080 mem: 3.36 + 04-04 07:46:19 | [555][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0650 ntime: 0072 mem: 3.36 + 04-04 07:46:25 | [555][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0430 ntime: 0085 mem: 3.36 + 04-04 07:46:31 | [555][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0366 ntime: 0084 mem: 3.36 + 04-04 07:46:38 | [555][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1121 ntime: 0082 mem: 3.36 + 04-04 07:46:44 | [555][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0802 ntime: 0079 mem: 3.36 + 04-04 07:46:52 | [555][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0078 mem: 3.36 + 04-04 07:46:58 | [555][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0244 ntime: 0076 mem: 3.36 + 04-04 07:47:05 | [555][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1178 ntime: 0078 mem: 3.36 + 04-04 07:47:13 | [555][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0212 ntime: 0085 mem: 3.36 + 04-04 07:47:20 | [555][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0554 ntime: 0085 mem: 3.36 + 04-04 07:47:28 | [555][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0685 ntime: 0082 mem: 3.36 + 04-04 07:47:34 | Time info >>>> elapsed: 624.19 mins remain: 498.46 mins + 04-04 07:47:35 | [556][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0700 ntime: 0076 mem: 3.36 + 04-04 07:47:41 | [556][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0379 ntime: 0084 mem: 3.36 + 04-04 07:47:48 | [556][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1051 ntime: 0081 mem: 3.36 + 04-04 07:47:55 | [556][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0089 ntime: 0083 mem: 3.36 + 04-04 07:48:01 | [556][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1061 ntime: 0083 mem: 3.36 + 04-04 07:48:08 | [556][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0076 mem: 3.36 + 04-04 07:48:14 | [556][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0823 ntime: 0085 mem: 3.36 + 04-04 07:48:19 | [556][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0788 ntime: 0078 mem: 3.36 + 04-04 07:48:27 | [556][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0494 ntime: 0087 mem: 3.36 + 04-04 07:48:34 | [556][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0046 ntime: 0085 mem: 3.36 + 04-04 07:48:39 | [556][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0575 ntime: 0075 mem: 3.36 + 04-04 07:48:49 | [556][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0973 ntime: 0078 mem: 3.36 + 04-04 07:48:57 | [556][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0883 ntime: 0084 mem: 3.36 + 04-04 07:49:06 | [556][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0727 ntime: 0078 mem: 3.36 + 04-04 07:49:14 | [556][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0726 ntime: 0079 mem: 3.36 + 04-04 07:49:23 | [556][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0859 ntime: 0085 mem: 3.36 + 04-04 07:49:29 | [556][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0196 ntime: 0078 mem: 3.36 + 04-04 07:49:38 | [556][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0992 ntime: 0076 mem: 3.36 + 04-04 07:49:42 | Time info >>>> elapsed: 626.33 mins remain: 498.14 mins + 04-04 07:49:44 | [557][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1220 ntime: 0076 mem: 3.36 + 04-04 07:49:50 | [557][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1179 ntime: 0087 mem: 3.36 + 04-04 07:49:59 | [557][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 07:50:05 | [557][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0076 mem: 3.36 + 04-04 07:50:12 | [557][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1121 ntime: 0080 mem: 3.36 + 04-04 07:50:19 | [557][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0363 ntime: 0083 mem: 3.36 + 04-04 07:50:25 | [557][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1156 ntime: 0086 mem: 3.36 + 04-04 07:50:32 | [557][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0915 ntime: 0082 mem: 3.36 + 04-04 07:50:38 | [557][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1325 ntime: 0082 mem: 3.36 + 04-04 07:50:45 | [557][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0600 ntime: 0082 mem: 3.36 + 04-04 07:50:51 | [557][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0854 ntime: 0082 mem: 3.36 + 04-04 07:50:58 | [557][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1321 ntime: 0079 mem: 3.36 + 04-04 07:51:06 | [557][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1398 ntime: 0076 mem: 3.36 + 04-04 07:51:16 | [557][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1454 ntime: 0087 mem: 3.36 + 04-04 07:51:24 | [557][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0598 ntime: 0079 mem: 3.36 + 04-04 07:51:33 | [557][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1126 ntime: 0079 mem: 3.36 + 04-04 07:51:42 | [557][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0815 ntime: 0075 mem: 3.36 + 04-04 07:51:50 | [557][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1044 ntime: 0079 mem: 3.36 + 04-04 07:51:55 | Time info >>>> elapsed: 628.55 mins remain: 497.89 mins + 04-04 07:51:57 | [558][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1280 ntime: 0082 mem: 3.36 + 04-04 07:52:03 | [558][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0076 mem: 3.36 + 04-04 07:52:11 | [558][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0085 mem: 3.36 + 04-04 07:52:20 | [558][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0962 ntime: 0085 mem: 3.36 + 04-04 07:52:27 | [558][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1398 ntime: 0082 mem: 3.36 + 04-04 07:52:33 | [558][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0099 ntime: 0081 mem: 3.36 + 04-04 07:52:40 | [558][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0080 mem: 3.36 + 04-04 07:52:47 | [558][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0566 ntime: 0078 mem: 3.36 + 04-04 07:52:53 | [558][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1300 ntime: 0078 mem: 3.36 + 04-04 07:53:00 | [558][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0599 ntime: 0083 mem: 3.36 + 04-04 07:53:08 | [558][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1018 ntime: 0089 mem: 3.36 + 04-04 07:53:16 | [558][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0527 ntime: 0079 mem: 3.36 + 04-04 07:53:22 | [558][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0197 ntime: 0082 mem: 3.36 + 04-04 07:53:31 | [558][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0075 mem: 3.36 + 04-04 07:53:40 | [558][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1097 ntime: 0080 mem: 3.36 + 04-04 07:53:47 | [558][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1357 ntime: 0070 mem: 3.36 + 04-04 07:53:54 | [558][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0850 ntime: 0085 mem: 3.36 + 04-04 07:54:01 | [558][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0627 ntime: 0084 mem: 3.36 + 04-04 07:54:07 | Time info >>>> elapsed: 630.74 mins remain: 497.60 mins + 04-04 07:54:07 | [559][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0085 ntime: 0083 mem: 3.36 + 04-04 07:54:13 | [559][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0634 ntime: 0086 mem: 3.36 + 04-04 07:54:17 | [559][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0054 mem: 3.36 + 04-04 07:54:24 | [559][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0559 ntime: 0079 mem: 3.36 + 04-04 07:54:29 | [559][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0877 ntime: 0081 mem: 3.36 + 04-04 07:54:34 | [559][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0153 ntime: 0081 mem: 3.36 + 04-04 07:54:42 | [559][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1279 ntime: 0079 mem: 3.36 + 04-04 07:54:50 | [559][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1046 ntime: 0085 mem: 3.36 + 04-04 07:54:56 | [559][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0556 ntime: 0080 mem: 3.36 + 04-04 07:55:03 | [559][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 07:55:12 | [559][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1351 ntime: 0076 mem: 3.36 + 04-04 07:55:19 | [559][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0064 ntime: 0077 mem: 3.36 + 04-04 07:55:26 | [559][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1177 ntime: 0080 mem: 3.36 + 04-04 07:55:31 | [559][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0073 ntime: 0078 mem: 3.36 + 04-04 07:55:38 | [559][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1168 ntime: 0079 mem: 3.36 + 04-04 07:55:42 | [559][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0920 ntime: 0082 mem: 3.36 + 04-04 07:55:50 | [559][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1059 ntime: 0085 mem: 3.36 + 04-04 07:55:59 | [559][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1192 ntime: 0089 mem: 3.36 + 04-04 07:56:04 | Time info >>>> elapsed: 632.69 mins remain: 497.11 mins + 04-04 07:56:04 | [560][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0090 ntime: 0081 mem: 3.36 + 04-04 07:56:10 | [560][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0734 ntime: 0080 mem: 3.36 + 04-04 07:56:17 | [560][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0770 ntime: 0081 mem: 3.36 + 04-04 07:56:24 | [560][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1167 ntime: 0075 mem: 3.36 + 04-04 07:56:31 | [560][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1149 ntime: 0080 mem: 3.36 + 04-04 07:56:39 | [560][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1180 ntime: 0087 mem: 3.36 + 04-04 07:56:45 | [560][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1033 ntime: 0079 mem: 3.36 + 04-04 07:56:52 | [560][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0064 ntime: 0075 mem: 3.36 + 04-04 07:57:00 | [560][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0872 ntime: 0088 mem: 3.36 + 04-04 07:57:10 | [560][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0846 ntime: 0087 mem: 3.36 + 04-04 07:57:18 | [560][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1161 ntime: 0087 mem: 3.36 + 04-04 07:57:23 | [560][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0080 ntime: 0075 mem: 3.36 + 04-04 07:57:30 | [560][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1069 ntime: 0074 mem: 3.36 + 04-04 07:57:37 | [560][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1043 ntime: 0078 mem: 3.36 + 04-04 07:57:44 | [560][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0627 ntime: 0082 mem: 3.36 + 04-04 07:57:49 | [560][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0073 ntime: 0075 mem: 3.36 + 04-04 07:57:56 | [560][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1090 ntime: 0087 mem: 3.36 + 04-04 07:58:06 | [560][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0080 mem: 3.36 + 04-04 07:58:12 | Time info >>>> elapsed: 634.83 mins remain: 496.77 mins + 04-04 07:58:12 | [561][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0046 ntime: 0083 mem: 3.36 + 04-04 07:58:20 | [561][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0942 ntime: 0087 mem: 3.36 + 04-04 07:58:28 | [561][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1243 ntime: 0083 mem: 3.36 + 04-04 07:58:36 | [561][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0620 ntime: 0080 mem: 3.36 + 04-04 07:58:44 | [561][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1026 ntime: 0084 mem: 3.36 + 04-04 07:58:53 | [561][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1437 ntime: 0077 mem: 3.36 + 04-04 07:58:59 | [561][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0397 ntime: 0089 mem: 3.36 + 04-04 07:59:07 | [561][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1298 ntime: 0081 mem: 3.36 + 04-04 07:59:15 | [561][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1315 ntime: 0081 mem: 3.36 + 04-04 07:59:21 | [561][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0159 ntime: 0076 mem: 3.36 + 04-04 07:59:29 | [561][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1228 ntime: 0072 mem: 3.36 + 04-04 07:59:36 | [561][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1154 ntime: 0081 mem: 3.36 + 04-04 07:59:50 | [561][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1736 ntime: 0079 mem: 3.36 + 04-04 08:00:00 | [561][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1328 ntime: 0079 mem: 3.36 + 04-04 08:00:11 | [561][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1194 ntime: 0078 mem: 3.36 + 04-04 08:00:19 | [561][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0417 ntime: 0080 mem: 3.36 + 04-04 08:00:28 | [561][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0267 ntime: 0092 mem: 3.36 + 04-04 08:00:37 | [561][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0843 ntime: 0090 mem: 3.36 + 04-04 08:00:45 | Time info >>>> elapsed: 637.38 mins remain: 496.75 mins + 04-04 08:00:47 | [562][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1430 ntime: 0076 mem: 3.36 + 04-04 08:00:57 | [562][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0216 ntime: 0073 mem: 3.36 + 04-04 08:01:08 | [562][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 2368 ntime: 0076 mem: 3.36 + 04-04 08:01:20 | [562][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1796 ntime: 0082 mem: 3.36 + 04-04 08:01:31 | [562][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1395 ntime: 0082 mem: 3.36 + 04-04 08:01:39 | [562][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0849 ntime: 0080 mem: 3.36 + 04-04 08:01:52 | [562][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1476 ntime: 0080 mem: 3.36 + 04-04 08:02:01 | [562][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0961 ntime: 0080 mem: 3.36 + 04-04 08:02:09 | [562][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0494 ntime: 0079 mem: 3.36 + 04-04 08:02:16 | [562][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0551 ntime: 0077 mem: 3.36 + 04-04 08:02:23 | [562][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1400 ntime: 0083 mem: 3.36 + 04-04 08:02:31 | [562][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1460 ntime: 0077 mem: 3.36 + 04-04 08:02:39 | [562][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0668 ntime: 0076 mem: 3.36 + 04-04 08:02:45 | [562][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0085 mem: 3.36 + 04-04 08:02:54 | [562][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0082 mem: 3.36 + 04-04 08:03:00 | [562][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0130 ntime: 0080 mem: 3.36 + 04-04 08:03:12 | [562][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0784 ntime: 0079 mem: 3.36 + 04-04 08:03:22 | [562][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0990 ntime: 0084 mem: 3.36 + 04-04 08:03:32 | Time info >>>> elapsed: 640.16 mins remain: 496.89 mins + 04-04 08:03:32 | [563][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 08:03:41 | [563][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1271 ntime: 0075 mem: 3.36 + 04-04 08:03:49 | [563][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0118 ntime: 0081 mem: 3.36 + 04-04 08:03:57 | [563][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1290 ntime: 0077 mem: 3.36 + 04-04 08:04:05 | [563][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1527 ntime: 0080 mem: 3.36 + 04-04 08:04:16 | [563][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1270 ntime: 0084 mem: 3.36 + 04-04 08:04:27 | [563][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1438 ntime: 0076 mem: 3.36 + 04-04 08:04:35 | [563][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0931 ntime: 0085 mem: 3.36 + 04-04 08:04:44 | [563][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0075 mem: 3.36 + 04-04 08:04:52 | [563][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0526 ntime: 0077 mem: 3.36 + 04-04 08:05:00 | [563][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0078 ntime: 0082 mem: 3.36 + 04-04 08:05:07 | [563][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0781 ntime: 0079 mem: 3.36 + 04-04 08:05:16 | [563][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0778 ntime: 0085 mem: 3.36 + 04-04 08:05:23 | [563][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0522 ntime: 0084 mem: 3.36 + 04-04 08:05:30 | [563][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1334 ntime: 0088 mem: 3.36 + 04-04 08:05:35 | [563][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0102 ntime: 0082 mem: 3.36 + 04-04 08:05:43 | [563][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0084 mem: 3.36 + 04-04 08:05:50 | [563][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0718 ntime: 0081 mem: 3.36 + 04-04 08:05:56 | Time info >>>> elapsed: 642.56 mins remain: 496.73 mins + 04-04 08:05:56 | [564][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0066 ntime: 0078 mem: 3.36 + 04-04 08:06:02 | [564][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0350 ntime: 0087 mem: 3.36 + 04-04 08:06:10 | [564][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0229 ntime: 0078 mem: 3.36 + 04-04 08:06:16 | [564][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0046 ntime: 0060 mem: 3.36 + 04-04 08:06:24 | [564][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0966 ntime: 0083 mem: 3.36 + 04-04 08:06:31 | [564][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0081 mem: 3.36 + 04-04 08:06:39 | [564][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0892 ntime: 0082 mem: 3.36 + 04-04 08:06:46 | [564][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0136 ntime: 0089 mem: 3.36 + 04-04 08:06:53 | [564][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1009 ntime: 0080 mem: 3.36 + 04-04 08:07:00 | [564][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1182 ntime: 0080 mem: 3.36 + 04-04 08:07:06 | [564][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1249 ntime: 0076 mem: 3.36 + 04-04 08:07:12 | [564][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0189 ntime: 0076 mem: 3.36 + 04-04 08:07:20 | [564][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0867 ntime: 0081 mem: 3.36 + 04-04 08:07:26 | [564][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1076 ntime: 0089 mem: 3.36 + 04-04 08:07:31 | [564][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0779 ntime: 0081 mem: 3.36 + 04-04 08:07:38 | [564][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0804 ntime: 0074 mem: 3.36 + 04-04 08:07:45 | [564][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0301 ntime: 0081 mem: 3.36 + 04-04 08:07:51 | [564][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0102 ntime: 0076 mem: 3.36 + 04-04 08:07:56 | Time info >>>> elapsed: 644.56 mins remain: 496.26 mins + 04-04 08:07:57 | [565][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1045 ntime: 0079 mem: 3.36 + 04-04 08:08:04 | [565][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0574 ntime: 0082 mem: 3.36 + 04-04 08:08:12 | [565][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0086 ntime: 0075 mem: 3.36 + 04-04 08:08:19 | [565][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0940 ntime: 0081 mem: 3.36 + 04-04 08:08:26 | [565][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0103 ntime: 0082 mem: 3.36 + 04-04 08:08:35 | [565][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1314 ntime: 0080 mem: 3.36 + 04-04 08:08:44 | [565][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1265 ntime: 0075 mem: 3.36 + 04-04 08:08:52 | [565][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0073 mem: 3.36 + 04-04 08:08:59 | [565][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0062 ntime: 0078 mem: 3.36 + 04-04 08:09:08 | [565][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0597 ntime: 0088 mem: 3.36 + 04-04 08:09:14 | [565][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0658 ntime: 0057 mem: 3.36 + 04-04 08:09:20 | [565][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 08:09:28 | [565][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1036 ntime: 0087 mem: 3.36 + 04-04 08:09:36 | [565][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0516 ntime: 0084 mem: 3.36 + 04-04 08:09:43 | [565][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1058 ntime: 0081 mem: 3.36 + 04-04 08:09:52 | [565][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1649 ntime: 0073 mem: 3.36 + 04-04 08:10:06 | [565][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1384 ntime: 0074 mem: 3.36 + 04-04 08:10:17 | [565][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0890 ntime: 0083 mem: 3.36 + 04-04 08:10:23 | Time info >>>> elapsed: 647.01 mins remain: 496.12 mins + 04-04 08:10:23 | [566][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0080 mem: 3.36 + 04-04 08:10:32 | [566][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0072 ntime: 0081 mem: 3.36 + 04-04 08:10:39 | [566][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0481 ntime: 0079 mem: 3.36 + 04-04 08:10:47 | [566][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0059 ntime: 0075 mem: 3.36 + 04-04 08:10:55 | [566][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1283 ntime: 0079 mem: 3.36 + 04-04 08:11:04 | [566][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1159 ntime: 0081 mem: 3.36 + 04-04 08:11:12 | [566][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0077 ntime: 0080 mem: 3.36 + 04-04 08:11:23 | [566][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1246 ntime: 0078 mem: 3.36 + 04-04 08:11:32 | [566][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0059 ntime: 0079 mem: 3.36 + 04-04 08:11:43 | [566][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1037 ntime: 0075 mem: 3.36 + 04-04 08:11:52 | [566][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0215 ntime: 0077 mem: 3.36 + 04-04 08:12:01 | [566][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0129 ntime: 0088 mem: 3.36 + 04-04 08:12:11 | [566][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 2169 ntime: 0076 mem: 3.36 + 04-04 08:12:20 | [566][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1522 ntime: 0055 mem: 3.36 + 04-04 08:12:27 | [566][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1467 ntime: 0079 mem: 3.36 + 04-04 08:12:35 | [566][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0132 ntime: 0076 mem: 3.36 + 04-04 08:12:43 | [566][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1080 ntime: 0081 mem: 3.36 + 04-04 08:12:51 | [566][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0756 ntime: 0080 mem: 3.36 + 04-04 08:12:58 | Time info >>>> elapsed: 649.59 mins remain: 496.07 mins + 04-04 08:12:58 | [567][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0053 ntime: 0074 mem: 3.36 + 04-04 08:13:05 | [567][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0624 ntime: 0082 mem: 3.36 + 04-04 08:13:13 | [567][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0257 ntime: 0082 mem: 3.36 + 04-04 08:13:21 | [567][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1109 ntime: 0085 mem: 3.36 + 04-04 08:13:31 | [567][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 08:13:40 | [567][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0982 ntime: 0088 mem: 3.36 + 04-04 08:13:46 | [567][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0075 mem: 3.36 + 04-04 08:13:53 | [567][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0815 ntime: 0084 mem: 3.36 + 04-04 08:13:57 | [567][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0627 ntime: 0081 mem: 3.36 + 04-04 08:14:06 | [567][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1249 ntime: 0079 mem: 3.36 + 04-04 08:14:13 | [567][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0769 ntime: 0079 mem: 3.36 + 04-04 08:14:20 | [567][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0619 ntime: 0080 mem: 3.36 + 04-04 08:14:28 | [567][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0222 ntime: 0080 mem: 3.36 + 04-04 08:14:33 | [567][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1374 ntime: 0075 mem: 3.36 + 04-04 08:14:40 | [567][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0673 ntime: 0076 mem: 3.36 + 04-04 08:14:48 | [567][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0046 ntime: 0078 mem: 3.36 + 04-04 08:14:56 | [567][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0908 ntime: 0077 mem: 3.36 + 04-04 08:15:00 | [567][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 08:15:05 | Time info >>>> elapsed: 651.71 mins remain: 495.67 mins + 04-04 08:15:06 | [568][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1140 ntime: 0085 mem: 3.36 + 04-04 08:15:10 | [568][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0952 ntime: 0084 mem: 3.36 + 04-04 08:15:16 | [568][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0417 ntime: 0058 mem: 3.36 + 04-04 08:15:21 | [568][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0436 ntime: 0079 mem: 3.36 + 04-04 08:15:28 | [568][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1282 ntime: 0081 mem: 3.36 + 04-04 08:15:34 | [568][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1122 ntime: 0080 mem: 3.36 + 04-04 08:15:38 | [568][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0074 mem: 3.36 + 04-04 08:15:45 | [568][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0080 ntime: 0077 mem: 3.36 + 04-04 08:15:52 | [568][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0124 ntime: 0078 mem: 3.36 + 04-04 08:16:00 | [568][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0105 ntime: 0079 mem: 3.36 + 04-04 08:16:09 | [568][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0071 ntime: 0073 mem: 3.36 + 04-04 08:16:18 | [568][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0688 ntime: 0088 mem: 3.36 + 04-04 08:16:25 | [568][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0086 ntime: 0087 mem: 3.36 + 04-04 08:16:34 | [568][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0090 ntime: 0080 mem: 3.36 + 04-04 08:16:41 | [568][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1246 ntime: 0081 mem: 3.36 + 04-04 08:16:48 | [568][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1034 ntime: 0072 mem: 3.36 + 04-04 08:16:55 | [568][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0961 ntime: 0079 mem: 3.36 + 04-04 08:17:01 | [568][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0568 ntime: 0090 mem: 3.36 + 04-04 08:17:08 | Time info >>>> elapsed: 653.76 mins remain: 495.21 mins + 04-04 08:17:08 | [569][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0135 ntime: 0078 mem: 3.36 + 04-04 08:17:17 | [569][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1431 ntime: 0087 mem: 3.36 + 04-04 08:17:24 | [569][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0530 ntime: 0085 mem: 3.36 + 04-04 08:17:32 | [569][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0793 ntime: 0084 mem: 3.36 + 04-04 08:17:41 | [569][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0973 ntime: 0079 mem: 3.36 + 04-04 08:17:47 | [569][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0834 ntime: 0079 mem: 3.36 + 04-04 08:17:53 | [569][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0451 ntime: 0082 mem: 3.36 + 04-04 08:18:02 | [569][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1106 ntime: 0080 mem: 3.36 + 04-04 08:18:09 | [569][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0142 ntime: 0065 mem: 3.36 + 04-04 08:18:16 | [569][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0212 ntime: 0073 mem: 3.36 + 04-04 08:18:22 | [569][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0763 ntime: 0083 mem: 3.36 + 04-04 08:18:31 | [569][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0845 ntime: 0081 mem: 3.36 + 04-04 08:18:36 | [569][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1250 ntime: 0079 mem: 3.36 + 04-04 08:18:41 | [569][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0333 ntime: 0081 mem: 3.36 + 04-04 08:18:49 | [569][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0818 ntime: 0087 mem: 3.36 + 04-04 08:18:55 | [569][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0596 ntime: 0080 mem: 3.36 + 04-04 08:19:01 | [569][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0372 ntime: 0075 mem: 3.36 + 04-04 08:19:08 | [569][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0612 ntime: 0087 mem: 3.36 + 04-04 08:19:14 | Time info >>>> elapsed: 655.85 mins remain: 494.77 mins + 04-04 08:19:14 | [570][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0049 ntime: 0090 mem: 3.36 + 04-04 08:19:20 | [570][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0333 ntime: 0080 mem: 3.36 + 04-04 08:19:28 | [570][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0511 ntime: 0089 mem: 3.36 + 04-04 08:19:37 | [570][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0696 ntime: 0086 mem: 3.36 + 04-04 08:19:43 | [570][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0380 ntime: 0079 mem: 3.36 + 04-04 08:19:49 | [570][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1016 ntime: 0081 mem: 3.36 + 04-04 08:19:58 | [570][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0440 ntime: 0080 mem: 3.36 + 04-04 08:20:06 | [570][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1342 ntime: 0078 mem: 3.36 + 04-04 08:20:16 | [570][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0989 ntime: 0080 mem: 3.36 + 04-04 08:20:23 | [570][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0497 ntime: 0075 mem: 3.36 + 04-04 08:20:30 | [570][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1279 ntime: 0077 mem: 3.36 + 04-04 08:20:38 | [570][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1280 ntime: 0077 mem: 3.36 + 04-04 08:20:45 | [570][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0114 ntime: 0076 mem: 3.36 + 04-04 08:20:53 | [570][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0100 ntime: 0080 mem: 3.36 + 04-04 08:21:00 | [570][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0529 ntime: 0084 mem: 3.36 + 04-04 08:21:07 | [570][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0088 ntime: 0074 mem: 3.36 + 04-04 08:21:14 | [570][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0076 mem: 3.36 + 04-04 08:21:22 | [570][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0547 ntime: 0082 mem: 3.36 + 04-04 08:21:28 | Time info >>>> elapsed: 658.09 mins remain: 494.43 mins + 04-04 08:21:28 | [571][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0598 ntime: 0075 mem: 3.36 + 04-04 08:21:35 | [571][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0109 ntime: 0081 mem: 3.36 + 04-04 08:21:42 | [571][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1314 ntime: 0082 mem: 3.36 + 04-04 08:21:49 | [571][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0081 ntime: 0084 mem: 3.36 + 04-04 08:21:54 | [571][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0643 ntime: 0088 mem: 3.36 + 04-04 08:22:02 | [571][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0943 ntime: 0083 mem: 3.36 + 04-04 08:22:10 | [571][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0075 ntime: 0080 mem: 3.36 + 04-04 08:22:17 | [571][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1269 ntime: 0075 mem: 3.36 + 04-04 08:22:24 | [571][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1268 ntime: 0079 mem: 3.36 + 04-04 08:22:30 | [571][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1158 ntime: 0084 mem: 3.36 + 04-04 08:22:37 | [571][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1026 ntime: 0079 mem: 3.36 + 04-04 08:22:45 | [571][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1194 ntime: 0078 mem: 3.36 + 04-04 08:22:50 | [571][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0796 ntime: 0084 mem: 3.36 + 04-04 08:22:58 | [571][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0852 ntime: 0088 mem: 3.36 + 04-04 08:23:06 | [571][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0520 ntime: 0079 mem: 3.36 + 04-04 08:23:13 | [571][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0082 ntime: 0078 mem: 3.36 + 04-04 08:23:19 | [571][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0077 mem: 3.36 + 04-04 08:23:28 | [571][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1119 ntime: 0084 mem: 3.36 + 04-04 08:23:32 | Time info >>>> elapsed: 660.17 mins remain: 493.97 mins + 04-04 08:23:33 | [572][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1028 ntime: 0078 mem: 3.36 + 04-04 08:23:41 | [572][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0101 ntime: 0083 mem: 3.36 + 04-04 08:23:49 | [572][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1164 ntime: 0087 mem: 3.36 + 04-04 08:23:56 | [572][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0128 ntime: 0075 mem: 3.36 + 04-04 08:24:04 | [572][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0081 ntime: 0060 mem: 3.36 + 04-04 08:24:10 | [572][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0677 ntime: 0074 mem: 3.36 + 04-04 08:24:16 | [572][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0567 ntime: 0075 mem: 3.36 + 04-04 08:24:21 | [572][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0066 ntime: 0078 mem: 3.36 + 04-04 08:24:27 | [572][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0703 ntime: 0088 mem: 3.36 + 04-04 08:24:37 | [572][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0105 ntime: 0079 mem: 3.36 + 04-04 08:24:49 | [572][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1554 ntime: 0082 mem: 3.36 + 04-04 08:24:59 | [572][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1081 ntime: 0079 mem: 3.36 + 04-04 08:25:06 | [572][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0766 ntime: 0095 mem: 3.36 + 04-04 08:25:15 | [572][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0069 ntime: 0077 mem: 3.36 + 04-04 08:25:25 | [572][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1151 ntime: 0076 mem: 3.36 + 04-04 08:25:32 | [572][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0081 mem: 3.36 + 04-04 08:25:39 | [572][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0959 ntime: 0081 mem: 3.36 + 04-04 08:25:48 | [572][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1372 ntime: 0078 mem: 3.36 + 04-04 08:25:55 | Time info >>>> elapsed: 662.54 mins remain: 493.72 mins + 04-04 08:25:56 | [573][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1270 ntime: 0078 mem: 3.36 + 04-04 08:26:05 | [573][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1406 ntime: 0084 mem: 3.36 + 04-04 08:26:11 | [573][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0160 ntime: 0097 mem: 3.36 + 04-04 08:26:19 | [573][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1470 ntime: 0082 mem: 3.36 + 04-04 08:26:28 | [573][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1461 ntime: 0084 mem: 3.36 + 04-04 08:26:37 | [573][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1315 ntime: 0078 mem: 3.36 + 04-04 08:26:45 | [573][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0981 ntime: 0087 mem: 3.36 + 04-04 08:26:52 | [573][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1128 ntime: 0082 mem: 3.36 + 04-04 08:27:01 | [573][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1259 ntime: 0072 mem: 3.36 + 04-04 08:27:13 | [573][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1385 ntime: 0075 mem: 3.36 + 04-04 08:27:21 | [573][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0983 ntime: 0082 mem: 3.36 + 04-04 08:27:30 | [573][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0079 mem: 3.36 + 04-04 08:27:39 | [573][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1146 ntime: 0075 mem: 3.36 + 04-04 08:27:48 | [573][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0186 ntime: 0076 mem: 3.36 + 04-04 08:27:58 | [573][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1333 ntime: 0080 mem: 3.36 + 04-04 08:28:07 | [573][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1032 ntime: 0087 mem: 3.36 + 04-04 08:28:14 | [573][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1045 ntime: 0077 mem: 3.36 + 04-04 08:28:21 | [573][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1069 ntime: 0079 mem: 3.36 + 04-04 08:28:27 | Time info >>>> elapsed: 665.09 mins remain: 493.60 mins + 04-04 08:28:28 | [574][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0077 mem: 3.36 + 04-04 08:28:36 | [574][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1095 ntime: 0077 mem: 3.36 + 04-04 08:28:45 | [574][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0736 ntime: 0086 mem: 3.36 + 04-04 08:28:53 | [574][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1117 ntime: 0083 mem: 3.36 + 04-04 08:29:03 | [574][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1373 ntime: 0075 mem: 3.36 + 04-04 08:29:11 | [574][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1325 ntime: 0077 mem: 3.36 + 04-04 08:29:18 | [574][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0090 ntime: 0081 mem: 3.36 + 04-04 08:29:25 | [574][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0070 ntime: 0084 mem: 3.36 + 04-04 08:29:33 | [574][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1045 ntime: 0084 mem: 3.36 + 04-04 08:29:43 | [574][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1259 ntime: 0075 mem: 3.36 + 04-04 08:29:53 | [574][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1114 ntime: 0080 mem: 3.36 + 04-04 08:30:05 | [574][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1373 ntime: 0060 mem: 3.36 + 04-04 08:30:12 | [574][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1181 ntime: 0074 mem: 3.36 + 04-04 08:30:22 | [574][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0945 ntime: 0077 mem: 3.36 + 04-04 08:30:29 | [574][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1361 ntime: 0077 mem: 3.36 + 04-04 08:30:36 | [574][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0059 ntime: 0082 mem: 3.36 + 04-04 08:30:45 | [574][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0973 ntime: 0082 mem: 3.36 + 04-04 08:30:53 | [574][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1474 ntime: 0077 mem: 3.36 + 04-04 08:30:58 | Time info >>>> elapsed: 667.60 mins remain: 493.45 mins + 04-04 08:30:59 | [575][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0117 ntime: 0075 mem: 3.36 + 04-04 08:31:06 | [575][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0934 ntime: 0076 mem: 3.36 + 04-04 08:31:14 | [575][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0123 ntime: 0077 mem: 3.36 + 04-04 08:31:22 | [575][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1522 ntime: 0085 mem: 3.36 + 04-04 08:31:32 | [575][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0120 ntime: 0082 mem: 3.36 + 04-04 08:31:43 | [575][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0125 ntime: 0077 mem: 3.36 + 04-04 08:31:52 | [575][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1250 ntime: 0079 mem: 3.36 + 04-04 08:31:58 | [575][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1037 ntime: 0079 mem: 3.36 + 04-04 08:32:04 | [575][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1164 ntime: 0080 mem: 3.36 + 04-04 08:32:12 | [575][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0079 mem: 3.36 + 04-04 08:32:22 | [575][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1288 ntime: 0084 mem: 3.36 + 04-04 08:32:27 | [575][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0084 mem: 3.36 + 04-04 08:32:35 | [575][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0889 ntime: 0081 mem: 3.36 + 04-04 08:32:43 | [575][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0729 ntime: 0087 mem: 3.36 + 04-04 08:32:49 | [575][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0144 ntime: 0082 mem: 3.36 + 04-04 08:32:57 | [575][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0743 ntime: 0084 mem: 3.36 + 04-04 08:33:07 | [575][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1332 ntime: 0076 mem: 3.36 + 04-04 08:33:16 | [575][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0986 ntime: 0061 mem: 3.36 + 04-04 08:33:23 | Time info >>>> elapsed: 670.01 mins remain: 493.20 mins + 04-04 08:33:23 | [576][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0054 ntime: 0087 mem: 3.36 + 04-04 08:33:30 | [576][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0753 ntime: 0079 mem: 3.36 + 04-04 08:33:37 | [576][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0662 ntime: 0086 mem: 3.36 + 04-04 08:33:46 | [576][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0905 ntime: 0078 mem: 3.36 + 04-04 08:33:56 | [576][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0150 ntime: 0082 mem: 3.36 + 04-04 08:34:06 | [576][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1311 ntime: 0077 mem: 3.36 + 04-04 08:34:14 | [576][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0891 ntime: 0076 mem: 3.36 + 04-04 08:34:20 | [576][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0138 ntime: 0083 mem: 3.36 + 04-04 08:34:29 | [576][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1160 ntime: 0081 mem: 3.36 + 04-04 08:34:38 | [576][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1196 ntime: 0079 mem: 3.36 + 04-04 08:34:46 | [576][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0826 ntime: 0074 mem: 3.36 + 04-04 08:34:56 | [576][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0144 ntime: 0082 mem: 3.36 + 04-04 08:35:05 | [576][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 08:35:13 | [576][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0753 ntime: 0086 mem: 3.36 + 04-04 08:35:22 | [576][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0729 ntime: 0082 mem: 3.36 + 04-04 08:35:29 | [576][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1372 ntime: 0073 mem: 3.36 + 04-04 08:35:38 | [576][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0776 ntime: 0088 mem: 3.36 + 04-04 08:35:45 | [576][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1267 ntime: 0075 mem: 3.36 + 04-04 08:35:52 | Time info >>>> elapsed: 672.49 mins remain: 493.00 mins + 04-04 08:35:53 | [577][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1105 ntime: 0081 mem: 3.36 + 04-04 08:36:02 | [577][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0535 ntime: 0083 mem: 3.36 + 04-04 08:36:09 | [577][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1290 ntime: 0086 mem: 3.36 + 04-04 08:36:17 | [577][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0098 ntime: 0076 mem: 3.36 + 04-04 08:36:25 | [577][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0672 ntime: 0082 mem: 3.36 + 04-04 08:36:33 | [577][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1263 ntime: 0076 mem: 3.36 + 04-04 08:36:42 | [577][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0089 ntime: 0076 mem: 3.36 + 04-04 08:36:49 | [577][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0098 ntime: 0080 mem: 3.36 + 04-04 08:36:58 | [577][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1147 ntime: 0083 mem: 3.36 + 04-04 08:37:04 | [577][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0078 mem: 3.36 + 04-04 08:37:13 | [577][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1052 ntime: 0084 mem: 3.36 + 04-04 08:37:20 | [577][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1035 ntime: 0085 mem: 3.36 + 04-04 08:37:27 | [577][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1109 ntime: 0091 mem: 3.36 + 04-04 08:37:35 | [577][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0076 mem: 3.36 + 04-04 08:37:45 | [577][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1108 ntime: 0088 mem: 3.36 + 04-04 08:37:51 | [577][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0987 ntime: 0080 mem: 3.36 + 04-04 08:37:59 | [577][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0667 ntime: 0076 mem: 3.36 + 04-04 08:38:08 | [577][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1479 ntime: 0079 mem: 3.36 + 04-04 08:38:13 | Time info >>>> elapsed: 674.85 mins remain: 492.71 mins + 04-04 08:38:15 | [578][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1244 ntime: 0069 mem: 3.36 + 04-04 08:38:22 | [578][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0071 ntime: 0084 mem: 3.36 + 04-04 08:38:31 | [578][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0064 ntime: 0087 mem: 3.36 + 04-04 08:38:39 | [578][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1144 ntime: 0079 mem: 3.36 + 04-04 08:38:48 | [578][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1349 ntime: 0083 mem: 3.36 + 04-04 08:38:57 | [578][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0082 ntime: 0091 mem: 3.36 + 04-04 08:39:05 | [578][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0094 ntime: 0079 mem: 3.36 + 04-04 08:39:14 | [578][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1180 ntime: 0080 mem: 3.36 + 04-04 08:39:22 | [578][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0165 ntime: 0082 mem: 3.36 + 04-04 08:39:30 | [578][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0070 ntime: 0084 mem: 3.36 + 04-04 08:39:40 | [578][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0089 mem: 3.36 + 04-04 08:39:49 | [578][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1308 ntime: 0086 mem: 3.36 + 04-04 08:39:58 | [578][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1353 ntime: 0083 mem: 3.36 + 04-04 08:40:06 | [578][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0935 ntime: 0074 mem: 3.36 + 04-04 08:40:14 | [578][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0680 ntime: 0079 mem: 3.36 + 04-04 08:40:23 | [578][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1466 ntime: 0088 mem: 3.36 + 04-04 08:40:33 | [578][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1254 ntime: 0080 mem: 3.36 + 04-04 08:40:40 | [578][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1111 ntime: 0076 mem: 3.36 + 04-04 08:40:44 | Time info >>>> elapsed: 677.36 mins remain: 492.52 mins + 04-04 08:40:45 | [579][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1016 ntime: 0082 mem: 3.36 + 04-04 08:40:53 | [579][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0081 mem: 3.36 + 04-04 08:41:05 | [579][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1521 ntime: 0088 mem: 3.36 + 04-04 08:41:14 | [579][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0066 ntime: 0072 mem: 3.36 + 04-04 08:41:26 | [579][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1230 ntime: 0075 mem: 3.36 + 04-04 08:41:38 | [579][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1522 ntime: 0084 mem: 3.36 + 04-04 08:41:47 | [579][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0958 ntime: 0087 mem: 3.36 + 04-04 08:41:55 | [579][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1051 ntime: 0082 mem: 3.36 + 04-04 08:42:03 | [579][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1343 ntime: 0088 mem: 3.36 + 04-04 08:42:11 | [579][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0777 ntime: 0076 mem: 3.36 + 04-04 08:42:19 | [579][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0817 ntime: 0082 mem: 3.36 + 04-04 08:42:28 | [579][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1369 ntime: 0081 mem: 3.36 + 04-04 08:42:38 | [579][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1005 ntime: 0088 mem: 3.36 + 04-04 08:42:46 | [579][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1050 ntime: 0079 mem: 3.36 + 04-04 08:42:54 | [579][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0804 ntime: 0082 mem: 3.36 + 04-04 08:43:01 | [579][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1225 ntime: 0077 mem: 3.36 + 04-04 08:43:11 | [579][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1009 ntime: 0089 mem: 3.36 + 04-04 08:43:19 | [579][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0078 ntime: 0079 mem: 3.36 + 04-04 08:43:26 | Time info >>>> elapsed: 680.07 mins remain: 492.46 mins + 04-04 08:43:27 | [580][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0978 ntime: 0073 mem: 3.36 + 04-04 08:43:38 | [580][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0148 ntime: 0077 mem: 3.36 + 04-04 08:43:48 | [580][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0507 ntime: 0082 mem: 3.36 + 04-04 08:43:57 | [580][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1357 ntime: 0080 mem: 3.36 + 04-04 08:44:05 | [580][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1186 ntime: 0075 mem: 3.36 + 04-04 08:44:14 | [580][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1042 ntime: 0073 mem: 3.36 + 04-04 08:44:21 | [580][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1197 ntime: 0086 mem: 3.36 + 04-04 08:44:29 | [580][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0461 ntime: 0077 mem: 3.36 + 04-04 08:44:36 | [580][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0083 mem: 3.36 + 04-04 08:44:45 | [580][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0107 ntime: 0084 mem: 3.36 + 04-04 08:44:54 | [580][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1271 ntime: 0086 mem: 3.36 + 04-04 08:45:02 | [580][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0652 ntime: 0078 mem: 3.36 + 04-04 08:45:11 | [580][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1300 ntime: 0083 mem: 3.36 + 04-04 08:45:19 | [580][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1504 ntime: 0079 mem: 3.36 + 04-04 08:45:29 | [580][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1228 ntime: 0084 mem: 3.36 + 04-04 08:45:37 | [580][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1163 ntime: 0078 mem: 3.36 + 04-04 08:45:47 | [580][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1402 ntime: 0074 mem: 3.36 + 04-04 08:45:57 | [580][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0059 ntime: 0082 mem: 3.36 + 04-04 08:46:04 | Time info >>>> elapsed: 682.69 mins remain: 492.34 mins + 04-04 08:46:05 | [581][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1318 ntime: 0082 mem: 3.36 + 04-04 08:46:13 | [581][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0959 ntime: 0082 mem: 3.36 + 04-04 08:46:19 | [581][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0921 ntime: 0081 mem: 3.36 + 04-04 08:46:27 | [581][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 08:46:35 | [581][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0079 mem: 3.36 + 04-04 08:46:44 | [581][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1429 ntime: 0087 mem: 3.36 + 04-04 08:46:51 | [581][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0679 ntime: 0080 mem: 3.36 + 04-04 08:47:01 | [581][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1374 ntime: 0080 mem: 3.36 + 04-04 08:47:13 | [581][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1488 ntime: 0078 mem: 3.36 + 04-04 08:47:23 | [581][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1380 ntime: 0079 mem: 3.36 + 04-04 08:47:33 | [581][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0529 ntime: 0081 mem: 3.36 + 04-04 08:47:45 | [581][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1209 ntime: 0080 mem: 3.36 + 04-04 08:47:54 | [581][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0064 mem: 3.36 + 04-04 08:48:03 | [581][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1439 ntime: 0082 mem: 3.36 + 04-04 08:48:14 | [581][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1177 ntime: 0073 mem: 3.36 + 04-04 08:48:24 | [581][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0198 ntime: 0075 mem: 3.36 + 04-04 08:48:33 | [581][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1516 ntime: 0080 mem: 3.36 + 04-04 08:48:43 | [581][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1373 ntime: 0082 mem: 3.36 + 04-04 08:48:50 | Time info >>>> elapsed: 685.47 mins remain: 492.31 mins + 04-04 08:48:52 | [582][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1297 ntime: 0085 mem: 3.36 + 04-04 08:49:00 | [582][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1501 ntime: 0075 mem: 3.36 + 04-04 08:49:09 | [582][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1381 ntime: 0073 mem: 3.36 + 04-04 08:49:19 | [582][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1117 ntime: 0080 mem: 3.36 + 04-04 08:49:27 | [582][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1399 ntime: 0081 mem: 3.36 + 04-04 08:49:34 | [582][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0065 ntime: 0087 mem: 3.36 + 04-04 08:49:44 | [582][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1262 ntime: 0098 mem: 3.36 + 04-04 08:49:53 | [582][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0166 ntime: 0083 mem: 3.36 + 04-04 08:50:03 | [582][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0899 ntime: 0083 mem: 3.36 + 04-04 08:50:14 | [582][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1376 ntime: 0079 mem: 3.36 + 04-04 08:50:22 | [582][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1202 ntime: 0073 mem: 3.36 + 04-04 08:50:30 | [582][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0089 ntime: 0085 mem: 3.36 + 04-04 08:50:42 | [582][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1024 ntime: 0081 mem: 3.36 + 04-04 08:50:51 | [582][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1286 ntime: 0083 mem: 3.36 + 04-04 08:51:04 | [582][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1475 ntime: 0080 mem: 3.36 + 04-04 08:51:14 | [582][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1399 ntime: 0078 mem: 3.36 + 04-04 08:51:23 | [582][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0412 ntime: 0077 mem: 3.36 + 04-04 08:51:33 | [582][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1323 ntime: 0075 mem: 3.36 + 04-04 08:51:39 | Time info >>>> elapsed: 688.28 mins remain: 492.30 mins + 04-04 08:51:41 | [583][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1211 ntime: 0081 mem: 3.36 + 04-04 08:51:49 | [583][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0073 ntime: 0077 mem: 3.36 + 04-04 08:51:59 | [583][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0579 ntime: 0079 mem: 3.36 + 04-04 08:52:07 | [583][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0070 ntime: 0081 mem: 3.36 + 04-04 08:52:18 | [583][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1229 ntime: 0084 mem: 3.36 + 04-04 08:52:27 | [583][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0047 ntime: 0069 mem: 3.36 + 04-04 08:52:35 | [583][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0924 ntime: 0084 mem: 3.36 + 04-04 08:52:43 | [583][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0080 ntime: 0080 mem: 3.36 + 04-04 08:52:51 | [583][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1404 ntime: 0084 mem: 3.36 + 04-04 08:52:59 | [583][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1310 ntime: 0082 mem: 3.36 + 04-04 08:53:08 | [583][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0101 ntime: 0083 mem: 3.36 + 04-04 08:53:16 | [583][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0134 ntime: 0076 mem: 3.36 + 04-04 08:53:26 | [583][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1289 ntime: 0085 mem: 3.36 + 04-04 08:53:34 | [583][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0120 ntime: 0079 mem: 3.36 + 04-04 08:53:41 | [583][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1113 ntime: 0076 mem: 3.36 + 04-04 08:53:50 | [583][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0831 ntime: 0082 mem: 3.36 + 04-04 08:53:57 | [583][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0058 mem: 3.36 + 04-04 08:54:06 | [583][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1407 ntime: 0080 mem: 3.36 + 04-04 08:54:14 | Time info >>>> elapsed: 690.87 mins remain: 492.12 mins + 04-04 08:54:16 | [584][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1327 ntime: 0075 mem: 3.36 + 04-04 08:54:24 | [584][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1325 ntime: 0080 mem: 3.36 + 04-04 08:54:34 | [584][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1694 ntime: 0088 mem: 3.36 + 04-04 08:54:43 | [584][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1357 ntime: 0084 mem: 3.36 + 04-04 08:54:51 | [584][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0062 ntime: 0079 mem: 3.36 + 04-04 08:54:58 | [584][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0367 ntime: 0078 mem: 3.36 + 04-04 08:55:06 | [584][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0671 ntime: 0076 mem: 3.36 + 04-04 08:55:16 | [584][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1454 ntime: 0084 mem: 3.36 + 04-04 08:55:24 | [584][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0809 ntime: 0078 mem: 3.36 + 04-04 08:55:34 | [584][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1253 ntime: 0083 mem: 3.36 + 04-04 08:55:43 | [584][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0193 ntime: 0076 mem: 3.36 + 04-04 08:55:54 | [584][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1313 ntime: 0078 mem: 3.36 + 04-04 08:56:05 | [584][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1218 ntime: 0084 mem: 3.36 + 04-04 08:56:14 | [584][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1446 ntime: 0083 mem: 3.36 + 04-04 08:56:23 | [584][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1557 ntime: 0081 mem: 3.36 + 04-04 08:56:33 | [584][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0903 ntime: 0081 mem: 3.36 + 04-04 08:56:43 | [584][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1246 ntime: 0081 mem: 3.36 + 04-04 08:56:51 | [584][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1425 ntime: 0089 mem: 3.36 + 04-04 08:56:57 | Time info >>>> elapsed: 693.58 mins remain: 492.03 mins + 04-04 08:56:57 | [585][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0075 mem: 3.36 + 04-04 08:57:06 | [585][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1345 ntime: 0072 mem: 3.36 + 04-04 08:57:14 | [585][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0698 ntime: 0080 mem: 3.36 + 04-04 08:57:21 | [585][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0080 mem: 3.36 + 04-04 08:57:30 | [585][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1331 ntime: 0078 mem: 3.36 + 04-04 08:57:40 | [585][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0105 ntime: 0082 mem: 3.36 + 04-04 08:57:48 | [585][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0080 mem: 3.36 + 04-04 08:57:59 | [585][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1539 ntime: 0081 mem: 3.36 + 04-04 08:58:05 | [585][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0106 ntime: 0079 mem: 3.36 + 04-04 08:58:13 | [585][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0082 mem: 3.36 + 04-04 08:58:20 | [585][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0078 ntime: 0076 mem: 3.36 + 04-04 08:58:28 | [585][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0077 mem: 3.36 + 04-04 08:58:35 | [585][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1299 ntime: 0084 mem: 3.36 + 04-04 08:58:44 | [585][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1152 ntime: 0076 mem: 3.36 + 04-04 08:58:50 | [585][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0245 ntime: 0081 mem: 3.36 + 04-04 08:58:58 | [585][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1281 ntime: 0082 mem: 3.36 + 04-04 08:59:05 | [585][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1302 ntime: 0080 mem: 3.36 + 04-04 08:59:12 | [585][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 08:59:19 | Time info >>>> elapsed: 695.94 mins remain: 491.67 mins + 04-04 08:59:20 | [586][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0900 ntime: 0078 mem: 3.36 + 04-04 08:59:27 | [586][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1292 ntime: 0077 mem: 3.36 + 04-04 08:59:33 | [586][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0057 ntime: 0084 mem: 3.36 + 04-04 08:59:41 | [586][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1530 ntime: 0086 mem: 3.36 + 04-04 08:59:47 | [586][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0072 ntime: 0064 mem: 3.36 + 04-04 08:59:55 | [586][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0078 ntime: 0082 mem: 3.36 + 04-04 09:00:00 | [586][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0081 ntime: 0077 mem: 3.36 + 04-04 09:00:07 | [586][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1133 ntime: 0079 mem: 3.36 + 04-04 09:00:13 | [586][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0077 ntime: 0076 mem: 3.36 + 04-04 09:00:19 | [586][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0752 ntime: 0078 mem: 3.36 + 04-04 09:00:26 | [586][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1113 ntime: 0086 mem: 3.36 + 04-04 09:00:33 | [586][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0114 ntime: 0081 mem: 3.36 + 04-04 09:00:39 | [586][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0386 ntime: 0076 mem: 3.36 + 04-04 09:00:46 | [586][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1320 ntime: 0076 mem: 3.36 + 04-04 09:00:54 | [586][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0673 ntime: 0075 mem: 3.36 + 04-04 09:01:02 | [586][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1039 ntime: 0076 mem: 3.36 + 04-04 09:01:10 | [586][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1128 ntime: 0078 mem: 3.36 + 04-04 09:01:18 | [586][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 09:01:24 | Time info >>>> elapsed: 698.03 mins remain: 491.12 mins + 04-04 09:01:26 | [587][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1332 ntime: 0076 mem: 3.36 + 04-04 09:01:35 | [587][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1553 ntime: 0079 mem: 3.36 + 04-04 09:01:44 | [587][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1443 ntime: 0079 mem: 3.36 + 04-04 09:01:50 | [587][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0062 ntime: 0077 mem: 3.36 + 04-04 09:01:59 | [587][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0970 ntime: 0083 mem: 3.36 + 04-04 09:02:07 | [587][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1114 ntime: 0086 mem: 3.36 + 04-04 09:02:13 | [587][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0078 mem: 3.36 + 04-04 09:02:22 | [587][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0834 ntime: 0089 mem: 3.36 + 04-04 09:02:30 | [587][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1062 ntime: 0078 mem: 3.36 + 04-04 09:02:39 | [587][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1202 ntime: 0069 mem: 3.36 + 04-04 09:02:46 | [587][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0856 ntime: 0082 mem: 3.36 + 04-04 09:02:53 | [587][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 09:03:02 | [587][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1125 ntime: 0082 mem: 3.36 + 04-04 09:03:09 | [587][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0076 mem: 3.36 + 04-04 09:03:16 | [587][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0076 mem: 3.36 + 04-04 09:03:24 | [587][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1275 ntime: 0078 mem: 3.36 + 04-04 09:03:32 | [587][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0943 ntime: 0078 mem: 3.36 + 04-04 09:03:40 | [587][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0835 ntime: 0087 mem: 3.36 + 04-04 09:03:45 | Time info >>>> elapsed: 700.38 mins remain: 490.75 mins + 04-04 09:03:47 | [588][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1416 ntime: 0077 mem: 3.36 + 04-04 09:03:55 | [588][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1105 ntime: 0081 mem: 3.36 + 04-04 09:04:03 | [588][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1091 ntime: 0089 mem: 3.36 + 04-04 09:04:11 | [588][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0109 ntime: 0081 mem: 3.36 + 04-04 09:04:19 | [588][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0282 ntime: 0083 mem: 3.36 + 04-04 09:04:26 | [588][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 09:04:34 | [588][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0982 ntime: 0083 mem: 3.36 + 04-04 09:04:43 | [588][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0995 ntime: 0082 mem: 3.36 + 04-04 09:04:49 | [588][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0862 ntime: 0080 mem: 3.36 + 04-04 09:04:55 | [588][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 09:05:03 | [588][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0058 ntime: 0087 mem: 3.36 + 04-04 09:05:12 | [588][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0732 ntime: 0081 mem: 3.36 + 04-04 09:05:19 | [588][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1315 ntime: 0078 mem: 3.36 + 04-04 09:05:28 | [588][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1487 ntime: 0078 mem: 3.36 + 04-04 09:05:34 | [588][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0047 ntime: 0059 mem: 3.36 + 04-04 09:05:41 | [588][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0101 ntime: 0083 mem: 3.36 + 04-04 09:05:48 | [588][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1252 ntime: 0077 mem: 3.36 + 04-04 09:05:55 | [588][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1342 ntime: 0091 mem: 3.36 + 04-04 09:06:03 | Time info >>>> elapsed: 702.67 mins remain: 490.32 mins + 04-04 09:06:04 | [589][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1058 ntime: 0078 mem: 3.36 + 04-04 09:06:11 | [589][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1460 ntime: 0074 mem: 3.36 + 04-04 09:06:18 | [589][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0162 ntime: 0076 mem: 3.36 + 04-04 09:06:26 | [589][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0530 ntime: 0078 mem: 3.36 + 04-04 09:06:33 | [589][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1506 ntime: 0082 mem: 3.36 + 04-04 09:06:40 | [589][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0837 ntime: 0087 mem: 3.36 + 04-04 09:06:46 | [589][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1178 ntime: 0082 mem: 3.36 + 04-04 09:06:54 | [589][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0600 ntime: 0085 mem: 3.36 + 04-04 09:06:59 | [589][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0583 ntime: 0079 mem: 3.36 + 04-04 09:07:06 | [589][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1355 ntime: 0083 mem: 3.36 + 04-04 09:07:12 | [589][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0075 ntime: 0082 mem: 3.36 + 04-04 09:07:20 | [589][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1351 ntime: 0085 mem: 3.36 + 04-04 09:07:28 | [589][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0642 ntime: 0081 mem: 3.36 + 04-04 09:07:34 | [589][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0169 ntime: 0083 mem: 3.36 + 04-04 09:07:44 | [589][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1350 ntime: 0084 mem: 3.36 + 04-04 09:07:52 | [589][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1689 ntime: 0082 mem: 3.36 + 04-04 09:07:58 | [589][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0094 ntime: 0086 mem: 3.36 + 04-04 09:08:04 | [589][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0068 ntime: 0081 mem: 3.36 + 04-04 09:08:11 | Time info >>>> elapsed: 704.81 mins remain: 489.79 mins + 04-04 09:08:12 | [590][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1097 ntime: 0082 mem: 3.36 + 04-04 09:08:19 | [590][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0077 mem: 3.36 + 04-04 09:08:24 | [590][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0075 ntime: 0092 mem: 3.36 + 04-04 09:08:33 | [590][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0067 ntime: 0083 mem: 3.36 + 04-04 09:08:40 | [590][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1402 ntime: 0075 mem: 3.36 + 04-04 09:08:46 | [590][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1271 ntime: 0076 mem: 3.36 + 04-04 09:08:54 | [590][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0086 ntime: 0088 mem: 3.36 + 04-04 09:09:01 | [590][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1056 ntime: 0083 mem: 3.36 + 04-04 09:09:08 | [590][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0098 ntime: 0089 mem: 3.36 + 04-04 09:09:13 | [590][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1014 ntime: 0079 mem: 3.36 + 04-04 09:09:19 | [590][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1250 ntime: 0080 mem: 3.36 + 04-04 09:09:27 | [590][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0730 ntime: 0086 mem: 3.36 + 04-04 09:09:35 | [590][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0835 ntime: 0076 mem: 3.36 + 04-04 09:09:41 | [590][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0085 mem: 3.36 + 04-04 09:09:47 | [590][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0070 ntime: 0089 mem: 3.36 + 04-04 09:09:54 | [590][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0772 ntime: 0077 mem: 3.36 + 04-04 09:10:00 | [590][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1257 ntime: 0086 mem: 3.36 + 04-04 09:10:06 | [590][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0881 ntime: 0074 mem: 3.36 + 04-04 09:10:12 | Time info >>>> elapsed: 706.82 mins remain: 489.15 mins + 04-04 09:10:12 | [591][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0142 ntime: 0076 mem: 3.36 + 04-04 09:10:18 | [591][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0077 mem: 3.36 + 04-04 09:10:26 | [591][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0080 ntime: 0082 mem: 3.36 + 04-04 09:10:32 | [591][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0092 ntime: 0087 mem: 3.36 + 04-04 09:10:41 | [591][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0126 ntime: 0081 mem: 3.36 + 04-04 09:10:49 | [591][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0376 ntime: 0084 mem: 3.36 + 04-04 09:10:56 | [591][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 09:11:03 | [591][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0580 ntime: 0082 mem: 3.36 + 04-04 09:11:10 | [591][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1178 ntime: 0078 mem: 3.36 + 04-04 09:11:19 | [591][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0926 ntime: 0079 mem: 3.36 + 04-04 09:11:26 | [591][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0642 ntime: 0074 mem: 3.36 + 04-04 09:11:33 | [591][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0084 ntime: 0088 mem: 3.36 + 04-04 09:11:42 | [591][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0711 ntime: 0077 mem: 3.36 + 04-04 09:11:48 | [591][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1021 ntime: 0079 mem: 3.36 + 04-04 09:11:56 | [591][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1174 ntime: 0077 mem: 3.36 + 04-04 09:12:02 | [591][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0070 ntime: 0072 mem: 3.36 + 04-04 09:12:09 | [591][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1156 ntime: 0079 mem: 3.36 + 04-04 09:12:17 | [591][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1580 ntime: 0079 mem: 3.36 + 04-04 09:12:23 | Time info >>>> elapsed: 709.01 mins remain: 488.64 mins + 04-04 09:12:23 | [592][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0152 ntime: 0083 mem: 3.36 + 04-04 09:12:31 | [592][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1199 ntime: 0075 mem: 3.36 + 04-04 09:12:38 | [592][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0523 ntime: 0075 mem: 3.36 + 04-04 09:12:45 | [592][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0081 mem: 3.36 + 04-04 09:12:53 | [592][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1109 ntime: 0077 mem: 3.36 + 04-04 09:12:58 | [592][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0082 mem: 3.36 + 04-04 09:13:05 | [592][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0083 ntime: 0081 mem: 3.36 + 04-04 09:13:14 | [592][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1307 ntime: 0084 mem: 3.36 + 04-04 09:13:20 | [592][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0167 ntime: 0078 mem: 3.36 + 04-04 09:13:28 | [592][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1362 ntime: 0089 mem: 3.36 + 04-04 09:13:34 | [592][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0049 ntime: 0077 mem: 3.36 + 04-04 09:13:41 | [592][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0047 ntime: 0084 mem: 3.36 + 04-04 09:13:50 | [592][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0801 ntime: 0077 mem: 3.36 + 04-04 09:13:56 | [592][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0047 ntime: 0080 mem: 3.36 + 04-04 09:14:04 | [592][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0282 ntime: 0080 mem: 3.36 + 04-04 09:14:10 | [592][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0115 ntime: 0078 mem: 3.36 + 04-04 09:14:17 | [592][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0070 ntime: 0078 mem: 3.36 + 04-04 09:14:23 | [592][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1006 ntime: 0086 mem: 3.36 + 04-04 09:14:29 | Time info >>>> elapsed: 711.11 mins remain: 488.06 mins + 04-04 09:14:29 | [593][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0581 ntime: 0077 mem: 3.36 + 04-04 09:14:37 | [593][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1653 ntime: 0076 mem: 3.36 + 04-04 09:14:46 | [593][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0582 ntime: 0079 mem: 3.36 + 04-04 09:14:52 | [593][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 09:14:59 | [593][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0060 ntime: 0080 mem: 3.36 + 04-04 09:15:05 | [593][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0089 mem: 3.36 + 04-04 09:15:15 | [593][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0064 ntime: 0080 mem: 3.36 + 04-04 09:15:21 | [593][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1055 ntime: 0081 mem: 3.36 + 04-04 09:15:28 | [593][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0052 ntime: 0090 mem: 3.36 + 04-04 09:15:37 | [593][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1393 ntime: 0078 mem: 3.36 + 04-04 09:15:44 | [593][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0065 ntime: 0078 mem: 3.36 + 04-04 09:15:50 | [593][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0077 ntime: 0075 mem: 3.36 + 04-04 09:15:58 | [593][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0782 ntime: 0078 mem: 3.36 + 04-04 09:16:04 | [593][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0787 ntime: 0087 mem: 3.36 + 04-04 09:16:14 | [593][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0918 ntime: 0084 mem: 3.36 + 04-04 09:16:21 | [593][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1039 ntime: 0078 mem: 3.36 + 04-04 09:16:30 | [593][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1245 ntime: 0084 mem: 3.36 + 04-04 09:16:38 | [593][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0606 ntime: 0079 mem: 3.36 + 04-04 09:16:43 | Time info >>>> elapsed: 713.35 mins remain: 487.57 mins + 04-04 09:16:45 | [594][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1314 ntime: 0076 mem: 3.36 + 04-04 09:16:52 | [594][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0486 ntime: 0079 mem: 3.36 + 04-04 09:17:00 | [594][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1117 ntime: 0085 mem: 3.36 + 04-04 09:17:07 | [594][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0056 ntime: 0079 mem: 3.36 + 04-04 09:17:15 | [594][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0118 ntime: 0081 mem: 3.36 + 04-04 09:17:23 | [594][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0066 mem: 3.36 + 04-04 09:17:30 | [594][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0174 ntime: 0086 mem: 3.36 + 04-04 09:17:37 | [594][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0858 ntime: 0076 mem: 3.36 + 04-04 09:17:43 | [594][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0051 ntime: 0084 mem: 3.36 + 04-04 09:17:50 | [594][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0075 mem: 3.36 + 04-04 09:17:59 | [594][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1409 ntime: 0077 mem: 3.36 + 04-04 09:18:09 | [594][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1229 ntime: 0081 mem: 3.36 + 04-04 09:18:18 | [594][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0088 ntime: 0081 mem: 3.36 + 04-04 09:18:28 | [594][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1050 ntime: 0082 mem: 3.36 + 04-04 09:18:35 | [594][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0666 ntime: 0082 mem: 3.36 + 04-04 09:18:44 | [594][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0073 ntime: 0074 mem: 3.36 + 04-04 09:18:55 | [594][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1560 ntime: 0084 mem: 3.36 + 04-04 09:19:03 | [594][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0196 ntime: 0085 mem: 3.36 + 04-04 09:19:12 | Time info >>>> elapsed: 715.83 mins remain: 487.25 mins + 04-04 09:19:12 | [595][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 0078 ntime: 0080 mem: 3.36 + 04-04 09:19:20 | [595][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0958 ntime: 0080 mem: 3.36 + 04-04 09:19:31 | [595][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1417 ntime: 0081 mem: 3.36 + 04-04 09:19:41 | [595][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1167 ntime: 0078 mem: 3.36 + 04-04 09:19:51 | [595][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1493 ntime: 0080 mem: 3.36 + 04-04 09:19:59 | [595][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0073 ntime: 0073 mem: 3.36 + 04-04 09:20:09 | [595][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0076 ntime: 0080 mem: 3.36 + 04-04 09:20:19 | [595][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1475 ntime: 0084 mem: 3.36 + 04-04 09:20:28 | [595][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0992 ntime: 0083 mem: 3.36 + 04-04 09:20:38 | [595][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0064 ntime: 0090 mem: 3.36 + 04-04 09:20:48 | [595][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1489 ntime: 0078 mem: 3.36 + 04-04 09:20:57 | [595][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0076 ntime: 0080 mem: 3.36 + 04-04 09:21:08 | [595][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0059 ntime: 0086 mem: 3.36 + 04-04 09:21:18 | [595][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0292 ntime: 0086 mem: 3.36 + 04-04 09:21:25 | [595][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0228 ntime: 0081 mem: 3.36 + 04-04 09:21:32 | [595][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1259 ntime: 0075 mem: 3.36 + 04-04 09:21:38 | [595][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0082 ntime: 0077 mem: 3.36 + 04-04 09:21:47 | [595][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0305 ntime: 0075 mem: 3.36 + 04-04 09:21:54 | Time info >>>> elapsed: 718.52 mins remain: 487.05 mins + 04-04 09:21:54 | [596][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0048 ntime: 0082 mem: 3.36 + 04-04 09:22:02 | [596][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1405 ntime: 0084 mem: 3.36 + 04-04 09:22:10 | [596][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1454 ntime: 0078 mem: 3.36 + 04-04 09:22:16 | [596][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0457 ntime: 0081 mem: 3.36 + 04-04 09:22:22 | [596][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0083 mem: 3.36 + 04-04 09:22:30 | [596][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0805 ntime: 0074 mem: 3.36 + 04-04 09:22:36 | [596][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1139 ntime: 0081 mem: 3.36 + 04-04 09:22:44 | [596][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1007 ntime: 0081 mem: 3.36 + 04-04 09:22:51 | [596][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1583 ntime: 0085 mem: 3.36 + 04-04 09:22:59 | [596][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0098 ntime: 0086 mem: 3.36 + 04-04 09:23:07 | [596][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1400 ntime: 0083 mem: 3.36 + 04-04 09:23:15 | [596][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1660 ntime: 0080 mem: 3.36 + 04-04 09:23:24 | [596][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0075 ntime: 0078 mem: 3.36 + 04-04 09:23:33 | [596][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1186 ntime: 0075 mem: 3.36 + 04-04 09:23:41 | [596][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1355 ntime: 0081 mem: 3.36 + 04-04 09:23:49 | [596][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0614 ntime: 0081 mem: 3.36 + 04-04 09:23:57 | [596][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0069 ntime: 0075 mem: 3.36 + 04-04 09:24:04 | [596][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1440 ntime: 0079 mem: 3.36 + 04-04 09:24:10 | Time info >>>> elapsed: 720.79 mins remain: 486.57 mins + 04-04 09:24:10 | [597][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0069 ntime: 0075 mem: 3.36 + 04-04 09:24:17 | [597][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0135 ntime: 0080 mem: 3.36 + 04-04 09:24:26 | [597][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1262 ntime: 0087 mem: 3.36 + 04-04 09:24:33 | [597][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1201 ntime: 0079 mem: 3.36 + 04-04 09:24:39 | [597][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0557 ntime: 0088 mem: 3.36 + 04-04 09:24:47 | [597][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1395 ntime: 0079 mem: 3.36 + 04-04 09:24:55 | [597][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0073 ntime: 0081 mem: 3.36 + 04-04 09:25:02 | [597][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0218 ntime: 0085 mem: 3.36 + 04-04 09:25:11 | [597][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1265 ntime: 0078 mem: 3.36 + 04-04 09:25:19 | [597][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 09:25:26 | [597][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1165 ntime: 0080 mem: 3.36 + 04-04 09:25:33 | [597][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0484 ntime: 0086 mem: 3.36 + 04-04 09:25:40 | [597][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0104 ntime: 0078 mem: 3.36 + 04-04 09:25:49 | [597][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1406 ntime: 0089 mem: 3.36 + 04-04 09:25:57 | [597][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1444 ntime: 0063 mem: 3.36 + 04-04 09:26:04 | [597][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0086 mem: 3.36 + 04-04 09:26:12 | [597][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1250 ntime: 0088 mem: 3.36 + 04-04 09:26:18 | [597][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0144 ntime: 0077 mem: 3.36 + 04-04 09:26:26 | Time info >>>> elapsed: 723.06 mins remain: 486.07 mins + 04-04 09:26:27 | [598][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1312 ntime: 0080 mem: 3.36 + 04-04 09:26:34 | [598][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0120 ntime: 0078 mem: 3.36 + 04-04 09:26:42 | [598][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1383 ntime: 0086 mem: 3.36 + 04-04 09:26:52 | [598][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1297 ntime: 0086 mem: 3.36 + 04-04 09:27:01 | [598][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1304 ntime: 0078 mem: 3.36 + 04-04 09:27:09 | [598][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1498 ntime: 0084 mem: 3.36 + 04-04 09:27:18 | [598][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1378 ntime: 0082 mem: 3.36 + 04-04 09:27:27 | [598][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1069 ntime: 0077 mem: 3.36 + 04-04 09:27:34 | [598][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1483 ntime: 0073 mem: 3.36 + 04-04 09:27:42 | [598][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0594 ntime: 0082 mem: 3.36 + 04-04 09:27:49 | [598][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1032 ntime: 0076 mem: 3.36 + 04-04 09:27:55 | [598][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0497 ntime: 0081 mem: 3.36 + 04-04 09:28:01 | [598][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0064 ntime: 0080 mem: 3.36 + 04-04 09:28:09 | [598][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1220 ntime: 0084 mem: 3.36 + 04-04 09:28:15 | [598][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0063 ntime: 0074 mem: 3.36 + 04-04 09:28:23 | [598][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0049 ntime: 0082 mem: 3.36 + 04-04 09:28:30 | [598][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1176 ntime: 0077 mem: 3.36 + 04-04 09:28:38 | [598][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1426 ntime: 0079 mem: 3.36 + 04-04 09:28:43 | Time info >>>> elapsed: 725.35 mins remain: 485.58 mins + 04-04 09:28:45 | [599][000/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1198 ntime: 0084 mem: 3.36 + 04-04 09:28:52 | [599][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1390 ntime: 0081 mem: 3.36 + 04-04 09:29:00 | [599][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0049 ntime: 0084 mem: 3.36 + 04-04 09:29:08 | [599][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0049 ntime: 0072 mem: 3.36 + 04-04 09:29:14 | [599][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0654 ntime: 0058 mem: 3.36 + 04-04 09:29:23 | [599][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1376 ntime: 0077 mem: 3.36 + 04-04 09:29:30 | [599][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1443 ntime: 0076 mem: 3.36 + 04-04 09:29:37 | [599][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1340 ntime: 0080 mem: 3.36 + 04-04 09:29:47 | [599][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1211 ntime: 0084 mem: 3.36 + 04-04 09:29:55 | [599][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1195 ntime: 0075 mem: 3.36 + 04-04 09:30:04 | [599][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0921 ntime: 0086 mem: 3.36 + 04-04 09:30:14 | [599][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1460 ntime: 0077 mem: 3.36 + 04-04 09:30:20 | [599][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0974 ntime: 0076 mem: 3.36 + 04-04 09:30:25 | [599][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0047 ntime: 0081 mem: 3.36 + 04-04 09:30:31 | [599][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0182 ntime: 0084 mem: 3.36 + 04-04 09:30:39 | [599][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0954 ntime: 0072 mem: 3.36 + 04-04 09:30:48 | [599][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1343 ntime: 0083 mem: 3.36 + 04-04 09:30:57 | [599][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0945 ntime: 0087 mem: 3.36 + 04-04 09:31:03 | Time info >>>> elapsed: 727.68 mins remain: 485.12 mins + 04-04 09:31:05 | [600][000/179] predict_x0_loss: 0.009 glr: 5.0e-07 dtime: 1323 ntime: 0078 mem: 3.36 + 04-04 09:31:13 | [600][010/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1474 ntime: 0083 mem: 3.36 + 04-04 09:31:19 | [600][020/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0251 ntime: 0080 mem: 3.36 + 04-04 09:31:26 | [600][030/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0054 ntime: 0079 mem: 3.36 + 04-04 09:31:35 | [600][040/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0863 ntime: 0081 mem: 3.36 + 04-04 09:31:41 | [600][050/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0061 ntime: 0081 mem: 3.36 + 04-04 09:31:50 | [600][060/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1409 ntime: 0086 mem: 3.36 + 04-04 09:31:57 | [600][070/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1220 ntime: 0081 mem: 3.36 + 04-04 09:32:05 | [600][080/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1025 ntime: 0084 mem: 3.36 + 04-04 09:32:12 | [600][090/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1364 ntime: 0082 mem: 3.36 + 04-04 09:32:17 | [600][100/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0078 ntime: 0081 mem: 3.36 + 04-04 09:32:24 | [600][110/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0959 ntime: 0080 mem: 3.36 + 04-04 09:32:30 | [600][120/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0870 ntime: 0071 mem: 3.36 + 04-04 09:32:38 | [600][130/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0936 ntime: 0087 mem: 3.36 + 04-04 09:32:48 | [600][140/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 1476 ntime: 0076 mem: 3.36 + 04-04 09:32:56 | [600][150/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0786 ntime: 0072 mem: 3.36 + 04-04 09:33:04 | [600][160/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0053 ntime: 0085 mem: 3.36 + 04-04 09:33:11 | [600][170/179] predict_x0_loss: 0.008 glr: 5.0e-07 dtime: 0764 ntime: 0082 mem: 3.36 + 04-04 09:33:20 | Time info >>>> elapsed: 729.97 mins remain: 484.62 mins + 04-04 09:33:22 | [601][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 1126 ntime: 0075 mem: 3.36 + 04-04 09:33:27 | [601][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0087 mem: 3.36 + 04-04 09:33:36 | [601][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1272 ntime: 0081 mem: 3.36 + 04-04 09:33:44 | [601][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1142 ntime: 0078 mem: 3.36 + 04-04 09:33:52 | [601][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1140 ntime: 0074 mem: 3.36 + 04-04 09:33:59 | [601][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0077 mem: 3.36 + 04-04 09:34:07 | [601][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0771 ntime: 0073 mem: 3.36 + 04-04 09:34:15 | [601][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1152 ntime: 0082 mem: 3.36 + 04-04 09:34:23 | [601][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0654 ntime: 0083 mem: 3.36 + 04-04 09:34:30 | [601][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1269 ntime: 0075 mem: 3.36 + 04-04 09:34:37 | [601][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1338 ntime: 0077 mem: 3.36 + 04-04 09:34:45 | [601][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1552 ntime: 0083 mem: 3.36 + 04-04 09:34:52 | [601][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1451 ntime: 0080 mem: 3.36 + 04-04 09:34:57 | [601][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0132 ntime: 0058 mem: 3.36 + 04-04 09:35:06 | [601][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1293 ntime: 0072 mem: 3.36 + 04-04 09:35:13 | [601][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1032 ntime: 0080 mem: 3.36 + 04-04 09:35:20 | [601][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0094 ntime: 0087 mem: 3.36 + 04-04 09:35:29 | [601][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1415 ntime: 0079 mem: 3.36 + 04-04 09:35:34 | Time info >>>> elapsed: 732.20 mins remain: 484.08 mins + 04-04 09:35:36 | [602][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1648 ntime: 0083 mem: 3.36 + 04-04 09:35:45 | [602][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0080 mem: 3.36 + 04-04 09:35:55 | [602][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1305 ntime: 0074 mem: 3.36 + 04-04 09:36:02 | [602][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0088 ntime: 0078 mem: 3.36 + 04-04 09:36:08 | [602][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0112 ntime: 0086 mem: 3.36 + 04-04 09:36:15 | [602][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0097 ntime: 0082 mem: 3.36 + 04-04 09:36:24 | [602][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1379 ntime: 0081 mem: 3.36 + 04-04 09:36:31 | [602][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0082 mem: 3.36 + 04-04 09:36:38 | [602][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0244 ntime: 0087 mem: 3.36 + 04-04 09:36:45 | [602][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0532 ntime: 0081 mem: 3.36 + 04-04 09:36:55 | [602][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1431 ntime: 0090 mem: 3.36 + 04-04 09:37:01 | [602][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0082 mem: 3.36 + 04-04 09:37:08 | [602][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1302 ntime: 0081 mem: 3.36 + 04-04 09:37:16 | [602][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0913 ntime: 0085 mem: 3.36 + 04-04 09:37:24 | [602][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0080 mem: 3.36 + 04-04 09:37:34 | [602][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0106 ntime: 0087 mem: 3.36 + 04-04 09:37:42 | [602][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0079 ntime: 0082 mem: 3.36 + 04-04 09:37:50 | [602][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0890 ntime: 0081 mem: 3.36 + 04-04 09:37:57 | Time info >>>> elapsed: 734.57 mins remain: 483.63 mins + 04-04 09:37:58 | [603][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1072 ntime: 0080 mem: 3.36 + 04-04 09:38:06 | [603][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0080 ntime: 0078 mem: 3.36 + 04-04 09:38:14 | [603][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0142 ntime: 0080 mem: 3.36 + 04-04 09:38:22 | [603][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1394 ntime: 0087 mem: 3.36 + 04-04 09:38:29 | [603][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1371 ntime: 0083 mem: 3.36 + 04-04 09:38:35 | [603][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0091 mem: 3.36 + 04-04 09:38:45 | [603][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1258 ntime: 0084 mem: 3.36 + 04-04 09:38:52 | [603][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0062 ntime: 0079 mem: 3.36 + 04-04 09:39:00 | [603][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0192 ntime: 0086 mem: 3.36 + 04-04 09:39:09 | [603][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0079 mem: 3.36 + 04-04 09:39:17 | [603][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1155 ntime: 0087 mem: 3.36 + 04-04 09:39:24 | [603][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0820 ntime: 0081 mem: 3.36 + 04-04 09:39:30 | [603][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0063 ntime: 0079 mem: 3.36 + 04-04 09:39:37 | [603][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0068 ntime: 0086 mem: 3.36 + 04-04 09:39:45 | [603][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0108 ntime: 0078 mem: 3.36 + 04-04 09:39:53 | [603][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0128 ntime: 0082 mem: 3.36 + 04-04 09:40:01 | [603][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1210 ntime: 0085 mem: 3.36 + 04-04 09:40:08 | [603][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0083 mem: 3.36 + 04-04 09:40:13 | Time info >>>> elapsed: 736.84 mins remain: 483.10 mins + 04-04 09:40:14 | [604][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1359 ntime: 0081 mem: 3.36 + 04-04 09:40:23 | [604][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0080 mem: 3.36 + 04-04 09:40:31 | [604][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0087 mem: 3.36 + 04-04 09:40:37 | [604][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0881 ntime: 0081 mem: 3.36 + 04-04 09:40:44 | [604][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1358 ntime: 0084 mem: 3.36 + 04-04 09:40:50 | [604][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0096 mem: 3.36 + 04-04 09:41:00 | [604][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1109 ntime: 0075 mem: 3.36 + 04-04 09:41:08 | [604][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0814 ntime: 0081 mem: 3.36 + 04-04 09:41:15 | [604][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1301 ntime: 0091 mem: 3.36 + 04-04 09:41:24 | [604][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1111 ntime: 0081 mem: 3.36 + 04-04 09:41:31 | [604][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0173 ntime: 0089 mem: 3.36 + 04-04 09:41:41 | [604][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1472 ntime: 0080 mem: 3.36 + 04-04 09:41:49 | [604][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0079 mem: 3.36 + 04-04 09:41:55 | [604][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0080 mem: 3.36 + 04-04 09:42:02 | [604][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0873 ntime: 0080 mem: 3.36 + 04-04 09:42:09 | [604][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0497 ntime: 0079 mem: 3.36 + 04-04 09:42:16 | [604][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0078 ntime: 0080 mem: 3.36 + 04-04 09:42:24 | [604][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 09:42:30 | Time info >>>> elapsed: 739.13 mins remain: 482.57 mins + 04-04 09:42:31 | [605][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0587 ntime: 0073 mem: 3.36 + 04-04 09:42:38 | [605][010/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 1378 ntime: 0086 mem: 3.36 + 04-04 09:42:44 | [605][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0081 mem: 3.36 + 04-04 09:42:53 | [605][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1130 ntime: 0077 mem: 3.36 + 04-04 09:43:00 | [605][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0151 ntime: 0085 mem: 3.36 + 04-04 09:43:08 | [605][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0067 ntime: 0076 mem: 3.36 + 04-04 09:43:14 | [605][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1425 ntime: 0084 mem: 3.36 + 04-04 09:43:22 | [605][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1332 ntime: 0083 mem: 3.36 + 04-04 09:43:30 | [605][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1309 ntime: 0076 mem: 3.36 + 04-04 09:43:37 | [605][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1256 ntime: 0078 mem: 3.36 + 04-04 09:43:45 | [605][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1253 ntime: 0065 mem: 3.36 + 04-04 09:43:52 | [605][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1334 ntime: 0078 mem: 3.36 + 04-04 09:44:01 | [605][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1395 ntime: 0083 mem: 3.36 + 04-04 09:44:06 | [605][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0088 ntime: 0071 mem: 3.36 + 04-04 09:44:13 | [605][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0801 ntime: 0077 mem: 3.36 + 04-04 09:44:20 | [605][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0081 mem: 3.36 + 04-04 09:44:28 | [605][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1093 ntime: 0082 mem: 3.36 + 04-04 09:44:36 | [605][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1185 ntime: 0083 mem: 3.36 + 04-04 09:44:42 | Time info >>>> elapsed: 741.33 mins remain: 481.99 mins + 04-04 09:44:42 | [606][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0123 ntime: 0076 mem: 3.36 + 04-04 09:44:49 | [606][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0562 ntime: 0084 mem: 3.36 + 04-04 09:44:56 | [606][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1314 ntime: 0084 mem: 3.36 + 04-04 09:45:03 | [606][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1193 ntime: 0078 mem: 3.36 + 04-04 09:45:10 | [606][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0105 ntime: 0085 mem: 3.36 + 04-04 09:45:20 | [606][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0089 ntime: 0083 mem: 3.36 + 04-04 09:45:28 | [606][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1136 ntime: 0077 mem: 3.36 + 04-04 09:45:35 | [606][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0907 ntime: 0078 mem: 3.36 + 04-04 09:45:43 | [606][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0047 ntime: 0079 mem: 3.36 + 04-04 09:45:48 | [606][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0062 ntime: 0080 mem: 3.36 + 04-04 09:45:58 | [606][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1368 ntime: 0083 mem: 3.36 + 04-04 09:46:07 | [606][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1551 ntime: 0076 mem: 3.36 + 04-04 09:46:14 | [606][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 09:46:22 | [606][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0105 ntime: 0086 mem: 3.36 + 04-04 09:46:28 | [606][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0077 ntime: 0086 mem: 3.36 + 04-04 09:46:35 | [606][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0826 ntime: 0082 mem: 3.36 + 04-04 09:46:42 | [606][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0892 ntime: 0083 mem: 3.36 + 04-04 09:46:50 | [606][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0074 mem: 3.36 + 04-04 09:46:59 | Time info >>>> elapsed: 743.60 mins remain: 481.44 mins + 04-04 09:46:59 | [607][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0689 ntime: 0075 mem: 3.36 + 04-04 09:47:07 | [607][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0063 ntime: 0086 mem: 3.36 + 04-04 09:47:16 | [607][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1380 ntime: 0082 mem: 3.36 + 04-04 09:47:23 | [607][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0077 mem: 3.36 + 04-04 09:47:29 | [607][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0469 ntime: 0081 mem: 3.36 + 04-04 09:47:36 | [607][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0114 ntime: 0086 mem: 3.36 + 04-04 09:47:44 | [607][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0829 ntime: 0076 mem: 3.36 + 04-04 09:47:52 | [607][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1409 ntime: 0077 mem: 3.36 + 04-04 09:47:58 | [607][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0720 ntime: 0081 mem: 3.36 + 04-04 09:48:05 | [607][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 09:48:13 | [607][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0085 ntime: 0083 mem: 3.36 + 04-04 09:48:23 | [607][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1410 ntime: 0079 mem: 3.36 + 04-04 09:48:32 | [607][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1260 ntime: 0074 mem: 3.36 + 04-04 09:48:39 | [607][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0083 mem: 3.36 + 04-04 09:48:44 | [607][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0189 ntime: 0078 mem: 3.36 + 04-04 09:48:52 | [607][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1278 ntime: 0076 mem: 3.36 + 04-04 09:48:58 | [607][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0075 mem: 3.36 + 04-04 09:49:09 | [607][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0086 mem: 3.36 + 04-04 09:49:14 | Time info >>>> elapsed: 745.86 mins remain: 480.88 mins + 04-04 09:49:14 | [608][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0085 ntime: 0090 mem: 3.36 + 04-04 09:49:23 | [608][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0757 ntime: 0080 mem: 3.36 + 04-04 09:49:28 | [608][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0045 ntime: 0067 mem: 3.36 + 04-04 09:49:35 | [608][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1004 ntime: 0072 mem: 3.36 + 04-04 09:49:42 | [608][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0078 mem: 3.36 + 04-04 09:49:48 | [608][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0364 ntime: 0081 mem: 3.36 + 04-04 09:49:56 | [608][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0076 mem: 3.36 + 04-04 09:50:01 | [608][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0600 ntime: 0077 mem: 3.36 + 04-04 09:50:09 | [608][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0860 ntime: 0084 mem: 3.36 + 04-04 09:50:16 | [608][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0076 mem: 3.36 + 04-04 09:50:23 | [608][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1289 ntime: 0080 mem: 3.36 + 04-04 09:50:32 | [608][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0082 mem: 3.36 + 04-04 09:50:43 | [608][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0768 ntime: 0077 mem: 3.36 + 04-04 09:50:50 | [608][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0078 mem: 3.36 + 04-04 09:50:58 | [608][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0084 mem: 3.36 + 04-04 09:51:06 | [608][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1003 ntime: 0081 mem: 3.36 + 04-04 09:51:13 | [608][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0077 mem: 3.36 + 04-04 09:51:20 | [608][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1196 ntime: 0076 mem: 3.36 + 04-04 09:51:25 | Time info >>>> elapsed: 748.04 mins remain: 480.27 mins + 04-04 09:51:26 | [609][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 1099 ntime: 0081 mem: 3.36 + 04-04 09:51:35 | [609][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1427 ntime: 0080 mem: 3.36 + 04-04 09:51:40 | [609][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1351 ntime: 0081 mem: 3.36 + 04-04 09:51:49 | [609][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1123 ntime: 0080 mem: 3.36 + 04-04 09:51:56 | [609][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0086 ntime: 0077 mem: 3.36 + 04-04 09:52:04 | [609][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0886 ntime: 0078 mem: 3.36 + 04-04 09:52:11 | [609][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0814 ntime: 0074 mem: 3.36 + 04-04 09:52:18 | [609][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0074 mem: 3.36 + 04-04 09:52:25 | [609][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0566 ntime: 0059 mem: 3.36 + 04-04 09:52:31 | [609][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0075 mem: 3.36 + 04-04 09:52:38 | [609][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0081 ntime: 0072 mem: 3.36 + 04-04 09:52:45 | [609][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0081 ntime: 0074 mem: 3.36 + 04-04 09:52:53 | [609][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1001 ntime: 0081 mem: 3.36 + 04-04 09:53:02 | [609][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1232 ntime: 0079 mem: 3.36 + 04-04 09:53:10 | [609][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0084 mem: 3.36 + 04-04 09:53:18 | [609][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0077 mem: 3.36 + 04-04 09:53:25 | [609][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0231 ntime: 0081 mem: 3.36 + 04-04 09:53:31 | [609][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0700 ntime: 0086 mem: 3.36 + 04-04 09:53:38 | Time info >>>> elapsed: 750.26 mins remain: 479.67 mins + 04-04 09:53:38 | [610][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0075 mem: 3.36 + 04-04 09:53:47 | [610][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0089 ntime: 0079 mem: 3.36 + 04-04 09:53:56 | [610][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0786 ntime: 0080 mem: 3.36 + 04-04 09:54:02 | [610][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0922 ntime: 0078 mem: 3.36 + 04-04 09:54:10 | [610][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1433 ntime: 0090 mem: 3.36 + 04-04 09:54:19 | [610][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1031 ntime: 0076 mem: 3.36 + 04-04 09:54:26 | [610][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0087 mem: 3.36 + 04-04 09:54:33 | [610][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0092 mem: 3.36 + 04-04 09:54:41 | [610][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0074 ntime: 0084 mem: 3.36 + 04-04 09:54:48 | [610][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1284 ntime: 0082 mem: 3.36 + 04-04 09:54:55 | [610][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0075 ntime: 0076 mem: 3.36 + 04-04 09:55:03 | [610][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0839 ntime: 0079 mem: 3.36 + 04-04 09:55:10 | [610][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0065 ntime: 0084 mem: 3.36 + 04-04 09:55:19 | [610][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0081 mem: 3.36 + 04-04 09:55:26 | [610][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1215 ntime: 0076 mem: 3.36 + 04-04 09:55:32 | [610][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0070 mem: 3.36 + 04-04 09:55:40 | [610][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1511 ntime: 0081 mem: 3.36 + 04-04 09:55:47 | [610][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0084 mem: 3.36 + 04-04 09:55:52 | Time info >>>> elapsed: 752.50 mins remain: 479.09 mins + 04-04 09:55:54 | [611][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1148 ntime: 0080 mem: 3.36 + 04-04 09:56:00 | [611][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0695 ntime: 0078 mem: 3.36 + 04-04 09:56:08 | [611][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0129 ntime: 0080 mem: 3.36 + 04-04 09:56:17 | [611][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1506 ntime: 0081 mem: 3.36 + 04-04 09:56:23 | [611][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0087 ntime: 0075 mem: 3.36 + 04-04 09:56:32 | [611][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0430 ntime: 0071 mem: 3.36 + 04-04 09:56:39 | [611][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0077 mem: 3.36 + 04-04 09:56:46 | [611][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0067 ntime: 0080 mem: 3.36 + 04-04 09:56:55 | [611][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1396 ntime: 0079 mem: 3.36 + 04-04 09:57:04 | [611][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1544 ntime: 0085 mem: 3.36 + 04-04 09:57:13 | [611][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0086 mem: 3.36 + 04-04 09:57:23 | [611][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1211 ntime: 0080 mem: 3.36 + 04-04 09:57:30 | [611][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1201 ntime: 0077 mem: 3.36 + 04-04 09:57:38 | [611][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0188 ntime: 0080 mem: 3.36 + 04-04 09:57:46 | [611][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1438 ntime: 0079 mem: 3.36 + 04-04 09:57:54 | [611][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1249 ntime: 0085 mem: 3.36 + 04-04 09:58:02 | [611][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0688 ntime: 0073 mem: 3.36 + 04-04 09:58:08 | [611][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0771 ntime: 0082 mem: 3.36 + 04-04 09:58:13 | Time info >>>> elapsed: 754.84 mins remain: 478.56 mins + 04-04 09:58:14 | [612][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 1153 ntime: 0076 mem: 3.36 + 04-04 09:58:22 | [612][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1189 ntime: 0077 mem: 3.36 + 04-04 09:58:30 | [612][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1299 ntime: 0077 mem: 3.36 + 04-04 09:58:39 | [612][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1400 ntime: 0077 mem: 3.36 + 04-04 09:58:48 | [612][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0998 ntime: 0076 mem: 3.36 + 04-04 09:58:53 | [612][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0083 ntime: 0084 mem: 3.36 + 04-04 09:59:02 | [612][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1101 ntime: 0076 mem: 3.36 + 04-04 09:59:07 | [612][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0078 mem: 3.36 + 04-04 09:59:15 | [612][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0143 ntime: 0087 mem: 3.36 + 04-04 09:59:24 | [612][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0374 ntime: 0084 mem: 3.36 + 04-04 09:59:31 | [612][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0077 ntime: 0083 mem: 3.36 + 04-04 09:59:36 | [612][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0084 ntime: 0084 mem: 3.36 + 04-04 09:59:44 | [612][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0062 ntime: 0080 mem: 3.36 + 04-04 09:59:52 | [612][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0082 mem: 3.36 + 04-04 10:00:00 | [612][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1422 ntime: 0081 mem: 3.36 + 04-04 10:00:06 | [612][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0085 mem: 3.36 + 04-04 10:00:13 | [612][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0079 ntime: 0080 mem: 3.36 + 04-04 10:00:22 | [612][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1082 ntime: 0088 mem: 3.36 + 04-04 10:00:27 | Time info >>>> elapsed: 757.08 mins remain: 477.96 mins + 04-04 10:00:29 | [613][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1428 ntime: 0085 mem: 3.36 + 04-04 10:00:34 | [613][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0769 ntime: 0080 mem: 3.36 + 04-04 10:00:43 | [613][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0657 ntime: 0083 mem: 3.36 + 04-04 10:00:50 | [613][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0083 ntime: 0085 mem: 3.36 + 04-04 10:00:58 | [613][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 10:01:06 | [613][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1393 ntime: 0075 mem: 3.36 + 04-04 10:01:15 | [613][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1308 ntime: 0055 mem: 3.36 + 04-04 10:01:20 | [613][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0090 mem: 3.36 + 04-04 10:01:27 | [613][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0941 ntime: 0085 mem: 3.36 + 04-04 10:01:35 | [613][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0726 ntime: 0085 mem: 3.36 + 04-04 10:01:42 | [613][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0088 mem: 3.36 + 04-04 10:01:51 | [613][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1341 ntime: 0086 mem: 3.36 + 04-04 10:02:00 | [613][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0787 ntime: 0081 mem: 3.36 + 04-04 10:02:07 | [613][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1448 ntime: 0085 mem: 3.36 + 04-04 10:02:14 | [613][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0070 ntime: 0090 mem: 3.36 + 04-04 10:02:22 | [613][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1283 ntime: 0078 mem: 3.36 + 04-04 10:02:31 | [613][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1379 ntime: 0080 mem: 3.36 + 04-04 10:02:38 | [613][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1080 ntime: 0084 mem: 3.36 + 04-04 10:02:45 | Time info >>>> elapsed: 759.38 mins remain: 477.40 mins + 04-04 10:02:46 | [614][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1168 ntime: 0077 mem: 3.36 + 04-04 10:02:52 | [614][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0070 ntime: 0091 mem: 3.36 + 04-04 10:03:01 | [614][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0695 ntime: 0084 mem: 3.36 + 04-04 10:03:08 | [614][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0082 mem: 3.36 + 04-04 10:03:15 | [614][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0692 ntime: 0078 mem: 3.36 + 04-04 10:03:22 | [614][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1153 ntime: 0076 mem: 3.36 + 04-04 10:03:28 | [614][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0084 ntime: 0079 mem: 3.36 + 04-04 10:03:35 | [614][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1008 ntime: 0082 mem: 3.36 + 04-04 10:03:43 | [614][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0647 ntime: 0079 mem: 3.36 + 04-04 10:03:50 | [614][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0819 ntime: 0080 mem: 3.36 + 04-04 10:03:57 | [614][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0769 ntime: 0075 mem: 3.36 + 04-04 10:04:03 | [614][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0076 mem: 3.36 + 04-04 10:04:08 | [614][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0868 ntime: 0074 mem: 3.36 + 04-04 10:04:14 | [614][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1154 ntime: 0076 mem: 3.36 + 04-04 10:04:20 | [614][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0644 ntime: 0080 mem: 3.36 + 04-04 10:04:29 | [614][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0948 ntime: 0079 mem: 3.36 + 04-04 10:04:37 | [614][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1789 ntime: 0080 mem: 3.36 + 04-04 10:04:44 | [614][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1087 ntime: 0087 mem: 3.36 + 04-04 10:04:50 | Time info >>>> elapsed: 761.47 mins remain: 476.69 mins + 04-04 10:04:51 | [615][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0244 ntime: 0078 mem: 3.36 + 04-04 10:04:57 | [615][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0077 mem: 3.36 + 04-04 10:05:05 | [615][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1085 ntime: 0075 mem: 3.36 + 04-04 10:05:11 | [615][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1344 ntime: 0083 mem: 3.36 + 04-04 10:05:17 | [615][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1016 ntime: 0077 mem: 3.36 + 04-04 10:05:26 | [615][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1517 ntime: 0084 mem: 3.36 + 04-04 10:05:34 | [615][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0160 ntime: 0076 mem: 3.36 + 04-04 10:05:42 | [615][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1264 ntime: 0081 mem: 3.36 + 04-04 10:05:49 | [615][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0979 ntime: 0082 mem: 3.36 + 04-04 10:05:57 | [615][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1413 ntime: 0088 mem: 3.36 + 04-04 10:06:04 | [615][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0738 ntime: 0085 mem: 3.36 + 04-04 10:06:12 | [615][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1359 ntime: 0079 mem: 3.36 + 04-04 10:06:18 | [615][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0261 ntime: 0087 mem: 3.36 + 04-04 10:06:27 | [615][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1485 ntime: 0084 mem: 3.36 + 04-04 10:06:34 | [615][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0074 ntime: 0086 mem: 3.36 + 04-04 10:06:41 | [615][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0080 mem: 3.36 + 04-04 10:06:48 | [615][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0992 ntime: 0091 mem: 3.36 + 04-04 10:06:57 | [615][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0067 ntime: 0072 mem: 3.36 + 04-04 10:07:01 | Time info >>>> elapsed: 763.65 mins remain: 476.04 mins + 04-04 10:07:02 | [616][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0176 ntime: 0058 mem: 3.36 + 04-04 10:07:08 | [616][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0647 ntime: 0077 mem: 3.36 + 04-04 10:07:15 | [616][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1084 ntime: 0086 mem: 3.36 + 04-04 10:07:20 | [616][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0112 ntime: 0082 mem: 3.36 + 04-04 10:07:28 | [616][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1382 ntime: 0084 mem: 3.36 + 04-04 10:07:34 | [616][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0092 mem: 3.36 + 04-04 10:07:41 | [616][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0086 mem: 3.36 + 04-04 10:07:49 | [616][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0494 ntime: 0080 mem: 3.36 + 04-04 10:07:56 | [616][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1168 ntime: 0080 mem: 3.36 + 04-04 10:08:04 | [616][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1611 ntime: 0076 mem: 3.36 + 04-04 10:08:10 | [616][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0083 mem: 3.36 + 04-04 10:08:19 | [616][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1048 ntime: 0076 mem: 3.36 + 04-04 10:08:27 | [616][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0984 ntime: 0080 mem: 3.36 + 04-04 10:08:33 | [616][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0119 ntime: 0081 mem: 3.36 + 04-04 10:08:42 | [616][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0083 mem: 3.36 + 04-04 10:08:49 | [616][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0813 ntime: 0086 mem: 3.36 + 04-04 10:08:57 | [616][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0587 ntime: 0076 mem: 3.36 + 04-04 10:09:04 | [616][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0728 ntime: 0071 mem: 3.36 + 04-04 10:09:08 | Time info >>>> elapsed: 765.76 mins remain: 475.34 mins + 04-04 10:09:09 | [617][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0257 ntime: 0075 mem: 3.36 + 04-04 10:09:15 | [617][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0069 ntime: 0084 mem: 3.36 + 04-04 10:09:21 | [617][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0085 mem: 3.36 + 04-04 10:09:28 | [617][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0638 ntime: 0092 mem: 3.36 + 04-04 10:09:33 | [617][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0080 mem: 3.36 + 04-04 10:09:39 | [617][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0852 ntime: 0080 mem: 3.36 + 04-04 10:09:45 | [617][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1387 ntime: 0088 mem: 3.36 + 04-04 10:09:51 | [617][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1092 ntime: 0084 mem: 3.36 + 04-04 10:09:58 | [617][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0982 ntime: 0060 mem: 3.36 + 04-04 10:10:06 | [617][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0076 mem: 3.36 + 04-04 10:10:14 | [617][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1178 ntime: 0078 mem: 3.36 + 04-04 10:10:20 | [617][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1319 ntime: 0086 mem: 3.36 + 04-04 10:10:28 | [617][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0640 ntime: 0082 mem: 3.36 + 04-04 10:10:36 | [617][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0082 mem: 3.36 + 04-04 10:10:42 | [617][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1540 ntime: 0083 mem: 3.36 + 04-04 10:10:49 | [617][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0327 ntime: 0077 mem: 3.36 + 04-04 10:10:56 | [617][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0085 mem: 3.36 + 04-04 10:11:04 | [617][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0088 ntime: 0081 mem: 3.36 + 04-04 10:11:10 | Time info >>>> elapsed: 767.79 mins remain: 474.59 mins + 04-04 10:11:10 | [618][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0105 ntime: 0079 mem: 3.36 + 04-04 10:11:17 | [618][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0956 ntime: 0078 mem: 3.36 + 04-04 10:11:26 | [618][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1093 ntime: 0083 mem: 3.36 + 04-04 10:11:31 | [618][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1448 ntime: 0080 mem: 3.36 + 04-04 10:11:37 | [618][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0324 ntime: 0083 mem: 3.36 + 04-04 10:11:46 | [618][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0937 ntime: 0085 mem: 3.36 + 04-04 10:11:53 | [618][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0088 ntime: 0078 mem: 3.36 + 04-04 10:11:59 | [618][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0741 ntime: 0086 mem: 3.36 + 04-04 10:12:07 | [618][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1171 ntime: 0085 mem: 3.36 + 04-04 10:12:13 | [618][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0742 ntime: 0077 mem: 3.36 + 04-04 10:12:20 | [618][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1114 ntime: 0085 mem: 3.36 + 04-04 10:12:28 | [618][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0078 mem: 3.36 + 04-04 10:12:35 | [618][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0435 ntime: 0082 mem: 3.36 + 04-04 10:12:43 | [618][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1179 ntime: 0077 mem: 3.36 + 04-04 10:12:49 | [618][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0954 ntime: 0079 mem: 3.36 + 04-04 10:12:56 | [618][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0901 ntime: 0078 mem: 3.36 + 04-04 10:13:03 | [618][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0088 mem: 3.36 + 04-04 10:13:12 | [618][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0089 ntime: 0087 mem: 3.36 + 04-04 10:13:16 | Time info >>>> elapsed: 769.90 mins remain: 473.88 mins + 04-04 10:13:17 | [619][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0565 ntime: 0074 mem: 3.36 + 04-04 10:13:23 | [619][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0086 mem: 3.36 + 04-04 10:13:30 | [619][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1351 ntime: 0085 mem: 3.36 + 04-04 10:13:39 | [619][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1109 ntime: 0081 mem: 3.36 + 04-04 10:13:46 | [619][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1412 ntime: 0079 mem: 3.36 + 04-04 10:13:51 | [619][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0084 mem: 3.36 + 04-04 10:13:59 | [619][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0662 ntime: 0077 mem: 3.36 + 04-04 10:14:08 | [619][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1036 ntime: 0080 mem: 3.36 + 04-04 10:14:15 | [619][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0085 mem: 3.36 + 04-04 10:14:23 | [619][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1203 ntime: 0078 mem: 3.36 + 04-04 10:14:31 | [619][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0928 ntime: 0075 mem: 3.36 + 04-04 10:14:37 | [619][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0078 mem: 3.36 + 04-04 10:14:45 | [619][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1419 ntime: 0081 mem: 3.36 + 04-04 10:14:53 | [619][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0146 ntime: 0077 mem: 3.36 + 04-04 10:14:59 | [619][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0160 ntime: 0081 mem: 3.36 + 04-04 10:15:06 | [619][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1642 ntime: 0080 mem: 3.36 + 04-04 10:15:12 | [619][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0426 ntime: 0085 mem: 3.36 + 04-04 10:15:18 | [619][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0075 ntime: 0075 mem: 3.36 + 04-04 10:15:24 | Time info >>>> elapsed: 772.03 mins remain: 473.18 mins + 04-04 10:15:24 | [620][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0106 ntime: 0076 mem: 3.36 + 04-04 10:15:32 | [620][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1253 ntime: 0090 mem: 3.36 + 04-04 10:15:40 | [620][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1427 ntime: 0079 mem: 3.36 + 04-04 10:15:45 | [620][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0865 ntime: 0080 mem: 3.36 + 04-04 10:15:53 | [620][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0704 ntime: 0080 mem: 3.36 + 04-04 10:15:59 | [620][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0084 mem: 3.36 + 04-04 10:16:07 | [620][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1380 ntime: 0078 mem: 3.36 + 04-04 10:16:11 | [620][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0498 ntime: 0082 mem: 3.36 + 04-04 10:16:19 | [620][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0955 ntime: 0085 mem: 3.36 + 04-04 10:16:25 | [620][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1346 ntime: 0079 mem: 3.36 + 04-04 10:16:32 | [620][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0098 ntime: 0075 mem: 3.36 + 04-04 10:16:39 | [620][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0086 ntime: 0083 mem: 3.36 + 04-04 10:16:48 | [620][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1325 ntime: 0083 mem: 3.36 + 04-04 10:16:54 | [620][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0100 ntime: 0085 mem: 3.36 + 04-04 10:17:03 | [620][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0193 ntime: 0086 mem: 3.36 + 04-04 10:17:11 | [620][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0153 ntime: 0085 mem: 3.36 + 04-04 10:17:20 | [620][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1408 ntime: 0078 mem: 3.36 + 04-04 10:17:28 | [620][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0543 ntime: 0079 mem: 3.36 + 04-04 10:17:34 | Time info >>>> elapsed: 774.20 mins remain: 472.50 mins + 04-04 10:17:35 | [621][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0082 ntime: 0076 mem: 3.36 + 04-04 10:17:41 | [621][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0115 ntime: 0076 mem: 3.36 + 04-04 10:17:50 | [621][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0079 mem: 3.36 + 04-04 10:17:57 | [621][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0075 mem: 3.36 + 04-04 10:18:04 | [621][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1251 ntime: 0080 mem: 3.36 + 04-04 10:18:11 | [621][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1280 ntime: 0081 mem: 3.36 + 04-04 10:18:16 | [621][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0082 mem: 3.36 + 04-04 10:18:23 | [621][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0979 ntime: 0077 mem: 3.36 + 04-04 10:18:30 | [621][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0077 ntime: 0088 mem: 3.36 + 04-04 10:18:39 | [621][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0083 mem: 3.36 + 04-04 10:18:45 | [621][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 10:18:51 | [621][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 10:19:00 | [621][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1205 ntime: 0075 mem: 3.36 + 04-04 10:19:09 | [621][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0204 ntime: 0081 mem: 3.36 + 04-04 10:19:15 | [621][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0139 ntime: 0081 mem: 3.36 + 04-04 10:19:22 | [621][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0085 mem: 3.36 + 04-04 10:19:30 | [621][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0270 ntime: 0085 mem: 3.36 + 04-04 10:19:38 | [621][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0770 ntime: 0079 mem: 3.36 + 04-04 10:19:43 | Time info >>>> elapsed: 776.34 mins remain: 471.79 mins + 04-04 10:19:44 | [622][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1537 ntime: 0087 mem: 3.36 + 04-04 10:19:51 | [622][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0095 ntime: 0074 mem: 3.36 + 04-04 10:19:58 | [622][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1263 ntime: 0082 mem: 3.36 + 04-04 10:20:05 | [622][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1024 ntime: 0082 mem: 3.36 + 04-04 10:20:11 | [622][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0511 ntime: 0077 mem: 3.36 + 04-04 10:20:18 | [622][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0087 ntime: 0079 mem: 3.36 + 04-04 10:20:25 | [622][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1133 ntime: 0077 mem: 3.36 + 04-04 10:20:33 | [622][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0073 ntime: 0080 mem: 3.36 + 04-04 10:20:40 | [622][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0079 mem: 3.36 + 04-04 10:20:50 | [622][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0146 ntime: 0081 mem: 3.36 + 04-04 10:20:58 | [622][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1345 ntime: 0082 mem: 3.36 + 04-04 10:21:06 | [622][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1107 ntime: 0079 mem: 3.36 + 04-04 10:21:12 | [622][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0584 ntime: 0079 mem: 3.36 + 04-04 10:21:18 | [622][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0084 mem: 3.36 + 04-04 10:21:26 | [622][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0810 ntime: 0078 mem: 3.36 + 04-04 10:21:32 | [622][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 10:21:39 | [622][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 10:21:47 | [622][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0085 ntime: 0085 mem: 3.36 + 04-04 10:21:55 | Time info >>>> elapsed: 778.55 mins remain: 471.13 mins + 04-04 10:21:55 | [623][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 10:22:02 | [623][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 10:22:08 | [623][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0099 ntime: 0087 mem: 3.36 + 04-04 10:22:15 | [623][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1358 ntime: 0078 mem: 3.36 + 04-04 10:22:24 | [623][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0130 ntime: 0081 mem: 3.36 + 04-04 10:22:30 | [623][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1257 ntime: 0082 mem: 3.36 + 04-04 10:22:36 | [623][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0081 mem: 3.36 + 04-04 10:22:44 | [623][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1646 ntime: 0079 mem: 3.36 + 04-04 10:22:52 | [623][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0563 ntime: 0079 mem: 3.36 + 04-04 10:22:59 | [623][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0868 ntime: 0079 mem: 3.36 + 04-04 10:23:07 | [623][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1458 ntime: 0081 mem: 3.36 + 04-04 10:23:15 | [623][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1404 ntime: 0084 mem: 3.36 + 04-04 10:23:22 | [623][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0109 ntime: 0093 mem: 3.36 + 04-04 10:23:29 | [623][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0755 ntime: 0090 mem: 3.36 + 04-04 10:23:37 | [623][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1067 ntime: 0075 mem: 3.36 + 04-04 10:23:44 | [623][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0088 mem: 3.36 + 04-04 10:23:51 | [623][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0075 mem: 3.36 + 04-04 10:23:58 | [623][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0985 ntime: 0080 mem: 3.36 + 04-04 10:24:03 | Time info >>>> elapsed: 780.68 mins remain: 470.41 mins + 04-04 10:24:05 | [624][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1348 ntime: 0077 mem: 3.36 + 04-04 10:24:13 | [624][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1581 ntime: 0080 mem: 3.36 + 04-04 10:24:20 | [624][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1228 ntime: 0078 mem: 3.36 + 04-04 10:24:27 | [624][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0861 ntime: 0080 mem: 3.36 + 04-04 10:24:34 | [624][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1101 ntime: 0070 mem: 3.36 + 04-04 10:24:41 | [624][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1279 ntime: 0080 mem: 3.36 + 04-04 10:24:47 | [624][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0651 ntime: 0078 mem: 3.36 + 04-04 10:24:53 | [624][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0109 ntime: 0073 mem: 3.36 + 04-04 10:25:01 | [624][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0613 ntime: 0076 mem: 3.36 + 04-04 10:25:09 | [624][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0541 ntime: 0080 mem: 3.36 + 04-04 10:25:16 | [624][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1162 ntime: 0072 mem: 3.36 + 04-04 10:25:23 | [624][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1219 ntime: 0083 mem: 3.36 + 04-04 10:25:31 | [624][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0078 mem: 3.36 + 04-04 10:25:38 | [624][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1243 ntime: 0080 mem: 3.36 + 04-04 10:25:47 | [624][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1173 ntime: 0076 mem: 3.36 + 04-04 10:25:54 | [624][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1145 ntime: 0084 mem: 3.36 + 04-04 10:26:01 | [624][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0871 ntime: 0078 mem: 3.36 + 04-04 10:26:09 | [624][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0081 mem: 3.36 + 04-04 10:26:17 | Time info >>>> elapsed: 782.90 mins remain: 469.74 mins + 04-04 10:26:17 | [625][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0099 ntime: 0080 mem: 3.36 + 04-04 10:26:24 | [625][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0286 ntime: 0082 mem: 3.36 + 04-04 10:26:34 | [625][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1612 ntime: 0073 mem: 3.36 + 04-04 10:26:42 | [625][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0203 ntime: 0075 mem: 3.36 + 04-04 10:26:49 | [625][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0079 mem: 3.36 + 04-04 10:26:57 | [625][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1071 ntime: 0086 mem: 3.36 + 04-04 10:27:04 | [625][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1541 ntime: 0080 mem: 3.36 + 04-04 10:27:12 | [625][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0086 mem: 3.36 + 04-04 10:27:19 | [625][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0150 ntime: 0086 mem: 3.36 + 04-04 10:27:27 | [625][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1103 ntime: 0073 mem: 3.36 + 04-04 10:27:36 | [625][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1398 ntime: 0087 mem: 3.36 + 04-04 10:27:44 | [625][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1539 ntime: 0081 mem: 3.36 + 04-04 10:27:50 | [625][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0635 ntime: 0079 mem: 3.36 + 04-04 10:27:59 | [625][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1527 ntime: 0080 mem: 3.36 + 04-04 10:28:07 | [625][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1339 ntime: 0081 mem: 3.36 + 04-04 10:28:12 | [625][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0071 ntime: 0094 mem: 3.36 + 04-04 10:28:19 | [625][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0812 ntime: 0082 mem: 3.36 + 04-04 10:28:26 | [625][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0872 ntime: 0083 mem: 3.36 + 04-04 10:28:31 | Time info >>>> elapsed: 785.15 mins remain: 469.08 mins + 04-04 10:28:32 | [626][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0081 ntime: 0070 mem: 3.36 + 04-04 10:28:38 | [626][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 10:28:46 | [626][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1600 ntime: 0083 mem: 3.36 + 04-04 10:28:54 | [626][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1375 ntime: 0073 mem: 3.36 + 04-04 10:29:02 | [626][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0846 ntime: 0072 mem: 3.36 + 04-04 10:29:09 | [626][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0198 ntime: 0086 mem: 3.36 + 04-04 10:29:19 | [626][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1340 ntime: 0080 mem: 3.36 + 04-04 10:29:27 | [626][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0082 mem: 3.36 + 04-04 10:29:35 | [626][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0075 mem: 3.36 + 04-04 10:29:43 | [626][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0925 ntime: 0086 mem: 3.36 + 04-04 10:29:52 | [626][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0723 ntime: 0077 mem: 3.36 + 04-04 10:30:01 | [626][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1103 ntime: 0088 mem: 3.36 + 04-04 10:30:09 | [626][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1529 ntime: 0084 mem: 3.36 + 04-04 10:30:16 | [626][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1385 ntime: 0088 mem: 3.36 + 04-04 10:30:23 | [626][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0047 ntime: 0071 mem: 3.36 + 04-04 10:30:29 | [626][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0176 ntime: 0087 mem: 3.36 + 04-04 10:30:37 | [626][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0302 ntime: 0078 mem: 3.36 + 04-04 10:30:47 | [626][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0114 ntime: 0072 mem: 3.36 + 04-04 10:30:54 | Time info >>>> elapsed: 787.52 mins remain: 468.49 mins + 04-04 10:30:54 | [627][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0130 ntime: 0077 mem: 3.36 + 04-04 10:31:01 | [627][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1161 ntime: 0079 mem: 3.36 + 04-04 10:31:09 | [627][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0063 ntime: 0080 mem: 3.36 + 04-04 10:31:16 | [627][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0078 ntime: 0078 mem: 3.36 + 04-04 10:31:24 | [627][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0062 ntime: 0090 mem: 3.36 + 04-04 10:31:32 | [627][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0084 mem: 3.36 + 04-04 10:31:41 | [627][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0923 ntime: 0082 mem: 3.36 + 04-04 10:31:49 | [627][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0086 ntime: 0088 mem: 3.36 + 04-04 10:31:58 | [627][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1393 ntime: 0082 mem: 3.36 + 04-04 10:32:07 | [627][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1656 ntime: 0084 mem: 3.36 + 04-04 10:32:16 | [627][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0085 mem: 3.36 + 04-04 10:32:24 | [627][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1452 ntime: 0080 mem: 3.36 + 04-04 10:32:31 | [627][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1118 ntime: 0084 mem: 3.36 + 04-04 10:32:39 | [627][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 10:32:45 | [627][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0067 ntime: 0077 mem: 3.36 + 04-04 10:32:53 | [627][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0116 ntime: 0080 mem: 3.36 + 04-04 10:33:01 | [627][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0083 mem: 3.36 + 04-04 10:33:07 | [627][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0308 ntime: 0085 mem: 3.36 + 04-04 10:33:12 | Time info >>>> elapsed: 789.83 mins remain: 467.86 mins + 04-04 10:33:14 | [628][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1334 ntime: 0079 mem: 3.36 + 04-04 10:33:22 | [628][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0086 mem: 3.36 + 04-04 10:33:27 | [628][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0087 ntime: 0084 mem: 3.36 + 04-04 10:33:35 | [628][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0152 ntime: 0078 mem: 3.36 + 04-04 10:33:43 | [628][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1353 ntime: 0079 mem: 3.36 + 04-04 10:33:52 | [628][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0078 mem: 3.36 + 04-04 10:33:58 | [628][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0047 ntime: 0077 mem: 3.36 + 04-04 10:34:07 | [628][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1303 ntime: 0078 mem: 3.36 + 04-04 10:34:16 | [628][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0939 ntime: 0077 mem: 3.36 + 04-04 10:34:23 | [628][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 10:34:31 | [628][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0620 ntime: 0078 mem: 3.36 + 04-04 10:34:38 | [628][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0633 ntime: 0077 mem: 3.36 + 04-04 10:34:46 | [628][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0086 ntime: 0079 mem: 3.36 + 04-04 10:34:54 | [628][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0082 mem: 3.36 + 04-04 10:35:03 | [628][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1390 ntime: 0084 mem: 3.36 + 04-04 10:35:10 | [628][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1434 ntime: 0086 mem: 3.36 + 04-04 10:35:18 | [628][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0613 ntime: 0077 mem: 3.36 + 04-04 10:35:23 | [628][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0068 ntime: 0083 mem: 3.36 + 04-04 10:35:30 | Time info >>>> elapsed: 792.13 mins remain: 467.22 mins + 04-04 10:35:31 | [629][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 10:35:37 | [629][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0077 mem: 3.36 + 04-04 10:35:45 | [629][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0083 mem: 3.36 + 04-04 10:35:54 | [629][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1075 ntime: 0078 mem: 3.36 + 04-04 10:35:59 | [629][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0059 mem: 3.36 + 04-04 10:36:08 | [629][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1370 ntime: 0076 mem: 3.36 + 04-04 10:36:17 | [629][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1076 ntime: 0081 mem: 3.36 + 04-04 10:36:24 | [629][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0076 mem: 3.36 + 04-04 10:36:31 | [629][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0870 ntime: 0084 mem: 3.36 + 04-04 10:36:39 | [629][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0568 ntime: 0077 mem: 3.36 + 04-04 10:36:47 | [629][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0062 ntime: 0080 mem: 3.36 + 04-04 10:36:54 | [629][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1134 ntime: 0075 mem: 3.36 + 04-04 10:37:01 | [629][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0523 ntime: 0077 mem: 3.36 + 04-04 10:37:09 | [629][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0082 mem: 3.36 + 04-04 10:37:16 | [629][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0991 ntime: 0079 mem: 3.36 + 04-04 10:37:25 | [629][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0077 mem: 3.36 + 04-04 10:37:34 | [629][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1182 ntime: 0084 mem: 3.36 + 04-04 10:37:43 | [629][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1516 ntime: 0088 mem: 3.36 + 04-04 10:37:48 | Time info >>>> elapsed: 794.44 mins remain: 466.57 mins + 04-04 10:37:49 | [630][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0081 mem: 3.36 + 04-04 10:37:58 | [630][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0144 ntime: 0081 mem: 3.36 + 04-04 10:38:06 | [630][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1273 ntime: 0078 mem: 3.36 + 04-04 10:38:15 | [630][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0073 ntime: 0078 mem: 3.36 + 04-04 10:38:23 | [630][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1100 ntime: 0081 mem: 3.36 + 04-04 10:38:31 | [630][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0080 mem: 3.36 + 04-04 10:38:41 | [630][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1386 ntime: 0085 mem: 3.36 + 04-04 10:38:50 | [630][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0266 ntime: 0088 mem: 3.36 + 04-04 10:38:58 | [630][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0926 ntime: 0089 mem: 3.36 + 04-04 10:39:07 | [630][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0081 mem: 3.36 + 04-04 10:39:14 | [630][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0081 mem: 3.36 + 04-04 10:39:22 | [630][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0080 mem: 3.36 + 04-04 10:39:30 | [630][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0079 mem: 3.36 + 04-04 10:39:39 | [630][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1298 ntime: 0080 mem: 3.36 + 04-04 10:39:47 | [630][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0062 ntime: 0079 mem: 3.36 + 04-04 10:39:57 | [630][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0111 ntime: 0079 mem: 3.36 + 04-04 10:40:08 | [630][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1375 ntime: 0082 mem: 3.36 + 04-04 10:40:18 | [630][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1429 ntime: 0079 mem: 3.36 + 04-04 10:40:26 | Time info >>>> elapsed: 797.06 mins remain: 466.11 mins + 04-04 10:40:26 | [631][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0077 ntime: 0077 mem: 3.36 + 04-04 10:40:34 | [631][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0094 ntime: 0076 mem: 3.36 + 04-04 10:40:44 | [631][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1015 ntime: 0086 mem: 3.36 + 04-04 10:40:52 | [631][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0116 ntime: 0074 mem: 3.36 + 04-04 10:41:02 | [631][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1413 ntime: 0082 mem: 3.36 + 04-04 10:41:11 | [631][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0105 ntime: 0080 mem: 3.36 + 04-04 10:41:21 | [631][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0087 mem: 3.36 + 04-04 10:41:30 | [631][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0082 mem: 3.36 + 04-04 10:41:39 | [631][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0974 ntime: 0088 mem: 3.36 + 04-04 10:41:47 | [631][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0182 ntime: 0077 mem: 3.36 + 04-04 10:41:54 | [631][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0880 ntime: 0056 mem: 3.36 + 04-04 10:42:04 | [631][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1292 ntime: 0079 mem: 3.36 + 04-04 10:42:14 | [631][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1394 ntime: 0083 mem: 3.36 + 04-04 10:42:23 | [631][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0076 mem: 3.36 + 04-04 10:42:33 | [631][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1193 ntime: 0077 mem: 3.36 + 04-04 10:42:41 | [631][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1400 ntime: 0081 mem: 3.36 + 04-04 10:42:50 | [631][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0085 mem: 3.36 + 04-04 10:42:59 | [631][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1053 ntime: 0081 mem: 3.36 + 04-04 10:43:08 | Time info >>>> elapsed: 799.76 mins remain: 465.68 mins + 04-04 10:43:09 | [632][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0883 ntime: 0077 mem: 3.36 + 04-04 10:43:18 | [632][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1506 ntime: 0086 mem: 3.36 + 04-04 10:43:27 | [632][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1186 ntime: 0078 mem: 3.36 + 04-04 10:43:37 | [632][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1205 ntime: 0084 mem: 3.36 + 04-04 10:43:45 | [632][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0073 mem: 3.36 + 04-04 10:43:55 | [632][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1212 ntime: 0078 mem: 3.36 + 04-04 10:44:06 | [632][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 10:44:15 | [632][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1353 ntime: 0079 mem: 3.36 + 04-04 10:44:22 | [632][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0087 ntime: 0074 mem: 3.36 + 04-04 10:44:32 | [632][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0085 ntime: 0082 mem: 3.36 + 04-04 10:44:42 | [632][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0076 ntime: 0077 mem: 3.36 + 04-04 10:44:51 | [632][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0076 mem: 3.36 + 04-04 10:45:01 | [632][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1472 ntime: 0080 mem: 3.36 + 04-04 10:45:12 | [632][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0103 ntime: 0083 mem: 3.36 + 04-04 10:45:22 | [632][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1382 ntime: 0079 mem: 3.36 + 04-04 10:45:31 | [632][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0199 ntime: 0082 mem: 3.36 + 04-04 10:45:40 | [632][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0065 ntime: 0080 mem: 3.36 + 04-04 10:45:51 | [632][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1545 ntime: 0081 mem: 3.36 + 04-04 10:45:57 | Time info >>>> elapsed: 802.57 mins remain: 465.31 mins + 04-04 10:45:58 | [633][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0800 ntime: 0075 mem: 3.36 + 04-04 10:46:08 | [633][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1419 ntime: 0085 mem: 3.36 + 04-04 10:46:18 | [633][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0125 ntime: 0081 mem: 3.36 + 04-04 10:46:29 | [633][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0586 ntime: 0088 mem: 3.36 + 04-04 10:46:37 | [633][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0400 ntime: 0092 mem: 3.36 + 04-04 10:46:46 | [633][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1501 ntime: 0079 mem: 3.36 + 04-04 10:46:57 | [633][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1207 ntime: 0076 mem: 3.36 + 04-04 10:47:07 | [633][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0085 ntime: 0077 mem: 3.36 + 04-04 10:47:18 | [633][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1334 ntime: 0078 mem: 3.36 + 04-04 10:47:28 | [633][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0848 ntime: 0083 mem: 3.36 + 04-04 10:47:36 | [633][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1430 ntime: 0073 mem: 3.36 + 04-04 10:47:43 | [633][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0164 ntime: 0075 mem: 3.36 + 04-04 10:47:53 | [633][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1403 ntime: 0082 mem: 3.36 + 04-04 10:48:02 | [633][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0081 mem: 3.36 + 04-04 10:48:12 | [633][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1530 ntime: 0079 mem: 3.36 + 04-04 10:48:21 | [633][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1347 ntime: 0089 mem: 3.36 + 04-04 10:48:31 | [633][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1325 ntime: 0081 mem: 3.36 + 04-04 10:48:42 | [633][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1215 ntime: 0080 mem: 3.36 + 04-04 10:48:50 | Time info >>>> elapsed: 805.45 mins remain: 464.98 mins + 04-04 10:48:50 | [634][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0180 ntime: 0076 mem: 3.36 + 04-04 10:49:00 | [634][010/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0107 ntime: 0075 mem: 3.36 + 04-04 10:49:10 | [634][020/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 1460 ntime: 0089 mem: 3.36 + 04-04 10:49:20 | [634][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1527 ntime: 0085 mem: 3.36 + 04-04 10:49:29 | [634][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1433 ntime: 0079 mem: 3.36 + 04-04 10:49:39 | [634][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1316 ntime: 0076 mem: 3.36 + 04-04 10:49:49 | [634][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0139 ntime: 0079 mem: 3.36 + 04-04 10:50:01 | [634][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0800 ntime: 0082 mem: 3.36 + 04-04 10:50:11 | [634][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0116 ntime: 0083 mem: 3.36 + 04-04 10:50:21 | [634][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1329 ntime: 0079 mem: 3.36 + 04-04 10:50:28 | [634][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0084 ntime: 0082 mem: 3.36 + 04-04 10:50:37 | [634][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0560 ntime: 0063 mem: 3.36 + 04-04 10:50:46 | [634][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0953 ntime: 0086 mem: 3.36 + 04-04 10:50:55 | [634][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0084 mem: 3.36 + 04-04 10:51:04 | [634][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0131 ntime: 0079 mem: 3.36 + 04-04 10:51:15 | [634][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1340 ntime: 0079 mem: 3.36 + 04-04 10:51:24 | [634][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1515 ntime: 0083 mem: 3.36 + 04-04 10:51:34 | [634][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1489 ntime: 0077 mem: 3.36 + 04-04 10:51:43 | Time info >>>> elapsed: 808.34 mins remain: 464.64 mins + 04-04 10:51:43 | [635][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0076 mem: 3.36 + 04-04 10:51:53 | [635][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1045 ntime: 0085 mem: 3.36 + 04-04 10:52:03 | [635][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1231 ntime: 0080 mem: 3.36 + 04-04 10:52:12 | [635][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1040 ntime: 0076 mem: 3.36 + 04-04 10:52:23 | [635][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0329 ntime: 0078 mem: 3.36 + 04-04 10:52:33 | [635][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0795 ntime: 0072 mem: 3.36 + 04-04 10:52:44 | [635][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1193 ntime: 0087 mem: 3.36 + 04-04 10:52:52 | [635][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1011 ntime: 0079 mem: 3.36 + 04-04 10:53:00 | [635][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0686 ntime: 0081 mem: 3.36 + 04-04 10:53:08 | [635][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0769 ntime: 0090 mem: 3.36 + 04-04 10:53:16 | [635][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1045 ntime: 0073 mem: 3.36 + 04-04 10:53:25 | [635][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1403 ntime: 0080 mem: 3.36 + 04-04 10:53:32 | [635][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1548 ntime: 0079 mem: 3.36 + 04-04 10:53:39 | [635][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 10:53:47 | [635][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0323 ntime: 0080 mem: 3.36 + 04-04 10:53:54 | [635][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0075 mem: 3.36 + 04-04 10:54:02 | [635][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0086 mem: 3.36 + 04-04 10:54:10 | [635][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0765 ntime: 0077 mem: 3.36 + 04-04 10:54:14 | Time info >>>> elapsed: 810.86 mins remain: 464.08 mins + 04-04 10:54:15 | [636][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1087 ntime: 0082 mem: 3.36 + 04-04 10:54:22 | [636][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0806 ntime: 0080 mem: 3.36 + 04-04 10:54:29 | [636][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0964 ntime: 0073 mem: 3.36 + 04-04 10:54:38 | [636][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1252 ntime: 0079 mem: 3.36 + 04-04 10:54:46 | [636][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1382 ntime: 0087 mem: 3.36 + 04-04 10:54:53 | [636][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0078 mem: 3.36 + 04-04 10:54:59 | [636][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0080 mem: 3.36 + 04-04 10:55:06 | [636][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1224 ntime: 0074 mem: 3.36 + 04-04 10:55:12 | [636][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0077 ntime: 0078 mem: 3.36 + 04-04 10:55:20 | [636][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0864 ntime: 0078 mem: 3.36 + 04-04 10:55:25 | [636][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 10:55:32 | [636][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0119 ntime: 0093 mem: 3.36 + 04-04 10:55:37 | [636][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1147 ntime: 0075 mem: 3.36 + 04-04 10:55:45 | [636][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0077 mem: 3.36 + 04-04 10:55:53 | [636][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1630 ntime: 0070 mem: 3.36 + 04-04 10:56:01 | [636][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1288 ntime: 0081 mem: 3.36 + 04-04 10:56:09 | [636][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1236 ntime: 0075 mem: 3.36 + 04-04 10:56:16 | [636][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0362 ntime: 0080 mem: 3.36 + 04-04 10:56:23 | Time info >>>> elapsed: 813.02 mins remain: 463.30 mins + 04-04 10:56:23 | [637][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0049 ntime: 0077 mem: 3.36 + 04-04 10:56:33 | [637][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0519 ntime: 0074 mem: 3.36 + 04-04 10:56:39 | [637][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1081 ntime: 0074 mem: 3.36 + 04-04 10:56:48 | [637][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1466 ntime: 0080 mem: 3.36 + 04-04 10:56:56 | [637][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0072 ntime: 0078 mem: 3.36 + 04-04 10:57:04 | [637][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0608 ntime: 0077 mem: 3.36 + 04-04 10:57:12 | [637][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1144 ntime: 0078 mem: 3.36 + 04-04 10:57:18 | [637][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0078 mem: 3.36 + 04-04 10:57:26 | [637][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1216 ntime: 0085 mem: 3.36 + 04-04 10:57:32 | [637][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0080 mem: 3.36 + 04-04 10:57:40 | [637][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0532 ntime: 0078 mem: 3.36 + 04-04 10:57:49 | [637][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1461 ntime: 0084 mem: 3.36 + 04-04 10:57:57 | [637][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1118 ntime: 0075 mem: 3.36 + 04-04 10:58:04 | [637][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0237 ntime: 0080 mem: 3.36 + 04-04 10:58:14 | [637][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1204 ntime: 0088 mem: 3.36 + 04-04 10:58:22 | [637][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 10:58:30 | [637][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0420 ntime: 0087 mem: 3.36 + 04-04 10:58:38 | [637][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1430 ntime: 0077 mem: 3.36 + 04-04 10:58:44 | Time info >>>> elapsed: 815.36 mins remain: 462.63 mins + 04-04 10:58:46 | [638][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1630 ntime: 0079 mem: 3.36 + 04-04 10:58:53 | [638][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0622 ntime: 0081 mem: 3.36 + 04-04 10:59:00 | [638][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1603 ntime: 0080 mem: 3.36 + 04-04 10:59:09 | [638][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1007 ntime: 0061 mem: 3.36 + 04-04 10:59:16 | [638][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0082 mem: 3.36 + 04-04 10:59:25 | [638][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0869 ntime: 0080 mem: 3.36 + 04-04 10:59:31 | [638][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0173 ntime: 0088 mem: 3.36 + 04-04 10:59:40 | [638][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1332 ntime: 0074 mem: 3.36 + 04-04 10:59:47 | [638][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0223 ntime: 0081 mem: 3.36 + 04-04 10:59:53 | [638][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0081 ntime: 0081 mem: 3.36 + 04-04 11:00:01 | [638][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0081 ntime: 0071 mem: 3.36 + 04-04 11:00:07 | [638][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0088 mem: 3.36 + 04-04 11:00:12 | [638][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0075 ntime: 0079 mem: 3.36 + 04-04 11:00:21 | [638][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1414 ntime: 0078 mem: 3.36 + 04-04 11:00:27 | [638][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0068 mem: 3.36 + 04-04 11:00:34 | [638][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1119 ntime: 0080 mem: 3.36 + 04-04 11:00:41 | [638][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0076 mem: 3.36 + 04-04 11:00:47 | [638][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0394 ntime: 0090 mem: 3.36 + 04-04 11:00:52 | Time info >>>> elapsed: 817.49 mins remain: 461.84 mins + 04-04 11:00:53 | [639][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0690 ntime: 0078 mem: 3.36 + 04-04 11:00:58 | [639][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0084 mem: 3.36 + 04-04 11:01:04 | [639][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0325 ntime: 0079 mem: 3.36 + 04-04 11:01:11 | [639][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0065 ntime: 0086 mem: 3.36 + 04-04 11:01:19 | [639][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1186 ntime: 0080 mem: 3.36 + 04-04 11:01:26 | [639][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1227 ntime: 0077 mem: 3.36 + 04-04 11:01:33 | [639][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0079 mem: 3.36 + 04-04 11:01:40 | [639][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0075 mem: 3.36 + 04-04 11:01:49 | [639][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1206 ntime: 0080 mem: 3.36 + 04-04 11:01:56 | [639][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0958 ntime: 0088 mem: 3.36 + 04-04 11:02:03 | [639][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1454 ntime: 0078 mem: 3.36 + 04-04 11:02:10 | [639][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0087 ntime: 0081 mem: 3.36 + 04-04 11:02:17 | [639][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0759 ntime: 0078 mem: 3.36 + 04-04 11:02:24 | [639][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1205 ntime: 0078 mem: 3.36 + 04-04 11:02:32 | [639][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0227 ntime: 0079 mem: 3.36 + 04-04 11:02:40 | [639][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1823 ntime: 0074 mem: 3.36 + 04-04 11:02:47 | [639][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0065 ntime: 0090 mem: 3.36 + 04-04 11:02:53 | [639][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0085 mem: 3.36 + 04-04 11:03:01 | Time info >>>> elapsed: 819.65 mins remain: 461.05 mins + 04-04 11:03:02 | [640][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0470 ntime: 0087 mem: 3.36 + 04-04 11:03:09 | [640][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0894 ntime: 0077 mem: 3.36 + 04-04 11:03:16 | [640][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1085 ntime: 0078 mem: 3.36 + 04-04 11:03:23 | [640][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0093 ntime: 0080 mem: 3.36 + 04-04 11:03:30 | [640][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0880 ntime: 0081 mem: 3.36 + 04-04 11:03:36 | [640][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1109 ntime: 0080 mem: 3.36 + 04-04 11:03:42 | [640][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1266 ntime: 0080 mem: 3.36 + 04-04 11:03:49 | [640][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0772 ntime: 0079 mem: 3.36 + 04-04 11:03:55 | [640][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1399 ntime: 0078 mem: 3.36 + 04-04 11:04:04 | [640][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0854 ntime: 0079 mem: 3.36 + 04-04 11:04:13 | [640][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1580 ntime: 0077 mem: 3.36 + 04-04 11:04:22 | [640][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0968 ntime: 0083 mem: 3.36 + 04-04 11:04:32 | [640][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0081 ntime: 0083 mem: 3.36 + 04-04 11:04:40 | [640][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0186 ntime: 0076 mem: 3.36 + 04-04 11:04:46 | [640][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0532 ntime: 0074 mem: 3.36 + 04-04 11:04:56 | [640][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1219 ntime: 0082 mem: 3.36 + 04-04 11:05:04 | [640][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0082 mem: 3.36 + 04-04 11:05:11 | [640][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1121 ntime: 0077 mem: 3.36 + 04-04 11:05:17 | Time info >>>> elapsed: 821.92 mins remain: 460.32 mins + 04-04 11:05:18 | [641][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0856 ntime: 0078 mem: 3.36 + 04-04 11:05:27 | [641][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1409 ntime: 0081 mem: 3.36 + 04-04 11:05:34 | [641][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1045 ntime: 0080 mem: 3.36 + 04-04 11:05:42 | [641][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1145 ntime: 0082 mem: 3.36 + 04-04 11:05:47 | [641][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0535 ntime: 0080 mem: 3.36 + 04-04 11:05:51 | [641][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0080 mem: 3.36 + 04-04 11:05:59 | [641][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1469 ntime: 0083 mem: 3.36 + 04-04 11:06:05 | [641][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0065 ntime: 0082 mem: 3.36 + 04-04 11:06:11 | [641][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0063 ntime: 0083 mem: 3.36 + 04-04 11:06:20 | [641][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0968 ntime: 0090 mem: 3.36 + 04-04 11:06:25 | [641][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0082 mem: 3.36 + 04-04 11:06:32 | [641][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0743 ntime: 0083 mem: 3.36 + 04-04 11:06:42 | [641][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0610 ntime: 0089 mem: 3.36 + 04-04 11:06:50 | [641][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1715 ntime: 0080 mem: 3.36 + 04-04 11:06:58 | [641][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1366 ntime: 0077 mem: 3.36 + 04-04 11:07:05 | [641][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0081 mem: 3.36 + 04-04 11:07:13 | [641][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1343 ntime: 0086 mem: 3.36 + 04-04 11:07:21 | [641][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0732 ntime: 0073 mem: 3.36 + 04-04 11:07:27 | Time info >>>> elapsed: 824.08 mins remain: 459.53 mins + 04-04 11:07:27 | [642][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0118 ntime: 0081 mem: 3.36 + 04-04 11:07:33 | [642][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 11:07:41 | [642][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0074 ntime: 0084 mem: 3.36 + 04-04 11:07:47 | [642][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1450 ntime: 0080 mem: 3.36 + 04-04 11:07:54 | [642][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0060 mem: 3.36 + 04-04 11:08:04 | [642][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0663 ntime: 0070 mem: 3.36 + 04-04 11:08:11 | [642][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0109 ntime: 0082 mem: 3.36 + 04-04 11:08:17 | [642][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0193 ntime: 0077 mem: 3.36 + 04-04 11:08:26 | [642][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0870 ntime: 0078 mem: 3.36 + 04-04 11:08:33 | [642][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1279 ntime: 0082 mem: 3.36 + 04-04 11:08:41 | [642][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1351 ntime: 0086 mem: 3.36 + 04-04 11:08:49 | [642][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1139 ntime: 0077 mem: 3.36 + 04-04 11:08:54 | [642][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0095 ntime: 0079 mem: 3.36 + 04-04 11:09:00 | [642][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0768 ntime: 0081 mem: 3.36 + 04-04 11:09:05 | [642][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0903 ntime: 0076 mem: 3.36 + 04-04 11:09:12 | [642][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1224 ntime: 0080 mem: 3.36 + 04-04 11:09:18 | [642][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0086 ntime: 0084 mem: 3.36 + 04-04 11:09:26 | [642][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0074 mem: 3.36 + 04-04 11:09:31 | Time info >>>> elapsed: 826.14 mins remain: 458.68 mins + 04-04 11:09:31 | [643][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0061 ntime: 0086 mem: 3.36 + 04-04 11:09:37 | [643][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0081 mem: 3.36 + 04-04 11:09:44 | [643][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1037 ntime: 0081 mem: 3.36 + 04-04 11:09:51 | [643][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0076 ntime: 0086 mem: 3.36 + 04-04 11:09:59 | [643][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0100 ntime: 0078 mem: 3.36 + 04-04 11:10:07 | [643][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1075 ntime: 0082 mem: 3.36 + 04-04 11:10:16 | [643][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0122 ntime: 0076 mem: 3.36 + 04-04 11:10:22 | [643][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0862 ntime: 0076 mem: 3.36 + 04-04 11:10:29 | [643][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1129 ntime: 0075 mem: 3.36 + 04-04 11:10:36 | [643][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1505 ntime: 0085 mem: 3.36 + 04-04 11:10:44 | [643][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1220 ntime: 0081 mem: 3.36 + 04-04 11:10:52 | [643][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0769 ntime: 0080 mem: 3.36 + 04-04 11:11:00 | [643][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1238 ntime: 0078 mem: 3.36 + 04-04 11:11:06 | [643][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0740 ntime: 0080 mem: 3.36 + 04-04 11:11:12 | [643][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1346 ntime: 0079 mem: 3.36 + 04-04 11:11:20 | [643][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0087 mem: 3.36 + 04-04 11:11:26 | [643][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0159 ntime: 0086 mem: 3.36 + 04-04 11:11:34 | [643][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1290 ntime: 0080 mem: 3.36 + 04-04 11:11:38 | Time info >>>> elapsed: 828.27 mins remain: 457.86 mins + 04-04 11:11:40 | [644][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 1332 ntime: 0092 mem: 3.36 + 04-04 11:11:46 | [644][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0155 ntime: 0082 mem: 3.36 + 04-04 11:11:53 | [644][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1395 ntime: 0080 mem: 3.36 + 04-04 11:12:01 | [644][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0653 ntime: 0073 mem: 3.36 + 04-04 11:12:07 | [644][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0078 mem: 3.36 + 04-04 11:12:15 | [644][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0115 ntime: 0084 mem: 3.36 + 04-04 11:12:23 | [644][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1261 ntime: 0081 mem: 3.36 + 04-04 11:12:29 | [644][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0085 mem: 3.36 + 04-04 11:12:37 | [644][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0767 ntime: 0077 mem: 3.36 + 04-04 11:12:44 | [644][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1147 ntime: 0081 mem: 3.36 + 04-04 11:12:53 | [644][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0881 ntime: 0082 mem: 3.36 + 04-04 11:12:59 | [644][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0108 ntime: 0079 mem: 3.36 + 04-04 11:13:07 | [644][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0923 ntime: 0083 mem: 3.36 + 04-04 11:13:13 | [644][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1224 ntime: 0080 mem: 3.36 + 04-04 11:13:20 | [644][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0115 ntime: 0078 mem: 3.36 + 04-04 11:13:28 | [644][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0345 ntime: 0081 mem: 3.36 + 04-04 11:13:37 | [644][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1478 ntime: 0079 mem: 3.36 + 04-04 11:13:42 | [644][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0267 ntime: 0094 mem: 3.36 + 04-04 11:13:47 | Time info >>>> elapsed: 830.42 mins remain: 457.05 mins + 04-04 11:13:47 | [645][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0057 ntime: 0080 mem: 3.36 + 04-04 11:13:56 | [645][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1441 ntime: 0083 mem: 3.36 + 04-04 11:14:02 | [645][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 11:14:09 | [645][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0889 ntime: 0077 mem: 3.36 + 04-04 11:14:16 | [645][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0678 ntime: 0090 mem: 3.36 + 04-04 11:14:23 | [645][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1464 ntime: 0077 mem: 3.36 + 04-04 11:14:30 | [645][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0077 mem: 3.36 + 04-04 11:14:40 | [645][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1116 ntime: 0083 mem: 3.36 + 04-04 11:14:47 | [645][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1183 ntime: 0077 mem: 3.36 + 04-04 11:14:54 | [645][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0244 ntime: 0088 mem: 3.36 + 04-04 11:15:00 | [645][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0098 ntime: 0080 mem: 3.36 + 04-04 11:15:09 | [645][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0067 ntime: 0079 mem: 3.36 + 04-04 11:15:16 | [645][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0820 ntime: 0081 mem: 3.36 + 04-04 11:15:22 | [645][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0139 ntime: 0079 mem: 3.36 + 04-04 11:15:28 | [645][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0989 ntime: 0080 mem: 3.36 + 04-04 11:15:34 | [645][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0183 ntime: 0085 mem: 3.36 + 04-04 11:15:41 | [645][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0685 ntime: 0077 mem: 3.36 + 04-04 11:15:49 | [645][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0133 ntime: 0081 mem: 3.36 + 04-04 11:15:54 | Time info >>>> elapsed: 832.52 mins remain: 456.21 mins + 04-04 11:15:54 | [646][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0076 ntime: 0080 mem: 3.36 + 04-04 11:16:01 | [646][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0090 mem: 3.36 + 04-04 11:16:10 | [646][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1065 ntime: 0086 mem: 3.36 + 04-04 11:16:16 | [646][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0695 ntime: 0084 mem: 3.36 + 04-04 11:16:23 | [646][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0634 ntime: 0086 mem: 3.36 + 04-04 11:16:28 | [646][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0552 ntime: 0084 mem: 3.36 + 04-04 11:16:36 | [646][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0179 ntime: 0081 mem: 3.36 + 04-04 11:16:44 | [646][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0155 ntime: 0062 mem: 3.36 + 04-04 11:16:52 | [646][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1166 ntime: 0076 mem: 3.36 + 04-04 11:16:57 | [646][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0078 ntime: 0080 mem: 3.36 + 04-04 11:17:05 | [646][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1125 ntime: 0086 mem: 3.36 + 04-04 11:17:11 | [646][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0627 ntime: 0079 mem: 3.36 + 04-04 11:17:18 | [646][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0296 ntime: 0087 mem: 3.36 + 04-04 11:17:26 | [646][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0924 ntime: 0087 mem: 3.36 + 04-04 11:17:32 | [646][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0118 ntime: 0082 mem: 3.36 + 04-04 11:17:41 | [646][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1059 ntime: 0077 mem: 3.36 + 04-04 11:17:48 | [646][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0148 ntime: 0081 mem: 3.36 + 04-04 11:17:56 | [646][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0082 mem: 3.36 + 04-04 11:18:01 | Time info >>>> elapsed: 834.65 mins remain: 455.38 mins + 04-04 11:18:02 | [647][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0763 ntime: 0083 mem: 3.36 + 04-04 11:18:09 | [647][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1103 ntime: 0086 mem: 3.36 + 04-04 11:18:16 | [647][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1046 ntime: 0086 mem: 3.36 + 04-04 11:18:21 | [647][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0081 mem: 3.36 + 04-04 11:18:29 | [647][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0377 ntime: 0079 mem: 3.36 + 04-04 11:18:37 | [647][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0120 ntime: 0080 mem: 3.36 + 04-04 11:18:45 | [647][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1082 ntime: 0083 mem: 3.36 + 04-04 11:18:51 | [647][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1276 ntime: 0081 mem: 3.36 + 04-04 11:18:58 | [647][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1241 ntime: 0077 mem: 3.36 + 04-04 11:19:06 | [647][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1236 ntime: 0081 mem: 3.36 + 04-04 11:19:14 | [647][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0889 ntime: 0090 mem: 3.36 + 04-04 11:19:21 | [647][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0768 ntime: 0079 mem: 3.36 + 04-04 11:19:26 | [647][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 11:19:33 | [647][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0078 mem: 3.36 + 04-04 11:19:40 | [647][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1523 ntime: 0088 mem: 3.36 + 04-04 11:19:49 | [647][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0523 ntime: 0074 mem: 3.36 + 04-04 11:19:56 | [647][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1586 ntime: 0085 mem: 3.36 + 04-04 11:20:05 | [647][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1092 ntime: 0080 mem: 3.36 + 04-04 11:20:11 | Time info >>>> elapsed: 836.81 mins remain: 454.56 mins + 04-04 11:20:12 | [648][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0832 ntime: 0081 mem: 3.36 + 04-04 11:20:20 | [648][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0860 ntime: 0082 mem: 3.36 + 04-04 11:20:27 | [648][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0046 ntime: 0082 mem: 3.36 + 04-04 11:20:33 | [648][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0099 ntime: 0079 mem: 3.36 + 04-04 11:20:41 | [648][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0080 mem: 3.36 + 04-04 11:20:48 | [648][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0081 ntime: 0079 mem: 3.36 + 04-04 11:20:55 | [648][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1392 ntime: 0083 mem: 3.36 + 04-04 11:20:59 | [648][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0094 ntime: 0081 mem: 3.36 + 04-04 11:21:08 | [648][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0981 ntime: 0087 mem: 3.36 + 04-04 11:21:15 | [648][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0114 ntime: 0082 mem: 3.36 + 04-04 11:21:21 | [648][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0572 ntime: 0080 mem: 3.36 + 04-04 11:21:29 | [648][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0807 ntime: 0082 mem: 3.36 + 04-04 11:21:35 | [648][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0710 ntime: 0076 mem: 3.36 + 04-04 11:21:44 | [648][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0950 ntime: 0085 mem: 3.36 + 04-04 11:21:50 | [648][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0079 mem: 3.36 + 04-04 11:21:58 | [648][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1113 ntime: 0080 mem: 3.36 + 04-04 11:22:05 | [648][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0994 ntime: 0082 mem: 3.36 + 04-04 11:22:11 | [648][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0253 ntime: 0081 mem: 3.36 + 04-04 11:22:17 | Time info >>>> elapsed: 838.91 mins remain: 453.71 mins + 04-04 11:22:17 | [649][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0067 ntime: 0081 mem: 3.36 + 04-04 11:22:25 | [649][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0224 ntime: 0088 mem: 3.36 + 04-04 11:22:33 | [649][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0441 ntime: 0085 mem: 3.36 + 04-04 11:22:40 | [649][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1565 ntime: 0075 mem: 3.36 + 04-04 11:22:46 | [649][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1018 ntime: 0084 mem: 3.36 + 04-04 11:22:54 | [649][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0332 ntime: 0078 mem: 3.36 + 04-04 11:23:01 | [649][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0105 ntime: 0079 mem: 3.36 + 04-04 11:23:08 | [649][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0112 ntime: 0077 mem: 3.36 + 04-04 11:23:12 | [649][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0145 ntime: 0080 mem: 3.36 + 04-04 11:23:18 | [649][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0098 ntime: 0084 mem: 3.36 + 04-04 11:23:24 | [649][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 11:23:33 | [649][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0986 ntime: 0080 mem: 3.36 + 04-04 11:23:40 | [649][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1000 ntime: 0090 mem: 3.36 + 04-04 11:23:47 | [649][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0666 ntime: 0081 mem: 3.36 + 04-04 11:23:57 | [649][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1131 ntime: 0089 mem: 3.36 + 04-04 11:24:04 | [649][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0961 ntime: 0080 mem: 3.36 + 04-04 11:24:10 | [649][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0084 mem: 3.36 + 04-04 11:24:16 | [649][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0404 ntime: 0079 mem: 3.36 + 04-04 11:24:21 | Time info >>>> elapsed: 840.98 mins remain: 452.84 mins + 04-04 11:24:22 | [650][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0780 ntime: 0085 mem: 3.36 + 04-04 11:24:29 | [650][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0121 ntime: 0078 mem: 3.36 + 04-04 11:24:37 | [650][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1626 ntime: 0084 mem: 3.36 + 04-04 11:24:47 | [650][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0878 ntime: 0084 mem: 3.36 + 04-04 11:24:52 | [650][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0362 ntime: 0089 mem: 3.36 + 04-04 11:24:57 | [650][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0611 ntime: 0079 mem: 3.36 + 04-04 11:25:03 | [650][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0821 ntime: 0083 mem: 3.36 + 04-04 11:25:11 | [650][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0988 ntime: 0083 mem: 3.36 + 04-04 11:25:16 | [650][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0799 ntime: 0082 mem: 3.36 + 04-04 11:25:23 | [650][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0449 ntime: 0078 mem: 3.36 + 04-04 11:25:30 | [650][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1311 ntime: 0070 mem: 3.36 + 04-04 11:25:37 | [650][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0207 ntime: 0082 mem: 3.36 + 04-04 11:25:44 | [650][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1479 ntime: 0076 mem: 3.36 + 04-04 11:25:49 | [650][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0077 mem: 3.36 + 04-04 11:25:58 | [650][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1147 ntime: 0081 mem: 3.36 + 04-04 11:26:03 | [650][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1104 ntime: 0084 mem: 3.36 + 04-04 11:26:11 | [650][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1047 ntime: 0083 mem: 3.36 + 04-04 11:26:18 | [650][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0075 mem: 3.36 + 04-04 11:26:23 | Time info >>>> elapsed: 843.01 mins remain: 451.94 mins + 04-04 11:26:24 | [651][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0922 ntime: 0079 mem: 3.36 + 04-04 11:26:30 | [651][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0092 ntime: 0076 mem: 3.36 + 04-04 11:26:36 | [651][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0144 ntime: 0070 mem: 3.36 + 04-04 11:26:42 | [651][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0450 ntime: 0080 mem: 3.36 + 04-04 11:26:50 | [651][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0553 ntime: 0082 mem: 3.36 + 04-04 11:26:55 | [651][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0078 ntime: 0083 mem: 3.36 + 04-04 11:27:03 | [651][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1345 ntime: 0087 mem: 3.36 + 04-04 11:27:09 | [651][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0910 ntime: 0083 mem: 3.36 + 04-04 11:27:15 | [651][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0077 mem: 3.36 + 04-04 11:27:20 | [651][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0086 mem: 3.36 + 04-04 11:27:26 | [651][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0852 ntime: 0077 mem: 3.36 + 04-04 11:27:32 | [651][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1385 ntime: 0086 mem: 3.36 + 04-04 11:27:39 | [651][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1218 ntime: 0086 mem: 3.36 + 04-04 11:27:44 | [651][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0086 mem: 3.36 + 04-04 11:27:52 | [651][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0167 ntime: 0080 mem: 3.36 + 04-04 11:27:59 | [651][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1165 ntime: 0080 mem: 3.36 + 04-04 11:28:06 | [651][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0123 ntime: 0074 mem: 3.36 + 04-04 11:28:14 | [651][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1730 ntime: 0078 mem: 3.36 + 04-04 11:28:20 | Time info >>>> elapsed: 844.97 mins remain: 450.99 mins + 04-04 11:28:20 | [652][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 11:28:27 | [652][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0079 ntime: 0085 mem: 3.36 + 04-04 11:28:35 | [652][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0092 ntime: 0082 mem: 3.36 + 04-04 11:28:42 | [652][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0603 ntime: 0076 mem: 3.36 + 04-04 11:28:49 | [652][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0449 ntime: 0087 mem: 3.36 + 04-04 11:28:57 | [652][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0077 mem: 3.36 + 04-04 11:29:02 | [652][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0602 ntime: 0088 mem: 3.36 + 04-04 11:29:09 | [652][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0859 ntime: 0078 mem: 3.36 + 04-04 11:29:14 | [652][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0876 ntime: 0074 mem: 3.36 + 04-04 11:29:20 | [652][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1295 ntime: 0085 mem: 3.36 + 04-04 11:29:27 | [652][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0087 ntime: 0086 mem: 3.36 + 04-04 11:29:33 | [652][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0696 ntime: 0071 mem: 3.36 + 04-04 11:29:40 | [652][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0135 ntime: 0085 mem: 3.36 + 04-04 11:29:47 | [652][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1125 ntime: 0083 mem: 3.36 + 04-04 11:29:54 | [652][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0146 ntime: 0079 mem: 3.36 + 04-04 11:30:00 | [652][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0235 ntime: 0081 mem: 3.36 + 04-04 11:30:06 | [652][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0231 ntime: 0077 mem: 3.36 + 04-04 11:30:12 | [652][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0452 ntime: 0081 mem: 3.36 + 04-04 11:30:19 | Time info >>>> elapsed: 846.95 mins remain: 450.06 mins + 04-04 11:30:20 | [653][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0084 ntime: 0083 mem: 3.36 + 04-04 11:30:26 | [653][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 11:30:35 | [653][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1259 ntime: 0081 mem: 3.36 + 04-04 11:30:43 | [653][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0731 ntime: 0079 mem: 3.36 + 04-04 11:30:51 | [653][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0090 ntime: 0081 mem: 3.36 + 04-04 11:30:57 | [653][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0118 ntime: 0079 mem: 3.36 + 04-04 11:31:04 | [653][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0077 mem: 3.36 + 04-04 11:31:11 | [653][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0128 ntime: 0081 mem: 3.36 + 04-04 11:31:19 | [653][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0373 ntime: 0080 mem: 3.36 + 04-04 11:31:25 | [653][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1101 ntime: 0085 mem: 3.36 + 04-04 11:31:31 | [653][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0078 mem: 3.36 + 04-04 11:31:37 | [653][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0987 ntime: 0087 mem: 3.36 + 04-04 11:31:45 | [653][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0141 ntime: 0084 mem: 3.36 + 04-04 11:31:50 | [653][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0099 ntime: 0085 mem: 3.36 + 04-04 11:31:57 | [653][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0529 ntime: 0085 mem: 3.36 + 04-04 11:32:05 | [653][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0742 ntime: 0091 mem: 3.36 + 04-04 11:32:11 | [653][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1477 ntime: 0076 mem: 3.36 + 04-04 11:32:18 | [653][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0074 mem: 3.36 + 04-04 11:32:24 | Time info >>>> elapsed: 849.02 mins remain: 449.18 mins + 04-04 11:32:24 | [654][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0843 ntime: 0087 mem: 3.36 + 04-04 11:32:30 | [654][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0201 ntime: 0080 mem: 3.36 + 04-04 11:32:38 | [654][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1118 ntime: 0074 mem: 3.36 + 04-04 11:32:44 | [654][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0114 ntime: 0082 mem: 3.36 + 04-04 11:32:51 | [654][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0099 ntime: 0077 mem: 3.36 + 04-04 11:32:59 | [654][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1225 ntime: 0081 mem: 3.36 + 04-04 11:33:06 | [654][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0116 ntime: 0079 mem: 3.36 + 04-04 11:33:12 | [654][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1321 ntime: 0082 mem: 3.36 + 04-04 11:33:18 | [654][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0647 ntime: 0080 mem: 3.36 + 04-04 11:33:24 | [654][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0070 mem: 3.36 + 04-04 11:33:30 | [654][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0604 ntime: 0080 mem: 3.36 + 04-04 11:33:37 | [654][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0850 ntime: 0082 mem: 3.36 + 04-04 11:33:43 | [654][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1155 ntime: 0083 mem: 3.36 + 04-04 11:33:50 | [654][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0166 ntime: 0085 mem: 3.36 + 04-04 11:33:58 | [654][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1394 ntime: 0073 mem: 3.36 + 04-04 11:34:04 | [654][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1073 ntime: 0076 mem: 3.36 + 04-04 11:34:12 | [654][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1465 ntime: 0079 mem: 3.36 + 04-04 11:34:18 | [654][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0125 ntime: 0080 mem: 3.36 + 04-04 11:34:23 | Time info >>>> elapsed: 851.01 mins remain: 448.24 mins + 04-04 11:34:24 | [655][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0896 ntime: 0077 mem: 3.36 + 04-04 11:34:30 | [655][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0787 ntime: 0079 mem: 3.36 + 04-04 11:34:37 | [655][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 11:34:43 | [655][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0074 mem: 3.36 + 04-04 11:34:49 | [655][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0155 ntime: 0079 mem: 3.36 + 04-04 11:34:54 | [655][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0085 ntime: 0080 mem: 3.36 + 04-04 11:35:02 | [655][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0851 ntime: 0079 mem: 3.36 + 04-04 11:35:08 | [655][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0092 ntime: 0083 mem: 3.36 + 04-04 11:35:14 | [655][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0533 ntime: 0076 mem: 3.36 + 04-04 11:35:23 | [655][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1053 ntime: 0084 mem: 3.36 + 04-04 11:35:31 | [655][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1233 ntime: 0081 mem: 3.36 + 04-04 11:35:38 | [655][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0192 ntime: 0081 mem: 3.36 + 04-04 11:35:45 | [655][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1058 ntime: 0082 mem: 3.36 + 04-04 11:35:52 | [655][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1668 ntime: 0080 mem: 3.36 + 04-04 11:35:59 | [655][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0224 ntime: 0084 mem: 3.36 + 04-04 11:36:05 | [655][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0092 ntime: 0080 mem: 3.36 + 04-04 11:36:12 | [655][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0127 ntime: 0079 mem: 3.36 + 04-04 11:36:17 | [655][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0550 ntime: 0078 mem: 3.36 + 04-04 11:36:23 | Time info >>>> elapsed: 853.01 mins remain: 447.31 mins + 04-04 11:36:23 | [656][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0169 ntime: 0072 mem: 3.36 + 04-04 11:36:29 | [656][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0114 ntime: 0079 mem: 3.36 + 04-04 11:36:36 | [656][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0072 mem: 3.36 + 04-04 11:36:44 | [656][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0814 ntime: 0080 mem: 3.36 + 04-04 11:36:52 | [656][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0905 ntime: 0080 mem: 3.36 + 04-04 11:36:58 | [656][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0080 mem: 3.36 + 04-04 11:37:06 | [656][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1431 ntime: 0085 mem: 3.36 + 04-04 11:37:14 | [656][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1358 ntime: 0085 mem: 3.36 + 04-04 11:37:23 | [656][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1568 ntime: 0080 mem: 3.36 + 04-04 11:37:29 | [656][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0458 ntime: 0075 mem: 3.36 + 04-04 11:37:35 | [656][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1047 ntime: 0077 mem: 3.36 + 04-04 11:37:42 | [656][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1535 ntime: 0078 mem: 3.36 + 04-04 11:37:49 | [656][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0489 ntime: 0081 mem: 3.36 + 04-04 11:37:56 | [656][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1066 ntime: 0082 mem: 3.36 + 04-04 11:38:03 | [656][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1107 ntime: 0077 mem: 3.36 + 04-04 11:38:10 | [656][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0801 ntime: 0082 mem: 3.36 + 04-04 11:38:17 | [656][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0869 ntime: 0080 mem: 3.36 + 04-04 11:38:24 | [656][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0607 ntime: 0078 mem: 3.36 + 04-04 11:38:30 | Time info >>>> elapsed: 855.13 mins remain: 446.44 mins + 04-04 11:38:31 | [657][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 1080 ntime: 0074 mem: 3.36 + 04-04 11:38:39 | [657][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0284 ntime: 0078 mem: 3.36 + 04-04 11:38:45 | [657][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1056 ntime: 0075 mem: 3.36 + 04-04 11:38:52 | [657][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1548 ntime: 0080 mem: 3.36 + 04-04 11:39:01 | [657][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1022 ntime: 0084 mem: 3.36 + 04-04 11:39:09 | [657][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0521 ntime: 0083 mem: 3.36 + 04-04 11:39:15 | [657][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0063 ntime: 0084 mem: 3.36 + 04-04 11:39:23 | [657][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 11:39:31 | [657][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0265 ntime: 0077 mem: 3.36 + 04-04 11:39:38 | [657][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 11:39:45 | [657][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0080 mem: 3.36 + 04-04 11:39:54 | [657][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1031 ntime: 0077 mem: 3.36 + 04-04 11:40:02 | [657][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1524 ntime: 0086 mem: 3.36 + 04-04 11:40:09 | [657][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0660 ntime: 0055 mem: 3.36 + 04-04 11:40:15 | [657][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0437 ntime: 0063 mem: 3.36 + 04-04 11:40:22 | [657][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0860 ntime: 0071 mem: 3.36 + 04-04 11:40:28 | [657][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0635 ntime: 0083 mem: 3.36 + 04-04 11:40:36 | [657][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1574 ntime: 0074 mem: 3.36 + 04-04 11:40:42 | Time info >>>> elapsed: 857.33 mins remain: 445.60 mins + 04-04 11:40:42 | [658][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0077 ntime: 0073 mem: 3.36 + 04-04 11:40:50 | [658][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0082 ntime: 0080 mem: 3.36 + 04-04 11:40:56 | [658][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0526 ntime: 0081 mem: 3.36 + 04-04 11:41:02 | [658][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0231 ntime: 0085 mem: 3.36 + 04-04 11:41:09 | [658][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1539 ntime: 0083 mem: 3.36 + 04-04 11:41:15 | [658][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0435 ntime: 0083 mem: 3.36 + 04-04 11:41:22 | [658][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0936 ntime: 0081 mem: 3.36 + 04-04 11:41:30 | [658][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0819 ntime: 0079 mem: 3.36 + 04-04 11:41:37 | [658][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0100 ntime: 0084 mem: 3.36 + 04-04 11:41:43 | [658][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0079 mem: 3.36 + 04-04 11:41:52 | [658][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1942 ntime: 0083 mem: 3.36 + 04-04 11:41:59 | [658][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0952 ntime: 0079 mem: 3.36 + 04-04 11:42:07 | [658][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1369 ntime: 0086 mem: 3.36 + 04-04 11:42:12 | [658][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0861 ntime: 0080 mem: 3.36 + 04-04 11:42:20 | [658][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0886 ntime: 0088 mem: 3.36 + 04-04 11:42:25 | [658][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0062 ntime: 0080 mem: 3.36 + 04-04 11:42:32 | [658][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0729 ntime: 0074 mem: 3.36 + 04-04 11:42:39 | [658][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0777 ntime: 0080 mem: 3.36 + 04-04 11:42:44 | Time info >>>> elapsed: 859.36 mins remain: 444.67 mins + 04-04 11:42:44 | [659][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0172 ntime: 0077 mem: 3.36 + 04-04 11:42:52 | [659][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0089 ntime: 0062 mem: 3.36 + 04-04 11:42:59 | [659][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0062 ntime: 0078 mem: 3.36 + 04-04 11:43:07 | [659][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0138 ntime: 0082 mem: 3.36 + 04-04 11:43:12 | [659][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0614 ntime: 0085 mem: 3.36 + 04-04 11:43:20 | [659][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0227 ntime: 0075 mem: 3.36 + 04-04 11:43:27 | [659][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0107 ntime: 0080 mem: 3.36 + 04-04 11:43:34 | [659][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0503 ntime: 0075 mem: 3.36 + 04-04 11:43:42 | [659][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1595 ntime: 0083 mem: 3.36 + 04-04 11:43:49 | [659][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0074 mem: 3.36 + 04-04 11:43:54 | [659][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0161 ntime: 0080 mem: 3.36 + 04-04 11:44:00 | [659][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0110 ntime: 0077 mem: 3.36 + 04-04 11:44:09 | [659][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0613 ntime: 0079 mem: 3.36 + 04-04 11:44:16 | [659][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1185 ntime: 0081 mem: 3.36 + 04-04 11:44:22 | [659][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0215 ntime: 0083 mem: 3.36 + 04-04 11:44:31 | [659][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0165 ntime: 0086 mem: 3.36 + 04-04 11:44:38 | [659][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1515 ntime: 0087 mem: 3.36 + 04-04 11:44:45 | [659][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0933 ntime: 0072 mem: 3.36 + 04-04 11:44:50 | Time info >>>> elapsed: 861.47 mins remain: 443.79 mins + 04-04 11:44:51 | [660][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0708 ntime: 0075 mem: 3.36 + 04-04 11:44:56 | [660][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0641 ntime: 0084 mem: 3.36 + 04-04 11:45:03 | [660][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0722 ntime: 0078 mem: 3.36 + 04-04 11:45:09 | [660][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0109 ntime: 0077 mem: 3.36 + 04-04 11:45:14 | [660][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1543 ntime: 0080 mem: 3.36 + 04-04 11:45:21 | [660][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1337 ntime: 0084 mem: 3.36 + 04-04 11:45:27 | [660][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0103 ntime: 0080 mem: 3.36 + 04-04 11:45:33 | [660][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 11:45:39 | [660][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0978 ntime: 0076 mem: 3.36 + 04-04 11:45:47 | [660][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0082 mem: 3.36 + 04-04 11:45:55 | [660][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1017 ntime: 0079 mem: 3.36 + 04-04 11:46:03 | [660][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1492 ntime: 0069 mem: 3.36 + 04-04 11:46:12 | [660][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1010 ntime: 0089 mem: 3.36 + 04-04 11:46:20 | [660][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1508 ntime: 0079 mem: 3.36 + 04-04 11:46:25 | [660][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0929 ntime: 0081 mem: 3.36 + 04-04 11:46:32 | [660][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0997 ntime: 0088 mem: 3.36 + 04-04 11:46:38 | [660][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1149 ntime: 0079 mem: 3.36 + 04-04 11:46:46 | [660][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1078 ntime: 0079 mem: 3.36 + 04-04 11:46:52 | Time info >>>> elapsed: 863.49 mins remain: 442.85 mins + 04-04 11:46:52 | [661][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0083 mem: 3.36 + 04-04 11:47:00 | [661][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1431 ntime: 0079 mem: 3.36 + 04-04 11:47:08 | [661][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1224 ntime: 0085 mem: 3.36 + 04-04 11:47:14 | [661][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0743 ntime: 0080 mem: 3.36 + 04-04 11:47:22 | [661][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0087 ntime: 0075 mem: 3.36 + 04-04 11:47:29 | [661][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1277 ntime: 0083 mem: 3.36 + 04-04 11:47:35 | [661][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0464 ntime: 0080 mem: 3.36 + 04-04 11:47:42 | [661][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0837 ntime: 0073 mem: 3.36 + 04-04 11:47:48 | [661][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0307 ntime: 0079 mem: 3.36 + 04-04 11:47:53 | [661][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1438 ntime: 0089 mem: 3.36 + 04-04 11:48:00 | [661][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0080 mem: 3.36 + 04-04 11:48:06 | [661][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0123 ntime: 0077 mem: 3.36 + 04-04 11:48:12 | [661][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0500 ntime: 0086 mem: 3.36 + 04-04 11:48:18 | [661][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0130 ntime: 0080 mem: 3.36 + 04-04 11:48:22 | [661][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1127 ntime: 0079 mem: 3.36 + 04-04 11:48:30 | [661][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0082 mem: 3.36 + 04-04 11:48:36 | [661][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0101 ntime: 0082 mem: 3.36 + 04-04 11:48:42 | [661][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0082 ntime: 0086 mem: 3.36 + 04-04 11:48:49 | Time info >>>> elapsed: 865.44 mins remain: 441.87 mins + 04-04 11:48:49 | [662][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0090 ntime: 0082 mem: 3.36 + 04-04 11:48:55 | [662][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0073 ntime: 0079 mem: 3.36 + 04-04 11:49:01 | [662][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0605 ntime: 0083 mem: 3.36 + 04-04 11:49:07 | [662][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0067 ntime: 0084 mem: 3.36 + 04-04 11:49:14 | [662][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1034 ntime: 0088 mem: 3.36 + 04-04 11:49:21 | [662][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0549 ntime: 0084 mem: 3.36 + 04-04 11:49:26 | [662][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0073 ntime: 0087 mem: 3.36 + 04-04 11:49:33 | [662][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0092 ntime: 0074 mem: 3.36 + 04-04 11:49:40 | [662][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0106 ntime: 0073 mem: 3.36 + 04-04 11:49:46 | [662][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0686 ntime: 0079 mem: 3.36 + 04-04 11:49:52 | [662][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1091 ntime: 0077 mem: 3.36 + 04-04 11:49:59 | [662][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1254 ntime: 0076 mem: 3.36 + 04-04 11:50:07 | [662][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1139 ntime: 0079 mem: 3.36 + 04-04 11:50:13 | [662][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0578 ntime: 0082 mem: 3.36 + 04-04 11:50:22 | [662][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1184 ntime: 0080 mem: 3.36 + 04-04 11:50:29 | [662][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0638 ntime: 0085 mem: 3.36 + 04-04 11:50:37 | [662][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1547 ntime: 0080 mem: 3.36 + 04-04 11:50:44 | [662][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0804 ntime: 0083 mem: 3.36 + 04-04 11:50:48 | Time info >>>> elapsed: 867.43 mins remain: 440.91 mins + 04-04 11:50:50 | [663][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1488 ntime: 0078 mem: 3.36 + 04-04 11:50:56 | [663][010/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0986 ntime: 0082 mem: 3.36 + 04-04 11:51:01 | [663][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0795 ntime: 0089 mem: 3.36 + 04-04 11:51:07 | [663][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0133 ntime: 0086 mem: 3.36 + 04-04 11:51:16 | [663][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1509 ntime: 0080 mem: 3.36 + 04-04 11:51:23 | [663][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1272 ntime: 0081 mem: 3.36 + 04-04 11:51:28 | [663][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0623 ntime: 0082 mem: 3.36 + 04-04 11:51:35 | [663][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0074 mem: 3.36 + 04-04 11:51:42 | [663][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0770 ntime: 0079 mem: 3.36 + 04-04 11:51:50 | [663][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0121 ntime: 0073 mem: 3.36 + 04-04 11:52:00 | [663][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1434 ntime: 0075 mem: 3.36 + 04-04 11:52:07 | [663][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0684 ntime: 0083 mem: 3.36 + 04-04 11:52:14 | [663][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1602 ntime: 0082 mem: 3.36 + 04-04 11:52:20 | [663][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0074 mem: 3.36 + 04-04 11:52:29 | [663][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1409 ntime: 0071 mem: 3.36 + 04-04 11:52:36 | [663][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0073 mem: 3.36 + 04-04 11:52:42 | [663][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1193 ntime: 0080 mem: 3.36 + 04-04 11:52:50 | [663][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0914 ntime: 0085 mem: 3.36 + 04-04 11:52:56 | Time info >>>> elapsed: 869.56 mins remain: 440.02 mins + 04-04 11:52:57 | [664][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1174 ntime: 0075 mem: 3.36 + 04-04 11:53:02 | [664][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0079 mem: 3.36 + 04-04 11:53:10 | [664][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1478 ntime: 0078 mem: 3.36 + 04-04 11:53:17 | [664][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0984 ntime: 0071 mem: 3.36 + 04-04 11:53:23 | [664][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0951 ntime: 0073 mem: 3.36 + 04-04 11:53:32 | [664][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0527 ntime: 0085 mem: 3.36 + 04-04 11:53:38 | [664][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0154 ntime: 0087 mem: 3.36 + 04-04 11:53:45 | [664][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0075 mem: 3.36 + 04-04 11:53:53 | [664][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0590 ntime: 0077 mem: 3.36 + 04-04 11:53:59 | [664][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0226 ntime: 0081 mem: 3.36 + 04-04 11:54:04 | [664][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0082 ntime: 0076 mem: 3.36 + 04-04 11:54:10 | [664][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0596 ntime: 0081 mem: 3.36 + 04-04 11:54:16 | [664][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0424 ntime: 0081 mem: 3.36 + 04-04 11:54:23 | [664][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0428 ntime: 0086 mem: 3.36 + 04-04 11:54:31 | [664][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1507 ntime: 0078 mem: 3.36 + 04-04 11:54:39 | [664][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1472 ntime: 0078 mem: 3.36 + 04-04 11:54:45 | [664][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0515 ntime: 0085 mem: 3.36 + 04-04 11:54:53 | [664][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0527 ntime: 0075 mem: 3.36 + 04-04 11:54:58 | Time info >>>> elapsed: 871.59 mins remain: 439.07 mins + 04-04 11:55:00 | [665][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1625 ntime: 0083 mem: 3.36 + 04-04 11:55:05 | [665][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 11:55:12 | [665][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0081 mem: 3.36 + 04-04 11:55:20 | [665][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0454 ntime: 0081 mem: 3.36 + 04-04 11:55:27 | [665][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1167 ntime: 0062 mem: 3.36 + 04-04 11:55:36 | [665][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1293 ntime: 0073 mem: 3.36 + 04-04 11:55:42 | [665][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0450 ntime: 0088 mem: 3.36 + 04-04 11:55:49 | [665][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1191 ntime: 0088 mem: 3.36 + 04-04 11:55:58 | [665][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0786 ntime: 0085 mem: 3.36 + 04-04 11:56:03 | [665][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 11:56:10 | [665][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1138 ntime: 0088 mem: 3.36 + 04-04 11:56:17 | [665][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1338 ntime: 0083 mem: 3.36 + 04-04 11:56:24 | [665][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0180 ntime: 0080 mem: 3.36 + 04-04 11:56:32 | [665][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0911 ntime: 0087 mem: 3.36 + 04-04 11:56:38 | [665][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0077 mem: 3.36 + 04-04 11:56:44 | [665][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0081 mem: 3.36 + 04-04 11:56:52 | [665][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0935 ntime: 0081 mem: 3.36 + 04-04 11:56:58 | [665][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0069 ntime: 0083 mem: 3.36 + 04-04 11:57:03 | Time info >>>> elapsed: 873.68 mins remain: 438.15 mins + 04-04 11:57:04 | [666][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0353 ntime: 0083 mem: 3.36 + 04-04 11:57:12 | [666][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0087 ntime: 0083 mem: 3.36 + 04-04 11:57:19 | [666][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0448 ntime: 0076 mem: 3.36 + 04-04 11:57:25 | [666][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0086 mem: 3.36 + 04-04 11:57:32 | [666][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1024 ntime: 0079 mem: 3.36 + 04-04 11:57:38 | [666][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0205 ntime: 0076 mem: 3.36 + 04-04 11:57:43 | [666][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0913 ntime: 0081 mem: 3.36 + 04-04 11:57:51 | [666][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1260 ntime: 0078 mem: 3.36 + 04-04 11:57:58 | [666][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0077 mem: 3.36 + 04-04 11:58:06 | [666][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0087 mem: 3.36 + 04-04 11:58:13 | [666][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1129 ntime: 0056 mem: 3.36 + 04-04 11:58:22 | [666][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0891 ntime: 0078 mem: 3.36 + 04-04 11:58:27 | [666][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0716 ntime: 0081 mem: 3.36 + 04-04 11:58:34 | [666][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0967 ntime: 0082 mem: 3.36 + 04-04 11:58:41 | [666][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0540 ntime: 0089 mem: 3.36 + 04-04 11:58:48 | [666][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0402 ntime: 0081 mem: 3.36 + 04-04 11:58:54 | [666][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0486 ntime: 0082 mem: 3.36 + 04-04 11:59:00 | [666][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0468 ntime: 0084 mem: 3.36 + 04-04 11:59:04 | Time info >>>> elapsed: 875.70 mins remain: 437.19 mins + 04-04 11:59:06 | [667][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1467 ntime: 0079 mem: 3.36 + 04-04 11:59:13 | [667][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1056 ntime: 0084 mem: 3.36 + 04-04 11:59:19 | [667][020/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 1126 ntime: 0074 mem: 3.36 + 04-04 11:59:26 | [667][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0914 ntime: 0082 mem: 3.36 + 04-04 11:59:32 | [667][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0999 ntime: 0081 mem: 3.36 + 04-04 11:59:36 | [667][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0118 ntime: 0086 mem: 3.36 + 04-04 11:59:43 | [667][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0110 ntime: 0081 mem: 3.36 + 04-04 11:59:48 | [667][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0104 ntime: 0081 mem: 3.36 + 04-04 11:59:55 | [667][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0768 ntime: 0077 mem: 3.36 + 04-04 12:00:00 | [667][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0078 ntime: 0078 mem: 3.36 + 04-04 12:00:07 | [667][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0194 ntime: 0082 mem: 3.36 + 04-04 12:00:15 | [667][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0165 ntime: 0077 mem: 3.36 + 04-04 12:00:22 | [667][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1127 ntime: 0087 mem: 3.36 + 04-04 12:00:29 | [667][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1100 ntime: 0088 mem: 3.36 + 04-04 12:00:37 | [667][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1622 ntime: 0071 mem: 3.36 + 04-04 12:00:43 | [667][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0093 ntime: 0083 mem: 3.36 + 04-04 12:00:51 | [667][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0137 ntime: 0079 mem: 3.36 + 04-04 12:00:59 | [667][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0645 ntime: 0081 mem: 3.36 + 04-04 12:01:03 | Time info >>>> elapsed: 877.68 mins remain: 436.21 mins + 04-04 12:01:04 | [668][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0834 ntime: 0081 mem: 3.36 + 04-04 12:01:13 | [668][010/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0954 ntime: 0075 mem: 3.36 + 04-04 12:01:20 | [668][020/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0046 ntime: 0074 mem: 3.36 + 04-04 12:01:30 | [668][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0646 ntime: 0080 mem: 3.36 + 04-04 12:01:36 | [668][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1208 ntime: 0087 mem: 3.36 + 04-04 12:01:43 | [668][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0157 ntime: 0081 mem: 3.36 + 04-04 12:01:51 | [668][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1261 ntime: 0079 mem: 3.36 + 04-04 12:01:57 | [668][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0046 ntime: 0081 mem: 3.36 + 04-04 12:02:03 | [668][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0625 ntime: 0078 mem: 3.36 + 04-04 12:02:09 | [668][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0605 ntime: 0080 mem: 3.36 + 04-04 12:02:14 | [668][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0076 mem: 3.36 + 04-04 12:02:21 | [668][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0080 mem: 3.36 + 04-04 12:02:29 | [668][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0809 ntime: 0078 mem: 3.36 + 04-04 12:02:35 | [668][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1101 ntime: 0085 mem: 3.36 + 04-04 12:02:41 | [668][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0799 ntime: 0080 mem: 3.36 + 04-04 12:02:48 | [668][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1590 ntime: 0088 mem: 3.36 + 04-04 12:02:58 | [668][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1561 ntime: 0081 mem: 3.36 + 04-04 12:03:05 | [668][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0133 ntime: 0078 mem: 3.36 + 04-04 12:03:10 | Time info >>>> elapsed: 879.79 mins remain: 435.29 mins + 04-04 12:03:11 | [669][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1029 ntime: 0085 mem: 3.36 + 04-04 12:03:18 | [669][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0089 mem: 3.36 + 04-04 12:03:25 | [669][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0528 ntime: 0077 mem: 3.36 + 04-04 12:03:33 | [669][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0695 ntime: 0078 mem: 3.36 + 04-04 12:03:41 | [669][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0974 ntime: 0079 mem: 3.36 + 04-04 12:03:47 | [669][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0121 ntime: 0081 mem: 3.36 + 04-04 12:03:55 | [669][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0984 ntime: 0086 mem: 3.36 + 04-04 12:04:01 | [669][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0434 ntime: 0084 mem: 3.36 + 04-04 12:04:09 | [669][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0120 ntime: 0078 mem: 3.36 + 04-04 12:04:16 | [669][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1298 ntime: 0072 mem: 3.36 + 04-04 12:04:21 | [669][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0757 ntime: 0089 mem: 3.36 + 04-04 12:04:27 | [669][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0912 ntime: 0089 mem: 3.36 + 04-04 12:04:35 | [669][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0101 ntime: 0076 mem: 3.36 + 04-04 12:04:41 | [669][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1050 ntime: 0085 mem: 3.36 + 04-04 12:04:49 | [669][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1529 ntime: 0078 mem: 3.36 + 04-04 12:04:55 | [669][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0117 ntime: 0081 mem: 3.36 + 04-04 12:05:03 | [669][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1714 ntime: 0080 mem: 3.36 + 04-04 12:05:12 | [669][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0122 ntime: 0076 mem: 3.36 + 04-04 12:05:17 | Time info >>>> elapsed: 881.91 mins remain: 434.37 mins + 04-04 12:05:17 | [670][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0093 ntime: 0077 mem: 3.36 + 04-04 12:05:25 | [670][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1153 ntime: 0078 mem: 3.36 + 04-04 12:05:32 | [670][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0888 ntime: 0077 mem: 3.36 + 04-04 12:05:37 | [670][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 12:05:45 | [670][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0935 ntime: 0080 mem: 3.36 + 04-04 12:05:52 | [670][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0082 mem: 3.36 + 04-04 12:06:00 | [670][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0090 ntime: 0089 mem: 3.36 + 04-04 12:06:07 | [670][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0893 ntime: 0079 mem: 3.36 + 04-04 12:06:12 | [670][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0534 ntime: 0086 mem: 3.36 + 04-04 12:06:19 | [670][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1060 ntime: 0081 mem: 3.36 + 04-04 12:06:24 | [670][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0102 ntime: 0085 mem: 3.36 + 04-04 12:06:31 | [670][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1322 ntime: 0083 mem: 3.36 + 04-04 12:06:38 | [670][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0502 ntime: 0087 mem: 3.36 + 04-04 12:06:44 | [670][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0083 mem: 3.36 + 04-04 12:06:52 | [670][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1712 ntime: 0081 mem: 3.36 + 04-04 12:06:58 | [670][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0077 mem: 3.36 + 04-04 12:07:10 | [670][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1030 ntime: 0092 mem: 3.36 + 04-04 12:07:16 | [670][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1268 ntime: 0081 mem: 3.36 + 04-04 12:07:21 | Time info >>>> elapsed: 883.98 mins remain: 433.43 mins + 04-04 12:07:22 | [671][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0345 ntime: 0074 mem: 3.36 + 04-04 12:07:27 | [671][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0863 ntime: 0085 mem: 3.36 + 04-04 12:07:33 | [671][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0345 ntime: 0081 mem: 3.36 + 04-04 12:07:42 | [671][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0080 mem: 3.36 + 04-04 12:07:50 | [671][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0120 ntime: 0079 mem: 3.36 + 04-04 12:07:57 | [671][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0832 ntime: 0082 mem: 3.36 + 04-04 12:08:05 | [671][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0913 ntime: 0080 mem: 3.36 + 04-04 12:08:10 | [671][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0553 ntime: 0075 mem: 3.36 + 04-04 12:08:16 | [671][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0082 mem: 3.36 + 04-04 12:08:23 | [671][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0134 ntime: 0080 mem: 3.36 + 04-04 12:08:29 | [671][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0427 ntime: 0076 mem: 3.36 + 04-04 12:08:38 | [671][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0083 mem: 3.36 + 04-04 12:08:45 | [671][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0047 ntime: 0080 mem: 3.36 + 04-04 12:08:53 | [671][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0669 ntime: 0081 mem: 3.36 + 04-04 12:08:58 | [671][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0741 ntime: 0076 mem: 3.36 + 04-04 12:09:07 | [671][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1046 ntime: 0079 mem: 3.36 + 04-04 12:09:13 | [671][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1158 ntime: 0077 mem: 3.36 + 04-04 12:09:20 | [671][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0065 ntime: 0086 mem: 3.36 + 04-04 12:09:24 | Time info >>>> elapsed: 886.03 mins remain: 432.47 mins + 04-04 12:09:25 | [672][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0833 ntime: 0080 mem: 3.36 + 04-04 12:09:32 | [672][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0384 ntime: 0078 mem: 3.36 + 04-04 12:09:37 | [672][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0080 mem: 3.36 + 04-04 12:09:44 | [672][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0778 ntime: 0087 mem: 3.36 + 04-04 12:09:51 | [672][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1406 ntime: 0077 mem: 3.36 + 04-04 12:09:58 | [672][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1210 ntime: 0056 mem: 3.36 + 04-04 12:10:03 | [672][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0081 mem: 3.36 + 04-04 12:10:10 | [672][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0149 ntime: 0081 mem: 3.36 + 04-04 12:10:16 | [672][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0136 ntime: 0076 mem: 3.36 + 04-04 12:10:23 | [672][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0579 ntime: 0083 mem: 3.36 + 04-04 12:10:30 | [672][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0080 ntime: 0087 mem: 3.36 + 04-04 12:10:36 | [672][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0104 ntime: 0086 mem: 3.36 + 04-04 12:10:43 | [672][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0619 ntime: 0083 mem: 3.36 + 04-04 12:10:49 | [672][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0171 ntime: 0087 mem: 3.36 + 04-04 12:10:56 | [672][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0545 ntime: 0077 mem: 3.36 + 04-04 12:11:04 | [672][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0082 ntime: 0080 mem: 3.36 + 04-04 12:11:10 | [672][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1106 ntime: 0085 mem: 3.36 + 04-04 12:11:18 | [672][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1484 ntime: 0060 mem: 3.36 + 04-04 12:11:22 | Time info >>>> elapsed: 887.99 mins remain: 431.46 mins + 04-04 12:11:22 | [673][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0404 ntime: 0076 mem: 3.36 + 04-04 12:11:29 | [673][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0913 ntime: 0081 mem: 3.36 + 04-04 12:11:36 | [673][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1559 ntime: 0081 mem: 3.36 + 04-04 12:11:43 | [673][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0657 ntime: 0077 mem: 3.36 + 04-04 12:11:50 | [673][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1356 ntime: 0079 mem: 3.36 + 04-04 12:11:56 | [673][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0302 ntime: 0088 mem: 3.36 + 04-04 12:12:03 | [673][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0914 ntime: 0086 mem: 3.36 + 04-04 12:12:08 | [673][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0838 ntime: 0082 mem: 3.36 + 04-04 12:12:13 | [673][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0091 ntime: 0078 mem: 3.36 + 04-04 12:12:18 | [673][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0084 ntime: 0078 mem: 3.36 + 04-04 12:12:22 | [673][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0168 ntime: 0078 mem: 3.36 + 04-04 12:12:28 | [673][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0081 mem: 3.36 + 04-04 12:12:35 | [673][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0765 ntime: 0085 mem: 3.36 + 04-04 12:12:42 | [673][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0408 ntime: 0080 mem: 3.36 + 04-04 12:12:49 | [673][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 12:12:54 | [673][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0075 ntime: 0080 mem: 3.36 + 04-04 12:13:02 | [673][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0081 mem: 3.36 + 04-04 12:13:09 | [673][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0159 ntime: 0079 mem: 3.36 + 04-04 12:13:15 | Time info >>>> elapsed: 889.87 mins remain: 430.41 mins + 04-04 12:13:15 | [674][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0079 ntime: 0078 mem: 3.36 + 04-04 12:13:22 | [674][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1040 ntime: 0086 mem: 3.36 + 04-04 12:13:27 | [674][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0105 ntime: 0081 mem: 3.36 + 04-04 12:13:32 | [674][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0774 ntime: 0081 mem: 3.36 + 04-04 12:13:37 | [674][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0197 ntime: 0074 mem: 3.36 + 04-04 12:13:44 | [674][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1675 ntime: 0086 mem: 3.36 + 04-04 12:13:51 | [674][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0093 ntime: 0087 mem: 3.36 + 04-04 12:13:58 | [674][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0352 ntime: 0072 mem: 3.36 + 04-04 12:14:05 | [674][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1014 ntime: 0082 mem: 3.36 + 04-04 12:14:11 | [674][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0994 ntime: 0072 mem: 3.36 + 04-04 12:14:19 | [674][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1510 ntime: 0086 mem: 3.36 + 04-04 12:14:28 | [674][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1158 ntime: 0081 mem: 3.36 + 04-04 12:14:32 | [674][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0645 ntime: 0080 mem: 3.36 + 04-04 12:14:37 | [674][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0084 ntime: 0089 mem: 3.36 + 04-04 12:14:42 | [674][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0094 ntime: 0082 mem: 3.36 + 04-04 12:14:48 | [674][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0638 ntime: 0074 mem: 3.36 + 04-04 12:14:54 | [674][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0182 ntime: 0085 mem: 3.36 + 04-04 12:15:00 | [674][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0698 ntime: 0078 mem: 3.36 + 04-04 12:15:04 | Time info >>>> elapsed: 891.70 mins remain: 429.34 mins + 04-04 12:15:04 | [675][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 12:15:11 | [675][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0229 ntime: 0082 mem: 3.36 + 04-04 12:15:17 | [675][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0080 mem: 3.36 + 04-04 12:15:25 | [675][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0892 ntime: 0083 mem: 3.36 + 04-04 12:15:31 | [675][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0819 ntime: 0075 mem: 3.36 + 04-04 12:15:37 | [675][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0616 ntime: 0082 mem: 3.36 + 04-04 12:15:43 | [675][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0926 ntime: 0088 mem: 3.36 + 04-04 12:15:49 | [675][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0745 ntime: 0086 mem: 3.36 + 04-04 12:15:53 | [675][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0182 ntime: 0082 mem: 3.36 + 04-04 12:16:01 | [675][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0102 ntime: 0073 mem: 3.36 + 04-04 12:16:07 | [675][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0062 ntime: 0074 mem: 3.36 + 04-04 12:16:13 | [675][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0074 ntime: 0081 mem: 3.36 + 04-04 12:16:19 | [675][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0080 mem: 3.36 + 04-04 12:16:25 | [675][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0889 ntime: 0081 mem: 3.36 + 04-04 12:16:30 | [675][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0250 ntime: 0079 mem: 3.36 + 04-04 12:16:36 | [675][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0154 ntime: 0075 mem: 3.36 + 04-04 12:16:40 | [675][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0103 ntime: 0087 mem: 3.36 + 04-04 12:16:48 | [675][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1427 ntime: 0077 mem: 3.36 + 04-04 12:16:53 | Time info >>>> elapsed: 893.51 mins remain: 428.25 mins + 04-04 12:16:54 | [676][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0714 ntime: 0083 mem: 3.36 + 04-04 12:17:02 | [676][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0920 ntime: 0076 mem: 3.36 + 04-04 12:17:08 | [676][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0714 ntime: 0080 mem: 3.36 + 04-04 12:17:13 | [676][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0093 ntime: 0085 mem: 3.36 + 04-04 12:17:19 | [676][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0070 ntime: 0080 mem: 3.36 + 04-04 12:17:25 | [676][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0442 ntime: 0085 mem: 3.36 + 04-04 12:17:30 | [676][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0367 ntime: 0083 mem: 3.36 + 04-04 12:17:36 | [676][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1027 ntime: 0086 mem: 3.36 + 04-04 12:17:42 | [676][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0919 ntime: 0084 mem: 3.36 + 04-04 12:17:47 | [676][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0828 ntime: 0084 mem: 3.36 + 04-04 12:17:53 | [676][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1439 ntime: 0080 mem: 3.36 + 04-04 12:17:57 | [676][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0698 ntime: 0081 mem: 3.36 + 04-04 12:18:04 | [676][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0669 ntime: 0080 mem: 3.36 + 04-04 12:18:09 | [676][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0081 mem: 3.36 + 04-04 12:18:15 | [676][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1211 ntime: 0076 mem: 3.36 + 04-04 12:18:22 | [676][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1070 ntime: 0085 mem: 3.36 + 04-04 12:18:26 | [676][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0082 ntime: 0079 mem: 3.36 + 04-04 12:18:33 | [676][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1022 ntime: 0085 mem: 3.36 + 04-04 12:18:36 | Time info >>>> elapsed: 895.23 mins remain: 427.12 mins + 04-04 12:18:37 | [677][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0085 ntime: 0082 mem: 3.36 + 04-04 12:18:43 | [677][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1120 ntime: 0082 mem: 3.36 + 04-04 12:18:49 | [677][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0129 ntime: 0078 mem: 3.36 + 04-04 12:18:55 | [677][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0514 ntime: 0080 mem: 3.36 + 04-04 12:19:02 | [677][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0858 ntime: 0081 mem: 3.36 + 04-04 12:19:07 | [677][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0156 ntime: 0076 mem: 3.36 + 04-04 12:19:12 | [677][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0355 ntime: 0077 mem: 3.36 + 04-04 12:19:17 | [677][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0079 mem: 3.36 + 04-04 12:19:24 | [677][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0691 ntime: 0080 mem: 3.36 + 04-04 12:19:31 | [677][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0085 mem: 3.36 + 04-04 12:19:37 | [677][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1097 ntime: 0084 mem: 3.36 + 04-04 12:19:42 | [677][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0116 ntime: 0079 mem: 3.36 + 04-04 12:19:48 | [677][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0079 mem: 3.36 + 04-04 12:19:55 | [677][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0140 ntime: 0079 mem: 3.36 + 04-04 12:20:02 | [677][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0119 ntime: 0081 mem: 3.36 + 04-04 12:20:08 | [677][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0865 ntime: 0077 mem: 3.36 + 04-04 12:20:13 | [677][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0085 ntime: 0087 mem: 3.36 + 04-04 12:20:20 | [677][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0921 ntime: 0078 mem: 3.36 + 04-04 12:20:24 | Time info >>>> elapsed: 897.02 mins remain: 426.02 mins + 04-04 12:20:24 | [678][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0667 ntime: 0081 mem: 3.36 + 04-04 12:20:30 | [678][010/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 1058 ntime: 0085 mem: 3.36 + 04-04 12:20:35 | [678][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0619 ntime: 0079 mem: 3.36 + 04-04 12:20:40 | [678][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0452 ntime: 0077 mem: 3.36 + 04-04 12:20:46 | [678][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0595 ntime: 0084 mem: 3.36 + 04-04 12:20:51 | [678][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1185 ntime: 0079 mem: 3.36 + 04-04 12:20:58 | [678][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0881 ntime: 0089 mem: 3.36 + 04-04 12:21:04 | [678][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0090 ntime: 0084 mem: 3.36 + 04-04 12:21:09 | [678][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0077 mem: 3.36 + 04-04 12:21:14 | [678][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0091 ntime: 0083 mem: 3.36 + 04-04 12:21:21 | [678][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0773 ntime: 0079 mem: 3.36 + 04-04 12:21:28 | [678][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1166 ntime: 0084 mem: 3.36 + 04-04 12:21:36 | [678][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1311 ntime: 0084 mem: 3.36 + 04-04 12:21:41 | [678][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0083 ntime: 0085 mem: 3.36 + 04-04 12:21:48 | [678][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1495 ntime: 0079 mem: 3.36 + 04-04 12:21:55 | [678][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0108 ntime: 0077 mem: 3.36 + 04-04 12:22:01 | [678][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0258 ntime: 0077 mem: 3.36 + 04-04 12:22:08 | [678][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0544 ntime: 0074 mem: 3.36 + 04-04 12:22:11 | Time info >>>> elapsed: 898.80 mins remain: 424.91 mins + 04-04 12:22:11 | [679][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0120 ntime: 0079 mem: 3.36 + 04-04 12:22:16 | [679][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 12:22:23 | [679][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0777 ntime: 0084 mem: 3.36 + 04-04 12:22:26 | [679][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0067 ntime: 0076 mem: 3.36 + 04-04 12:22:33 | [679][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0087 ntime: 0081 mem: 3.36 + 04-04 12:22:38 | [679][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0070 ntime: 0083 mem: 3.36 + 04-04 12:22:45 | [679][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0712 ntime: 0082 mem: 3.36 + 04-04 12:22:49 | [679][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0182 ntime: 0079 mem: 3.36 + 04-04 12:22:55 | [679][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0995 ntime: 0080 mem: 3.36 + 04-04 12:23:00 | [679][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1313 ntime: 0084 mem: 3.36 + 04-04 12:23:05 | [679][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0081 mem: 3.36 + 04-04 12:23:11 | [679][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0077 mem: 3.36 + 04-04 12:23:16 | [679][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0718 ntime: 0088 mem: 3.36 + 04-04 12:23:21 | [679][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0081 ntime: 0077 mem: 3.36 + 04-04 12:23:27 | [679][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0679 ntime: 0085 mem: 3.36 + 04-04 12:23:34 | [679][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0077 ntime: 0077 mem: 3.36 + 04-04 12:23:39 | [679][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 12:23:45 | [679][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0717 ntime: 0082 mem: 3.36 + 04-04 12:23:48 | Time info >>>> elapsed: 900.42 mins remain: 423.73 mins + 04-04 12:23:48 | [680][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0071 ntime: 0082 mem: 3.36 + 04-04 12:23:53 | [680][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0931 ntime: 0084 mem: 3.36 + 04-04 12:24:00 | [680][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0726 ntime: 0088 mem: 3.36 + 04-04 12:24:06 | [680][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0082 mem: 3.36 + 04-04 12:24:11 | [680][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0063 ntime: 0078 mem: 3.36 + 04-04 12:24:17 | [680][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0083 mem: 3.36 + 04-04 12:24:23 | [680][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1000 ntime: 0082 mem: 3.36 + 04-04 12:24:30 | [680][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1107 ntime: 0081 mem: 3.36 + 04-04 12:24:35 | [680][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 12:24:41 | [680][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0816 ntime: 0057 mem: 3.36 + 04-04 12:24:48 | [680][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0131 ntime: 0076 mem: 3.36 + 04-04 12:24:55 | [680][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0129 ntime: 0078 mem: 3.36 + 04-04 12:25:02 | [680][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1053 ntime: 0081 mem: 3.36 + 04-04 12:25:09 | [680][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0996 ntime: 0084 mem: 3.36 + 04-04 12:25:17 | [680][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0192 ntime: 0082 mem: 3.36 + 04-04 12:25:23 | [680][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0313 ntime: 0081 mem: 3.36 + 04-04 12:25:29 | [680][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0654 ntime: 0089 mem: 3.36 + 04-04 12:25:33 | [680][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0063 ntime: 0072 mem: 3.36 + 04-04 12:25:39 | Time info >>>> elapsed: 902.28 mins remain: 422.65 mins + 04-04 12:25:40 | [681][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1113 ntime: 0080 mem: 3.36 + 04-04 12:25:47 | [681][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0085 mem: 3.36 + 04-04 12:25:53 | [681][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0758 ntime: 0081 mem: 3.36 + 04-04 12:25:58 | [681][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0079 mem: 3.36 + 04-04 12:26:03 | [681][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0132 ntime: 0084 mem: 3.36 + 04-04 12:26:10 | [681][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0163 ntime: 0083 mem: 3.36 + 04-04 12:26:17 | [681][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0739 ntime: 0077 mem: 3.36 + 04-04 12:26:23 | [681][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0681 ntime: 0082 mem: 3.36 + 04-04 12:26:29 | [681][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0983 ntime: 0083 mem: 3.36 + 04-04 12:26:35 | [681][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0110 ntime: 0078 mem: 3.36 + 04-04 12:26:41 | [681][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0600 ntime: 0080 mem: 3.36 + 04-04 12:26:46 | [681][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0378 ntime: 0085 mem: 3.36 + 04-04 12:26:53 | [681][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1345 ntime: 0086 mem: 3.36 + 04-04 12:26:58 | [681][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0080 mem: 3.36 + 04-04 12:27:05 | [681][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0483 ntime: 0078 mem: 3.36 + 04-04 12:27:09 | [681][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0084 ntime: 0088 mem: 3.36 + 04-04 12:27:15 | [681][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0462 ntime: 0075 mem: 3.36 + 04-04 12:27:23 | [681][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0971 ntime: 0080 mem: 3.36 + 04-04 12:27:28 | Time info >>>> elapsed: 904.09 mins remain: 421.55 mins + 04-04 12:27:28 | [682][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0178 ntime: 0081 mem: 3.36 + 04-04 12:27:35 | [682][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1291 ntime: 0058 mem: 3.36 + 04-04 12:27:43 | [682][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1061 ntime: 0081 mem: 3.36 + 04-04 12:27:49 | [682][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0129 ntime: 0074 mem: 3.36 + 04-04 12:27:55 | [682][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 12:28:02 | [682][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0136 ntime: 0077 mem: 3.36 + 04-04 12:28:07 | [682][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0080 ntime: 0080 mem: 3.36 + 04-04 12:28:13 | [682][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0097 ntime: 0085 mem: 3.36 + 04-04 12:28:18 | [682][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0121 ntime: 0083 mem: 3.36 + 04-04 12:28:23 | [682][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0110 ntime: 0081 mem: 3.36 + 04-04 12:28:30 | [682][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0121 ntime: 0082 mem: 3.36 + 04-04 12:28:36 | [682][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0971 ntime: 0079 mem: 3.36 + 04-04 12:28:43 | [682][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1110 ntime: 0078 mem: 3.36 + 04-04 12:28:50 | [682][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1056 ntime: 0086 mem: 3.36 + 04-04 12:28:55 | [682][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0129 ntime: 0085 mem: 3.36 + 04-04 12:29:02 | [682][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0733 ntime: 0082 mem: 3.36 + 04-04 12:29:07 | [682][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0526 ntime: 0073 mem: 3.36 + 04-04 12:29:13 | [682][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1444 ntime: 0078 mem: 3.36 + 04-04 12:29:17 | Time info >>>> elapsed: 905.91 mins remain: 420.46 mins + 04-04 12:29:17 | [683][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0129 ntime: 0079 mem: 3.36 + 04-04 12:29:23 | [683][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0077 ntime: 0080 mem: 3.36 + 04-04 12:29:29 | [683][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0077 mem: 3.36 + 04-04 12:29:34 | [683][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0076 mem: 3.36 + 04-04 12:29:40 | [683][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0154 ntime: 0082 mem: 3.36 + 04-04 12:29:44 | [683][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0516 ntime: 0080 mem: 3.36 + 04-04 12:29:50 | [683][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0574 ntime: 0079 mem: 3.36 + 04-04 12:29:57 | [683][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1559 ntime: 0072 mem: 3.36 + 04-04 12:30:02 | [683][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0720 ntime: 0078 mem: 3.36 + 04-04 12:30:08 | [683][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1048 ntime: 0084 mem: 3.36 + 04-04 12:30:15 | [683][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0818 ntime: 0079 mem: 3.36 + 04-04 12:30:20 | [683][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0957 ntime: 0081 mem: 3.36 + 04-04 12:30:25 | [683][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0073 ntime: 0079 mem: 3.36 + 04-04 12:30:30 | [683][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0584 ntime: 0059 mem: 3.36 + 04-04 12:30:36 | [683][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0047 ntime: 0083 mem: 3.36 + 04-04 12:30:41 | [683][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0059 mem: 3.36 + 04-04 12:30:47 | [683][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0082 mem: 3.36 + 04-04 12:30:53 | [683][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0895 ntime: 0084 mem: 3.36 + 04-04 12:30:58 | Time info >>>> elapsed: 907.60 mins remain: 419.30 mins + 04-04 12:30:58 | [684][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0116 ntime: 0080 mem: 3.36 + 04-04 12:31:06 | [684][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0633 ntime: 0083 mem: 3.36 + 04-04 12:31:10 | [684][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0078 mem: 3.36 + 04-04 12:31:17 | [684][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0148 ntime: 0081 mem: 3.36 + 04-04 12:31:22 | [684][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0103 ntime: 0076 mem: 3.36 + 04-04 12:31:29 | [684][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0082 mem: 3.36 + 04-04 12:31:35 | [684][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1098 ntime: 0078 mem: 3.36 + 04-04 12:31:40 | [684][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0285 ntime: 0081 mem: 3.36 + 04-04 12:31:47 | [684][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0674 ntime: 0083 mem: 3.36 + 04-04 12:31:55 | [684][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0099 ntime: 0081 mem: 3.36 + 04-04 12:32:01 | [684][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0561 ntime: 0079 mem: 3.36 + 04-04 12:32:07 | [684][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0702 ntime: 0079 mem: 3.36 + 04-04 12:32:13 | [684][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0713 ntime: 0085 mem: 3.36 + 04-04 12:32:19 | [684][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1035 ntime: 0075 mem: 3.36 + 04-04 12:32:24 | [684][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0710 ntime: 0075 mem: 3.36 + 04-04 12:32:30 | [684][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0093 ntime: 0081 mem: 3.36 + 04-04 12:32:35 | [684][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0073 mem: 3.36 + 04-04 12:32:42 | [684][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0067 ntime: 0080 mem: 3.36 + 04-04 12:32:47 | Time info >>>> elapsed: 909.41 mins remain: 418.20 mins + 04-04 12:32:47 | [685][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0093 ntime: 0086 mem: 3.36 + 04-04 12:32:53 | [685][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0140 ntime: 0084 mem: 3.36 + 04-04 12:32:58 | [685][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0502 ntime: 0085 mem: 3.36 + 04-04 12:33:02 | [685][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0739 ntime: 0081 mem: 3.36 + 04-04 12:33:09 | [685][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0154 ntime: 0082 mem: 3.36 + 04-04 12:33:17 | [685][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0751 ntime: 0083 mem: 3.36 + 04-04 12:33:24 | [685][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0906 ntime: 0089 mem: 3.36 + 04-04 12:33:30 | [685][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0287 ntime: 0075 mem: 3.36 + 04-04 12:33:35 | [685][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0785 ntime: 0073 mem: 3.36 + 04-04 12:33:42 | [685][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0065 ntime: 0084 mem: 3.36 + 04-04 12:33:47 | [685][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0065 ntime: 0079 mem: 3.36 + 04-04 12:33:53 | [685][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1109 ntime: 0072 mem: 3.36 + 04-04 12:33:59 | [685][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0981 ntime: 0084 mem: 3.36 + 04-04 12:34:03 | [685][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0106 ntime: 0086 mem: 3.36 + 04-04 12:34:09 | [685][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0654 ntime: 0082 mem: 3.36 + 04-04 12:34:15 | [685][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0634 ntime: 0085 mem: 3.36 + 04-04 12:34:22 | [685][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0219 ntime: 0074 mem: 3.36 + 04-04 12:34:27 | [685][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0905 ntime: 0080 mem: 3.36 + 04-04 12:34:32 | Time info >>>> elapsed: 911.17 mins remain: 417.07 mins + 04-04 12:34:33 | [686][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0630 ntime: 0077 mem: 3.36 + 04-04 12:34:40 | [686][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1202 ntime: 0081 mem: 3.36 + 04-04 12:34:46 | [686][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0678 ntime: 0085 mem: 3.36 + 04-04 12:34:52 | [686][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1069 ntime: 0085 mem: 3.36 + 04-04 12:34:57 | [686][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0635 ntime: 0078 mem: 3.36 + 04-04 12:35:03 | [686][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0816 ntime: 0078 mem: 3.36 + 04-04 12:35:09 | [686][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0114 ntime: 0076 mem: 3.36 + 04-04 12:35:15 | [686][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0475 ntime: 0083 mem: 3.36 + 04-04 12:35:21 | [686][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0555 ntime: 0080 mem: 3.36 + 04-04 12:35:27 | [686][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0109 ntime: 0081 mem: 3.36 + 04-04 12:35:34 | [686][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0351 ntime: 0084 mem: 3.36 + 04-04 12:35:39 | [686][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0502 ntime: 0082 mem: 3.36 + 04-04 12:35:43 | [686][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0116 ntime: 0086 mem: 3.36 + 04-04 12:35:50 | [686][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1067 ntime: 0088 mem: 3.36 + 04-04 12:35:56 | [686][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0162 ntime: 0082 mem: 3.36 + 04-04 12:36:01 | [686][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0679 ntime: 0086 mem: 3.36 + 04-04 12:36:07 | [686][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0656 ntime: 0075 mem: 3.36 + 04-04 12:36:12 | [686][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0073 mem: 3.36 + 04-04 12:36:17 | Time info >>>> elapsed: 912.92 mins remain: 415.93 mins + 04-04 12:36:18 | [687][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0872 ntime: 0080 mem: 3.36 + 04-04 12:36:23 | [687][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0625 ntime: 0082 mem: 3.36 + 04-04 12:36:28 | [687][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0986 ntime: 0060 mem: 3.36 + 04-04 12:36:34 | [687][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0146 ntime: 0056 mem: 3.36 + 04-04 12:36:40 | [687][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0667 ntime: 0081 mem: 3.36 + 04-04 12:36:45 | [687][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0109 ntime: 0078 mem: 3.36 + 04-04 12:36:52 | [687][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0947 ntime: 0076 mem: 3.36 + 04-04 12:36:57 | [687][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0082 mem: 3.36 + 04-04 12:37:04 | [687][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0551 ntime: 0079 mem: 3.36 + 04-04 12:37:08 | [687][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0299 ntime: 0082 mem: 3.36 + 04-04 12:37:15 | [687][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1366 ntime: 0078 mem: 3.36 + 04-04 12:37:21 | [687][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0516 ntime: 0094 mem: 3.36 + 04-04 12:37:26 | [687][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0075 ntime: 0078 mem: 3.36 + 04-04 12:37:33 | [687][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0081 mem: 3.36 + 04-04 12:37:39 | [687][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1000 ntime: 0084 mem: 3.36 + 04-04 12:37:44 | [687][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0095 ntime: 0087 mem: 3.36 + 04-04 12:37:50 | [687][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1260 ntime: 0076 mem: 3.36 + 04-04 12:37:56 | [687][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0774 ntime: 0077 mem: 3.36 + 04-04 12:38:02 | Time info >>>> elapsed: 914.66 mins remain: 414.79 mins + 04-04 12:38:02 | [688][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 12:38:08 | [688][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0926 ntime: 0072 mem: 3.36 + 04-04 12:38:15 | [688][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1056 ntime: 0080 mem: 3.36 + 04-04 12:38:22 | [688][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0384 ntime: 0079 mem: 3.36 + 04-04 12:38:28 | [688][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0950 ntime: 0083 mem: 3.36 + 04-04 12:38:34 | [688][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0652 ntime: 0082 mem: 3.36 + 04-04 12:38:38 | [688][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0076 mem: 3.36 + 04-04 12:38:44 | [688][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 12:38:51 | [688][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1062 ntime: 0087 mem: 3.36 + 04-04 12:38:55 | [688][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1218 ntime: 0081 mem: 3.36 + 04-04 12:39:01 | [688][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0144 ntime: 0077 mem: 3.36 + 04-04 12:39:08 | [688][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0729 ntime: 0081 mem: 3.36 + 04-04 12:39:14 | [688][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0919 ntime: 0084 mem: 3.36 + 04-04 12:39:21 | [688][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0916 ntime: 0078 mem: 3.36 + 04-04 12:39:27 | [688][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0096 ntime: 0082 mem: 3.36 + 04-04 12:39:33 | [688][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0068 ntime: 0077 mem: 3.36 + 04-04 12:39:41 | [688][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0068 ntime: 0074 mem: 3.36 + 04-04 12:39:48 | [688][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0069 ntime: 0080 mem: 3.36 + 04-04 12:39:53 | Time info >>>> elapsed: 916.51 mins remain: 413.69 mins + 04-04 12:39:53 | [689][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0055 ntime: 0081 mem: 3.36 + 04-04 12:40:01 | [689][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0735 ntime: 0084 mem: 3.36 + 04-04 12:40:08 | [689][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1599 ntime: 0083 mem: 3.36 + 04-04 12:40:15 | [689][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1105 ntime: 0083 mem: 3.36 + 04-04 12:40:22 | [689][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0293 ntime: 0084 mem: 3.36 + 04-04 12:40:28 | [689][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0155 ntime: 0081 mem: 3.36 + 04-04 12:40:35 | [689][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0109 ntime: 0081 mem: 3.36 + 04-04 12:40:41 | [689][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1219 ntime: 0082 mem: 3.36 + 04-04 12:40:47 | [689][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0517 ntime: 0080 mem: 3.36 + 04-04 12:40:55 | [689][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1079 ntime: 0082 mem: 3.36 + 04-04 12:41:00 | [689][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 12:41:06 | [689][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0505 ntime: 0077 mem: 3.36 + 04-04 12:41:12 | [689][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0125 ntime: 0078 mem: 3.36 + 04-04 12:41:17 | [689][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1106 ntime: 0080 mem: 3.36 + 04-04 12:41:24 | [689][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1377 ntime: 0083 mem: 3.36 + 04-04 12:41:30 | [689][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0843 ntime: 0074 mem: 3.36 + 04-04 12:41:37 | [689][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0094 ntime: 0080 mem: 3.36 + 04-04 12:41:42 | [689][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0110 ntime: 0084 mem: 3.36 + 04-04 12:41:47 | Time info >>>> elapsed: 918.41 mins remain: 412.62 mins + 04-04 12:41:47 | [690][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0083 ntime: 0083 mem: 3.36 + 04-04 12:41:55 | [690][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1213 ntime: 0078 mem: 3.36 + 04-04 12:42:01 | [690][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0090 ntime: 0090 mem: 3.36 + 04-04 12:42:04 | [690][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0046 ntime: 0074 mem: 3.36 + 04-04 12:42:10 | [690][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1030 ntime: 0082 mem: 3.36 + 04-04 12:42:15 | [690][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0921 ntime: 0089 mem: 3.36 + 04-04 12:42:21 | [690][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0646 ntime: 0074 mem: 3.36 + 04-04 12:42:27 | [690][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0073 ntime: 0078 mem: 3.36 + 04-04 12:42:33 | [690][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0658 ntime: 0080 mem: 3.36 + 04-04 12:42:40 | [690][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1047 ntime: 0080 mem: 3.36 + 04-04 12:42:45 | [690][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0150 ntime: 0080 mem: 3.36 + 04-04 12:42:53 | [690][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0981 ntime: 0077 mem: 3.36 + 04-04 12:42:59 | [690][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1170 ntime: 0083 mem: 3.36 + 04-04 12:43:05 | [690][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0675 ntime: 0081 mem: 3.36 + 04-04 12:43:11 | [690][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0591 ntime: 0084 mem: 3.36 + 04-04 12:43:16 | [690][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0309 ntime: 0075 mem: 3.36 + 04-04 12:43:23 | [690][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0119 ntime: 0087 mem: 3.36 + 04-04 12:43:29 | [690][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0182 ntime: 0085 mem: 3.36 + 04-04 12:43:36 | Time info >>>> elapsed: 920.22 mins remain: 411.50 mins + 04-04 12:43:36 | [691][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0102 ntime: 0087 mem: 3.36 + 04-04 12:43:45 | [691][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1147 ntime: 0096 mem: 3.36 + 04-04 12:43:51 | [691][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0075 mem: 3.36 + 04-04 12:43:57 | [691][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0647 ntime: 0075 mem: 3.36 + 04-04 12:44:04 | [691][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1556 ntime: 0084 mem: 3.36 + 04-04 12:44:11 | [691][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1088 ntime: 0087 mem: 3.36 + 04-04 12:44:16 | [691][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0118 ntime: 0083 mem: 3.36 + 04-04 12:44:22 | [691][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0101 ntime: 0076 mem: 3.36 + 04-04 12:44:28 | [691][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0083 mem: 3.36 + 04-04 12:44:34 | [691][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0067 ntime: 0083 mem: 3.36 + 04-04 12:44:40 | [691][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0081 mem: 3.36 + 04-04 12:44:48 | [691][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0081 ntime: 0079 mem: 3.36 + 04-04 12:44:54 | [691][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0568 ntime: 0072 mem: 3.36 + 04-04 12:45:00 | [691][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0736 ntime: 0084 mem: 3.36 + 04-04 12:45:08 | [691][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1394 ntime: 0089 mem: 3.36 + 04-04 12:45:14 | [691][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0731 ntime: 0083 mem: 3.36 + 04-04 12:45:19 | [691][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0656 ntime: 0076 mem: 3.36 + 04-04 12:45:23 | [691][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0070 ntime: 0076 mem: 3.36 + 04-04 12:45:28 | Time info >>>> elapsed: 922.09 mins remain: 410.41 mins + 04-04 12:45:28 | [692][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0080 mem: 3.36 + 04-04 12:45:33 | [692][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0105 ntime: 0085 mem: 3.36 + 04-04 12:45:39 | [692][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0550 ntime: 0083 mem: 3.36 + 04-04 12:45:45 | [692][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0140 ntime: 0083 mem: 3.36 + 04-04 12:45:51 | [692][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0264 ntime: 0081 mem: 3.36 + 04-04 12:45:56 | [692][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0968 ntime: 0081 mem: 3.36 + 04-04 12:46:03 | [692][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0799 ntime: 0080 mem: 3.36 + 04-04 12:46:10 | [692][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0973 ntime: 0084 mem: 3.36 + 04-04 12:46:16 | [692][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0280 ntime: 0079 mem: 3.36 + 04-04 12:46:22 | [692][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1286 ntime: 0089 mem: 3.36 + 04-04 12:46:27 | [692][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0063 ntime: 0092 mem: 3.36 + 04-04 12:46:33 | [692][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0073 ntime: 0078 mem: 3.36 + 04-04 12:46:38 | [692][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0604 ntime: 0081 mem: 3.36 + 04-04 12:46:45 | [692][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0065 ntime: 0073 mem: 3.36 + 04-04 12:46:52 | [692][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1161 ntime: 0084 mem: 3.36 + 04-04 12:46:59 | [692][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1443 ntime: 0069 mem: 3.36 + 04-04 12:47:06 | [692][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0716 ntime: 0085 mem: 3.36 + 04-04 12:47:13 | [692][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1000 ntime: 0079 mem: 3.36 + 04-04 12:47:18 | Time info >>>> elapsed: 923.92 mins remain: 409.30 mins + 04-04 12:47:18 | [693][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0080 mem: 3.36 + 04-04 12:47:24 | [693][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0956 ntime: 0077 mem: 3.36 + 04-04 12:47:31 | [693][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0078 mem: 3.36 + 04-04 12:47:38 | [693][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0119 ntime: 0082 mem: 3.36 + 04-04 12:47:44 | [693][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0916 ntime: 0080 mem: 3.36 + 04-04 12:47:51 | [693][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0878 ntime: 0079 mem: 3.36 + 04-04 12:47:56 | [693][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0546 ntime: 0078 mem: 3.36 + 04-04 12:48:02 | [693][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0864 ntime: 0090 mem: 3.36 + 04-04 12:48:08 | [693][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0630 ntime: 0076 mem: 3.36 + 04-04 12:48:16 | [693][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0567 ntime: 0075 mem: 3.36 + 04-04 12:48:20 | [693][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0065 ntime: 0085 mem: 3.36 + 04-04 12:48:26 | [693][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0147 ntime: 0081 mem: 3.36 + 04-04 12:48:33 | [693][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0880 ntime: 0077 mem: 3.36 + 04-04 12:48:38 | [693][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0123 ntime: 0075 mem: 3.36 + 04-04 12:48:43 | [693][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0558 ntime: 0088 mem: 3.36 + 04-04 12:48:48 | [693][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0076 mem: 3.36 + 04-04 12:48:55 | [693][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1077 ntime: 0085 mem: 3.36 + 04-04 12:49:01 | [693][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1181 ntime: 0075 mem: 3.36 + 04-04 12:49:05 | Time info >>>> elapsed: 925.72 mins remain: 408.17 mins + 04-04 12:49:05 | [694][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0089 ntime: 0079 mem: 3.36 + 04-04 12:49:10 | [694][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 12:49:15 | [694][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0080 mem: 3.36 + 04-04 12:49:22 | [694][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0890 ntime: 0077 mem: 3.36 + 04-04 12:49:29 | [694][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1120 ntime: 0081 mem: 3.36 + 04-04 12:49:35 | [694][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0626 ntime: 0083 mem: 3.36 + 04-04 12:49:41 | [694][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0176 ntime: 0084 mem: 3.36 + 04-04 12:49:47 | [694][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0077 mem: 3.36 + 04-04 12:49:53 | [694][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0084 mem: 3.36 + 04-04 12:50:00 | [694][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0081 ntime: 0089 mem: 3.36 + 04-04 12:50:06 | [694][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0089 mem: 3.36 + 04-04 12:50:12 | [694][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0062 ntime: 0074 mem: 3.36 + 04-04 12:50:16 | [694][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0079 ntime: 0079 mem: 3.36 + 04-04 12:50:23 | [694][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0117 ntime: 0081 mem: 3.36 + 04-04 12:50:28 | [694][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0078 ntime: 0079 mem: 3.36 + 04-04 12:50:34 | [694][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0861 ntime: 0082 mem: 3.36 + 04-04 12:50:38 | [694][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0467 ntime: 0077 mem: 3.36 + 04-04 12:50:44 | [694][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0047 ntime: 0076 mem: 3.36 + 04-04 12:50:49 | Time info >>>> elapsed: 927.44 mins remain: 407.01 mins + 04-04 12:50:49 | [695][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0078 mem: 3.36 + 04-04 12:50:55 | [695][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 12:51:00 | [695][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0108 ntime: 0088 mem: 3.36 + 04-04 12:51:05 | [695][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0074 mem: 3.36 + 04-04 12:51:11 | [695][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0136 ntime: 0078 mem: 3.36 + 04-04 12:51:17 | [695][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0761 ntime: 0079 mem: 3.36 + 04-04 12:51:23 | [695][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0529 ntime: 0076 mem: 3.36 + 04-04 12:51:28 | [695][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0080 mem: 3.36 + 04-04 12:51:32 | [695][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1476 ntime: 0086 mem: 3.36 + 04-04 12:51:37 | [695][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0073 ntime: 0083 mem: 3.36 + 04-04 12:51:42 | [695][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0292 ntime: 0074 mem: 3.36 + 04-04 12:51:48 | [695][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0645 ntime: 0079 mem: 3.36 + 04-04 12:51:54 | [695][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0633 ntime: 0076 mem: 3.36 + 04-04 12:52:01 | [695][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0636 ntime: 0086 mem: 3.36 + 04-04 12:52:08 | [695][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 12:52:15 | [695][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0945 ntime: 0077 mem: 3.36 + 04-04 12:52:19 | [695][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0046 ntime: 0076 mem: 3.36 + 04-04 12:52:25 | [695][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0076 ntime: 0079 mem: 3.36 + 04-04 12:52:30 | Time info >>>> elapsed: 929.13 mins remain: 405.83 mins + 04-04 12:52:30 | [696][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0144 ntime: 0078 mem: 3.36 + 04-04 12:52:37 | [696][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0076 mem: 3.36 + 04-04 12:52:42 | [696][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0145 ntime: 0085 mem: 3.36 + 04-04 12:52:49 | [696][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1169 ntime: 0085 mem: 3.36 + 04-04 12:52:55 | [696][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 12:53:00 | [696][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0081 mem: 3.36 + 04-04 12:53:09 | [696][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0412 ntime: 0092 mem: 3.36 + 04-04 12:53:14 | [696][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0046 ntime: 0079 mem: 3.36 + 04-04 12:53:21 | [696][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0198 ntime: 0080 mem: 3.36 + 04-04 12:53:26 | [696][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1016 ntime: 0081 mem: 3.36 + 04-04 12:53:32 | [696][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1295 ntime: 0087 mem: 3.36 + 04-04 12:53:38 | [696][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0139 ntime: 0086 mem: 3.36 + 04-04 12:53:44 | [696][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0259 ntime: 0078 mem: 3.36 + 04-04 12:53:50 | [696][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0694 ntime: 0081 mem: 3.36 + 04-04 12:53:57 | [696][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0082 ntime: 0082 mem: 3.36 + 04-04 12:54:04 | [696][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0097 ntime: 0081 mem: 3.36 + 04-04 12:54:10 | [696][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1122 ntime: 0077 mem: 3.36 + 04-04 12:54:15 | [696][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0776 ntime: 0080 mem: 3.36 + 04-04 12:54:21 | Time info >>>> elapsed: 930.98 mins remain: 404.72 mins + 04-04 12:54:21 | [697][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0081 ntime: 0081 mem: 3.36 + 04-04 12:54:28 | [697][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0613 ntime: 0084 mem: 3.36 + 04-04 12:54:35 | [697][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0726 ntime: 0085 mem: 3.36 + 04-04 12:54:41 | [697][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0100 ntime: 0078 mem: 3.36 + 04-04 12:54:46 | [697][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0526 ntime: 0076 mem: 3.36 + 04-04 12:54:51 | [697][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0073 ntime: 0084 mem: 3.36 + 04-04 12:54:57 | [697][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0111 ntime: 0079 mem: 3.36 + 04-04 12:55:03 | [697][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0125 ntime: 0080 mem: 3.36 + 04-04 12:55:09 | [697][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1182 ntime: 0083 mem: 3.36 + 04-04 12:55:15 | [697][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0106 ntime: 0079 mem: 3.36 + 04-04 12:55:21 | [697][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0081 mem: 3.36 + 04-04 12:55:29 | [697][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0142 ntime: 0087 mem: 3.36 + 04-04 12:55:36 | [697][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0486 ntime: 0080 mem: 3.36 + 04-04 12:55:41 | [697][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1037 ntime: 0077 mem: 3.36 + 04-04 12:55:50 | [697][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0335 ntime: 0081 mem: 3.36 + 04-04 12:55:57 | [697][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0367 ntime: 0083 mem: 3.36 + 04-04 12:56:04 | [697][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0123 ntime: 0081 mem: 3.36 + 04-04 12:56:10 | [697][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0702 ntime: 0084 mem: 3.36 + 04-04 12:56:16 | Time info >>>> elapsed: 932.89 mins remain: 403.63 mins + 04-04 12:56:17 | [698][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0973 ntime: 0080 mem: 3.36 + 04-04 12:56:24 | [698][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0134 ntime: 0081 mem: 3.36 + 04-04 12:56:29 | [698][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0090 ntime: 0081 mem: 3.36 + 04-04 12:56:34 | [698][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0541 ntime: 0080 mem: 3.36 + 04-04 12:56:40 | [698][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0357 ntime: 0077 mem: 3.36 + 04-04 12:56:46 | [698][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1109 ntime: 0082 mem: 3.36 + 04-04 12:56:52 | [698][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 12:56:58 | [698][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0077 ntime: 0084 mem: 3.36 + 04-04 12:57:04 | [698][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1116 ntime: 0078 mem: 3.36 + 04-04 12:57:09 | [698][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0080 mem: 3.36 + 04-04 12:57:14 | [698][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0078 mem: 3.36 + 04-04 12:57:21 | [698][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0522 ntime: 0081 mem: 3.36 + 04-04 12:57:26 | [698][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0248 ntime: 0086 mem: 3.36 + 04-04 12:57:33 | [698][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1070 ntime: 0079 mem: 3.36 + 04-04 12:57:40 | [698][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1017 ntime: 0087 mem: 3.36 + 04-04 12:57:46 | [698][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0542 ntime: 0083 mem: 3.36 + 04-04 12:57:54 | [698][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1142 ntime: 0083 mem: 3.36 + 04-04 12:58:00 | [698][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0300 ntime: 0076 mem: 3.36 + 04-04 12:58:04 | Time info >>>> elapsed: 934.70 mins remain: 402.50 mins + 04-04 12:58:05 | [699][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0531 ntime: 0076 mem: 3.36 + 04-04 12:58:11 | [699][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0115 ntime: 0076 mem: 3.36 + 04-04 12:58:17 | [699][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0622 ntime: 0083 mem: 3.36 + 04-04 12:58:22 | [699][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0807 ntime: 0083 mem: 3.36 + 04-04 12:58:28 | [699][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0818 ntime: 0081 mem: 3.36 + 04-04 12:58:33 | [699][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0409 ntime: 0087 mem: 3.36 + 04-04 12:58:38 | [699][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0092 ntime: 0083 mem: 3.36 + 04-04 12:58:45 | [699][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0067 ntime: 0080 mem: 3.36 + 04-04 12:58:51 | [699][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0742 ntime: 0088 mem: 3.36 + 04-04 12:58:55 | [699][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0382 ntime: 0079 mem: 3.36 + 04-04 12:59:02 | [699][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0080 mem: 3.36 + 04-04 12:59:08 | [699][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0389 ntime: 0079 mem: 3.36 + 04-04 12:59:14 | [699][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0046 ntime: 0087 mem: 3.36 + 04-04 12:59:20 | [699][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0163 ntime: 0078 mem: 3.36 + 04-04 12:59:27 | [699][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0200 ntime: 0087 mem: 3.36 + 04-04 12:59:33 | [699][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1313 ntime: 0080 mem: 3.36 + 04-04 12:59:41 | [699][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0834 ntime: 0081 mem: 3.36 + 04-04 12:59:44 | [699][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0080 mem: 3.36 + 04-04 12:59:50 | Time info >>>> elapsed: 936.46 mins remain: 401.34 mins + 04-04 12:59:50 | [700][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0077 mem: 3.36 + 04-04 12:59:56 | [700][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0153 ntime: 0080 mem: 3.36 + 04-04 13:00:03 | [700][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 13:00:08 | [700][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1107 ntime: 0078 mem: 3.36 + 04-04 13:00:15 | [700][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0361 ntime: 0080 mem: 3.36 + 04-04 13:00:21 | [700][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0994 ntime: 0078 mem: 3.36 + 04-04 13:00:24 | [700][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0092 ntime: 0080 mem: 3.36 + 04-04 13:00:30 | [700][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0080 mem: 3.36 + 04-04 13:00:37 | [700][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0078 mem: 3.36 + 04-04 13:00:46 | [700][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1033 ntime: 0079 mem: 3.36 + 04-04 13:00:51 | [700][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0065 ntime: 0075 mem: 3.36 + 04-04 13:00:57 | [700][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0744 ntime: 0079 mem: 3.36 + 04-04 13:01:02 | [700][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0865 ntime: 0079 mem: 3.36 + 04-04 13:01:08 | [700][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0842 ntime: 0083 mem: 3.36 + 04-04 13:01:12 | [700][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0102 ntime: 0082 mem: 3.36 + 04-04 13:01:18 | [700][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0799 ntime: 0090 mem: 3.36 + 04-04 13:01:23 | [700][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0268 ntime: 0081 mem: 3.36 + 04-04 13:01:29 | [700][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0934 ntime: 0076 mem: 3.36 + 04-04 13:01:33 | Time info >>>> elapsed: 938.18 mins remain: 400.17 mins + 04-04 13:01:35 | [701][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1472 ntime: 0080 mem: 3.36 + 04-04 13:01:41 | [701][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0083 mem: 3.36 + 04-04 13:01:47 | [701][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1172 ntime: 0080 mem: 3.36 + 04-04 13:01:52 | [701][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0117 ntime: 0083 mem: 3.36 + 04-04 13:01:57 | [701][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 13:02:05 | [701][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1084 ntime: 0078 mem: 3.36 + 04-04 13:02:10 | [701][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0484 ntime: 0080 mem: 3.36 + 04-04 13:02:15 | [701][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0133 ntime: 0081 mem: 3.36 + 04-04 13:02:21 | [701][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0525 ntime: 0080 mem: 3.36 + 04-04 13:02:27 | [701][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1051 ntime: 0086 mem: 3.36 + 04-04 13:02:36 | [701][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0471 ntime: 0086 mem: 3.36 + 04-04 13:02:41 | [701][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0086 mem: 3.36 + 04-04 13:02:48 | [701][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1059 ntime: 0079 mem: 3.36 + 04-04 13:02:54 | [701][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0079 mem: 3.36 + 04-04 13:03:00 | [701][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0090 ntime: 0075 mem: 3.36 + 04-04 13:03:06 | [701][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0151 ntime: 0086 mem: 3.36 + 04-04 13:03:13 | [701][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0740 ntime: 0076 mem: 3.36 + 04-04 13:03:21 | [701][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1082 ntime: 0079 mem: 3.36 + 04-04 13:03:25 | Time info >>>> elapsed: 940.04 mins remain: 399.05 mins + 04-04 13:03:25 | [702][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0145 ntime: 0084 mem: 3.36 + 04-04 13:03:31 | [702][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1027 ntime: 0077 mem: 3.36 + 04-04 13:03:38 | [702][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1177 ntime: 0087 mem: 3.36 + 04-04 13:03:45 | [702][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1233 ntime: 0086 mem: 3.36 + 04-04 13:03:51 | [702][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1107 ntime: 0081 mem: 3.36 + 04-04 13:03:58 | [702][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1001 ntime: 0080 mem: 3.36 + 04-04 13:04:03 | [702][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0080 mem: 3.36 + 04-04 13:04:10 | [702][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0167 ntime: 0080 mem: 3.36 + 04-04 13:04:15 | [702][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0081 mem: 3.36 + 04-04 13:04:22 | [702][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0159 ntime: 0081 mem: 3.36 + 04-04 13:04:27 | [702][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0084 ntime: 0079 mem: 3.36 + 04-04 13:04:31 | [702][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0079 mem: 3.36 + 04-04 13:04:36 | [702][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0099 ntime: 0082 mem: 3.36 + 04-04 13:04:43 | [702][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1673 ntime: 0083 mem: 3.36 + 04-04 13:04:48 | [702][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0087 mem: 3.36 + 04-04 13:04:55 | [702][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0737 ntime: 0076 mem: 3.36 + 04-04 13:05:01 | [702][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0968 ntime: 0081 mem: 3.36 + 04-04 13:05:06 | [702][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0083 ntime: 0083 mem: 3.36 + 04-04 13:05:11 | Time info >>>> elapsed: 941.80 mins remain: 397.89 mins + 04-04 13:05:11 | [703][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0539 ntime: 0084 mem: 3.36 + 04-04 13:05:18 | [703][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1018 ntime: 0084 mem: 3.36 + 04-04 13:05:25 | [703][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0702 ntime: 0084 mem: 3.36 + 04-04 13:05:29 | [703][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0532 ntime: 0080 mem: 3.36 + 04-04 13:05:34 | [703][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0085 mem: 3.36 + 04-04 13:05:41 | [703][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0826 ntime: 0082 mem: 3.36 + 04-04 13:05:49 | [703][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0111 ntime: 0084 mem: 3.36 + 04-04 13:05:55 | [703][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1163 ntime: 0083 mem: 3.36 + 04-04 13:06:01 | [703][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0753 ntime: 0085 mem: 3.36 + 04-04 13:06:07 | [703][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0104 ntime: 0079 mem: 3.36 + 04-04 13:06:12 | [703][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0133 ntime: 0081 mem: 3.36 + 04-04 13:06:23 | [703][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0140 ntime: 0079 mem: 3.36 + 04-04 13:06:31 | [703][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1128 ntime: 0075 mem: 3.36 + 04-04 13:06:38 | [703][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0089 mem: 3.36 + 04-04 13:06:46 | [703][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0123 ntime: 0079 mem: 3.36 + 04-04 13:06:52 | [703][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0133 ntime: 0081 mem: 3.36 + 04-04 13:06:58 | [703][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0913 ntime: 0083 mem: 3.36 + 04-04 13:07:07 | [703][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0824 ntime: 0076 mem: 3.36 + 04-04 13:07:13 | Time info >>>> elapsed: 943.84 mins remain: 396.84 mins + 04-04 13:07:13 | [704][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0081 mem: 3.36 + 04-04 13:07:23 | [704][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0809 ntime: 0079 mem: 3.36 + 04-04 13:07:31 | [704][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 13:07:36 | [704][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0084 mem: 3.36 + 04-04 13:07:43 | [704][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0137 ntime: 0075 mem: 3.36 + 04-04 13:07:50 | [704][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 13:07:57 | [704][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 13:08:03 | [704][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0327 ntime: 0078 mem: 3.36 + 04-04 13:08:09 | [704][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0597 ntime: 0088 mem: 3.36 + 04-04 13:08:19 | [704][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0979 ntime: 0083 mem: 3.36 + 04-04 13:08:27 | [704][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0820 ntime: 0085 mem: 3.36 + 04-04 13:08:33 | [704][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0081 ntime: 0077 mem: 3.36 + 04-04 13:08:40 | [704][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0200 ntime: 0091 mem: 3.36 + 04-04 13:08:47 | [704][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1452 ntime: 0071 mem: 3.36 + 04-04 13:08:54 | [704][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1400 ntime: 0080 mem: 3.36 + 04-04 13:09:00 | [704][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0065 ntime: 0077 mem: 3.36 + 04-04 13:09:07 | [704][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0105 ntime: 0078 mem: 3.36 + 04-04 13:09:13 | [704][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0167 ntime: 0080 mem: 3.36 + 04-04 13:09:19 | Time info >>>> elapsed: 945.95 mins remain: 395.82 mins + 04-04 13:09:20 | [705][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0714 ntime: 0076 mem: 3.36 + 04-04 13:09:27 | [705][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1197 ntime: 0085 mem: 3.36 + 04-04 13:09:32 | [705][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1201 ntime: 0081 mem: 3.36 + 04-04 13:09:39 | [705][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0089 mem: 3.36 + 04-04 13:09:45 | [705][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0565 ntime: 0086 mem: 3.36 + 04-04 13:09:52 | [705][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0103 ntime: 0075 mem: 3.36 + 04-04 13:10:00 | [705][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0895 ntime: 0088 mem: 3.36 + 04-04 13:10:08 | [705][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1616 ntime: 0083 mem: 3.36 + 04-04 13:10:15 | [705][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1419 ntime: 0076 mem: 3.36 + 04-04 13:10:22 | [705][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1149 ntime: 0086 mem: 3.36 + 04-04 13:10:29 | [705][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1441 ntime: 0079 mem: 3.36 + 04-04 13:10:37 | [705][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0752 ntime: 0080 mem: 3.36 + 04-04 13:10:42 | [705][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1099 ntime: 0089 mem: 3.36 + 04-04 13:10:49 | [705][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0082 mem: 3.36 + 04-04 13:10:56 | [705][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0768 ntime: 0076 mem: 3.36 + 04-04 13:11:03 | [705][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0912 ntime: 0078 mem: 3.36 + 04-04 13:11:10 | [705][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0515 ntime: 0078 mem: 3.36 + 04-04 13:11:17 | [705][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0103 ntime: 0086 mem: 3.36 + 04-04 13:11:23 | Time info >>>> elapsed: 948.01 mins remain: 394.78 mins + 04-04 13:11:23 | [706][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0074 mem: 3.36 + 04-04 13:11:30 | [706][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0087 mem: 3.36 + 04-04 13:11:38 | [706][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1479 ntime: 0078 mem: 3.36 + 04-04 13:11:46 | [706][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0216 ntime: 0076 mem: 3.36 + 04-04 13:11:53 | [706][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0492 ntime: 0078 mem: 3.36 + 04-04 13:11:59 | [706][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0127 ntime: 0079 mem: 3.36 + 04-04 13:12:04 | [706][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0075 mem: 3.36 + 04-04 13:12:10 | [706][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0078 ntime: 0087 mem: 3.36 + 04-04 13:12:18 | [706][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1272 ntime: 0088 mem: 3.36 + 04-04 13:12:24 | [706][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0423 ntime: 0092 mem: 3.36 + 04-04 13:12:32 | [706][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0080 ntime: 0079 mem: 3.36 + 04-04 13:12:37 | [706][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0668 ntime: 0085 mem: 3.36 + 04-04 13:12:45 | [706][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 13:12:53 | [706][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1350 ntime: 0076 mem: 3.36 + 04-04 13:12:58 | [706][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0073 ntime: 0084 mem: 3.36 + 04-04 13:13:05 | [706][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0932 ntime: 0078 mem: 3.36 + 04-04 13:13:11 | [706][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1210 ntime: 0077 mem: 3.36 + 04-04 13:13:16 | [706][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0799 ntime: 0077 mem: 3.36 + 04-04 13:13:23 | Time info >>>> elapsed: 950.02 mins remain: 393.71 mins + 04-04 13:13:24 | [707][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0121 ntime: 0088 mem: 3.36 + 04-04 13:13:31 | [707][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1521 ntime: 0081 mem: 3.36 + 04-04 13:13:40 | [707][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1543 ntime: 0083 mem: 3.36 + 04-04 13:13:47 | [707][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1451 ntime: 0089 mem: 3.36 + 04-04 13:13:55 | [707][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0082 mem: 3.36 + 04-04 13:14:03 | [707][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1575 ntime: 0086 mem: 3.36 + 04-04 13:14:12 | [707][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0940 ntime: 0081 mem: 3.36 + 04-04 13:14:19 | [707][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1058 ntime: 0081 mem: 3.36 + 04-04 13:14:25 | [707][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0082 ntime: 0081 mem: 3.36 + 04-04 13:14:34 | [707][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0693 ntime: 0081 mem: 3.36 + 04-04 13:14:40 | [707][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0719 ntime: 0090 mem: 3.36 + 04-04 13:14:48 | [707][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1091 ntime: 0076 mem: 3.36 + 04-04 13:14:55 | [707][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1508 ntime: 0082 mem: 3.36 + 04-04 13:14:59 | [707][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0130 ntime: 0079 mem: 3.36 + 04-04 13:15:08 | [707][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0107 ntime: 0080 mem: 3.36 + 04-04 13:15:15 | [707][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1283 ntime: 0086 mem: 3.36 + 04-04 13:15:23 | [707][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1007 ntime: 0082 mem: 3.36 + 04-04 13:15:32 | [707][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0681 ntime: 0080 mem: 3.36 + 04-04 13:15:37 | Time info >>>> elapsed: 952.25 mins remain: 392.74 mins + 04-04 13:15:37 | [708][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0083 mem: 3.36 + 04-04 13:15:43 | [708][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0087 mem: 3.36 + 04-04 13:15:51 | [708][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0076 mem: 3.36 + 04-04 13:15:58 | [708][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1558 ntime: 0075 mem: 3.36 + 04-04 13:16:04 | [708][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 13:16:12 | [708][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1310 ntime: 0081 mem: 3.36 + 04-04 13:16:19 | [708][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0114 ntime: 0089 mem: 3.36 + 04-04 13:16:27 | [708][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0196 ntime: 0085 mem: 3.36 + 04-04 13:16:35 | [708][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1361 ntime: 0077 mem: 3.36 + 04-04 13:16:41 | [708][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1215 ntime: 0085 mem: 3.36 + 04-04 13:16:47 | [708][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0811 ntime: 0082 mem: 3.36 + 04-04 13:16:54 | [708][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1269 ntime: 0079 mem: 3.36 + 04-04 13:17:02 | [708][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1332 ntime: 0080 mem: 3.36 + 04-04 13:17:09 | [708][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0071 mem: 3.36 + 04-04 13:17:18 | [708][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1040 ntime: 0089 mem: 3.36 + 04-04 13:17:25 | [708][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0090 ntime: 0080 mem: 3.36 + 04-04 13:17:33 | [708][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0540 ntime: 0077 mem: 3.36 + 04-04 13:17:39 | [708][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1168 ntime: 0081 mem: 3.36 + 04-04 13:17:47 | Time info >>>> elapsed: 954.41 mins remain: 391.72 mins + 04-04 13:17:48 | [709][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1027 ntime: 0074 mem: 3.36 + 04-04 13:17:55 | [709][010/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0708 ntime: 0073 mem: 3.36 + 04-04 13:18:01 | [709][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0728 ntime: 0082 mem: 3.36 + 04-04 13:18:09 | [709][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0666 ntime: 0080 mem: 3.36 + 04-04 13:18:17 | [709][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1501 ntime: 0078 mem: 3.36 + 04-04 13:18:23 | [709][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0443 ntime: 0077 mem: 3.36 + 04-04 13:18:31 | [709][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1031 ntime: 0084 mem: 3.36 + 04-04 13:18:39 | [709][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0902 ntime: 0084 mem: 3.36 + 04-04 13:18:46 | [709][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1093 ntime: 0078 mem: 3.36 + 04-04 13:18:53 | [709][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1316 ntime: 0085 mem: 3.36 + 04-04 13:19:00 | [709][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0079 mem: 3.36 + 04-04 13:19:10 | [709][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1290 ntime: 0086 mem: 3.36 + 04-04 13:19:19 | [709][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0908 ntime: 0078 mem: 3.36 + 04-04 13:19:25 | [709][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0505 ntime: 0069 mem: 3.36 + 04-04 13:19:33 | [709][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0078 mem: 3.36 + 04-04 13:19:41 | [709][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0100 ntime: 0079 mem: 3.36 + 04-04 13:19:50 | [709][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1578 ntime: 0078 mem: 3.36 + 04-04 13:19:58 | [709][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0693 ntime: 0078 mem: 3.36 + 04-04 13:20:05 | Time info >>>> elapsed: 956.70 mins remain: 390.77 mins + 04-04 13:20:05 | [710][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0604 ntime: 0083 mem: 3.36 + 04-04 13:20:11 | [710][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0084 ntime: 0086 mem: 3.36 + 04-04 13:20:18 | [710][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0123 ntime: 0084 mem: 3.36 + 04-04 13:20:26 | [710][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1180 ntime: 0079 mem: 3.36 + 04-04 13:20:32 | [710][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0485 ntime: 0086 mem: 3.36 + 04-04 13:20:40 | [710][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0682 ntime: 0075 mem: 3.36 + 04-04 13:20:47 | [710][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0495 ntime: 0079 mem: 3.36 + 04-04 13:20:55 | [710][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1498 ntime: 0093 mem: 3.36 + 04-04 13:21:03 | [710][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0110 ntime: 0087 mem: 3.36 + 04-04 13:21:09 | [710][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0184 ntime: 0076 mem: 3.36 + 04-04 13:21:13 | [710][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0086 ntime: 0078 mem: 3.36 + 04-04 13:21:20 | [710][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1427 ntime: 0077 mem: 3.36 + 04-04 13:21:26 | [710][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1659 ntime: 0083 mem: 3.36 + 04-04 13:21:33 | [710][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0525 ntime: 0072 mem: 3.36 + 04-04 13:21:37 | [710][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0087 ntime: 0085 mem: 3.36 + 04-04 13:21:45 | [710][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1259 ntime: 0082 mem: 3.36 + 04-04 13:21:49 | [710][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0075 mem: 3.36 + 04-04 13:21:55 | [710][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0073 mem: 3.36 + 04-04 13:22:00 | Time info >>>> elapsed: 958.62 mins remain: 389.65 mins + 04-04 13:22:01 | [711][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1161 ntime: 0077 mem: 3.36 + 04-04 13:22:06 | [711][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1164 ntime: 0083 mem: 3.36 + 04-04 13:22:13 | [711][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1269 ntime: 0079 mem: 3.36 + 04-04 13:22:19 | [711][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1090 ntime: 0082 mem: 3.36 + 04-04 13:22:27 | [711][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1549 ntime: 0081 mem: 3.36 + 04-04 13:22:34 | [711][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0716 ntime: 0090 mem: 3.36 + 04-04 13:22:41 | [711][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0080 mem: 3.36 + 04-04 13:22:48 | [711][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0079 mem: 3.36 + 04-04 13:22:55 | [711][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0347 ntime: 0079 mem: 3.36 + 04-04 13:23:03 | [711][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0761 ntime: 0081 mem: 3.36 + 04-04 13:23:10 | [711][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0067 ntime: 0084 mem: 3.36 + 04-04 13:23:16 | [711][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0332 ntime: 0078 mem: 3.36 + 04-04 13:23:24 | [711][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0786 ntime: 0079 mem: 3.36 + 04-04 13:23:31 | [711][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0085 ntime: 0081 mem: 3.36 + 04-04 13:23:37 | [711][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0108 ntime: 0058 mem: 3.36 + 04-04 13:23:44 | [711][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0675 ntime: 0075 mem: 3.36 + 04-04 13:23:52 | [711][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0077 mem: 3.36 + 04-04 13:24:00 | [711][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0760 ntime: 0071 mem: 3.36 + 04-04 13:24:06 | Time info >>>> elapsed: 960.73 mins remain: 388.61 mins + 04-04 13:24:06 | [712][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0077 mem: 3.36 + 04-04 13:24:13 | [712][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0676 ntime: 0082 mem: 3.36 + 04-04 13:24:18 | [712][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1351 ntime: 0076 mem: 3.36 + 04-04 13:24:25 | [712][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1088 ntime: 0080 mem: 3.36 + 04-04 13:24:33 | [712][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1326 ntime: 0082 mem: 3.36 + 04-04 13:24:39 | [712][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0694 ntime: 0089 mem: 3.36 + 04-04 13:24:46 | [712][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0336 ntime: 0073 mem: 3.36 + 04-04 13:24:54 | [712][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0870 ntime: 0087 mem: 3.36 + 04-04 13:25:01 | [712][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0690 ntime: 0088 mem: 3.36 + 04-04 13:25:07 | [712][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0081 mem: 3.36 + 04-04 13:25:14 | [712][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0881 ntime: 0082 mem: 3.36 + 04-04 13:25:20 | [712][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1193 ntime: 0080 mem: 3.36 + 04-04 13:25:26 | [712][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0675 ntime: 0084 mem: 3.36 + 04-04 13:25:30 | [712][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0068 ntime: 0075 mem: 3.36 + 04-04 13:25:39 | [712][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1304 ntime: 0081 mem: 3.36 + 04-04 13:25:46 | [712][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0223 ntime: 0077 mem: 3.36 + 04-04 13:25:52 | [712][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0153 ntime: 0081 mem: 3.36 + 04-04 13:26:01 | [712][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1348 ntime: 0085 mem: 3.36 + 04-04 13:26:06 | Time info >>>> elapsed: 962.72 mins remain: 387.52 mins + 04-04 13:26:06 | [713][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0091 mem: 3.36 + 04-04 13:26:11 | [713][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0080 mem: 3.36 + 04-04 13:26:16 | [713][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1021 ntime: 0083 mem: 3.36 + 04-04 13:26:23 | [713][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0292 ntime: 0076 mem: 3.36 + 04-04 13:26:29 | [713][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0081 mem: 3.36 + 04-04 13:26:36 | [713][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0196 ntime: 0059 mem: 3.36 + 04-04 13:26:43 | [713][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1671 ntime: 0088 mem: 3.36 + 04-04 13:26:48 | [713][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0070 ntime: 0073 mem: 3.36 + 04-04 13:26:55 | [713][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0622 ntime: 0082 mem: 3.36 + 04-04 13:27:01 | [713][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0140 ntime: 0079 mem: 3.36 + 04-04 13:27:09 | [713][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0122 ntime: 0078 mem: 3.36 + 04-04 13:27:15 | [713][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0504 ntime: 0073 mem: 3.36 + 04-04 13:27:22 | [713][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1070 ntime: 0074 mem: 3.36 + 04-04 13:27:30 | [713][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0079 mem: 3.36 + 04-04 13:27:37 | [713][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0552 ntime: 0078 mem: 3.36 + 04-04 13:27:42 | [713][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0078 mem: 3.36 + 04-04 13:27:49 | [713][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0713 ntime: 0077 mem: 3.36 + 04-04 13:27:56 | [713][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0085 mem: 3.36 + 04-04 13:28:01 | Time info >>>> elapsed: 964.64 mins remain: 386.40 mins + 04-04 13:28:01 | [714][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0096 ntime: 0080 mem: 3.36 + 04-04 13:28:06 | [714][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0146 ntime: 0078 mem: 3.36 + 04-04 13:28:14 | [714][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0488 ntime: 0084 mem: 3.36 + 04-04 13:28:19 | [714][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0808 ntime: 0089 mem: 3.36 + 04-04 13:28:25 | [714][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0092 ntime: 0079 mem: 3.36 + 04-04 13:28:32 | [714][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1065 ntime: 0081 mem: 3.36 + 04-04 13:28:39 | [714][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1324 ntime: 0079 mem: 3.36 + 04-04 13:28:45 | [714][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0865 ntime: 0085 mem: 3.36 + 04-04 13:28:53 | [714][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1303 ntime: 0081 mem: 3.36 + 04-04 13:29:01 | [714][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0930 ntime: 0073 mem: 3.36 + 04-04 13:29:07 | [714][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0062 ntime: 0079 mem: 3.36 + 04-04 13:29:14 | [714][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0079 mem: 3.36 + 04-04 13:29:20 | [714][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0983 ntime: 0083 mem: 3.36 + 04-04 13:29:27 | [714][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1074 ntime: 0085 mem: 3.36 + 04-04 13:29:34 | [714][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1328 ntime: 0059 mem: 3.36 + 04-04 13:29:39 | [714][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0177 ntime: 0082 mem: 3.36 + 04-04 13:29:47 | [714][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1027 ntime: 0079 mem: 3.36 + 04-04 13:29:53 | [714][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0065 ntime: 0081 mem: 3.36 + 04-04 13:29:58 | Time info >>>> elapsed: 966.60 mins remain: 385.29 mins + 04-04 13:29:59 | [715][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1024 ntime: 0079 mem: 3.36 + 04-04 13:30:07 | [715][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0855 ntime: 0079 mem: 3.36 + 04-04 13:30:14 | [715][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0764 ntime: 0080 mem: 3.36 + 04-04 13:30:23 | [715][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0766 ntime: 0087 mem: 3.36 + 04-04 13:30:31 | [715][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0912 ntime: 0077 mem: 3.36 + 04-04 13:30:37 | [715][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0099 ntime: 0081 mem: 3.36 + 04-04 13:30:44 | [715][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0082 ntime: 0076 mem: 3.36 + 04-04 13:30:50 | [715][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0115 ntime: 0087 mem: 3.36 + 04-04 13:30:59 | [715][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1526 ntime: 0078 mem: 3.36 + 04-04 13:31:07 | [715][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1195 ntime: 0077 mem: 3.36 + 04-04 13:31:14 | [715][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0100 ntime: 0079 mem: 3.36 + 04-04 13:31:21 | [715][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0180 ntime: 0080 mem: 3.36 + 04-04 13:31:29 | [715][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1077 ntime: 0078 mem: 3.36 + 04-04 13:31:35 | [715][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0149 ntime: 0076 mem: 3.36 + 04-04 13:31:42 | [715][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0924 ntime: 0084 mem: 3.36 + 04-04 13:31:46 | [715][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0266 ntime: 0079 mem: 3.36 + 04-04 13:31:54 | [715][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0100 ntime: 0084 mem: 3.36 + 04-04 13:31:59 | [715][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1122 ntime: 0078 mem: 3.36 + 04-04 13:32:04 | Time info >>>> elapsed: 968.70 mins remain: 384.23 mins + 04-04 13:32:05 | [716][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0349 ntime: 0090 mem: 3.36 + 04-04 13:32:12 | [716][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 13:32:18 | [716][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0831 ntime: 0082 mem: 3.36 + 04-04 13:32:24 | [716][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0130 ntime: 0077 mem: 3.36 + 04-04 13:32:31 | [716][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0734 ntime: 0074 mem: 3.36 + 04-04 13:32:39 | [716][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1035 ntime: 0076 mem: 3.36 + 04-04 13:32:46 | [716][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0825 ntime: 0072 mem: 3.36 + 04-04 13:32:53 | [716][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0118 ntime: 0076 mem: 3.36 + 04-04 13:32:58 | [716][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0712 ntime: 0079 mem: 3.36 + 04-04 13:33:06 | [716][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0860 ntime: 0078 mem: 3.36 + 04-04 13:33:12 | [716][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0703 ntime: 0082 mem: 3.36 + 04-04 13:33:18 | [716][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0095 ntime: 0079 mem: 3.36 + 04-04 13:33:25 | [716][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0071 ntime: 0079 mem: 3.36 + 04-04 13:33:31 | [716][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0088 ntime: 0079 mem: 3.36 + 04-04 13:33:37 | [716][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0141 ntime: 0078 mem: 3.36 + 04-04 13:33:44 | [716][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0090 mem: 3.36 + 04-04 13:33:53 | [716][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1447 ntime: 0086 mem: 3.36 + 04-04 13:33:59 | [716][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0966 ntime: 0080 mem: 3.36 + 04-04 13:34:05 | Time info >>>> elapsed: 970.71 mins remain: 383.14 mins + 04-04 13:34:06 | [717][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1405 ntime: 0072 mem: 3.36 + 04-04 13:34:15 | [717][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1056 ntime: 0076 mem: 3.36 + 04-04 13:34:21 | [717][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0879 ntime: 0078 mem: 3.36 + 04-04 13:34:29 | [717][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1775 ntime: 0073 mem: 3.36 + 04-04 13:34:35 | [717][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0076 ntime: 0080 mem: 3.36 + 04-04 13:34:42 | [717][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0450 ntime: 0085 mem: 3.36 + 04-04 13:34:48 | [717][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0087 ntime: 0080 mem: 3.36 + 04-04 13:34:56 | [717][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0075 ntime: 0081 mem: 3.36 + 04-04 13:35:04 | [717][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0843 ntime: 0083 mem: 3.36 + 04-04 13:35:10 | [717][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0829 ntime: 0089 mem: 3.36 + 04-04 13:35:15 | [717][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0121 ntime: 0079 mem: 3.36 + 04-04 13:35:22 | [717][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0457 ntime: 0081 mem: 3.36 + 04-04 13:35:29 | [717][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0074 ntime: 0085 mem: 3.36 + 04-04 13:35:35 | [717][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1009 ntime: 0080 mem: 3.36 + 04-04 13:35:40 | [717][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0098 ntime: 0084 mem: 3.36 + 04-04 13:35:46 | [717][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1013 ntime: 0078 mem: 3.36 + 04-04 13:35:52 | [717][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0905 ntime: 0075 mem: 3.36 + 04-04 13:35:59 | [717][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0809 ntime: 0076 mem: 3.36 + 04-04 13:36:04 | Time info >>>> elapsed: 972.70 mins remain: 382.04 mins + 04-04 13:36:05 | [718][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0544 ntime: 0073 mem: 3.36 + 04-04 13:36:12 | [718][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0953 ntime: 0080 mem: 3.36 + 04-04 13:36:17 | [718][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0077 mem: 3.36 + 04-04 13:36:24 | [718][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0803 ntime: 0075 mem: 3.36 + 04-04 13:36:30 | [718][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0097 ntime: 0084 mem: 3.36 + 04-04 13:36:36 | [718][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0081 mem: 3.36 + 04-04 13:36:42 | [718][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0618 ntime: 0076 mem: 3.36 + 04-04 13:36:50 | [718][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1377 ntime: 0080 mem: 3.36 + 04-04 13:36:55 | [718][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0044 ntime: 0078 mem: 3.36 + 04-04 13:37:02 | [718][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1539 ntime: 0082 mem: 3.36 + 04-04 13:37:08 | [718][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0107 ntime: 0078 mem: 3.36 + 04-04 13:37:15 | [718][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 13:37:21 | [718][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0130 ntime: 0080 mem: 3.36 + 04-04 13:37:28 | [718][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0184 ntime: 0077 mem: 3.36 + 04-04 13:37:40 | [718][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0083 mem: 3.36 + 04-04 13:37:48 | [718][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1155 ntime: 0081 mem: 3.36 + 04-04 13:37:53 | [718][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0085 mem: 3.36 + 04-04 13:38:00 | [718][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0139 ntime: 0090 mem: 3.36 + 04-04 13:38:07 | Time info >>>> elapsed: 974.74 mins remain: 380.95 mins + 04-04 13:38:07 | [719][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0080 mem: 3.36 + 04-04 13:38:13 | [719][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1059 ntime: 0079 mem: 3.36 + 04-04 13:38:21 | [719][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1205 ntime: 0078 mem: 3.36 + 04-04 13:38:29 | [719][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0084 mem: 3.36 + 04-04 13:38:34 | [719][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0143 ntime: 0077 mem: 3.36 + 04-04 13:38:40 | [719][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0115 ntime: 0085 mem: 3.36 + 04-04 13:38:48 | [719][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0103 ntime: 0080 mem: 3.36 + 04-04 13:38:54 | [719][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0306 ntime: 0085 mem: 3.36 + 04-04 13:39:01 | [719][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0087 ntime: 0079 mem: 3.36 + 04-04 13:39:05 | [719][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0880 ntime: 0085 mem: 3.36 + 04-04 13:39:12 | [719][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0105 ntime: 0079 mem: 3.36 + 04-04 13:39:18 | [719][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0407 ntime: 0078 mem: 3.36 + 04-04 13:39:25 | [719][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1374 ntime: 0083 mem: 3.36 + 04-04 13:39:32 | [719][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1093 ntime: 0079 mem: 3.36 + 04-04 13:39:38 | [719][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0234 ntime: 0078 mem: 3.36 + 04-04 13:39:44 | [719][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0062 ntime: 0085 mem: 3.36 + 04-04 13:39:51 | [719][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0080 ntime: 0088 mem: 3.36 + 04-04 13:39:57 | [719][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1094 ntime: 0087 mem: 3.36 + 04-04 13:40:02 | Time info >>>> elapsed: 976.66 mins remain: 379.81 mins + 04-04 13:40:03 | [720][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 1015 ntime: 0085 mem: 3.36 + 04-04 13:40:09 | [720][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1564 ntime: 0078 mem: 3.36 + 04-04 13:40:16 | [720][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0939 ntime: 0077 mem: 3.36 + 04-04 13:40:21 | [720][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0128 ntime: 0080 mem: 3.36 + 04-04 13:40:30 | [720][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1086 ntime: 0077 mem: 3.36 + 04-04 13:40:35 | [720][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0167 ntime: 0077 mem: 3.36 + 04-04 13:40:42 | [720][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0538 ntime: 0083 mem: 3.36 + 04-04 13:40:50 | [720][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0108 ntime: 0081 mem: 3.36 + 04-04 13:40:56 | [720][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0164 ntime: 0083 mem: 3.36 + 04-04 13:41:02 | [720][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 13:41:08 | [720][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0699 ntime: 0086 mem: 3.36 + 04-04 13:41:14 | [720][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0713 ntime: 0077 mem: 3.36 + 04-04 13:41:20 | [720][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0972 ntime: 0089 mem: 3.36 + 04-04 13:41:25 | [720][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0092 ntime: 0083 mem: 3.36 + 04-04 13:41:31 | [720][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0809 ntime: 0078 mem: 3.36 + 04-04 13:41:38 | [720][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1428 ntime: 0084 mem: 3.36 + 04-04 13:41:44 | [720][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0940 ntime: 0080 mem: 3.36 + 04-04 13:41:50 | [720][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0817 ntime: 0077 mem: 3.36 + 04-04 13:41:57 | Time info >>>> elapsed: 978.57 mins remain: 378.67 mins + 04-04 13:41:58 | [721][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0912 ntime: 0087 mem: 3.36 + 04-04 13:42:04 | [721][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0670 ntime: 0078 mem: 3.36 + 04-04 13:42:09 | [721][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0811 ntime: 0070 mem: 3.36 + 04-04 13:42:14 | [721][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0099 ntime: 0078 mem: 3.36 + 04-04 13:42:20 | [721][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0137 ntime: 0083 mem: 3.36 + 04-04 13:42:25 | [721][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0079 mem: 3.36 + 04-04 13:42:30 | [721][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0310 ntime: 0080 mem: 3.36 + 04-04 13:42:37 | [721][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0076 mem: 3.36 + 04-04 13:42:43 | [721][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0077 mem: 3.36 + 04-04 13:42:47 | [721][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0079 ntime: 0086 mem: 3.36 + 04-04 13:42:54 | [721][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0086 mem: 3.36 + 04-04 13:43:00 | [721][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0108 ntime: 0079 mem: 3.36 + 04-04 13:43:07 | [721][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1055 ntime: 0084 mem: 3.36 + 04-04 13:43:13 | [721][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0700 ntime: 0077 mem: 3.36 + 04-04 13:43:19 | [721][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1060 ntime: 0080 mem: 3.36 + 04-04 13:43:23 | [721][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0107 ntime: 0086 mem: 3.36 + 04-04 13:43:28 | [721][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0068 ntime: 0081 mem: 3.36 + 04-04 13:43:34 | [721][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0399 ntime: 0085 mem: 3.36 + 04-04 13:43:38 | Time info >>>> elapsed: 980.25 mins remain: 377.44 mins + 04-04 13:43:38 | [722][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0601 ntime: 0077 mem: 3.36 + 04-04 13:43:46 | [722][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0094 ntime: 0077 mem: 3.36 + 04-04 13:43:51 | [722][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0068 ntime: 0088 mem: 3.36 + 04-04 13:43:57 | [722][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0420 ntime: 0080 mem: 3.36 + 04-04 13:44:03 | [722][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0878 ntime: 0078 mem: 3.36 + 04-04 13:44:08 | [722][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0102 ntime: 0080 mem: 3.36 + 04-04 13:44:14 | [722][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1273 ntime: 0084 mem: 3.36 + 04-04 13:44:20 | [722][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1197 ntime: 0086 mem: 3.36 + 04-04 13:44:26 | [722][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1025 ntime: 0088 mem: 3.36 + 04-04 13:44:32 | [722][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0099 ntime: 0076 mem: 3.36 + 04-04 13:44:39 | [722][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0080 mem: 3.36 + 04-04 13:44:45 | [722][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0079 mem: 3.36 + 04-04 13:44:53 | [722][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1094 ntime: 0082 mem: 3.36 + 04-04 13:44:57 | [722][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0085 mem: 3.36 + 04-04 13:45:04 | [722][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1410 ntime: 0082 mem: 3.36 + 04-04 13:45:09 | [722][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0771 ntime: 0083 mem: 3.36 + 04-04 13:45:16 | [722][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0098 ntime: 0081 mem: 3.36 + 04-04 13:45:23 | [722][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1562 ntime: 0083 mem: 3.36 + 04-04 13:45:27 | Time info >>>> elapsed: 982.07 mins remain: 376.26 mins + 04-04 13:45:27 | [723][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0133 ntime: 0086 mem: 3.36 + 04-04 13:45:33 | [723][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0988 ntime: 0077 mem: 3.36 + 04-04 13:45:37 | [723][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1351 ntime: 0080 mem: 3.36 + 04-04 13:45:42 | [723][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0598 ntime: 0086 mem: 3.36 + 04-04 13:45:48 | [723][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0995 ntime: 0083 mem: 3.36 + 04-04 13:45:53 | [723][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1023 ntime: 0080 mem: 3.36 + 04-04 13:45:59 | [723][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0623 ntime: 0084 mem: 3.36 + 04-04 13:46:04 | [723][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0157 ntime: 0083 mem: 3.36 + 04-04 13:46:10 | [723][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0559 ntime: 0086 mem: 3.36 + 04-04 13:46:15 | [723][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0078 ntime: 0080 mem: 3.36 + 04-04 13:46:20 | [723][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1686 ntime: 0076 mem: 3.36 + 04-04 13:46:26 | [723][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0494 ntime: 0086 mem: 3.36 + 04-04 13:46:34 | [723][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1237 ntime: 0078 mem: 3.36 + 04-04 13:46:39 | [723][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0078 mem: 3.36 + 04-04 13:46:45 | [723][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0126 ntime: 0079 mem: 3.36 + 04-04 13:46:51 | [723][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0608 ntime: 0081 mem: 3.36 + 04-04 13:46:57 | [723][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1098 ntime: 0078 mem: 3.36 + 04-04 13:47:03 | [723][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0076 mem: 3.36 + 04-04 13:47:08 | Time info >>>> elapsed: 983.76 mins remain: 375.02 mins + 04-04 13:47:08 | [724][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0641 ntime: 0081 mem: 3.36 + 04-04 13:47:13 | [724][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0143 ntime: 0077 mem: 3.36 + 04-04 13:47:19 | [724][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0119 ntime: 0080 mem: 3.36 + 04-04 13:47:24 | [724][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0068 ntime: 0086 mem: 3.36 + 04-04 13:47:31 | [724][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0769 ntime: 0058 mem: 3.36 + 04-04 13:47:39 | [724][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0526 ntime: 0080 mem: 3.36 + 04-04 13:47:46 | [724][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0135 ntime: 0080 mem: 3.36 + 04-04 13:47:53 | [724][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0337 ntime: 0083 mem: 3.36 + 04-04 13:47:57 | [724][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0934 ntime: 0079 mem: 3.36 + 04-04 13:48:04 | [724][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0084 mem: 3.36 + 04-04 13:48:11 | [724][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1213 ntime: 0083 mem: 3.36 + 04-04 13:48:18 | [724][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0161 ntime: 0080 mem: 3.36 + 04-04 13:48:24 | [724][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0094 ntime: 0080 mem: 3.36 + 04-04 13:48:29 | [724][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0413 ntime: 0085 mem: 3.36 + 04-04 13:48:34 | [724][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1135 ntime: 0084 mem: 3.36 + 04-04 13:48:39 | [724][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0158 ntime: 0083 mem: 3.36 + 04-04 13:48:46 | [724][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0609 ntime: 0087 mem: 3.36 + 04-04 13:48:52 | [724][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0892 ntime: 0082 mem: 3.36 + 04-04 13:48:57 | Time info >>>> elapsed: 985.58 mins remain: 373.84 mins + 04-04 13:48:57 | [725][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0061 ntime: 0083 mem: 3.36 + 04-04 13:49:03 | [725][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0142 ntime: 0078 mem: 3.36 + 04-04 13:49:08 | [725][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0087 mem: 3.36 + 04-04 13:49:15 | [725][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0496 ntime: 0084 mem: 3.36 + 04-04 13:49:21 | [725][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1202 ntime: 0080 mem: 3.36 + 04-04 13:49:26 | [725][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0083 mem: 3.36 + 04-04 13:49:34 | [725][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1021 ntime: 0086 mem: 3.36 + 04-04 13:49:41 | [725][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0846 ntime: 0079 mem: 3.36 + 04-04 13:49:46 | [725][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1095 ntime: 0086 mem: 3.36 + 04-04 13:49:52 | [725][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0095 ntime: 0087 mem: 3.36 + 04-04 13:50:00 | [725][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1078 ntime: 0079 mem: 3.36 + 04-04 13:50:06 | [725][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 13:50:11 | [725][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0094 ntime: 0073 mem: 3.36 + 04-04 13:50:19 | [725][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0753 ntime: 0081 mem: 3.36 + 04-04 13:50:25 | [725][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1435 ntime: 0082 mem: 3.36 + 04-04 13:50:31 | [725][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0080 mem: 3.36 + 04-04 13:50:39 | [725][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1436 ntime: 0078 mem: 3.36 + 04-04 13:50:47 | [725][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1329 ntime: 0076 mem: 3.36 + 04-04 13:50:51 | Time info >>>> elapsed: 987.48 mins remain: 372.69 mins + 04-04 13:50:52 | [726][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0698 ntime: 0073 mem: 3.36 + 04-04 13:50:59 | [726][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0112 ntime: 0085 mem: 3.36 + 04-04 13:51:08 | [726][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0537 ntime: 0083 mem: 3.36 + 04-04 13:51:14 | [726][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1102 ntime: 0078 mem: 3.36 + 04-04 13:51:20 | [726][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0531 ntime: 0084 mem: 3.36 + 04-04 13:51:25 | [726][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0656 ntime: 0083 mem: 3.36 + 04-04 13:51:31 | [726][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1123 ntime: 0084 mem: 3.36 + 04-04 13:51:36 | [726][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0488 ntime: 0076 mem: 3.36 + 04-04 13:51:43 | [726][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0957 ntime: 0080 mem: 3.36 + 04-04 13:51:48 | [726][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0945 ntime: 0077 mem: 3.36 + 04-04 13:51:54 | [726][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 13:51:58 | [726][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0077 mem: 3.36 + 04-04 13:52:07 | [726][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1298 ntime: 0077 mem: 3.36 + 04-04 13:52:13 | [726][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0950 ntime: 0080 mem: 3.36 + 04-04 13:52:18 | [726][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1301 ntime: 0078 mem: 3.36 + 04-04 13:52:23 | [726][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0082 mem: 3.36 + 04-04 13:52:28 | [726][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0725 ntime: 0085 mem: 3.36 + 04-04 13:52:34 | [726][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0046 ntime: 0078 mem: 3.36 + 04-04 13:52:38 | Time info >>>> elapsed: 989.26 mins remain: 371.48 mins + 04-04 13:52:39 | [727][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0885 ntime: 0082 mem: 3.36 + 04-04 13:52:44 | [727][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0642 ntime: 0085 mem: 3.36 + 04-04 13:52:49 | [727][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1006 ntime: 0079 mem: 3.36 + 04-04 13:52:52 | [727][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0171 ntime: 0080 mem: 3.36 + 04-04 13:52:58 | [727][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0081 ntime: 0074 mem: 3.36 + 04-04 13:53:03 | [727][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0084 ntime: 0086 mem: 3.36 + 04-04 13:53:09 | [727][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1375 ntime: 0089 mem: 3.36 + 04-04 13:53:14 | [727][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1274 ntime: 0078 mem: 3.36 + 04-04 13:53:19 | [727][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0409 ntime: 0080 mem: 3.36 + 04-04 13:53:25 | [727][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0423 ntime: 0081 mem: 3.36 + 04-04 13:53:31 | [727][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1009 ntime: 0079 mem: 3.36 + 04-04 13:53:38 | [727][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0083 mem: 3.36 + 04-04 13:53:45 | [727][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0140 ntime: 0085 mem: 3.36 + 04-04 13:53:50 | [727][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 13:53:56 | [727][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0995 ntime: 0078 mem: 3.36 + 04-04 13:54:02 | [727][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0610 ntime: 0079 mem: 3.36 + 04-04 13:54:09 | [727][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1229 ntime: 0082 mem: 3.36 + 04-04 13:54:14 | [727][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0080 mem: 3.36 + 04-04 13:54:17 | Time info >>>> elapsed: 990.92 mins remain: 370.23 mins + 04-04 13:54:18 | [728][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 13:54:23 | [728][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0577 ntime: 0084 mem: 3.36 + 04-04 13:54:28 | [728][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0634 ntime: 0085 mem: 3.36 + 04-04 13:54:35 | [728][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0482 ntime: 0084 mem: 3.36 + 04-04 13:54:43 | [728][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1195 ntime: 0087 mem: 3.36 + 04-04 13:54:46 | [728][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 13:54:52 | [728][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0127 ntime: 0078 mem: 3.36 + 04-04 13:54:58 | [728][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0081 mem: 3.36 + 04-04 13:55:04 | [728][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0083 mem: 3.36 + 04-04 13:55:09 | [728][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0892 ntime: 0086 mem: 3.36 + 04-04 13:55:13 | [728][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0072 ntime: 0086 mem: 3.36 + 04-04 13:55:17 | [728][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0086 ntime: 0082 mem: 3.36 + 04-04 13:55:24 | [728][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0103 ntime: 0082 mem: 3.36 + 04-04 13:55:32 | [728][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1500 ntime: 0083 mem: 3.36 + 04-04 13:55:36 | [728][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0128 ntime: 0084 mem: 3.36 + 04-04 13:55:43 | [728][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0376 ntime: 0086 mem: 3.36 + 04-04 13:55:48 | [728][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0072 mem: 3.36 + 04-04 13:55:53 | [728][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0074 ntime: 0078 mem: 3.36 + 04-04 13:55:58 | Time info >>>> elapsed: 992.60 mins remain: 368.99 mins + 04-04 13:55:59 | [729][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0914 ntime: 0065 mem: 3.36 + 04-04 13:56:05 | [729][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0737 ntime: 0084 mem: 3.36 + 04-04 13:56:09 | [729][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0556 ntime: 0077 mem: 3.36 + 04-04 13:56:15 | [729][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 13:56:21 | [729][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0545 ntime: 0079 mem: 3.36 + 04-04 13:56:28 | [729][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0111 ntime: 0082 mem: 3.36 + 04-04 13:56:35 | [729][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0781 ntime: 0080 mem: 3.36 + 04-04 13:56:42 | [729][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0075 ntime: 0078 mem: 3.36 + 04-04 13:56:49 | [729][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0122 ntime: 0075 mem: 3.36 + 04-04 13:56:54 | [729][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0077 mem: 3.36 + 04-04 13:57:00 | [729][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0088 ntime: 0077 mem: 3.36 + 04-04 13:57:05 | [729][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0094 ntime: 0082 mem: 3.36 + 04-04 13:57:10 | [729][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0834 ntime: 0082 mem: 3.36 + 04-04 13:57:16 | [729][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0096 ntime: 0085 mem: 3.36 + 04-04 13:57:21 | [729][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0121 ntime: 0080 mem: 3.36 + 04-04 13:57:27 | [729][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0149 ntime: 0083 mem: 3.36 + 04-04 13:57:32 | [729][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0753 ntime: 0076 mem: 3.36 + 04-04 13:57:37 | [729][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0078 ntime: 0081 mem: 3.36 + 04-04 13:57:43 | Time info >>>> elapsed: 994.34 mins remain: 367.77 mins + 04-04 13:57:44 | [730][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0569 ntime: 0076 mem: 3.36 + 04-04 13:57:51 | [730][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1361 ntime: 0076 mem: 3.36 + 04-04 13:57:55 | [730][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0332 ntime: 0080 mem: 3.36 + 04-04 13:57:59 | [730][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0626 ntime: 0087 mem: 3.36 + 04-04 13:58:07 | [730][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0083 mem: 3.36 + 04-04 13:58:14 | [730][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0113 ntime: 0081 mem: 3.36 + 04-04 13:58:21 | [730][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0085 mem: 3.36 + 04-04 13:58:27 | [730][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0394 ntime: 0085 mem: 3.36 + 04-04 13:58:34 | [730][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0071 mem: 3.36 + 04-04 13:58:40 | [730][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1055 ntime: 0079 mem: 3.36 + 04-04 13:58:49 | [730][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0083 ntime: 0081 mem: 3.36 + 04-04 13:58:54 | [730][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0610 ntime: 0078 mem: 3.36 + 04-04 13:59:01 | [730][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0077 mem: 3.36 + 04-04 13:59:09 | [730][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1534 ntime: 0076 mem: 3.36 + 04-04 13:59:15 | [730][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0091 ntime: 0078 mem: 3.36 + 04-04 13:59:22 | [730][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0705 ntime: 0078 mem: 3.36 + 04-04 13:59:28 | [730][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0547 ntime: 0079 mem: 3.36 + 04-04 13:59:33 | [730][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0358 ntime: 0074 mem: 3.36 + 04-04 13:59:37 | Time info >>>> elapsed: 996.24 mins remain: 366.61 mins + 04-04 13:59:37 | [731][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0156 ntime: 0081 mem: 3.36 + 04-04 13:59:42 | [731][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0085 mem: 3.36 + 04-04 13:59:46 | [731][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0063 ntime: 0077 mem: 3.36 + 04-04 13:59:53 | [731][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0957 ntime: 0083 mem: 3.36 + 04-04 13:59:59 | [731][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0821 ntime: 0082 mem: 3.36 + 04-04 14:00:05 | [731][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0077 ntime: 0082 mem: 3.36 + 04-04 14:00:10 | [731][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0083 ntime: 0078 mem: 3.36 + 04-04 14:00:14 | [731][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0081 mem: 3.36 + 04-04 14:00:19 | [731][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0129 ntime: 0076 mem: 3.36 + 04-04 14:00:25 | [731][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1099 ntime: 0087 mem: 3.36 + 04-04 14:00:30 | [731][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0081 mem: 3.36 + 04-04 14:00:36 | [731][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0131 ntime: 0074 mem: 3.36 + 04-04 14:00:41 | [731][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0486 ntime: 0082 mem: 3.36 + 04-04 14:00:45 | [731][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0124 ntime: 0079 mem: 3.36 + 04-04 14:00:51 | [731][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0933 ntime: 0086 mem: 3.36 + 04-04 14:00:56 | [731][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0956 ntime: 0078 mem: 3.36 + 04-04 14:01:03 | [731][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0085 mem: 3.36 + 04-04 14:01:08 | [731][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0892 ntime: 0075 mem: 3.36 + 04-04 14:01:14 | Time info >>>> elapsed: 997.87 mins remain: 365.34 mins + 04-04 14:01:16 | [732][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1242 ntime: 0080 mem: 3.36 + 04-04 14:01:22 | [732][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0568 ntime: 0082 mem: 3.36 + 04-04 14:01:28 | [732][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0533 ntime: 0078 mem: 3.36 + 04-04 14:01:32 | [732][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0102 ntime: 0073 mem: 3.36 + 04-04 14:01:39 | [732][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1073 ntime: 0078 mem: 3.36 + 04-04 14:01:44 | [732][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0082 mem: 3.36 + 04-04 14:01:50 | [732][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0457 ntime: 0058 mem: 3.36 + 04-04 14:01:57 | [732][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0071 ntime: 0077 mem: 3.36 + 04-04 14:02:09 | [732][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0585 ntime: 0080 mem: 3.36 + 04-04 14:02:17 | [732][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1156 ntime: 0078 mem: 3.36 + 04-04 14:02:23 | [732][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0570 ntime: 0079 mem: 3.36 + 04-04 14:02:28 | [732][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0098 ntime: 0080 mem: 3.36 + 04-04 14:02:34 | [732][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0082 ntime: 0083 mem: 3.36 + 04-04 14:02:41 | [732][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0944 ntime: 0078 mem: 3.36 + 04-04 14:02:48 | [732][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1114 ntime: 0078 mem: 3.36 + 04-04 14:02:56 | [732][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1507 ntime: 0074 mem: 3.36 + 04-04 14:03:04 | [732][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0187 ntime: 0085 mem: 3.36 + 04-04 14:03:10 | [732][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0652 ntime: 0083 mem: 3.36 + 04-04 14:03:17 | Time info >>>> elapsed: 999.91 mins remain: 364.22 mins + 04-04 14:03:17 | [733][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0075 ntime: 0077 mem: 3.36 + 04-04 14:03:25 | [733][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0757 ntime: 0086 mem: 3.36 + 04-04 14:03:33 | [733][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0132 ntime: 0076 mem: 3.36 + 04-04 14:03:40 | [733][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0080 mem: 3.36 + 04-04 14:03:48 | [733][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0637 ntime: 0083 mem: 3.36 + 04-04 14:03:55 | [733][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0995 ntime: 0083 mem: 3.36 + 04-04 14:04:00 | [733][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0176 ntime: 0076 mem: 3.36 + 04-04 14:04:04 | [733][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0740 ntime: 0076 mem: 3.36 + 04-04 14:04:11 | [733][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0082 ntime: 0084 mem: 3.36 + 04-04 14:04:17 | [733][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0115 ntime: 0086 mem: 3.36 + 04-04 14:04:24 | [733][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0459 ntime: 0089 mem: 3.36 + 04-04 14:04:30 | [733][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1029 ntime: 0075 mem: 3.36 + 04-04 14:04:38 | [733][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1499 ntime: 0086 mem: 3.36 + 04-04 14:04:44 | [733][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0533 ntime: 0078 mem: 3.36 + 04-04 14:04:51 | [733][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0080 mem: 3.36 + 04-04 14:04:57 | [733][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0076 mem: 3.36 + 04-04 14:05:05 | [733][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0084 ntime: 0079 mem: 3.36 + 04-04 14:05:11 | [733][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0616 ntime: 0078 mem: 3.36 + 04-04 14:05:16 | Time info >>>> elapsed: 1001.89 mins remain: 363.08 mins + 04-04 14:05:16 | [734][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0162 ntime: 0073 mem: 3.36 + 04-04 14:05:23 | [734][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0797 ntime: 0080 mem: 3.36 + 04-04 14:05:27 | [734][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0108 ntime: 0084 mem: 3.36 + 04-04 14:05:34 | [734][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0710 ntime: 0079 mem: 3.36 + 04-04 14:05:40 | [734][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0128 ntime: 0083 mem: 3.36 + 04-04 14:05:48 | [734][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1595 ntime: 0082 mem: 3.36 + 04-04 14:05:54 | [734][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0780 ntime: 0078 mem: 3.36 + 04-04 14:06:02 | [734][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 14:06:10 | [734][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0079 ntime: 0078 mem: 3.36 + 04-04 14:06:20 | [734][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0164 ntime: 0077 mem: 3.36 + 04-04 14:06:27 | [734][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1293 ntime: 0077 mem: 3.36 + 04-04 14:06:32 | [734][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1392 ntime: 0079 mem: 3.36 + 04-04 14:06:38 | [734][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0946 ntime: 0083 mem: 3.36 + 04-04 14:06:44 | [734][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0695 ntime: 0078 mem: 3.36 + 04-04 14:06:52 | [734][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0089 ntime: 0077 mem: 3.36 + 04-04 14:06:58 | [734][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0090 ntime: 0082 mem: 3.36 + 04-04 14:07:05 | [734][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1002 ntime: 0079 mem: 3.36 + 04-04 14:07:12 | [734][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1239 ntime: 0083 mem: 3.36 + 04-04 14:07:16 | Time info >>>> elapsed: 1003.90 mins remain: 361.95 mins + 04-04 14:07:17 | [735][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0118 ntime: 0080 mem: 3.36 + 04-04 14:07:22 | [735][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0106 ntime: 0089 mem: 3.36 + 04-04 14:07:27 | [735][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0132 ntime: 0077 mem: 3.36 + 04-04 14:07:34 | [735][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0086 mem: 3.36 + 04-04 14:07:39 | [735][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1112 ntime: 0076 mem: 3.36 + 04-04 14:07:45 | [735][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0904 ntime: 0080 mem: 3.36 + 04-04 14:07:52 | [735][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0714 ntime: 0074 mem: 3.36 + 04-04 14:07:59 | [735][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0136 ntime: 0081 mem: 3.36 + 04-04 14:08:06 | [735][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0046 ntime: 0078 mem: 3.36 + 04-04 14:08:12 | [735][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0976 ntime: 0081 mem: 3.36 + 04-04 14:08:18 | [735][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1287 ntime: 0090 mem: 3.36 + 04-04 14:08:24 | [735][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0136 ntime: 0075 mem: 3.36 + 04-04 14:08:31 | [735][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1416 ntime: 0080 mem: 3.36 + 04-04 14:08:38 | [735][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0739 ntime: 0080 mem: 3.36 + 04-04 14:08:44 | [735][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0077 ntime: 0082 mem: 3.36 + 04-04 14:08:51 | [735][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0421 ntime: 0085 mem: 3.36 + 04-04 14:08:56 | [735][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0124 ntime: 0080 mem: 3.36 + 04-04 14:09:02 | [735][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0079 mem: 3.36 + 04-04 14:09:09 | Time info >>>> elapsed: 1005.77 mins remain: 360.77 mins + 04-04 14:09:09 | [736][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0082 mem: 3.36 + 04-04 14:09:16 | [736][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1311 ntime: 0080 mem: 3.36 + 04-04 14:09:21 | [736][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0078 mem: 3.36 + 04-04 14:09:29 | [736][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0083 mem: 3.36 + 04-04 14:09:38 | [736][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1414 ntime: 0086 mem: 3.36 + 04-04 14:09:44 | [736][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0089 mem: 3.36 + 04-04 14:09:50 | [736][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0078 mem: 3.36 + 04-04 14:09:56 | [736][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0085 mem: 3.36 + 04-04 14:10:02 | [736][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0955 ntime: 0077 mem: 3.36 + 04-04 14:10:08 | [736][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0177 ntime: 0083 mem: 3.36 + 04-04 14:10:15 | [736][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0972 ntime: 0079 mem: 3.36 + 04-04 14:10:22 | [736][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1038 ntime: 0087 mem: 3.36 + 04-04 14:10:29 | [736][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0999 ntime: 0080 mem: 3.36 + 04-04 14:10:36 | [736][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0682 ntime: 0072 mem: 3.36 + 04-04 14:10:43 | [736][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0065 ntime: 0082 mem: 3.36 + 04-04 14:10:49 | [736][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0998 ntime: 0075 mem: 3.36 + 04-04 14:10:56 | [736][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0690 ntime: 0070 mem: 3.36 + 04-04 14:11:02 | [736][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0110 ntime: 0081 mem: 3.36 + 04-04 14:11:07 | Time info >>>> elapsed: 1007.74 mins remain: 359.61 mins + 04-04 14:11:07 | [737][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0621 ntime: 0080 mem: 3.36 + 04-04 14:11:14 | [737][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0131 ntime: 0082 mem: 3.36 + 04-04 14:11:21 | [737][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0703 ntime: 0081 mem: 3.36 + 04-04 14:11:26 | [737][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0264 ntime: 0077 mem: 3.36 + 04-04 14:11:35 | [737][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0690 ntime: 0078 mem: 3.36 + 04-04 14:11:41 | [737][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0073 ntime: 0076 mem: 3.36 + 04-04 14:11:49 | [737][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1562 ntime: 0081 mem: 3.36 + 04-04 14:11:55 | [737][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0743 ntime: 0078 mem: 3.36 + 04-04 14:12:03 | [737][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0562 ntime: 0074 mem: 3.36 + 04-04 14:12:11 | [737][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1123 ntime: 0083 mem: 3.36 + 04-04 14:12:15 | [737][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0587 ntime: 0078 mem: 3.36 + 04-04 14:12:21 | [737][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0092 ntime: 0077 mem: 3.36 + 04-04 14:12:28 | [737][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1307 ntime: 0078 mem: 3.36 + 04-04 14:12:36 | [737][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0346 ntime: 0077 mem: 3.36 + 04-04 14:12:41 | [737][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0155 ntime: 0077 mem: 3.36 + 04-04 14:12:50 | [737][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1109 ntime: 0086 mem: 3.36 + 04-04 14:12:56 | [737][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0683 ntime: 0075 mem: 3.36 + 04-04 14:13:01 | [737][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0636 ntime: 0085 mem: 3.36 + 04-04 14:13:07 | Time info >>>> elapsed: 1009.74 mins remain: 358.47 mins + 04-04 14:13:08 | [738][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0804 ntime: 0080 mem: 3.36 + 04-04 14:13:13 | [738][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0829 ntime: 0078 mem: 3.36 + 04-04 14:13:21 | [738][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0766 ntime: 0077 mem: 3.36 + 04-04 14:13:28 | [738][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1111 ntime: 0083 mem: 3.36 + 04-04 14:13:36 | [738][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0872 ntime: 0055 mem: 3.36 + 04-04 14:13:44 | [738][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1532 ntime: 0075 mem: 3.36 + 04-04 14:13:53 | [738][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1099 ntime: 0071 mem: 3.36 + 04-04 14:13:59 | [738][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0139 ntime: 0085 mem: 3.36 + 04-04 14:14:06 | [738][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 14:14:15 | [738][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1445 ntime: 0078 mem: 3.36 + 04-04 14:14:23 | [738][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0998 ntime: 0077 mem: 3.36 + 04-04 14:14:29 | [738][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0081 mem: 3.36 + 04-04 14:14:35 | [738][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0082 mem: 3.36 + 04-04 14:14:41 | [738][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 14:14:46 | [738][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0063 ntime: 0083 mem: 3.36 + 04-04 14:14:53 | [738][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0078 ntime: 0075 mem: 3.36 + 04-04 14:14:59 | [738][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0106 ntime: 0080 mem: 3.36 + 04-04 14:15:05 | [738][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0789 ntime: 0085 mem: 3.36 + 04-04 14:15:10 | Time info >>>> elapsed: 1011.79 mins remain: 357.34 mins + 04-04 14:15:11 | [739][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0945 ntime: 0082 mem: 3.36 + 04-04 14:15:18 | [739][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0168 ntime: 0081 mem: 3.36 + 04-04 14:15:25 | [739][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0120 ntime: 0078 mem: 3.36 + 04-04 14:15:32 | [739][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0727 ntime: 0076 mem: 3.36 + 04-04 14:15:37 | [739][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0143 ntime: 0082 mem: 3.36 + 04-04 14:15:44 | [739][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1083 ntime: 0081 mem: 3.36 + 04-04 14:15:49 | [739][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0716 ntime: 0083 mem: 3.36 + 04-04 14:15:56 | [739][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1063 ntime: 0082 mem: 3.36 + 04-04 14:16:03 | [739][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0652 ntime: 0076 mem: 3.36 + 04-04 14:16:08 | [739][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0251 ntime: 0074 mem: 3.36 + 04-04 14:16:16 | [739][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0710 ntime: 0077 mem: 3.36 + 04-04 14:16:24 | [739][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1000 ntime: 0080 mem: 3.36 + 04-04 14:16:31 | [739][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 14:16:40 | [739][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0404 ntime: 0073 mem: 3.36 + 04-04 14:16:49 | [739][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0070 ntime: 0082 mem: 3.36 + 04-04 14:16:57 | [739][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0162 ntime: 0078 mem: 3.36 + 04-04 14:17:02 | [739][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 14:17:10 | [739][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0076 ntime: 0087 mem: 3.36 + 04-04 14:17:17 | Time info >>>> elapsed: 1013.91 mins remain: 356.24 mins + 04-04 14:17:17 | [740][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0083 ntime: 0084 mem: 3.36 + 04-04 14:17:24 | [740][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0947 ntime: 0081 mem: 3.36 + 04-04 14:17:33 | [740][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0625 ntime: 0077 mem: 3.36 + 04-04 14:17:41 | [740][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1200 ntime: 0084 mem: 3.36 + 04-04 14:17:47 | [740][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0123 ntime: 0077 mem: 3.36 + 04-04 14:17:56 | [740][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0667 ntime: 0080 mem: 3.36 + 04-04 14:18:02 | [740][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0732 ntime: 0085 mem: 3.36 + 04-04 14:18:09 | [740][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0095 ntime: 0080 mem: 3.36 + 04-04 14:18:16 | [740][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0949 ntime: 0078 mem: 3.36 + 04-04 14:18:20 | [740][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0140 ntime: 0079 mem: 3.36 + 04-04 14:18:27 | [740][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0219 ntime: 0084 mem: 3.36 + 04-04 14:18:33 | [740][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0838 ntime: 0076 mem: 3.36 + 04-04 14:18:38 | [740][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0599 ntime: 0081 mem: 3.36 + 04-04 14:18:44 | [740][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1267 ntime: 0088 mem: 3.36 + 04-04 14:18:51 | [740][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0596 ntime: 0083 mem: 3.36 + 04-04 14:18:58 | [740][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0983 ntime: 0077 mem: 3.36 + 04-04 14:19:06 | [740][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1568 ntime: 0079 mem: 3.36 + 04-04 14:19:10 | [740][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1513 ntime: 0080 mem: 3.36 + 04-04 14:19:17 | Time info >>>> elapsed: 1015.91 mins remain: 355.09 mins + 04-04 14:19:17 | [741][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 14:19:24 | [741][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1518 ntime: 0079 mem: 3.36 + 04-04 14:19:31 | [741][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0076 mem: 3.36 + 04-04 14:19:37 | [741][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0075 mem: 3.36 + 04-04 14:19:45 | [741][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1396 ntime: 0076 mem: 3.36 + 04-04 14:19:54 | [741][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1559 ntime: 0076 mem: 3.36 + 04-04 14:19:59 | [741][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0133 ntime: 0085 mem: 3.36 + 04-04 14:20:07 | [741][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0086 ntime: 0078 mem: 3.36 + 04-04 14:20:13 | [741][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 14:20:20 | [741][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0165 ntime: 0078 mem: 3.36 + 04-04 14:20:26 | [741][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0868 ntime: 0079 mem: 3.36 + 04-04 14:20:33 | [741][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1443 ntime: 0079 mem: 3.36 + 04-04 14:20:39 | [741][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0157 ntime: 0073 mem: 3.36 + 04-04 14:20:47 | [741][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0547 ntime: 0084 mem: 3.36 + 04-04 14:20:55 | [741][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1102 ntime: 0077 mem: 3.36 + 04-04 14:21:01 | [741][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0319 ntime: 0082 mem: 3.36 + 04-04 14:21:12 | [741][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1505 ntime: 0077 mem: 3.36 + 04-04 14:21:26 | [741][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0647 ntime: 0080 mem: 3.36 + 04-04 14:21:29 | Time info >>>> elapsed: 1018.11 mins remain: 354.01 mins + 04-04 14:21:30 | [742][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0661 ntime: 0072 mem: 3.36 + 04-04 14:21:38 | [742][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1193 ntime: 0082 mem: 3.36 + 04-04 14:21:45 | [742][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1568 ntime: 0079 mem: 3.36 + 04-04 14:21:53 | [742][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1252 ntime: 0086 mem: 3.36 + 04-04 14:22:01 | [742][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0989 ntime: 0080 mem: 3.36 + 04-04 14:22:07 | [742][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0523 ntime: 0087 mem: 3.36 + 04-04 14:22:14 | [742][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0795 ntime: 0084 mem: 3.36 + 04-04 14:22:23 | [742][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0102 ntime: 0078 mem: 3.36 + 04-04 14:22:34 | [742][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0473 ntime: 0074 mem: 3.36 + 04-04 14:22:43 | [742][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0075 mem: 3.36 + 04-04 14:22:54 | [742][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1486 ntime: 0074 mem: 3.36 + 04-04 14:23:06 | [742][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0919 ntime: 0077 mem: 3.36 + 04-04 14:23:16 | [742][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1720 ntime: 0081 mem: 3.36 + 04-04 14:23:22 | [742][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0641 ntime: 0079 mem: 3.36 + 04-04 14:23:28 | [742][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0112 ntime: 0079 mem: 3.36 + 04-04 14:23:39 | [742][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0182 ntime: 0081 mem: 3.36 + 04-04 14:23:46 | [742][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1114 ntime: 0085 mem: 3.36 + 04-04 14:23:52 | [742][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0683 ntime: 0080 mem: 3.36 + 04-04 14:23:58 | Time info >>>> elapsed: 1020.59 mins remain: 353.02 mins + 04-04 14:23:58 | [743][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 14:24:06 | [743][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0989 ntime: 0079 mem: 3.36 + 04-04 14:24:13 | [743][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1026 ntime: 0074 mem: 3.36 + 04-04 14:24:20 | [743][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1201 ntime: 0078 mem: 3.36 + 04-04 14:24:27 | [743][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0183 ntime: 0083 mem: 3.36 + 04-04 14:24:34 | [743][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0091 ntime: 0090 mem: 3.36 + 04-04 14:24:40 | [743][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0046 ntime: 0078 mem: 3.36 + 04-04 14:24:50 | [743][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1647 ntime: 0075 mem: 3.36 + 04-04 14:24:54 | [743][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0638 ntime: 0081 mem: 3.36 + 04-04 14:25:01 | [743][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0169 ntime: 0078 mem: 3.36 + 04-04 14:25:09 | [743][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1363 ntime: 0081 mem: 3.36 + 04-04 14:25:18 | [743][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1017 ntime: 0073 mem: 3.36 + 04-04 14:25:25 | [743][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0872 ntime: 0085 mem: 3.36 + 04-04 14:25:33 | [743][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0149 ntime: 0075 mem: 3.36 + 04-04 14:25:44 | [743][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1092 ntime: 0081 mem: 3.36 + 04-04 14:25:53 | [743][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0116 ntime: 0078 mem: 3.36 + 04-04 14:26:03 | [743][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1251 ntime: 0076 mem: 3.36 + 04-04 14:26:10 | [743][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0083 mem: 3.36 + 04-04 14:26:16 | Time info >>>> elapsed: 1022.89 mins remain: 351.96 mins + 04-04 14:26:17 | [744][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1093 ntime: 0083 mem: 3.36 + 04-04 14:26:26 | [744][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0937 ntime: 0082 mem: 3.36 + 04-04 14:26:35 | [744][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1236 ntime: 0078 mem: 3.36 + 04-04 14:26:42 | [744][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0086 ntime: 0085 mem: 3.36 + 04-04 14:26:49 | [744][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0545 ntime: 0088 mem: 3.36 + 04-04 14:26:57 | [744][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0484 ntime: 0077 mem: 3.36 + 04-04 14:27:04 | [744][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0164 ntime: 0078 mem: 3.36 + 04-04 14:27:11 | [744][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0085 ntime: 0077 mem: 3.36 + 04-04 14:27:17 | [744][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0568 ntime: 0087 mem: 3.36 + 04-04 14:27:25 | [744][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0151 ntime: 0079 mem: 3.36 + 04-04 14:27:33 | [744][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0491 ntime: 0081 mem: 3.36 + 04-04 14:27:41 | [744][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0709 ntime: 0075 mem: 3.36 + 04-04 14:27:52 | [744][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1576 ntime: 0081 mem: 3.36 + 04-04 14:28:00 | [744][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1497 ntime: 0085 mem: 3.36 + 04-04 14:28:10 | [744][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0157 ntime: 0077 mem: 3.36 + 04-04 14:28:17 | [744][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0101 ntime: 0083 mem: 3.36 + 04-04 14:28:26 | [744][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1745 ntime: 0079 mem: 3.36 + 04-04 14:28:34 | [744][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1085 ntime: 0082 mem: 3.36 + 04-04 14:28:39 | Time info >>>> elapsed: 1025.28 mins remain: 350.93 mins + 04-04 14:28:40 | [745][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1234 ntime: 0075 mem: 3.36 + 04-04 14:28:48 | [745][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1290 ntime: 0077 mem: 3.36 + 04-04 14:28:54 | [745][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0088 ntime: 0078 mem: 3.36 + 04-04 14:29:00 | [745][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0200 ntime: 0084 mem: 3.36 + 04-04 14:29:09 | [745][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0955 ntime: 0077 mem: 3.36 + 04-04 14:29:15 | [745][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0116 ntime: 0078 mem: 3.36 + 04-04 14:29:20 | [745][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0133 ntime: 0077 mem: 3.36 + 04-04 14:29:26 | [745][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0105 ntime: 0088 mem: 3.36 + 04-04 14:29:34 | [745][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0362 ntime: 0083 mem: 3.36 + 04-04 14:29:41 | [745][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0068 ntime: 0087 mem: 3.36 + 04-04 14:29:50 | [745][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1084 ntime: 0087 mem: 3.36 + 04-04 14:29:56 | [745][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0081 mem: 3.36 + 04-04 14:30:03 | [745][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0660 ntime: 0075 mem: 3.36 + 04-04 14:30:11 | [745][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1251 ntime: 0081 mem: 3.36 + 04-04 14:30:17 | [745][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0074 mem: 3.36 + 04-04 14:30:23 | [745][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1593 ntime: 0087 mem: 3.36 + 04-04 14:30:32 | [745][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1139 ntime: 0084 mem: 3.36 + 04-04 14:30:40 | [745][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1596 ntime: 0083 mem: 3.36 + 04-04 14:30:47 | Time info >>>> elapsed: 1027.41 mins remain: 349.81 mins + 04-04 14:30:48 | [746][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 1251 ntime: 0084 mem: 3.36 + 04-04 14:30:55 | [746][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0069 ntime: 0082 mem: 3.36 + 04-04 14:31:02 | [746][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0451 ntime: 0080 mem: 3.36 + 04-04 14:31:12 | [746][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0732 ntime: 0076 mem: 3.36 + 04-04 14:31:19 | [746][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1619 ntime: 0085 mem: 3.36 + 04-04 14:31:28 | [746][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0945 ntime: 0083 mem: 3.36 + 04-04 14:31:37 | [746][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1581 ntime: 0078 mem: 3.36 + 04-04 14:31:45 | [746][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0879 ntime: 0085 mem: 3.36 + 04-04 14:31:54 | [746][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0129 ntime: 0078 mem: 3.36 + 04-04 14:32:05 | [746][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1996 ntime: 0086 mem: 3.36 + 04-04 14:32:12 | [746][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0097 ntime: 0082 mem: 3.36 + 04-04 14:32:22 | [746][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0094 ntime: 0079 mem: 3.36 + 04-04 14:32:28 | [746][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0076 mem: 3.36 + 04-04 14:32:39 | [746][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1004 ntime: 0085 mem: 3.36 + 04-04 14:32:46 | [746][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0081 mem: 3.36 + 04-04 14:32:55 | [746][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0069 ntime: 0079 mem: 3.36 + 04-04 14:33:05 | [746][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1210 ntime: 0084 mem: 3.36 + 04-04 14:33:12 | [746][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1008 ntime: 0081 mem: 3.36 + 04-04 14:33:18 | Time info >>>> elapsed: 1029.93 mins remain: 348.83 mins + 04-04 14:33:20 | [747][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 1453 ntime: 0081 mem: 3.36 + 04-04 14:33:28 | [747][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0111 ntime: 0078 mem: 3.36 + 04-04 14:33:36 | [747][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1037 ntime: 0084 mem: 3.36 + 04-04 14:33:43 | [747][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0632 ntime: 0076 mem: 3.36 + 04-04 14:33:51 | [747][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1568 ntime: 0080 mem: 3.36 + 04-04 14:33:59 | [747][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0596 ntime: 0078 mem: 3.36 + 04-04 14:34:11 | [747][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0828 ntime: 0080 mem: 3.36 + 04-04 14:34:17 | [747][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0225 ntime: 0083 mem: 3.36 + 04-04 14:34:24 | [747][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0658 ntime: 0084 mem: 3.36 + 04-04 14:34:30 | [747][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1418 ntime: 0079 mem: 3.36 + 04-04 14:34:37 | [747][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1165 ntime: 0081 mem: 3.36 + 04-04 14:34:42 | [747][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0201 ntime: 0079 mem: 3.36 + 04-04 14:34:48 | [747][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0086 ntime: 0078 mem: 3.36 + 04-04 14:34:55 | [747][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0447 ntime: 0083 mem: 3.36 + 04-04 14:35:01 | [747][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0817 ntime: 0081 mem: 3.36 + 04-04 14:35:05 | [747][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0063 ntime: 0080 mem: 3.36 + 04-04 14:35:11 | [747][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0824 ntime: 0078 mem: 3.36 + 04-04 14:35:16 | [747][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0749 ntime: 0083 mem: 3.36 + 04-04 14:35:21 | Time info >>>> elapsed: 1031.97 mins remain: 347.67 mins + 04-04 14:35:21 | [748][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0079 ntime: 0072 mem: 3.36 + 04-04 14:35:27 | [748][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0728 ntime: 0080 mem: 3.36 + 04-04 14:35:34 | [748][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0329 ntime: 0086 mem: 3.36 + 04-04 14:35:39 | [748][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0085 ntime: 0081 mem: 3.36 + 04-04 14:35:47 | [748][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0115 ntime: 0085 mem: 3.36 + 04-04 14:35:53 | [748][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0949 ntime: 0083 mem: 3.36 + 04-04 14:35:59 | [748][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0839 ntime: 0075 mem: 3.36 + 04-04 14:36:05 | [748][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0574 ntime: 0086 mem: 3.36 + 04-04 14:36:11 | [748][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1285 ntime: 0083 mem: 3.36 + 04-04 14:36:19 | [748][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1639 ntime: 0078 mem: 3.36 + 04-04 14:36:24 | [748][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0099 ntime: 0079 mem: 3.36 + 04-04 14:36:29 | [748][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0063 ntime: 0082 mem: 3.36 + 04-04 14:36:35 | [748][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0085 mem: 3.36 + 04-04 14:36:40 | [748][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0075 ntime: 0079 mem: 3.36 + 04-04 14:36:48 | [748][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1095 ntime: 0088 mem: 3.36 + 04-04 14:36:57 | [748][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1040 ntime: 0081 mem: 3.36 + 04-04 14:37:03 | [748][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0908 ntime: 0085 mem: 3.36 + 04-04 14:37:11 | [748][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0080 ntime: 0078 mem: 3.36 + 04-04 14:37:15 | Time info >>>> elapsed: 1033.88 mins remain: 346.47 mins + 04-04 14:37:16 | [749][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1097 ntime: 0054 mem: 3.36 + 04-04 14:37:23 | [749][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0085 mem: 3.36 + 04-04 14:37:30 | [749][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1404 ntime: 0078 mem: 3.36 + 04-04 14:37:34 | [749][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0107 ntime: 0079 mem: 3.36 + 04-04 14:37:40 | [749][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1035 ntime: 0084 mem: 3.36 + 04-04 14:37:47 | [749][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0945 ntime: 0081 mem: 3.36 + 04-04 14:37:54 | [749][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0140 ntime: 0083 mem: 3.36 + 04-04 14:38:03 | [749][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0703 ntime: 0076 mem: 3.36 + 04-04 14:38:08 | [749][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0937 ntime: 0081 mem: 3.36 + 04-04 14:38:13 | [749][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0076 mem: 3.36 + 04-04 14:38:19 | [749][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0168 ntime: 0075 mem: 3.36 + 04-04 14:38:27 | [749][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1241 ntime: 0078 mem: 3.36 + 04-04 14:38:32 | [749][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 14:38:39 | [749][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0797 ntime: 0080 mem: 3.36 + 04-04 14:38:46 | [749][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0084 ntime: 0079 mem: 3.36 + 04-04 14:38:55 | [749][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0816 ntime: 0080 mem: 3.36 + 04-04 14:39:04 | [749][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1107 ntime: 0088 mem: 3.36 + 04-04 14:39:09 | [749][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0152 ntime: 0080 mem: 3.36 + 04-04 14:39:14 | Time info >>>> elapsed: 1035.86 mins remain: 345.29 mins + 04-04 14:39:15 | [750][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 1141 ntime: 0075 mem: 3.36 + 04-04 14:39:23 | [750][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1083 ntime: 0060 mem: 3.36 + 04-04 14:39:30 | [750][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1064 ntime: 0082 mem: 3.36 + 04-04 14:39:36 | [750][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0161 ntime: 0075 mem: 3.36 + 04-04 14:39:43 | [750][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0069 ntime: 0080 mem: 3.36 + 04-04 14:39:48 | [750][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0325 ntime: 0080 mem: 3.36 + 04-04 14:39:56 | [750][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0962 ntime: 0084 mem: 3.36 + 04-04 14:40:02 | [750][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0123 ntime: 0085 mem: 3.36 + 04-04 14:40:08 | [750][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0055 mem: 3.36 + 04-04 14:40:14 | [750][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0529 ntime: 0079 mem: 3.36 + 04-04 14:40:20 | [750][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0065 mem: 3.36 + 04-04 14:40:28 | [750][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1299 ntime: 0078 mem: 3.36 + 04-04 14:40:34 | [750][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0411 ntime: 0087 mem: 3.36 + 04-04 14:40:40 | [750][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1429 ntime: 0083 mem: 3.36 + 04-04 14:40:48 | [750][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1203 ntime: 0073 mem: 3.36 + 04-04 14:40:53 | [750][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0572 ntime: 0085 mem: 3.36 + 04-04 14:40:57 | [750][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0178 ntime: 0079 mem: 3.36 + 04-04 14:41:03 | [750][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1159 ntime: 0078 mem: 3.36 + 04-04 14:41:08 | Time info >>>> elapsed: 1037.76 mins remain: 344.08 mins + 04-04 14:41:08 | [751][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0067 ntime: 0078 mem: 3.36 + 04-04 14:41:13 | [751][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1054 ntime: 0082 mem: 3.36 + 04-04 14:41:19 | [751][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0411 ntime: 0076 mem: 3.36 + 04-04 14:41:24 | [751][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1234 ntime: 0079 mem: 3.36 + 04-04 14:41:32 | [751][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0991 ntime: 0080 mem: 3.36 + 04-04 14:41:39 | [751][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0562 ntime: 0089 mem: 3.36 + 04-04 14:41:45 | [751][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0649 ntime: 0079 mem: 3.36 + 04-04 14:41:51 | [751][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0584 ntime: 0078 mem: 3.36 + 04-04 14:41:58 | [751][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0080 ntime: 0083 mem: 3.36 + 04-04 14:42:05 | [751][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1138 ntime: 0079 mem: 3.36 + 04-04 14:42:11 | [751][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0086 ntime: 0084 mem: 3.36 + 04-04 14:42:18 | [751][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0127 ntime: 0083 mem: 3.36 + 04-04 14:42:28 | [751][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1554 ntime: 0078 mem: 3.36 + 04-04 14:42:36 | [751][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0868 ntime: 0081 mem: 3.36 + 04-04 14:42:42 | [751][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0793 ntime: 0083 mem: 3.36 + 04-04 14:42:49 | [751][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1592 ntime: 0072 mem: 3.36 + 04-04 14:42:56 | [751][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0419 ntime: 0080 mem: 3.36 + 04-04 14:43:04 | [751][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1285 ntime: 0082 mem: 3.36 + 04-04 14:43:10 | Time info >>>> elapsed: 1039.79 mins remain: 342.91 mins + 04-04 14:43:11 | [752][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1361 ntime: 0081 mem: 3.36 + 04-04 14:43:23 | [752][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0135 ntime: 0083 mem: 3.36 + 04-04 14:43:35 | [752][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1464 ntime: 0085 mem: 3.36 + 04-04 14:43:44 | [752][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0115 ntime: 0087 mem: 3.36 + 04-04 14:43:53 | [752][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1566 ntime: 0080 mem: 3.36 + 04-04 14:44:01 | [752][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0983 ntime: 0082 mem: 3.36 + 04-04 14:44:15 | [752][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 2136 ntime: 0085 mem: 3.36 + 04-04 14:44:23 | [752][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0829 ntime: 0078 mem: 3.36 + 04-04 14:44:30 | [752][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0313 ntime: 0079 mem: 3.36 + 04-04 14:44:40 | [752][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1482 ntime: 0079 mem: 3.36 + 04-04 14:44:46 | [752][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1044 ntime: 0081 mem: 3.36 + 04-04 14:44:54 | [752][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0707 ntime: 0089 mem: 3.36 + 04-04 14:45:01 | [752][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0992 ntime: 0077 mem: 3.36 + 04-04 14:45:11 | [752][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0107 ntime: 0077 mem: 3.36 + 04-04 14:45:19 | [752][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1338 ntime: 0071 mem: 3.36 + 04-04 14:45:26 | [752][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0076 ntime: 0074 mem: 3.36 + 04-04 14:45:33 | [752][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1374 ntime: 0081 mem: 3.36 + 04-04 14:45:39 | [752][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0566 ntime: 0085 mem: 3.36 + 04-04 14:45:45 | Time info >>>> elapsed: 1042.38 mins remain: 341.92 mins + 04-04 14:45:45 | [753][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0052 ntime: 0085 mem: 3.36 + 04-04 14:45:53 | [753][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1279 ntime: 0084 mem: 3.36 + 04-04 14:46:01 | [753][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0941 ntime: 0088 mem: 3.36 + 04-04 14:46:09 | [753][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0939 ntime: 0074 mem: 3.36 + 04-04 14:46:16 | [753][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1223 ntime: 0077 mem: 3.36 + 04-04 14:46:23 | [753][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0775 ntime: 0080 mem: 3.36 + 04-04 14:46:31 | [753][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0211 ntime: 0074 mem: 3.36 + 04-04 14:46:38 | [753][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0671 ntime: 0075 mem: 3.36 + 04-04 14:46:46 | [753][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0087 ntime: 0077 mem: 3.36 + 04-04 14:46:53 | [753][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0067 ntime: 0087 mem: 3.36 + 04-04 14:47:00 | [753][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1274 ntime: 0086 mem: 3.36 + 04-04 14:47:07 | [753][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0289 ntime: 0079 mem: 3.36 + 04-04 14:47:13 | [753][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0077 mem: 3.36 + 04-04 14:47:21 | [753][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0491 ntime: 0090 mem: 3.36 + 04-04 14:47:27 | [753][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0080 mem: 3.36 + 04-04 14:47:37 | [753][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1543 ntime: 0076 mem: 3.36 + 04-04 14:47:44 | [753][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0430 ntime: 0073 mem: 3.36 + 04-04 14:47:50 | [753][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0460 ntime: 0081 mem: 3.36 + 04-04 14:47:56 | Time info >>>> elapsed: 1044.57 mins remain: 340.80 mins + 04-04 14:47:56 | [754][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0098 ntime: 0083 mem: 3.36 + 04-04 14:48:04 | [754][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0781 ntime: 0108 mem: 3.36 + 04-04 14:48:10 | [754][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0749 ntime: 0079 mem: 3.36 + 04-04 14:48:18 | [754][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0394 ntime: 0076 mem: 3.36 + 04-04 14:48:24 | [754][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0309 ntime: 0084 mem: 3.36 + 04-04 14:48:32 | [754][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 14:48:41 | [754][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1528 ntime: 0082 mem: 3.36 + 04-04 14:48:48 | [754][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0156 ntime: 0086 mem: 3.36 + 04-04 14:48:57 | [754][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1053 ntime: 0080 mem: 3.36 + 04-04 14:49:07 | [754][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1321 ntime: 0058 mem: 3.36 + 04-04 14:49:16 | [754][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1370 ntime: 0080 mem: 3.36 + 04-04 14:49:25 | [754][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1487 ntime: 0093 mem: 3.36 + 04-04 14:49:32 | [754][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0089 ntime: 0080 mem: 3.36 + 04-04 14:49:40 | [754][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0838 ntime: 0077 mem: 3.36 + 04-04 14:49:47 | [754][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1361 ntime: 0082 mem: 3.36 + 04-04 14:49:58 | [754][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1554 ntime: 0077 mem: 3.36 + 04-04 14:50:03 | [754][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0429 ntime: 0068 mem: 3.36 + 04-04 14:50:12 | [754][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0081 mem: 3.36 + 04-04 14:50:19 | Time info >>>> elapsed: 1046.95 mins remain: 339.74 mins + 04-04 14:50:21 | [755][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1081 ntime: 0076 mem: 3.36 + 04-04 14:50:28 | [755][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0073 ntime: 0074 mem: 3.36 + 04-04 14:50:35 | [755][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0081 mem: 3.36 + 04-04 14:50:44 | [755][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0128 ntime: 0078 mem: 3.36 + 04-04 14:50:51 | [755][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 14:51:00 | [755][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1659 ntime: 0081 mem: 3.36 + 04-04 14:51:07 | [755][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0062 ntime: 0084 mem: 3.36 + 04-04 14:51:16 | [755][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1064 ntime: 0073 mem: 3.36 + 04-04 14:51:24 | [755][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0078 ntime: 0087 mem: 3.36 + 04-04 14:51:32 | [755][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1179 ntime: 0077 mem: 3.36 + 04-04 14:51:40 | [755][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1285 ntime: 0087 mem: 3.36 + 04-04 14:51:47 | [755][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0095 ntime: 0081 mem: 3.36 + 04-04 14:51:54 | [755][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0059 mem: 3.36 + 04-04 14:52:00 | [755][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1108 ntime: 0072 mem: 3.36 + 04-04 14:52:08 | [755][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0413 ntime: 0079 mem: 3.36 + 04-04 14:52:17 | [755][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1147 ntime: 0074 mem: 3.36 + 04-04 14:52:27 | [755][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0078 ntime: 0079 mem: 3.36 + 04-04 14:52:35 | [755][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0988 ntime: 0079 mem: 3.36 + 04-04 14:52:42 | Time info >>>> elapsed: 1049.32 mins remain: 338.67 mins + 04-04 14:52:42 | [756][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0085 ntime: 0078 mem: 3.36 + 04-04 14:52:53 | [756][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0578 ntime: 0088 mem: 3.36 + 04-04 14:53:03 | [756][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1718 ntime: 0077 mem: 3.36 + 04-04 14:53:09 | [756][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1594 ntime: 0085 mem: 3.36 + 04-04 14:53:18 | [756][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1392 ntime: 0083 mem: 3.36 + 04-04 14:53:27 | [756][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1600 ntime: 0082 mem: 3.36 + 04-04 14:53:34 | [756][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0555 ntime: 0084 mem: 3.36 + 04-04 14:53:40 | [756][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0366 ntime: 0072 mem: 3.36 + 04-04 14:53:48 | [756][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0946 ntime: 0084 mem: 3.36 + 04-04 14:53:54 | [756][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0081 mem: 3.36 + 04-04 14:54:03 | [756][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1050 ntime: 0078 mem: 3.36 + 04-04 14:54:10 | [756][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1452 ntime: 0074 mem: 3.36 + 04-04 14:54:18 | [756][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0743 ntime: 0080 mem: 3.36 + 04-04 14:54:25 | [756][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0075 ntime: 0078 mem: 3.36 + 04-04 14:54:34 | [756][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0494 ntime: 0079 mem: 3.36 + 04-04 14:54:44 | [756][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1660 ntime: 0083 mem: 3.36 + 04-04 14:54:52 | [756][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0105 ntime: 0056 mem: 3.36 + 04-04 14:54:59 | [756][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0086 ntime: 0077 mem: 3.36 + 04-04 14:55:06 | Time info >>>> elapsed: 1051.72 mins remain: 337.61 mins + 04-04 14:55:07 | [757][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1511 ntime: 0078 mem: 3.36 + 04-04 14:55:15 | [757][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0089 ntime: 0085 mem: 3.36 + 04-04 14:55:23 | [757][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 14:55:32 | [757][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0720 ntime: 0079 mem: 3.36 + 04-04 14:55:38 | [757][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1403 ntime: 0079 mem: 3.36 + 04-04 14:55:48 | [757][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1551 ntime: 0086 mem: 3.36 + 04-04 14:55:56 | [757][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0667 ntime: 0077 mem: 3.36 + 04-04 14:56:03 | [757][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0090 ntime: 0076 mem: 3.36 + 04-04 14:56:12 | [757][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1294 ntime: 0079 mem: 3.36 + 04-04 14:56:20 | [757][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0077 mem: 3.36 + 04-04 14:56:29 | [757][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0138 ntime: 0083 mem: 3.36 + 04-04 14:56:37 | [757][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0642 ntime: 0080 mem: 3.36 + 04-04 14:56:46 | [757][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0108 ntime: 0075 mem: 3.36 + 04-04 14:56:58 | [757][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0088 mem: 3.36 + 04-04 14:57:07 | [757][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1379 ntime: 0076 mem: 3.36 + 04-04 14:57:14 | [757][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0596 ntime: 0079 mem: 3.36 + 04-04 14:57:20 | [757][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1085 ntime: 0077 mem: 3.36 + 04-04 14:57:26 | [757][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0646 ntime: 0086 mem: 3.36 + 04-04 14:57:33 | Time info >>>> elapsed: 1054.18 mins remain: 336.56 mins + 04-04 14:57:34 | [758][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0958 ntime: 0083 mem: 3.36 + 04-04 14:57:39 | [758][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0071 ntime: 0081 mem: 3.36 + 04-04 14:57:47 | [758][020/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 1544 ntime: 0085 mem: 3.36 + 04-04 14:57:54 | [758][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0232 ntime: 0075 mem: 3.36 + 04-04 14:58:03 | [758][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0696 ntime: 0079 mem: 3.36 + 04-04 14:58:10 | [758][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0250 ntime: 0077 mem: 3.36 + 04-04 14:58:18 | [758][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0133 ntime: 0077 mem: 3.36 + 04-04 14:58:26 | [758][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0713 ntime: 0077 mem: 3.36 + 04-04 14:58:34 | [758][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0087 ntime: 0082 mem: 3.36 + 04-04 14:58:41 | [758][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1176 ntime: 0075 mem: 3.36 + 04-04 14:58:46 | [758][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0165 ntime: 0081 mem: 3.36 + 04-04 14:58:52 | [758][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0083 ntime: 0078 mem: 3.36 + 04-04 14:58:58 | [758][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0141 ntime: 0087 mem: 3.36 + 04-04 14:59:05 | [758][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1048 ntime: 0082 mem: 3.36 + 04-04 14:59:11 | [758][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0808 ntime: 0082 mem: 3.36 + 04-04 14:59:17 | [758][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0578 ntime: 0084 mem: 3.36 + 04-04 14:59:24 | [758][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0388 ntime: 0076 mem: 3.36 + 04-04 14:59:31 | [758][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0073 mem: 3.36 + 04-04 14:59:36 | Time info >>>> elapsed: 1056.23 mins remain: 335.38 mins + 04-04 14:59:37 | [759][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0913 ntime: 0075 mem: 3.36 + 04-04 14:59:46 | [759][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0067 ntime: 0076 mem: 3.36 + 04-04 14:59:53 | [759][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1282 ntime: 0078 mem: 3.36 + 04-04 15:00:00 | [759][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1027 ntime: 0081 mem: 3.36 + 04-04 15:00:07 | [759][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0927 ntime: 0077 mem: 3.36 + 04-04 15:00:14 | [759][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0101 ntime: 0078 mem: 3.36 + 04-04 15:00:21 | [759][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0971 ntime: 0080 mem: 3.36 + 04-04 15:00:28 | [759][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0076 mem: 3.36 + 04-04 15:00:35 | [759][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0826 ntime: 0072 mem: 3.36 + 04-04 15:00:43 | [759][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0076 mem: 3.36 + 04-04 15:00:51 | [759][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0100 ntime: 0080 mem: 3.36 + 04-04 15:00:59 | [759][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1529 ntime: 0081 mem: 3.36 + 04-04 15:01:07 | [759][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1597 ntime: 0085 mem: 3.36 + 04-04 15:01:14 | [759][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0963 ntime: 0084 mem: 3.36 + 04-04 15:01:23 | [759][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0974 ntime: 0077 mem: 3.36 + 04-04 15:01:31 | [759][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0075 mem: 3.36 + 04-04 15:01:39 | [759][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0981 ntime: 0080 mem: 3.36 + 04-04 15:01:45 | [759][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0065 ntime: 0080 mem: 3.36 + 04-04 15:01:50 | Time info >>>> elapsed: 1058.46 mins remain: 334.25 mins + 04-04 15:01:51 | [760][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0914 ntime: 0081 mem: 3.36 + 04-04 15:01:59 | [760][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0081 mem: 3.36 + 04-04 15:02:06 | [760][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0076 mem: 3.36 + 04-04 15:02:13 | [760][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0109 ntime: 0095 mem: 3.36 + 04-04 15:02:22 | [760][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1265 ntime: 0085 mem: 3.36 + 04-04 15:02:28 | [760][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0089 ntime: 0079 mem: 3.36 + 04-04 15:02:35 | [760][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0584 ntime: 0081 mem: 3.36 + 04-04 15:02:41 | [760][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0481 ntime: 0079 mem: 3.36 + 04-04 15:02:47 | [760][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0475 ntime: 0077 mem: 3.36 + 04-04 15:02:55 | [760][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0121 ntime: 0078 mem: 3.36 + 04-04 15:03:02 | [760][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0285 ntime: 0077 mem: 3.36 + 04-04 15:03:08 | [760][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0937 ntime: 0082 mem: 3.36 + 04-04 15:03:14 | [760][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0151 ntime: 0078 mem: 3.36 + 04-04 15:03:21 | [760][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0103 ntime: 0082 mem: 3.36 + 04-04 15:03:30 | [760][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1611 ntime: 0071 mem: 3.36 + 04-04 15:03:39 | [760][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0140 ntime: 0076 mem: 3.36 + 04-04 15:03:47 | [760][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1077 ntime: 0072 mem: 3.36 + 04-04 15:03:54 | [760][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1167 ntime: 0077 mem: 3.36 + 04-04 15:03:58 | Time info >>>> elapsed: 1060.59 mins remain: 333.09 mins + 04-04 15:03:58 | [761][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0079 mem: 3.36 + 04-04 15:04:05 | [761][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0654 ntime: 0080 mem: 3.36 + 04-04 15:04:12 | [761][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0129 ntime: 0078 mem: 3.36 + 04-04 15:04:20 | [761][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0473 ntime: 0079 mem: 3.36 + 04-04 15:04:27 | [761][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0062 ntime: 0080 mem: 3.36 + 04-04 15:04:34 | [761][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0062 ntime: 0082 mem: 3.36 + 04-04 15:04:42 | [761][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1029 ntime: 0079 mem: 3.36 + 04-04 15:04:49 | [761][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0600 ntime: 0081 mem: 3.36 + 04-04 15:04:53 | [761][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0595 ntime: 0081 mem: 3.36 + 04-04 15:04:59 | [761][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0727 ntime: 0081 mem: 3.36 + 04-04 15:05:05 | [761][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0088 ntime: 0084 mem: 3.36 + 04-04 15:05:12 | [761][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0139 ntime: 0081 mem: 3.36 + 04-04 15:05:18 | [761][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0082 ntime: 0083 mem: 3.36 + 04-04 15:05:26 | [761][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0897 ntime: 0081 mem: 3.36 + 04-04 15:05:34 | [761][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0639 ntime: 0083 mem: 3.36 + 04-04 15:05:41 | [761][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1196 ntime: 0081 mem: 3.36 + 04-04 15:05:50 | [761][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1077 ntime: 0084 mem: 3.36 + 04-04 15:05:56 | [761][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0122 ntime: 0079 mem: 3.36 + 04-04 15:06:02 | Time info >>>> elapsed: 1062.66 mins remain: 331.91 mins + 04-04 15:06:03 | [762][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1323 ntime: 0080 mem: 3.36 + 04-04 15:06:11 | [762][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0956 ntime: 0073 mem: 3.36 + 04-04 15:06:18 | [762][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0063 ntime: 0082 mem: 3.36 + 04-04 15:06:24 | [762][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0115 ntime: 0078 mem: 3.36 + 04-04 15:06:31 | [762][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0100 ntime: 0094 mem: 3.36 + 04-04 15:06:38 | [762][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0334 ntime: 0080 mem: 3.36 + 04-04 15:06:46 | [762][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0796 ntime: 0078 mem: 3.36 + 04-04 15:06:53 | [762][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1487 ntime: 0083 mem: 3.36 + 04-04 15:06:58 | [762][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1135 ntime: 0080 mem: 3.36 + 04-04 15:07:04 | [762][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0100 ntime: 0084 mem: 3.36 + 04-04 15:07:11 | [762][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0074 ntime: 0080 mem: 3.36 + 04-04 15:07:18 | [762][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0888 ntime: 0082 mem: 3.36 + 04-04 15:07:24 | [762][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1416 ntime: 0089 mem: 3.36 + 04-04 15:07:33 | [762][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1127 ntime: 0086 mem: 3.36 + 04-04 15:07:41 | [762][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1048 ntime: 0079 mem: 3.36 + 04-04 15:07:47 | [762][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0836 ntime: 0082 mem: 3.36 + 04-04 15:07:52 | [762][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0521 ntime: 0078 mem: 3.36 + 04-04 15:07:59 | [762][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0072 ntime: 0080 mem: 3.36 + 04-04 15:08:06 | Time info >>>> elapsed: 1064.72 mins remain: 330.72 mins + 04-04 15:08:06 | [763][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0672 ntime: 0085 mem: 3.36 + 04-04 15:08:13 | [763][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0077 mem: 3.36 + 04-04 15:08:21 | [763][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0724 ntime: 0083 mem: 3.36 + 04-04 15:08:29 | [763][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0990 ntime: 0078 mem: 3.36 + 04-04 15:08:36 | [763][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1446 ntime: 0090 mem: 3.36 + 04-04 15:08:43 | [763][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0089 ntime: 0085 mem: 3.36 + 04-04 15:08:50 | [763][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0080 mem: 3.36 + 04-04 15:08:55 | [763][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0073 mem: 3.36 + 04-04 15:09:02 | [763][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0641 ntime: 0079 mem: 3.36 + 04-04 15:09:08 | [763][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0880 ntime: 0080 mem: 3.36 + 04-04 15:09:15 | [763][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0086 mem: 3.36 + 04-04 15:09:24 | [763][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1605 ntime: 0088 mem: 3.36 + 04-04 15:09:32 | [763][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0073 mem: 3.36 + 04-04 15:09:39 | [763][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1133 ntime: 0080 mem: 3.36 + 04-04 15:09:46 | [763][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0129 ntime: 0083 mem: 3.36 + 04-04 15:09:53 | [763][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0072 ntime: 0075 mem: 3.36 + 04-04 15:09:59 | [763][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0586 ntime: 0088 mem: 3.36 + 04-04 15:10:05 | [763][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0082 mem: 3.36 + 04-04 15:10:10 | Time info >>>> elapsed: 1066.80 mins remain: 329.53 mins + 04-04 15:10:10 | [764][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0061 ntime: 0075 mem: 3.36 + 04-04 15:10:17 | [764][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0784 ntime: 0078 mem: 3.36 + 04-04 15:10:23 | [764][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0618 ntime: 0080 mem: 3.36 + 04-04 15:10:32 | [764][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0109 ntime: 0078 mem: 3.36 + 04-04 15:10:39 | [764][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0887 ntime: 0080 mem: 3.36 + 04-04 15:10:46 | [764][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1499 ntime: 0077 mem: 3.36 + 04-04 15:10:52 | [764][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0083 ntime: 0075 mem: 3.36 + 04-04 15:11:02 | [764][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0536 ntime: 0090 mem: 3.36 + 04-04 15:11:10 | [764][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1076 ntime: 0079 mem: 3.36 + 04-04 15:11:17 | [764][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0126 ntime: 0077 mem: 3.36 + 04-04 15:11:28 | [764][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0819 ntime: 0080 mem: 3.36 + 04-04 15:11:34 | [764][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0082 mem: 3.36 + 04-04 15:11:41 | [764][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0087 mem: 3.36 + 04-04 15:11:49 | [764][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1083 ntime: 0088 mem: 3.36 + 04-04 15:11:57 | [764][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1301 ntime: 0077 mem: 3.36 + 04-04 15:12:02 | [764][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0603 ntime: 0080 mem: 3.36 + 04-04 15:12:06 | [764][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0127 ntime: 0087 mem: 3.36 + 04-04 15:12:12 | [764][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0081 ntime: 0082 mem: 3.36 + 04-04 15:12:17 | Time info >>>> elapsed: 1068.91 mins remain: 328.36 mins + 04-04 15:12:18 | [765][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1211 ntime: 0062 mem: 3.36 + 04-04 15:12:25 | [765][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0185 ntime: 0083 mem: 3.36 + 04-04 15:12:34 | [765][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1608 ntime: 0078 mem: 3.36 + 04-04 15:12:41 | [765][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1106 ntime: 0086 mem: 3.36 + 04-04 15:12:46 | [765][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 15:12:52 | [765][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0077 ntime: 0077 mem: 3.36 + 04-04 15:12:58 | [765][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0074 mem: 3.36 + 04-04 15:13:04 | [765][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0085 ntime: 0076 mem: 3.36 + 04-04 15:13:11 | [765][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1033 ntime: 0085 mem: 3.36 + 04-04 15:13:18 | [765][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0096 ntime: 0084 mem: 3.36 + 04-04 15:13:26 | [765][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0167 ntime: 0079 mem: 3.36 + 04-04 15:13:32 | [765][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0385 ntime: 0087 mem: 3.36 + 04-04 15:13:39 | [765][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0834 ntime: 0082 mem: 3.36 + 04-04 15:13:46 | [765][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1199 ntime: 0084 mem: 3.36 + 04-04 15:13:52 | [765][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0504 ntime: 0087 mem: 3.36 + 04-04 15:13:59 | [765][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0084 ntime: 0081 mem: 3.36 + 04-04 15:14:08 | [765][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0092 ntime: 0080 mem: 3.36 + 04-04 15:14:17 | [765][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1014 ntime: 0077 mem: 3.36 + 04-04 15:14:22 | Time info >>>> elapsed: 1071.00 mins remain: 327.17 mins + 04-04 15:14:23 | [766][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0166 ntime: 0077 mem: 3.36 + 04-04 15:14:29 | [766][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0086 ntime: 0084 mem: 3.36 + 04-04 15:14:37 | [766][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0230 ntime: 0074 mem: 3.36 + 04-04 15:14:45 | [766][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1092 ntime: 0084 mem: 3.36 + 04-04 15:14:52 | [766][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0181 ntime: 0077 mem: 3.36 + 04-04 15:15:01 | [766][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0138 ntime: 0084 mem: 3.36 + 04-04 15:15:13 | [766][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0096 ntime: 0078 mem: 3.36 + 04-04 15:15:22 | [766][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1443 ntime: 0079 mem: 3.36 + 04-04 15:15:30 | [766][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1225 ntime: 0084 mem: 3.36 + 04-04 15:15:37 | [766][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1295 ntime: 0080 mem: 3.36 + 04-04 15:15:43 | [766][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0189 ntime: 0074 mem: 3.36 + 04-04 15:15:52 | [766][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0705 ntime: 0080 mem: 3.36 + 04-04 15:16:00 | [766][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0077 mem: 3.36 + 04-04 15:16:09 | [766][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0712 ntime: 0083 mem: 3.36 + 04-04 15:16:16 | [766][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1090 ntime: 0088 mem: 3.36 + 04-04 15:16:23 | [766][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0072 mem: 3.36 + 04-04 15:16:37 | [766][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1280 ntime: 0077 mem: 3.36 + 04-04 15:16:46 | [766][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0426 ntime: 0078 mem: 3.36 + 04-04 15:16:52 | Time info >>>> elapsed: 1073.49 mins remain: 326.11 mins + 04-04 15:16:52 | [767][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0082 ntime: 0080 mem: 3.36 + 04-04 15:16:57 | [767][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 15:17:06 | [767][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0837 ntime: 0085 mem: 3.36 + 04-04 15:17:14 | [767][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0178 ntime: 0077 mem: 3.36 + 04-04 15:17:21 | [767][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0501 ntime: 0079 mem: 3.36 + 04-04 15:17:29 | [767][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1356 ntime: 0084 mem: 3.36 + 04-04 15:17:38 | [767][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0177 ntime: 0079 mem: 3.36 + 04-04 15:17:46 | [767][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1166 ntime: 0096 mem: 3.36 + 04-04 15:17:53 | [767][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0691 ntime: 0079 mem: 3.36 + 04-04 15:18:00 | [767][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0124 ntime: 0079 mem: 3.36 + 04-04 15:18:08 | [767][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1172 ntime: 0086 mem: 3.36 + 04-04 15:18:15 | [767][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1247 ntime: 0081 mem: 3.36 + 04-04 15:18:21 | [767][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0588 ntime: 0071 mem: 3.36 + 04-04 15:18:28 | [767][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0078 ntime: 0078 mem: 3.36 + 04-04 15:18:37 | [767][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1062 ntime: 0089 mem: 3.36 + 04-04 15:18:44 | [767][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1240 ntime: 0081 mem: 3.36 + 04-04 15:18:53 | [767][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1056 ntime: 0086 mem: 3.36 + 04-04 15:18:59 | [767][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0087 ntime: 0078 mem: 3.36 + 04-04 15:19:04 | Time info >>>> elapsed: 1075.69 mins remain: 324.95 mins + 04-04 15:19:04 | [768][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0079 ntime: 0079 mem: 3.36 + 04-04 15:19:11 | [768][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0089 ntime: 0084 mem: 3.36 + 04-04 15:19:17 | [768][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0071 ntime: 0080 mem: 3.36 + 04-04 15:19:23 | [768][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0979 ntime: 0076 mem: 3.36 + 04-04 15:19:30 | [768][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 15:19:37 | [768][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0103 ntime: 0081 mem: 3.36 + 04-04 15:19:45 | [768][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0069 mem: 3.36 + 04-04 15:19:51 | [768][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0646 ntime: 0078 mem: 3.36 + 04-04 15:19:58 | [768][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0194 ntime: 0080 mem: 3.36 + 04-04 15:20:06 | [768][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1028 ntime: 0082 mem: 3.36 + 04-04 15:20:13 | [768][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0103 ntime: 0078 mem: 3.36 + 04-04 15:20:21 | [768][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0274 ntime: 0079 mem: 3.36 + 04-04 15:20:27 | [768][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0157 ntime: 0083 mem: 3.36 + 04-04 15:20:35 | [768][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1561 ntime: 0075 mem: 3.36 + 04-04 15:20:44 | [768][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1230 ntime: 0078 mem: 3.36 + 04-04 15:20:51 | [768][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0091 ntime: 0076 mem: 3.36 + 04-04 15:20:56 | [768][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0078 mem: 3.36 + 04-04 15:21:02 | [768][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0658 ntime: 0086 mem: 3.36 + 04-04 15:21:08 | Time info >>>> elapsed: 1077.76 mins remain: 323.75 mins + 04-04 15:21:08 | [769][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0168 ntime: 0078 mem: 3.36 + 04-04 15:21:15 | [769][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1699 ntime: 0074 mem: 3.36 + 04-04 15:21:20 | [769][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0070 ntime: 0085 mem: 3.36 + 04-04 15:21:27 | [769][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0109 ntime: 0076 mem: 3.36 + 04-04 15:21:34 | [769][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0956 ntime: 0081 mem: 3.36 + 04-04 15:21:41 | [769][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0096 ntime: 0076 mem: 3.36 + 04-04 15:21:47 | [769][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1107 ntime: 0081 mem: 3.36 + 04-04 15:21:53 | [769][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0103 ntime: 0075 mem: 3.36 + 04-04 15:22:02 | [769][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0881 ntime: 0076 mem: 3.36 + 04-04 15:22:09 | [769][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0888 ntime: 0082 mem: 3.36 + 04-04 15:22:16 | [769][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 15:22:24 | [769][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1022 ntime: 0079 mem: 3.36 + 04-04 15:22:31 | [769][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1459 ntime: 0076 mem: 3.36 + 04-04 15:22:36 | [769][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0085 mem: 3.36 + 04-04 15:22:42 | [769][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1160 ntime: 0084 mem: 3.36 + 04-04 15:22:51 | [769][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0397 ntime: 0081 mem: 3.36 + 04-04 15:23:00 | [769][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1473 ntime: 0077 mem: 3.36 + 04-04 15:23:08 | [769][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0164 ntime: 0080 mem: 3.36 + 04-04 15:23:12 | Time info >>>> elapsed: 1079.83 mins remain: 322.55 mins + 04-04 15:23:14 | [770][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1449 ntime: 0080 mem: 3.36 + 04-04 15:23:20 | [770][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0588 ntime: 0074 mem: 3.36 + 04-04 15:23:29 | [770][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0131 ntime: 0078 mem: 3.36 + 04-04 15:23:37 | [770][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0547 ntime: 0083 mem: 3.36 + 04-04 15:23:42 | [770][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0081 mem: 3.36 + 04-04 15:23:50 | [770][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0680 ntime: 0084 mem: 3.36 + 04-04 15:23:58 | [770][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0519 ntime: 0081 mem: 3.36 + 04-04 15:24:06 | [770][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0083 mem: 3.36 + 04-04 15:24:14 | [770][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1084 ntime: 0080 mem: 3.36 + 04-04 15:24:23 | [770][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1165 ntime: 0079 mem: 3.36 + 04-04 15:24:31 | [770][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0902 ntime: 0085 mem: 3.36 + 04-04 15:24:38 | [770][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1395 ntime: 0086 mem: 3.36 + 04-04 15:24:44 | [770][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0257 ntime: 0081 mem: 3.36 + 04-04 15:24:52 | [770][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0830 ntime: 0081 mem: 3.36 + 04-04 15:24:59 | [770][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0939 ntime: 0079 mem: 3.36 + 04-04 15:25:06 | [770][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1124 ntime: 0085 mem: 3.36 + 04-04 15:25:14 | [770][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1088 ntime: 0078 mem: 3.36 + 04-04 15:25:22 | [770][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0556 ntime: 0079 mem: 3.36 + 04-04 15:25:27 | Time info >>>> elapsed: 1082.08 mins remain: 321.40 mins + 04-04 15:25:27 | [771][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0186 ntime: 0081 mem: 3.36 + 04-04 15:25:34 | [771][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1030 ntime: 0077 mem: 3.36 + 04-04 15:25:40 | [771][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0541 ntime: 0077 mem: 3.36 + 04-04 15:25:47 | [771][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0082 mem: 3.36 + 04-04 15:25:55 | [771][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0066 mem: 3.36 + 04-04 15:26:02 | [771][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0854 ntime: 0075 mem: 3.36 + 04-04 15:26:10 | [771][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1206 ntime: 0086 mem: 3.36 + 04-04 15:26:17 | [771][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1402 ntime: 0079 mem: 3.36 + 04-04 15:26:23 | [771][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0124 ntime: 0076 mem: 3.36 + 04-04 15:26:30 | [771][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0151 ntime: 0081 mem: 3.36 + 04-04 15:26:38 | [771][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0475 ntime: 0078 mem: 3.36 + 04-04 15:26:44 | [771][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1335 ntime: 0085 mem: 3.36 + 04-04 15:26:53 | [771][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0978 ntime: 0076 mem: 3.36 + 04-04 15:26:57 | [771][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0135 ntime: 0081 mem: 3.36 + 04-04 15:27:05 | [771][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0913 ntime: 0086 mem: 3.36 + 04-04 15:27:13 | [771][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0167 ntime: 0080 mem: 3.36 + 04-04 15:27:22 | [771][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1225 ntime: 0085 mem: 3.36 + 04-04 15:27:29 | [771][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0272 ntime: 0083 mem: 3.36 + 04-04 15:27:34 | Time info >>>> elapsed: 1084.20 mins remain: 320.20 mins + 04-04 15:27:35 | [772][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0698 ntime: 0081 mem: 3.36 + 04-04 15:27:42 | [772][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1445 ntime: 0078 mem: 3.36 + 04-04 15:27:50 | [772][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0336 ntime: 0072 mem: 3.36 + 04-04 15:27:56 | [772][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0702 ntime: 0078 mem: 3.36 + 04-04 15:28:05 | [772][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1143 ntime: 0081 mem: 3.36 + 04-04 15:28:11 | [772][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0758 ntime: 0080 mem: 3.36 + 04-04 15:28:18 | [772][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1442 ntime: 0086 mem: 3.36 + 04-04 15:28:24 | [772][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0082 ntime: 0074 mem: 3.36 + 04-04 15:28:32 | [772][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1486 ntime: 0082 mem: 3.36 + 04-04 15:28:37 | [772][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0284 ntime: 0080 mem: 3.36 + 04-04 15:28:44 | [772][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0879 ntime: 0075 mem: 3.36 + 04-04 15:28:50 | [772][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0566 ntime: 0082 mem: 3.36 + 04-04 15:28:57 | [772][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0119 ntime: 0078 mem: 3.36 + 04-04 15:29:05 | [772][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0465 ntime: 0076 mem: 3.36 + 04-04 15:29:13 | [772][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0570 ntime: 0077 mem: 3.36 + 04-04 15:29:21 | [772][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0073 ntime: 0082 mem: 3.36 + 04-04 15:29:28 | [772][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0870 ntime: 0080 mem: 3.36 + 04-04 15:29:35 | [772][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0998 ntime: 0086 mem: 3.36 + 04-04 15:29:39 | Time info >>>> elapsed: 1086.28 mins remain: 319.00 mins + 04-04 15:29:40 | [773][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0717 ntime: 0075 mem: 3.36 + 04-04 15:29:45 | [773][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0086 mem: 3.36 + 04-04 15:29:51 | [773][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0088 mem: 3.36 + 04-04 15:29:56 | [773][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0469 ntime: 0083 mem: 3.36 + 04-04 15:30:04 | [773][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0858 ntime: 0082 mem: 3.36 + 04-04 15:30:09 | [773][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0675 ntime: 0077 mem: 3.36 + 04-04 15:30:15 | [773][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0835 ntime: 0083 mem: 3.36 + 04-04 15:30:21 | [773][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0120 ntime: 0077 mem: 3.36 + 04-04 15:30:28 | [773][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1316 ntime: 0086 mem: 3.36 + 04-04 15:30:35 | [773][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0079 ntime: 0089 mem: 3.36 + 04-04 15:30:41 | [773][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0716 ntime: 0078 mem: 3.36 + 04-04 15:30:47 | [773][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0806 ntime: 0089 mem: 3.36 + 04-04 15:30:53 | [773][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0626 ntime: 0082 mem: 3.36 + 04-04 15:30:58 | [773][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1009 ntime: 0077 mem: 3.36 + 04-04 15:31:06 | [773][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0086 ntime: 0077 mem: 3.36 + 04-04 15:31:13 | [773][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0502 ntime: 0081 mem: 3.36 + 04-04 15:31:18 | [773][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0078 mem: 3.36 + 04-04 15:31:24 | [773][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0107 ntime: 0081 mem: 3.36 + 04-04 15:31:29 | Time info >>>> elapsed: 1088.12 mins remain: 317.72 mins + 04-04 15:31:30 | [774][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0067 ntime: 0078 mem: 3.36 + 04-04 15:31:36 | [774][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0190 ntime: 0082 mem: 3.36 + 04-04 15:31:41 | [774][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0547 ntime: 0076 mem: 3.36 + 04-04 15:31:51 | [774][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1581 ntime: 0085 mem: 3.36 + 04-04 15:31:58 | [774][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0068 ntime: 0081 mem: 3.36 + 04-04 15:32:03 | [774][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0177 ntime: 0086 mem: 3.36 + 04-04 15:32:10 | [774][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1303 ntime: 0085 mem: 3.36 + 04-04 15:32:17 | [774][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0104 ntime: 0081 mem: 3.36 + 04-04 15:32:24 | [774][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0092 ntime: 0076 mem: 3.36 + 04-04 15:32:29 | [774][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0084 ntime: 0078 mem: 3.36 + 04-04 15:32:38 | [774][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0675 ntime: 0084 mem: 3.36 + 04-04 15:32:45 | [774][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1256 ntime: 0080 mem: 3.36 + 04-04 15:32:52 | [774][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0625 ntime: 0060 mem: 3.36 + 04-04 15:32:58 | [774][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0078 ntime: 0079 mem: 3.36 + 04-04 15:33:05 | [774][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0456 ntime: 0071 mem: 3.36 + 04-04 15:33:11 | [774][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0072 ntime: 0081 mem: 3.36 + 04-04 15:33:19 | [774][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1150 ntime: 0079 mem: 3.36 + 04-04 15:33:24 | [774][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0107 ntime: 0080 mem: 3.36 + 04-04 15:33:29 | Time info >>>> elapsed: 1090.11 mins remain: 316.48 mins + 04-04 15:33:29 | [775][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0225 ntime: 0080 mem: 3.36 + 04-04 15:33:36 | [775][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0109 ntime: 0077 mem: 3.36 + 04-04 15:33:43 | [775][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0077 ntime: 0072 mem: 3.36 + 04-04 15:33:51 | [775][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0740 ntime: 0079 mem: 3.36 + 04-04 15:33:57 | [775][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0132 ntime: 0082 mem: 3.36 + 04-04 15:34:03 | [775][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 15:34:09 | [775][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0691 ntime: 0102 mem: 3.36 + 04-04 15:34:14 | [775][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1200 ntime: 0081 mem: 3.36 + 04-04 15:34:20 | [775][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0078 ntime: 0089 mem: 3.36 + 04-04 15:34:26 | [775][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0113 ntime: 0078 mem: 3.36 + 04-04 15:34:33 | [775][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0706 ntime: 0086 mem: 3.36 + 04-04 15:34:40 | [775][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1034 ntime: 0082 mem: 3.36 + 04-04 15:34:46 | [775][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0165 ntime: 0078 mem: 3.36 + 04-04 15:34:53 | [775][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0569 ntime: 0075 mem: 3.36 + 04-04 15:35:02 | [775][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0980 ntime: 0077 mem: 3.36 + 04-04 15:35:08 | [775][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0080 mem: 3.36 + 04-04 15:35:16 | [775][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1278 ntime: 0078 mem: 3.36 + 04-04 15:35:23 | [775][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0626 ntime: 0080 mem: 3.36 + 04-04 15:35:28 | Time info >>>> elapsed: 1092.10 mins remain: 315.24 mins + 04-04 15:35:28 | [776][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0074 mem: 3.36 + 04-04 15:35:35 | [776][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0091 ntime: 0080 mem: 3.36 + 04-04 15:35:42 | [776][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1120 ntime: 0076 mem: 3.36 + 04-04 15:35:50 | [776][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1235 ntime: 0086 mem: 3.36 + 04-04 15:35:57 | [776][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1080 ntime: 0078 mem: 3.36 + 04-04 15:36:02 | [776][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0933 ntime: 0073 mem: 3.36 + 04-04 15:36:09 | [776][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0903 ntime: 0079 mem: 3.36 + 04-04 15:36:15 | [776][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0169 ntime: 0078 mem: 3.36 + 04-04 15:36:21 | [776][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0119 ntime: 0079 mem: 3.36 + 04-04 15:36:29 | [776][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0156 ntime: 0081 mem: 3.36 + 04-04 15:36:37 | [776][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1191 ntime: 0087 mem: 3.36 + 04-04 15:36:46 | [776][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1102 ntime: 0084 mem: 3.36 + 04-04 15:36:52 | [776][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0182 ntime: 0080 mem: 3.36 + 04-04 15:37:00 | [776][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1111 ntime: 0078 mem: 3.36 + 04-04 15:37:07 | [776][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1072 ntime: 0077 mem: 3.36 + 04-04 15:37:14 | [776][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0715 ntime: 0070 mem: 3.36 + 04-04 15:37:21 | [776][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0138 ntime: 0080 mem: 3.36 + 04-04 15:37:28 | [776][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1531 ntime: 0083 mem: 3.36 + 04-04 15:37:34 | Time info >>>> elapsed: 1094.19 mins remain: 314.04 mins + 04-04 15:37:34 | [777][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0054 ntime: 0075 mem: 3.36 + 04-04 15:37:42 | [777][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1428 ntime: 0081 mem: 3.36 + 04-04 15:37:48 | [777][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0106 ntime: 0078 mem: 3.36 + 04-04 15:37:54 | [777][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0075 mem: 3.36 + 04-04 15:37:59 | [777][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1242 ntime: 0075 mem: 3.36 + 04-04 15:38:07 | [777][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0965 ntime: 0075 mem: 3.36 + 04-04 15:38:14 | [777][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0603 ntime: 0078 mem: 3.36 + 04-04 15:38:23 | [777][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1785 ntime: 0078 mem: 3.36 + 04-04 15:38:31 | [777][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0866 ntime: 0083 mem: 3.36 + 04-04 15:38:37 | [777][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0866 ntime: 0072 mem: 3.36 + 04-04 15:38:44 | [777][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0853 ntime: 0079 mem: 3.36 + 04-04 15:38:53 | [777][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1605 ntime: 0090 mem: 3.36 + 04-04 15:38:58 | [777][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0118 ntime: 0078 mem: 3.36 + 04-04 15:39:10 | [777][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1245 ntime: 0079 mem: 3.36 + 04-04 15:39:17 | [777][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0142 ntime: 0077 mem: 3.36 + 04-04 15:39:24 | [777][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0657 ntime: 0074 mem: 3.36 + 04-04 15:39:31 | [777][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1259 ntime: 0077 mem: 3.36 + 04-04 15:39:37 | [777][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0101 ntime: 0081 mem: 3.36 + 04-04 15:39:43 | Time info >>>> elapsed: 1096.34 mins remain: 312.84 mins + 04-04 15:39:44 | [778][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1034 ntime: 0087 mem: 3.36 + 04-04 15:39:52 | [778][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1453 ntime: 0085 mem: 3.36 + 04-04 15:39:58 | [778][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0657 ntime: 0081 mem: 3.36 + 04-04 15:40:05 | [778][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0521 ntime: 0079 mem: 3.36 + 04-04 15:40:15 | [778][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0710 ntime: 0084 mem: 3.36 + 04-04 15:40:23 | [778][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0112 ntime: 0085 mem: 3.36 + 04-04 15:40:29 | [778][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0167 ntime: 0081 mem: 3.36 + 04-04 15:40:34 | [778][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0976 ntime: 0080 mem: 3.36 + 04-04 15:40:40 | [778][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1007 ntime: 0086 mem: 3.36 + 04-04 15:40:47 | [778][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1065 ntime: 0077 mem: 3.36 + 04-04 15:40:54 | [778][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1586 ntime: 0082 mem: 3.36 + 04-04 15:41:01 | [778][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1113 ntime: 0076 mem: 3.36 + 04-04 15:41:09 | [778][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1028 ntime: 0076 mem: 3.36 + 04-04 15:41:16 | [778][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0981 ntime: 0079 mem: 3.36 + 04-04 15:41:24 | [778][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1215 ntime: 0075 mem: 3.36 + 04-04 15:41:31 | [778][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0127 ntime: 0081 mem: 3.36 + 04-04 15:41:38 | [778][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0095 ntime: 0079 mem: 3.36 + 04-04 15:41:45 | [778][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1119 ntime: 0057 mem: 3.36 + 04-04 15:41:48 | Time info >>>> elapsed: 1098.43 mins remain: 311.62 mins + 04-04 15:41:50 | [779][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1518 ntime: 0061 mem: 3.36 + 04-04 15:41:56 | [779][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0720 ntime: 0079 mem: 3.36 + 04-04 15:42:05 | [779][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1569 ntime: 0076 mem: 3.36 + 04-04 15:42:15 | [779][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1207 ntime: 0081 mem: 3.36 + 04-04 15:42:22 | [779][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0116 ntime: 0078 mem: 3.36 + 04-04 15:42:32 | [779][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1387 ntime: 0078 mem: 3.36 + 04-04 15:42:39 | [779][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0186 ntime: 0077 mem: 3.36 + 04-04 15:42:47 | [779][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0732 ntime: 0077 mem: 3.36 + 04-04 15:42:54 | [779][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0523 ntime: 0085 mem: 3.36 + 04-04 15:43:03 | [779][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1195 ntime: 0079 mem: 3.36 + 04-04 15:43:11 | [779][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0126 ntime: 0080 mem: 3.36 + 04-04 15:43:18 | [779][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1166 ntime: 0080 mem: 3.36 + 04-04 15:43:23 | [779][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0139 ntime: 0085 mem: 3.36 + 04-04 15:43:35 | [779][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1614 ntime: 0079 mem: 3.36 + 04-04 15:43:41 | [779][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0137 ntime: 0082 mem: 3.36 + 04-04 15:43:53 | [779][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1431 ntime: 0074 mem: 3.36 + 04-04 15:44:02 | [779][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0135 ntime: 0085 mem: 3.36 + 04-04 15:44:12 | [779][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1872 ntime: 0078 mem: 3.36 + 04-04 15:44:18 | Time info >>>> elapsed: 1100.92 mins remain: 310.52 mins + 04-04 15:44:19 | [780][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0781 ntime: 0088 mem: 3.36 + 04-04 15:44:24 | [780][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0075 ntime: 0084 mem: 3.36 + 04-04 15:44:35 | [780][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1503 ntime: 0080 mem: 3.36 + 04-04 15:44:42 | [780][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0607 ntime: 0059 mem: 3.36 + 04-04 15:44:53 | [780][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 2024 ntime: 0086 mem: 3.36 + 04-04 15:45:00 | [780][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0776 ntime: 0080 mem: 3.36 + 04-04 15:45:09 | [780][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1391 ntime: 0075 mem: 3.36 + 04-04 15:45:16 | [780][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0047 ntime: 0074 mem: 3.36 + 04-04 15:45:24 | [780][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0082 mem: 3.36 + 04-04 15:45:31 | [780][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0086 ntime: 0077 mem: 3.36 + 04-04 15:45:37 | [780][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0862 ntime: 0077 mem: 3.36 + 04-04 15:45:45 | [780][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1448 ntime: 0083 mem: 3.36 + 04-04 15:45:51 | [780][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0074 mem: 3.36 + 04-04 15:45:58 | [780][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0886 ntime: 0084 mem: 3.36 + 04-04 15:46:02 | [780][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0161 ntime: 0074 mem: 3.36 + 04-04 15:46:10 | [780][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0317 ntime: 0082 mem: 3.36 + 04-04 15:46:17 | [780][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0856 ntime: 0081 mem: 3.36 + 04-04 15:46:26 | [780][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0199 ntime: 0081 mem: 3.36 + 04-04 15:46:32 | Time info >>>> elapsed: 1103.15 mins remain: 309.34 mins + 04-04 15:46:32 | [781][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0076 mem: 3.36 + 04-04 15:46:40 | [781][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1288 ntime: 0077 mem: 3.36 + 04-04 15:46:45 | [781][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0119 ntime: 0086 mem: 3.36 + 04-04 15:46:51 | [781][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0120 ntime: 0078 mem: 3.36 + 04-04 15:46:57 | [781][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0077 mem: 3.36 + 04-04 15:47:04 | [781][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0069 ntime: 0084 mem: 3.36 + 04-04 15:47:09 | [781][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 15:47:19 | [781][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1243 ntime: 0075 mem: 3.36 + 04-04 15:47:23 | [781][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0065 ntime: 0078 mem: 3.36 + 04-04 15:47:30 | [781][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0079 ntime: 0082 mem: 3.36 + 04-04 15:47:37 | [781][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1294 ntime: 0088 mem: 3.36 + 04-04 15:47:43 | [781][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0163 ntime: 0055 mem: 3.36 + 04-04 15:47:51 | [781][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1481 ntime: 0084 mem: 3.36 + 04-04 15:47:56 | [781][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0647 ntime: 0058 mem: 3.36 + 04-04 15:48:04 | [781][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0660 ntime: 0080 mem: 3.36 + 04-04 15:48:09 | [781][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0661 ntime: 0076 mem: 3.36 + 04-04 15:48:17 | [781][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0054 ntime: 0079 mem: 3.36 + 04-04 15:48:24 | [781][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0082 ntime: 0086 mem: 3.36 + 04-04 15:48:31 | Time info >>>> elapsed: 1105.15 mins remain: 308.08 mins + 04-04 15:48:32 | [782][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0267 ntime: 0082 mem: 3.36 + 04-04 15:48:39 | [782][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0990 ntime: 0084 mem: 3.36 + 04-04 15:48:46 | [782][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1320 ntime: 0081 mem: 3.36 + 04-04 15:48:51 | [782][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0087 ntime: 0083 mem: 3.36 + 04-04 15:48:55 | [782][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0150 ntime: 0081 mem: 3.36 + 04-04 15:49:02 | [782][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0082 mem: 3.36 + 04-04 15:49:10 | [782][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0715 ntime: 0075 mem: 3.36 + 04-04 15:49:18 | [782][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1058 ntime: 0079 mem: 3.36 + 04-04 15:49:27 | [782][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1474 ntime: 0078 mem: 3.36 + 04-04 15:49:31 | [782][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0783 ntime: 0077 mem: 3.36 + 04-04 15:49:38 | [782][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0174 ntime: 0079 mem: 3.36 + 04-04 15:49:46 | [782][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0450 ntime: 0077 mem: 3.36 + 04-04 15:49:51 | [782][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0077 mem: 3.36 + 04-04 15:50:00 | [782][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0938 ntime: 0080 mem: 3.36 + 04-04 15:50:05 | [782][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0162 ntime: 0079 mem: 3.36 + 04-04 15:50:08 | [782][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0084 mem: 3.36 + 04-04 15:50:16 | [782][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0047 ntime: 0079 mem: 3.36 + 04-04 15:50:23 | [782][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1641 ntime: 0081 mem: 3.36 + 04-04 15:50:26 | Time info >>>> elapsed: 1107.07 mins remain: 306.81 mins + 04-04 15:50:28 | [783][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1452 ntime: 0084 mem: 3.36 + 04-04 15:50:35 | [783][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0068 ntime: 0080 mem: 3.36 + 04-04 15:50:41 | [783][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1062 ntime: 0086 mem: 3.36 + 04-04 15:50:48 | [783][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0535 ntime: 0081 mem: 3.36 + 04-04 15:50:54 | [783][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0617 ntime: 0071 mem: 3.36 + 04-04 15:51:00 | [783][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0570 ntime: 0080 mem: 3.36 + 04-04 15:51:14 | [783][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1006 ntime: 0079 mem: 3.36 + 04-04 15:51:24 | [783][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1019 ntime: 0079 mem: 3.36 + 04-04 15:51:31 | [783][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0220 ntime: 0077 mem: 3.36 + 04-04 15:51:38 | [783][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1034 ntime: 0078 mem: 3.36 + 04-04 15:51:44 | [783][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1248 ntime: 0082 mem: 3.36 + 04-04 15:51:51 | [783][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0275 ntime: 0076 mem: 3.36 + 04-04 15:51:57 | [783][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1159 ntime: 0073 mem: 3.36 + 04-04 15:52:04 | [783][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1453 ntime: 0086 mem: 3.36 + 04-04 15:52:12 | [783][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0825 ntime: 0079 mem: 3.36 + 04-04 15:52:18 | [783][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0066 ntime: 0070 mem: 3.36 + 04-04 15:52:25 | [783][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0701 ntime: 0082 mem: 3.36 + 04-04 15:52:32 | [783][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0110 ntime: 0076 mem: 3.36 + 04-04 15:52:38 | Time info >>>> elapsed: 1109.27 mins remain: 305.61 mins + 04-04 15:52:40 | [784][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1000 ntime: 0080 mem: 3.36 + 04-04 15:52:46 | [784][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1089 ntime: 0080 mem: 3.36 + 04-04 15:52:52 | [784][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0047 ntime: 0079 mem: 3.36 + 04-04 15:52:59 | [784][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0081 ntime: 0078 mem: 3.36 + 04-04 15:53:04 | [784][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0046 ntime: 0080 mem: 3.36 + 04-04 15:53:11 | [784][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 15:53:17 | [784][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1469 ntime: 0083 mem: 3.36 + 04-04 15:53:23 | [784][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0768 ntime: 0079 mem: 3.36 + 04-04 15:53:31 | [784][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0156 ntime: 0079 mem: 3.36 + 04-04 15:53:39 | [784][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0144 ntime: 0081 mem: 3.36 + 04-04 15:53:44 | [784][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0080 mem: 3.36 + 04-04 15:53:52 | [784][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0083 ntime: 0080 mem: 3.36 + 04-04 15:53:57 | [784][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0935 ntime: 0083 mem: 3.36 + 04-04 15:54:03 | [784][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0973 ntime: 0077 mem: 3.36 + 04-04 15:54:10 | [784][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0062 ntime: 0078 mem: 3.36 + 04-04 15:54:15 | [784][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0354 ntime: 0075 mem: 3.36 + 04-04 15:54:22 | [784][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1147 ntime: 0087 mem: 3.36 + 04-04 15:54:30 | [784][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0823 ntime: 0077 mem: 3.36 + 04-04 15:54:34 | Time info >>>> elapsed: 1111.19 mins remain: 304.34 mins + 04-04 15:54:34 | [785][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0397 ntime: 0077 mem: 3.36 + 04-04 15:54:42 | [785][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 2039 ntime: 0082 mem: 3.36 + 04-04 15:54:49 | [785][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1205 ntime: 0085 mem: 3.36 + 04-04 15:54:55 | [785][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1118 ntime: 0080 mem: 3.36 + 04-04 15:55:01 | [785][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0077 mem: 3.36 + 04-04 15:55:05 | [785][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0082 mem: 3.36 + 04-04 15:55:12 | [785][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1060 ntime: 0081 mem: 3.36 + 04-04 15:55:19 | [785][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0453 ntime: 0079 mem: 3.36 + 04-04 15:55:24 | [785][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0090 ntime: 0084 mem: 3.36 + 04-04 15:55:30 | [785][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1229 ntime: 0085 mem: 3.36 + 04-04 15:55:36 | [785][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0793 ntime: 0071 mem: 3.36 + 04-04 15:55:42 | [785][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0754 ntime: 0081 mem: 3.36 + 04-04 15:55:49 | [785][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1318 ntime: 0075 mem: 3.36 + 04-04 15:55:55 | [785][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1285 ntime: 0085 mem: 3.36 + 04-04 15:56:01 | [785][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0535 ntime: 0083 mem: 3.36 + 04-04 15:56:06 | [785][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0182 ntime: 0079 mem: 3.36 + 04-04 15:56:14 | [785][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0277 ntime: 0082 mem: 3.36 + 04-04 15:56:20 | [785][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0874 ntime: 0075 mem: 3.36 + 04-04 15:56:25 | Time info >>>> elapsed: 1113.04 mins remain: 303.04 mins + 04-04 15:56:26 | [786][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0733 ntime: 0085 mem: 3.36 + 04-04 15:56:30 | [786][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0080 ntime: 0077 mem: 3.36 + 04-04 15:56:37 | [786][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0866 ntime: 0081 mem: 3.36 + 04-04 15:56:43 | [786][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0615 ntime: 0080 mem: 3.36 + 04-04 15:56:48 | [786][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0175 ntime: 0057 mem: 3.36 + 04-04 15:56:55 | [786][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0634 ntime: 0079 mem: 3.36 + 04-04 15:57:03 | [786][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1128 ntime: 0083 mem: 3.36 + 04-04 15:57:10 | [786][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0883 ntime: 0071 mem: 3.36 + 04-04 15:57:14 | [786][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0093 ntime: 0076 mem: 3.36 + 04-04 15:57:21 | [786][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1177 ntime: 0080 mem: 3.36 + 04-04 15:57:27 | [786][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0553 ntime: 0082 mem: 3.36 + 04-04 15:57:32 | [786][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0085 mem: 3.36 + 04-04 15:57:40 | [786][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0093 ntime: 0091 mem: 3.36 + 04-04 15:57:47 | [786][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 15:57:54 | [786][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1325 ntime: 0087 mem: 3.36 + 04-04 15:58:01 | [786][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0968 ntime: 0077 mem: 3.36 + 04-04 15:58:04 | [786][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0081 ntime: 0080 mem: 3.36 + 04-04 15:58:11 | [786][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0080 ntime: 0079 mem: 3.36 + 04-04 15:58:17 | Time info >>>> elapsed: 1114.91 mins remain: 301.75 mins + 04-04 15:58:17 | [787][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0074 ntime: 0079 mem: 3.36 + 04-04 15:58:24 | [787][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0827 ntime: 0083 mem: 3.36 + 04-04 15:58:30 | [787][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0754 ntime: 0076 mem: 3.36 + 04-04 15:58:36 | [787][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1345 ntime: 0076 mem: 3.36 + 04-04 15:58:45 | [787][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1750 ntime: 0077 mem: 3.36 + 04-04 15:58:54 | [787][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0412 ntime: 0074 mem: 3.36 + 04-04 15:59:00 | [787][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 15:59:06 | [787][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0580 ntime: 0073 mem: 3.36 + 04-04 15:59:12 | [787][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0768 ntime: 0082 mem: 3.36 + 04-04 15:59:19 | [787][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0072 mem: 3.36 + 04-04 15:59:24 | [787][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0984 ntime: 0078 mem: 3.36 + 04-04 15:59:33 | [787][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0640 ntime: 0079 mem: 3.36 + 04-04 15:59:39 | [787][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0867 ntime: 0080 mem: 3.36 + 04-04 15:59:44 | [787][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1034 ntime: 0080 mem: 3.36 + 04-04 15:59:50 | [787][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0209 ntime: 0076 mem: 3.36 + 04-04 15:59:55 | [787][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0904 ntime: 0058 mem: 3.36 + 04-04 16:00:00 | [787][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0085 mem: 3.36 + 04-04 16:00:07 | [787][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0164 ntime: 0072 mem: 3.36 + 04-04 16:00:13 | Time info >>>> elapsed: 1116.84 mins remain: 300.47 mins + 04-04 16:00:13 | [788][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0849 ntime: 0077 mem: 3.36 + 04-04 16:00:18 | [788][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0074 ntime: 0084 mem: 3.36 + 04-04 16:00:24 | [788][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0099 ntime: 0078 mem: 3.36 + 04-04 16:00:31 | [788][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0662 ntime: 0080 mem: 3.36 + 04-04 16:00:36 | [788][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0676 ntime: 0084 mem: 3.36 + 04-04 16:00:43 | [788][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1398 ntime: 0079 mem: 3.36 + 04-04 16:00:50 | [788][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1018 ntime: 0080 mem: 3.36 + 04-04 16:00:54 | [788][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0115 ntime: 0081 mem: 3.36 + 04-04 16:01:02 | [788][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1079 ntime: 0086 mem: 3.36 + 04-04 16:01:07 | [788][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0052 ntime: 0080 mem: 3.36 + 04-04 16:01:13 | [788][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 16:01:19 | [788][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0642 ntime: 0079 mem: 3.36 + 04-04 16:01:24 | [788][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0046 ntime: 0084 mem: 3.36 + 04-04 16:01:31 | [788][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1345 ntime: 0081 mem: 3.36 + 04-04 16:01:36 | [788][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0076 mem: 3.36 + 04-04 16:01:42 | [788][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0155 ntime: 0080 mem: 3.36 + 04-04 16:01:49 | [788][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1015 ntime: 0080 mem: 3.36 + 04-04 16:01:54 | [788][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1212 ntime: 0079 mem: 3.36 + 04-04 16:02:01 | Time info >>>> elapsed: 1118.64 mins remain: 299.15 mins + 04-04 16:02:01 | [789][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0124 ntime: 0077 mem: 3.36 + 04-04 16:02:07 | [789][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0650 ntime: 0077 mem: 3.36 + 04-04 16:02:13 | [789][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0272 ntime: 0078 mem: 3.36 + 04-04 16:02:21 | [789][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0105 ntime: 0083 mem: 3.36 + 04-04 16:02:28 | [789][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1164 ntime: 0082 mem: 3.36 + 04-04 16:02:35 | [789][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1410 ntime: 0073 mem: 3.36 + 04-04 16:02:41 | [789][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0901 ntime: 0077 mem: 3.36 + 04-04 16:02:45 | [789][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0192 ntime: 0083 mem: 3.36 + 04-04 16:02:52 | [789][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0980 ntime: 0082 mem: 3.36 + 04-04 16:02:58 | [789][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0635 ntime: 0087 mem: 3.36 + 04-04 16:03:04 | [789][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1182 ntime: 0082 mem: 3.36 + 04-04 16:03:11 | [789][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0322 ntime: 0071 mem: 3.36 + 04-04 16:03:18 | [789][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0080 mem: 3.36 + 04-04 16:03:24 | [789][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0082 mem: 3.36 + 04-04 16:03:31 | [789][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0173 ntime: 0073 mem: 3.36 + 04-04 16:03:38 | [789][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0145 ntime: 0077 mem: 3.36 + 04-04 16:03:44 | [789][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1046 ntime: 0078 mem: 3.36 + 04-04 16:03:52 | [789][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0514 ntime: 0078 mem: 3.36 + 04-04 16:03:55 | Time info >>>> elapsed: 1120.54 mins remain: 297.86 mins + 04-04 16:03:55 | [790][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0134 ntime: 0079 mem: 3.36 + 04-04 16:04:02 | [790][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1399 ntime: 0074 mem: 3.36 + 04-04 16:04:09 | [790][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0053 ntime: 0080 mem: 3.36 + 04-04 16:04:15 | [790][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0070 ntime: 0077 mem: 3.36 + 04-04 16:04:21 | [790][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0779 ntime: 0079 mem: 3.36 + 04-04 16:04:26 | [790][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0196 ntime: 0083 mem: 3.36 + 04-04 16:04:30 | [790][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0095 ntime: 0078 mem: 3.36 + 04-04 16:04:37 | [790][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1266 ntime: 0081 mem: 3.36 + 04-04 16:04:43 | [790][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1267 ntime: 0087 mem: 3.36 + 04-04 16:04:48 | [790][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0080 mem: 3.36 + 04-04 16:04:55 | [790][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0594 ntime: 0085 mem: 3.36 + 04-04 16:05:02 | [790][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0091 ntime: 0083 mem: 3.36 + 04-04 16:05:08 | [790][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0932 ntime: 0075 mem: 3.36 + 04-04 16:05:14 | [790][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1077 ntime: 0086 mem: 3.36 + 04-04 16:05:20 | [790][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1859 ntime: 0089 mem: 3.36 + 04-04 16:05:25 | [790][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1495 ntime: 0084 mem: 3.36 + 04-04 16:05:32 | [790][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0717 ntime: 0076 mem: 3.36 + 04-04 16:05:37 | [790][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0155 ntime: 0079 mem: 3.36 + 04-04 16:05:43 | Time info >>>> elapsed: 1122.35 mins remain: 296.55 mins + 04-04 16:05:44 | [791][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0188 ntime: 0080 mem: 3.36 + 04-04 16:05:48 | [791][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0244 ntime: 0081 mem: 3.36 + 04-04 16:05:55 | [791][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1183 ntime: 0082 mem: 3.36 + 04-04 16:06:07 | [791][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1067 ntime: 0083 mem: 3.36 + 04-04 16:06:11 | [791][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0368 ntime: 0085 mem: 3.36 + 04-04 16:06:16 | [791][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0211 ntime: 0081 mem: 3.36 + 04-04 16:06:24 | [791][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0104 ntime: 0084 mem: 3.36 + 04-04 16:06:30 | [791][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0110 ntime: 0084 mem: 3.36 + 04-04 16:06:37 | [791][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1019 ntime: 0087 mem: 3.36 + 04-04 16:06:41 | [791][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0690 ntime: 0080 mem: 3.36 + 04-04 16:06:46 | [791][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0811 ntime: 0086 mem: 3.36 + 04-04 16:06:53 | [791][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0795 ntime: 0081 mem: 3.36 + 04-04 16:06:59 | [791][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1110 ntime: 0079 mem: 3.36 + 04-04 16:07:04 | [791][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0078 mem: 3.36 + 04-04 16:07:08 | [791][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0088 mem: 3.36 + 04-04 16:07:15 | [791][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0956 ntime: 0081 mem: 3.36 + 04-04 16:07:19 | [791][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0819 ntime: 0082 mem: 3.36 + 04-04 16:07:25 | [791][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0250 ntime: 0088 mem: 3.36 + 04-04 16:07:29 | Time info >>>> elapsed: 1124.11 mins remain: 295.22 mins + 04-04 16:07:29 | [792][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0088 ntime: 0085 mem: 3.36 + 04-04 16:07:36 | [792][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0049 ntime: 0080 mem: 3.36 + 04-04 16:07:45 | [792][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0096 ntime: 0089 mem: 3.36 + 04-04 16:07:51 | [792][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1370 ntime: 0079 mem: 3.36 + 04-04 16:07:57 | [792][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0137 ntime: 0078 mem: 3.36 + 04-04 16:08:03 | [792][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1205 ntime: 0082 mem: 3.36 + 04-04 16:08:11 | [792][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1092 ntime: 0088 mem: 3.36 + 04-04 16:08:19 | [792][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1647 ntime: 0078 mem: 3.36 + 04-04 16:08:25 | [792][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0147 ntime: 0081 mem: 3.36 + 04-04 16:08:33 | [792][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0608 ntime: 0086 mem: 3.36 + 04-04 16:08:38 | [792][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0462 ntime: 0077 mem: 3.36 + 04-04 16:08:45 | [792][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0130 ntime: 0079 mem: 3.36 + 04-04 16:08:53 | [792][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0723 ntime: 0079 mem: 3.36 + 04-04 16:09:00 | [792][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1400 ntime: 0085 mem: 3.36 + 04-04 16:09:08 | [792][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0138 ntime: 0086 mem: 3.36 + 04-04 16:09:11 | [792][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0620 ntime: 0084 mem: 3.36 + 04-04 16:09:17 | [792][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1251 ntime: 0079 mem: 3.36 + 04-04 16:09:24 | [792][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0080 mem: 3.36 + 04-04 16:09:29 | Time info >>>> elapsed: 1126.11 mins remain: 293.95 mins + 04-04 16:09:30 | [793][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1130 ntime: 0079 mem: 3.36 + 04-04 16:09:35 | [793][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0050 ntime: 0086 mem: 3.36 + 04-04 16:09:40 | [793][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0163 ntime: 0078 mem: 3.36 + 04-04 16:09:46 | [793][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1080 ntime: 0085 mem: 3.36 + 04-04 16:09:52 | [793][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1159 ntime: 0072 mem: 3.36 + 04-04 16:10:00 | [793][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0067 ntime: 0080 mem: 3.36 + 04-04 16:10:05 | [793][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0237 ntime: 0084 mem: 3.36 + 04-04 16:10:11 | [793][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0254 ntime: 0077 mem: 3.36 + 04-04 16:10:16 | [793][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0158 ntime: 0078 mem: 3.36 + 04-04 16:10:22 | [793][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0090 ntime: 0076 mem: 3.36 + 04-04 16:10:28 | [793][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0613 ntime: 0081 mem: 3.36 + 04-04 16:10:34 | [793][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0664 ntime: 0079 mem: 3.36 + 04-04 16:10:40 | [793][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0091 ntime: 0083 mem: 3.36 + 04-04 16:10:48 | [793][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0839 ntime: 0079 mem: 3.36 + 04-04 16:10:55 | [793][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0094 ntime: 0078 mem: 3.36 + 04-04 16:11:00 | [793][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0445 ntime: 0078 mem: 3.36 + 04-04 16:11:09 | [793][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0781 ntime: 0078 mem: 3.36 + 04-04 16:11:16 | [793][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1167 ntime: 0080 mem: 3.36 + 04-04 16:11:20 | Time info >>>> elapsed: 1127.95 mins remain: 292.64 mins + 04-04 16:11:20 | [794][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0814 ntime: 0081 mem: 3.36 + 04-04 16:11:25 | [794][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1059 ntime: 0083 mem: 3.36 + 04-04 16:11:30 | [794][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0103 ntime: 0074 mem: 3.36 + 04-04 16:11:37 | [794][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0533 ntime: 0082 mem: 3.36 + 04-04 16:11:42 | [794][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0729 ntime: 0085 mem: 3.36 + 04-04 16:11:48 | [794][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1583 ntime: 0092 mem: 3.36 + 04-04 16:11:53 | [794][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0076 mem: 3.36 + 04-04 16:11:59 | [794][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0542 ntime: 0078 mem: 3.36 + 04-04 16:12:05 | [794][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0070 ntime: 0082 mem: 3.36 + 04-04 16:12:12 | [794][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0986 ntime: 0082 mem: 3.36 + 04-04 16:12:18 | [794][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0244 ntime: 0080 mem: 3.36 + 04-04 16:12:23 | [794][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0094 ntime: 0086 mem: 3.36 + 04-04 16:12:29 | [794][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0410 ntime: 0079 mem: 3.36 + 04-04 16:12:34 | [794][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1455 ntime: 0081 mem: 3.36 + 04-04 16:12:43 | [794][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0115 ntime: 0077 mem: 3.36 + 04-04 16:12:52 | [794][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1036 ntime: 0084 mem: 3.36 + 04-04 16:13:01 | [794][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1456 ntime: 0081 mem: 3.36 + 04-04 16:13:06 | [794][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0086 mem: 3.36 + 04-04 16:13:12 | Time info >>>> elapsed: 1129.82 mins remain: 291.34 mins + 04-04 16:13:13 | [795][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0918 ntime: 0081 mem: 3.36 + 04-04 16:13:19 | [795][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0091 ntime: 0077 mem: 3.36 + 04-04 16:13:25 | [795][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0072 ntime: 0084 mem: 3.36 + 04-04 16:13:32 | [795][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0045 ntime: 0078 mem: 3.36 + 04-04 16:13:39 | [795][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1451 ntime: 0084 mem: 3.36 + 04-04 16:13:45 | [795][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0091 ntime: 0089 mem: 3.36 + 04-04 16:13:52 | [795][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0598 ntime: 0081 mem: 3.36 + 04-04 16:13:57 | [795][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0426 ntime: 0086 mem: 3.36 + 04-04 16:14:05 | [795][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1317 ntime: 0080 mem: 3.36 + 04-04 16:14:11 | [795][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0464 ntime: 0079 mem: 3.36 + 04-04 16:14:17 | [795][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0129 ntime: 0076 mem: 3.36 + 04-04 16:14:23 | [795][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1178 ntime: 0079 mem: 3.36 + 04-04 16:14:29 | [795][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0087 ntime: 0085 mem: 3.36 + 04-04 16:14:40 | [795][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0103 ntime: 0054 mem: 3.36 + 04-04 16:14:46 | [795][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0529 ntime: 0080 mem: 3.36 + 04-04 16:14:51 | [795][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0064 ntime: 0085 mem: 3.36 + 04-04 16:14:57 | [795][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0101 ntime: 0077 mem: 3.36 + 04-04 16:15:03 | [795][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0781 ntime: 0081 mem: 3.36 + 04-04 16:15:06 | Time info >>>> elapsed: 1131.73 mins remain: 290.04 mins + 04-04 16:15:07 | [796][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1339 ntime: 0082 mem: 3.36 + 04-04 16:15:14 | [796][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0805 ntime: 0079 mem: 3.36 + 04-04 16:15:20 | [796][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0618 ntime: 0083 mem: 3.36 + 04-04 16:15:27 | [796][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0077 mem: 3.36 + 04-04 16:15:33 | [796][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0654 ntime: 0074 mem: 3.36 + 04-04 16:15:40 | [796][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0074 ntime: 0081 mem: 3.36 + 04-04 16:15:44 | [796][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0076 mem: 3.36 + 04-04 16:15:51 | [796][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0055 ntime: 0081 mem: 3.36 + 04-04 16:15:58 | [796][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0075 ntime: 0081 mem: 3.36 + 04-04 16:16:04 | [796][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0065 ntime: 0081 mem: 3.36 + 04-04 16:16:10 | [796][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0342 ntime: 0079 mem: 3.36 + 04-04 16:16:15 | [796][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0685 ntime: 0089 mem: 3.36 + 04-04 16:16:23 | [796][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0955 ntime: 0086 mem: 3.36 + 04-04 16:16:28 | [796][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0096 ntime: 0087 mem: 3.36 + 04-04 16:16:35 | [796][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0635 ntime: 0087 mem: 3.36 + 04-04 16:16:43 | [796][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1002 ntime: 0072 mem: 3.36 + 04-04 16:16:49 | [796][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0932 ntime: 0086 mem: 3.36 + 04-04 16:16:57 | [796][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1084 ntime: 0086 mem: 3.36 + 04-04 16:17:02 | Time info >>>> elapsed: 1133.67 mins remain: 288.75 mins + 04-04 16:17:03 | [797][000/179] predict_x0_loss: 0.009 glr: 5.0e-08 dtime: 0187 ntime: 0082 mem: 3.36 + 04-04 16:17:09 | [797][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0082 mem: 3.36 + 04-04 16:17:15 | [797][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0743 ntime: 0085 mem: 3.36 + 04-04 16:17:22 | [797][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0646 ntime: 0084 mem: 3.36 + 04-04 16:17:28 | [797][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0085 mem: 3.36 + 04-04 16:17:36 | [797][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0089 ntime: 0086 mem: 3.36 + 04-04 16:17:44 | [797][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0112 ntime: 0074 mem: 3.36 + 04-04 16:17:51 | [797][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0078 mem: 3.36 + 04-04 16:17:57 | [797][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0073 ntime: 0077 mem: 3.36 + 04-04 16:18:03 | [797][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0116 ntime: 0078 mem: 3.36 + 04-04 16:18:09 | [797][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1546 ntime: 0083 mem: 3.36 + 04-04 16:18:14 | [797][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0057 ntime: 0076 mem: 3.36 + 04-04 16:18:21 | [797][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0107 ntime: 0080 mem: 3.36 + 04-04 16:18:25 | [797][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0156 ntime: 0077 mem: 3.36 + 04-04 16:18:32 | [797][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0666 ntime: 0078 mem: 3.36 + 04-04 16:18:38 | [797][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0515 ntime: 0087 mem: 3.36 + 04-04 16:18:45 | [797][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0245 ntime: 0083 mem: 3.36 + 04-04 16:18:52 | [797][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1023 ntime: 0088 mem: 3.36 + 04-04 16:18:57 | Time info >>>> elapsed: 1135.57 mins remain: 287.45 mins + 04-04 16:18:57 | [798][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0080 mem: 3.36 + 04-04 16:19:03 | [798][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0072 mem: 3.36 + 04-04 16:19:10 | [798][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0058 ntime: 0079 mem: 3.36 + 04-04 16:19:15 | [798][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0138 ntime: 0082 mem: 3.36 + 04-04 16:19:21 | [798][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1494 ntime: 0079 mem: 3.36 + 04-04 16:19:27 | [798][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1245 ntime: 0084 mem: 3.36 + 04-04 16:19:33 | [798][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0137 ntime: 0080 mem: 3.36 + 04-04 16:19:38 | [798][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0924 ntime: 0078 mem: 3.36 + 04-04 16:19:45 | [798][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0067 ntime: 0081 mem: 3.36 + 04-04 16:19:52 | [798][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0870 ntime: 0082 mem: 3.36 + 04-04 16:19:57 | [798][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0101 ntime: 0078 mem: 3.36 + 04-04 16:20:03 | [798][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0838 ntime: 0086 mem: 3.36 + 04-04 16:20:09 | [798][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1061 ntime: 0082 mem: 3.36 + 04-04 16:20:16 | [798][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0475 ntime: 0086 mem: 3.36 + 04-04 16:20:21 | [798][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0056 ntime: 0086 mem: 3.36 + 04-04 16:20:28 | [798][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1111 ntime: 0083 mem: 3.36 + 04-04 16:20:34 | [798][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0060 ntime: 0076 mem: 3.36 + 04-04 16:20:40 | [798][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0992 ntime: 0072 mem: 3.36 + 04-04 16:20:45 | Time info >>>> elapsed: 1137.38 mins remain: 286.12 mins + 04-04 16:20:47 | [799][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1621 ntime: 0074 mem: 3.36 + 04-04 16:20:56 | [799][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0987 ntime: 0079 mem: 3.36 + 04-04 16:21:00 | [799][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1169 ntime: 0069 mem: 3.36 + 04-04 16:21:06 | [799][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0735 ntime: 0077 mem: 3.36 + 04-04 16:21:13 | [799][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0063 ntime: 0080 mem: 3.36 + 04-04 16:21:21 | [799][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0440 ntime: 0083 mem: 3.36 + 04-04 16:21:27 | [799][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1078 ntime: 0080 mem: 3.36 + 04-04 16:21:32 | [799][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0074 ntime: 0080 mem: 3.36 + 04-04 16:21:39 | [799][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0802 ntime: 0075 mem: 3.36 + 04-04 16:21:46 | [799][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0059 ntime: 0078 mem: 3.36 + 04-04 16:21:53 | [799][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0385 ntime: 0081 mem: 3.36 + 04-04 16:22:01 | [799][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1080 ntime: 0078 mem: 3.36 + 04-04 16:22:07 | [799][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0048 ntime: 0075 mem: 3.36 + 04-04 16:22:15 | [799][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0622 ntime: 0076 mem: 3.36 + 04-04 16:22:21 | [799][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0072 ntime: 0085 mem: 3.36 + 04-04 16:22:29 | [799][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1075 ntime: 0078 mem: 3.36 + 04-04 16:22:35 | [799][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0105 ntime: 0092 mem: 3.36 + 04-04 16:22:42 | [799][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1095 ntime: 0083 mem: 3.36 + 04-04 16:22:46 | Time info >>>> elapsed: 1139.40 mins remain: 284.85 mins + 04-04 16:22:48 | [800][000/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1217 ntime: 0086 mem: 3.36 + 04-04 16:22:54 | [800][010/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0092 ntime: 0084 mem: 3.36 + 04-04 16:23:04 | [800][020/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1630 ntime: 0081 mem: 3.36 + 04-04 16:23:08 | [800][030/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0076 ntime: 0073 mem: 3.36 + 04-04 16:23:14 | [800][040/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0083 ntime: 0075 mem: 3.36 + 04-04 16:23:21 | [800][050/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1436 ntime: 0082 mem: 3.36 + 04-04 16:23:27 | [800][060/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0061 ntime: 0078 mem: 3.36 + 04-04 16:23:36 | [800][070/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1470 ntime: 0058 mem: 3.36 + 04-04 16:23:42 | [800][080/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0824 ntime: 0085 mem: 3.36 + 04-04 16:23:48 | [800][090/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0212 ntime: 0076 mem: 3.36 + 04-04 16:23:56 | [800][100/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0046 ntime: 0081 mem: 3.36 + 04-04 16:24:03 | [800][110/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0079 ntime: 0088 mem: 3.36 + 04-04 16:24:11 | [800][120/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0069 ntime: 0084 mem: 3.36 + 04-04 16:24:15 | [800][130/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0078 ntime: 0077 mem: 3.36 + 04-04 16:24:22 | [800][140/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1222 ntime: 0082 mem: 3.36 + 04-04 16:24:28 | [800][150/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 1124 ntime: 0087 mem: 3.36 + 04-04 16:24:33 | [800][160/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0042 ntime: 0083 mem: 3.36 + 04-04 16:24:40 | [800][170/179] predict_x0_loss: 0.008 glr: 5.0e-08 dtime: 0062 ntime: 0076 mem: 3.36 + 04-04 16:24:46 | Time info >>>> elapsed: 1141.39 mins remain: 283.57 mins + 04-04 16:24:47 | [801][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0464 ntime: 0076 mem: 3.36 + 04-04 16:24:51 | [801][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0961 ntime: 0083 mem: 3.36 + 04-04 16:24:58 | [801][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0143 ntime: 0077 mem: 3.36 + 04-04 16:25:04 | [801][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 16:25:11 | [801][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1566 ntime: 0078 mem: 3.36 + 04-04 16:25:19 | [801][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1220 ntime: 0083 mem: 3.36 + 04-04 16:25:24 | [801][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0178 ntime: 0079 mem: 3.36 + 04-04 16:25:28 | [801][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0433 ntime: 0078 mem: 3.36 + 04-04 16:25:35 | [801][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0078 mem: 3.36 + 04-04 16:25:43 | [801][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0956 ntime: 0070 mem: 3.36 + 04-04 16:25:51 | [801][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0079 mem: 3.36 + 04-04 16:25:58 | [801][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0098 ntime: 0076 mem: 3.36 + 04-04 16:26:03 | [801][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0659 ntime: 0075 mem: 3.36 + 04-04 16:26:10 | [801][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1573 ntime: 0079 mem: 3.36 + 04-04 16:26:18 | [801][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1379 ntime: 0078 mem: 3.36 + 04-04 16:26:28 | [801][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1370 ntime: 0091 mem: 3.36 + 04-04 16:26:33 | [801][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0110 ntime: 0088 mem: 3.36 + 04-04 16:26:38 | [801][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0082 mem: 3.36 + 04-04 16:26:46 | Time info >>>> elapsed: 1143.39 mins remain: 282.28 mins + 04-04 16:26:46 | [802][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0078 mem: 3.36 + 04-04 16:26:51 | [802][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1345 ntime: 0080 mem: 3.36 + 04-04 16:26:58 | [802][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 16:27:03 | [802][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0929 ntime: 0084 mem: 3.36 + 04-04 16:27:09 | [802][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0074 mem: 3.36 + 04-04 16:27:14 | [802][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0132 ntime: 0086 mem: 3.36 + 04-04 16:27:18 | [802][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0080 mem: 3.36 + 04-04 16:27:24 | [802][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0708 ntime: 0091 mem: 3.36 + 04-04 16:27:28 | [802][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0088 mem: 3.36 + 04-04 16:27:36 | [802][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1413 ntime: 0087 mem: 3.36 + 04-04 16:27:42 | [802][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0380 ntime: 0076 mem: 3.36 + 04-04 16:27:48 | [802][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0468 ntime: 0081 mem: 3.36 + 04-04 16:27:54 | [802][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0895 ntime: 0080 mem: 3.36 + 04-04 16:27:58 | [802][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0085 mem: 3.36 + 04-04 16:28:02 | [802][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0087 ntime: 0080 mem: 3.36 + 04-04 16:28:09 | [802][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0306 ntime: 0085 mem: 3.36 + 04-04 16:28:15 | [802][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1059 ntime: 0083 mem: 3.36 + 04-04 16:28:18 | [802][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0112 ntime: 0080 mem: 3.36 + 04-04 16:28:23 | Time info >>>> elapsed: 1145.00 mins remain: 280.90 mins + 04-04 16:28:23 | [803][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0583 ntime: 0081 mem: 3.36 + 04-04 16:28:29 | [803][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0755 ntime: 0078 mem: 3.36 + 04-04 16:28:35 | [803][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0106 ntime: 0083 mem: 3.36 + 04-04 16:28:41 | [803][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1413 ntime: 0079 mem: 3.36 + 04-04 16:28:48 | [803][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1131 ntime: 0083 mem: 3.36 + 04-04 16:28:52 | [803][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0906 ntime: 0081 mem: 3.36 + 04-04 16:28:57 | [803][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0182 ntime: 0079 mem: 3.36 + 04-04 16:29:04 | [803][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0718 ntime: 0077 mem: 3.36 + 04-04 16:29:09 | [803][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0230 ntime: 0084 mem: 3.36 + 04-04 16:29:15 | [803][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0732 ntime: 0079 mem: 3.36 + 04-04 16:29:23 | [803][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1359 ntime: 0081 mem: 3.36 + 04-04 16:29:28 | [803][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0264 ntime: 0058 mem: 3.36 + 04-04 16:29:34 | [803][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0747 ntime: 0057 mem: 3.36 + 04-04 16:29:38 | [803][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0099 ntime: 0077 mem: 3.36 + 04-04 16:29:43 | [803][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0799 ntime: 0070 mem: 3.36 + 04-04 16:29:47 | [803][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0172 ntime: 0079 mem: 3.36 + 04-04 16:29:53 | [803][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0088 ntime: 0078 mem: 3.36 + 04-04 16:30:00 | [803][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0194 ntime: 0083 mem: 3.36 + 04-04 16:30:07 | Time info >>>> elapsed: 1146.74 mins remain: 279.55 mins + 04-04 16:30:07 | [804][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0061 ntime: 0090 mem: 3.36 + 04-04 16:30:11 | [804][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0068 ntime: 0077 mem: 3.36 + 04-04 16:30:17 | [804][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0357 ntime: 0080 mem: 3.36 + 04-04 16:30:22 | [804][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0096 ntime: 0079 mem: 3.36 + 04-04 16:30:27 | [804][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0080 mem: 3.36 + 04-04 16:30:32 | [804][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0072 ntime: 0075 mem: 3.36 + 04-04 16:30:39 | [804][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0080 mem: 3.36 + 04-04 16:30:45 | [804][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0955 ntime: 0079 mem: 3.36 + 04-04 16:30:49 | [804][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0083 ntime: 0077 mem: 3.36 + 04-04 16:30:57 | [804][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0830 ntime: 0082 mem: 3.36 + 04-04 16:31:05 | [804][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1575 ntime: 0085 mem: 3.36 + 04-04 16:31:10 | [804][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0080 mem: 3.36 + 04-04 16:31:18 | [804][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0208 ntime: 0085 mem: 3.36 + 04-04 16:31:26 | [804][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0193 ntime: 0081 mem: 3.36 + 04-04 16:31:32 | [804][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0917 ntime: 0079 mem: 3.36 + 04-04 16:31:38 | [804][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1151 ntime: 0085 mem: 3.36 + 04-04 16:31:43 | [804][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1398 ntime: 0077 mem: 3.36 + 04-04 16:31:47 | [804][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0085 mem: 3.36 + 04-04 16:31:54 | Time info >>>> elapsed: 1148.52 mins remain: 278.21 mins + 04-04 16:31:54 | [805][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0141 ntime: 0072 mem: 3.36 + 04-04 16:31:59 | [805][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0450 ntime: 0078 mem: 3.36 + 04-04 16:32:04 | [805][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0505 ntime: 0086 mem: 3.36 + 04-04 16:32:09 | [805][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0078 mem: 3.36 + 04-04 16:32:14 | [805][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0099 ntime: 0080 mem: 3.36 + 04-04 16:32:19 | [805][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0071 mem: 3.36 + 04-04 16:32:25 | [805][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0084 mem: 3.36 + 04-04 16:32:30 | [805][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0206 ntime: 0084 mem: 3.36 + 04-04 16:32:36 | [805][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0059 mem: 3.36 + 04-04 16:32:42 | [805][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2111 ntime: 0079 mem: 3.36 + 04-04 16:32:48 | [805][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0063 ntime: 0076 mem: 3.36 + 04-04 16:32:55 | [805][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1078 ntime: 0079 mem: 3.36 + 04-04 16:33:04 | [805][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1299 ntime: 0086 mem: 3.36 + 04-04 16:33:11 | [805][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0862 ntime: 0078 mem: 3.36 + 04-04 16:33:18 | [805][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1123 ntime: 0081 mem: 3.36 + 04-04 16:33:24 | [805][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0086 mem: 3.36 + 04-04 16:33:32 | [805][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0205 ntime: 0076 mem: 3.36 + 04-04 16:33:39 | [805][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1140 ntime: 0080 mem: 3.36 + 04-04 16:33:43 | Time info >>>> elapsed: 1150.35 mins remain: 276.88 mins + 04-04 16:33:44 | [806][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0075 ntime: 0076 mem: 3.36 + 04-04 16:33:51 | [806][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0469 ntime: 0084 mem: 3.36 + 04-04 16:33:57 | [806][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0136 ntime: 0082 mem: 3.36 + 04-04 16:34:03 | [806][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0923 ntime: 0080 mem: 3.36 + 04-04 16:34:11 | [806][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0713 ntime: 0076 mem: 3.36 + 04-04 16:34:18 | [806][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1228 ntime: 0073 mem: 3.36 + 04-04 16:34:24 | [806][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0068 ntime: 0081 mem: 3.36 + 04-04 16:34:30 | [806][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0332 ntime: 0082 mem: 3.36 + 04-04 16:34:37 | [806][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0678 ntime: 0080 mem: 3.36 + 04-04 16:34:44 | [806][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0225 ntime: 0082 mem: 3.36 + 04-04 16:34:49 | [806][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0167 ntime: 0077 mem: 3.36 + 04-04 16:34:54 | [806][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0138 ntime: 0078 mem: 3.36 + 04-04 16:35:01 | [806][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1022 ntime: 0081 mem: 3.36 + 04-04 16:35:07 | [806][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0072 ntime: 0088 mem: 3.36 + 04-04 16:35:13 | [806][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0658 ntime: 0076 mem: 3.36 + 04-04 16:35:19 | [806][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0336 ntime: 0083 mem: 3.36 + 04-04 16:35:25 | [806][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0400 ntime: 0081 mem: 3.36 + 04-04 16:35:30 | [806][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0931 ntime: 0078 mem: 3.36 + 04-04 16:35:37 | Time info >>>> elapsed: 1152.24 mins remain: 275.57 mins + 04-04 16:35:38 | [807][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0896 ntime: 0071 mem: 3.36 + 04-04 16:35:44 | [807][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0789 ntime: 0076 mem: 3.36 + 04-04 16:35:51 | [807][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 16:35:59 | [807][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0086 mem: 3.36 + 04-04 16:36:05 | [807][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0190 ntime: 0078 mem: 3.36 + 04-04 16:36:11 | [807][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0183 ntime: 0085 mem: 3.36 + 04-04 16:36:18 | [807][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1546 ntime: 0084 mem: 3.36 + 04-04 16:36:23 | [807][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0781 ntime: 0083 mem: 3.36 + 04-04 16:36:29 | [807][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0490 ntime: 0078 mem: 3.36 + 04-04 16:36:34 | [807][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0084 mem: 3.36 + 04-04 16:36:41 | [807][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0780 ntime: 0076 mem: 3.36 + 04-04 16:36:49 | [807][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1299 ntime: 0080 mem: 3.36 + 04-04 16:36:54 | [807][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0111 ntime: 0078 mem: 3.36 + 04-04 16:37:02 | [807][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0839 ntime: 0080 mem: 3.36 + 04-04 16:37:10 | [807][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0966 ntime: 0078 mem: 3.36 + 04-04 16:37:14 | [807][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0094 ntime: 0081 mem: 3.36 + 04-04 16:37:23 | [807][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1435 ntime: 0080 mem: 3.36 + 04-04 16:37:29 | [807][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0085 mem: 3.36 + 04-04 16:37:33 | Time info >>>> elapsed: 1154.18 mins remain: 274.26 mins + 04-04 16:37:34 | [808][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0485 ntime: 0078 mem: 3.36 + 04-04 16:37:39 | [808][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0102 ntime: 0080 mem: 3.36 + 04-04 16:37:45 | [808][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0706 ntime: 0082 mem: 3.36 + 04-04 16:37:51 | [808][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0154 ntime: 0075 mem: 3.36 + 04-04 16:37:58 | [808][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0355 ntime: 0084 mem: 3.36 + 04-04 16:38:05 | [808][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0079 mem: 3.36 + 04-04 16:38:11 | [808][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0081 mem: 3.36 + 04-04 16:38:17 | [808][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0134 ntime: 0084 mem: 3.36 + 04-04 16:38:25 | [808][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0950 ntime: 0082 mem: 3.36 + 04-04 16:38:31 | [808][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0065 ntime: 0079 mem: 3.36 + 04-04 16:38:37 | [808][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1097 ntime: 0075 mem: 3.36 + 04-04 16:38:41 | [808][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0083 mem: 3.36 + 04-04 16:38:51 | [808][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0853 ntime: 0080 mem: 3.36 + 04-04 16:38:57 | [808][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0080 ntime: 0078 mem: 3.36 + 04-04 16:39:04 | [808][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0139 ntime: 0071 mem: 3.36 + 04-04 16:39:09 | [808][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0768 ntime: 0081 mem: 3.36 + 04-04 16:39:15 | [808][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1115 ntime: 0084 mem: 3.36 + 04-04 16:39:20 | [808][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0076 mem: 3.36 + 04-04 16:39:26 | Time info >>>> elapsed: 1156.05 mins remain: 272.94 mins + 04-04 16:39:26 | [809][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0125 ntime: 0077 mem: 3.36 + 04-04 16:39:31 | [809][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0130 ntime: 0079 mem: 3.36 + 04-04 16:39:37 | [809][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1527 ntime: 0080 mem: 3.36 + 04-04 16:39:43 | [809][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1204 ntime: 0079 mem: 3.36 + 04-04 16:39:49 | [809][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0127 ntime: 0084 mem: 3.36 + 04-04 16:39:55 | [809][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0373 ntime: 0083 mem: 3.36 + 04-04 16:40:02 | [809][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1197 ntime: 0086 mem: 3.36 + 04-04 16:40:08 | [809][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0088 ntime: 0081 mem: 3.36 + 04-04 16:40:14 | [809][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0076 ntime: 0081 mem: 3.36 + 04-04 16:40:20 | [809][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 16:40:27 | [809][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0068 ntime: 0082 mem: 3.36 + 04-04 16:40:34 | [809][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0073 ntime: 0079 mem: 3.36 + 04-04 16:40:40 | [809][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0081 mem: 3.36 + 04-04 16:40:46 | [809][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0477 ntime: 0078 mem: 3.36 + 04-04 16:40:52 | [809][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1546 ntime: 0082 mem: 3.36 + 04-04 16:40:57 | [809][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0863 ntime: 0081 mem: 3.36 + 04-04 16:41:02 | [809][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0076 mem: 3.36 + 04-04 16:41:07 | [809][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0981 ntime: 0083 mem: 3.36 + 04-04 16:41:12 | Time info >>>> elapsed: 1157.82 mins remain: 271.59 mins + 04-04 16:41:12 | [810][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0523 ntime: 0082 mem: 3.36 + 04-04 16:41:19 | [810][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1126 ntime: 0086 mem: 3.36 + 04-04 16:41:23 | [810][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0508 ntime: 0077 mem: 3.36 + 04-04 16:41:30 | [810][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1175 ntime: 0084 mem: 3.36 + 04-04 16:41:37 | [810][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0603 ntime: 0077 mem: 3.36 + 04-04 16:41:41 | [810][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 16:41:46 | [810][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0782 ntime: 0077 mem: 3.36 + 04-04 16:41:52 | [810][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1210 ntime: 0079 mem: 3.36 + 04-04 16:41:58 | [810][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0260 ntime: 0088 mem: 3.36 + 04-04 16:42:03 | [810][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0694 ntime: 0074 mem: 3.36 + 04-04 16:42:08 | [810][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0912 ntime: 0081 mem: 3.36 + 04-04 16:42:14 | [810][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0822 ntime: 0083 mem: 3.36 + 04-04 16:42:19 | [810][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0134 ntime: 0079 mem: 3.36 + 04-04 16:42:26 | [810][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1094 ntime: 0079 mem: 3.36 + 04-04 16:42:33 | [810][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0733 ntime: 0077 mem: 3.36 + 04-04 16:42:38 | [810][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0767 ntime: 0079 mem: 3.36 + 04-04 16:42:43 | [810][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0084 mem: 3.36 + 04-04 16:42:50 | [810][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0078 ntime: 0081 mem: 3.36 + 04-04 16:42:55 | Time info >>>> elapsed: 1159.54 mins remain: 270.23 mins + 04-04 16:42:56 | [811][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1489 ntime: 0080 mem: 3.36 + 04-04 16:43:02 | [811][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1077 ntime: 0083 mem: 3.36 + 04-04 16:43:06 | [811][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0905 ntime: 0079 mem: 3.36 + 04-04 16:43:12 | [811][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0920 ntime: 0087 mem: 3.36 + 04-04 16:43:20 | [811][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1337 ntime: 0079 mem: 3.36 + 04-04 16:43:26 | [811][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0084 mem: 3.36 + 04-04 16:43:33 | [811][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0108 ntime: 0082 mem: 3.36 + 04-04 16:43:38 | [811][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1005 ntime: 0080 mem: 3.36 + 04-04 16:43:43 | [811][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0079 mem: 3.36 + 04-04 16:43:48 | [811][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0590 ntime: 0082 mem: 3.36 + 04-04 16:43:56 | [811][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0790 ntime: 0076 mem: 3.36 + 04-04 16:44:02 | [811][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0995 ntime: 0078 mem: 3.36 + 04-04 16:44:08 | [811][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0559 ntime: 0079 mem: 3.36 + 04-04 16:44:13 | [811][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 16:44:19 | [811][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0089 mem: 3.36 + 04-04 16:44:25 | [811][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0165 ntime: 0067 mem: 3.36 + 04-04 16:44:33 | [811][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0084 mem: 3.36 + 04-04 16:44:38 | [811][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0588 ntime: 0077 mem: 3.36 + 04-04 16:44:42 | Time info >>>> elapsed: 1161.33 mins remain: 268.88 mins + 04-04 16:44:43 | [812][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0076 mem: 3.36 + 04-04 16:44:49 | [812][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 16:44:55 | [812][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1307 ntime: 0079 mem: 3.36 + 04-04 16:45:02 | [812][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0093 ntime: 0083 mem: 3.36 + 04-04 16:45:08 | [812][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 16:45:14 | [812][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0157 ntime: 0081 mem: 3.36 + 04-04 16:45:20 | [812][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0938 ntime: 0077 mem: 3.36 + 04-04 16:45:26 | [812][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0811 ntime: 0079 mem: 3.36 + 04-04 16:45:33 | [812][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1078 ntime: 0086 mem: 3.36 + 04-04 16:45:39 | [812][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0988 ntime: 0086 mem: 3.36 + 04-04 16:45:44 | [812][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0075 ntime: 0079 mem: 3.36 + 04-04 16:45:54 | [812][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1295 ntime: 0077 mem: 3.36 + 04-04 16:46:00 | [812][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0664 ntime: 0082 mem: 3.36 + 04-04 16:46:07 | [812][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1366 ntime: 0079 mem: 3.36 + 04-04 16:46:16 | [812][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1626 ntime: 0081 mem: 3.36 + 04-04 16:46:22 | [812][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1550 ntime: 0083 mem: 3.36 + 04-04 16:46:35 | [812][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1883 ntime: 0078 mem: 3.36 + 04-04 16:46:45 | [812][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0084 mem: 3.36 + 04-04 16:46:51 | Time info >>>> elapsed: 1163.48 mins remain: 267.61 mins + 04-04 16:46:51 | [813][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0189 ntime: 0084 mem: 3.36 + 04-04 16:46:57 | [813][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0780 ntime: 0084 mem: 3.36 + 04-04 16:47:04 | [813][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0114 ntime: 0075 mem: 3.36 + 04-04 16:47:10 | [813][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0086 ntime: 0057 mem: 3.36 + 04-04 16:47:17 | [813][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1069 ntime: 0079 mem: 3.36 + 04-04 16:47:25 | [813][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0556 ntime: 0060 mem: 3.36 + 04-04 16:47:32 | [813][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0720 ntime: 0077 mem: 3.36 + 04-04 16:47:40 | [813][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0746 ntime: 0081 mem: 3.36 + 04-04 16:47:49 | [813][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1041 ntime: 0083 mem: 3.36 + 04-04 16:47:57 | [813][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1658 ntime: 0084 mem: 3.36 + 04-04 16:48:05 | [813][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1189 ntime: 0072 mem: 3.36 + 04-04 16:48:12 | [813][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0127 ntime: 0078 mem: 3.36 + 04-04 16:48:23 | [813][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1310 ntime: 0081 mem: 3.36 + 04-04 16:48:32 | [813][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0087 mem: 3.36 + 04-04 16:48:39 | [813][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0142 ntime: 0073 mem: 3.36 + 04-04 16:48:45 | [813][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0847 ntime: 0074 mem: 3.36 + 04-04 16:48:49 | [813][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0597 ntime: 0088 mem: 3.36 + 04-04 16:48:56 | [813][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0417 ntime: 0078 mem: 3.36 + 04-04 16:49:01 | Time info >>>> elapsed: 1165.64 mins remain: 266.35 mins + 04-04 16:49:02 | [814][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1493 ntime: 0084 mem: 3.36 + 04-04 16:49:09 | [814][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0321 ntime: 0081 mem: 3.36 + 04-04 16:49:15 | [814][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1639 ntime: 0086 mem: 3.36 + 04-04 16:49:24 | [814][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1029 ntime: 0087 mem: 3.36 + 04-04 16:49:29 | [814][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0113 ntime: 0081 mem: 3.36 + 04-04 16:49:36 | [814][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0811 ntime: 0082 mem: 3.36 + 04-04 16:49:43 | [814][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0104 ntime: 0078 mem: 3.36 + 04-04 16:49:50 | [814][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0116 ntime: 0074 mem: 3.36 + 04-04 16:50:00 | [814][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1113 ntime: 0075 mem: 3.36 + 04-04 16:50:07 | [814][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 16:50:15 | [814][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0113 ntime: 0078 mem: 3.36 + 04-04 16:50:20 | [814][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0726 ntime: 0080 mem: 3.36 + 04-04 16:50:26 | [814][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1138 ntime: 0079 mem: 3.36 + 04-04 16:50:33 | [814][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1289 ntime: 0081 mem: 3.36 + 04-04 16:50:41 | [814][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0540 ntime: 0081 mem: 3.36 + 04-04 16:50:48 | [814][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0686 ntime: 0086 mem: 3.36 + 04-04 16:50:53 | [814][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0202 ntime: 0089 mem: 3.36 + 04-04 16:51:00 | [814][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0867 ntime: 0080 mem: 3.36 + 04-04 16:51:05 | Time info >>>> elapsed: 1167.72 mins remain: 265.07 mins + 04-04 16:51:07 | [815][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0950 ntime: 0081 mem: 3.36 + 04-04 16:51:13 | [815][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1680 ntime: 0083 mem: 3.36 + 04-04 16:51:20 | [815][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0084 mem: 3.36 + 04-04 16:51:26 | [815][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1121 ntime: 0076 mem: 3.36 + 04-04 16:51:34 | [815][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0771 ntime: 0082 mem: 3.36 + 04-04 16:51:41 | [815][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0887 ntime: 0089 mem: 3.36 + 04-04 16:51:52 | [815][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0184 ntime: 0082 mem: 3.36 + 04-04 16:52:01 | [815][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1390 ntime: 0082 mem: 3.36 + 04-04 16:52:06 | [815][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0174 ntime: 0080 mem: 3.36 + 04-04 16:52:14 | [815][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1528 ntime: 0084 mem: 3.36 + 04-04 16:52:22 | [815][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1022 ntime: 0084 mem: 3.36 + 04-04 16:52:29 | [815][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0845 ntime: 0076 mem: 3.36 + 04-04 16:52:36 | [815][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0087 ntime: 0075 mem: 3.36 + 04-04 16:52:43 | [815][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0122 ntime: 0087 mem: 3.36 + 04-04 16:52:49 | [815][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0184 ntime: 0078 mem: 3.36 + 04-04 16:52:55 | [815][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0656 ntime: 0076 mem: 3.36 + 04-04 16:53:04 | [815][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0076 ntime: 0090 mem: 3.36 + 04-04 16:53:12 | [815][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1383 ntime: 0083 mem: 3.36 + 04-04 16:53:18 | Time info >>>> elapsed: 1169.92 mins remain: 263.81 mins + 04-04 16:53:18 | [816][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 16:53:26 | [816][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0868 ntime: 0077 mem: 3.36 + 04-04 16:53:35 | [816][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1444 ntime: 0075 mem: 3.36 + 04-04 16:53:41 | [816][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0975 ntime: 0080 mem: 3.36 + 04-04 16:53:47 | [816][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0786 ntime: 0078 mem: 3.36 + 04-04 16:53:54 | [816][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0650 ntime: 0076 mem: 3.36 + 04-04 16:54:01 | [816][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0086 mem: 3.36 + 04-04 16:54:08 | [816][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0308 ntime: 0083 mem: 3.36 + 04-04 16:54:15 | [816][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0077 mem: 3.36 + 04-04 16:54:22 | [816][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 16:54:30 | [816][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 16:54:37 | [816][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0075 mem: 3.36 + 04-04 16:54:44 | [816][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1572 ntime: 0085 mem: 3.36 + 04-04 16:54:51 | [816][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 16:54:58 | [816][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0960 ntime: 0088 mem: 3.36 + 04-04 16:55:03 | [816][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0944 ntime: 0078 mem: 3.36 + 04-04 16:55:09 | [816][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0086 mem: 3.36 + 04-04 16:55:18 | [816][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0610 ntime: 0079 mem: 3.36 + 04-04 16:55:23 | Time info >>>> elapsed: 1172.02 mins remain: 262.52 mins + 04-04 16:55:24 | [817][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0871 ntime: 0079 mem: 3.36 + 04-04 16:55:31 | [817][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0935 ntime: 0082 mem: 3.36 + 04-04 16:55:36 | [817][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0079 mem: 3.36 + 04-04 16:55:46 | [817][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1109 ntime: 0084 mem: 3.36 + 04-04 16:55:53 | [817][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0376 ntime: 0084 mem: 3.36 + 04-04 16:55:59 | [817][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1129 ntime: 0086 mem: 3.36 + 04-04 16:56:04 | [817][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0103 ntime: 0082 mem: 3.36 + 04-04 16:56:10 | [817][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1395 ntime: 0077 mem: 3.36 + 04-04 16:56:16 | [817][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0143 ntime: 0077 mem: 3.36 + 04-04 16:56:22 | [817][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0187 ntime: 0073 mem: 3.36 + 04-04 16:56:33 | [817][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1254 ntime: 0078 mem: 3.36 + 04-04 16:56:39 | [817][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0384 ntime: 0080 mem: 3.36 + 04-04 16:56:44 | [817][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0082 mem: 3.36 + 04-04 16:56:50 | [817][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0903 ntime: 0089 mem: 3.36 + 04-04 16:56:57 | [817][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0065 ntime: 0082 mem: 3.36 + 04-04 16:57:03 | [817][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0078 mem: 3.36 + 04-04 16:57:08 | [817][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0083 mem: 3.36 + 04-04 16:57:14 | [817][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0948 ntime: 0086 mem: 3.36 + 04-04 16:57:19 | Time info >>>> elapsed: 1173.94 mins remain: 261.20 mins + 04-04 16:57:19 | [818][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0075 ntime: 0081 mem: 3.36 + 04-04 16:57:24 | [818][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0766 ntime: 0078 mem: 3.36 + 04-04 16:57:32 | [818][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1295 ntime: 0075 mem: 3.36 + 04-04 16:57:39 | [818][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 16:57:44 | [818][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0429 ntime: 0077 mem: 3.36 + 04-04 16:57:52 | [818][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0913 ntime: 0075 mem: 3.36 + 04-04 16:58:02 | [818][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1912 ntime: 0089 mem: 3.36 + 04-04 16:58:10 | [818][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0194 ntime: 0080 mem: 3.36 + 04-04 16:58:21 | [818][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1048 ntime: 0079 mem: 3.36 + 04-04 16:58:29 | [818][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0075 mem: 3.36 + 04-04 16:58:36 | [818][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0065 ntime: 0079 mem: 3.36 + 04-04 16:58:44 | [818][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0116 ntime: 0077 mem: 3.36 + 04-04 16:58:53 | [818][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0580 ntime: 0074 mem: 3.36 + 04-04 16:59:01 | [818][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1129 ntime: 0081 mem: 3.36 + 04-04 16:59:09 | [818][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0105 ntime: 0076 mem: 3.36 + 04-04 16:59:16 | [818][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0369 ntime: 0092 mem: 3.36 + 04-04 16:59:24 | [818][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 16:59:31 | [818][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1367 ntime: 0070 mem: 3.36 + 04-04 16:59:37 | Time info >>>> elapsed: 1176.25 mins remain: 259.95 mins + 04-04 16:59:38 | [819][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0516 ntime: 0072 mem: 3.36 + 04-04 16:59:49 | [819][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1559 ntime: 0082 mem: 3.36 + 04-04 16:59:56 | [819][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1443 ntime: 0085 mem: 3.36 + 04-04 17:00:03 | [819][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1398 ntime: 0082 mem: 3.36 + 04-04 17:00:12 | [819][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1232 ntime: 0088 mem: 3.36 + 04-04 17:00:19 | [819][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0137 ntime: 0078 mem: 3.36 + 04-04 17:00:26 | [819][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0121 ntime: 0082 mem: 3.36 + 04-04 17:00:32 | [819][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0596 ntime: 0079 mem: 3.36 + 04-04 17:00:38 | [819][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0124 ntime: 0079 mem: 3.36 + 04-04 17:00:47 | [819][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0113 ntime: 0078 mem: 3.36 + 04-04 17:00:54 | [819][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0134 ntime: 0079 mem: 3.36 + 04-04 17:01:00 | [819][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0080 mem: 3.36 + 04-04 17:01:05 | [819][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0073 mem: 3.36 + 04-04 17:01:13 | [819][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0081 mem: 3.36 + 04-04 17:01:19 | [819][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1421 ntime: 0080 mem: 3.36 + 04-04 17:01:25 | [819][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0142 ntime: 0086 mem: 3.36 + 04-04 17:01:35 | [819][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1399 ntime: 0082 mem: 3.36 + 04-04 17:01:42 | [819][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0993 ntime: 0081 mem: 3.36 + 04-04 17:01:48 | Time info >>>> elapsed: 1178.42 mins remain: 258.68 mins + 04-04 17:01:49 | [820][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1126 ntime: 0076 mem: 3.36 + 04-04 17:01:54 | [820][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0098 ntime: 0080 mem: 3.36 + 04-04 17:02:05 | [820][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1739 ntime: 0081 mem: 3.36 + 04-04 17:02:11 | [820][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0103 ntime: 0079 mem: 3.36 + 04-04 17:02:16 | [820][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0652 ntime: 0086 mem: 3.36 + 04-04 17:02:23 | [820][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0464 ntime: 0077 mem: 3.36 + 04-04 17:02:28 | [820][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0148 ntime: 0072 mem: 3.36 + 04-04 17:02:36 | [820][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0935 ntime: 0080 mem: 3.36 + 04-04 17:02:41 | [820][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0087 mem: 3.36 + 04-04 17:02:48 | [820][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0072 mem: 3.36 + 04-04 17:02:53 | [820][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0243 ntime: 0077 mem: 3.36 + 04-04 17:02:58 | [820][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0083 mem: 3.36 + 04-04 17:03:09 | [820][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0828 ntime: 0079 mem: 3.36 + 04-04 17:03:20 | [820][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0658 ntime: 0079 mem: 3.36 + 04-04 17:03:26 | [820][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0100 ntime: 0079 mem: 3.36 + 04-04 17:03:31 | [820][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0736 ntime: 0080 mem: 3.36 + 04-04 17:03:39 | [820][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0874 ntime: 0081 mem: 3.36 + 04-04 17:03:46 | [820][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0660 ntime: 0078 mem: 3.36 + 04-04 17:03:51 | Time info >>>> elapsed: 1180.48 mins remain: 257.38 mins + 04-04 17:03:51 | [821][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0127 ntime: 0078 mem: 3.36 + 04-04 17:03:58 | [821][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1100 ntime: 0076 mem: 3.36 + 04-04 17:04:02 | [821][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1071 ntime: 0078 mem: 3.36 + 04-04 17:04:08 | [821][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0142 ntime: 0079 mem: 3.36 + 04-04 17:04:14 | [821][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0621 ntime: 0079 mem: 3.36 + 04-04 17:04:19 | [821][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0131 ntime: 0084 mem: 3.36 + 04-04 17:04:26 | [821][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0847 ntime: 0081 mem: 3.36 + 04-04 17:04:33 | [821][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0203 ntime: 0074 mem: 3.36 + 04-04 17:04:38 | [821][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0079 ntime: 0082 mem: 3.36 + 04-04 17:04:44 | [821][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0142 ntime: 0082 mem: 3.36 + 04-04 17:04:49 | [821][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0095 ntime: 0083 mem: 3.36 + 04-04 17:04:55 | [821][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1041 ntime: 0072 mem: 3.36 + 04-04 17:05:02 | [821][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0080 mem: 3.36 + 04-04 17:05:07 | [821][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0799 ntime: 0083 mem: 3.36 + 04-04 17:05:14 | [821][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0797 ntime: 0080 mem: 3.36 + 04-04 17:05:23 | [821][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2003 ntime: 0085 mem: 3.36 + 04-04 17:05:30 | [821][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0979 ntime: 0081 mem: 3.36 + 04-04 17:05:38 | [821][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0101 ntime: 0080 mem: 3.36 + 04-04 17:05:46 | Time info >>>> elapsed: 1182.39 mins remain: 256.04 mins + 04-04 17:05:48 | [822][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2579 ntime: 0081 mem: 3.36 + 04-04 17:06:01 | [822][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1149 ntime: 0083 mem: 3.36 + 04-04 17:06:09 | [822][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0675 ntime: 0077 mem: 3.36 + 04-04 17:06:14 | [822][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0192 ntime: 0072 mem: 3.36 + 04-04 17:06:22 | [822][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0978 ntime: 0077 mem: 3.36 + 04-04 17:06:27 | [822][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0133 ntime: 0079 mem: 3.36 + 04-04 17:06:35 | [822][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0738 ntime: 0080 mem: 3.36 + 04-04 17:06:44 | [822][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1051 ntime: 0085 mem: 3.36 + 04-04 17:06:52 | [822][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1118 ntime: 0083 mem: 3.36 + 04-04 17:06:58 | [822][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0067 ntime: 0077 mem: 3.36 + 04-04 17:07:04 | [822][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0181 ntime: 0084 mem: 3.36 + 04-04 17:07:11 | [822][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1017 ntime: 0084 mem: 3.36 + 04-04 17:07:19 | [822][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0768 ntime: 0075 mem: 3.36 + 04-04 17:07:27 | [822][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1081 ntime: 0082 mem: 3.36 + 04-04 17:07:34 | [822][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1039 ntime: 0075 mem: 3.36 + 04-04 17:07:40 | [822][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0870 ntime: 0087 mem: 3.36 + 04-04 17:07:49 | [822][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1225 ntime: 0084 mem: 3.36 + 04-04 17:07:55 | [822][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0071 mem: 3.36 + 04-04 17:08:03 | Time info >>>> elapsed: 1184.68 mins remain: 254.79 mins + 04-04 17:08:03 | [823][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0082 mem: 3.36 + 04-04 17:08:11 | [823][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0698 ntime: 0085 mem: 3.36 + 04-04 17:08:19 | [823][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0783 ntime: 0079 mem: 3.36 + 04-04 17:08:25 | [823][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1088 ntime: 0077 mem: 3.36 + 04-04 17:08:33 | [823][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1363 ntime: 0077 mem: 3.36 + 04-04 17:08:40 | [823][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0814 ntime: 0087 mem: 3.36 + 04-04 17:08:48 | [823][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1467 ntime: 0077 mem: 3.36 + 04-04 17:08:55 | [823][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1099 ntime: 0077 mem: 3.36 + 04-04 17:09:02 | [823][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1486 ntime: 0086 mem: 3.36 + 04-04 17:09:10 | [823][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0943 ntime: 0082 mem: 3.36 + 04-04 17:09:16 | [823][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0077 mem: 3.36 + 04-04 17:09:21 | [823][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0507 ntime: 0081 mem: 3.36 + 04-04 17:09:28 | [823][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1152 ntime: 0082 mem: 3.36 + 04-04 17:09:34 | [823][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1049 ntime: 0075 mem: 3.36 + 04-04 17:09:41 | [823][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1249 ntime: 0079 mem: 3.36 + 04-04 17:09:46 | [823][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0141 ntime: 0089 mem: 3.36 + 04-04 17:09:53 | [823][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0067 ntime: 0076 mem: 3.36 + 04-04 17:10:01 | [823][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0083 mem: 3.36 + 04-04 17:10:08 | Time info >>>> elapsed: 1186.77 mins remain: 253.48 mins + 04-04 17:10:10 | [824][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 1235 ntime: 0078 mem: 3.36 + 04-04 17:10:16 | [824][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1012 ntime: 0078 mem: 3.36 + 04-04 17:10:24 | [824][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0945 ntime: 0085 mem: 3.36 + 04-04 17:10:30 | [824][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0781 ntime: 0077 mem: 3.36 + 04-04 17:10:36 | [824][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1150 ntime: 0058 mem: 3.36 + 04-04 17:10:45 | [824][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0084 mem: 3.36 + 04-04 17:10:52 | [824][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1130 ntime: 0083 mem: 3.36 + 04-04 17:10:58 | [824][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0629 ntime: 0080 mem: 3.36 + 04-04 17:11:05 | [824][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0707 ntime: 0084 mem: 3.36 + 04-04 17:11:14 | [824][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0109 ntime: 0076 mem: 3.36 + 04-04 17:11:21 | [824][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1426 ntime: 0077 mem: 3.36 + 04-04 17:11:28 | [824][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0074 mem: 3.36 + 04-04 17:11:36 | [824][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1319 ntime: 0083 mem: 3.36 + 04-04 17:11:45 | [824][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1392 ntime: 0079 mem: 3.36 + 04-04 17:11:51 | [824][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0123 ntime: 0091 mem: 3.36 + 04-04 17:11:58 | [824][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0915 ntime: 0073 mem: 3.36 + 04-04 17:12:07 | [824][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1062 ntime: 0083 mem: 3.36 + 04-04 17:12:13 | [824][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1364 ntime: 0074 mem: 3.36 + 04-04 17:12:19 | Time info >>>> elapsed: 1188.94 mins remain: 252.20 mins + 04-04 17:12:19 | [825][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0602 ntime: 0076 mem: 3.36 + 04-04 17:12:28 | [825][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1366 ntime: 0079 mem: 3.36 + 04-04 17:12:37 | [825][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0291 ntime: 0085 mem: 3.36 + 04-04 17:12:45 | [825][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1075 ntime: 0080 mem: 3.36 + 04-04 17:12:52 | [825][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0110 ntime: 0083 mem: 3.36 + 04-04 17:13:01 | [825][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1369 ntime: 0081 mem: 3.36 + 04-04 17:13:08 | [825][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0092 ntime: 0083 mem: 3.36 + 04-04 17:13:15 | [825][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0329 ntime: 0075 mem: 3.36 + 04-04 17:13:23 | [825][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1138 ntime: 0086 mem: 3.36 + 04-04 17:13:28 | [825][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0117 ntime: 0080 mem: 3.36 + 04-04 17:13:37 | [825][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0918 ntime: 0082 mem: 3.36 + 04-04 17:13:45 | [825][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1155 ntime: 0080 mem: 3.36 + 04-04 17:13:53 | [825][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0776 ntime: 0078 mem: 3.36 + 04-04 17:13:58 | [825][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1206 ntime: 0087 mem: 3.36 + 04-04 17:14:07 | [825][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0200 ntime: 0082 mem: 3.36 + 04-04 17:14:14 | [825][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0413 ntime: 0078 mem: 3.36 + 04-04 17:14:23 | [825][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1994 ntime: 0079 mem: 3.36 + 04-04 17:14:31 | [825][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0559 ntime: 0082 mem: 3.36 + 04-04 17:14:39 | Time info >>>> elapsed: 1191.28 mins remain: 250.95 mins + 04-04 17:14:39 | [826][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0080 ntime: 0079 mem: 3.36 + 04-04 17:14:47 | [826][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0139 ntime: 0077 mem: 3.36 + 04-04 17:14:52 | [826][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0072 ntime: 0082 mem: 3.36 + 04-04 17:15:02 | [826][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1671 ntime: 0074 mem: 3.36 + 04-04 17:15:10 | [826][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0958 ntime: 0072 mem: 3.36 + 04-04 17:15:18 | [826][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1067 ntime: 0081 mem: 3.36 + 04-04 17:15:25 | [826][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0141 ntime: 0082 mem: 3.36 + 04-04 17:15:35 | [826][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0079 mem: 3.36 + 04-04 17:15:43 | [826][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0121 ntime: 0080 mem: 3.36 + 04-04 17:15:50 | [826][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0854 ntime: 0077 mem: 3.36 + 04-04 17:15:57 | [826][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0117 ntime: 0077 mem: 3.36 + 04-04 17:16:04 | [826][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0949 ntime: 0085 mem: 3.36 + 04-04 17:16:12 | [826][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1204 ntime: 0078 mem: 3.36 + 04-04 17:16:19 | [826][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0765 ntime: 0080 mem: 3.36 + 04-04 17:16:26 | [826][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0068 ntime: 0088 mem: 3.36 + 04-04 17:16:34 | [826][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0575 ntime: 0081 mem: 3.36 + 04-04 17:16:40 | [826][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0295 ntime: 0085 mem: 3.36 + 04-04 17:16:46 | [826][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1137 ntime: 0080 mem: 3.36 + 04-04 17:16:52 | Time info >>>> elapsed: 1193.49 mins remain: 249.67 mins + 04-04 17:16:53 | [827][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1689 ntime: 0079 mem: 3.36 + 04-04 17:17:01 | [827][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1048 ntime: 0084 mem: 3.36 + 04-04 17:17:10 | [827][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0871 ntime: 0055 mem: 3.36 + 04-04 17:17:16 | [827][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0533 ntime: 0080 mem: 3.36 + 04-04 17:17:23 | [827][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1551 ntime: 0055 mem: 3.36 + 04-04 17:17:31 | [827][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0469 ntime: 0090 mem: 3.36 + 04-04 17:17:38 | [827][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1072 ntime: 0082 mem: 3.36 + 04-04 17:17:46 | [827][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0822 ntime: 0084 mem: 3.36 + 04-04 17:17:52 | [827][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0907 ntime: 0078 mem: 3.36 + 04-04 17:18:02 | [827][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1078 ntime: 0069 mem: 3.36 + 04-04 17:18:08 | [827][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0075 mem: 3.36 + 04-04 17:18:16 | [827][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0137 ntime: 0089 mem: 3.36 + 04-04 17:18:30 | [827][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1385 ntime: 0079 mem: 3.36 + 04-04 17:18:36 | [827][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0100 ntime: 0084 mem: 3.36 + 04-04 17:18:42 | [827][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0111 ntime: 0075 mem: 3.36 + 04-04 17:18:49 | [827][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0527 ntime: 0085 mem: 3.36 + 04-04 17:18:56 | [827][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0317 ntime: 0081 mem: 3.36 + 04-04 17:19:06 | [827][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0174 ntime: 0077 mem: 3.36 + 04-04 17:19:11 | Time info >>>> elapsed: 1195.80 mins remain: 248.40 mins + 04-04 17:19:11 | [828][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0737 ntime: 0083 mem: 3.36 + 04-04 17:19:17 | [828][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0084 mem: 3.36 + 04-04 17:19:22 | [828][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0826 ntime: 0079 mem: 3.36 + 04-04 17:19:29 | [828][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1107 ntime: 0079 mem: 3.36 + 04-04 17:19:35 | [828][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0079 mem: 3.36 + 04-04 17:19:42 | [828][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1385 ntime: 0082 mem: 3.36 + 04-04 17:19:49 | [828][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0711 ntime: 0079 mem: 3.36 + 04-04 17:19:54 | [828][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0128 ntime: 0075 mem: 3.36 + 04-04 17:20:01 | [828][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0617 ntime: 0081 mem: 3.36 + 04-04 17:20:08 | [828][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0863 ntime: 0077 mem: 3.36 + 04-04 17:20:16 | [828][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1076 ntime: 0057 mem: 3.36 + 04-04 17:20:24 | [828][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0210 ntime: 0086 mem: 3.36 + 04-04 17:20:30 | [828][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0696 ntime: 0076 mem: 3.36 + 04-04 17:20:37 | [828][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1455 ntime: 0081 mem: 3.36 + 04-04 17:20:45 | [828][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0163 ntime: 0081 mem: 3.36 + 04-04 17:20:52 | [828][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0725 ntime: 0086 mem: 3.36 + 04-04 17:20:58 | [828][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0123 ntime: 0078 mem: 3.36 + 04-04 17:21:04 | [828][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0896 ntime: 0086 mem: 3.36 + 04-04 17:21:10 | Time info >>>> elapsed: 1197.79 mins remain: 247.07 mins + 04-04 17:21:11 | [829][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1000 ntime: 0081 mem: 3.36 + 04-04 17:21:18 | [829][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0085 mem: 3.36 + 04-04 17:21:25 | [829][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1164 ntime: 0086 mem: 3.36 + 04-04 17:21:28 | [829][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0116 ntime: 0080 mem: 3.36 + 04-04 17:21:33 | [829][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0722 ntime: 0085 mem: 3.36 + 04-04 17:21:39 | [829][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0115 ntime: 0079 mem: 3.36 + 04-04 17:21:44 | [829][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0621 ntime: 0082 mem: 3.36 + 04-04 17:21:50 | [829][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1090 ntime: 0084 mem: 3.36 + 04-04 17:21:58 | [829][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0280 ntime: 0085 mem: 3.36 + 04-04 17:22:05 | [829][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0100 ntime: 0077 mem: 3.36 + 04-04 17:22:11 | [829][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1569 ntime: 0086 mem: 3.36 + 04-04 17:22:17 | [829][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1572 ntime: 0081 mem: 3.36 + 04-04 17:22:21 | [829][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0087 mem: 3.36 + 04-04 17:22:27 | [829][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0080 ntime: 0076 mem: 3.36 + 04-04 17:22:34 | [829][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0678 ntime: 0082 mem: 3.36 + 04-04 17:22:39 | [829][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0630 ntime: 0078 mem: 3.36 + 04-04 17:22:44 | [829][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0235 ntime: 0077 mem: 3.36 + 04-04 17:22:53 | [829][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0135 ntime: 0080 mem: 3.36 + 04-04 17:23:00 | Time info >>>> elapsed: 1199.63 mins remain: 245.71 mins + 04-04 17:23:01 | [830][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0530 ntime: 0083 mem: 3.36 + 04-04 17:23:12 | [830][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1415 ntime: 0081 mem: 3.36 + 04-04 17:23:21 | [830][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0125 ntime: 0081 mem: 3.36 + 04-04 17:23:31 | [830][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1059 ntime: 0078 mem: 3.36 + 04-04 17:23:37 | [830][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0115 ntime: 0085 mem: 3.36 + 04-04 17:23:44 | [830][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0084 mem: 3.36 + 04-04 17:23:52 | [830][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0820 ntime: 0079 mem: 3.36 + 04-04 17:23:59 | [830][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1178 ntime: 0083 mem: 3.36 + 04-04 17:24:06 | [830][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0118 ntime: 0084 mem: 3.36 + 04-04 17:24:13 | [830][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0809 ntime: 0077 mem: 3.36 + 04-04 17:24:21 | [830][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0533 ntime: 0088 mem: 3.36 + 04-04 17:24:28 | [830][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1442 ntime: 0079 mem: 3.36 + 04-04 17:24:36 | [830][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1511 ntime: 0083 mem: 3.36 + 04-04 17:24:44 | [830][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0309 ntime: 0082 mem: 3.36 + 04-04 17:24:52 | [830][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0075 ntime: 0077 mem: 3.36 + 04-04 17:24:59 | [830][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0104 ntime: 0078 mem: 3.36 + 04-04 17:25:05 | [830][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0069 ntime: 0087 mem: 3.36 + 04-04 17:25:12 | [830][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0361 ntime: 0079 mem: 3.36 + 04-04 17:25:17 | Time info >>>> elapsed: 1201.92 mins remain: 244.43 mins + 04-04 17:25:18 | [831][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0517 ntime: 0082 mem: 3.36 + 04-04 17:25:24 | [831][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0119 ntime: 0077 mem: 3.36 + 04-04 17:25:32 | [831][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0079 mem: 3.36 + 04-04 17:25:37 | [831][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0085 mem: 3.36 + 04-04 17:25:45 | [831][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0094 ntime: 0081 mem: 3.36 + 04-04 17:25:50 | [831][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0444 ntime: 0077 mem: 3.36 + 04-04 17:25:56 | [831][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0875 ntime: 0083 mem: 3.36 + 04-04 17:26:04 | [831][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1089 ntime: 0086 mem: 3.36 + 04-04 17:26:09 | [831][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 17:26:15 | [831][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1037 ntime: 0082 mem: 3.36 + 04-04 17:26:21 | [831][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0843 ntime: 0059 mem: 3.36 + 04-04 17:26:26 | [831][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0426 ntime: 0079 mem: 3.36 + 04-04 17:26:33 | [831][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0207 ntime: 0078 mem: 3.36 + 04-04 17:26:47 | [831][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0104 ntime: 0072 mem: 3.36 + 04-04 17:26:53 | [831][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0624 ntime: 0080 mem: 3.36 + 04-04 17:27:00 | [831][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1554 ntime: 0087 mem: 3.36 + 04-04 17:27:04 | [831][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0487 ntime: 0075 mem: 3.36 + 04-04 17:27:10 | [831][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1635 ntime: 0081 mem: 3.36 + 04-04 17:27:15 | Time info >>>> elapsed: 1203.88 mins remain: 243.09 mins + 04-04 17:27:16 | [832][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0047 ntime: 0082 mem: 3.36 + 04-04 17:27:22 | [832][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0777 ntime: 0080 mem: 3.36 + 04-04 17:27:30 | [832][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1445 ntime: 0086 mem: 3.36 + 04-04 17:27:36 | [832][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0777 ntime: 0084 mem: 3.36 + 04-04 17:27:42 | [832][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0519 ntime: 0078 mem: 3.36 + 04-04 17:27:52 | [832][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1544 ntime: 0072 mem: 3.36 + 04-04 17:27:59 | [832][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1334 ntime: 0085 mem: 3.36 + 04-04 17:28:04 | [832][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0471 ntime: 0082 mem: 3.36 + 04-04 17:28:13 | [832][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0081 mem: 3.36 + 04-04 17:28:21 | [832][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0655 ntime: 0078 mem: 3.36 + 04-04 17:28:28 | [832][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1311 ntime: 0080 mem: 3.36 + 04-04 17:28:35 | [832][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0509 ntime: 0084 mem: 3.36 + 04-04 17:28:39 | [832][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 17:28:48 | [832][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0908 ntime: 0084 mem: 3.36 + 04-04 17:28:54 | [832][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0242 ntime: 0084 mem: 3.36 + 04-04 17:29:01 | [832][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1111 ntime: 0082 mem: 3.36 + 04-04 17:29:08 | [832][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1221 ntime: 0083 mem: 3.36 + 04-04 17:29:13 | [832][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0488 ntime: 0082 mem: 3.36 + 04-04 17:29:18 | Time info >>>> elapsed: 1205.92 mins remain: 241.76 mins + 04-04 17:29:18 | [833][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0098 ntime: 0079 mem: 3.36 + 04-04 17:29:24 | [833][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1199 ntime: 0076 mem: 3.36 + 04-04 17:29:33 | [833][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0755 ntime: 0072 mem: 3.36 + 04-04 17:29:39 | [833][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0077 ntime: 0079 mem: 3.36 + 04-04 17:29:44 | [833][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0081 mem: 3.36 + 04-04 17:29:49 | [833][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0080 mem: 3.36 + 04-04 17:29:57 | [833][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0886 ntime: 0078 mem: 3.36 + 04-04 17:30:04 | [833][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1061 ntime: 0085 mem: 3.36 + 04-04 17:30:09 | [833][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0737 ntime: 0077 mem: 3.36 + 04-04 17:30:15 | [833][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0346 ntime: 0077 mem: 3.36 + 04-04 17:30:21 | [833][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0513 ntime: 0086 mem: 3.36 + 04-04 17:30:26 | [833][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0812 ntime: 0086 mem: 3.36 + 04-04 17:30:32 | [833][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1649 ntime: 0078 mem: 3.36 + 04-04 17:30:42 | [833][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0134 ntime: 0078 mem: 3.36 + 04-04 17:30:48 | [833][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0623 ntime: 0080 mem: 3.36 + 04-04 17:30:55 | [833][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0629 ntime: 0084 mem: 3.36 + 04-04 17:31:02 | [833][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1202 ntime: 0083 mem: 3.36 + 04-04 17:31:08 | [833][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0579 ntime: 0082 mem: 3.36 + 04-04 17:31:12 | Time info >>>> elapsed: 1207.84 mins remain: 240.41 mins + 04-04 17:31:13 | [834][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0075 mem: 3.36 + 04-04 17:31:17 | [834][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0084 mem: 3.36 + 04-04 17:31:25 | [834][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0812 ntime: 0080 mem: 3.36 + 04-04 17:31:30 | [834][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0138 ntime: 0079 mem: 3.36 + 04-04 17:31:36 | [834][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0069 ntime: 0079 mem: 3.36 + 04-04 17:31:42 | [834][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0067 ntime: 0081 mem: 3.36 + 04-04 17:31:48 | [834][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1688 ntime: 0073 mem: 3.36 + 04-04 17:31:55 | [834][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0914 ntime: 0077 mem: 3.36 + 04-04 17:31:59 | [834][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0075 mem: 3.36 + 04-04 17:32:06 | [834][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0386 ntime: 0059 mem: 3.36 + 04-04 17:32:12 | [834][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1456 ntime: 0080 mem: 3.36 + 04-04 17:32:17 | [834][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1380 ntime: 0083 mem: 3.36 + 04-04 17:32:25 | [834][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0694 ntime: 0079 mem: 3.36 + 04-04 17:32:32 | [834][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0981 ntime: 0078 mem: 3.36 + 04-04 17:32:37 | [834][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0308 ntime: 0082 mem: 3.36 + 04-04 17:32:44 | [834][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0822 ntime: 0082 mem: 3.36 + 04-04 17:32:50 | [834][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0877 ntime: 0078 mem: 3.36 + 04-04 17:32:56 | [834][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0919 ntime: 0082 mem: 3.36 + 04-04 17:33:02 | Time info >>>> elapsed: 1209.66 mins remain: 239.03 mins + 04-04 17:33:02 | [835][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0079 mem: 3.36 + 04-04 17:33:10 | [835][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0083 mem: 3.36 + 04-04 17:33:17 | [835][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0703 ntime: 0081 mem: 3.36 + 04-04 17:33:22 | [835][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0939 ntime: 0081 mem: 3.36 + 04-04 17:33:29 | [835][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0138 ntime: 0082 mem: 3.36 + 04-04 17:33:36 | [835][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0143 ntime: 0086 mem: 3.36 + 04-04 17:33:42 | [835][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0083 mem: 3.36 + 04-04 17:33:50 | [835][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0087 mem: 3.36 + 04-04 17:33:56 | [835][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0440 ntime: 0085 mem: 3.36 + 04-04 17:34:01 | [835][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0069 ntime: 0083 mem: 3.36 + 04-04 17:34:07 | [835][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0447 ntime: 0081 mem: 3.36 + 04-04 17:34:13 | [835][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1127 ntime: 0085 mem: 3.36 + 04-04 17:34:20 | [835][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1233 ntime: 0085 mem: 3.36 + 04-04 17:34:26 | [835][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1163 ntime: 0081 mem: 3.36 + 04-04 17:34:31 | [835][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0811 ntime: 0088 mem: 3.36 + 04-04 17:34:40 | [835][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0083 ntime: 0085 mem: 3.36 + 04-04 17:34:47 | [835][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0812 ntime: 0079 mem: 3.36 + 04-04 17:34:53 | [835][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0122 ntime: 0081 mem: 3.36 + 04-04 17:34:57 | Time info >>>> elapsed: 1211.58 mins remain: 237.68 mins + 04-04 17:34:57 | [836][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0133 ntime: 0078 mem: 3.36 + 04-04 17:35:04 | [836][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0097 ntime: 0075 mem: 3.36 + 04-04 17:35:10 | [836][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0196 ntime: 0082 mem: 3.36 + 04-04 17:35:18 | [836][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 17:35:27 | [836][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0082 ntime: 0077 mem: 3.36 + 04-04 17:35:33 | [836][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0533 ntime: 0085 mem: 3.36 + 04-04 17:35:39 | [836][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0585 ntime: 0087 mem: 3.36 + 04-04 17:35:47 | [836][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1243 ntime: 0084 mem: 3.36 + 04-04 17:35:54 | [836][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0079 ntime: 0074 mem: 3.36 + 04-04 17:36:00 | [836][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0464 ntime: 0080 mem: 3.36 + 04-04 17:36:06 | [836][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1230 ntime: 0082 mem: 3.36 + 04-04 17:36:13 | [836][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0076 ntime: 0080 mem: 3.36 + 04-04 17:36:17 | [836][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0096 ntime: 0086 mem: 3.36 + 04-04 17:36:25 | [836][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0914 ntime: 0078 mem: 3.36 + 04-04 17:36:31 | [836][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0093 ntime: 0080 mem: 3.36 + 04-04 17:36:36 | [836][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0547 ntime: 0088 mem: 3.36 + 04-04 17:36:42 | [836][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0331 ntime: 0076 mem: 3.36 + 04-04 17:36:47 | [836][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0102 ntime: 0085 mem: 3.36 + 04-04 17:36:54 | Time info >>>> elapsed: 1213.53 mins remain: 236.33 mins + 04-04 17:36:55 | [837][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0794 ntime: 0085 mem: 3.36 + 04-04 17:37:01 | [837][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0113 ntime: 0083 mem: 3.36 + 04-04 17:37:07 | [837][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 17:37:14 | [837][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1273 ntime: 0087 mem: 3.36 + 04-04 17:37:21 | [837][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0148 ntime: 0079 mem: 3.36 + 04-04 17:37:27 | [837][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0082 mem: 3.36 + 04-04 17:37:36 | [837][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0084 mem: 3.36 + 04-04 17:37:42 | [837][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0708 ntime: 0081 mem: 3.36 + 04-04 17:37:47 | [837][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0230 ntime: 0085 mem: 3.36 + 04-04 17:37:53 | [837][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1101 ntime: 0079 mem: 3.36 + 04-04 17:37:59 | [837][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0587 ntime: 0080 mem: 3.36 + 04-04 17:38:08 | [837][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1615 ntime: 0077 mem: 3.36 + 04-04 17:38:17 | [837][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0081 ntime: 0077 mem: 3.36 + 04-04 17:38:24 | [837][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0073 ntime: 0082 mem: 3.36 + 04-04 17:38:31 | [837][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0179 ntime: 0084 mem: 3.36 + 04-04 17:38:38 | [837][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0078 mem: 3.36 + 04-04 17:38:46 | [837][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0538 ntime: 0089 mem: 3.36 + 04-04 17:38:54 | [837][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0688 ntime: 0077 mem: 3.36 + 04-04 17:39:00 | Time info >>>> elapsed: 1215.63 mins remain: 235.00 mins + 04-04 17:39:00 | [838][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0065 ntime: 0078 mem: 3.36 + 04-04 17:39:07 | [838][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0216 ntime: 0085 mem: 3.36 + 04-04 17:39:17 | [838][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0081 mem: 3.36 + 04-04 17:39:23 | [838][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0096 ntime: 0081 mem: 3.36 + 04-04 17:39:29 | [838][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1403 ntime: 0079 mem: 3.36 + 04-04 17:39:35 | [838][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0108 ntime: 0075 mem: 3.36 + 04-04 17:39:42 | [838][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0975 ntime: 0075 mem: 3.36 + 04-04 17:39:48 | [838][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0701 ntime: 0079 mem: 3.36 + 04-04 17:39:55 | [838][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0077 ntime: 0077 mem: 3.36 + 04-04 17:40:00 | [838][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0217 ntime: 0081 mem: 3.36 + 04-04 17:40:06 | [838][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0068 ntime: 0077 mem: 3.36 + 04-04 17:40:13 | [838][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0087 ntime: 0075 mem: 3.36 + 04-04 17:40:17 | [838][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0128 ntime: 0076 mem: 3.36 + 04-04 17:40:25 | [838][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1219 ntime: 0081 mem: 3.36 + 04-04 17:40:31 | [838][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0078 mem: 3.36 + 04-04 17:40:39 | [838][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1081 ntime: 0078 mem: 3.36 + 04-04 17:40:46 | [838][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 17:40:52 | [838][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1524 ntime: 0082 mem: 3.36 + 04-04 17:40:57 | Time info >>>> elapsed: 1217.59 mins remain: 233.65 mins + 04-04 17:40:59 | [839][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 1051 ntime: 0077 mem: 3.36 + 04-04 17:41:07 | [839][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1294 ntime: 0082 mem: 3.36 + 04-04 17:41:14 | [839][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0145 ntime: 0075 mem: 3.36 + 04-04 17:41:20 | [839][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0445 ntime: 0078 mem: 3.36 + 04-04 17:41:25 | [839][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0920 ntime: 0081 mem: 3.36 + 04-04 17:41:31 | [839][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1421 ntime: 0086 mem: 3.36 + 04-04 17:41:39 | [839][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1000 ntime: 0085 mem: 3.36 + 04-04 17:41:44 | [839][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0079 mem: 3.36 + 04-04 17:41:49 | [839][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0829 ntime: 0086 mem: 3.36 + 04-04 17:41:54 | [839][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0782 ntime: 0079 mem: 3.36 + 04-04 17:42:01 | [839][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1833 ntime: 0081 mem: 3.36 + 04-04 17:42:11 | [839][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1403 ntime: 0087 mem: 3.36 + 04-04 17:42:17 | [839][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0807 ntime: 0087 mem: 3.36 + 04-04 17:42:22 | [839][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0250 ntime: 0080 mem: 3.36 + 04-04 17:42:30 | [839][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1283 ntime: 0082 mem: 3.36 + 04-04 17:42:38 | [839][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1121 ntime: 0085 mem: 3.36 + 04-04 17:42:44 | [839][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0956 ntime: 0080 mem: 3.36 + 04-04 17:42:51 | [839][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0076 mem: 3.36 + 04-04 17:42:55 | Time info >>>> elapsed: 1219.55 mins remain: 232.30 mins + 04-04 17:42:56 | [840][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0053 ntime: 0081 mem: 3.36 + 04-04 17:43:03 | [840][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0755 ntime: 0078 mem: 3.36 + 04-04 17:43:10 | [840][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0398 ntime: 0087 mem: 3.36 + 04-04 17:43:15 | [840][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0115 ntime: 0082 mem: 3.36 + 04-04 17:43:22 | [840][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0111 ntime: 0087 mem: 3.36 + 04-04 17:43:28 | [840][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0797 ntime: 0079 mem: 3.36 + 04-04 17:43:34 | [840][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0714 ntime: 0075 mem: 3.36 + 04-04 17:43:39 | [840][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0078 ntime: 0080 mem: 3.36 + 04-04 17:43:47 | [840][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0653 ntime: 0081 mem: 3.36 + 04-04 17:43:51 | [840][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0085 ntime: 0056 mem: 3.36 + 04-04 17:43:57 | [840][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0585 ntime: 0079 mem: 3.36 + 04-04 17:44:02 | [840][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0798 ntime: 0075 mem: 3.36 + 04-04 17:44:08 | [840][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0125 ntime: 0071 mem: 3.36 + 04-04 17:44:16 | [840][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1791 ntime: 0079 mem: 3.36 + 04-04 17:44:24 | [840][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0080 ntime: 0081 mem: 3.36 + 04-04 17:44:30 | [840][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0174 ntime: 0077 mem: 3.36 + 04-04 17:44:35 | [840][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0065 ntime: 0079 mem: 3.36 + 04-04 17:44:41 | [840][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0686 ntime: 0085 mem: 3.36 + 04-04 17:44:45 | Time info >>>> elapsed: 1221.39 mins remain: 230.92 mins + 04-04 17:44:46 | [841][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0086 ntime: 0078 mem: 3.36 + 04-04 17:44:52 | [841][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0594 ntime: 0080 mem: 3.36 + 04-04 17:44:58 | [841][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0668 ntime: 0084 mem: 3.36 + 04-04 17:45:03 | [841][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0339 ntime: 0079 mem: 3.36 + 04-04 17:45:09 | [841][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0622 ntime: 0073 mem: 3.36 + 04-04 17:45:15 | [841][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0311 ntime: 0086 mem: 3.36 + 04-04 17:45:22 | [841][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0083 ntime: 0083 mem: 3.36 + 04-04 17:45:28 | [841][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0081 mem: 3.36 + 04-04 17:45:37 | [841][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1279 ntime: 0081 mem: 3.36 + 04-04 17:45:45 | [841][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0286 ntime: 0080 mem: 3.36 + 04-04 17:45:52 | [841][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1347 ntime: 0086 mem: 3.36 + 04-04 17:46:00 | [841][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1509 ntime: 0082 mem: 3.36 + 04-04 17:46:07 | [841][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0071 ntime: 0082 mem: 3.36 + 04-04 17:46:15 | [841][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0850 ntime: 0080 mem: 3.36 + 04-04 17:46:22 | [841][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0085 ntime: 0087 mem: 3.36 + 04-04 17:46:29 | [841][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0575 ntime: 0076 mem: 3.36 + 04-04 17:46:36 | [841][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0546 ntime: 0081 mem: 3.36 + 04-04 17:46:46 | [841][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0061 mem: 3.36 + 04-04 17:46:54 | Time info >>>> elapsed: 1223.52 mins remain: 229.59 mins + 04-04 17:46:54 | [842][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0348 ntime: 0087 mem: 3.36 + 04-04 17:47:08 | [842][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0937 ntime: 0076 mem: 3.36 + 04-04 17:47:19 | [842][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0098 ntime: 0084 mem: 3.36 + 04-04 17:47:24 | [842][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0081 mem: 3.36 + 04-04 17:47:28 | [842][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0389 ntime: 0074 mem: 3.36 + 04-04 17:47:34 | [842][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0072 ntime: 0079 mem: 3.36 + 04-04 17:47:39 | [842][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0079 mem: 3.36 + 04-04 17:47:49 | [842][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0978 ntime: 0081 mem: 3.36 + 04-04 17:47:54 | [842][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0897 ntime: 0080 mem: 3.36 + 04-04 17:48:00 | [842][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1337 ntime: 0073 mem: 3.36 + 04-04 17:48:07 | [842][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0837 ntime: 0085 mem: 3.36 + 04-04 17:48:12 | [842][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0085 mem: 3.36 + 04-04 17:48:18 | [842][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0178 ntime: 0082 mem: 3.36 + 04-04 17:48:23 | [842][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0087 mem: 3.36 + 04-04 17:48:30 | [842][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0858 ntime: 0086 mem: 3.36 + 04-04 17:48:39 | [842][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1517 ntime: 0076 mem: 3.36 + 04-04 17:48:44 | [842][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0088 ntime: 0081 mem: 3.36 + 04-04 17:48:51 | [842][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0438 ntime: 0077 mem: 3.36 + 04-04 17:48:56 | Time info >>>> elapsed: 1225.57 mins remain: 228.25 mins + 04-04 17:48:57 | [843][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0089 ntime: 0089 mem: 3.36 + 04-04 17:49:05 | [843][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0747 ntime: 0075 mem: 3.36 + 04-04 17:49:11 | [843][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0472 ntime: 0085 mem: 3.36 + 04-04 17:49:18 | [843][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0101 ntime: 0077 mem: 3.36 + 04-04 17:49:25 | [843][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0118 ntime: 0085 mem: 3.36 + 04-04 17:49:32 | [843][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0131 ntime: 0076 mem: 3.36 + 04-04 17:49:38 | [843][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0198 ntime: 0076 mem: 3.36 + 04-04 17:49:43 | [843][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0122 ntime: 0082 mem: 3.36 + 04-04 17:49:52 | [843][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0810 ntime: 0080 mem: 3.36 + 04-04 17:49:59 | [843][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0095 ntime: 0079 mem: 3.36 + 04-04 17:50:04 | [843][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0977 ntime: 0079 mem: 3.36 + 04-04 17:50:10 | [843][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0753 ntime: 0080 mem: 3.36 + 04-04 17:50:14 | [843][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0510 ntime: 0072 mem: 3.36 + 04-04 17:50:21 | [843][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1189 ntime: 0084 mem: 3.36 + 04-04 17:50:27 | [843][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0839 ntime: 0081 mem: 3.36 + 04-04 17:50:33 | [843][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0072 mem: 3.36 + 04-04 17:50:39 | [843][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0112 ntime: 0078 mem: 3.36 + 04-04 17:50:45 | [843][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0121 ntime: 0081 mem: 3.36 + 04-04 17:50:51 | Time info >>>> elapsed: 1227.48 mins remain: 226.88 mins + 04-04 17:50:52 | [844][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1028 ntime: 0087 mem: 3.36 + 04-04 17:50:59 | [844][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1182 ntime: 0085 mem: 3.36 + 04-04 17:51:05 | [844][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0140 ntime: 0082 mem: 3.36 + 04-04 17:51:12 | [844][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0119 ntime: 0079 mem: 3.36 + 04-04 17:51:18 | [844][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 17:51:24 | [844][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0416 ntime: 0082 mem: 3.36 + 04-04 17:51:29 | [844][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 17:51:34 | [844][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0086 ntime: 0083 mem: 3.36 + 04-04 17:51:42 | [844][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0904 ntime: 0085 mem: 3.36 + 04-04 17:51:47 | [844][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0194 ntime: 0074 mem: 3.36 + 04-04 17:51:52 | [844][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0201 ntime: 0087 mem: 3.36 + 04-04 17:51:58 | [844][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0681 ntime: 0083 mem: 3.36 + 04-04 17:52:03 | [844][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0160 ntime: 0081 mem: 3.36 + 04-04 17:52:09 | [844][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0114 ntime: 0092 mem: 3.36 + 04-04 17:52:14 | [844][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0076 mem: 3.36 + 04-04 17:52:24 | [844][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0613 ntime: 0078 mem: 3.36 + 04-04 17:52:30 | [844][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0082 mem: 3.36 + 04-04 17:52:36 | [844][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0164 ntime: 0079 mem: 3.36 + 04-04 17:52:39 | Time info >>>> elapsed: 1229.28 mins remain: 225.49 mins + 04-04 17:52:40 | [845][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0167 ntime: 0080 mem: 3.36 + 04-04 17:52:47 | [845][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0097 ntime: 0088 mem: 3.36 + 04-04 17:52:54 | [845][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0077 mem: 3.36 + 04-04 17:53:01 | [845][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0875 ntime: 0077 mem: 3.36 + 04-04 17:53:07 | [845][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0692 ntime: 0076 mem: 3.36 + 04-04 17:53:13 | [845][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1265 ntime: 0072 mem: 3.36 + 04-04 17:53:20 | [845][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1090 ntime: 0076 mem: 3.36 + 04-04 17:53:27 | [845][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0138 ntime: 0083 mem: 3.36 + 04-04 17:53:34 | [845][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 17:53:40 | [845][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0068 ntime: 0082 mem: 3.36 + 04-04 17:53:45 | [845][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0899 ntime: 0082 mem: 3.36 + 04-04 17:53:49 | [845][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0185 ntime: 0073 mem: 3.36 + 04-04 17:53:56 | [845][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0838 ntime: 0086 mem: 3.36 + 04-04 17:54:01 | [845][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0725 ntime: 0080 mem: 3.36 + 04-04 17:54:09 | [845][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0448 ntime: 0079 mem: 3.36 + 04-04 17:54:14 | [845][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0072 mem: 3.36 + 04-04 17:54:21 | [845][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0705 ntime: 0080 mem: 3.36 + 04-04 17:54:27 | [845][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1006 ntime: 0075 mem: 3.36 + 04-04 17:54:34 | Time info >>>> elapsed: 1231.19 mins remain: 224.12 mins + 04-04 17:54:34 | [846][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0201 ntime: 0076 mem: 3.36 + 04-04 17:54:40 | [846][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0286 ntime: 0086 mem: 3.36 + 04-04 17:54:46 | [846][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0468 ntime: 0078 mem: 3.36 + 04-04 17:54:52 | [846][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0118 ntime: 0081 mem: 3.36 + 04-04 17:54:58 | [846][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1110 ntime: 0082 mem: 3.36 + 04-04 17:55:04 | [846][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0084 ntime: 0080 mem: 3.36 + 04-04 17:55:10 | [846][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0568 ntime: 0077 mem: 3.36 + 04-04 17:55:15 | [846][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0086 mem: 3.36 + 04-04 17:55:21 | [846][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0100 ntime: 0081 mem: 3.36 + 04-04 17:55:26 | [846][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0154 ntime: 0077 mem: 3.36 + 04-04 17:55:33 | [846][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0074 mem: 3.36 + 04-04 17:55:37 | [846][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0077 mem: 3.36 + 04-04 17:55:45 | [846][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0156 ntime: 0081 mem: 3.36 + 04-04 17:55:52 | [846][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0119 ntime: 0079 mem: 3.36 + 04-04 17:56:00 | [846][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1359 ntime: 0074 mem: 3.36 + 04-04 17:56:10 | [846][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0932 ntime: 0075 mem: 3.36 + 04-04 17:56:19 | [846][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0771 ntime: 0090 mem: 3.36 + 04-04 17:56:26 | [846][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0463 ntime: 0077 mem: 3.36 + 04-04 17:56:30 | Time info >>>> elapsed: 1233.13 mins remain: 222.75 mins + 04-04 17:56:31 | [847][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0607 ntime: 0078 mem: 3.36 + 04-04 17:56:37 | [847][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1405 ntime: 0080 mem: 3.36 + 04-04 17:56:45 | [847][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0919 ntime: 0081 mem: 3.36 + 04-04 17:56:52 | [847][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0940 ntime: 0085 mem: 3.36 + 04-04 17:56:58 | [847][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0259 ntime: 0078 mem: 3.36 + 04-04 17:57:04 | [847][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0151 ntime: 0082 mem: 3.36 + 04-04 17:57:11 | [847][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0911 ntime: 0083 mem: 3.36 + 04-04 17:57:18 | [847][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0080 mem: 3.36 + 04-04 17:57:24 | [847][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1032 ntime: 0078 mem: 3.36 + 04-04 17:57:31 | [847][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1325 ntime: 0086 mem: 3.36 + 04-04 17:57:38 | [847][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0915 ntime: 0087 mem: 3.36 + 04-04 17:57:44 | [847][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1021 ntime: 0080 mem: 3.36 + 04-04 17:57:53 | [847][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0658 ntime: 0085 mem: 3.36 + 04-04 17:58:00 | [847][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0072 ntime: 0078 mem: 3.36 + 04-04 17:58:07 | [847][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0184 ntime: 0083 mem: 3.36 + 04-04 17:58:13 | [847][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0922 ntime: 0082 mem: 3.36 + 04-04 17:58:19 | [847][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0133 ntime: 0078 mem: 3.36 + 04-04 17:58:25 | [847][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0083 ntime: 0079 mem: 3.36 + 04-04 17:58:29 | Time info >>>> elapsed: 1235.10 mins remain: 221.39 mins + 04-04 17:58:29 | [848][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0100 ntime: 0072 mem: 3.36 + 04-04 17:58:36 | [848][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1027 ntime: 0083 mem: 3.36 + 04-04 17:58:41 | [848][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0083 mem: 3.36 + 04-04 17:58:47 | [848][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1530 ntime: 0077 mem: 3.36 + 04-04 17:58:52 | [848][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0075 mem: 3.36 + 04-04 17:59:00 | [848][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0617 ntime: 0076 mem: 3.36 + 04-04 17:59:07 | [848][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0935 ntime: 0078 mem: 3.36 + 04-04 17:59:14 | [848][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0978 ntime: 0078 mem: 3.36 + 04-04 17:59:21 | [848][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1295 ntime: 0075 mem: 3.36 + 04-04 17:59:27 | [848][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0138 ntime: 0080 mem: 3.36 + 04-04 17:59:33 | [848][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0363 ntime: 0078 mem: 3.36 + 04-04 17:59:37 | [848][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0165 ntime: 0078 mem: 3.36 + 04-04 17:59:44 | [848][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1130 ntime: 0080 mem: 3.36 + 04-04 17:59:53 | [848][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1498 ntime: 0078 mem: 3.36 + 04-04 18:00:01 | [848][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0624 ntime: 0082 mem: 3.36 + 04-04 18:00:08 | [848][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1296 ntime: 0083 mem: 3.36 + 04-04 18:00:16 | [848][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1080 ntime: 0076 mem: 3.36 + 04-04 18:00:23 | [848][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1055 ntime: 0083 mem: 3.36 + 04-04 18:00:28 | Time info >>>> elapsed: 1237.09 mins remain: 220.02 mins + 04-04 18:00:28 | [849][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0156 ntime: 0077 mem: 3.36 + 04-04 18:00:34 | [849][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0063 ntime: 0084 mem: 3.36 + 04-04 18:00:39 | [849][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0616 ntime: 0085 mem: 3.36 + 04-04 18:00:44 | [849][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0073 mem: 3.36 + 04-04 18:00:51 | [849][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1240 ntime: 0088 mem: 3.36 + 04-04 18:00:56 | [849][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0109 ntime: 0076 mem: 3.36 + 04-04 18:01:03 | [849][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0878 ntime: 0085 mem: 3.36 + 04-04 18:01:08 | [849][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0534 ntime: 0079 mem: 3.36 + 04-04 18:01:13 | [849][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1015 ntime: 0080 mem: 3.36 + 04-04 18:01:23 | [849][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0161 ntime: 0085 mem: 3.36 + 04-04 18:01:29 | [849][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0114 ntime: 0057 mem: 3.36 + 04-04 18:01:36 | [849][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0907 ntime: 0085 mem: 3.36 + 04-04 18:01:42 | [849][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0805 ntime: 0082 mem: 3.36 + 04-04 18:01:47 | [849][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1126 ntime: 0076 mem: 3.36 + 04-04 18:01:53 | [849][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0891 ntime: 0075 mem: 3.36 + 04-04 18:01:59 | [849][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0091 ntime: 0080 mem: 3.36 + 04-04 18:02:05 | [849][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0080 mem: 3.36 + 04-04 18:02:11 | [849][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0738 ntime: 0085 mem: 3.36 + 04-04 18:02:15 | Time info >>>> elapsed: 1238.88 mins remain: 218.63 mins + 04-04 18:02:15 | [850][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0110 ntime: 0077 mem: 3.36 + 04-04 18:02:21 | [850][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0090 mem: 3.36 + 04-04 18:02:28 | [850][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0814 ntime: 0078 mem: 3.36 + 04-04 18:02:33 | [850][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0084 mem: 3.36 + 04-04 18:02:39 | [850][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1082 ntime: 0079 mem: 3.36 + 04-04 18:02:43 | [850][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0110 ntime: 0078 mem: 3.36 + 04-04 18:02:50 | [850][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0932 ntime: 0082 mem: 3.36 + 04-04 18:02:56 | [850][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1139 ntime: 0088 mem: 3.36 + 04-04 18:03:04 | [850][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1017 ntime: 0090 mem: 3.36 + 04-04 18:03:09 | [850][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0129 ntime: 0080 mem: 3.36 + 04-04 18:03:15 | [850][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1136 ntime: 0080 mem: 3.36 + 04-04 18:03:21 | [850][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1479 ntime: 0084 mem: 3.36 + 04-04 18:03:26 | [850][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0646 ntime: 0083 mem: 3.36 + 04-04 18:03:31 | [850][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0892 ntime: 0082 mem: 3.36 + 04-04 18:03:37 | [850][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1114 ntime: 0080 mem: 3.36 + 04-04 18:03:42 | [850][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0079 mem: 3.36 + 04-04 18:03:47 | [850][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0228 ntime: 0073 mem: 3.36 + 04-04 18:03:54 | [850][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0104 ntime: 0077 mem: 3.36 + 04-04 18:03:58 | Time info >>>> elapsed: 1240.60 mins remain: 217.21 mins + 04-04 18:03:59 | [851][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0465 ntime: 0080 mem: 3.36 + 04-04 18:04:04 | [851][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0090 ntime: 0079 mem: 3.36 + 04-04 18:04:11 | [851][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0421 ntime: 0080 mem: 3.36 + 04-04 18:04:18 | [851][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0956 ntime: 0080 mem: 3.36 + 04-04 18:04:24 | [851][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0087 ntime: 0090 mem: 3.36 + 04-04 18:04:29 | [851][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0062 mem: 3.36 + 04-04 18:04:36 | [851][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0766 ntime: 0080 mem: 3.36 + 04-04 18:04:41 | [851][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0097 ntime: 0078 mem: 3.36 + 04-04 18:04:46 | [851][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0090 ntime: 0075 mem: 3.36 + 04-04 18:04:50 | [851][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0087 mem: 3.36 + 04-04 18:04:57 | [851][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0091 ntime: 0078 mem: 3.36 + 04-04 18:05:02 | [851][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0151 ntime: 0079 mem: 3.36 + 04-04 18:05:10 | [851][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0068 ntime: 0078 mem: 3.36 + 04-04 18:05:15 | [851][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0087 ntime: 0072 mem: 3.36 + 04-04 18:05:24 | [851][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0091 mem: 3.36 + 04-04 18:05:32 | [851][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0748 ntime: 0077 mem: 3.36 + 04-04 18:05:38 | [851][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0859 ntime: 0091 mem: 3.36 + 04-04 18:05:43 | [851][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0195 ntime: 0082 mem: 3.36 + 04-04 18:05:49 | Time info >>>> elapsed: 1242.45 mins remain: 215.82 mins + 04-04 18:05:49 | [852][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0061 ntime: 0084 mem: 3.36 + 04-04 18:05:56 | [852][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0125 ntime: 0078 mem: 3.36 + 04-04 18:06:04 | [852][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 18:06:10 | [852][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0102 ntime: 0082 mem: 3.36 + 04-04 18:06:16 | [852][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0952 ntime: 0077 mem: 3.36 + 04-04 18:06:21 | [852][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0141 ntime: 0074 mem: 3.36 + 04-04 18:06:26 | [852][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0771 ntime: 0084 mem: 3.36 + 04-04 18:06:33 | [852][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0102 ntime: 0077 mem: 3.36 + 04-04 18:06:40 | [852][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0142 ntime: 0083 mem: 3.36 + 04-04 18:06:47 | [852][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1055 ntime: 0082 mem: 3.36 + 04-04 18:06:54 | [852][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0129 ntime: 0079 mem: 3.36 + 04-04 18:07:00 | [852][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0120 ntime: 0076 mem: 3.36 + 04-04 18:07:06 | [852][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0129 ntime: 0079 mem: 3.36 + 04-04 18:07:11 | [852][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1470 ntime: 0084 mem: 3.36 + 04-04 18:07:17 | [852][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0654 ntime: 0080 mem: 3.36 + 04-04 18:07:25 | [852][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0744 ntime: 0086 mem: 3.36 + 04-04 18:07:32 | [852][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0479 ntime: 0081 mem: 3.36 + 04-04 18:07:37 | [852][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1187 ntime: 0082 mem: 3.36 + 04-04 18:07:43 | Time info >>>> elapsed: 1244.34 mins remain: 214.44 mins + 04-04 18:07:43 | [853][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0404 ntime: 0078 mem: 3.36 + 04-04 18:07:51 | [853][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1324 ntime: 0079 mem: 3.36 + 04-04 18:07:57 | [853][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1404 ntime: 0075 mem: 3.36 + 04-04 18:08:03 | [853][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0442 ntime: 0087 mem: 3.36 + 04-04 18:08:09 | [853][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0898 ntime: 0077 mem: 3.36 + 04-04 18:08:16 | [853][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0738 ntime: 0086 mem: 3.36 + 04-04 18:08:24 | [853][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0172 ntime: 0081 mem: 3.36 + 04-04 18:08:31 | [853][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0147 ntime: 0082 mem: 3.36 + 04-04 18:08:37 | [853][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0271 ntime: 0082 mem: 3.36 + 04-04 18:08:44 | [853][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0680 ntime: 0086 mem: 3.36 + 04-04 18:08:51 | [853][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0079 mem: 3.36 + 04-04 18:08:57 | [853][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0067 ntime: 0086 mem: 3.36 + 04-04 18:09:04 | [853][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0293 ntime: 0085 mem: 3.36 + 04-04 18:09:09 | [853][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0127 ntime: 0080 mem: 3.36 + 04-04 18:09:15 | [853][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0083 mem: 3.36 + 04-04 18:09:24 | [853][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1504 ntime: 0083 mem: 3.36 + 04-04 18:09:30 | [853][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1269 ntime: 0084 mem: 3.36 + 04-04 18:09:37 | [853][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0129 ntime: 0080 mem: 3.36 + 04-04 18:09:41 | Time info >>>> elapsed: 1246.31 mins remain: 213.07 mins + 04-04 18:09:42 | [854][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0999 ntime: 0083 mem: 3.36 + 04-04 18:09:49 | [854][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0161 ntime: 0081 mem: 3.36 + 04-04 18:09:56 | [854][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0104 ntime: 0083 mem: 3.36 + 04-04 18:10:03 | [854][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0084 mem: 3.36 + 04-04 18:10:10 | [854][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1097 ntime: 0080 mem: 3.36 + 04-04 18:10:15 | [854][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0626 ntime: 0080 mem: 3.36 + 04-04 18:10:21 | [854][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0677 ntime: 0077 mem: 3.36 + 04-04 18:10:29 | [854][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1006 ntime: 0081 mem: 3.36 + 04-04 18:10:36 | [854][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1200 ntime: 0085 mem: 3.36 + 04-04 18:10:41 | [854][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1331 ntime: 0072 mem: 3.36 + 04-04 18:10:47 | [854][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0764 ntime: 0086 mem: 3.36 + 04-04 18:10:54 | [854][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1050 ntime: 0076 mem: 3.36 + 04-04 18:11:01 | [854][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0093 ntime: 0086 mem: 3.36 + 04-04 18:11:07 | [854][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0983 ntime: 0080 mem: 3.36 + 04-04 18:11:14 | [854][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0481 ntime: 0080 mem: 3.36 + 04-04 18:11:19 | [854][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0943 ntime: 0080 mem: 3.36 + 04-04 18:11:26 | [854][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0678 ntime: 0082 mem: 3.36 + 04-04 18:11:34 | [854][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0159 ntime: 0080 mem: 3.36 + 04-04 18:11:39 | Time info >>>> elapsed: 1248.28 mins remain: 211.70 mins + 04-04 18:11:39 | [855][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0118 ntime: 0078 mem: 3.36 + 04-04 18:11:46 | [855][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0750 ntime: 0079 mem: 3.36 + 04-04 18:11:52 | [855][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0387 ntime: 0075 mem: 3.36 + 04-04 18:11:58 | [855][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0908 ntime: 0076 mem: 3.36 + 04-04 18:12:06 | [855][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0978 ntime: 0079 mem: 3.36 + 04-04 18:12:13 | [855][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1282 ntime: 0086 mem: 3.36 + 04-04 18:12:18 | [855][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0126 ntime: 0079 mem: 3.36 + 04-04 18:12:24 | [855][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0098 mem: 3.36 + 04-04 18:12:30 | [855][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0085 mem: 3.36 + 04-04 18:12:36 | [855][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0109 ntime: 0079 mem: 3.36 + 04-04 18:12:40 | [855][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0095 ntime: 0082 mem: 3.36 + 04-04 18:12:46 | [855][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1455 ntime: 0072 mem: 3.36 + 04-04 18:12:56 | [855][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1295 ntime: 0083 mem: 3.36 + 04-04 18:13:02 | [855][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0768 ntime: 0081 mem: 3.36 + 04-04 18:13:09 | [855][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0403 ntime: 0079 mem: 3.36 + 04-04 18:13:14 | [855][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0082 mem: 3.36 + 04-04 18:13:20 | [855][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0082 mem: 3.36 + 04-04 18:13:26 | [855][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0992 ntime: 0057 mem: 3.36 + 04-04 18:13:33 | Time info >>>> elapsed: 1250.18 mins remain: 210.31 mins + 04-04 18:13:33 | [856][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0081 mem: 3.36 + 04-04 18:13:40 | [856][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 18:13:48 | [856][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0725 ntime: 0079 mem: 3.36 + 04-04 18:13:55 | [856][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1039 ntime: 0079 mem: 3.36 + 04-04 18:14:00 | [856][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0630 ntime: 0083 mem: 3.36 + 04-04 18:14:07 | [856][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0974 ntime: 0084 mem: 3.36 + 04-04 18:14:15 | [856][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0097 ntime: 0083 mem: 3.36 + 04-04 18:14:21 | [856][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0950 ntime: 0081 mem: 3.36 + 04-04 18:14:27 | [856][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0883 ntime: 0083 mem: 3.36 + 04-04 18:14:33 | [856][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0874 ntime: 0081 mem: 3.36 + 04-04 18:14:38 | [856][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0084 mem: 3.36 + 04-04 18:14:45 | [856][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0137 ntime: 0081 mem: 3.36 + 04-04 18:14:51 | [856][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0091 ntime: 0080 mem: 3.36 + 04-04 18:14:55 | [856][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0063 ntime: 0088 mem: 3.36 + 04-04 18:15:02 | [856][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0090 ntime: 0091 mem: 3.36 + 04-04 18:15:08 | [856][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0148 ntime: 0077 mem: 3.36 + 04-04 18:15:15 | [856][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0125 ntime: 0079 mem: 3.36 + 04-04 18:15:21 | [856][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0184 ntime: 0082 mem: 3.36 + 04-04 18:15:27 | Time info >>>> elapsed: 1252.07 mins remain: 208.92 mins + 04-04 18:15:27 | [857][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0608 ntime: 0085 mem: 3.36 + 04-04 18:15:36 | [857][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1542 ntime: 0077 mem: 3.36 + 04-04 18:15:41 | [857][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0791 ntime: 0079 mem: 3.36 + 04-04 18:15:46 | [857][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0080 mem: 3.36 + 04-04 18:15:53 | [857][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0101 ntime: 0082 mem: 3.36 + 04-04 18:16:01 | [857][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1622 ntime: 0074 mem: 3.36 + 04-04 18:16:06 | [857][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0900 ntime: 0080 mem: 3.36 + 04-04 18:16:13 | [857][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1168 ntime: 0080 mem: 3.36 + 04-04 18:16:18 | [857][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0083 mem: 3.36 + 04-04 18:16:26 | [857][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0761 ntime: 0079 mem: 3.36 + 04-04 18:16:33 | [857][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0754 ntime: 0074 mem: 3.36 + 04-04 18:16:41 | [857][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0081 mem: 3.36 + 04-04 18:16:49 | [857][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1182 ntime: 0084 mem: 3.36 + 04-04 18:16:55 | [857][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0152 ntime: 0077 mem: 3.36 + 04-04 18:17:01 | [857][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0091 ntime: 0079 mem: 3.36 + 04-04 18:17:08 | [857][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0085 mem: 3.36 + 04-04 18:17:15 | [857][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1494 ntime: 0083 mem: 3.36 + 04-04 18:17:22 | [857][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0115 ntime: 0081 mem: 3.36 + 04-04 18:17:26 | Time info >>>> elapsed: 1254.06 mins remain: 207.55 mins + 04-04 18:17:26 | [858][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0099 ntime: 0079 mem: 3.36 + 04-04 18:17:31 | [858][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0129 ntime: 0078 mem: 3.36 + 04-04 18:17:38 | [858][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1400 ntime: 0081 mem: 3.36 + 04-04 18:17:43 | [858][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0071 ntime: 0077 mem: 3.36 + 04-04 18:17:50 | [858][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0647 ntime: 0084 mem: 3.36 + 04-04 18:17:56 | [858][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0081 mem: 3.36 + 04-04 18:18:03 | [858][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0907 ntime: 0084 mem: 3.36 + 04-04 18:18:10 | [858][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0412 ntime: 0080 mem: 3.36 + 04-04 18:18:16 | [858][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0599 ntime: 0082 mem: 3.36 + 04-04 18:18:20 | [858][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0889 ntime: 0084 mem: 3.36 + 04-04 18:18:28 | [858][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1146 ntime: 0083 mem: 3.36 + 04-04 18:18:35 | [858][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1248 ntime: 0077 mem: 3.36 + 04-04 18:18:43 | [858][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0096 ntime: 0073 mem: 3.36 + 04-04 18:18:51 | [858][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0092 ntime: 0084 mem: 3.36 + 04-04 18:18:58 | [858][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0122 ntime: 0083 mem: 3.36 + 04-04 18:19:02 | [858][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0074 mem: 3.36 + 04-04 18:19:07 | [858][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0088 mem: 3.36 + 04-04 18:19:13 | [858][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1036 ntime: 0093 mem: 3.36 + 04-04 18:19:17 | Time info >>>> elapsed: 1255.91 mins remain: 206.15 mins + 04-04 18:19:17 | [859][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0099 ntime: 0076 mem: 3.36 + 04-04 18:19:25 | [859][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0773 ntime: 0085 mem: 3.36 + 04-04 18:19:31 | [859][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1092 ntime: 0085 mem: 3.36 + 04-04 18:19:39 | [859][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1166 ntime: 0078 mem: 3.36 + 04-04 18:19:44 | [859][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1097 ntime: 0076 mem: 3.36 + 04-04 18:19:51 | [859][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0946 ntime: 0084 mem: 3.36 + 04-04 18:19:58 | [859][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1091 ntime: 0077 mem: 3.36 + 04-04 18:20:04 | [859][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0400 ntime: 0078 mem: 3.36 + 04-04 18:20:10 | [859][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1136 ntime: 0077 mem: 3.36 + 04-04 18:20:15 | [859][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0778 ntime: 0074 mem: 3.36 + 04-04 18:20:20 | [859][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0087 ntime: 0078 mem: 3.36 + 04-04 18:20:26 | [859][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0696 ntime: 0079 mem: 3.36 + 04-04 18:20:36 | [859][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0165 ntime: 0076 mem: 3.36 + 04-04 18:20:43 | [859][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1001 ntime: 0062 mem: 3.36 + 04-04 18:20:49 | [859][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0964 ntime: 0079 mem: 3.36 + 04-04 18:20:56 | [859][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0132 ntime: 0083 mem: 3.36 + 04-04 18:21:03 | [859][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0525 ntime: 0082 mem: 3.36 + 04-04 18:21:10 | [859][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0805 ntime: 0079 mem: 3.36 + 04-04 18:21:15 | Time info >>>> elapsed: 1257.88 mins remain: 204.77 mins + 04-04 18:21:17 | [860][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 1438 ntime: 0072 mem: 3.36 + 04-04 18:21:24 | [860][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0086 mem: 3.36 + 04-04 18:21:31 | [860][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0092 ntime: 0078 mem: 3.36 + 04-04 18:21:39 | [860][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1032 ntime: 0084 mem: 3.36 + 04-04 18:21:46 | [860][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1301 ntime: 0076 mem: 3.36 + 04-04 18:21:54 | [860][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1602 ntime: 0071 mem: 3.36 + 04-04 18:22:02 | [860][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0621 ntime: 0083 mem: 3.36 + 04-04 18:22:10 | [860][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0102 ntime: 0081 mem: 3.36 + 04-04 18:22:16 | [860][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0587 ntime: 0079 mem: 3.36 + 04-04 18:22:22 | [860][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0526 ntime: 0078 mem: 3.36 + 04-04 18:22:28 | [860][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0168 ntime: 0081 mem: 3.36 + 04-04 18:22:35 | [860][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0567 ntime: 0087 mem: 3.36 + 04-04 18:22:40 | [860][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0114 ntime: 0072 mem: 3.36 + 04-04 18:22:49 | [860][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0110 ntime: 0075 mem: 3.36 + 04-04 18:22:54 | [860][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0105 ntime: 0082 mem: 3.36 + 04-04 18:23:01 | [860][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0814 ntime: 0075 mem: 3.36 + 04-04 18:23:08 | [860][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1298 ntime: 0075 mem: 3.36 + 04-04 18:23:14 | [860][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0167 ntime: 0084 mem: 3.36 + 04-04 18:23:21 | Time info >>>> elapsed: 1259.97 mins remain: 203.41 mins + 04-04 18:23:21 | [861][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0467 ntime: 0077 mem: 3.36 + 04-04 18:23:27 | [861][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0656 ntime: 0083 mem: 3.36 + 04-04 18:23:32 | [861][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 18:23:36 | [861][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0114 ntime: 0080 mem: 3.36 + 04-04 18:23:42 | [861][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0078 mem: 3.36 + 04-04 18:23:48 | [861][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0099 ntime: 0087 mem: 3.36 + 04-04 18:23:55 | [861][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0115 ntime: 0079 mem: 3.36 + 04-04 18:24:03 | [861][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0971 ntime: 0078 mem: 3.36 + 04-04 18:24:09 | [861][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0612 ntime: 0071 mem: 3.36 + 04-04 18:24:13 | [861][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0075 mem: 3.36 + 04-04 18:24:19 | [861][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1419 ntime: 0081 mem: 3.36 + 04-04 18:24:24 | [861][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1298 ntime: 0082 mem: 3.36 + 04-04 18:24:31 | [861][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0084 ntime: 0084 mem: 3.36 + 04-04 18:24:39 | [861][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1205 ntime: 0074 mem: 3.36 + 04-04 18:24:45 | [861][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0812 ntime: 0081 mem: 3.36 + 04-04 18:24:50 | [861][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0866 ntime: 0078 mem: 3.36 + 04-04 18:24:57 | [861][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0074 mem: 3.36 + 04-04 18:25:05 | [861][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1359 ntime: 0075 mem: 3.36 + 04-04 18:25:11 | Time info >>>> elapsed: 1261.81 mins remain: 202.01 mins + 04-04 18:25:12 | [862][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0790 ntime: 0077 mem: 3.36 + 04-04 18:25:20 | [862][010/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0977 ntime: 0073 mem: 3.36 + 04-04 18:25:23 | [862][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0072 mem: 3.36 + 04-04 18:25:30 | [862][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0485 ntime: 0082 mem: 3.36 + 04-04 18:25:37 | [862][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 18:25:44 | [862][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0690 ntime: 0084 mem: 3.36 + 04-04 18:25:50 | [862][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0905 ntime: 0082 mem: 3.36 + 04-04 18:25:57 | [862][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0742 ntime: 0078 mem: 3.36 + 04-04 18:26:04 | [862][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1129 ntime: 0078 mem: 3.36 + 04-04 18:26:11 | [862][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0080 mem: 3.36 + 04-04 18:26:19 | [862][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1016 ntime: 0084 mem: 3.36 + 04-04 18:26:24 | [862][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0986 ntime: 0082 mem: 3.36 + 04-04 18:26:32 | [862][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0621 ntime: 0083 mem: 3.36 + 04-04 18:26:38 | [862][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1306 ntime: 0078 mem: 3.36 + 04-04 18:26:45 | [862][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0480 ntime: 0086 mem: 3.36 + 04-04 18:26:51 | [862][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1058 ntime: 0083 mem: 3.36 + 04-04 18:26:57 | [862][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0535 ntime: 0083 mem: 3.36 + 04-04 18:27:04 | [862][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0637 ntime: 0078 mem: 3.36 + 04-04 18:27:09 | Time info >>>> elapsed: 1263.77 mins remain: 200.62 mins + 04-04 18:27:09 | [863][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0760 ntime: 0079 mem: 3.36 + 04-04 18:27:16 | [863][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0785 ntime: 0081 mem: 3.36 + 04-04 18:27:23 | [863][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0424 ntime: 0077 mem: 3.36 + 04-04 18:27:30 | [863][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1262 ntime: 0082 mem: 3.36 + 04-04 18:27:37 | [863][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1075 ntime: 0078 mem: 3.36 + 04-04 18:27:44 | [863][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0728 ntime: 0080 mem: 3.36 + 04-04 18:27:50 | [863][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0507 ntime: 0081 mem: 3.36 + 04-04 18:27:55 | [863][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0977 ntime: 0086 mem: 3.36 + 04-04 18:28:01 | [863][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0576 ntime: 0085 mem: 3.36 + 04-04 18:28:08 | [863][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0087 ntime: 0056 mem: 3.36 + 04-04 18:28:15 | [863][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0074 mem: 3.36 + 04-04 18:28:25 | [863][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0090 ntime: 0080 mem: 3.36 + 04-04 18:28:29 | [863][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0755 ntime: 0087 mem: 3.36 + 04-04 18:28:36 | [863][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0702 ntime: 0077 mem: 3.36 + 04-04 18:28:41 | [863][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 18:28:48 | [863][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0977 ntime: 0080 mem: 3.36 + 04-04 18:28:55 | [863][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0743 ntime: 0079 mem: 3.36 + 04-04 18:29:01 | [863][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0356 ntime: 0073 mem: 3.36 + 04-04 18:29:06 | Time info >>>> elapsed: 1265.73 mins remain: 199.24 mins + 04-04 18:29:07 | [864][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0645 ntime: 0084 mem: 3.36 + 04-04 18:29:13 | [864][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0399 ntime: 0079 mem: 3.36 + 04-04 18:29:19 | [864][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0343 ntime: 0077 mem: 3.36 + 04-04 18:29:24 | [864][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0078 ntime: 0084 mem: 3.36 + 04-04 18:29:30 | [864][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0101 ntime: 0079 mem: 3.36 + 04-04 18:29:36 | [864][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0086 mem: 3.36 + 04-04 18:29:43 | [864][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0668 ntime: 0081 mem: 3.36 + 04-04 18:29:48 | [864][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0069 ntime: 0074 mem: 3.36 + 04-04 18:29:53 | [864][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0076 mem: 3.36 + 04-04 18:30:00 | [864][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 18:30:08 | [864][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1205 ntime: 0080 mem: 3.36 + 04-04 18:30:15 | [864][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0114 ntime: 0079 mem: 3.36 + 04-04 18:30:22 | [864][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0398 ntime: 0082 mem: 3.36 + 04-04 18:30:28 | [864][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0107 ntime: 0082 mem: 3.36 + 04-04 18:30:34 | [864][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0918 ntime: 0078 mem: 3.36 + 04-04 18:30:39 | [864][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0087 mem: 3.36 + 04-04 18:30:45 | [864][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1409 ntime: 0085 mem: 3.36 + 04-04 18:30:50 | [864][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0355 ntime: 0077 mem: 3.36 + 04-04 18:30:55 | Time info >>>> elapsed: 1267.54 mins remain: 197.82 mins + 04-04 18:30:56 | [865][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0735 ntime: 0081 mem: 3.36 + 04-04 18:31:01 | [865][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0611 ntime: 0083 mem: 3.36 + 04-04 18:31:09 | [865][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0841 ntime: 0080 mem: 3.36 + 04-04 18:31:15 | [865][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0701 ntime: 0079 mem: 3.36 + 04-04 18:31:21 | [865][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0065 ntime: 0077 mem: 3.36 + 04-04 18:31:28 | [865][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0542 ntime: 0079 mem: 3.36 + 04-04 18:31:35 | [865][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1428 ntime: 0075 mem: 3.36 + 04-04 18:31:41 | [865][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0094 ntime: 0079 mem: 3.36 + 04-04 18:31:49 | [865][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0100 ntime: 0077 mem: 3.36 + 04-04 18:31:54 | [865][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0076 mem: 3.36 + 04-04 18:32:01 | [865][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1750 ntime: 0081 mem: 3.36 + 04-04 18:32:09 | [865][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0308 ntime: 0078 mem: 3.36 + 04-04 18:32:17 | [865][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0093 ntime: 0082 mem: 3.36 + 04-04 18:32:24 | [865][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0072 ntime: 0079 mem: 3.36 + 04-04 18:32:31 | [865][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1013 ntime: 0078 mem: 3.36 + 04-04 18:32:37 | [865][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0122 ntime: 0080 mem: 3.36 + 04-04 18:32:44 | [865][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0150 ntime: 0087 mem: 3.36 + 04-04 18:32:51 | [865][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1089 ntime: 0083 mem: 3.36 + 04-04 18:32:56 | Time info >>>> elapsed: 1269.57 mins remain: 196.45 mins + 04-04 18:32:58 | [866][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1324 ntime: 0077 mem: 3.36 + 04-04 18:33:02 | [866][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0149 ntime: 0078 mem: 3.36 + 04-04 18:33:08 | [866][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0080 mem: 3.36 + 04-04 18:33:15 | [866][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0139 ntime: 0080 mem: 3.36 + 04-04 18:33:21 | [866][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1066 ntime: 0078 mem: 3.36 + 04-04 18:33:27 | [866][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0080 mem: 3.36 + 04-04 18:33:32 | [866][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1061 ntime: 0082 mem: 3.36 + 04-04 18:33:39 | [866][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1051 ntime: 0079 mem: 3.36 + 04-04 18:33:45 | [866][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0079 mem: 3.36 + 04-04 18:33:52 | [866][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0082 mem: 3.36 + 04-04 18:33:57 | [866][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0125 ntime: 0082 mem: 3.36 + 04-04 18:34:03 | [866][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0483 ntime: 0080 mem: 3.36 + 04-04 18:34:09 | [866][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1214 ntime: 0084 mem: 3.36 + 04-04 18:34:16 | [866][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 18:34:22 | [866][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0088 mem: 3.36 + 04-04 18:34:29 | [866][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0088 ntime: 0079 mem: 3.36 + 04-04 18:34:35 | [866][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0125 ntime: 0082 mem: 3.36 + 04-04 18:34:40 | [866][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0080 mem: 3.36 + 04-04 18:34:46 | Time info >>>> elapsed: 1271.40 mins remain: 195.04 mins + 04-04 18:34:46 | [867][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0157 ntime: 0079 mem: 3.36 + 04-04 18:34:53 | [867][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0077 ntime: 0088 mem: 3.36 + 04-04 18:35:00 | [867][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0844 ntime: 0087 mem: 3.36 + 04-04 18:35:06 | [867][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1532 ntime: 0081 mem: 3.36 + 04-04 18:35:12 | [867][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1138 ntime: 0079 mem: 3.36 + 04-04 18:35:18 | [867][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0639 ntime: 0079 mem: 3.36 + 04-04 18:35:24 | [867][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0402 ntime: 0082 mem: 3.36 + 04-04 18:35:29 | [867][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0083 ntime: 0081 mem: 3.36 + 04-04 18:35:36 | [867][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0142 ntime: 0086 mem: 3.36 + 04-04 18:35:41 | [867][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0774 ntime: 0081 mem: 3.36 + 04-04 18:35:47 | [867][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0616 ntime: 0080 mem: 3.36 + 04-04 18:35:54 | [867][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 18:36:01 | [867][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0075 ntime: 0088 mem: 3.36 + 04-04 18:36:06 | [867][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0654 ntime: 0081 mem: 3.36 + 04-04 18:36:13 | [867][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0810 ntime: 0083 mem: 3.36 + 04-04 18:36:20 | [867][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0977 ntime: 0086 mem: 3.36 + 04-04 18:36:27 | [867][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0063 ntime: 0086 mem: 3.36 + 04-04 18:36:34 | [867][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0082 ntime: 0079 mem: 3.36 + 04-04 18:36:41 | Time info >>>> elapsed: 1273.30 mins remain: 193.64 mins + 04-04 18:36:41 | [868][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0121 ntime: 0085 mem: 3.36 + 04-04 18:36:46 | [868][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0223 ntime: 0080 mem: 3.36 + 04-04 18:36:52 | [868][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0093 ntime: 0060 mem: 3.36 + 04-04 18:36:59 | [868][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0513 ntime: 0084 mem: 3.36 + 04-04 18:37:06 | [868][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0152 ntime: 0079 mem: 3.36 + 04-04 18:37:13 | [868][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0692 ntime: 0080 mem: 3.36 + 04-04 18:37:18 | [868][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0121 ntime: 0081 mem: 3.36 + 04-04 18:37:25 | [868][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1517 ntime: 0087 mem: 3.36 + 04-04 18:37:30 | [868][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0134 ntime: 0077 mem: 3.36 + 04-04 18:37:36 | [868][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0071 ntime: 0083 mem: 3.36 + 04-04 18:37:43 | [868][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0095 ntime: 0092 mem: 3.36 + 04-04 18:37:49 | [868][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0882 ntime: 0083 mem: 3.36 + 04-04 18:37:55 | [868][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0084 mem: 3.36 + 04-04 18:38:02 | [868][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1412 ntime: 0078 mem: 3.36 + 04-04 18:38:08 | [868][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0565 ntime: 0081 mem: 3.36 + 04-04 18:38:15 | [868][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0135 ntime: 0085 mem: 3.36 + 04-04 18:38:23 | [868][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1193 ntime: 0079 mem: 3.36 + 04-04 18:38:28 | [868][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0899 ntime: 0079 mem: 3.36 + 04-04 18:38:33 | Time info >>>> elapsed: 1275.18 mins remain: 192.23 mins + 04-04 18:38:34 | [869][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0086 ntime: 0081 mem: 3.36 + 04-04 18:38:40 | [869][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0097 ntime: 0086 mem: 3.36 + 04-04 18:38:46 | [869][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0769 ntime: 0080 mem: 3.36 + 04-04 18:38:51 | [869][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0081 mem: 3.36 + 04-04 18:39:00 | [869][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0717 ntime: 0088 mem: 3.36 + 04-04 18:39:07 | [869][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1472 ntime: 0085 mem: 3.36 + 04-04 18:39:12 | [869][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0569 ntime: 0079 mem: 3.36 + 04-04 18:39:19 | [869][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1451 ntime: 0082 mem: 3.36 + 04-04 18:39:26 | [869][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1449 ntime: 0081 mem: 3.36 + 04-04 18:39:32 | [869][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0097 ntime: 0077 mem: 3.36 + 04-04 18:39:36 | [869][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0106 ntime: 0079 mem: 3.36 + 04-04 18:39:43 | [869][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0419 ntime: 0082 mem: 3.36 + 04-04 18:39:51 | [869][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0198 ntime: 0076 mem: 3.36 + 04-04 18:39:58 | [869][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1144 ntime: 0076 mem: 3.36 + 04-04 18:40:05 | [869][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0103 ntime: 0081 mem: 3.36 + 04-04 18:40:12 | [869][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0886 ntime: 0081 mem: 3.36 + 04-04 18:40:17 | [869][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0476 ntime: 0080 mem: 3.36 + 04-04 18:40:23 | [869][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0615 ntime: 0085 mem: 3.36 + 04-04 18:40:26 | Time info >>>> elapsed: 1277.07 mins remain: 190.83 mins + 04-04 18:40:27 | [870][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1060 ntime: 0079 mem: 3.36 + 04-04 18:40:35 | [870][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1252 ntime: 0080 mem: 3.36 + 04-04 18:40:42 | [870][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0546 ntime: 0081 mem: 3.36 + 04-04 18:40:48 | [870][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0464 ntime: 0085 mem: 3.36 + 04-04 18:40:53 | [870][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0138 ntime: 0086 mem: 3.36 + 04-04 18:41:01 | [870][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0103 ntime: 0078 mem: 3.36 + 04-04 18:41:08 | [870][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0712 ntime: 0085 mem: 3.36 + 04-04 18:41:13 | [870][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0072 mem: 3.36 + 04-04 18:41:20 | [870][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0940 ntime: 0083 mem: 3.36 + 04-04 18:41:26 | [870][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1231 ntime: 0079 mem: 3.36 + 04-04 18:41:32 | [870][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0119 ntime: 0083 mem: 3.36 + 04-04 18:41:39 | [870][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0119 ntime: 0083 mem: 3.36 + 04-04 18:41:47 | [870][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0076 ntime: 0088 mem: 3.36 + 04-04 18:41:51 | [870][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0160 ntime: 0079 mem: 3.36 + 04-04 18:41:58 | [870][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0635 ntime: 0088 mem: 3.36 + 04-04 18:42:03 | [870][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0076 ntime: 0076 mem: 3.36 + 04-04 18:42:10 | [870][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0111 ntime: 0078 mem: 3.36 + 04-04 18:42:15 | [870][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0185 ntime: 0078 mem: 3.36 + 04-04 18:42:20 | Time info >>>> elapsed: 1278.96 mins remain: 189.42 mins + 04-04 18:42:21 | [871][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1447 ntime: 0074 mem: 3.36 + 04-04 18:42:27 | [871][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0151 ntime: 0081 mem: 3.36 + 04-04 18:42:37 | [871][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0559 ntime: 0080 mem: 3.36 + 04-04 18:42:41 | [871][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0242 ntime: 0079 mem: 3.36 + 04-04 18:42:48 | [871][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0075 mem: 3.36 + 04-04 18:42:54 | [871][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0993 ntime: 0078 mem: 3.36 + 04-04 18:42:58 | [871][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0079 mem: 3.36 + 04-04 18:43:05 | [871][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1051 ntime: 0077 mem: 3.36 + 04-04 18:43:10 | [871][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0077 ntime: 0077 mem: 3.36 + 04-04 18:43:16 | [871][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1332 ntime: 0080 mem: 3.36 + 04-04 18:43:23 | [871][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0958 ntime: 0077 mem: 3.36 + 04-04 18:43:28 | [871][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0082 mem: 3.36 + 04-04 18:43:34 | [871][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0103 ntime: 0080 mem: 3.36 + 04-04 18:43:41 | [871][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0141 ntime: 0079 mem: 3.36 + 04-04 18:43:48 | [871][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0082 mem: 3.36 + 04-04 18:43:55 | [871][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0091 ntime: 0078 mem: 3.36 + 04-04 18:44:01 | [871][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0994 ntime: 0083 mem: 3.36 + 04-04 18:44:06 | [871][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0290 ntime: 0080 mem: 3.36 + 04-04 18:44:10 | Time info >>>> elapsed: 1280.80 mins remain: 188.01 mins + 04-04 18:44:11 | [872][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0537 ntime: 0081 mem: 3.36 + 04-04 18:44:16 | [872][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0568 ntime: 0090 mem: 3.36 + 04-04 18:44:21 | [872][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0666 ntime: 0085 mem: 3.36 + 04-04 18:44:26 | [872][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0079 mem: 3.36 + 04-04 18:44:33 | [872][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0091 ntime: 0079 mem: 3.36 + 04-04 18:44:40 | [872][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0591 ntime: 0081 mem: 3.36 + 04-04 18:44:47 | [872][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0082 mem: 3.36 + 04-04 18:44:53 | [872][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0079 ntime: 0081 mem: 3.36 + 04-04 18:44:58 | [872][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1010 ntime: 0078 mem: 3.36 + 04-04 18:45:03 | [872][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0072 ntime: 0087 mem: 3.36 + 04-04 18:45:07 | [872][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0086 ntime: 0082 mem: 3.36 + 04-04 18:45:14 | [872][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0364 ntime: 0084 mem: 3.36 + 04-04 18:45:20 | [872][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1277 ntime: 0077 mem: 3.36 + 04-04 18:45:25 | [872][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0068 ntime: 0079 mem: 3.36 + 04-04 18:45:32 | [872][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0106 ntime: 0089 mem: 3.36 + 04-04 18:45:37 | [872][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 18:45:44 | [872][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1134 ntime: 0080 mem: 3.36 + 04-04 18:45:50 | [872][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0074 mem: 3.36 + 04-04 18:45:55 | Time info >>>> elapsed: 1282.54 mins remain: 186.58 mins + 04-04 18:45:56 | [873][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 1116 ntime: 0076 mem: 3.36 + 04-04 18:46:03 | [873][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0068 ntime: 0080 mem: 3.36 + 04-04 18:46:10 | [873][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0084 mem: 3.36 + 04-04 18:46:17 | [873][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0323 ntime: 0082 mem: 3.36 + 04-04 18:46:23 | [873][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0072 ntime: 0075 mem: 3.36 + 04-04 18:46:28 | [873][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0661 ntime: 0076 mem: 3.36 + 04-04 18:46:33 | [873][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1154 ntime: 0084 mem: 3.36 + 04-04 18:46:41 | [873][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1351 ntime: 0084 mem: 3.36 + 04-04 18:46:46 | [873][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0147 ntime: 0082 mem: 3.36 + 04-04 18:46:53 | [873][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0263 ntime: 0082 mem: 3.36 + 04-04 18:47:00 | [873][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0900 ntime: 0081 mem: 3.36 + 04-04 18:47:07 | [873][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0080 ntime: 0085 mem: 3.36 + 04-04 18:47:14 | [873][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 18:47:20 | [873][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0080 mem: 3.36 + 04-04 18:47:33 | [873][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0906 ntime: 0085 mem: 3.36 + 04-04 18:47:39 | [873][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0072 ntime: 0078 mem: 3.36 + 04-04 18:47:44 | [873][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0075 ntime: 0080 mem: 3.36 + 04-04 18:47:51 | [873][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1267 ntime: 0077 mem: 3.36 + 04-04 18:47:56 | Time info >>>> elapsed: 1284.56 mins remain: 185.19 mins + 04-04 18:47:56 | [874][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0155 ntime: 0081 mem: 3.36 + 04-04 18:48:01 | [874][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0109 ntime: 0088 mem: 3.36 + 04-04 18:48:07 | [874][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 18:48:16 | [874][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1365 ntime: 0080 mem: 3.36 + 04-04 18:48:21 | [874][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0105 ntime: 0078 mem: 3.36 + 04-04 18:48:26 | [874][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0152 ntime: 0079 mem: 3.36 + 04-04 18:48:34 | [874][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0135 ntime: 0082 mem: 3.36 + 04-04 18:48:39 | [874][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 18:48:49 | [874][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1078 ntime: 0100 mem: 3.36 + 04-04 18:48:54 | [874][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0074 mem: 3.36 + 04-04 18:49:02 | [874][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1028 ntime: 0084 mem: 3.36 + 04-04 18:49:08 | [874][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0118 ntime: 0076 mem: 3.36 + 04-04 18:49:15 | [874][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0307 ntime: 0078 mem: 3.36 + 04-04 18:49:22 | [874][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0396 ntime: 0084 mem: 3.36 + 04-04 18:49:26 | [874][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0094 ntime: 0077 mem: 3.36 + 04-04 18:49:34 | [874][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0775 ntime: 0078 mem: 3.36 + 04-04 18:49:41 | [874][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0891 ntime: 0086 mem: 3.36 + 04-04 18:49:47 | [874][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0907 ntime: 0081 mem: 3.36 + 04-04 18:49:52 | Time info >>>> elapsed: 1286.49 mins remain: 183.78 mins + 04-04 18:49:53 | [875][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1019 ntime: 0079 mem: 3.36 + 04-04 18:50:01 | [875][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1474 ntime: 0079 mem: 3.36 + 04-04 18:50:08 | [875][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0079 mem: 3.36 + 04-04 18:50:14 | [875][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1247 ntime: 0082 mem: 3.36 + 04-04 18:50:22 | [875][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0904 ntime: 0078 mem: 3.36 + 04-04 18:50:29 | [875][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1032 ntime: 0079 mem: 3.36 + 04-04 18:50:34 | [875][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1224 ntime: 0084 mem: 3.36 + 04-04 18:50:41 | [875][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0837 ntime: 0082 mem: 3.36 + 04-04 18:50:46 | [875][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0229 ntime: 0074 mem: 3.36 + 04-04 18:50:52 | [875][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1050 ntime: 0079 mem: 3.36 + 04-04 18:50:59 | [875][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1104 ntime: 0085 mem: 3.36 + 04-04 18:51:05 | [875][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0331 ntime: 0081 mem: 3.36 + 04-04 18:51:11 | [875][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0752 ntime: 0069 mem: 3.36 + 04-04 18:51:23 | [875][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2328 ntime: 0077 mem: 3.36 + 04-04 18:51:31 | [875][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1086 ntime: 0086 mem: 3.36 + 04-04 18:51:37 | [875][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0111 ntime: 0079 mem: 3.36 + 04-04 18:51:41 | [875][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0085 mem: 3.36 + 04-04 18:51:48 | [875][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1201 ntime: 0079 mem: 3.36 + 04-04 18:51:52 | Time info >>>> elapsed: 1288.49 mins remain: 182.39 mins + 04-04 18:51:53 | [876][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1148 ntime: 0081 mem: 3.36 + 04-04 18:52:01 | [876][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0077 mem: 3.36 + 04-04 18:52:08 | [876][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0753 ntime: 0081 mem: 3.36 + 04-04 18:52:13 | [876][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0237 ntime: 0076 mem: 3.36 + 04-04 18:52:18 | [876][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0117 ntime: 0083 mem: 3.36 + 04-04 18:52:24 | [876][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0744 ntime: 0083 mem: 3.36 + 04-04 18:52:31 | [876][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0079 mem: 3.36 + 04-04 18:52:38 | [876][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0545 ntime: 0087 mem: 3.36 + 04-04 18:52:44 | [876][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0077 mem: 3.36 + 04-04 18:52:50 | [876][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1039 ntime: 0083 mem: 3.36 + 04-04 18:52:55 | [876][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0105 ntime: 0080 mem: 3.36 + 04-04 18:53:01 | [876][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0114 ntime: 0079 mem: 3.36 + 04-04 18:53:08 | [876][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0569 ntime: 0084 mem: 3.36 + 04-04 18:53:15 | [876][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0867 ntime: 0076 mem: 3.36 + 04-04 18:53:21 | [876][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1299 ntime: 0082 mem: 3.36 + 04-04 18:53:27 | [876][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0693 ntime: 0081 mem: 3.36 + 04-04 18:53:32 | [876][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1042 ntime: 0074 mem: 3.36 + 04-04 18:53:37 | [876][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0133 ntime: 0078 mem: 3.36 + 04-04 18:53:44 | Time info >>>> elapsed: 1290.36 mins remain: 180.97 mins + 04-04 18:53:45 | [877][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0723 ntime: 0081 mem: 3.36 + 04-04 18:53:50 | [877][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1225 ntime: 0077 mem: 3.36 + 04-04 18:53:57 | [877][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0875 ntime: 0080 mem: 3.36 + 04-04 18:54:02 | [877][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 18:54:08 | [877][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0132 ntime: 0079 mem: 3.36 + 04-04 18:54:14 | [877][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0945 ntime: 0084 mem: 3.36 + 04-04 18:54:19 | [877][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0084 ntime: 0081 mem: 3.36 + 04-04 18:54:25 | [877][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0535 ntime: 0057 mem: 3.36 + 04-04 18:54:31 | [877][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0074 mem: 3.36 + 04-04 18:54:36 | [877][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0652 ntime: 0088 mem: 3.36 + 04-04 18:54:42 | [877][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0091 ntime: 0079 mem: 3.36 + 04-04 18:54:50 | [877][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0944 ntime: 0081 mem: 3.36 + 04-04 18:54:55 | [877][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0084 mem: 3.36 + 04-04 18:55:02 | [877][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0800 ntime: 0081 mem: 3.36 + 04-04 18:55:08 | [877][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0083 ntime: 0081 mem: 3.36 + 04-04 18:55:15 | [877][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0074 mem: 3.36 + 04-04 18:55:22 | [877][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0407 ntime: 0081 mem: 3.36 + 04-04 18:55:27 | [877][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1216 ntime: 0086 mem: 3.36 + 04-04 18:55:31 | Time info >>>> elapsed: 1292.14 mins remain: 179.55 mins + 04-04 18:55:32 | [878][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0894 ntime: 0079 mem: 3.36 + 04-04 18:55:38 | [878][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0112 ntime: 0083 mem: 3.36 + 04-04 18:55:47 | [878][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1134 ntime: 0077 mem: 3.36 + 04-04 18:55:54 | [878][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1354 ntime: 0078 mem: 3.36 + 04-04 18:56:01 | [878][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1130 ntime: 0076 mem: 3.36 + 04-04 18:56:06 | [878][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0764 ntime: 0080 mem: 3.36 + 04-04 18:56:12 | [878][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0967 ntime: 0084 mem: 3.36 + 04-04 18:56:18 | [878][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0111 ntime: 0082 mem: 3.36 + 04-04 18:56:25 | [878][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1300 ntime: 0083 mem: 3.36 + 04-04 18:56:31 | [878][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1321 ntime: 0084 mem: 3.36 + 04-04 18:56:37 | [878][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0075 ntime: 0084 mem: 3.36 + 04-04 18:56:42 | [878][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0547 ntime: 0083 mem: 3.36 + 04-04 18:56:49 | [878][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0122 ntime: 0086 mem: 3.36 + 04-04 18:56:56 | [878][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1458 ntime: 0080 mem: 3.36 + 04-04 18:57:02 | [878][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1197 ntime: 0069 mem: 3.36 + 04-04 18:57:09 | [878][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0936 ntime: 0079 mem: 3.36 + 04-04 18:57:15 | [878][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0077 mem: 3.36 + 04-04 18:57:22 | [878][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0727 ntime: 0085 mem: 3.36 + 04-04 18:57:28 | Time info >>>> elapsed: 1294.09 mins remain: 178.14 mins + 04-04 18:57:28 | [879][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0125 ntime: 0077 mem: 3.36 + 04-04 18:57:34 | [879][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0830 ntime: 0083 mem: 3.36 + 04-04 18:57:42 | [879][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0245 ntime: 0082 mem: 3.36 + 04-04 18:57:49 | [879][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0090 ntime: 0086 mem: 3.36 + 04-04 18:57:54 | [879][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0610 ntime: 0081 mem: 3.36 + 04-04 18:58:01 | [879][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0116 ntime: 0080 mem: 3.36 + 04-04 18:58:06 | [879][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1321 ntime: 0086 mem: 3.36 + 04-04 18:58:14 | [879][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0097 ntime: 0080 mem: 3.36 + 04-04 18:58:20 | [879][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0615 ntime: 0079 mem: 3.36 + 04-04 18:58:26 | [879][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0172 ntime: 0078 mem: 3.36 + 04-04 18:58:33 | [879][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0128 ntime: 0075 mem: 3.36 + 04-04 18:58:39 | [879][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0083 ntime: 0084 mem: 3.36 + 04-04 18:58:48 | [879][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0738 ntime: 0082 mem: 3.36 + 04-04 18:58:54 | [879][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0079 mem: 3.36 + 04-04 18:59:03 | [879][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1444 ntime: 0082 mem: 3.36 + 04-04 18:59:11 | [879][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0205 ntime: 0088 mem: 3.36 + 04-04 18:59:22 | [879][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2152 ntime: 0080 mem: 3.36 + 04-04 18:59:30 | [879][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0946 ntime: 0094 mem: 3.36 + 04-04 18:59:40 | Time info >>>> elapsed: 1296.30 mins remain: 176.77 mins + 04-04 18:59:40 | [880][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0079 mem: 3.36 + 04-04 18:59:50 | [880][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0088 ntime: 0080 mem: 3.36 + 04-04 18:59:58 | [880][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0931 ntime: 0085 mem: 3.36 + 04-04 19:00:03 | [880][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0892 ntime: 0078 mem: 3.36 + 04-04 19:00:09 | [880][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0193 ntime: 0077 mem: 3.36 + 04-04 19:00:14 | [880][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0726 ntime: 0081 mem: 3.36 + 04-04 19:00:20 | [880][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0994 ntime: 0085 mem: 3.36 + 04-04 19:00:27 | [880][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0894 ntime: 0083 mem: 3.36 + 04-04 19:00:34 | [880][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0331 ntime: 0081 mem: 3.36 + 04-04 19:00:39 | [880][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0099 ntime: 0082 mem: 3.36 + 04-04 19:00:47 | [880][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1261 ntime: 0079 mem: 3.36 + 04-04 19:00:56 | [880][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1302 ntime: 0074 mem: 3.36 + 04-04 19:01:03 | [880][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0129 ntime: 0076 mem: 3.36 + 04-04 19:01:11 | [880][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0547 ntime: 0088 mem: 3.36 + 04-04 19:01:21 | [880][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0159 ntime: 0078 mem: 3.36 + 04-04 19:01:27 | [880][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 19:01:34 | [880][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1050 ntime: 0081 mem: 3.36 + 04-04 19:01:40 | [880][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1295 ntime: 0073 mem: 3.36 + 04-04 19:01:46 | Time info >>>> elapsed: 1298.40 mins remain: 175.38 mins + 04-04 19:01:46 | [881][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0101 ntime: 0077 mem: 3.36 + 04-04 19:01:51 | [881][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0395 ntime: 0080 mem: 3.36 + 04-04 19:01:58 | [881][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0522 ntime: 0080 mem: 3.36 + 04-04 19:02:03 | [881][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0081 ntime: 0085 mem: 3.36 + 04-04 19:02:12 | [881][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1195 ntime: 0082 mem: 3.36 + 04-04 19:02:19 | [881][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1160 ntime: 0086 mem: 3.36 + 04-04 19:02:26 | [881][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1305 ntime: 0076 mem: 3.36 + 04-04 19:02:33 | [881][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0084 mem: 3.36 + 04-04 19:02:40 | [881][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1070 ntime: 0086 mem: 3.36 + 04-04 19:02:46 | [881][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0067 ntime: 0078 mem: 3.36 + 04-04 19:02:56 | [881][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1106 ntime: 0083 mem: 3.36 + 04-04 19:03:08 | [881][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1269 ntime: 0078 mem: 3.36 + 04-04 19:03:13 | [881][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0081 mem: 3.36 + 04-04 19:03:19 | [881][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0596 ntime: 0084 mem: 3.36 + 04-04 19:03:27 | [881][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1435 ntime: 0078 mem: 3.36 + 04-04 19:03:35 | [881][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 19:03:40 | [881][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0338 ntime: 0071 mem: 3.36 + 04-04 19:03:48 | [881][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0723 ntime: 0087 mem: 3.36 + 04-04 19:03:54 | Time info >>>> elapsed: 1300.52 mins remain: 173.99 mins + 04-04 19:03:54 | [882][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0082 ntime: 0083 mem: 3.36 + 04-04 19:03:59 | [882][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0136 ntime: 0084 mem: 3.36 + 04-04 19:04:05 | [882][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0110 ntime: 0076 mem: 3.36 + 04-04 19:04:12 | [882][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0939 ntime: 0083 mem: 3.36 + 04-04 19:04:18 | [882][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0820 ntime: 0083 mem: 3.36 + 04-04 19:04:23 | [882][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0730 ntime: 0083 mem: 3.36 + 04-04 19:04:28 | [882][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0111 ntime: 0079 mem: 3.36 + 04-04 19:04:34 | [882][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0864 ntime: 0083 mem: 3.36 + 04-04 19:04:41 | [882][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0078 mem: 3.36 + 04-04 19:04:48 | [882][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0845 ntime: 0078 mem: 3.36 + 04-04 19:04:54 | [882][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0134 ntime: 0080 mem: 3.36 + 04-04 19:04:59 | [882][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 19:05:07 | [882][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0067 ntime: 0087 mem: 3.36 + 04-04 19:05:12 | [882][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0697 ntime: 0077 mem: 3.36 + 04-04 19:05:17 | [882][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0160 ntime: 0081 mem: 3.36 + 04-04 19:05:24 | [882][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0836 ntime: 0079 mem: 3.36 + 04-04 19:05:30 | [882][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0729 ntime: 0081 mem: 3.36 + 04-04 19:05:37 | [882][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1000 ntime: 0084 mem: 3.36 + 04-04 19:05:44 | Time info >>>> elapsed: 1302.37 mins remain: 172.57 mins + 04-04 19:05:46 | [883][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1629 ntime: 0083 mem: 3.36 + 04-04 19:05:54 | [883][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0767 ntime: 0074 mem: 3.36 + 04-04 19:06:02 | [883][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1486 ntime: 0083 mem: 3.36 + 04-04 19:06:10 | [883][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1013 ntime: 0057 mem: 3.36 + 04-04 19:06:22 | [883][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0917 ntime: 0065 mem: 3.36 + 04-04 19:06:33 | [883][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1408 ntime: 0087 mem: 3.36 + 04-04 19:06:45 | [883][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0808 ntime: 0078 mem: 3.36 + 04-04 19:06:53 | [883][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1376 ntime: 0075 mem: 3.36 + 04-04 19:07:00 | [883][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0096 ntime: 0080 mem: 3.36 + 04-04 19:07:11 | [883][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0995 ntime: 0087 mem: 3.36 + 04-04 19:07:22 | [883][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1422 ntime: 0079 mem: 3.36 + 04-04 19:07:26 | [883][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0425 ntime: 0078 mem: 3.36 + 04-04 19:07:33 | [883][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1025 ntime: 0083 mem: 3.36 + 04-04 19:07:39 | [883][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0895 ntime: 0082 mem: 3.36 + 04-04 19:07:48 | [883][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1238 ntime: 0073 mem: 3.36 + 04-04 19:07:57 | [883][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2085 ntime: 0082 mem: 3.36 + 04-04 19:08:04 | [883][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1280 ntime: 0078 mem: 3.36 + 04-04 19:08:10 | [883][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0499 ntime: 0075 mem: 3.36 + 04-04 19:08:16 | Time info >>>> elapsed: 1304.89 mins remain: 171.23 mins + 04-04 19:08:16 | [884][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0194 ntime: 0077 mem: 3.36 + 04-04 19:08:22 | [884][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1006 ntime: 0086 mem: 3.36 + 04-04 19:08:27 | [884][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0116 ntime: 0080 mem: 3.36 + 04-04 19:08:36 | [884][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1599 ntime: 0071 mem: 3.36 + 04-04 19:08:42 | [884][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1290 ntime: 0086 mem: 3.36 + 04-04 19:08:49 | [884][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0146 ntime: 0080 mem: 3.36 + 04-04 19:08:54 | [884][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0087 ntime: 0083 mem: 3.36 + 04-04 19:08:59 | [884][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0130 ntime: 0079 mem: 3.36 + 04-04 19:09:06 | [884][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1051 ntime: 0084 mem: 3.36 + 04-04 19:09:13 | [884][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1202 ntime: 0082 mem: 3.36 + 04-04 19:09:17 | [884][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0067 ntime: 0057 mem: 3.36 + 04-04 19:09:24 | [884][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0987 ntime: 0078 mem: 3.36 + 04-04 19:09:29 | [884][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0113 ntime: 0069 mem: 3.36 + 04-04 19:09:36 | [884][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0085 mem: 3.36 + 04-04 19:09:42 | [884][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0143 ntime: 0079 mem: 3.36 + 04-04 19:09:48 | [884][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0103 ntime: 0081 mem: 3.36 + 04-04 19:09:55 | [884][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0417 ntime: 0081 mem: 3.36 + 04-04 19:10:02 | [884][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1382 ntime: 0087 mem: 3.36 + 04-04 19:10:07 | Time info >>>> elapsed: 1306.75 mins remain: 169.80 mins + 04-04 19:10:08 | [885][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0123 ntime: 0073 mem: 3.36 + 04-04 19:10:14 | [885][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0103 ntime: 0084 mem: 3.36 + 04-04 19:10:20 | [885][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0086 mem: 3.36 + 04-04 19:10:26 | [885][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0075 ntime: 0078 mem: 3.36 + 04-04 19:10:32 | [885][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1032 ntime: 0072 mem: 3.36 + 04-04 19:10:39 | [885][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0077 mem: 3.36 + 04-04 19:10:44 | [885][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0079 ntime: 0083 mem: 3.36 + 04-04 19:10:53 | [885][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1322 ntime: 0080 mem: 3.36 + 04-04 19:11:01 | [885][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0091 ntime: 0084 mem: 3.36 + 04-04 19:11:08 | [885][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0131 ntime: 0082 mem: 3.36 + 04-04 19:11:13 | [885][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0166 ntime: 0083 mem: 3.36 + 04-04 19:11:24 | [885][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1689 ntime: 0072 mem: 3.36 + 04-04 19:11:38 | [885][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0079 mem: 3.36 + 04-04 19:11:42 | [885][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1165 ntime: 0083 mem: 3.36 + 04-04 19:11:48 | [885][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0937 ntime: 0093 mem: 3.36 + 04-04 19:11:54 | [885][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0589 ntime: 0079 mem: 3.36 + 04-04 19:12:03 | [885][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0854 ntime: 0079 mem: 3.36 + 04-04 19:12:09 | [885][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 19:12:15 | Time info >>>> elapsed: 1308.88 mins remain: 168.41 mins + 04-04 19:12:17 | [886][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1706 ntime: 0058 mem: 3.36 + 04-04 19:12:24 | [886][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1089 ntime: 0081 mem: 3.36 + 04-04 19:12:34 | [886][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1743 ntime: 0077 mem: 3.36 + 04-04 19:12:42 | [886][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0079 ntime: 0078 mem: 3.36 + 04-04 19:12:50 | [886][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1407 ntime: 0085 mem: 3.36 + 04-04 19:12:59 | [886][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0297 ntime: 0076 mem: 3.36 + 04-04 19:13:06 | [886][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0154 ntime: 0080 mem: 3.36 + 04-04 19:13:12 | [886][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0075 mem: 3.36 + 04-04 19:13:20 | [886][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0100 ntime: 0081 mem: 3.36 + 04-04 19:13:26 | [886][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0764 ntime: 0073 mem: 3.36 + 04-04 19:13:33 | [886][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0449 ntime: 0072 mem: 3.36 + 04-04 19:13:40 | [886][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0080 mem: 3.36 + 04-04 19:13:46 | [886][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0096 ntime: 0078 mem: 3.36 + 04-04 19:13:51 | [886][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0097 ntime: 0078 mem: 3.36 + 04-04 19:13:59 | [886][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0724 ntime: 0082 mem: 3.36 + 04-04 19:14:06 | [886][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0069 ntime: 0080 mem: 3.36 + 04-04 19:14:12 | [886][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0830 ntime: 0085 mem: 3.36 + 04-04 19:14:17 | [886][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0071 ntime: 0086 mem: 3.36 + 04-04 19:14:23 | Time info >>>> elapsed: 1311.01 mins remain: 167.02 mins + 04-04 19:14:24 | [887][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1040 ntime: 0086 mem: 3.36 + 04-04 19:14:29 | [887][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0395 ntime: 0075 mem: 3.36 + 04-04 19:14:34 | [887][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0086 mem: 3.36 + 04-04 19:14:41 | [887][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0118 ntime: 0086 mem: 3.36 + 04-04 19:14:51 | [887][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2250 ntime: 0081 mem: 3.36 + 04-04 19:15:00 | [887][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0076 mem: 3.36 + 04-04 19:15:09 | [887][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1221 ntime: 0073 mem: 3.36 + 04-04 19:15:16 | [887][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0574 ntime: 0055 mem: 3.36 + 04-04 19:15:23 | [887][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1131 ntime: 0075 mem: 3.36 + 04-04 19:15:33 | [887][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0075 ntime: 0056 mem: 3.36 + 04-04 19:15:44 | [887][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1061 ntime: 0075 mem: 3.36 + 04-04 19:15:50 | [887][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1182 ntime: 0075 mem: 3.36 + 04-04 19:15:56 | [887][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0764 ntime: 0080 mem: 3.36 + 04-04 19:16:04 | [887][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0756 ntime: 0089 mem: 3.36 + 04-04 19:16:12 | [887][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0077 ntime: 0088 mem: 3.36 + 04-04 19:16:16 | [887][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0228 ntime: 0082 mem: 3.36 + 04-04 19:16:22 | [887][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0891 ntime: 0077 mem: 3.36 + 04-04 19:16:28 | [887][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0085 ntime: 0079 mem: 3.36 + 04-04 19:16:34 | Time info >>>> elapsed: 1313.19 mins remain: 165.63 mins + 04-04 19:16:34 | [888][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0078 mem: 3.36 + 04-04 19:16:41 | [888][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1304 ntime: 0085 mem: 3.36 + 04-04 19:16:47 | [888][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0137 ntime: 0088 mem: 3.36 + 04-04 19:16:52 | [888][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0198 ntime: 0079 mem: 3.36 + 04-04 19:16:58 | [888][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0078 mem: 3.36 + 04-04 19:17:04 | [888][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0075 ntime: 0084 mem: 3.36 + 04-04 19:17:10 | [888][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0081 mem: 3.36 + 04-04 19:17:18 | [888][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1126 ntime: 0084 mem: 3.36 + 04-04 19:17:26 | [888][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1417 ntime: 0078 mem: 3.36 + 04-04 19:17:32 | [888][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0081 mem: 3.36 + 04-04 19:17:38 | [888][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0880 ntime: 0079 mem: 3.36 + 04-04 19:17:42 | [888][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0144 ntime: 0073 mem: 3.36 + 04-04 19:17:49 | [888][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0700 ntime: 0079 mem: 3.36 + 04-04 19:17:54 | [888][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0069 ntime: 0087 mem: 3.36 + 04-04 19:18:00 | [888][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0075 ntime: 0080 mem: 3.36 + 04-04 19:18:05 | [888][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0277 ntime: 0078 mem: 3.36 + 04-04 19:18:12 | [888][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1224 ntime: 0078 mem: 3.36 + 04-04 19:18:17 | [888][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0081 mem: 3.36 + 04-04 19:18:24 | Time info >>>> elapsed: 1315.03 mins remain: 164.19 mins + 04-04 19:18:24 | [889][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0489 ntime: 0085 mem: 3.36 + 04-04 19:18:30 | [889][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0519 ntime: 0073 mem: 3.36 + 04-04 19:18:38 | [889][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1356 ntime: 0082 mem: 3.36 + 04-04 19:18:45 | [889][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1263 ntime: 0077 mem: 3.36 + 04-04 19:18:52 | [889][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0249 ntime: 0082 mem: 3.36 + 04-04 19:18:59 | [889][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1204 ntime: 0080 mem: 3.36 + 04-04 19:19:04 | [889][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0130 ntime: 0089 mem: 3.36 + 04-04 19:19:11 | [889][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1096 ntime: 0069 mem: 3.36 + 04-04 19:19:16 | [889][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0086 mem: 3.36 + 04-04 19:19:23 | [889][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0851 ntime: 0071 mem: 3.36 + 04-04 19:19:31 | [889][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1234 ntime: 0083 mem: 3.36 + 04-04 19:19:35 | [889][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0692 ntime: 0077 mem: 3.36 + 04-04 19:19:41 | [889][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0078 mem: 3.36 + 04-04 19:19:47 | [889][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0793 ntime: 0081 mem: 3.36 + 04-04 19:19:56 | [889][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0096 ntime: 0084 mem: 3.36 + 04-04 19:20:02 | [889][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0192 ntime: 0072 mem: 3.36 + 04-04 19:20:09 | [889][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1021 ntime: 0076 mem: 3.36 + 04-04 19:20:12 | [889][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0088 mem: 3.36 + 04-04 19:20:16 | Time info >>>> elapsed: 1316.90 mins remain: 162.76 mins + 04-04 19:20:18 | [890][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1237 ntime: 0078 mem: 3.36 + 04-04 19:20:25 | [890][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0081 mem: 3.36 + 04-04 19:20:32 | [890][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0119 ntime: 0078 mem: 3.36 + 04-04 19:20:38 | [890][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0759 ntime: 0082 mem: 3.36 + 04-04 19:20:44 | [890][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0074 mem: 3.36 + 04-04 19:20:50 | [890][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 19:20:56 | [890][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0798 ntime: 0081 mem: 3.36 + 04-04 19:21:01 | [890][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0068 ntime: 0077 mem: 3.36 + 04-04 19:21:06 | [890][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0083 mem: 3.36 + 04-04 19:21:12 | [890][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1030 ntime: 0083 mem: 3.36 + 04-04 19:21:19 | [890][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1009 ntime: 0083 mem: 3.36 + 04-04 19:21:23 | [890][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0084 mem: 3.36 + 04-04 19:21:28 | [890][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0068 ntime: 0080 mem: 3.36 + 04-04 19:21:34 | [890][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0515 ntime: 0080 mem: 3.36 + 04-04 19:21:40 | [890][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0091 ntime: 0084 mem: 3.36 + 04-04 19:21:48 | [890][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 19:21:56 | [890][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0825 ntime: 0082 mem: 3.36 + 04-04 19:22:00 | [890][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0082 mem: 3.36 + 04-04 19:22:05 | Time info >>>> elapsed: 1318.72 mins remain: 161.32 mins + 04-04 19:22:07 | [891][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 1629 ntime: 0078 mem: 3.36 + 04-04 19:22:13 | [891][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0695 ntime: 0081 mem: 3.36 + 04-04 19:22:19 | [891][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0858 ntime: 0080 mem: 3.36 + 04-04 19:22:25 | [891][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0073 mem: 3.36 + 04-04 19:22:31 | [891][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0910 ntime: 0078 mem: 3.36 + 04-04 19:22:36 | [891][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0525 ntime: 0079 mem: 3.36 + 04-04 19:22:43 | [891][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1437 ntime: 0086 mem: 3.36 + 04-04 19:22:49 | [891][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0082 mem: 3.36 + 04-04 19:22:56 | [891][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1283 ntime: 0078 mem: 3.36 + 04-04 19:23:01 | [891][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0432 ntime: 0080 mem: 3.36 + 04-04 19:23:08 | [891][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0322 ntime: 0081 mem: 3.36 + 04-04 19:23:15 | [891][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1037 ntime: 0080 mem: 3.36 + 04-04 19:23:20 | [891][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0082 ntime: 0082 mem: 3.36 + 04-04 19:23:26 | [891][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0119 ntime: 0076 mem: 3.36 + 04-04 19:23:31 | [891][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0139 ntime: 0080 mem: 3.36 + 04-04 19:23:35 | [891][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0521 ntime: 0083 mem: 3.36 + 04-04 19:23:40 | [891][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0751 ntime: 0081 mem: 3.36 + 04-04 19:23:46 | [891][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0106 ntime: 0084 mem: 3.36 + 04-04 19:23:50 | Time info >>>> elapsed: 1320.46 mins remain: 159.88 mins + 04-04 19:23:50 | [892][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0083 ntime: 0079 mem: 3.36 + 04-04 19:23:58 | [892][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0636 ntime: 0083 mem: 3.36 + 04-04 19:24:05 | [892][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0110 ntime: 0085 mem: 3.36 + 04-04 19:24:10 | [892][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0079 mem: 3.36 + 04-04 19:24:15 | [892][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1184 ntime: 0088 mem: 3.36 + 04-04 19:24:22 | [892][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1124 ntime: 0077 mem: 3.36 + 04-04 19:24:29 | [892][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0738 ntime: 0087 mem: 3.36 + 04-04 19:24:33 | [892][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0082 mem: 3.36 + 04-04 19:24:42 | [892][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1075 ntime: 0080 mem: 3.36 + 04-04 19:24:48 | [892][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1739 ntime: 0086 mem: 3.36 + 04-04 19:24:54 | [892][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0537 ntime: 0075 mem: 3.36 + 04-04 19:25:00 | [892][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0536 ntime: 0084 mem: 3.36 + 04-04 19:25:05 | [892][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0123 ntime: 0083 mem: 3.36 + 04-04 19:25:12 | [892][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1369 ntime: 0085 mem: 3.36 + 04-04 19:25:17 | [892][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0885 ntime: 0078 mem: 3.36 + 04-04 19:25:22 | [892][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0458 ntime: 0074 mem: 3.36 + 04-04 19:25:30 | [892][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1378 ntime: 0081 mem: 3.36 + 04-04 19:25:38 | [892][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0388 ntime: 0081 mem: 3.36 + 04-04 19:25:43 | Time info >>>> elapsed: 1322.35 mins remain: 158.45 mins + 04-04 19:25:44 | [893][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0650 ntime: 0080 mem: 3.36 + 04-04 19:25:50 | [893][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0695 ntime: 0080 mem: 3.36 + 04-04 19:25:55 | [893][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0565 ntime: 0080 mem: 3.36 + 04-04 19:26:02 | [893][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1055 ntime: 0088 mem: 3.36 + 04-04 19:26:08 | [893][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0091 ntime: 0077 mem: 3.36 + 04-04 19:26:13 | [893][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0756 ntime: 0077 mem: 3.36 + 04-04 19:26:22 | [893][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2005 ntime: 0079 mem: 3.36 + 04-04 19:26:28 | [893][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1065 ntime: 0084 mem: 3.36 + 04-04 19:26:35 | [893][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1175 ntime: 0079 mem: 3.36 + 04-04 19:26:41 | [893][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0968 ntime: 0079 mem: 3.36 + 04-04 19:26:47 | [893][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0655 ntime: 0071 mem: 3.36 + 04-04 19:26:52 | [893][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0131 ntime: 0087 mem: 3.36 + 04-04 19:26:58 | [893][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0610 ntime: 0079 mem: 3.36 + 04-04 19:27:03 | [893][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0104 ntime: 0074 mem: 3.36 + 04-04 19:27:07 | [893][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0054 mem: 3.36 + 04-04 19:27:12 | [893][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0106 ntime: 0078 mem: 3.36 + 04-04 19:27:19 | [893][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0071 ntime: 0077 mem: 3.36 + 04-04 19:27:25 | [893][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0110 ntime: 0074 mem: 3.36 + 04-04 19:27:31 | Time info >>>> elapsed: 1324.15 mins remain: 157.00 mins + 04-04 19:27:31 | [894][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0074 mem: 3.36 + 04-04 19:27:37 | [894][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0082 mem: 3.36 + 04-04 19:27:43 | [894][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0119 ntime: 0084 mem: 3.36 + 04-04 19:27:50 | [894][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0790 ntime: 0086 mem: 3.36 + 04-04 19:27:54 | [894][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0340 ntime: 0090 mem: 3.36 + 04-04 19:28:01 | [894][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0500 ntime: 0084 mem: 3.36 + 04-04 19:28:08 | [894][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0933 ntime: 0086 mem: 3.36 + 04-04 19:28:13 | [894][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0467 ntime: 0081 mem: 3.36 + 04-04 19:28:18 | [894][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0968 ntime: 0091 mem: 3.36 + 04-04 19:28:23 | [894][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0106 ntime: 0075 mem: 3.36 + 04-04 19:28:30 | [894][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0788 ntime: 0081 mem: 3.36 + 04-04 19:28:35 | [894][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1182 ntime: 0078 mem: 3.36 + 04-04 19:28:42 | [894][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0108 ntime: 0081 mem: 3.36 + 04-04 19:28:48 | [894][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0874 ntime: 0077 mem: 3.36 + 04-04 19:28:53 | [894][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0103 ntime: 0086 mem: 3.36 + 04-04 19:28:58 | [894][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1239 ntime: 0084 mem: 3.36 + 04-04 19:29:04 | [894][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1145 ntime: 0079 mem: 3.36 + 04-04 19:29:12 | [894][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0840 ntime: 0076 mem: 3.36 + 04-04 19:29:17 | Time info >>>> elapsed: 1325.91 mins remain: 155.55 mins + 04-04 19:29:17 | [895][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0076 mem: 3.36 + 04-04 19:29:23 | [895][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0072 mem: 3.36 + 04-04 19:29:29 | [895][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0118 ntime: 0084 mem: 3.36 + 04-04 19:29:35 | [895][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1295 ntime: 0075 mem: 3.36 + 04-04 19:29:42 | [895][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1350 ntime: 0073 mem: 3.36 + 04-04 19:29:52 | [895][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0918 ntime: 0075 mem: 3.36 + 04-04 19:30:00 | [895][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1356 ntime: 0072 mem: 3.36 + 04-04 19:30:05 | [895][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0638 ntime: 0086 mem: 3.36 + 04-04 19:30:11 | [895][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0160 ntime: 0061 mem: 3.36 + 04-04 19:30:18 | [895][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1596 ntime: 0072 mem: 3.36 + 04-04 19:30:25 | [895][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0102 ntime: 0081 mem: 3.36 + 04-04 19:30:32 | [895][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0094 ntime: 0076 mem: 3.36 + 04-04 19:30:37 | [895][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0091 ntime: 0077 mem: 3.36 + 04-04 19:30:44 | [895][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0750 ntime: 0073 mem: 3.36 + 04-04 19:30:50 | [895][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0718 ntime: 0077 mem: 3.36 + 04-04 19:30:56 | [895][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0983 ntime: 0069 mem: 3.36 + 04-04 19:31:01 | [895][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0112 ntime: 0092 mem: 3.36 + 04-04 19:31:07 | [895][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0606 ntime: 0073 mem: 3.36 + 04-04 19:31:12 | Time info >>>> elapsed: 1327.83 mins remain: 154.12 mins + 04-04 19:31:12 | [896][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0121 ntime: 0079 mem: 3.36 + 04-04 19:31:17 | [896][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0157 ntime: 0084 mem: 3.36 + 04-04 19:31:24 | [896][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1097 ntime: 0083 mem: 3.36 + 04-04 19:31:29 | [896][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0776 ntime: 0081 mem: 3.36 + 04-04 19:31:35 | [896][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0077 mem: 3.36 + 04-04 19:31:42 | [896][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0921 ntime: 0078 mem: 3.36 + 04-04 19:31:53 | [896][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0788 ntime: 0080 mem: 3.36 + 04-04 19:32:02 | [896][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0077 mem: 3.36 + 04-04 19:32:08 | [896][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0076 mem: 3.36 + 04-04 19:32:14 | [896][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0082 mem: 3.36 + 04-04 19:32:21 | [896][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0077 mem: 3.36 + 04-04 19:32:29 | [896][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1229 ntime: 0080 mem: 3.36 + 04-04 19:32:37 | [896][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0889 ntime: 0086 mem: 3.36 + 04-04 19:32:44 | [896][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0777 ntime: 0077 mem: 3.36 + 04-04 19:32:51 | [896][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0078 mem: 3.36 + 04-04 19:32:58 | [896][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0118 ntime: 0089 mem: 3.36 + 04-04 19:33:05 | [896][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1605 ntime: 0057 mem: 3.36 + 04-04 19:33:11 | [896][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0610 ntime: 0077 mem: 3.36 + 04-04 19:33:17 | Time info >>>> elapsed: 1329.91 mins remain: 152.71 mins + 04-04 19:33:17 | [897][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0084 ntime: 0077 mem: 3.36 + 04-04 19:33:25 | [897][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1120 ntime: 0071 mem: 3.36 + 04-04 19:33:31 | [897][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0087 ntime: 0076 mem: 3.36 + 04-04 19:33:37 | [897][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0832 ntime: 0081 mem: 3.36 + 04-04 19:33:43 | [897][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0076 mem: 3.36 + 04-04 19:33:50 | [897][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1238 ntime: 0080 mem: 3.36 + 04-04 19:33:58 | [897][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1680 ntime: 0079 mem: 3.36 + 04-04 19:34:08 | [897][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1429 ntime: 0081 mem: 3.36 + 04-04 19:34:15 | [897][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0486 ntime: 0081 mem: 3.36 + 04-04 19:34:22 | [897][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 19:34:29 | [897][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1358 ntime: 0079 mem: 3.36 + 04-04 19:34:36 | [897][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1169 ntime: 0079 mem: 3.36 + 04-04 19:34:43 | [897][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0829 ntime: 0074 mem: 3.36 + 04-04 19:34:51 | [897][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1587 ntime: 0075 mem: 3.36 + 04-04 19:34:58 | [897][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0276 ntime: 0077 mem: 3.36 + 04-04 19:35:06 | [897][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1202 ntime: 0076 mem: 3.36 + 04-04 19:35:14 | [897][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1413 ntime: 0081 mem: 3.36 + 04-04 19:35:20 | [897][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1118 ntime: 0086 mem: 3.36 + 04-04 19:35:25 | Time info >>>> elapsed: 1332.05 mins remain: 151.30 mins + 04-04 19:35:26 | [898][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0132 ntime: 0078 mem: 3.36 + 04-04 19:35:32 | [898][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0137 ntime: 0077 mem: 3.36 + 04-04 19:35:38 | [898][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0120 ntime: 0078 mem: 3.36 + 04-04 19:35:43 | [898][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0072 mem: 3.36 + 04-04 19:35:51 | [898][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1510 ntime: 0072 mem: 3.36 + 04-04 19:35:57 | [898][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0129 ntime: 0072 mem: 3.36 + 04-04 19:36:09 | [898][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1400 ntime: 0076 mem: 3.36 + 04-04 19:36:15 | [898][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0204 ntime: 0079 mem: 3.36 + 04-04 19:36:21 | [898][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0151 ntime: 0088 mem: 3.36 + 04-04 19:36:29 | [898][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 19:36:37 | [898][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0890 ntime: 0068 mem: 3.36 + 04-04 19:36:44 | [898][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0077 mem: 3.36 + 04-04 19:36:52 | [898][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0125 ntime: 0074 mem: 3.36 + 04-04 19:37:01 | [898][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1422 ntime: 0080 mem: 3.36 + 04-04 19:37:07 | [898][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0102 ntime: 0078 mem: 3.36 + 04-04 19:37:14 | [898][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1595 ntime: 0079 mem: 3.36 + 04-04 19:37:22 | [898][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0538 ntime: 0074 mem: 3.36 + 04-04 19:37:30 | [898][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1239 ntime: 0078 mem: 3.36 + 04-04 19:37:33 | Time info >>>> elapsed: 1334.18 mins remain: 149.89 mins + 04-04 19:37:34 | [899][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1094 ntime: 0076 mem: 3.36 + 04-04 19:37:42 | [899][010/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0835 ntime: 0074 mem: 3.36 + 04-04 19:37:51 | [899][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0077 ntime: 0079 mem: 3.36 + 04-04 19:37:58 | [899][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0082 mem: 3.36 + 04-04 19:38:05 | [899][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1191 ntime: 0077 mem: 3.36 + 04-04 19:38:11 | [899][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0636 ntime: 0079 mem: 3.36 + 04-04 19:38:18 | [899][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0562 ntime: 0076 mem: 3.36 + 04-04 19:38:25 | [899][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1091 ntime: 0082 mem: 3.36 + 04-04 19:38:31 | [899][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0287 ntime: 0072 mem: 3.36 + 04-04 19:38:38 | [899][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0206 ntime: 0078 mem: 3.36 + 04-04 19:38:47 | [899][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1145 ntime: 0080 mem: 3.36 + 04-04 19:38:56 | [899][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1101 ntime: 0086 mem: 3.36 + 04-04 19:39:01 | [899][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0963 ntime: 0079 mem: 3.36 + 04-04 19:39:08 | [899][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1301 ntime: 0081 mem: 3.36 + 04-04 19:39:15 | [899][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0707 ntime: 0078 mem: 3.36 + 04-04 19:39:22 | [899][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0075 mem: 3.36 + 04-04 19:39:30 | [899][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0078 mem: 3.36 + 04-04 19:39:37 | [899][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0795 ntime: 0080 mem: 3.36 + 04-04 19:39:40 | Time info >>>> elapsed: 1336.30 mins remain: 148.48 mins + 04-04 19:39:41 | [900][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0076 mem: 3.36 + 04-04 19:39:48 | [900][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0192 ntime: 0075 mem: 3.36 + 04-04 19:39:56 | [900][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0109 ntime: 0074 mem: 3.36 + 04-04 19:40:04 | [900][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1565 ntime: 0072 mem: 3.36 + 04-04 19:40:11 | [900][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1313 ntime: 0080 mem: 3.36 + 04-04 19:40:17 | [900][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0075 mem: 3.36 + 04-04 19:40:23 | [900][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0075 mem: 3.36 + 04-04 19:40:31 | [900][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0819 ntime: 0073 mem: 3.36 + 04-04 19:40:37 | [900][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 19:40:44 | [900][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1085 ntime: 0084 mem: 3.36 + 04-04 19:40:52 | [900][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0500 ntime: 0071 mem: 3.36 + 04-04 19:41:02 | [900][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2411 ntime: 0080 mem: 3.36 + 04-04 19:41:09 | [900][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0090 ntime: 0083 mem: 3.36 + 04-04 19:41:15 | [900][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0124 ntime: 0076 mem: 3.36 + 04-04 19:41:23 | [900][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0462 ntime: 0075 mem: 3.36 + 04-04 19:41:30 | [900][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0076 mem: 3.36 + 04-04 19:41:40 | [900][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 19:41:46 | [900][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0732 ntime: 0079 mem: 3.36 + 04-04 19:41:51 | Time info >>>> elapsed: 1338.47 mins remain: 147.07 mins + 04-04 19:41:52 | [901][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0946 ntime: 0077 mem: 3.36 + 04-04 19:41:58 | [901][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0562 ntime: 0086 mem: 3.36 + 04-04 19:42:04 | [901][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0601 ntime: 0086 mem: 3.36 + 04-04 19:42:13 | [901][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1395 ntime: 0077 mem: 3.36 + 04-04 19:42:19 | [901][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0323 ntime: 0073 mem: 3.36 + 04-04 19:42:27 | [901][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1449 ntime: 0082 mem: 3.36 + 04-04 19:42:32 | [901][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0783 ntime: 0078 mem: 3.36 + 04-04 19:42:39 | [901][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0564 ntime: 0074 mem: 3.36 + 04-04 19:42:48 | [901][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1710 ntime: 0075 mem: 3.36 + 04-04 19:42:57 | [901][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1300 ntime: 0058 mem: 3.36 + 04-04 19:43:03 | [901][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0109 ntime: 0074 mem: 3.36 + 04-04 19:43:11 | [901][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0812 ntime: 0080 mem: 3.36 + 04-04 19:43:16 | [901][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0821 ntime: 0079 mem: 3.36 + 04-04 19:43:23 | [901][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0079 ntime: 0079 mem: 3.36 + 04-04 19:43:29 | [901][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0668 ntime: 0078 mem: 3.36 + 04-04 19:43:36 | [901][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0989 ntime: 0083 mem: 3.36 + 04-04 19:43:42 | [901][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0075 ntime: 0082 mem: 3.36 + 04-04 19:43:52 | [901][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1677 ntime: 0079 mem: 3.36 + 04-04 19:43:58 | Time info >>>> elapsed: 1340.60 mins remain: 145.65 mins + 04-04 19:43:58 | [902][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0132 ntime: 0081 mem: 3.36 + 04-04 19:44:05 | [902][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0162 ntime: 0078 mem: 3.36 + 04-04 19:44:14 | [902][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1349 ntime: 0082 mem: 3.36 + 04-04 19:44:20 | [902][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1028 ntime: 0076 mem: 3.36 + 04-04 19:44:26 | [902][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0580 ntime: 0080 mem: 3.36 + 04-04 19:44:30 | [902][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0601 ntime: 0072 mem: 3.36 + 04-04 19:44:36 | [902][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0819 ntime: 0080 mem: 3.36 + 04-04 19:44:41 | [902][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0601 ntime: 0076 mem: 3.36 + 04-04 19:44:49 | [902][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0127 ntime: 0078 mem: 3.36 + 04-04 19:44:55 | [902][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 19:45:02 | [902][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1461 ntime: 0078 mem: 3.36 + 04-04 19:45:07 | [902][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0133 ntime: 0081 mem: 3.36 + 04-04 19:45:13 | [902][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0132 ntime: 0082 mem: 3.36 + 04-04 19:45:20 | [902][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0721 ntime: 0079 mem: 3.36 + 04-04 19:45:26 | [902][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0841 ntime: 0079 mem: 3.36 + 04-04 19:45:34 | [902][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0345 ntime: 0080 mem: 3.36 + 04-04 19:45:42 | [902][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0081 mem: 3.36 + 04-04 19:45:50 | [902][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0063 ntime: 0078 mem: 3.36 + 04-04 19:45:55 | Time info >>>> elapsed: 1342.55 mins remain: 144.22 mins + 04-04 19:45:55 | [903][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0042 ntime: 0078 mem: 3.36 + 04-04 19:46:02 | [903][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0093 ntime: 0077 mem: 3.36 + 04-04 19:46:11 | [903][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0786 ntime: 0075 mem: 3.36 + 04-04 19:46:17 | [903][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0074 mem: 3.36 + 04-04 19:46:23 | [903][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1272 ntime: 0072 mem: 3.36 + 04-04 19:46:29 | [903][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0079 mem: 3.36 + 04-04 19:46:36 | [903][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1597 ntime: 0078 mem: 3.36 + 04-04 19:46:43 | [903][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0934 ntime: 0079 mem: 3.36 + 04-04 19:46:51 | [903][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0878 ntime: 0078 mem: 3.36 + 04-04 19:46:57 | [903][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1011 ntime: 0077 mem: 3.36 + 04-04 19:47:05 | [903][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0300 ntime: 0069 mem: 3.36 + 04-04 19:47:12 | [903][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0090 ntime: 0079 mem: 3.36 + 04-04 19:47:20 | [903][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0503 ntime: 0081 mem: 3.36 + 04-04 19:47:27 | [903][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0763 ntime: 0075 mem: 3.36 + 04-04 19:47:33 | [903][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0198 ntime: 0081 mem: 3.36 + 04-04 19:47:40 | [903][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0080 mem: 3.36 + 04-04 19:47:46 | [903][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0105 ntime: 0081 mem: 3.36 + 04-04 19:47:55 | [903][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1053 ntime: 0078 mem: 3.36 + 04-04 19:48:00 | Time info >>>> elapsed: 1344.63 mins remain: 142.79 mins + 04-04 19:48:02 | [904][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1190 ntime: 0078 mem: 3.36 + 04-04 19:48:08 | [904][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1410 ntime: 0078 mem: 3.36 + 04-04 19:48:16 | [904][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1446 ntime: 0079 mem: 3.36 + 04-04 19:48:24 | [904][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1304 ntime: 0077 mem: 3.36 + 04-04 19:48:32 | [904][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1053 ntime: 0081 mem: 3.36 + 04-04 19:48:39 | [904][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0654 ntime: 0085 mem: 3.36 + 04-04 19:48:48 | [904][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1352 ntime: 0078 mem: 3.36 + 04-04 19:48:57 | [904][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1685 ntime: 0071 mem: 3.36 + 04-04 19:49:04 | [904][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0145 ntime: 0077 mem: 3.36 + 04-04 19:49:11 | [904][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0499 ntime: 0080 mem: 3.36 + 04-04 19:49:17 | [904][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0085 mem: 3.36 + 04-04 19:49:25 | [904][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1617 ntime: 0071 mem: 3.36 + 04-04 19:49:34 | [904][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1450 ntime: 0079 mem: 3.36 + 04-04 19:49:44 | [904][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1361 ntime: 0080 mem: 3.36 + 04-04 19:49:53 | [904][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0928 ntime: 0071 mem: 3.36 + 04-04 19:50:02 | [904][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0830 ntime: 0083 mem: 3.36 + 04-04 19:50:09 | [904][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0882 ntime: 0077 mem: 3.36 + 04-04 19:50:16 | [904][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1039 ntime: 0078 mem: 3.36 + 04-04 19:50:22 | Time info >>>> elapsed: 1346.99 mins remain: 141.40 mins + 04-04 19:50:22 | [905][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0538 ntime: 0076 mem: 3.36 + 04-04 19:50:30 | [905][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0063 ntime: 0077 mem: 3.36 + 04-04 19:50:38 | [905][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0664 ntime: 0073 mem: 3.36 + 04-04 19:50:44 | [905][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0076 mem: 3.36 + 04-04 19:50:51 | [905][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0072 mem: 3.36 + 04-04 19:50:59 | [905][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0081 mem: 3.36 + 04-04 19:51:06 | [905][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0875 ntime: 0076 mem: 3.36 + 04-04 19:51:12 | [905][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0108 ntime: 0079 mem: 3.36 + 04-04 19:51:19 | [905][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1200 ntime: 0080 mem: 3.36 + 04-04 19:51:25 | [905][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0095 ntime: 0082 mem: 3.36 + 04-04 19:51:33 | [905][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0593 ntime: 0071 mem: 3.36 + 04-04 19:51:41 | [905][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1059 ntime: 0081 mem: 3.36 + 04-04 19:51:49 | [905][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1058 ntime: 0086 mem: 3.36 + 04-04 19:51:55 | [905][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0256 ntime: 0077 mem: 3.36 + 04-04 19:52:04 | [905][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1210 ntime: 0073 mem: 3.36 + 04-04 19:52:13 | [905][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1301 ntime: 0060 mem: 3.36 + 04-04 19:52:19 | [905][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0076 mem: 3.36 + 04-04 19:52:27 | [905][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0194 ntime: 0079 mem: 3.36 + 04-04 19:52:33 | Time info >>>> elapsed: 1349.18 mins remain: 139.98 mins + 04-04 19:52:34 | [906][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 1139 ntime: 0074 mem: 3.36 + 04-04 19:52:43 | [906][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0224 ntime: 0075 mem: 3.36 + 04-04 19:52:49 | [906][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1229 ntime: 0077 mem: 3.36 + 04-04 19:52:57 | [906][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1817 ntime: 0078 mem: 3.36 + 04-04 19:53:05 | [906][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1140 ntime: 0084 mem: 3.36 + 04-04 19:53:11 | [906][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0117 ntime: 0080 mem: 3.36 + 04-04 19:53:17 | [906][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1533 ntime: 0070 mem: 3.36 + 04-04 19:53:26 | [906][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0074 mem: 3.36 + 04-04 19:53:35 | [906][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0416 ntime: 0076 mem: 3.36 + 04-04 19:53:42 | [906][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0764 ntime: 0080 mem: 3.36 + 04-04 19:53:48 | [906][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0197 ntime: 0077 mem: 3.36 + 04-04 19:53:55 | [906][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0900 ntime: 0078 mem: 3.36 + 04-04 19:54:03 | [906][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1239 ntime: 0081 mem: 3.36 + 04-04 19:54:11 | [906][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0715 ntime: 0075 mem: 3.36 + 04-04 19:54:24 | [906][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1806 ntime: 0087 mem: 3.36 + 04-04 19:54:36 | [906][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1093 ntime: 0076 mem: 3.36 + 04-04 19:54:47 | [906][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0076 mem: 3.36 + 04-04 19:54:55 | [906][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1424 ntime: 0080 mem: 3.36 + 04-04 19:55:05 | Time info >>>> elapsed: 1351.70 mins remain: 138.60 mins + 04-04 19:55:05 | [907][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0211 ntime: 0095 mem: 3.36 + 04-04 19:55:15 | [907][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1306 ntime: 0081 mem: 3.36 + 04-04 19:55:24 | [907][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0920 ntime: 0076 mem: 3.36 + 04-04 19:55:34 | [907][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0176 ntime: 0089 mem: 3.36 + 04-04 19:55:44 | [907][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1669 ntime: 0074 mem: 3.36 + 04-04 19:55:55 | [907][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1444 ntime: 0072 mem: 3.36 + 04-04 19:56:05 | [907][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1612 ntime: 0077 mem: 3.36 + 04-04 19:56:15 | [907][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1495 ntime: 0080 mem: 3.36 + 04-04 19:56:25 | [907][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0087 ntime: 0072 mem: 3.36 + 04-04 19:56:34 | [907][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0122 ntime: 0075 mem: 3.36 + 04-04 19:56:44 | [907][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1644 ntime: 0075 mem: 3.36 + 04-04 19:56:55 | [907][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1241 ntime: 0074 mem: 3.36 + 04-04 19:57:07 | [907][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0073 mem: 3.36 + 04-04 19:57:15 | [907][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1891 ntime: 0081 mem: 3.36 + 04-04 19:57:26 | [907][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0084 mem: 3.36 + 04-04 19:57:36 | [907][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0166 ntime: 0075 mem: 3.36 + 04-04 19:57:46 | [907][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1453 ntime: 0083 mem: 3.36 + 04-04 19:57:55 | [907][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0783 ntime: 0081 mem: 3.36 + 04-04 19:57:59 | Time info >>>> elapsed: 1354.61 mins remain: 137.25 mins + 04-04 19:57:59 | [908][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0088 mem: 3.36 + 04-04 19:58:07 | [908][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0985 ntime: 0078 mem: 3.36 + 04-04 19:58:12 | [908][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 19:58:17 | [908][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0079 mem: 3.36 + 04-04 19:58:23 | [908][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0077 mem: 3.36 + 04-04 19:58:28 | [908][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0096 ntime: 0077 mem: 3.36 + 04-04 19:58:34 | [908][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0794 ntime: 0078 mem: 3.36 + 04-04 19:58:39 | [908][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0081 mem: 3.36 + 04-04 19:58:44 | [908][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0152 ntime: 0057 mem: 3.36 + 04-04 19:58:50 | [908][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0105 ntime: 0076 mem: 3.36 + 04-04 19:58:55 | [908][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0080 ntime: 0076 mem: 3.36 + 04-04 19:59:00 | [908][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0075 mem: 3.36 + 04-04 19:59:06 | [908][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0344 ntime: 0079 mem: 3.36 + 04-04 19:59:10 | [908][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0099 ntime: 0077 mem: 3.36 + 04-04 19:59:18 | [908][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 19:59:23 | [908][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0570 ntime: 0078 mem: 3.36 + 04-04 19:59:30 | [908][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0087 ntime: 0079 mem: 3.36 + 04-04 19:59:36 | [908][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1161 ntime: 0079 mem: 3.36 + 04-04 19:59:40 | Time info >>>> elapsed: 1356.29 mins remain: 135.78 mins + 04-04 19:59:40 | [909][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0083 mem: 3.36 + 04-04 19:59:46 | [909][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1250 ntime: 0081 mem: 3.36 + 04-04 19:59:52 | [909][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0078 mem: 3.36 + 04-04 19:59:58 | [909][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1179 ntime: 0079 mem: 3.36 + 04-04 20:00:03 | [909][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0488 ntime: 0079 mem: 3.36 + 04-04 20:00:06 | [909][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0827 ntime: 0084 mem: 3.36 + 04-04 20:00:11 | [909][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0769 ntime: 0077 mem: 3.36 + 04-04 20:00:17 | [909][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0077 mem: 3.36 + 04-04 20:00:22 | [909][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1280 ntime: 0082 mem: 3.36 + 04-04 20:00:27 | [909][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0106 ntime: 0081 mem: 3.36 + 04-04 20:00:33 | [909][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0204 ntime: 0079 mem: 3.36 + 04-04 20:00:40 | [909][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0612 ntime: 0078 mem: 3.36 + 04-04 20:00:49 | [909][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0670 ntime: 0074 mem: 3.36 + 04-04 20:00:56 | [909][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0876 ntime: 0078 mem: 3.36 + 04-04 20:01:00 | [909][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0518 ntime: 0084 mem: 3.36 + 04-04 20:01:06 | [909][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0078 mem: 3.36 + 04-04 20:01:11 | [909][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0073 mem: 3.36 + 04-04 20:01:16 | [909][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0077 mem: 3.36 + 04-04 20:01:23 | Time info >>>> elapsed: 1358.01 mins remain: 134.31 mins + 04-04 20:01:24 | [910][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0713 ntime: 0076 mem: 3.36 + 04-04 20:01:30 | [910][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0502 ntime: 0077 mem: 3.36 + 04-04 20:01:35 | [910][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1056 ntime: 0078 mem: 3.36 + 04-04 20:01:41 | [910][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0509 ntime: 0078 mem: 3.36 + 04-04 20:01:46 | [910][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0058 mem: 3.36 + 04-04 20:01:52 | [910][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 20:01:58 | [910][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0596 ntime: 0078 mem: 3.36 + 04-04 20:02:04 | [910][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0081 mem: 3.36 + 04-04 20:02:08 | [910][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0079 mem: 3.36 + 04-04 20:02:13 | [910][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0108 ntime: 0077 mem: 3.36 + 04-04 20:02:18 | [910][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0163 ntime: 0081 mem: 3.36 + 04-04 20:02:24 | [910][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0647 ntime: 0077 mem: 3.36 + 04-04 20:02:30 | [910][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1218 ntime: 0070 mem: 3.36 + 04-04 20:02:35 | [910][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0139 ntime: 0072 mem: 3.36 + 04-04 20:02:42 | [910][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0101 ntime: 0081 mem: 3.36 + 04-04 20:02:48 | [910][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0153 ntime: 0073 mem: 3.36 + 04-04 20:02:55 | [910][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0372 ntime: 0079 mem: 3.36 + 04-04 20:02:59 | [910][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0683 ntime: 0083 mem: 3.36 + 04-04 20:03:03 | Time info >>>> elapsed: 1359.67 mins remain: 132.83 mins + 04-04 20:03:03 | [911][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0620 ntime: 0080 mem: 3.36 + 04-04 20:03:09 | [911][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0141 ntime: 0083 mem: 3.36 + 04-04 20:03:14 | [911][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0096 ntime: 0081 mem: 3.36 + 04-04 20:03:20 | [911][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1020 ntime: 0081 mem: 3.36 + 04-04 20:03:26 | [911][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0712 ntime: 0078 mem: 3.36 + 04-04 20:03:32 | [911][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0671 ntime: 0075 mem: 3.36 + 04-04 20:03:36 | [911][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0754 ntime: 0084 mem: 3.36 + 04-04 20:03:42 | [911][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0822 ntime: 0079 mem: 3.36 + 04-04 20:03:47 | [911][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0755 ntime: 0080 mem: 3.36 + 04-04 20:03:53 | [911][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0084 mem: 3.36 + 04-04 20:04:01 | [911][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1246 ntime: 0077 mem: 3.36 + 04-04 20:04:06 | [911][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1050 ntime: 0076 mem: 3.36 + 04-04 20:04:11 | [911][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0145 ntime: 0079 mem: 3.36 + 04-04 20:04:18 | [911][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1075 ntime: 0083 mem: 3.36 + 04-04 20:04:22 | [911][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0135 ntime: 0078 mem: 3.36 + 04-04 20:04:28 | [911][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0078 mem: 3.36 + 04-04 20:04:33 | [911][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0189 ntime: 0075 mem: 3.36 + 04-04 20:04:40 | [911][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0096 ntime: 0081 mem: 3.36 + 04-04 20:04:45 | Time info >>>> elapsed: 1361.37 mins remain: 131.36 mins + 04-04 20:04:45 | [912][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0738 ntime: 0078 mem: 3.36 + 04-04 20:04:52 | [912][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0083 ntime: 0062 mem: 3.36 + 04-04 20:04:57 | [912][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0076 mem: 3.36 + 04-04 20:05:02 | [912][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0556 ntime: 0078 mem: 3.36 + 04-04 20:05:09 | [912][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1655 ntime: 0075 mem: 3.36 + 04-04 20:05:15 | [912][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0112 ntime: 0079 mem: 3.36 + 04-04 20:05:22 | [912][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0920 ntime: 0074 mem: 3.36 + 04-04 20:05:27 | [912][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0681 ntime: 0077 mem: 3.36 + 04-04 20:05:33 | [912][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0078 mem: 3.36 + 04-04 20:05:39 | [912][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0588 ntime: 0078 mem: 3.36 + 04-04 20:05:46 | [912][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0820 ntime: 0087 mem: 3.36 + 04-04 20:05:51 | [912][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0084 mem: 3.36 + 04-04 20:05:57 | [912][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0883 ntime: 0075 mem: 3.36 + 04-04 20:06:02 | [912][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0088 ntime: 0074 mem: 3.36 + 04-04 20:06:07 | [912][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0109 ntime: 0072 mem: 3.36 + 04-04 20:06:11 | [912][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0628 ntime: 0076 mem: 3.36 + 04-04 20:06:15 | [912][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 20:06:22 | [912][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0947 ntime: 0076 mem: 3.36 + 04-04 20:06:26 | Time info >>>> elapsed: 1363.06 mins remain: 129.89 mins + 04-04 20:06:26 | [913][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 20:06:32 | [913][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0099 ntime: 0082 mem: 3.36 + 04-04 20:06:38 | [913][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0758 ntime: 0074 mem: 3.36 + 04-04 20:06:42 | [913][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0238 ntime: 0078 mem: 3.36 + 04-04 20:06:46 | [913][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0102 ntime: 0080 mem: 3.36 + 04-04 20:06:52 | [913][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0074 mem: 3.36 + 04-04 20:06:58 | [913][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0102 ntime: 0077 mem: 3.36 + 04-04 20:07:03 | [913][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0080 mem: 3.36 + 04-04 20:07:10 | [913][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1403 ntime: 0074 mem: 3.36 + 04-04 20:07:14 | [913][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0073 ntime: 0088 mem: 3.36 + 04-04 20:07:20 | [913][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0091 ntime: 0086 mem: 3.36 + 04-04 20:07:25 | [913][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0077 mem: 3.36 + 04-04 20:07:30 | [913][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 20:07:35 | [913][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0077 ntime: 0078 mem: 3.36 + 04-04 20:07:40 | [913][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0096 ntime: 0074 mem: 3.36 + 04-04 20:07:46 | [913][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0509 ntime: 0095 mem: 3.36 + 04-04 20:07:53 | [913][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0812 ntime: 0078 mem: 3.36 + 04-04 20:07:58 | [913][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0067 mem: 3.36 + 04-04 20:08:03 | Time info >>>> elapsed: 1364.68 mins remain: 128.41 mins + 04-04 20:08:04 | [914][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0558 ntime: 0076 mem: 3.36 + 04-04 20:08:08 | [914][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0166 ntime: 0076 mem: 3.36 + 04-04 20:08:15 | [914][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0061 mem: 3.36 + 04-04 20:08:20 | [914][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0562 ntime: 0071 mem: 3.36 + 04-04 20:08:23 | [914][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0074 mem: 3.36 + 04-04 20:08:28 | [914][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0981 ntime: 0078 mem: 3.36 + 04-04 20:08:37 | [914][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1447 ntime: 0079 mem: 3.36 + 04-04 20:08:41 | [914][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0080 mem: 3.36 + 04-04 20:08:47 | [914][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0112 ntime: 0080 mem: 3.36 + 04-04 20:08:53 | [914][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0892 ntime: 0080 mem: 3.36 + 04-04 20:08:58 | [914][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0648 ntime: 0077 mem: 3.36 + 04-04 20:09:05 | [914][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1020 ntime: 0078 mem: 3.36 + 04-04 20:09:11 | [914][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0098 ntime: 0079 mem: 3.36 + 04-04 20:09:17 | [914][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0200 ntime: 0074 mem: 3.36 + 04-04 20:09:23 | [914][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1011 ntime: 0076 mem: 3.36 + 04-04 20:09:30 | [914][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1091 ntime: 0081 mem: 3.36 + 04-04 20:09:35 | [914][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0687 ntime: 0078 mem: 3.36 + 04-04 20:09:41 | [914][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1334 ntime: 0076 mem: 3.36 + 04-04 20:09:45 | Time info >>>> elapsed: 1366.38 mins remain: 126.93 mins + 04-04 20:09:45 | [915][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0134 ntime: 0080 mem: 3.36 + 04-04 20:09:52 | [915][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0788 ntime: 0086 mem: 3.36 + 04-04 20:09:59 | [915][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1004 ntime: 0081 mem: 3.36 + 04-04 20:10:05 | [915][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0412 ntime: 0081 mem: 3.36 + 04-04 20:10:12 | [915][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0118 ntime: 0074 mem: 3.36 + 04-04 20:10:16 | [915][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0068 ntime: 0075 mem: 3.36 + 04-04 20:10:21 | [915][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0077 mem: 3.36 + 04-04 20:10:26 | [915][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1118 ntime: 0078 mem: 3.36 + 04-04 20:10:32 | [915][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0084 ntime: 0077 mem: 3.36 + 04-04 20:10:39 | [915][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0799 ntime: 0079 mem: 3.36 + 04-04 20:10:45 | [915][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0083 ntime: 0080 mem: 3.36 + 04-04 20:10:53 | [915][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0611 ntime: 0077 mem: 3.36 + 04-04 20:10:58 | [915][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0085 ntime: 0080 mem: 3.36 + 04-04 20:11:05 | [915][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1681 ntime: 0078 mem: 3.36 + 04-04 20:11:10 | [915][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0074 mem: 3.36 + 04-04 20:11:18 | [915][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0763 ntime: 0084 mem: 3.36 + 04-04 20:11:23 | [915][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0071 mem: 3.36 + 04-04 20:11:30 | [915][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0384 ntime: 0070 mem: 3.36 + 04-04 20:11:36 | Time info >>>> elapsed: 1368.22 mins remain: 125.47 mins + 04-04 20:11:37 | [916][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0784 ntime: 0084 mem: 3.36 + 04-04 20:11:43 | [916][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0199 ntime: 0079 mem: 3.36 + 04-04 20:11:48 | [916][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1586 ntime: 0078 mem: 3.36 + 04-04 20:11:54 | [916][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0469 ntime: 0080 mem: 3.36 + 04-04 20:11:59 | [916][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0106 ntime: 0071 mem: 3.36 + 04-04 20:12:05 | [916][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1168 ntime: 0083 mem: 3.36 + 04-04 20:12:11 | [916][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0575 ntime: 0078 mem: 3.36 + 04-04 20:12:17 | [916][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 20:12:23 | [916][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0573 ntime: 0078 mem: 3.36 + 04-04 20:12:29 | [916][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0078 ntime: 0086 mem: 3.36 + 04-04 20:12:36 | [916][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0086 ntime: 0090 mem: 3.36 + 04-04 20:12:41 | [916][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0102 ntime: 0078 mem: 3.36 + 04-04 20:12:47 | [916][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0639 ntime: 0082 mem: 3.36 + 04-04 20:12:53 | [916][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0154 ntime: 0076 mem: 3.36 + 04-04 20:12:58 | [916][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0186 ntime: 0083 mem: 3.36 + 04-04 20:13:04 | [916][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0084 mem: 3.36 + 04-04 20:13:10 | [916][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0131 ntime: 0078 mem: 3.36 + 04-04 20:13:15 | [916][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0839 ntime: 0076 mem: 3.36 + 04-04 20:13:19 | Time info >>>> elapsed: 1369.95 mins remain: 124.00 mins + 04-04 20:13:19 | [917][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0104 ntime: 0077 mem: 3.36 + 04-04 20:13:25 | [917][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0827 ntime: 0084 mem: 3.36 + 04-04 20:13:30 | [917][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0254 ntime: 0078 mem: 3.36 + 04-04 20:13:33 | [917][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0075 mem: 3.36 + 04-04 20:13:39 | [917][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0932 ntime: 0078 mem: 3.36 + 04-04 20:13:44 | [917][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0671 ntime: 0083 mem: 3.36 + 04-04 20:13:51 | [917][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0151 ntime: 0079 mem: 3.36 + 04-04 20:13:59 | [917][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0994 ntime: 0077 mem: 3.36 + 04-04 20:14:05 | [917][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0828 ntime: 0077 mem: 3.36 + 04-04 20:14:11 | [917][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1232 ntime: 0073 mem: 3.36 + 04-04 20:14:18 | [917][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0069 ntime: 0078 mem: 3.36 + 04-04 20:14:23 | [917][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 20:14:32 | [917][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1639 ntime: 0079 mem: 3.36 + 04-04 20:14:37 | [917][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0081 mem: 3.36 + 04-04 20:14:42 | [917][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 20:14:48 | [917][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1059 ntime: 0079 mem: 3.36 + 04-04 20:14:52 | [917][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0699 ntime: 0077 mem: 3.36 + 04-04 20:14:57 | [917][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0075 ntime: 0076 mem: 3.36 + 04-04 20:15:03 | Time info >>>> elapsed: 1371.67 mins remain: 122.52 mins + 04-04 20:15:03 | [918][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0184 ntime: 0070 mem: 3.36 + 04-04 20:15:09 | [918][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0955 ntime: 0079 mem: 3.36 + 04-04 20:15:15 | [918][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1113 ntime: 0079 mem: 3.36 + 04-04 20:15:20 | [918][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0198 ntime: 0084 mem: 3.36 + 04-04 20:15:25 | [918][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0792 ntime: 0075 mem: 3.36 + 04-04 20:15:29 | [918][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0723 ntime: 0078 mem: 3.36 + 04-04 20:15:34 | [918][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0069 ntime: 0079 mem: 3.36 + 04-04 20:15:40 | [918][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0813 ntime: 0074 mem: 3.36 + 04-04 20:15:45 | [918][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 20:15:50 | [918][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0091 ntime: 0080 mem: 3.36 + 04-04 20:15:56 | [918][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0812 ntime: 0082 mem: 3.36 + 04-04 20:16:01 | [918][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0085 mem: 3.36 + 04-04 20:16:07 | [918][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0080 mem: 3.36 + 04-04 20:16:14 | [918][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1140 ntime: 0077 mem: 3.36 + 04-04 20:16:19 | [918][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0082 mem: 3.36 + 04-04 20:16:24 | [918][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0885 ntime: 0071 mem: 3.36 + 04-04 20:16:31 | [918][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0882 ntime: 0083 mem: 3.36 + 04-04 20:16:36 | [918][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1009 ntime: 0079 mem: 3.36 + 04-04 20:16:44 | Time info >>>> elapsed: 1373.36 mins remain: 121.05 mins + 04-04 20:16:44 | [919][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0060 ntime: 0078 mem: 3.36 + 04-04 20:16:51 | [919][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0796 ntime: 0073 mem: 3.36 + 04-04 20:16:57 | [919][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0740 ntime: 0085 mem: 3.36 + 04-04 20:17:04 | [919][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0601 ntime: 0080 mem: 3.36 + 04-04 20:17:11 | [919][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0091 ntime: 0077 mem: 3.36 + 04-04 20:17:17 | [919][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0686 ntime: 0073 mem: 3.36 + 04-04 20:17:24 | [919][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1205 ntime: 0080 mem: 3.36 + 04-04 20:17:30 | [919][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1510 ntime: 0079 mem: 3.36 + 04-04 20:17:37 | [919][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1646 ntime: 0082 mem: 3.36 + 04-04 20:17:43 | [919][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0985 ntime: 0078 mem: 3.36 + 04-04 20:17:49 | [919][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0117 ntime: 0082 mem: 3.36 + 04-04 20:17:56 | [919][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0664 ntime: 0078 mem: 3.36 + 04-04 20:18:01 | [919][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0593 ntime: 0083 mem: 3.36 + 04-04 20:18:07 | [919][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1569 ntime: 0079 mem: 3.36 + 04-04 20:18:12 | [919][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0208 ntime: 0076 mem: 3.36 + 04-04 20:18:17 | [919][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0614 ntime: 0077 mem: 3.36 + 04-04 20:18:22 | [919][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0796 ntime: 0074 mem: 3.36 + 04-04 20:18:28 | [919][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0972 ntime: 0074 mem: 3.36 + 04-04 20:18:31 | Time info >>>> elapsed: 1375.15 mins remain: 119.58 mins + 04-04 20:18:33 | [920][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1296 ntime: 0077 mem: 3.36 + 04-04 20:18:39 | [920][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0080 mem: 3.36 + 04-04 20:18:44 | [920][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0581 ntime: 0079 mem: 3.36 + 04-04 20:18:49 | [920][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0079 mem: 3.36 + 04-04 20:18:56 | [920][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1150 ntime: 0077 mem: 3.36 + 04-04 20:19:02 | [920][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0092 ntime: 0082 mem: 3.36 + 04-04 20:19:08 | [920][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0544 ntime: 0078 mem: 3.36 + 04-04 20:19:14 | [920][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0130 ntime: 0079 mem: 3.36 + 04-04 20:19:20 | [920][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0105 ntime: 0076 mem: 3.36 + 04-04 20:19:25 | [920][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0078 mem: 3.36 + 04-04 20:19:30 | [920][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0071 mem: 3.36 + 04-04 20:19:36 | [920][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0607 ntime: 0082 mem: 3.36 + 04-04 20:19:40 | [920][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0317 ntime: 0081 mem: 3.36 + 04-04 20:19:45 | [920][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0814 ntime: 0081 mem: 3.36 + 04-04 20:19:52 | [920][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0193 ntime: 0090 mem: 3.36 + 04-04 20:19:57 | [920][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0125 ntime: 0081 mem: 3.36 + 04-04 20:20:00 | [920][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0975 ntime: 0080 mem: 3.36 + 04-04 20:20:08 | [920][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0628 ntime: 0071 mem: 3.36 + 04-04 20:20:15 | Time info >>>> elapsed: 1376.87 mins remain: 118.10 mins + 04-04 20:20:15 | [921][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0044 ntime: 0056 mem: 3.36 + 04-04 20:20:20 | [921][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1451 ntime: 0075 mem: 3.36 + 04-04 20:20:25 | [921][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0204 ntime: 0083 mem: 3.36 + 04-04 20:20:32 | [921][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0420 ntime: 0095 mem: 3.36 + 04-04 20:20:39 | [921][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0080 mem: 3.36 + 04-04 20:20:45 | [921][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0074 mem: 3.36 + 04-04 20:20:53 | [921][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0262 ntime: 0079 mem: 3.36 + 04-04 20:21:00 | [921][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0839 ntime: 0072 mem: 3.36 + 04-04 20:21:06 | [921][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0669 ntime: 0078 mem: 3.36 + 04-04 20:21:14 | [921][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0439 ntime: 0072 mem: 3.36 + 04-04 20:21:22 | [921][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0133 ntime: 0080 mem: 3.36 + 04-04 20:21:28 | [921][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0125 ntime: 0078 mem: 3.36 + 04-04 20:21:37 | [921][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1041 ntime: 0087 mem: 3.36 + 04-04 20:21:43 | [921][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0134 ntime: 0088 mem: 3.36 + 04-04 20:21:51 | [921][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1026 ntime: 0077 mem: 3.36 + 04-04 20:21:59 | [921][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0111 ntime: 0080 mem: 3.36 + 04-04 20:22:06 | [921][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1750 ntime: 0082 mem: 3.36 + 04-04 20:22:11 | [921][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0080 mem: 3.36 + 04-04 20:22:16 | Time info >>>> elapsed: 1378.90 mins remain: 116.65 mins + 04-04 20:22:16 | [922][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0048 ntime: 0077 mem: 3.36 + 04-04 20:22:23 | [922][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1069 ntime: 0071 mem: 3.36 + 04-04 20:22:30 | [922][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0842 ntime: 0069 mem: 3.36 + 04-04 20:22:35 | [922][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0125 ntime: 0078 mem: 3.36 + 04-04 20:22:41 | [922][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0132 ntime: 0086 mem: 3.36 + 04-04 20:22:46 | [922][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0086 mem: 3.36 + 04-04 20:22:52 | [922][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0150 ntime: 0073 mem: 3.36 + 04-04 20:22:57 | [922][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0068 ntime: 0077 mem: 3.36 + 04-04 20:23:02 | [922][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0529 ntime: 0083 mem: 3.36 + 04-04 20:23:08 | [922][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1253 ntime: 0079 mem: 3.36 + 04-04 20:23:13 | [922][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0564 ntime: 0079 mem: 3.36 + 04-04 20:23:18 | [922][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0139 ntime: 0077 mem: 3.36 + 04-04 20:23:24 | [922][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0816 ntime: 0078 mem: 3.36 + 04-04 20:23:28 | [922][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0105 ntime: 0063 mem: 3.36 + 04-04 20:23:34 | [922][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0166 ntime: 0080 mem: 3.36 + 04-04 20:23:39 | [922][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0790 ntime: 0075 mem: 3.36 + 04-04 20:23:46 | [922][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0645 ntime: 0074 mem: 3.36 + 04-04 20:23:52 | [922][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0071 mem: 3.36 + 04-04 20:23:56 | Time info >>>> elapsed: 1380.56 mins remain: 115.17 mins + 04-04 20:23:56 | [923][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0066 ntime: 0077 mem: 3.36 + 04-04 20:24:01 | [923][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0176 ntime: 0081 mem: 3.36 + 04-04 20:24:07 | [923][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0085 mem: 3.36 + 04-04 20:24:13 | [923][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0063 ntime: 0079 mem: 3.36 + 04-04 20:24:20 | [923][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0159 ntime: 0076 mem: 3.36 + 04-04 20:24:26 | [923][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0556 ntime: 0071 mem: 3.36 + 04-04 20:24:32 | [923][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1268 ntime: 0071 mem: 3.36 + 04-04 20:24:38 | [923][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0174 ntime: 0089 mem: 3.36 + 04-04 20:24:43 | [923][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0107 ntime: 0078 mem: 3.36 + 04-04 20:24:50 | [923][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0999 ntime: 0078 mem: 3.36 + 04-04 20:24:57 | [923][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0078 mem: 3.36 + 04-04 20:25:01 | [923][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0176 ntime: 0076 mem: 3.36 + 04-04 20:25:07 | [923][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0112 ntime: 0079 mem: 3.36 + 04-04 20:25:13 | [923][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0497 ntime: 0081 mem: 3.36 + 04-04 20:25:18 | [923][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0900 ntime: 0074 mem: 3.36 + 04-04 20:25:23 | [923][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0117 ntime: 0076 mem: 3.36 + 04-04 20:25:28 | [923][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0576 ntime: 0080 mem: 3.36 + 04-04 20:25:32 | [923][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0084 ntime: 0074 mem: 3.36 + 04-04 20:25:38 | Time info >>>> elapsed: 1382.25 mins remain: 113.69 mins + 04-04 20:25:38 | [924][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0074 mem: 3.36 + 04-04 20:25:44 | [924][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1198 ntime: 0078 mem: 3.36 + 04-04 20:25:49 | [924][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1050 ntime: 0075 mem: 3.36 + 04-04 20:25:56 | [924][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1482 ntime: 0080 mem: 3.36 + 04-04 20:26:03 | [924][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0077 mem: 3.36 + 04-04 20:26:07 | [924][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0133 ntime: 0078 mem: 3.36 + 04-04 20:26:13 | [924][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0808 ntime: 0078 mem: 3.36 + 04-04 20:26:20 | [924][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0598 ntime: 0080 mem: 3.36 + 04-04 20:26:24 | [924][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0085 mem: 3.36 + 04-04 20:26:30 | [924][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0768 ntime: 0082 mem: 3.36 + 04-04 20:26:36 | [924][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0917 ntime: 0056 mem: 3.36 + 04-04 20:26:41 | [924][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0655 ntime: 0084 mem: 3.36 + 04-04 20:26:46 | [924][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0707 ntime: 0083 mem: 3.36 + 04-04 20:26:53 | [924][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0781 ntime: 0083 mem: 3.36 + 04-04 20:26:58 | [924][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0724 ntime: 0075 mem: 3.36 + 04-04 20:27:03 | [924][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0117 ntime: 0077 mem: 3.36 + 04-04 20:27:07 | [924][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1022 ntime: 0081 mem: 3.36 + 04-04 20:27:13 | [924][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0405 ntime: 0088 mem: 3.36 + 04-04 20:27:17 | Time info >>>> elapsed: 1383.92 mins remain: 112.21 mins + 04-04 20:27:18 | [925][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 20:27:24 | [925][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0305 ntime: 0072 mem: 3.36 + 04-04 20:27:30 | [925][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1268 ntime: 0077 mem: 3.36 + 04-04 20:27:37 | [925][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1324 ntime: 0081 mem: 3.36 + 04-04 20:27:41 | [925][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0340 ntime: 0084 mem: 3.36 + 04-04 20:27:47 | [925][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0076 ntime: 0080 mem: 3.36 + 04-04 20:27:54 | [925][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0067 ntime: 0081 mem: 3.36 + 04-04 20:28:01 | [925][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0710 ntime: 0075 mem: 3.36 + 04-04 20:28:07 | [925][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0076 mem: 3.36 + 04-04 20:28:16 | [925][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1117 ntime: 0073 mem: 3.36 + 04-04 20:28:26 | [925][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0087 ntime: 0083 mem: 3.36 + 04-04 20:28:33 | [925][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1522 ntime: 0072 mem: 3.36 + 04-04 20:28:41 | [925][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0073 ntime: 0080 mem: 3.36 + 04-04 20:28:48 | [925][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0772 ntime: 0075 mem: 3.36 + 04-04 20:28:58 | [925][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0071 ntime: 0077 mem: 3.36 + 04-04 20:29:05 | [925][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0083 ntime: 0083 mem: 3.36 + 04-04 20:29:12 | [925][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0134 ntime: 0077 mem: 3.36 + 04-04 20:29:20 | [925][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0867 ntime: 0078 mem: 3.36 + 04-04 20:29:26 | Time info >>>> elapsed: 1386.06 mins remain: 110.76 mins + 04-04 20:29:26 | [926][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0106 ntime: 0078 mem: 3.36 + 04-04 20:29:33 | [926][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0120 ntime: 0078 mem: 3.36 + 04-04 20:29:38 | [926][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0116 ntime: 0080 mem: 3.36 + 04-04 20:29:46 | [926][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0658 ntime: 0079 mem: 3.36 + 04-04 20:29:52 | [926][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0421 ntime: 0072 mem: 3.36 + 04-04 20:29:59 | [926][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1036 ntime: 0078 mem: 3.36 + 04-04 20:30:07 | [926][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0163 ntime: 0077 mem: 3.36 + 04-04 20:30:15 | [926][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 20:30:22 | [926][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1338 ntime: 0079 mem: 3.36 + 04-04 20:30:27 | [926][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1052 ntime: 0075 mem: 3.36 + 04-04 20:30:35 | [926][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0184 ntime: 0076 mem: 3.36 + 04-04 20:30:42 | [926][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0993 ntime: 0076 mem: 3.36 + 04-04 20:30:50 | [926][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0075 mem: 3.36 + 04-04 20:30:56 | [926][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0708 ntime: 0077 mem: 3.36 + 04-04 20:31:02 | [926][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0765 ntime: 0083 mem: 3.36 + 04-04 20:31:09 | [926][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1156 ntime: 0081 mem: 3.36 + 04-04 20:31:19 | [926][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1596 ntime: 0078 mem: 3.36 + 04-04 20:31:23 | [926][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0067 ntime: 0085 mem: 3.36 + 04-04 20:31:28 | Time info >>>> elapsed: 1388.10 mins remain: 109.31 mins + 04-04 20:31:28 | [927][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0060 ntime: 0073 mem: 3.36 + 04-04 20:31:36 | [927][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0903 ntime: 0084 mem: 3.36 + 04-04 20:31:43 | [927][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1524 ntime: 0078 mem: 3.36 + 04-04 20:31:50 | [927][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0076 mem: 3.36 + 04-04 20:31:59 | [927][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0953 ntime: 0082 mem: 3.36 + 04-04 20:32:05 | [927][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0090 ntime: 0076 mem: 3.36 + 04-04 20:32:12 | [927][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0714 ntime: 0079 mem: 3.36 + 04-04 20:32:20 | [927][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0881 ntime: 0069 mem: 3.36 + 04-04 20:32:26 | [927][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0121 ntime: 0080 mem: 3.36 + 04-04 20:32:33 | [927][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1082 ntime: 0074 mem: 3.36 + 04-04 20:32:40 | [927][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0737 ntime: 0074 mem: 3.36 + 04-04 20:32:44 | [927][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0651 ntime: 0084 mem: 3.36 + 04-04 20:32:51 | [927][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0128 ntime: 0055 mem: 3.36 + 04-04 20:32:56 | [927][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0916 ntime: 0055 mem: 3.36 + 04-04 20:33:01 | [927][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0440 ntime: 0076 mem: 3.36 + 04-04 20:33:07 | [927][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0475 ntime: 0080 mem: 3.36 + 04-04 20:33:12 | [927][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0065 ntime: 0077 mem: 3.36 + 04-04 20:33:19 | [927][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0117 ntime: 0077 mem: 3.36 + 04-04 20:33:24 | Time info >>>> elapsed: 1390.02 mins remain: 107.85 mins + 04-04 20:33:25 | [928][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1072 ntime: 0076 mem: 3.36 + 04-04 20:33:42 | [928][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1619 ntime: 0076 mem: 3.36 + 04-04 20:33:52 | [928][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1510 ntime: 0071 mem: 3.36 + 04-04 20:33:58 | [928][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 20:34:09 | [928][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0139 ntime: 0082 mem: 3.36 + 04-04 20:34:19 | [928][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0077 mem: 3.36 + 04-04 20:34:30 | [928][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1687 ntime: 0074 mem: 3.36 + 04-04 20:34:36 | [928][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0720 ntime: 0074 mem: 3.36 + 04-04 20:34:45 | [928][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0819 ntime: 0080 mem: 3.36 + 04-04 20:34:54 | [928][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0347 ntime: 0079 mem: 3.36 + 04-04 20:35:01 | [928][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0078 mem: 3.36 + 04-04 20:35:11 | [928][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0088 mem: 3.36 + 04-04 20:35:19 | [928][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0610 ntime: 0073 mem: 3.36 + 04-04 20:35:29 | [928][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2363 ntime: 0079 mem: 3.36 + 04-04 20:35:38 | [928][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0763 ntime: 0074 mem: 3.36 + 04-04 20:35:45 | [928][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0094 ntime: 0081 mem: 3.36 + 04-04 20:35:55 | [928][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1077 ntime: 0079 mem: 3.36 + 04-04 20:36:05 | [928][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0898 ntime: 0082 mem: 3.36 + 04-04 20:36:13 | Time info >>>> elapsed: 1392.84 mins remain: 106.45 mins + 04-04 20:36:14 | [929][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1156 ntime: 0080 mem: 3.36 + 04-04 20:36:22 | [929][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1543 ntime: 0057 mem: 3.36 + 04-04 20:36:32 | [929][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1683 ntime: 0082 mem: 3.36 + 04-04 20:36:41 | [929][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0338 ntime: 0080 mem: 3.36 + 04-04 20:36:47 | [929][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0155 ntime: 0080 mem: 3.36 + 04-04 20:36:53 | [929][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0953 ntime: 0078 mem: 3.36 + 04-04 20:37:00 | [929][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0613 ntime: 0079 mem: 3.36 + 04-04 20:37:06 | [929][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0077 mem: 3.36 + 04-04 20:37:12 | [929][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0099 ntime: 0080 mem: 3.36 + 04-04 20:37:20 | [929][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0685 ntime: 0085 mem: 3.36 + 04-04 20:37:28 | [929][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1396 ntime: 0081 mem: 3.36 + 04-04 20:37:36 | [929][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1145 ntime: 0074 mem: 3.36 + 04-04 20:37:42 | [929][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0982 ntime: 0075 mem: 3.36 + 04-04 20:37:48 | [929][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1144 ntime: 0081 mem: 3.36 + 04-04 20:37:55 | [929][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0149 ntime: 0084 mem: 3.36 + 04-04 20:38:00 | [929][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0360 ntime: 0079 mem: 3.36 + 04-04 20:38:06 | [929][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0923 ntime: 0078 mem: 3.36 + 04-04 20:38:13 | [929][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1175 ntime: 0072 mem: 3.36 + 04-04 20:38:18 | Time info >>>> elapsed: 1394.92 mins remain: 104.99 mins + 04-04 20:38:18 | [930][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0070 mem: 3.36 + 04-04 20:38:25 | [930][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0113 ntime: 0072 mem: 3.36 + 04-04 20:38:32 | [930][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0115 ntime: 0077 mem: 3.36 + 04-04 20:38:38 | [930][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0078 mem: 3.36 + 04-04 20:38:45 | [930][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0098 ntime: 0081 mem: 3.36 + 04-04 20:38:56 | [930][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1088 ntime: 0072 mem: 3.36 + 04-04 20:39:03 | [930][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0348 ntime: 0079 mem: 3.36 + 04-04 20:39:08 | [930][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0372 ntime: 0073 mem: 3.36 + 04-04 20:39:14 | [930][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0085 mem: 3.36 + 04-04 20:39:20 | [930][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0992 ntime: 0078 mem: 3.36 + 04-04 20:39:26 | [930][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0949 ntime: 0079 mem: 3.36 + 04-04 20:39:34 | [930][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0930 ntime: 0078 mem: 3.36 + 04-04 20:39:40 | [930][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0441 ntime: 0075 mem: 3.36 + 04-04 20:39:47 | [930][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1990 ntime: 0081 mem: 3.36 + 04-04 20:39:53 | [930][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0542 ntime: 0080 mem: 3.36 + 04-04 20:40:01 | [930][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0719 ntime: 0082 mem: 3.36 + 04-04 20:40:09 | [930][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0075 mem: 3.36 + 04-04 20:40:15 | [930][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0406 ntime: 0077 mem: 3.36 + 04-04 20:40:19 | Time info >>>> elapsed: 1396.94 mins remain: 103.53 mins + 04-04 20:40:20 | [931][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0665 ntime: 0076 mem: 3.36 + 04-04 20:40:26 | [931][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0138 ntime: 0087 mem: 3.36 + 04-04 20:40:30 | [931][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0807 ntime: 0072 mem: 3.36 + 04-04 20:40:38 | [931][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0893 ntime: 0076 mem: 3.36 + 04-04 20:40:45 | [931][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0065 ntime: 0080 mem: 3.36 + 04-04 20:40:51 | [931][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0078 mem: 3.36 + 04-04 20:40:58 | [931][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0955 ntime: 0081 mem: 3.36 + 04-04 20:41:05 | [931][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1053 ntime: 0080 mem: 3.36 + 04-04 20:41:11 | [931][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0738 ntime: 0070 mem: 3.36 + 04-04 20:41:17 | [931][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0082 mem: 3.36 + 04-04 20:41:23 | [931][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1540 ntime: 0084 mem: 3.36 + 04-04 20:41:31 | [931][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0117 ntime: 0071 mem: 3.36 + 04-04 20:41:38 | [931][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0781 ntime: 0075 mem: 3.36 + 04-04 20:41:43 | [931][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1056 ntime: 0079 mem: 3.36 + 04-04 20:41:48 | [931][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0514 ntime: 0072 mem: 3.36 + 04-04 20:41:56 | [931][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0965 ntime: 0078 mem: 3.36 + 04-04 20:42:02 | [931][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0172 ntime: 0077 mem: 3.36 + 04-04 20:42:08 | [931][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0761 ntime: 0077 mem: 3.36 + 04-04 20:42:12 | Time info >>>> elapsed: 1398.83 mins remain: 102.06 mins + 04-04 20:42:13 | [932][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0080 ntime: 0078 mem: 3.36 + 04-04 20:42:18 | [932][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0065 ntime: 0077 mem: 3.36 + 04-04 20:42:23 | [932][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0078 mem: 3.36 + 04-04 20:42:30 | [932][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0703 ntime: 0076 mem: 3.36 + 04-04 20:42:37 | [932][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0934 ntime: 0076 mem: 3.36 + 04-04 20:42:43 | [932][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1171 ntime: 0071 mem: 3.36 + 04-04 20:42:49 | [932][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0157 ntime: 0081 mem: 3.36 + 04-04 20:42:58 | [932][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0635 ntime: 0078 mem: 3.36 + 04-04 20:43:05 | [932][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1111 ntime: 0071 mem: 3.36 + 04-04 20:43:11 | [932][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0666 ntime: 0076 mem: 3.36 + 04-04 20:43:17 | [932][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1332 ntime: 0085 mem: 3.36 + 04-04 20:43:22 | [932][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0165 ntime: 0080 mem: 3.36 + 04-04 20:43:30 | [932][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0077 mem: 3.36 + 04-04 20:43:37 | [932][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1070 ntime: 0074 mem: 3.36 + 04-04 20:43:43 | [932][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0076 ntime: 0077 mem: 3.36 + 04-04 20:43:48 | [932][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0082 mem: 3.36 + 04-04 20:43:55 | [932][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0948 ntime: 0079 mem: 3.36 + 04-04 20:44:00 | [932][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0580 ntime: 0078 mem: 3.36 + 04-04 20:44:05 | Time info >>>> elapsed: 1400.70 mins remain: 100.59 mins + 04-04 20:44:05 | [933][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0597 ntime: 0079 mem: 3.36 + 04-04 20:44:10 | [933][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0742 ntime: 0089 mem: 3.36 + 04-04 20:44:18 | [933][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0104 ntime: 0079 mem: 3.36 + 04-04 20:44:25 | [933][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0141 ntime: 0074 mem: 3.36 + 04-04 20:44:30 | [933][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0080 mem: 3.36 + 04-04 20:44:36 | [933][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0073 ntime: 0080 mem: 3.36 + 04-04 20:44:42 | [933][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0146 ntime: 0078 mem: 3.36 + 04-04 20:44:50 | [933][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1917 ntime: 0084 mem: 3.36 + 04-04 20:44:57 | [933][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1187 ntime: 0076 mem: 3.36 + 04-04 20:45:03 | [933][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1331 ntime: 0084 mem: 3.36 + 04-04 20:45:08 | [933][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0075 mem: 3.36 + 04-04 20:45:16 | [933][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0682 ntime: 0078 mem: 3.36 + 04-04 20:45:22 | [933][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1127 ntime: 0081 mem: 3.36 + 04-04 20:45:28 | [933][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0080 mem: 3.36 + 04-04 20:45:34 | [933][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0072 ntime: 0080 mem: 3.36 + 04-04 20:45:40 | [933][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0541 ntime: 0080 mem: 3.36 + 04-04 20:45:46 | [933][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1306 ntime: 0077 mem: 3.36 + 04-04 20:45:53 | [933][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0055 mem: 3.36 + 04-04 20:45:56 | Time info >>>> elapsed: 1402.57 mins remain: 99.11 mins + 04-04 20:45:58 | [934][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1353 ntime: 0076 mem: 3.36 + 04-04 20:46:05 | [934][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0334 ntime: 0077 mem: 3.36 + 04-04 20:46:13 | [934][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1314 ntime: 0076 mem: 3.36 + 04-04 20:46:18 | [934][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0076 mem: 3.36 + 04-04 20:46:23 | [934][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0044 ntime: 0077 mem: 3.36 + 04-04 20:46:30 | [934][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0762 ntime: 0076 mem: 3.36 + 04-04 20:46:36 | [934][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0776 ntime: 0085 mem: 3.36 + 04-04 20:46:43 | [934][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0078 mem: 3.36 + 04-04 20:46:49 | [934][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0077 mem: 3.36 + 04-04 20:46:56 | [934][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1173 ntime: 0084 mem: 3.36 + 04-04 20:47:01 | [934][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0470 ntime: 0073 mem: 3.36 + 04-04 20:47:08 | [934][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0620 ntime: 0076 mem: 3.36 + 04-04 20:47:15 | [934][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0898 ntime: 0073 mem: 3.36 + 04-04 20:47:21 | [934][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0641 ntime: 0080 mem: 3.36 + 04-04 20:47:28 | [934][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1136 ntime: 0076 mem: 3.36 + 04-04 20:47:33 | [934][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0447 ntime: 0078 mem: 3.36 + 04-04 20:47:39 | [934][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0292 ntime: 0079 mem: 3.36 + 04-04 20:47:47 | [934][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0735 ntime: 0075 mem: 3.36 + 04-04 20:47:53 | Time info >>>> elapsed: 1404.50 mins remain: 97.64 mins + 04-04 20:47:53 | [935][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0673 ntime: 0077 mem: 3.36 + 04-04 20:48:00 | [935][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0095 ntime: 0087 mem: 3.36 + 04-04 20:48:08 | [935][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0075 mem: 3.36 + 04-04 20:48:16 | [935][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0098 ntime: 0075 mem: 3.36 + 04-04 20:48:21 | [935][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0076 mem: 3.36 + 04-04 20:48:28 | [935][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1075 ntime: 0078 mem: 3.36 + 04-04 20:48:33 | [935][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0075 mem: 3.36 + 04-04 20:48:37 | [935][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0171 ntime: 0078 mem: 3.36 + 04-04 20:48:43 | [935][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0086 ntime: 0073 mem: 3.36 + 04-04 20:48:49 | [935][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0535 ntime: 0072 mem: 3.36 + 04-04 20:48:55 | [935][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0128 ntime: 0076 mem: 3.36 + 04-04 20:49:02 | [935][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0904 ntime: 0071 mem: 3.36 + 04-04 20:49:09 | [935][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1231 ntime: 0079 mem: 3.36 + 04-04 20:49:14 | [935][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0883 ntime: 0081 mem: 3.36 + 04-04 20:49:22 | [935][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1086 ntime: 0078 mem: 3.36 + 04-04 20:49:27 | [935][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0093 ntime: 0078 mem: 3.36 + 04-04 20:49:33 | [935][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0156 ntime: 0076 mem: 3.36 + 04-04 20:49:38 | [935][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0084 ntime: 0078 mem: 3.36 + 04-04 20:49:44 | Time info >>>> elapsed: 1406.36 mins remain: 96.16 mins + 04-04 20:49:45 | [936][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1125 ntime: 0079 mem: 3.36 + 04-04 20:49:50 | [936][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0121 ntime: 0079 mem: 3.36 + 04-04 20:49:57 | [936][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0655 ntime: 0081 mem: 3.36 + 04-04 20:50:04 | [936][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0094 ntime: 0084 mem: 3.36 + 04-04 20:50:10 | [936][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1169 ntime: 0083 mem: 3.36 + 04-04 20:50:17 | [936][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0810 ntime: 0073 mem: 3.36 + 04-04 20:50:26 | [936][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1497 ntime: 0075 mem: 3.36 + 04-04 20:50:34 | [936][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1376 ntime: 0072 mem: 3.36 + 04-04 20:50:39 | [936][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0771 ntime: 0079 mem: 3.36 + 04-04 20:50:45 | [936][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0118 ntime: 0074 mem: 3.36 + 04-04 20:50:52 | [936][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0737 ntime: 0071 mem: 3.36 + 04-04 20:50:58 | [936][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0080 mem: 3.36 + 04-04 20:51:07 | [936][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1306 ntime: 0080 mem: 3.36 + 04-04 20:51:14 | [936][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0079 mem: 3.36 + 04-04 20:51:21 | [936][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0075 mem: 3.36 + 04-04 20:51:28 | [936][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1093 ntime: 0081 mem: 3.36 + 04-04 20:51:35 | [936][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1811 ntime: 0081 mem: 3.36 + 04-04 20:51:41 | [936][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0188 ntime: 0082 mem: 3.36 + 04-04 20:51:45 | Time info >>>> elapsed: 1408.38 mins remain: 94.69 mins + 04-04 20:51:45 | [937][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0049 ntime: 0081 mem: 3.36 + 04-04 20:51:50 | [937][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0065 ntime: 0081 mem: 3.36 + 04-04 20:51:58 | [937][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0841 ntime: 0057 mem: 3.36 + 04-04 20:52:02 | [937][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0079 mem: 3.36 + 04-04 20:52:09 | [937][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0603 ntime: 0072 mem: 3.36 + 04-04 20:52:15 | [937][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 20:52:22 | [937][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0935 ntime: 0073 mem: 3.36 + 04-04 20:52:31 | [937][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1008 ntime: 0082 mem: 3.36 + 04-04 20:52:37 | [937][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1233 ntime: 0077 mem: 3.36 + 04-04 20:52:42 | [937][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0072 ntime: 0075 mem: 3.36 + 04-04 20:52:46 | [937][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0463 ntime: 0074 mem: 3.36 + 04-04 20:52:55 | [937][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1088 ntime: 0075 mem: 3.36 + 04-04 20:53:00 | [937][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0889 ntime: 0078 mem: 3.36 + 04-04 20:53:07 | [937][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0951 ntime: 0080 mem: 3.36 + 04-04 20:53:12 | [937][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0078 mem: 3.36 + 04-04 20:53:20 | [937][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1204 ntime: 0083 mem: 3.36 + 04-04 20:53:28 | [937][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0739 ntime: 0079 mem: 3.36 + 04-04 20:53:33 | [937][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0082 ntime: 0075 mem: 3.36 + 04-04 20:53:41 | Time info >>>> elapsed: 1410.30 mins remain: 93.22 mins + 04-04 20:53:41 | [938][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0120 ntime: 0084 mem: 3.36 + 04-04 20:53:47 | [938][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0896 ntime: 0078 mem: 3.36 + 04-04 20:53:54 | [938][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0109 ntime: 0074 mem: 3.36 + 04-04 20:53:59 | [938][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0099 ntime: 0071 mem: 3.36 + 04-04 20:54:05 | [938][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0077 mem: 3.36 + 04-04 20:54:11 | [938][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0086 ntime: 0079 mem: 3.36 + 04-04 20:54:17 | [938][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0133 ntime: 0080 mem: 3.36 + 04-04 20:54:24 | [938][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0076 mem: 3.36 + 04-04 20:54:30 | [938][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0836 ntime: 0080 mem: 3.36 + 04-04 20:54:39 | [938][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0401 ntime: 0084 mem: 3.36 + 04-04 20:54:44 | [938][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0757 ntime: 0076 mem: 3.36 + 04-04 20:54:50 | [938][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0900 ntime: 0079 mem: 3.36 + 04-04 20:54:58 | [938][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0893 ntime: 0072 mem: 3.36 + 04-04 20:55:03 | [938][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1354 ntime: 0074 mem: 3.36 + 04-04 20:55:09 | [938][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1469 ntime: 0074 mem: 3.36 + 04-04 20:55:16 | [938][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0902 ntime: 0074 mem: 3.36 + 04-04 20:55:24 | [938][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0328 ntime: 0055 mem: 3.36 + 04-04 20:55:31 | [938][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1627 ntime: 0074 mem: 3.36 + 04-04 20:55:35 | Time info >>>> elapsed: 1412.21 mins remain: 91.74 mins + 04-04 20:55:35 | [939][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0096 ntime: 0076 mem: 3.36 + 04-04 20:55:40 | [939][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0644 ntime: 0077 mem: 3.36 + 04-04 20:55:46 | [939][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1401 ntime: 0076 mem: 3.36 + 04-04 20:55:52 | [939][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0082 ntime: 0076 mem: 3.36 + 04-04 20:55:57 | [939][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0607 ntime: 0077 mem: 3.36 + 04-04 20:56:02 | [939][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0110 ntime: 0079 mem: 3.36 + 04-04 20:56:10 | [939][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1084 ntime: 0074 mem: 3.36 + 04-04 20:56:15 | [939][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0092 ntime: 0077 mem: 3.36 + 04-04 20:56:22 | [939][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0632 ntime: 0072 mem: 3.36 + 04-04 20:56:26 | [939][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0110 ntime: 0079 mem: 3.36 + 04-04 20:56:33 | [939][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0083 mem: 3.36 + 04-04 20:56:40 | [939][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0895 ntime: 0077 mem: 3.36 + 04-04 20:56:46 | [939][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 20:56:50 | [939][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0921 ntime: 0079 mem: 3.36 + 04-04 20:56:57 | [939][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0424 ntime: 0076 mem: 3.36 + 04-04 20:57:00 | [939][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0419 ntime: 0076 mem: 3.36 + 04-04 20:57:06 | [939][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0708 ntime: 0072 mem: 3.36 + 04-04 20:57:11 | [939][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0179 ntime: 0081 mem: 3.36 + 04-04 20:57:17 | Time info >>>> elapsed: 1413.91 mins remain: 90.25 mins + 04-04 20:57:18 | [940][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0855 ntime: 0069 mem: 3.36 + 04-04 20:57:34 | [940][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1982 ntime: 0080 mem: 3.36 + 04-04 20:57:41 | [940][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0072 ntime: 0073 mem: 3.36 + 04-04 20:57:51 | [940][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0711 ntime: 0082 mem: 3.36 + 04-04 20:57:57 | [940][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0903 ntime: 0071 mem: 3.36 + 04-04 20:58:05 | [940][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0559 ntime: 0079 mem: 3.36 + 04-04 20:58:13 | [940][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0137 ntime: 0079 mem: 3.36 + 04-04 20:58:22 | [940][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0510 ntime: 0078 mem: 3.36 + 04-04 20:58:27 | [940][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0087 ntime: 0080 mem: 3.36 + 04-04 20:58:34 | [940][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0464 ntime: 0076 mem: 3.36 + 04-04 20:58:40 | [940][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0187 ntime: 0074 mem: 3.36 + 04-04 20:58:49 | [940][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0834 ntime: 0072 mem: 3.36 + 04-04 20:58:58 | [940][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0101 ntime: 0090 mem: 3.36 + 04-04 20:59:05 | [940][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1704 ntime: 0074 mem: 3.36 + 04-04 20:59:13 | [940][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1250 ntime: 0082 mem: 3.36 + 04-04 20:59:24 | [940][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1403 ntime: 0072 mem: 3.36 + 04-04 20:59:33 | [940][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0147 ntime: 0076 mem: 3.36 + 04-04 20:59:44 | [940][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1267 ntime: 0081 mem: 3.36 + 04-04 20:59:51 | Time info >>>> elapsed: 1416.47 mins remain: 88.81 mins + 04-04 20:59:52 | [941][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1622 ntime: 0080 mem: 3.36 + 04-04 21:00:00 | [941][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 21:00:07 | [941][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0085 ntime: 0076 mem: 3.36 + 04-04 21:00:14 | [941][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0601 ntime: 0077 mem: 3.36 + 04-04 21:00:21 | [941][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0078 mem: 3.36 + 04-04 21:00:30 | [941][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0958 ntime: 0077 mem: 3.36 + 04-04 21:00:37 | [941][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 21:00:42 | [941][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0080 mem: 3.36 + 04-04 21:00:49 | [941][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0688 ntime: 0075 mem: 3.36 + 04-04 21:00:55 | [941][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0983 ntime: 0072 mem: 3.36 + 04-04 21:01:02 | [941][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1507 ntime: 0073 mem: 3.36 + 04-04 21:01:07 | [941][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0081 mem: 3.36 + 04-04 21:01:13 | [941][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0614 ntime: 0078 mem: 3.36 + 04-04 21:01:19 | [941][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0380 ntime: 0081 mem: 3.36 + 04-04 21:01:24 | [941][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0679 ntime: 0078 mem: 3.36 + 04-04 21:01:29 | [941][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0076 mem: 3.36 + 04-04 21:01:34 | [941][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0140 ntime: 0056 mem: 3.36 + 04-04 21:01:41 | [941][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0333 ntime: 0077 mem: 3.36 + 04-04 21:01:47 | Time info >>>> elapsed: 1418.41 mins remain: 87.33 mins + 04-04 21:01:47 | [942][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0164 ntime: 0080 mem: 3.36 + 04-04 21:01:53 | [942][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0846 ntime: 0074 mem: 3.36 + 04-04 21:01:59 | [942][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0426 ntime: 0076 mem: 3.36 + 04-04 21:02:06 | [942][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0118 ntime: 0079 mem: 3.36 + 04-04 21:02:12 | [942][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0094 ntime: 0075 mem: 3.36 + 04-04 21:02:17 | [942][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0502 ntime: 0087 mem: 3.36 + 04-04 21:02:22 | [942][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1351 ntime: 0081 mem: 3.36 + 04-04 21:02:29 | [942][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1088 ntime: 0080 mem: 3.36 + 04-04 21:02:33 | [942][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0076 mem: 3.36 + 04-04 21:02:38 | [942][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0080 ntime: 0076 mem: 3.36 + 04-04 21:02:44 | [942][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0172 ntime: 0075 mem: 3.36 + 04-04 21:02:52 | [942][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0086 ntime: 0080 mem: 3.36 + 04-04 21:02:59 | [942][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0072 mem: 3.36 + 04-04 21:03:05 | [942][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1007 ntime: 0078 mem: 3.36 + 04-04 21:03:14 | [942][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 3569 ntime: 0085 mem: 3.36 + 04-04 21:03:23 | [942][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 21:03:29 | [942][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1206 ntime: 0078 mem: 3.36 + 04-04 21:03:37 | [942][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0908 ntime: 0079 mem: 3.36 + 04-04 21:03:42 | Time info >>>> elapsed: 1420.33 mins remain: 85.85 mins + 04-04 21:03:43 | [943][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0712 ntime: 0071 mem: 3.36 + 04-04 21:03:48 | [943][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0096 ntime: 0077 mem: 3.36 + 04-04 21:03:54 | [943][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0093 ntime: 0086 mem: 3.36 + 04-04 21:04:01 | [943][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0707 ntime: 0076 mem: 3.36 + 04-04 21:04:06 | [943][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1375 ntime: 0078 mem: 3.36 + 04-04 21:04:12 | [943][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0176 ntime: 0086 mem: 3.36 + 04-04 21:04:16 | [943][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0563 ntime: 0076 mem: 3.36 + 04-04 21:04:22 | [943][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0528 ntime: 0077 mem: 3.36 + 04-04 21:04:28 | [943][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0078 mem: 3.36 + 04-04 21:04:34 | [943][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0709 ntime: 0079 mem: 3.36 + 04-04 21:04:40 | [943][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1105 ntime: 0080 mem: 3.36 + 04-04 21:04:46 | [943][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0968 ntime: 0083 mem: 3.36 + 04-04 21:04:52 | [943][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0922 ntime: 0074 mem: 3.36 + 04-04 21:04:57 | [943][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0077 mem: 3.36 + 04-04 21:05:03 | [943][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0704 ntime: 0077 mem: 3.36 + 04-04 21:05:07 | [943][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0549 ntime: 0080 mem: 3.36 + 04-04 21:05:13 | [943][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1059 ntime: 0084 mem: 3.36 + 04-04 21:05:19 | [943][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1055 ntime: 0079 mem: 3.36 + 04-04 21:05:22 | Time info >>>> elapsed: 1422.00 mins remain: 84.36 mins + 04-04 21:05:24 | [944][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1257 ntime: 0077 mem: 3.36 + 04-04 21:05:28 | [944][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0730 ntime: 0082 mem: 3.36 + 04-04 21:05:32 | [944][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0158 ntime: 0079 mem: 3.36 + 04-04 21:05:37 | [944][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0111 ntime: 0080 mem: 3.36 + 04-04 21:05:43 | [944][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0775 ntime: 0082 mem: 3.36 + 04-04 21:05:48 | [944][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0623 ntime: 0073 mem: 3.36 + 04-04 21:05:53 | [944][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0086 mem: 3.36 + 04-04 21:06:00 | [944][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0648 ntime: 0084 mem: 3.36 + 04-04 21:06:03 | [944][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0189 ntime: 0076 mem: 3.36 + 04-04 21:06:09 | [944][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0966 ntime: 0084 mem: 3.36 + 04-04 21:06:14 | [944][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1037 ntime: 0078 mem: 3.36 + 04-04 21:06:20 | [944][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0069 ntime: 0084 mem: 3.36 + 04-04 21:06:26 | [944][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0497 ntime: 0077 mem: 3.36 + 04-04 21:06:32 | [944][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1273 ntime: 0071 mem: 3.36 + 04-04 21:06:38 | [944][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0803 ntime: 0073 mem: 3.36 + 04-04 21:06:43 | [944][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0072 mem: 3.36 + 04-04 21:06:48 | [944][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0578 ntime: 0077 mem: 3.36 + 04-04 21:06:54 | [944][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1219 ntime: 0079 mem: 3.36 + 04-04 21:06:58 | Time info >>>> elapsed: 1423.59 mins remain: 82.85 mins + 04-04 21:06:59 | [945][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1033 ntime: 0087 mem: 3.36 + 04-04 21:07:05 | [945][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1082 ntime: 0083 mem: 3.36 + 04-04 21:07:10 | [945][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0123 ntime: 0079 mem: 3.36 + 04-04 21:07:15 | [945][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1042 ntime: 0078 mem: 3.36 + 04-04 21:07:20 | [945][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1059 ntime: 0072 mem: 3.36 + 04-04 21:07:27 | [945][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1094 ntime: 0080 mem: 3.36 + 04-04 21:07:32 | [945][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0065 ntime: 0081 mem: 3.36 + 04-04 21:07:39 | [945][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0097 ntime: 0081 mem: 3.36 + 04-04 21:07:45 | [945][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0525 ntime: 0080 mem: 3.36 + 04-04 21:07:51 | [945][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1126 ntime: 0073 mem: 3.36 + 04-04 21:07:59 | [945][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0149 ntime: 0082 mem: 3.36 + 04-04 21:08:04 | [945][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0121 ntime: 0077 mem: 3.36 + 04-04 21:08:08 | [945][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0072 mem: 3.36 + 04-04 21:08:14 | [945][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0121 ntime: 0073 mem: 3.36 + 04-04 21:08:30 | [945][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2029 ntime: 0081 mem: 3.36 + 04-04 21:08:40 | [945][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1992 ntime: 0081 mem: 3.36 + 04-04 21:08:48 | [945][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1952 ntime: 0089 mem: 3.36 + 04-04 21:08:54 | [945][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0702 ntime: 0081 mem: 3.36 + 04-04 21:09:00 | Time info >>>> elapsed: 1425.63 mins remain: 81.38 mins + 04-04 21:09:01 | [946][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0949 ntime: 0076 mem: 3.36 + 04-04 21:09:08 | [946][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1132 ntime: 0080 mem: 3.36 + 04-04 21:09:15 | [946][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0158 ntime: 0076 mem: 3.36 + 04-04 21:09:23 | [946][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1213 ntime: 0079 mem: 3.36 + 04-04 21:09:30 | [946][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1092 ntime: 0077 mem: 3.36 + 04-04 21:09:36 | [946][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0065 ntime: 0083 mem: 3.36 + 04-04 21:09:43 | [946][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1043 ntime: 0079 mem: 3.36 + 04-04 21:09:51 | [946][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1412 ntime: 0083 mem: 3.36 + 04-04 21:09:59 | [946][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1246 ntime: 0080 mem: 3.36 + 04-04 21:10:06 | [946][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2173 ntime: 0083 mem: 3.36 + 04-04 21:10:14 | [946][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1736 ntime: 0076 mem: 3.36 + 04-04 21:10:20 | [946][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0634 ntime: 0079 mem: 3.36 + 04-04 21:10:28 | [946][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1015 ntime: 0077 mem: 3.36 + 04-04 21:10:36 | [946][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0122 ntime: 0078 mem: 3.36 + 04-04 21:10:42 | [946][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0083 mem: 3.36 + 04-04 21:10:50 | [946][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0075 mem: 3.36 + 04-04 21:10:57 | [946][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0082 mem: 3.36 + 04-04 21:11:04 | [946][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0075 mem: 3.36 + 04-04 21:11:08 | Time info >>>> elapsed: 1427.76 mins remain: 79.91 mins + 04-04 21:11:09 | [947][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 1172 ntime: 0064 mem: 3.36 + 04-04 21:11:16 | [947][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0530 ntime: 0078 mem: 3.36 + 04-04 21:11:21 | [947][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0077 ntime: 0079 mem: 3.36 + 04-04 21:11:26 | [947][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0642 ntime: 0078 mem: 3.36 + 04-04 21:11:31 | [947][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1035 ntime: 0077 mem: 3.36 + 04-04 21:11:37 | [947][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1246 ntime: 0081 mem: 3.36 + 04-04 21:11:42 | [947][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 21:11:48 | [947][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0782 ntime: 0077 mem: 3.36 + 04-04 21:11:54 | [947][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0723 ntime: 0080 mem: 3.36 + 04-04 21:11:58 | [947][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1277 ntime: 0074 mem: 3.36 + 04-04 21:12:02 | [947][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0077 mem: 3.36 + 04-04 21:12:09 | [947][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0606 ntime: 0078 mem: 3.36 + 04-04 21:12:14 | [947][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0925 ntime: 0076 mem: 3.36 + 04-04 21:12:19 | [947][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0076 ntime: 0085 mem: 3.36 + 04-04 21:12:23 | [947][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1077 ntime: 0074 mem: 3.36 + 04-04 21:12:27 | [947][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0083 ntime: 0079 mem: 3.36 + 04-04 21:12:33 | [947][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0119 ntime: 0084 mem: 3.36 + 04-04 21:12:38 | [947][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0330 ntime: 0081 mem: 3.36 + 04-04 21:12:42 | Time info >>>> elapsed: 1429.33 mins remain: 78.40 mins + 04-04 21:12:43 | [948][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0077 mem: 3.36 + 04-04 21:12:50 | [948][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0080 mem: 3.36 + 04-04 21:12:53 | [948][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0310 ntime: 0079 mem: 3.36 + 04-04 21:12:58 | [948][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0868 ntime: 0080 mem: 3.36 + 04-04 21:13:02 | [948][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0798 ntime: 0085 mem: 3.36 + 04-04 21:13:06 | [948][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0065 ntime: 0072 mem: 3.36 + 04-04 21:13:10 | [948][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0288 ntime: 0077 mem: 3.36 + 04-04 21:13:15 | [948][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0812 ntime: 0074 mem: 3.36 + 04-04 21:13:21 | [948][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0642 ntime: 0074 mem: 3.36 + 04-04 21:13:25 | [948][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0082 ntime: 0084 mem: 3.36 + 04-04 21:13:31 | [948][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0119 ntime: 0083 mem: 3.36 + 04-04 21:13:35 | [948][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0135 ntime: 0078 mem: 3.36 + 04-04 21:13:41 | [948][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1409 ntime: 0081 mem: 3.36 + 04-04 21:13:46 | [948][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0149 ntime: 0083 mem: 3.36 + 04-04 21:13:50 | [948][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0100 ntime: 0080 mem: 3.36 + 04-04 21:13:55 | [948][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 21:13:59 | [948][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0644 ntime: 0085 mem: 3.36 + 04-04 21:14:04 | [948][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0067 ntime: 0079 mem: 3.36 + 04-04 21:14:08 | Time info >>>> elapsed: 1430.76 mins remain: 76.89 mins + 04-04 21:14:08 | [949][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0104 ntime: 0075 mem: 3.36 + 04-04 21:14:12 | [949][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0162 ntime: 0082 mem: 3.36 + 04-04 21:14:17 | [949][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0279 ntime: 0082 mem: 3.36 + 04-04 21:14:24 | [949][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0077 ntime: 0074 mem: 3.36 + 04-04 21:14:30 | [949][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0965 ntime: 0077 mem: 3.36 + 04-04 21:14:34 | [949][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0079 mem: 3.36 + 04-04 21:14:39 | [949][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0128 ntime: 0073 mem: 3.36 + 04-04 21:14:44 | [949][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0802 ntime: 0081 mem: 3.36 + 04-04 21:14:51 | [949][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1382 ntime: 0082 mem: 3.36 + 04-04 21:14:56 | [949][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0096 ntime: 0079 mem: 3.36 + 04-04 21:15:01 | [949][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0518 ntime: 0086 mem: 3.36 + 04-04 21:15:05 | [949][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0341 ntime: 0079 mem: 3.36 + 04-04 21:15:11 | [949][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0401 ntime: 0083 mem: 3.36 + 04-04 21:15:16 | [949][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0656 ntime: 0073 mem: 3.36 + 04-04 21:15:21 | [949][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0071 ntime: 0079 mem: 3.36 + 04-04 21:15:26 | [949][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0582 ntime: 0077 mem: 3.36 + 04-04 21:15:34 | [949][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1371 ntime: 0081 mem: 3.36 + 04-04 21:15:40 | [949][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1072 ntime: 0076 mem: 3.36 + 04-04 21:15:43 | Time info >>>> elapsed: 1432.35 mins remain: 75.39 mins + 04-04 21:15:44 | [950][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0329 ntime: 0071 mem: 3.36 + 04-04 21:15:51 | [950][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0576 ntime: 0084 mem: 3.36 + 04-04 21:15:58 | [950][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0071 mem: 3.36 + 04-04 21:16:04 | [950][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0081 ntime: 0075 mem: 3.36 + 04-04 21:16:10 | [950][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0063 ntime: 0077 mem: 3.36 + 04-04 21:16:15 | [950][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0076 mem: 3.36 + 04-04 21:16:21 | [950][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0080 mem: 3.36 + 04-04 21:16:27 | [950][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1094 ntime: 0078 mem: 3.36 + 04-04 21:16:35 | [950][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0095 ntime: 0082 mem: 3.36 + 04-04 21:16:43 | [950][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0971 ntime: 0076 mem: 3.36 + 04-04 21:16:50 | [950][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0829 ntime: 0073 mem: 3.36 + 04-04 21:16:58 | [950][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1413 ntime: 0079 mem: 3.36 + 04-04 21:17:07 | [950][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0100 ntime: 0085 mem: 3.36 + 04-04 21:17:14 | [950][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0568 ntime: 0079 mem: 3.36 + 04-04 21:17:21 | [950][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0086 ntime: 0057 mem: 3.36 + 04-04 21:17:27 | [950][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0087 ntime: 0068 mem: 3.36 + 04-04 21:17:34 | [950][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0138 ntime: 0072 mem: 3.36 + 04-04 21:17:41 | [950][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0072 mem: 3.36 + 04-04 21:17:46 | Time info >>>> elapsed: 1434.39 mins remain: 73.91 mins + 04-04 21:17:47 | [951][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1286 ntime: 0078 mem: 3.36 + 04-04 21:17:54 | [951][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1106 ntime: 0076 mem: 3.36 + 04-04 21:18:02 | [951][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0078 mem: 3.36 + 04-04 21:18:07 | [951][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0193 ntime: 0078 mem: 3.36 + 04-04 21:18:13 | [951][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0127 ntime: 0079 mem: 3.36 + 04-04 21:18:20 | [951][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0138 ntime: 0076 mem: 3.36 + 04-04 21:18:26 | [951][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0712 ntime: 0077 mem: 3.36 + 04-04 21:18:33 | [951][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0112 ntime: 0074 mem: 3.36 + 04-04 21:18:39 | [951][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0079 mem: 3.36 + 04-04 21:18:51 | [951][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2254 ntime: 0081 mem: 3.36 + 04-04 21:18:56 | [951][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0725 ntime: 0081 mem: 3.36 + 04-04 21:19:03 | [951][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0563 ntime: 0077 mem: 3.36 + 04-04 21:19:10 | [951][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0114 ntime: 0077 mem: 3.36 + 04-04 21:19:19 | [951][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1081 ntime: 0077 mem: 3.36 + 04-04 21:19:27 | [951][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1272 ntime: 0080 mem: 3.36 + 04-04 21:19:37 | [951][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1346 ntime: 0074 mem: 3.36 + 04-04 21:19:47 | [951][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1655 ntime: 0074 mem: 3.36 + 04-04 21:20:00 | [951][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1781 ntime: 0080 mem: 3.36 + 04-04 21:20:08 | Time info >>>> elapsed: 1436.76 mins remain: 72.44 mins + 04-04 21:20:09 | [952][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1279 ntime: 0085 mem: 3.36 + 04-04 21:20:14 | [952][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0850 ntime: 0057 mem: 3.36 + 04-04 21:20:22 | [952][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0073 mem: 3.36 + 04-04 21:20:35 | [952][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2440 ntime: 0078 mem: 3.36 + 04-04 21:20:44 | [952][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1049 ntime: 0073 mem: 3.36 + 04-04 21:20:52 | [952][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1471 ntime: 0071 mem: 3.36 + 04-04 21:20:57 | [952][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0618 ntime: 0070 mem: 3.36 + 04-04 21:21:06 | [952][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1040 ntime: 0075 mem: 3.36 + 04-04 21:21:12 | [952][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0875 ntime: 0082 mem: 3.36 + 04-04 21:21:19 | [952][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0581 ntime: 0076 mem: 3.36 + 04-04 21:21:26 | [952][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1410 ntime: 0081 mem: 3.36 + 04-04 21:21:32 | [952][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0079 ntime: 0092 mem: 3.36 + 04-04 21:21:38 | [952][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0720 ntime: 0079 mem: 3.36 + 04-04 21:21:45 | [952][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0115 ntime: 0079 mem: 3.36 + 04-04 21:21:54 | [952][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1443 ntime: 0082 mem: 3.36 + 04-04 21:21:59 | [952][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0077 mem: 3.36 + 04-04 21:22:10 | [952][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1634 ntime: 0078 mem: 3.36 + 04-04 21:22:17 | [952][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 21:22:24 | Time info >>>> elapsed: 1439.03 mins remain: 70.97 mins + 04-04 21:22:26 | [953][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1511 ntime: 0074 mem: 3.36 + 04-04 21:22:32 | [953][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0072 mem: 3.36 + 04-04 21:22:41 | [953][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1678 ntime: 0079 mem: 3.36 + 04-04 21:22:50 | [953][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0455 ntime: 0079 mem: 3.36 + 04-04 21:22:57 | [953][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0876 ntime: 0082 mem: 3.36 + 04-04 21:23:04 | [953][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0299 ntime: 0076 mem: 3.36 + 04-04 21:23:11 | [953][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1416 ntime: 0083 mem: 3.36 + 04-04 21:23:19 | [953][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0462 ntime: 0074 mem: 3.36 + 04-04 21:23:29 | [953][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0718 ntime: 0074 mem: 3.36 + 04-04 21:23:35 | [953][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0186 ntime: 0077 mem: 3.36 + 04-04 21:23:41 | [953][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0703 ntime: 0077 mem: 3.36 + 04-04 21:23:49 | [953][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0139 ntime: 0080 mem: 3.36 + 04-04 21:23:56 | [953][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0085 mem: 3.36 + 04-04 21:24:04 | [953][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0088 mem: 3.36 + 04-04 21:24:10 | [953][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1192 ntime: 0077 mem: 3.36 + 04-04 21:24:17 | [953][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1215 ntime: 0078 mem: 3.36 + 04-04 21:24:24 | [953][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0276 ntime: 0075 mem: 3.36 + 04-04 21:24:31 | [953][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1514 ntime: 0075 mem: 3.36 + 04-04 21:24:39 | Time info >>>> elapsed: 1441.27 mins remain: 69.50 mins + 04-04 21:24:39 | [954][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0082 mem: 3.36 + 04-04 21:24:46 | [954][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0170 ntime: 0076 mem: 3.36 + 04-04 21:24:51 | [954][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0152 ntime: 0078 mem: 3.36 + 04-04 21:24:58 | [954][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0068 ntime: 0081 mem: 3.36 + 04-04 21:25:06 | [954][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0147 ntime: 0077 mem: 3.36 + 04-04 21:25:12 | [954][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1189 ntime: 0082 mem: 3.36 + 04-04 21:25:20 | [954][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0792 ntime: 0084 mem: 3.36 + 04-04 21:25:27 | [954][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0663 ntime: 0081 mem: 3.36 + 04-04 21:25:36 | [954][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0790 ntime: 0075 mem: 3.36 + 04-04 21:25:46 | [954][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1494 ntime: 0073 mem: 3.36 + 04-04 21:25:54 | [954][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1190 ntime: 0084 mem: 3.36 + 04-04 21:26:01 | [954][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0698 ntime: 0076 mem: 3.36 + 04-04 21:26:09 | [954][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0090 ntime: 0095 mem: 3.36 + 04-04 21:26:18 | [954][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1146 ntime: 0073 mem: 3.36 + 04-04 21:26:25 | [954][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1077 ntime: 0077 mem: 3.36 + 04-04 21:26:32 | [954][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0626 ntime: 0078 mem: 3.36 + 04-04 21:26:41 | [954][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1390 ntime: 0083 mem: 3.36 + 04-04 21:26:50 | [954][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0063 ntime: 0087 mem: 3.36 + 04-04 21:26:53 | Time info >>>> elapsed: 1443.52 mins remain: 68.02 mins + 04-04 21:26:55 | [955][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1558 ntime: 0076 mem: 3.36 + 04-04 21:27:02 | [955][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1069 ntime: 0074 mem: 3.36 + 04-04 21:27:09 | [955][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1386 ntime: 0078 mem: 3.36 + 04-04 21:27:15 | [955][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0147 ntime: 0079 mem: 3.36 + 04-04 21:27:24 | [955][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1348 ntime: 0071 mem: 3.36 + 04-04 21:27:32 | [955][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1084 ntime: 0077 mem: 3.36 + 04-04 21:27:38 | [955][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0934 ntime: 0076 mem: 3.36 + 04-04 21:27:42 | [955][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0098 ntime: 0079 mem: 3.36 + 04-04 21:27:47 | [955][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0079 mem: 3.36 + 04-04 21:27:54 | [955][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0080 mem: 3.36 + 04-04 21:28:03 | [955][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1082 ntime: 0081 mem: 3.36 + 04-04 21:28:10 | [955][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1582 ntime: 0075 mem: 3.36 + 04-04 21:28:15 | [955][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0078 mem: 3.36 + 04-04 21:28:21 | [955][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0963 ntime: 0077 mem: 3.36 + 04-04 21:28:28 | [955][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1001 ntime: 0080 mem: 3.36 + 04-04 21:28:36 | [955][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1511 ntime: 0077 mem: 3.36 + 04-04 21:28:44 | [955][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1098 ntime: 0079 mem: 3.36 + 04-04 21:28:51 | [955][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1756 ntime: 0077 mem: 3.36 + 04-04 21:28:58 | Time info >>>> elapsed: 1445.59 mins remain: 66.53 mins + 04-04 21:28:58 | [956][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 21:29:07 | [956][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1835 ntime: 0079 mem: 3.36 + 04-04 21:29:13 | [956][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1107 ntime: 0076 mem: 3.36 + 04-04 21:29:20 | [956][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0706 ntime: 0079 mem: 3.36 + 04-04 21:29:25 | [956][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0073 mem: 3.36 + 04-04 21:29:34 | [956][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1301 ntime: 0079 mem: 3.36 + 04-04 21:29:41 | [956][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1008 ntime: 0074 mem: 3.36 + 04-04 21:29:46 | [956][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0226 ntime: 0082 mem: 3.36 + 04-04 21:29:52 | [956][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1552 ntime: 0073 mem: 3.36 + 04-04 21:30:00 | [956][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1679 ntime: 0083 mem: 3.36 + 04-04 21:30:08 | [956][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1250 ntime: 0078 mem: 3.36 + 04-04 21:30:17 | [956][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0121 ntime: 0076 mem: 3.36 + 04-04 21:30:22 | [956][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1148 ntime: 0076 mem: 3.36 + 04-04 21:30:28 | [956][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0079 ntime: 0073 mem: 3.36 + 04-04 21:30:37 | [956][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1157 ntime: 0078 mem: 3.36 + 04-04 21:30:45 | [956][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1282 ntime: 0071 mem: 3.36 + 04-04 21:30:53 | [956][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0958 ntime: 0079 mem: 3.36 + 04-04 21:31:00 | [956][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0082 mem: 3.36 + 04-04 21:31:05 | Time info >>>> elapsed: 1447.70 mins remain: 65.05 mins + 04-04 21:31:05 | [957][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0067 ntime: 0075 mem: 3.36 + 04-04 21:31:13 | [957][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1370 ntime: 0075 mem: 3.36 + 04-04 21:31:19 | [957][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0080 mem: 3.36 + 04-04 21:31:28 | [957][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0083 mem: 3.36 + 04-04 21:31:35 | [957][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1347 ntime: 0080 mem: 3.36 + 04-04 21:31:39 | [957][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0073 mem: 3.36 + 04-04 21:31:46 | [957][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0886 ntime: 0080 mem: 3.36 + 04-04 21:31:56 | [957][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0088 ntime: 0077 mem: 3.36 + 04-04 21:32:04 | [957][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1121 ntime: 0081 mem: 3.36 + 04-04 21:32:09 | [957][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0671 ntime: 0073 mem: 3.36 + 04-04 21:32:14 | [957][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0083 ntime: 0078 mem: 3.36 + 04-04 21:32:22 | [957][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1433 ntime: 0079 mem: 3.36 + 04-04 21:32:28 | [957][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0331 ntime: 0071 mem: 3.36 + 04-04 21:32:36 | [957][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 21:32:43 | [957][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0087 mem: 3.36 + 04-04 21:32:48 | [957][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0080 mem: 3.36 + 04-04 21:32:54 | [957][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0915 ntime: 0079 mem: 3.36 + 04-04 21:33:00 | [957][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0092 ntime: 0082 mem: 3.36 + 04-04 21:33:05 | Time info >>>> elapsed: 1449.71 mins remain: 63.56 mins + 04-04 21:33:05 | [958][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0112 ntime: 0075 mem: 3.36 + 04-04 21:33:11 | [958][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0124 ntime: 0078 mem: 3.36 + 04-04 21:33:15 | [958][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0067 ntime: 0088 mem: 3.36 + 04-04 21:33:20 | [958][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0642 ntime: 0079 mem: 3.36 + 04-04 21:33:25 | [958][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0607 ntime: 0075 mem: 3.36 + 04-04 21:33:31 | [958][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0148 ntime: 0073 mem: 3.36 + 04-04 21:33:37 | [958][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1563 ntime: 0072 mem: 3.36 + 04-04 21:33:43 | [958][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0171 ntime: 0070 mem: 3.36 + 04-04 21:33:50 | [958][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0098 ntime: 0085 mem: 3.36 + 04-04 21:33:58 | [958][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1319 ntime: 0081 mem: 3.36 + 04-04 21:34:07 | [958][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1158 ntime: 0082 mem: 3.36 + 04-04 21:34:12 | [958][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0896 ntime: 0077 mem: 3.36 + 04-04 21:34:19 | [958][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1125 ntime: 0094 mem: 3.36 + 04-04 21:34:25 | [958][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 21:34:32 | [958][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0077 mem: 3.36 + 04-04 21:34:39 | [958][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0722 ntime: 0071 mem: 3.36 + 04-04 21:34:46 | [958][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1937 ntime: 0072 mem: 3.36 + 04-04 21:34:52 | [958][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0075 ntime: 0069 mem: 3.36 + 04-04 21:34:56 | Time info >>>> elapsed: 1451.56 mins remain: 62.06 mins + 04-04 21:34:56 | [959][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0451 ntime: 0074 mem: 3.36 + 04-04 21:35:05 | [959][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0611 ntime: 0074 mem: 3.36 + 04-04 21:35:11 | [959][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1125 ntime: 0077 mem: 3.36 + 04-04 21:35:16 | [959][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0923 ntime: 0083 mem: 3.36 + 04-04 21:35:25 | [959][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1294 ntime: 0079 mem: 3.36 + 04-04 21:35:31 | [959][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0074 mem: 3.36 + 04-04 21:35:38 | [959][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0630 ntime: 0081 mem: 3.36 + 04-04 21:35:45 | [959][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0518 ntime: 0076 mem: 3.36 + 04-04 21:35:53 | [959][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0146 ntime: 0078 mem: 3.36 + 04-04 21:36:02 | [959][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1468 ntime: 0074 mem: 3.36 + 04-04 21:36:09 | [959][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1159 ntime: 0071 mem: 3.36 + 04-04 21:36:13 | [959][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0147 ntime: 0079 mem: 3.36 + 04-04 21:36:19 | [959][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0562 ntime: 0075 mem: 3.36 + 04-04 21:36:25 | [959][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0972 ntime: 0082 mem: 3.36 + 04-04 21:36:32 | [959][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0635 ntime: 0083 mem: 3.36 + 04-04 21:36:38 | [959][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0504 ntime: 0079 mem: 3.36 + 04-04 21:36:45 | [959][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1625 ntime: 0076 mem: 3.36 + 04-04 21:36:52 | [959][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0864 ntime: 0076 mem: 3.36 + 04-04 21:36:57 | Time info >>>> elapsed: 1453.57 mins remain: 60.57 mins + 04-04 21:36:58 | [960][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0743 ntime: 0085 mem: 3.36 + 04-04 21:37:06 | [960][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2189 ntime: 0086 mem: 3.36 + 04-04 21:37:12 | [960][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 21:37:18 | [960][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0149 ntime: 0073 mem: 3.36 + 04-04 21:37:24 | [960][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0084 ntime: 0074 mem: 3.36 + 04-04 21:37:33 | [960][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0044 ntime: 0079 mem: 3.36 + 04-04 21:37:40 | [960][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1909 ntime: 0090 mem: 3.36 + 04-04 21:37:49 | [960][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0078 mem: 3.36 + 04-04 21:37:57 | [960][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1255 ntime: 0079 mem: 3.36 + 04-04 21:38:03 | [960][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0097 ntime: 0077 mem: 3.36 + 04-04 21:38:14 | [960][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0127 ntime: 0077 mem: 3.36 + 04-04 21:38:22 | [960][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0094 ntime: 0079 mem: 3.36 + 04-04 21:38:33 | [960][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0892 ntime: 0074 mem: 3.36 + 04-04 21:38:41 | [960][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0942 ntime: 0076 mem: 3.36 + 04-04 21:38:50 | [960][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0789 ntime: 0080 mem: 3.36 + 04-04 21:38:56 | [960][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0078 mem: 3.36 + 04-04 21:39:00 | [960][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0095 ntime: 0075 mem: 3.36 + 04-04 21:39:10 | [960][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1018 ntime: 0072 mem: 3.36 + 04-04 21:39:26 | Time info >>>> elapsed: 1456.06 mins remain: 59.09 mins + 04-04 21:39:27 | [961][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1007 ntime: 0076 mem: 3.36 + 04-04 21:39:34 | [961][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0123 ntime: 0079 mem: 3.36 + 04-04 21:39:42 | [961][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 21:39:50 | [961][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0179 ntime: 0076 mem: 3.36 + 04-04 21:40:05 | [961][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2093 ntime: 0084 mem: 3.36 + 04-04 21:40:14 | [961][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 21:40:24 | [961][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1796 ntime: 0072 mem: 3.36 + 04-04 21:40:30 | [961][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1525 ntime: 0079 mem: 3.36 + 04-04 21:40:40 | [961][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1289 ntime: 0072 mem: 3.36 + 04-04 21:40:49 | [961][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1619 ntime: 0076 mem: 3.36 + 04-04 21:40:59 | [961][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1338 ntime: 0078 mem: 3.36 + 04-04 21:41:09 | [961][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1676 ntime: 0081 mem: 3.36 + 04-04 21:41:19 | [961][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2105 ntime: 0079 mem: 3.36 + 04-04 21:41:28 | [961][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1750 ntime: 0077 mem: 3.36 + 04-04 21:41:37 | [961][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1658 ntime: 0078 mem: 3.36 + 04-04 21:41:44 | [961][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0730 ntime: 0075 mem: 3.36 + 04-04 21:41:53 | [961][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0084 mem: 3.36 + 04-04 21:42:01 | [961][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1935 ntime: 0073 mem: 3.36 + 04-04 21:42:08 | Time info >>>> elapsed: 1458.77 mins remain: 57.62 mins + 04-04 21:42:10 | [962][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1267 ntime: 0072 mem: 3.36 + 04-04 21:42:18 | [962][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1071 ntime: 0079 mem: 3.36 + 04-04 21:42:25 | [962][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1076 ntime: 0082 mem: 3.36 + 04-04 21:42:32 | [962][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0626 ntime: 0078 mem: 3.36 + 04-04 21:42:40 | [962][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0101 ntime: 0076 mem: 3.36 + 04-04 21:42:49 | [962][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1141 ntime: 0083 mem: 3.36 + 04-04 21:42:57 | [962][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0077 ntime: 0078 mem: 3.36 + 04-04 21:43:06 | [962][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1375 ntime: 0083 mem: 3.36 + 04-04 21:43:13 | [962][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0831 ntime: 0074 mem: 3.36 + 04-04 21:43:21 | [962][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0687 ntime: 0074 mem: 3.36 + 04-04 21:43:30 | [962][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0104 ntime: 0088 mem: 3.36 + 04-04 21:43:39 | [962][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0850 ntime: 0074 mem: 3.36 + 04-04 21:43:47 | [962][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1235 ntime: 0077 mem: 3.36 + 04-04 21:43:56 | [962][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1381 ntime: 0082 mem: 3.36 + 04-04 21:44:04 | [962][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0911 ntime: 0078 mem: 3.36 + 04-04 21:44:13 | [962][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1034 ntime: 0085 mem: 3.36 + 04-04 21:44:20 | [962][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0103 ntime: 0083 mem: 3.36 + 04-04 21:44:28 | [962][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0128 ntime: 0075 mem: 3.36 + 04-04 21:44:35 | Time info >>>> elapsed: 1461.21 mins remain: 56.14 mins + 04-04 21:44:36 | [963][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 1361 ntime: 0075 mem: 3.36 + 04-04 21:44:44 | [963][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1604 ntime: 0075 mem: 3.36 + 04-04 21:44:49 | [963][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0653 ntime: 0079 mem: 3.36 + 04-04 21:44:53 | [963][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0920 ntime: 0075 mem: 3.36 + 04-04 21:45:01 | [963][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0669 ntime: 0077 mem: 3.36 + 04-04 21:45:08 | [963][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0812 ntime: 0077 mem: 3.36 + 04-04 21:45:15 | [963][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1134 ntime: 0075 mem: 3.36 + 04-04 21:45:24 | [963][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1517 ntime: 0070 mem: 3.36 + 04-04 21:45:32 | [963][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1730 ntime: 0075 mem: 3.36 + 04-04 21:45:45 | [963][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0738 ntime: 0076 mem: 3.36 + 04-04 21:45:54 | [963][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1078 ntime: 0094 mem: 3.36 + 04-04 21:46:02 | [963][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0043 ntime: 0079 mem: 3.36 + 04-04 21:46:13 | [963][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0202 ntime: 0075 mem: 3.36 + 04-04 21:46:21 | [963][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0126 ntime: 0070 mem: 3.36 + 04-04 21:46:29 | [963][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1489 ntime: 0071 mem: 3.36 + 04-04 21:46:38 | [963][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1035 ntime: 0073 mem: 3.36 + 04-04 21:46:49 | [963][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0587 ntime: 0070 mem: 3.36 + 04-04 21:46:59 | [963][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0753 ntime: 0080 mem: 3.36 + 04-04 21:47:05 | Time info >>>> elapsed: 1463.71 mins remain: 54.66 mins + 04-04 21:47:06 | [964][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0478 ntime: 0080 mem: 3.36 + 04-04 21:47:13 | [964][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0987 ntime: 0075 mem: 3.36 + 04-04 21:47:21 | [964][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0207 ntime: 0074 mem: 3.36 + 04-04 21:47:31 | [964][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1074 ntime: 0074 mem: 3.36 + 04-04 21:47:39 | [964][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1467 ntime: 0079 mem: 3.36 + 04-04 21:47:45 | [964][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0078 mem: 3.36 + 04-04 21:47:54 | [964][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1487 ntime: 0077 mem: 3.36 + 04-04 21:48:02 | [964][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0085 mem: 3.36 + 04-04 21:48:09 | [964][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1484 ntime: 0076 mem: 3.36 + 04-04 21:48:17 | [964][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0854 ntime: 0078 mem: 3.36 + 04-04 21:48:26 | [964][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1666 ntime: 0077 mem: 3.36 + 04-04 21:48:33 | [964][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0936 ntime: 0083 mem: 3.36 + 04-04 21:48:41 | [964][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0856 ntime: 0078 mem: 3.36 + 04-04 21:48:48 | [964][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0318 ntime: 0080 mem: 3.36 + 04-04 21:48:56 | [964][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0922 ntime: 0076 mem: 3.36 + 04-04 21:49:02 | [964][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0630 ntime: 0074 mem: 3.36 + 04-04 21:49:09 | [964][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1273 ntime: 0083 mem: 3.36 + 04-04 21:49:18 | [964][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0739 ntime: 0075 mem: 3.36 + 04-04 21:49:23 | Time info >>>> elapsed: 1466.02 mins remain: 53.17 mins + 04-04 21:49:24 | [965][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0147 ntime: 0080 mem: 3.36 + 04-04 21:49:30 | [965][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0164 ntime: 0077 mem: 3.36 + 04-04 21:49:37 | [965][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0126 ntime: 0074 mem: 3.36 + 04-04 21:49:45 | [965][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1090 ntime: 0080 mem: 3.36 + 04-04 21:49:53 | [965][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0732 ntime: 0080 mem: 3.36 + 04-04 21:50:00 | [965][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1510 ntime: 0075 mem: 3.36 + 04-04 21:50:07 | [965][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0079 mem: 3.36 + 04-04 21:50:16 | [965][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0798 ntime: 0077 mem: 3.36 + 04-04 21:50:26 | [965][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1562 ntime: 0082 mem: 3.36 + 04-04 21:50:33 | [965][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0358 ntime: 0075 mem: 3.36 + 04-04 21:50:40 | [965][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0478 ntime: 0076 mem: 3.36 + 04-04 21:50:46 | [965][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0374 ntime: 0076 mem: 3.36 + 04-04 21:50:51 | [965][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0078 ntime: 0079 mem: 3.36 + 04-04 21:50:58 | [965][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0102 ntime: 0079 mem: 3.36 + 04-04 21:51:05 | [965][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0697 ntime: 0078 mem: 3.36 + 04-04 21:51:12 | [965][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1612 ntime: 0083 mem: 3.36 + 04-04 21:51:17 | [965][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0093 ntime: 0070 mem: 3.36 + 04-04 21:51:25 | [965][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0085 mem: 3.36 + 04-04 21:51:30 | Time info >>>> elapsed: 1468.12 mins remain: 51.67 mins + 04-04 21:51:30 | [966][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0077 mem: 3.36 + 04-04 21:51:37 | [966][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0944 ntime: 0080 mem: 3.36 + 04-04 21:51:44 | [966][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0633 ntime: 0075 mem: 3.36 + 04-04 21:51:54 | [966][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1469 ntime: 0076 mem: 3.36 + 04-04 21:52:01 | [966][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0071 ntime: 0081 mem: 3.36 + 04-04 21:52:09 | [966][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1114 ntime: 0084 mem: 3.36 + 04-04 21:52:15 | [966][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0426 ntime: 0082 mem: 3.36 + 04-04 21:52:22 | [966][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1293 ntime: 0077 mem: 3.36 + 04-04 21:52:27 | [966][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0772 ntime: 0071 mem: 3.36 + 04-04 21:52:34 | [966][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0154 ntime: 0073 mem: 3.36 + 04-04 21:52:40 | [966][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0187 ntime: 0084 mem: 3.36 + 04-04 21:52:48 | [966][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2229 ntime: 0077 mem: 3.36 + 04-04 21:52:56 | [966][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0908 ntime: 0083 mem: 3.36 + 04-04 21:53:01 | [966][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0099 ntime: 0082 mem: 3.36 + 04-04 21:53:09 | [966][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 21:53:14 | [966][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0086 mem: 3.36 + 04-04 21:53:21 | [966][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0043 ntime: 0072 mem: 3.36 + 04-04 21:53:28 | [966][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0985 ntime: 0080 mem: 3.36 + 04-04 21:53:33 | Time info >>>> elapsed: 1470.18 mins remain: 50.17 mins + 04-04 21:53:34 | [967][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0457 ntime: 0069 mem: 3.36 + 04-04 21:53:40 | [967][010/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0813 ntime: 0082 mem: 3.36 + 04-04 21:53:48 | [967][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1045 ntime: 0079 mem: 3.36 + 04-04 21:53:58 | [967][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0078 mem: 3.36 + 04-04 21:54:06 | [967][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1178 ntime: 0080 mem: 3.36 + 04-04 21:54:13 | [967][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0088 ntime: 0080 mem: 3.36 + 04-04 21:54:18 | [967][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0216 ntime: 0077 mem: 3.36 + 04-04 21:54:25 | [967][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0094 ntime: 0085 mem: 3.36 + 04-04 21:54:32 | [967][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0762 ntime: 0078 mem: 3.36 + 04-04 21:54:38 | [967][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0719 ntime: 0076 mem: 3.36 + 04-04 21:54:45 | [967][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0137 ntime: 0072 mem: 3.36 + 04-04 21:54:53 | [967][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0683 ntime: 0077 mem: 3.36 + 04-04 21:55:00 | [967][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1034 ntime: 0079 mem: 3.36 + 04-04 21:55:06 | [967][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0073 mem: 3.36 + 04-04 21:55:11 | [967][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0078 mem: 3.36 + 04-04 21:55:16 | [967][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0082 mem: 3.36 + 04-04 21:55:22 | [967][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0247 ntime: 0076 mem: 3.36 + 04-04 21:55:33 | [967][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0082 mem: 3.36 + 04-04 21:55:38 | Time info >>>> elapsed: 1472.26 mins remain: 48.67 mins + 04-04 21:55:39 | [968][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0846 ntime: 0075 mem: 3.36 + 04-04 21:55:45 | [968][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0935 ntime: 0078 mem: 3.36 + 04-04 21:55:51 | [968][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0335 ntime: 0079 mem: 3.36 + 04-04 21:55:59 | [968][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0082 ntime: 0071 mem: 3.36 + 04-04 21:56:05 | [968][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0084 ntime: 0071 mem: 3.36 + 04-04 21:56:12 | [968][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0079 mem: 3.36 + 04-04 21:56:20 | [968][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0391 ntime: 0077 mem: 3.36 + 04-04 21:56:26 | [968][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1172 ntime: 0081 mem: 3.36 + 04-04 21:56:32 | [968][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1064 ntime: 0060 mem: 3.36 + 04-04 21:56:40 | [968][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0107 ntime: 0075 mem: 3.36 + 04-04 21:56:46 | [968][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1126 ntime: 0079 mem: 3.36 + 04-04 21:56:59 | [968][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1754 ntime: 0071 mem: 3.36 + 04-04 21:57:04 | [968][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0073 mem: 3.36 + 04-04 21:57:11 | [968][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0471 ntime: 0084 mem: 3.36 + 04-04 21:57:18 | [968][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0512 ntime: 0076 mem: 3.36 + 04-04 21:57:25 | [968][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0100 ntime: 0081 mem: 3.36 + 04-04 21:57:30 | [968][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0114 ntime: 0074 mem: 3.36 + 04-04 21:57:35 | [968][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0850 ntime: 0076 mem: 3.36 + 04-04 21:57:39 | Time info >>>> elapsed: 1474.27 mins remain: 47.16 mins + 04-04 21:57:39 | [969][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0148 ntime: 0080 mem: 3.36 + 04-04 21:57:44 | [969][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0783 ntime: 0086 mem: 3.36 + 04-04 21:57:47 | [969][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0311 ntime: 0079 mem: 3.36 + 04-04 21:57:52 | [969][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0985 ntime: 0076 mem: 3.36 + 04-04 21:57:56 | [969][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0086 mem: 3.36 + 04-04 21:58:02 | [969][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0079 mem: 3.36 + 04-04 21:58:07 | [969][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0087 ntime: 0085 mem: 3.36 + 04-04 21:58:11 | [969][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0485 ntime: 0080 mem: 3.36 + 04-04 21:58:16 | [969][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0097 ntime: 0075 mem: 3.36 + 04-04 21:58:21 | [969][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0660 ntime: 0081 mem: 3.36 + 04-04 21:58:27 | [969][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0088 ntime: 0079 mem: 3.36 + 04-04 21:58:31 | [969][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0486 ntime: 0081 mem: 3.36 + 04-04 21:58:35 | [969][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0661 ntime: 0077 mem: 3.36 + 04-04 21:58:40 | [969][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0910 ntime: 0077 mem: 3.36 + 04-04 21:58:45 | [969][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0129 ntime: 0076 mem: 3.36 + 04-04 21:58:49 | [969][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0103 ntime: 0078 mem: 3.36 + 04-04 21:58:57 | [969][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1518 ntime: 0073 mem: 3.36 + 04-04 21:59:02 | [969][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0772 ntime: 0072 mem: 3.36 + 04-04 21:59:05 | Time info >>>> elapsed: 1475.71 mins remain: 45.64 mins + 04-04 21:59:05 | [970][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0218 ntime: 0076 mem: 3.36 + 04-04 21:59:11 | [970][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0112 ntime: 0071 mem: 3.36 + 04-04 21:59:15 | [970][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0063 ntime: 0078 mem: 3.36 + 04-04 21:59:20 | [970][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0082 mem: 3.36 + 04-04 21:59:26 | [970][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0209 ntime: 0079 mem: 3.36 + 04-04 21:59:34 | [970][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1402 ntime: 0079 mem: 3.36 + 04-04 21:59:38 | [970][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1061 ntime: 0075 mem: 3.36 + 04-04 21:59:44 | [970][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0091 mem: 3.36 + 04-04 21:59:49 | [970][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0100 ntime: 0077 mem: 3.36 + 04-04 21:59:54 | [970][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0828 ntime: 0076 mem: 3.36 + 04-04 21:59:58 | [970][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0105 ntime: 0077 mem: 3.36 + 04-04 22:00:04 | [970][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0435 ntime: 0083 mem: 3.36 + 04-04 22:00:08 | [970][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0640 ntime: 0081 mem: 3.36 + 04-04 22:00:12 | [970][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1439 ntime: 0078 mem: 3.36 + 04-04 22:00:19 | [970][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1369 ntime: 0084 mem: 3.36 + 04-04 22:00:23 | [970][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0081 mem: 3.36 + 04-04 22:00:27 | [970][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0303 ntime: 0078 mem: 3.36 + 04-04 22:00:32 | [970][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0065 ntime: 0080 mem: 3.36 + 04-04 22:00:37 | Time info >>>> elapsed: 1477.24 mins remain: 44.12 mins + 04-04 22:00:38 | [971][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0857 ntime: 0074 mem: 3.36 + 04-04 22:00:41 | [971][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0081 mem: 3.36 + 04-04 22:00:46 | [971][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0492 ntime: 0082 mem: 3.36 + 04-04 22:00:53 | [971][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0623 ntime: 0084 mem: 3.36 + 04-04 22:00:57 | [971][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0742 ntime: 0076 mem: 3.36 + 04-04 22:01:03 | [971][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0309 ntime: 0085 mem: 3.36 + 04-04 22:01:07 | [971][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0897 ntime: 0085 mem: 3.36 + 04-04 22:01:14 | [971][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0438 ntime: 0084 mem: 3.36 + 04-04 22:01:19 | [971][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1425 ntime: 0078 mem: 3.36 + 04-04 22:01:23 | [971][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0110 ntime: 0078 mem: 3.36 + 04-04 22:01:27 | [971][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0113 ntime: 0078 mem: 3.36 + 04-04 22:01:32 | [971][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0837 ntime: 0079 mem: 3.36 + 04-04 22:01:37 | [971][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1124 ntime: 0077 mem: 3.36 + 04-04 22:01:42 | [971][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0926 ntime: 0082 mem: 3.36 + 04-04 22:01:46 | [971][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0087 ntime: 0079 mem: 3.36 + 04-04 22:01:51 | [971][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0063 ntime: 0074 mem: 3.36 + 04-04 22:01:55 | [971][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0206 ntime: 0072 mem: 3.36 + 04-04 22:02:00 | [971][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0847 ntime: 0081 mem: 3.36 + 04-04 22:02:04 | Time info >>>> elapsed: 1478.70 mins remain: 42.60 mins + 04-04 22:02:05 | [972][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0955 ntime: 0072 mem: 3.36 + 04-04 22:02:11 | [972][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0571 ntime: 0079 mem: 3.36 + 04-04 22:02:16 | [972][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0111 ntime: 0076 mem: 3.36 + 04-04 22:02:21 | [972][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0735 ntime: 0080 mem: 3.36 + 04-04 22:02:25 | [972][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0732 ntime: 0076 mem: 3.36 + 04-04 22:02:29 | [972][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0156 ntime: 0075 mem: 3.36 + 04-04 22:02:36 | [972][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0786 ntime: 0079 mem: 3.36 + 04-04 22:02:41 | [972][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0754 ntime: 0080 mem: 3.36 + 04-04 22:02:44 | [972][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0135 ntime: 0078 mem: 3.36 + 04-04 22:02:49 | [972][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0067 ntime: 0076 mem: 3.36 + 04-04 22:02:52 | [972][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0253 ntime: 0075 mem: 3.36 + 04-04 22:02:57 | [972][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0113 ntime: 0073 mem: 3.36 + 04-04 22:03:01 | [972][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0484 ntime: 0080 mem: 3.36 + 04-04 22:03:05 | [972][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0146 ntime: 0079 mem: 3.36 + 04-04 22:03:10 | [972][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 22:03:16 | [972][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0951 ntime: 0079 mem: 3.36 + 04-04 22:03:22 | [972][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0215 ntime: 0083 mem: 3.36 + 04-04 22:03:25 | [972][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0581 ntime: 0078 mem: 3.36 + 04-04 22:03:27 | Time info >>>> elapsed: 1480.08 mins remain: 41.07 mins + 04-04 22:03:29 | [973][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1110 ntime: 0081 mem: 3.36 + 04-04 22:03:32 | [973][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0494 ntime: 0083 mem: 3.36 + 04-04 22:03:36 | [973][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0503 ntime: 0075 mem: 3.36 + 04-04 22:03:41 | [973][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0592 ntime: 0076 mem: 3.36 + 04-04 22:03:46 | [973][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0714 ntime: 0078 mem: 3.36 + 04-04 22:03:49 | [973][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0063 ntime: 0079 mem: 3.36 + 04-04 22:03:55 | [973][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0201 ntime: 0078 mem: 3.36 + 04-04 22:04:00 | [973][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0311 ntime: 0079 mem: 3.36 + 04-04 22:04:04 | [973][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0078 mem: 3.36 + 04-04 22:04:09 | [973][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0932 ntime: 0081 mem: 3.36 + 04-04 22:04:13 | [973][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0083 mem: 3.36 + 04-04 22:04:19 | [973][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1057 ntime: 0078 mem: 3.36 + 04-04 22:04:23 | [973][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0072 mem: 3.36 + 04-04 22:04:27 | [973][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0079 mem: 3.36 + 04-04 22:04:32 | [973][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0078 mem: 3.36 + 04-04 22:04:37 | [973][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0643 ntime: 0074 mem: 3.36 + 04-04 22:04:43 | [973][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0067 ntime: 0073 mem: 3.36 + 04-04 22:04:46 | [973][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0096 ntime: 0078 mem: 3.36 + 04-04 22:04:50 | Time info >>>> elapsed: 1481.46 mins remain: 39.55 mins + 04-04 22:04:50 | [974][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0048 ntime: 0076 mem: 3.36 + 04-04 22:04:55 | [974][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0507 ntime: 0080 mem: 3.36 + 04-04 22:04:59 | [974][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0108 ntime: 0078 mem: 3.36 + 04-04 22:05:04 | [974][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0129 ntime: 0079 mem: 3.36 + 04-04 22:05:08 | [974][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0607 ntime: 0075 mem: 3.36 + 04-04 22:05:15 | [974][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1112 ntime: 0093 mem: 3.36 + 04-04 22:05:19 | [974][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1287 ntime: 0079 mem: 3.36 + 04-04 22:05:23 | [974][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0088 ntime: 0074 mem: 3.36 + 04-04 22:05:28 | [974][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0790 ntime: 0077 mem: 3.36 + 04-04 22:05:32 | [974][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0094 ntime: 0076 mem: 3.36 + 04-04 22:05:38 | [974][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0191 ntime: 0077 mem: 3.36 + 04-04 22:05:44 | [974][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1726 ntime: 0076 mem: 3.36 + 04-04 22:05:54 | [974][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1575 ntime: 0077 mem: 3.36 + 04-04 22:06:01 | [974][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0424 ntime: 0083 mem: 3.36 + 04-04 22:06:08 | [974][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 2014 ntime: 0088 mem: 3.36 + 04-04 22:06:11 | [974][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0077 mem: 3.36 + 04-04 22:06:14 | [974][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0076 mem: 3.36 + 04-04 22:06:20 | [974][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 22:06:24 | Time info >>>> elapsed: 1483.03 mins remain: 38.03 mins + 04-04 22:06:24 | [975][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0088 ntime: 0074 mem: 3.36 + 04-04 22:06:31 | [975][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0107 ntime: 0077 mem: 3.36 + 04-04 22:06:37 | [975][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0118 ntime: 0073 mem: 3.36 + 04-04 22:06:42 | [975][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0601 ntime: 0083 mem: 3.36 + 04-04 22:06:47 | [975][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0690 ntime: 0083 mem: 3.36 + 04-04 22:06:52 | [975][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0937 ntime: 0082 mem: 3.36 + 04-04 22:06:56 | [975][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0132 ntime: 0079 mem: 3.36 + 04-04 22:07:00 | [975][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0079 mem: 3.36 + 04-04 22:07:04 | [975][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0636 ntime: 0084 mem: 3.36 + 04-04 22:07:14 | [975][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1170 ntime: 0089 mem: 3.36 + 04-04 22:07:20 | [975][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0454 ntime: 0077 mem: 3.36 + 04-04 22:07:27 | [975][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0344 ntime: 0077 mem: 3.36 + 04-04 22:07:30 | [975][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0489 ntime: 0055 mem: 3.36 + 04-04 22:07:34 | [975][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0094 ntime: 0055 mem: 3.36 + 04-04 22:07:37 | [975][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0453 ntime: 0072 mem: 3.36 + 04-04 22:07:43 | [975][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0074 ntime: 0077 mem: 3.36 + 04-04 22:07:47 | [975][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0109 ntime: 0057 mem: 3.36 + 04-04 22:07:52 | [975][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0829 ntime: 0077 mem: 3.36 + 04-04 22:07:55 | Time info >>>> elapsed: 1484.55 mins remain: 36.51 mins + 04-04 22:07:56 | [976][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0482 ntime: 0078 mem: 3.36 + 04-04 22:08:00 | [976][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0532 ntime: 0083 mem: 3.36 + 04-04 22:08:05 | [976][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0081 ntime: 0075 mem: 3.36 + 04-04 22:08:10 | [976][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0927 ntime: 0080 mem: 3.36 + 04-04 22:08:14 | [976][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0522 ntime: 0074 mem: 3.36 + 04-04 22:08:18 | [976][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1038 ntime: 0085 mem: 3.36 + 04-04 22:08:23 | [976][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0508 ntime: 0081 mem: 3.36 + 04-04 22:08:29 | [976][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0044 ntime: 0085 mem: 3.36 + 04-04 22:08:34 | [976][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0604 ntime: 0082 mem: 3.36 + 04-04 22:08:37 | [976][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0076 mem: 3.36 + 04-04 22:08:42 | [976][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0078 ntime: 0075 mem: 3.36 + 04-04 22:08:46 | [976][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0074 mem: 3.36 + 04-04 22:08:51 | [976][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0320 ntime: 0083 mem: 3.36 + 04-04 22:08:56 | [976][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0968 ntime: 0072 mem: 3.36 + 04-04 22:09:01 | [976][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0083 ntime: 0085 mem: 3.36 + 04-04 22:09:07 | [976][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0983 ntime: 0083 mem: 3.36 + 04-04 22:09:14 | [976][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0082 mem: 3.36 + 04-04 22:09:23 | [976][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0078 mem: 3.36 + 04-04 22:09:26 | Time info >>>> elapsed: 1486.06 mins remain: 34.98 mins + 04-04 22:09:27 | [977][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0943 ntime: 0081 mem: 3.36 + 04-04 22:09:31 | [977][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0392 ntime: 0077 mem: 3.36 + 04-04 22:09:36 | [977][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0076 mem: 3.36 + 04-04 22:09:44 | [977][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0076 mem: 3.36 + 04-04 22:09:49 | [977][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0897 ntime: 0066 mem: 3.36 + 04-04 22:09:53 | [977][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0044 ntime: 0078 mem: 3.36 + 04-04 22:09:57 | [977][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0077 mem: 3.36 + 04-04 22:10:01 | [977][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0533 ntime: 0077 mem: 3.36 + 04-04 22:10:05 | [977][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0148 ntime: 0084 mem: 3.36 + 04-04 22:10:10 | [977][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0073 ntime: 0075 mem: 3.36 + 04-04 22:10:15 | [977][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 22:10:23 | [977][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0993 ntime: 0082 mem: 3.36 + 04-04 22:10:28 | [977][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0080 ntime: 0081 mem: 3.36 + 04-04 22:10:33 | [977][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1412 ntime: 0079 mem: 3.36 + 04-04 22:10:37 | [977][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0645 ntime: 0076 mem: 3.36 + 04-04 22:10:41 | [977][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0875 ntime: 0078 mem: 3.36 + 04-04 22:10:45 | [977][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0102 ntime: 0078 mem: 3.36 + 04-04 22:10:51 | [977][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0562 ntime: 0077 mem: 3.36 + 04-04 22:10:56 | Time info >>>> elapsed: 1487.56 mins remain: 33.46 mins + 04-04 22:10:56 | [978][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0145 ntime: 0078 mem: 3.36 + 04-04 22:11:03 | [978][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0100 ntime: 0079 mem: 3.36 + 04-04 22:11:09 | [978][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0091 ntime: 0079 mem: 3.36 + 04-04 22:11:13 | [978][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0079 mem: 3.36 + 04-04 22:11:19 | [978][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0980 ntime: 0084 mem: 3.36 + 04-04 22:11:24 | [978][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0080 mem: 3.36 + 04-04 22:11:29 | [978][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0081 ntime: 0079 mem: 3.36 + 04-04 22:11:33 | [978][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0084 ntime: 0080 mem: 3.36 + 04-04 22:11:38 | [978][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0659 ntime: 0077 mem: 3.36 + 04-04 22:11:41 | [978][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0302 ntime: 0077 mem: 3.36 + 04-04 22:11:45 | [978][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0085 ntime: 0077 mem: 3.36 + 04-04 22:11:49 | [978][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0122 ntime: 0080 mem: 3.36 + 04-04 22:11:53 | [978][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0144 ntime: 0082 mem: 3.36 + 04-04 22:11:57 | [978][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0078 mem: 3.36 + 04-04 22:12:03 | [978][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 22:12:07 | [978][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0604 ntime: 0078 mem: 3.36 + 04-04 22:12:13 | [978][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0451 ntime: 0072 mem: 3.36 + 04-04 22:12:17 | [978][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0083 ntime: 0072 mem: 3.36 + 04-04 22:12:21 | Time info >>>> elapsed: 1488.97 mins remain: 31.94 mins + 04-04 22:12:21 | [979][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0082 mem: 3.36 + 04-04 22:12:25 | [979][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0068 ntime: 0077 mem: 3.36 + 04-04 22:12:30 | [979][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1070 ntime: 0075 mem: 3.36 + 04-04 22:12:34 | [979][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0754 ntime: 0078 mem: 3.36 + 04-04 22:12:38 | [979][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0072 mem: 3.36 + 04-04 22:12:42 | [979][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0311 ntime: 0079 mem: 3.36 + 04-04 22:12:47 | [979][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0619 ntime: 0078 mem: 3.36 + 04-04 22:12:52 | [979][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0807 ntime: 0085 mem: 3.36 + 04-04 22:12:55 | [979][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0145 ntime: 0079 mem: 3.36 + 04-04 22:13:01 | [979][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0081 mem: 3.36 + 04-04 22:13:06 | [979][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0076 mem: 3.36 + 04-04 22:13:11 | [979][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0726 ntime: 0080 mem: 3.36 + 04-04 22:13:15 | [979][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0084 mem: 3.36 + 04-04 22:13:19 | [979][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0710 ntime: 0079 mem: 3.36 + 04-04 22:13:24 | [979][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0075 mem: 3.36 + 04-04 22:13:28 | [979][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0374 ntime: 0055 mem: 3.36 + 04-04 22:13:33 | [979][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0055 mem: 3.36 + 04-04 22:13:37 | [979][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0114 ntime: 0080 mem: 3.36 + 04-04 22:13:42 | Time info >>>> elapsed: 1490.32 mins remain: 30.41 mins + 04-04 22:13:42 | [980][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0249 ntime: 0075 mem: 3.36 + 04-04 22:13:46 | [980][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0615 ntime: 0075 mem: 3.36 + 04-04 22:13:51 | [980][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0347 ntime: 0077 mem: 3.36 + 04-04 22:13:56 | [980][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0108 ntime: 0077 mem: 3.36 + 04-04 22:14:02 | [980][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0638 ntime: 0078 mem: 3.36 + 04-04 22:14:09 | [980][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0439 ntime: 0075 mem: 3.36 + 04-04 22:14:14 | [980][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0158 ntime: 0077 mem: 3.36 + 04-04 22:14:18 | [980][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0506 ntime: 0081 mem: 3.36 + 04-04 22:14:23 | [980][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 22:14:27 | [980][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0072 ntime: 0077 mem: 3.36 + 04-04 22:14:32 | [980][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0362 ntime: 0083 mem: 3.36 + 04-04 22:14:38 | [980][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0079 mem: 3.36 + 04-04 22:14:45 | [980][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0606 ntime: 0083 mem: 3.36 + 04-04 22:14:53 | [980][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1068 ntime: 0077 mem: 3.36 + 04-04 22:14:58 | [980][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0648 ntime: 0082 mem: 3.36 + 04-04 22:15:07 | [980][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0534 ntime: 0082 mem: 3.36 + 04-04 22:15:16 | [980][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0092 ntime: 0075 mem: 3.36 + 04-04 22:15:19 | [980][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0077 mem: 3.36 + 04-04 22:15:22 | Time info >>>> elapsed: 1492.00 mins remain: 28.90 mins + 04-04 22:15:22 | [981][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0106 ntime: 0076 mem: 3.36 + 04-04 22:15:26 | [981][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0074 mem: 3.36 + 04-04 22:15:33 | [981][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0100 ntime: 0075 mem: 3.36 + 04-04 22:15:40 | [981][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0077 mem: 3.36 + 04-04 22:15:47 | [981][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0311 ntime: 0078 mem: 3.36 + 04-04 22:15:55 | [981][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0514 ntime: 0083 mem: 3.36 + 04-04 22:15:59 | [981][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0195 ntime: 0073 mem: 3.36 + 04-04 22:16:04 | [981][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0121 ntime: 0078 mem: 3.36 + 04-04 22:16:10 | [981][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0695 ntime: 0075 mem: 3.36 + 04-04 22:16:17 | [981][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1391 ntime: 0082 mem: 3.36 + 04-04 22:16:23 | [981][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0072 ntime: 0082 mem: 3.36 + 04-04 22:16:29 | [981][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1006 ntime: 0080 mem: 3.36 + 04-04 22:16:34 | [981][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0590 ntime: 0081 mem: 3.36 + 04-04 22:16:38 | [981][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0582 ntime: 0078 mem: 3.36 + 04-04 22:16:44 | [981][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0073 mem: 3.36 + 04-04 22:16:49 | [981][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0107 ntime: 0079 mem: 3.36 + 04-04 22:16:54 | [981][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1139 ntime: 0080 mem: 3.36 + 04-04 22:16:58 | [981][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0138 ntime: 0079 mem: 3.36 + 04-04 22:17:02 | Time info >>>> elapsed: 1493.65 mins remain: 27.38 mins + 04-04 22:17:02 | [982][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0681 ntime: 0072 mem: 3.36 + 04-04 22:17:07 | [982][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0727 ntime: 0073 mem: 3.36 + 04-04 22:17:11 | [982][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0079 mem: 3.36 + 04-04 22:17:15 | [982][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0328 ntime: 0081 mem: 3.36 + 04-04 22:17:19 | [982][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0086 ntime: 0072 mem: 3.36 + 04-04 22:17:24 | [982][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0077 mem: 3.36 + 04-04 22:17:29 | [982][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0612 ntime: 0075 mem: 3.36 + 04-04 22:17:33 | [982][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0873 ntime: 0079 mem: 3.36 + 04-04 22:17:37 | [982][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0395 ntime: 0077 mem: 3.36 + 04-04 22:17:41 | [982][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0078 mem: 3.36 + 04-04 22:17:45 | [982][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0109 ntime: 0079 mem: 3.36 + 04-04 22:17:50 | [982][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0095 ntime: 0080 mem: 3.36 + 04-04 22:17:55 | [982][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0781 ntime: 0083 mem: 3.36 + 04-04 22:17:59 | [982][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0091 ntime: 0079 mem: 3.36 + 04-04 22:18:06 | [982][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0637 ntime: 0078 mem: 3.36 + 04-04 22:18:11 | [982][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1122 ntime: 0076 mem: 3.36 + 04-04 22:18:15 | [982][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0921 ntime: 0076 mem: 3.36 + 04-04 22:18:18 | [982][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0077 mem: 3.36 + 04-04 22:18:21 | Time info >>>> elapsed: 1494.98 mins remain: 25.85 mins + 04-04 22:18:22 | [983][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0543 ntime: 0082 mem: 3.36 + 04-04 22:18:27 | [983][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0085 mem: 3.36 + 04-04 22:18:34 | [983][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1495 ntime: 0074 mem: 3.36 + 04-04 22:18:39 | [983][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 22:18:45 | [983][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0958 ntime: 0075 mem: 3.36 + 04-04 22:18:51 | [983][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0163 ntime: 0078 mem: 3.36 + 04-04 22:18:58 | [983][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0967 ntime: 0077 mem: 3.36 + 04-04 22:19:06 | [983][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1481 ntime: 0073 mem: 3.36 + 04-04 22:19:12 | [983][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0897 ntime: 0078 mem: 3.36 + 04-04 22:19:18 | [983][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0043 ntime: 0077 mem: 3.36 + 04-04 22:19:23 | [983][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1102 ntime: 0078 mem: 3.36 + 04-04 22:19:30 | [983][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0104 ntime: 0086 mem: 3.36 + 04-04 22:19:36 | [983][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0574 ntime: 0077 mem: 3.36 + 04-04 22:19:40 | [983][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0566 ntime: 0078 mem: 3.36 + 04-04 22:19:44 | [983][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0080 mem: 3.36 + 04-04 22:19:48 | [983][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0073 mem: 3.36 + 04-04 22:19:54 | [983][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0568 ntime: 0076 mem: 3.36 + 04-04 22:19:57 | [983][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0080 mem: 3.36 + 04-04 22:20:01 | Time info >>>> elapsed: 1496.64 mins remain: 24.34 mins + 04-04 22:20:01 | [984][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0134 ntime: 0074 mem: 3.36 + 04-04 22:20:04 | [984][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0120 ntime: 0076 mem: 3.36 + 04-04 22:20:08 | [984][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0696 ntime: 0081 mem: 3.36 + 04-04 22:20:12 | [984][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0504 ntime: 0078 mem: 3.36 + 04-04 22:20:17 | [984][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0710 ntime: 0078 mem: 3.36 + 04-04 22:20:21 | [984][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 22:20:27 | [984][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0119 ntime: 0080 mem: 3.36 + 04-04 22:20:32 | [984][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0126 ntime: 0080 mem: 3.36 + 04-04 22:20:35 | [984][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0802 ntime: 0076 mem: 3.36 + 04-04 22:20:42 | [984][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1243 ntime: 0072 mem: 3.36 + 04-04 22:20:47 | [984][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0083 mem: 3.36 + 04-04 22:20:52 | [984][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0063 ntime: 0075 mem: 3.36 + 04-04 22:20:56 | [984][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0161 ntime: 0085 mem: 3.36 + 04-04 22:21:01 | [984][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0075 mem: 3.36 + 04-04 22:21:06 | [984][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0076 mem: 3.36 + 04-04 22:21:13 | [984][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0113 ntime: 0073 mem: 3.36 + 04-04 22:21:17 | [984][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0115 ntime: 0080 mem: 3.36 + 04-04 22:21:24 | [984][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0332 ntime: 0082 mem: 3.36 + 04-04 22:21:26 | Time info >>>> elapsed: 1498.07 mins remain: 22.81 mins + 04-04 22:21:27 | [985][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0389 ntime: 0079 mem: 3.36 + 04-04 22:21:33 | [985][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0215 ntime: 0074 mem: 3.36 + 04-04 22:21:39 | [985][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0602 ntime: 0076 mem: 3.36 + 04-04 22:21:43 | [985][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0095 ntime: 0086 mem: 3.36 + 04-04 22:21:49 | [985][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0082 mem: 3.36 + 04-04 22:21:56 | [985][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1779 ntime: 0080 mem: 3.36 + 04-04 22:22:03 | [985][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0776 ntime: 0085 mem: 3.36 + 04-04 22:22:09 | [985][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0511 ntime: 0078 mem: 3.36 + 04-04 22:22:13 | [985][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0372 ntime: 0078 mem: 3.36 + 04-04 22:22:19 | [985][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1510 ntime: 0075 mem: 3.36 + 04-04 22:22:25 | [985][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0638 ntime: 0092 mem: 3.36 + 04-04 22:22:31 | [985][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0539 ntime: 0083 mem: 3.36 + 04-04 22:22:39 | [985][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0219 ntime: 0077 mem: 3.36 + 04-04 22:22:45 | [985][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0784 ntime: 0085 mem: 3.36 + 04-04 22:22:50 | [985][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0076 mem: 3.36 + 04-04 22:22:54 | [985][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0212 ntime: 0078 mem: 3.36 + 04-04 22:23:00 | [985][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0886 ntime: 0080 mem: 3.36 + 04-04 22:23:04 | [985][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0100 ntime: 0076 mem: 3.36 + 04-04 22:23:08 | Time info >>>> elapsed: 1499.76 mins remain: 21.29 mins + 04-04 22:23:08 | [986][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0121 ntime: 0080 mem: 3.36 + 04-04 22:23:14 | [986][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0727 ntime: 0074 mem: 3.36 + 04-04 22:23:20 | [986][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0139 ntime: 0078 mem: 3.36 + 04-04 22:23:26 | [986][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0085 ntime: 0078 mem: 3.36 + 04-04 22:23:32 | [986][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0078 mem: 3.36 + 04-04 22:23:39 | [986][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0404 ntime: 0077 mem: 3.36 + 04-04 22:23:45 | [986][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0099 ntime: 0077 mem: 3.36 + 04-04 22:23:51 | [986][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 1190 ntime: 0079 mem: 3.36 + 04-04 22:23:56 | [986][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0098 ntime: 0076 mem: 3.36 + 04-04 22:24:01 | [986][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0249 ntime: 0073 mem: 3.36 + 04-04 22:24:08 | [986][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0492 ntime: 0077 mem: 3.36 + 04-04 22:24:14 | [986][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0167 ntime: 0080 mem: 3.36 + 04-04 22:24:19 | [986][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0077 mem: 3.36 + 04-04 22:24:26 | [986][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0154 ntime: 0075 mem: 3.36 + 04-04 22:24:32 | [986][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0924 ntime: 0075 mem: 3.36 + 04-04 22:24:37 | [986][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0076 mem: 3.36 + 04-04 22:24:42 | [986][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0651 ntime: 0083 mem: 3.36 + 04-04 22:24:45 | [986][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0683 ntime: 0082 mem: 3.36 + 04-04 22:24:49 | Time info >>>> elapsed: 1501.45 mins remain: 19.78 mins + 04-04 22:24:50 | [987][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0665 ntime: 0086 mem: 3.36 + 04-04 22:24:55 | [987][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0076 mem: 3.36 + 04-04 22:24:59 | [987][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0090 ntime: 0082 mem: 3.36 + 04-04 22:25:05 | [987][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0248 ntime: 0079 mem: 3.36 + 04-04 22:25:11 | [987][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0077 mem: 3.36 + 04-04 22:25:16 | [987][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0930 ntime: 0085 mem: 3.36 + 04-04 22:25:22 | [987][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0072 ntime: 0078 mem: 3.36 + 04-04 22:25:27 | [987][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0079 mem: 3.36 + 04-04 22:25:33 | [987][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0738 ntime: 0056 mem: 3.36 + 04-04 22:25:37 | [987][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 22:25:43 | [987][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0657 ntime: 0077 mem: 3.36 + 04-04 22:25:48 | [987][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0838 ntime: 0081 mem: 3.36 + 04-04 22:25:52 | [987][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0079 mem: 3.36 + 04-04 22:25:56 | [987][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0075 mem: 3.36 + 04-04 22:26:00 | [987][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0408 ntime: 0072 mem: 3.36 + 04-04 22:26:07 | [987][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0061 ntime: 0077 mem: 3.36 + 04-04 22:26:12 | [987][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0391 ntime: 0080 mem: 3.36 + 04-04 22:26:17 | [987][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0082 mem: 3.36 + 04-04 22:26:22 | Time info >>>> elapsed: 1502.99 mins remain: 18.25 mins + 04-04 22:26:22 | [988][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0072 mem: 3.36 + 04-04 22:26:28 | [988][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0625 ntime: 0078 mem: 3.36 + 04-04 22:26:33 | [988][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0262 ntime: 0075 mem: 3.36 + 04-04 22:26:39 | [988][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0508 ntime: 0077 mem: 3.36 + 04-04 22:26:42 | [988][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0089 ntime: 0078 mem: 3.36 + 04-04 22:26:46 | [988][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0186 ntime: 0080 mem: 3.36 + 04-04 22:26:51 | [988][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0739 ntime: 0078 mem: 3.36 + 04-04 22:26:56 | [988][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0542 ntime: 0087 mem: 3.36 + 04-04 22:27:00 | [988][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0075 ntime: 0076 mem: 3.36 + 04-04 22:27:05 | [988][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0611 ntime: 0078 mem: 3.36 + 04-04 22:27:10 | [988][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0605 ntime: 0078 mem: 3.36 + 04-04 22:27:14 | [988][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0079 mem: 3.36 + 04-04 22:27:19 | [988][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0221 ntime: 0081 mem: 3.36 + 04-04 22:27:25 | [988][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0792 ntime: 0080 mem: 3.36 + 04-04 22:27:29 | [988][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0428 ntime: 0077 mem: 3.36 + 04-04 22:27:34 | [988][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0080 ntime: 0086 mem: 3.36 + 04-04 22:27:40 | [988][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0710 ntime: 0082 mem: 3.36 + 04-04 22:27:43 | [988][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0122 ntime: 0076 mem: 3.36 + 04-04 22:27:45 | Time info >>>> elapsed: 1504.37 mins remain: 16.73 mins + 04-04 22:27:45 | [989][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0082 ntime: 0084 mem: 3.36 + 04-04 22:27:46 | [989][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0156 ntime: 0083 mem: 3.36 + 04-04 22:27:48 | [989][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0211 ntime: 0083 mem: 3.36 + 04-04 22:27:49 | [989][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0076 mem: 3.36 + 04-04 22:27:51 | [989][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0081 ntime: 0079 mem: 3.36 + 04-04 22:27:52 | [989][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0044 ntime: 0076 mem: 3.36 + 04-04 22:27:54 | [989][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0083 mem: 3.36 + 04-04 22:27:55 | [989][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0083 mem: 3.36 + 04-04 22:27:57 | [989][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0082 mem: 3.36 + 04-04 22:27:58 | [989][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0134 ntime: 0083 mem: 3.36 + 04-04 22:28:00 | [989][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0063 ntime: 0078 mem: 3.36 + 04-04 22:28:01 | [989][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0082 mem: 3.36 + 04-04 22:28:03 | [989][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0118 ntime: 0073 mem: 3.36 + 04-04 22:28:04 | [989][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0081 mem: 3.36 + 04-04 22:28:06 | [989][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0075 ntime: 0087 mem: 3.36 + 04-04 22:28:08 | [989][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0091 ntime: 0087 mem: 3.36 + 04-04 22:28:09 | [989][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 22:28:11 | [989][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0085 mem: 3.36 + 04-04 22:28:12 | Time info >>>> elapsed: 1504.83 mins remain: 15.20 mins + 04-04 22:28:12 | [990][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0097 ntime: 0075 mem: 3.36 + 04-04 22:28:14 | [990][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0082 mem: 3.36 + 04-04 22:28:15 | [990][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0077 mem: 3.36 + 04-04 22:28:17 | [990][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0072 mem: 3.36 + 04-04 22:28:18 | [990][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0077 mem: 3.36 + 04-04 22:28:20 | [990][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0087 mem: 3.36 + 04-04 22:28:21 | [990][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0084 mem: 3.36 + 04-04 22:28:23 | [990][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0133 ntime: 0079 mem: 3.36 + 04-04 22:28:24 | [990][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0073 ntime: 0077 mem: 3.36 + 04-04 22:28:26 | [990][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0108 ntime: 0087 mem: 3.36 + 04-04 22:28:27 | [990][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0076 mem: 3.36 + 04-04 22:28:29 | [990][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0079 mem: 3.36 + 04-04 22:28:30 | [990][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0082 mem: 3.36 + 04-04 22:28:32 | [990][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0077 mem: 3.36 + 04-04 22:28:33 | [990][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0094 ntime: 0083 mem: 3.36 + 04-04 22:28:35 | [990][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0039 ntime: 0058 mem: 3.36 + 04-04 22:28:36 | [990][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0105 ntime: 0086 mem: 3.36 + 04-04 22:28:38 | [990][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0080 mem: 3.36 + 04-04 22:28:39 | Time info >>>> elapsed: 1505.27 mins remain: 13.67 mins + 04-04 22:28:39 | [991][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0072 ntime: 0077 mem: 3.36 + 04-04 22:28:41 | [991][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0078 mem: 3.36 + 04-04 22:28:42 | [991][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0115 ntime: 0075 mem: 3.36 + 04-04 22:28:44 | [991][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0072 mem: 3.36 + 04-04 22:28:45 | [991][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0124 ntime: 0082 mem: 3.36 + 04-04 22:28:47 | [991][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0073 ntime: 0080 mem: 3.36 + 04-04 22:28:48 | [991][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0077 mem: 3.36 + 04-04 22:28:50 | [991][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0072 ntime: 0080 mem: 3.36 + 04-04 22:28:51 | [991][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0079 mem: 3.36 + 04-04 22:28:52 | [991][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0068 mem: 3.36 + 04-04 22:28:54 | [991][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0067 ntime: 0079 mem: 3.36 + 04-04 22:28:55 | [991][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 22:28:57 | [991][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0137 ntime: 0078 mem: 3.36 + 04-04 22:28:58 | [991][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0072 mem: 3.36 + 04-04 22:29:00 | [991][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0081 ntime: 0092 mem: 3.36 + 04-04 22:29:01 | [991][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0082 mem: 3.36 + 04-04 22:29:03 | [991][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0065 ntime: 0084 mem: 3.36 + 04-04 22:29:04 | [991][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0080 mem: 3.36 + 04-04 22:29:06 | Time info >>>> elapsed: 1505.72 mins remain: 12.14 mins + 04-04 22:29:06 | [992][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0121 ntime: 0072 mem: 3.36 + 04-04 22:29:07 | [992][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0065 ntime: 0076 mem: 3.36 + 04-04 22:29:09 | [992][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0092 ntime: 0082 mem: 3.36 + 04-04 22:29:11 | [992][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0085 ntime: 0083 mem: 3.36 + 04-04 22:29:12 | [992][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0044 ntime: 0078 mem: 3.36 + 04-04 22:29:14 | [992][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0095 ntime: 0085 mem: 3.36 + 04-04 22:29:15 | [992][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0118 ntime: 0084 mem: 3.36 + 04-04 22:29:17 | [992][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0076 mem: 3.36 + 04-04 22:29:18 | [992][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0076 mem: 3.36 + 04-04 22:29:20 | [992][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0100 ntime: 0081 mem: 3.36 + 04-04 22:29:22 | [992][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0079 mem: 3.36 + 04-04 22:29:23 | [992][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0198 ntime: 0076 mem: 3.36 + 04-04 22:29:25 | [992][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0078 ntime: 0083 mem: 3.36 + 04-04 22:29:27 | [992][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0083 mem: 3.36 + 04-04 22:29:28 | [992][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0077 mem: 3.36 + 04-04 22:29:30 | [992][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0076 mem: 3.36 + 04-04 22:29:31 | [992][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0052 ntime: 0087 mem: 3.36 + 04-04 22:29:33 | [992][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0077 mem: 3.36 + 04-04 22:29:34 | Time info >>>> elapsed: 1506.20 mins remain: 10.62 mins + 04-04 22:29:34 | [993][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0075 ntime: 0079 mem: 3.36 + 04-04 22:29:36 | [993][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0095 ntime: 0087 mem: 3.36 + 04-04 22:29:37 | [993][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 22:29:39 | [993][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0108 ntime: 0079 mem: 3.36 + 04-04 22:29:40 | [993][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0082 mem: 3.36 + 04-04 22:29:42 | [993][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0078 mem: 3.36 + 04-04 22:29:43 | [993][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0079 mem: 3.36 + 04-04 22:29:45 | [993][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0071 ntime: 0079 mem: 3.36 + 04-04 22:29:46 | [993][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0075 mem: 3.36 + 04-04 22:29:48 | [993][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0099 ntime: 0078 mem: 3.36 + 04-04 22:29:49 | [993][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0081 mem: 3.36 + 04-04 22:29:51 | [993][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0079 ntime: 0078 mem: 3.36 + 04-04 22:29:53 | [993][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0075 mem: 3.36 + 04-04 22:29:54 | [993][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0084 ntime: 0073 mem: 3.36 + 04-04 22:29:56 | [993][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0078 mem: 3.36 + 04-04 22:29:57 | [993][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0058 ntime: 0082 mem: 3.36 + 04-04 22:29:59 | [993][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 22:30:00 | [993][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0086 mem: 3.36 + 04-04 22:30:01 | Time info >>>> elapsed: 1506.65 mins remain: 9.09 mins + 04-04 22:30:01 | [994][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0145 ntime: 0079 mem: 3.36 + 04-04 22:30:03 | [994][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0077 mem: 3.36 + 04-04 22:30:05 | [994][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0117 ntime: 0083 mem: 3.36 + 04-04 22:30:06 | [994][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0084 mem: 3.36 + 04-04 22:30:08 | [994][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0124 ntime: 0083 mem: 3.36 + 04-04 22:30:09 | [994][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0079 ntime: 0086 mem: 3.36 + 04-04 22:30:11 | [994][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0079 mem: 3.36 + 04-04 22:30:12 | [994][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0080 mem: 3.36 + 04-04 22:30:14 | [994][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0082 ntime: 0085 mem: 3.36 + 04-04 22:30:15 | [994][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0072 mem: 3.36 + 04-04 22:30:17 | [994][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0081 mem: 3.36 + 04-04 22:30:18 | [994][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0099 ntime: 0080 mem: 3.36 + 04-04 22:30:20 | [994][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0076 mem: 3.36 + 04-04 22:30:21 | [994][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0075 mem: 3.36 + 04-04 22:30:23 | [994][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0076 mem: 3.36 + 04-04 22:30:24 | [994][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0077 mem: 3.36 + 04-04 22:30:25 | [994][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0078 mem: 3.36 + 04-04 22:30:27 | [994][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0075 mem: 3.36 + 04-04 22:30:28 | Time info >>>> elapsed: 1507.10 mins remain: 7.57 mins + 04-04 22:30:28 | [995][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0050 ntime: 0085 mem: 3.36 + 04-04 22:30:30 | [995][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0083 mem: 3.36 + 04-04 22:30:31 | [995][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0083 mem: 3.36 + 04-04 22:30:33 | [995][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0053 ntime: 0082 mem: 3.36 + 04-04 22:30:34 | [995][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0080 mem: 3.36 + 04-04 22:30:36 | [995][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0123 ntime: 0079 mem: 3.36 + 04-04 22:30:38 | [995][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0076 mem: 3.36 + 04-04 22:30:39 | [995][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0078 mem: 3.36 + 04-04 22:30:41 | [995][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0071 ntime: 0084 mem: 3.36 + 04-04 22:30:42 | [995][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0044 ntime: 0085 mem: 3.36 + 04-04 22:30:44 | [995][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0069 ntime: 0081 mem: 3.36 + 04-04 22:30:46 | [995][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0101 ntime: 0089 mem: 3.36 + 04-04 22:30:47 | [995][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0083 mem: 3.36 + 04-04 22:30:49 | [995][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0070 ntime: 0082 mem: 3.36 + 04-04 22:30:50 | [995][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0087 mem: 3.36 + 04-04 22:30:52 | [995][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0080 mem: 3.36 + 04-04 22:30:54 | [995][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0080 mem: 3.36 + 04-04 22:30:55 | [995][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0057 ntime: 0076 mem: 3.36 + 04-04 22:30:57 | Time info >>>> elapsed: 1507.57 mins remain: 6.05 mins + 04-04 22:30:57 | [996][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0064 ntime: 0078 mem: 3.36 + 04-04 22:30:58 | [996][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0079 mem: 3.36 + 04-04 22:31:00 | [996][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0083 mem: 3.36 + 04-04 22:31:01 | [996][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0089 mem: 3.36 + 04-04 22:31:03 | [996][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0073 mem: 3.36 + 04-04 22:31:05 | [996][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0115 ntime: 0081 mem: 3.36 + 04-04 22:31:06 | [996][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0080 ntime: 0077 mem: 3.36 + 04-04 22:31:08 | [996][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0077 mem: 3.36 + 04-04 22:31:09 | [996][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0077 mem: 3.36 + 04-04 22:31:11 | [996][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0079 mem: 3.36 + 04-04 22:31:12 | [996][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0080 mem: 3.36 + 04-04 22:31:14 | [996][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0065 ntime: 0083 mem: 3.36 + 04-04 22:31:16 | [996][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0078 mem: 3.36 + 04-04 22:31:17 | [996][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0080 mem: 3.36 + 04-04 22:31:19 | [996][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0154 ntime: 0081 mem: 3.36 + 04-04 22:31:20 | [996][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0086 mem: 3.36 + 04-04 22:31:22 | [996][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0081 mem: 3.36 + 04-04 22:31:23 | [996][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0056 ntime: 0078 mem: 3.36 + 04-04 22:31:24 | Time info >>>> elapsed: 1508.03 mins remain: 4.54 mins + 04-04 22:31:24 | [997][000/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0049 ntime: 0085 mem: 3.36 + 04-04 22:31:26 | [997][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0192 ntime: 0076 mem: 3.36 + 04-04 22:31:28 | [997][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0080 mem: 3.36 + 04-04 22:31:29 | [997][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0078 mem: 3.36 + 04-04 22:31:30 | [997][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0077 ntime: 0077 mem: 3.36 + 04-04 22:31:32 | [997][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0065 mem: 3.36 + 04-04 22:31:33 | [997][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0073 mem: 3.36 + 04-04 22:31:35 | [997][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0085 mem: 3.36 + 04-04 22:31:36 | [997][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0079 mem: 3.36 + 04-04 22:31:38 | [997][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0083 mem: 3.36 + 04-04 22:31:39 | [997][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0060 ntime: 0082 mem: 3.36 + 04-04 22:31:41 | [997][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0054 ntime: 0083 mem: 3.36 + 04-04 22:31:43 | [997][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0077 mem: 3.36 + 04-04 22:31:44 | [997][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0114 ntime: 0076 mem: 3.36 + 04-04 22:31:46 | [997][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0098 ntime: 0086 mem: 3.36 + 04-04 22:31:47 | [997][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0051 ntime: 0078 mem: 3.36 + 04-04 22:31:49 | [997][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0087 ntime: 0080 mem: 3.36 + 04-04 22:31:50 | [997][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0064 mem: 3.36 + 04-04 22:31:51 | Time info >>>> elapsed: 1508.48 mins remain: 3.02 mins + 04-04 22:31:51 | [998][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0061 ntime: 0075 mem: 3.36 + 04-04 22:31:53 | [998][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0136 ntime: 0072 mem: 3.36 + 04-04 22:31:54 | [998][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0069 ntime: 0079 mem: 3.36 + 04-04 22:31:56 | [998][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0092 ntime: 0088 mem: 3.36 + 04-04 22:31:58 | [998][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0199 ntime: 0085 mem: 3.36 + 04-04 22:31:59 | [998][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0081 mem: 3.36 + 04-04 22:32:01 | [998][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0116 ntime: 0079 mem: 3.36 + 04-04 22:32:03 | [998][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0078 mem: 3.36 + 04-04 22:32:04 | [998][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0174 ntime: 0085 mem: 3.36 + 04-04 22:32:06 | [998][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0066 ntime: 0077 mem: 3.36 + 04-04 22:32:08 | [998][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0189 ntime: 0079 mem: 3.36 + 04-04 22:32:09 | [998][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0077 mem: 3.36 + 04-04 22:32:10 | [998][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0077 mem: 3.36 + 04-04 22:32:12 | [998][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0095 ntime: 0078 mem: 3.36 + 04-04 22:32:13 | [998][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0108 ntime: 0081 mem: 3.36 + 04-04 22:32:15 | [998][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0103 ntime: 0085 mem: 3.36 + 04-04 22:32:16 | [998][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0080 mem: 3.36 + 04-04 22:32:18 | [998][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0044 ntime: 0075 mem: 3.36 + 04-04 22:32:19 | Time info >>>> elapsed: 1508.95 mins remain: 1.51 mins + 04-04 22:32:19 | [999][000/179] predict_x0_loss: 0.009 glr: 5.0e-09 dtime: 0070 ntime: 0080 mem: 3.36 + 04-04 22:32:21 | [999][010/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0050 ntime: 0086 mem: 3.36 + 04-04 22:32:22 | [999][020/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0062 ntime: 0080 mem: 3.36 + 04-04 22:32:24 | [999][030/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0076 mem: 3.36 + 04-04 22:32:25 | [999][040/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0078 mem: 3.36 + 04-04 22:32:27 | [999][050/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0073 ntime: 0088 mem: 3.36 + 04-04 22:32:28 | [999][060/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0059 ntime: 0080 mem: 3.36 + 04-04 22:32:30 | [999][070/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0113 ntime: 0084 mem: 3.36 + 04-04 22:32:31 | [999][080/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0086 mem: 3.36 + 04-04 22:32:33 | [999][090/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0140 ntime: 0079 mem: 3.36 + 04-04 22:32:34 | [999][100/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0048 ntime: 0078 mem: 3.36 + 04-04 22:32:36 | [999][110/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0122 ntime: 0083 mem: 3.36 + 04-04 22:32:38 | [999][120/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0139 ntime: 0079 mem: 3.36 + 04-04 22:32:39 | [999][130/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0080 mem: 3.36 + 04-04 22:32:41 | [999][140/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0055 ntime: 0074 mem: 3.36 + 04-04 22:32:42 | [999][150/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0047 ntime: 0082 mem: 3.36 + 04-04 22:32:44 | [999][160/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0045 ntime: 0080 mem: 3.36 + 04-04 22:32:45 | [999][170/179] predict_x0_loss: 0.008 glr: 5.0e-09 dtime: 0046 ntime: 0081 mem: 3.36 + 04-04 22:32:47 | Time info >>>> elapsed: 1509.41 mins remain: -0.00 mins + 04-04 22:32:47 | An error has been caught in function '', process 'MainProcess' (640214), thread 'MainThread' (140176887994176): +Traceback (most recent call last): + +> File "/mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/train.py", line 310, in + main_worker(0, 1, args) + │ └ Namespace(config='configs/diffusion_rvqvae_128.yaml', project='s2g', stat='ts', csv_name='a2g_0', notes='', trainer='diffusio... + └ + + File "/mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/train.py", line 289, in main_worker + other_tools.record_trial(args, trainer.tracker) + │ │ │ │ └ + │ │ │ └ + │ │ └ Namespace(config='configs/diffusion_rvqvae_128.yaml', project='s2g', stat='ts', csv_name='a2g_0', notes='', trainer='diffusio... + │ └ + └ + + File "/mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/utils/other_tools.py", line 863, in record_trial + df_aligned = df_existing.append(df_new).fillna("") + │ └ config project stat ... predict_x0_loss_test_last_epoch predict_x0_loss_test_best predict_x0_l... + └ config project stat csv_name ... latent_self_test_last latent_self_test_last_epoch latent_self_test_best l... + + File "/home/chenbohong/miniconda3/envs/gdc/lib/python3.10/site-packages/pandas/core/generic.py", line 5989, in __getattr__ + return object.__getattribute__(self, name) + │ └ 'append' + └ config project stat csv_name ... latent_self_test_last latent_self_test_last_epoch latent_self_test_best l... + +AttributeError: 'DataFrame' object has no attribute 'append' diff --git a/ckpt/beatx2_cospeech_diffusion/0403_212319_diffusion_rvqvae_128.yaml b/ckpt/beatx2_cospeech_diffusion/0403_212319_diffusion_rvqvae_128.yaml new file mode 100644 index 0000000000000000000000000000000000000000..abab2e7d2cd5b6f192224d32435a2fcc68b057d1 --- /dev/null +++ b/ckpt/beatx2_cospeech_diffusion/0403_212319_diffusion_rvqvae_128.yaml @@ -0,0 +1,54 @@ +{a_encoder: null, a_fix_pre: false, a_pre_encoder: null, acc: 1, acc_weight: 0.0, + additional_data: false, adv_weight: 20.0, ali_weight: 0.0, amsgrad: false, apex: false, + asmr: 0.0, atcont: 0.0, atmr: 0.0, aud_prob: 1.0, audio_dims: 1, audio_f: 256, audio_fps: 16000, + audio_norm: false, audio_rep: onset+amplitude, audio_sr: 16000, batch_size: 40, + beat_align: true, benchmark: true, cache_only: false, cache_path: datasets/beat_cache/beat_smplx_en_emage_2_128/, + cf: 0.0, ch: 1.0, cl: 1.0, clean_final_seconds: 0, clean_first_seconds: 0, commit: 0.02, + config: configs/diffusion_rvqvae_128.yaml, csv_name: a2g_0, cu: 1.0, cudnn_enabled: true, + d_lr_weight: 0.2, d_name: null, data_path: /mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/, + data_path_1: /mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/datasets/hub/, + dataset: beat_sep_lower, ddp: false, debug: false, decay_epochs: 200, decay_rate: 0.1, + decode_fusion: null, depth: 3, deterministic: true, dilation_growth_rate: 3, disable_filtering: false, + div_reg_weight: 0.0, downs_t: [3], dropout_prob: 0.3, e_name: VAESKConv, e_path: weights/AESKConv_240_100.bin, + emb_width: 512, emo_rep: null, emotion_dims: 8, emotion_f: 0, epoch_stage: 0, epochs: 1000, + eval_model: motion_representation, f_encoder: 'null', f_fix_pre: false, f_pre_encoder: 'null', + fac_prob: 1.0, facial_dims: 100, facial_f: 0, facial_fps: 15, facial_norm: false, + facial_rep: smplxflame_30, fid_weight: 0.0, finger_net: original, freeze_wordembed: false, + fsmr: 0.0, ftmr: 0.0, fusion_mode: sum, g_name: MDM, gap_weight: 0.0, gpus: [0], + grad_norm: 0.99, hidden_size: 768, hvqvae_multipliers: [1], id_rep: onehot, input_context: both, + is_train: true, ita_weight: 0.0, iwa_weight: 0.0, joint_channel: 3, kld_aud_weight: 0.0, + kld_fac_weight: 0.0, kld_weight: 0.0, l: 4, l_bins: 512, l_mu: 0.99, levels: 1, + lf: 3.0, lh: 3.0, ll: 3.0, loader_workers: 0, log_period: 10, loss_contrastive_neg_weight: 0.005, + loss_contrastive_pos_weight: 0.2, loss_gan_weight: 5.0, loss_kld_weight: 0.1, loss_physical_weight: 0.0, + loss_reg_weight: 0.05, loss_regression_weight: 70.0, lr_base: 5.0e-05, lr_min: 1.0e-07, + lr_policy: step, lu: 3.0, m_conv: 1.0, m_decoder: null, m_encoder: 'null', m_fix_pre: false, + m_pre_encoder: 'null', mean_pose_path: /mnt/fu09a/chenbohong/PantoMatrix/beatx_2_330_mean.npy, + mean_trans_path: /mnt/fu09a/chenbohong/PantoMatrix/beatx_2_trans_mean.npy, model: denoiser, + momentum: 0.8, motion_f: 256, msmr: 0.0, mtmr: 0.0, multi_length_training: [1.0], + n_layer: 1, n_poses: 34, n_pre_poses: 4, name: 0403_212319_diffusion_rvqvae_128, + nesterov: true, new_cache: false, no_adv_epoch: 999, notes: '', opt: adam, opt_betas: [ + 0.5, 0.999], ori_joints: beat_smplx_joints, out_path: /mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/outputs/audio2pose/, + pos_encoding_type: sin, pos_prob: 1.0, pose_dims: 330, pose_fps: 30, pose_length: 128, + pose_norm: true, pose_rep: smplxflame_30, pre_frames: 4, pre_type: zero, pretrain: false, + project: s2g, queue_size: 1024, random_seed: 2021, rec_aud_weight: 0.0, rec_fac_weight: 0.0, + rec_pos_weight: 0.0, rec_txt_weight: 0.0, rec_ver_weight: 0.0, rec_weight: 1.0, + root_path: /mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/, root_weight: 1.0, + rot6d: true, sample_length: 34, sem_rep: null, sparse: 1, speaker_dims: 4, speaker_f: 0, + speaker_id: onehot, stat: ts, std_pose_path: /mnt/fu09a/chenbohong/PantoMatrix/beatx_2_330_std.npy, + std_trans_path: /mnt/fu09a/chenbohong/PantoMatrix/beatx_2_trans_std.npy, stride: 20, + strides_t: [2], t_encoder: 'null', t_fix_pre: false, t_pre_encoder: fasttext, tar_joints: beat_smplx_full, + test_ckpt: /mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/outputs/audio2pose/custom/0330_140056_diffusion_rvqvae/last_300.bin, + test_data_path: /datasets/trinity/test/, test_length: 128, test_period: 20, train_data_path: /datasets/trinity/train/, + train_trans: true, trainer: diffusion_rvqvae, training_speakers: [2], tsmr: 0.0, + ttmr: 0.0, txt_prob: 1.0, use_amass: false, use_aug: false, use_bottleneck: true, + use_trans: true, vae_codebook_size: 256, vae_grow: [1, 1, 2, 1], vae_layer: 4, vae_length: 240, + vae_quantizer_lambda: 1.0, vae_test_dim: 330, vae_test_len: 32, vae_test_stride: 20, + val_data_path: /datasets/trinity/val/, variational: false, vel: 1, vel_weight: 0.0, + vqvae_ckpt: null, vqvae_hands_path: /mnt/fu09a/chenbohong/gdc/T2M-GPT/output_beatx2/RVQVAE_hands/net_300000.pth, + vqvae_latent_scale: 5.0, vqvae_lower_path: /mnt/fu09a/chenbohong/gdc/T2M-GPT/output_beatx2/RVQVAE_lower/net_300000.pth, + vqvae_lower_trans_path: /mnt/fu09a/chenbohong/gdc/T2M-GPT/output_beatx2/RVQVAE_lower_trans/net_300000.pth, + vqvae_reverse_decoder_dilation: true, vqvae_squeeze_scale: 4, vqvae_type: rvqvae, + vqvae_upper_path: /mnt/fu09a/chenbohong/gdc/T2M-GPT/output_beatx2/RVQVAE_upper/net_300000.pth, + warmup_epochs: 0, warmup_lr: 0.0005, wei_weight: 0.0, weight_decay: 0.0, width: 512, + word_cache: false, word_dims: 300, word_f: 256, word_index_num: 11195, word_rep: textgrid, + z_type: speaker} diff --git a/ckpt/beatx2_cospeech_diffusion/1001_203942_diffusion_rvqvae_128_gaps-210-0.txt b/ckpt/beatx2_cospeech_diffusion/1001_203942_diffusion_rvqvae_128_gaps-210-0.txt new file mode 100644 index 0000000000000000000000000000000000000000..6a7282c374140b0c70b2e272d9ebe69ce9712938 --- /dev/null +++ b/ckpt/beatx2_cospeech_diffusion/1001_203942_diffusion_rvqvae_128_gaps-210-0.txt @@ -0,0 +1,451 @@ + 10-01 20:39:43 | {'a_encoder': None, + 'a_fix_pre': False, + 'a_pre_encoder': None, + 'acc': 1, + 'acc_weight': 0.0, + 'additional_data': False, + 'adv_weight': 20.0, + 'ali_weight': 0.0, + 'amsgrad': False, + 'apex': False, + 'asmr': 0.0, + 'atcont': 0.0, + 'atmr': 0.0, + 'aud_prob': 1.0, + 'audio_dims': 1, + 'audio_f': 256, + 'audio_fps': 16000, + 'audio_norm': False, + 'audio_rep': 'onset+amplitude', + 'audio_sr': 16000, + 'batch_size': 40, + 'beat_align': True, + 'benchmark': True, + 'cache_only': False, + 'cache_path': 'datasets/beat_cache/beat_smplx_en_emage_2_128/', + 'cf': 0.0, + 'ch': 1.0, + 'cl': 1.0, + 'clean_final_seconds': 0, + 'clean_first_seconds': 0, + 'commit': 0.02, + 'config': 'configs/diffusion_rvqvae_128_gaps-210-0.yaml', + 'csv_name': 'a2g_0', + 'cu': 1.0, + 'cudnn_enabled': True, + 'd_lr_weight': 0.2, + 'd_name': None, + 'data_path': './datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/', + 'data_path_1': './datasets/hub/', + 'dataset': 'beat_sep_lower', + 'ddp': False, + 'debug': False, + 'decay_epochs': 500, + 'decay_rate': 0.1, + 'decode_fusion': None, + 'depth': 3, + 'deterministic': True, + 'dilation_growth_rate': 3, + 'disable_filtering': False, + 'div_reg_weight': 0.0, + 'downs_t': [3], + 'dropout_prob': 0.3, + 'e_name': 'VAESKConv', + 'e_path': 'weights/AESKConv_240_100.bin', + 'emb_width': 512, + 'emo_rep': None, + 'emotion_dims': 8, + 'emotion_f': 0, + 'epoch_stage': 0, + 'epochs': 2000, + 'eval_model': 'motion_representation', + 'f_encoder': 'null', + 'f_fix_pre': False, + 'f_pre_encoder': 'null', + 'fac_prob': 1.0, + 'facial_dims': 100, + 'facial_f': 0, + 'facial_fps': 15, + 'facial_norm': False, + 'facial_rep': 'smplxflame_30', + 'fid_weight': 0.0, + 'finger_net': 'original', + 'freeze_wordembed': False, + 'fsmr': 0.0, + 'ftmr': 0.0, + 'fusion_mode': 'sum', + 'g_name': 'MDM', + 'gap_weight': 0.0, + 'gpus': [0], + 'grad_norm': 0.99, + 'hidden_size': 768, + 'hvqvae_multipliers': [1], + 'id_rep': 'onehot', + 'input_context': 'both', + 'is_train': True, + 'ita_weight': 0.0, + 'iwa_weight': 0.0, + 'joint_channel': 3, + 'kld_aud_weight': 0.0, + 'kld_fac_weight': 0.0, + 'kld_weight': 0.0, + 'l': 4, + 'l_bins': 512, + 'l_mu': 0.99, + 'levels': 1, + 'lf': 3.0, + 'lh': 3.0, + 'll': 3.0, + 'loader_workers': 0, + 'log_period': 10, + 'loss_contrastive_neg_weight': 0.005, + 'loss_contrastive_pos_weight': 0.2, + 'loss_gan_weight': 5.0, + 'loss_kld_weight': 0.1, + 'loss_physical_weight': 0.0, + 'loss_reg_weight': 0.05, + 'loss_regression_weight': 70.0, + 'lr_base': 5e-05, + 'lr_min': 1e-07, + 'lr_policy': 'step', + 'lu': 3.0, + 'm_conv': 1.0, + 'm_decoder': None, + 'm_encoder': 'null', + 'm_fix_pre': False, + 'm_pre_encoder': 'null', + 'mean_pose_path': '../../beatx_2_330_mean.npy', + 'mean_trans_path': '../../beatx_2_trans_mean.npy', + 'model': 'denoiser', + 'momentum': 0.8, + 'motion_f': 256, + 'msmr': 0.0, + 'mtmr': 0.0, + 'multi_length_training': [1.0], + 'n_layer': 1, + 'n_poses': 34, + 'n_pre_poses': 4, + 'name': '1001_203942_diffusion_rvqvae_128_gaps-210-0', + 'nesterov': True, + 'new_cache': False, + 'no_adv_epoch': 999, + 'notes': '', + 'opt': 'adam', + 'opt_betas': [0.5, 0.999], + 'ori_joints': 'beat_smplx_joints', + 'out_path': './outputs/audio2pose/', + 'pos_encoding_type': 'sin', + 'pos_prob': 1.0, + 'pose_dims': 330, + 'pose_fps': 30, + 'pose_length': 128, + 'pose_norm': True, + 'pose_rep': 'smplxflame_30', + 'pre_frames': 4, + 'pre_type': 'zero', + 'pretrain': False, + 'project': 's2g', + 'queue_size': 1024, + 'random_seed': 2021, + 'rec_aud_weight': 0.0, + 'rec_fac_weight': 0.0, + 'rec_pos_weight': 0.0, + 'rec_txt_weight': 0.0, + 'rec_ver_weight': 0.0, + 'rec_weight': 1.0, + 'root_path': './', + 'root_weight': 1.0, + 'rot6d': True, + 'sample_length': 34, + 'sem_rep': None, + 'sparse': 1, + 'speaker_dims': 4, + 'speaker_f': 0, + 'speaker_id': 'onehot', + 'stat': 'ts', + 'std_pose_path': '../../beatx_2_330_std.npy', + 'std_trans_path': '../../beatx_2_trans_std.npy', + 'stride': 20, + 'strides_t': [2], + 't_encoder': 'null', + 't_fix_pre': False, + 't_pre_encoder': 'fasttext', + 'tar_joints': 'beat_smplx_full', + 'test_ckpt': './outputs/audio2pose/custom/0403_212319_diffusion_rvqvae_128/last_500.bin', + 'test_data_path': '/datasets/trinity/test/', + 'test_length': 128, + 'test_period': 20, + 'train_data_path': '/datasets/trinity/train/', + 'train_trans': True, + 'trainer': 'diffusion_rvqvae', + 'training_speakers': [2], + 'tsmr': 0.0, + 'ttmr': 0.0, + 'txt_prob': 1.0, + 'use_amass': False, + 'use_aug': False, + 'use_bottleneck': True, + 'use_motionclip': False, + 'use_trans': True, + 'vae_codebook_size': 256, + 'vae_grow': [1, 1, 2, 1], + 'vae_layer': 4, + 'vae_length': 240, + 'vae_quantizer_lambda': 1.0, + 'vae_test_dim': 330, + 'vae_test_len': 32, + 'vae_test_stride': 20, + 'val_data_path': '/datasets/trinity/val/', + 'variational': False, + 'vel': 1, + 'vel_weight': 0.0, + 'vqvae_ckpt': None, + 'vqvae_hands_path': './datasets/hub/output_beatx2/RVQVAE_hands/net_300000.pth', + 'vqvae_latent_scale': 5.0, + 'vqvae_lower_path': './datasets/hub/output_beatx2/RVQVAE_lower/net_300000.pth', + 'vqvae_lower_trans_path': './datasets/hub/output_beatx2/RVQVAE_lower_trans/net_300000.pth', + 'vqvae_reverse_decoder_dilation': True, + 'vqvae_squeeze_scale': 4, + 'vqvae_type': 'rvqvae', + 'vqvae_upper_path': './datasets/hub/output_beatx2/RVQVAE_upper/net_300000.pth', + 'warmup_epochs': 0, + 'warmup_lr': 0.0005, + 'wei_weight': 0.0, + 'weight_decay': 0.0, + 'width': 512, + 'word_cache': False, + 'word_dims': 300, + 'word_f': 256, + 'word_index_num': 11195, + 'word_rep': 'textgrid', + 'z_type': 'speaker'} + 10-01 20:39:43 | # ------------ 1001_203942_diffusion_rvqvae_128_gaps-210-0 ----------- # + 10-01 20:39:43 | PyTorch version: 2.4.1+cu121 + 10-01 20:39:43 | CUDA version: 12.1 + 10-01 20:39:43 | 1 GPUs + 10-01 20:39:43 | Random Seed: 2021 + 10-01 20:39:46 | Audio bit rate: 16000 + 10-01 20:39:46 | Reading data './datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/'... + 10-01 20:39:46 | Creating the dataset cache... + 10-01 20:39:46 | Found the cache ./datasets/beat_cache/beat_smplx_en_emage_2_128/train/smplxflame_30_cache + 10-01 20:39:46 | Init train dataloader success + 10-01 20:39:46 | Init val dataloader success + 10-01 20:39:46 | Audio bit rate: 16000 + 10-01 20:39:46 | Reading data './datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/'... + 10-01 20:39:46 | Creating the dataset cache... + 10-01 20:39:46 | Found the cache ./datasets/beat_cache/beat_smplx_en_emage_2_128/test/smplxflame_30_cache + 10-01 20:39:46 | Init test dataloader success + 10-01 20:39:46 | DataParallel( + (module): MDM( + (WavEncoder): WavEncoder( + (feat_extractor): Sequential( + (0): BasicBlock( + (conv1): Conv1d(2, 64, kernel_size=(15,), stride=(5,), padding=(1700,)) + (bn1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act1): LeakyReLU(negative_slope=0.01, inplace=True) + (conv2): Conv1d(64, 64, kernel_size=(15,), stride=(1,), padding=(7,)) + (bn2): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act2): LeakyReLU(negative_slope=0.01, inplace=True) + (downsample): Sequential( + (0): Conv1d(2, 64, kernel_size=(15,), stride=(5,), padding=(1700,)) + (1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (1): BasicBlock( + (conv1): Conv1d(64, 64, kernel_size=(15,), stride=(6,)) + (bn1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act1): LeakyReLU(negative_slope=0.01, inplace=True) + (conv2): Conv1d(64, 64, kernel_size=(15,), stride=(1,), padding=(7,)) + (bn2): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act2): LeakyReLU(negative_slope=0.01, inplace=True) + (downsample): Sequential( + (0): Conv1d(64, 64, kernel_size=(15,), stride=(6,)) + (1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (2): BasicBlock( + (conv1): Conv1d(64, 64, kernel_size=(15,), stride=(1,), padding=(7,)) + (bn1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act1): LeakyReLU(negative_slope=0.01, inplace=True) + (conv2): Conv1d(64, 64, kernel_size=(15,), stride=(1,), padding=(7,)) + (bn2): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act2): LeakyReLU(negative_slope=0.01, inplace=True) + ) + (3): BasicBlock( + (conv1): Conv1d(64, 128, kernel_size=(15,), stride=(6,)) + (bn1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act1): LeakyReLU(negative_slope=0.01, inplace=True) + (conv2): Conv1d(128, 128, kernel_size=(15,), stride=(1,), padding=(7,)) + (bn2): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act2): LeakyReLU(negative_slope=0.01, inplace=True) + (downsample): Sequential( + (0): Conv1d(64, 128, kernel_size=(15,), stride=(6,)) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (4): BasicBlock( + (conv1): Conv1d(128, 128, kernel_size=(15,), stride=(1,), padding=(7,)) + (bn1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act1): LeakyReLU(negative_slope=0.01, inplace=True) + (conv2): Conv1d(128, 128, kernel_size=(15,), stride=(1,), padding=(7,)) + (bn2): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act2): LeakyReLU(negative_slope=0.01, inplace=True) + ) + (5): BasicBlock( + (conv1): Conv1d(128, 256, kernel_size=(15,), stride=(3,)) + (bn1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act1): LeakyReLU(negative_slope=0.01, inplace=True) + (conv2): Conv1d(256, 256, kernel_size=(15,), stride=(1,), padding=(7,)) + (bn2): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (act2): LeakyReLU(negative_slope=0.01, inplace=True) + (downsample): Sequential( + (0): Conv1d(128, 256, kernel_size=(15,), stride=(3,)) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + ) + ) + (text_encoder_body): Linear(in_features=300, out_features=256, bias=True) + (text_pre_encoder_body): Embedding(11195, 300) + (sequence_pos_encoder): PositionalEncoding( + (dropout): Dropout(p=0.1, inplace=False) + ) + (mytimmblocks): ModuleList( + (0-7): 8 x Block( + (norm1): LayerNorm((512,), eps=1e-05, elementwise_affine=True) + (attn): Attention( + (qkv): Linear(in_features=512, out_features=1536, bias=False) + (q_norm): Identity() + (k_norm): Identity() + (attn_drop): Dropout(p=0.0, inplace=False) + (proj): Linear(in_features=512, out_features=512, bias=True) + (proj_drop): Dropout(p=0.0, inplace=False) + ) + (ls1): Identity() + (drop_path1): DropPath(drop_prob=0.100) + (norm2): LayerNorm((512,), eps=1e-05, elementwise_affine=True) + (mlp): Mlp( + (fc1): Linear(in_features=512, out_features=1024, bias=True) + (act): GELU(approximate='none') + (drop1): Dropout(p=0.0, inplace=False) + (norm): Identity() + (fc2): Linear(in_features=1024, out_features=512, bias=True) + (drop2): Dropout(p=0.0, inplace=False) + ) + (ls2): Identity() + (drop_path2): DropPath(drop_prob=0.100) + ) + ) + (embed_timestep): TimestepEmbedder( + (sequence_pos_encoder): PositionalEncoding( + (dropout): Dropout(p=0.1, inplace=False) + ) + (time_embed): Sequential( + (0): Linear(in_features=512, out_features=512, bias=True) + (1): SiLU() + (2): Linear(in_features=512, out_features=512, bias=True) + ) + ) + (embed_style): Linear(in_features=6, out_features=64, bias=True) + (embed_text): Linear(in_features=6144, out_features=512, bias=True) + (output_process): OutputProcess( + (poseFinal): Linear(in_features=512, out_features=1536, bias=True) + ) + (rel_pos): SinusoidalEmbeddings() + (input_process): InputProcess( + (poseEmbedding): Linear(in_features=1536, out_features=512, bias=True) + ) + (input_process2): Linear(in_features=1280, out_features=512, bias=True) + (mix_audio_text): Linear(in_features=512, out_features=256, bias=True) + ) +) + 10-01 20:39:46 | init MDM success + 10-01 20:39:46 | load self-pretrained checkpoints for VAESKConv + 10-01 20:39:46 | load self-pretrained checkpoints for VAESKConv + 10-01 20:39:46 | VAESKConv( + (encoder): LocalEncoder( + (layers): ModuleList( + (0): Sequential( + (0): SkeletonResidual( + (residual): Sequential( + (0): SkeletonConv() + (1): GroupNorm(10, 330, eps=1e-05, affine=True) + ) + (shortcut): SkeletonConv() + (common): Sequential( + (0): SkeletonPool() + (1): Tanh() + ) + ) + ) + (1): Sequential( + (0): SkeletonResidual( + (residual): Sequential( + (0): SkeletonConv() + (1): GroupNorm(10, 210, eps=1e-05, affine=True) + ) + (shortcut): SkeletonConv() + (common): Sequential( + (0): SkeletonPool() + (1): Tanh() + ) + ) + ) + (2-3): 2 x Sequential( + (0): SkeletonResidual( + (residual): Sequential( + (0): SkeletonConv() + (1): GroupNorm(10, 240, eps=1e-05, affine=True) + ) + (shortcut): SkeletonConv() + (common): Sequential( + (0): Tanh() + ) + ) + ) + ) + ) + (decoder): VQDecoderV3( + (main): Sequential( + (0): ResBlock( + (model): Sequential( + (0): Conv1d(240, 240, kernel_size=(3,), stride=(1,), padding=(1,)) + (1): LeakyReLU(negative_slope=0.2, inplace=True) + (2): Conv1d(240, 240, kernel_size=(3,), stride=(1,), padding=(1,)) + ) + ) + (1): ResBlock( + (model): Sequential( + (0): Conv1d(240, 240, kernel_size=(3,), stride=(1,), padding=(1,)) + (1): LeakyReLU(negative_slope=0.2, inplace=True) + (2): Conv1d(240, 240, kernel_size=(3,), stride=(1,), padding=(1,)) + ) + ) + (2): Upsample(scale_factor=2.0, mode='nearest') + (3): Conv1d(240, 240, kernel_size=(3,), stride=(1,), padding=(1,)) + (4): LeakyReLU(negative_slope=0.2, inplace=True) + (5): Upsample(scale_factor=2.0, mode='nearest') + (6): Conv1d(240, 240, kernel_size=(3,), stride=(1,), padding=(1,)) + (7): LeakyReLU(negative_slope=0.2, inplace=True) + (8): Upsample(scale_factor=2.0, mode='nearest') + (9): Conv1d(240, 240, kernel_size=(3,), stride=(1,), padding=(1,)) + (10): LeakyReLU(negative_slope=0.2, inplace=True) + (11): Upsample(scale_factor=2.0, mode='nearest') + (12): Conv1d(240, 330, kernel_size=(3,), stride=(1,), padding=(1,)) + (13): LeakyReLU(negative_slope=0.2, inplace=True) + (14): Conv1d(330, 330, kernel_size=(3,), stride=(1,), padding=(1,)) + ) + ) + (fc_mu): Linear(in_features=240, out_features=240, bias=True) + (fc_logvar): Linear(in_features=240, out_features=240, bias=True) +) + 10-01 20:39:46 | init VAESKConv success + 10-01 20:39:47 | load self-pretrained checkpoints for VAESKConv + 10-01 20:39:47 | load self-pretrained checkpoints for VAESKConv + 10-01 20:39:48 | load self-pretrained checkpoints for MDM + 10-01 21:08:57 | l2 loss: 0.0 + 10-01 21:08:57 | lvel loss: 0.0 + 10-01 21:08:58 | fid score: 0.46525881529758983 + 10-01 21:08:58 | align score: 0.7361291368819373 + 10-01 21:08:58 | l1div score: 12.30848217010498 + 10-01 21:08:58 | total inference time: 1749 s for 945 s motion diff --git a/ckpt/beatx2_cospeech_diffusion/last_500.bin b/ckpt/beatx2_cospeech_diffusion/last_500.bin new file mode 100644 index 0000000000000000000000000000000000000000..dfe3dd7c32ca03fd7427181b10b971bc27f285d9 --- /dev/null +++ b/ckpt/beatx2_cospeech_diffusion/last_500.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d79b6fd3e412f7e3cb61eb6795ff686f6cdf80d32ce2bf941cd985d8cae24cc1 +size 128770342 diff --git a/ckpt/beatx2_rvqvae/RVQVAE_hands/net_300000.pth b/ckpt/beatx2_rvqvae/RVQVAE_hands/net_300000.pth new file mode 100644 index 0000000000000000000000000000000000000000..9839d1b99620e6a1d3d015e183e286aeb9bac2d1 --- /dev/null +++ b/ckpt/beatx2_rvqvae/RVQVAE_hands/net_300000.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f4eb84ff69009be0b3e68419c5382aa10443b73739dfe2e2928b046e2db59a8b +size 83048747 diff --git a/ckpt/beatx2_rvqvae/RVQVAE_hands/run.log b/ckpt/beatx2_rvqvae/RVQVAE_hands/run.log new file mode 100644 index 0000000000000000000000000000000000000000..d2fadaabc688717408e7924a23053da2727a13b9 --- /dev/null +++ b/ckpt/beatx2_rvqvae/RVQVAE_hands/run.log @@ -0,0 +1,1630 @@ +2024-03-16 22:21:01,432 INFO { + "batch_size": 256, + "beta": 1.0, + "body_part": "hands", + "code_dim": 512, + "commit": 0.02, + "dataname": "kit", + "depth": 3, + "dilation_growth_rate": 3, + "down_t": 2, + "eval_iter": 1000, + "exp_name": "RVQVAE", + "gamma": 0.05, + "loss_vel": 0.5, + "lr": 0.0002, + "lr_scheduler": [ + 200000 + ], + "mu": 0.99, + "nb_code": 512, + "nb_vis": 20, + "out_dir": "output_beatx2/RVQVAE_hands", + "output_emb_width": 512, + "print_iter": 200, + "quantizer": "ema_reset", + "recons_loss": "l1_smooth", + "results_dir": "visual_results/", + "resume_gpt": null, + "resume_pth": null, + "seed": 123, + "stride_t": 2, + "total_iter": 300000, + "vis_gt": false, + "visual_name": "baseline", + "vq_act": "relu", + "vq_norm": null, + "warm_up_iter": 1000, + "weight_decay": 0.0, + "width": 512, + "window_size": 64 +} +2024-03-16 22:21:01,442 INFO Training on kit, motions are with 21 joints +2024-03-16 22:21:11,139 INFO { + "batch_size": 256, + "beta": 1.0, + "body_part": "hands", + "code_dim": 512, + "commit": 0.02, + "dataname": "kit", + "depth": 3, + "dilation_growth_rate": 3, + "down_t": 2, + "eval_iter": 1000, + "exp_name": "RVQVAE", + "gamma": 0.05, + "loss_vel": 0.5, + "lr": 0.0002, + "lr_scheduler": [ + 200000 + ], + "mu": 0.99, + "nb_code": 512, + "nb_vis": 20, + "out_dir": "output_beatx2/RVQVAE_hands", + "output_emb_width": 512, + "print_iter": 200, + "quantizer": "ema_reset", + "recons_loss": "l1_smooth", + "results_dir": "visual_results/", + "resume_gpt": null, + "resume_pth": null, + "seed": 123, + "stride_t": 2, + "total_iter": 300000, + "vis_gt": false, + "visual_name": "baseline", + "vq_act": "relu", + "vq_norm": null, + "warm_up_iter": 1000, + "weight_decay": 0.0, + "width": 512, + "window_size": 64 +} +2024-03-16 22:21:11,147 INFO Training on kit, motions are with 21 joints +2024-03-16 22:21:17,353 INFO { + "batch_size": 256, + "beta": 1.0, + "body_part": "hands", + "code_dim": 512, + "commit": 0.02, + "dataname": "kit", + "depth": 3, + "dilation_growth_rate": 3, + "down_t": 2, + "eval_iter": 1000, + "exp_name": "RVQVAE", + "gamma": 0.05, + "loss_vel": 0.5, + "lr": 0.0002, + "lr_scheduler": [ + 200000 + ], + "mu": 0.99, + "nb_code": 512, + "nb_vis": 20, + "out_dir": "output_beatx2/RVQVAE_hands", + "output_emb_width": 512, + "print_iter": 200, + "quantizer": "ema_reset", + "recons_loss": "l1_smooth", + "results_dir": "visual_results/", + "resume_gpt": null, + "resume_pth": null, + "seed": 123, + "stride_t": 2, + "total_iter": 300000, + "vis_gt": false, + "visual_name": "baseline", + "vq_act": "relu", + "vq_norm": null, + "warm_up_iter": 1000, + "weight_decay": 0.0, + "width": 512, + "window_size": 64 +} +2024-03-16 22:21:17,360 INFO Training on kit, motions are with 21 joints +2024-03-16 22:21:38,618 INFO Warmup. Iter 200 : lr 0.00004 Commit. 0.03481 PPL. 269.55 Recons. 0.31527 +2024-03-16 22:21:58,504 INFO Warmup. Iter 400 : lr 0.00008 Commit. 0.14665 PPL. 193.02 Recons. 0.18248 +2024-03-16 22:22:18,394 INFO Warmup. Iter 600 : lr 0.00012 Commit. 0.61930 PPL. 264.56 Recons. 0.12713 +2024-03-16 22:22:39,103 INFO Warmup. Iter 800 : lr 0.00016 Commit. 1.05793 PPL. 395.25 Recons. 0.09953 +2024-03-16 22:23:21,209 INFO Train. Iter 200 : Commit. 1.81709 PPL. 423.22 Recons. 0.06633 +2024-03-16 22:23:42,151 INFO Train. Iter 400 : Commit. 2.15557 PPL. 421.68 Recons. 0.05860 +2024-03-16 22:24:02,991 INFO Train. Iter 600 : Commit. 2.43316 PPL. 420.29 Recons. 0.05215 +2024-03-16 22:24:24,126 INFO Train. Iter 800 : Commit. 2.64115 PPL. 417.73 Recons. 0.04831 +2024-03-16 22:24:44,625 INFO Train. Iter 1000 : Commit. 2.70340 PPL. 419.06 Recons. 0.04192 +2024-03-16 22:25:06,043 INFO Train. Iter 1200 : Commit. 2.80954 PPL. 417.33 Recons. 0.03967 +2024-03-16 22:25:27,147 INFO Train. Iter 1400 : Commit. 2.92757 PPL. 413.53 Recons. 0.03951 +2024-03-16 22:25:49,449 INFO Train. Iter 1600 : Commit. 2.89680 PPL. 415.75 Recons. 0.03495 +2024-03-16 22:26:11,415 INFO Train. Iter 1800 : Commit. 3.00099 PPL. 412.46 Recons. 0.03768 +2024-03-16 22:26:32,955 INFO Train. Iter 2000 : Commit. 2.94512 PPL. 414.95 Recons. 0.03336 +2024-03-16 22:26:54,507 INFO Train. Iter 2200 : Commit. 2.89864 PPL. 416.76 Recons. 0.03028 +2024-03-16 22:27:16,191 INFO Train. Iter 2400 : Commit. 2.96991 PPL. 414.59 Recons. 0.03305 +2024-03-16 22:27:37,326 INFO Train. Iter 2600 : Commit. 2.95484 PPL. 415.74 Recons. 0.03041 +2024-03-16 22:27:57,532 INFO Train. Iter 2800 : Commit. 2.91535 PPL. 416.29 Recons. 0.02882 +2024-03-16 22:28:19,185 INFO Train. Iter 3000 : Commit. 2.91660 PPL. 415.48 Recons. 0.02841 +2024-03-16 22:28:40,719 INFO Train. Iter 3200 : Commit. 2.90406 PPL. 415.84 Recons. 0.02754 +2024-03-16 22:29:02,335 INFO Train. Iter 3400 : Commit. 2.84054 PPL. 416.26 Recons. 0.02712 +2024-03-16 22:29:24,180 INFO Train. Iter 3600 : Commit. 2.78504 PPL. 418.18 Recons. 0.02502 +2024-03-16 22:29:45,340 INFO Train. Iter 3800 : Commit. 2.82243 PPL. 416.37 Recons. 0.02635 +2024-03-16 22:30:06,961 INFO Train. Iter 4000 : Commit. 2.86772 PPL. 414.63 Recons. 0.02892 +2024-03-16 22:30:28,442 INFO Train. Iter 4200 : Commit. 2.87609 PPL. 414.72 Recons. 0.02781 +2024-03-16 22:30:50,120 INFO Train. Iter 4400 : Commit. 2.78671 PPL. 416.26 Recons. 0.02506 +2024-03-16 22:31:11,020 INFO Train. Iter 4600 : Commit. 2.75930 PPL. 416.30 Recons. 0.02623 +2024-03-16 22:31:31,369 INFO Train. Iter 4800 : Commit. 2.79248 PPL. 415.10 Recons. 0.02644 +2024-03-16 22:31:52,570 INFO Train. Iter 5000 : Commit. 2.77515 PPL. 416.17 Recons. 0.02625 +2024-03-16 22:32:14,673 INFO Train. Iter 5200 : Commit. 2.73056 PPL. 416.48 Recons. 0.02531 +2024-03-16 22:32:36,470 INFO Train. Iter 5400 : Commit. 2.67148 PPL. 417.74 Recons. 0.02328 +2024-03-16 22:32:58,263 INFO Train. Iter 5600 : Commit. 2.66978 PPL. 417.28 Recons. 0.02484 +2024-03-16 22:33:20,242 INFO Train. Iter 5800 : Commit. 2.62274 PPL. 417.62 Recons. 0.02315 +2024-03-16 22:33:42,314 INFO Train. Iter 6000 : Commit. 2.60634 PPL. 417.34 Recons. 0.02277 +2024-03-16 22:34:04,786 INFO Train. Iter 6200 : Commit. 2.58308 PPL. 417.33 Recons. 0.02405 +2024-03-16 22:34:27,067 INFO Train. Iter 6400 : Commit. 2.53522 PPL. 418.81 Recons. 0.02107 +2024-03-16 22:34:49,382 INFO Train. Iter 6600 : Commit. 2.55555 PPL. 418.17 Recons. 0.02246 +2024-03-16 22:35:09,528 INFO Train. Iter 6800 : Commit. 2.60122 PPL. 416.76 Recons. 0.02290 +2024-03-16 22:35:31,027 INFO Train. Iter 7000 : Commit. 2.63115 PPL. 415.53 Recons. 0.02577 +2024-03-16 22:35:52,475 INFO Train. Iter 7200 : Commit. 2.49839 PPL. 418.41 Recons. 0.02108 +2024-03-16 22:36:14,130 INFO Train. Iter 7400 : Commit. 2.50491 PPL. 417.25 Recons. 0.02288 +2024-03-16 22:36:35,011 INFO Train. Iter 7600 : Commit. 2.46762 PPL. 418.54 Recons. 0.02208 +2024-03-16 22:36:55,993 INFO Train. Iter 7800 : Commit. 2.32344 PPL. 421.97 Recons. 0.01812 +2024-03-16 22:37:17,022 INFO Train. Iter 8000 : Commit. 2.38876 PPL. 419.54 Recons. 0.02033 +2024-03-16 22:37:37,986 INFO Train. Iter 8200 : Commit. 2.43170 PPL. 419.15 Recons. 0.02189 +2024-03-16 22:37:58,976 INFO Train. Iter 8400 : Commit. 2.47125 PPL. 417.35 Recons. 0.02348 +2024-03-16 22:38:18,849 INFO Train. Iter 8600 : Commit. 2.42552 PPL. 419.02 Recons. 0.02024 +2024-03-16 22:38:40,477 INFO Train. Iter 8800 : Commit. 2.39589 PPL. 418.30 Recons. 0.02143 +2024-03-16 22:39:02,047 INFO Train. Iter 9000 : Commit. 2.32040 PPL. 419.96 Recons. 0.01920 +2024-03-16 22:39:23,467 INFO Train. Iter 9200 : Commit. 2.41454 PPL. 417.15 Recons. 0.02251 +2024-03-16 22:39:45,543 INFO Train. Iter 9400 : Commit. 2.28712 PPL. 421.33 Recons. 0.01811 +2024-03-16 22:40:07,101 INFO Train. Iter 9600 : Commit. 2.28247 PPL. 419.64 Recons. 0.01955 +2024-03-16 22:40:28,868 INFO Train. Iter 9800 : Commit. 2.28924 PPL. 421.26 Recons. 0.01816 +2024-03-16 22:40:51,034 INFO Train. Iter 10000 : Commit. 2.30242 PPL. 419.14 Recons. 0.02053 +2024-03-16 22:41:12,927 INFO Train. Iter 10200 : Commit. 2.30551 PPL. 418.94 Recons. 0.02155 +2024-03-16 22:41:35,102 INFO Train. Iter 10400 : Commit. 2.24866 PPL. 420.11 Recons. 0.01921 +2024-03-16 22:41:56,148 INFO Train. Iter 10600 : Commit. 2.22738 PPL. 420.27 Recons. 0.01784 +2024-03-16 22:42:17,517 INFO Train. Iter 10800 : Commit. 2.16196 PPL. 421.66 Recons. 0.01741 +2024-03-16 22:42:38,857 INFO Train. Iter 11000 : Commit. 2.18228 PPL. 421.26 Recons. 0.01784 +2024-03-16 22:43:00,599 INFO Train. Iter 11200 : Commit. 2.18826 PPL. 421.31 Recons. 0.01833 +2024-03-16 22:43:22,518 INFO Train. Iter 11400 : Commit. 2.17849 PPL. 420.44 Recons. 0.01903 +2024-03-16 22:43:44,232 INFO Train. Iter 11600 : Commit. 2.16630 PPL. 421.19 Recons. 0.01789 +2024-03-16 22:44:06,743 INFO Train. Iter 11800 : Commit. 2.13796 PPL. 422.42 Recons. 0.01728 +2024-03-16 22:44:28,934 INFO Train. Iter 12000 : Commit. 2.17549 PPL. 419.96 Recons. 0.01891 +2024-03-16 22:44:50,609 INFO Train. Iter 12200 : Commit. 2.14946 PPL. 421.45 Recons. 0.01788 +2024-03-16 22:45:12,771 INFO Train. Iter 12400 : Commit. 2.10443 PPL. 421.70 Recons. 0.01706 +2024-03-16 22:45:33,663 INFO Train. Iter 12600 : Commit. 2.09760 PPL. 421.99 Recons. 0.01663 +2024-03-16 22:45:56,338 INFO Train. Iter 12800 : Commit. 2.04306 PPL. 422.91 Recons. 0.01583 +2024-03-16 22:46:19,358 INFO Train. Iter 13000 : Commit. 2.09008 PPL. 421.32 Recons. 0.01774 +2024-03-16 22:46:41,828 INFO Train. Iter 13200 : Commit. 2.13393 PPL. 421.19 Recons. 0.01885 +2024-03-16 22:47:04,555 INFO Train. Iter 13400 : Commit. 2.08858 PPL. 421.72 Recons. 0.01721 +2024-03-16 22:47:26,697 INFO Train. Iter 13600 : Commit. 2.03244 PPL. 421.32 Recons. 0.01674 +2024-03-16 22:47:48,755 INFO Train. Iter 13800 : Commit. 2.03620 PPL. 423.01 Recons. 0.01609 +2024-03-16 22:48:11,796 INFO Train. Iter 14000 : Commit. 2.04502 PPL. 422.67 Recons. 0.01615 +2024-03-16 22:48:34,877 INFO Train. Iter 14200 : Commit. 2.04087 PPL. 420.92 Recons. 0.01765 +2024-03-16 22:48:55,781 INFO Train. Iter 14400 : Commit. 1.91749 PPL. 425.12 Recons. 0.01341 +2024-03-16 22:49:17,968 INFO Train. Iter 14600 : Commit. 2.04713 PPL. 420.75 Recons. 0.01787 +2024-03-16 22:49:39,804 INFO Train. Iter 14800 : Commit. 1.93286 PPL. 424.27 Recons. 0.01440 +2024-03-16 22:50:02,025 INFO Train. Iter 15000 : Commit. 2.06428 PPL. 420.60 Recons. 0.01826 +2024-03-16 22:50:24,594 INFO Train. Iter 15200 : Commit. 1.97884 PPL. 422.09 Recons. 0.01612 +2024-03-16 22:50:47,186 INFO Train. Iter 15400 : Commit. 2.06764 PPL. 420.19 Recons. 0.01891 +2024-03-16 22:51:13,677 INFO Train. Iter 15600 : Commit. 2.00389 PPL. 421.24 Recons. 0.01662 +2024-03-16 22:51:47,710 INFO Train. Iter 15800 : Commit. 2.06641 PPL. 420.75 Recons. 0.01788 +2024-03-16 22:52:21,356 INFO Train. Iter 16000 : Commit. 1.87790 PPL. 425.08 Recons. 0.01321 +2024-03-16 22:52:55,453 INFO Train. Iter 16200 : Commit. 1.92346 PPL. 421.07 Recons. 0.01603 +2024-03-16 22:53:28,110 INFO Train. Iter 16400 : Commit. 1.93800 PPL. 422.33 Recons. 0.01631 +2024-03-16 22:54:01,182 INFO Train. Iter 16600 : Commit. 1.93623 PPL. 420.52 Recons. 0.01584 +2024-03-16 22:54:34,659 INFO Train. Iter 16800 : Commit. 1.93628 PPL. 422.68 Recons. 0.01550 +2024-03-16 22:55:08,834 INFO Train. Iter 17000 : Commit. 1.91535 PPL. 421.91 Recons. 0.01582 +2024-03-16 22:55:42,971 INFO Train. Iter 17200 : Commit. 1.91006 PPL. 422.59 Recons. 0.01514 +2024-03-16 22:56:17,111 INFO Train. Iter 17400 : Commit. 1.87931 PPL. 423.34 Recons. 0.01465 +2024-03-16 22:56:51,333 INFO Train. Iter 17600 : Commit. 1.92440 PPL. 422.47 Recons. 0.01590 +2024-03-16 22:57:26,109 INFO Train. Iter 17800 : Commit. 1.82688 PPL. 423.85 Recons. 0.01397 +2024-03-16 22:58:00,749 INFO Train. Iter 18000 : Commit. 1.87452 PPL. 422.77 Recons. 0.01556 +2024-03-16 22:58:31,851 INFO Train. Iter 18200 : Commit. 1.86861 PPL. 422.92 Recons. 0.01474 +2024-03-16 22:58:54,160 INFO Train. Iter 18400 : Commit. 1.84654 PPL. 423.22 Recons. 0.01458 +2024-03-16 22:59:17,495 INFO Train. Iter 18600 : Commit. 1.90099 PPL. 421.33 Recons. 0.01607 +2024-03-16 22:59:40,987 INFO Train. Iter 18800 : Commit. 1.81519 PPL. 422.62 Recons. 0.01376 +2024-03-16 23:00:04,000 INFO Train. Iter 19000 : Commit. 1.78894 PPL. 424.73 Recons. 0.01316 +2024-03-16 23:00:27,007 INFO Train. Iter 19200 : Commit. 1.85764 PPL. 422.09 Recons. 0.01590 +2024-03-16 23:00:49,224 INFO Train. Iter 19400 : Commit. 1.86252 PPL. 422.13 Recons. 0.01510 +2024-03-16 23:01:11,542 INFO Train. Iter 19600 : Commit. 1.73901 PPL. 424.10 Recons. 0.01278 +2024-03-16 23:01:34,426 INFO Train. Iter 19800 : Commit. 1.89336 PPL. 418.67 Recons. 0.01638 +2024-03-16 23:01:57,058 INFO Train. Iter 20000 : Commit. 1.73883 PPL. 424.47 Recons. 0.01274 +2024-03-16 23:02:18,359 INFO Train. Iter 20200 : Commit. 1.84198 PPL. 419.83 Recons. 0.01613 +2024-03-16 23:02:40,586 INFO Train. Iter 20400 : Commit. 1.82652 PPL. 422.82 Recons. 0.01444 +2024-03-16 23:03:03,134 INFO Train. Iter 20600 : Commit. 1.83452 PPL. 420.24 Recons. 0.01597 +2024-03-16 23:03:25,991 INFO Train. Iter 20800 : Commit. 1.75791 PPL. 423.59 Recons. 0.01345 +2024-03-16 23:03:48,750 INFO Train. Iter 21000 : Commit. 1.78087 PPL. 421.57 Recons. 0.01504 +2024-03-16 23:04:11,200 INFO Train. Iter 21200 : Commit. 1.77684 PPL. 422.42 Recons. 0.01429 +2024-03-16 23:04:33,674 INFO Train. Iter 21400 : Commit. 1.74596 PPL. 422.47 Recons. 0.01369 +2024-03-16 23:04:56,148 INFO Train. Iter 21600 : Commit. 1.83492 PPL. 420.47 Recons. 0.01563 +2024-03-16 23:05:18,522 INFO Train. Iter 21800 : Commit. 1.74222 PPL. 422.94 Recons. 0.01329 +2024-03-16 23:05:41,451 INFO Train. Iter 22000 : Commit. 1.68384 PPL. 423.97 Recons. 0.01244 +2024-03-16 23:06:03,426 INFO Train. Iter 22200 : Commit. 1.75716 PPL. 422.01 Recons. 0.01449 +2024-03-16 23:06:26,283 INFO Train. Iter 22400 : Commit. 1.77039 PPL. 421.53 Recons. 0.01459 +2024-03-16 23:06:49,104 INFO Train. Iter 22600 : Commit. 1.83135 PPL. 420.07 Recons. 0.01621 +2024-03-16 23:07:11,756 INFO Train. Iter 22800 : Commit. 1.72128 PPL. 423.45 Recons. 0.01336 +2024-03-16 23:07:34,611 INFO Train. Iter 23000 : Commit. 1.67530 PPL. 422.54 Recons. 0.01254 +2024-03-16 23:07:57,158 INFO Train. Iter 23200 : Commit. 1.80075 PPL. 420.45 Recons. 0.01535 +2024-03-16 23:08:20,257 INFO Train. Iter 23400 : Commit. 1.70373 PPL. 423.23 Recons. 0.01316 +2024-03-16 23:08:42,628 INFO Train. Iter 23600 : Commit. 1.78824 PPL. 419.58 Recons. 0.01550 +2024-03-16 23:09:05,341 INFO Train. Iter 23800 : Commit. 1.80647 PPL. 420.81 Recons. 0.01537 +2024-03-16 23:09:27,831 INFO Train. Iter 24000 : Commit. 1.69553 PPL. 421.96 Recons. 0.01364 +2024-03-16 23:09:48,682 INFO Train. Iter 24200 : Commit. 1.69901 PPL. 422.76 Recons. 0.01297 +2024-03-16 23:10:10,447 INFO Train. Iter 24400 : Commit. 1.72323 PPL. 422.06 Recons. 0.01365 +2024-03-16 23:10:32,233 INFO Train. Iter 24600 : Commit. 1.67975 PPL. 421.21 Recons. 0.01383 +2024-03-16 23:10:54,065 INFO Train. Iter 24800 : Commit. 1.66258 PPL. 423.76 Recons. 0.01272 +2024-03-16 23:11:15,622 INFO Train. Iter 25000 : Commit. 1.74137 PPL. 421.34 Recons. 0.01420 +2024-03-16 23:11:38,409 INFO Train. Iter 25200 : Commit. 1.71997 PPL. 421.54 Recons. 0.01428 +2024-03-16 23:12:01,130 INFO Train. Iter 25400 : Commit. 1.73530 PPL. 420.75 Recons. 0.01454 +2024-03-16 23:12:24,199 INFO Train. Iter 25600 : Commit. 1.64677 PPL. 422.57 Recons. 0.01300 +2024-03-16 23:12:46,929 INFO Train. Iter 25800 : Commit. 1.61623 PPL. 423.76 Recons. 0.01190 +2024-03-16 23:13:08,790 INFO Train. Iter 26000 : Commit. 1.61810 PPL. 423.75 Recons. 0.01185 +2024-03-16 23:13:31,015 INFO Train. Iter 26200 : Commit. 1.67044 PPL. 422.52 Recons. 0.01341 +2024-03-16 23:13:53,761 INFO Train. Iter 26400 : Commit. 1.60528 PPL. 423.39 Recons. 0.01192 +2024-03-16 23:14:16,281 INFO Train. Iter 26600 : Commit. 1.68753 PPL. 419.93 Recons. 0.01413 +2024-03-16 23:14:39,714 INFO Train. Iter 26800 : Commit. 1.70679 PPL. 419.02 Recons. 0.01485 +2024-03-16 23:15:02,714 INFO Train. Iter 27000 : Commit. 1.64156 PPL. 421.90 Recons. 0.01280 +2024-03-16 23:15:25,170 INFO Train. Iter 27200 : Commit. 1.59367 PPL. 423.30 Recons. 0.01227 +2024-03-16 23:15:48,676 INFO Train. Iter 27400 : Commit. 1.58414 PPL. 422.61 Recons. 0.01190 +2024-03-16 23:16:11,526 INFO Train. Iter 27600 : Commit. 1.66369 PPL. 421.83 Recons. 0.01321 +2024-03-16 23:16:34,285 INFO Train. Iter 27800 : Commit. 1.62119 PPL. 421.51 Recons. 0.01254 +2024-03-16 23:16:55,691 INFO Train. Iter 28000 : Commit. 1.66019 PPL. 420.91 Recons. 0.01356 +2024-03-16 23:17:18,058 INFO Train. Iter 28200 : Commit. 1.63569 PPL. 421.79 Recons. 0.01312 +2024-03-16 23:17:40,856 INFO Train. Iter 28400 : Commit. 1.64194 PPL. 420.76 Recons. 0.01321 +2024-03-16 23:18:04,171 INFO Train. Iter 28600 : Commit. 1.64524 PPL. 419.79 Recons. 0.01355 +2024-03-16 23:18:26,987 INFO Train. Iter 28800 : Commit. 1.57346 PPL. 422.91 Recons. 0.01143 +2024-03-16 23:18:49,501 INFO Train. Iter 29000 : Commit. 1.68697 PPL. 418.49 Recons. 0.01480 +2024-03-16 23:19:12,112 INFO Train. Iter 29200 : Commit. 1.57872 PPL. 422.59 Recons. 0.01160 +2024-03-16 23:19:34,461 INFO Train. Iter 29400 : Commit. 1.57691 PPL. 421.84 Recons. 0.01231 +2024-03-16 23:19:57,243 INFO Train. Iter 29600 : Commit. 1.54113 PPL. 423.14 Recons. 0.01088 +2024-03-16 23:20:20,357 INFO Train. Iter 29800 : Commit. 1.61340 PPL. 418.77 Recons. 0.01376 +2024-03-16 23:20:41,592 INFO Train. Iter 30000 : Commit. 1.57429 PPL. 422.93 Recons. 0.01127 +2024-03-16 23:21:04,584 INFO Train. Iter 30200 : Commit. 1.56275 PPL. 421.65 Recons. 0.01173 +2024-03-16 23:21:27,641 INFO Train. Iter 30400 : Commit. 1.63669 PPL. 419.38 Recons. 0.01369 +2024-03-16 23:21:50,988 INFO Train. Iter 30600 : Commit. 1.56990 PPL. 420.79 Recons. 0.01222 +2024-03-16 23:22:13,070 INFO Train. Iter 30800 : Commit. 1.62029 PPL. 418.58 Recons. 0.01360 +2024-03-16 23:22:35,678 INFO Train. Iter 31000 : Commit. 1.63561 PPL. 418.52 Recons. 0.01351 +2024-03-16 23:22:58,824 INFO Train. Iter 31200 : Commit. 1.59206 PPL. 420.96 Recons. 0.01242 +2024-03-16 23:23:22,146 INFO Train. Iter 31400 : Commit. 1.60196 PPL. 418.39 Recons. 0.01325 +2024-03-16 23:23:45,033 INFO Train. Iter 31600 : Commit. 1.61259 PPL. 419.81 Recons. 0.01300 +2024-03-16 23:24:07,218 INFO Train. Iter 31800 : Commit. 1.62790 PPL. 418.55 Recons. 0.01373 +2024-03-16 23:24:29,832 INFO Train. Iter 32000 : Commit. 1.50763 PPL. 421.89 Recons. 0.01098 +2024-03-16 23:24:52,184 INFO Train. Iter 32200 : Commit. 1.57819 PPL. 419.76 Recons. 0.01270 +2024-03-16 23:25:15,400 INFO Train. Iter 32400 : Commit. 1.58387 PPL. 418.73 Recons. 0.01297 +2024-03-16 23:25:38,158 INFO Train. Iter 32600 : Commit. 1.61789 PPL. 419.38 Recons. 0.01318 +2024-03-16 23:26:01,249 INFO Train. Iter 32800 : Commit. 1.63929 PPL. 418.59 Recons. 0.01390 +2024-03-16 23:26:23,682 INFO Train. Iter 33000 : Commit. 1.55901 PPL. 420.29 Recons. 0.01214 +2024-03-16 23:26:46,312 INFO Train. Iter 33200 : Commit. 1.58089 PPL. 419.91 Recons. 0.01230 +2024-03-16 23:27:09,062 INFO Train. Iter 33400 : Commit. 1.50937 PPL. 420.13 Recons. 0.01156 +2024-03-16 23:27:31,892 INFO Train. Iter 33600 : Commit. 1.60592 PPL. 419.39 Recons. 0.01302 +2024-03-16 23:27:53,201 INFO Train. Iter 33800 : Commit. 1.48190 PPL. 422.31 Recons. 0.01026 +2024-03-16 23:28:15,726 INFO Train. Iter 34000 : Commit. 1.53060 PPL. 420.81 Recons. 0.01185 +2024-03-16 23:28:37,829 INFO Train. Iter 34200 : Commit. 1.54666 PPL. 419.63 Recons. 0.01206 +2024-03-16 23:29:00,562 INFO Train. Iter 34400 : Commit. 1.55841 PPL. 418.04 Recons. 0.01270 +2024-03-16 23:29:22,986 INFO Train. Iter 34600 : Commit. 1.53525 PPL. 420.22 Recons. 0.01188 +2024-03-16 23:29:45,024 INFO Train. Iter 34800 : Commit. 1.64573 PPL. 417.42 Recons. 0.01411 +2024-03-16 23:30:07,459 INFO Train. Iter 35000 : Commit. 1.57114 PPL. 418.95 Recons. 0.01255 +2024-03-16 23:30:30,228 INFO Train. Iter 35200 : Commit. 1.52628 PPL. 419.29 Recons. 0.01186 +2024-03-16 23:30:52,482 INFO Train. Iter 35400 : Commit. 1.56094 PPL. 418.70 Recons. 0.01251 +2024-03-16 23:31:15,235 INFO Train. Iter 35600 : Commit. 1.48499 PPL. 419.34 Recons. 0.01123 +2024-03-16 23:31:36,934 INFO Train. Iter 35800 : Commit. 1.50430 PPL. 419.63 Recons. 0.01171 +2024-03-16 23:31:59,380 INFO Train. Iter 36000 : Commit. 1.53344 PPL. 419.79 Recons. 0.01212 +2024-03-16 23:32:21,461 INFO Train. Iter 36200 : Commit. 1.56482 PPL. 416.29 Recons. 0.01323 +2024-03-16 23:32:44,151 INFO Train. Iter 36400 : Commit. 1.52263 PPL. 419.75 Recons. 0.01163 +2024-03-16 23:33:06,787 INFO Train. Iter 36600 : Commit. 1.57836 PPL. 417.51 Recons. 0.01335 +2024-03-16 23:33:29,540 INFO Train. Iter 36800 : Commit. 1.54796 PPL. 415.33 Recons. 0.01274 +2024-03-16 23:33:51,959 INFO Train. Iter 37000 : Commit. 1.52437 PPL. 420.10 Recons. 0.01135 +2024-03-16 23:34:14,686 INFO Train. Iter 37200 : Commit. 1.48945 PPL. 418.33 Recons. 0.01167 +2024-03-16 23:34:37,062 INFO Train. Iter 37400 : Commit. 1.54229 PPL. 419.11 Recons. 0.01197 +2024-03-16 23:34:59,018 INFO Train. Iter 37600 : Commit. 1.46594 PPL. 421.15 Recons. 0.01060 +2024-03-16 23:35:21,694 INFO Train. Iter 37800 : Commit. 1.50021 PPL. 417.11 Recons. 0.01228 +2024-03-16 23:35:44,055 INFO Train. Iter 38000 : Commit. 1.49042 PPL. 419.07 Recons. 0.01134 +2024-03-16 23:36:06,000 INFO Train. Iter 38200 : Commit. 1.50659 PPL. 419.39 Recons. 0.01172 +2024-03-16 23:36:28,207 INFO Train. Iter 38400 : Commit. 1.55265 PPL. 417.74 Recons. 0.01219 +2024-03-16 23:36:49,776 INFO Train. Iter 38600 : Commit. 1.48306 PPL. 419.17 Recons. 0.01142 +2024-03-16 23:37:11,412 INFO Train. Iter 38800 : Commit. 1.51917 PPL. 418.32 Recons. 0.01248 +2024-03-16 23:37:33,349 INFO Train. Iter 39000 : Commit. 1.50725 PPL. 418.74 Recons. 0.01197 +2024-03-16 23:37:55,202 INFO Train. Iter 39200 : Commit. 1.49058 PPL. 417.20 Recons. 0.01246 +2024-03-16 23:38:17,362 INFO Train. Iter 39400 : Commit. 1.49193 PPL. 418.56 Recons. 0.01152 +2024-03-16 23:38:38,669 INFO Train. Iter 39600 : Commit. 1.53956 PPL. 418.98 Recons. 0.01193 +2024-03-16 23:39:00,961 INFO Train. Iter 39800 : Commit. 1.47335 PPL. 419.74 Recons. 0.01117 +2024-03-16 23:39:23,403 INFO Train. Iter 40000 : Commit. 1.48879 PPL. 418.05 Recons. 0.01192 +2024-03-16 23:39:46,294 INFO Train. Iter 40200 : Commit. 1.44681 PPL. 418.99 Recons. 0.01092 +2024-03-16 23:40:09,172 INFO Train. Iter 40400 : Commit. 1.43370 PPL. 420.20 Recons. 0.01034 +2024-03-16 23:40:31,710 INFO Train. Iter 40600 : Commit. 1.49931 PPL. 417.41 Recons. 0.01214 +2024-03-16 23:40:54,284 INFO Train. Iter 40800 : Commit. 1.50749 PPL. 417.99 Recons. 0.01201 +2024-03-16 23:41:16,106 INFO Train. Iter 41000 : Commit. 1.42601 PPL. 418.95 Recons. 0.01054 +2024-03-16 23:41:37,316 INFO Train. Iter 41200 : Commit. 1.45971 PPL. 418.06 Recons. 0.01107 +2024-03-16 23:41:58,489 INFO Train. Iter 41400 : Commit. 1.48976 PPL. 417.35 Recons. 0.01202 +2024-03-16 23:42:18,606 INFO Train. Iter 41600 : Commit. 1.50738 PPL. 415.78 Recons. 0.01263 +2024-03-16 23:42:40,084 INFO Train. Iter 41800 : Commit. 1.43381 PPL. 419.12 Recons. 0.01051 +2024-03-16 23:43:01,459 INFO Train. Iter 42000 : Commit. 1.43243 PPL. 419.08 Recons. 0.01072 +2024-03-16 23:43:23,095 INFO Train. Iter 42200 : Commit. 1.52335 PPL. 416.67 Recons. 0.01265 +2024-03-16 23:43:44,783 INFO Train. Iter 42400 : Commit. 1.49065 PPL. 417.09 Recons. 0.01202 +2024-03-16 23:44:05,956 INFO Train. Iter 42600 : Commit. 1.46489 PPL. 417.83 Recons. 0.01129 +2024-03-16 23:44:27,684 INFO Train. Iter 42800 : Commit. 1.43929 PPL. 419.95 Recons. 0.01012 +2024-03-16 23:44:49,152 INFO Train. Iter 43000 : Commit. 1.38615 PPL. 420.12 Recons. 0.00945 +2024-03-16 23:45:10,801 INFO Train. Iter 43200 : Commit. 1.41519 PPL. 419.82 Recons. 0.01011 +2024-03-16 23:45:31,299 INFO Train. Iter 43400 : Commit. 1.44047 PPL. 418.43 Recons. 0.01089 +2024-03-16 23:45:52,906 INFO Train. Iter 43600 : Commit. 1.40028 PPL. 417.97 Recons. 0.01084 +2024-03-16 23:46:14,878 INFO Train. Iter 43800 : Commit. 1.43793 PPL. 418.08 Recons. 0.01116 +2024-03-16 23:46:36,588 INFO Train. Iter 44000 : Commit. 1.41214 PPL. 418.52 Recons. 0.01056 +2024-03-16 23:46:57,988 INFO Train. Iter 44200 : Commit. 1.42360 PPL. 416.54 Recons. 0.01107 +2024-03-16 23:47:19,205 INFO Train. Iter 44400 : Commit. 1.48759 PPL. 416.78 Recons. 0.01214 +2024-03-16 23:47:40,370 INFO Train. Iter 44600 : Commit. 1.47114 PPL. 418.09 Recons. 0.01132 +2024-03-16 23:48:01,867 INFO Train. Iter 44800 : Commit. 1.42983 PPL. 419.09 Recons. 0.01054 +2024-03-16 23:48:23,067 INFO Train. Iter 45000 : Commit. 1.48359 PPL. 415.78 Recons. 0.01215 +2024-03-16 23:48:44,634 INFO Train. Iter 45200 : Commit. 1.42243 PPL. 417.80 Recons. 0.01058 +2024-03-16 23:49:05,222 INFO Train. Iter 45400 : Commit. 1.42029 PPL. 417.23 Recons. 0.01100 +2024-03-16 23:49:26,787 INFO Train. Iter 45600 : Commit. 1.43266 PPL. 417.51 Recons. 0.01094 +2024-03-16 23:49:48,653 INFO Train. Iter 45800 : Commit. 1.43345 PPL. 418.19 Recons. 0.01071 +2024-03-16 23:50:10,297 INFO Train. Iter 46000 : Commit. 1.45180 PPL. 417.96 Recons. 0.01109 +2024-03-16 23:50:31,532 INFO Train. Iter 46200 : Commit. 1.44348 PPL. 416.12 Recons. 0.01179 +2024-03-16 23:50:53,037 INFO Train. Iter 46400 : Commit. 1.45968 PPL. 417.57 Recons. 0.01094 +2024-03-16 23:51:14,722 INFO Train. Iter 46600 : Commit. 1.43431 PPL. 417.91 Recons. 0.01074 +2024-03-16 23:51:36,118 INFO Train. Iter 46800 : Commit. 1.40742 PPL. 418.25 Recons. 0.01041 +2024-03-16 23:51:58,047 INFO Train. Iter 47000 : Commit. 1.37533 PPL. 419.49 Recons. 0.00952 +2024-03-16 23:52:19,644 INFO Train. Iter 47200 : Commit. 1.39630 PPL. 417.59 Recons. 0.01056 +2024-03-16 23:52:40,389 INFO Train. Iter 47400 : Commit. 1.49064 PPL. 413.71 Recons. 0.01309 +2024-03-16 23:53:01,708 INFO Train. Iter 47600 : Commit. 1.46923 PPL. 417.12 Recons. 0.01128 +2024-03-16 23:53:23,514 INFO Train. Iter 47800 : Commit. 1.42654 PPL. 418.01 Recons. 0.01047 +2024-03-16 23:53:45,277 INFO Train. Iter 48000 : Commit. 1.40467 PPL. 416.25 Recons. 0.01093 +2024-03-16 23:54:06,981 INFO Train. Iter 48200 : Commit. 1.43957 PPL. 416.27 Recons. 0.01152 +2024-03-16 23:54:28,989 INFO Train. Iter 48400 : Commit. 1.42236 PPL. 417.81 Recons. 0.01070 +2024-03-16 23:54:50,704 INFO Train. Iter 48600 : Commit. 1.39323 PPL. 417.03 Recons. 0.01047 +2024-03-16 23:55:12,238 INFO Train. Iter 48800 : Commit. 1.39433 PPL. 417.33 Recons. 0.01044 +2024-03-16 23:55:34,232 INFO Train. Iter 49000 : Commit. 1.37678 PPL. 418.60 Recons. 0.00989 +2024-03-16 23:55:54,651 INFO Train. Iter 49200 : Commit. 1.45034 PPL. 415.29 Recons. 0.01180 +2024-03-16 23:56:15,724 INFO Train. Iter 49400 : Commit. 1.38728 PPL. 417.69 Recons. 0.01013 +2024-03-16 23:56:37,068 INFO Train. Iter 49600 : Commit. 1.42087 PPL. 416.42 Recons. 0.01081 +2024-03-16 23:56:58,325 INFO Train. Iter 49800 : Commit. 1.45152 PPL. 414.31 Recons. 0.01184 +2024-03-16 23:57:19,666 INFO Train. Iter 50000 : Commit. 1.42840 PPL. 416.70 Recons. 0.01085 +2024-03-16 23:57:40,808 INFO Train. Iter 50200 : Commit. 1.34844 PPL. 418.23 Recons. 0.00934 +2024-03-16 23:58:02,313 INFO Train. Iter 50400 : Commit. 1.35724 PPL. 416.66 Recons. 0.01012 +2024-03-16 23:58:23,758 INFO Train. Iter 50600 : Commit. 1.39691 PPL. 415.71 Recons. 0.01096 +2024-03-16 23:58:45,175 INFO Train. Iter 50800 : Commit. 1.41466 PPL. 416.71 Recons. 0.01070 +2024-03-16 23:59:06,487 INFO Train. Iter 51000 : Commit. 1.34340 PPL. 418.37 Recons. 0.00937 +2024-03-16 23:59:26,614 INFO Train. Iter 51200 : Commit. 1.40108 PPL. 415.44 Recons. 0.01085 +2024-03-16 23:59:48,138 INFO Train. Iter 51400 : Commit. 1.33080 PPL. 418.51 Recons. 0.00927 +2024-03-17 00:00:09,298 INFO Train. Iter 51600 : Commit. 1.37541 PPL. 417.01 Recons. 0.01040 +2024-03-17 00:00:30,883 INFO Train. Iter 51800 : Commit. 1.41946 PPL. 414.45 Recons. 0.01162 +2024-03-17 00:00:52,118 INFO Train. Iter 52000 : Commit. 1.35067 PPL. 417.98 Recons. 0.00948 +2024-03-17 00:01:13,525 INFO Train. Iter 52200 : Commit. 1.40638 PPL. 416.49 Recons. 0.01035 +2024-03-17 00:01:35,133 INFO Train. Iter 52400 : Commit. 1.40738 PPL. 414.90 Recons. 0.01083 +2024-03-17 00:01:56,862 INFO Train. Iter 52600 : Commit. 1.34264 PPL. 417.49 Recons. 0.00964 +2024-03-17 00:02:18,187 INFO Train. Iter 52800 : Commit. 1.39814 PPL. 414.37 Recons. 0.01129 +2024-03-17 00:02:39,604 INFO Train. Iter 53000 : Commit. 1.40056 PPL. 416.04 Recons. 0.01060 +2024-03-17 00:03:00,343 INFO Train. Iter 53200 : Commit. 1.35868 PPL. 416.85 Recons. 0.00987 +2024-03-17 00:03:21,564 INFO Train. Iter 53400 : Commit. 1.41990 PPL. 414.04 Recons. 0.01167 +2024-03-17 00:03:43,077 INFO Train. Iter 53600 : Commit. 1.39797 PPL. 416.51 Recons. 0.01037 +2024-03-17 00:04:04,103 INFO Train. Iter 53800 : Commit. 1.38738 PPL. 415.85 Recons. 0.01025 +2024-03-17 00:04:24,941 INFO Train. Iter 54000 : Commit. 1.39937 PPL. 415.28 Recons. 0.01093 +2024-03-17 00:04:46,306 INFO Train. Iter 54200 : Commit. 1.35075 PPL. 416.57 Recons. 0.00974 +2024-03-17 00:05:07,629 INFO Train. Iter 54400 : Commit. 1.29904 PPL. 418.16 Recons. 0.00862 +2024-03-17 00:05:29,128 INFO Train. Iter 54600 : Commit. 1.36250 PPL. 416.15 Recons. 0.01057 +2024-03-17 00:05:50,542 INFO Train. Iter 54800 : Commit. 1.37215 PPL. 417.00 Recons. 0.01007 +2024-03-17 00:06:10,959 INFO Train. Iter 55000 : Commit. 1.36291 PPL. 417.14 Recons. 0.00971 +2024-03-17 00:06:32,034 INFO Train. Iter 55200 : Commit. 1.40686 PPL. 414.19 Recons. 0.01144 +2024-03-17 00:06:53,607 INFO Train. Iter 55400 : Commit. 1.37166 PPL. 414.81 Recons. 0.01050 +2024-03-17 00:07:14,792 INFO Train. Iter 55600 : Commit. 1.33636 PPL. 414.78 Recons. 0.00974 +2024-03-17 00:07:36,424 INFO Train. Iter 55800 : Commit. 1.34532 PPL. 416.58 Recons. 0.00974 +2024-03-17 00:07:58,071 INFO Train. Iter 56000 : Commit. 1.35597 PPL. 415.52 Recons. 0.01030 +2024-03-17 00:08:19,438 INFO Train. Iter 56200 : Commit. 1.34113 PPL. 415.12 Recons. 0.00980 +2024-03-17 00:08:40,541 INFO Train. Iter 56400 : Commit. 1.34760 PPL. 415.87 Recons. 0.00983 +2024-03-17 00:09:01,882 INFO Train. Iter 56600 : Commit. 1.34072 PPL. 416.10 Recons. 0.00994 +2024-03-17 00:09:22,817 INFO Train. Iter 56800 : Commit. 1.38867 PPL. 414.76 Recons. 0.01081 +2024-03-17 00:09:42,976 INFO Train. Iter 57000 : Commit. 1.31115 PPL. 415.93 Recons. 0.00919 +2024-03-17 00:10:04,716 INFO Train. Iter 57200 : Commit. 1.31150 PPL. 416.79 Recons. 0.00923 +2024-03-17 00:10:25,879 INFO Train. Iter 57400 : Commit. 1.24779 PPL. 418.27 Recons. 0.00788 +2024-03-17 00:10:47,420 INFO Train. Iter 57600 : Commit. 1.35954 PPL. 414.01 Recons. 0.01108 +2024-03-17 00:11:08,761 INFO Train. Iter 57800 : Commit. 1.32802 PPL. 416.28 Recons. 0.00927 +2024-03-17 00:11:30,328 INFO Train. Iter 58000 : Commit. 1.34735 PPL. 413.87 Recons. 0.01038 +2024-03-17 00:11:51,699 INFO Train. Iter 58200 : Commit. 1.33540 PPL. 415.06 Recons. 0.01020 +2024-03-17 00:12:12,981 INFO Train. Iter 58400 : Commit. 1.35822 PPL. 414.64 Recons. 0.00995 +2024-03-17 00:12:34,178 INFO Train. Iter 58600 : Commit. 1.37017 PPL. 413.90 Recons. 0.01092 +2024-03-17 00:12:56,047 INFO Train. Iter 58800 : Commit. 1.36030 PPL. 415.60 Recons. 0.00982 +2024-03-17 00:13:16,431 INFO Train. Iter 59000 : Commit. 1.39641 PPL. 412.95 Recons. 0.01091 +2024-03-17 00:13:37,684 INFO Train. Iter 59200 : Commit. 1.31021 PPL. 416.78 Recons. 0.00884 +2024-03-17 00:13:59,287 INFO Train. Iter 59400 : Commit. 1.35542 PPL. 413.07 Recons. 0.01058 +2024-03-17 00:14:20,874 INFO Train. Iter 59600 : Commit. 1.32606 PPL. 415.86 Recons. 0.00951 +2024-03-17 00:14:42,583 INFO Train. Iter 59800 : Commit. 1.33927 PPL. 416.20 Recons. 0.00951 +2024-03-17 00:15:03,749 INFO Train. Iter 60000 : Commit. 1.33273 PPL. 414.73 Recons. 0.00992 +2024-03-17 00:15:25,125 INFO Train. Iter 60200 : Commit. 1.37364 PPL. 414.69 Recons. 0.01023 +2024-03-17 00:15:46,680 INFO Train. Iter 60400 : Commit. 1.31600 PPL. 415.84 Recons. 0.00919 +2024-03-17 00:16:08,218 INFO Train. Iter 60600 : Commit. 1.42355 PPL. 410.41 Recons. 0.01199 +2024-03-17 00:16:28,868 INFO Train. Iter 60800 : Commit. 1.35148 PPL. 414.25 Recons. 0.00998 +2024-03-17 00:16:50,948 INFO Train. Iter 61000 : Commit. 1.30286 PPL. 416.20 Recons. 0.00896 +2024-03-17 00:17:12,675 INFO Train. Iter 61200 : Commit. 1.35555 PPL. 413.94 Recons. 0.01039 +2024-03-17 00:17:34,064 INFO Train. Iter 61400 : Commit. 1.30862 PPL. 415.58 Recons. 0.00945 +2024-03-17 00:17:55,655 INFO Train. Iter 61600 : Commit. 1.31806 PPL. 416.36 Recons. 0.00919 +2024-03-17 00:18:16,773 INFO Train. Iter 61800 : Commit. 1.36635 PPL. 414.07 Recons. 0.01052 +2024-03-17 00:18:38,058 INFO Train. Iter 62000 : Commit. 1.32212 PPL. 414.74 Recons. 0.00982 +2024-03-17 00:19:00,022 INFO Train. Iter 62200 : Commit. 1.36713 PPL. 413.62 Recons. 0.01085 +2024-03-17 00:19:22,227 INFO Train. Iter 62400 : Commit. 1.36427 PPL. 413.38 Recons. 0.01044 +2024-03-17 00:19:44,313 INFO Train. Iter 62600 : Commit. 1.34025 PPL. 414.87 Recons. 0.00978 +2024-03-17 00:20:05,041 INFO Train. Iter 62800 : Commit. 1.32799 PPL. 415.35 Recons. 0.00961 +2024-03-17 00:20:26,829 INFO Train. Iter 63000 : Commit. 1.38048 PPL. 412.59 Recons. 0.01092 +2024-03-17 00:20:48,776 INFO Train. Iter 63200 : Commit. 1.33729 PPL. 413.20 Recons. 0.01003 +2024-03-17 00:21:10,536 INFO Train. Iter 63400 : Commit. 1.32906 PPL. 413.48 Recons. 0.01001 +2024-03-17 00:21:32,375 INFO Train. Iter 63600 : Commit. 1.30570 PPL. 415.78 Recons. 0.00905 +2024-03-17 00:21:53,818 INFO Train. Iter 63800 : Commit. 1.32294 PPL. 413.69 Recons. 0.01003 +2024-03-17 00:22:15,294 INFO Train. Iter 64000 : Commit. 1.33539 PPL. 414.25 Recons. 0.01000 +2024-03-17 00:22:37,154 INFO Train. Iter 64200 : Commit. 1.28910 PPL. 415.00 Recons. 0.00914 +2024-03-17 00:22:59,387 INFO Train. Iter 64400 : Commit. 1.32421 PPL. 413.96 Recons. 0.00980 +2024-03-17 00:23:21,267 INFO Train. Iter 64600 : Commit. 1.33885 PPL. 413.89 Recons. 0.00977 +2024-03-17 00:23:41,652 INFO Train. Iter 64800 : Commit. 1.34884 PPL. 413.24 Recons. 0.01009 +2024-03-17 00:24:03,251 INFO Train. Iter 65000 : Commit. 1.29850 PPL. 415.19 Recons. 0.00894 +2024-03-17 00:24:24,738 INFO Train. Iter 65200 : Commit. 1.29903 PPL. 414.86 Recons. 0.00944 +2024-03-17 00:24:46,178 INFO Train. Iter 65400 : Commit. 1.30996 PPL. 414.64 Recons. 0.00941 +2024-03-17 00:25:07,006 INFO Train. Iter 65600 : Commit. 1.37749 PPL. 411.24 Recons. 0.01127 +2024-03-17 00:25:28,166 INFO Train. Iter 65800 : Commit. 1.25609 PPL. 416.40 Recons. 0.00806 +2024-03-17 00:25:49,782 INFO Train. Iter 66000 : Commit. 1.26881 PPL. 415.15 Recons. 0.00887 +2024-03-17 00:26:11,448 INFO Train. Iter 66200 : Commit. 1.27998 PPL. 414.73 Recons. 0.00916 +2024-03-17 00:26:33,108 INFO Train. Iter 66400 : Commit. 1.26093 PPL. 415.85 Recons. 0.00861 +2024-03-17 00:26:53,458 INFO Train. Iter 66600 : Commit. 1.32412 PPL. 412.71 Recons. 0.01008 +2024-03-17 00:27:15,134 INFO Train. Iter 66800 : Commit. 1.28798 PPL. 412.70 Recons. 0.00947 +2024-03-17 00:27:37,588 INFO Train. Iter 67000 : Commit. 1.28664 PPL. 415.11 Recons. 0.00874 +2024-03-17 00:27:59,222 INFO Train. Iter 67200 : Commit. 1.33547 PPL. 410.69 Recons. 0.01137 +2024-03-17 00:28:20,589 INFO Train. Iter 67400 : Commit. 1.32210 PPL. 414.44 Recons. 0.00924 +2024-03-17 00:28:42,003 INFO Train. Iter 67600 : Commit. 1.30854 PPL. 413.87 Recons. 0.00947 +2024-03-17 00:29:03,514 INFO Train. Iter 67800 : Commit. 1.33333 PPL. 414.11 Recons. 0.00980 +2024-03-17 00:29:24,722 INFO Train. Iter 68000 : Commit. 1.33060 PPL. 413.29 Recons. 0.01008 +2024-03-17 00:29:46,400 INFO Train. Iter 68200 : Commit. 1.30660 PPL. 413.24 Recons. 0.00922 +2024-03-17 00:30:08,248 INFO Train. Iter 68400 : Commit. 1.43893 PPL. 405.48 Recons. 0.01273 +2024-03-17 00:30:28,931 INFO Train. Iter 68600 : Commit. 1.37246 PPL. 412.70 Recons. 0.01017 +2024-03-17 00:30:51,026 INFO Train. Iter 68800 : Commit. 1.32514 PPL. 411.75 Recons. 0.01001 +2024-03-17 00:31:12,434 INFO Train. Iter 69000 : Commit. 1.28374 PPL. 415.06 Recons. 0.00860 +2024-03-17 00:31:33,722 INFO Train. Iter 69200 : Commit. 1.29762 PPL. 413.70 Recons. 0.00935 +2024-03-17 00:31:55,209 INFO Train. Iter 69400 : Commit. 1.34106 PPL. 412.66 Recons. 0.01042 +2024-03-17 00:32:16,587 INFO Train. Iter 69600 : Commit. 1.29727 PPL. 413.37 Recons. 0.00921 +2024-03-17 00:32:38,083 INFO Train. Iter 69800 : Commit. 1.26727 PPL. 413.19 Recons. 0.00905 +2024-03-17 00:32:59,566 INFO Train. Iter 70000 : Commit. 1.29859 PPL. 412.68 Recons. 0.00954 +2024-03-17 00:33:20,915 INFO Train. Iter 70200 : Commit. 1.25346 PPL. 415.78 Recons. 0.00836 +2024-03-17 00:33:42,365 INFO Train. Iter 70400 : Commit. 1.34295 PPL. 410.97 Recons. 0.01085 +2024-03-17 00:34:02,904 INFO Train. Iter 70600 : Commit. 1.34255 PPL. 412.41 Recons. 0.01036 +2024-03-17 00:34:23,931 INFO Train. Iter 70800 : Commit. 1.27177 PPL. 414.29 Recons. 0.00881 +2024-03-17 00:34:45,557 INFO Train. Iter 71000 : Commit. 1.29988 PPL. 413.23 Recons. 0.00977 +2024-03-17 00:35:06,805 INFO Train. Iter 71200 : Commit. 1.30377 PPL. 410.79 Recons. 0.01009 +2024-03-17 00:35:27,963 INFO Train. Iter 71400 : Commit. 1.32040 PPL. 412.65 Recons. 0.00973 +2024-03-17 00:35:50,005 INFO Train. Iter 71600 : Commit. 1.24772 PPL. 414.76 Recons. 0.00823 +2024-03-17 00:36:11,534 INFO Train. Iter 71800 : Commit. 1.30577 PPL. 413.36 Recons. 0.00944 +2024-03-17 00:36:33,044 INFO Train. Iter 72000 : Commit. 1.27657 PPL. 412.62 Recons. 0.00924 +2024-03-17 00:36:54,768 INFO Train. Iter 72200 : Commit. 1.27842 PPL. 411.24 Recons. 0.00969 +2024-03-17 00:37:15,640 INFO Train. Iter 72400 : Commit. 1.30297 PPL. 413.02 Recons. 0.00920 +2024-03-17 00:37:37,853 INFO Train. Iter 72600 : Commit. 1.23330 PPL. 414.69 Recons. 0.00808 +2024-03-17 00:38:00,370 INFO Train. Iter 72800 : Commit. 1.28739 PPL. 413.06 Recons. 0.00972 +2024-03-17 00:38:22,648 INFO Train. Iter 73000 : Commit. 1.28384 PPL. 412.55 Recons. 0.00906 +2024-03-17 00:38:45,142 INFO Train. Iter 73200 : Commit. 1.31256 PPL. 410.86 Recons. 0.00972 +2024-03-17 00:39:06,763 INFO Train. Iter 73400 : Commit. 1.27538 PPL. 412.38 Recons. 0.00930 +2024-03-17 00:39:28,096 INFO Train. Iter 73600 : Commit. 1.23469 PPL. 414.09 Recons. 0.00808 +2024-03-17 00:39:49,630 INFO Train. Iter 73800 : Commit. 1.26618 PPL. 413.17 Recons. 0.00899 +2024-03-17 00:40:10,987 INFO Train. Iter 74000 : Commit. 1.27054 PPL. 413.40 Recons. 0.00917 +2024-03-17 00:40:32,389 INFO Train. Iter 74200 : Commit. 1.23781 PPL. 414.71 Recons. 0.00806 +2024-03-17 00:40:52,490 INFO Train. Iter 74400 : Commit. 1.29658 PPL. 412.95 Recons. 0.00962 +2024-03-17 00:41:13,688 INFO Train. Iter 74600 : Commit. 1.30553 PPL. 413.26 Recons. 0.00940 +2024-03-17 00:41:34,939 INFO Train. Iter 74800 : Commit. 1.32193 PPL. 409.51 Recons. 0.01004 +2024-03-17 00:41:56,575 INFO Train. Iter 75000 : Commit. 1.29822 PPL. 413.20 Recons. 0.00922 +2024-03-17 00:42:17,941 INFO Train. Iter 75200 : Commit. 1.30213 PPL. 413.03 Recons. 0.00944 +2024-03-17 00:42:39,601 INFO Train. Iter 75400 : Commit. 1.30098 PPL. 412.12 Recons. 0.00952 +2024-03-17 00:43:01,122 INFO Train. Iter 75600 : Commit. 1.25951 PPL. 413.88 Recons. 0.00861 +2024-03-17 00:43:22,745 INFO Train. Iter 75800 : Commit. 1.28485 PPL. 411.79 Recons. 0.00968 +2024-03-17 00:43:43,959 INFO Train. Iter 76000 : Commit. 1.26693 PPL. 412.45 Recons. 0.00912 +2024-03-17 00:44:05,750 INFO Train. Iter 76200 : Commit. 1.30835 PPL. 411.13 Recons. 0.00958 +2024-03-17 00:44:25,866 INFO Train. Iter 76400 : Commit. 1.22612 PPL. 413.73 Recons. 0.00807 +2024-03-17 00:44:46,963 INFO Train. Iter 76600 : Commit. 1.28348 PPL. 412.45 Recons. 0.00916 +2024-03-17 00:45:08,169 INFO Train. Iter 76800 : Commit. 1.23788 PPL. 413.88 Recons. 0.00821 +2024-03-17 00:45:29,469 INFO Train. Iter 77000 : Commit. 1.30554 PPL. 410.60 Recons. 0.01024 +2024-03-17 00:45:50,803 INFO Train. Iter 77200 : Commit. 1.29612 PPL. 412.77 Recons. 0.00921 +2024-03-17 00:46:12,331 INFO Train. Iter 77400 : Commit. 1.29223 PPL. 411.87 Recons. 0.00939 +2024-03-17 00:46:33,483 INFO Train. Iter 77600 : Commit. 1.30327 PPL. 411.79 Recons. 0.00978 +2024-03-17 00:46:55,208 INFO Train. Iter 77800 : Commit. 1.26736 PPL. 410.80 Recons. 0.00904 +2024-03-17 00:47:16,263 INFO Train. Iter 78000 : Commit. 1.25607 PPL. 412.63 Recons. 0.00886 +2024-03-17 00:47:36,865 INFO Train. Iter 78200 : Commit. 1.25216 PPL. 412.70 Recons. 0.00858 +2024-03-17 00:47:58,035 INFO Train. Iter 78400 : Commit. 1.31624 PPL. 407.72 Recons. 0.01031 +2024-03-17 00:48:19,665 INFO Train. Iter 78600 : Commit. 1.24439 PPL. 413.57 Recons. 0.00811 +2024-03-17 00:48:40,939 INFO Train. Iter 78800 : Commit. 1.31178 PPL. 410.50 Recons. 0.01048 +2024-03-17 00:49:02,378 INFO Train. Iter 79000 : Commit. 1.26554 PPL. 413.27 Recons. 0.00859 +2024-03-17 00:49:24,014 INFO Train. Iter 79200 : Commit. 1.23640 PPL. 411.97 Recons. 0.00851 +2024-03-17 00:49:45,930 INFO Train. Iter 79400 : Commit. 1.34984 PPL. 407.68 Recons. 0.01055 +2024-03-17 00:50:08,314 INFO Train. Iter 79600 : Commit. 1.24554 PPL. 413.01 Recons. 0.00824 +2024-03-17 00:50:29,945 INFO Train. Iter 79800 : Commit. 1.30316 PPL. 409.04 Recons. 0.01020 +2024-03-17 00:50:51,185 INFO Train. Iter 80000 : Commit. 1.28930 PPL. 412.80 Recons. 0.00899 +2024-03-17 00:51:11,326 INFO Train. Iter 80200 : Commit. 1.19952 PPL. 413.87 Recons. 0.00760 +2024-03-17 00:51:32,648 INFO Train. Iter 80400 : Commit. 1.28861 PPL. 410.62 Recons. 0.00973 +2024-03-17 00:51:53,951 INFO Train. Iter 80600 : Commit. 1.27798 PPL. 410.39 Recons. 0.00940 +2024-03-17 00:52:15,070 INFO Train. Iter 80800 : Commit. 1.30364 PPL. 409.75 Recons. 0.00963 +2024-03-17 00:52:36,957 INFO Train. Iter 81000 : Commit. 1.21899 PPL. 412.07 Recons. 0.00841 +2024-03-17 00:52:58,945 INFO Train. Iter 81200 : Commit. 1.33557 PPL. 409.99 Recons. 0.01025 +2024-03-17 00:53:20,851 INFO Train. Iter 81400 : Commit. 1.25062 PPL. 412.84 Recons. 0.00842 +2024-03-17 00:53:42,521 INFO Train. Iter 81600 : Commit. 1.26224 PPL. 411.57 Recons. 0.00895 +2024-03-17 00:54:04,109 INFO Train. Iter 81800 : Commit. 1.29878 PPL. 407.80 Recons. 0.01059 +2024-03-17 00:54:25,655 INFO Train. Iter 82000 : Commit. 1.24190 PPL. 413.08 Recons. 0.00820 +2024-03-17 00:54:45,782 INFO Train. Iter 82200 : Commit. 1.23053 PPL. 413.05 Recons. 0.00819 +2024-03-17 00:55:07,248 INFO Train. Iter 82400 : Commit. 1.24839 PPL. 411.52 Recons. 0.00905 +2024-03-17 00:55:28,903 INFO Train. Iter 82600 : Commit. 1.30877 PPL. 408.71 Recons. 0.01045 +2024-03-17 00:55:50,383 INFO Train. Iter 82800 : Commit. 1.25642 PPL. 412.42 Recons. 0.00851 +2024-03-17 00:56:11,881 INFO Train. Iter 83000 : Commit. 1.25663 PPL. 412.04 Recons. 0.00868 +2024-03-17 00:56:33,312 INFO Train. Iter 83200 : Commit. 1.25547 PPL. 411.43 Recons. 0.00876 +2024-03-17 00:56:54,549 INFO Train. Iter 83400 : Commit. 1.26401 PPL. 410.99 Recons. 0.00903 +2024-03-17 00:57:15,949 INFO Train. Iter 83600 : Commit. 1.24324 PPL. 411.31 Recons. 0.00893 +2024-03-17 00:57:37,279 INFO Train. Iter 83800 : Commit. 1.23299 PPL. 412.20 Recons. 0.00830 +2024-03-17 00:57:57,798 INFO Train. Iter 84000 : Commit. 1.32002 PPL. 408.75 Recons. 0.01066 +2024-03-17 00:58:19,860 INFO Train. Iter 84200 : Commit. 1.27488 PPL. 412.75 Recons. 0.00875 +2024-03-17 00:58:41,833 INFO Train. Iter 84400 : Commit. 1.26395 PPL. 412.18 Recons. 0.00877 +2024-03-17 00:59:03,452 INFO Train. Iter 84600 : Commit. 1.27333 PPL. 409.88 Recons. 0.00977 +2024-03-17 00:59:25,099 INFO Train. Iter 84800 : Commit. 1.22994 PPL. 411.43 Recons. 0.00833 +2024-03-17 00:59:46,936 INFO Train. Iter 85000 : Commit. 1.27203 PPL. 411.07 Recons. 0.00923 +2024-03-17 01:00:08,301 INFO Train. Iter 85200 : Commit. 1.25960 PPL. 411.80 Recons. 0.00869 +2024-03-17 01:00:30,110 INFO Train. Iter 85400 : Commit. 1.31351 PPL. 408.11 Recons. 0.01048 +2024-03-17 01:00:51,129 INFO Train. Iter 85600 : Commit. 1.24322 PPL. 412.64 Recons. 0.00826 +2024-03-17 01:01:12,629 INFO Train. Iter 85800 : Commit. 1.22266 PPL. 411.83 Recons. 0.00822 +2024-03-17 01:01:32,867 INFO Train. Iter 86000 : Commit. 1.26440 PPL. 411.39 Recons. 0.00893 +2024-03-17 01:01:54,317 INFO Train. Iter 86200 : Commit. 1.26997 PPL. 409.25 Recons. 0.00932 +2024-03-17 01:02:15,798 INFO Train. Iter 86400 : Commit. 1.29364 PPL. 406.86 Recons. 0.01010 +2024-03-17 01:02:37,567 INFO Train. Iter 86600 : Commit. 1.29227 PPL. 412.03 Recons. 0.00877 +2024-03-17 01:02:59,214 INFO Train. Iter 86800 : Commit. 1.27444 PPL. 411.06 Recons. 0.00904 +2024-03-17 01:03:20,448 INFO Train. Iter 87000 : Commit. 1.21924 PPL. 411.72 Recons. 0.00812 +2024-03-17 01:03:41,619 INFO Train. Iter 87200 : Commit. 1.24682 PPL. 409.64 Recons. 0.00896 +2024-03-17 01:04:03,349 INFO Train. Iter 87400 : Commit. 1.27301 PPL. 409.43 Recons. 0.00941 +2024-03-17 01:04:24,629 INFO Train. Iter 87600 : Commit. 1.21840 PPL. 412.57 Recons. 0.00784 +2024-03-17 01:04:45,819 INFO Train. Iter 87800 : Commit. 1.24680 PPL. 410.39 Recons. 0.00908 +2024-03-17 01:05:06,050 INFO Train. Iter 88000 : Commit. 1.23237 PPL. 411.29 Recons. 0.00838 +2024-03-17 01:05:27,413 INFO Train. Iter 88200 : Commit. 1.21855 PPL. 411.68 Recons. 0.00810 +2024-03-17 01:05:49,091 INFO Train. Iter 88400 : Commit. 1.24761 PPL. 409.79 Recons. 0.00897 +2024-03-17 01:06:09,990 INFO Train. Iter 88600 : Commit. 1.17839 PPL. 412.48 Recons. 0.00746 +2024-03-17 01:06:31,022 INFO Train. Iter 88800 : Commit. 1.21198 PPL. 411.50 Recons. 0.00803 +2024-03-17 01:06:52,126 INFO Train. Iter 89000 : Commit. 1.27911 PPL. 407.60 Recons. 0.00982 +2024-03-17 01:07:13,393 INFO Train. Iter 89200 : Commit. 1.28546 PPL. 411.24 Recons. 0.00929 +2024-03-17 01:07:35,090 INFO Train. Iter 89400 : Commit. 1.24974 PPL. 411.38 Recons. 0.00872 +2024-03-17 01:07:56,848 INFO Train. Iter 89600 : Commit. 1.23072 PPL. 411.68 Recons. 0.00837 +2024-03-17 01:08:16,959 INFO Train. Iter 89800 : Commit. 1.23120 PPL. 411.01 Recons. 0.00846 +2024-03-17 01:08:38,795 INFO Train. Iter 90000 : Commit. 1.21923 PPL. 411.98 Recons. 0.00811 +2024-03-17 01:09:00,608 INFO Train. Iter 90200 : Commit. 1.26127 PPL. 410.07 Recons. 0.00928 +2024-03-17 01:09:21,745 INFO Train. Iter 90400 : Commit. 1.24258 PPL. 412.07 Recons. 0.00818 +2024-03-17 01:09:42,883 INFO Train. Iter 90600 : Commit. 1.27397 PPL. 409.60 Recons. 0.00945 +2024-03-17 01:10:04,601 INFO Train. Iter 90800 : Commit. 1.21620 PPL. 411.02 Recons. 0.00810 +2024-03-17 01:10:26,312 INFO Train. Iter 91000 : Commit. 1.29022 PPL. 408.08 Recons. 0.00980 +2024-03-17 01:10:47,973 INFO Train. Iter 91200 : Commit. 1.20254 PPL. 412.86 Recons. 0.00750 +2024-03-17 01:11:09,410 INFO Train. Iter 91400 : Commit. 1.27854 PPL. 408.68 Recons. 0.00947 +2024-03-17 01:11:30,599 INFO Train. Iter 91600 : Commit. 1.23742 PPL. 411.41 Recons. 0.00851 +2024-03-17 01:11:50,939 INFO Train. Iter 91800 : Commit. 1.27511 PPL. 408.86 Recons. 0.00976 +2024-03-17 01:12:12,195 INFO Train. Iter 92000 : Commit. 1.27128 PPL. 408.73 Recons. 0.00968 +2024-03-17 01:12:33,403 INFO Train. Iter 92200 : Commit. 1.25255 PPL. 410.19 Recons. 0.00864 +2024-03-17 01:12:54,877 INFO Train. Iter 92400 : Commit. 1.25314 PPL. 409.68 Recons. 0.00938 +2024-03-17 01:13:16,183 INFO Train. Iter 92600 : Commit. 1.22086 PPL. 410.96 Recons. 0.00831 +2024-03-17 01:13:37,344 INFO Train. Iter 92800 : Commit. 1.24814 PPL. 408.73 Recons. 0.00924 +2024-03-17 01:13:58,661 INFO Train. Iter 93000 : Commit. 1.21724 PPL. 411.07 Recons. 0.00799 +2024-03-17 01:14:20,762 INFO Train. Iter 93200 : Commit. 1.28138 PPL. 409.37 Recons. 0.00968 +2024-03-17 01:14:42,090 INFO Train. Iter 93400 : Commit. 1.20417 PPL. 412.46 Recons. 0.00772 +2024-03-17 01:15:04,074 INFO Train. Iter 93600 : Commit. 1.22001 PPL. 409.92 Recons. 0.00841 +2024-03-17 01:15:24,448 INFO Train. Iter 93800 : Commit. 1.22957 PPL. 410.27 Recons. 0.00883 +2024-03-17 01:15:45,605 INFO Train. Iter 94000 : Commit. 1.23950 PPL. 407.17 Recons. 0.00933 +2024-03-17 01:16:06,966 INFO Train. Iter 94200 : Commit. 1.28848 PPL. 409.79 Recons. 0.00940 +2024-03-17 01:16:28,596 INFO Train. Iter 94400 : Commit. 1.21574 PPL. 412.12 Recons. 0.00794 +2024-03-17 01:16:50,009 INFO Train. Iter 94600 : Commit. 1.16846 PPL. 412.00 Recons. 0.00745 +2024-03-17 01:17:11,277 INFO Train. Iter 94800 : Commit. 1.24761 PPL. 409.75 Recons. 0.00907 +2024-03-17 01:17:32,579 INFO Train. Iter 95000 : Commit. 1.18798 PPL. 411.47 Recons. 0.00761 +2024-03-17 01:17:54,152 INFO Train. Iter 95200 : Commit. 1.20944 PPL. 410.54 Recons. 0.00828 +2024-03-17 01:18:15,550 INFO Train. Iter 95400 : Commit. 1.22212 PPL. 410.10 Recons. 0.00858 +2024-03-17 01:18:36,152 INFO Train. Iter 95600 : Commit. 1.20734 PPL. 410.82 Recons. 0.00798 +2024-03-17 01:18:57,278 INFO Train. Iter 95800 : Commit. 1.25221 PPL. 407.91 Recons. 0.00928 +2024-03-17 01:19:18,584 INFO Train. Iter 96000 : Commit. 1.28580 PPL. 408.14 Recons. 0.00955 +2024-03-17 01:19:39,752 INFO Train. Iter 96200 : Commit. 1.22134 PPL. 409.96 Recons. 0.00823 +2024-03-17 01:20:01,572 INFO Train. Iter 96400 : Commit. 1.23751 PPL. 409.04 Recons. 0.00888 +2024-03-17 01:20:23,183 INFO Train. Iter 96600 : Commit. 1.23993 PPL. 408.25 Recons. 0.00893 +2024-03-17 01:20:45,154 INFO Train. Iter 96800 : Commit. 1.23345 PPL. 409.46 Recons. 0.00871 +2024-03-17 01:21:07,032 INFO Train. Iter 97000 : Commit. 1.23819 PPL. 409.92 Recons. 0.00870 +2024-03-17 01:21:28,611 INFO Train. Iter 97200 : Commit. 1.23703 PPL. 410.45 Recons. 0.00848 +2024-03-17 01:21:49,957 INFO Train. Iter 97400 : Commit. 1.23289 PPL. 409.77 Recons. 0.00872 +2024-03-17 01:22:10,260 INFO Train. Iter 97600 : Commit. 1.25666 PPL. 407.47 Recons. 0.00920 +2024-03-17 01:22:31,353 INFO Train. Iter 97800 : Commit. 1.22703 PPL. 410.17 Recons. 0.00825 +2024-03-17 01:22:52,619 INFO Train. Iter 98000 : Commit. 1.23082 PPL. 410.12 Recons. 0.00856 +2024-03-17 01:23:14,270 INFO Train. Iter 98200 : Commit. 1.20450 PPL. 411.65 Recons. 0.00774 +2024-03-17 01:23:35,963 INFO Train. Iter 98400 : Commit. 1.23723 PPL. 409.14 Recons. 0.00893 +2024-03-17 01:23:58,051 INFO Train. Iter 98600 : Commit. 1.20712 PPL. 411.35 Recons. 0.00783 +2024-03-17 01:24:19,795 INFO Train. Iter 98800 : Commit. 1.28011 PPL. 407.57 Recons. 0.00966 +2024-03-17 01:24:41,729 INFO Train. Iter 99000 : Commit. 1.22969 PPL. 409.63 Recons. 0.00849 +2024-03-17 01:25:03,595 INFO Train. Iter 99200 : Commit. 1.24134 PPL. 407.70 Recons. 0.00906 +2024-03-17 01:25:25,237 INFO Train. Iter 99400 : Commit. 1.20880 PPL. 410.92 Recons. 0.00788 +2024-03-17 01:25:46,083 INFO Train. Iter 99600 : Commit. 1.19367 PPL. 410.45 Recons. 0.00796 +2024-03-17 01:26:07,394 INFO Train. Iter 99800 : Commit. 1.22918 PPL. 409.84 Recons. 0.00840 +2024-03-17 01:26:28,736 INFO Train. Iter 100000 : Commit. 1.23486 PPL. 409.75 Recons. 0.00869 +2024-03-17 01:26:50,327 INFO Train. Iter 100200 : Commit. 1.26989 PPL. 406.39 Recons. 0.00974 +2024-03-17 01:27:11,920 INFO Train. Iter 100400 : Commit. 1.19571 PPL. 409.98 Recons. 0.00773 +2024-03-17 01:27:33,167 INFO Train. Iter 100600 : Commit. 1.22839 PPL. 408.01 Recons. 0.00943 +2024-03-17 01:27:54,666 INFO Train. Iter 100800 : Commit. 1.27299 PPL. 409.06 Recons. 0.00935 +2024-03-17 01:28:16,677 INFO Train. Iter 101000 : Commit. 1.21770 PPL. 409.80 Recons. 0.00833 +2024-03-17 01:28:38,543 INFO Train. Iter 101200 : Commit. 1.21954 PPL. 409.75 Recons. 0.00843 +2024-03-17 01:28:58,803 INFO Train. Iter 101400 : Commit. 1.27773 PPL. 407.86 Recons. 0.00946 +2024-03-17 01:29:20,525 INFO Train. Iter 101600 : Commit. 1.24680 PPL. 408.87 Recons. 0.00892 +2024-03-17 01:29:41,840 INFO Train. Iter 101800 : Commit. 1.22823 PPL. 409.01 Recons. 0.00891 +2024-03-17 01:30:04,047 INFO Train. Iter 102000 : Commit. 1.23574 PPL. 409.81 Recons. 0.00850 +2024-03-17 01:30:25,741 INFO Train. Iter 102200 : Commit. 1.25445 PPL. 408.35 Recons. 0.00942 +2024-03-17 01:30:47,330 INFO Train. Iter 102400 : Commit. 1.26473 PPL. 407.13 Recons. 0.00927 +2024-03-17 01:31:09,101 INFO Train. Iter 102600 : Commit. 1.21220 PPL. 410.47 Recons. 0.00796 +2024-03-17 01:31:30,534 INFO Train. Iter 102800 : Commit. 1.24973 PPL. 408.31 Recons. 0.00897 +2024-03-17 01:31:52,343 INFO Train. Iter 103000 : Commit. 1.21889 PPL. 409.05 Recons. 0.00845 +2024-03-17 01:32:14,111 INFO Train. Iter 103200 : Commit. 1.22014 PPL. 409.07 Recons. 0.00849 +2024-03-17 01:32:34,724 INFO Train. Iter 103400 : Commit. 1.25560 PPL. 409.06 Recons. 0.00887 +2024-03-17 01:32:55,821 INFO Train. Iter 103600 : Commit. 1.21224 PPL. 409.18 Recons. 0.00820 +2024-03-17 01:33:17,182 INFO Train. Iter 103800 : Commit. 1.21347 PPL. 409.86 Recons. 0.00827 +2024-03-17 01:33:38,885 INFO Train. Iter 104000 : Commit. 1.23318 PPL. 408.42 Recons. 0.00867 +2024-03-17 01:34:00,396 INFO Train. Iter 104200 : Commit. 1.27245 PPL. 405.74 Recons. 0.01007 +2024-03-17 01:34:21,982 INFO Train. Iter 104400 : Commit. 1.20998 PPL. 410.65 Recons. 0.00777 +2024-03-17 01:34:43,650 INFO Train. Iter 104600 : Commit. 1.29674 PPL. 405.30 Recons. 0.01070 +2024-03-17 01:35:05,852 INFO Train. Iter 104800 : Commit. 1.21656 PPL. 410.18 Recons. 0.00775 +2024-03-17 01:35:27,773 INFO Train. Iter 105000 : Commit. 1.20124 PPL. 410.61 Recons. 0.00776 +2024-03-17 01:35:49,210 INFO Train. Iter 105200 : Commit. 1.23931 PPL. 408.14 Recons. 0.00920 +2024-03-17 01:36:09,531 INFO Train. Iter 105400 : Commit. 1.25831 PPL. 407.85 Recons. 0.00919 +2024-03-17 01:36:30,762 INFO Train. Iter 105600 : Commit. 1.19731 PPL. 410.79 Recons. 0.00764 +2024-03-17 01:36:52,113 INFO Train. Iter 105800 : Commit. 1.23186 PPL. 407.16 Recons. 0.00913 +2024-03-17 01:37:13,888 INFO Train. Iter 106000 : Commit. 1.24717 PPL. 408.70 Recons. 0.00868 +2024-03-17 01:37:35,031 INFO Train. Iter 106200 : Commit. 1.21501 PPL. 408.56 Recons. 0.00820 +2024-03-17 01:37:56,391 INFO Train. Iter 106400 : Commit. 1.21855 PPL. 410.74 Recons. 0.00808 +2024-03-17 01:38:17,895 INFO Train. Iter 106600 : Commit. 1.22616 PPL. 407.85 Recons. 0.00884 +2024-03-17 01:38:39,301 INFO Train. Iter 106800 : Commit. 1.26438 PPL. 406.72 Recons. 0.00944 +2024-03-17 01:39:01,045 INFO Train. Iter 107000 : Commit. 1.21329 PPL. 409.01 Recons. 0.00824 +2024-03-17 01:39:21,370 INFO Train. Iter 107200 : Commit. 1.19528 PPL. 410.12 Recons. 0.00762 +2024-03-17 01:39:42,972 INFO Train. Iter 107400 : Commit. 1.22160 PPL. 408.41 Recons. 0.00860 +2024-03-17 01:40:04,589 INFO Train. Iter 107600 : Commit. 1.17818 PPL. 410.56 Recons. 0.00751 +2024-03-17 01:40:26,033 INFO Train. Iter 107800 : Commit. 1.20412 PPL. 409.32 Recons. 0.00818 +2024-03-17 01:40:48,125 INFO Train. Iter 108000 : Commit. 1.23111 PPL. 407.97 Recons. 0.00886 +2024-03-17 01:41:09,764 INFO Train. Iter 108200 : Commit. 1.19832 PPL. 410.75 Recons. 0.00765 +2024-03-17 01:41:31,264 INFO Train. Iter 108400 : Commit. 1.24146 PPL. 406.98 Recons. 0.00907 +2024-03-17 01:41:53,034 INFO Train. Iter 108600 : Commit. 1.20407 PPL. 407.56 Recons. 0.00852 +2024-03-17 01:42:15,189 INFO Train. Iter 108800 : Commit. 1.25162 PPL. 410.81 Recons. 0.00840 +2024-03-17 01:42:36,752 INFO Train. Iter 109000 : Commit. 1.25435 PPL. 407.76 Recons. 0.00895 +2024-03-17 01:42:57,321 INFO Train. Iter 109200 : Commit. 1.19444 PPL. 410.40 Recons. 0.00762 +2024-03-17 01:43:19,050 INFO Train. Iter 109400 : Commit. 1.33017 PPL. 402.28 Recons. 0.01110 +2024-03-17 01:43:40,531 INFO Train. Iter 109600 : Commit. 1.23644 PPL. 407.71 Recons. 0.00863 +2024-03-17 01:44:02,015 INFO Train. Iter 109800 : Commit. 1.21571 PPL. 409.38 Recons. 0.00806 +2024-03-17 01:44:23,833 INFO Train. Iter 110000 : Commit. 1.23478 PPL. 408.93 Recons. 0.00856 +2024-03-17 01:44:45,265 INFO Train. Iter 110200 : Commit. 1.23141 PPL. 408.00 Recons. 0.00863 +2024-03-17 01:45:06,983 INFO Train. Iter 110400 : Commit. 1.19239 PPL. 409.77 Recons. 0.00772 +2024-03-17 01:45:28,503 INFO Train. Iter 110600 : Commit. 1.19037 PPL. 409.04 Recons. 0.00788 +2024-03-17 01:45:50,001 INFO Train. Iter 110800 : Commit. 1.24194 PPL. 407.00 Recons. 0.00912 +2024-03-17 01:46:11,757 INFO Train. Iter 111000 : Commit. 1.18191 PPL. 410.64 Recons. 0.00738 +2024-03-17 01:46:32,193 INFO Train. Iter 111200 : Commit. 1.21692 PPL. 408.31 Recons. 0.00854 +2024-03-17 01:46:53,849 INFO Train. Iter 111400 : Commit. 1.19442 PPL. 409.07 Recons. 0.00803 +2024-03-17 01:47:15,724 INFO Train. Iter 111600 : Commit. 1.26290 PPL. 405.37 Recons. 0.00964 +2024-03-17 01:47:36,849 INFO Train. Iter 111800 : Commit. 1.17679 PPL. 410.78 Recons. 0.00723 +2024-03-17 01:47:58,113 INFO Train. Iter 112000 : Commit. 1.25741 PPL. 407.37 Recons. 0.00924 +2024-03-17 01:48:20,055 INFO Train. Iter 112200 : Commit. 1.24985 PPL. 405.50 Recons. 0.00950 +2024-03-17 01:48:41,760 INFO Train. Iter 112400 : Commit. 1.20571 PPL. 410.56 Recons. 0.00753 +2024-03-17 01:49:03,347 INFO Train. Iter 112600 : Commit. 1.17939 PPL. 408.69 Recons. 0.00769 +2024-03-17 01:49:24,764 INFO Train. Iter 112800 : Commit. 1.21281 PPL. 408.41 Recons. 0.00853 +2024-03-17 01:49:44,989 INFO Train. Iter 113000 : Commit. 1.19835 PPL. 409.88 Recons. 0.00793 +2024-03-17 01:50:06,545 INFO Train. Iter 113200 : Commit. 1.19155 PPL. 409.43 Recons. 0.00792 +2024-03-17 01:50:27,854 INFO Train. Iter 113400 : Commit. 1.23455 PPL. 406.16 Recons. 0.00903 +2024-03-17 01:50:49,463 INFO Train. Iter 113600 : Commit. 1.26130 PPL. 405.77 Recons. 0.00940 +2024-03-17 01:51:10,725 INFO Train. Iter 113800 : Commit. 1.23516 PPL. 408.06 Recons. 0.00858 +2024-03-17 01:51:32,457 INFO Train. Iter 114000 : Commit. 1.18742 PPL. 409.47 Recons. 0.00779 +2024-03-17 01:51:54,590 INFO Train. Iter 114200 : Commit. 1.24547 PPL. 407.42 Recons. 0.00874 +2024-03-17 01:52:16,684 INFO Train. Iter 114400 : Commit. 1.21944 PPL. 406.97 Recons. 0.00886 +2024-03-17 01:52:39,021 INFO Train. Iter 114600 : Commit. 1.20549 PPL. 408.50 Recons. 0.00822 +2024-03-17 01:53:00,444 INFO Train. Iter 114800 : Commit. 1.21860 PPL. 407.72 Recons. 0.00856 +2024-03-17 01:53:20,515 INFO Train. Iter 115000 : Commit. 1.16678 PPL. 410.08 Recons. 0.00735 +2024-03-17 01:53:41,509 INFO Train. Iter 115200 : Commit. 1.21655 PPL. 407.93 Recons. 0.00857 +2024-03-17 01:54:03,492 INFO Train. Iter 115400 : Commit. 1.18253 PPL. 408.99 Recons. 0.00773 +2024-03-17 01:54:25,321 INFO Train. Iter 115600 : Commit. 1.20255 PPL. 408.25 Recons. 0.00811 +2024-03-17 01:54:46,555 INFO Train. Iter 115800 : Commit. 1.17614 PPL. 408.11 Recons. 0.00780 +2024-03-17 01:55:07,834 INFO Train. Iter 116000 : Commit. 1.19288 PPL. 409.67 Recons. 0.00784 +2024-03-17 01:55:28,820 INFO Train. Iter 116200 : Commit. 1.21953 PPL. 407.79 Recons. 0.00849 +2024-03-17 01:55:50,693 INFO Train. Iter 116400 : Commit. 1.13900 PPL. 409.64 Recons. 0.00685 +2024-03-17 01:56:12,275 INFO Train. Iter 116600 : Commit. 1.17668 PPL. 408.41 Recons. 0.00779 +2024-03-17 01:56:34,368 INFO Train. Iter 116800 : Commit. 1.22742 PPL. 407.44 Recons. 0.00881 +2024-03-17 01:56:54,723 INFO Train. Iter 117000 : Commit. 1.19298 PPL. 406.96 Recons. 0.00832 +2024-03-17 01:57:16,655 INFO Train. Iter 117200 : Commit. 1.23170 PPL. 407.99 Recons. 0.00865 +2024-03-17 01:57:37,795 INFO Train. Iter 117400 : Commit. 1.15247 PPL. 409.64 Recons. 0.00706 +2024-03-17 01:57:58,915 INFO Train. Iter 117600 : Commit. 1.20784 PPL. 407.97 Recons. 0.00833 +2024-03-17 01:58:19,958 INFO Train. Iter 117800 : Commit. 1.22947 PPL. 407.04 Recons. 0.00881 +2024-03-17 01:58:40,968 INFO Train. Iter 118000 : Commit. 1.18839 PPL. 409.72 Recons. 0.00760 +2024-03-17 01:59:02,591 INFO Train. Iter 118200 : Commit. 1.18887 PPL. 409.30 Recons. 0.00779 +2024-03-17 01:59:24,031 INFO Train. Iter 118400 : Commit. 1.22425 PPL. 405.57 Recons. 0.00885 +2024-03-17 01:59:45,217 INFO Train. Iter 118600 : Commit. 1.23972 PPL. 407.85 Recons. 0.00862 +2024-03-17 02:00:05,637 INFO Train. Iter 118800 : Commit. 1.17002 PPL. 410.20 Recons. 0.00725 +2024-03-17 02:00:27,261 INFO Train. Iter 119000 : Commit. 1.16692 PPL. 409.36 Recons. 0.00743 +2024-03-17 02:00:49,129 INFO Train. Iter 119200 : Commit. 1.17687 PPL. 407.63 Recons. 0.00801 +2024-03-17 02:01:11,213 INFO Train. Iter 119400 : Commit. 1.22376 PPL. 406.61 Recons. 0.00885 +2024-03-17 02:01:32,850 INFO Train. Iter 119600 : Commit. 1.18474 PPL. 409.33 Recons. 0.00760 +2024-03-17 02:01:54,308 INFO Train. Iter 119800 : Commit. 1.20079 PPL. 408.86 Recons. 0.00812 +2024-03-17 02:02:15,846 INFO Train. Iter 120000 : Commit. 1.24026 PPL. 407.24 Recons. 0.00906 +2024-03-17 02:02:37,511 INFO Train. Iter 120200 : Commit. 1.21733 PPL. 408.39 Recons. 0.00842 +2024-03-17 02:02:58,565 INFO Train. Iter 120400 : Commit. 1.21223 PPL. 408.50 Recons. 0.00830 +2024-03-17 02:03:19,943 INFO Train. Iter 120600 : Commit. 1.24388 PPL. 405.40 Recons. 0.00920 +2024-03-17 02:03:40,290 INFO Train. Iter 120800 : Commit. 1.21071 PPL. 407.56 Recons. 0.00832 +2024-03-17 02:04:01,543 INFO Train. Iter 121000 : Commit. 1.20014 PPL. 407.59 Recons. 0.00820 +2024-03-17 02:04:22,555 INFO Train. Iter 121200 : Commit. 1.16417 PPL. 408.92 Recons. 0.00738 +2024-03-17 02:04:43,650 INFO Train. Iter 121400 : Commit. 1.22724 PPL. 408.16 Recons. 0.00874 +2024-03-17 02:05:04,925 INFO Train. Iter 121600 : Commit. 1.23927 PPL. 406.85 Recons. 0.00895 +2024-03-17 02:05:26,159 INFO Train. Iter 121800 : Commit. 1.18921 PPL. 409.69 Recons. 0.00758 +2024-03-17 02:05:47,775 INFO Train. Iter 122000 : Commit. 1.18905 PPL. 408.90 Recons. 0.00784 +2024-03-17 02:06:09,483 INFO Train. Iter 122200 : Commit. 1.14561 PPL. 409.74 Recons. 0.00709 +2024-03-17 02:06:31,360 INFO Train. Iter 122400 : Commit. 1.23805 PPL. 404.70 Recons. 0.00938 +2024-03-17 02:06:52,849 INFO Train. Iter 122600 : Commit. 1.13773 PPL. 409.67 Recons. 0.00682 +2024-03-17 02:07:13,390 INFO Train. Iter 122800 : Commit. 1.21586 PPL. 405.84 Recons. 0.00878 +2024-03-17 02:07:34,653 INFO Train. Iter 123000 : Commit. 1.22110 PPL. 405.46 Recons. 0.00875 +2024-03-17 02:07:56,467 INFO Train. Iter 123200 : Commit. 1.18883 PPL. 408.42 Recons. 0.00765 +2024-03-17 02:08:17,618 INFO Train. Iter 123400 : Commit. 1.21795 PPL. 407.03 Recons. 0.00842 +2024-03-17 02:08:39,151 INFO Train. Iter 123600 : Commit. 1.20553 PPL. 407.03 Recons. 0.00827 +2024-03-17 02:09:00,542 INFO Train. Iter 123800 : Commit. 1.15570 PPL. 409.04 Recons. 0.00721 +2024-03-17 02:09:21,973 INFO Train. Iter 124000 : Commit. 1.24509 PPL. 405.80 Recons. 0.00919 +2024-03-17 02:09:43,272 INFO Train. Iter 124200 : Commit. 1.22834 PPL. 406.41 Recons. 0.00862 +2024-03-17 02:10:04,534 INFO Train. Iter 124400 : Commit. 1.19036 PPL. 406.81 Recons. 0.00828 +2024-03-17 02:10:25,553 INFO Train. Iter 124600 : Commit. 1.21886 PPL. 407.42 Recons. 0.00847 +2024-03-17 02:10:47,447 INFO Train. Iter 124800 : Commit. 1.19302 PPL. 407.62 Recons. 0.00793 +2024-03-17 02:11:09,143 INFO Train. Iter 125000 : Commit. 1.19253 PPL. 406.53 Recons. 0.00830 +2024-03-17 02:11:31,253 INFO Train. Iter 125200 : Commit. 1.16945 PPL. 407.51 Recons. 0.00740 +2024-03-17 02:11:53,293 INFO Train. Iter 125400 : Commit. 1.20457 PPL. 405.12 Recons. 0.00850 +2024-03-17 02:12:14,839 INFO Train. Iter 125600 : Commit. 1.24982 PPL. 407.14 Recons. 0.00905 +2024-03-17 02:12:36,761 INFO Train. Iter 125800 : Commit. 1.20068 PPL. 406.12 Recons. 0.00808 +2024-03-17 02:12:58,900 INFO Train. Iter 126000 : Commit. 1.21935 PPL. 405.04 Recons. 0.00856 +2024-03-17 02:13:20,480 INFO Train. Iter 126200 : Commit. 1.20731 PPL. 408.66 Recons. 0.00743 +2024-03-17 02:13:42,333 INFO Train. Iter 126400 : Commit. 1.17896 PPL. 407.63 Recons. 0.00789 +2024-03-17 02:14:02,923 INFO Train. Iter 126600 : Commit. 1.19954 PPL. 407.12 Recons. 0.00813 +2024-03-17 02:14:24,900 INFO Train. Iter 126800 : Commit. 1.19693 PPL. 407.83 Recons. 0.00793 +2024-03-17 02:14:46,408 INFO Train. Iter 127000 : Commit. 1.20381 PPL. 405.10 Recons. 0.00849 +2024-03-17 02:15:08,701 INFO Train. Iter 127200 : Commit. 1.25022 PPL. 405.78 Recons. 0.00926 +2024-03-17 02:15:30,674 INFO Train. Iter 127400 : Commit. 1.21342 PPL. 406.18 Recons. 0.00848 +2024-03-17 02:15:52,292 INFO Train. Iter 127600 : Commit. 1.21322 PPL. 407.06 Recons. 0.00831 +2024-03-17 02:16:14,001 INFO Train. Iter 127800 : Commit. 1.17947 PPL. 407.16 Recons. 0.00780 +2024-03-17 02:16:35,569 INFO Train. Iter 128000 : Commit. 1.14097 PPL. 409.74 Recons. 0.00689 +2024-03-17 02:16:56,918 INFO Train. Iter 128200 : Commit. 1.17258 PPL. 407.27 Recons. 0.00789 +2024-03-17 02:17:17,926 INFO Train. Iter 128400 : Commit. 1.18583 PPL. 407.79 Recons. 0.00774 +2024-03-17 02:17:37,963 INFO Train. Iter 128600 : Commit. 1.21556 PPL. 405.07 Recons. 0.00853 +2024-03-17 02:17:59,243 INFO Train. Iter 128800 : Commit. 1.15384 PPL. 406.81 Recons. 0.00738 +2024-03-17 02:18:21,432 INFO Train. Iter 129000 : Commit. 1.20196 PPL. 407.50 Recons. 0.00791 +2024-03-17 02:18:44,025 INFO Train. Iter 129200 : Commit. 1.15323 PPL. 408.23 Recons. 0.00732 +2024-03-17 02:19:06,105 INFO Train. Iter 129400 : Commit. 1.23009 PPL. 405.21 Recons. 0.00913 +2024-03-17 02:19:28,503 INFO Train. Iter 129600 : Commit. 1.12846 PPL. 409.10 Recons. 0.00651 +2024-03-17 02:19:51,031 INFO Train. Iter 129800 : Commit. 1.17186 PPL. 406.93 Recons. 0.00793 +2024-03-17 02:20:13,302 INFO Train. Iter 130000 : Commit. 1.18868 PPL. 407.53 Recons. 0.00788 +2024-03-17 02:20:35,679 INFO Train. Iter 130200 : Commit. 1.20331 PPL. 406.93 Recons. 0.00815 +2024-03-17 02:20:57,191 INFO Train. Iter 130400 : Commit. 1.15481 PPL. 408.77 Recons. 0.00721 +2024-03-17 02:21:19,885 INFO Train. Iter 130600 : Commit. 1.17622 PPL. 406.46 Recons. 0.00784 +2024-03-17 02:21:42,733 INFO Train. Iter 130800 : Commit. 1.20314 PPL. 406.03 Recons. 0.00833 +2024-03-17 02:22:05,328 INFO Train. Iter 131000 : Commit. 1.17382 PPL. 408.30 Recons. 0.00748 +2024-03-17 02:22:28,155 INFO Train. Iter 131200 : Commit. 1.23462 PPL. 406.66 Recons. 0.00873 +2024-03-17 02:22:51,241 INFO Train. Iter 131400 : Commit. 1.17404 PPL. 407.41 Recons. 0.00771 +2024-03-17 02:23:14,020 INFO Train. Iter 131600 : Commit. 1.24281 PPL. 405.26 Recons. 0.00892 +2024-03-17 02:23:36,636 INFO Train. Iter 131800 : Commit. 1.16038 PPL. 407.56 Recons. 0.00739 +2024-03-17 02:23:59,097 INFO Train. Iter 132000 : Commit. 1.20049 PPL. 406.71 Recons. 0.00839 +2024-03-17 02:24:22,312 INFO Train. Iter 132200 : Commit. 1.14258 PPL. 408.08 Recons. 0.00718 +2024-03-17 02:24:43,561 INFO Train. Iter 132400 : Commit. 1.16795 PPL. 407.66 Recons. 0.00769 +2024-03-17 02:25:05,306 INFO Train. Iter 132600 : Commit. 1.18619 PPL. 408.02 Recons. 0.00786 +2024-03-17 02:25:27,809 INFO Train. Iter 132800 : Commit. 1.24519 PPL. 400.77 Recons. 0.00966 +2024-03-17 02:25:50,565 INFO Train. Iter 133000 : Commit. 1.18219 PPL. 407.66 Recons. 0.00748 +2024-03-17 02:26:13,597 INFO Train. Iter 133200 : Commit. 1.16253 PPL. 408.66 Recons. 0.00722 +2024-03-17 02:26:36,407 INFO Train. Iter 133400 : Commit. 1.18209 PPL. 407.04 Recons. 0.00787 +2024-03-17 02:26:59,265 INFO Train. Iter 133600 : Commit. 1.19695 PPL. 404.77 Recons. 0.00838 +2024-03-17 02:27:21,803 INFO Train. Iter 133800 : Commit. 1.22379 PPL. 405.41 Recons. 0.00884 +2024-03-17 02:27:44,622 INFO Train. Iter 134000 : Commit. 1.20780 PPL. 406.15 Recons. 0.00821 +2024-03-17 02:28:07,196 INFO Train. Iter 134200 : Commit. 1.16448 PPL. 407.78 Recons. 0.00735 +2024-03-17 02:28:28,420 INFO Train. Iter 134400 : Commit. 1.21234 PPL. 404.81 Recons. 0.00852 +2024-03-17 02:28:51,710 INFO Train. Iter 134600 : Commit. 1.15164 PPL. 408.42 Recons. 0.00709 +2024-03-17 02:29:13,955 INFO Train. Iter 134800 : Commit. 1.17960 PPL. 406.62 Recons. 0.00797 +2024-03-17 02:29:36,901 INFO Train. Iter 135000 : Commit. 1.18163 PPL. 407.22 Recons. 0.00782 +2024-03-17 02:29:59,687 INFO Train. Iter 135200 : Commit. 1.14800 PPL. 408.08 Recons. 0.00716 +2024-03-17 02:30:22,162 INFO Train. Iter 135400 : Commit. 1.22227 PPL. 405.67 Recons. 0.00884 +2024-03-17 02:30:44,414 INFO Train. Iter 135600 : Commit. 1.20096 PPL. 404.70 Recons. 0.00841 +2024-03-17 02:31:07,005 INFO Train. Iter 135800 : Commit. 1.15376 PPL. 406.70 Recons. 0.00722 +2024-03-17 02:31:29,880 INFO Train. Iter 136000 : Commit. 1.17661 PPL. 407.87 Recons. 0.00759 +2024-03-17 02:31:51,655 INFO Train. Iter 136200 : Commit. 1.22112 PPL. 404.24 Recons. 0.00867 +2024-03-17 02:32:14,549 INFO Train. Iter 136400 : Commit. 1.18377 PPL. 406.90 Recons. 0.00768 +2024-03-17 02:32:36,747 INFO Train. Iter 136600 : Commit. 1.16578 PPL. 407.56 Recons. 0.00750 +2024-03-17 02:32:59,282 INFO Train. Iter 136800 : Commit. 1.19198 PPL. 407.07 Recons. 0.00816 +2024-03-17 02:33:21,763 INFO Train. Iter 137000 : Commit. 1.20488 PPL. 404.76 Recons. 0.00865 +2024-03-17 02:33:44,988 INFO Train. Iter 137200 : Commit. 1.16668 PPL. 407.54 Recons. 0.00739 +2024-03-17 02:34:07,225 INFO Train. Iter 137400 : Commit. 1.18819 PPL. 405.34 Recons. 0.00808 +2024-03-17 02:34:30,061 INFO Train. Iter 137600 : Commit. 1.19304 PPL. 406.16 Recons. 0.00826 +2024-03-17 02:34:51,891 INFO Train. Iter 137800 : Commit. 1.23838 PPL. 405.29 Recons. 0.00883 +2024-03-17 02:35:13,908 INFO Train. Iter 138000 : Commit. 1.17337 PPL. 406.88 Recons. 0.00765 +2024-03-17 02:35:34,372 INFO Train. Iter 138200 : Commit. 1.18632 PPL. 405.56 Recons. 0.00812 +2024-03-17 02:35:56,606 INFO Train. Iter 138400 : Commit. 1.16861 PPL. 407.16 Recons. 0.00746 +2024-03-17 02:36:18,850 INFO Train. Iter 138600 : Commit. 1.17738 PPL. 407.19 Recons. 0.00776 +2024-03-17 02:36:40,932 INFO Train. Iter 138800 : Commit. 1.21327 PPL. 405.49 Recons. 0.00861 +2024-03-17 02:37:03,972 INFO Train. Iter 139000 : Commit. 1.16796 PPL. 407.27 Recons. 0.00745 +2024-03-17 02:37:26,608 INFO Train. Iter 139200 : Commit. 1.26233 PPL. 403.13 Recons. 0.00890 +2024-03-17 02:37:49,100 INFO Train. Iter 139400 : Commit. 1.16648 PPL. 407.06 Recons. 0.00741 +2024-03-17 02:38:11,571 INFO Train. Iter 139600 : Commit. 1.15992 PPL. 406.19 Recons. 0.00757 +2024-03-17 02:38:34,511 INFO Train. Iter 139800 : Commit. 1.18474 PPL. 405.62 Recons. 0.00801 +2024-03-17 02:38:57,151 INFO Train. Iter 140000 : Commit. 1.18523 PPL. 405.66 Recons. 0.00786 +2024-03-17 02:39:19,204 INFO Train. Iter 140200 : Commit. 1.14045 PPL. 407.39 Recons. 0.00693 +2024-03-17 02:39:41,741 INFO Train. Iter 140400 : Commit. 1.16165 PPL. 406.67 Recons. 0.00750 +2024-03-17 02:40:04,720 INFO Train. Iter 140600 : Commit. 1.21272 PPL. 406.39 Recons. 0.00851 +2024-03-17 02:40:27,644 INFO Train. Iter 140800 : Commit. 1.15638 PPL. 407.35 Recons. 0.00711 +2024-03-17 02:40:50,546 INFO Train. Iter 141000 : Commit. 1.21058 PPL. 405.05 Recons. 0.00867 +2024-03-17 02:41:12,497 INFO Train. Iter 141200 : Commit. 1.19058 PPL. 405.23 Recons. 0.00798 +2024-03-17 02:41:35,070 INFO Train. Iter 141400 : Commit. 1.20217 PPL. 404.79 Recons. 0.00821 +2024-03-17 02:41:57,948 INFO Train. Iter 141600 : Commit. 1.21088 PPL. 405.05 Recons. 0.00823 +2024-03-17 02:42:20,445 INFO Train. Iter 141800 : Commit. 1.17984 PPL. 405.32 Recons. 0.00795 +2024-03-17 02:42:41,830 INFO Train. Iter 142000 : Commit. 1.14593 PPL. 406.69 Recons. 0.00727 +2024-03-17 02:43:04,860 INFO Train. Iter 142200 : Commit. 1.17208 PPL. 406.60 Recons. 0.00753 +2024-03-17 02:43:27,394 INFO Train. Iter 142400 : Commit. 1.18196 PPL. 405.59 Recons. 0.00797 +2024-03-17 02:43:50,107 INFO Train. Iter 142600 : Commit. 1.17592 PPL. 405.98 Recons. 0.00768 +2024-03-17 02:44:12,541 INFO Train. Iter 142800 : Commit. 1.19519 PPL. 404.82 Recons. 0.00815 +2024-03-17 02:44:35,059 INFO Train. Iter 143000 : Commit. 1.17666 PPL. 404.01 Recons. 0.00805 +2024-03-17 02:44:57,947 INFO Train. Iter 143200 : Commit. 1.14295 PPL. 407.77 Recons. 0.00676 +2024-03-17 02:45:20,803 INFO Train. Iter 143400 : Commit. 1.17560 PPL. 405.02 Recons. 0.00798 +2024-03-17 02:45:43,469 INFO Train. Iter 143600 : Commit. 1.18855 PPL. 405.96 Recons. 0.00806 +2024-03-17 02:46:05,903 INFO Train. Iter 143800 : Commit. 1.19515 PPL. 404.97 Recons. 0.00818 +2024-03-17 02:46:27,435 INFO Train. Iter 144000 : Commit. 1.16233 PPL. 406.88 Recons. 0.00727 +2024-03-17 02:46:50,160 INFO Train. Iter 144200 : Commit. 1.22671 PPL. 404.60 Recons. 0.00883 +2024-03-17 02:47:13,126 INFO Train. Iter 144400 : Commit. 1.15365 PPL. 406.94 Recons. 0.00713 +2024-03-17 02:47:36,466 INFO Train. Iter 144600 : Commit. 1.16868 PPL. 406.09 Recons. 0.00772 +2024-03-17 02:47:59,011 INFO Train. Iter 144800 : Commit. 1.18842 PPL. 404.15 Recons. 0.00792 +2024-03-17 02:48:22,249 INFO Train. Iter 145000 : Commit. 1.22604 PPL. 404.15 Recons. 0.00891 +2024-03-17 02:48:45,376 INFO Train. Iter 145200 : Commit. 1.20698 PPL. 405.48 Recons. 0.00817 +2024-03-17 02:49:07,509 INFO Train. Iter 145400 : Commit. 1.15664 PPL. 403.72 Recons. 0.00771 +2024-03-17 02:49:29,727 INFO Train. Iter 145600 : Commit. 1.25600 PPL. 403.06 Recons. 0.00898 +2024-03-17 02:49:52,075 INFO Train. Iter 145800 : Commit. 1.15498 PPL. 406.00 Recons. 0.00722 +2024-03-17 02:50:13,565 INFO Train. Iter 146000 : Commit. 1.16480 PPL. 405.98 Recons. 0.00752 +2024-03-17 02:50:36,948 INFO Train. Iter 146200 : Commit. 1.21069 PPL. 403.02 Recons. 0.00851 +2024-03-17 02:50:59,866 INFO Train. Iter 146400 : Commit. 1.16315 PPL. 405.79 Recons. 0.00752 +2024-03-17 02:51:22,745 INFO Train. Iter 146600 : Commit. 1.17564 PPL. 405.66 Recons. 0.00760 +2024-03-17 02:51:45,301 INFO Train. Iter 146800 : Commit. 1.20484 PPL. 404.46 Recons. 0.00833 +2024-03-17 02:52:08,218 INFO Train. Iter 147000 : Commit. 1.15297 PPL. 405.41 Recons. 0.00735 +2024-03-17 02:52:31,066 INFO Train. Iter 147200 : Commit. 1.16693 PPL. 406.29 Recons. 0.00748 +2024-03-17 02:52:53,441 INFO Train. Iter 147400 : Commit. 1.18538 PPL. 405.40 Recons. 0.00795 +2024-03-17 02:53:15,696 INFO Train. Iter 147600 : Commit. 1.15613 PPL. 406.08 Recons. 0.00728 +2024-03-17 02:53:37,484 INFO Train. Iter 147800 : Commit. 1.26089 PPL. 400.49 Recons. 0.00973 +2024-03-17 02:54:00,579 INFO Train. Iter 148000 : Commit. 1.15464 PPL. 406.04 Recons. 0.00684 +2024-03-17 02:54:22,707 INFO Train. Iter 148200 : Commit. 1.17945 PPL. 405.63 Recons. 0.00793 +2024-03-17 02:54:45,430 INFO Train. Iter 148400 : Commit. 1.17086 PPL. 404.20 Recons. 0.00800 +2024-03-17 02:55:08,120 INFO Train. Iter 148600 : Commit. 1.12188 PPL. 406.64 Recons. 0.00660 +2024-03-17 02:55:31,167 INFO Train. Iter 148800 : Commit. 1.16784 PPL. 403.91 Recons. 0.00792 +2024-03-17 02:55:53,975 INFO Train. Iter 149000 : Commit. 1.18174 PPL. 404.89 Recons. 0.00808 +2024-03-17 02:56:16,480 INFO Train. Iter 149200 : Commit. 1.20274 PPL. 404.94 Recons. 0.00825 +2024-03-17 02:56:38,932 INFO Train. Iter 149400 : Commit. 1.14486 PPL. 406.52 Recons. 0.00710 +2024-03-17 02:57:01,419 INFO Train. Iter 149600 : Commit. 1.16839 PPL. 405.48 Recons. 0.00758 +2024-03-17 02:57:22,974 INFO Train. Iter 149800 : Commit. 1.18698 PPL. 402.89 Recons. 0.00841 +2024-03-17 02:57:45,852 INFO Train. Iter 150000 : Commit. 1.15371 PPL. 406.20 Recons. 0.00701 +2024-03-17 02:58:08,184 INFO Train. Iter 150200 : Commit. 1.14844 PPL. 404.88 Recons. 0.00725 +2024-03-17 02:58:30,808 INFO Train. Iter 150400 : Commit. 1.20527 PPL. 402.49 Recons. 0.00815 +2024-03-17 02:58:53,227 INFO Train. Iter 150600 : Commit. 1.18953 PPL. 403.05 Recons. 0.00834 +2024-03-17 02:59:16,129 INFO Train. Iter 150800 : Commit. 1.14210 PPL. 406.58 Recons. 0.00672 +2024-03-17 02:59:39,179 INFO Train. Iter 151000 : Commit. 1.15355 PPL. 405.74 Recons. 0.00752 +2024-03-17 03:00:02,339 INFO Train. Iter 151200 : Commit. 1.20450 PPL. 404.09 Recons. 0.00864 +2024-03-17 03:00:25,207 INFO Train. Iter 151400 : Commit. 1.14825 PPL. 404.35 Recons. 0.00742 +2024-03-17 03:00:48,087 INFO Train. Iter 151600 : Commit. 1.18124 PPL. 403.94 Recons. 0.00815 +2024-03-17 03:01:09,194 INFO Train. Iter 151800 : Commit. 1.14638 PPL. 405.92 Recons. 0.00725 +2024-03-17 03:01:31,911 INFO Train. Iter 152000 : Commit. 1.15943 PPL. 405.07 Recons. 0.00751 +2024-03-17 03:01:54,997 INFO Train. Iter 152200 : Commit. 1.15895 PPL. 405.48 Recons. 0.00734 +2024-03-17 03:02:17,723 INFO Train. Iter 152400 : Commit. 1.15591 PPL. 406.25 Recons. 0.00722 +2024-03-17 03:02:40,346 INFO Train. Iter 152600 : Commit. 1.17264 PPL. 404.78 Recons. 0.00779 +2024-03-17 03:03:02,574 INFO Train. Iter 152800 : Commit. 1.13193 PPL. 405.64 Recons. 0.00700 +2024-03-17 03:03:24,947 INFO Train. Iter 153000 : Commit. 1.16878 PPL. 404.73 Recons. 0.00756 +2024-03-17 03:03:47,643 INFO Train. Iter 153200 : Commit. 1.15687 PPL. 404.61 Recons. 0.00732 +2024-03-17 03:04:09,572 INFO Train. Iter 153400 : Commit. 1.16786 PPL. 403.75 Recons. 0.00772 +2024-03-17 03:04:30,988 INFO Train. Iter 153600 : Commit. 1.14382 PPL. 405.82 Recons. 0.00704 +2024-03-17 03:04:53,651 INFO Train. Iter 153800 : Commit. 1.14819 PPL. 405.03 Recons. 0.00725 +2024-03-17 03:05:16,212 INFO Train. Iter 154000 : Commit. 1.20434 PPL. 401.74 Recons. 0.00894 +2024-03-17 03:05:39,608 INFO Train. Iter 154200 : Commit. 1.17248 PPL. 405.35 Recons. 0.00738 +2024-03-17 03:06:02,322 INFO Train. Iter 154400 : Commit. 1.18630 PPL. 403.05 Recons. 0.00810 +2024-03-17 03:06:24,698 INFO Train. Iter 154600 : Commit. 1.19302 PPL. 404.22 Recons. 0.00783 +2024-03-17 03:06:47,169 INFO Train. Iter 154800 : Commit. 1.20134 PPL. 403.52 Recons. 0.00826 +2024-03-17 03:07:09,531 INFO Train. Iter 155000 : Commit. 1.16416 PPL. 403.71 Recons. 0.00749 +2024-03-17 03:07:32,594 INFO Train. Iter 155200 : Commit. 1.17823 PPL. 404.69 Recons. 0.00766 +2024-03-17 03:07:56,020 INFO Train. Iter 155400 : Commit. 1.14162 PPL. 405.31 Recons. 0.00709 +2024-03-17 03:08:17,981 INFO Train. Iter 155600 : Commit. 1.17326 PPL. 404.21 Recons. 0.00778 +2024-03-17 03:08:40,444 INFO Train. Iter 155800 : Commit. 1.17214 PPL. 403.56 Recons. 0.00757 +2024-03-17 03:09:02,733 INFO Train. Iter 156000 : Commit. 1.12793 PPL. 405.55 Recons. 0.00682 +2024-03-17 03:09:25,340 INFO Train. Iter 156200 : Commit. 1.19037 PPL. 401.74 Recons. 0.00850 +2024-03-17 03:09:47,754 INFO Train. Iter 156400 : Commit. 1.19401 PPL. 402.58 Recons. 0.00820 +2024-03-17 03:10:11,536 INFO Train. Iter 156600 : Commit. 1.16916 PPL. 404.02 Recons. 0.00747 +2024-03-17 03:10:34,361 INFO Train. Iter 156800 : Commit. 1.15757 PPL. 405.34 Recons. 0.00727 +2024-03-17 03:10:56,806 INFO Train. Iter 157000 : Commit. 1.14480 PPL. 406.04 Recons. 0.00697 +2024-03-17 03:11:19,268 INFO Train. Iter 157200 : Commit. 1.12582 PPL. 405.72 Recons. 0.00705 +2024-03-17 03:11:40,858 INFO Train. Iter 157400 : Commit. 1.15628 PPL. 404.76 Recons. 0.00750 +2024-03-17 03:12:01,726 INFO Train. Iter 157600 : Commit. 1.13331 PPL. 405.72 Recons. 0.00695 +2024-03-17 03:12:24,043 INFO Train. Iter 157800 : Commit. 1.16942 PPL. 401.66 Recons. 0.00832 +2024-03-17 03:12:46,864 INFO Train. Iter 158000 : Commit. 1.14679 PPL. 405.59 Recons. 0.00683 +2024-03-17 03:13:09,740 INFO Train. Iter 158200 : Commit. 1.20908 PPL. 401.91 Recons. 0.00850 +2024-03-17 03:13:32,483 INFO Train. Iter 158400 : Commit. 1.15507 PPL. 403.65 Recons. 0.00749 +2024-03-17 03:13:55,070 INFO Train. Iter 158600 : Commit. 1.16711 PPL. 403.17 Recons. 0.00759 +2024-03-17 03:14:17,323 INFO Train. Iter 158800 : Commit. 1.23153 PPL. 401.80 Recons. 0.00887 +2024-03-17 03:14:39,488 INFO Train. Iter 159000 : Commit. 1.18426 PPL. 403.16 Recons. 0.00781 +2024-03-17 03:15:03,201 INFO Train. Iter 159200 : Commit. 1.17617 PPL. 403.81 Recons. 0.00771 +2024-03-17 03:15:24,678 INFO Train. Iter 159400 : Commit. 1.16007 PPL. 404.89 Recons. 0.00730 +2024-03-17 03:15:47,526 INFO Train. Iter 159600 : Commit. 1.15645 PPL. 404.33 Recons. 0.00746 +2024-03-17 03:16:10,456 INFO Train. Iter 159800 : Commit. 1.18767 PPL. 402.50 Recons. 0.00801 +2024-03-17 03:16:33,269 INFO Train. Iter 160000 : Commit. 1.14507 PPL. 405.32 Recons. 0.00699 +2024-03-17 03:16:56,162 INFO Train. Iter 160200 : Commit. 1.16743 PPL. 401.70 Recons. 0.00788 +2024-03-17 03:17:18,712 INFO Train. Iter 160400 : Commit. 1.19055 PPL. 403.62 Recons. 0.00798 +2024-03-17 03:17:41,331 INFO Train. Iter 160600 : Commit. 1.14594 PPL. 404.08 Recons. 0.00721 +2024-03-17 03:18:04,361 INFO Train. Iter 160800 : Commit. 1.15180 PPL. 403.85 Recons. 0.00723 +2024-03-17 03:18:27,491 INFO Train. Iter 161000 : Commit. 1.16683 PPL. 404.08 Recons. 0.00759 +2024-03-17 03:18:50,140 INFO Train. Iter 161200 : Commit. 1.15383 PPL. 403.20 Recons. 0.00755 +2024-03-17 03:19:11,968 INFO Train. Iter 161400 : Commit. 1.18161 PPL. 402.44 Recons. 0.00803 +2024-03-17 03:19:34,367 INFO Train. Iter 161600 : Commit. 1.16002 PPL. 403.83 Recons. 0.00741 +2024-03-17 03:19:57,127 INFO Train. Iter 161800 : Commit. 1.12664 PPL. 404.78 Recons. 0.00692 +2024-03-17 03:20:20,431 INFO Train. Iter 162000 : Commit. 1.16823 PPL. 404.91 Recons. 0.00762 +2024-03-17 03:20:43,183 INFO Train. Iter 162200 : Commit. 1.20987 PPL. 401.45 Recons. 0.00873 +2024-03-17 03:21:05,807 INFO Train. Iter 162400 : Commit. 1.14664 PPL. 405.06 Recons. 0.00702 +2024-03-17 03:21:28,594 INFO Train. Iter 162600 : Commit. 1.17000 PPL. 402.70 Recons. 0.00773 +2024-03-17 03:21:51,902 INFO Train. Iter 162800 : Commit. 1.14142 PPL. 405.56 Recons. 0.00696 +2024-03-17 03:22:15,211 INFO Train. Iter 163000 : Commit. 1.14581 PPL. 403.63 Recons. 0.00737 +2024-03-17 03:22:38,158 INFO Train. Iter 163200 : Commit. 1.14091 PPL. 404.13 Recons. 0.00712 +2024-03-17 03:23:00,099 INFO Train. Iter 163400 : Commit. 1.14624 PPL. 403.94 Recons. 0.00715 +2024-03-17 03:23:22,671 INFO Train. Iter 163600 : Commit. 1.13978 PPL. 404.11 Recons. 0.00708 +2024-03-17 03:23:45,135 INFO Train. Iter 163800 : Commit. 1.16274 PPL. 402.36 Recons. 0.00755 +2024-03-17 03:24:07,157 INFO Train. Iter 164000 : Commit. 1.17012 PPL. 404.05 Recons. 0.00762 +2024-03-17 03:24:29,452 INFO Train. Iter 164200 : Commit. 1.11563 PPL. 404.59 Recons. 0.00651 +2024-03-17 03:24:52,211 INFO Train. Iter 164400 : Commit. 1.15864 PPL. 404.82 Recons. 0.00745 +2024-03-17 03:25:14,669 INFO Train. Iter 164600 : Commit. 1.12364 PPL. 403.84 Recons. 0.00687 +2024-03-17 03:25:37,136 INFO Train. Iter 164800 : Commit. 1.15838 PPL. 402.62 Recons. 0.00768 +2024-03-17 03:25:59,791 INFO Train. Iter 165000 : Commit. 1.14133 PPL. 403.45 Recons. 0.00716 +2024-03-17 03:26:20,917 INFO Train. Iter 165200 : Commit. 1.18590 PPL. 402.63 Recons. 0.00828 +2024-03-17 03:26:43,673 INFO Train. Iter 165400 : Commit. 1.17122 PPL. 403.86 Recons. 0.00740 +2024-03-17 03:27:05,800 INFO Train. Iter 165600 : Commit. 1.11390 PPL. 404.17 Recons. 0.00666 +2024-03-17 03:27:28,612 INFO Train. Iter 165800 : Commit. 1.15356 PPL. 402.85 Recons. 0.00747 +2024-03-17 03:27:51,520 INFO Train. Iter 166000 : Commit. 1.12522 PPL. 404.43 Recons. 0.00672 +2024-03-17 03:28:14,306 INFO Train. Iter 166200 : Commit. 1.15527 PPL. 403.14 Recons. 0.00766 +2024-03-17 03:28:36,814 INFO Train. Iter 166400 : Commit. 1.14274 PPL. 404.71 Recons. 0.00708 +2024-03-17 03:28:59,806 INFO Train. Iter 166600 : Commit. 1.19396 PPL. 398.42 Recons. 0.00807 +2024-03-17 03:29:22,253 INFO Train. Iter 166800 : Commit. 1.19299 PPL. 400.82 Recons. 0.00837 +2024-03-17 03:29:45,219 INFO Train. Iter 167000 : Commit. 1.16266 PPL. 401.49 Recons. 0.00763 +2024-03-17 03:30:06,927 INFO Train. Iter 167200 : Commit. 1.15116 PPL. 402.57 Recons. 0.00721 +2024-03-17 03:30:29,951 INFO Train. Iter 167400 : Commit. 1.14871 PPL. 403.98 Recons. 0.00699 +2024-03-17 03:30:52,601 INFO Train. Iter 167600 : Commit. 1.17926 PPL. 402.57 Recons. 0.00810 +2024-03-17 03:31:14,563 INFO Train. Iter 167800 : Commit. 1.12751 PPL. 404.08 Recons. 0.00665 +2024-03-17 03:31:37,126 INFO Train. Iter 168000 : Commit. 1.17863 PPL. 402.46 Recons. 0.00801 +2024-03-17 03:32:00,568 INFO Train. Iter 168200 : Commit. 1.15742 PPL. 403.61 Recons. 0.00737 +2024-03-17 03:32:22,969 INFO Train. Iter 168400 : Commit. 1.14930 PPL. 401.52 Recons. 0.00759 +2024-03-17 03:32:45,687 INFO Train. Iter 168600 : Commit. 1.12391 PPL. 403.36 Recons. 0.00672 +2024-03-17 03:33:08,536 INFO Train. Iter 168800 : Commit. 1.13878 PPL. 403.38 Recons. 0.00708 +2024-03-17 03:33:31,282 INFO Train. Iter 169000 : Commit. 1.20853 PPL. 400.49 Recons. 0.00871 +2024-03-17 03:33:53,078 INFO Train. Iter 169200 : Commit. 1.18092 PPL. 402.24 Recons. 0.00782 +2024-03-17 03:34:15,527 INFO Train. Iter 169400 : Commit. 1.19924 PPL. 401.45 Recons. 0.00844 +2024-03-17 03:34:38,592 INFO Train. Iter 169600 : Commit. 1.20657 PPL. 402.04 Recons. 0.00813 +2024-03-17 03:35:00,938 INFO Train. Iter 169800 : Commit. 1.17898 PPL. 404.01 Recons. 0.00764 +2024-03-17 03:35:23,436 INFO Train. Iter 170000 : Commit. 1.12837 PPL. 403.88 Recons. 0.00685 +2024-03-17 03:35:46,238 INFO Train. Iter 170200 : Commit. 1.20091 PPL. 397.55 Recons. 0.00876 +2024-03-17 03:36:08,730 INFO Train. Iter 170400 : Commit. 1.20486 PPL. 402.00 Recons. 0.00782 +2024-03-17 03:36:31,342 INFO Train. Iter 170600 : Commit. 1.16006 PPL. 402.95 Recons. 0.00748 +2024-03-17 03:36:54,074 INFO Train. Iter 170800 : Commit. 1.19820 PPL. 400.38 Recons. 0.00815 +2024-03-17 03:37:15,775 INFO Train. Iter 171000 : Commit. 1.18056 PPL. 401.14 Recons. 0.00787 +2024-03-17 03:37:38,976 INFO Train. Iter 171200 : Commit. 1.19150 PPL. 402.26 Recons. 0.00796 +2024-03-17 03:38:01,623 INFO Train. Iter 171400 : Commit. 1.20237 PPL. 401.91 Recons. 0.00812 +2024-03-17 03:38:24,747 INFO Train. Iter 171600 : Commit. 1.13386 PPL. 403.71 Recons. 0.00688 +2024-03-17 03:38:47,960 INFO Train. Iter 171800 : Commit. 1.19044 PPL. 401.17 Recons. 0.00839 +2024-03-17 03:39:10,458 INFO Train. Iter 172000 : Commit. 1.13506 PPL. 402.11 Recons. 0.00705 +2024-03-17 03:39:32,935 INFO Train. Iter 172200 : Commit. 1.19747 PPL. 400.00 Recons. 0.00816 +2024-03-17 03:39:55,470 INFO Train. Iter 172400 : Commit. 1.17584 PPL. 402.54 Recons. 0.00781 +2024-03-17 03:40:18,598 INFO Train. Iter 172600 : Commit. 1.17674 PPL. 399.08 Recons. 0.00786 +2024-03-17 03:40:41,759 INFO Train. Iter 172800 : Commit. 1.18218 PPL. 401.10 Recons. 0.00811 +2024-03-17 03:41:03,291 INFO Train. Iter 173000 : Commit. 1.16824 PPL. 402.46 Recons. 0.00773 +2024-03-17 03:41:26,210 INFO Train. Iter 173200 : Commit. 1.12805 PPL. 404.02 Recons. 0.00662 +2024-03-17 03:41:49,395 INFO Train. Iter 173400 : Commit. 1.16886 PPL. 401.47 Recons. 0.00772 +2024-03-17 03:42:12,468 INFO Train. Iter 173600 : Commit. 1.15950 PPL. 403.29 Recons. 0.00720 +2024-03-17 03:42:35,195 INFO Train. Iter 173800 : Commit. 1.15152 PPL. 401.63 Recons. 0.00744 +2024-03-17 03:42:57,614 INFO Train. Iter 174000 : Commit. 1.16840 PPL. 401.40 Recons. 0.00790 +2024-03-17 03:43:19,995 INFO Train. Iter 174200 : Commit. 1.15269 PPL. 401.05 Recons. 0.00754 +2024-03-17 03:43:42,555 INFO Train. Iter 174400 : Commit. 1.14484 PPL. 402.90 Recons. 0.00685 +2024-03-17 03:44:04,956 INFO Train. Iter 174600 : Commit. 1.16401 PPL. 401.60 Recons. 0.00770 +2024-03-17 03:44:27,171 INFO Train. Iter 174800 : Commit. 1.16591 PPL. 401.17 Recons. 0.00778 +2024-03-17 03:44:49,064 INFO Train. Iter 175000 : Commit. 1.19802 PPL. 401.99 Recons. 0.00777 +2024-03-17 03:45:12,221 INFO Train. Iter 175200 : Commit. 1.14092 PPL. 402.44 Recons. 0.00721 +2024-03-17 03:45:35,033 INFO Train. Iter 175400 : Commit. 1.15187 PPL. 401.71 Recons. 0.00720 +2024-03-17 03:45:58,251 INFO Train. Iter 175600 : Commit. 1.16365 PPL. 402.00 Recons. 0.00764 +2024-03-17 03:46:21,254 INFO Train. Iter 175800 : Commit. 1.22432 PPL. 398.58 Recons. 0.00888 +2024-03-17 03:46:43,722 INFO Train. Iter 176000 : Commit. 1.14932 PPL. 402.53 Recons. 0.00723 +2024-03-17 03:47:06,183 INFO Train. Iter 176200 : Commit. 1.13368 PPL. 402.77 Recons. 0.00694 +2024-03-17 03:47:28,802 INFO Train. Iter 176400 : Commit. 1.16613 PPL. 399.78 Recons. 0.00821 +2024-03-17 03:47:51,500 INFO Train. Iter 176600 : Commit. 1.15327 PPL. 401.46 Recons. 0.00714 +2024-03-17 03:48:13,121 INFO Train. Iter 176800 : Commit. 1.12980 PPL. 403.76 Recons. 0.00668 +2024-03-17 03:48:35,374 INFO Train. Iter 177000 : Commit. 1.13405 PPL. 402.13 Recons. 0.00720 +2024-03-17 03:48:57,682 INFO Train. Iter 177200 : Commit. 1.19620 PPL. 398.16 Recons. 0.00850 +2024-03-17 03:49:20,294 INFO Train. Iter 177400 : Commit. 1.14629 PPL. 402.75 Recons. 0.00699 +2024-03-17 03:49:43,741 INFO Train. Iter 177600 : Commit. 1.16569 PPL. 400.83 Recons. 0.00783 +2024-03-17 03:50:06,993 INFO Train. Iter 177800 : Commit. 1.18545 PPL. 399.88 Recons. 0.00814 +2024-03-17 03:50:29,576 INFO Train. Iter 178000 : Commit. 1.13245 PPL. 402.76 Recons. 0.00679 +2024-03-17 03:50:51,521 INFO Train. Iter 178200 : Commit. 1.15367 PPL. 400.47 Recons. 0.00750 +2024-03-17 03:51:13,640 INFO Train. Iter 178400 : Commit. 1.12885 PPL. 401.83 Recons. 0.00688 +2024-03-17 03:51:36,395 INFO Train. Iter 178600 : Commit. 1.11517 PPL. 402.33 Recons. 0.00661 +2024-03-17 03:51:58,259 INFO Train. Iter 178800 : Commit. 1.16159 PPL. 399.36 Recons. 0.00785 +2024-03-17 03:52:20,523 INFO Train. Iter 179000 : Commit. 1.15218 PPL. 402.35 Recons. 0.00716 +2024-03-17 03:52:42,902 INFO Train. Iter 179200 : Commit. 1.13121 PPL. 400.80 Recons. 0.00696 +2024-03-17 03:53:05,795 INFO Train. Iter 179400 : Commit. 1.13846 PPL. 401.51 Recons. 0.00702 +2024-03-17 03:53:28,414 INFO Train. Iter 179600 : Commit. 1.14213 PPL. 400.70 Recons. 0.00728 +2024-03-17 03:53:51,060 INFO Train. Iter 179800 : Commit. 1.13594 PPL. 402.29 Recons. 0.00684 +2024-03-17 03:54:13,593 INFO Train. Iter 180000 : Commit. 1.14150 PPL. 400.95 Recons. 0.00732 +2024-03-17 03:54:36,664 INFO Train. Iter 180200 : Commit. 1.14898 PPL. 400.79 Recons. 0.00707 +2024-03-17 03:54:59,770 INFO Train. Iter 180400 : Commit. 1.18939 PPL. 398.66 Recons. 0.00833 +2024-03-17 03:55:22,554 INFO Train. Iter 180600 : Commit. 1.15376 PPL. 403.13 Recons. 0.00688 +2024-03-17 03:55:44,002 INFO Train. Iter 180800 : Commit. 1.15952 PPL. 399.32 Recons. 0.00765 +2024-03-17 03:56:06,779 INFO Train. Iter 181000 : Commit. 1.14594 PPL. 402.58 Recons. 0.00694 +2024-03-17 03:56:29,955 INFO Train. Iter 181200 : Commit. 1.18355 PPL. 398.16 Recons. 0.00834 +2024-03-17 03:56:53,069 INFO Train. Iter 181400 : Commit. 1.14247 PPL. 401.89 Recons. 0.00675 +2024-03-17 03:57:16,028 INFO Train. Iter 181600 : Commit. 1.11724 PPL. 402.20 Recons. 0.00666 +2024-03-17 03:57:38,899 INFO Train. Iter 181800 : Commit. 1.17127 PPL. 399.89 Recons. 0.00783 +2024-03-17 03:58:00,911 INFO Train. Iter 182000 : Commit. 1.16182 PPL. 400.75 Recons. 0.00734 +2024-03-17 03:58:23,267 INFO Train. Iter 182200 : Commit. 1.17602 PPL. 400.92 Recons. 0.00773 +2024-03-17 03:58:46,146 INFO Train. Iter 182400 : Commit. 1.15756 PPL. 399.68 Recons. 0.00738 +2024-03-17 03:59:08,234 INFO Train. Iter 182600 : Commit. 1.13225 PPL. 401.31 Recons. 0.00668 +2024-03-17 03:59:31,158 INFO Train. Iter 182800 : Commit. 1.15028 PPL. 401.29 Recons. 0.00730 +2024-03-17 03:59:53,052 INFO Train. Iter 183000 : Commit. 1.20508 PPL. 399.72 Recons. 0.00847 +2024-03-17 04:00:15,414 INFO Train. Iter 183200 : Commit. 1.18472 PPL. 399.75 Recons. 0.00807 +2024-03-17 04:00:38,280 INFO Train. Iter 183400 : Commit. 1.18716 PPL. 400.87 Recons. 0.00786 +2024-03-17 04:01:00,607 INFO Train. Iter 183600 : Commit. 1.15619 PPL. 400.53 Recons. 0.00747 +2024-03-17 04:01:23,440 INFO Train. Iter 183800 : Commit. 1.19563 PPL. 400.37 Recons. 0.00808 +2024-03-17 04:01:46,544 INFO Train. Iter 184000 : Commit. 1.17215 PPL. 400.70 Recons. 0.00764 +2024-03-17 04:02:08,881 INFO Train. Iter 184200 : Commit. 1.16088 PPL. 398.98 Recons. 0.00739 +2024-03-17 04:02:31,588 INFO Train. Iter 184400 : Commit. 1.12288 PPL. 401.72 Recons. 0.00669 +2024-03-17 04:02:53,217 INFO Train. Iter 184600 : Commit. 1.13534 PPL. 402.48 Recons. 0.00691 +2024-03-17 04:03:15,541 INFO Train. Iter 184800 : Commit. 1.13076 PPL. 401.96 Recons. 0.00690 +2024-03-17 04:03:37,907 INFO Train. Iter 185000 : Commit. 1.15250 PPL. 399.45 Recons. 0.00758 +2024-03-17 04:04:01,120 INFO Train. Iter 185200 : Commit. 1.14400 PPL. 400.78 Recons. 0.00717 +2024-03-17 04:04:23,609 INFO Train. Iter 185400 : Commit. 1.18510 PPL. 398.87 Recons. 0.00795 +2024-03-17 04:04:46,399 INFO Train. Iter 185600 : Commit. 1.18222 PPL. 400.17 Recons. 0.00783 +2024-03-17 04:05:08,942 INFO Train. Iter 185800 : Commit. 1.15788 PPL. 400.56 Recons. 0.00742 +2024-03-17 04:05:31,518 INFO Train. Iter 186000 : Commit. 1.12158 PPL. 402.02 Recons. 0.00659 +2024-03-17 04:05:54,126 INFO Train. Iter 186200 : Commit. 1.17568 PPL. 398.02 Recons. 0.00825 +2024-03-17 04:06:16,824 INFO Train. Iter 186400 : Commit. 1.16967 PPL. 398.15 Recons. 0.00774 +2024-03-17 04:06:38,195 INFO Train. Iter 186600 : Commit. 1.14000 PPL. 402.18 Recons. 0.00672 +2024-03-17 04:07:01,487 INFO Train. Iter 186800 : Commit. 1.17189 PPL. 400.51 Recons. 0.00780 +2024-03-17 04:07:24,422 INFO Train. Iter 187000 : Commit. 1.17287 PPL. 398.10 Recons. 0.00783 +2024-03-17 04:07:46,810 INFO Train. Iter 187200 : Commit. 1.16836 PPL. 399.81 Recons. 0.00747 +2024-03-17 04:08:09,081 INFO Train. Iter 187400 : Commit. 1.17009 PPL. 400.35 Recons. 0.00768 +2024-03-17 04:08:31,476 INFO Train. Iter 187600 : Commit. 1.16519 PPL. 399.34 Recons. 0.00727 +2024-03-17 04:08:54,330 INFO Train. Iter 187800 : Commit. 1.14700 PPL. 400.48 Recons. 0.00720 +2024-03-17 04:09:17,278 INFO Train. Iter 188000 : Commit. 1.12532 PPL. 401.04 Recons. 0.00670 +2024-03-17 04:09:39,538 INFO Train. Iter 188200 : Commit. 1.14897 PPL. 400.33 Recons. 0.00740 +2024-03-17 04:10:00,764 INFO Train. Iter 188400 : Commit. 1.13346 PPL. 400.81 Recons. 0.00683 +2024-03-17 04:10:23,824 INFO Train. Iter 188600 : Commit. 1.13180 PPL. 401.30 Recons. 0.00689 +2024-03-17 04:10:46,749 INFO Train. Iter 188800 : Commit. 1.09651 PPL. 401.88 Recons. 0.00622 +2024-03-17 04:11:09,281 INFO Train. Iter 189000 : Commit. 1.11584 PPL. 400.76 Recons. 0.00667 +2024-03-17 04:11:31,387 INFO Train. Iter 189200 : Commit. 1.12724 PPL. 400.86 Recons. 0.00682 +2024-03-17 04:11:53,680 INFO Train. Iter 189400 : Commit. 1.14687 PPL. 400.16 Recons. 0.00715 +2024-03-17 04:12:15,318 INFO Train. Iter 189600 : Commit. 1.17209 PPL. 399.50 Recons. 0.00791 +2024-03-17 04:12:37,417 INFO Train. Iter 189800 : Commit. 1.14855 PPL. 401.40 Recons. 0.00710 +2024-03-17 04:12:59,919 INFO Train. Iter 190000 : Commit. 1.17450 PPL. 396.95 Recons. 0.00787 +2024-03-17 04:13:23,013 INFO Train. Iter 190200 : Commit. 1.13979 PPL. 400.50 Recons. 0.00687 +2024-03-17 04:13:44,501 INFO Train. Iter 190400 : Commit. 1.16573 PPL. 397.57 Recons. 0.00794 +2024-03-17 04:14:07,475 INFO Train. Iter 190600 : Commit. 1.14442 PPL. 401.56 Recons. 0.00669 +2024-03-17 04:14:30,549 INFO Train. Iter 190800 : Commit. 1.16100 PPL. 399.84 Recons. 0.00757 +2024-03-17 04:14:53,104 INFO Train. Iter 191000 : Commit. 1.16079 PPL. 398.99 Recons. 0.00753 +2024-03-17 04:15:15,563 INFO Train. Iter 191200 : Commit. 1.14392 PPL. 399.95 Recons. 0.00711 +2024-03-17 04:15:38,124 INFO Train. Iter 191400 : Commit. 1.15090 PPL. 399.45 Recons. 0.00737 +2024-03-17 04:16:00,928 INFO Train. Iter 191600 : Commit. 1.12349 PPL. 400.18 Recons. 0.00672 +2024-03-17 04:16:24,152 INFO Train. Iter 191800 : Commit. 1.15489 PPL. 399.85 Recons. 0.00737 +2024-03-17 04:16:46,720 INFO Train. Iter 192000 : Commit. 1.13877 PPL. 398.94 Recons. 0.00703 +2024-03-17 04:17:09,136 INFO Train. Iter 192200 : Commit. 1.15579 PPL. 398.07 Recons. 0.00736 +2024-03-17 04:17:30,327 INFO Train. Iter 192400 : Commit. 1.18721 PPL. 397.48 Recons. 0.00805 +2024-03-17 04:17:53,059 INFO Train. Iter 192600 : Commit. 1.15668 PPL. 399.76 Recons. 0.00727 +2024-03-17 04:18:15,607 INFO Train. Iter 192800 : Commit. 1.13240 PPL. 399.88 Recons. 0.00692 +2024-03-17 04:18:38,423 INFO Train. Iter 193000 : Commit. 1.15610 PPL. 398.72 Recons. 0.00750 +2024-03-17 04:19:01,352 INFO Train. Iter 193200 : Commit. 1.12185 PPL. 399.57 Recons. 0.00667 +2024-03-17 04:19:24,286 INFO Train. Iter 193400 : Commit. 1.11438 PPL. 400.00 Recons. 0.00663 +2024-03-17 04:19:47,112 INFO Train. Iter 193600 : Commit. 1.11756 PPL. 400.49 Recons. 0.00664 +2024-03-17 04:20:10,105 INFO Train. Iter 193800 : Commit. 1.15922 PPL. 398.95 Recons. 0.00745 +2024-03-17 04:20:33,218 INFO Train. Iter 194000 : Commit. 1.15390 PPL. 400.45 Recons. 0.00722 +2024-03-17 04:20:54,236 INFO Train. Iter 194200 : Commit. 1.14606 PPL. 400.56 Recons. 0.00700 +2024-03-17 04:21:16,752 INFO Train. Iter 194400 : Commit. 1.17863 PPL. 396.82 Recons. 0.00841 +2024-03-17 04:21:39,711 INFO Train. Iter 194600 : Commit. 1.18163 PPL. 399.28 Recons. 0.00755 +2024-03-17 04:22:02,790 INFO Train. Iter 194800 : Commit. 1.12897 PPL. 398.67 Recons. 0.00698 +2024-03-17 04:22:25,692 INFO Train. Iter 195000 : Commit. 1.10249 PPL. 400.94 Recons. 0.00629 +2024-03-17 04:22:48,176 INFO Train. Iter 195200 : Commit. 1.15992 PPL. 397.88 Recons. 0.00744 +2024-03-17 04:23:11,483 INFO Train. Iter 195400 : Commit. 1.14220 PPL. 400.37 Recons. 0.00705 +2024-03-17 04:23:34,575 INFO Train. Iter 195600 : Commit. 1.14737 PPL. 397.95 Recons. 0.00729 +2024-03-17 04:23:57,348 INFO Train. Iter 195800 : Commit. 1.17850 PPL. 397.97 Recons. 0.00797 +2024-03-17 04:24:20,070 INFO Train. Iter 196000 : Commit. 1.15134 PPL. 398.75 Recons. 0.00699 +2024-03-17 04:24:41,585 INFO Train. Iter 196200 : Commit. 1.13792 PPL. 398.00 Recons. 0.00715 +2024-03-17 04:25:04,390 INFO Train. Iter 196400 : Commit. 1.14285 PPL. 399.25 Recons. 0.00702 +2024-03-17 04:25:26,871 INFO Train. Iter 196600 : Commit. 1.16473 PPL. 398.61 Recons. 0.00756 +2024-03-17 04:25:49,737 INFO Train. Iter 196800 : Commit. 1.16022 PPL. 399.35 Recons. 0.00733 +2024-03-17 04:26:12,543 INFO Train. Iter 197000 : Commit. 1.12257 PPL. 399.78 Recons. 0.00674 +2024-03-17 04:26:35,088 INFO Train. Iter 197200 : Commit. 1.14374 PPL. 398.02 Recons. 0.00724 +2024-03-17 04:26:58,242 INFO Train. Iter 197400 : Commit. 1.13146 PPL. 399.85 Recons. 0.00681 +2024-03-17 04:27:21,054 INFO Train. Iter 197600 : Commit. 1.17946 PPL. 396.93 Recons. 0.00800 +2024-03-17 04:27:43,116 INFO Train. Iter 197800 : Commit. 1.14308 PPL. 398.54 Recons. 0.00709 +2024-03-17 04:28:05,469 INFO Train. Iter 198000 : Commit. 1.15604 PPL. 397.66 Recons. 0.00736 +2024-03-17 04:28:27,394 INFO Train. Iter 198200 : Commit. 1.16883 PPL. 399.57 Recons. 0.00746 +2024-03-17 04:28:49,836 INFO Train. Iter 198400 : Commit. 1.13961 PPL. 398.63 Recons. 0.00707 +2024-03-17 04:29:12,640 INFO Train. Iter 198600 : Commit. 1.12773 PPL. 399.65 Recons. 0.00671 +2024-03-17 04:29:35,657 INFO Train. Iter 198800 : Commit. 1.11859 PPL. 399.30 Recons. 0.00662 +2024-03-17 04:29:58,596 INFO Train. Iter 199000 : Commit. 1.16112 PPL. 397.32 Recons. 0.00791 +2024-03-17 04:30:21,369 INFO Train. Iter 199200 : Commit. 1.12116 PPL. 399.47 Recons. 0.00634 +2024-03-17 04:30:43,816 INFO Train. Iter 199400 : Commit. 1.12430 PPL. 399.01 Recons. 0.00678 +2024-03-17 04:31:06,690 INFO Train. Iter 199600 : Commit. 1.17423 PPL. 398.56 Recons. 0.00772 +2024-03-17 04:31:29,890 INFO Train. Iter 199800 : Commit. 1.14285 PPL. 398.02 Recons. 0.00725 +2024-03-17 04:31:51,918 INFO Train. Iter 200000 : Commit. 1.14741 PPL. 399.12 Recons. 0.00693 +2024-03-17 04:32:13,697 INFO Train. Iter 200200 : Commit. 1.12865 PPL. 400.56 Recons. 0.00636 +2024-03-17 04:32:34,840 INFO Train. Iter 200400 : Commit. 1.08806 PPL. 399.99 Recons. 0.00584 +2024-03-17 04:32:56,714 INFO Train. Iter 200600 : Commit. 1.08628 PPL. 400.79 Recons. 0.00566 +2024-03-17 04:33:19,069 INFO Train. Iter 200800 : Commit. 1.08548 PPL. 399.43 Recons. 0.00584 +2024-03-17 04:33:41,132 INFO Train. Iter 201000 : Commit. 1.06567 PPL. 400.06 Recons. 0.00555 +2024-03-17 04:34:02,798 INFO Train. Iter 201200 : Commit. 1.07680 PPL. 400.09 Recons. 0.00563 +2024-03-17 04:34:24,330 INFO Train. Iter 201400 : Commit. 1.08201 PPL. 399.58 Recons. 0.00581 +2024-03-17 04:34:45,547 INFO Train. Iter 201600 : Commit. 1.06188 PPL. 399.67 Recons. 0.00542 +2024-03-17 04:35:06,907 INFO Train. Iter 201800 : Commit. 1.07905 PPL. 400.33 Recons. 0.00572 +2024-03-17 04:35:27,433 INFO Train. Iter 202000 : Commit. 1.07023 PPL. 399.02 Recons. 0.00570 +2024-03-17 04:35:49,188 INFO Train. Iter 202200 : Commit. 1.07225 PPL. 399.04 Recons. 0.00565 +2024-03-17 04:36:10,503 INFO Train. Iter 202400 : Commit. 1.07652 PPL. 398.54 Recons. 0.00574 +2024-03-17 04:36:31,499 INFO Train. Iter 202600 : Commit. 1.09200 PPL. 399.24 Recons. 0.00598 +2024-03-17 04:36:52,929 INFO Train. Iter 202800 : Commit. 1.06884 PPL. 398.45 Recons. 0.00565 +2024-03-17 04:37:14,099 INFO Train. Iter 203000 : Commit. 1.07852 PPL. 398.37 Recons. 0.00581 +2024-03-17 04:37:35,495 INFO Train. Iter 203200 : Commit. 1.06901 PPL. 399.27 Recons. 0.00545 +2024-03-17 04:37:56,967 INFO Train. Iter 203400 : Commit. 1.06475 PPL. 399.55 Recons. 0.00540 +2024-03-17 04:38:18,462 INFO Train. Iter 203600 : Commit. 1.07001 PPL. 399.10 Recons. 0.00554 +2024-03-17 04:38:39,875 INFO Train. Iter 203800 : Commit. 1.08916 PPL. 398.61 Recons. 0.00593 +2024-03-17 04:39:00,879 INFO Train. Iter 204000 : Commit. 1.06815 PPL. 398.38 Recons. 0.00562 +2024-03-17 04:39:22,460 INFO Train. Iter 204200 : Commit. 1.07118 PPL. 398.58 Recons. 0.00551 +2024-03-17 04:39:44,135 INFO Train. Iter 204400 : Commit. 1.07643 PPL. 398.51 Recons. 0.00561 +2024-03-17 04:40:05,542 INFO Train. Iter 204600 : Commit. 1.07444 PPL. 398.02 Recons. 0.00566 +2024-03-17 04:40:26,882 INFO Train. Iter 204800 : Commit. 1.07860 PPL. 398.59 Recons. 0.00562 +2024-03-17 04:40:48,563 INFO Train. Iter 205000 : Commit. 1.09342 PPL. 398.08 Recons. 0.00606 +2024-03-17 04:41:09,993 INFO Train. Iter 205200 : Commit. 1.08077 PPL. 397.71 Recons. 0.00579 +2024-03-17 04:41:31,181 INFO Train. Iter 205400 : Commit. 1.07366 PPL. 397.79 Recons. 0.00561 +2024-03-17 04:41:52,542 INFO Train. Iter 205600 : Commit. 1.07440 PPL. 397.82 Recons. 0.00558 +2024-03-17 04:42:13,414 INFO Train. Iter 205800 : Commit. 1.05809 PPL. 398.26 Recons. 0.00529 +2024-03-17 04:42:35,087 INFO Train. Iter 206000 : Commit. 1.09630 PPL. 396.66 Recons. 0.00618 +2024-03-17 04:42:56,303 INFO Train. Iter 206200 : Commit. 1.08930 PPL. 398.83 Recons. 0.00570 +2024-03-17 04:43:17,666 INFO Train. Iter 206400 : Commit. 1.07269 PPL. 398.36 Recons. 0.00540 +2024-03-17 04:43:38,723 INFO Train. Iter 206600 : Commit. 1.09064 PPL. 398.56 Recons. 0.00580 +2024-03-17 04:43:59,933 INFO Train. Iter 206800 : Commit. 1.08326 PPL. 397.88 Recons. 0.00575 +2024-03-17 04:44:21,372 INFO Train. Iter 207000 : Commit. 1.08610 PPL. 397.49 Recons. 0.00573 +2024-03-17 04:44:42,865 INFO Train. Iter 207200 : Commit. 1.09171 PPL. 397.49 Recons. 0.00592 +2024-03-17 04:45:03,929 INFO Train. Iter 207400 : Commit. 1.07072 PPL. 398.54 Recons. 0.00538 +2024-03-17 04:45:25,355 INFO Train. Iter 207600 : Commit. 1.07578 PPL. 398.40 Recons. 0.00556 +2024-03-17 04:45:45,863 INFO Train. Iter 207800 : Commit. 1.08684 PPL. 397.63 Recons. 0.00577 +2024-03-17 04:46:07,442 INFO Train. Iter 208000 : Commit. 1.08854 PPL. 398.11 Recons. 0.00580 +2024-03-17 04:46:29,164 INFO Train. Iter 208200 : Commit. 1.06243 PPL. 397.62 Recons. 0.00538 +2024-03-17 04:46:50,515 INFO Train. Iter 208400 : Commit. 1.08298 PPL. 398.46 Recons. 0.00563 +2024-03-17 04:47:12,069 INFO Train. Iter 208600 : Commit. 1.08366 PPL. 398.31 Recons. 0.00560 +2024-03-17 04:47:33,534 INFO Train. Iter 208800 : Commit. 1.09208 PPL. 397.74 Recons. 0.00590 +2024-03-17 04:47:54,725 INFO Train. Iter 209000 : Commit. 1.06453 PPL. 397.51 Recons. 0.00540 +2024-03-17 04:48:15,861 INFO Train. Iter 209200 : Commit. 1.05837 PPL. 398.63 Recons. 0.00515 +2024-03-17 04:48:37,255 INFO Train. Iter 209400 : Commit. 1.08540 PPL. 397.25 Recons. 0.00584 +2024-03-17 04:48:58,890 INFO Train. Iter 209600 : Commit. 1.07403 PPL. 397.89 Recons. 0.00549 +2024-03-17 04:49:19,217 INFO Train. Iter 209800 : Commit. 1.09406 PPL. 397.28 Recons. 0.00591 +2024-03-17 04:49:40,787 INFO Train. Iter 210000 : Commit. 1.06311 PPL. 397.30 Recons. 0.00533 +2024-03-17 04:50:02,560 INFO Train. Iter 210200 : Commit. 1.08496 PPL. 397.16 Recons. 0.00577 +2024-03-17 04:50:24,049 INFO Train. Iter 210400 : Commit. 1.07329 PPL. 397.80 Recons. 0.00546 +2024-03-17 04:50:45,207 INFO Train. Iter 210600 : Commit. 1.09869 PPL. 397.49 Recons. 0.00593 +2024-03-17 04:51:06,415 INFO Train. Iter 210800 : Commit. 1.06492 PPL. 397.35 Recons. 0.00540 +2024-03-17 04:51:27,731 INFO Train. Iter 211000 : Commit. 1.06995 PPL. 397.71 Recons. 0.00537 +2024-03-17 04:51:49,288 INFO Train. Iter 211200 : Commit. 1.06845 PPL. 397.27 Recons. 0.00539 +2024-03-17 04:52:11,574 INFO Train. Iter 211400 : Commit. 1.08298 PPL. 397.26 Recons. 0.00571 +2024-03-17 04:52:32,022 INFO Train. Iter 211600 : Commit. 1.08363 PPL. 396.42 Recons. 0.00578 +2024-03-17 04:52:53,106 INFO Train. Iter 211800 : Commit. 1.05742 PPL. 396.67 Recons. 0.00525 +2024-03-17 04:53:15,275 INFO Train. Iter 212000 : Commit. 1.08855 PPL. 396.52 Recons. 0.00590 +2024-03-17 04:53:36,470 INFO Train. Iter 212200 : Commit. 1.06900 PPL. 397.94 Recons. 0.00531 +2024-03-17 04:53:57,976 INFO Train. Iter 212400 : Commit. 1.08372 PPL. 397.61 Recons. 0.00557 +2024-03-17 04:54:19,551 INFO Train. Iter 212600 : Commit. 1.05636 PPL. 397.24 Recons. 0.00513 +2024-03-17 04:54:41,166 INFO Train. Iter 212800 : Commit. 1.10203 PPL. 397.82 Recons. 0.00593 +2024-03-17 04:55:02,698 INFO Train. Iter 213000 : Commit. 1.07870 PPL. 396.88 Recons. 0.00560 +2024-03-17 04:55:24,007 INFO Train. Iter 213200 : Commit. 1.06562 PPL. 396.91 Recons. 0.00532 +2024-03-17 04:55:45,652 INFO Train. Iter 213400 : Commit. 1.07531 PPL. 396.68 Recons. 0.00562 +2024-03-17 04:56:06,392 INFO Train. Iter 213600 : Commit. 1.08640 PPL. 396.38 Recons. 0.00582 +2024-03-17 04:56:28,599 INFO Train. Iter 213800 : Commit. 1.06069 PPL. 396.16 Recons. 0.00532 +2024-03-17 04:56:50,209 INFO Train. Iter 214000 : Commit. 1.06759 PPL. 397.25 Recons. 0.00533 +2024-03-17 04:57:11,796 INFO Train. Iter 214200 : Commit. 1.07124 PPL. 396.89 Recons. 0.00538 +2024-03-17 04:57:32,994 INFO Train. Iter 214400 : Commit. 1.08556 PPL. 397.29 Recons. 0.00562 +2024-03-17 04:57:54,212 INFO Train. Iter 214600 : Commit. 1.08228 PPL. 396.55 Recons. 0.00571 +2024-03-17 04:58:15,863 INFO Train. Iter 214800 : Commit. 1.09135 PPL. 397.58 Recons. 0.00575 +2024-03-17 04:58:37,483 INFO Train. Iter 215000 : Commit. 1.06409 PPL. 397.03 Recons. 0.00525 +2024-03-17 04:58:59,210 INFO Train. Iter 215200 : Commit. 1.06820 PPL. 396.76 Recons. 0.00534 +2024-03-17 04:59:21,148 INFO Train. Iter 215400 : Commit. 1.07516 PPL. 397.12 Recons. 0.00550 +2024-03-17 04:59:42,106 INFO Train. Iter 215600 : Commit. 1.07426 PPL. 396.40 Recons. 0.00548 +2024-03-17 05:00:03,555 INFO Train. Iter 215800 : Commit. 1.08884 PPL. 396.83 Recons. 0.00570 +2024-03-17 05:00:24,976 INFO Train. Iter 216000 : Commit. 1.08381 PPL. 395.87 Recons. 0.00571 +2024-03-17 05:00:46,765 INFO Train. Iter 216200 : Commit. 1.07499 PPL. 396.30 Recons. 0.00549 +2024-03-17 05:01:08,273 INFO Train. Iter 216400 : Commit. 1.08156 PPL. 396.49 Recons. 0.00566 +2024-03-17 05:01:30,792 INFO Train. Iter 216600 : Commit. 1.07538 PPL. 396.37 Recons. 0.00556 +2024-03-17 05:01:52,585 INFO Train. Iter 216800 : Commit. 1.06853 PPL. 395.36 Recons. 0.00555 +2024-03-17 05:02:13,898 INFO Train. Iter 217000 : Commit. 1.08969 PPL. 397.16 Recons. 0.00570 +2024-03-17 05:02:35,224 INFO Train. Iter 217200 : Commit. 1.07599 PPL. 395.83 Recons. 0.00557 +2024-03-17 05:02:55,703 INFO Train. Iter 217400 : Commit. 1.07629 PPL. 395.87 Recons. 0.00562 +2024-03-17 05:03:17,291 INFO Train. Iter 217600 : Commit. 1.08147 PPL. 397.16 Recons. 0.00555 +2024-03-17 05:03:38,593 INFO Train. Iter 217800 : Commit. 1.07687 PPL. 396.29 Recons. 0.00556 +2024-03-17 05:04:00,081 INFO Train. Iter 218000 : Commit. 1.07618 PPL. 396.09 Recons. 0.00560 +2024-03-17 05:04:21,591 INFO Train. Iter 218200 : Commit. 1.07128 PPL. 396.39 Recons. 0.00552 +2024-03-17 05:04:43,002 INFO Train. Iter 218400 : Commit. 1.08977 PPL. 396.69 Recons. 0.00568 +2024-03-17 05:05:04,930 INFO Train. Iter 218600 : Commit. 1.08912 PPL. 396.85 Recons. 0.00574 +2024-03-17 05:05:26,574 INFO Train. Iter 218800 : Commit. 1.09410 PPL. 395.59 Recons. 0.00597 +2024-03-17 05:05:48,888 INFO Train. Iter 219000 : Commit. 1.06543 PPL. 396.16 Recons. 0.00534 +2024-03-17 05:06:10,534 INFO Train. Iter 219200 : Commit. 1.06843 PPL. 396.50 Recons. 0.00537 +2024-03-17 05:06:30,678 INFO Train. Iter 219400 : Commit. 1.06902 PPL. 396.05 Recons. 0.00547 +2024-03-17 05:06:52,173 INFO Train. Iter 219600 : Commit. 1.06809 PPL. 396.12 Recons. 0.00536 +2024-03-17 05:07:13,444 INFO Train. Iter 219800 : Commit. 1.07237 PPL. 396.83 Recons. 0.00547 +2024-03-17 05:07:35,040 INFO Train. Iter 220000 : Commit. 1.07094 PPL. 397.38 Recons. 0.00542 +2024-03-17 05:07:56,972 INFO Train. Iter 220200 : Commit. 1.06041 PPL. 396.60 Recons. 0.00523 +2024-03-17 05:08:19,105 INFO Train. Iter 220400 : Commit. 1.06995 PPL. 396.34 Recons. 0.00546 +2024-03-17 05:08:40,702 INFO Train. Iter 220600 : Commit. 1.08364 PPL. 396.12 Recons. 0.00578 +2024-03-17 05:09:02,465 INFO Train. Iter 220800 : Commit. 1.07894 PPL. 397.58 Recons. 0.00548 +2024-03-17 05:09:23,899 INFO Train. Iter 221000 : Commit. 1.06517 PPL. 396.86 Recons. 0.00531 +2024-03-17 05:09:45,470 INFO Train. Iter 221200 : Commit. 1.07049 PPL. 396.69 Recons. 0.00542 +2024-03-17 05:10:06,019 INFO Train. Iter 221400 : Commit. 1.06765 PPL. 396.56 Recons. 0.00539 +2024-03-17 05:10:27,623 INFO Train. Iter 221600 : Commit. 1.05872 PPL. 395.58 Recons. 0.00540 +2024-03-17 05:10:49,047 INFO Train. Iter 221800 : Commit. 1.06508 PPL. 396.07 Recons. 0.00543 +2024-03-17 05:11:10,317 INFO Train. Iter 222000 : Commit. 1.07916 PPL. 396.55 Recons. 0.00560 +2024-03-17 05:11:31,762 INFO Train. Iter 222200 : Commit. 1.07256 PPL. 396.81 Recons. 0.00538 +2024-03-17 05:11:53,938 INFO Train. Iter 222400 : Commit. 1.05843 PPL. 396.63 Recons. 0.00521 +2024-03-17 05:12:15,470 INFO Train. Iter 222600 : Commit. 1.06797 PPL. 397.40 Recons. 0.00533 +2024-03-17 05:12:37,032 INFO Train. Iter 222800 : Commit. 1.05889 PPL. 396.01 Recons. 0.00528 +2024-03-17 05:12:58,198 INFO Train. Iter 223000 : Commit. 1.05904 PPL. 397.24 Recons. 0.00521 +2024-03-17 05:13:18,639 INFO Train. Iter 223200 : Commit. 1.07895 PPL. 397.19 Recons. 0.00551 +2024-03-17 05:13:40,609 INFO Train. Iter 223400 : Commit. 1.05100 PPL. 396.46 Recons. 0.00512 +2024-03-17 05:14:02,520 INFO Train. Iter 223600 : Commit. 1.06465 PPL. 396.74 Recons. 0.00536 +2024-03-17 05:14:24,142 INFO Train. Iter 223800 : Commit. 1.06770 PPL. 396.71 Recons. 0.00540 +2024-03-17 05:14:45,663 INFO Train. Iter 224000 : Commit. 1.07760 PPL. 395.88 Recons. 0.00568 +2024-03-17 05:15:07,003 INFO Train. Iter 224200 : Commit. 1.07754 PPL. 396.73 Recons. 0.00557 +2024-03-17 05:15:27,885 INFO Train. Iter 224400 : Commit. 1.08361 PPL. 395.61 Recons. 0.00584 +2024-03-17 05:15:48,938 INFO Train. Iter 224600 : Commit. 1.05669 PPL. 396.81 Recons. 0.00508 +2024-03-17 05:16:10,510 INFO Train. Iter 224800 : Commit. 1.06515 PPL. 396.45 Recons. 0.00541 +2024-03-17 05:16:31,966 INFO Train. Iter 225000 : Commit. 1.07448 PPL. 396.86 Recons. 0.00546 +2024-03-17 05:16:52,372 INFO Train. Iter 225200 : Commit. 1.07359 PPL. 395.82 Recons. 0.00560 +2024-03-17 05:17:13,780 INFO Train. Iter 225400 : Commit. 1.06434 PPL. 397.22 Recons. 0.00529 +2024-03-17 05:17:35,267 INFO Train. Iter 225600 : Commit. 1.06840 PPL. 396.85 Recons. 0.00541 +2024-03-17 05:17:56,861 INFO Train. Iter 225800 : Commit. 1.04806 PPL. 396.80 Recons. 0.00505 +2024-03-17 05:18:18,312 INFO Train. Iter 226000 : Commit. 1.06989 PPL. 396.36 Recons. 0.00551 +2024-03-17 05:18:39,780 INFO Train. Iter 226200 : Commit. 1.06296 PPL. 395.83 Recons. 0.00541 +2024-03-17 05:19:01,126 INFO Train. Iter 226400 : Commit. 1.09362 PPL. 397.11 Recons. 0.00578 +2024-03-17 05:19:22,829 INFO Train. Iter 226600 : Commit. 1.08465 PPL. 396.16 Recons. 0.00578 +2024-03-17 05:19:44,533 INFO Train. Iter 226800 : Commit. 1.07714 PPL. 395.92 Recons. 0.00570 +2024-03-17 05:20:05,830 INFO Train. Iter 227000 : Commit. 1.08126 PPL. 396.20 Recons. 0.00568 +2024-03-17 05:20:26,060 INFO Train. Iter 227200 : Commit. 1.07568 PPL. 395.73 Recons. 0.00567 +2024-03-17 05:20:47,466 INFO Train. Iter 227400 : Commit. 1.06834 PPL. 395.59 Recons. 0.00548 +2024-03-17 05:21:09,363 INFO Train. Iter 227600 : Commit. 1.06335 PPL. 396.81 Recons. 0.00529 +2024-03-17 05:21:31,491 INFO Train. Iter 227800 : Commit. 1.06155 PPL. 396.39 Recons. 0.00525 +2024-03-17 05:21:53,316 INFO Train. Iter 228000 : Commit. 1.07813 PPL. 395.47 Recons. 0.00570 +2024-03-17 05:22:15,195 INFO Train. Iter 228200 : Commit. 1.06202 PPL. 396.24 Recons. 0.00526 +2024-03-17 05:22:36,117 INFO Train. Iter 228400 : Commit. 1.08096 PPL. 395.19 Recons. 0.00581 +2024-03-17 05:22:57,714 INFO Train. Iter 228600 : Commit. 1.06188 PPL. 395.20 Recons. 0.00538 +2024-03-17 05:23:19,080 INFO Train. Iter 228800 : Commit. 1.06034 PPL. 395.69 Recons. 0.00533 +2024-03-17 05:23:39,520 INFO Train. Iter 229000 : Commit. 1.07250 PPL. 395.09 Recons. 0.00563 +2024-03-17 05:24:01,133 INFO Train. Iter 229200 : Commit. 1.06192 PPL. 395.70 Recons. 0.00535 +2024-03-17 05:24:22,460 INFO Train. Iter 229400 : Commit. 1.06897 PPL. 396.29 Recons. 0.00548 +2024-03-17 05:24:43,901 INFO Train. Iter 229600 : Commit. 1.06731 PPL. 395.68 Recons. 0.00547 +2024-03-17 05:25:04,958 INFO Train. Iter 229800 : Commit. 1.07491 PPL. 396.05 Recons. 0.00558 +2024-03-17 05:25:26,583 INFO Train. Iter 230000 : Commit. 1.05777 PPL. 394.73 Recons. 0.00534 +2024-03-17 05:25:48,180 INFO Train. Iter 230200 : Commit. 1.06207 PPL. 395.71 Recons. 0.00536 +2024-03-17 05:26:09,547 INFO Train. Iter 230400 : Commit. 1.05338 PPL. 395.90 Recons. 0.00514 +2024-03-17 05:26:30,720 INFO Train. Iter 230600 : Commit. 1.08170 PPL. 396.22 Recons. 0.00571 +2024-03-17 05:26:52,036 INFO Train. Iter 230800 : Commit. 1.07314 PPL. 395.61 Recons. 0.00547 +2024-03-17 05:27:12,078 INFO Train. Iter 231000 : Commit. 1.06477 PPL. 395.12 Recons. 0.00542 +2024-03-17 05:27:33,615 INFO Train. Iter 231200 : Commit. 1.07802 PPL. 395.25 Recons. 0.00570 +2024-03-17 05:27:55,166 INFO Train. Iter 231400 : Commit. 1.05802 PPL. 394.95 Recons. 0.00531 +2024-03-17 05:28:16,364 INFO Train. Iter 231600 : Commit. 1.08070 PPL. 395.66 Recons. 0.00566 +2024-03-17 05:28:37,518 INFO Train. Iter 231800 : Commit. 1.07030 PPL. 394.54 Recons. 0.00557 +2024-03-17 05:28:58,565 INFO Train. Iter 232000 : Commit. 1.06354 PPL. 395.83 Recons. 0.00534 +2024-03-17 05:29:19,359 INFO Train. Iter 232200 : Commit. 1.08978 PPL. 395.04 Recons. 0.00591 +2024-03-17 05:29:40,808 INFO Train. Iter 232400 : Commit. 1.07204 PPL. 394.35 Recons. 0.00559 +2024-03-17 05:30:02,338 INFO Train. Iter 232600 : Commit. 1.06775 PPL. 395.52 Recons. 0.00541 +2024-03-17 05:30:23,801 INFO Train. Iter 232800 : Commit. 1.07490 PPL. 395.31 Recons. 0.00562 +2024-03-17 05:30:44,203 INFO Train. Iter 233000 : Commit. 1.06262 PPL. 394.93 Recons. 0.00535 +2024-03-17 05:31:05,622 INFO Train. Iter 233200 : Commit. 1.05175 PPL. 395.44 Recons. 0.00519 +2024-03-17 05:31:27,189 INFO Train. Iter 233400 : Commit. 1.06962 PPL. 395.81 Recons. 0.00547 +2024-03-17 05:31:49,145 INFO Train. Iter 233600 : Commit. 1.06815 PPL. 394.71 Recons. 0.00547 +2024-03-17 05:32:11,168 INFO Train. Iter 233800 : Commit. 1.07108 PPL. 394.63 Recons. 0.00559 +2024-03-17 05:32:33,224 INFO Train. Iter 234000 : Commit. 1.05030 PPL. 394.68 Recons. 0.00517 +2024-03-17 05:32:55,037 INFO Train. Iter 234200 : Commit. 1.06806 PPL. 395.42 Recons. 0.00547 +2024-03-17 05:33:16,291 INFO Train. Iter 234400 : Commit. 1.05059 PPL. 394.24 Recons. 0.00524 +2024-03-17 05:33:37,919 INFO Train. Iter 234600 : Commit. 1.04686 PPL. 394.90 Recons. 0.00513 +2024-03-17 05:33:58,235 INFO Train. Iter 234800 : Commit. 1.07318 PPL. 394.91 Recons. 0.00568 +2024-03-17 05:34:20,624 INFO Train. Iter 235000 : Commit. 1.08091 PPL. 395.54 Recons. 0.00568 +2024-03-17 05:34:42,507 INFO Train. Iter 235200 : Commit. 1.04810 PPL. 394.90 Recons. 0.00516 +2024-03-17 05:35:04,218 INFO Train. Iter 235400 : Commit. 1.06595 PPL. 395.33 Recons. 0.00544 +2024-03-17 05:35:25,659 INFO Train. Iter 235600 : Commit. 1.07677 PPL. 395.22 Recons. 0.00568 +2024-03-17 05:35:47,486 INFO Train. Iter 235800 : Commit. 1.07131 PPL. 394.57 Recons. 0.00564 +2024-03-17 05:36:09,625 INFO Train. Iter 236000 : Commit. 1.05694 PPL. 395.80 Recons. 0.00524 +2024-03-17 05:36:30,915 INFO Train. Iter 236200 : Commit. 1.05553 PPL. 394.62 Recons. 0.00534 +2024-03-17 05:36:52,634 INFO Train. Iter 236400 : Commit. 1.07368 PPL. 395.02 Recons. 0.00559 +2024-03-17 05:37:14,045 INFO Train. Iter 236600 : Commit. 1.06672 PPL. 394.90 Recons. 0.00551 +2024-03-17 05:37:34,492 INFO Train. Iter 236800 : Commit. 1.06260 PPL. 395.61 Recons. 0.00536 +2024-03-17 05:37:55,647 INFO Train. Iter 237000 : Commit. 1.06828 PPL. 395.56 Recons. 0.00547 +2024-03-17 05:38:16,836 INFO Train. Iter 237200 : Commit. 1.08585 PPL. 394.54 Recons. 0.00597 +2024-03-17 05:38:38,493 INFO Train. Iter 237400 : Commit. 1.06302 PPL. 394.67 Recons. 0.00545 +2024-03-17 05:38:59,649 INFO Train. Iter 237600 : Commit. 1.06711 PPL. 394.01 Recons. 0.00565 +2024-03-17 05:39:22,197 INFO Train. Iter 237800 : Commit. 1.05236 PPL. 395.20 Recons. 0.00522 +2024-03-17 05:39:43,609 INFO Train. Iter 238000 : Commit. 1.06781 PPL. 395.99 Recons. 0.00541 +2024-03-17 05:40:05,580 INFO Train. Iter 238200 : Commit. 1.04844 PPL. 394.94 Recons. 0.00518 +2024-03-17 05:40:27,123 INFO Train. Iter 238400 : Commit. 1.06713 PPL. 395.26 Recons. 0.00546 +2024-03-17 05:40:48,449 INFO Train. Iter 238600 : Commit. 1.06108 PPL. 395.06 Recons. 0.00534 +2024-03-17 05:41:09,329 INFO Train. Iter 238800 : Commit. 1.06917 PPL. 394.93 Recons. 0.00557 +2024-03-17 05:41:31,041 INFO Train. Iter 239000 : Commit. 1.06019 PPL. 394.58 Recons. 0.00545 +2024-03-17 05:41:52,860 INFO Train. Iter 239200 : Commit. 1.05935 PPL. 395.23 Recons. 0.00533 +2024-03-17 05:42:14,576 INFO Train. Iter 239400 : Commit. 1.05801 PPL. 395.08 Recons. 0.00536 +2024-03-17 05:42:36,499 INFO Train. Iter 239600 : Commit. 1.05620 PPL. 395.46 Recons. 0.00526 +2024-03-17 05:42:57,931 INFO Train. Iter 239800 : Commit. 1.06543 PPL. 394.75 Recons. 0.00549 +2024-03-17 05:43:19,138 INFO Train. Iter 240000 : Commit. 1.06331 PPL. 394.44 Recons. 0.00545 +2024-03-17 05:43:40,786 INFO Train. Iter 240200 : Commit. 1.05448 PPL. 394.43 Recons. 0.00533 +2024-03-17 05:44:02,136 INFO Train. Iter 240400 : Commit. 1.06053 PPL. 394.41 Recons. 0.00548 +2024-03-17 05:44:22,047 INFO Train. Iter 240600 : Commit. 1.07995 PPL. 394.59 Recons. 0.00573 +2024-03-17 05:44:43,539 INFO Train. Iter 240800 : Commit. 1.06201 PPL. 394.18 Recons. 0.00547 +2024-03-17 05:45:04,691 INFO Train. Iter 241000 : Commit. 1.05075 PPL. 394.48 Recons. 0.00519 +2024-03-17 05:45:26,315 INFO Train. Iter 241200 : Commit. 1.05379 PPL. 394.37 Recons. 0.00531 +2024-03-17 05:45:47,789 INFO Train. Iter 241400 : Commit. 1.05978 PPL. 394.49 Recons. 0.00536 +2024-03-17 05:46:09,175 INFO Train. Iter 241600 : Commit. 1.06163 PPL. 393.66 Recons. 0.00550 +2024-03-17 05:46:30,609 INFO Train. Iter 241800 : Commit. 1.06027 PPL. 393.54 Recons. 0.00551 +2024-03-17 05:46:51,916 INFO Train. Iter 242000 : Commit. 1.05869 PPL. 394.08 Recons. 0.00540 +2024-03-17 05:47:13,009 INFO Train. Iter 242200 : Commit. 1.05933 PPL. 394.21 Recons. 0.00538 +2024-03-17 05:47:35,125 INFO Train. Iter 242400 : Commit. 1.05680 PPL. 394.98 Recons. 0.00524 +2024-03-17 05:47:55,973 INFO Train. Iter 242600 : Commit. 1.06640 PPL. 393.09 Recons. 0.00565 +2024-03-17 05:48:17,406 INFO Train. Iter 242800 : Commit. 1.07340 PPL. 394.62 Recons. 0.00558 +2024-03-17 05:48:39,150 INFO Train. Iter 243000 : Commit. 1.08450 PPL. 394.38 Recons. 0.00576 +2024-03-17 05:49:00,504 INFO Train. Iter 243200 : Commit. 1.07908 PPL. 394.12 Recons. 0.00574 +2024-03-17 05:49:21,440 INFO Train. Iter 243400 : Commit. 1.06629 PPL. 393.53 Recons. 0.00560 +2024-03-17 05:49:42,770 INFO Train. Iter 243600 : Commit. 1.04913 PPL. 393.08 Recons. 0.00531 +2024-03-17 05:50:04,169 INFO Train. Iter 243800 : Commit. 1.06727 PPL. 394.33 Recons. 0.00552 +2024-03-17 05:50:25,901 INFO Train. Iter 244000 : Commit. 1.07009 PPL. 394.33 Recons. 0.00557 +2024-03-17 05:50:47,955 INFO Train. Iter 244200 : Commit. 1.05146 PPL. 393.16 Recons. 0.00530 +2024-03-17 05:51:08,898 INFO Train. Iter 244400 : Commit. 1.07591 PPL. 394.31 Recons. 0.00563 +2024-03-17 05:51:29,766 INFO Train. Iter 244600 : Commit. 1.05313 PPL. 394.02 Recons. 0.00525 +2024-03-17 05:51:51,143 INFO Train. Iter 244800 : Commit. 1.05686 PPL. 393.51 Recons. 0.00538 +2024-03-17 05:52:12,605 INFO Train. Iter 245000 : Commit. 1.06529 PPL. 393.50 Recons. 0.00549 +2024-03-17 05:52:34,123 INFO Train. Iter 245200 : Commit. 1.05727 PPL. 393.95 Recons. 0.00531 +2024-03-17 05:52:56,040 INFO Train. Iter 245400 : Commit. 1.05735 PPL. 394.05 Recons. 0.00531 +2024-03-17 05:53:17,753 INFO Train. Iter 245600 : Commit. 1.05605 PPL. 394.61 Recons. 0.00525 +2024-03-17 05:53:39,616 INFO Train. Iter 245800 : Commit. 1.04894 PPL. 394.28 Recons. 0.00513 +2024-03-17 05:54:01,993 INFO Train. Iter 246000 : Commit. 1.05343 PPL. 393.60 Recons. 0.00529 +2024-03-17 05:54:23,243 INFO Train. Iter 246200 : Commit. 1.05708 PPL. 394.41 Recons. 0.00529 +2024-03-17 05:54:43,437 INFO Train. Iter 246400 : Commit. 1.06728 PPL. 394.11 Recons. 0.00553 +2024-03-17 05:55:04,883 INFO Train. Iter 246600 : Commit. 1.04345 PPL. 393.19 Recons. 0.00508 +2024-03-17 05:55:26,260 INFO Train. Iter 246800 : Commit. 1.05921 PPL. 392.72 Recons. 0.00550 +2024-03-17 05:55:47,726 INFO Train. Iter 247000 : Commit. 1.05423 PPL. 392.99 Recons. 0.00532 +2024-03-17 05:56:09,613 INFO Train. Iter 247200 : Commit. 1.06728 PPL. 392.67 Recons. 0.00565 +2024-03-17 05:56:31,701 INFO Train. Iter 247400 : Commit. 1.08126 PPL. 393.77 Recons. 0.00576 +2024-03-17 05:56:53,494 INFO Train. Iter 247600 : Commit. 1.05534 PPL. 393.20 Recons. 0.00533 +2024-03-17 05:57:14,985 INFO Train. Iter 247800 : Commit. 1.07932 PPL. 393.29 Recons. 0.00582 +2024-03-17 05:57:36,561 INFO Train. Iter 248000 : Commit. 1.04589 PPL. 393.33 Recons. 0.00518 +2024-03-17 05:57:58,144 INFO Train. Iter 248200 : Commit. 1.06296 PPL. 392.99 Recons. 0.00552 +2024-03-17 05:58:18,481 INFO Train. Iter 248400 : Commit. 1.05758 PPL. 393.07 Recons. 0.00540 +2024-03-17 05:58:39,953 INFO Train. Iter 248600 : Commit. 1.06232 PPL. 394.38 Recons. 0.00534 +2024-03-17 05:59:01,705 INFO Train. Iter 248800 : Commit. 1.06655 PPL. 394.07 Recons. 0.00546 +2024-03-17 05:59:22,717 INFO Train. Iter 249000 : Commit. 1.06403 PPL. 394.29 Recons. 0.00541 +2024-03-17 05:59:44,022 INFO Train. Iter 249200 : Commit. 1.06696 PPL. 393.14 Recons. 0.00556 +2024-03-17 06:00:05,614 INFO Train. Iter 249400 : Commit. 1.05323 PPL. 393.10 Recons. 0.00530 +2024-03-17 06:00:27,333 INFO Train. Iter 249600 : Commit. 1.06512 PPL. 393.60 Recons. 0.00551 +2024-03-17 06:00:49,151 INFO Train. Iter 249800 : Commit. 1.06509 PPL. 392.69 Recons. 0.00554 +2024-03-17 06:01:10,509 INFO Train. Iter 250000 : Commit. 1.05805 PPL. 392.22 Recons. 0.00547 +2024-03-17 06:01:31,970 INFO Train. Iter 250200 : Commit. 1.06451 PPL. 392.28 Recons. 0.00563 +2024-03-17 06:01:52,233 INFO Train. Iter 250400 : Commit. 1.03331 PPL. 392.52 Recons. 0.00497 +2024-03-17 06:02:13,546 INFO Train. Iter 250600 : Commit. 1.05800 PPL. 391.38 Recons. 0.00551 +2024-03-17 06:02:34,706 INFO Train. Iter 250800 : Commit. 1.06173 PPL. 393.16 Recons. 0.00541 +2024-03-17 06:02:56,364 INFO Train. Iter 251000 : Commit. 1.06712 PPL. 393.76 Recons. 0.00540 +2024-03-17 06:03:18,569 INFO Train. Iter 251200 : Commit. 1.06322 PPL. 393.12 Recons. 0.00546 +2024-03-17 06:03:40,186 INFO Train. Iter 251400 : Commit. 1.05481 PPL. 391.55 Recons. 0.00543 +2024-03-17 06:04:02,511 INFO Train. Iter 251600 : Commit. 1.06817 PPL. 393.13 Recons. 0.00550 +2024-03-17 06:04:23,722 INFO Train. Iter 251800 : Commit. 1.06649 PPL. 392.88 Recons. 0.00554 +2024-03-17 06:04:44,963 INFO Train. Iter 252000 : Commit. 1.04835 PPL. 392.23 Recons. 0.00523 +2024-03-17 06:05:05,445 INFO Train. Iter 252200 : Commit. 1.04364 PPL. 393.58 Recons. 0.00498 +2024-03-17 06:05:26,538 INFO Train. Iter 252400 : Commit. 1.06056 PPL. 392.31 Recons. 0.00554 +2024-03-17 06:05:47,810 INFO Train. Iter 252600 : Commit. 1.07928 PPL. 392.96 Recons. 0.00575 +2024-03-17 06:06:09,105 INFO Train. Iter 252800 : Commit. 1.07516 PPL. 393.34 Recons. 0.00566 +2024-03-17 06:06:30,695 INFO Train. Iter 253000 : Commit. 1.06073 PPL. 392.81 Recons. 0.00543 +2024-03-17 06:06:52,087 INFO Train. Iter 253200 : Commit. 1.04951 PPL. 393.06 Recons. 0.00520 +2024-03-17 06:07:13,722 INFO Train. Iter 253400 : Commit. 1.04021 PPL. 392.27 Recons. 0.00513 +2024-03-17 06:07:35,501 INFO Train. Iter 253600 : Commit. 1.06837 PPL. 392.38 Recons. 0.00572 +2024-03-17 06:07:56,794 INFO Train. Iter 253800 : Commit. 1.04441 PPL. 393.15 Recons. 0.00511 +2024-03-17 06:08:18,916 INFO Train. Iter 254000 : Commit. 1.05735 PPL. 392.83 Recons. 0.00536 +2024-03-17 06:08:39,369 INFO Train. Iter 254200 : Commit. 1.05410 PPL. 391.25 Recons. 0.00546 +2024-03-17 06:09:01,212 INFO Train. Iter 254400 : Commit. 1.05550 PPL. 392.10 Recons. 0.00541 +2024-03-17 06:09:22,847 INFO Train. Iter 254600 : Commit. 1.06538 PPL. 393.12 Recons. 0.00553 +2024-03-17 06:09:44,354 INFO Train. Iter 254800 : Commit. 1.06876 PPL. 391.62 Recons. 0.00567 +2024-03-17 06:10:06,058 INFO Train. Iter 255000 : Commit. 1.04932 PPL. 393.26 Recons. 0.00521 +2024-03-17 06:10:27,550 INFO Train. Iter 255200 : Commit. 1.07021 PPL. 393.07 Recons. 0.00565 +2024-03-17 06:10:48,703 INFO Train. Iter 255400 : Commit. 1.04297 PPL. 392.63 Recons. 0.00510 +2024-03-17 06:11:09,968 INFO Train. Iter 255600 : Commit. 1.05517 PPL. 392.58 Recons. 0.00534 +2024-03-17 06:11:31,454 INFO Train. Iter 255800 : Commit. 1.05765 PPL. 392.45 Recons. 0.00539 +2024-03-17 06:11:53,306 INFO Train. Iter 256000 : Commit. 1.04976 PPL. 391.26 Recons. 0.00535 +2024-03-17 06:12:14,283 INFO Train. Iter 256200 : Commit. 1.05783 PPL. 392.38 Recons. 0.00542 +2024-03-17 06:12:36,305 INFO Train. Iter 256400 : Commit. 1.04546 PPL. 392.72 Recons. 0.00510 +2024-03-17 06:12:57,870 INFO Train. Iter 256600 : Commit. 1.05524 PPL. 391.10 Recons. 0.00548 +2024-03-17 06:13:19,049 INFO Train. Iter 256800 : Commit. 1.04975 PPL. 392.51 Recons. 0.00520 +2024-03-17 06:13:40,021 INFO Train. Iter 257000 : Commit. 1.06891 PPL. 391.92 Recons. 0.00564 +2024-03-17 06:14:01,152 INFO Train. Iter 257200 : Commit. 1.04478 PPL. 392.12 Recons. 0.00517 +2024-03-17 06:14:22,427 INFO Train. Iter 257400 : Commit. 1.05282 PPL. 391.88 Recons. 0.00533 +2024-03-17 06:14:43,834 INFO Train. Iter 257600 : Commit. 1.08065 PPL. 393.41 Recons. 0.00569 +2024-03-17 06:15:05,227 INFO Train. Iter 257800 : Commit. 1.05312 PPL. 391.82 Recons. 0.00537 +2024-03-17 06:15:25,197 INFO Train. Iter 258000 : Commit. 1.05651 PPL. 392.09 Recons. 0.00541 +2024-03-17 06:15:47,313 INFO Train. Iter 258200 : Commit. 1.04740 PPL. 392.39 Recons. 0.00522 +2024-03-17 06:16:09,061 INFO Train. Iter 258400 : Commit. 1.04497 PPL. 392.65 Recons. 0.00513 +2024-03-17 06:16:30,625 INFO Train. Iter 258600 : Commit. 1.05155 PPL. 391.34 Recons. 0.00541 +2024-03-17 06:16:52,370 INFO Train. Iter 258800 : Commit. 1.05023 PPL. 392.16 Recons. 0.00528 +2024-03-17 06:17:13,610 INFO Train. Iter 259000 : Commit. 1.06628 PPL. 391.41 Recons. 0.00572 +2024-03-17 06:17:35,416 INFO Train. Iter 259200 : Commit. 1.05389 PPL. 391.01 Recons. 0.00547 +2024-03-17 06:17:56,835 INFO Train. Iter 259400 : Commit. 1.07606 PPL. 392.06 Recons. 0.00576 +2024-03-17 06:18:18,010 INFO Train. Iter 259600 : Commit. 1.05690 PPL. 392.14 Recons. 0.00533 +2024-03-17 06:18:39,869 INFO Train. Iter 259800 : Commit. 1.04860 PPL. 390.84 Recons. 0.00535 +2024-03-17 06:19:00,674 INFO Train. Iter 260000 : Commit. 1.05334 PPL. 392.11 Recons. 0.00530 +2024-03-17 06:19:22,250 INFO Train. Iter 260200 : Commit. 1.04518 PPL. 391.44 Recons. 0.00519 +2024-03-17 06:19:43,755 INFO Train. Iter 260400 : Commit. 1.03777 PPL. 391.89 Recons. 0.00505 +2024-03-17 06:20:05,302 INFO Train. Iter 260600 : Commit. 1.03945 PPL. 391.03 Recons. 0.00516 +2024-03-17 06:20:27,264 INFO Train. Iter 260800 : Commit. 1.05662 PPL. 392.92 Recons. 0.00529 +2024-03-17 06:20:49,291 INFO Train. Iter 261000 : Commit. 1.03258 PPL. 391.99 Recons. 0.00496 +2024-03-17 06:21:11,120 INFO Train. Iter 261200 : Commit. 1.05809 PPL. 392.81 Recons. 0.00536 +2024-03-17 06:21:32,322 INFO Train. Iter 261400 : Commit. 1.04180 PPL. 390.52 Recons. 0.00525 +2024-03-17 06:21:53,654 INFO Train. Iter 261600 : Commit. 1.05023 PPL. 391.28 Recons. 0.00538 +2024-03-17 06:22:14,500 INFO Train. Iter 261800 : Commit. 1.05621 PPL. 391.89 Recons. 0.00541 +2024-03-17 06:22:34,386 INFO Train. Iter 262000 : Commit. 1.05186 PPL. 391.79 Recons. 0.00532 +2024-03-17 06:22:55,819 INFO Train. Iter 262200 : Commit. 1.05259 PPL. 391.05 Recons. 0.00537 +2024-03-17 06:23:16,762 INFO Train. Iter 262400 : Commit. 1.04798 PPL. 390.49 Recons. 0.00537 +2024-03-17 06:23:37,940 INFO Train. Iter 262600 : Commit. 1.06882 PPL. 390.82 Recons. 0.00573 +2024-03-17 06:23:59,644 INFO Train. Iter 262800 : Commit. 1.04720 PPL. 391.05 Recons. 0.00527 +2024-03-17 06:24:21,138 INFO Train. Iter 263000 : Commit. 1.06052 PPL. 391.93 Recons. 0.00543 +2024-03-17 06:24:42,612 INFO Train. Iter 263200 : Commit. 1.06891 PPL. 391.95 Recons. 0.00565 +2024-03-17 06:25:03,855 INFO Train. Iter 263400 : Commit. 1.06679 PPL. 392.04 Recons. 0.00558 +2024-03-17 06:25:25,202 INFO Train. Iter 263600 : Commit. 1.05240 PPL. 392.11 Recons. 0.00525 +2024-03-17 06:25:45,911 INFO Train. Iter 263800 : Commit. 1.05858 PPL. 392.20 Recons. 0.00540 +2024-03-17 06:26:06,695 INFO Train. Iter 264000 : Commit. 1.07255 PPL. 390.99 Recons. 0.00582 +2024-03-17 06:26:28,257 INFO Train. Iter 264200 : Commit. 1.05763 PPL. 391.68 Recons. 0.00540 +2024-03-17 06:26:49,970 INFO Train. Iter 264400 : Commit. 1.06385 PPL. 391.42 Recons. 0.00561 +2024-03-17 06:27:11,464 INFO Train. Iter 264600 : Commit. 1.04150 PPL. 391.23 Recons. 0.00518 +2024-03-17 06:27:33,156 INFO Train. Iter 264800 : Commit. 1.02748 PPL. 391.78 Recons. 0.00487 +2024-03-17 06:27:54,989 INFO Train. Iter 265000 : Commit. 1.04261 PPL. 391.76 Recons. 0.00515 +2024-03-17 06:28:16,067 INFO Train. Iter 265200 : Commit. 1.04172 PPL. 390.92 Recons. 0.00527 +2024-03-17 06:28:37,364 INFO Train. Iter 265400 : Commit. 1.04114 PPL. 391.63 Recons. 0.00516 +2024-03-17 06:28:58,631 INFO Train. Iter 265600 : Commit. 1.04176 PPL. 391.17 Recons. 0.00519 +2024-03-17 06:29:18,732 INFO Train. Iter 265800 : Commit. 1.05006 PPL. 390.94 Recons. 0.00538 +2024-03-17 06:29:39,660 INFO Train. Iter 266000 : Commit. 1.04720 PPL. 390.47 Recons. 0.00540 +2024-03-17 06:30:00,973 INFO Train. Iter 266200 : Commit. 1.06195 PPL. 391.32 Recons. 0.00553 +2024-03-17 06:30:22,393 INFO Train. Iter 266400 : Commit. 1.04328 PPL. 391.53 Recons. 0.00518 +2024-03-17 06:30:43,734 INFO Train. Iter 266600 : Commit. 1.04854 PPL. 391.42 Recons. 0.00529 +2024-03-17 06:31:05,498 INFO Train. Iter 266800 : Commit. 1.08451 PPL. 392.06 Recons. 0.00590 +2024-03-17 06:31:27,131 INFO Train. Iter 267000 : Commit. 1.06259 PPL. 391.25 Recons. 0.00552 +2024-03-17 06:31:48,505 INFO Train. Iter 267200 : Commit. 1.06609 PPL. 391.71 Recons. 0.00561 +2024-03-17 06:32:10,124 INFO Train. Iter 267400 : Commit. 1.03945 PPL. 390.94 Recons. 0.00512 +2024-03-17 06:32:31,399 INFO Train. Iter 267600 : Commit. 1.07210 PPL. 391.15 Recons. 0.00580 +2024-03-17 06:32:51,944 INFO Train. Iter 267800 : Commit. 1.08348 PPL. 390.45 Recons. 0.00607 +2024-03-17 06:33:12,859 INFO Train. Iter 268000 : Commit. 1.07126 PPL. 392.25 Recons. 0.00558 +2024-03-17 06:33:34,742 INFO Train. Iter 268200 : Commit. 1.05980 PPL. 390.24 Recons. 0.00556 +2024-03-17 06:33:56,141 INFO Train. Iter 268400 : Commit. 1.06621 PPL. 391.67 Recons. 0.00560 +2024-03-17 06:34:17,570 INFO Train. Iter 268600 : Commit. 1.03616 PPL. 390.37 Recons. 0.00513 +2024-03-17 06:34:39,243 INFO Train. Iter 268800 : Commit. 1.05841 PPL. 391.03 Recons. 0.00557 +2024-03-17 06:35:00,790 INFO Train. Iter 269000 : Commit. 1.04019 PPL. 391.07 Recons. 0.00514 +2024-03-17 06:35:22,255 INFO Train. Iter 269200 : Commit. 1.05302 PPL. 391.04 Recons. 0.00539 +2024-03-17 06:35:43,683 INFO Train. Iter 269400 : Commit. 1.06157 PPL. 390.29 Recons. 0.00558 +2024-03-17 06:36:04,157 INFO Train. Iter 269600 : Commit. 1.07095 PPL. 392.43 Recons. 0.00560 +2024-03-17 06:36:25,422 INFO Train. Iter 269800 : Commit. 1.06529 PPL. 391.55 Recons. 0.00557 +2024-03-17 06:36:46,526 INFO Train. Iter 270000 : Commit. 1.05773 PPL. 391.08 Recons. 0.00544 +2024-03-17 06:37:08,114 INFO Train. Iter 270200 : Commit. 1.05342 PPL. 390.86 Recons. 0.00541 +2024-03-17 06:37:29,972 INFO Train. Iter 270400 : Commit. 1.05602 PPL. 390.83 Recons. 0.00544 +2024-03-17 06:37:51,718 INFO Train. Iter 270600 : Commit. 1.08835 PPL. 391.22 Recons. 0.00600 +2024-03-17 06:38:13,145 INFO Train. Iter 270800 : Commit. 1.05211 PPL. 391.56 Recons. 0.00531 +2024-03-17 06:38:34,739 INFO Train. Iter 271000 : Commit. 1.04284 PPL. 390.82 Recons. 0.00520 +2024-03-17 06:38:56,745 INFO Train. Iter 271200 : Commit. 1.04759 PPL. 390.93 Recons. 0.00529 +2024-03-17 06:39:18,236 INFO Train. Iter 271400 : Commit. 1.06181 PPL. 391.04 Recons. 0.00557 +2024-03-17 06:39:38,686 INFO Train. Iter 271600 : Commit. 1.05768 PPL. 391.26 Recons. 0.00540 +2024-03-17 06:40:00,427 INFO Train. Iter 271800 : Commit. 1.05663 PPL. 390.49 Recons. 0.00543 +2024-03-17 06:40:22,177 INFO Train. Iter 272000 : Commit. 1.04090 PPL. 389.85 Recons. 0.00524 +2024-03-17 06:40:43,218 INFO Train. Iter 272200 : Commit. 1.06705 PPL. 391.74 Recons. 0.00556 +2024-03-17 06:41:04,418 INFO Train. Iter 272400 : Commit. 1.05603 PPL. 389.80 Recons. 0.00558 +2024-03-17 06:41:25,763 INFO Train. Iter 272600 : Commit. 1.05531 PPL. 391.15 Recons. 0.00540 +2024-03-17 06:41:46,754 INFO Train. Iter 272800 : Commit. 1.05357 PPL. 391.02 Recons. 0.00534 +2024-03-17 06:42:07,748 INFO Train. Iter 273000 : Commit. 1.05669 PPL. 390.27 Recons. 0.00550 +2024-03-17 06:42:29,002 INFO Train. Iter 273200 : Commit. 1.05934 PPL. 390.42 Recons. 0.00548 +2024-03-17 06:42:49,880 INFO Train. Iter 273400 : Commit. 1.06594 PPL. 390.64 Recons. 0.00563 +2024-03-17 06:43:10,087 INFO Train. Iter 273600 : Commit. 1.06560 PPL. 391.05 Recons. 0.00558 +2024-03-17 06:43:31,627 INFO Train. Iter 273800 : Commit. 1.04553 PPL. 389.88 Recons. 0.00529 +2024-03-17 06:43:52,764 INFO Train. Iter 274000 : Commit. 1.04970 PPL. 390.27 Recons. 0.00537 +2024-03-17 06:44:13,859 INFO Train. Iter 274200 : Commit. 1.06417 PPL. 390.02 Recons. 0.00558 +2024-03-17 06:44:35,376 INFO Train. Iter 274400 : Commit. 1.05411 PPL. 390.16 Recons. 0.00541 +2024-03-17 06:44:57,297 INFO Train. Iter 274600 : Commit. 1.06761 PPL. 390.25 Recons. 0.00573 +2024-03-17 06:45:19,466 INFO Train. Iter 274800 : Commit. 1.05876 PPL. 390.46 Recons. 0.00546 +2024-03-17 06:45:41,060 INFO Train. Iter 275000 : Commit. 1.06139 PPL. 391.23 Recons. 0.00546 +2024-03-17 06:46:02,514 INFO Train. Iter 275200 : Commit. 1.04354 PPL. 389.82 Recons. 0.00526 +2024-03-17 06:46:23,169 INFO Train. Iter 275400 : Commit. 1.06123 PPL. 389.72 Recons. 0.00559 +2024-03-17 06:46:44,153 INFO Train. Iter 275600 : Commit. 1.05431 PPL. 390.16 Recons. 0.00538 +2024-03-17 06:47:05,273 INFO Train. Iter 275800 : Commit. 1.03946 PPL. 389.95 Recons. 0.00517 +2024-03-17 06:47:26,408 INFO Train. Iter 276000 : Commit. 1.05637 PPL. 390.74 Recons. 0.00541 +2024-03-17 06:47:47,617 INFO Train. Iter 276200 : Commit. 1.06117 PPL. 390.73 Recons. 0.00553 +2024-03-17 06:48:08,909 INFO Train. Iter 276400 : Commit. 1.04807 PPL. 389.53 Recons. 0.00533 +2024-03-17 06:48:30,393 INFO Train. Iter 276600 : Commit. 1.05896 PPL. 390.16 Recons. 0.00552 +2024-03-17 06:48:52,017 INFO Train. Iter 276800 : Commit. 1.06676 PPL. 390.51 Recons. 0.00566 +2024-03-17 06:49:14,032 INFO Train. Iter 277000 : Commit. 1.05892 PPL. 390.42 Recons. 0.00546 +2024-03-17 06:49:35,723 INFO Train. Iter 277200 : Commit. 1.02267 PPL. 389.96 Recons. 0.00485 +2024-03-17 06:49:56,207 INFO Train. Iter 277400 : Commit. 1.05336 PPL. 389.05 Recons. 0.00553 +2024-03-17 06:50:17,284 INFO Train. Iter 277600 : Commit. 1.05622 PPL. 390.31 Recons. 0.00548 +2024-03-17 06:50:38,336 INFO Train. Iter 277800 : Commit. 1.06508 PPL. 390.75 Recons. 0.00557 +2024-03-17 06:50:59,888 INFO Train. Iter 278000 : Commit. 1.04735 PPL. 389.34 Recons. 0.00532 +2024-03-17 06:51:20,979 INFO Train. Iter 278200 : Commit. 1.03064 PPL. 388.74 Recons. 0.00506 +2024-03-17 06:51:42,360 INFO Train. Iter 278400 : Commit. 1.04887 PPL. 391.27 Recons. 0.00516 +2024-03-17 06:52:03,623 INFO Train. Iter 278600 : Commit. 1.04813 PPL. 389.87 Recons. 0.00531 +2024-03-17 06:52:25,221 INFO Train. Iter 278800 : Commit. 1.06861 PPL. 390.14 Recons. 0.00565 +2024-03-17 06:52:46,745 INFO Train. Iter 279000 : Commit. 1.05540 PPL. 390.25 Recons. 0.00536 +2024-03-17 06:53:08,050 INFO Train. Iter 279200 : Commit. 1.05045 PPL. 388.89 Recons. 0.00549 +2024-03-17 06:53:28,015 INFO Train. Iter 279400 : Commit. 1.06308 PPL. 389.88 Recons. 0.00557 +2024-03-17 06:53:49,615 INFO Train. Iter 279600 : Commit. 1.04707 PPL. 390.53 Recons. 0.00524 +2024-03-17 06:54:11,214 INFO Train. Iter 279800 : Commit. 1.06705 PPL. 389.90 Recons. 0.00567 +2024-03-17 06:54:32,840 INFO Train. Iter 280000 : Commit. 1.02709 PPL. 389.75 Recons. 0.00495 +2024-03-17 06:54:54,656 INFO Train. Iter 280200 : Commit. 1.05559 PPL. 390.97 Recons. 0.00538 +2024-03-17 06:55:16,289 INFO Train. Iter 280400 : Commit. 1.05332 PPL. 389.80 Recons. 0.00540 +2024-03-17 06:55:37,793 INFO Train. Iter 280600 : Commit. 1.04813 PPL. 390.21 Recons. 0.00533 +2024-03-17 06:55:59,301 INFO Train. Iter 280800 : Commit. 1.04330 PPL. 389.39 Recons. 0.00527 +2024-03-17 06:56:20,730 INFO Train. Iter 281000 : Commit. 1.05436 PPL. 389.70 Recons. 0.00549 +2024-03-17 06:56:41,111 INFO Train. Iter 281200 : Commit. 1.05825 PPL. 390.65 Recons. 0.00544 +2024-03-17 06:57:02,456 INFO Train. Iter 281400 : Commit. 1.07199 PPL. 391.34 Recons. 0.00561 +2024-03-17 06:57:23,821 INFO Train. Iter 281600 : Commit. 1.03352 PPL. 389.56 Recons. 0.00507 +2024-03-17 06:57:45,775 INFO Train. Iter 281800 : Commit. 1.05130 PPL. 390.92 Recons. 0.00526 +2024-03-17 06:58:07,600 INFO Train. Iter 282000 : Commit. 1.05080 PPL. 389.02 Recons. 0.00549 +2024-03-17 06:58:28,867 INFO Train. Iter 282200 : Commit. 1.03691 PPL. 389.16 Recons. 0.00516 +2024-03-17 06:58:50,430 INFO Train. Iter 282400 : Commit. 1.04566 PPL. 389.34 Recons. 0.00529 +2024-03-17 06:59:11,784 INFO Train. Iter 282600 : Commit. 1.05821 PPL. 390.24 Recons. 0.00544 +2024-03-17 06:59:33,569 INFO Train. Iter 282800 : Commit. 1.05739 PPL. 389.69 Recons. 0.00551 +2024-03-17 06:59:55,068 INFO Train. Iter 283000 : Commit. 1.05507 PPL. 389.92 Recons. 0.00544 +2024-03-17 07:00:16,067 INFO Train. Iter 283200 : Commit. 1.03119 PPL. 389.12 Recons. 0.00503 +2024-03-17 07:00:38,723 INFO Train. Iter 283400 : Commit. 1.03283 PPL. 388.10 Recons. 0.00519 +2024-03-17 07:01:00,341 INFO Train. Iter 283600 : Commit. 1.04992 PPL. 390.53 Recons. 0.00529 +2024-03-17 07:01:22,115 INFO Train. Iter 283800 : Commit. 1.03759 PPL. 388.89 Recons. 0.00518 +2024-03-17 07:01:43,351 INFO Train. Iter 284000 : Commit. 1.04429 PPL. 388.72 Recons. 0.00535 +2024-03-17 07:02:04,732 INFO Train. Iter 284200 : Commit. 1.04448 PPL. 390.18 Recons. 0.00519 +2024-03-17 07:02:26,210 INFO Train. Iter 284400 : Commit. 1.05831 PPL. 389.52 Recons. 0.00554 +2024-03-17 07:02:47,647 INFO Train. Iter 284600 : Commit. 1.06834 PPL. 389.19 Recons. 0.00578 +2024-03-17 07:03:08,754 INFO Train. Iter 284800 : Commit. 1.03986 PPL. 388.97 Recons. 0.00518 +2024-03-17 07:03:30,147 INFO Train. Iter 285000 : Commit. 1.04946 PPL. 389.32 Recons. 0.00535 +2024-03-17 07:03:50,498 INFO Train. Iter 285200 : Commit. 1.04953 PPL. 390.12 Recons. 0.00531 +2024-03-17 07:04:11,932 INFO Train. Iter 285400 : Commit. 1.05383 PPL. 389.61 Recons. 0.00537 +2024-03-17 07:04:33,877 INFO Train. Iter 285600 : Commit. 1.02962 PPL. 388.27 Recons. 0.00502 +2024-03-17 07:04:55,357 INFO Train. Iter 285800 : Commit. 1.03649 PPL. 389.55 Recons. 0.00501 +2024-03-17 07:05:16,851 INFO Train. Iter 286000 : Commit. 1.04401 PPL. 388.11 Recons. 0.00535 +2024-03-17 07:05:38,154 INFO Train. Iter 286200 : Commit. 1.04387 PPL. 388.57 Recons. 0.00536 +2024-03-17 07:05:59,824 INFO Train. Iter 286400 : Commit. 1.04615 PPL. 389.14 Recons. 0.00529 +2024-03-17 07:06:21,397 INFO Train. Iter 286600 : Commit. 1.06038 PPL. 389.91 Recons. 0.00546 +2024-03-17 07:06:43,182 INFO Train. Iter 286800 : Commit. 1.03902 PPL. 388.29 Recons. 0.00520 +2024-03-17 07:07:03,739 INFO Train. Iter 287000 : Commit. 1.05255 PPL. 389.56 Recons. 0.00543 +2024-03-17 07:07:25,875 INFO Train. Iter 287200 : Commit. 1.03811 PPL. 389.21 Recons. 0.00516 +2024-03-17 07:07:47,948 INFO Train. Iter 287400 : Commit. 1.05516 PPL. 389.06 Recons. 0.00541 +2024-03-17 07:08:09,642 INFO Train. Iter 287600 : Commit. 1.05490 PPL. 388.61 Recons. 0.00546 +2024-03-17 07:08:31,607 INFO Train. Iter 287800 : Commit. 1.05036 PPL. 391.16 Recons. 0.00518 +2024-03-17 07:08:52,955 INFO Train. Iter 288000 : Commit. 1.03908 PPL. 389.42 Recons. 0.00515 +2024-03-17 07:09:14,272 INFO Train. Iter 288200 : Commit. 1.05492 PPL. 389.35 Recons. 0.00548 +2024-03-17 07:09:36,082 INFO Train. Iter 288400 : Commit. 1.04863 PPL. 387.89 Recons. 0.00544 +2024-03-17 07:09:57,534 INFO Train. Iter 288600 : Commit. 1.06459 PPL. 389.39 Recons. 0.00565 +2024-03-17 07:10:18,733 INFO Train. Iter 288800 : Commit. 1.02483 PPL. 388.20 Recons. 0.00496 +2024-03-17 07:10:39,183 INFO Train. Iter 289000 : Commit. 1.04146 PPL. 389.06 Recons. 0.00525 +2024-03-17 07:11:00,516 INFO Train. Iter 289200 : Commit. 1.04245 PPL. 388.69 Recons. 0.00531 +2024-03-17 07:11:22,007 INFO Train. Iter 289400 : Commit. 1.05253 PPL. 389.26 Recons. 0.00539 +2024-03-17 07:11:44,209 INFO Train. Iter 289600 : Commit. 1.03737 PPL. 389.18 Recons. 0.00512 +2024-03-17 07:12:06,309 INFO Train. Iter 289800 : Commit. 1.06389 PPL. 389.62 Recons. 0.00563 +2024-03-17 07:12:28,045 INFO Train. Iter 290000 : Commit. 1.05702 PPL. 389.37 Recons. 0.00548 +2024-03-17 07:12:49,527 INFO Train. Iter 290200 : Commit. 1.02969 PPL. 388.18 Recons. 0.00503 +2024-03-17 07:13:11,480 INFO Train. Iter 290400 : Commit. 1.05688 PPL. 388.34 Recons. 0.00555 +2024-03-17 07:13:33,065 INFO Train. Iter 290600 : Commit. 1.05477 PPL. 389.87 Recons. 0.00534 +2024-03-17 07:13:54,574 INFO Train. Iter 290800 : Commit. 1.04880 PPL. 389.72 Recons. 0.00531 +2024-03-17 07:14:15,169 INFO Train. Iter 291000 : Commit. 1.04944 PPL. 387.28 Recons. 0.00554 +2024-03-17 07:14:36,371 INFO Train. Iter 291200 : Commit. 1.05079 PPL. 388.89 Recons. 0.00540 +2024-03-17 07:14:58,145 INFO Train. Iter 291400 : Commit. 1.02974 PPL. 388.98 Recons. 0.00499 +2024-03-17 07:15:19,540 INFO Train. Iter 291600 : Commit. 1.02755 PPL. 387.63 Recons. 0.00511 +2024-03-17 07:15:40,819 INFO Train. Iter 291800 : Commit. 1.03786 PPL. 389.61 Recons. 0.00512 +2024-03-17 07:16:02,869 INFO Train. Iter 292000 : Commit. 1.02121 PPL. 388.20 Recons. 0.00491 +2024-03-17 07:16:24,805 INFO Train. Iter 292200 : Commit. 1.03943 PPL. 388.17 Recons. 0.00521 +2024-03-17 07:16:46,599 INFO Train. Iter 292400 : Commit. 1.03915 PPL. 388.90 Recons. 0.00515 +2024-03-17 07:17:07,737 INFO Train. Iter 292600 : Commit. 1.05601 PPL. 388.39 Recons. 0.00561 +2024-03-17 07:17:28,265 INFO Train. Iter 292800 : Commit. 1.03461 PPL. 388.65 Recons. 0.00509 +2024-03-17 07:17:50,202 INFO Train. Iter 293000 : Commit. 1.05423 PPL. 389.10 Recons. 0.00543 +2024-03-17 07:18:11,243 INFO Train. Iter 293200 : Commit. 1.05125 PPL. 388.28 Recons. 0.00541 +2024-03-17 07:18:32,793 INFO Train. Iter 293400 : Commit. 1.04009 PPL. 388.46 Recons. 0.00517 +2024-03-17 07:18:54,416 INFO Train. Iter 293600 : Commit. 1.04126 PPL. 387.12 Recons. 0.00538 +2024-03-17 07:19:16,072 INFO Train. Iter 293800 : Commit. 1.04565 PPL. 388.24 Recons. 0.00530 +2024-03-17 07:19:37,491 INFO Train. Iter 294000 : Commit. 1.04151 PPL. 388.96 Recons. 0.00521 +2024-03-17 07:19:59,103 INFO Train. Iter 294200 : Commit. 1.05057 PPL. 388.62 Recons. 0.00538 +2024-03-17 07:20:20,404 INFO Train. Iter 294400 : Commit. 1.04929 PPL. 387.82 Recons. 0.00549 +2024-03-17 07:20:42,489 INFO Train. Iter 294600 : Commit. 1.04034 PPL. 387.67 Recons. 0.00527 +2024-03-17 07:21:02,796 INFO Train. Iter 294800 : Commit. 1.05370 PPL. 388.39 Recons. 0.00546 +2024-03-17 07:21:24,421 INFO Train. Iter 295000 : Commit. 1.06826 PPL. 389.57 Recons. 0.00561 +2024-03-17 07:21:45,692 INFO Train. Iter 295200 : Commit. 1.04973 PPL. 389.36 Recons. 0.00533 +2024-03-17 07:22:06,079 INFO Train. Iter 295400 : Commit. 1.05425 PPL. 387.63 Recons. 0.00560 +2024-03-17 07:22:26,497 INFO Train. Iter 295600 : Commit. 1.04173 PPL. 387.61 Recons. 0.00527 +2024-03-17 07:22:46,915 INFO Train. Iter 295800 : Commit. 1.03605 PPL. 389.10 Recons. 0.00506 +2024-03-17 07:23:06,967 INFO Train. Iter 296000 : Commit. 1.03953 PPL. 387.92 Recons. 0.00532 +2024-03-17 07:23:26,748 INFO Train. Iter 296200 : Commit. 1.04238 PPL. 388.89 Recons. 0.00523 +2024-03-17 07:23:46,705 INFO Train. Iter 296400 : Commit. 1.05370 PPL. 388.14 Recons. 0.00548 +2024-03-17 07:24:06,477 INFO Train. Iter 296600 : Commit. 1.04000 PPL. 388.48 Recons. 0.00520 +2024-03-17 07:24:25,436 INFO Train. Iter 296800 : Commit. 1.03484 PPL. 386.74 Recons. 0.00522 +2024-03-17 07:24:45,675 INFO Train. Iter 297000 : Commit. 1.04548 PPL. 388.68 Recons. 0.00525 +2024-03-17 07:25:06,024 INFO Train. Iter 297200 : Commit. 1.04026 PPL. 387.29 Recons. 0.00535 +2024-03-17 07:25:26,510 INFO Train. Iter 297400 : Commit. 1.05119 PPL. 388.61 Recons. 0.00540 +2024-03-17 07:25:46,902 INFO Train. Iter 297600 : Commit. 1.04667 PPL. 388.07 Recons. 0.00538 +2024-03-17 07:26:07,210 INFO Train. Iter 297800 : Commit. 1.05707 PPL. 387.25 Recons. 0.00561 +2024-03-17 07:26:27,539 INFO Train. Iter 298000 : Commit. 1.04192 PPL. 388.18 Recons. 0.00526 +2024-03-17 07:26:47,722 INFO Train. Iter 298200 : Commit. 1.04738 PPL. 388.07 Recons. 0.00539 +2024-03-17 07:27:08,153 INFO Train. Iter 298400 : Commit. 1.05153 PPL. 387.76 Recons. 0.00548 +2024-03-17 07:27:27,491 INFO Train. Iter 298600 : Commit. 1.03378 PPL. 388.02 Recons. 0.00508 +2024-03-17 07:27:47,896 INFO Train. Iter 298800 : Commit. 1.05276 PPL. 388.39 Recons. 0.00545 +2024-03-17 07:28:08,357 INFO Train. Iter 299000 : Commit. 1.04697 PPL. 388.43 Recons. 0.00536 +2024-03-17 07:28:28,732 INFO Train. Iter 299200 : Commit. 1.04590 PPL. 387.65 Recons. 0.00539 +2024-03-17 07:28:48,673 INFO Train. Iter 299400 : Commit. 1.04643 PPL. 387.84 Recons. 0.00536 +2024-03-17 07:29:08,999 INFO Train. Iter 299600 : Commit. 1.05002 PPL. 388.42 Recons. 0.00541 +2024-03-17 07:29:29,552 INFO Train. Iter 299800 : Commit. 1.03580 PPL. 387.43 Recons. 0.00515 +2024-03-17 07:29:50,066 INFO Train. Iter 300000 : Commit. 1.04249 PPL. 387.22 Recons. 0.00536 diff --git a/ckpt/beatx2_rvqvae/RVQVAE_lower/net_300000.pth b/ckpt/beatx2_rvqvae/RVQVAE_lower/net_300000.pth new file mode 100644 index 0000000000000000000000000000000000000000..2663b38e505b966d4984d87a1fe5521dfa9c7b60 --- /dev/null +++ b/ckpt/beatx2_rvqvae/RVQVAE_lower/net_300000.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a29217af4f33b7b50ae9aebfdfc2bf2c0e80bed48316ab218cdc40043bb03d20 +size 81499947 diff --git a/ckpt/beatx2_rvqvae/RVQVAE_lower/run.log b/ckpt/beatx2_rvqvae/RVQVAE_lower/run.log new file mode 100644 index 0000000000000000000000000000000000000000..56f24791668a72fdc8e5851d16d1a9a5888cf955 --- /dev/null +++ b/ckpt/beatx2_rvqvae/RVQVAE_lower/run.log @@ -0,0 +1,1546 @@ +2024-03-16 22:22:23,240 INFO { + "batch_size": 256, + "beta": 1.0, + "body_part": "lower", + "code_dim": 512, + "commit": 0.02, + "dataname": "kit", + "depth": 3, + "dilation_growth_rate": 3, + "down_t": 2, + "eval_iter": 1000, + "exp_name": "RVQVAE", + "gamma": 0.05, + "loss_vel": 0.5, + "lr": 0.0002, + "lr_scheduler": [ + 200000 + ], + "mu": 0.99, + "nb_code": 512, + "nb_vis": 20, + "out_dir": "output_beatx2/RVQVAE_lower", + "output_emb_width": 512, + "print_iter": 200, + "quantizer": "ema_reset", + "recons_loss": "l1_smooth", + "results_dir": "visual_results/", + "resume_gpt": null, + "resume_pth": null, + "seed": 123, + "stride_t": 2, + "total_iter": 300000, + "vis_gt": false, + "visual_name": "baseline", + "vq_act": "relu", + "vq_norm": null, + "warm_up_iter": 1000, + "weight_decay": 0.0, + "width": 512, + "window_size": 64 +} +2024-03-16 22:22:23,247 INFO Training on kit, motions are with 21 joints +2024-03-16 22:22:46,941 INFO Warmup. Iter 200 : lr 0.00004 Commit. 0.03632 PPL. 281.38 Recons. 0.29623 +2024-03-16 22:23:08,046 INFO Warmup. Iter 400 : lr 0.00008 Commit. 0.32469 PPL. 157.60 Recons. 0.12597 +2024-03-16 22:23:29,221 INFO Warmup. Iter 600 : lr 0.00012 Commit. 0.67509 PPL. 310.94 Recons. 0.07468 +2024-03-16 22:23:50,290 INFO Warmup. Iter 800 : lr 0.00016 Commit. 0.85791 PPL. 409.21 Recons. 0.04913 +2024-03-16 22:24:31,869 INFO Train. Iter 200 : Commit. 1.01528 PPL. 405.90 Recons. 0.03009 +2024-03-16 22:24:53,464 INFO Train. Iter 400 : Commit. 0.99972 PPL. 404.29 Recons. 0.02545 +2024-03-16 22:25:14,492 INFO Train. Iter 600 : Commit. 0.99347 PPL. 404.71 Recons. 0.02108 +2024-03-16 22:25:35,674 INFO Train. Iter 800 : Commit. 0.98879 PPL. 405.87 Recons. 0.01883 +2024-03-16 22:25:56,170 INFO Train. Iter 1000 : Commit. 0.97120 PPL. 405.82 Recons. 0.01820 +2024-03-16 22:26:17,246 INFO Train. Iter 1200 : Commit. 0.98959 PPL. 409.54 Recons. 0.01849 +2024-03-16 22:26:38,269 INFO Train. Iter 1400 : Commit. 0.98559 PPL. 409.75 Recons. 0.01720 +2024-03-16 22:26:59,354 INFO Train. Iter 1600 : Commit. 0.96971 PPL. 410.59 Recons. 0.01663 +2024-03-16 22:27:20,918 INFO Train. Iter 1800 : Commit. 0.94975 PPL. 412.82 Recons. 0.01425 +2024-03-16 22:27:41,929 INFO Train. Iter 2000 : Commit. 0.98705 PPL. 410.40 Recons. 0.01766 +2024-03-16 22:28:02,856 INFO Train. Iter 2200 : Commit. 0.90901 PPL. 416.23 Recons. 0.01198 +2024-03-16 22:28:24,209 INFO Train. Iter 2400 : Commit. 0.88701 PPL. 416.08 Recons. 0.01070 +2024-03-16 22:28:45,502 INFO Train. Iter 2600 : Commit. 0.95768 PPL. 411.78 Recons. 0.01606 +2024-03-16 22:29:05,942 INFO Train. Iter 2800 : Commit. 0.89706 PPL. 416.49 Recons. 0.01183 +2024-03-16 22:29:27,211 INFO Train. Iter 3000 : Commit. 0.86985 PPL. 416.99 Recons. 0.01119 +2024-03-16 22:29:48,272 INFO Train. Iter 3200 : Commit. 0.86082 PPL. 416.06 Recons. 0.01127 +2024-03-16 22:30:09,266 INFO Train. Iter 3400 : Commit. 0.84946 PPL. 417.35 Recons. 0.01102 +2024-03-16 22:30:30,530 INFO Train. Iter 3600 : Commit. 0.85271 PPL. 416.25 Recons. 0.01154 +2024-03-16 22:30:51,547 INFO Train. Iter 3800 : Commit. 0.82511 PPL. 418.25 Recons. 0.00969 +2024-03-16 22:31:12,877 INFO Train. Iter 4000 : Commit. 0.83510 PPL. 416.98 Recons. 0.01195 +2024-03-16 22:31:34,164 INFO Train. Iter 4200 : Commit. 0.82537 PPL. 416.75 Recons. 0.01160 +2024-03-16 22:31:54,997 INFO Train. Iter 4400 : Commit. 0.81342 PPL. 419.41 Recons. 0.01039 +2024-03-16 22:32:16,300 INFO Train. Iter 4600 : Commit. 0.82761 PPL. 415.94 Recons. 0.01199 +2024-03-16 22:32:36,204 INFO Train. Iter 4800 : Commit. 0.83680 PPL. 418.09 Recons. 0.01193 +2024-03-16 22:32:57,566 INFO Train. Iter 5000 : Commit. 0.79971 PPL. 419.49 Recons. 0.00988 +2024-03-16 22:33:18,718 INFO Train. Iter 5200 : Commit. 0.83281 PPL. 415.88 Recons. 0.01239 +2024-03-16 22:33:40,169 INFO Train. Iter 5400 : Commit. 0.79402 PPL. 420.64 Recons. 0.01048 +2024-03-16 22:34:01,777 INFO Train. Iter 5600 : Commit. 0.79905 PPL. 416.86 Recons. 0.01140 +2024-03-16 22:34:22,806 INFO Train. Iter 5800 : Commit. 0.76826 PPL. 420.59 Recons. 0.01021 +2024-03-16 22:34:43,913 INFO Train. Iter 6000 : Commit. 0.79164 PPL. 419.70 Recons. 0.01100 +2024-03-16 22:35:05,183 INFO Train. Iter 6200 : Commit. 0.80298 PPL. 416.86 Recons. 0.01137 +2024-03-16 22:35:25,909 INFO Train. Iter 6400 : Commit. 0.75271 PPL. 420.97 Recons. 0.00921 +2024-03-16 22:35:46,733 INFO Train. Iter 6600 : Commit. 0.75780 PPL. 421.09 Recons. 0.00978 +2024-03-16 22:36:06,614 INFO Train. Iter 6800 : Commit. 0.76229 PPL. 418.10 Recons. 0.01125 +2024-03-16 22:36:27,730 INFO Train. Iter 7000 : Commit. 0.71655 PPL. 422.87 Recons. 0.00855 +2024-03-16 22:36:48,763 INFO Train. Iter 7200 : Commit. 0.70328 PPL. 420.63 Recons. 0.00871 +2024-03-16 22:37:09,911 INFO Train. Iter 7400 : Commit. 0.72369 PPL. 421.88 Recons. 0.00996 +2024-03-16 22:37:30,520 INFO Train. Iter 7600 : Commit. 0.71883 PPL. 422.97 Recons. 0.00894 +2024-03-16 22:37:51,686 INFO Train. Iter 7800 : Commit. 0.68412 PPL. 420.38 Recons. 0.00820 +2024-03-16 22:38:12,962 INFO Train. Iter 8000 : Commit. 0.64273 PPL. 423.26 Recons. 0.00725 +2024-03-16 22:38:34,591 INFO Train. Iter 8200 : Commit. 0.65748 PPL. 423.18 Recons. 0.00809 +2024-03-16 22:38:56,387 INFO Train. Iter 8400 : Commit. 0.68420 PPL. 422.41 Recons. 0.00878 +2024-03-16 22:39:17,308 INFO Train. Iter 8600 : Commit. 0.66536 PPL. 421.26 Recons. 0.00875 +2024-03-16 22:39:39,527 INFO Train. Iter 8800 : Commit. 0.68011 PPL. 423.06 Recons. 0.00828 +2024-03-16 22:40:01,826 INFO Train. Iter 9000 : Commit. 0.65192 PPL. 423.70 Recons. 0.00781 +2024-03-16 22:40:23,568 INFO Train. Iter 9200 : Commit. 0.66993 PPL. 419.95 Recons. 0.00948 +2024-03-16 22:40:44,931 INFO Train. Iter 9400 : Commit. 0.63062 PPL. 424.80 Recons. 0.00684 +2024-03-16 22:41:06,607 INFO Train. Iter 9600 : Commit. 0.60551 PPL. 424.63 Recons. 0.00658 +2024-03-16 22:41:28,298 INFO Train. Iter 9800 : Commit. 0.61417 PPL. 422.25 Recons. 0.00819 +2024-03-16 22:41:50,208 INFO Train. Iter 10000 : Commit. 0.64399 PPL. 423.97 Recons. 0.00885 +2024-03-16 22:42:12,230 INFO Train. Iter 10200 : Commit. 0.61611 PPL. 425.62 Recons. 0.00692 +2024-03-16 22:42:33,641 INFO Train. Iter 10400 : Commit. 0.64584 PPL. 422.89 Recons. 0.00912 +2024-03-16 22:42:54,269 INFO Train. Iter 10600 : Commit. 0.62773 PPL. 424.62 Recons. 0.00723 +2024-03-16 22:43:16,223 INFO Train. Iter 10800 : Commit. 0.63000 PPL. 424.66 Recons. 0.00801 +2024-03-16 22:43:38,059 INFO Train. Iter 11000 : Commit. 0.58771 PPL. 423.98 Recons. 0.00721 +2024-03-16 22:43:59,787 INFO Train. Iter 11200 : Commit. 0.65067 PPL. 420.48 Recons. 0.00988 +2024-03-16 22:44:21,443 INFO Train. Iter 11400 : Commit. 0.61512 PPL. 424.34 Recons. 0.00778 +2024-03-16 22:44:43,360 INFO Train. Iter 11600 : Commit. 0.60189 PPL. 424.72 Recons. 0.00785 +2024-03-16 22:45:05,237 INFO Train. Iter 11800 : Commit. 0.63315 PPL. 422.60 Recons. 0.00971 +2024-03-16 22:45:26,889 INFO Train. Iter 12000 : Commit. 0.59940 PPL. 424.75 Recons. 0.00703 +2024-03-16 22:45:48,907 INFO Train. Iter 12200 : Commit. 0.60795 PPL. 425.96 Recons. 0.00743 +2024-03-16 22:46:11,327 INFO Train. Iter 12400 : Commit. 0.58186 PPL. 425.25 Recons. 0.00686 +2024-03-16 22:46:32,096 INFO Train. Iter 12600 : Commit. 0.62265 PPL. 420.98 Recons. 0.00986 +2024-03-16 22:46:54,392 INFO Train. Iter 12800 : Commit. 0.58873 PPL. 423.56 Recons. 0.00763 +2024-03-16 22:47:16,337 INFO Train. Iter 13000 : Commit. 0.56215 PPL. 425.88 Recons. 0.00669 +2024-03-16 22:47:37,746 INFO Train. Iter 13200 : Commit. 0.57014 PPL. 427.20 Recons. 0.00679 +2024-03-16 22:47:59,780 INFO Train. Iter 13400 : Commit. 0.58282 PPL. 424.32 Recons. 0.00821 +2024-03-16 22:48:21,763 INFO Train. Iter 13600 : Commit. 0.57709 PPL. 423.57 Recons. 0.00787 +2024-03-16 22:48:43,646 INFO Train. Iter 13800 : Commit. 0.60471 PPL. 425.01 Recons. 0.00837 +2024-03-16 22:49:05,549 INFO Train. Iter 14000 : Commit. 0.60302 PPL. 425.38 Recons. 0.00795 +2024-03-16 22:49:27,227 INFO Train. Iter 14200 : Commit. 0.59145 PPL. 425.62 Recons. 0.00701 +2024-03-16 22:49:47,705 INFO Train. Iter 14400 : Commit. 0.58543 PPL. 422.48 Recons. 0.00807 +2024-03-16 22:50:09,287 INFO Train. Iter 14600 : Commit. 0.53517 PPL. 426.98 Recons. 0.00560 +2024-03-16 22:50:31,624 INFO Train. Iter 14800 : Commit. 0.57187 PPL. 426.16 Recons. 0.00738 +2024-03-16 22:50:53,840 INFO Train. Iter 15000 : Commit. 0.58902 PPL. 424.99 Recons. 0.00790 +2024-03-16 22:51:16,424 INFO Train. Iter 15200 : Commit. 0.57504 PPL. 427.13 Recons. 0.00682 +2024-03-16 22:51:38,390 INFO Train. Iter 15400 : Commit. 0.56480 PPL. 425.57 Recons. 0.00695 +2024-03-16 22:52:00,211 INFO Train. Iter 15600 : Commit. 0.58617 PPL. 425.87 Recons. 0.00820 +2024-03-16 22:52:21,922 INFO Train. Iter 15800 : Commit. 0.57876 PPL. 426.74 Recons. 0.00734 +2024-03-16 22:52:44,085 INFO Train. Iter 16000 : Commit. 0.55244 PPL. 424.62 Recons. 0.00710 +2024-03-16 22:53:06,456 INFO Train. Iter 16200 : Commit. 0.52863 PPL. 427.77 Recons. 0.00615 +2024-03-16 22:53:27,285 INFO Train. Iter 16400 : Commit. 0.55367 PPL. 424.26 Recons. 0.00788 +2024-03-16 22:53:49,395 INFO Train. Iter 16600 : Commit. 0.50180 PPL. 427.00 Recons. 0.00598 +2024-03-16 22:54:11,905 INFO Train. Iter 16800 : Commit. 0.51817 PPL. 425.82 Recons. 0.00722 +2024-03-16 22:54:33,581 INFO Train. Iter 17000 : Commit. 0.51973 PPL. 427.90 Recons. 0.00602 +2024-03-16 22:54:55,143 INFO Train. Iter 17200 : Commit. 0.53249 PPL. 427.22 Recons. 0.00669 +2024-03-16 22:55:17,389 INFO Train. Iter 17400 : Commit. 0.50735 PPL. 427.87 Recons. 0.00608 +2024-03-16 22:55:39,035 INFO Train. Iter 17600 : Commit. 0.53079 PPL. 428.32 Recons. 0.00631 +2024-03-16 22:56:00,534 INFO Train. Iter 17800 : Commit. 0.49032 PPL. 429.05 Recons. 0.00514 +2024-03-16 22:56:22,649 INFO Train. Iter 18000 : Commit. 0.52566 PPL. 425.24 Recons. 0.00718 +2024-03-16 22:56:44,535 INFO Train. Iter 18200 : Commit. 0.52902 PPL. 425.31 Recons. 0.00690 +2024-03-16 22:57:05,626 INFO Train. Iter 18400 : Commit. 0.49299 PPL. 428.40 Recons. 0.00541 +2024-03-16 22:57:28,014 INFO Train. Iter 18600 : Commit. 0.51558 PPL. 424.20 Recons. 0.00770 +2024-03-16 22:57:49,429 INFO Train. Iter 18800 : Commit. 0.53580 PPL. 426.35 Recons. 0.00711 +2024-03-16 22:58:11,431 INFO Train. Iter 19000 : Commit. 0.49360 PPL. 428.03 Recons. 0.00540 +2024-03-16 22:58:33,528 INFO Train. Iter 19200 : Commit. 0.53710 PPL. 423.65 Recons. 0.00816 +2024-03-16 22:58:57,035 INFO Train. Iter 19400 : Commit. 0.52389 PPL. 425.48 Recons. 0.00697 +2024-03-16 22:59:20,248 INFO Train. Iter 19600 : Commit. 0.52348 PPL. 427.65 Recons. 0.00632 +2024-03-16 22:59:42,658 INFO Train. Iter 19800 : Commit. 0.47001 PPL. 429.12 Recons. 0.00470 +2024-03-16 23:00:04,892 INFO Train. Iter 20000 : Commit. 0.50208 PPL. 425.85 Recons. 0.00722 +2024-03-16 23:00:25,901 INFO Train. Iter 20200 : Commit. 0.48252 PPL. 426.24 Recons. 0.00634 +2024-03-16 23:00:47,718 INFO Train. Iter 20400 : Commit. 0.47323 PPL. 428.27 Recons. 0.00541 +2024-03-16 23:01:09,459 INFO Train. Iter 20600 : Commit. 0.44634 PPL. 430.59 Recons. 0.00414 +2024-03-16 23:01:32,221 INFO Train. Iter 20800 : Commit. 0.47040 PPL. 426.09 Recons. 0.00581 +2024-03-16 23:01:54,417 INFO Train. Iter 21000 : Commit. 0.47648 PPL. 426.63 Recons. 0.00642 +2024-03-16 23:02:16,227 INFO Train. Iter 21200 : Commit. 0.46551 PPL. 427.65 Recons. 0.00547 +2024-03-16 23:02:38,101 INFO Train. Iter 21400 : Commit. 0.47801 PPL. 427.54 Recons. 0.00605 +2024-03-16 23:02:59,815 INFO Train. Iter 21600 : Commit. 0.48404 PPL. 425.47 Recons. 0.00699 +2024-03-16 23:03:21,637 INFO Train. Iter 21800 : Commit. 0.46880 PPL. 428.85 Recons. 0.00525 +2024-03-16 23:03:43,956 INFO Train. Iter 22000 : Commit. 0.46561 PPL. 426.56 Recons. 0.00577 +2024-03-16 23:04:05,171 INFO Train. Iter 22200 : Commit. 0.46124 PPL. 428.26 Recons. 0.00560 +2024-03-16 23:04:27,177 INFO Train. Iter 22400 : Commit. 0.47759 PPL. 427.47 Recons. 0.00657 +2024-03-16 23:04:48,808 INFO Train. Iter 22600 : Commit. 0.46940 PPL. 429.69 Recons. 0.00467 +2024-03-16 23:05:10,952 INFO Train. Iter 22800 : Commit. 0.49642 PPL. 426.41 Recons. 0.00692 +2024-03-16 23:05:32,803 INFO Train. Iter 23000 : Commit. 0.45458 PPL. 426.84 Recons. 0.00544 +2024-03-16 23:05:55,247 INFO Train. Iter 23200 : Commit. 0.44674 PPL. 427.53 Recons. 0.00530 +2024-03-16 23:06:17,400 INFO Train. Iter 23400 : Commit. 0.46532 PPL. 424.88 Recons. 0.00647 +2024-03-16 23:06:39,551 INFO Train. Iter 23600 : Commit. 0.42410 PPL. 430.28 Recons. 0.00435 +2024-03-16 23:07:01,836 INFO Train. Iter 23800 : Commit. 0.45096 PPL. 427.36 Recons. 0.00588 +2024-03-16 23:07:23,844 INFO Train. Iter 24000 : Commit. 0.42862 PPL. 428.35 Recons. 0.00496 +2024-03-16 23:07:45,077 INFO Train. Iter 24200 : Commit. 0.46297 PPL. 427.77 Recons. 0.00636 +2024-03-16 23:08:07,407 INFO Train. Iter 24400 : Commit. 0.45702 PPL. 428.29 Recons. 0.00523 +2024-03-16 23:08:29,711 INFO Train. Iter 24600 : Commit. 0.43350 PPL. 430.06 Recons. 0.00438 +2024-03-16 23:08:51,928 INFO Train. Iter 24800 : Commit. 0.44663 PPL. 427.87 Recons. 0.00550 +2024-03-16 23:09:13,542 INFO Train. Iter 25000 : Commit. 0.43596 PPL. 428.46 Recons. 0.00497 +2024-03-16 23:09:35,936 INFO Train. Iter 25200 : Commit. 0.42993 PPL. 430.22 Recons. 0.00479 +2024-03-16 23:09:57,700 INFO Train. Iter 25400 : Commit. 0.45309 PPL. 426.85 Recons. 0.00600 +2024-03-16 23:10:20,399 INFO Train. Iter 25600 : Commit. 0.46649 PPL. 425.62 Recons. 0.00651 +2024-03-16 23:10:42,396 INFO Train. Iter 25800 : Commit. 0.45279 PPL. 428.27 Recons. 0.00517 +2024-03-16 23:11:03,147 INFO Train. Iter 26000 : Commit. 0.42626 PPL. 426.69 Recons. 0.00490 +2024-03-16 23:11:24,321 INFO Train. Iter 26200 : Commit. 0.41724 PPL. 428.43 Recons. 0.00510 +2024-03-16 23:11:46,394 INFO Train. Iter 26400 : Commit. 0.45614 PPL. 425.66 Recons. 0.00672 +2024-03-16 23:12:08,006 INFO Train. Iter 26600 : Commit. 0.42499 PPL. 428.67 Recons. 0.00439 +2024-03-16 23:12:30,027 INFO Train. Iter 26800 : Commit. 0.45231 PPL. 425.62 Recons. 0.00604 +2024-03-16 23:12:51,946 INFO Train. Iter 27000 : Commit. 0.44005 PPL. 429.32 Recons. 0.00475 +2024-03-16 23:13:14,636 INFO Train. Iter 27200 : Commit. 0.43448 PPL. 428.94 Recons. 0.00474 +2024-03-16 23:13:36,386 INFO Train. Iter 27400 : Commit. 0.43859 PPL. 427.85 Recons. 0.00555 +2024-03-16 23:13:58,122 INFO Train. Iter 27600 : Commit. 0.39082 PPL. 431.12 Recons. 0.00339 +2024-03-16 23:14:20,109 INFO Train. Iter 27800 : Commit. 0.41439 PPL. 428.09 Recons. 0.00518 +2024-03-16 23:14:40,963 INFO Train. Iter 28000 : Commit. 0.42482 PPL. 426.77 Recons. 0.00554 +2024-03-16 23:15:03,855 INFO Train. Iter 28200 : Commit. 0.40549 PPL. 429.38 Recons. 0.00423 +2024-03-16 23:15:26,189 INFO Train. Iter 28400 : Commit. 0.41034 PPL. 429.26 Recons. 0.00490 +2024-03-16 23:15:48,973 INFO Train. Iter 28600 : Commit. 0.39865 PPL. 428.49 Recons. 0.00446 +2024-03-16 23:16:11,381 INFO Train. Iter 28800 : Commit. 0.41967 PPL. 428.27 Recons. 0.00495 +2024-03-16 23:16:33,483 INFO Train. Iter 29000 : Commit. 0.43445 PPL. 427.95 Recons. 0.00550 +2024-03-16 23:16:55,336 INFO Train. Iter 29200 : Commit. 0.43334 PPL. 427.87 Recons. 0.00522 +2024-03-16 23:17:17,506 INFO Train. Iter 29400 : Commit. 0.41664 PPL. 426.29 Recons. 0.00522 +2024-03-16 23:17:39,695 INFO Train. Iter 29600 : Commit. 0.41505 PPL. 428.42 Recons. 0.00499 +2024-03-16 23:18:01,686 INFO Train. Iter 29800 : Commit. 0.42501 PPL. 427.57 Recons. 0.00502 +2024-03-16 23:18:22,528 INFO Train. Iter 30000 : Commit. 0.40115 PPL. 427.13 Recons. 0.00483 +2024-03-16 23:18:44,735 INFO Train. Iter 30200 : Commit. 0.39620 PPL. 427.67 Recons. 0.00469 +2024-03-16 23:19:06,832 INFO Train. Iter 30400 : Commit. 0.40644 PPL. 428.77 Recons. 0.00460 +2024-03-16 23:19:29,465 INFO Train. Iter 30600 : Commit. 0.40487 PPL. 428.08 Recons. 0.00479 +2024-03-16 23:19:51,708 INFO Train. Iter 30800 : Commit. 0.39819 PPL. 429.20 Recons. 0.00430 +2024-03-16 23:20:13,977 INFO Train. Iter 31000 : Commit. 0.42650 PPL. 425.83 Recons. 0.00600 +2024-03-16 23:20:36,205 INFO Train. Iter 31200 : Commit. 0.43334 PPL. 425.53 Recons. 0.00571 +2024-03-16 23:20:58,242 INFO Train. Iter 31400 : Commit. 0.38939 PPL. 428.64 Recons. 0.00393 +2024-03-16 23:21:20,095 INFO Train. Iter 31600 : Commit. 0.38368 PPL. 429.20 Recons. 0.00386 +2024-03-16 23:21:40,399 INFO Train. Iter 31800 : Commit. 0.36991 PPL. 429.78 Recons. 0.00375 +2024-03-16 23:22:02,350 INFO Train. Iter 32000 : Commit. 0.39576 PPL. 428.40 Recons. 0.00476 +2024-03-16 23:22:23,868 INFO Train. Iter 32200 : Commit. 0.40477 PPL. 425.45 Recons. 0.00513 +2024-03-16 23:22:46,109 INFO Train. Iter 32400 : Commit. 0.37555 PPL. 429.07 Recons. 0.00382 +2024-03-16 23:23:08,248 INFO Train. Iter 32600 : Commit. 0.38875 PPL. 425.77 Recons. 0.00507 +2024-03-16 23:23:30,053 INFO Train. Iter 32800 : Commit. 0.41287 PPL. 426.41 Recons. 0.00552 +2024-03-16 23:23:52,093 INFO Train. Iter 33000 : Commit. 0.38979 PPL. 426.10 Recons. 0.00473 +2024-03-16 23:24:13,558 INFO Train. Iter 33200 : Commit. 0.38705 PPL. 428.16 Recons. 0.00449 +2024-03-16 23:24:35,540 INFO Train. Iter 33400 : Commit. 0.36517 PPL. 429.94 Recons. 0.00357 +2024-03-16 23:24:57,786 INFO Train. Iter 33600 : Commit. 0.37901 PPL. 427.20 Recons. 0.00468 +2024-03-16 23:25:19,228 INFO Train. Iter 33800 : Commit. 0.38567 PPL. 428.87 Recons. 0.00427 +2024-03-16 23:25:41,668 INFO Train. Iter 34000 : Commit. 0.38598 PPL. 427.92 Recons. 0.00440 +2024-03-16 23:26:03,358 INFO Train. Iter 34200 : Commit. 0.39688 PPL. 425.44 Recons. 0.00541 +2024-03-16 23:26:25,219 INFO Train. Iter 34400 : Commit. 0.40510 PPL. 424.81 Recons. 0.00564 +2024-03-16 23:26:47,393 INFO Train. Iter 34600 : Commit. 0.41294 PPL. 427.83 Recons. 0.00498 +2024-03-16 23:27:09,491 INFO Train. Iter 34800 : Commit. 0.40742 PPL. 425.17 Recons. 0.00587 +2024-03-16 23:27:31,655 INFO Train. Iter 35000 : Commit. 0.37769 PPL. 429.42 Recons. 0.00373 +2024-03-16 23:27:53,993 INFO Train. Iter 35200 : Commit. 0.40334 PPL. 426.62 Recons. 0.00545 +2024-03-16 23:28:16,107 INFO Train. Iter 35400 : Commit. 0.41004 PPL. 427.43 Recons. 0.00496 +2024-03-16 23:28:38,200 INFO Train. Iter 35600 : Commit. 0.38281 PPL. 428.30 Recons. 0.00398 +2024-03-16 23:28:59,013 INFO Train. Iter 35800 : Commit. 0.38068 PPL. 428.19 Recons. 0.00444 +2024-03-16 23:29:20,850 INFO Train. Iter 36000 : Commit. 0.34625 PPL. 429.70 Recons. 0.00313 +2024-03-16 23:29:43,289 INFO Train. Iter 36200 : Commit. 0.37901 PPL. 425.60 Recons. 0.00500 +2024-03-16 23:30:05,593 INFO Train. Iter 36400 : Commit. 0.37161 PPL. 426.32 Recons. 0.00441 +2024-03-16 23:30:27,838 INFO Train. Iter 36600 : Commit. 0.38091 PPL. 425.35 Recons. 0.00497 +2024-03-16 23:30:50,080 INFO Train. Iter 36800 : Commit. 0.37372 PPL. 428.76 Recons. 0.00411 +2024-03-16 23:31:11,959 INFO Train. Iter 37000 : Commit. 0.39198 PPL. 426.50 Recons. 0.00474 +2024-03-16 23:31:33,577 INFO Train. Iter 37200 : Commit. 0.38150 PPL. 426.22 Recons. 0.00502 +2024-03-16 23:31:55,578 INFO Train. Iter 37400 : Commit. 0.37891 PPL. 426.79 Recons. 0.00457 +2024-03-16 23:32:16,957 INFO Train. Iter 37600 : Commit. 0.37890 PPL. 427.37 Recons. 0.00478 +2024-03-16 23:32:38,983 INFO Train. Iter 37800 : Commit. 0.36267 PPL. 426.74 Recons. 0.00427 +2024-03-16 23:33:00,527 INFO Train. Iter 38000 : Commit. 0.36963 PPL. 426.33 Recons. 0.00452 +2024-03-16 23:33:22,954 INFO Train. Iter 38200 : Commit. 0.36101 PPL. 428.47 Recons. 0.00396 +2024-03-16 23:33:44,718 INFO Train. Iter 38400 : Commit. 0.35999 PPL. 426.60 Recons. 0.00443 +2024-03-16 23:34:06,692 INFO Train. Iter 38600 : Commit. 0.35900 PPL. 427.99 Recons. 0.00390 +2024-03-16 23:34:28,928 INFO Train. Iter 38800 : Commit. 0.37619 PPL. 424.87 Recons. 0.00479 +2024-03-16 23:34:51,048 INFO Train. Iter 39000 : Commit. 0.36818 PPL. 424.50 Recons. 0.00489 +2024-03-16 23:35:13,108 INFO Train. Iter 39200 : Commit. 0.34897 PPL. 427.88 Recons. 0.00363 +2024-03-16 23:35:35,261 INFO Train. Iter 39400 : Commit. 0.35356 PPL. 426.66 Recons. 0.00416 +2024-03-16 23:35:56,187 INFO Train. Iter 39600 : Commit. 0.37997 PPL. 427.46 Recons. 0.00468 +2024-03-16 23:36:17,331 INFO Train. Iter 39800 : Commit. 0.36413 PPL. 428.13 Recons. 0.00384 +2024-03-16 23:36:39,093 INFO Train. Iter 40000 : Commit. 0.35219 PPL. 428.79 Recons. 0.00340 +2024-03-16 23:37:01,295 INFO Train. Iter 40200 : Commit. 0.38584 PPL. 423.74 Recons. 0.00510 +2024-03-16 23:37:22,773 INFO Train. Iter 40400 : Commit. 0.36693 PPL. 426.53 Recons. 0.00430 +2024-03-16 23:37:44,428 INFO Train. Iter 40600 : Commit. 0.37225 PPL. 424.34 Recons. 0.00471 +2024-03-16 23:38:06,002 INFO Train. Iter 40800 : Commit. 0.37235 PPL. 427.02 Recons. 0.00431 +2024-03-16 23:38:27,861 INFO Train. Iter 41000 : Commit. 0.35683 PPL. 428.38 Recons. 0.00391 +2024-03-16 23:38:49,098 INFO Train. Iter 41200 : Commit. 0.35689 PPL. 426.74 Recons. 0.00449 +2024-03-16 23:39:10,119 INFO Train. Iter 41400 : Commit. 0.35528 PPL. 426.55 Recons. 0.00441 +2024-03-16 23:39:30,430 INFO Train. Iter 41600 : Commit. 0.34460 PPL. 428.98 Recons. 0.00345 +2024-03-16 23:39:52,194 INFO Train. Iter 41800 : Commit. 0.34848 PPL. 428.68 Recons. 0.00376 +2024-03-16 23:40:13,638 INFO Train. Iter 42000 : Commit. 0.34944 PPL. 426.71 Recons. 0.00408 +2024-03-16 23:40:35,027 INFO Train. Iter 42200 : Commit. 0.33663 PPL. 427.28 Recons. 0.00386 +2024-03-16 23:40:56,256 INFO Train. Iter 42400 : Commit. 0.36061 PPL. 424.08 Recons. 0.00503 +2024-03-16 23:41:17,638 INFO Train. Iter 42600 : Commit. 0.35337 PPL. 427.95 Recons. 0.00378 +2024-03-16 23:41:38,876 INFO Train. Iter 42800 : Commit. 0.36179 PPL. 424.86 Recons. 0.00489 +2024-03-16 23:42:00,712 INFO Train. Iter 43000 : Commit. 0.34760 PPL. 428.28 Recons. 0.00342 +2024-03-16 23:42:21,768 INFO Train. Iter 43200 : Commit. 0.36377 PPL. 425.54 Recons. 0.00488 +2024-03-16 23:42:42,297 INFO Train. Iter 43400 : Commit. 0.35019 PPL. 427.25 Recons. 0.00379 +2024-03-16 23:43:03,459 INFO Train. Iter 43600 : Commit. 0.34601 PPL. 427.95 Recons. 0.00365 +2024-03-16 23:43:24,586 INFO Train. Iter 43800 : Commit. 0.35916 PPL. 425.65 Recons. 0.00478 +2024-03-16 23:43:46,317 INFO Train. Iter 44000 : Commit. 0.33773 PPL. 427.28 Recons. 0.00367 +2024-03-16 23:44:08,204 INFO Train. Iter 44200 : Commit. 0.35075 PPL. 426.44 Recons. 0.00434 +2024-03-16 23:44:29,447 INFO Train. Iter 44400 : Commit. 0.34559 PPL. 427.67 Recons. 0.00375 +2024-03-16 23:44:51,143 INFO Train. Iter 44600 : Commit. 0.33772 PPL. 428.20 Recons. 0.00352 +2024-03-16 23:45:12,204 INFO Train. Iter 44800 : Commit. 0.34349 PPL. 426.20 Recons. 0.00419 +2024-03-16 23:45:33,728 INFO Train. Iter 45000 : Commit. 0.33571 PPL. 428.17 Recons. 0.00332 +2024-03-16 23:45:55,490 INFO Train. Iter 45200 : Commit. 0.33007 PPL. 427.10 Recons. 0.00387 +2024-03-16 23:46:15,947 INFO Train. Iter 45400 : Commit. 0.34229 PPL. 426.02 Recons. 0.00424 +2024-03-16 23:46:37,494 INFO Train. Iter 45600 : Commit. 0.33368 PPL. 427.08 Recons. 0.00362 +2024-03-16 23:46:59,050 INFO Train. Iter 45800 : Commit. 0.34282 PPL. 425.71 Recons. 0.00420 +2024-03-16 23:47:20,494 INFO Train. Iter 46000 : Commit. 0.33884 PPL. 426.38 Recons. 0.00375 +2024-03-16 23:47:41,956 INFO Train. Iter 46200 : Commit. 0.34180 PPL. 426.61 Recons. 0.00385 +2024-03-16 23:48:03,381 INFO Train. Iter 46400 : Commit. 0.34302 PPL. 425.62 Recons. 0.00412 +2024-03-16 23:48:24,330 INFO Train. Iter 46600 : Commit. 0.34600 PPL. 426.09 Recons. 0.00406 +2024-03-16 23:48:45,609 INFO Train. Iter 46800 : Commit. 0.33497 PPL. 426.64 Recons. 0.00389 +2024-03-16 23:49:07,375 INFO Train. Iter 47000 : Commit. 0.34520 PPL. 426.40 Recons. 0.00394 +2024-03-16 23:49:28,945 INFO Train. Iter 47200 : Commit. 0.32930 PPL. 426.91 Recons. 0.00382 +2024-03-16 23:49:49,802 INFO Train. Iter 47400 : Commit. 0.35042 PPL. 426.15 Recons. 0.00422 +2024-03-16 23:50:11,739 INFO Train. Iter 47600 : Commit. 0.35024 PPL. 425.97 Recons. 0.00435 +2024-03-16 23:50:33,257 INFO Train. Iter 47800 : Commit. 0.34299 PPL. 427.30 Recons. 0.00360 +2024-03-16 23:50:54,799 INFO Train. Iter 48000 : Commit. 0.34683 PPL. 425.86 Recons. 0.00410 +2024-03-16 23:51:16,117 INFO Train. Iter 48200 : Commit. 0.33252 PPL. 427.38 Recons. 0.00356 +2024-03-16 23:51:37,571 INFO Train. Iter 48400 : Commit. 0.34021 PPL. 424.29 Recons. 0.00450 +2024-03-16 23:51:59,141 INFO Train. Iter 48600 : Commit. 0.33753 PPL. 427.60 Recons. 0.00361 +2024-03-16 23:52:20,482 INFO Train. Iter 48800 : Commit. 0.34689 PPL. 424.65 Recons. 0.00443 +2024-03-16 23:52:42,325 INFO Train. Iter 49000 : Commit. 0.33115 PPL. 427.04 Recons. 0.00354 +2024-03-16 23:53:02,354 INFO Train. Iter 49200 : Commit. 0.33894 PPL. 425.42 Recons. 0.00444 +2024-03-16 23:53:24,056 INFO Train. Iter 49400 : Commit. 0.32594 PPL. 428.39 Recons. 0.00311 +2024-03-16 23:53:45,449 INFO Train. Iter 49600 : Commit. 0.32648 PPL. 425.45 Recons. 0.00383 +2024-03-16 23:54:06,920 INFO Train. Iter 49800 : Commit. 0.32422 PPL. 427.29 Recons. 0.00344 +2024-03-16 23:54:28,758 INFO Train. Iter 50000 : Commit. 0.33942 PPL. 424.96 Recons. 0.00441 +2024-03-16 23:54:50,328 INFO Train. Iter 50200 : Commit. 0.33388 PPL. 426.84 Recons. 0.00372 +2024-03-16 23:55:11,801 INFO Train. Iter 50400 : Commit. 0.31985 PPL. 427.03 Recons. 0.00341 +2024-03-16 23:55:33,294 INFO Train. Iter 50600 : Commit. 0.33058 PPL. 425.18 Recons. 0.00406 +2024-03-16 23:55:54,631 INFO Train. Iter 50800 : Commit. 0.32595 PPL. 425.51 Recons. 0.00403 +2024-03-16 23:56:15,643 INFO Train. Iter 51000 : Commit. 0.31796 PPL. 427.56 Recons. 0.00338 +2024-03-16 23:56:35,958 INFO Train. Iter 51200 : Commit. 0.33179 PPL. 426.38 Recons. 0.00396 +2024-03-16 23:56:57,254 INFO Train. Iter 51400 : Commit. 0.32981 PPL. 426.97 Recons. 0.00373 +2024-03-16 23:57:18,401 INFO Train. Iter 51600 : Commit. 0.33076 PPL. 425.45 Recons. 0.00408 +2024-03-16 23:57:39,843 INFO Train. Iter 51800 : Commit. 0.32198 PPL. 426.55 Recons. 0.00357 +2024-03-16 23:58:00,991 INFO Train. Iter 52000 : Commit. 0.33342 PPL. 426.76 Recons. 0.00410 +2024-03-16 23:58:22,700 INFO Train. Iter 52200 : Commit. 0.33335 PPL. 426.65 Recons. 0.00373 +2024-03-16 23:58:43,922 INFO Train. Iter 52400 : Commit. 0.33055 PPL. 423.69 Recons. 0.00431 +2024-03-16 23:59:05,595 INFO Train. Iter 52600 : Commit. 0.32366 PPL. 426.16 Recons. 0.00354 +2024-03-16 23:59:26,768 INFO Train. Iter 52800 : Commit. 0.31630 PPL. 427.14 Recons. 0.00346 +2024-03-16 23:59:48,104 INFO Train. Iter 53000 : Commit. 0.30642 PPL. 428.88 Recons. 0.00276 +2024-03-17 00:00:08,265 INFO Train. Iter 53200 : Commit. 0.31952 PPL. 426.56 Recons. 0.00381 +2024-03-17 00:00:29,853 INFO Train. Iter 53400 : Commit. 0.31646 PPL. 425.54 Recons. 0.00367 +2024-03-17 00:00:51,137 INFO Train. Iter 53600 : Commit. 0.31074 PPL. 427.26 Recons. 0.00314 +2024-03-17 00:01:12,647 INFO Train. Iter 53800 : Commit. 0.31120 PPL. 425.18 Recons. 0.00342 +2024-03-17 00:01:33,696 INFO Train. Iter 54000 : Commit. 0.33636 PPL. 422.90 Recons. 0.00465 +2024-03-17 00:01:55,483 INFO Train. Iter 54200 : Commit. 0.32624 PPL. 424.41 Recons. 0.00372 +2024-03-17 00:02:16,936 INFO Train. Iter 54400 : Commit. 0.33169 PPL. 426.47 Recons. 0.00391 +2024-03-17 00:02:38,292 INFO Train. Iter 54600 : Commit. 0.31032 PPL. 424.93 Recons. 0.00370 +2024-03-17 00:02:59,747 INFO Train. Iter 54800 : Commit. 0.32089 PPL. 426.44 Recons. 0.00368 +2024-03-17 00:03:20,090 INFO Train. Iter 55000 : Commit. 0.31265 PPL. 425.55 Recons. 0.00373 +2024-03-17 00:03:41,353 INFO Train. Iter 55200 : Commit. 0.31939 PPL. 423.84 Recons. 0.00390 +2024-03-17 00:04:02,485 INFO Train. Iter 55400 : Commit. 0.31000 PPL. 427.06 Recons. 0.00315 +2024-03-17 00:04:23,781 INFO Train. Iter 55600 : Commit. 0.30695 PPL. 428.41 Recons. 0.00287 +2024-03-17 00:04:45,433 INFO Train. Iter 55800 : Commit. 0.31398 PPL. 425.44 Recons. 0.00378 +2024-03-17 00:05:07,338 INFO Train. Iter 56000 : Commit. 0.31767 PPL. 427.02 Recons. 0.00328 +2024-03-17 00:05:29,644 INFO Train. Iter 56200 : Commit. 0.30254 PPL. 427.33 Recons. 0.00314 +2024-03-17 00:05:51,783 INFO Train. Iter 56400 : Commit. 0.30341 PPL. 425.38 Recons. 0.00337 +2024-03-17 00:06:13,742 INFO Train. Iter 56600 : Commit. 0.30628 PPL. 426.25 Recons. 0.00329 +2024-03-17 00:06:34,927 INFO Train. Iter 56800 : Commit. 0.31321 PPL. 422.98 Recons. 0.00383 +2024-03-17 00:06:55,006 INFO Train. Iter 57000 : Commit. 0.30884 PPL. 426.60 Recons. 0.00346 +2024-03-17 00:07:16,556 INFO Train. Iter 57200 : Commit. 0.29509 PPL. 428.08 Recons. 0.00268 +2024-03-17 00:07:37,831 INFO Train. Iter 57400 : Commit. 0.31226 PPL. 424.61 Recons. 0.00391 +2024-03-17 00:07:59,281 INFO Train. Iter 57600 : Commit. 0.31054 PPL. 426.09 Recons. 0.00348 +2024-03-17 00:08:20,832 INFO Train. Iter 57800 : Commit. 0.31828 PPL. 423.09 Recons. 0.00417 +2024-03-17 00:08:41,930 INFO Train. Iter 58000 : Commit. 0.30872 PPL. 425.51 Recons. 0.00355 +2024-03-17 00:09:02,863 INFO Train. Iter 58200 : Commit. 0.30060 PPL. 427.02 Recons. 0.00286 +2024-03-17 00:09:24,173 INFO Train. Iter 58400 : Commit. 0.29951 PPL. 426.03 Recons. 0.00318 +2024-03-17 00:09:45,581 INFO Train. Iter 58600 : Commit. 0.31303 PPL. 425.30 Recons. 0.00374 +2024-03-17 00:10:07,276 INFO Train. Iter 58800 : Commit. 0.30482 PPL. 425.95 Recons. 0.00321 +2024-03-17 00:10:27,561 INFO Train. Iter 59000 : Commit. 0.31600 PPL. 426.55 Recons. 0.00340 +2024-03-17 00:10:49,005 INFO Train. Iter 59200 : Commit. 0.29628 PPL. 428.35 Recons. 0.00255 +2024-03-17 00:11:10,708 INFO Train. Iter 59400 : Commit. 0.30996 PPL. 423.78 Recons. 0.00394 +2024-03-17 00:11:32,300 INFO Train. Iter 59600 : Commit. 0.31807 PPL. 425.20 Recons. 0.00381 +2024-03-17 00:11:53,761 INFO Train. Iter 59800 : Commit. 0.31272 PPL. 425.44 Recons. 0.00350 +2024-03-17 00:12:15,773 INFO Train. Iter 60000 : Commit. 0.31595 PPL. 425.27 Recons. 0.00362 +2024-03-17 00:12:37,491 INFO Train. Iter 60200 : Commit. 0.29355 PPL. 427.37 Recons. 0.00279 +2024-03-17 00:12:58,875 INFO Train. Iter 60400 : Commit. 0.30534 PPL. 426.08 Recons. 0.00358 +2024-03-17 00:13:20,398 INFO Train. Iter 60600 : Commit. 0.30325 PPL. 426.41 Recons. 0.00326 +2024-03-17 00:13:40,475 INFO Train. Iter 60800 : Commit. 0.29292 PPL. 425.95 Recons. 0.00295 +2024-03-17 00:14:01,655 INFO Train. Iter 61000 : Commit. 0.31454 PPL. 424.12 Recons. 0.00387 +2024-03-17 00:14:23,135 INFO Train. Iter 61200 : Commit. 0.30363 PPL. 425.62 Recons. 0.00322 +2024-03-17 00:14:44,997 INFO Train. Iter 61400 : Commit. 0.29802 PPL. 427.68 Recons. 0.00293 +2024-03-17 00:15:06,400 INFO Train. Iter 61600 : Commit. 0.30777 PPL. 425.65 Recons. 0.00362 +2024-03-17 00:15:27,504 INFO Train. Iter 61800 : Commit. 0.30448 PPL. 424.30 Recons. 0.00347 +2024-03-17 00:15:49,332 INFO Train. Iter 62000 : Commit. 0.28902 PPL. 426.51 Recons. 0.00303 +2024-03-17 00:16:10,658 INFO Train. Iter 62200 : Commit. 0.30189 PPL. 425.87 Recons. 0.00321 +2024-03-17 00:16:32,002 INFO Train. Iter 62400 : Commit. 0.30994 PPL. 424.32 Recons. 0.00380 +2024-03-17 00:16:53,251 INFO Train. Iter 62600 : Commit. 0.28413 PPL. 427.96 Recons. 0.00248 +2024-03-17 00:17:13,575 INFO Train. Iter 62800 : Commit. 0.30649 PPL. 423.48 Recons. 0.00427 +2024-03-17 00:17:35,482 INFO Train. Iter 63000 : Commit. 0.29609 PPL. 426.63 Recons. 0.00297 +2024-03-17 00:17:56,839 INFO Train. Iter 63200 : Commit. 0.29304 PPL. 426.80 Recons. 0.00314 +2024-03-17 00:18:17,965 INFO Train. Iter 63400 : Commit. 0.30103 PPL. 425.61 Recons. 0.00330 +2024-03-17 00:18:39,011 INFO Train. Iter 63600 : Commit. 0.30031 PPL. 425.58 Recons. 0.00321 +2024-03-17 00:19:00,421 INFO Train. Iter 63800 : Commit. 0.29153 PPL. 426.30 Recons. 0.00288 +2024-03-17 00:19:22,560 INFO Train. Iter 64000 : Commit. 0.29393 PPL. 425.61 Recons. 0.00306 +2024-03-17 00:19:44,772 INFO Train. Iter 64200 : Commit. 0.29372 PPL. 424.43 Recons. 0.00341 +2024-03-17 00:20:06,676 INFO Train. Iter 64400 : Commit. 0.27910 PPL. 427.35 Recons. 0.00252 +2024-03-17 00:20:28,753 INFO Train. Iter 64600 : Commit. 0.28116 PPL. 427.86 Recons. 0.00255 +2024-03-17 00:20:49,891 INFO Train. Iter 64800 : Commit. 0.28762 PPL. 425.86 Recons. 0.00298 +2024-03-17 00:21:12,351 INFO Train. Iter 65000 : Commit. 0.29560 PPL. 425.36 Recons. 0.00333 +2024-03-17 00:21:34,100 INFO Train. Iter 65200 : Commit. 0.28471 PPL. 426.21 Recons. 0.00267 +2024-03-17 00:21:56,325 INFO Train. Iter 65400 : Commit. 0.30224 PPL. 422.40 Recons. 0.00397 +2024-03-17 00:22:17,585 INFO Train. Iter 65600 : Commit. 0.30268 PPL. 427.33 Recons. 0.00288 +2024-03-17 00:22:39,904 INFO Train. Iter 65800 : Commit. 0.28603 PPL. 426.06 Recons. 0.00274 +2024-03-17 00:23:02,343 INFO Train. Iter 66000 : Commit. 0.28769 PPL. 427.13 Recons. 0.00291 +2024-03-17 00:23:24,827 INFO Train. Iter 66200 : Commit. 0.29296 PPL. 425.80 Recons. 0.00315 +2024-03-17 00:23:46,556 INFO Train. Iter 66400 : Commit. 0.29889 PPL. 424.83 Recons. 0.00369 +2024-03-17 00:24:07,496 INFO Train. Iter 66600 : Commit. 0.29098 PPL. 426.45 Recons. 0.00284 +2024-03-17 00:24:29,866 INFO Train. Iter 66800 : Commit. 0.28567 PPL. 424.19 Recons. 0.00323 +2024-03-17 00:24:50,972 INFO Train. Iter 67000 : Commit. 0.28543 PPL. 426.90 Recons. 0.00289 +2024-03-17 00:25:12,609 INFO Train. Iter 67200 : Commit. 0.28151 PPL. 427.15 Recons. 0.00276 +2024-03-17 00:25:33,848 INFO Train. Iter 67400 : Commit. 0.28150 PPL. 427.01 Recons. 0.00287 +2024-03-17 00:25:54,885 INFO Train. Iter 67600 : Commit. 0.29837 PPL. 424.18 Recons. 0.00381 +2024-03-17 00:26:16,114 INFO Train. Iter 67800 : Commit. 0.29332 PPL. 425.11 Recons. 0.00317 +2024-03-17 00:26:37,195 INFO Train. Iter 68000 : Commit. 0.29371 PPL. 425.02 Recons. 0.00332 +2024-03-17 00:26:58,228 INFO Train. Iter 68200 : Commit. 0.29521 PPL. 423.87 Recons. 0.00341 +2024-03-17 00:27:19,654 INFO Train. Iter 68400 : Commit. 0.29675 PPL. 424.92 Recons. 0.00335 +2024-03-17 00:27:39,957 INFO Train. Iter 68600 : Commit. 0.28235 PPL. 426.62 Recons. 0.00268 +2024-03-17 00:28:01,094 INFO Train. Iter 68800 : Commit. 0.29669 PPL. 424.44 Recons. 0.00360 +2024-03-17 00:28:22,137 INFO Train. Iter 69000 : Commit. 0.28363 PPL. 426.84 Recons. 0.00261 +2024-03-17 00:28:43,187 INFO Train. Iter 69200 : Commit. 0.28687 PPL. 425.66 Recons. 0.00304 +2024-03-17 00:29:04,570 INFO Train. Iter 69400 : Commit. 0.29912 PPL. 422.57 Recons. 0.00376 +2024-03-17 00:29:25,906 INFO Train. Iter 69600 : Commit. 0.29375 PPL. 425.46 Recons. 0.00318 +2024-03-17 00:29:47,127 INFO Train. Iter 69800 : Commit. 0.28446 PPL. 427.12 Recons. 0.00271 +2024-03-17 00:30:08,557 INFO Train. Iter 70000 : Commit. 0.29346 PPL. 422.67 Recons. 0.00370 +2024-03-17 00:30:29,895 INFO Train. Iter 70200 : Commit. 0.27410 PPL. 427.22 Recons. 0.00243 +2024-03-17 00:30:51,331 INFO Train. Iter 70400 : Commit. 0.27895 PPL. 426.08 Recons. 0.00283 +2024-03-17 00:31:11,459 INFO Train. Iter 70600 : Commit. 0.29294 PPL. 423.74 Recons. 0.00357 +2024-03-17 00:31:32,849 INFO Train. Iter 70800 : Commit. 0.29588 PPL. 424.42 Recons. 0.00340 +2024-03-17 00:31:54,288 INFO Train. Iter 71000 : Commit. 0.29197 PPL. 424.24 Recons. 0.00328 +2024-03-17 00:32:15,991 INFO Train. Iter 71200 : Commit. 0.29050 PPL. 425.18 Recons. 0.00308 +2024-03-17 00:32:37,282 INFO Train. Iter 71400 : Commit. 0.28593 PPL. 425.78 Recons. 0.00301 +2024-03-17 00:32:59,175 INFO Train. Iter 71600 : Commit. 0.29864 PPL. 424.35 Recons. 0.00344 +2024-03-17 00:33:20,386 INFO Train. Iter 71800 : Commit. 0.28923 PPL. 424.68 Recons. 0.00322 +2024-03-17 00:33:41,835 INFO Train. Iter 72000 : Commit. 0.29405 PPL. 423.17 Recons. 0.00369 +2024-03-17 00:34:02,984 INFO Train. Iter 72200 : Commit. 0.27601 PPL. 425.95 Recons. 0.00292 +2024-03-17 00:34:23,104 INFO Train. Iter 72400 : Commit. 0.28588 PPL. 425.12 Recons. 0.00315 +2024-03-17 00:34:44,166 INFO Train. Iter 72600 : Commit. 0.28925 PPL. 424.99 Recons. 0.00308 +2024-03-17 00:35:05,856 INFO Train. Iter 72800 : Commit. 0.28435 PPL. 425.87 Recons. 0.00305 +2024-03-17 00:35:27,407 INFO Train. Iter 73000 : Commit. 0.28587 PPL. 426.38 Recons. 0.00279 +2024-03-17 00:35:49,068 INFO Train. Iter 73200 : Commit. 0.28186 PPL. 424.39 Recons. 0.00319 +2024-03-17 00:36:10,683 INFO Train. Iter 73400 : Commit. 0.28463 PPL. 424.30 Recons. 0.00335 +2024-03-17 00:36:31,776 INFO Train. Iter 73600 : Commit. 0.26917 PPL. 426.72 Recons. 0.00233 +2024-03-17 00:36:53,150 INFO Train. Iter 73800 : Commit. 0.28465 PPL. 425.09 Recons. 0.00327 +2024-03-17 00:37:14,680 INFO Train. Iter 74000 : Commit. 0.27843 PPL. 426.08 Recons. 0.00276 +2024-03-17 00:37:36,525 INFO Train. Iter 74200 : Commit. 0.28170 PPL. 424.92 Recons. 0.00317 +2024-03-17 00:37:57,605 INFO Train. Iter 74400 : Commit. 0.28993 PPL. 421.92 Recons. 0.00382 +2024-03-17 00:38:19,760 INFO Train. Iter 74600 : Commit. 0.27464 PPL. 426.51 Recons. 0.00242 +2024-03-17 00:38:41,367 INFO Train. Iter 74800 : Commit. 0.27920 PPL. 424.71 Recons. 0.00296 +2024-03-17 00:39:02,889 INFO Train. Iter 75000 : Commit. 0.26877 PPL. 426.55 Recons. 0.00240 +2024-03-17 00:39:23,879 INFO Train. Iter 75200 : Commit. 0.27115 PPL. 426.24 Recons. 0.00261 +2024-03-17 00:39:45,121 INFO Train. Iter 75400 : Commit. 0.28620 PPL. 423.60 Recons. 0.00369 +2024-03-17 00:40:06,608 INFO Train. Iter 75600 : Commit. 0.28032 PPL. 424.76 Recons. 0.00287 +2024-03-17 00:40:28,041 INFO Train. Iter 75800 : Commit. 0.28191 PPL. 423.67 Recons. 0.00321 +2024-03-17 00:40:49,500 INFO Train. Iter 76000 : Commit. 0.27492 PPL. 426.73 Recons. 0.00256 +2024-03-17 00:41:10,874 INFO Train. Iter 76200 : Commit. 0.27021 PPL. 426.49 Recons. 0.00250 +2024-03-17 00:41:30,943 INFO Train. Iter 76400 : Commit. 0.27753 PPL. 424.79 Recons. 0.00294 +2024-03-17 00:41:52,601 INFO Train. Iter 76600 : Commit. 0.27982 PPL. 424.42 Recons. 0.00306 +2024-03-17 00:42:14,124 INFO Train. Iter 76800 : Commit. 0.28119 PPL. 424.16 Recons. 0.00313 +2024-03-17 00:42:35,790 INFO Train. Iter 77000 : Commit. 0.27652 PPL. 425.45 Recons. 0.00278 +2024-03-17 00:42:57,120 INFO Train. Iter 77200 : Commit. 0.28173 PPL. 423.48 Recons. 0.00317 +2024-03-17 00:43:18,089 INFO Train. Iter 77400 : Commit. 0.27103 PPL. 425.07 Recons. 0.00273 +2024-03-17 00:43:39,781 INFO Train. Iter 77600 : Commit. 0.26339 PPL. 426.22 Recons. 0.00239 +2024-03-17 00:44:01,072 INFO Train. Iter 77800 : Commit. 0.26642 PPL. 424.77 Recons. 0.00283 +2024-03-17 00:44:22,548 INFO Train. Iter 78000 : Commit. 0.26886 PPL. 425.51 Recons. 0.00249 +2024-03-17 00:44:42,121 INFO Train. Iter 78200 : Commit. 0.27390 PPL. 422.96 Recons. 0.00309 +2024-03-17 00:45:03,320 INFO Train. Iter 78400 : Commit. 0.27813 PPL. 424.64 Recons. 0.00298 +2024-03-17 00:45:24,764 INFO Train. Iter 78600 : Commit. 0.27143 PPL. 425.90 Recons. 0.00239 +2024-03-17 00:45:46,908 INFO Train. Iter 78800 : Commit. 0.27568 PPL. 423.76 Recons. 0.00305 +2024-03-17 00:46:08,792 INFO Train. Iter 79000 : Commit. 0.28152 PPL. 424.60 Recons. 0.00314 +2024-03-17 00:46:30,270 INFO Train. Iter 79200 : Commit. 0.26171 PPL. 426.47 Recons. 0.00238 +2024-03-17 00:46:51,520 INFO Train. Iter 79400 : Commit. 0.28634 PPL. 423.15 Recons. 0.00363 +2024-03-17 00:47:13,341 INFO Train. Iter 79600 : Commit. 0.28003 PPL. 422.67 Recons. 0.00320 +2024-03-17 00:47:35,352 INFO Train. Iter 79800 : Commit. 0.27943 PPL. 425.57 Recons. 0.00262 +2024-03-17 00:47:57,052 INFO Train. Iter 80000 : Commit. 0.28996 PPL. 421.08 Recons. 0.00404 +2024-03-17 00:48:17,800 INFO Train. Iter 80200 : Commit. 0.27995 PPL. 424.48 Recons. 0.00299 +2024-03-17 00:48:39,911 INFO Train. Iter 80400 : Commit. 0.27182 PPL. 426.40 Recons. 0.00244 +2024-03-17 00:49:01,789 INFO Train. Iter 80600 : Commit. 0.27347 PPL. 424.71 Recons. 0.00309 +2024-03-17 00:49:23,976 INFO Train. Iter 80800 : Commit. 0.28208 PPL. 423.91 Recons. 0.00321 +2024-03-17 00:49:46,356 INFO Train. Iter 81000 : Commit. 0.27335 PPL. 425.44 Recons. 0.00263 +2024-03-17 00:50:08,774 INFO Train. Iter 81200 : Commit. 0.28184 PPL. 424.89 Recons. 0.00308 +2024-03-17 00:50:30,860 INFO Train. Iter 81400 : Commit. 0.27506 PPL. 426.23 Recons. 0.00262 +2024-03-17 00:50:52,288 INFO Train. Iter 81600 : Commit. 0.28374 PPL. 423.00 Recons. 0.00342 +2024-03-17 00:51:13,725 INFO Train. Iter 81800 : Commit. 0.28429 PPL. 424.69 Recons. 0.00295 +2024-03-17 00:51:35,762 INFO Train. Iter 82000 : Commit. 0.29027 PPL. 421.76 Recons. 0.00367 +2024-03-17 00:51:56,611 INFO Train. Iter 82200 : Commit. 0.27892 PPL. 424.01 Recons. 0.00293 +2024-03-17 00:52:17,999 INFO Train. Iter 82400 : Commit. 0.27434 PPL. 425.41 Recons. 0.00264 +2024-03-17 00:52:39,689 INFO Train. Iter 82600 : Commit. 0.28547 PPL. 422.95 Recons. 0.00337 +2024-03-17 00:53:01,860 INFO Train. Iter 82800 : Commit. 0.27940 PPL. 425.06 Recons. 0.00292 +2024-03-17 00:53:24,380 INFO Train. Iter 83000 : Commit. 0.27999 PPL. 424.19 Recons. 0.00306 +2024-03-17 00:53:46,703 INFO Train. Iter 83200 : Commit. 0.27397 PPL. 425.72 Recons. 0.00273 +2024-03-17 00:54:08,625 INFO Train. Iter 83400 : Commit. 0.26704 PPL. 426.54 Recons. 0.00253 +2024-03-17 00:54:30,331 INFO Train. Iter 83600 : Commit. 0.27725 PPL. 424.57 Recons. 0.00307 +2024-03-17 00:54:52,376 INFO Train. Iter 83800 : Commit. 0.28306 PPL. 423.15 Recons. 0.00328 +2024-03-17 00:55:13,163 INFO Train. Iter 84000 : Commit. 0.27448 PPL. 423.96 Recons. 0.00311 +2024-03-17 00:55:35,537 INFO Train. Iter 84200 : Commit. 0.26042 PPL. 425.61 Recons. 0.00244 +2024-03-17 00:55:56,978 INFO Train. Iter 84400 : Commit. 0.28124 PPL. 422.22 Recons. 0.00322 +2024-03-17 00:56:19,020 INFO Train. Iter 84600 : Commit. 0.28425 PPL. 424.46 Recons. 0.00298 +2024-03-17 00:56:40,955 INFO Train. Iter 84800 : Commit. 0.27620 PPL. 424.83 Recons. 0.00291 +2024-03-17 00:57:02,722 INFO Train. Iter 85000 : Commit. 0.27771 PPL. 423.91 Recons. 0.00308 +2024-03-17 00:57:24,359 INFO Train. Iter 85200 : Commit. 0.26645 PPL. 425.80 Recons. 0.00249 +2024-03-17 00:57:46,308 INFO Train. Iter 85400 : Commit. 0.26609 PPL. 426.02 Recons. 0.00250 +2024-03-17 00:58:08,204 INFO Train. Iter 85600 : Commit. 0.28282 PPL. 421.33 Recons. 0.00353 +2024-03-17 00:58:30,427 INFO Train. Iter 85800 : Commit. 0.26702 PPL. 426.05 Recons. 0.00240 +2024-03-17 00:58:51,837 INFO Train. Iter 86000 : Commit. 0.27675 PPL. 422.01 Recons. 0.00315 +2024-03-17 00:59:14,109 INFO Train. Iter 86200 : Commit. 0.27769 PPL. 422.48 Recons. 0.00310 +2024-03-17 00:59:36,557 INFO Train. Iter 86400 : Commit. 0.27576 PPL. 424.24 Recons. 0.00306 +2024-03-17 00:59:58,029 INFO Train. Iter 86600 : Commit. 0.27315 PPL. 424.45 Recons. 0.00278 +2024-03-17 01:00:19,278 INFO Train. Iter 86800 : Commit. 0.27065 PPL. 425.35 Recons. 0.00268 +2024-03-17 01:00:40,972 INFO Train. Iter 87000 : Commit. 0.28055 PPL. 423.76 Recons. 0.00330 +2024-03-17 01:01:02,694 INFO Train. Iter 87200 : Commit. 0.27198 PPL. 424.01 Recons. 0.00295 +2024-03-17 01:01:24,285 INFO Train. Iter 87400 : Commit. 0.26836 PPL. 425.73 Recons. 0.00260 +2024-03-17 01:01:45,898 INFO Train. Iter 87600 : Commit. 0.27626 PPL. 424.34 Recons. 0.00290 +2024-03-17 01:02:07,412 INFO Train. Iter 87800 : Commit. 0.27291 PPL. 423.85 Recons. 0.00297 +2024-03-17 01:02:27,889 INFO Train. Iter 88000 : Commit. 0.27860 PPL. 423.27 Recons. 0.00305 +2024-03-17 01:02:49,431 INFO Train. Iter 88200 : Commit. 0.28331 PPL. 422.69 Recons. 0.00340 +2024-03-17 01:03:10,709 INFO Train. Iter 88400 : Commit. 0.27181 PPL. 423.92 Recons. 0.00290 +2024-03-17 01:03:32,044 INFO Train. Iter 88600 : Commit. 0.27069 PPL. 424.63 Recons. 0.00276 +2024-03-17 01:03:53,185 INFO Train. Iter 88800 : Commit. 0.27259 PPL. 424.46 Recons. 0.00291 +2024-03-17 01:04:14,747 INFO Train. Iter 89000 : Commit. 0.26130 PPL. 425.58 Recons. 0.00250 +2024-03-17 01:04:36,017 INFO Train. Iter 89200 : Commit. 0.27292 PPL. 423.34 Recons. 0.00313 +2024-03-17 01:04:57,158 INFO Train. Iter 89400 : Commit. 0.26438 PPL. 425.41 Recons. 0.00250 +2024-03-17 01:05:18,432 INFO Train. Iter 89600 : Commit. 0.29084 PPL. 423.56 Recons. 0.00346 +2024-03-17 01:05:38,472 INFO Train. Iter 89800 : Commit. 0.26195 PPL. 426.13 Recons. 0.00225 +2024-03-17 01:05:59,848 INFO Train. Iter 90000 : Commit. 0.25655 PPL. 424.24 Recons. 0.00244 +2024-03-17 01:06:21,120 INFO Train. Iter 90200 : Commit. 0.28345 PPL. 422.42 Recons. 0.00363 +2024-03-17 01:06:42,525 INFO Train. Iter 90400 : Commit. 0.25824 PPL. 426.33 Recons. 0.00219 +2024-03-17 01:07:04,093 INFO Train. Iter 90600 : Commit. 0.25808 PPL. 426.33 Recons. 0.00239 +2024-03-17 01:07:25,566 INFO Train. Iter 90800 : Commit. 0.26949 PPL. 423.71 Recons. 0.00320 +2024-03-17 01:07:47,939 INFO Train. Iter 91000 : Commit. 0.26604 PPL. 425.24 Recons. 0.00269 +2024-03-17 01:08:09,519 INFO Train. Iter 91200 : Commit. 0.26400 PPL. 425.21 Recons. 0.00265 +2024-03-17 01:08:31,418 INFO Train. Iter 91400 : Commit. 0.26789 PPL. 425.17 Recons. 0.00265 +2024-03-17 01:08:53,411 INFO Train. Iter 91600 : Commit. 0.27000 PPL. 424.49 Recons. 0.00287 +2024-03-17 01:09:14,045 INFO Train. Iter 91800 : Commit. 0.27154 PPL. 422.59 Recons. 0.00318 +2024-03-17 01:09:35,588 INFO Train. Iter 92000 : Commit. 0.26713 PPL. 425.73 Recons. 0.00257 +2024-03-17 01:09:57,235 INFO Train. Iter 92200 : Commit. 0.26273 PPL. 425.34 Recons. 0.00262 +2024-03-17 01:10:18,762 INFO Train. Iter 92400 : Commit. 0.25947 PPL. 424.97 Recons. 0.00257 +2024-03-17 01:10:40,324 INFO Train. Iter 92600 : Commit. 0.27030 PPL. 425.14 Recons. 0.00265 +2024-03-17 01:11:01,862 INFO Train. Iter 92800 : Commit. 0.27065 PPL. 422.79 Recons. 0.00328 +2024-03-17 01:11:23,447 INFO Train. Iter 93000 : Commit. 0.26309 PPL. 425.94 Recons. 0.00233 +2024-03-17 01:11:44,916 INFO Train. Iter 93200 : Commit. 0.26253 PPL. 423.57 Recons. 0.00281 +2024-03-17 01:12:06,525 INFO Train. Iter 93400 : Commit. 0.26106 PPL. 424.94 Recons. 0.00249 +2024-03-17 01:12:27,531 INFO Train. Iter 93600 : Commit. 0.26175 PPL. 422.92 Recons. 0.00280 +2024-03-17 01:12:47,745 INFO Train. Iter 93800 : Commit. 0.25352 PPL. 426.05 Recons. 0.00222 +2024-03-17 01:13:09,326 INFO Train. Iter 94000 : Commit. 0.25586 PPL. 424.64 Recons. 0.00270 +2024-03-17 01:13:30,470 INFO Train. Iter 94200 : Commit. 0.26511 PPL. 424.18 Recons. 0.00289 +2024-03-17 01:13:51,905 INFO Train. Iter 94400 : Commit. 0.26553 PPL. 424.44 Recons. 0.00283 +2024-03-17 01:14:13,459 INFO Train. Iter 94600 : Commit. 0.26372 PPL. 424.15 Recons. 0.00286 +2024-03-17 01:14:35,132 INFO Train. Iter 94800 : Commit. 0.25805 PPL. 425.24 Recons. 0.00235 +2024-03-17 01:14:56,528 INFO Train. Iter 95000 : Commit. 0.25832 PPL. 424.28 Recons. 0.00280 +2024-03-17 01:15:18,178 INFO Train. Iter 95200 : Commit. 0.25775 PPL. 424.87 Recons. 0.00251 +2024-03-17 01:15:39,199 INFO Train. Iter 95400 : Commit. 0.26613 PPL. 423.44 Recons. 0.00299 +2024-03-17 01:15:59,554 INFO Train. Iter 95600 : Commit. 0.26353 PPL. 423.17 Recons. 0.00290 +2024-03-17 01:16:21,031 INFO Train. Iter 95800 : Commit. 0.27011 PPL. 422.69 Recons. 0.00301 +2024-03-17 01:16:42,936 INFO Train. Iter 96000 : Commit. 0.26744 PPL. 424.32 Recons. 0.00261 +2024-03-17 01:17:04,056 INFO Train. Iter 96200 : Commit. 0.25995 PPL. 425.19 Recons. 0.00244 +2024-03-17 01:17:25,874 INFO Train. Iter 96400 : Commit. 0.26513 PPL. 422.87 Recons. 0.00317 +2024-03-17 01:17:47,793 INFO Train. Iter 96600 : Commit. 0.26075 PPL. 424.36 Recons. 0.00256 +2024-03-17 01:18:09,348 INFO Train. Iter 96800 : Commit. 0.25487 PPL. 424.63 Recons. 0.00238 +2024-03-17 01:18:31,317 INFO Train. Iter 97000 : Commit. 0.27190 PPL. 422.11 Recons. 0.00332 +2024-03-17 01:18:53,422 INFO Train. Iter 97200 : Commit. 0.25952 PPL. 423.36 Recons. 0.00259 +2024-03-17 01:19:15,869 INFO Train. Iter 97400 : Commit. 0.26189 PPL. 421.93 Recons. 0.00302 +2024-03-17 01:19:36,718 INFO Train. Iter 97600 : Commit. 0.26080 PPL. 424.00 Recons. 0.00259 +2024-03-17 01:19:58,517 INFO Train. Iter 97800 : Commit. 0.25870 PPL. 423.63 Recons. 0.00264 +2024-03-17 01:20:20,126 INFO Train. Iter 98000 : Commit. 0.24840 PPL. 425.97 Recons. 0.00208 +2024-03-17 01:20:42,235 INFO Train. Iter 98200 : Commit. 0.26452 PPL. 424.29 Recons. 0.00274 +2024-03-17 01:21:04,529 INFO Train. Iter 98400 : Commit. 0.26860 PPL. 421.87 Recons. 0.00318 +2024-03-17 01:21:26,607 INFO Train. Iter 98600 : Commit. 0.26332 PPL. 423.58 Recons. 0.00256 +2024-03-17 01:21:48,257 INFO Train. Iter 98800 : Commit. 0.25191 PPL. 424.88 Recons. 0.00223 +2024-03-17 01:22:09,958 INFO Train. Iter 99000 : Commit. 0.25897 PPL. 423.66 Recons. 0.00266 +2024-03-17 01:22:31,262 INFO Train. Iter 99200 : Commit. 0.25409 PPL. 423.58 Recons. 0.00257 +2024-03-17 01:22:52,587 INFO Train. Iter 99400 : Commit. 0.26354 PPL. 419.52 Recons. 0.00299 +2024-03-17 01:23:12,657 INFO Train. Iter 99600 : Commit. 0.25727 PPL. 423.24 Recons. 0.00257 +2024-03-17 01:23:33,748 INFO Train. Iter 99800 : Commit. 0.26060 PPL. 423.54 Recons. 0.00263 +2024-03-17 01:23:55,489 INFO Train. Iter 100000 : Commit. 0.25919 PPL. 424.08 Recons. 0.00262 +2024-03-17 01:24:16,899 INFO Train. Iter 100200 : Commit. 0.25250 PPL. 424.07 Recons. 0.00231 +2024-03-17 01:24:38,852 INFO Train. Iter 100400 : Commit. 0.24825 PPL. 424.71 Recons. 0.00221 +2024-03-17 01:25:00,355 INFO Train. Iter 100600 : Commit. 0.26326 PPL. 419.68 Recons. 0.00321 +2024-03-17 01:25:21,466 INFO Train. Iter 100800 : Commit. 0.25889 PPL. 423.27 Recons. 0.00259 +2024-03-17 01:25:42,841 INFO Train. Iter 101000 : Commit. 0.26078 PPL. 422.37 Recons. 0.00288 +2024-03-17 01:26:04,364 INFO Train. Iter 101200 : Commit. 0.26781 PPL. 422.28 Recons. 0.00270 +2024-03-17 01:26:24,524 INFO Train. Iter 101400 : Commit. 0.25069 PPL. 425.07 Recons. 0.00211 +2024-03-17 01:26:45,959 INFO Train. Iter 101600 : Commit. 0.25418 PPL. 423.66 Recons. 0.00250 +2024-03-17 01:27:07,688 INFO Train. Iter 101800 : Commit. 0.25679 PPL. 423.40 Recons. 0.00257 +2024-03-17 01:27:29,056 INFO Train. Iter 102000 : Commit. 0.26687 PPL. 423.14 Recons. 0.00294 +2024-03-17 01:27:50,826 INFO Train. Iter 102200 : Commit. 0.25188 PPL. 423.88 Recons. 0.00220 +2024-03-17 01:28:12,937 INFO Train. Iter 102400 : Commit. 0.26526 PPL. 422.14 Recons. 0.00315 +2024-03-17 01:28:34,793 INFO Train. Iter 102600 : Commit. 0.26852 PPL. 422.19 Recons. 0.00298 +2024-03-17 01:28:56,760 INFO Train. Iter 102800 : Commit. 0.26001 PPL. 423.32 Recons. 0.00244 +2024-03-17 01:29:18,579 INFO Train. Iter 103000 : Commit. 0.25391 PPL. 423.16 Recons. 0.00248 +2024-03-17 01:29:40,557 INFO Train. Iter 103200 : Commit. 0.25232 PPL. 423.35 Recons. 0.00253 +2024-03-17 01:30:01,372 INFO Train. Iter 103400 : Commit. 0.26285 PPL. 420.77 Recons. 0.00309 +2024-03-17 01:30:24,083 INFO Train. Iter 103600 : Commit. 0.25737 PPL. 422.38 Recons. 0.00245 +2024-03-17 01:30:46,135 INFO Train. Iter 103800 : Commit. 0.25582 PPL. 423.65 Recons. 0.00231 +2024-03-17 01:31:07,702 INFO Train. Iter 104000 : Commit. 0.25088 PPL. 422.64 Recons. 0.00240 +2024-03-17 01:31:29,720 INFO Train. Iter 104200 : Commit. 0.26465 PPL. 420.02 Recons. 0.00319 +2024-03-17 01:31:51,537 INFO Train. Iter 104400 : Commit. 0.25312 PPL. 422.54 Recons. 0.00237 +2024-03-17 01:32:13,358 INFO Train. Iter 104600 : Commit. 0.25836 PPL. 421.85 Recons. 0.00278 +2024-03-17 01:32:34,939 INFO Train. Iter 104800 : Commit. 0.26586 PPL. 420.33 Recons. 0.00329 +2024-03-17 01:32:56,191 INFO Train. Iter 105000 : Commit. 0.25868 PPL. 424.15 Recons. 0.00222 +2024-03-17 01:33:17,564 INFO Train. Iter 105200 : Commit. 0.25298 PPL. 423.28 Recons. 0.00251 +2024-03-17 01:33:37,986 INFO Train. Iter 105400 : Commit. 0.25624 PPL. 421.85 Recons. 0.00256 +2024-03-17 01:33:59,125 INFO Train. Iter 105600 : Commit. 0.25338 PPL. 422.46 Recons. 0.00275 +2024-03-17 01:34:20,658 INFO Train. Iter 105800 : Commit. 0.26325 PPL. 419.95 Recons. 0.00307 +2024-03-17 01:34:41,493 INFO Train. Iter 106000 : Commit. 0.26417 PPL. 421.01 Recons. 0.00293 +2024-03-17 01:35:02,794 INFO Train. Iter 106200 : Commit. 0.25590 PPL. 422.18 Recons. 0.00249 +2024-03-17 01:35:23,934 INFO Train. Iter 106400 : Commit. 0.25669 PPL. 422.20 Recons. 0.00251 +2024-03-17 01:35:45,274 INFO Train. Iter 106600 : Commit. 0.24893 PPL. 422.10 Recons. 0.00233 +2024-03-17 01:36:06,358 INFO Train. Iter 106800 : Commit. 0.24894 PPL. 423.16 Recons. 0.00233 +2024-03-17 01:36:27,561 INFO Train. Iter 107000 : Commit. 0.25020 PPL. 421.54 Recons. 0.00266 +2024-03-17 01:36:47,917 INFO Train. Iter 107200 : Commit. 0.24967 PPL. 423.27 Recons. 0.00217 +2024-03-17 01:37:09,434 INFO Train. Iter 107400 : Commit. 0.26754 PPL. 419.73 Recons. 0.00310 +2024-03-17 01:37:30,809 INFO Train. Iter 107600 : Commit. 0.25775 PPL. 422.49 Recons. 0.00268 +2024-03-17 01:37:52,376 INFO Train. Iter 107800 : Commit. 0.25529 PPL. 422.14 Recons. 0.00258 +2024-03-17 01:38:14,147 INFO Train. Iter 108000 : Commit. 0.25560 PPL. 420.96 Recons. 0.00265 +2024-03-17 01:38:34,860 INFO Train. Iter 108200 : Commit. 0.25352 PPL. 422.48 Recons. 0.00231 +2024-03-17 01:38:56,196 INFO Train. Iter 108400 : Commit. 0.26287 PPL. 420.38 Recons. 0.00314 +2024-03-17 01:39:17,292 INFO Train. Iter 108600 : Commit. 0.25289 PPL. 421.99 Recons. 0.00233 +2024-03-17 01:39:39,209 INFO Train. Iter 108800 : Commit. 0.24377 PPL. 422.92 Recons. 0.00212 +2024-03-17 01:40:00,402 INFO Train. Iter 109000 : Commit. 0.25817 PPL. 420.38 Recons. 0.00276 +2024-03-17 01:40:20,569 INFO Train. Iter 109200 : Commit. 0.25566 PPL. 421.13 Recons. 0.00260 +2024-03-17 01:40:42,462 INFO Train. Iter 109400 : Commit. 0.25860 PPL. 421.58 Recons. 0.00264 +2024-03-17 01:41:04,357 INFO Train. Iter 109600 : Commit. 0.25344 PPL. 420.21 Recons. 0.00277 +2024-03-17 01:41:26,091 INFO Train. Iter 109800 : Commit. 0.25827 PPL. 420.95 Recons. 0.00268 +2024-03-17 01:41:47,445 INFO Train. Iter 110000 : Commit. 0.25314 PPL. 422.32 Recons. 0.00235 +2024-03-17 01:42:09,256 INFO Train. Iter 110200 : Commit. 0.24626 PPL. 423.19 Recons. 0.00224 +2024-03-17 01:42:30,683 INFO Train. Iter 110400 : Commit. 0.25121 PPL. 421.82 Recons. 0.00249 +2024-03-17 01:42:52,078 INFO Train. Iter 110600 : Commit. 0.26083 PPL. 420.08 Recons. 0.00290 +2024-03-17 01:43:13,374 INFO Train. Iter 110800 : Commit. 0.25781 PPL. 422.16 Recons. 0.00251 +2024-03-17 01:43:34,919 INFO Train. Iter 111000 : Commit. 0.24743 PPL. 421.95 Recons. 0.00238 +2024-03-17 01:43:55,152 INFO Train. Iter 111200 : Commit. 0.25578 PPL. 422.46 Recons. 0.00254 +2024-03-17 01:44:16,313 INFO Train. Iter 111400 : Commit. 0.25135 PPL. 419.98 Recons. 0.00277 +2024-03-17 01:44:37,795 INFO Train. Iter 111600 : Commit. 0.25245 PPL. 422.07 Recons. 0.00246 +2024-03-17 01:44:59,381 INFO Train. Iter 111800 : Commit. 0.24951 PPL. 421.97 Recons. 0.00238 +2024-03-17 01:45:21,035 INFO Train. Iter 112000 : Commit. 0.24887 PPL. 423.05 Recons. 0.00228 +2024-03-17 01:45:42,077 INFO Train. Iter 112200 : Commit. 0.25535 PPL. 420.51 Recons. 0.00279 +2024-03-17 01:46:03,269 INFO Train. Iter 112400 : Commit. 0.24277 PPL. 422.92 Recons. 0.00209 +2024-03-17 01:46:24,998 INFO Train. Iter 112600 : Commit. 0.25572 PPL. 420.59 Recons. 0.00283 +2024-03-17 01:46:46,707 INFO Train. Iter 112800 : Commit. 0.24134 PPL. 423.33 Recons. 0.00195 +2024-03-17 01:47:07,025 INFO Train. Iter 113000 : Commit. 0.24701 PPL. 419.40 Recons. 0.00274 +2024-03-17 01:47:28,984 INFO Train. Iter 113200 : Commit. 0.25407 PPL. 421.04 Recons. 0.00259 +2024-03-17 01:47:51,174 INFO Train. Iter 113400 : Commit. 0.25268 PPL. 421.42 Recons. 0.00248 +2024-03-17 01:48:13,138 INFO Train. Iter 113600 : Commit. 0.24643 PPL. 422.15 Recons. 0.00216 +2024-03-17 01:48:34,936 INFO Train. Iter 113800 : Commit. 0.26093 PPL. 420.25 Recons. 0.00303 +2024-03-17 01:48:57,478 INFO Train. Iter 114000 : Commit. 0.25622 PPL. 420.42 Recons. 0.00275 +2024-03-17 01:49:19,686 INFO Train. Iter 114200 : Commit. 0.25176 PPL. 418.86 Recons. 0.00276 +2024-03-17 01:49:41,276 INFO Train. Iter 114400 : Commit. 0.25481 PPL. 420.55 Recons. 0.00255 +2024-03-17 01:50:02,955 INFO Train. Iter 114600 : Commit. 0.25663 PPL. 419.85 Recons. 0.00286 +2024-03-17 01:50:25,143 INFO Train. Iter 114800 : Commit. 0.25530 PPL. 419.73 Recons. 0.00274 +2024-03-17 01:50:45,708 INFO Train. Iter 115000 : Commit. 0.24450 PPL. 422.35 Recons. 0.00211 +2024-03-17 01:51:07,184 INFO Train. Iter 115200 : Commit. 0.25399 PPL. 419.44 Recons. 0.00285 +2024-03-17 01:51:28,860 INFO Train. Iter 115400 : Commit. 0.25315 PPL. 421.85 Recons. 0.00227 +2024-03-17 01:51:50,782 INFO Train. Iter 115600 : Commit. 0.24063 PPL. 422.30 Recons. 0.00203 +2024-03-17 01:52:12,943 INFO Train. Iter 115800 : Commit. 0.24656 PPL. 421.53 Recons. 0.00239 +2024-03-17 01:52:34,945 INFO Train. Iter 116000 : Commit. 0.25589 PPL. 421.36 Recons. 0.00261 +2024-03-17 01:52:56,851 INFO Train. Iter 116200 : Commit. 0.25343 PPL. 419.15 Recons. 0.00290 +2024-03-17 01:53:18,464 INFO Train. Iter 116400 : Commit. 0.25932 PPL. 419.69 Recons. 0.00285 +2024-03-17 01:53:40,156 INFO Train. Iter 116600 : Commit. 0.24574 PPL. 421.77 Recons. 0.00212 +2024-03-17 01:54:02,347 INFO Train. Iter 116800 : Commit. 0.24442 PPL. 421.20 Recons. 0.00224 +2024-03-17 01:54:23,617 INFO Train. Iter 117000 : Commit. 0.25571 PPL. 418.78 Recons. 0.00290 +2024-03-17 01:54:45,624 INFO Train. Iter 117200 : Commit. 0.24216 PPL. 421.48 Recons. 0.00204 +2024-03-17 01:55:07,051 INFO Train. Iter 117400 : Commit. 0.24900 PPL. 420.60 Recons. 0.00258 +2024-03-17 01:55:28,325 INFO Train. Iter 117600 : Commit. 0.25057 PPL. 419.36 Recons. 0.00265 +2024-03-17 01:55:49,952 INFO Train. Iter 117800 : Commit. 0.24713 PPL. 421.03 Recons. 0.00226 +2024-03-17 01:56:11,711 INFO Train. Iter 118000 : Commit. 0.25268 PPL. 419.67 Recons. 0.00274 +2024-03-17 01:56:33,839 INFO Train. Iter 118200 : Commit. 0.25115 PPL. 421.09 Recons. 0.00255 +2024-03-17 01:56:55,278 INFO Train. Iter 118400 : Commit. 0.25958 PPL. 420.44 Recons. 0.00277 +2024-03-17 01:57:17,497 INFO Train. Iter 118600 : Commit. 0.24256 PPL. 421.81 Recons. 0.00202 +2024-03-17 01:57:38,152 INFO Train. Iter 118800 : Commit. 0.24282 PPL. 420.38 Recons. 0.00243 +2024-03-17 01:57:59,824 INFO Train. Iter 119000 : Commit. 0.25422 PPL. 420.19 Recons. 0.00264 +2024-03-17 01:58:21,514 INFO Train. Iter 119200 : Commit. 0.24571 PPL. 421.30 Recons. 0.00229 +2024-03-17 01:58:43,345 INFO Train. Iter 119400 : Commit. 0.24477 PPL. 422.40 Recons. 0.00218 +2024-03-17 01:59:05,458 INFO Train. Iter 119600 : Commit. 0.24852 PPL. 420.57 Recons. 0.00239 +2024-03-17 01:59:26,798 INFO Train. Iter 119800 : Commit. 0.25735 PPL. 420.34 Recons. 0.00247 +2024-03-17 01:59:48,193 INFO Train. Iter 120000 : Commit. 0.24439 PPL. 420.75 Recons. 0.00223 +2024-03-17 02:00:10,207 INFO Train. Iter 120200 : Commit. 0.24200 PPL. 422.04 Recons. 0.00210 +2024-03-17 02:00:32,416 INFO Train. Iter 120400 : Commit. 0.25782 PPL. 418.47 Recons. 0.00305 +2024-03-17 02:00:54,762 INFO Train. Iter 120600 : Commit. 0.25291 PPL. 419.47 Recons. 0.00260 +2024-03-17 02:01:15,986 INFO Train. Iter 120800 : Commit. 0.24535 PPL. 420.79 Recons. 0.00218 +2024-03-17 02:01:37,516 INFO Train. Iter 121000 : Commit. 0.24883 PPL. 419.89 Recons. 0.00257 +2024-03-17 02:01:59,098 INFO Train. Iter 121200 : Commit. 0.26384 PPL. 417.78 Recons. 0.00316 +2024-03-17 02:02:20,831 INFO Train. Iter 121400 : Commit. 0.24572 PPL. 420.70 Recons. 0.00210 +2024-03-17 02:02:42,218 INFO Train. Iter 121600 : Commit. 0.25058 PPL. 419.47 Recons. 0.00261 +2024-03-17 02:03:03,272 INFO Train. Iter 121800 : Commit. 0.24520 PPL. 420.07 Recons. 0.00234 +2024-03-17 02:03:24,448 INFO Train. Iter 122000 : Commit. 0.25526 PPL. 420.60 Recons. 0.00272 +2024-03-17 02:03:45,876 INFO Train. Iter 122200 : Commit. 0.24589 PPL. 420.37 Recons. 0.00251 +2024-03-17 02:04:06,866 INFO Train. Iter 122400 : Commit. 0.25034 PPL. 420.67 Recons. 0.00244 +2024-03-17 02:04:28,298 INFO Train. Iter 122600 : Commit. 0.24921 PPL. 420.74 Recons. 0.00230 +2024-03-17 02:04:48,626 INFO Train. Iter 122800 : Commit. 0.25005 PPL. 418.75 Recons. 0.00266 +2024-03-17 02:05:09,868 INFO Train. Iter 123000 : Commit. 0.24532 PPL. 419.78 Recons. 0.00232 +2024-03-17 02:05:30,561 INFO Train. Iter 123200 : Commit. 0.25033 PPL. 419.32 Recons. 0.00295 +2024-03-17 02:05:51,992 INFO Train. Iter 123400 : Commit. 0.24733 PPL. 420.48 Recons. 0.00221 +2024-03-17 02:06:13,676 INFO Train. Iter 123600 : Commit. 0.24844 PPL. 418.56 Recons. 0.00276 +2024-03-17 02:06:35,934 INFO Train. Iter 123800 : Commit. 0.23952 PPL. 422.09 Recons. 0.00193 +2024-03-17 02:06:57,682 INFO Train. Iter 124000 : Commit. 0.24994 PPL. 418.55 Recons. 0.00284 +2024-03-17 02:07:19,242 INFO Train. Iter 124200 : Commit. 0.24441 PPL. 420.75 Recons. 0.00232 +2024-03-17 02:07:41,451 INFO Train. Iter 124400 : Commit. 0.23971 PPL. 420.69 Recons. 0.00217 +2024-03-17 02:08:01,713 INFO Train. Iter 124600 : Commit. 0.25962 PPL. 417.29 Recons. 0.00315 +2024-03-17 02:08:23,045 INFO Train. Iter 124800 : Commit. 0.25243 PPL. 420.51 Recons. 0.00239 +2024-03-17 02:08:44,268 INFO Train. Iter 125000 : Commit. 0.24919 PPL. 421.31 Recons. 0.00233 +2024-03-17 02:09:05,845 INFO Train. Iter 125200 : Commit. 0.24137 PPL. 420.86 Recons. 0.00221 +2024-03-17 02:09:26,868 INFO Train. Iter 125400 : Commit. 0.24063 PPL. 420.83 Recons. 0.00222 +2024-03-17 02:09:48,317 INFO Train. Iter 125600 : Commit. 0.24302 PPL. 418.37 Recons. 0.00257 +2024-03-17 02:10:09,952 INFO Train. Iter 125800 : Commit. 0.23886 PPL. 420.09 Recons. 0.00205 +2024-03-17 02:10:32,053 INFO Train. Iter 126000 : Commit. 0.24637 PPL. 420.31 Recons. 0.00235 +2024-03-17 02:10:53,980 INFO Train. Iter 126200 : Commit. 0.25393 PPL. 417.96 Recons. 0.00288 +2024-03-17 02:11:15,771 INFO Train. Iter 126400 : Commit. 0.24951 PPL. 419.87 Recons. 0.00245 +2024-03-17 02:11:36,729 INFO Train. Iter 126600 : Commit. 0.25111 PPL. 418.67 Recons. 0.00267 +2024-03-17 02:11:58,556 INFO Train. Iter 126800 : Commit. 0.24079 PPL. 421.03 Recons. 0.00201 +2024-03-17 02:12:20,652 INFO Train. Iter 127000 : Commit. 0.25076 PPL. 418.18 Recons. 0.00298 +2024-03-17 02:12:42,451 INFO Train. Iter 127200 : Commit. 0.24575 PPL. 420.31 Recons. 0.00240 +2024-03-17 02:13:04,067 INFO Train. Iter 127400 : Commit. 0.25130 PPL. 419.42 Recons. 0.00272 +2024-03-17 02:13:25,787 INFO Train. Iter 127600 : Commit. 0.24556 PPL. 417.74 Recons. 0.00260 +2024-03-17 02:13:47,931 INFO Train. Iter 127800 : Commit. 0.24690 PPL. 419.19 Recons. 0.00239 +2024-03-17 02:14:09,929 INFO Train. Iter 128000 : Commit. 0.23867 PPL. 420.58 Recons. 0.00210 +2024-03-17 02:14:31,518 INFO Train. Iter 128200 : Commit. 0.24703 PPL. 417.48 Recons. 0.00271 +2024-03-17 02:14:53,427 INFO Train. Iter 128400 : Commit. 0.24414 PPL. 419.30 Recons. 0.00281 +2024-03-17 02:15:13,724 INFO Train. Iter 128600 : Commit. 0.24455 PPL. 420.48 Recons. 0.00218 +2024-03-17 02:15:35,757 INFO Train. Iter 128800 : Commit. 0.25374 PPL. 419.78 Recons. 0.00265 +2024-03-17 02:15:57,117 INFO Train. Iter 129000 : Commit. 0.23749 PPL. 419.17 Recons. 0.00213 +2024-03-17 02:16:18,760 INFO Train. Iter 129200 : Commit. 0.25053 PPL. 418.30 Recons. 0.00271 +2024-03-17 02:16:40,067 INFO Train. Iter 129400 : Commit. 0.23928 PPL. 421.60 Recons. 0.00196 +2024-03-17 02:17:01,683 INFO Train. Iter 129600 : Commit. 0.24123 PPL. 418.75 Recons. 0.00241 +2024-03-17 02:17:22,867 INFO Train. Iter 129800 : Commit. 0.24029 PPL. 420.34 Recons. 0.00213 +2024-03-17 02:17:44,022 INFO Train. Iter 130000 : Commit. 0.23524 PPL. 419.79 Recons. 0.00207 +2024-03-17 02:18:05,268 INFO Train. Iter 130200 : Commit. 0.25316 PPL. 415.84 Recons. 0.00305 +2024-03-17 02:18:25,533 INFO Train. Iter 130400 : Commit. 0.24540 PPL. 420.47 Recons. 0.00242 +2024-03-17 02:18:46,992 INFO Train. Iter 130600 : Commit. 0.24320 PPL. 419.55 Recons. 0.00222 +2024-03-17 02:19:08,243 INFO Train. Iter 130800 : Commit. 0.24874 PPL. 416.62 Recons. 0.00303 +2024-03-17 02:19:29,707 INFO Train. Iter 131000 : Commit. 0.25207 PPL. 418.97 Recons. 0.00259 +2024-03-17 02:19:51,315 INFO Train. Iter 131200 : Commit. 0.24501 PPL. 420.18 Recons. 0.00225 +2024-03-17 02:20:12,805 INFO Train. Iter 131400 : Commit. 0.24017 PPL. 420.05 Recons. 0.00221 +2024-03-17 02:20:34,325 INFO Train. Iter 131600 : Commit. 0.24781 PPL. 417.71 Recons. 0.00273 +2024-03-17 02:20:56,493 INFO Train. Iter 131800 : Commit. 0.24380 PPL. 419.42 Recons. 0.00238 +2024-03-17 02:21:18,629 INFO Train. Iter 132000 : Commit. 0.24423 PPL. 419.97 Recons. 0.00230 +2024-03-17 02:21:40,527 INFO Train. Iter 132200 : Commit. 0.23585 PPL. 419.47 Recons. 0.00213 +2024-03-17 02:22:00,592 INFO Train. Iter 132400 : Commit. 0.25805 PPL. 416.88 Recons. 0.00291 +2024-03-17 02:22:22,366 INFO Train. Iter 132600 : Commit. 0.24783 PPL. 419.80 Recons. 0.00229 +2024-03-17 02:22:44,876 INFO Train. Iter 132800 : Commit. 0.24464 PPL. 419.38 Recons. 0.00247 +2024-03-17 02:23:06,912 INFO Train. Iter 133000 : Commit. 0.23806 PPL. 419.98 Recons. 0.00217 +2024-03-17 02:23:28,709 INFO Train. Iter 133200 : Commit. 0.24232 PPL. 419.17 Recons. 0.00257 +2024-03-17 02:23:50,309 INFO Train. Iter 133400 : Commit. 0.24012 PPL. 419.35 Recons. 0.00213 +2024-03-17 02:24:12,245 INFO Train. Iter 133600 : Commit. 0.23770 PPL. 419.06 Recons. 0.00228 +2024-03-17 02:24:34,311 INFO Train. Iter 133800 : Commit. 0.24053 PPL. 419.37 Recons. 0.00215 +2024-03-17 02:24:56,375 INFO Train. Iter 134000 : Commit. 0.24652 PPL. 417.45 Recons. 0.00287 +2024-03-17 02:25:17,865 INFO Train. Iter 134200 : Commit. 0.23707 PPL. 420.02 Recons. 0.00201 +2024-03-17 02:25:38,857 INFO Train. Iter 134400 : Commit. 0.24408 PPL. 418.30 Recons. 0.00256 +2024-03-17 02:26:00,897 INFO Train. Iter 134600 : Commit. 0.24267 PPL. 420.41 Recons. 0.00210 +2024-03-17 02:26:23,202 INFO Train. Iter 134800 : Commit. 0.24539 PPL. 418.29 Recons. 0.00260 +2024-03-17 02:26:45,699 INFO Train. Iter 135000 : Commit. 0.24485 PPL. 418.13 Recons. 0.00243 +2024-03-17 02:27:07,545 INFO Train. Iter 135200 : Commit. 0.23899 PPL. 419.04 Recons. 0.00230 +2024-03-17 02:27:29,427 INFO Train. Iter 135400 : Commit. 0.24511 PPL. 418.82 Recons. 0.00238 +2024-03-17 02:27:51,289 INFO Train. Iter 135600 : Commit. 0.23199 PPL. 420.86 Recons. 0.00186 +2024-03-17 02:28:13,228 INFO Train. Iter 135800 : Commit. 0.24933 PPL. 417.24 Recons. 0.00289 +2024-03-17 02:28:35,175 INFO Train. Iter 136000 : Commit. 0.24234 PPL. 420.29 Recons. 0.00209 +2024-03-17 02:28:56,039 INFO Train. Iter 136200 : Commit. 0.23812 PPL. 420.18 Recons. 0.00206 +2024-03-17 02:29:17,804 INFO Train. Iter 136400 : Commit. 0.25061 PPL. 417.42 Recons. 0.00285 +2024-03-17 02:29:40,172 INFO Train. Iter 136600 : Commit. 0.24160 PPL. 419.38 Recons. 0.00209 +2024-03-17 02:30:02,165 INFO Train. Iter 136800 : Commit. 0.24773 PPL. 417.43 Recons. 0.00277 +2024-03-17 02:30:23,618 INFO Train. Iter 137000 : Commit. 0.23149 PPL. 420.64 Recons. 0.00184 +2024-03-17 02:30:45,792 INFO Train. Iter 137200 : Commit. 0.24090 PPL. 419.33 Recons. 0.00227 +2024-03-17 02:31:08,044 INFO Train. Iter 137400 : Commit. 0.25013 PPL. 417.26 Recons. 0.00283 +2024-03-17 02:31:30,529 INFO Train. Iter 137600 : Commit. 0.24896 PPL. 417.02 Recons. 0.00277 +2024-03-17 02:31:52,911 INFO Train. Iter 137800 : Commit. 0.23924 PPL. 418.99 Recons. 0.00209 +2024-03-17 02:32:14,372 INFO Train. Iter 138000 : Commit. 0.23615 PPL. 420.36 Recons. 0.00204 +2024-03-17 02:32:35,090 INFO Train. Iter 138200 : Commit. 0.24432 PPL. 417.38 Recons. 0.00250 +2024-03-17 02:32:56,928 INFO Train. Iter 138400 : Commit. 0.23462 PPL. 419.89 Recons. 0.00200 +2024-03-17 02:33:19,168 INFO Train. Iter 138600 : Commit. 0.23806 PPL. 418.26 Recons. 0.00242 +2024-03-17 02:33:41,656 INFO Train. Iter 138800 : Commit. 0.24083 PPL. 418.84 Recons. 0.00226 +2024-03-17 02:34:03,691 INFO Train. Iter 139000 : Commit. 0.23998 PPL. 419.83 Recons. 0.00218 +2024-03-17 02:34:25,589 INFO Train. Iter 139200 : Commit. 0.24456 PPL. 417.64 Recons. 0.00262 +2024-03-17 02:34:47,313 INFO Train. Iter 139400 : Commit. 0.23942 PPL. 419.59 Recons. 0.00220 +2024-03-17 02:35:09,518 INFO Train. Iter 139600 : Commit. 0.24258 PPL. 417.80 Recons. 0.00250 +2024-03-17 02:35:31,187 INFO Train. Iter 139800 : Commit. 0.24235 PPL. 419.45 Recons. 0.00209 +2024-03-17 02:35:53,302 INFO Train. Iter 140000 : Commit. 0.24218 PPL. 418.96 Recons. 0.00235 +2024-03-17 02:36:14,247 INFO Train. Iter 140200 : Commit. 0.23415 PPL. 418.90 Recons. 0.00226 +2024-03-17 02:36:35,863 INFO Train. Iter 140400 : Commit. 0.23429 PPL. 420.59 Recons. 0.00195 +2024-03-17 02:36:57,771 INFO Train. Iter 140600 : Commit. 0.24809 PPL. 416.39 Recons. 0.00289 +2024-03-17 02:37:20,003 INFO Train. Iter 140800 : Commit. 0.25373 PPL. 418.57 Recons. 0.00248 +2024-03-17 02:37:41,950 INFO Train. Iter 141000 : Commit. 0.24263 PPL. 418.78 Recons. 0.00219 +2024-03-17 02:38:03,500 INFO Train. Iter 141200 : Commit. 0.23922 PPL. 418.53 Recons. 0.00221 +2024-03-17 02:38:25,423 INFO Train. Iter 141400 : Commit. 0.23724 PPL. 417.49 Recons. 0.00235 +2024-03-17 02:38:46,903 INFO Train. Iter 141600 : Commit. 0.23744 PPL. 418.96 Recons. 0.00218 +2024-03-17 02:39:08,744 INFO Train. Iter 141800 : Commit. 0.24186 PPL. 418.19 Recons. 0.00253 +2024-03-17 02:39:29,256 INFO Train. Iter 142000 : Commit. 0.23738 PPL. 418.70 Recons. 0.00211 +2024-03-17 02:39:52,129 INFO Train. Iter 142200 : Commit. 0.23025 PPL. 420.57 Recons. 0.00191 +2024-03-17 02:40:14,631 INFO Train. Iter 142400 : Commit. 0.25005 PPL. 417.03 Recons. 0.00268 +2024-03-17 02:40:36,533 INFO Train. Iter 142600 : Commit. 0.24206 PPL. 418.08 Recons. 0.00238 +2024-03-17 02:40:58,321 INFO Train. Iter 142800 : Commit. 0.23758 PPL. 418.42 Recons. 0.00206 +2024-03-17 02:41:20,245 INFO Train. Iter 143000 : Commit. 0.24173 PPL. 416.20 Recons. 0.00235 +2024-03-17 02:41:42,185 INFO Train. Iter 143200 : Commit. 0.24633 PPL. 416.72 Recons. 0.00256 +2024-03-17 02:42:04,622 INFO Train. Iter 143400 : Commit. 0.24028 PPL. 419.04 Recons. 0.00212 +2024-03-17 02:42:26,482 INFO Train. Iter 143600 : Commit. 0.23959 PPL. 418.43 Recons. 0.00230 +2024-03-17 02:42:49,003 INFO Train. Iter 143800 : Commit. 0.24177 PPL. 417.61 Recons. 0.00249 +2024-03-17 02:43:10,050 INFO Train. Iter 144000 : Commit. 0.24028 PPL. 417.43 Recons. 0.00241 +2024-03-17 02:43:31,577 INFO Train. Iter 144200 : Commit. 0.23657 PPL. 419.52 Recons. 0.00205 +2024-03-17 02:43:53,239 INFO Train. Iter 144400 : Commit. 0.25123 PPL. 416.67 Recons. 0.00311 +2024-03-17 02:44:14,751 INFO Train. Iter 144600 : Commit. 0.23994 PPL. 419.92 Recons. 0.00194 +2024-03-17 02:44:35,894 INFO Train. Iter 144800 : Commit. 0.24472 PPL. 417.00 Recons. 0.00251 +2024-03-17 02:44:57,731 INFO Train. Iter 145000 : Commit. 0.23630 PPL. 419.58 Recons. 0.00202 +2024-03-17 02:45:19,587 INFO Train. Iter 145200 : Commit. 0.23575 PPL. 418.88 Recons. 0.00213 +2024-03-17 02:45:41,834 INFO Train. Iter 145400 : Commit. 0.23183 PPL. 419.27 Recons. 0.00208 +2024-03-17 02:46:03,960 INFO Train. Iter 145600 : Commit. 0.24441 PPL. 417.17 Recons. 0.00255 +2024-03-17 02:46:26,487 INFO Train. Iter 145800 : Commit. 0.24716 PPL. 416.53 Recons. 0.00275 +2024-03-17 02:46:47,830 INFO Train. Iter 146000 : Commit. 0.23981 PPL. 418.71 Recons. 0.00215 +2024-03-17 02:47:09,515 INFO Train. Iter 146200 : Commit. 0.24142 PPL. 417.86 Recons. 0.00231 +2024-03-17 02:47:31,662 INFO Train. Iter 146400 : Commit. 0.23675 PPL. 419.09 Recons. 0.00212 +2024-03-17 02:47:53,183 INFO Train. Iter 146600 : Commit. 0.22520 PPL. 419.50 Recons. 0.00179 +2024-03-17 02:48:15,664 INFO Train. Iter 146800 : Commit. 0.23525 PPL. 417.38 Recons. 0.00235 +2024-03-17 02:48:37,581 INFO Train. Iter 147000 : Commit. 0.24363 PPL. 416.99 Recons. 0.00252 +2024-03-17 02:48:59,551 INFO Train. Iter 147200 : Commit. 0.23932 PPL. 417.18 Recons. 0.00232 +2024-03-17 02:49:21,367 INFO Train. Iter 147400 : Commit. 0.24145 PPL. 417.06 Recons. 0.00246 +2024-03-17 02:49:43,759 INFO Train. Iter 147600 : Commit. 0.23390 PPL. 418.36 Recons. 0.00206 +2024-03-17 02:50:04,879 INFO Train. Iter 147800 : Commit. 0.23991 PPL. 417.20 Recons. 0.00231 +2024-03-17 02:50:27,387 INFO Train. Iter 148000 : Commit. 0.23100 PPL. 418.33 Recons. 0.00205 +2024-03-17 02:50:50,097 INFO Train. Iter 148200 : Commit. 0.23773 PPL. 417.70 Recons. 0.00243 +2024-03-17 02:51:12,410 INFO Train. Iter 148400 : Commit. 0.23357 PPL. 418.56 Recons. 0.00208 +2024-03-17 02:51:34,096 INFO Train. Iter 148600 : Commit. 0.24143 PPL. 416.91 Recons. 0.00254 +2024-03-17 02:51:55,672 INFO Train. Iter 148800 : Commit. 0.23525 PPL. 418.23 Recons. 0.00211 +2024-03-17 02:52:17,567 INFO Train. Iter 149000 : Commit. 0.23709 PPL. 417.44 Recons. 0.00235 +2024-03-17 02:52:39,075 INFO Train. Iter 149200 : Commit. 0.24014 PPL. 416.70 Recons. 0.00234 +2024-03-17 02:53:01,186 INFO Train. Iter 149400 : Commit. 0.24133 PPL. 417.07 Recons. 0.00233 +2024-03-17 02:53:23,054 INFO Train. Iter 149600 : Commit. 0.23628 PPL. 418.78 Recons. 0.00216 +2024-03-17 02:53:43,888 INFO Train. Iter 149800 : Commit. 0.23383 PPL. 417.74 Recons. 0.00213 +2024-03-17 02:54:05,851 INFO Train. Iter 150000 : Commit. 0.23239 PPL. 418.77 Recons. 0.00196 +2024-03-17 02:54:28,020 INFO Train. Iter 150200 : Commit. 0.23601 PPL. 417.31 Recons. 0.00240 +2024-03-17 02:54:49,841 INFO Train. Iter 150400 : Commit. 0.23718 PPL. 416.20 Recons. 0.00249 +2024-03-17 02:55:11,363 INFO Train. Iter 150600 : Commit. 0.23439 PPL. 416.57 Recons. 0.00193 +2024-03-17 02:55:32,788 INFO Train. Iter 150800 : Commit. 0.23572 PPL. 417.38 Recons. 0.00215 +2024-03-17 02:55:54,472 INFO Train. Iter 151000 : Commit. 0.24060 PPL. 417.42 Recons. 0.00227 +2024-03-17 02:56:16,124 INFO Train. Iter 151200 : Commit. 0.23894 PPL. 417.62 Recons. 0.00225 +2024-03-17 02:56:37,540 INFO Train. Iter 151400 : Commit. 0.24820 PPL. 415.34 Recons. 0.00276 +2024-03-17 02:56:59,331 INFO Train. Iter 151600 : Commit. 0.23904 PPL. 418.80 Recons. 0.00220 +2024-03-17 02:57:20,250 INFO Train. Iter 151800 : Commit. 0.23937 PPL. 417.60 Recons. 0.00228 +2024-03-17 02:57:42,517 INFO Train. Iter 152000 : Commit. 0.24429 PPL. 416.33 Recons. 0.00281 +2024-03-17 02:58:04,608 INFO Train. Iter 152200 : Commit. 0.22849 PPL. 419.56 Recons. 0.00177 +2024-03-17 02:58:26,229 INFO Train. Iter 152400 : Commit. 0.23356 PPL. 418.95 Recons. 0.00215 +2024-03-17 02:58:48,353 INFO Train. Iter 152600 : Commit. 0.23150 PPL. 418.71 Recons. 0.00203 +2024-03-17 02:59:10,005 INFO Train. Iter 152800 : Commit. 0.23435 PPL. 417.08 Recons. 0.00230 +2024-03-17 02:59:31,650 INFO Train. Iter 153000 : Commit. 0.23439 PPL. 417.59 Recons. 0.00198 +2024-03-17 02:59:53,430 INFO Train. Iter 153200 : Commit. 0.23280 PPL. 417.99 Recons. 0.00202 +2024-03-17 03:00:15,095 INFO Train. Iter 153400 : Commit. 0.23681 PPL. 416.96 Recons. 0.00227 +2024-03-17 03:00:35,490 INFO Train. Iter 153600 : Commit. 0.24268 PPL. 415.25 Recons. 0.00275 +2024-03-17 03:00:57,185 INFO Train. Iter 153800 : Commit. 0.24344 PPL. 419.13 Recons. 0.00217 +2024-03-17 03:01:18,675 INFO Train. Iter 154000 : Commit. 0.24903 PPL. 416.49 Recons. 0.00274 +2024-03-17 03:01:41,406 INFO Train. Iter 154200 : Commit. 0.22290 PPL. 418.89 Recons. 0.00167 +2024-03-17 03:02:03,428 INFO Train. Iter 154400 : Commit. 0.23830 PPL. 416.61 Recons. 0.00238 +2024-03-17 03:02:25,619 INFO Train. Iter 154600 : Commit. 0.24023 PPL. 415.80 Recons. 0.00257 +2024-03-17 03:02:47,632 INFO Train. Iter 154800 : Commit. 0.24119 PPL. 418.72 Recons. 0.00221 +2024-03-17 03:03:09,477 INFO Train. Iter 155000 : Commit. 0.24190 PPL. 416.88 Recons. 0.00249 +2024-03-17 03:03:31,331 INFO Train. Iter 155200 : Commit. 0.23392 PPL. 418.20 Recons. 0.00206 +2024-03-17 03:03:52,955 INFO Train. Iter 155400 : Commit. 0.22850 PPL. 418.31 Recons. 0.00192 +2024-03-17 03:04:13,316 INFO Train. Iter 155600 : Commit. 0.22814 PPL. 418.40 Recons. 0.00202 +2024-03-17 03:04:35,873 INFO Train. Iter 155800 : Commit. 0.23464 PPL. 418.45 Recons. 0.00204 +2024-03-17 03:04:58,248 INFO Train. Iter 156000 : Commit. 0.23370 PPL. 417.73 Recons. 0.00220 +2024-03-17 03:05:19,785 INFO Train. Iter 156200 : Commit. 0.23668 PPL. 416.47 Recons. 0.00235 +2024-03-17 03:05:41,958 INFO Train. Iter 156400 : Commit. 0.23615 PPL. 417.41 Recons. 0.00232 +2024-03-17 03:06:03,445 INFO Train. Iter 156600 : Commit. 0.23366 PPL. 418.26 Recons. 0.00197 +2024-03-17 03:06:25,077 INFO Train. Iter 156800 : Commit. 0.24501 PPL. 415.34 Recons. 0.00272 +2024-03-17 03:06:46,398 INFO Train. Iter 157000 : Commit. 0.22683 PPL. 418.69 Recons. 0.00170 +2024-03-17 03:07:07,901 INFO Train. Iter 157200 : Commit. 0.24446 PPL. 417.43 Recons. 0.00254 +2024-03-17 03:07:30,233 INFO Train. Iter 157400 : Commit. 0.23079 PPL. 418.52 Recons. 0.00187 +2024-03-17 03:07:50,959 INFO Train. Iter 157600 : Commit. 0.23690 PPL. 415.16 Recons. 0.00263 +2024-03-17 03:08:12,873 INFO Train. Iter 157800 : Commit. 0.23709 PPL. 416.94 Recons. 0.00224 +2024-03-17 03:08:34,930 INFO Train. Iter 158000 : Commit. 0.23860 PPL. 415.64 Recons. 0.00252 +2024-03-17 03:08:56,504 INFO Train. Iter 158200 : Commit. 0.23529 PPL. 418.19 Recons. 0.00201 +2024-03-17 03:09:17,976 INFO Train. Iter 158400 : Commit. 0.24379 PPL. 416.39 Recons. 0.00266 +2024-03-17 03:09:39,489 INFO Train. Iter 158600 : Commit. 0.23821 PPL. 416.79 Recons. 0.00223 +2024-03-17 03:10:01,333 INFO Train. Iter 158800 : Commit. 0.23104 PPL. 418.22 Recons. 0.00191 +2024-03-17 03:10:22,949 INFO Train. Iter 159000 : Commit. 0.23296 PPL. 417.80 Recons. 0.00209 +2024-03-17 03:10:44,613 INFO Train. Iter 159200 : Commit. 0.23687 PPL. 416.93 Recons. 0.00240 +2024-03-17 03:11:05,132 INFO Train. Iter 159400 : Commit. 0.23271 PPL. 417.19 Recons. 0.00214 +2024-03-17 03:11:26,578 INFO Train. Iter 159600 : Commit. 0.23255 PPL. 417.18 Recons. 0.00209 +2024-03-17 03:11:47,988 INFO Train. Iter 159800 : Commit. 0.23241 PPL. 417.54 Recons. 0.00205 +2024-03-17 03:12:09,397 INFO Train. Iter 160000 : Commit. 0.23249 PPL. 416.36 Recons. 0.00239 +2024-03-17 03:12:31,065 INFO Train. Iter 160200 : Commit. 0.22667 PPL. 418.67 Recons. 0.00180 +2024-03-17 03:12:52,666 INFO Train. Iter 160400 : Commit. 0.23581 PPL. 416.14 Recons. 0.00236 +2024-03-17 03:13:14,320 INFO Train. Iter 160600 : Commit. 0.23668 PPL. 417.17 Recons. 0.00219 +2024-03-17 03:13:35,631 INFO Train. Iter 160800 : Commit. 0.23199 PPL. 418.32 Recons. 0.00194 +2024-03-17 03:13:57,063 INFO Train. Iter 161000 : Commit. 0.23665 PPL. 416.90 Recons. 0.00239 +2024-03-17 03:14:18,225 INFO Train. Iter 161200 : Commit. 0.23264 PPL. 417.78 Recons. 0.00193 +2024-03-17 03:14:38,550 INFO Train. Iter 161400 : Commit. 0.23212 PPL. 418.13 Recons. 0.00196 +2024-03-17 03:15:00,613 INFO Train. Iter 161600 : Commit. 0.23304 PPL. 416.55 Recons. 0.00227 +2024-03-17 03:15:22,590 INFO Train. Iter 161800 : Commit. 0.24019 PPL. 417.06 Recons. 0.00234 +2024-03-17 03:15:44,408 INFO Train. Iter 162000 : Commit. 0.24111 PPL. 416.21 Recons. 0.00260 +2024-03-17 03:16:06,720 INFO Train. Iter 162200 : Commit. 0.23070 PPL. 417.94 Recons. 0.00184 +2024-03-17 03:16:29,121 INFO Train. Iter 162400 : Commit. 0.22718 PPL. 417.95 Recons. 0.00190 +2024-03-17 03:16:51,219 INFO Train. Iter 162600 : Commit. 0.22998 PPL. 415.35 Recons. 0.00220 +2024-03-17 03:17:13,311 INFO Train. Iter 162800 : Commit. 0.22613 PPL. 417.21 Recons. 0.00185 +2024-03-17 03:17:35,360 INFO Train. Iter 163000 : Commit. 0.24423 PPL. 416.23 Recons. 0.00259 +2024-03-17 03:17:57,688 INFO Train. Iter 163200 : Commit. 0.24187 PPL. 417.23 Recons. 0.00233 +2024-03-17 03:18:19,010 INFO Train. Iter 163400 : Commit. 0.23677 PPL. 417.30 Recons. 0.00215 +2024-03-17 03:18:41,324 INFO Train. Iter 163600 : Commit. 0.23036 PPL. 416.59 Recons. 0.00223 +2024-03-17 03:19:03,188 INFO Train. Iter 163800 : Commit. 0.23071 PPL. 418.44 Recons. 0.00199 +2024-03-17 03:19:24,880 INFO Train. Iter 164000 : Commit. 0.23871 PPL. 416.80 Recons. 0.00237 +2024-03-17 03:19:47,234 INFO Train. Iter 164200 : Commit. 0.22503 PPL. 417.69 Recons. 0.00188 +2024-03-17 03:20:09,277 INFO Train. Iter 164400 : Commit. 0.23179 PPL. 417.16 Recons. 0.00218 +2024-03-17 03:20:31,493 INFO Train. Iter 164600 : Commit. 0.22969 PPL. 416.24 Recons. 0.00218 +2024-03-17 03:20:53,424 INFO Train. Iter 164800 : Commit. 0.22932 PPL. 417.25 Recons. 0.00202 +2024-03-17 03:21:15,794 INFO Train. Iter 165000 : Commit. 0.23215 PPL. 417.41 Recons. 0.00215 +2024-03-17 03:21:36,593 INFO Train. Iter 165200 : Commit. 0.22519 PPL. 417.81 Recons. 0.00186 +2024-03-17 03:21:58,775 INFO Train. Iter 165400 : Commit. 0.23344 PPL. 415.00 Recons. 0.00232 +2024-03-17 03:22:20,578 INFO Train. Iter 165600 : Commit. 0.22591 PPL. 416.71 Recons. 0.00187 +2024-03-17 03:22:43,078 INFO Train. Iter 165800 : Commit. 0.22388 PPL. 418.17 Recons. 0.00187 +2024-03-17 03:23:05,626 INFO Train. Iter 166000 : Commit. 0.23165 PPL. 416.51 Recons. 0.00224 +2024-03-17 03:23:27,211 INFO Train. Iter 166200 : Commit. 0.23593 PPL. 416.43 Recons. 0.00224 +2024-03-17 03:23:49,241 INFO Train. Iter 166400 : Commit. 0.23999 PPL. 415.56 Recons. 0.00245 +2024-03-17 03:24:11,184 INFO Train. Iter 166600 : Commit. 0.23201 PPL. 417.29 Recons. 0.00211 +2024-03-17 03:24:32,909 INFO Train. Iter 166800 : Commit. 0.24004 PPL. 416.24 Recons. 0.00237 +2024-03-17 03:24:54,566 INFO Train. Iter 167000 : Commit. 0.23059 PPL. 416.97 Recons. 0.00202 +2024-03-17 03:25:15,380 INFO Train. Iter 167200 : Commit. 0.22783 PPL. 416.84 Recons. 0.00207 +2024-03-17 03:25:37,060 INFO Train. Iter 167400 : Commit. 0.23889 PPL. 416.24 Recons. 0.00243 +2024-03-17 03:25:59,460 INFO Train. Iter 167600 : Commit. 0.22637 PPL. 416.58 Recons. 0.00194 +2024-03-17 03:26:21,829 INFO Train. Iter 167800 : Commit. 0.24356 PPL. 413.78 Recons. 0.00265 +2024-03-17 03:26:43,719 INFO Train. Iter 168000 : Commit. 0.22965 PPL. 417.61 Recons. 0.00185 +2024-03-17 03:27:05,846 INFO Train. Iter 168200 : Commit. 0.21666 PPL. 417.95 Recons. 0.00170 +2024-03-17 03:27:27,531 INFO Train. Iter 168400 : Commit. 0.24210 PPL. 412.55 Recons. 0.00303 +2024-03-17 03:27:49,569 INFO Train. Iter 168600 : Commit. 0.24160 PPL. 416.62 Recons. 0.00245 +2024-03-17 03:28:11,420 INFO Train. Iter 168800 : Commit. 0.22091 PPL. 418.12 Recons. 0.00165 +2024-03-17 03:28:33,939 INFO Train. Iter 169000 : Commit. 0.23715 PPL. 416.31 Recons. 0.00236 +2024-03-17 03:28:54,654 INFO Train. Iter 169200 : Commit. 0.22990 PPL. 416.22 Recons. 0.00208 +2024-03-17 03:29:16,607 INFO Train. Iter 169400 : Commit. 0.22561 PPL. 416.67 Recons. 0.00201 +2024-03-17 03:29:38,447 INFO Train. Iter 169600 : Commit. 0.23208 PPL. 416.33 Recons. 0.00204 +2024-03-17 03:30:00,173 INFO Train. Iter 169800 : Commit. 0.23426 PPL. 416.18 Recons. 0.00227 +2024-03-17 03:30:22,047 INFO Train. Iter 170000 : Commit. 0.24051 PPL. 414.73 Recons. 0.00247 +2024-03-17 03:30:43,897 INFO Train. Iter 170200 : Commit. 0.23307 PPL. 417.11 Recons. 0.00205 +2024-03-17 03:31:05,346 INFO Train. Iter 170400 : Commit. 0.23158 PPL. 417.81 Recons. 0.00190 +2024-03-17 03:31:27,279 INFO Train. Iter 170600 : Commit. 0.24029 PPL. 413.14 Recons. 0.00255 +2024-03-17 03:31:49,450 INFO Train. Iter 170800 : Commit. 0.23762 PPL. 416.51 Recons. 0.00222 +2024-03-17 03:32:10,924 INFO Train. Iter 171000 : Commit. 0.22850 PPL. 417.81 Recons. 0.00186 +2024-03-17 03:32:32,594 INFO Train. Iter 171200 : Commit. 0.23509 PPL. 416.21 Recons. 0.00221 +2024-03-17 03:32:54,166 INFO Train. Iter 171400 : Commit. 0.23324 PPL. 416.17 Recons. 0.00204 +2024-03-17 03:33:16,228 INFO Train. Iter 171600 : Commit. 0.23173 PPL. 416.90 Recons. 0.00200 +2024-03-17 03:33:38,097 INFO Train. Iter 171800 : Commit. 0.24410 PPL. 414.96 Recons. 0.00243 +2024-03-17 03:33:59,930 INFO Train. Iter 172000 : Commit. 0.23969 PPL. 414.73 Recons. 0.00256 +2024-03-17 03:34:21,119 INFO Train. Iter 172200 : Commit. 0.23464 PPL. 416.23 Recons. 0.00226 +2024-03-17 03:34:42,925 INFO Train. Iter 172400 : Commit. 0.22840 PPL. 417.58 Recons. 0.00187 +2024-03-17 03:35:04,506 INFO Train. Iter 172600 : Commit. 0.23592 PPL. 415.00 Recons. 0.00243 +2024-03-17 03:35:26,360 INFO Train. Iter 172800 : Commit. 0.23325 PPL. 415.22 Recons. 0.00231 +2024-03-17 03:35:46,918 INFO Train. Iter 173000 : Commit. 0.22381 PPL. 416.96 Recons. 0.00190 +2024-03-17 03:36:08,565 INFO Train. Iter 173200 : Commit. 0.23121 PPL. 416.21 Recons. 0.00208 +2024-03-17 03:36:29,761 INFO Train. Iter 173400 : Commit. 0.23599 PPL. 416.00 Recons. 0.00238 +2024-03-17 03:36:51,617 INFO Train. Iter 173600 : Commit. 0.22151 PPL. 417.40 Recons. 0.00174 +2024-03-17 03:37:12,848 INFO Train. Iter 173800 : Commit. 0.22479 PPL. 416.92 Recons. 0.00194 +2024-03-17 03:37:34,731 INFO Train. Iter 174000 : Commit. 0.22665 PPL. 416.41 Recons. 0.00196 +2024-03-17 03:37:56,420 INFO Train. Iter 174200 : Commit. 0.22545 PPL. 415.78 Recons. 0.00197 +2024-03-17 03:38:18,038 INFO Train. Iter 174400 : Commit. 0.23277 PPL. 416.06 Recons. 0.00214 +2024-03-17 03:38:40,008 INFO Train. Iter 174600 : Commit. 0.22566 PPL. 416.46 Recons. 0.00196 +2024-03-17 03:39:01,603 INFO Train. Iter 174800 : Commit. 0.23775 PPL. 413.68 Recons. 0.00251 +2024-03-17 03:39:21,815 INFO Train. Iter 175000 : Commit. 0.23432 PPL. 415.18 Recons. 0.00230 +2024-03-17 03:39:43,268 INFO Train. Iter 175200 : Commit. 0.22653 PPL. 416.17 Recons. 0.00196 +2024-03-17 03:40:05,133 INFO Train. Iter 175400 : Commit. 0.22678 PPL. 416.51 Recons. 0.00189 +2024-03-17 03:40:26,399 INFO Train. Iter 175600 : Commit. 0.23799 PPL. 415.22 Recons. 0.00241 +2024-03-17 03:40:48,120 INFO Train. Iter 175800 : Commit. 0.23243 PPL. 416.07 Recons. 0.00218 +2024-03-17 03:41:09,757 INFO Train. Iter 176000 : Commit. 0.22936 PPL. 414.98 Recons. 0.00212 +2024-03-17 03:41:31,621 INFO Train. Iter 176200 : Commit. 0.23135 PPL. 415.84 Recons. 0.00212 +2024-03-17 03:41:53,466 INFO Train. Iter 176400 : Commit. 0.23846 PPL. 414.56 Recons. 0.00246 +2024-03-17 03:42:15,126 INFO Train. Iter 176600 : Commit. 0.22208 PPL. 416.70 Recons. 0.00179 +2024-03-17 03:42:35,421 INFO Train. Iter 176800 : Commit. 0.23064 PPL. 416.24 Recons. 0.00213 +2024-03-17 03:42:56,972 INFO Train. Iter 177000 : Commit. 0.23369 PPL. 415.02 Recons. 0.00234 +2024-03-17 03:43:18,687 INFO Train. Iter 177200 : Commit. 0.22955 PPL. 416.26 Recons. 0.00203 +2024-03-17 03:43:40,568 INFO Train. Iter 177400 : Commit. 0.23393 PPL. 413.79 Recons. 0.00232 +2024-03-17 03:44:02,224 INFO Train. Iter 177600 : Commit. 0.22881 PPL. 416.46 Recons. 0.00191 +2024-03-17 03:44:23,561 INFO Train. Iter 177800 : Commit. 0.23404 PPL. 416.28 Recons. 0.00215 +2024-03-17 03:44:45,608 INFO Train. Iter 178000 : Commit. 0.22237 PPL. 417.94 Recons. 0.00174 +2024-03-17 03:45:07,118 INFO Train. Iter 178200 : Commit. 0.23621 PPL. 414.33 Recons. 0.00257 +2024-03-17 03:45:28,704 INFO Train. Iter 178400 : Commit. 0.22125 PPL. 417.20 Recons. 0.00159 +2024-03-17 03:45:51,174 INFO Train. Iter 178600 : Commit. 0.23358 PPL. 414.21 Recons. 0.00229 +2024-03-17 03:46:11,770 INFO Train. Iter 178800 : Commit. 0.22374 PPL. 415.85 Recons. 0.00192 +2024-03-17 03:46:33,404 INFO Train. Iter 179000 : Commit. 0.23801 PPL. 415.25 Recons. 0.00233 +2024-03-17 03:46:54,934 INFO Train. Iter 179200 : Commit. 0.22937 PPL. 415.80 Recons. 0.00210 +2024-03-17 03:47:16,501 INFO Train. Iter 179400 : Commit. 0.22903 PPL. 415.76 Recons. 0.00202 +2024-03-17 03:47:37,828 INFO Train. Iter 179600 : Commit. 0.23469 PPL. 415.47 Recons. 0.00217 +2024-03-17 03:47:59,331 INFO Train. Iter 179800 : Commit. 0.22761 PPL. 415.87 Recons. 0.00205 +2024-03-17 03:48:20,916 INFO Train. Iter 180000 : Commit. 0.23233 PPL. 414.63 Recons. 0.00210 +2024-03-17 03:48:42,449 INFO Train. Iter 180200 : Commit. 0.22874 PPL. 416.90 Recons. 0.00193 +2024-03-17 03:49:04,136 INFO Train. Iter 180400 : Commit. 0.22007 PPL. 416.69 Recons. 0.00172 +2024-03-17 03:49:26,390 INFO Train. Iter 180600 : Commit. 0.23507 PPL. 415.31 Recons. 0.00237 +2024-03-17 03:49:46,921 INFO Train. Iter 180800 : Commit. 0.22448 PPL. 416.25 Recons. 0.00185 +2024-03-17 03:50:08,553 INFO Train. Iter 181000 : Commit. 0.23046 PPL. 415.06 Recons. 0.00218 +2024-03-17 03:50:30,228 INFO Train. Iter 181200 : Commit. 0.23115 PPL. 414.98 Recons. 0.00220 +2024-03-17 03:50:51,505 INFO Train. Iter 181400 : Commit. 0.23424 PPL. 415.72 Recons. 0.00216 +2024-03-17 03:51:13,115 INFO Train. Iter 181600 : Commit. 0.22340 PPL. 415.35 Recons. 0.00184 +2024-03-17 03:51:35,056 INFO Train. Iter 181800 : Commit. 0.23538 PPL. 414.77 Recons. 0.00247 +2024-03-17 03:51:56,883 INFO Train. Iter 182000 : Commit. 0.23546 PPL. 412.86 Recons. 0.00236 +2024-03-17 03:52:18,648 INFO Train. Iter 182200 : Commit. 0.23084 PPL. 415.88 Recons. 0.00196 +2024-03-17 03:52:40,025 INFO Train. Iter 182400 : Commit. 0.22657 PPL. 416.09 Recons. 0.00188 +2024-03-17 03:53:00,744 INFO Train. Iter 182600 : Commit. 0.22644 PPL. 415.80 Recons. 0.00198 +2024-03-17 03:53:22,244 INFO Train. Iter 182800 : Commit. 0.23021 PPL. 415.39 Recons. 0.00200 +2024-03-17 03:53:43,607 INFO Train. Iter 183000 : Commit. 0.22961 PPL. 414.51 Recons. 0.00228 +2024-03-17 03:54:05,213 INFO Train. Iter 183200 : Commit. 0.22494 PPL. 415.58 Recons. 0.00186 +2024-03-17 03:54:27,227 INFO Train. Iter 183400 : Commit. 0.23124 PPL. 413.58 Recons. 0.00224 +2024-03-17 03:54:48,993 INFO Train. Iter 183600 : Commit. 0.22671 PPL. 416.26 Recons. 0.00185 +2024-03-17 03:55:10,473 INFO Train. Iter 183800 : Commit. 0.23412 PPL. 413.43 Recons. 0.00257 +2024-03-17 03:55:32,065 INFO Train. Iter 184000 : Commit. 0.22591 PPL. 415.40 Recons. 0.00203 +2024-03-17 03:55:53,652 INFO Train. Iter 184200 : Commit. 0.22643 PPL. 416.94 Recons. 0.00178 +2024-03-17 03:56:15,668 INFO Train. Iter 184400 : Commit. 0.23703 PPL. 413.34 Recons. 0.00262 +2024-03-17 03:56:36,440 INFO Train. Iter 184600 : Commit. 0.23422 PPL. 416.40 Recons. 0.00206 +2024-03-17 03:56:57,997 INFO Train. Iter 184800 : Commit. 0.22403 PPL. 416.59 Recons. 0.00183 +2024-03-17 03:57:19,642 INFO Train. Iter 185000 : Commit. 0.22221 PPL. 416.71 Recons. 0.00181 +2024-03-17 03:57:41,458 INFO Train. Iter 185200 : Commit. 0.23299 PPL. 414.18 Recons. 0.00241 +2024-03-17 03:58:03,331 INFO Train. Iter 185400 : Commit. 0.23142 PPL. 414.30 Recons. 0.00224 +2024-03-17 03:58:25,073 INFO Train. Iter 185600 : Commit. 0.22267 PPL. 416.13 Recons. 0.00182 +2024-03-17 03:58:47,347 INFO Train. Iter 185800 : Commit. 0.22262 PPL. 415.61 Recons. 0.00190 +2024-03-17 03:59:09,967 INFO Train. Iter 186000 : Commit. 0.22623 PPL. 414.92 Recons. 0.00200 +2024-03-17 03:59:31,356 INFO Train. Iter 186200 : Commit. 0.22830 PPL. 415.47 Recons. 0.00206 +2024-03-17 03:59:53,499 INFO Train. Iter 186400 : Commit. 0.22927 PPL. 414.65 Recons. 0.00206 +2024-03-17 04:00:14,040 INFO Train. Iter 186600 : Commit. 0.23198 PPL. 412.78 Recons. 0.00240 +2024-03-17 04:00:36,176 INFO Train. Iter 186800 : Commit. 0.23334 PPL. 414.05 Recons. 0.00225 +2024-03-17 04:00:58,439 INFO Train. Iter 187000 : Commit. 0.22846 PPL. 414.61 Recons. 0.00203 +2024-03-17 04:01:20,363 INFO Train. Iter 187200 : Commit. 0.22715 PPL. 415.78 Recons. 0.00200 +2024-03-17 04:01:41,952 INFO Train. Iter 187400 : Commit. 0.22266 PPL. 415.74 Recons. 0.00190 +2024-03-17 04:02:03,378 INFO Train. Iter 187600 : Commit. 0.21998 PPL. 416.17 Recons. 0.00174 +2024-03-17 04:02:24,980 INFO Train. Iter 187800 : Commit. 0.22773 PPL. 415.76 Recons. 0.00198 +2024-03-17 04:02:46,680 INFO Train. Iter 188000 : Commit. 0.23350 PPL. 412.32 Recons. 0.00243 +2024-03-17 04:03:08,530 INFO Train. Iter 188200 : Commit. 0.23221 PPL. 414.41 Recons. 0.00219 +2024-03-17 04:03:29,110 INFO Train. Iter 188400 : Commit. 0.22620 PPL. 413.68 Recons. 0.00211 +2024-03-17 04:03:50,809 INFO Train. Iter 188600 : Commit. 0.23017 PPL. 415.75 Recons. 0.00196 +2024-03-17 04:04:12,660 INFO Train. Iter 188800 : Commit. 0.22192 PPL. 415.77 Recons. 0.00183 +2024-03-17 04:04:34,551 INFO Train. Iter 189000 : Commit. 0.22743 PPL. 414.93 Recons. 0.00204 +2024-03-17 04:04:56,855 INFO Train. Iter 189200 : Commit. 0.23457 PPL. 414.63 Recons. 0.00218 +2024-03-17 04:05:18,686 INFO Train. Iter 189400 : Commit. 0.22979 PPL. 414.45 Recons. 0.00206 +2024-03-17 04:05:40,856 INFO Train. Iter 189600 : Commit. 0.22314 PPL. 416.40 Recons. 0.00178 +2024-03-17 04:06:02,989 INFO Train. Iter 189800 : Commit. 0.22319 PPL. 415.60 Recons. 0.00183 +2024-03-17 04:06:24,942 INFO Train. Iter 190000 : Commit. 0.23903 PPL. 413.52 Recons. 0.00247 +2024-03-17 04:06:47,008 INFO Train. Iter 190200 : Commit. 0.23303 PPL. 414.08 Recons. 0.00229 +2024-03-17 04:07:08,038 INFO Train. Iter 190400 : Commit. 0.22184 PPL. 416.69 Recons. 0.00165 +2024-03-17 04:07:29,675 INFO Train. Iter 190600 : Commit. 0.23379 PPL. 413.48 Recons. 0.00250 +2024-03-17 04:07:51,798 INFO Train. Iter 190800 : Commit. 0.22997 PPL. 415.45 Recons. 0.00202 +2024-03-17 04:08:13,525 INFO Train. Iter 191000 : Commit. 0.22874 PPL. 413.52 Recons. 0.00218 +2024-03-17 04:08:35,973 INFO Train. Iter 191200 : Commit. 0.23479 PPL. 414.46 Recons. 0.00221 +2024-03-17 04:08:58,147 INFO Train. Iter 191400 : Commit. 0.22366 PPL. 416.18 Recons. 0.00178 +2024-03-17 04:09:19,744 INFO Train. Iter 191600 : Commit. 0.22780 PPL. 413.51 Recons. 0.00226 +2024-03-17 04:09:41,555 INFO Train. Iter 191800 : Commit. 0.22348 PPL. 414.47 Recons. 0.00188 +2024-03-17 04:10:03,120 INFO Train. Iter 192000 : Commit. 0.22298 PPL. 415.01 Recons. 0.00186 +2024-03-17 04:10:25,077 INFO Train. Iter 192200 : Commit. 0.23154 PPL. 413.50 Recons. 0.00228 +2024-03-17 04:10:46,106 INFO Train. Iter 192400 : Commit. 0.23265 PPL. 415.11 Recons. 0.00206 +2024-03-17 04:11:08,097 INFO Train. Iter 192600 : Commit. 0.22799 PPL. 414.49 Recons. 0.00212 +2024-03-17 04:11:30,101 INFO Train. Iter 192800 : Commit. 0.22757 PPL. 415.91 Recons. 0.00194 +2024-03-17 04:11:51,823 INFO Train. Iter 193000 : Commit. 0.23498 PPL. 413.22 Recons. 0.00241 +2024-03-17 04:12:13,700 INFO Train. Iter 193200 : Commit. 0.23015 PPL. 414.30 Recons. 0.00212 +2024-03-17 04:12:35,564 INFO Train. Iter 193400 : Commit. 0.21906 PPL. 416.58 Recons. 0.00167 +2024-03-17 04:12:57,778 INFO Train. Iter 193600 : Commit. 0.22836 PPL. 413.95 Recons. 0.00220 +2024-03-17 04:13:19,525 INFO Train. Iter 193800 : Commit. 0.22899 PPL. 416.54 Recons. 0.00198 +2024-03-17 04:13:41,068 INFO Train. Iter 194000 : Commit. 0.22983 PPL. 414.68 Recons. 0.00228 +2024-03-17 04:14:01,669 INFO Train. Iter 194200 : Commit. 0.22125 PPL. 415.03 Recons. 0.00177 +2024-03-17 04:14:23,288 INFO Train. Iter 194400 : Commit. 0.22587 PPL. 415.77 Recons. 0.00187 +2024-03-17 04:14:44,537 INFO Train. Iter 194600 : Commit. 0.22268 PPL. 415.25 Recons. 0.00197 +2024-03-17 04:15:05,875 INFO Train. Iter 194800 : Commit. 0.22976 PPL. 413.92 Recons. 0.00217 +2024-03-17 04:15:27,784 INFO Train. Iter 195000 : Commit. 0.22140 PPL. 414.24 Recons. 0.00190 +2024-03-17 04:15:49,250 INFO Train. Iter 195200 : Commit. 0.23956 PPL. 414.91 Recons. 0.00237 +2024-03-17 04:16:11,503 INFO Train. Iter 195400 : Commit. 0.22349 PPL. 413.78 Recons. 0.00194 +2024-03-17 04:16:33,450 INFO Train. Iter 195600 : Commit. 0.22958 PPL. 415.27 Recons. 0.00220 +2024-03-17 04:16:55,245 INFO Train. Iter 195800 : Commit. 0.22564 PPL. 413.01 Recons. 0.00206 +2024-03-17 04:17:16,902 INFO Train. Iter 196000 : Commit. 0.22242 PPL. 415.57 Recons. 0.00196 +2024-03-17 04:17:37,557 INFO Train. Iter 196200 : Commit. 0.22921 PPL. 415.62 Recons. 0.00202 +2024-03-17 04:17:59,988 INFO Train. Iter 196400 : Commit. 0.23132 PPL. 414.08 Recons. 0.00235 +2024-03-17 04:18:22,420 INFO Train. Iter 196600 : Commit. 0.22538 PPL. 414.40 Recons. 0.00191 +2024-03-17 04:18:44,638 INFO Train. Iter 196800 : Commit. 0.22708 PPL. 414.55 Recons. 0.00206 +2024-03-17 04:19:06,758 INFO Train. Iter 197000 : Commit. 0.22309 PPL. 415.58 Recons. 0.00181 +2024-03-17 04:19:28,739 INFO Train. Iter 197200 : Commit. 0.22749 PPL. 414.83 Recons. 0.00213 +2024-03-17 04:19:50,971 INFO Train. Iter 197400 : Commit. 0.22694 PPL. 415.72 Recons. 0.00194 +2024-03-17 04:20:13,524 INFO Train. Iter 197600 : Commit. 0.22401 PPL. 414.54 Recons. 0.00199 +2024-03-17 04:20:35,890 INFO Train. Iter 197800 : Commit. 0.23581 PPL. 413.18 Recons. 0.00226 +2024-03-17 04:20:57,869 INFO Train. Iter 198000 : Commit. 0.22083 PPL. 414.85 Recons. 0.00185 +2024-03-17 04:21:18,538 INFO Train. Iter 198200 : Commit. 0.22552 PPL. 414.23 Recons. 0.00206 +2024-03-17 04:21:40,270 INFO Train. Iter 198400 : Commit. 0.22067 PPL. 415.30 Recons. 0.00181 +2024-03-17 04:22:02,082 INFO Train. Iter 198600 : Commit. 0.22814 PPL. 414.63 Recons. 0.00216 +2024-03-17 04:22:24,022 INFO Train. Iter 198800 : Commit. 0.22918 PPL. 413.13 Recons. 0.00216 +2024-03-17 04:22:45,916 INFO Train. Iter 199000 : Commit. 0.22234 PPL. 414.85 Recons. 0.00182 +2024-03-17 04:23:08,394 INFO Train. Iter 199200 : Commit. 0.22870 PPL. 414.33 Recons. 0.00217 +2024-03-17 04:23:30,613 INFO Train. Iter 199400 : Commit. 0.22983 PPL. 413.57 Recons. 0.00220 +2024-03-17 04:23:52,535 INFO Train. Iter 199600 : Commit. 0.22673 PPL. 414.88 Recons. 0.00193 +2024-03-17 04:24:14,416 INFO Train. Iter 199800 : Commit. 0.23548 PPL. 413.82 Recons. 0.00229 +2024-03-17 04:24:35,053 INFO Train. Iter 200000 : Commit. 0.23467 PPL. 414.38 Recons. 0.00217 +2024-03-17 04:24:57,367 INFO Train. Iter 200200 : Commit. 0.23018 PPL. 417.03 Recons. 0.00162 +2024-03-17 04:25:18,616 INFO Train. Iter 200400 : Commit. 0.22519 PPL. 416.17 Recons. 0.00146 +2024-03-17 04:25:40,673 INFO Train. Iter 200600 : Commit. 0.22197 PPL. 416.41 Recons. 0.00137 +2024-03-17 04:26:02,418 INFO Train. Iter 200800 : Commit. 0.21841 PPL. 416.28 Recons. 0.00132 +2024-03-17 04:26:23,910 INFO Train. Iter 201000 : Commit. 0.21867 PPL. 416.31 Recons. 0.00131 +2024-03-17 04:26:45,913 INFO Train. Iter 201200 : Commit. 0.21314 PPL. 415.60 Recons. 0.00125 +2024-03-17 04:27:07,623 INFO Train. Iter 201400 : Commit. 0.21324 PPL. 416.13 Recons. 0.00123 +2024-03-17 04:27:29,496 INFO Train. Iter 201600 : Commit. 0.22033 PPL. 415.81 Recons. 0.00132 +2024-03-17 04:27:50,976 INFO Train. Iter 201800 : Commit. 0.21619 PPL. 415.54 Recons. 0.00128 +2024-03-17 04:28:11,565 INFO Train. Iter 202000 : Commit. 0.21260 PPL. 415.38 Recons. 0.00123 +2024-03-17 04:28:32,948 INFO Train. Iter 202200 : Commit. 0.21722 PPL. 414.91 Recons. 0.00131 +2024-03-17 04:28:54,090 INFO Train. Iter 202400 : Commit. 0.21636 PPL. 415.45 Recons. 0.00128 +2024-03-17 04:29:16,176 INFO Train. Iter 202600 : Commit. 0.21570 PPL. 415.36 Recons. 0.00126 +2024-03-17 04:29:37,855 INFO Train. Iter 202800 : Commit. 0.21920 PPL. 415.44 Recons. 0.00132 +2024-03-17 04:29:59,433 INFO Train. Iter 203000 : Commit. 0.21641 PPL. 415.54 Recons. 0.00127 +2024-03-17 04:30:20,927 INFO Train. Iter 203200 : Commit. 0.20963 PPL. 415.43 Recons. 0.00117 +2024-03-17 04:30:42,841 INFO Train. Iter 203400 : Commit. 0.21457 PPL. 415.02 Recons. 0.00125 +2024-03-17 04:31:04,245 INFO Train. Iter 203600 : Commit. 0.21850 PPL. 414.53 Recons. 0.00132 +2024-03-17 04:31:26,361 INFO Train. Iter 203800 : Commit. 0.21266 PPL. 415.02 Recons. 0.00121 +2024-03-17 04:31:47,368 INFO Train. Iter 204000 : Commit. 0.21296 PPL. 415.30 Recons. 0.00121 +2024-03-17 04:32:08,688 INFO Train. Iter 204200 : Commit. 0.21470 PPL. 414.85 Recons. 0.00124 +2024-03-17 04:32:30,646 INFO Train. Iter 204400 : Commit. 0.21529 PPL. 415.20 Recons. 0.00125 +2024-03-17 04:32:52,243 INFO Train. Iter 204600 : Commit. 0.21046 PPL. 415.29 Recons. 0.00117 +2024-03-17 04:33:14,222 INFO Train. Iter 204800 : Commit. 0.21151 PPL. 414.11 Recons. 0.00121 +2024-03-17 04:33:35,914 INFO Train. Iter 205000 : Commit. 0.21234 PPL. 415.10 Recons. 0.00119 +2024-03-17 04:33:57,536 INFO Train. Iter 205200 : Commit. 0.21809 PPL. 415.37 Recons. 0.00128 +2024-03-17 04:34:19,181 INFO Train. Iter 205400 : Commit. 0.20930 PPL. 415.35 Recons. 0.00114 +2024-03-17 04:34:40,573 INFO Train. Iter 205600 : Commit. 0.21464 PPL. 415.35 Recons. 0.00121 +2024-03-17 04:35:00,929 INFO Train. Iter 205800 : Commit. 0.21209 PPL. 414.83 Recons. 0.00119 +2024-03-17 04:35:22,182 INFO Train. Iter 206000 : Commit. 0.21579 PPL. 414.11 Recons. 0.00126 +2024-03-17 04:35:44,244 INFO Train. Iter 206200 : Commit. 0.21576 PPL. 414.89 Recons. 0.00124 +2024-03-17 04:36:05,403 INFO Train. Iter 206400 : Commit. 0.21465 PPL. 414.37 Recons. 0.00124 +2024-03-17 04:36:27,244 INFO Train. Iter 206600 : Commit. 0.21576 PPL. 415.05 Recons. 0.00123 +2024-03-17 04:36:48,377 INFO Train. Iter 206800 : Commit. 0.21516 PPL. 414.46 Recons. 0.00124 +2024-03-17 04:37:09,664 INFO Train. Iter 207000 : Commit. 0.21508 PPL. 414.63 Recons. 0.00123 +2024-03-17 04:37:30,787 INFO Train. Iter 207200 : Commit. 0.21203 PPL. 414.55 Recons. 0.00118 +2024-03-17 04:37:52,335 INFO Train. Iter 207400 : Commit. 0.21389 PPL. 414.60 Recons. 0.00122 +2024-03-17 04:38:13,965 INFO Train. Iter 207600 : Commit. 0.21506 PPL. 414.79 Recons. 0.00122 +2024-03-17 04:38:34,157 INFO Train. Iter 207800 : Commit. 0.22162 PPL. 414.40 Recons. 0.00133 +2024-03-17 04:38:56,500 INFO Train. Iter 208000 : Commit. 0.21363 PPL. 414.42 Recons. 0.00122 +2024-03-17 04:39:18,779 INFO Train. Iter 208200 : Commit. 0.21574 PPL. 414.70 Recons. 0.00125 +2024-03-17 04:39:40,248 INFO Train. Iter 208400 : Commit. 0.21357 PPL. 413.88 Recons. 0.00122 +2024-03-17 04:40:01,958 INFO Train. Iter 208600 : Commit. 0.21733 PPL. 413.82 Recons. 0.00128 +2024-03-17 04:40:23,798 INFO Train. Iter 208800 : Commit. 0.21642 PPL. 414.55 Recons. 0.00125 +2024-03-17 04:40:45,413 INFO Train. Iter 209000 : Commit. 0.21170 PPL. 414.78 Recons. 0.00118 +2024-03-17 04:41:07,037 INFO Train. Iter 209200 : Commit. 0.21870 PPL. 414.91 Recons. 0.00128 +2024-03-17 04:41:28,440 INFO Train. Iter 209400 : Commit. 0.21687 PPL. 415.15 Recons. 0.00125 +2024-03-17 04:41:50,157 INFO Train. Iter 209600 : Commit. 0.21730 PPL. 414.40 Recons. 0.00127 +2024-03-17 04:42:10,255 INFO Train. Iter 209800 : Commit. 0.21007 PPL. 414.80 Recons. 0.00116 +2024-03-17 04:42:31,879 INFO Train. Iter 210000 : Commit. 0.21655 PPL. 414.28 Recons. 0.00127 +2024-03-17 04:42:54,183 INFO Train. Iter 210200 : Commit. 0.21631 PPL. 414.52 Recons. 0.00125 +2024-03-17 04:43:15,575 INFO Train. Iter 210400 : Commit. 0.21330 PPL. 415.53 Recons. 0.00119 +2024-03-17 04:43:36,824 INFO Train. Iter 210600 : Commit. 0.21485 PPL. 413.89 Recons. 0.00125 +2024-03-17 04:43:57,724 INFO Train. Iter 210800 : Commit. 0.21170 PPL. 414.90 Recons. 0.00118 +2024-03-17 04:44:19,199 INFO Train. Iter 211000 : Commit. 0.21297 PPL. 414.31 Recons. 0.00120 +2024-03-17 04:44:40,935 INFO Train. Iter 211200 : Commit. 0.21199 PPL. 413.77 Recons. 0.00121 +2024-03-17 04:45:02,337 INFO Train. Iter 211400 : Commit. 0.21031 PPL. 414.59 Recons. 0.00117 +2024-03-17 04:45:22,733 INFO Train. Iter 211600 : Commit. 0.21157 PPL. 413.33 Recons. 0.00122 +2024-03-17 04:45:44,277 INFO Train. Iter 211800 : Commit. 0.21373 PPL. 413.76 Recons. 0.00123 +2024-03-17 04:46:05,466 INFO Train. Iter 212000 : Commit. 0.21909 PPL. 414.59 Recons. 0.00128 +2024-03-17 04:46:26,833 INFO Train. Iter 212200 : Commit. 0.21258 PPL. 413.95 Recons. 0.00119 +2024-03-17 04:46:48,432 INFO Train. Iter 212400 : Commit. 0.21715 PPL. 414.54 Recons. 0.00125 +2024-03-17 04:47:09,815 INFO Train. Iter 212600 : Commit. 0.21274 PPL. 415.40 Recons. 0.00117 +2024-03-17 04:47:31,120 INFO Train. Iter 212800 : Commit. 0.21517 PPL. 414.96 Recons. 0.00122 +2024-03-17 04:47:52,679 INFO Train. Iter 213000 : Commit. 0.21618 PPL. 414.70 Recons. 0.00125 +2024-03-17 04:48:13,558 INFO Train. Iter 213200 : Commit. 0.21311 PPL. 414.62 Recons. 0.00120 +2024-03-17 04:48:34,602 INFO Train. Iter 213400 : Commit. 0.21084 PPL. 414.19 Recons. 0.00118 +2024-03-17 04:48:54,773 INFO Train. Iter 213600 : Commit. 0.21916 PPL. 414.65 Recons. 0.00128 +2024-03-17 04:49:16,142 INFO Train. Iter 213800 : Commit. 0.21027 PPL. 414.23 Recons. 0.00117 +2024-03-17 04:49:37,710 INFO Train. Iter 214000 : Commit. 0.22027 PPL. 413.45 Recons. 0.00133 +2024-03-17 04:49:59,113 INFO Train. Iter 214200 : Commit. 0.21817 PPL. 414.61 Recons. 0.00127 +2024-03-17 04:50:20,116 INFO Train. Iter 214400 : Commit. 0.21218 PPL. 413.75 Recons. 0.00121 +2024-03-17 04:50:41,264 INFO Train. Iter 214600 : Commit. 0.21583 PPL. 414.21 Recons. 0.00125 +2024-03-17 04:51:02,369 INFO Train. Iter 214800 : Commit. 0.20992 PPL. 413.94 Recons. 0.00116 +2024-03-17 04:51:23,809 INFO Train. Iter 215000 : Commit. 0.21427 PPL. 413.85 Recons. 0.00123 +2024-03-17 04:51:45,180 INFO Train. Iter 215200 : Commit. 0.21249 PPL. 413.89 Recons. 0.00120 +2024-03-17 04:52:06,979 INFO Train. Iter 215400 : Commit. 0.21601 PPL. 414.17 Recons. 0.00125 +2024-03-17 04:52:27,211 INFO Train. Iter 215600 : Commit. 0.21570 PPL. 415.35 Recons. 0.00123 +2024-03-17 04:52:48,811 INFO Train. Iter 215800 : Commit. 0.20771 PPL. 414.00 Recons. 0.00113 +2024-03-17 04:53:10,785 INFO Train. Iter 216000 : Commit. 0.21280 PPL. 414.67 Recons. 0.00119 +2024-03-17 04:53:32,627 INFO Train. Iter 216200 : Commit. 0.21636 PPL. 414.32 Recons. 0.00125 +2024-03-17 04:53:54,528 INFO Train. Iter 216400 : Commit. 0.21575 PPL. 413.82 Recons. 0.00125 +2024-03-17 04:54:16,249 INFO Train. Iter 216600 : Commit. 0.21037 PPL. 414.86 Recons. 0.00115 +2024-03-17 04:54:37,988 INFO Train. Iter 216800 : Commit. 0.21514 PPL. 414.49 Recons. 0.00123 +2024-03-17 04:54:59,097 INFO Train. Iter 217000 : Commit. 0.21802 PPL. 414.52 Recons. 0.00127 +2024-03-17 04:55:21,256 INFO Train. Iter 217200 : Commit. 0.21585 PPL. 414.38 Recons. 0.00124 +2024-03-17 04:55:41,446 INFO Train. Iter 217400 : Commit. 0.21049 PPL. 414.00 Recons. 0.00118 +2024-03-17 04:56:03,141 INFO Train. Iter 217600 : Commit. 0.21201 PPL. 414.34 Recons. 0.00119 +2024-03-17 04:56:24,657 INFO Train. Iter 217800 : Commit. 0.20967 PPL. 414.05 Recons. 0.00117 +2024-03-17 04:56:45,559 INFO Train. Iter 218000 : Commit. 0.21579 PPL. 414.47 Recons. 0.00123 +2024-03-17 04:57:07,180 INFO Train. Iter 218200 : Commit. 0.21210 PPL. 413.91 Recons. 0.00118 +2024-03-17 04:57:28,555 INFO Train. Iter 218400 : Commit. 0.21470 PPL. 413.89 Recons. 0.00122 +2024-03-17 04:57:49,948 INFO Train. Iter 218600 : Commit. 0.21352 PPL. 414.91 Recons. 0.00119 +2024-03-17 04:58:11,452 INFO Train. Iter 218800 : Commit. 0.21206 PPL. 413.58 Recons. 0.00120 +2024-03-17 04:58:33,275 INFO Train. Iter 219000 : Commit. 0.20839 PPL. 414.07 Recons. 0.00114 +2024-03-17 04:58:55,109 INFO Train. Iter 219200 : Commit. 0.21767 PPL. 413.46 Recons. 0.00127 +2024-03-17 04:59:16,350 INFO Train. Iter 219400 : Commit. 0.20678 PPL. 413.61 Recons. 0.00112 +2024-03-17 04:59:38,537 INFO Train. Iter 219600 : Commit. 0.20970 PPL. 413.76 Recons. 0.00115 +2024-03-17 04:59:59,653 INFO Train. Iter 219800 : Commit. 0.20847 PPL. 414.51 Recons. 0.00112 +2024-03-17 05:00:21,094 INFO Train. Iter 220000 : Commit. 0.21499 PPL. 414.68 Recons. 0.00121 +2024-03-17 05:00:42,660 INFO Train. Iter 220200 : Commit. 0.21618 PPL. 413.98 Recons. 0.00124 +2024-03-17 05:01:04,389 INFO Train. Iter 220400 : Commit. 0.21650 PPL. 414.89 Recons. 0.00122 +2024-03-17 05:01:26,764 INFO Train. Iter 220600 : Commit. 0.21408 PPL. 413.32 Recons. 0.00122 +2024-03-17 05:01:48,577 INFO Train. Iter 220800 : Commit. 0.21203 PPL. 414.02 Recons. 0.00118 +2024-03-17 05:02:10,068 INFO Train. Iter 221000 : Commit. 0.21762 PPL. 413.74 Recons. 0.00127 +2024-03-17 05:02:31,841 INFO Train. Iter 221200 : Commit. 0.21303 PPL. 414.20 Recons. 0.00120 +2024-03-17 05:02:52,228 INFO Train. Iter 221400 : Commit. 0.21555 PPL. 413.59 Recons. 0.00124 +2024-03-17 05:03:13,511 INFO Train. Iter 221600 : Commit. 0.21243 PPL. 414.04 Recons. 0.00119 +2024-03-17 05:03:34,736 INFO Train. Iter 221800 : Commit. 0.21150 PPL. 413.82 Recons. 0.00119 +2024-03-17 05:03:56,149 INFO Train. Iter 222000 : Commit. 0.21100 PPL. 413.97 Recons. 0.00117 +2024-03-17 05:04:18,121 INFO Train. Iter 222200 : Commit. 0.21002 PPL. 413.77 Recons. 0.00116 +2024-03-17 05:04:40,092 INFO Train. Iter 222400 : Commit. 0.21034 PPL. 413.96 Recons. 0.00116 +2024-03-17 05:05:02,699 INFO Train. Iter 222600 : Commit. 0.21268 PPL. 413.43 Recons. 0.00120 +2024-03-17 05:05:24,808 INFO Train. Iter 222800 : Commit. 0.21228 PPL. 413.43 Recons. 0.00120 +2024-03-17 05:05:47,097 INFO Train. Iter 223000 : Commit. 0.20893 PPL. 413.95 Recons. 0.00114 +2024-03-17 05:06:07,160 INFO Train. Iter 223200 : Commit. 0.20787 PPL. 413.59 Recons. 0.00115 +2024-03-17 05:06:28,601 INFO Train. Iter 223400 : Commit. 0.21192 PPL. 413.30 Recons. 0.00121 +2024-03-17 05:06:49,545 INFO Train. Iter 223600 : Commit. 0.21049 PPL. 413.87 Recons. 0.00118 +2024-03-17 05:07:11,126 INFO Train. Iter 223800 : Commit. 0.20538 PPL. 413.18 Recons. 0.00112 +2024-03-17 05:07:32,133 INFO Train. Iter 224000 : Commit. 0.20752 PPL. 414.05 Recons. 0.00112 +2024-03-17 05:07:53,422 INFO Train. Iter 224200 : Commit. 0.20854 PPL. 414.21 Recons. 0.00114 +2024-03-17 05:08:15,036 INFO Train. Iter 224400 : Commit. 0.21536 PPL. 413.24 Recons. 0.00124 +2024-03-17 05:08:36,472 INFO Train. Iter 224600 : Commit. 0.21319 PPL. 413.91 Recons. 0.00121 +2024-03-17 05:08:57,685 INFO Train. Iter 224800 : Commit. 0.21633 PPL. 413.99 Recons. 0.00125 +2024-03-17 05:09:19,015 INFO Train. Iter 225000 : Commit. 0.21350 PPL. 414.56 Recons. 0.00120 +2024-03-17 05:09:39,307 INFO Train. Iter 225200 : Commit. 0.20966 PPL. 413.94 Recons. 0.00116 +2024-03-17 05:10:00,576 INFO Train. Iter 225400 : Commit. 0.20633 PPL. 413.47 Recons. 0.00112 +2024-03-17 05:10:22,561 INFO Train. Iter 225600 : Commit. 0.21078 PPL. 413.40 Recons. 0.00118 +2024-03-17 05:10:43,860 INFO Train. Iter 225800 : Commit. 0.21024 PPL. 413.59 Recons. 0.00118 +2024-03-17 05:11:04,916 INFO Train. Iter 226000 : Commit. 0.21422 PPL. 413.17 Recons. 0.00123 +2024-03-17 05:11:26,106 INFO Train. Iter 226200 : Commit. 0.21061 PPL. 413.76 Recons. 0.00117 +2024-03-17 05:11:47,755 INFO Train. Iter 226400 : Commit. 0.20995 PPL. 413.64 Recons. 0.00117 +2024-03-17 05:12:09,518 INFO Train. Iter 226600 : Commit. 0.20949 PPL. 413.62 Recons. 0.00116 +2024-03-17 05:12:30,893 INFO Train. Iter 226800 : Commit. 0.21019 PPL. 413.34 Recons. 0.00117 +2024-03-17 05:12:51,449 INFO Train. Iter 227000 : Commit. 0.21475 PPL. 413.42 Recons. 0.00122 +2024-03-17 05:13:11,666 INFO Train. Iter 227200 : Commit. 0.21268 PPL. 413.86 Recons. 0.00119 +2024-03-17 05:13:32,747 INFO Train. Iter 227400 : Commit. 0.21352 PPL. 413.21 Recons. 0.00123 +2024-03-17 05:13:54,702 INFO Train. Iter 227600 : Commit. 0.21000 PPL. 414.17 Recons. 0.00115 +2024-03-17 05:14:16,101 INFO Train. Iter 227800 : Commit. 0.21305 PPL. 413.15 Recons. 0.00121 +2024-03-17 05:14:37,564 INFO Train. Iter 228000 : Commit. 0.21197 PPL. 414.01 Recons. 0.00117 +2024-03-17 05:14:59,975 INFO Train. Iter 228200 : Commit. 0.21133 PPL. 414.31 Recons. 0.00116 +2024-03-17 05:15:21,754 INFO Train. Iter 228400 : Commit. 0.21408 PPL. 413.64 Recons. 0.00121 +2024-03-17 05:15:43,710 INFO Train. Iter 228600 : Commit. 0.21233 PPL. 413.89 Recons. 0.00119 +2024-03-17 05:16:05,153 INFO Train. Iter 228800 : Commit. 0.20909 PPL. 412.56 Recons. 0.00118 +2024-03-17 05:16:25,668 INFO Train. Iter 229000 : Commit. 0.21115 PPL. 414.38 Recons. 0.00116 +2024-03-17 05:16:46,637 INFO Train. Iter 229200 : Commit. 0.20860 PPL. 413.60 Recons. 0.00115 +2024-03-17 05:17:08,196 INFO Train. Iter 229400 : Commit. 0.21347 PPL. 413.81 Recons. 0.00121 +2024-03-17 05:17:29,430 INFO Train. Iter 229600 : Commit. 0.21234 PPL. 412.84 Recons. 0.00122 +2024-03-17 05:17:51,104 INFO Train. Iter 229800 : Commit. 0.21119 PPL. 413.53 Recons. 0.00120 +2024-03-17 05:18:12,551 INFO Train. Iter 230000 : Commit. 0.21361 PPL. 412.86 Recons. 0.00124 +2024-03-17 05:18:33,715 INFO Train. Iter 230200 : Commit. 0.21110 PPL. 413.47 Recons. 0.00119 +2024-03-17 05:18:55,325 INFO Train. Iter 230400 : Commit. 0.20717 PPL. 413.47 Recons. 0.00113 +2024-03-17 05:19:16,476 INFO Train. Iter 230600 : Commit. 0.22152 PPL. 412.82 Recons. 0.00135 +2024-03-17 05:19:37,891 INFO Train. Iter 230800 : Commit. 0.21040 PPL. 413.01 Recons. 0.00119 +2024-03-17 05:19:58,065 INFO Train. Iter 231000 : Commit. 0.20650 PPL. 413.57 Recons. 0.00113 +2024-03-17 05:20:19,343 INFO Train. Iter 231200 : Commit. 0.21230 PPL. 414.58 Recons. 0.00118 +2024-03-17 05:20:40,738 INFO Train. Iter 231400 : Commit. 0.21218 PPL. 413.27 Recons. 0.00122 +2024-03-17 05:21:02,431 INFO Train. Iter 231600 : Commit. 0.21193 PPL. 413.61 Recons. 0.00121 +2024-03-17 05:21:24,513 INFO Train. Iter 231800 : Commit. 0.21126 PPL. 413.01 Recons. 0.00120 +2024-03-17 05:21:46,519 INFO Train. Iter 232000 : Commit. 0.21181 PPL. 413.20 Recons. 0.00121 +2024-03-17 05:22:08,810 INFO Train. Iter 232200 : Commit. 0.21288 PPL. 413.25 Recons. 0.00123 +2024-03-17 05:22:30,894 INFO Train. Iter 232400 : Commit. 0.21260 PPL. 413.16 Recons. 0.00122 +2024-03-17 05:22:52,794 INFO Train. Iter 232600 : Commit. 0.20884 PPL. 413.70 Recons. 0.00116 +2024-03-17 05:23:13,960 INFO Train. Iter 232800 : Commit. 0.21098 PPL. 413.49 Recons. 0.00119 +2024-03-17 05:23:35,154 INFO Train. Iter 233000 : Commit. 0.20974 PPL. 412.86 Recons. 0.00118 +2024-03-17 05:23:57,215 INFO Train. Iter 233200 : Commit. 0.20563 PPL. 414.23 Recons. 0.00110 +2024-03-17 05:24:18,803 INFO Train. Iter 233400 : Commit. 0.20645 PPL. 413.14 Recons. 0.00113 +2024-03-17 05:24:40,867 INFO Train. Iter 233600 : Commit. 0.20968 PPL. 413.51 Recons. 0.00117 +2024-03-17 05:25:02,412 INFO Train. Iter 233800 : Commit. 0.21097 PPL. 412.94 Recons. 0.00120 +2024-03-17 05:25:24,319 INFO Train. Iter 234000 : Commit. 0.20805 PPL. 412.84 Recons. 0.00115 +2024-03-17 05:25:46,076 INFO Train. Iter 234200 : Commit. 0.20787 PPL. 413.53 Recons. 0.00113 +2024-03-17 05:26:07,572 INFO Train. Iter 234400 : Commit. 0.21327 PPL. 413.69 Recons. 0.00121 +2024-03-17 05:26:28,673 INFO Train. Iter 234600 : Commit. 0.20979 PPL. 413.16 Recons. 0.00116 +2024-03-17 05:26:48,761 INFO Train. Iter 234800 : Commit. 0.21327 PPL. 413.07 Recons. 0.00121 +2024-03-17 05:27:10,468 INFO Train. Iter 235000 : Commit. 0.21244 PPL. 413.44 Recons. 0.00119 +2024-03-17 05:27:32,491 INFO Train. Iter 235200 : Commit. 0.21074 PPL. 412.61 Recons. 0.00119 +2024-03-17 05:27:54,693 INFO Train. Iter 235400 : Commit. 0.20951 PPL. 413.59 Recons. 0.00116 +2024-03-17 05:28:16,411 INFO Train. Iter 235600 : Commit. 0.21057 PPL. 413.59 Recons. 0.00117 +2024-03-17 05:28:37,978 INFO Train. Iter 235800 : Commit. 0.21283 PPL. 413.64 Recons. 0.00121 +2024-03-17 05:28:59,468 INFO Train. Iter 236000 : Commit. 0.20654 PPL. 413.73 Recons. 0.00112 +2024-03-17 05:29:20,505 INFO Train. Iter 236200 : Commit. 0.21576 PPL. 413.30 Recons. 0.00125 +2024-03-17 05:29:42,001 INFO Train. Iter 236400 : Commit. 0.20465 PPL. 412.36 Recons. 0.00112 +2024-03-17 05:30:03,633 INFO Train. Iter 236600 : Commit. 0.20514 PPL. 412.78 Recons. 0.00112 +2024-03-17 05:30:24,069 INFO Train. Iter 236800 : Commit. 0.20744 PPL. 412.81 Recons. 0.00114 +2024-03-17 05:30:45,634 INFO Train. Iter 237000 : Commit. 0.20638 PPL. 412.89 Recons. 0.00113 +2024-03-17 05:31:07,453 INFO Train. Iter 237200 : Commit. 0.21046 PPL. 413.24 Recons. 0.00117 +2024-03-17 05:31:29,173 INFO Train. Iter 237400 : Commit. 0.21108 PPL. 412.29 Recons. 0.00121 +2024-03-17 05:31:51,130 INFO Train. Iter 237600 : Commit. 0.20446 PPL. 412.85 Recons. 0.00111 +2024-03-17 05:32:13,130 INFO Train. Iter 237800 : Commit. 0.20898 PPL. 413.00 Recons. 0.00117 +2024-03-17 05:32:34,973 INFO Train. Iter 238000 : Commit. 0.20991 PPL. 412.73 Recons. 0.00118 +2024-03-17 05:32:56,667 INFO Train. Iter 238200 : Commit. 0.20821 PPL. 413.05 Recons. 0.00114 +2024-03-17 05:33:18,592 INFO Train. Iter 238400 : Commit. 0.20758 PPL. 413.34 Recons. 0.00114 +2024-03-17 05:33:40,058 INFO Train. Iter 238600 : Commit. 0.21015 PPL. 413.53 Recons. 0.00117 +2024-03-17 05:34:00,634 INFO Train. Iter 238800 : Commit. 0.20844 PPL. 413.09 Recons. 0.00115 +2024-03-17 05:34:22,852 INFO Train. Iter 239000 : Commit. 0.20789 PPL. 413.84 Recons. 0.00112 +2024-03-17 05:34:44,882 INFO Train. Iter 239200 : Commit. 0.21086 PPL. 413.04 Recons. 0.00118 +2024-03-17 05:35:06,598 INFO Train. Iter 239400 : Commit. 0.20717 PPL. 412.91 Recons. 0.00113 +2024-03-17 05:35:28,359 INFO Train. Iter 239600 : Commit. 0.21304 PPL. 413.32 Recons. 0.00120 +2024-03-17 05:35:50,420 INFO Train. Iter 239800 : Commit. 0.20721 PPL. 412.40 Recons. 0.00116 +2024-03-17 05:36:12,935 INFO Train. Iter 240000 : Commit. 0.20869 PPL. 412.66 Recons. 0.00116 +2024-03-17 05:36:34,884 INFO Train. Iter 240200 : Commit. 0.20783 PPL. 413.04 Recons. 0.00114 +2024-03-17 05:36:56,168 INFO Train. Iter 240400 : Commit. 0.20805 PPL. 412.90 Recons. 0.00114 +2024-03-17 05:37:16,182 INFO Train. Iter 240600 : Commit. 0.20611 PPL. 413.40 Recons. 0.00111 +2024-03-17 05:37:37,726 INFO Train. Iter 240800 : Commit. 0.20476 PPL. 412.36 Recons. 0.00112 +2024-03-17 05:37:59,161 INFO Train. Iter 241000 : Commit. 0.20391 PPL. 412.32 Recons. 0.00110 +2024-03-17 05:38:20,956 INFO Train. Iter 241200 : Commit. 0.20807 PPL. 412.13 Recons. 0.00117 +2024-03-17 05:38:42,201 INFO Train. Iter 241400 : Commit. 0.20876 PPL. 412.98 Recons. 0.00116 +2024-03-17 05:39:03,571 INFO Train. Iter 241600 : Commit. 0.20665 PPL. 412.00 Recons. 0.00115 +2024-03-17 05:39:25,477 INFO Train. Iter 241800 : Commit. 0.21141 PPL. 412.80 Recons. 0.00120 +2024-03-17 05:39:47,118 INFO Train. Iter 242000 : Commit. 0.21030 PPL. 413.11 Recons. 0.00118 +2024-03-17 05:40:08,600 INFO Train. Iter 242200 : Commit. 0.20982 PPL. 412.87 Recons. 0.00117 +2024-03-17 05:40:30,918 INFO Train. Iter 242400 : Commit. 0.21173 PPL. 413.16 Recons. 0.00119 +2024-03-17 05:40:50,920 INFO Train. Iter 242600 : Commit. 0.20468 PPL. 412.43 Recons. 0.00112 +2024-03-17 05:41:11,999 INFO Train. Iter 242800 : Commit. 0.20895 PPL. 412.88 Recons. 0.00116 +2024-03-17 05:41:33,345 INFO Train. Iter 243000 : Commit. 0.20793 PPL. 412.29 Recons. 0.00116 +2024-03-17 05:41:54,810 INFO Train. Iter 243200 : Commit. 0.21581 PPL. 413.01 Recons. 0.00126 +2024-03-17 05:42:16,458 INFO Train. Iter 243400 : Commit. 0.20441 PPL. 412.74 Recons. 0.00111 +2024-03-17 05:42:37,710 INFO Train. Iter 243600 : Commit. 0.20567 PPL. 412.30 Recons. 0.00113 +2024-03-17 05:42:59,182 INFO Train. Iter 243800 : Commit. 0.21081 PPL. 413.60 Recons. 0.00117 +2024-03-17 05:43:20,679 INFO Train. Iter 244000 : Commit. 0.20184 PPL. 412.16 Recons. 0.00108 +2024-03-17 05:43:41,879 INFO Train. Iter 244200 : Commit. 0.21015 PPL. 413.41 Recons. 0.00118 +2024-03-17 05:44:03,415 INFO Train. Iter 244400 : Commit. 0.20971 PPL. 413.10 Recons. 0.00117 +2024-03-17 05:44:23,908 INFO Train. Iter 244600 : Commit. 0.20928 PPL. 412.16 Recons. 0.00119 +2024-03-17 05:44:44,733 INFO Train. Iter 244800 : Commit. 0.21319 PPL. 411.81 Recons. 0.00125 +2024-03-17 05:45:05,520 INFO Train. Iter 245000 : Commit. 0.20596 PPL. 413.25 Recons. 0.00113 +2024-03-17 05:45:26,902 INFO Train. Iter 245200 : Commit. 0.20948 PPL. 412.37 Recons. 0.00119 +2024-03-17 05:45:48,511 INFO Train. Iter 245400 : Commit. 0.21022 PPL. 412.31 Recons. 0.00120 +2024-03-17 05:46:09,918 INFO Train. Iter 245600 : Commit. 0.20550 PPL. 413.13 Recons. 0.00112 +2024-03-17 05:46:31,154 INFO Train. Iter 245800 : Commit. 0.20920 PPL. 413.29 Recons. 0.00117 +2024-03-17 05:46:52,613 INFO Train. Iter 246000 : Commit. 0.20758 PPL. 412.64 Recons. 0.00115 +2024-03-17 05:47:13,914 INFO Train. Iter 246200 : Commit. 0.21680 PPL. 413.06 Recons. 0.00128 +2024-03-17 05:47:33,945 INFO Train. Iter 246400 : Commit. 0.20539 PPL. 412.53 Recons. 0.00113 +2024-03-17 05:47:55,143 INFO Train. Iter 246600 : Commit. 0.20705 PPL. 412.15 Recons. 0.00116 +2024-03-17 05:48:16,479 INFO Train. Iter 246800 : Commit. 0.20755 PPL. 412.55 Recons. 0.00117 +2024-03-17 05:48:37,666 INFO Train. Iter 247000 : Commit. 0.20714 PPL. 412.34 Recons. 0.00116 +2024-03-17 05:48:58,406 INFO Train. Iter 247200 : Commit. 0.20733 PPL. 412.72 Recons. 0.00115 +2024-03-17 05:49:19,503 INFO Train. Iter 247400 : Commit. 0.20618 PPL. 412.99 Recons. 0.00114 +2024-03-17 05:49:40,822 INFO Train. Iter 247600 : Commit. 0.20857 PPL. 413.06 Recons. 0.00115 +2024-03-17 05:50:02,454 INFO Train. Iter 247800 : Commit. 0.20482 PPL. 412.87 Recons. 0.00111 +2024-03-17 05:50:24,641 INFO Train. Iter 248000 : Commit. 0.20413 PPL. 412.56 Recons. 0.00111 +2024-03-17 05:50:46,446 INFO Train. Iter 248200 : Commit. 0.20551 PPL. 412.43 Recons. 0.00112 +2024-03-17 05:51:07,216 INFO Train. Iter 248400 : Commit. 0.21005 PPL. 413.07 Recons. 0.00117 +2024-03-17 05:51:29,030 INFO Train. Iter 248600 : Commit. 0.20443 PPL. 413.01 Recons. 0.00110 +2024-03-17 05:51:51,136 INFO Train. Iter 248800 : Commit. 0.20598 PPL. 413.04 Recons. 0.00112 +2024-03-17 05:52:13,378 INFO Train. Iter 249000 : Commit. 0.20643 PPL. 413.62 Recons. 0.00112 +2024-03-17 05:52:35,367 INFO Train. Iter 249200 : Commit. 0.20708 PPL. 413.02 Recons. 0.00113 +2024-03-17 05:52:57,398 INFO Train. Iter 249400 : Commit. 0.21195 PPL. 412.70 Recons. 0.00120 +2024-03-17 05:53:19,127 INFO Train. Iter 249600 : Commit. 0.20725 PPL. 411.65 Recons. 0.00116 +2024-03-17 05:53:40,829 INFO Train. Iter 249800 : Commit. 0.21194 PPL. 413.23 Recons. 0.00120 +2024-03-17 05:54:02,898 INFO Train. Iter 250000 : Commit. 0.20965 PPL. 411.95 Recons. 0.00120 +2024-03-17 05:54:24,328 INFO Train. Iter 250200 : Commit. 0.21030 PPL. 412.94 Recons. 0.00117 +2024-03-17 05:54:44,980 INFO Train. Iter 250400 : Commit. 0.21272 PPL. 412.80 Recons. 0.00122 +2024-03-17 05:55:06,481 INFO Train. Iter 250600 : Commit. 0.20504 PPL. 412.58 Recons. 0.00112 +2024-03-17 05:55:28,029 INFO Train. Iter 250800 : Commit. 0.20951 PPL. 412.08 Recons. 0.00118 +2024-03-17 05:55:49,919 INFO Train. Iter 251000 : Commit. 0.20603 PPL. 413.13 Recons. 0.00111 +2024-03-17 05:56:11,671 INFO Train. Iter 251200 : Commit. 0.21167 PPL. 413.14 Recons. 0.00120 +2024-03-17 05:56:33,739 INFO Train. Iter 251400 : Commit. 0.21450 PPL. 412.65 Recons. 0.00123 +2024-03-17 05:56:54,850 INFO Train. Iter 251600 : Commit. 0.20799 PPL. 412.08 Recons. 0.00117 +2024-03-17 05:57:16,285 INFO Train. Iter 251800 : Commit. 0.20750 PPL. 412.70 Recons. 0.00116 +2024-03-17 05:57:37,727 INFO Train. Iter 252000 : Commit. 0.20883 PPL. 413.01 Recons. 0.00116 +2024-03-17 05:57:58,360 INFO Train. Iter 252200 : Commit. 0.20896 PPL. 412.10 Recons. 0.00118 +2024-03-17 05:58:19,695 INFO Train. Iter 252400 : Commit. 0.20645 PPL. 413.14 Recons. 0.00113 +2024-03-17 05:58:41,449 INFO Train. Iter 252600 : Commit. 0.20895 PPL. 412.16 Recons. 0.00117 +2024-03-17 05:59:02,465 INFO Train. Iter 252800 : Commit. 0.20838 PPL. 412.88 Recons. 0.00116 +2024-03-17 05:59:23,993 INFO Train. Iter 253000 : Commit. 0.20863 PPL. 412.44 Recons. 0.00118 +2024-03-17 05:59:45,582 INFO Train. Iter 253200 : Commit. 0.21153 PPL. 411.88 Recons. 0.00122 +2024-03-17 06:00:07,096 INFO Train. Iter 253400 : Commit. 0.20960 PPL. 412.41 Recons. 0.00119 +2024-03-17 06:00:29,266 INFO Train. Iter 253600 : Commit. 0.20940 PPL. 412.76 Recons. 0.00118 +2024-03-17 06:00:51,376 INFO Train. Iter 253800 : Commit. 0.20638 PPL. 413.03 Recons. 0.00113 +2024-03-17 06:01:12,612 INFO Train. Iter 254000 : Commit. 0.20540 PPL. 412.86 Recons. 0.00112 +2024-03-17 06:01:33,001 INFO Train. Iter 254200 : Commit. 0.20830 PPL. 413.70 Recons. 0.00114 +2024-03-17 06:01:54,172 INFO Train. Iter 254400 : Commit. 0.21036 PPL. 413.42 Recons. 0.00118 +2024-03-17 06:02:15,603 INFO Train. Iter 254600 : Commit. 0.20920 PPL. 413.17 Recons. 0.00117 +2024-03-17 06:02:37,229 INFO Train. Iter 254800 : Commit. 0.21097 PPL. 412.84 Recons. 0.00120 +2024-03-17 06:02:58,701 INFO Train. Iter 255000 : Commit. 0.20623 PPL. 411.87 Recons. 0.00116 +2024-03-17 06:03:20,606 INFO Train. Iter 255200 : Commit. 0.20666 PPL. 412.29 Recons. 0.00115 +2024-03-17 06:03:41,909 INFO Train. Iter 255400 : Commit. 0.20402 PPL. 412.11 Recons. 0.00112 +2024-03-17 06:04:03,323 INFO Train. Iter 255600 : Commit. 0.20854 PPL. 412.60 Recons. 0.00117 +2024-03-17 06:04:24,777 INFO Train. Iter 255800 : Commit. 0.20108 PPL. 412.53 Recons. 0.00107 +2024-03-17 06:04:46,115 INFO Train. Iter 256000 : Commit. 0.20718 PPL. 413.13 Recons. 0.00114 +2024-03-17 06:05:06,649 INFO Train. Iter 256200 : Commit. 0.20516 PPL. 411.86 Recons. 0.00114 +2024-03-17 06:05:28,182 INFO Train. Iter 256400 : Commit. 0.20517 PPL. 412.51 Recons. 0.00113 +2024-03-17 06:05:50,178 INFO Train. Iter 256600 : Commit. 0.20845 PPL. 412.04 Recons. 0.00119 +2024-03-17 06:06:12,102 INFO Train. Iter 256800 : Commit. 0.20467 PPL. 411.18 Recons. 0.00115 +2024-03-17 06:06:34,225 INFO Train. Iter 257000 : Commit. 0.20234 PPL. 412.17 Recons. 0.00110 +2024-03-17 06:06:56,301 INFO Train. Iter 257200 : Commit. 0.20835 PPL. 412.23 Recons. 0.00119 +2024-03-17 06:07:18,355 INFO Train. Iter 257400 : Commit. 0.20061 PPL. 411.71 Recons. 0.00108 +2024-03-17 06:07:40,281 INFO Train. Iter 257600 : Commit. 0.20088 PPL. 411.58 Recons. 0.00109 +2024-03-17 06:08:01,744 INFO Train. Iter 257800 : Commit. 0.20836 PPL. 412.02 Recons. 0.00118 +2024-03-17 06:08:22,312 INFO Train. Iter 258000 : Commit. 0.20540 PPL. 412.37 Recons. 0.00114 +2024-03-17 06:08:44,552 INFO Train. Iter 258200 : Commit. 0.20861 PPL. 412.08 Recons. 0.00118 +2024-03-17 06:09:06,604 INFO Train. Iter 258400 : Commit. 0.20374 PPL. 412.74 Recons. 0.00111 +2024-03-17 06:09:27,776 INFO Train. Iter 258600 : Commit. 0.20378 PPL. 411.92 Recons. 0.00112 +2024-03-17 06:09:49,056 INFO Train. Iter 258800 : Commit. 0.20685 PPL. 412.22 Recons. 0.00115 +2024-03-17 06:10:10,207 INFO Train. Iter 259000 : Commit. 0.20755 PPL. 412.54 Recons. 0.00116 +2024-03-17 06:10:31,123 INFO Train. Iter 259200 : Commit. 0.20417 PPL. 411.67 Recons. 0.00113 +2024-03-17 06:10:52,370 INFO Train. Iter 259400 : Commit. 0.20387 PPL. 412.76 Recons. 0.00111 +2024-03-17 06:11:13,544 INFO Train. Iter 259600 : Commit. 0.20485 PPL. 412.58 Recons. 0.00112 +2024-03-17 06:11:34,534 INFO Train. Iter 259800 : Commit. 0.21460 PPL. 411.83 Recons. 0.00127 +2024-03-17 06:11:55,432 INFO Train. Iter 260000 : Commit. 0.20714 PPL. 411.99 Recons. 0.00116 +2024-03-17 06:12:16,703 INFO Train. Iter 260200 : Commit. 0.20142 PPL. 411.90 Recons. 0.00109 +2024-03-17 06:12:38,125 INFO Train. Iter 260400 : Commit. 0.20314 PPL. 412.01 Recons. 0.00111 +2024-03-17 06:12:59,345 INFO Train. Iter 260600 : Commit. 0.20181 PPL. 411.55 Recons. 0.00110 +2024-03-17 06:13:20,511 INFO Train. Iter 260800 : Commit. 0.20853 PPL. 412.10 Recons. 0.00117 +2024-03-17 06:13:41,771 INFO Train. Iter 261000 : Commit. 0.20576 PPL. 412.41 Recons. 0.00113 +2024-03-17 06:14:03,362 INFO Train. Iter 261200 : Commit. 0.20060 PPL. 411.70 Recons. 0.00107 +2024-03-17 06:14:25,096 INFO Train. Iter 261400 : Commit. 0.20607 PPL. 412.40 Recons. 0.00114 +2024-03-17 06:14:46,606 INFO Train. Iter 261600 : Commit. 0.20917 PPL. 412.37 Recons. 0.00118 +2024-03-17 06:15:08,495 INFO Train. Iter 261800 : Commit. 0.20829 PPL. 411.55 Recons. 0.00118 +2024-03-17 06:15:28,361 INFO Train. Iter 262000 : Commit. 0.20170 PPL. 411.18 Recons. 0.00110 +2024-03-17 06:15:50,281 INFO Train. Iter 262200 : Commit. 0.20285 PPL. 412.36 Recons. 0.00109 +2024-03-17 06:16:12,012 INFO Train. Iter 262400 : Commit. 0.20158 PPL. 411.87 Recons. 0.00109 +2024-03-17 06:16:33,445 INFO Train. Iter 262600 : Commit. 0.20384 PPL. 412.38 Recons. 0.00111 +2024-03-17 06:16:54,714 INFO Train. Iter 262800 : Commit. 0.20788 PPL. 412.44 Recons. 0.00116 +2024-03-17 06:17:16,200 INFO Train. Iter 263000 : Commit. 0.20678 PPL. 411.23 Recons. 0.00116 +2024-03-17 06:17:37,267 INFO Train. Iter 263200 : Commit. 0.21250 PPL. 412.41 Recons. 0.00120 +2024-03-17 06:17:58,471 INFO Train. Iter 263400 : Commit. 0.20384 PPL. 411.82 Recons. 0.00111 +2024-03-17 06:18:19,995 INFO Train. Iter 263600 : Commit. 0.20639 PPL. 411.87 Recons. 0.00115 +2024-03-17 06:18:40,452 INFO Train. Iter 263800 : Commit. 0.20845 PPL. 412.07 Recons. 0.00116 +2024-03-17 06:19:01,935 INFO Train. Iter 264000 : Commit. 0.20454 PPL. 411.16 Recons. 0.00114 +2024-03-17 06:19:23,502 INFO Train. Iter 264200 : Commit. 0.21057 PPL. 411.98 Recons. 0.00119 +2024-03-17 06:19:44,843 INFO Train. Iter 264400 : Commit. 0.20638 PPL. 412.53 Recons. 0.00113 +2024-03-17 06:20:06,285 INFO Train. Iter 264600 : Commit. 0.20713 PPL. 411.34 Recons. 0.00117 +2024-03-17 06:20:27,216 INFO Train. Iter 264800 : Commit. 0.20397 PPL. 411.65 Recons. 0.00112 +2024-03-17 06:20:48,685 INFO Train. Iter 265000 : Commit. 0.20560 PPL. 412.57 Recons. 0.00113 +2024-03-17 06:21:10,563 INFO Train. Iter 265200 : Commit. 0.20498 PPL. 411.16 Recons. 0.00114 +2024-03-17 06:21:31,587 INFO Train. Iter 265400 : Commit. 0.20431 PPL. 411.79 Recons. 0.00112 +2024-03-17 06:21:52,479 INFO Train. Iter 265600 : Commit. 0.20917 PPL. 413.10 Recons. 0.00116 +2024-03-17 06:22:12,436 INFO Train. Iter 265800 : Commit. 0.20860 PPL. 411.90 Recons. 0.00118 +2024-03-17 06:22:33,354 INFO Train. Iter 266000 : Commit. 0.20273 PPL. 412.48 Recons. 0.00109 +2024-03-17 06:22:54,486 INFO Train. Iter 266200 : Commit. 0.20419 PPL. 411.99 Recons. 0.00112 +2024-03-17 06:23:15,736 INFO Train. Iter 266400 : Commit. 0.20543 PPL. 411.94 Recons. 0.00114 +2024-03-17 06:23:36,929 INFO Train. Iter 266600 : Commit. 0.20768 PPL. 411.87 Recons. 0.00116 +2024-03-17 06:23:58,855 INFO Train. Iter 266800 : Commit. 0.21076 PPL. 412.90 Recons. 0.00118 +2024-03-17 06:24:20,260 INFO Train. Iter 267000 : Commit. 0.20539 PPL. 411.85 Recons. 0.00114 +2024-03-17 06:24:41,515 INFO Train. Iter 267200 : Commit. 0.20743 PPL. 411.84 Recons. 0.00116 +2024-03-17 06:25:02,559 INFO Train. Iter 267400 : Commit. 0.20363 PPL. 411.54 Recons. 0.00112 +2024-03-17 06:25:23,972 INFO Train. Iter 267600 : Commit. 0.19963 PPL. 411.50 Recons. 0.00107 +2024-03-17 06:25:44,147 INFO Train. Iter 267800 : Commit. 0.20595 PPL. 412.49 Recons. 0.00113 +2024-03-17 06:26:05,112 INFO Train. Iter 268000 : Commit. 0.20655 PPL. 412.20 Recons. 0.00114 +2024-03-17 06:26:26,966 INFO Train. Iter 268200 : Commit. 0.20660 PPL. 411.47 Recons. 0.00115 +2024-03-17 06:26:48,647 INFO Train. Iter 268400 : Commit. 0.20575 PPL. 412.42 Recons. 0.00113 +2024-03-17 06:27:10,223 INFO Train. Iter 268600 : Commit. 0.20683 PPL. 411.13 Recons. 0.00116 +2024-03-17 06:27:31,573 INFO Train. Iter 268800 : Commit. 0.20455 PPL. 411.85 Recons. 0.00112 +2024-03-17 06:27:52,728 INFO Train. Iter 269000 : Commit. 0.20367 PPL. 411.01 Recons. 0.00112 +2024-03-17 06:28:13,788 INFO Train. Iter 269200 : Commit. 0.20579 PPL. 411.51 Recons. 0.00114 +2024-03-17 06:28:34,741 INFO Train. Iter 269400 : Commit. 0.20453 PPL. 411.48 Recons. 0.00112 +2024-03-17 06:28:55,186 INFO Train. Iter 269600 : Commit. 0.20739 PPL. 411.57 Recons. 0.00116 +2024-03-17 06:29:16,331 INFO Train. Iter 269800 : Commit. 0.20450 PPL. 411.64 Recons. 0.00113 +2024-03-17 06:29:37,761 INFO Train. Iter 270000 : Commit. 0.20567 PPL. 411.24 Recons. 0.00114 +2024-03-17 06:29:59,296 INFO Train. Iter 270200 : Commit. 0.20657 PPL. 411.69 Recons. 0.00114 +2024-03-17 06:30:21,214 INFO Train. Iter 270400 : Commit. 0.20549 PPL. 411.51 Recons. 0.00113 +2024-03-17 06:30:43,360 INFO Train. Iter 270600 : Commit. 0.20461 PPL. 411.97 Recons. 0.00111 +2024-03-17 06:31:04,536 INFO Train. Iter 270800 : Commit. 0.20439 PPL. 410.33 Recons. 0.00114 +2024-03-17 06:31:26,355 INFO Train. Iter 271000 : Commit. 0.20801 PPL. 411.59 Recons. 0.00116 +2024-03-17 06:31:48,104 INFO Train. Iter 271200 : Commit. 0.20469 PPL. 410.75 Recons. 0.00114 +2024-03-17 06:32:09,724 INFO Train. Iter 271400 : Commit. 0.20114 PPL. 411.36 Recons. 0.00107 +2024-03-17 06:32:30,801 INFO Train. Iter 271600 : Commit. 0.20148 PPL. 411.40 Recons. 0.00108 +2024-03-17 06:32:52,801 INFO Train. Iter 271800 : Commit. 0.20783 PPL. 411.18 Recons. 0.00117 +2024-03-17 06:33:14,393 INFO Train. Iter 272000 : Commit. 0.20712 PPL. 410.67 Recons. 0.00116 +2024-03-17 06:33:36,057 INFO Train. Iter 272200 : Commit. 0.20802 PPL. 411.63 Recons. 0.00116 +2024-03-17 06:33:57,106 INFO Train. Iter 272400 : Commit. 0.21240 PPL. 411.08 Recons. 0.00123 +2024-03-17 06:34:19,020 INFO Train. Iter 272600 : Commit. 0.20943 PPL. 409.68 Recons. 0.00121 +2024-03-17 06:34:40,385 INFO Train. Iter 272800 : Commit. 0.20248 PPL. 411.60 Recons. 0.00108 +2024-03-17 06:35:01,459 INFO Train. Iter 273000 : Commit. 0.20385 PPL. 410.78 Recons. 0.00112 +2024-03-17 06:35:22,874 INFO Train. Iter 273200 : Commit. 0.20562 PPL. 411.17 Recons. 0.00114 +2024-03-17 06:35:44,728 INFO Train. Iter 273400 : Commit. 0.20201 PPL. 410.56 Recons. 0.00111 +2024-03-17 06:36:05,092 INFO Train. Iter 273600 : Commit. 0.20290 PPL. 411.39 Recons. 0.00110 +2024-03-17 06:36:26,137 INFO Train. Iter 273800 : Commit. 0.20497 PPL. 411.85 Recons. 0.00111 +2024-03-17 06:36:47,078 INFO Train. Iter 274000 : Commit. 0.20955 PPL. 411.04 Recons. 0.00120 +2024-03-17 06:37:08,401 INFO Train. Iter 274200 : Commit. 0.20397 PPL. 410.62 Recons. 0.00112 +2024-03-17 06:37:30,244 INFO Train. Iter 274400 : Commit. 0.20049 PPL. 410.86 Recons. 0.00108 +2024-03-17 06:37:51,865 INFO Train. Iter 274600 : Commit. 0.20729 PPL. 411.53 Recons. 0.00116 +2024-03-17 06:38:13,057 INFO Train. Iter 274800 : Commit. 0.20809 PPL. 410.94 Recons. 0.00118 +2024-03-17 06:38:34,828 INFO Train. Iter 275000 : Commit. 0.20168 PPL. 411.19 Recons. 0.00108 +2024-03-17 06:38:56,266 INFO Train. Iter 275200 : Commit. 0.20689 PPL. 411.17 Recons. 0.00116 +2024-03-17 06:39:16,844 INFO Train. Iter 275400 : Commit. 0.20587 PPL. 410.01 Recons. 0.00116 +2024-03-17 06:39:38,846 INFO Train. Iter 275600 : Commit. 0.19595 PPL. 410.66 Recons. 0.00102 +2024-03-17 06:40:00,452 INFO Train. Iter 275800 : Commit. 0.19947 PPL. 410.56 Recons. 0.00107 +2024-03-17 06:40:22,253 INFO Train. Iter 276000 : Commit. 0.20252 PPL. 410.68 Recons. 0.00111 +2024-03-17 06:40:43,906 INFO Train. Iter 276200 : Commit. 0.20865 PPL. 411.79 Recons. 0.00116 +2024-03-17 06:41:04,916 INFO Train. Iter 276400 : Commit. 0.20138 PPL. 410.70 Recons. 0.00108 +2024-03-17 06:41:26,030 INFO Train. Iter 276600 : Commit. 0.20207 PPL. 410.83 Recons. 0.00109 +2024-03-17 06:41:47,886 INFO Train. Iter 276800 : Commit. 0.19839 PPL. 410.82 Recons. 0.00104 +2024-03-17 06:42:08,928 INFO Train. Iter 277000 : Commit. 0.20482 PPL. 410.99 Recons. 0.00113 +2024-03-17 06:42:29,992 INFO Train. Iter 277200 : Commit. 0.20314 PPL. 410.76 Recons. 0.00112 +2024-03-17 06:42:50,310 INFO Train. Iter 277400 : Commit. 0.20554 PPL. 410.84 Recons. 0.00114 +2024-03-17 06:43:11,486 INFO Train. Iter 277600 : Commit. 0.20206 PPL. 410.60 Recons. 0.00111 +2024-03-17 06:43:32,903 INFO Train. Iter 277800 : Commit. 0.20518 PPL. 410.35 Recons. 0.00115 +2024-03-17 06:43:54,286 INFO Train. Iter 278000 : Commit. 0.21031 PPL. 410.85 Recons. 0.00121 +2024-03-17 06:44:15,777 INFO Train. Iter 278200 : Commit. 0.20178 PPL. 410.73 Recons. 0.00110 +2024-03-17 06:44:37,486 INFO Train. Iter 278400 : Commit. 0.19831 PPL. 410.17 Recons. 0.00105 +2024-03-17 06:44:59,394 INFO Train. Iter 278600 : Commit. 0.20290 PPL. 410.86 Recons. 0.00110 +2024-03-17 06:45:21,066 INFO Train. Iter 278800 : Commit. 0.20727 PPL. 411.45 Recons. 0.00115 +2024-03-17 06:45:42,578 INFO Train. Iter 279000 : Commit. 0.20654 PPL. 411.28 Recons. 0.00115 +2024-03-17 06:46:03,703 INFO Train. Iter 279200 : Commit. 0.20705 PPL. 410.24 Recons. 0.00117 +2024-03-17 06:46:24,039 INFO Train. Iter 279400 : Commit. 0.20093 PPL. 410.42 Recons. 0.00110 +2024-03-17 06:46:45,765 INFO Train. Iter 279600 : Commit. 0.20507 PPL. 410.90 Recons. 0.00114 +2024-03-17 06:47:06,828 INFO Train. Iter 279800 : Commit. 0.20083 PPL. 410.10 Recons. 0.00109 +2024-03-17 06:47:28,331 INFO Train. Iter 280000 : Commit. 0.20731 PPL. 410.94 Recons. 0.00116 +2024-03-17 06:47:49,834 INFO Train. Iter 280200 : Commit. 0.20609 PPL. 411.35 Recons. 0.00114 +2024-03-17 06:48:10,846 INFO Train. Iter 280400 : Commit. 0.20204 PPL. 410.81 Recons. 0.00110 +2024-03-17 06:48:32,738 INFO Train. Iter 280600 : Commit. 0.20574 PPL. 411.48 Recons. 0.00113 +2024-03-17 06:48:54,429 INFO Train. Iter 280800 : Commit. 0.20130 PPL. 410.86 Recons. 0.00109 +2024-03-17 06:49:15,787 INFO Train. Iter 281000 : Commit. 0.20663 PPL. 410.85 Recons. 0.00114 +2024-03-17 06:49:35,974 INFO Train. Iter 281200 : Commit. 0.20357 PPL. 410.13 Recons. 0.00113 +2024-03-17 06:49:57,219 INFO Train. Iter 281400 : Commit. 0.20015 PPL. 410.30 Recons. 0.00108 +2024-03-17 06:50:18,413 INFO Train. Iter 281600 : Commit. 0.20450 PPL. 410.32 Recons. 0.00113 +2024-03-17 06:50:39,529 INFO Train. Iter 281800 : Commit. 0.19988 PPL. 410.41 Recons. 0.00107 +2024-03-17 06:51:00,638 INFO Train. Iter 282000 : Commit. 0.20492 PPL. 410.81 Recons. 0.00114 +2024-03-17 06:51:22,477 INFO Train. Iter 282200 : Commit. 0.20405 PPL. 410.40 Recons. 0.00113 +2024-03-17 06:51:43,966 INFO Train. Iter 282400 : Commit. 0.20330 PPL. 410.27 Recons. 0.00112 +2024-03-17 06:52:05,231 INFO Train. Iter 282600 : Commit. 0.20724 PPL. 410.30 Recons. 0.00119 +2024-03-17 06:52:26,782 INFO Train. Iter 282800 : Commit. 0.20216 PPL. 410.33 Recons. 0.00111 +2024-03-17 06:52:48,241 INFO Train. Iter 283000 : Commit. 0.20453 PPL. 410.68 Recons. 0.00113 +2024-03-17 06:53:08,768 INFO Train. Iter 283200 : Commit. 0.19893 PPL. 409.56 Recons. 0.00107 +2024-03-17 06:53:30,665 INFO Train. Iter 283400 : Commit. 0.20223 PPL. 410.52 Recons. 0.00111 +2024-03-17 06:53:52,252 INFO Train. Iter 283600 : Commit. 0.20380 PPL. 410.70 Recons. 0.00112 +2024-03-17 06:54:13,362 INFO Train. Iter 283800 : Commit. 0.20687 PPL. 411.34 Recons. 0.00114 +2024-03-17 06:54:34,529 INFO Train. Iter 284000 : Commit. 0.20918 PPL. 411.28 Recons. 0.00118 +2024-03-17 06:54:56,275 INFO Train. Iter 284200 : Commit. 0.20044 PPL. 410.99 Recons. 0.00107 +2024-03-17 06:55:17,525 INFO Train. Iter 284400 : Commit. 0.20106 PPL. 410.22 Recons. 0.00109 +2024-03-17 06:55:38,695 INFO Train. Iter 284600 : Commit. 0.20485 PPL. 410.74 Recons. 0.00114 +2024-03-17 06:56:00,469 INFO Train. Iter 284800 : Commit. 0.20482 PPL. 410.70 Recons. 0.00113 +2024-03-17 06:56:22,085 INFO Train. Iter 285000 : Commit. 0.20543 PPL. 409.94 Recons. 0.00116 +2024-03-17 06:56:42,723 INFO Train. Iter 285200 : Commit. 0.20310 PPL. 410.46 Recons. 0.00111 +2024-03-17 06:57:04,430 INFO Train. Iter 285400 : Commit. 0.20151 PPL. 410.68 Recons. 0.00109 +2024-03-17 06:57:25,711 INFO Train. Iter 285600 : Commit. 0.20401 PPL. 410.17 Recons. 0.00113 +2024-03-17 06:57:46,836 INFO Train. Iter 285800 : Commit. 0.19996 PPL. 410.30 Recons. 0.00107 +2024-03-17 06:58:08,801 INFO Train. Iter 286000 : Commit. 0.20984 PPL. 410.55 Recons. 0.00121 +2024-03-17 06:58:30,366 INFO Train. Iter 286200 : Commit. 0.19783 PPL. 409.69 Recons. 0.00106 +2024-03-17 06:58:51,881 INFO Train. Iter 286400 : Commit. 0.20223 PPL. 410.61 Recons. 0.00111 +2024-03-17 06:59:13,460 INFO Train. Iter 286600 : Commit. 0.20324 PPL. 410.58 Recons. 0.00111 +2024-03-17 06:59:35,185 INFO Train. Iter 286800 : Commit. 0.20722 PPL. 410.84 Recons. 0.00116 +2024-03-17 06:59:55,406 INFO Train. Iter 287000 : Commit. 0.20080 PPL. 410.64 Recons. 0.00109 +2024-03-17 07:00:16,520 INFO Train. Iter 287200 : Commit. 0.20588 PPL. 410.41 Recons. 0.00116 +2024-03-17 07:00:38,384 INFO Train. Iter 287400 : Commit. 0.20107 PPL. 410.16 Recons. 0.00110 +2024-03-17 07:00:59,607 INFO Train. Iter 287600 : Commit. 0.20689 PPL. 410.59 Recons. 0.00117 +2024-03-17 07:01:20,785 INFO Train. Iter 287800 : Commit. 0.20443 PPL. 410.74 Recons. 0.00113 +2024-03-17 07:01:41,648 INFO Train. Iter 288000 : Commit. 0.20435 PPL. 409.78 Recons. 0.00115 +2024-03-17 07:02:02,984 INFO Train. Iter 288200 : Commit. 0.20562 PPL. 410.78 Recons. 0.00115 +2024-03-17 07:02:24,288 INFO Train. Iter 288400 : Commit. 0.20106 PPL. 410.67 Recons. 0.00109 +2024-03-17 07:02:45,632 INFO Train. Iter 288600 : Commit. 0.20676 PPL. 409.76 Recons. 0.00119 +2024-03-17 07:03:06,895 INFO Train. Iter 288800 : Commit. 0.20324 PPL. 410.25 Recons. 0.00112 +2024-03-17 07:03:27,236 INFO Train. Iter 289000 : Commit. 0.20428 PPL. 410.43 Recons. 0.00114 +2024-03-17 07:03:49,177 INFO Train. Iter 289200 : Commit. 0.20101 PPL. 411.26 Recons. 0.00107 +2024-03-17 07:04:10,728 INFO Train. Iter 289400 : Commit. 0.20313 PPL. 410.54 Recons. 0.00111 +2024-03-17 07:04:32,208 INFO Train. Iter 289600 : Commit. 0.20755 PPL. 410.36 Recons. 0.00117 +2024-03-17 07:04:53,751 INFO Train. Iter 289800 : Commit. 0.20369 PPL. 409.43 Recons. 0.00115 +2024-03-17 07:05:15,408 INFO Train. Iter 290000 : Commit. 0.20050 PPL. 409.65 Recons. 0.00110 +2024-03-17 07:05:36,716 INFO Train. Iter 290200 : Commit. 0.20118 PPL. 410.72 Recons. 0.00109 +2024-03-17 07:05:58,448 INFO Train. Iter 290400 : Commit. 0.20040 PPL. 411.08 Recons. 0.00107 +2024-03-17 07:06:20,291 INFO Train. Iter 290600 : Commit. 0.20238 PPL. 410.30 Recons. 0.00112 +2024-03-17 07:06:42,100 INFO Train. Iter 290800 : Commit. 0.20311 PPL. 410.16 Recons. 0.00112 +2024-03-17 07:07:03,013 INFO Train. Iter 291000 : Commit. 0.19749 PPL. 409.95 Recons. 0.00104 +2024-03-17 07:07:25,203 INFO Train. Iter 291200 : Commit. 0.20223 PPL. 409.63 Recons. 0.00112 +2024-03-17 07:07:47,424 INFO Train. Iter 291400 : Commit. 0.20226 PPL. 410.30 Recons. 0.00110 +2024-03-17 07:08:09,350 INFO Train. Iter 291600 : Commit. 0.20442 PPL. 409.81 Recons. 0.00114 +2024-03-17 07:08:31,091 INFO Train. Iter 291800 : Commit. 0.19755 PPL. 410.12 Recons. 0.00105 +2024-03-17 07:08:52,588 INFO Train. Iter 292000 : Commit. 0.20342 PPL. 410.40 Recons. 0.00112 +2024-03-17 07:09:14,097 INFO Train. Iter 292200 : Commit. 0.20309 PPL. 409.52 Recons. 0.00113 +2024-03-17 07:09:35,296 INFO Train. Iter 292400 : Commit. 0.20758 PPL. 410.09 Recons. 0.00118 +2024-03-17 07:09:56,591 INFO Train. Iter 292600 : Commit. 0.20112 PPL. 411.23 Recons. 0.00107 +2024-03-17 07:10:16,495 INFO Train. Iter 292800 : Commit. 0.20105 PPL. 410.60 Recons. 0.00108 +2024-03-17 07:10:37,857 INFO Train. Iter 293000 : Commit. 0.19810 PPL. 410.38 Recons. 0.00104 +2024-03-17 07:10:59,000 INFO Train. Iter 293200 : Commit. 0.20724 PPL. 410.01 Recons. 0.00119 +2024-03-17 07:11:20,135 INFO Train. Iter 293400 : Commit. 0.20385 PPL. 410.40 Recons. 0.00113 +2024-03-17 07:11:41,533 INFO Train. Iter 293600 : Commit. 0.20655 PPL. 410.55 Recons. 0.00115 +2024-03-17 07:12:03,114 INFO Train. Iter 293800 : Commit. 0.19964 PPL. 410.03 Recons. 0.00107 +2024-03-17 07:12:24,765 INFO Train. Iter 294000 : Commit. 0.19568 PPL. 409.43 Recons. 0.00104 +2024-03-17 07:12:46,548 INFO Train. Iter 294200 : Commit. 0.20012 PPL. 409.95 Recons. 0.00108 +2024-03-17 07:13:08,364 INFO Train. Iter 294400 : Commit. 0.19911 PPL. 408.68 Recons. 0.00110 +2024-03-17 07:13:29,612 INFO Train. Iter 294600 : Commit. 0.20301 PPL. 410.18 Recons. 0.00112 +2024-03-17 07:13:49,926 INFO Train. Iter 294800 : Commit. 0.20284 PPL. 410.02 Recons. 0.00112 +2024-03-17 07:14:12,127 INFO Train. Iter 295000 : Commit. 0.20295 PPL. 410.28 Recons. 0.00112 +2024-03-17 07:14:33,383 INFO Train. Iter 295200 : Commit. 0.20469 PPL. 410.83 Recons. 0.00112 +2024-03-17 07:14:55,006 INFO Train. Iter 295400 : Commit. 0.20495 PPL. 410.42 Recons. 0.00113 +2024-03-17 07:15:16,492 INFO Train. Iter 295600 : Commit. 0.20554 PPL. 411.38 Recons. 0.00113 +2024-03-17 07:15:38,137 INFO Train. Iter 295800 : Commit. 0.20547 PPL. 410.15 Recons. 0.00115 +2024-03-17 07:15:59,739 INFO Train. Iter 296000 : Commit. 0.20207 PPL. 409.88 Recons. 0.00112 +2024-03-17 07:16:20,924 INFO Train. Iter 296200 : Commit. 0.19862 PPL. 410.20 Recons. 0.00105 +2024-03-17 07:16:42,530 INFO Train. Iter 296400 : Commit. 0.20170 PPL. 409.90 Recons. 0.00110 +2024-03-17 07:17:03,616 INFO Train. Iter 296600 : Commit. 0.20365 PPL. 410.74 Recons. 0.00111 +2024-03-17 07:17:23,807 INFO Train. Iter 296800 : Commit. 0.19685 PPL. 409.87 Recons. 0.00104 +2024-03-17 07:17:44,922 INFO Train. Iter 297000 : Commit. 0.20363 PPL. 410.79 Recons. 0.00112 +2024-03-17 07:18:06,437 INFO Train. Iter 297200 : Commit. 0.20450 PPL. 410.89 Recons. 0.00113 +2024-03-17 07:18:27,802 INFO Train. Iter 297400 : Commit. 0.19695 PPL. 409.76 Recons. 0.00104 +2024-03-17 07:18:49,520 INFO Train. Iter 297600 : Commit. 0.19958 PPL. 410.13 Recons. 0.00107 +2024-03-17 07:19:11,078 INFO Train. Iter 297800 : Commit. 0.20297 PPL. 410.83 Recons. 0.00110 +2024-03-17 07:19:32,676 INFO Train. Iter 298000 : Commit. 0.20161 PPL. 409.53 Recons. 0.00111 +2024-03-17 07:19:54,016 INFO Train. Iter 298200 : Commit. 0.20415 PPL. 409.79 Recons. 0.00113 +2024-03-17 07:20:15,440 INFO Train. Iter 298400 : Commit. 0.20218 PPL. 410.32 Recons. 0.00110 +2024-03-17 07:20:35,879 INFO Train. Iter 298600 : Commit. 0.20141 PPL. 410.65 Recons. 0.00109 +2024-03-17 07:20:57,178 INFO Train. Iter 298800 : Commit. 0.20615 PPL. 409.61 Recons. 0.00117 +2024-03-17 07:21:18,582 INFO Train. Iter 299000 : Commit. 0.20157 PPL. 409.98 Recons. 0.00110 +2024-03-17 07:21:39,697 INFO Train. Iter 299200 : Commit. 0.20345 PPL. 410.80 Recons. 0.00111 +2024-03-17 07:21:59,780 INFO Train. Iter 299400 : Commit. 0.20302 PPL. 410.40 Recons. 0.00111 +2024-03-17 07:22:20,474 INFO Train. Iter 299600 : Commit. 0.19896 PPL. 409.81 Recons. 0.00106 +2024-03-17 07:22:40,644 INFO Train. Iter 299800 : Commit. 0.20008 PPL. 409.51 Recons. 0.00109 +2024-03-17 07:23:00,782 INFO Train. Iter 300000 : Commit. 0.20063 PPL. 410.12 Recons. 0.00108 diff --git a/ckpt/beatx2_rvqvae/RVQVAE_lower_trans/net_300000.pth b/ckpt/beatx2_rvqvae/RVQVAE_lower_trans/net_300000.pth new file mode 100644 index 0000000000000000000000000000000000000000..ec0bb0f2ee793db7e17a78775e0e9d52ed62ec93 --- /dev/null +++ b/ckpt/beatx2_rvqvae/RVQVAE_lower_trans/net_300000.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cb81af9ebd6c34b473db39e4c343e76fb3b30e4dbbab60d56460544f9cea7f6f +size 81536811 diff --git a/ckpt/beatx2_rvqvae/RVQVAE_lower_trans/run.log b/ckpt/beatx2_rvqvae/RVQVAE_lower_trans/run.log new file mode 100644 index 0000000000000000000000000000000000000000..4debdaaadb16609e6d12371a2746a0ab33241db4 --- /dev/null +++ b/ckpt/beatx2_rvqvae/RVQVAE_lower_trans/run.log @@ -0,0 +1,1546 @@ +2024-03-29 22:47:31,675 INFO { + "batch_size": 256, + "beta": 1.0, + "body_part": "lower_trans", + "code_dim": 512, + "commit": 0.02, + "dataname": "kit", + "depth": 3, + "dilation_growth_rate": 3, + "down_t": 2, + "eval_iter": 1000, + "exp_name": "RVQVAE", + "gamma": 0.05, + "loss_vel": 0.5, + "lr": 0.0002, + "lr_scheduler": [ + 200000 + ], + "mu": 0.99, + "nb_code": 512, + "nb_vis": 20, + "out_dir": "output_beatx2/RVQVAE_lower_trans", + "output_emb_width": 512, + "print_iter": 200, + "quantizer": "ema_reset", + "recons_loss": "l1_smooth", + "results_dir": "visual_results/", + "resume_gpt": null, + "resume_pth": null, + "seed": 123, + "stride_t": 2, + "total_iter": 300000, + "vis_gt": false, + "visual_name": "baseline", + "vq_act": "relu", + "vq_norm": null, + "warm_up_iter": 1000, + "weight_decay": 0.0, + "width": 512, + "window_size": 64 +} +2024-03-29 22:47:31,683 INFO Training on kit, motions are with 21 joints +2024-03-29 22:47:58,506 INFO Warmup. Iter 200 : lr 0.00004 Commit. 0.04082 PPL. 285.49 Recons. 0.29989 +2024-03-29 22:48:23,633 INFO Warmup. Iter 400 : lr 0.00008 Commit. 0.36833 PPL. 156.41 Recons. 0.12814 +2024-03-29 22:48:48,528 INFO Warmup. Iter 600 : lr 0.00012 Commit. 0.62618 PPL. 338.75 Recons. 0.07369 +2024-03-29 22:49:13,260 INFO Warmup. Iter 800 : lr 0.00016 Commit. 0.90612 PPL. 412.95 Recons. 0.04982 +2024-03-29 22:50:02,392 INFO Train. Iter 200 : Commit. 1.06469 PPL. 409.46 Recons. 0.03120 +2024-03-29 22:50:27,034 INFO Train. Iter 400 : Commit. 1.04439 PPL. 406.06 Recons. 0.02697 +2024-03-29 22:50:51,635 INFO Train. Iter 600 : Commit. 1.07187 PPL. 405.73 Recons. 0.02511 +2024-03-29 22:51:16,411 INFO Train. Iter 800 : Commit. 1.08602 PPL. 406.67 Recons. 0.02193 +2024-03-29 22:51:40,120 INFO Train. Iter 1000 : Commit. 1.09110 PPL. 407.39 Recons. 0.02053 +2024-03-29 22:52:05,075 INFO Train. Iter 1200 : Commit. 1.09175 PPL. 409.76 Recons. 0.01823 +2024-03-29 22:52:29,844 INFO Train. Iter 1400 : Commit. 1.10665 PPL. 410.57 Recons. 0.01781 +2024-03-29 22:52:54,279 INFO Train. Iter 1600 : Commit. 1.09611 PPL. 411.29 Recons. 0.01823 +2024-03-29 22:53:18,808 INFO Train. Iter 1800 : Commit. 1.06101 PPL. 414.17 Recons. 0.01530 +2024-03-29 22:53:43,190 INFO Train. Iter 2000 : Commit. 1.07372 PPL. 413.28 Recons. 0.01518 +2024-03-29 22:54:08,219 INFO Train. Iter 2200 : Commit. 1.07348 PPL. 412.27 Recons. 0.01605 +2024-03-29 22:54:32,531 INFO Train. Iter 2400 : Commit. 1.08075 PPL. 413.45 Recons. 0.01436 +2024-03-29 22:54:56,907 INFO Train. Iter 2600 : Commit. 0.99605 PPL. 414.52 Recons. 0.01251 +2024-03-29 22:55:21,197 INFO Train. Iter 2800 : Commit. 1.01090 PPL. 413.68 Recons. 0.01303 +2024-03-29 22:55:46,179 INFO Train. Iter 3000 : Commit. 0.99491 PPL. 414.69 Recons. 0.01279 +2024-03-29 22:56:10,852 INFO Train. Iter 3200 : Commit. 1.00036 PPL. 414.07 Recons. 0.01365 +2024-03-29 22:56:35,260 INFO Train. Iter 3400 : Commit. 0.97994 PPL. 416.05 Recons. 0.01179 +2024-03-29 22:57:00,040 INFO Train. Iter 3600 : Commit. 0.92994 PPL. 416.31 Recons. 0.01017 +2024-03-29 22:57:24,399 INFO Train. Iter 3800 : Commit. 0.95308 PPL. 413.78 Recons. 0.01186 +2024-03-29 22:57:49,034 INFO Train. Iter 4000 : Commit. 0.97696 PPL. 415.09 Recons. 0.01202 +2024-03-29 22:58:13,441 INFO Train. Iter 4200 : Commit. 0.92479 PPL. 415.73 Recons. 0.01119 +2024-03-29 22:58:37,931 INFO Train. Iter 4400 : Commit. 0.99361 PPL. 410.06 Recons. 0.01480 +2024-03-29 22:59:02,698 INFO Train. Iter 4600 : Commit. 0.92854 PPL. 416.88 Recons. 0.01080 +2024-03-29 22:59:26,459 INFO Train. Iter 4800 : Commit. 0.93824 PPL. 414.11 Recons. 0.01270 +2024-03-29 22:59:50,891 INFO Train. Iter 5000 : Commit. 0.91682 PPL. 417.95 Recons. 0.01052 +2024-03-29 23:00:15,443 INFO Train. Iter 5200 : Commit. 0.85092 PPL. 419.24 Recons. 0.00896 +2024-03-29 23:00:39,640 INFO Train. Iter 5400 : Commit. 0.91149 PPL. 416.13 Recons. 0.01202 +2024-03-29 23:01:04,311 INFO Train. Iter 5600 : Commit. 0.90237 PPL. 416.15 Recons. 0.01169 +2024-03-29 23:01:29,143 INFO Train. Iter 5800 : Commit. 0.88908 PPL. 417.57 Recons. 0.01077 +2024-03-29 23:01:53,495 INFO Train. Iter 6000 : Commit. 0.88396 PPL. 418.38 Recons. 0.01139 +2024-03-29 23:02:18,224 INFO Train. Iter 6200 : Commit. 0.89922 PPL. 418.96 Recons. 0.01136 +2024-03-29 23:02:43,066 INFO Train. Iter 6400 : Commit. 0.82657 PPL. 420.67 Recons. 0.00863 +2024-03-29 23:03:07,539 INFO Train. Iter 6600 : Commit. 0.81250 PPL. 416.82 Recons. 0.00967 +2024-03-29 23:03:31,244 INFO Train. Iter 6800 : Commit. 0.90549 PPL. 417.69 Recons. 0.01276 +2024-03-29 23:03:56,025 INFO Train. Iter 7000 : Commit. 0.85289 PPL. 419.43 Recons. 0.00961 +2024-03-29 23:04:20,408 INFO Train. Iter 7200 : Commit. 0.85369 PPL. 418.17 Recons. 0.01128 +2024-03-29 23:04:44,820 INFO Train. Iter 7400 : Commit. 0.82419 PPL. 420.29 Recons. 0.00956 +2024-03-29 23:05:09,437 INFO Train. Iter 7600 : Commit. 0.86604 PPL. 417.51 Recons. 0.01124 +2024-03-29 23:05:34,155 INFO Train. Iter 7800 : Commit. 0.79042 PPL. 421.47 Recons. 0.00833 +2024-03-29 23:05:58,618 INFO Train. Iter 8000 : Commit. 0.80673 PPL. 419.15 Recons. 0.01041 +2024-03-29 23:06:23,227 INFO Train. Iter 8200 : Commit. 0.80963 PPL. 418.34 Recons. 0.01059 +2024-03-29 23:06:47,567 INFO Train. Iter 8400 : Commit. 0.78862 PPL. 419.41 Recons. 0.00965 +2024-03-29 23:07:11,199 INFO Train. Iter 8600 : Commit. 0.71610 PPL. 423.47 Recons. 0.00647 +2024-03-29 23:07:35,769 INFO Train. Iter 8800 : Commit. 0.73670 PPL. 420.77 Recons. 0.00866 +2024-03-29 23:08:00,898 INFO Train. Iter 9000 : Commit. 0.78324 PPL. 418.08 Recons. 0.01111 +2024-03-29 23:08:25,176 INFO Train. Iter 9200 : Commit. 0.71269 PPL. 423.93 Recons. 0.00695 +2024-03-29 23:08:49,763 INFO Train. Iter 9400 : Commit. 0.78936 PPL. 418.17 Recons. 0.01105 +2024-03-29 23:09:14,692 INFO Train. Iter 9600 : Commit. 0.75678 PPL. 421.41 Recons. 0.00884 +2024-03-29 23:09:39,462 INFO Train. Iter 9800 : Commit. 0.72795 PPL. 423.00 Recons. 0.00791 +2024-03-29 23:10:04,027 INFO Train. Iter 10000 : Commit. 0.74081 PPL. 420.81 Recons. 0.00907 +2024-03-29 23:10:28,851 INFO Train. Iter 10200 : Commit. 0.72565 PPL. 419.05 Recons. 0.00943 +2024-03-29 23:10:53,927 INFO Train. Iter 10400 : Commit. 0.71866 PPL. 422.76 Recons. 0.00848 +2024-03-29 23:11:17,385 INFO Train. Iter 10600 : Commit. 0.72663 PPL. 422.18 Recons. 0.00930 +2024-03-29 23:11:42,180 INFO Train. Iter 10800 : Commit. 0.72746 PPL. 422.35 Recons. 0.00831 +2024-03-29 23:12:07,185 INFO Train. Iter 11000 : Commit. 0.68393 PPL. 423.82 Recons. 0.00706 +2024-03-29 23:12:31,980 INFO Train. Iter 11200 : Commit. 0.69302 PPL. 422.03 Recons. 0.00798 +2024-03-29 23:12:56,980 INFO Train. Iter 11400 : Commit. 0.69014 PPL. 422.66 Recons. 0.00788 +2024-03-29 23:13:21,622 INFO Train. Iter 11600 : Commit. 0.73251 PPL. 421.09 Recons. 0.00997 +2024-03-29 23:13:46,159 INFO Train. Iter 11800 : Commit. 0.72590 PPL. 423.39 Recons. 0.00816 +2024-03-29 23:14:10,551 INFO Train. Iter 12000 : Commit. 0.70202 PPL. 419.41 Recons. 0.00922 +2024-03-29 23:14:34,799 INFO Train. Iter 12200 : Commit. 0.66515 PPL. 422.45 Recons. 0.00781 +2024-03-29 23:14:59,236 INFO Train. Iter 12400 : Commit. 0.66902 PPL. 424.48 Recons. 0.00705 +2024-03-29 23:15:22,904 INFO Train. Iter 12600 : Commit. 0.66212 PPL. 423.40 Recons. 0.00784 +2024-03-29 23:15:47,734 INFO Train. Iter 12800 : Commit. 0.71934 PPL. 418.23 Recons. 0.01078 +2024-03-29 23:16:13,117 INFO Train. Iter 13000 : Commit. 0.70131 PPL. 422.50 Recons. 0.00939 +2024-03-29 23:16:38,445 INFO Train. Iter 13200 : Commit. 0.66042 PPL. 422.53 Recons. 0.00728 +2024-03-29 23:17:03,164 INFO Train. Iter 13400 : Commit. 0.66834 PPL. 422.90 Recons. 0.00759 +2024-03-29 23:17:28,316 INFO Train. Iter 13600 : Commit. 0.60773 PPL. 424.98 Recons. 0.00651 +2024-03-29 23:17:53,980 INFO Train. Iter 13800 : Commit. 0.62101 PPL. 422.96 Recons. 0.00755 +2024-03-29 23:18:18,890 INFO Train. Iter 14000 : Commit. 0.64131 PPL. 423.91 Recons. 0.00720 +2024-03-29 23:18:43,221 INFO Train. Iter 14200 : Commit. 0.63621 PPL. 423.25 Recons. 0.00846 +2024-03-29 23:19:06,480 INFO Train. Iter 14400 : Commit. 0.63985 PPL. 424.37 Recons. 0.00748 +2024-03-29 23:19:30,814 INFO Train. Iter 14600 : Commit. 0.59655 PPL. 424.11 Recons. 0.00636 +2024-03-29 23:19:55,582 INFO Train. Iter 14800 : Commit. 0.60412 PPL. 426.06 Recons. 0.00619 +2024-03-29 23:20:20,703 INFO Train. Iter 15000 : Commit. 0.57795 PPL. 425.80 Recons. 0.00569 +2024-03-29 23:20:45,834 INFO Train. Iter 15200 : Commit. 0.61780 PPL. 421.77 Recons. 0.00868 +2024-03-29 23:21:10,482 INFO Train. Iter 15400 : Commit. 0.65797 PPL. 421.51 Recons. 0.00927 +2024-03-29 23:21:35,337 INFO Train. Iter 15600 : Commit. 0.60923 PPL. 424.84 Recons. 0.00661 +2024-03-29 23:22:00,414 INFO Train. Iter 15800 : Commit. 0.60590 PPL. 424.99 Recons. 0.00643 +2024-03-29 23:22:25,302 INFO Train. Iter 16000 : Commit. 0.61993 PPL. 422.21 Recons. 0.00802 +2024-03-29 23:22:50,536 INFO Train. Iter 16200 : Commit. 0.63137 PPL. 423.30 Recons. 0.00810 +2024-03-29 23:23:15,117 INFO Train. Iter 16400 : Commit. 0.61457 PPL. 425.03 Recons. 0.00648 +2024-03-29 23:23:41,276 INFO Train. Iter 16600 : Commit. 0.61666 PPL. 423.48 Recons. 0.00735 +2024-03-29 23:24:07,125 INFO Train. Iter 16800 : Commit. 0.61253 PPL. 421.49 Recons. 0.00858 +2024-03-29 23:24:33,054 INFO Train. Iter 17000 : Commit. 0.63571 PPL. 422.82 Recons. 0.00807 +2024-03-29 23:24:58,366 INFO Train. Iter 17200 : Commit. 0.63841 PPL. 424.68 Recons. 0.00779 +2024-03-29 23:25:23,882 INFO Train. Iter 17400 : Commit. 0.63450 PPL. 422.69 Recons. 0.00803 +2024-03-29 23:25:49,615 INFO Train. Iter 17600 : Commit. 0.59113 PPL. 424.18 Recons. 0.00641 +2024-03-29 23:26:15,462 INFO Train. Iter 17800 : Commit. 0.55783 PPL. 423.82 Recons. 0.00609 +2024-03-29 23:26:40,988 INFO Train. Iter 18000 : Commit. 0.56138 PPL. 422.37 Recons. 0.00753 +2024-03-29 23:27:06,619 INFO Train. Iter 18200 : Commit. 0.58411 PPL. 423.04 Recons. 0.00722 +2024-03-29 23:27:30,928 INFO Train. Iter 18400 : Commit. 0.55291 PPL. 426.74 Recons. 0.00549 +2024-03-29 23:27:56,099 INFO Train. Iter 18600 : Commit. 0.55536 PPL. 422.36 Recons. 0.00691 +2024-03-29 23:28:21,461 INFO Train. Iter 18800 : Commit. 0.54616 PPL. 425.99 Recons. 0.00586 +2024-03-29 23:28:47,103 INFO Train. Iter 19000 : Commit. 0.56301 PPL. 422.71 Recons. 0.00704 +2024-03-29 23:29:12,502 INFO Train. Iter 19200 : Commit. 0.53147 PPL. 425.94 Recons. 0.00547 +2024-03-29 23:29:38,013 INFO Train. Iter 19400 : Commit. 0.57069 PPL. 422.36 Recons. 0.00756 +2024-03-29 23:30:03,676 INFO Train. Iter 19600 : Commit. 0.53293 PPL. 424.88 Recons. 0.00572 +2024-03-29 23:30:29,861 INFO Train. Iter 19800 : Commit. 0.53680 PPL. 425.25 Recons. 0.00621 +2024-03-29 23:30:55,890 INFO Train. Iter 20000 : Commit. 0.58252 PPL. 422.89 Recons. 0.00762 +2024-03-29 23:31:20,165 INFO Train. Iter 20200 : Commit. 0.54292 PPL. 425.46 Recons. 0.00531 +2024-03-29 23:31:45,943 INFO Train. Iter 20400 : Commit. 0.53259 PPL. 424.43 Recons. 0.00612 +2024-03-29 23:32:11,794 INFO Train. Iter 20600 : Commit. 0.52886 PPL. 425.35 Recons. 0.00592 +2024-03-29 23:32:37,461 INFO Train. Iter 20800 : Commit. 0.55037 PPL. 421.80 Recons. 0.00757 +2024-03-29 23:33:03,740 INFO Train. Iter 21000 : Commit. 0.53907 PPL. 423.43 Recons. 0.00685 +2024-03-29 23:33:29,440 INFO Train. Iter 21200 : Commit. 0.48212 PPL. 427.65 Recons. 0.00413 +2024-03-29 23:33:54,899 INFO Train. Iter 21400 : Commit. 0.56824 PPL. 422.36 Recons. 0.00802 +2024-03-29 23:34:21,132 INFO Train. Iter 21600 : Commit. 0.55271 PPL. 421.54 Recons. 0.00723 +2024-03-29 23:34:46,388 INFO Train. Iter 21800 : Commit. 0.54383 PPL. 425.17 Recons. 0.00592 +2024-03-29 23:35:11,734 INFO Train. Iter 22000 : Commit. 0.53666 PPL. 425.19 Recons. 0.00635 +2024-03-29 23:35:36,200 INFO Train. Iter 22200 : Commit. 0.48863 PPL. 427.13 Recons. 0.00447 +2024-03-29 23:36:01,633 INFO Train. Iter 22400 : Commit. 0.50931 PPL. 424.33 Recons. 0.00616 +2024-03-29 23:36:27,088 INFO Train. Iter 22600 : Commit. 0.54442 PPL. 422.97 Recons. 0.00688 +2024-03-29 23:36:52,252 INFO Train. Iter 22800 : Commit. 0.54199 PPL. 425.57 Recons. 0.00578 +2024-03-29 23:37:17,948 INFO Train. Iter 23000 : Commit. 0.48977 PPL. 426.99 Recons. 0.00446 +2024-03-29 23:37:43,532 INFO Train. Iter 23200 : Commit. 0.54320 PPL. 420.80 Recons. 0.00780 +2024-03-29 23:38:08,992 INFO Train. Iter 23400 : Commit. 0.51906 PPL. 425.41 Recons. 0.00566 +2024-03-29 23:38:34,882 INFO Train. Iter 23600 : Commit. 0.54174 PPL. 423.14 Recons. 0.00704 +2024-03-29 23:39:00,724 INFO Train. Iter 23800 : Commit. 0.51433 PPL. 424.26 Recons. 0.00587 +2024-03-29 23:39:26,352 INFO Train. Iter 24000 : Commit. 0.57004 PPL. 421.47 Recons. 0.00782 +2024-03-29 23:39:51,273 INFO Train. Iter 24200 : Commit. 0.49890 PPL. 426.98 Recons. 0.00438 +2024-03-29 23:40:17,001 INFO Train. Iter 24400 : Commit. 0.48699 PPL. 425.68 Recons. 0.00496 +2024-03-29 23:40:42,680 INFO Train. Iter 24600 : Commit. 0.52255 PPL. 424.52 Recons. 0.00613 +2024-03-29 23:41:08,461 INFO Train. Iter 24800 : Commit. 0.50816 PPL. 423.01 Recons. 0.00608 +2024-03-29 23:41:33,943 INFO Train. Iter 25000 : Commit. 0.51516 PPL. 424.43 Recons. 0.00631 +2024-03-29 23:41:59,369 INFO Train. Iter 25200 : Commit. 0.48902 PPL. 426.38 Recons. 0.00473 +2024-03-29 23:42:25,504 INFO Train. Iter 25400 : Commit. 0.51508 PPL. 421.07 Recons. 0.00715 +2024-03-29 23:42:51,028 INFO Train. Iter 25600 : Commit. 0.51313 PPL. 422.37 Recons. 0.00698 +2024-03-29 23:43:17,448 INFO Train. Iter 25800 : Commit. 0.50735 PPL. 425.70 Recons. 0.00548 +2024-03-29 23:43:42,556 INFO Train. Iter 26000 : Commit. 0.46800 PPL. 426.88 Recons. 0.00447 +2024-03-29 23:44:08,608 INFO Train. Iter 26200 : Commit. 0.47927 PPL. 423.70 Recons. 0.00602 +2024-03-29 23:44:34,243 INFO Train. Iter 26400 : Commit. 0.51870 PPL. 423.02 Recons. 0.00659 +2024-03-29 23:44:59,645 INFO Train. Iter 26600 : Commit. 0.49709 PPL. 423.74 Recons. 0.00584 +2024-03-29 23:45:25,360 INFO Train. Iter 26800 : Commit. 0.52206 PPL. 422.99 Recons. 0.00687 +2024-03-29 23:45:51,643 INFO Train. Iter 27000 : Commit. 0.46696 PPL. 425.60 Recons. 0.00481 +2024-03-29 23:46:17,097 INFO Train. Iter 27200 : Commit. 0.47282 PPL. 425.61 Recons. 0.00500 +2024-03-29 23:46:42,738 INFO Train. Iter 27400 : Commit. 0.46222 PPL. 426.15 Recons. 0.00483 +2024-03-29 23:47:08,460 INFO Train. Iter 27600 : Commit. 0.46094 PPL. 424.64 Recons. 0.00512 +2024-03-29 23:47:34,490 INFO Train. Iter 27800 : Commit. 0.46344 PPL. 424.23 Recons. 0.00530 +2024-03-29 23:47:59,192 INFO Train. Iter 28000 : Commit. 0.46684 PPL. 424.04 Recons. 0.00503 +2024-03-29 23:48:24,895 INFO Train. Iter 28200 : Commit. 0.48939 PPL. 421.86 Recons. 0.00644 +2024-03-29 23:48:50,552 INFO Train. Iter 28400 : Commit. 0.50610 PPL. 421.81 Recons. 0.00724 +2024-03-29 23:49:16,206 INFO Train. Iter 28600 : Commit. 0.48683 PPL. 423.31 Recons. 0.00561 +2024-03-29 23:49:41,652 INFO Train. Iter 28800 : Commit. 0.46463 PPL. 423.90 Recons. 0.00534 +2024-03-29 23:50:08,042 INFO Train. Iter 29000 : Commit. 0.47575 PPL. 425.46 Recons. 0.00493 +2024-03-29 23:50:33,840 INFO Train. Iter 29200 : Commit. 0.47029 PPL. 424.55 Recons. 0.00517 +2024-03-29 23:50:59,072 INFO Train. Iter 29400 : Commit. 0.46223 PPL. 424.06 Recons. 0.00522 +2024-03-29 23:51:25,338 INFO Train. Iter 29600 : Commit. 0.45541 PPL. 424.37 Recons. 0.00522 +2024-03-29 23:51:50,906 INFO Train. Iter 29800 : Commit. 0.46388 PPL. 423.02 Recons. 0.00551 +2024-03-29 23:52:15,685 INFO Train. Iter 30000 : Commit. 0.47278 PPL. 422.50 Recons. 0.00591 +2024-03-29 23:52:42,219 INFO Train. Iter 30200 : Commit. 0.44306 PPL. 426.18 Recons. 0.00432 +2024-03-29 23:53:07,754 INFO Train. Iter 30400 : Commit. 0.44759 PPL. 425.22 Recons. 0.00476 +2024-03-29 23:53:33,323 INFO Train. Iter 30600 : Commit. 0.44657 PPL. 423.21 Recons. 0.00550 +2024-03-29 23:53:59,116 INFO Train. Iter 30800 : Commit. 0.45355 PPL. 424.27 Recons. 0.00509 +2024-03-29 23:54:24,068 INFO Train. Iter 31000 : Commit. 0.44905 PPL. 423.56 Recons. 0.00528 +2024-03-29 23:54:49,167 INFO Train. Iter 31200 : Commit. 0.45497 PPL. 424.33 Recons. 0.00524 +2024-03-29 23:55:14,251 INFO Train. Iter 31400 : Commit. 0.46527 PPL. 423.17 Recons. 0.00566 +2024-03-29 23:55:38,892 INFO Train. Iter 31600 : Commit. 0.47728 PPL. 423.48 Recons. 0.00608 +2024-03-29 23:56:02,333 INFO Train. Iter 31800 : Commit. 0.44884 PPL. 421.41 Recons. 0.00573 +2024-03-29 23:56:27,112 INFO Train. Iter 32000 : Commit. 0.46691 PPL. 425.21 Recons. 0.00511 +2024-03-29 23:56:52,470 INFO Train. Iter 32200 : Commit. 0.44603 PPL. 424.48 Recons. 0.00495 +2024-03-29 23:57:17,459 INFO Train. Iter 32400 : Commit. 0.44588 PPL. 424.94 Recons. 0.00531 +2024-03-29 23:57:42,512 INFO Train. Iter 32600 : Commit. 0.46805 PPL. 421.59 Recons. 0.00597 +2024-03-29 23:58:07,826 INFO Train. Iter 32800 : Commit. 0.43851 PPL. 425.56 Recons. 0.00454 +2024-03-29 23:58:32,980 INFO Train. Iter 33000 : Commit. 0.43294 PPL. 421.88 Recons. 0.00506 +2024-03-29 23:58:57,773 INFO Train. Iter 33200 : Commit. 0.42676 PPL. 424.96 Recons. 0.00450 +2024-03-29 23:59:22,899 INFO Train. Iter 33400 : Commit. 0.42759 PPL. 425.64 Recons. 0.00447 +2024-03-29 23:59:48,108 INFO Train. Iter 33600 : Commit. 0.44183 PPL. 422.18 Recons. 0.00577 +2024-03-30 00:00:11,684 INFO Train. Iter 33800 : Commit. 0.42129 PPL. 424.46 Recons. 0.00426 +2024-03-30 00:00:36,958 INFO Train. Iter 34000 : Commit. 0.41462 PPL. 424.61 Recons. 0.00452 +2024-03-30 00:01:03,287 INFO Train. Iter 34200 : Commit. 0.43188 PPL. 423.98 Recons. 0.00483 +2024-03-30 00:01:29,113 INFO Train. Iter 34400 : Commit. 0.41869 PPL. 424.97 Recons. 0.00436 +2024-03-30 00:01:54,815 INFO Train. Iter 34600 : Commit. 0.44481 PPL. 421.46 Recons. 0.00564 +2024-03-30 00:02:20,384 INFO Train. Iter 34800 : Commit. 0.43112 PPL. 424.88 Recons. 0.00468 +2024-03-30 00:02:45,935 INFO Train. Iter 35000 : Commit. 0.42604 PPL. 423.08 Recons. 0.00520 +2024-03-30 00:03:11,403 INFO Train. Iter 35200 : Commit. 0.41410 PPL. 424.56 Recons. 0.00449 +2024-03-30 00:03:36,851 INFO Train. Iter 35400 : Commit. 0.41066 PPL. 426.07 Recons. 0.00397 +2024-03-30 00:04:02,489 INFO Train. Iter 35600 : Commit. 0.43206 PPL. 423.26 Recons. 0.00502 +2024-03-30 00:04:27,021 INFO Train. Iter 35800 : Commit. 0.42596 PPL. 425.05 Recons. 0.00441 +2024-03-30 00:04:53,152 INFO Train. Iter 36000 : Commit. 0.42990 PPL. 421.74 Recons. 0.00566 +2024-03-30 00:05:18,927 INFO Train. Iter 36200 : Commit. 0.43645 PPL. 424.15 Recons. 0.00484 +2024-03-30 00:05:44,392 INFO Train. Iter 36400 : Commit. 0.40806 PPL. 424.26 Recons. 0.00447 +2024-03-30 00:06:09,712 INFO Train. Iter 36600 : Commit. 0.43762 PPL. 423.27 Recons. 0.00520 +2024-03-30 00:06:35,930 INFO Train. Iter 36800 : Commit. 0.43143 PPL. 422.66 Recons. 0.00526 +2024-03-30 00:07:01,717 INFO Train. Iter 37000 : Commit. 0.41605 PPL. 422.17 Recons. 0.00505 +2024-03-30 00:07:27,805 INFO Train. Iter 37200 : Commit. 0.40698 PPL. 425.61 Recons. 0.00403 +2024-03-30 00:07:53,810 INFO Train. Iter 37400 : Commit. 0.41564 PPL. 423.37 Recons. 0.00486 +2024-03-30 00:08:18,688 INFO Train. Iter 37600 : Commit. 0.40140 PPL. 425.76 Recons. 0.00380 +2024-03-30 00:08:44,111 INFO Train. Iter 37800 : Commit. 0.40410 PPL. 424.35 Recons. 0.00458 +2024-03-30 00:09:10,149 INFO Train. Iter 38000 : Commit. 0.42541 PPL. 424.10 Recons. 0.00469 +2024-03-30 00:09:36,002 INFO Train. Iter 38200 : Commit. 0.41481 PPL. 422.32 Recons. 0.00507 +2024-03-30 00:10:02,148 INFO Train. Iter 38400 : Commit. 0.42924 PPL. 423.19 Recons. 0.00501 +2024-03-30 00:10:28,244 INFO Train. Iter 38600 : Commit. 0.42214 PPL. 421.58 Recons. 0.00533 +2024-03-30 00:10:53,660 INFO Train. Iter 38800 : Commit. 0.40530 PPL. 425.66 Recons. 0.00379 +2024-03-30 00:11:19,113 INFO Train. Iter 39000 : Commit. 0.39903 PPL. 423.67 Recons. 0.00468 +2024-03-30 00:11:44,072 INFO Train. Iter 39200 : Commit. 0.38638 PPL. 424.26 Recons. 0.00403 +2024-03-30 00:12:10,252 INFO Train. Iter 39400 : Commit. 0.41674 PPL. 424.25 Recons. 0.00461 +2024-03-30 00:12:35,137 INFO Train. Iter 39600 : Commit. 0.41105 PPL. 422.05 Recons. 0.00502 +2024-03-30 00:13:00,782 INFO Train. Iter 39800 : Commit. 0.40349 PPL. 423.96 Recons. 0.00438 +2024-03-30 00:13:26,283 INFO Train. Iter 40000 : Commit. 0.40199 PPL. 421.22 Recons. 0.00518 +2024-03-30 00:13:52,227 INFO Train. Iter 40200 : Commit. 0.40817 PPL. 423.89 Recons. 0.00432 +2024-03-30 00:14:18,221 INFO Train. Iter 40400 : Commit. 0.41525 PPL. 421.52 Recons. 0.00530 +2024-03-30 00:14:43,750 INFO Train. Iter 40600 : Commit. 0.41821 PPL. 421.26 Recons. 0.00497 +2024-03-30 00:15:08,933 INFO Train. Iter 40800 : Commit. 0.42311 PPL. 421.35 Recons. 0.00564 +2024-03-30 00:15:34,718 INFO Train. Iter 41000 : Commit. 0.40645 PPL. 425.37 Recons. 0.00375 +2024-03-30 00:16:00,070 INFO Train. Iter 41200 : Commit. 0.41963 PPL. 420.85 Recons. 0.00538 +2024-03-30 00:16:25,861 INFO Train. Iter 41400 : Commit. 0.39435 PPL. 423.90 Recons. 0.00402 +2024-03-30 00:16:50,408 INFO Train. Iter 41600 : Commit. 0.39051 PPL. 423.15 Recons. 0.00425 +2024-03-30 00:17:16,091 INFO Train. Iter 41800 : Commit. 0.41476 PPL. 422.76 Recons. 0.00483 +2024-03-30 00:17:42,283 INFO Train. Iter 42000 : Commit. 0.39821 PPL. 423.01 Recons. 0.00441 +2024-03-30 00:18:07,836 INFO Train. Iter 42200 : Commit. 0.39030 PPL. 422.24 Recons. 0.00447 +2024-03-30 00:18:33,579 INFO Train. Iter 42400 : Commit. 0.39779 PPL. 424.29 Recons. 0.00411 +2024-03-30 00:18:59,205 INFO Train. Iter 42600 : Commit. 0.38392 PPL. 423.56 Recons. 0.00419 +2024-03-30 00:19:25,468 INFO Train. Iter 42800 : Commit. 0.39320 PPL. 421.93 Recons. 0.00472 +2024-03-30 00:19:51,127 INFO Train. Iter 43000 : Commit. 0.39164 PPL. 424.07 Recons. 0.00396 +2024-03-30 00:20:16,952 INFO Train. Iter 43200 : Commit. 0.37292 PPL. 424.07 Recons. 0.00379 +2024-03-30 00:20:41,895 INFO Train. Iter 43400 : Commit. 0.38549 PPL. 420.21 Recons. 0.00478 +2024-03-30 00:21:07,891 INFO Train. Iter 43600 : Commit. 0.37983 PPL. 425.34 Recons. 0.00355 +2024-03-30 00:21:33,394 INFO Train. Iter 43800 : Commit. 0.39200 PPL. 422.75 Recons. 0.00470 +2024-03-30 00:21:59,436 INFO Train. Iter 44000 : Commit. 0.37147 PPL. 421.65 Recons. 0.00426 +2024-03-30 00:22:25,158 INFO Train. Iter 44200 : Commit. 0.38167 PPL. 423.69 Recons. 0.00410 +2024-03-30 00:22:51,231 INFO Train. Iter 44400 : Commit. 0.37003 PPL. 422.80 Recons. 0.00393 +2024-03-30 00:23:16,808 INFO Train. Iter 44600 : Commit. 0.39098 PPL. 421.20 Recons. 0.00476 +2024-03-30 00:23:42,281 INFO Train. Iter 44800 : Commit. 0.37159 PPL. 423.77 Recons. 0.00393 +2024-03-30 00:24:07,818 INFO Train. Iter 45000 : Commit. 0.38230 PPL. 422.35 Recons. 0.00433 +2024-03-30 00:24:33,478 INFO Train. Iter 45200 : Commit. 0.39409 PPL. 422.12 Recons. 0.00471 +2024-03-30 00:24:57,528 INFO Train. Iter 45400 : Commit. 0.41337 PPL. 420.17 Recons. 0.00528 +2024-03-30 00:25:23,311 INFO Train. Iter 45600 : Commit. 0.41308 PPL. 420.95 Recons. 0.00529 +2024-03-30 00:25:49,556 INFO Train. Iter 45800 : Commit. 0.38502 PPL. 422.32 Recons. 0.00435 +2024-03-30 00:26:14,796 INFO Train. Iter 46000 : Commit. 0.36280 PPL. 424.71 Recons. 0.00328 +2024-03-30 00:26:40,495 INFO Train. Iter 46200 : Commit. 0.36853 PPL. 422.23 Recons. 0.00417 +2024-03-30 00:27:05,741 INFO Train. Iter 46400 : Commit. 0.36916 PPL. 424.28 Recons. 0.00366 +2024-03-30 00:27:30,958 INFO Train. Iter 46600 : Commit. 0.37667 PPL. 423.55 Recons. 0.00394 +2024-03-30 00:27:56,182 INFO Train. Iter 46800 : Commit. 0.38290 PPL. 418.83 Recons. 0.00498 +2024-03-30 00:28:21,678 INFO Train. Iter 47000 : Commit. 0.36663 PPL. 421.85 Recons. 0.00394 +2024-03-30 00:28:47,509 INFO Train. Iter 47200 : Commit. 0.39280 PPL. 421.41 Recons. 0.00469 +2024-03-30 00:29:11,621 INFO Train. Iter 47400 : Commit. 0.38207 PPL. 422.45 Recons. 0.00442 +2024-03-30 00:29:37,191 INFO Train. Iter 47600 : Commit. 0.37870 PPL. 422.72 Recons. 0.00419 +2024-03-30 00:30:02,941 INFO Train. Iter 47800 : Commit. 0.35693 PPL. 423.96 Recons. 0.00353 +2024-03-30 00:30:28,384 INFO Train. Iter 48000 : Commit. 0.35953 PPL. 423.91 Recons. 0.00351 +2024-03-30 00:30:53,777 INFO Train. Iter 48200 : Commit. 0.39790 PPL. 417.72 Recons. 0.00547 +2024-03-30 00:31:18,871 INFO Train. Iter 48400 : Commit. 0.38997 PPL. 421.91 Recons. 0.00474 +2024-03-30 00:31:44,520 INFO Train. Iter 48600 : Commit. 0.37003 PPL. 422.22 Recons. 0.00416 +2024-03-30 00:32:09,893 INFO Train. Iter 48800 : Commit. 0.36208 PPL. 423.04 Recons. 0.00373 +2024-03-30 00:32:35,325 INFO Train. Iter 49000 : Commit. 0.36860 PPL. 421.67 Recons. 0.00404 +2024-03-30 00:32:59,905 INFO Train. Iter 49200 : Commit. 0.38378 PPL. 421.25 Recons. 0.00485 +2024-03-30 00:33:25,741 INFO Train. Iter 49400 : Commit. 0.36735 PPL. 423.57 Recons. 0.00389 +2024-03-30 00:33:51,131 INFO Train. Iter 49600 : Commit. 0.37333 PPL. 422.16 Recons. 0.00419 +2024-03-30 00:34:16,949 INFO Train. Iter 49800 : Commit. 0.35927 PPL. 423.52 Recons. 0.00367 +2024-03-30 00:34:42,768 INFO Train. Iter 50000 : Commit. 0.35710 PPL. 422.23 Recons. 0.00408 +2024-03-30 00:35:08,902 INFO Train. Iter 50200 : Commit. 0.36781 PPL. 423.08 Recons. 0.00382 +2024-03-30 00:35:34,946 INFO Train. Iter 50400 : Commit. 0.36678 PPL. 422.34 Recons. 0.00398 +2024-03-30 00:35:59,728 INFO Train. Iter 50600 : Commit. 0.36102 PPL. 423.04 Recons. 0.00369 +2024-03-30 00:36:25,154 INFO Train. Iter 50800 : Commit. 0.35774 PPL. 423.23 Recons. 0.00383 +2024-03-30 00:36:51,157 INFO Train. Iter 51000 : Commit. 0.36181 PPL. 422.43 Recons. 0.00396 +2024-03-30 00:37:15,044 INFO Train. Iter 51200 : Commit. 0.36894 PPL. 421.82 Recons. 0.00420 +2024-03-30 00:37:40,195 INFO Train. Iter 51400 : Commit. 0.37454 PPL. 421.29 Recons. 0.00450 +2024-03-30 00:38:04,990 INFO Train. Iter 51600 : Commit. 0.37279 PPL. 421.09 Recons. 0.00402 +2024-03-30 00:38:30,215 INFO Train. Iter 51800 : Commit. 0.35705 PPL. 421.85 Recons. 0.00406 +2024-03-30 00:38:55,536 INFO Train. Iter 52000 : Commit. 0.36385 PPL. 422.06 Recons. 0.00422 +2024-03-30 00:39:20,911 INFO Train. Iter 52200 : Commit. 0.35651 PPL. 422.68 Recons. 0.00359 +2024-03-30 00:39:46,896 INFO Train. Iter 52400 : Commit. 0.36617 PPL. 422.66 Recons. 0.00423 +2024-03-30 00:40:12,517 INFO Train. Iter 52600 : Commit. 0.36087 PPL. 422.43 Recons. 0.00371 +2024-03-30 00:40:38,157 INFO Train. Iter 52800 : Commit. 0.37597 PPL. 421.68 Recons. 0.00423 +2024-03-30 00:41:04,060 INFO Train. Iter 53000 : Commit. 0.35914 PPL. 420.94 Recons. 0.00399 +2024-03-30 00:41:28,499 INFO Train. Iter 53200 : Commit. 0.35210 PPL. 423.28 Recons. 0.00341 +2024-03-30 00:41:53,864 INFO Train. Iter 53400 : Commit. 0.33355 PPL. 423.10 Recons. 0.00321 +2024-03-30 00:42:19,642 INFO Train. Iter 53600 : Commit. 0.36234 PPL. 421.91 Recons. 0.00414 +2024-03-30 00:42:45,093 INFO Train. Iter 53800 : Commit. 0.33960 PPL. 420.59 Recons. 0.00394 +2024-03-30 00:43:10,785 INFO Train. Iter 54000 : Commit. 0.36081 PPL. 421.89 Recons. 0.00394 +2024-03-30 00:43:36,372 INFO Train. Iter 54200 : Commit. 0.33812 PPL. 423.04 Recons. 0.00341 +2024-03-30 00:44:02,620 INFO Train. Iter 54400 : Commit. 0.35457 PPL. 421.38 Recons. 0.00382 +2024-03-30 00:44:28,674 INFO Train. Iter 54600 : Commit. 0.36592 PPL. 419.30 Recons. 0.00443 +2024-03-30 00:44:54,707 INFO Train. Iter 54800 : Commit. 0.33918 PPL. 423.83 Recons. 0.00288 +2024-03-30 00:45:19,402 INFO Train. Iter 55000 : Commit. 0.34935 PPL. 421.60 Recons. 0.00399 +2024-03-30 00:45:45,285 INFO Train. Iter 55200 : Commit. 0.37364 PPL. 420.83 Recons. 0.00436 +2024-03-30 00:46:11,570 INFO Train. Iter 55400 : Commit. 0.35931 PPL. 421.26 Recons. 0.00412 +2024-03-30 00:46:37,437 INFO Train. Iter 55600 : Commit. 0.37112 PPL. 420.94 Recons. 0.00442 +2024-03-30 00:47:03,224 INFO Train. Iter 55800 : Commit. 0.35703 PPL. 422.88 Recons. 0.00359 +2024-03-30 00:47:28,839 INFO Train. Iter 56000 : Commit. 0.35129 PPL. 422.22 Recons. 0.00365 +2024-03-30 00:47:54,466 INFO Train. Iter 56200 : Commit. 0.35181 PPL. 421.04 Recons. 0.00410 +2024-03-30 00:48:20,628 INFO Train. Iter 56400 : Commit. 0.34601 PPL. 423.55 Recons. 0.00329 +2024-03-30 00:48:46,238 INFO Train. Iter 56600 : Commit. 0.35632 PPL. 419.91 Recons. 0.00420 +2024-03-30 00:49:12,149 INFO Train. Iter 56800 : Commit. 0.33344 PPL. 423.67 Recons. 0.00294 +2024-03-30 00:49:36,884 INFO Train. Iter 57000 : Commit. 0.36840 PPL. 419.51 Recons. 0.00493 +2024-03-30 00:50:02,059 INFO Train. Iter 57200 : Commit. 0.35271 PPL. 421.64 Recons. 0.00379 +2024-03-30 00:50:27,435 INFO Train. Iter 57400 : Commit. 0.35339 PPL. 421.53 Recons. 0.00375 +2024-03-30 00:50:53,244 INFO Train. Iter 57600 : Commit. 0.33395 PPL. 423.46 Recons. 0.00305 +2024-03-30 00:51:18,341 INFO Train. Iter 57800 : Commit. 0.36356 PPL. 420.09 Recons. 0.00440 +2024-03-30 00:51:43,850 INFO Train. Iter 58000 : Commit. 0.35112 PPL. 421.18 Recons. 0.00390 +2024-03-30 00:52:09,219 INFO Train. Iter 58200 : Commit. 0.36132 PPL. 420.37 Recons. 0.00412 +2024-03-30 00:52:34,771 INFO Train. Iter 58400 : Commit. 0.34962 PPL. 421.00 Recons. 0.00407 +2024-03-30 00:53:00,319 INFO Train. Iter 58600 : Commit. 0.33736 PPL. 422.97 Recons. 0.00310 +2024-03-30 00:53:25,825 INFO Train. Iter 58800 : Commit. 0.34817 PPL. 421.03 Recons. 0.00410 +2024-03-30 00:53:49,948 INFO Train. Iter 59000 : Commit. 0.35222 PPL. 420.84 Recons. 0.00390 +2024-03-30 00:54:15,596 INFO Train. Iter 59200 : Commit. 0.33766 PPL. 423.72 Recons. 0.00310 +2024-03-30 00:54:40,540 INFO Train. Iter 59400 : Commit. 0.33845 PPL. 421.01 Recons. 0.00372 +2024-03-30 00:55:05,769 INFO Train. Iter 59600 : Commit. 0.34283 PPL. 421.15 Recons. 0.00399 +2024-03-30 00:55:30,391 INFO Train. Iter 59800 : Commit. 0.32798 PPL. 423.51 Recons. 0.00289 +2024-03-30 00:55:55,372 INFO Train. Iter 60000 : Commit. 0.32607 PPL. 421.80 Recons. 0.00359 +2024-03-30 00:56:20,714 INFO Train. Iter 60200 : Commit. 0.35064 PPL. 421.00 Recons. 0.00367 +2024-03-30 00:56:45,820 INFO Train. Iter 60400 : Commit. 0.33689 PPL. 420.99 Recons. 0.00351 +2024-03-30 00:57:10,856 INFO Train. Iter 60600 : Commit. 0.33562 PPL. 422.23 Recons. 0.00343 +2024-03-30 00:57:34,515 INFO Train. Iter 60800 : Commit. 0.34201 PPL. 421.06 Recons. 0.00383 +2024-03-30 00:57:59,129 INFO Train. Iter 61000 : Commit. 0.35675 PPL. 419.33 Recons. 0.00430 +2024-03-30 00:58:23,919 INFO Train. Iter 61200 : Commit. 0.33907 PPL. 421.73 Recons. 0.00348 +2024-03-30 00:58:48,718 INFO Train. Iter 61400 : Commit. 0.35380 PPL. 418.70 Recons. 0.00436 +2024-03-30 00:59:12,781 INFO Train. Iter 61600 : Commit. 0.34029 PPL. 421.46 Recons. 0.00346 +2024-03-30 00:59:38,451 INFO Train. Iter 61800 : Commit. 0.34304 PPL. 420.43 Recons. 0.00392 +2024-03-30 01:00:03,829 INFO Train. Iter 62000 : Commit. 0.34494 PPL. 420.60 Recons. 0.00388 +2024-03-30 01:00:28,513 INFO Train. Iter 62200 : Commit. 0.34716 PPL. 420.40 Recons. 0.00420 +2024-03-30 01:00:54,049 INFO Train. Iter 62400 : Commit. 0.33890 PPL. 423.19 Recons. 0.00318 +2024-03-30 01:01:19,262 INFO Train. Iter 62600 : Commit. 0.33708 PPL. 422.65 Recons. 0.00325 +2024-03-30 01:01:43,728 INFO Train. Iter 62800 : Commit. 0.35254 PPL. 419.89 Recons. 0.00431 +2024-03-30 01:02:09,277 INFO Train. Iter 63000 : Commit. 0.33060 PPL. 421.84 Recons. 0.00326 +2024-03-30 01:02:34,165 INFO Train. Iter 63200 : Commit. 0.33307 PPL. 420.77 Recons. 0.00367 +2024-03-30 01:02:59,131 INFO Train. Iter 63400 : Commit. 0.34167 PPL. 422.25 Recons. 0.00335 +2024-03-30 01:03:24,209 INFO Train. Iter 63600 : Commit. 0.33775 PPL. 420.50 Recons. 0.00359 +2024-03-30 01:03:49,841 INFO Train. Iter 63800 : Commit. 0.33603 PPL. 421.29 Recons. 0.00353 +2024-03-30 01:04:15,023 INFO Train. Iter 64000 : Commit. 0.31842 PPL. 422.27 Recons. 0.00297 +2024-03-30 01:04:40,649 INFO Train. Iter 64200 : Commit. 0.34668 PPL. 417.34 Recons. 0.00480 +2024-03-30 01:05:05,729 INFO Train. Iter 64400 : Commit. 0.33878 PPL. 422.39 Recons. 0.00320 +2024-03-30 01:05:30,874 INFO Train. Iter 64600 : Commit. 0.33242 PPL. 419.27 Recons. 0.00400 +2024-03-30 01:05:54,676 INFO Train. Iter 64800 : Commit. 0.34183 PPL. 420.18 Recons. 0.00370 +2024-03-30 01:06:19,959 INFO Train. Iter 65000 : Commit. 0.34268 PPL. 421.24 Recons. 0.00361 +2024-03-30 01:06:44,647 INFO Train. Iter 65200 : Commit. 0.33834 PPL. 421.34 Recons. 0.00357 +2024-03-30 01:07:09,439 INFO Train. Iter 65400 : Commit. 0.34050 PPL. 420.68 Recons. 0.00375 +2024-03-30 01:07:34,293 INFO Train. Iter 65600 : Commit. 0.32349 PPL. 420.57 Recons. 0.00312 +2024-03-30 01:08:00,101 INFO Train. Iter 65800 : Commit. 0.34913 PPL. 420.86 Recons. 0.00393 +2024-03-30 01:08:24,949 INFO Train. Iter 66000 : Commit. 0.33362 PPL. 420.82 Recons. 0.00333 +2024-03-30 01:08:49,839 INFO Train. Iter 66200 : Commit. 0.35549 PPL. 419.44 Recons. 0.00413 +2024-03-30 01:09:14,883 INFO Train. Iter 66400 : Commit. 0.32903 PPL. 420.73 Recons. 0.00332 +2024-03-30 01:09:38,828 INFO Train. Iter 66600 : Commit. 0.31565 PPL. 423.17 Recons. 0.00268 +2024-03-30 01:10:03,551 INFO Train. Iter 66800 : Commit. 0.32371 PPL. 421.32 Recons. 0.00343 +2024-03-30 01:10:28,279 INFO Train. Iter 67000 : Commit. 0.33384 PPL. 420.90 Recons. 0.00362 +2024-03-30 01:10:53,703 INFO Train. Iter 67200 : Commit. 0.32808 PPL. 421.68 Recons. 0.00320 +2024-03-30 01:11:19,031 INFO Train. Iter 67400 : Commit. 0.33833 PPL. 417.02 Recons. 0.00415 +2024-03-30 01:11:44,179 INFO Train. Iter 67600 : Commit. 0.32610 PPL. 422.43 Recons. 0.00301 +2024-03-30 01:12:09,983 INFO Train. Iter 67800 : Commit. 0.35187 PPL. 418.15 Recons. 0.00422 +2024-03-30 01:12:34,678 INFO Train. Iter 68000 : Commit. 0.33010 PPL. 422.14 Recons. 0.00323 +2024-03-30 01:12:59,329 INFO Train. Iter 68200 : Commit. 0.33769 PPL. 418.15 Recons. 0.00424 +2024-03-30 01:13:24,077 INFO Train. Iter 68400 : Commit. 0.32492 PPL. 422.29 Recons. 0.00302 +2024-03-30 01:13:47,775 INFO Train. Iter 68600 : Commit. 0.33082 PPL. 420.63 Recons. 0.00356 +2024-03-30 01:14:11,897 INFO Train. Iter 68800 : Commit. 0.33151 PPL. 420.96 Recons. 0.00350 +2024-03-30 01:14:36,630 INFO Train. Iter 69000 : Commit. 0.34040 PPL. 418.55 Recons. 0.00415 +2024-03-30 01:15:01,000 INFO Train. Iter 69200 : Commit. 0.32624 PPL. 420.38 Recons. 0.00328 +2024-03-30 01:15:25,534 INFO Train. Iter 69400 : Commit. 0.32587 PPL. 421.56 Recons. 0.00305 +2024-03-30 01:15:49,302 INFO Train. Iter 69600 : Commit. 0.32990 PPL. 419.51 Recons. 0.00388 +2024-03-30 01:16:13,205 INFO Train. Iter 69800 : Commit. 0.32470 PPL. 421.65 Recons. 0.00323 +2024-03-30 01:16:36,872 INFO Train. Iter 70000 : Commit. 0.30812 PPL. 422.82 Recons. 0.00259 +2024-03-30 01:17:00,891 INFO Train. Iter 70200 : Commit. 0.34525 PPL. 417.18 Recons. 0.00466 +2024-03-30 01:17:25,266 INFO Train. Iter 70400 : Commit. 0.33826 PPL. 420.78 Recons. 0.00323 +2024-03-30 01:17:49,343 INFO Train. Iter 70600 : Commit. 0.31695 PPL. 421.02 Recons. 0.00305 +2024-03-30 01:18:14,350 INFO Train. Iter 70800 : Commit. 0.32618 PPL. 419.07 Recons. 0.00382 +2024-03-30 01:18:39,243 INFO Train. Iter 71000 : Commit. 0.32607 PPL. 422.47 Recons. 0.00285 +2024-03-30 01:19:04,555 INFO Train. Iter 71200 : Commit. 0.33382 PPL. 420.83 Recons. 0.00366 +2024-03-30 01:19:29,560 INFO Train. Iter 71400 : Commit. 0.33163 PPL. 419.96 Recons. 0.00350 +2024-03-30 01:19:54,512 INFO Train. Iter 71600 : Commit. 0.32747 PPL. 419.45 Recons. 0.00348 +2024-03-30 01:20:20,085 INFO Train. Iter 71800 : Commit. 0.33324 PPL. 419.98 Recons. 0.00377 +2024-03-30 01:20:45,796 INFO Train. Iter 72000 : Commit. 0.34287 PPL. 418.69 Recons. 0.00392 +2024-03-30 01:21:11,163 INFO Train. Iter 72200 : Commit. 0.31732 PPL. 421.33 Recons. 0.00303 +2024-03-30 01:21:35,450 INFO Train. Iter 72400 : Commit. 0.32060 PPL. 420.48 Recons. 0.00305 +2024-03-30 01:22:01,074 INFO Train. Iter 72600 : Commit. 0.32974 PPL. 419.50 Recons. 0.00361 +2024-03-30 01:22:26,121 INFO Train. Iter 72800 : Commit. 0.32802 PPL. 420.23 Recons. 0.00354 +2024-03-30 01:22:51,103 INFO Train. Iter 73000 : Commit. 0.32483 PPL. 421.49 Recons. 0.00314 +2024-03-30 01:23:16,319 INFO Train. Iter 73200 : Commit. 0.33178 PPL. 418.81 Recons. 0.00364 +2024-03-30 01:23:41,676 INFO Train. Iter 73400 : Commit. 0.31754 PPL. 421.66 Recons. 0.00285 +2024-03-30 01:24:07,038 INFO Train. Iter 73600 : Commit. 0.31787 PPL. 420.88 Recons. 0.00313 +2024-03-30 01:24:31,909 INFO Train. Iter 73800 : Commit. 0.32038 PPL. 420.24 Recons. 0.00331 +2024-03-30 01:24:56,604 INFO Train. Iter 74000 : Commit. 0.33156 PPL. 419.48 Recons. 0.00362 +2024-03-30 01:25:21,174 INFO Train. Iter 74200 : Commit. 0.32451 PPL. 419.93 Recons. 0.00324 +2024-03-30 01:25:45,648 INFO Train. Iter 74400 : Commit. 0.32241 PPL. 419.70 Recons. 0.00347 +2024-03-30 01:26:11,208 INFO Train. Iter 74600 : Commit. 0.31549 PPL. 421.03 Recons. 0.00299 +2024-03-30 01:26:37,173 INFO Train. Iter 74800 : Commit. 0.31700 PPL. 420.26 Recons. 0.00317 +2024-03-30 01:27:02,216 INFO Train. Iter 75000 : Commit. 0.31526 PPL. 420.53 Recons. 0.00324 +2024-03-30 01:27:27,578 INFO Train. Iter 75200 : Commit. 0.32277 PPL. 419.20 Recons. 0.00359 +2024-03-30 01:27:53,160 INFO Train. Iter 75400 : Commit. 0.31032 PPL. 421.49 Recons. 0.00277 +2024-03-30 01:28:18,588 INFO Train. Iter 75600 : Commit. 0.31146 PPL. 420.34 Recons. 0.00314 +2024-03-30 01:28:44,547 INFO Train. Iter 75800 : Commit. 0.31404 PPL. 421.23 Recons. 0.00296 +2024-03-30 01:29:10,825 INFO Train. Iter 76000 : Commit. 0.31790 PPL. 419.39 Recons. 0.00322 +2024-03-30 01:29:36,071 INFO Train. Iter 76200 : Commit. 0.32471 PPL. 419.34 Recons. 0.00361 +2024-03-30 01:30:01,109 INFO Train. Iter 76400 : Commit. 0.34032 PPL. 417.96 Recons. 0.00428 +2024-03-30 01:30:26,492 INFO Train. Iter 76600 : Commit. 0.30597 PPL. 422.26 Recons. 0.00254 +2024-03-30 01:30:51,868 INFO Train. Iter 76800 : Commit. 0.31158 PPL. 420.13 Recons. 0.00305 +2024-03-30 01:31:18,269 INFO Train. Iter 77000 : Commit. 0.33070 PPL. 419.03 Recons. 0.00401 +2024-03-30 01:31:43,461 INFO Train. Iter 77200 : Commit. 0.31352 PPL. 422.53 Recons. 0.00263 +2024-03-30 01:32:08,758 INFO Train. Iter 77400 : Commit. 0.32366 PPL. 418.27 Recons. 0.00366 +2024-03-30 01:32:34,214 INFO Train. Iter 77600 : Commit. 0.32484 PPL. 418.49 Recons. 0.00358 +2024-03-30 01:32:59,014 INFO Train. Iter 77800 : Commit. 0.31898 PPL. 420.50 Recons. 0.00315 +2024-03-30 01:33:23,664 INFO Train. Iter 78000 : Commit. 0.31902 PPL. 417.67 Recons. 0.00367 +2024-03-30 01:33:47,302 INFO Train. Iter 78200 : Commit. 0.31942 PPL. 419.43 Recons. 0.00315 +2024-03-30 01:34:12,454 INFO Train. Iter 78400 : Commit. 0.31512 PPL. 420.25 Recons. 0.00324 +2024-03-30 01:34:37,002 INFO Train. Iter 78600 : Commit. 0.30688 PPL. 420.09 Recons. 0.00296 +2024-03-30 01:35:01,826 INFO Train. Iter 78800 : Commit. 0.30684 PPL. 420.98 Recons. 0.00284 +2024-03-30 01:35:27,598 INFO Train. Iter 79000 : Commit. 0.32718 PPL. 418.98 Recons. 0.00367 +2024-03-30 01:35:52,995 INFO Train. Iter 79200 : Commit. 0.31340 PPL. 418.97 Recons. 0.00320 +2024-03-30 01:36:18,447 INFO Train. Iter 79400 : Commit. 0.31111 PPL. 420.45 Recons. 0.00296 +2024-03-30 01:36:44,163 INFO Train. Iter 79600 : Commit. 0.30499 PPL. 420.30 Recons. 0.00291 +2024-03-30 01:37:09,610 INFO Train. Iter 79800 : Commit. 0.31634 PPL. 420.00 Recons. 0.00313 +2024-03-30 01:37:35,010 INFO Train. Iter 80000 : Commit. 0.32091 PPL. 417.36 Recons. 0.00380 +2024-03-30 01:37:59,981 INFO Train. Iter 80200 : Commit. 0.30788 PPL. 421.79 Recons. 0.00259 +2024-03-30 01:38:25,792 INFO Train. Iter 80400 : Commit. 0.31362 PPL. 419.63 Recons. 0.00315 +2024-03-30 01:38:51,434 INFO Train. Iter 80600 : Commit. 0.30567 PPL. 420.95 Recons. 0.00279 +2024-03-30 01:39:17,139 INFO Train. Iter 80800 : Commit. 0.30688 PPL. 418.27 Recons. 0.00370 +2024-03-30 01:39:42,227 INFO Train. Iter 81000 : Commit. 0.32266 PPL. 419.21 Recons. 0.00340 +2024-03-30 01:40:07,287 INFO Train. Iter 81200 : Commit. 0.32568 PPL. 419.04 Recons. 0.00355 +2024-03-30 01:40:31,953 INFO Train. Iter 81400 : Commit. 0.31595 PPL. 418.99 Recons. 0.00336 +2024-03-30 01:40:57,426 INFO Train. Iter 81600 : Commit. 0.31354 PPL. 419.28 Recons. 0.00298 +2024-03-30 01:41:22,934 INFO Train. Iter 81800 : Commit. 0.31701 PPL. 419.64 Recons. 0.00321 +2024-03-30 01:41:48,270 INFO Train. Iter 82000 : Commit. 0.30616 PPL. 417.89 Recons. 0.00326 +2024-03-30 01:42:12,619 INFO Train. Iter 82200 : Commit. 0.30476 PPL. 417.19 Recons. 0.00338 +2024-03-30 01:42:37,974 INFO Train. Iter 82400 : Commit. 0.31254 PPL. 420.72 Recons. 0.00304 +2024-03-30 01:43:03,102 INFO Train. Iter 82600 : Commit. 0.29225 PPL. 420.94 Recons. 0.00255 +2024-03-30 01:43:28,135 INFO Train. Iter 82800 : Commit. 0.30779 PPL. 418.50 Recons. 0.00331 +2024-03-30 01:43:53,181 INFO Train. Iter 83000 : Commit. 0.31386 PPL. 419.33 Recons. 0.00344 +2024-03-30 01:44:18,226 INFO Train. Iter 83200 : Commit. 0.29837 PPL. 420.64 Recons. 0.00271 +2024-03-30 01:44:43,065 INFO Train. Iter 83400 : Commit. 0.30580 PPL. 418.81 Recons. 0.00328 +2024-03-30 01:45:08,430 INFO Train. Iter 83600 : Commit. 0.30734 PPL. 419.20 Recons. 0.00330 +2024-03-30 01:45:33,940 INFO Train. Iter 83800 : Commit. 0.31119 PPL. 419.50 Recons. 0.00330 +2024-03-30 01:45:58,469 INFO Train. Iter 84000 : Commit. 0.31450 PPL. 418.89 Recons. 0.00343 +2024-03-30 01:46:24,519 INFO Train. Iter 84200 : Commit. 0.30476 PPL. 420.69 Recons. 0.00269 +2024-03-30 01:46:50,716 INFO Train. Iter 84400 : Commit. 0.31042 PPL. 420.04 Recons. 0.00301 +2024-03-30 01:47:15,813 INFO Train. Iter 84600 : Commit. 0.30772 PPL. 419.94 Recons. 0.00296 +2024-03-30 01:47:41,753 INFO Train. Iter 84800 : Commit. 0.31370 PPL. 417.71 Recons. 0.00347 +2024-03-30 01:48:07,445 INFO Train. Iter 85000 : Commit. 0.32674 PPL. 414.82 Recons. 0.00427 +2024-03-30 01:48:33,052 INFO Train. Iter 85200 : Commit. 0.30475 PPL. 420.85 Recons. 0.00252 +2024-03-30 01:48:58,656 INFO Train. Iter 85400 : Commit. 0.31241 PPL. 416.69 Recons. 0.00332 +2024-03-30 01:49:24,958 INFO Train. Iter 85600 : Commit. 0.30208 PPL. 420.36 Recons. 0.00274 +2024-03-30 01:49:51,111 INFO Train. Iter 85800 : Commit. 0.31655 PPL. 418.98 Recons. 0.00337 +2024-03-30 01:50:15,439 INFO Train. Iter 86000 : Commit. 0.29765 PPL. 420.68 Recons. 0.00272 +2024-03-30 01:50:41,105 INFO Train. Iter 86200 : Commit. 0.30478 PPL. 419.20 Recons. 0.00291 +2024-03-30 01:51:06,078 INFO Train. Iter 86400 : Commit. 0.30559 PPL. 419.06 Recons. 0.00311 +2024-03-30 01:51:31,342 INFO Train. Iter 86600 : Commit. 0.31115 PPL. 418.78 Recons. 0.00342 +2024-03-30 01:51:55,796 INFO Train. Iter 86800 : Commit. 0.30826 PPL. 417.83 Recons. 0.00330 +2024-03-30 01:52:20,282 INFO Train. Iter 87000 : Commit. 0.31376 PPL. 418.15 Recons. 0.00329 +2024-03-30 01:52:43,873 INFO Train. Iter 87200 : Commit. 0.31092 PPL. 418.90 Recons. 0.00340 +2024-03-30 01:53:07,900 INFO Train. Iter 87400 : Commit. 0.30943 PPL. 420.07 Recons. 0.00287 +2024-03-30 01:53:32,819 INFO Train. Iter 87600 : Commit. 0.29868 PPL. 420.60 Recons. 0.00269 +2024-03-30 01:53:56,756 INFO Train. Iter 87800 : Commit. 0.29220 PPL. 419.28 Recons. 0.00289 +2024-03-30 01:54:20,043 INFO Train. Iter 88000 : Commit. 0.29643 PPL. 419.92 Recons. 0.00266 +2024-03-30 01:54:44,706 INFO Train. Iter 88200 : Commit. 0.30433 PPL. 417.16 Recons. 0.00335 +2024-03-30 01:55:09,445 INFO Train. Iter 88400 : Commit. 0.31029 PPL. 418.77 Recons. 0.00330 +2024-03-30 01:55:33,723 INFO Train. Iter 88600 : Commit. 0.30937 PPL. 417.90 Recons. 0.00335 +2024-03-30 01:55:59,066 INFO Train. Iter 88800 : Commit. 0.30895 PPL. 418.40 Recons. 0.00327 +2024-03-30 01:56:24,609 INFO Train. Iter 89000 : Commit. 0.29354 PPL. 420.25 Recons. 0.00258 +2024-03-30 01:56:49,980 INFO Train. Iter 89200 : Commit. 0.30489 PPL. 418.57 Recons. 0.00312 +2024-03-30 01:57:14,741 INFO Train. Iter 89400 : Commit. 0.31344 PPL. 417.73 Recons. 0.00359 +2024-03-30 01:57:40,138 INFO Train. Iter 89600 : Commit. 0.30333 PPL. 419.09 Recons. 0.00279 +2024-03-30 01:58:04,005 INFO Train. Iter 89800 : Commit. 0.29356 PPL. 420.59 Recons. 0.00257 +2024-03-30 01:58:28,965 INFO Train. Iter 90000 : Commit. 0.30444 PPL. 418.44 Recons. 0.00329 +2024-03-30 01:58:54,093 INFO Train. Iter 90200 : Commit. 0.31798 PPL. 416.06 Recons. 0.00368 +2024-03-30 01:59:19,632 INFO Train. Iter 90400 : Commit. 0.31221 PPL. 417.87 Recons. 0.00342 +2024-03-30 01:59:44,998 INFO Train. Iter 90600 : Commit. 0.30470 PPL. 419.88 Recons. 0.00283 +2024-03-30 02:00:10,473 INFO Train. Iter 90800 : Commit. 0.31276 PPL. 417.93 Recons. 0.00342 +2024-03-30 02:00:36,032 INFO Train. Iter 91000 : Commit. 0.29114 PPL. 420.02 Recons. 0.00249 +2024-03-30 02:01:01,567 INFO Train. Iter 91200 : Commit. 0.29844 PPL. 419.12 Recons. 0.00293 +2024-03-30 02:01:26,552 INFO Train. Iter 91400 : Commit. 0.31037 PPL. 417.98 Recons. 0.00337 +2024-03-30 02:01:51,606 INFO Train. Iter 91600 : Commit. 0.30953 PPL. 416.80 Recons. 0.00364 +2024-03-30 02:02:15,230 INFO Train. Iter 91800 : Commit. 0.31911 PPL. 419.11 Recons. 0.00319 +2024-03-30 02:02:40,409 INFO Train. Iter 92000 : Commit. 0.30139 PPL. 418.94 Recons. 0.00304 +2024-03-30 02:03:05,889 INFO Train. Iter 92200 : Commit. 0.30759 PPL. 418.26 Recons. 0.00318 +2024-03-30 02:03:30,831 INFO Train. Iter 92400 : Commit. 0.30213 PPL. 417.61 Recons. 0.00305 +2024-03-30 02:03:55,773 INFO Train. Iter 92600 : Commit. 0.30329 PPL. 418.40 Recons. 0.00322 +2024-03-30 02:04:20,584 INFO Train. Iter 92800 : Commit. 0.31506 PPL. 417.89 Recons. 0.00337 +2024-03-30 02:04:45,412 INFO Train. Iter 93000 : Commit. 0.29956 PPL. 419.47 Recons. 0.00288 +2024-03-30 02:05:10,247 INFO Train. Iter 93200 : Commit. 0.30020 PPL. 417.90 Recons. 0.00310 +2024-03-30 02:05:34,916 INFO Train. Iter 93400 : Commit. 0.29591 PPL. 417.90 Recons. 0.00303 +2024-03-30 02:06:00,257 INFO Train. Iter 93600 : Commit. 0.30172 PPL. 417.14 Recons. 0.00332 +2024-03-30 02:06:24,388 INFO Train. Iter 93800 : Commit. 0.30385 PPL. 418.88 Recons. 0.00283 +2024-03-30 02:06:49,446 INFO Train. Iter 94000 : Commit. 0.31098 PPL. 416.44 Recons. 0.00360 +2024-03-30 02:07:14,220 INFO Train. Iter 94200 : Commit. 0.30395 PPL. 419.03 Recons. 0.00290 +2024-03-30 02:07:38,804 INFO Train. Iter 94400 : Commit. 0.29479 PPL. 418.28 Recons. 0.00281 +2024-03-30 02:08:03,884 INFO Train. Iter 94600 : Commit. 0.29324 PPL. 416.73 Recons. 0.00297 +2024-03-30 02:08:29,544 INFO Train. Iter 94800 : Commit. 0.29350 PPL. 419.07 Recons. 0.00265 +2024-03-30 02:08:54,136 INFO Train. Iter 95000 : Commit. 0.31095 PPL. 415.80 Recons. 0.00356 +2024-03-30 02:09:19,305 INFO Train. Iter 95200 : Commit. 0.28952 PPL. 419.41 Recons. 0.00249 +2024-03-30 02:09:44,497 INFO Train. Iter 95400 : Commit. 0.30342 PPL. 418.65 Recons. 0.00297 +2024-03-30 02:10:08,657 INFO Train. Iter 95600 : Commit. 0.29147 PPL. 418.36 Recons. 0.00298 +2024-03-30 02:10:33,786 INFO Train. Iter 95800 : Commit. 0.29532 PPL. 417.91 Recons. 0.00282 +2024-03-30 02:10:58,784 INFO Train. Iter 96000 : Commit. 0.30176 PPL. 418.44 Recons. 0.00295 +2024-03-30 02:11:23,553 INFO Train. Iter 96200 : Commit. 0.28795 PPL. 418.41 Recons. 0.00275 +2024-03-30 02:11:48,505 INFO Train. Iter 96400 : Commit. 0.29441 PPL. 417.48 Recons. 0.00310 +2024-03-30 02:12:13,736 INFO Train. Iter 96600 : Commit. 0.30163 PPL. 416.46 Recons. 0.00321 +2024-03-30 02:12:38,657 INFO Train. Iter 96800 : Commit. 0.30402 PPL. 417.19 Recons. 0.00325 +2024-03-30 02:13:03,740 INFO Train. Iter 97000 : Commit. 0.29318 PPL. 419.48 Recons. 0.00252 +2024-03-30 02:13:29,404 INFO Train. Iter 97200 : Commit. 0.29647 PPL. 417.18 Recons. 0.00295 +2024-03-30 02:13:55,170 INFO Train. Iter 97400 : Commit. 0.29073 PPL. 419.00 Recons. 0.00258 +2024-03-30 02:14:19,599 INFO Train. Iter 97600 : Commit. 0.29289 PPL. 417.46 Recons. 0.00293 +2024-03-30 02:14:45,203 INFO Train. Iter 97800 : Commit. 0.30192 PPL. 416.71 Recons. 0.00328 +2024-03-30 02:15:11,286 INFO Train. Iter 98000 : Commit. 0.29464 PPL. 417.90 Recons. 0.00284 +2024-03-30 02:15:37,144 INFO Train. Iter 98200 : Commit. 0.29352 PPL. 418.68 Recons. 0.00269 +2024-03-30 02:16:02,636 INFO Train. Iter 98400 : Commit. 0.29202 PPL. 417.47 Recons. 0.00297 +2024-03-30 02:16:27,680 INFO Train. Iter 98600 : Commit. 0.29115 PPL. 418.10 Recons. 0.00273 +2024-03-30 02:16:52,864 INFO Train. Iter 98800 : Commit. 0.29364 PPL. 417.32 Recons. 0.00301 +2024-03-30 02:17:17,885 INFO Train. Iter 99000 : Commit. 0.28925 PPL. 417.79 Recons. 0.00268 +2024-03-30 02:17:43,903 INFO Train. Iter 99200 : Commit. 0.28621 PPL. 418.59 Recons. 0.00248 +2024-03-30 02:18:09,380 INFO Train. Iter 99400 : Commit. 0.29474 PPL. 414.24 Recons. 0.00318 +2024-03-30 02:18:33,671 INFO Train. Iter 99600 : Commit. 0.29187 PPL. 417.42 Recons. 0.00286 +2024-03-30 02:18:59,015 INFO Train. Iter 99800 : Commit. 0.29302 PPL. 417.99 Recons. 0.00272 +2024-03-30 02:19:24,314 INFO Train. Iter 100000 : Commit. 0.29430 PPL. 418.34 Recons. 0.00272 +2024-03-30 02:19:50,136 INFO Train. Iter 100200 : Commit. 0.28157 PPL. 418.87 Recons. 0.00241 +2024-03-30 02:20:15,778 INFO Train. Iter 100400 : Commit. 0.30712 PPL. 413.64 Recons. 0.00367 +2024-03-30 02:20:40,972 INFO Train. Iter 100600 : Commit. 0.29353 PPL. 418.02 Recons. 0.00273 +2024-03-30 02:21:06,171 INFO Train. Iter 100800 : Commit. 0.29344 PPL. 415.86 Recons. 0.00311 +2024-03-30 02:21:31,260 INFO Train. Iter 101000 : Commit. 0.29308 PPL. 417.75 Recons. 0.00278 +2024-03-30 02:21:56,458 INFO Train. Iter 101200 : Commit. 0.30420 PPL. 416.10 Recons. 0.00323 +2024-03-30 02:22:19,932 INFO Train. Iter 101400 : Commit. 0.29332 PPL. 418.15 Recons. 0.00268 +2024-03-30 02:22:44,652 INFO Train. Iter 101600 : Commit. 0.29760 PPL. 416.67 Recons. 0.00309 +2024-03-30 02:23:10,108 INFO Train. Iter 101800 : Commit. 0.28366 PPL. 418.43 Recons. 0.00238 +2024-03-30 02:23:35,181 INFO Train. Iter 102000 : Commit. 0.30062 PPL. 417.17 Recons. 0.00311 +2024-03-30 02:24:00,861 INFO Train. Iter 102200 : Commit. 0.29765 PPL. 416.67 Recons. 0.00310 +2024-03-30 02:24:26,345 INFO Train. Iter 102400 : Commit. 0.29352 PPL. 416.00 Recons. 0.00279 +2024-03-30 02:24:51,907 INFO Train. Iter 102600 : Commit. 0.30238 PPL. 418.26 Recons. 0.00297 +2024-03-30 02:25:16,960 INFO Train. Iter 102800 : Commit. 0.28249 PPL. 418.74 Recons. 0.00230 +2024-03-30 02:25:42,428 INFO Train. Iter 103000 : Commit. 0.30233 PPL. 414.80 Recons. 0.00340 +2024-03-30 02:26:07,499 INFO Train. Iter 103200 : Commit. 0.28357 PPL. 418.05 Recons. 0.00240 +2024-03-30 02:26:31,291 INFO Train. Iter 103400 : Commit. 0.30692 PPL. 415.10 Recons. 0.00361 +2024-03-30 02:26:56,425 INFO Train. Iter 103600 : Commit. 0.29827 PPL. 413.57 Recons. 0.00315 +2024-03-30 02:27:21,536 INFO Train. Iter 103800 : Commit. 0.29008 PPL. 416.15 Recons. 0.00282 +2024-03-30 02:27:46,534 INFO Train. Iter 104000 : Commit. 0.29863 PPL. 417.32 Recons. 0.00277 +2024-03-30 02:28:11,882 INFO Train. Iter 104200 : Commit. 0.30583 PPL. 415.90 Recons. 0.00341 +2024-03-30 02:28:37,086 INFO Train. Iter 104400 : Commit. 0.28711 PPL. 417.73 Recons. 0.00243 +2024-03-30 02:29:01,986 INFO Train. Iter 104600 : Commit. 0.28974 PPL. 415.58 Recons. 0.00291 +2024-03-30 02:29:26,861 INFO Train. Iter 104800 : Commit. 0.28438 PPL. 417.93 Recons. 0.00256 +2024-03-30 02:29:52,027 INFO Train. Iter 105000 : Commit. 0.29057 PPL. 415.71 Recons. 0.00315 +2024-03-30 02:30:17,873 INFO Train. Iter 105200 : Commit. 0.27804 PPL. 418.60 Recons. 0.00228 +2024-03-30 02:30:42,002 INFO Train. Iter 105400 : Commit. 0.29161 PPL. 416.80 Recons. 0.00307 +2024-03-30 02:31:06,604 INFO Train. Iter 105600 : Commit. 0.27296 PPL. 418.48 Recons. 0.00221 +2024-03-30 02:31:31,462 INFO Train. Iter 105800 : Commit. 0.29216 PPL. 416.45 Recons. 0.00299 +2024-03-30 02:31:56,211 INFO Train. Iter 106000 : Commit. 0.28675 PPL. 415.20 Recons. 0.00280 +2024-03-30 02:32:20,851 INFO Train. Iter 106200 : Commit. 0.29332 PPL. 416.51 Recons. 0.00296 +2024-03-30 02:32:46,012 INFO Train. Iter 106400 : Commit. 0.28893 PPL. 417.38 Recons. 0.00286 +2024-03-30 02:33:11,177 INFO Train. Iter 106600 : Commit. 0.28969 PPL. 417.35 Recons. 0.00269 +2024-03-30 02:33:36,057 INFO Train. Iter 106800 : Commit. 0.29093 PPL. 415.19 Recons. 0.00310 +2024-03-30 02:34:00,998 INFO Train. Iter 107000 : Commit. 0.29152 PPL. 416.48 Recons. 0.00275 +2024-03-30 02:34:25,175 INFO Train. Iter 107200 : Commit. 0.28881 PPL. 416.68 Recons. 0.00279 +2024-03-30 02:34:50,203 INFO Train. Iter 107400 : Commit. 0.30033 PPL. 415.64 Recons. 0.00316 +2024-03-30 02:35:16,212 INFO Train. Iter 107600 : Commit. 0.28548 PPL. 417.94 Recons. 0.00252 +2024-03-30 02:35:41,699 INFO Train. Iter 107800 : Commit. 0.29569 PPL. 414.62 Recons. 0.00317 +2024-03-30 02:36:07,256 INFO Train. Iter 108000 : Commit. 0.29075 PPL. 415.75 Recons. 0.00306 +2024-03-30 02:36:32,902 INFO Train. Iter 108200 : Commit. 0.30425 PPL. 414.09 Recons. 0.00331 +2024-03-30 02:36:57,750 INFO Train. Iter 108400 : Commit. 0.29097 PPL. 416.15 Recons. 0.00281 +2024-03-30 02:37:22,655 INFO Train. Iter 108600 : Commit. 0.30659 PPL. 414.08 Recons. 0.00351 +2024-03-30 02:37:47,653 INFO Train. Iter 108800 : Commit. 0.28283 PPL. 417.79 Recons. 0.00243 +2024-03-30 02:38:12,355 INFO Train. Iter 109000 : Commit. 0.28239 PPL. 417.07 Recons. 0.00260 +2024-03-30 02:38:36,029 INFO Train. Iter 109200 : Commit. 0.28237 PPL. 416.83 Recons. 0.00259 +2024-03-30 02:39:00,493 INFO Train. Iter 109400 : Commit. 0.29313 PPL. 414.00 Recons. 0.00319 +2024-03-30 02:39:25,303 INFO Train. Iter 109600 : Commit. 0.28354 PPL. 417.15 Recons. 0.00257 +2024-03-30 02:39:50,616 INFO Train. Iter 109800 : Commit. 0.29314 PPL. 416.64 Recons. 0.00286 +2024-03-30 02:40:15,139 INFO Train. Iter 110000 : Commit. 0.28912 PPL. 415.31 Recons. 0.00307 +2024-03-30 02:40:39,550 INFO Train. Iter 110200 : Commit. 0.28692 PPL. 416.61 Recons. 0.00271 +2024-03-30 02:41:04,669 INFO Train. Iter 110400 : Commit. 0.29115 PPL. 416.36 Recons. 0.00294 +2024-03-30 02:41:29,345 INFO Train. Iter 110600 : Commit. 0.29321 PPL. 414.99 Recons. 0.00315 +2024-03-30 02:41:54,025 INFO Train. Iter 110800 : Commit. 0.28450 PPL. 416.01 Recons. 0.00265 +2024-03-30 02:42:19,724 INFO Train. Iter 111000 : Commit. 0.29215 PPL. 415.69 Recons. 0.00279 +2024-03-30 02:42:44,209 INFO Train. Iter 111200 : Commit. 0.28925 PPL. 416.58 Recons. 0.00284 +2024-03-30 02:43:10,086 INFO Train. Iter 111400 : Commit. 0.28805 PPL. 417.08 Recons. 0.00274 +2024-03-30 02:43:35,943 INFO Train. Iter 111600 : Commit. 0.28636 PPL. 417.14 Recons. 0.00265 +2024-03-30 02:44:01,626 INFO Train. Iter 111800 : Commit. 0.28580 PPL. 415.76 Recons. 0.00292 +2024-03-30 02:44:28,061 INFO Train. Iter 112000 : Commit. 0.28199 PPL. 417.11 Recons. 0.00259 +2024-03-30 02:44:53,771 INFO Train. Iter 112200 : Commit. 0.29490 PPL. 415.72 Recons. 0.00307 +2024-03-30 02:45:19,198 INFO Train. Iter 112400 : Commit. 0.29080 PPL. 415.97 Recons. 0.00288 +2024-03-30 02:45:43,885 INFO Train. Iter 112600 : Commit. 0.28758 PPL. 415.60 Recons. 0.00280 +2024-03-30 02:46:08,539 INFO Train. Iter 112800 : Commit. 0.27521 PPL. 417.20 Recons. 0.00231 +2024-03-30 02:46:32,171 INFO Train. Iter 113000 : Commit. 0.28909 PPL. 412.27 Recons. 0.00336 +2024-03-30 02:46:56,398 INFO Train. Iter 113200 : Commit. 0.29611 PPL. 415.82 Recons. 0.00308 +2024-03-30 02:47:21,321 INFO Train. Iter 113400 : Commit. 0.28296 PPL. 416.73 Recons. 0.00255 +2024-03-30 02:47:46,243 INFO Train. Iter 113600 : Commit. 0.27982 PPL. 416.84 Recons. 0.00261 +2024-03-30 02:48:11,397 INFO Train. Iter 113800 : Commit. 0.27932 PPL. 416.83 Recons. 0.00248 +2024-03-30 02:48:36,443 INFO Train. Iter 114000 : Commit. 0.29325 PPL. 414.60 Recons. 0.00300 +2024-03-30 02:49:01,169 INFO Train. Iter 114200 : Commit. 0.28176 PPL. 417.09 Recons. 0.00238 +2024-03-30 02:49:26,853 INFO Train. Iter 114400 : Commit. 0.29006 PPL. 413.88 Recons. 0.00321 +2024-03-30 02:49:51,438 INFO Train. Iter 114600 : Commit. 0.28038 PPL. 416.67 Recons. 0.00250 +2024-03-30 02:50:16,356 INFO Train. Iter 114800 : Commit. 0.28502 PPL. 416.61 Recons. 0.00256 +2024-03-30 02:50:40,635 INFO Train. Iter 115000 : Commit. 0.28822 PPL. 414.15 Recons. 0.00305 +2024-03-30 02:51:05,425 INFO Train. Iter 115200 : Commit. 0.30261 PPL. 414.27 Recons. 0.00337 +2024-03-30 02:51:30,121 INFO Train. Iter 115400 : Commit. 0.29306 PPL. 415.03 Recons. 0.00286 +2024-03-30 02:51:54,950 INFO Train. Iter 115600 : Commit. 0.28330 PPL. 417.10 Recons. 0.00242 +2024-03-30 02:52:20,095 INFO Train. Iter 115800 : Commit. 0.27421 PPL. 416.55 Recons. 0.00231 +2024-03-30 02:52:45,318 INFO Train. Iter 116000 : Commit. 0.27241 PPL. 416.49 Recons. 0.00233 +2024-03-30 02:53:10,122 INFO Train. Iter 116200 : Commit. 0.28119 PPL. 415.41 Recons. 0.00275 +2024-03-30 02:53:35,235 INFO Train. Iter 116400 : Commit. 0.27916 PPL. 415.33 Recons. 0.00261 +2024-03-30 02:54:00,303 INFO Train. Iter 116600 : Commit. 0.29173 PPL. 414.02 Recons. 0.00306 +2024-03-30 02:54:25,195 INFO Train. Iter 116800 : Commit. 0.28506 PPL. 414.56 Recons. 0.00276 +2024-03-30 02:54:49,246 INFO Train. Iter 117000 : Commit. 0.28086 PPL. 415.23 Recons. 0.00257 +2024-03-30 02:55:14,531 INFO Train. Iter 117200 : Commit. 0.28673 PPL. 414.86 Recons. 0.00293 +2024-03-30 02:55:40,123 INFO Train. Iter 117400 : Commit. 0.29716 PPL. 414.62 Recons. 0.00312 +2024-03-30 02:56:05,463 INFO Train. Iter 117600 : Commit. 0.28740 PPL. 415.43 Recons. 0.00270 +2024-03-30 02:56:31,064 INFO Train. Iter 117800 : Commit. 0.28616 PPL. 414.54 Recons. 0.00286 +2024-03-30 02:56:56,401 INFO Train. Iter 118000 : Commit. 0.29117 PPL. 413.63 Recons. 0.00301 +2024-03-30 02:57:22,088 INFO Train. Iter 118200 : Commit. 0.28533 PPL. 414.99 Recons. 0.00278 +2024-03-30 02:57:47,976 INFO Train. Iter 118400 : Commit. 0.28881 PPL. 414.86 Recons. 0.00286 +2024-03-30 02:58:13,442 INFO Train. Iter 118600 : Commit. 0.29239 PPL. 413.13 Recons. 0.00313 +2024-03-30 02:58:38,022 INFO Train. Iter 118800 : Commit. 0.27508 PPL. 416.44 Recons. 0.00219 +2024-03-30 02:59:03,796 INFO Train. Iter 119000 : Commit. 0.28597 PPL. 414.13 Recons. 0.00287 +2024-03-30 02:59:29,368 INFO Train. Iter 119200 : Commit. 0.28087 PPL. 414.22 Recons. 0.00257 +2024-03-30 02:59:54,850 INFO Train. Iter 119400 : Commit. 0.29105 PPL. 413.55 Recons. 0.00303 +2024-03-30 03:00:20,153 INFO Train. Iter 119600 : Commit. 0.28637 PPL. 416.16 Recons. 0.00242 +2024-03-30 03:00:45,434 INFO Train. Iter 119800 : Commit. 0.29437 PPL. 413.18 Recons. 0.00328 +2024-03-30 03:01:10,596 INFO Train. Iter 120000 : Commit. 0.28405 PPL. 414.58 Recons. 0.00258 +2024-03-30 03:01:36,052 INFO Train. Iter 120200 : Commit. 0.28557 PPL. 415.53 Recons. 0.00258 +2024-03-30 03:02:01,904 INFO Train. Iter 120400 : Commit. 0.28411 PPL. 415.62 Recons. 0.00259 +2024-03-30 03:02:27,781 INFO Train. Iter 120600 : Commit. 0.28651 PPL. 413.08 Recons. 0.00295 +2024-03-30 03:02:51,636 INFO Train. Iter 120800 : Commit. 0.28743 PPL. 413.53 Recons. 0.00295 +2024-03-30 03:03:16,788 INFO Train. Iter 121000 : Commit. 0.28363 PPL. 414.36 Recons. 0.00286 +2024-03-30 03:03:41,832 INFO Train. Iter 121200 : Commit. 0.28532 PPL. 414.88 Recons. 0.00254 +2024-03-30 03:04:06,791 INFO Train. Iter 121400 : Commit. 0.28129 PPL. 414.04 Recons. 0.00272 +2024-03-30 03:04:31,935 INFO Train. Iter 121600 : Commit. 0.27797 PPL. 413.92 Recons. 0.00268 +2024-03-30 03:04:57,916 INFO Train. Iter 121800 : Commit. 0.28571 PPL. 413.04 Recons. 0.00325 +2024-03-30 03:05:23,167 INFO Train. Iter 122000 : Commit. 0.29010 PPL. 415.05 Recons. 0.00267 +2024-03-30 03:05:48,445 INFO Train. Iter 122200 : Commit. 0.27516 PPL. 414.84 Recons. 0.00242 +2024-03-30 03:06:14,514 INFO Train. Iter 122400 : Commit. 0.27319 PPL. 414.80 Recons. 0.00255 +2024-03-30 03:06:40,284 INFO Train. Iter 122600 : Commit. 0.28304 PPL. 411.82 Recons. 0.00323 +2024-03-30 03:07:05,266 INFO Train. Iter 122800 : Commit. 0.27095 PPL. 415.35 Recons. 0.00223 +2024-03-30 03:07:31,127 INFO Train. Iter 123000 : Commit. 0.28742 PPL. 413.06 Recons. 0.00305 +2024-03-30 03:07:56,840 INFO Train. Iter 123200 : Commit. 0.28500 PPL. 414.64 Recons. 0.00272 +2024-03-30 03:08:22,808 INFO Train. Iter 123400 : Commit. 0.27784 PPL. 415.19 Recons. 0.00240 +2024-03-30 03:08:48,608 INFO Train. Iter 123600 : Commit. 0.28709 PPL. 413.54 Recons. 0.00300 +2024-03-30 03:09:13,483 INFO Train. Iter 123800 : Commit. 0.28522 PPL. 413.60 Recons. 0.00277 +2024-03-30 03:09:38,024 INFO Train. Iter 124000 : Commit. 0.28325 PPL. 413.48 Recons. 0.00279 +2024-03-30 03:10:02,274 INFO Train. Iter 124200 : Commit. 0.28129 PPL. 414.16 Recons. 0.00270 +2024-03-30 03:10:26,382 INFO Train. Iter 124400 : Commit. 0.27973 PPL. 415.02 Recons. 0.00254 +2024-03-30 03:10:49,476 INFO Train. Iter 124600 : Commit. 0.28155 PPL. 413.54 Recons. 0.00267 +2024-03-30 03:11:13,750 INFO Train. Iter 124800 : Commit. 0.28583 PPL. 412.18 Recons. 0.00299 +2024-03-30 03:11:37,857 INFO Train. Iter 125000 : Commit. 0.27231 PPL. 414.96 Recons. 0.00228 +2024-03-30 03:12:01,804 INFO Train. Iter 125200 : Commit. 0.28669 PPL. 414.01 Recons. 0.00284 +2024-03-30 03:12:25,814 INFO Train. Iter 125400 : Commit. 0.27251 PPL. 414.90 Recons. 0.00224 +2024-03-30 03:12:50,437 INFO Train. Iter 125600 : Commit. 0.27536 PPL. 410.26 Recons. 0.00292 +2024-03-30 03:13:15,256 INFO Train. Iter 125800 : Commit. 0.29470 PPL. 413.53 Recons. 0.00294 +2024-03-30 03:13:40,112 INFO Train. Iter 126000 : Commit. 0.27980 PPL. 412.52 Recons. 0.00276 +2024-03-30 03:14:04,637 INFO Train. Iter 126200 : Commit. 0.28115 PPL. 412.65 Recons. 0.00276 +2024-03-30 03:14:29,556 INFO Train. Iter 126400 : Commit. 0.26829 PPL. 414.30 Recons. 0.00220 +2024-03-30 03:14:54,392 INFO Train. Iter 126600 : Commit. 0.27084 PPL. 414.27 Recons. 0.00240 +2024-03-30 03:15:20,305 INFO Train. Iter 126800 : Commit. 0.28632 PPL. 412.92 Recons. 0.00295 +2024-03-30 03:15:46,379 INFO Train. Iter 127000 : Commit. 0.28575 PPL. 413.11 Recons. 0.00306 +2024-03-30 03:16:11,581 INFO Train. Iter 127200 : Commit. 0.27952 PPL. 414.46 Recons. 0.00246 +2024-03-30 03:16:36,948 INFO Train. Iter 127400 : Commit. 0.28250 PPL. 412.54 Recons. 0.00277 +2024-03-30 03:17:02,183 INFO Train. Iter 127600 : Commit. 0.27674 PPL. 413.49 Recons. 0.00265 +2024-03-30 03:17:26,064 INFO Train. Iter 127800 : Commit. 0.27594 PPL. 413.49 Recons. 0.00261 +2024-03-30 03:17:51,314 INFO Train. Iter 128000 : Commit. 0.27554 PPL. 413.56 Recons. 0.00260 +2024-03-30 03:18:15,876 INFO Train. Iter 128200 : Commit. 0.28287 PPL. 413.18 Recons. 0.00273 +2024-03-30 03:18:41,260 INFO Train. Iter 128400 : Commit. 0.27798 PPL. 412.48 Recons. 0.00266 +2024-03-30 03:19:05,169 INFO Train. Iter 128600 : Commit. 0.28045 PPL. 413.62 Recons. 0.00254 +2024-03-30 03:19:29,741 INFO Train. Iter 128800 : Commit. 0.28928 PPL. 412.11 Recons. 0.00294 +2024-03-30 03:19:54,693 INFO Train. Iter 129000 : Commit. 0.27699 PPL. 413.54 Recons. 0.00254 +2024-03-30 03:20:20,361 INFO Train. Iter 129200 : Commit. 0.27394 PPL. 414.20 Recons. 0.00242 +2024-03-30 03:20:45,371 INFO Train. Iter 129400 : Commit. 0.27776 PPL. 414.01 Recons. 0.00251 +2024-03-30 03:21:10,069 INFO Train. Iter 129600 : Commit. 0.27946 PPL. 412.70 Recons. 0.00279 +2024-03-30 03:21:34,797 INFO Train. Iter 129800 : Commit. 0.28099 PPL. 412.85 Recons. 0.00275 +2024-03-30 03:21:59,819 INFO Train. Iter 130000 : Commit. 0.28881 PPL. 413.13 Recons. 0.00290 +2024-03-30 03:22:24,617 INFO Train. Iter 130200 : Commit. 0.27394 PPL. 412.86 Recons. 0.00253 +2024-03-30 03:22:48,619 INFO Train. Iter 130400 : Commit. 0.27449 PPL. 414.24 Recons. 0.00250 +2024-03-30 03:23:14,197 INFO Train. Iter 130600 : Commit. 0.27591 PPL. 414.33 Recons. 0.00253 +2024-03-30 03:23:39,482 INFO Train. Iter 130800 : Commit. 0.28616 PPL. 411.41 Recons. 0.00321 +2024-03-30 03:24:04,769 INFO Train. Iter 131000 : Commit. 0.28395 PPL. 413.32 Recons. 0.00273 +2024-03-30 03:24:30,393 INFO Train. Iter 131200 : Commit. 0.27914 PPL. 414.27 Recons. 0.00262 +2024-03-30 03:24:56,203 INFO Train. Iter 131400 : Commit. 0.28134 PPL. 413.41 Recons. 0.00272 +2024-03-30 03:25:21,807 INFO Train. Iter 131600 : Commit. 0.27661 PPL. 412.98 Recons. 0.00268 +2024-03-30 03:25:47,517 INFO Train. Iter 131800 : Commit. 0.27257 PPL. 413.16 Recons. 0.00265 +2024-03-30 03:26:13,300 INFO Train. Iter 132000 : Commit. 0.27061 PPL. 413.58 Recons. 0.00242 +2024-03-30 03:26:38,934 INFO Train. Iter 132200 : Commit. 0.27469 PPL. 413.50 Recons. 0.00256 +2024-03-30 03:27:03,175 INFO Train. Iter 132400 : Commit. 0.26932 PPL. 413.70 Recons. 0.00234 +2024-03-30 03:27:27,784 INFO Train. Iter 132600 : Commit. 0.28706 PPL. 412.31 Recons. 0.00302 +2024-03-30 03:27:52,491 INFO Train. Iter 132800 : Commit. 0.27877 PPL. 412.98 Recons. 0.00265 +2024-03-30 03:28:17,511 INFO Train. Iter 133000 : Commit. 0.26664 PPL. 413.73 Recons. 0.00216 +2024-03-30 03:28:42,754 INFO Train. Iter 133200 : Commit. 0.27422 PPL. 412.76 Recons. 0.00273 +2024-03-30 03:29:07,560 INFO Train. Iter 133400 : Commit. 0.28265 PPL. 412.43 Recons. 0.00265 +2024-03-30 03:29:32,567 INFO Train. Iter 133600 : Commit. 0.27015 PPL. 412.92 Recons. 0.00245 +2024-03-30 03:29:57,809 INFO Train. Iter 133800 : Commit. 0.27713 PPL. 411.74 Recons. 0.00260 +2024-03-30 03:30:23,500 INFO Train. Iter 134000 : Commit. 0.27141 PPL. 414.34 Recons. 0.00233 +2024-03-30 03:30:48,621 INFO Train. Iter 134200 : Commit. 0.27327 PPL. 412.52 Recons. 0.00248 +2024-03-30 03:31:12,820 INFO Train. Iter 134400 : Commit. 0.27701 PPL. 411.43 Recons. 0.00271 +2024-03-30 03:31:38,773 INFO Train. Iter 134600 : Commit. 0.27582 PPL. 412.83 Recons. 0.00267 +2024-03-30 03:32:04,245 INFO Train. Iter 134800 : Commit. 0.26583 PPL. 413.99 Recons. 0.00218 +2024-03-30 03:32:29,734 INFO Train. Iter 135000 : Commit. 0.28333 PPL. 412.08 Recons. 0.00304 +2024-03-30 03:32:54,708 INFO Train. Iter 135200 : Commit. 0.27713 PPL. 413.05 Recons. 0.00242 +2024-03-30 03:33:20,114 INFO Train. Iter 135400 : Commit. 0.27785 PPL. 412.31 Recons. 0.00268 +2024-03-30 03:33:45,415 INFO Train. Iter 135600 : Commit. 0.28132 PPL. 412.09 Recons. 0.00272 +2024-03-30 03:34:11,127 INFO Train. Iter 135800 : Commit. 0.26988 PPL. 411.78 Recons. 0.00256 +2024-03-30 03:34:36,603 INFO Train. Iter 136000 : Commit. 0.27142 PPL. 412.66 Recons. 0.00232 +2024-03-30 03:35:00,937 INFO Train. Iter 136200 : Commit. 0.27843 PPL. 410.43 Recons. 0.00274 +2024-03-30 03:35:26,133 INFO Train. Iter 136400 : Commit. 0.27830 PPL. 413.17 Recons. 0.00251 +2024-03-30 03:35:51,544 INFO Train. Iter 136600 : Commit. 0.27547 PPL. 412.87 Recons. 0.00254 +2024-03-30 03:36:16,786 INFO Train. Iter 136800 : Commit. 0.27917 PPL. 411.97 Recons. 0.00274 +2024-03-30 03:36:41,815 INFO Train. Iter 137000 : Commit. 0.27829 PPL. 414.05 Recons. 0.00250 +2024-03-30 03:37:07,172 INFO Train. Iter 137200 : Commit. 0.27373 PPL. 412.48 Recons. 0.00256 +2024-03-30 03:37:32,009 INFO Train. Iter 137400 : Commit. 0.27703 PPL. 411.81 Recons. 0.00283 +2024-03-30 03:37:57,181 INFO Train. Iter 137600 : Commit. 0.27970 PPL. 412.23 Recons. 0.00280 +2024-03-30 03:38:22,468 INFO Train. Iter 137800 : Commit. 0.27066 PPL. 411.67 Recons. 0.00261 +2024-03-30 03:38:47,692 INFO Train. Iter 138000 : Commit. 0.27217 PPL. 411.88 Recons. 0.00254 +2024-03-30 03:39:11,777 INFO Train. Iter 138200 : Commit. 0.27018 PPL. 413.38 Recons. 0.00223 +2024-03-30 03:39:36,927 INFO Train. Iter 138400 : Commit. 0.27691 PPL. 409.81 Recons. 0.00312 +2024-03-30 03:40:02,158 INFO Train. Iter 138600 : Commit. 0.26908 PPL. 413.60 Recons. 0.00216 +2024-03-30 03:40:27,702 INFO Train. Iter 138800 : Commit. 0.28318 PPL. 411.26 Recons. 0.00297 +2024-03-30 03:40:52,709 INFO Train. Iter 139000 : Commit. 0.27156 PPL. 412.41 Recons. 0.00243 +2024-03-30 03:41:17,699 INFO Train. Iter 139200 : Commit. 0.27122 PPL. 413.32 Recons. 0.00234 +2024-03-30 03:41:42,949 INFO Train. Iter 139400 : Commit. 0.26992 PPL. 412.33 Recons. 0.00257 +2024-03-30 03:42:08,200 INFO Train. Iter 139600 : Commit. 0.26654 PPL. 413.17 Recons. 0.00229 +2024-03-30 03:42:33,367 INFO Train. Iter 139800 : Commit. 0.27419 PPL. 408.92 Recons. 0.00282 +2024-03-30 03:42:58,160 INFO Train. Iter 140000 : Commit. 0.27511 PPL. 411.18 Recons. 0.00264 +2024-03-30 03:43:22,347 INFO Train. Iter 140200 : Commit. 0.27440 PPL. 412.96 Recons. 0.00247 +2024-03-30 03:43:46,886 INFO Train. Iter 140400 : Commit. 0.27033 PPL. 413.16 Recons. 0.00228 +2024-03-30 03:44:12,587 INFO Train. Iter 140600 : Commit. 0.26954 PPL. 412.75 Recons. 0.00251 +2024-03-30 03:44:37,635 INFO Train. Iter 140800 : Commit. 0.27657 PPL. 412.17 Recons. 0.00261 +2024-03-30 03:45:03,282 INFO Train. Iter 141000 : Commit. 0.26004 PPL. 412.83 Recons. 0.00208 +2024-03-30 03:45:28,386 INFO Train. Iter 141200 : Commit. 0.27935 PPL. 409.88 Recons. 0.00271 +2024-03-30 03:45:53,648 INFO Train. Iter 141400 : Commit. 0.27795 PPL. 410.30 Recons. 0.00290 +2024-03-30 03:46:19,305 INFO Train. Iter 141600 : Commit. 0.26861 PPL. 411.06 Recons. 0.00252 +2024-03-30 03:46:44,303 INFO Train. Iter 141800 : Commit. 0.28133 PPL. 411.83 Recons. 0.00280 +2024-03-30 03:47:08,523 INFO Train. Iter 142000 : Commit. 0.27554 PPL. 412.19 Recons. 0.00263 +2024-03-30 03:47:33,974 INFO Train. Iter 142200 : Commit. 0.27980 PPL. 411.50 Recons. 0.00287 +2024-03-30 03:47:58,451 INFO Train. Iter 142400 : Commit. 0.27916 PPL. 411.14 Recons. 0.00277 +2024-03-30 03:48:23,612 INFO Train. Iter 142600 : Commit. 0.26517 PPL. 413.20 Recons. 0.00213 +2024-03-30 03:48:48,245 INFO Train. Iter 142800 : Commit. 0.27825 PPL. 410.43 Recons. 0.00281 +2024-03-30 03:49:13,016 INFO Train. Iter 143000 : Commit. 0.27353 PPL. 411.28 Recons. 0.00265 +2024-03-30 03:49:37,980 INFO Train. Iter 143200 : Commit. 0.27830 PPL. 411.42 Recons. 0.00275 +2024-03-30 03:50:02,903 INFO Train. Iter 143400 : Commit. 0.27012 PPL. 412.11 Recons. 0.00242 +2024-03-30 03:50:27,627 INFO Train. Iter 143600 : Commit. 0.26712 PPL. 413.43 Recons. 0.00219 +2024-03-30 03:50:52,721 INFO Train. Iter 143800 : Commit. 0.27021 PPL. 412.41 Recons. 0.00246 +2024-03-30 03:51:17,022 INFO Train. Iter 144000 : Commit. 0.26960 PPL. 411.67 Recons. 0.00263 +2024-03-30 03:51:42,638 INFO Train. Iter 144200 : Commit. 0.27147 PPL. 411.74 Recons. 0.00247 +2024-03-30 03:52:08,762 INFO Train. Iter 144400 : Commit. 0.26461 PPL. 411.88 Recons. 0.00228 +2024-03-30 03:52:34,246 INFO Train. Iter 144600 : Commit. 0.27983 PPL. 411.34 Recons. 0.00272 +2024-03-30 03:53:00,239 INFO Train. Iter 144800 : Commit. 0.27606 PPL. 411.73 Recons. 0.00270 +2024-03-30 03:53:25,642 INFO Train. Iter 145000 : Commit. 0.26832 PPL. 411.79 Recons. 0.00243 +2024-03-30 03:53:51,247 INFO Train. Iter 145200 : Commit. 0.27639 PPL. 411.51 Recons. 0.00266 +2024-03-30 03:54:17,191 INFO Train. Iter 145400 : Commit. 0.27062 PPL. 411.27 Recons. 0.00256 +2024-03-30 03:54:42,719 INFO Train. Iter 145600 : Commit. 0.26783 PPL. 412.04 Recons. 0.00253 +2024-03-30 03:55:07,841 INFO Train. Iter 145800 : Commit. 0.26729 PPL. 412.53 Recons. 0.00226 +2024-03-30 03:55:31,406 INFO Train. Iter 146000 : Commit. 0.28359 PPL. 410.56 Recons. 0.00287 +2024-03-30 03:55:56,054 INFO Train. Iter 146200 : Commit. 0.26684 PPL. 412.40 Recons. 0.00235 +2024-03-30 03:56:20,665 INFO Train. Iter 146400 : Commit. 0.28103 PPL. 409.47 Recons. 0.00301 +2024-03-30 03:56:46,514 INFO Train. Iter 146600 : Commit. 0.27988 PPL. 409.92 Recons. 0.00300 +2024-03-30 03:57:12,072 INFO Train. Iter 146800 : Commit. 0.28027 PPL. 410.80 Recons. 0.00280 +2024-03-30 03:57:37,439 INFO Train. Iter 147000 : Commit. 0.26715 PPL. 411.91 Recons. 0.00226 +2024-03-30 03:58:03,485 INFO Train. Iter 147200 : Commit. 0.27048 PPL. 410.28 Recons. 0.00271 +2024-03-30 03:58:28,776 INFO Train. Iter 147400 : Commit. 0.26836 PPL. 412.27 Recons. 0.00238 +2024-03-30 03:58:54,038 INFO Train. Iter 147600 : Commit. 0.25944 PPL. 413.22 Recons. 0.00209 +2024-03-30 03:59:18,221 INFO Train. Iter 147800 : Commit. 0.27499 PPL. 409.45 Recons. 0.00282 +2024-03-30 03:59:43,032 INFO Train. Iter 148000 : Commit. 0.26388 PPL. 412.11 Recons. 0.00227 +2024-03-30 04:00:07,848 INFO Train. Iter 148200 : Commit. 0.28027 PPL. 410.91 Recons. 0.00286 +2024-03-30 04:00:33,129 INFO Train. Iter 148400 : Commit. 0.27318 PPL. 412.48 Recons. 0.00239 +2024-03-30 04:00:58,381 INFO Train. Iter 148600 : Commit. 0.26676 PPL. 412.34 Recons. 0.00225 +2024-03-30 04:01:23,560 INFO Train. Iter 148800 : Commit. 0.26751 PPL. 412.04 Recons. 0.00235 +2024-03-30 04:01:48,174 INFO Train. Iter 149000 : Commit. 0.27044 PPL. 410.13 Recons. 0.00271 +2024-03-30 04:02:14,047 INFO Train. Iter 149200 : Commit. 0.26408 PPL. 412.95 Recons. 0.00204 +2024-03-30 04:02:39,278 INFO Train. Iter 149400 : Commit. 0.28126 PPL. 410.35 Recons. 0.00287 +2024-03-30 04:03:05,024 INFO Train. Iter 149600 : Commit. 0.26380 PPL. 411.96 Recons. 0.00219 +2024-03-30 04:03:29,982 INFO Train. Iter 149800 : Commit. 0.26537 PPL. 411.49 Recons. 0.00234 +2024-03-30 04:03:55,878 INFO Train. Iter 150000 : Commit. 0.27995 PPL. 410.21 Recons. 0.00301 +2024-03-30 04:04:21,527 INFO Train. Iter 150200 : Commit. 0.26789 PPL. 412.65 Recons. 0.00216 +2024-03-30 04:04:47,225 INFO Train. Iter 150400 : Commit. 0.27247 PPL. 409.16 Recons. 0.00281 +2024-03-30 04:05:12,629 INFO Train. Iter 150600 : Commit. 0.27068 PPL. 410.77 Recons. 0.00250 +2024-03-30 04:05:38,427 INFO Train. Iter 150800 : Commit. 0.26956 PPL. 411.25 Recons. 0.00240 +2024-03-30 04:06:04,745 INFO Train. Iter 151000 : Commit. 0.27998 PPL. 411.01 Recons. 0.00266 +2024-03-30 04:06:30,581 INFO Train. Iter 151200 : Commit. 0.26523 PPL. 411.28 Recons. 0.00236 +2024-03-30 04:06:56,102 INFO Train. Iter 151400 : Commit. 0.27190 PPL. 410.78 Recons. 0.00245 +2024-03-30 04:07:21,194 INFO Train. Iter 151600 : Commit. 0.26883 PPL. 411.72 Recons. 0.00245 +2024-03-30 04:07:44,974 INFO Train. Iter 151800 : Commit. 0.26886 PPL. 410.92 Recons. 0.00248 +2024-03-30 04:08:09,853 INFO Train. Iter 152000 : Commit. 0.26478 PPL. 412.19 Recons. 0.00218 +2024-03-30 04:08:34,820 INFO Train. Iter 152200 : Commit. 0.27264 PPL. 409.92 Recons. 0.00263 +2024-03-30 04:09:00,333 INFO Train. Iter 152400 : Commit. 0.26420 PPL. 412.44 Recons. 0.00213 +2024-03-30 04:09:26,007 INFO Train. Iter 152600 : Commit. 0.26976 PPL. 410.91 Recons. 0.00261 +2024-03-30 04:09:50,763 INFO Train. Iter 152800 : Commit. 0.26134 PPL. 411.41 Recons. 0.00237 +2024-03-30 04:10:16,155 INFO Train. Iter 153000 : Commit. 0.27546 PPL. 409.96 Recons. 0.00260 +2024-03-30 04:10:41,150 INFO Train. Iter 153200 : Commit. 0.25975 PPL. 412.40 Recons. 0.00211 +2024-03-30 04:11:06,221 INFO Train. Iter 153400 : Commit. 0.26795 PPL. 412.74 Recons. 0.00229 +2024-03-30 04:11:30,226 INFO Train. Iter 153600 : Commit. 0.26835 PPL. 409.46 Recons. 0.00255 +2024-03-30 04:11:55,595 INFO Train. Iter 153800 : Commit. 0.26791 PPL. 410.68 Recons. 0.00245 +2024-03-30 04:12:21,004 INFO Train. Iter 154000 : Commit. 0.25937 PPL. 411.07 Recons. 0.00230 +2024-03-30 04:12:46,203 INFO Train. Iter 154200 : Commit. 0.27108 PPL. 409.37 Recons. 0.00276 +2024-03-30 04:13:11,559 INFO Train. Iter 154400 : Commit. 0.26242 PPL. 412.11 Recons. 0.00218 +2024-03-30 04:13:36,520 INFO Train. Iter 154600 : Commit. 0.26539 PPL. 411.05 Recons. 0.00235 +2024-03-30 04:14:01,666 INFO Train. Iter 154800 : Commit. 0.26562 PPL. 411.15 Recons. 0.00242 +2024-03-30 04:14:26,999 INFO Train. Iter 155000 : Commit. 0.26727 PPL. 410.78 Recons. 0.00250 +2024-03-30 04:14:51,639 INFO Train. Iter 155200 : Commit. 0.27099 PPL. 411.34 Recons. 0.00243 +2024-03-30 04:15:16,269 INFO Train. Iter 155400 : Commit. 0.26066 PPL. 412.18 Recons. 0.00217 +2024-03-30 04:15:40,111 INFO Train. Iter 155600 : Commit. 0.26269 PPL. 410.95 Recons. 0.00251 +2024-03-30 04:16:05,712 INFO Train. Iter 155800 : Commit. 0.26416 PPL. 411.95 Recons. 0.00229 +2024-03-30 04:16:30,993 INFO Train. Iter 156000 : Commit. 0.27893 PPL. 409.79 Recons. 0.00293 +2024-03-30 04:16:56,038 INFO Train. Iter 156200 : Commit. 0.27373 PPL. 411.84 Recons. 0.00249 +2024-03-30 04:17:21,848 INFO Train. Iter 156400 : Commit. 0.26671 PPL. 411.93 Recons. 0.00227 +2024-03-30 04:17:46,464 INFO Train. Iter 156600 : Commit. 0.26589 PPL. 410.60 Recons. 0.00253 +2024-03-30 04:18:11,569 INFO Train. Iter 156800 : Commit. 0.27567 PPL. 411.35 Recons. 0.00273 +2024-03-30 04:18:36,636 INFO Train. Iter 157000 : Commit. 0.26068 PPL. 411.29 Recons. 0.00222 +2024-03-30 04:19:01,744 INFO Train. Iter 157200 : Commit. 0.25590 PPL. 412.54 Recons. 0.00200 +2024-03-30 04:19:26,555 INFO Train. Iter 157400 : Commit. 0.26554 PPL. 410.07 Recons. 0.00255 +2024-03-30 04:19:50,475 INFO Train. Iter 157600 : Commit. 0.26531 PPL. 411.20 Recons. 0.00233 +2024-03-30 04:20:15,936 INFO Train. Iter 157800 : Commit. 0.26081 PPL. 411.24 Recons. 0.00230 +2024-03-30 04:20:41,809 INFO Train. Iter 158000 : Commit. 0.26193 PPL. 410.73 Recons. 0.00238 +2024-03-30 04:21:06,984 INFO Train. Iter 158200 : Commit. 0.27407 PPL. 410.00 Recons. 0.00271 +2024-03-30 04:21:32,366 INFO Train. Iter 158400 : Commit. 0.26443 PPL. 411.47 Recons. 0.00223 +2024-03-30 04:21:57,092 INFO Train. Iter 158600 : Commit. 0.26061 PPL. 411.65 Recons. 0.00235 +2024-03-30 04:22:21,465 INFO Train. Iter 158800 : Commit. 0.25407 PPL. 411.57 Recons. 0.00197 +2024-03-30 04:22:46,379 INFO Train. Iter 159000 : Commit. 0.27668 PPL. 407.88 Recons. 0.00303 +2024-03-30 04:23:11,755 INFO Train. Iter 159200 : Commit. 0.26734 PPL. 411.89 Recons. 0.00225 +2024-03-30 04:23:35,766 INFO Train. Iter 159400 : Commit. 0.27211 PPL. 410.59 Recons. 0.00265 +2024-03-30 04:24:00,835 INFO Train. Iter 159600 : Commit. 0.26479 PPL. 410.48 Recons. 0.00245 +2024-03-30 04:24:26,136 INFO Train. Iter 159800 : Commit. 0.25864 PPL. 411.76 Recons. 0.00215 +2024-03-30 04:24:50,655 INFO Train. Iter 160000 : Commit. 0.27213 PPL. 409.24 Recons. 0.00286 +2024-03-30 04:25:15,632 INFO Train. Iter 160200 : Commit. 0.26874 PPL. 410.29 Recons. 0.00259 +2024-03-30 04:25:40,798 INFO Train. Iter 160400 : Commit. 0.26744 PPL. 411.55 Recons. 0.00240 +2024-03-30 04:26:04,821 INFO Train. Iter 160600 : Commit. 0.26998 PPL. 410.24 Recons. 0.00256 +2024-03-30 04:26:28,754 INFO Train. Iter 160800 : Commit. 0.27168 PPL. 409.93 Recons. 0.00267 +2024-03-30 04:26:52,594 INFO Train. Iter 161000 : Commit. 0.26462 PPL. 411.86 Recons. 0.00229 +2024-03-30 04:27:16,233 INFO Train. Iter 161200 : Commit. 0.26325 PPL. 409.72 Recons. 0.00242 +2024-03-30 04:27:39,622 INFO Train. Iter 161400 : Commit. 0.26207 PPL. 410.59 Recons. 0.00227 +2024-03-30 04:28:03,961 INFO Train. Iter 161600 : Commit. 0.26852 PPL. 410.57 Recons. 0.00252 +2024-03-30 04:28:27,912 INFO Train. Iter 161800 : Commit. 0.26159 PPL. 411.77 Recons. 0.00216 +2024-03-30 04:28:51,855 INFO Train. Iter 162000 : Commit. 0.26705 PPL. 410.90 Recons. 0.00245 +2024-03-30 04:29:15,958 INFO Train. Iter 162200 : Commit. 0.27319 PPL. 411.20 Recons. 0.00256 +2024-03-30 04:29:40,906 INFO Train. Iter 162400 : Commit. 0.26774 PPL. 410.76 Recons. 0.00250 +2024-03-30 04:30:06,071 INFO Train. Iter 162600 : Commit. 0.26371 PPL. 410.39 Recons. 0.00233 +2024-03-30 04:30:31,388 INFO Train. Iter 162800 : Commit. 0.25767 PPL. 411.61 Recons. 0.00218 +2024-03-30 04:30:56,684 INFO Train. Iter 163000 : Commit. 0.26176 PPL. 410.10 Recons. 0.00230 +2024-03-30 04:31:21,705 INFO Train. Iter 163200 : Commit. 0.27231 PPL. 411.11 Recons. 0.00245 +2024-03-30 04:31:45,724 INFO Train. Iter 163400 : Commit. 0.26362 PPL. 409.66 Recons. 0.00225 +2024-03-30 04:32:10,881 INFO Train. Iter 163600 : Commit. 0.26828 PPL. 411.57 Recons. 0.00233 +2024-03-30 04:32:36,885 INFO Train. Iter 163800 : Commit. 0.26088 PPL. 410.36 Recons. 0.00225 +2024-03-30 04:33:02,782 INFO Train. Iter 164000 : Commit. 0.26320 PPL. 409.67 Recons. 0.00250 +2024-03-30 04:33:27,237 INFO Train. Iter 164200 : Commit. 0.26644 PPL. 411.03 Recons. 0.00238 +2024-03-30 04:33:52,307 INFO Train. Iter 164400 : Commit. 0.25810 PPL. 411.43 Recons. 0.00214 +2024-03-30 04:34:17,175 INFO Train. Iter 164600 : Commit. 0.26308 PPL. 410.95 Recons. 0.00234 +2024-03-30 04:34:41,696 INFO Train. Iter 164800 : Commit. 0.25923 PPL. 410.95 Recons. 0.00218 +2024-03-30 04:35:06,714 INFO Train. Iter 165000 : Commit. 0.27237 PPL. 410.20 Recons. 0.00261 +2024-03-30 04:35:30,569 INFO Train. Iter 165200 : Commit. 0.26761 PPL. 409.57 Recons. 0.00254 +2024-03-30 04:35:55,376 INFO Train. Iter 165400 : Commit. 0.26671 PPL. 411.35 Recons. 0.00234 +2024-03-30 04:36:20,449 INFO Train. Iter 165600 : Commit. 0.25865 PPL. 411.82 Recons. 0.00211 +2024-03-30 04:36:45,234 INFO Train. Iter 165800 : Commit. 0.26345 PPL. 411.04 Recons. 0.00239 +2024-03-30 04:37:11,051 INFO Train. Iter 166000 : Commit. 0.26411 PPL. 410.84 Recons. 0.00232 +2024-03-30 04:37:35,792 INFO Train. Iter 166200 : Commit. 0.26022 PPL. 411.44 Recons. 0.00221 +2024-03-30 04:37:59,924 INFO Train. Iter 166400 : Commit. 0.25439 PPL. 411.47 Recons. 0.00207 +2024-03-30 04:38:24,646 INFO Train. Iter 166600 : Commit. 0.25865 PPL. 411.38 Recons. 0.00221 +2024-03-30 04:38:49,309 INFO Train. Iter 166800 : Commit. 0.26317 PPL. 411.14 Recons. 0.00232 +2024-03-30 04:39:14,042 INFO Train. Iter 167000 : Commit. 0.26855 PPL. 410.45 Recons. 0.00253 +2024-03-30 04:39:38,681 INFO Train. Iter 167200 : Commit. 0.25642 PPL. 410.91 Recons. 0.00216 +2024-03-30 04:40:03,740 INFO Train. Iter 167400 : Commit. 0.26591 PPL. 408.67 Recons. 0.00272 +2024-03-30 04:40:28,549 INFO Train. Iter 167600 : Commit. 0.25997 PPL. 411.47 Recons. 0.00219 +2024-03-30 04:40:53,185 INFO Train. Iter 167800 : Commit. 0.25988 PPL. 410.52 Recons. 0.00225 +2024-03-30 04:41:18,310 INFO Train. Iter 168000 : Commit. 0.27309 PPL. 409.09 Recons. 0.00280 +2024-03-30 04:41:42,684 INFO Train. Iter 168200 : Commit. 0.26291 PPL. 410.65 Recons. 0.00225 +2024-03-30 04:42:07,770 INFO Train. Iter 168400 : Commit. 0.26477 PPL. 410.43 Recons. 0.00258 +2024-03-30 04:42:32,937 INFO Train. Iter 168600 : Commit. 0.25832 PPL. 410.71 Recons. 0.00210 +2024-03-30 04:42:57,638 INFO Train. Iter 168800 : Commit. 0.27176 PPL. 409.65 Recons. 0.00262 +2024-03-30 04:43:23,283 INFO Train. Iter 169000 : Commit. 0.26388 PPL. 410.68 Recons. 0.00239 +2024-03-30 04:43:47,539 INFO Train. Iter 169200 : Commit. 0.26202 PPL. 411.75 Recons. 0.00226 +2024-03-30 04:44:13,151 INFO Train. Iter 169400 : Commit. 0.26547 PPL. 409.70 Recons. 0.00266 +2024-03-30 04:44:39,122 INFO Train. Iter 169600 : Commit. 0.26442 PPL. 411.14 Recons. 0.00244 +2024-03-30 04:45:04,328 INFO Train. Iter 169800 : Commit. 0.26071 PPL. 409.69 Recons. 0.00259 +2024-03-30 04:45:29,944 INFO Train. Iter 170000 : Commit. 0.25011 PPL. 410.94 Recons. 0.00197 +2024-03-30 04:45:55,787 INFO Train. Iter 170200 : Commit. 0.26438 PPL. 409.99 Recons. 0.00257 +2024-03-30 04:46:20,336 INFO Train. Iter 170400 : Commit. 0.26074 PPL. 410.82 Recons. 0.00223 +2024-03-30 04:46:46,169 INFO Train. Iter 170600 : Commit. 0.27242 PPL. 409.03 Recons. 0.00288 +2024-03-30 04:47:11,481 INFO Train. Iter 170800 : Commit. 0.26003 PPL. 409.52 Recons. 0.00227 +2024-03-30 04:47:35,291 INFO Train. Iter 171000 : Commit. 0.26397 PPL. 411.51 Recons. 0.00218 +2024-03-30 04:48:00,691 INFO Train. Iter 171200 : Commit. 0.26404 PPL. 411.63 Recons. 0.00218 +2024-03-30 04:48:25,095 INFO Train. Iter 171400 : Commit. 0.27759 PPL. 408.71 Recons. 0.00305 +2024-03-30 04:48:50,062 INFO Train. Iter 171600 : Commit. 0.26493 PPL. 410.70 Recons. 0.00227 +2024-03-30 04:49:14,845 INFO Train. Iter 171800 : Commit. 0.25458 PPL. 411.38 Recons. 0.00203 +2024-03-30 04:49:39,971 INFO Train. Iter 172000 : Commit. 0.25871 PPL. 411.50 Recons. 0.00221 +2024-03-30 04:50:05,211 INFO Train. Iter 172200 : Commit. 0.26309 PPL. 409.94 Recons. 0.00245 +2024-03-30 04:50:30,950 INFO Train. Iter 172400 : Commit. 0.26404 PPL. 408.79 Recons. 0.00242 +2024-03-30 04:50:56,491 INFO Train. Iter 172600 : Commit. 0.26558 PPL. 409.99 Recons. 0.00241 +2024-03-30 04:51:22,428 INFO Train. Iter 172800 : Commit. 0.26365 PPL. 408.38 Recons. 0.00250 +2024-03-30 04:51:47,411 INFO Train. Iter 173000 : Commit. 0.25812 PPL. 411.19 Recons. 0.00205 +2024-03-30 04:52:12,891 INFO Train. Iter 173200 : Commit. 0.27248 PPL. 409.14 Recons. 0.00279 +2024-03-30 04:52:38,005 INFO Train. Iter 173400 : Commit. 0.27221 PPL. 410.03 Recons. 0.00247 +2024-03-30 04:53:02,928 INFO Train. Iter 173600 : Commit. 0.25964 PPL. 409.95 Recons. 0.00229 +2024-03-30 04:53:28,057 INFO Train. Iter 173800 : Commit. 0.25916 PPL. 411.04 Recons. 0.00219 +2024-03-30 04:53:53,958 INFO Train. Iter 174000 : Commit. 0.26460 PPL. 410.58 Recons. 0.00245 +2024-03-30 04:54:19,677 INFO Train. Iter 174200 : Commit. 0.25912 PPL. 410.16 Recons. 0.00234 +2024-03-30 04:54:44,703 INFO Train. Iter 174400 : Commit. 0.26360 PPL. 409.76 Recons. 0.00227 +2024-03-30 04:55:10,192 INFO Train. Iter 174600 : Commit. 0.25618 PPL. 411.18 Recons. 0.00218 +2024-03-30 04:55:35,793 INFO Train. Iter 174800 : Commit. 0.26446 PPL. 408.91 Recons. 0.00256 +2024-03-30 04:55:59,829 INFO Train. Iter 175000 : Commit. 0.26710 PPL. 410.28 Recons. 0.00235 +2024-03-30 04:56:25,001 INFO Train. Iter 175200 : Commit. 0.26267 PPL. 410.55 Recons. 0.00234 +2024-03-30 04:56:50,294 INFO Train. Iter 175400 : Commit. 0.26393 PPL. 410.19 Recons. 0.00242 +2024-03-30 04:57:16,246 INFO Train. Iter 175600 : Commit. 0.26006 PPL. 410.14 Recons. 0.00229 +2024-03-30 04:57:41,615 INFO Train. Iter 175800 : Commit. 0.25779 PPL. 410.28 Recons. 0.00218 +2024-03-30 04:58:07,799 INFO Train. Iter 176000 : Commit. 0.25660 PPL. 410.66 Recons. 0.00204 +2024-03-30 04:58:33,651 INFO Train. Iter 176200 : Commit. 0.25454 PPL. 410.61 Recons. 0.00212 +2024-03-30 04:58:59,387 INFO Train. Iter 176400 : Commit. 0.25158 PPL. 410.02 Recons. 0.00210 +2024-03-30 04:59:25,770 INFO Train. Iter 176600 : Commit. 0.26546 PPL. 408.54 Recons. 0.00269 +2024-03-30 04:59:50,680 INFO Train. Iter 176800 : Commit. 0.26115 PPL. 411.59 Recons. 0.00203 +2024-03-30 05:00:15,627 INFO Train. Iter 177000 : Commit. 0.26046 PPL. 411.30 Recons. 0.00216 +2024-03-30 05:00:40,484 INFO Train. Iter 177200 : Commit. 0.25676 PPL. 408.87 Recons. 0.00233 +2024-03-30 05:01:05,337 INFO Train. Iter 177400 : Commit. 0.25854 PPL. 410.73 Recons. 0.00207 +2024-03-30 05:01:29,630 INFO Train. Iter 177600 : Commit. 0.25441 PPL. 410.17 Recons. 0.00209 +2024-03-30 05:01:54,621 INFO Train. Iter 177800 : Commit. 0.26662 PPL. 409.34 Recons. 0.00279 +2024-03-30 05:02:19,248 INFO Train. Iter 178000 : Commit. 0.25533 PPL. 411.15 Recons. 0.00194 +2024-03-30 05:02:44,175 INFO Train. Iter 178200 : Commit. 0.25867 PPL. 408.80 Recons. 0.00241 +2024-03-30 05:03:09,070 INFO Train. Iter 178400 : Commit. 0.25961 PPL. 409.32 Recons. 0.00226 +2024-03-30 05:03:34,306 INFO Train. Iter 178600 : Commit. 0.27209 PPL. 408.51 Recons. 0.00273 +2024-03-30 05:03:58,672 INFO Train. Iter 178800 : Commit. 0.25103 PPL. 410.25 Recons. 0.00193 +2024-03-30 05:04:23,862 INFO Train. Iter 179000 : Commit. 0.26522 PPL. 409.51 Recons. 0.00250 +2024-03-30 05:04:48,362 INFO Train. Iter 179200 : Commit. 0.25729 PPL. 409.27 Recons. 0.00217 +2024-03-30 05:05:13,185 INFO Train. Iter 179400 : Commit. 0.25543 PPL. 411.26 Recons. 0.00201 +2024-03-30 05:05:38,352 INFO Train. Iter 179600 : Commit. 0.26458 PPL. 408.39 Recons. 0.00253 +2024-03-30 05:06:03,354 INFO Train. Iter 179800 : Commit. 0.25687 PPL. 408.08 Recons. 0.00237 +2024-03-30 05:06:27,478 INFO Train. Iter 180000 : Commit. 0.25647 PPL. 409.94 Recons. 0.00211 +2024-03-30 05:06:51,429 INFO Train. Iter 180200 : Commit. 0.26086 PPL. 410.07 Recons. 0.00226 +2024-03-30 05:07:15,282 INFO Train. Iter 180400 : Commit. 0.25789 PPL. 410.60 Recons. 0.00214 +2024-03-30 05:07:39,681 INFO Train. Iter 180600 : Commit. 0.25208 PPL. 409.58 Recons. 0.00212 +2024-03-30 05:08:03,415 INFO Train. Iter 180800 : Commit. 0.26135 PPL. 409.02 Recons. 0.00228 +2024-03-30 05:08:28,261 INFO Train. Iter 181000 : Commit. 0.25404 PPL. 410.84 Recons. 0.00191 +2024-03-30 05:08:53,447 INFO Train. Iter 181200 : Commit. 0.26474 PPL. 410.03 Recons. 0.00239 +2024-03-30 05:09:18,946 INFO Train. Iter 181400 : Commit. 0.25915 PPL. 409.22 Recons. 0.00231 +2024-03-30 05:09:44,946 INFO Train. Iter 181600 : Commit. 0.26488 PPL. 410.22 Recons. 0.00233 +2024-03-30 05:10:10,883 INFO Train. Iter 181800 : Commit. 0.26122 PPL. 409.40 Recons. 0.00227 +2024-03-30 05:10:37,416 INFO Train. Iter 182000 : Commit. 0.26173 PPL. 409.56 Recons. 0.00228 +2024-03-30 05:11:03,714 INFO Train. Iter 182200 : Commit. 0.26256 PPL. 408.43 Recons. 0.00241 +2024-03-30 05:11:29,443 INFO Train. Iter 182400 : Commit. 0.26114 PPL. 409.18 Recons. 0.00231 +2024-03-30 05:11:52,994 INFO Train. Iter 182600 : Commit. 0.26478 PPL. 408.40 Recons. 0.00251 +2024-03-30 05:12:17,876 INFO Train. Iter 182800 : Commit. 0.24958 PPL. 410.59 Recons. 0.00183 +2024-03-30 05:12:42,475 INFO Train. Iter 183000 : Commit. 0.26002 PPL. 408.54 Recons. 0.00234 +2024-03-30 05:13:08,072 INFO Train. Iter 183200 : Commit. 0.26187 PPL. 407.68 Recons. 0.00254 +2024-03-30 05:13:33,063 INFO Train. Iter 183400 : Commit. 0.25787 PPL. 410.18 Recons. 0.00199 +2024-03-30 05:13:58,353 INFO Train. Iter 183600 : Commit. 0.26510 PPL. 407.27 Recons. 0.00256 +2024-03-30 05:14:24,022 INFO Train. Iter 183800 : Commit. 0.25848 PPL. 409.65 Recons. 0.00218 +2024-03-30 05:14:49,805 INFO Train. Iter 184000 : Commit. 0.26293 PPL. 409.38 Recons. 0.00230 +2024-03-30 05:15:15,510 INFO Train. Iter 184200 : Commit. 0.25174 PPL. 408.96 Recons. 0.00209 +2024-03-30 05:15:41,548 INFO Train. Iter 184400 : Commit. 0.25903 PPL. 409.61 Recons. 0.00225 +2024-03-30 05:16:05,740 INFO Train. Iter 184600 : Commit. 0.25826 PPL. 409.99 Recons. 0.00223 +2024-03-30 05:16:30,310 INFO Train. Iter 184800 : Commit. 0.26219 PPL. 409.19 Recons. 0.00234 +2024-03-30 05:16:55,180 INFO Train. Iter 185000 : Commit. 0.25789 PPL. 408.30 Recons. 0.00224 +2024-03-30 05:17:20,696 INFO Train. Iter 185200 : Commit. 0.26250 PPL. 408.02 Recons. 0.00241 +2024-03-30 05:17:45,214 INFO Train. Iter 185400 : Commit. 0.25286 PPL. 409.58 Recons. 0.00211 +2024-03-30 05:18:10,087 INFO Train. Iter 185600 : Commit. 0.25844 PPL. 409.97 Recons. 0.00222 +2024-03-30 05:18:35,040 INFO Train. Iter 185800 : Commit. 0.26393 PPL. 408.10 Recons. 0.00257 +2024-03-30 05:19:00,029 INFO Train. Iter 186000 : Commit. 0.25448 PPL. 409.82 Recons. 0.00201 +2024-03-30 05:19:24,591 INFO Train. Iter 186200 : Commit. 0.27488 PPL. 405.60 Recons. 0.00308 +2024-03-30 05:19:49,430 INFO Train. Iter 186400 : Commit. 0.25802 PPL. 410.06 Recons. 0.00205 +2024-03-30 05:20:13,422 INFO Train. Iter 186600 : Commit. 0.25586 PPL. 409.16 Recons. 0.00220 +2024-03-30 05:20:38,674 INFO Train. Iter 186800 : Commit. 0.26863 PPL. 408.24 Recons. 0.00264 +2024-03-30 05:21:03,941 INFO Train. Iter 187000 : Commit. 0.25109 PPL. 410.84 Recons. 0.00192 +2024-03-30 05:21:29,682 INFO Train. Iter 187200 : Commit. 0.25799 PPL. 408.66 Recons. 0.00237 +2024-03-30 05:21:55,310 INFO Train. Iter 187400 : Commit. 0.25506 PPL. 409.26 Recons. 0.00230 +2024-03-30 05:22:20,886 INFO Train. Iter 187600 : Commit. 0.25668 PPL. 409.24 Recons. 0.00223 +2024-03-30 05:22:46,916 INFO Train. Iter 187800 : Commit. 0.25052 PPL. 408.11 Recons. 0.00207 +2024-03-30 05:23:12,179 INFO Train. Iter 188000 : Commit. 0.25357 PPL. 409.46 Recons. 0.00209 +2024-03-30 05:23:37,450 INFO Train. Iter 188200 : Commit. 0.26433 PPL. 407.57 Recons. 0.00260 +2024-03-30 05:24:01,937 INFO Train. Iter 188400 : Commit. 0.24884 PPL. 408.94 Recons. 0.00216 +2024-03-30 05:24:27,539 INFO Train. Iter 188600 : Commit. 0.25340 PPL. 408.92 Recons. 0.00201 +2024-03-30 05:24:53,047 INFO Train. Iter 188800 : Commit. 0.25672 PPL. 407.46 Recons. 0.00260 +2024-03-30 05:25:18,685 INFO Train. Iter 189000 : Commit. 0.25546 PPL. 409.44 Recons. 0.00198 +2024-03-30 05:25:44,254 INFO Train. Iter 189200 : Commit. 0.25875 PPL. 408.24 Recons. 0.00232 +2024-03-30 05:26:09,575 INFO Train. Iter 189400 : Commit. 0.25334 PPL. 408.41 Recons. 0.00211 +2024-03-30 05:26:34,428 INFO Train. Iter 189600 : Commit. 0.26464 PPL. 407.69 Recons. 0.00249 +2024-03-30 05:26:59,622 INFO Train. Iter 189800 : Commit. 0.25676 PPL. 409.42 Recons. 0.00214 +2024-03-30 05:27:25,523 INFO Train. Iter 190000 : Commit. 0.25705 PPL. 407.19 Recons. 0.00221 +2024-03-30 05:27:50,986 INFO Train. Iter 190200 : Commit. 0.26410 PPL. 407.91 Recons. 0.00260 +2024-03-30 05:28:15,075 INFO Train. Iter 190400 : Commit. 0.25759 PPL. 409.27 Recons. 0.00216 +2024-03-30 05:28:40,184 INFO Train. Iter 190600 : Commit. 0.25777 PPL. 409.56 Recons. 0.00221 +2024-03-30 05:29:05,858 INFO Train. Iter 190800 : Commit. 0.25680 PPL. 409.24 Recons. 0.00207 +2024-03-30 05:29:31,518 INFO Train. Iter 191000 : Commit. 0.25757 PPL. 409.38 Recons. 0.00217 +2024-03-30 05:29:56,402 INFO Train. Iter 191200 : Commit. 0.25598 PPL. 408.98 Recons. 0.00218 +2024-03-30 05:30:21,283 INFO Train. Iter 191400 : Commit. 0.26283 PPL. 408.10 Recons. 0.00242 +2024-03-30 05:30:46,253 INFO Train. Iter 191600 : Commit. 0.25638 PPL. 409.08 Recons. 0.00213 +2024-03-30 05:31:12,205 INFO Train. Iter 191800 : Commit. 0.25160 PPL. 408.01 Recons. 0.00210 +2024-03-30 05:31:37,493 INFO Train. Iter 192000 : Commit. 0.25081 PPL. 408.74 Recons. 0.00204 +2024-03-30 05:32:02,847 INFO Train. Iter 192200 : Commit. 0.26183 PPL. 408.84 Recons. 0.00241 +2024-03-30 05:32:27,477 INFO Train. Iter 192400 : Commit. 0.24697 PPL. 410.25 Recons. 0.00187 +2024-03-30 05:32:52,859 INFO Train. Iter 192600 : Commit. 0.26796 PPL. 406.47 Recons. 0.00272 +2024-03-30 05:33:18,262 INFO Train. Iter 192800 : Commit. 0.25503 PPL. 409.42 Recons. 0.00213 +2024-03-30 05:33:43,683 INFO Train. Iter 193000 : Commit. 0.26072 PPL. 409.05 Recons. 0.00219 +2024-03-30 05:34:08,960 INFO Train. Iter 193200 : Commit. 0.25815 PPL. 409.23 Recons. 0.00220 +2024-03-30 05:34:34,399 INFO Train. Iter 193400 : Commit. 0.25297 PPL. 409.66 Recons. 0.00205 +2024-03-30 05:34:59,147 INFO Train. Iter 193600 : Commit. 0.26086 PPL. 409.75 Recons. 0.00221 +2024-03-30 05:35:24,292 INFO Train. Iter 193800 : Commit. 0.25703 PPL. 408.79 Recons. 0.00221 +2024-03-30 05:35:49,473 INFO Train. Iter 194000 : Commit. 0.25901 PPL. 409.24 Recons. 0.00223 +2024-03-30 05:36:13,334 INFO Train. Iter 194200 : Commit. 0.26828 PPL. 405.23 Recons. 0.00298 +2024-03-30 05:36:37,715 INFO Train. Iter 194400 : Commit. 0.26164 PPL. 408.10 Recons. 0.00229 +2024-03-30 05:37:02,735 INFO Train. Iter 194600 : Commit. 0.25840 PPL. 407.65 Recons. 0.00224 +2024-03-30 05:37:27,484 INFO Train. Iter 194800 : Commit. 0.26627 PPL. 407.41 Recons. 0.00280 +2024-03-30 05:37:52,620 INFO Train. Iter 195000 : Commit. 0.25677 PPL. 408.46 Recons. 0.00208 +2024-03-30 05:38:17,722 INFO Train. Iter 195200 : Commit. 0.25136 PPL. 408.38 Recons. 0.00214 +2024-03-30 05:38:42,843 INFO Train. Iter 195400 : Commit. 0.25143 PPL. 409.06 Recons. 0.00199 +2024-03-30 05:39:07,575 INFO Train. Iter 195600 : Commit. 0.26329 PPL. 407.43 Recons. 0.00251 +2024-03-30 05:39:32,345 INFO Train. Iter 195800 : Commit. 0.26396 PPL. 406.99 Recons. 0.00258 +2024-03-30 05:39:57,604 INFO Train. Iter 196000 : Commit. 0.25914 PPL. 408.87 Recons. 0.00216 +2024-03-30 05:40:21,017 INFO Train. Iter 196200 : Commit. 0.26509 PPL. 407.74 Recons. 0.00247 +2024-03-30 05:40:45,797 INFO Train. Iter 196400 : Commit. 0.25853 PPL. 408.69 Recons. 0.00222 +2024-03-30 05:41:10,794 INFO Train. Iter 196600 : Commit. 0.26205 PPL. 406.84 Recons. 0.00248 +2024-03-30 05:41:35,929 INFO Train. Iter 196800 : Commit. 0.26070 PPL. 406.02 Recons. 0.00247 +2024-03-30 05:42:01,146 INFO Train. Iter 197000 : Commit. 0.25313 PPL. 409.02 Recons. 0.00194 +2024-03-30 05:42:26,472 INFO Train. Iter 197200 : Commit. 0.25738 PPL. 408.99 Recons. 0.00221 +2024-03-30 05:42:51,034 INFO Train. Iter 197400 : Commit. 0.24976 PPL. 408.43 Recons. 0.00208 +2024-03-30 05:43:15,954 INFO Train. Iter 197600 : Commit. 0.25459 PPL. 407.18 Recons. 0.00218 +2024-03-30 05:43:39,955 INFO Train. Iter 197800 : Commit. 0.26348 PPL. 407.03 Recons. 0.00250 +2024-03-30 05:44:04,456 INFO Train. Iter 198000 : Commit. 0.25095 PPL. 409.33 Recons. 0.00197 +2024-03-30 05:44:27,286 INFO Train. Iter 198200 : Commit. 0.26166 PPL. 408.27 Recons. 0.00249 +2024-03-30 05:44:50,962 INFO Train. Iter 198400 : Commit. 0.25679 PPL. 406.48 Recons. 0.00248 +2024-03-30 05:45:15,217 INFO Train. Iter 198600 : Commit. 0.25133 PPL. 408.77 Recons. 0.00199 +2024-03-30 05:45:39,279 INFO Train. Iter 198800 : Commit. 0.25652 PPL. 406.98 Recons. 0.00254 +2024-03-30 05:46:04,362 INFO Train. Iter 199000 : Commit. 0.26308 PPL. 407.73 Recons. 0.00231 +2024-03-30 05:46:29,473 INFO Train. Iter 199200 : Commit. 0.25332 PPL. 407.93 Recons. 0.00213 +2024-03-30 05:46:54,366 INFO Train. Iter 199400 : Commit. 0.26944 PPL. 407.73 Recons. 0.00257 +2024-03-30 05:47:19,441 INFO Train. Iter 199600 : Commit. 0.25457 PPL. 407.58 Recons. 0.00215 +2024-03-30 05:47:44,604 INFO Train. Iter 199800 : Commit. 0.25765 PPL. 407.86 Recons. 0.00222 +2024-03-30 05:48:08,960 INFO Train. Iter 200000 : Commit. 0.25771 PPL. 407.96 Recons. 0.00228 +2024-03-30 05:48:33,774 INFO Train. Iter 200200 : Commit. 0.25403 PPL. 410.44 Recons. 0.00163 +2024-03-30 05:48:58,673 INFO Train. Iter 200400 : Commit. 0.24806 PPL. 410.05 Recons. 0.00150 +2024-03-30 05:49:24,088 INFO Train. Iter 200600 : Commit. 0.24716 PPL. 409.33 Recons. 0.00150 +2024-03-30 05:49:49,409 INFO Train. Iter 200800 : Commit. 0.24402 PPL. 409.22 Recons. 0.00144 +2024-03-30 05:50:14,423 INFO Train. Iter 201000 : Commit. 0.24987 PPL. 409.84 Recons. 0.00152 +2024-03-30 05:50:39,819 INFO Train. Iter 201200 : Commit. 0.24819 PPL. 409.27 Recons. 0.00149 +2024-03-30 05:51:04,777 INFO Train. Iter 201400 : Commit. 0.24603 PPL. 409.10 Recons. 0.00146 +2024-03-30 05:51:29,494 INFO Train. Iter 201600 : Commit. 0.24257 PPL. 409.40 Recons. 0.00139 +2024-03-30 05:51:54,442 INFO Train. Iter 201800 : Commit. 0.24280 PPL. 409.78 Recons. 0.00139 +2024-03-30 05:52:18,831 INFO Train. Iter 202000 : Commit. 0.24502 PPL. 408.72 Recons. 0.00146 +2024-03-30 05:52:44,249 INFO Train. Iter 202200 : Commit. 0.24698 PPL. 408.71 Recons. 0.00149 +2024-03-30 05:53:10,320 INFO Train. Iter 202400 : Commit. 0.24758 PPL. 409.10 Recons. 0.00148 +2024-03-30 05:53:35,846 INFO Train. Iter 202600 : Commit. 0.24372 PPL. 409.49 Recons. 0.00138 +2024-03-30 05:54:01,243 INFO Train. Iter 202800 : Commit. 0.24273 PPL. 408.76 Recons. 0.00139 +2024-03-30 05:54:27,445 INFO Train. Iter 203000 : Commit. 0.24182 PPL. 409.03 Recons. 0.00138 +2024-03-30 05:54:52,847 INFO Train. Iter 203200 : Commit. 0.23855 PPL. 408.18 Recons. 0.00136 +2024-03-30 05:55:18,351 INFO Train. Iter 203400 : Commit. 0.24847 PPL. 408.88 Recons. 0.00149 +2024-03-30 05:55:43,389 INFO Train. Iter 203600 : Commit. 0.24266 PPL. 408.56 Recons. 0.00140 +2024-03-30 05:56:08,261 INFO Train. Iter 203800 : Commit. 0.24277 PPL. 408.50 Recons. 0.00141 +2024-03-30 05:56:33,409 INFO Train. Iter 204000 : Commit. 0.23822 PPL. 408.08 Recons. 0.00134 +2024-03-30 05:56:58,859 INFO Train. Iter 204200 : Commit. 0.24369 PPL. 408.49 Recons. 0.00143 +2024-03-30 05:57:24,580 INFO Train. Iter 204400 : Commit. 0.23915 PPL. 408.47 Recons. 0.00134 +2024-03-30 05:57:50,078 INFO Train. Iter 204600 : Commit. 0.24011 PPL. 408.68 Recons. 0.00136 +2024-03-30 05:58:15,444 INFO Train. Iter 204800 : Commit. 0.24646 PPL. 408.81 Recons. 0.00145 +2024-03-30 05:58:40,929 INFO Train. Iter 205000 : Commit. 0.23919 PPL. 408.13 Recons. 0.00134 +2024-03-30 05:59:06,333 INFO Train. Iter 205200 : Commit. 0.24648 PPL. 408.70 Recons. 0.00145 +2024-03-30 05:59:31,768 INFO Train. Iter 205400 : Commit. 0.24426 PPL. 409.27 Recons. 0.00139 +2024-03-30 05:59:57,533 INFO Train. Iter 205600 : Commit. 0.24619 PPL. 408.46 Recons. 0.00145 +2024-03-30 06:00:22,046 INFO Train. Iter 205800 : Commit. 0.24893 PPL. 408.41 Recons. 0.00148 +2024-03-30 06:00:48,089 INFO Train. Iter 206000 : Commit. 0.23864 PPL. 409.19 Recons. 0.00131 +2024-03-30 06:01:14,385 INFO Train. Iter 206200 : Commit. 0.23562 PPL. 407.99 Recons. 0.00129 +2024-03-30 06:01:39,823 INFO Train. Iter 206400 : Commit. 0.24363 PPL. 408.41 Recons. 0.00140 +2024-03-30 06:02:05,686 INFO Train. Iter 206600 : Commit. 0.24143 PPL. 408.55 Recons. 0.00136 +2024-03-30 06:02:31,494 INFO Train. Iter 206800 : Commit. 0.24300 PPL. 408.17 Recons. 0.00139 +2024-03-30 06:02:57,357 INFO Train. Iter 207000 : Commit. 0.24176 PPL. 407.77 Recons. 0.00137 +2024-03-30 06:03:23,643 INFO Train. Iter 207200 : Commit. 0.24463 PPL. 408.57 Recons. 0.00141 +2024-03-30 06:03:49,130 INFO Train. Iter 207400 : Commit. 0.24382 PPL. 407.96 Recons. 0.00142 +2024-03-30 06:04:14,476 INFO Train. Iter 207600 : Commit. 0.24733 PPL. 407.40 Recons. 0.00149 +2024-03-30 06:04:39,506 INFO Train. Iter 207800 : Commit. 0.23876 PPL. 408.19 Recons. 0.00133 +2024-03-30 06:05:04,722 INFO Train. Iter 208000 : Commit. 0.24586 PPL. 407.54 Recons. 0.00146 +2024-03-30 06:05:30,136 INFO Train. Iter 208200 : Commit. 0.24450 PPL. 407.37 Recons. 0.00144 +2024-03-30 06:05:55,650 INFO Train. Iter 208400 : Commit. 0.23933 PPL. 408.11 Recons. 0.00133 +2024-03-30 06:06:22,144 INFO Train. Iter 208600 : Commit. 0.23605 PPL. 408.14 Recons. 0.00128 +2024-03-30 06:06:48,570 INFO Train. Iter 208800 : Commit. 0.23707 PPL. 407.98 Recons. 0.00130 +2024-03-30 06:07:14,013 INFO Train. Iter 209000 : Commit. 0.24262 PPL. 408.41 Recons. 0.00137 +2024-03-30 06:07:39,074 INFO Train. Iter 209200 : Commit. 0.24693 PPL. 408.06 Recons. 0.00146 +2024-03-30 06:08:04,816 INFO Train. Iter 209400 : Commit. 0.24328 PPL. 408.28 Recons. 0.00139 +2024-03-30 06:08:30,069 INFO Train. Iter 209600 : Commit. 0.24543 PPL. 408.09 Recons. 0.00142 +2024-03-30 06:08:54,360 INFO Train. Iter 209800 : Commit. 0.24421 PPL. 408.60 Recons. 0.00140 +2024-03-30 06:09:19,145 INFO Train. Iter 210000 : Commit. 0.24415 PPL. 407.84 Recons. 0.00142 +2024-03-30 06:09:44,655 INFO Train. Iter 210200 : Commit. 0.24427 PPL. 407.27 Recons. 0.00144 +2024-03-30 06:10:09,610 INFO Train. Iter 210400 : Commit. 0.24325 PPL. 407.99 Recons. 0.00141 +2024-03-30 06:10:35,137 INFO Train. Iter 210600 : Commit. 0.23584 PPL. 407.39 Recons. 0.00130 +2024-03-30 06:11:00,678 INFO Train. Iter 210800 : Commit. 0.24387 PPL. 407.04 Recons. 0.00142 +2024-03-30 06:11:25,966 INFO Train. Iter 211000 : Commit. 0.23900 PPL. 407.23 Recons. 0.00134 +2024-03-30 06:11:51,312 INFO Train. Iter 211200 : Commit. 0.23836 PPL. 407.36 Recons. 0.00131 +2024-03-30 06:12:16,249 INFO Train. Iter 211400 : Commit. 0.23478 PPL. 407.02 Recons. 0.00128 +2024-03-30 06:12:40,017 INFO Train. Iter 211600 : Commit. 0.24167 PPL. 406.97 Recons. 0.00138 +2024-03-30 06:13:05,600 INFO Train. Iter 211800 : Commit. 0.23897 PPL. 407.41 Recons. 0.00133 +2024-03-30 06:13:30,899 INFO Train. Iter 212000 : Commit. 0.24396 PPL. 407.40 Recons. 0.00140 +2024-03-30 06:13:56,642 INFO Train. Iter 212200 : Commit. 0.23926 PPL. 407.55 Recons. 0.00133 +2024-03-30 06:14:22,372 INFO Train. Iter 212400 : Commit. 0.23892 PPL. 407.05 Recons. 0.00134 +2024-03-30 06:14:47,183 INFO Train. Iter 212600 : Commit. 0.24895 PPL. 407.20 Recons. 0.00148 +2024-03-30 06:15:12,466 INFO Train. Iter 212800 : Commit. 0.24032 PPL. 407.22 Recons. 0.00135 +2024-03-30 06:15:37,388 INFO Train. Iter 213000 : Commit. 0.24353 PPL. 408.04 Recons. 0.00138 +2024-03-30 06:16:03,021 INFO Train. Iter 213200 : Commit. 0.24544 PPL. 407.87 Recons. 0.00141 +2024-03-30 06:16:28,216 INFO Train. Iter 213400 : Commit. 0.23517 PPL. 407.34 Recons. 0.00127 +2024-03-30 06:16:51,763 INFO Train. Iter 213600 : Commit. 0.23386 PPL. 407.00 Recons. 0.00126 +2024-03-30 06:17:16,982 INFO Train. Iter 213800 : Commit. 0.24228 PPL. 408.13 Recons. 0.00137 +2024-03-30 06:17:42,369 INFO Train. Iter 214000 : Commit. 0.23818 PPL. 407.17 Recons. 0.00133 +2024-03-30 06:18:06,799 INFO Train. Iter 214200 : Commit. 0.23967 PPL. 407.71 Recons. 0.00134 +2024-03-30 06:18:32,632 INFO Train. Iter 214400 : Commit. 0.24395 PPL. 407.51 Recons. 0.00141 +2024-03-30 06:18:58,158 INFO Train. Iter 214600 : Commit. 0.23580 PPL. 406.10 Recons. 0.00131 +2024-03-30 06:19:23,183 INFO Train. Iter 214800 : Commit. 0.24120 PPL. 407.70 Recons. 0.00134 +2024-03-30 06:19:48,481 INFO Train. Iter 215000 : Commit. 0.24191 PPL. 407.09 Recons. 0.00139 +2024-03-30 06:20:14,064 INFO Train. Iter 215200 : Commit. 0.24077 PPL. 406.48 Recons. 0.00139 +2024-03-30 06:20:39,818 INFO Train. Iter 215400 : Commit. 0.24011 PPL. 407.56 Recons. 0.00135 +2024-03-30 06:21:03,997 INFO Train. Iter 215600 : Commit. 0.24183 PPL. 407.39 Recons. 0.00137 +2024-03-30 06:21:28,813 INFO Train. Iter 215800 : Commit. 0.23758 PPL. 408.11 Recons. 0.00128 +2024-03-30 06:21:53,771 INFO Train. Iter 216000 : Commit. 0.24085 PPL. 407.10 Recons. 0.00136 +2024-03-30 06:22:19,002 INFO Train. Iter 216200 : Commit. 0.24276 PPL. 407.52 Recons. 0.00138 +2024-03-30 06:22:43,706 INFO Train. Iter 216400 : Commit. 0.24522 PPL. 407.04 Recons. 0.00145 +2024-03-30 06:23:08,059 INFO Train. Iter 216600 : Commit. 0.23876 PPL. 407.04 Recons. 0.00134 +2024-03-30 06:23:32,188 INFO Train. Iter 216800 : Commit. 0.23928 PPL. 407.48 Recons. 0.00135 +2024-03-30 06:23:56,456 INFO Train. Iter 217000 : Commit. 0.23733 PPL. 406.46 Recons. 0.00133 +2024-03-30 06:24:21,476 INFO Train. Iter 217200 : Commit. 0.23915 PPL. 407.36 Recons. 0.00133 +2024-03-30 06:24:45,999 INFO Train. Iter 217400 : Commit. 0.24232 PPL. 406.72 Recons. 0.00140 +2024-03-30 06:25:11,562 INFO Train. Iter 217600 : Commit. 0.24060 PPL. 407.28 Recons. 0.00135 +2024-03-30 06:25:37,813 INFO Train. Iter 217800 : Commit. 0.24101 PPL. 407.06 Recons. 0.00136 +2024-03-30 06:26:03,063 INFO Train. Iter 218000 : Commit. 0.23936 PPL. 406.68 Recons. 0.00136 +2024-03-30 06:26:28,003 INFO Train. Iter 218200 : Commit. 0.24259 PPL. 406.71 Recons. 0.00141 +2024-03-30 06:26:52,817 INFO Train. Iter 218400 : Commit. 0.23677 PPL. 407.22 Recons. 0.00130 +2024-03-30 06:27:17,737 INFO Train. Iter 218600 : Commit. 0.24054 PPL. 406.81 Recons. 0.00137 +2024-03-30 06:27:42,649 INFO Train. Iter 218800 : Commit. 0.24362 PPL. 406.66 Recons. 0.00143 +2024-03-30 06:28:07,821 INFO Train. Iter 219000 : Commit. 0.23907 PPL. 406.85 Recons. 0.00134 +2024-03-30 06:28:32,483 INFO Train. Iter 219200 : Commit. 0.23811 PPL. 406.32 Recons. 0.00134 +2024-03-30 06:28:56,770 INFO Train. Iter 219400 : Commit. 0.23965 PPL. 407.66 Recons. 0.00134 +2024-03-30 06:29:21,381 INFO Train. Iter 219600 : Commit. 0.24116 PPL. 406.51 Recons. 0.00139 +2024-03-30 06:29:46,515 INFO Train. Iter 219800 : Commit. 0.23844 PPL. 406.77 Recons. 0.00135 +2024-03-30 06:30:11,926 INFO Train. Iter 220000 : Commit. 0.23948 PPL. 406.51 Recons. 0.00137 +2024-03-30 06:30:37,554 INFO Train. Iter 220200 : Commit. 0.23814 PPL. 406.54 Recons. 0.00133 +2024-03-30 06:31:02,684 INFO Train. Iter 220400 : Commit. 0.24039 PPL. 407.31 Recons. 0.00135 +2024-03-30 06:31:27,437 INFO Train. Iter 220600 : Commit. 0.23682 PPL. 407.08 Recons. 0.00130 +2024-03-30 06:31:53,026 INFO Train. Iter 220800 : Commit. 0.23499 PPL. 406.54 Recons. 0.00128 +2024-03-30 06:32:18,343 INFO Train. Iter 221000 : Commit. 0.23039 PPL. 406.39 Recons. 0.00122 +2024-03-30 06:32:44,146 INFO Train. Iter 221200 : Commit. 0.23448 PPL. 406.45 Recons. 0.00129 +2024-03-30 06:33:09,236 INFO Train. Iter 221400 : Commit. 0.23677 PPL. 406.43 Recons. 0.00132 +2024-03-30 06:33:34,872 INFO Train. Iter 221600 : Commit. 0.23990 PPL. 406.88 Recons. 0.00134 +2024-03-30 06:34:00,391 INFO Train. Iter 221800 : Commit. 0.23461 PPL. 405.96 Recons. 0.00130 +2024-03-30 06:34:26,151 INFO Train. Iter 222000 : Commit. 0.23975 PPL. 406.80 Recons. 0.00134 +2024-03-30 06:34:52,251 INFO Train. Iter 222200 : Commit. 0.24463 PPL. 406.88 Recons. 0.00142 +2024-03-30 06:35:17,982 INFO Train. Iter 222400 : Commit. 0.23799 PPL. 406.22 Recons. 0.00133 +2024-03-30 06:35:42,682 INFO Train. Iter 222600 : Commit. 0.24144 PPL. 406.88 Recons. 0.00138 +2024-03-30 06:36:07,899 INFO Train. Iter 222800 : Commit. 0.24070 PPL. 406.56 Recons. 0.00137 +2024-03-30 06:36:33,440 INFO Train. Iter 223000 : Commit. 0.23678 PPL. 406.86 Recons. 0.00130 +2024-03-30 06:36:57,629 INFO Train. Iter 223200 : Commit. 0.23847 PPL. 406.42 Recons. 0.00133 +2024-03-30 06:37:23,244 INFO Train. Iter 223400 : Commit. 0.23753 PPL. 406.65 Recons. 0.00131 +2024-03-30 06:37:48,127 INFO Train. Iter 223600 : Commit. 0.23917 PPL. 406.59 Recons. 0.00134 +2024-03-30 06:38:12,835 INFO Train. Iter 223800 : Commit. 0.24168 PPL. 406.71 Recons. 0.00137 +2024-03-30 06:38:37,344 INFO Train. Iter 224000 : Commit. 0.23849 PPL. 407.06 Recons. 0.00131 +2024-03-30 06:39:01,734 INFO Train. Iter 224200 : Commit. 0.24265 PPL. 407.44 Recons. 0.00136 +2024-03-30 06:39:27,362 INFO Train. Iter 224400 : Commit. 0.23772 PPL. 406.72 Recons. 0.00131 +2024-03-30 06:39:52,786 INFO Train. Iter 224600 : Commit. 0.23579 PPL. 406.50 Recons. 0.00130 +2024-03-30 06:40:18,622 INFO Train. Iter 224800 : Commit. 0.23562 PPL. 406.45 Recons. 0.00130 +2024-03-30 06:40:44,344 INFO Train. Iter 225000 : Commit. 0.23656 PPL. 407.02 Recons. 0.00130 +2024-03-30 06:41:08,581 INFO Train. Iter 225200 : Commit. 0.24004 PPL. 406.69 Recons. 0.00137 +2024-03-30 06:41:33,418 INFO Train. Iter 225400 : Commit. 0.24175 PPL. 407.27 Recons. 0.00137 +2024-03-30 06:41:59,402 INFO Train. Iter 225600 : Commit. 0.23596 PPL. 406.01 Recons. 0.00132 +2024-03-30 06:42:24,618 INFO Train. Iter 225800 : Commit. 0.24257 PPL. 407.32 Recons. 0.00138 +2024-03-30 06:42:50,174 INFO Train. Iter 226000 : Commit. 0.23708 PPL. 406.90 Recons. 0.00131 +2024-03-30 06:43:15,349 INFO Train. Iter 226200 : Commit. 0.23782 PPL. 406.20 Recons. 0.00134 +2024-03-30 06:43:40,944 INFO Train. Iter 226400 : Commit. 0.23964 PPL. 406.69 Recons. 0.00135 +2024-03-30 06:44:06,415 INFO Train. Iter 226600 : Commit. 0.23392 PPL. 406.34 Recons. 0.00127 +2024-03-30 06:44:32,268 INFO Train. Iter 226800 : Commit. 0.24103 PPL. 406.93 Recons. 0.00137 +2024-03-30 06:44:57,731 INFO Train. Iter 227000 : Commit. 0.23896 PPL. 406.14 Recons. 0.00135 +2024-03-30 06:45:22,917 INFO Train. Iter 227200 : Commit. 0.23688 PPL. 406.73 Recons. 0.00131 +2024-03-30 06:45:48,812 INFO Train. Iter 227400 : Commit. 0.23611 PPL. 406.23 Recons. 0.00131 +2024-03-30 06:46:13,510 INFO Train. Iter 227600 : Commit. 0.24018 PPL. 406.74 Recons. 0.00136 +2024-03-30 06:46:39,070 INFO Train. Iter 227800 : Commit. 0.23479 PPL. 406.67 Recons. 0.00128 +2024-03-30 06:47:03,652 INFO Train. Iter 228000 : Commit. 0.24113 PPL. 406.91 Recons. 0.00137 +2024-03-30 06:47:28,751 INFO Train. Iter 228200 : Commit. 0.23554 PPL. 406.23 Recons. 0.00131 +2024-03-30 06:47:53,512 INFO Train. Iter 228400 : Commit. 0.23756 PPL. 405.89 Recons. 0.00133 +2024-03-30 06:48:17,954 INFO Train. Iter 228600 : Commit. 0.24495 PPL. 406.47 Recons. 0.00143 +2024-03-30 06:48:42,321 INFO Train. Iter 228800 : Commit. 0.23485 PPL. 406.11 Recons. 0.00129 +2024-03-30 06:49:06,442 INFO Train. Iter 229000 : Commit. 0.23829 PPL. 406.68 Recons. 0.00133 +2024-03-30 06:49:31,970 INFO Train. Iter 229200 : Commit. 0.24036 PPL. 406.85 Recons. 0.00135 +2024-03-30 06:49:57,447 INFO Train. Iter 229400 : Commit. 0.23297 PPL. 405.87 Recons. 0.00126 +2024-03-30 06:50:23,304 INFO Train. Iter 229600 : Commit. 0.23505 PPL. 405.94 Recons. 0.00129 +2024-03-30 06:50:49,416 INFO Train. Iter 229800 : Commit. 0.23818 PPL. 405.67 Recons. 0.00135 +2024-03-30 06:51:15,305 INFO Train. Iter 230000 : Commit. 0.23633 PPL. 406.96 Recons. 0.00128 +2024-03-30 06:51:41,241 INFO Train. Iter 230200 : Commit. 0.23744 PPL. 405.51 Recons. 0.00135 +2024-03-30 06:52:07,520 INFO Train. Iter 230400 : Commit. 0.23458 PPL. 405.81 Recons. 0.00129 +2024-03-30 06:52:33,593 INFO Train. Iter 230600 : Commit. 0.24072 PPL. 406.30 Recons. 0.00137 +2024-03-30 06:52:59,430 INFO Train. Iter 230800 : Commit. 0.23447 PPL. 405.76 Recons. 0.00128 +2024-03-30 06:53:23,929 INFO Train. Iter 231000 : Commit. 0.24229 PPL. 405.90 Recons. 0.00141 +2024-03-30 06:53:49,224 INFO Train. Iter 231200 : Commit. 0.23394 PPL. 405.47 Recons. 0.00129 +2024-03-30 06:54:14,662 INFO Train. Iter 231400 : Commit. 0.24044 PPL. 405.97 Recons. 0.00139 +2024-03-30 06:54:40,482 INFO Train. Iter 231600 : Commit. 0.23594 PPL. 405.98 Recons. 0.00131 +2024-03-30 06:55:06,271 INFO Train. Iter 231800 : Commit. 0.23358 PPL. 405.51 Recons. 0.00129 +2024-03-30 06:55:32,109 INFO Train. Iter 232000 : Commit. 0.23906 PPL. 405.95 Recons. 0.00136 +2024-03-30 06:55:57,917 INFO Train. Iter 232200 : Commit. 0.24030 PPL. 405.95 Recons. 0.00137 +2024-03-30 06:56:23,702 INFO Train. Iter 232400 : Commit. 0.23552 PPL. 406.45 Recons. 0.00129 +2024-03-30 06:56:49,647 INFO Train. Iter 232600 : Commit. 0.23088 PPL. 405.51 Recons. 0.00124 +2024-03-30 06:57:15,408 INFO Train. Iter 232800 : Commit. 0.24043 PPL. 406.18 Recons. 0.00138 +2024-03-30 06:57:39,659 INFO Train. Iter 233000 : Commit. 0.24144 PPL. 406.27 Recons. 0.00138 +2024-03-30 06:58:05,415 INFO Train. Iter 233200 : Commit. 0.23899 PPL. 406.09 Recons. 0.00135 +2024-03-30 06:58:31,160 INFO Train. Iter 233400 : Commit. 0.24126 PPL. 406.74 Recons. 0.00137 +2024-03-30 06:58:57,108 INFO Train. Iter 233600 : Commit. 0.23535 PPL. 406.16 Recons. 0.00130 +2024-03-30 06:59:22,490 INFO Train. Iter 233800 : Commit. 0.22761 PPL. 404.43 Recons. 0.00122 +2024-03-30 06:59:47,304 INFO Train. Iter 234000 : Commit. 0.23397 PPL. 405.46 Recons. 0.00128 +2024-03-30 07:00:11,955 INFO Train. Iter 234200 : Commit. 0.23846 PPL. 405.09 Recons. 0.00137 +2024-03-30 07:00:36,672 INFO Train. Iter 234400 : Commit. 0.23718 PPL. 405.17 Recons. 0.00135 +2024-03-30 07:01:01,021 INFO Train. Iter 234600 : Commit. 0.23782 PPL. 405.35 Recons. 0.00134 +2024-03-30 07:01:24,359 INFO Train. Iter 234800 : Commit. 0.23108 PPL. 405.44 Recons. 0.00124 +2024-03-30 07:01:49,441 INFO Train. Iter 235000 : Commit. 0.23329 PPL. 405.27 Recons. 0.00128 +2024-03-30 07:02:14,316 INFO Train. Iter 235200 : Commit. 0.23598 PPL. 405.68 Recons. 0.00131 +2024-03-30 07:02:39,438 INFO Train. Iter 235400 : Commit. 0.23172 PPL. 405.14 Recons. 0.00125 +2024-03-30 07:03:05,162 INFO Train. Iter 235600 : Commit. 0.23567 PPL. 406.14 Recons. 0.00129 +2024-03-30 07:03:30,972 INFO Train. Iter 235800 : Commit. 0.23671 PPL. 405.70 Recons. 0.00130 +2024-03-30 07:03:56,697 INFO Train. Iter 236000 : Commit. 0.24078 PPL. 406.00 Recons. 0.00137 +2024-03-30 07:04:21,492 INFO Train. Iter 236200 : Commit. 0.23945 PPL. 405.46 Recons. 0.00136 +2024-03-30 07:04:46,508 INFO Train. Iter 236400 : Commit. 0.23602 PPL. 404.92 Recons. 0.00131 +2024-03-30 07:05:11,614 INFO Train. Iter 236600 : Commit. 0.23602 PPL. 405.75 Recons. 0.00129 +2024-03-30 07:05:35,144 INFO Train. Iter 236800 : Commit. 0.23777 PPL. 405.09 Recons. 0.00135 +2024-03-30 07:06:00,545 INFO Train. Iter 237000 : Commit. 0.23914 PPL. 405.71 Recons. 0.00136 +2024-03-30 07:06:26,636 INFO Train. Iter 237200 : Commit. 0.22896 PPL. 405.09 Recons. 0.00122 +2024-03-30 07:06:51,538 INFO Train. Iter 237400 : Commit. 0.23535 PPL. 405.84 Recons. 0.00130 +2024-03-30 07:07:17,358 INFO Train. Iter 237600 : Commit. 0.23689 PPL. 405.18 Recons. 0.00132 +2024-03-30 07:07:42,001 INFO Train. Iter 237800 : Commit. 0.23867 PPL. 406.05 Recons. 0.00134 +2024-03-30 07:08:06,831 INFO Train. Iter 238000 : Commit. 0.23526 PPL. 405.68 Recons. 0.00129 +2024-03-30 07:08:31,905 INFO Train. Iter 238200 : Commit. 0.23584 PPL. 405.98 Recons. 0.00129 +2024-03-30 07:08:56,730 INFO Train. Iter 238400 : Commit. 0.23435 PPL. 405.71 Recons. 0.00127 +2024-03-30 07:09:21,641 INFO Train. Iter 238600 : Commit. 0.23963 PPL. 405.83 Recons. 0.00135 +2024-03-30 07:09:45,644 INFO Train. Iter 238800 : Commit. 0.23749 PPL. 405.59 Recons. 0.00132 +2024-03-30 07:10:11,456 INFO Train. Iter 239000 : Commit. 0.23568 PPL. 405.91 Recons. 0.00128 +2024-03-30 07:10:36,929 INFO Train. Iter 239200 : Commit. 0.23481 PPL. 406.09 Recons. 0.00127 +2024-03-30 07:11:02,994 INFO Train. Iter 239400 : Commit. 0.23865 PPL. 405.28 Recons. 0.00136 +2024-03-30 07:11:28,388 INFO Train. Iter 239600 : Commit. 0.24060 PPL. 406.07 Recons. 0.00136 +2024-03-30 07:11:53,355 INFO Train. Iter 239800 : Commit. 0.23766 PPL. 404.68 Recons. 0.00136 +2024-03-30 07:12:19,218 INFO Train. Iter 240000 : Commit. 0.23885 PPL. 405.35 Recons. 0.00137 +2024-03-30 07:12:45,710 INFO Train. Iter 240200 : Commit. 0.23640 PPL. 405.61 Recons. 0.00131 +2024-03-30 07:13:11,733 INFO Train. Iter 240400 : Commit. 0.24025 PPL. 405.36 Recons. 0.00138 +2024-03-30 07:13:35,819 INFO Train. Iter 240600 : Commit. 0.23822 PPL. 405.79 Recons. 0.00134 +2024-03-30 07:14:00,387 INFO Train. Iter 240800 : Commit. 0.23880 PPL. 405.65 Recons. 0.00135 +2024-03-30 07:14:25,644 INFO Train. Iter 241000 : Commit. 0.23385 PPL. 405.50 Recons. 0.00128 +2024-03-30 07:14:50,584 INFO Train. Iter 241200 : Commit. 0.23134 PPL. 405.70 Recons. 0.00124 +2024-03-30 07:15:15,613 INFO Train. Iter 241400 : Commit. 0.23309 PPL. 406.09 Recons. 0.00126 +2024-03-30 07:15:41,652 INFO Train. Iter 241600 : Commit. 0.23473 PPL. 405.80 Recons. 0.00129 +2024-03-30 07:16:06,897 INFO Train. Iter 241800 : Commit. 0.23068 PPL. 404.85 Recons. 0.00126 +2024-03-30 07:16:32,208 INFO Train. Iter 242000 : Commit. 0.23563 PPL. 405.73 Recons. 0.00131 +2024-03-30 07:16:57,116 INFO Train. Iter 242200 : Commit. 0.23608 PPL. 405.74 Recons. 0.00130 +2024-03-30 07:17:22,368 INFO Train. Iter 242400 : Commit. 0.23395 PPL. 405.14 Recons. 0.00130 +2024-03-30 07:17:46,087 INFO Train. Iter 242600 : Commit. 0.23875 PPL. 406.40 Recons. 0.00133 +2024-03-30 07:18:10,726 INFO Train. Iter 242800 : Commit. 0.23539 PPL. 404.16 Recons. 0.00134 +2024-03-30 07:18:35,622 INFO Train. Iter 243000 : Commit. 0.23434 PPL. 406.37 Recons. 0.00127 +2024-03-30 07:19:00,122 INFO Train. Iter 243200 : Commit. 0.23778 PPL. 405.33 Recons. 0.00135 +2024-03-30 07:19:25,039 INFO Train. Iter 243400 : Commit. 0.23717 PPL. 406.24 Recons. 0.00131 +2024-03-30 07:19:50,025 INFO Train. Iter 243600 : Commit. 0.23775 PPL. 406.93 Recons. 0.00131 +2024-03-30 07:20:15,441 INFO Train. Iter 243800 : Commit. 0.23029 PPL. 405.27 Recons. 0.00122 +2024-03-30 07:20:40,931 INFO Train. Iter 244000 : Commit. 0.23266 PPL. 405.31 Recons. 0.00127 +2024-03-30 07:21:06,234 INFO Train. Iter 244200 : Commit. 0.23394 PPL. 405.45 Recons. 0.00128 +2024-03-30 07:21:31,558 INFO Train. Iter 244400 : Commit. 0.23893 PPL. 405.25 Recons. 0.00136 +2024-03-30 07:21:55,849 INFO Train. Iter 244600 : Commit. 0.23719 PPL. 404.77 Recons. 0.00136 +2024-03-30 07:22:21,813 INFO Train. Iter 244800 : Commit. 0.24397 PPL. 405.85 Recons. 0.00142 +2024-03-30 07:22:47,816 INFO Train. Iter 245000 : Commit. 0.23089 PPL. 405.18 Recons. 0.00124 +2024-03-30 07:23:13,674 INFO Train. Iter 245200 : Commit. 0.22984 PPL. 404.94 Recons. 0.00124 +2024-03-30 07:23:39,232 INFO Train. Iter 245400 : Commit. 0.23509 PPL. 404.88 Recons. 0.00132 +2024-03-30 07:24:04,880 INFO Train. Iter 245600 : Commit. 0.23745 PPL. 406.20 Recons. 0.00131 +2024-03-30 07:24:30,061 INFO Train. Iter 245800 : Commit. 0.23239 PPL. 405.70 Recons. 0.00125 +2024-03-30 07:24:56,092 INFO Train. Iter 246000 : Commit. 0.23547 PPL. 405.46 Recons. 0.00131 +2024-03-30 07:25:21,900 INFO Train. Iter 246200 : Commit. 0.23573 PPL. 404.82 Recons. 0.00131 +2024-03-30 07:25:45,864 INFO Train. Iter 246400 : Commit. 0.23752 PPL. 405.96 Recons. 0.00131 +2024-03-30 07:26:11,364 INFO Train. Iter 246600 : Commit. 0.23470 PPL. 405.20 Recons. 0.00128 +2024-03-30 07:26:36,562 INFO Train. Iter 246800 : Commit. 0.23702 PPL. 405.57 Recons. 0.00132 +2024-03-30 07:27:01,110 INFO Train. Iter 247000 : Commit. 0.23639 PPL. 405.39 Recons. 0.00132 +2024-03-30 07:27:25,925 INFO Train. Iter 247200 : Commit. 0.23476 PPL. 405.43 Recons. 0.00130 +2024-03-30 07:27:50,009 INFO Train. Iter 247400 : Commit. 0.23224 PPL. 404.42 Recons. 0.00128 +2024-03-30 07:28:15,378 INFO Train. Iter 247600 : Commit. 0.23803 PPL. 404.66 Recons. 0.00136 +2024-03-30 07:28:40,358 INFO Train. Iter 247800 : Commit. 0.22757 PPL. 404.62 Recons. 0.00120 +2024-03-30 07:29:04,975 INFO Train. Iter 248000 : Commit. 0.23589 PPL. 405.84 Recons. 0.00131 +2024-03-30 07:29:29,693 INFO Train. Iter 248200 : Commit. 0.24035 PPL. 405.44 Recons. 0.00138 +2024-03-30 07:29:53,983 INFO Train. Iter 248400 : Commit. 0.23222 PPL. 405.04 Recons. 0.00126 +2024-03-30 07:30:19,173 INFO Train. Iter 248600 : Commit. 0.23117 PPL. 405.06 Recons. 0.00125 +2024-03-30 07:30:44,604 INFO Train. Iter 248800 : Commit. 0.23252 PPL. 405.65 Recons. 0.00126 +2024-03-30 07:31:09,579 INFO Train. Iter 249000 : Commit. 0.22801 PPL. 404.92 Recons. 0.00120 +2024-03-30 07:31:34,138 INFO Train. Iter 249200 : Commit. 0.23480 PPL. 404.83 Recons. 0.00130 +2024-03-30 07:31:59,765 INFO Train. Iter 249400 : Commit. 0.23292 PPL. 405.43 Recons. 0.00126 +2024-03-30 07:32:25,072 INFO Train. Iter 249600 : Commit. 0.22884 PPL. 403.98 Recons. 0.00123 +2024-03-30 07:32:49,731 INFO Train. Iter 249800 : Commit. 0.23604 PPL. 404.51 Recons. 0.00133 +2024-03-30 07:33:14,881 INFO Train. Iter 250000 : Commit. 0.23851 PPL. 405.01 Recons. 0.00137 +2024-03-30 07:33:40,580 INFO Train. Iter 250200 : Commit. 0.23776 PPL. 405.18 Recons. 0.00133 +2024-03-30 07:34:04,230 INFO Train. Iter 250400 : Commit. 0.23644 PPL. 404.88 Recons. 0.00133 +2024-03-30 07:34:29,405 INFO Train. Iter 250600 : Commit. 0.23453 PPL. 404.59 Recons. 0.00130 +2024-03-30 07:34:54,148 INFO Train. Iter 250800 : Commit. 0.23718 PPL. 404.75 Recons. 0.00134 +2024-03-30 07:35:19,725 INFO Train. Iter 251000 : Commit. 0.23199 PPL. 404.36 Recons. 0.00127 +2024-03-30 07:35:45,004 INFO Train. Iter 251200 : Commit. 0.22885 PPL. 403.47 Recons. 0.00124 +2024-03-30 07:36:10,683 INFO Train. Iter 251400 : Commit. 0.23014 PPL. 404.20 Recons. 0.00124 +2024-03-30 07:36:35,621 INFO Train. Iter 251600 : Commit. 0.24056 PPL. 404.71 Recons. 0.00140 +2024-03-30 07:37:00,453 INFO Train. Iter 251800 : Commit. 0.23714 PPL. 404.83 Recons. 0.00134 +2024-03-30 07:37:25,208 INFO Train. Iter 252000 : Commit. 0.22858 PPL. 404.25 Recons. 0.00122 +2024-03-30 07:37:49,041 INFO Train. Iter 252200 : Commit. 0.23998 PPL. 405.45 Recons. 0.00137 +2024-03-30 07:38:14,293 INFO Train. Iter 252400 : Commit. 0.23120 PPL. 404.57 Recons. 0.00126 +2024-03-30 07:38:39,120 INFO Train. Iter 252600 : Commit. 0.23174 PPL. 403.97 Recons. 0.00127 +2024-03-30 07:39:03,932 INFO Train. Iter 252800 : Commit. 0.23552 PPL. 404.52 Recons. 0.00133 +2024-03-30 07:39:29,295 INFO Train. Iter 253000 : Commit. 0.23389 PPL. 404.55 Recons. 0.00130 +2024-03-30 07:39:54,554 INFO Train. Iter 253200 : Commit. 0.23268 PPL. 404.18 Recons. 0.00129 +2024-03-30 07:40:19,240 INFO Train. Iter 253400 : Commit. 0.23372 PPL. 405.08 Recons. 0.00128 +2024-03-30 07:40:44,099 INFO Train. Iter 253600 : Commit. 0.23276 PPL. 404.58 Recons. 0.00128 +2024-03-30 07:41:08,569 INFO Train. Iter 253800 : Commit. 0.23351 PPL. 404.11 Recons. 0.00129 +2024-03-30 07:41:33,964 INFO Train. Iter 254000 : Commit. 0.23467 PPL. 404.41 Recons. 0.00131 +2024-03-30 07:41:58,569 INFO Train. Iter 254200 : Commit. 0.23377 PPL. 404.17 Recons. 0.00130 +2024-03-30 07:42:24,299 INFO Train. Iter 254400 : Commit. 0.22776 PPL. 404.33 Recons. 0.00121 +2024-03-30 07:42:49,730 INFO Train. Iter 254600 : Commit. 0.23401 PPL. 404.58 Recons. 0.00130 +2024-03-30 07:43:15,027 INFO Train. Iter 254800 : Commit. 0.23286 PPL. 403.94 Recons. 0.00130 +2024-03-30 07:43:39,470 INFO Train. Iter 255000 : Commit. 0.23728 PPL. 403.78 Recons. 0.00136 +2024-03-30 07:44:04,441 INFO Train. Iter 255200 : Commit. 0.23753 PPL. 404.62 Recons. 0.00135 +2024-03-30 07:44:29,390 INFO Train. Iter 255400 : Commit. 0.23378 PPL. 404.79 Recons. 0.00129 +2024-03-30 07:44:54,336 INFO Train. Iter 255600 : Commit. 0.23943 PPL. 404.50 Recons. 0.00138 +2024-03-30 07:45:19,387 INFO Train. Iter 255800 : Commit. 0.23353 PPL. 404.19 Recons. 0.00130 +2024-03-30 07:45:44,846 INFO Train. Iter 256000 : Commit. 0.22923 PPL. 404.03 Recons. 0.00124 +2024-03-30 07:46:09,803 INFO Train. Iter 256200 : Commit. 0.23450 PPL. 404.77 Recons. 0.00131 +2024-03-30 07:46:36,628 INFO Train. Iter 256400 : Commit. 0.23446 PPL. 404.31 Recons. 0.00131 +2024-03-30 07:47:03,634 INFO Train. Iter 256600 : Commit. 0.23411 PPL. 404.01 Recons. 0.00132 +2024-03-30 07:47:30,345 INFO Train. Iter 256800 : Commit. 0.23398 PPL. 404.23 Recons. 0.00132 +2024-03-30 07:47:57,725 INFO Train. Iter 257000 : Commit. 0.23193 PPL. 403.68 Recons. 0.00129 +2024-03-30 07:48:25,068 INFO Train. Iter 257200 : Commit. 0.23426 PPL. 404.15 Recons. 0.00131 +2024-03-30 07:48:52,435 INFO Train. Iter 257400 : Commit. 0.23850 PPL. 405.29 Recons. 0.00135 +2024-03-30 07:49:19,443 INFO Train. Iter 257600 : Commit. 0.23267 PPL. 403.92 Recons. 0.00129 +2024-03-30 07:49:46,219 INFO Train. Iter 257800 : Commit. 0.23636 PPL. 404.58 Recons. 0.00134 +2024-03-30 07:50:11,493 INFO Train. Iter 258000 : Commit. 0.23694 PPL. 405.19 Recons. 0.00132 +2024-03-30 07:50:39,137 INFO Train. Iter 258200 : Commit. 0.22852 PPL. 404.24 Recons. 0.00123 +2024-03-30 07:51:06,639 INFO Train. Iter 258400 : Commit. 0.23267 PPL. 403.92 Recons. 0.00130 +2024-03-30 07:51:33,687 INFO Train. Iter 258600 : Commit. 0.23301 PPL. 404.34 Recons. 0.00129 +2024-03-30 07:52:00,788 INFO Train. Iter 258800 : Commit. 0.23049 PPL. 403.98 Recons. 0.00126 +2024-03-30 07:52:27,655 INFO Train. Iter 259000 : Commit. 0.23350 PPL. 403.76 Recons. 0.00131 +2024-03-30 07:52:55,952 INFO Train. Iter 259200 : Commit. 0.22795 PPL. 403.80 Recons. 0.00123 +2024-03-30 07:53:25,601 INFO Train. Iter 259400 : Commit. 0.23167 PPL. 404.33 Recons. 0.00127 +2024-03-30 07:53:55,989 INFO Train. Iter 259600 : Commit. 0.23268 PPL. 403.99 Recons. 0.00128 +2024-03-30 07:54:26,393 INFO Train. Iter 259800 : Commit. 0.23480 PPL. 404.53 Recons. 0.00131 +2024-03-30 07:54:56,033 INFO Train. Iter 260000 : Commit. 0.23028 PPL. 404.00 Recons. 0.00124 +2024-03-30 07:55:26,539 INFO Train. Iter 260200 : Commit. 0.23582 PPL. 404.15 Recons. 0.00132 +2024-03-30 07:55:58,870 INFO Train. Iter 260400 : Commit. 0.23242 PPL. 404.51 Recons. 0.00126 +2024-03-30 07:56:32,531 INFO Train. Iter 260600 : Commit. 0.23658 PPL. 403.47 Recons. 0.00136 +2024-03-30 07:57:06,116 INFO Train. Iter 260800 : Commit. 0.22661 PPL. 403.66 Recons. 0.00119 +2024-03-30 07:57:39,610 INFO Train. Iter 261000 : Commit. 0.23363 PPL. 404.16 Recons. 0.00127 +2024-03-30 07:58:11,716 INFO Train. Iter 261200 : Commit. 0.23076 PPL. 404.46 Recons. 0.00123 +2024-03-30 07:58:42,571 INFO Train. Iter 261400 : Commit. 0.22977 PPL. 403.88 Recons. 0.00124 +2024-03-30 07:59:10,854 INFO Train. Iter 261600 : Commit. 0.23390 PPL. 403.56 Recons. 0.00132 +2024-03-30 07:59:39,758 INFO Train. Iter 261800 : Commit. 0.23722 PPL. 403.29 Recons. 0.00138 +2024-03-30 08:00:08,038 INFO Train. Iter 262000 : Commit. 0.23507 PPL. 403.52 Recons. 0.00133 +2024-03-30 08:00:37,982 INFO Train. Iter 262200 : Commit. 0.23228 PPL. 403.82 Recons. 0.00127 +2024-03-30 08:01:07,674 INFO Train. Iter 262400 : Commit. 0.23078 PPL. 403.89 Recons. 0.00127 +2024-03-30 08:01:37,245 INFO Train. Iter 262600 : Commit. 0.23306 PPL. 404.08 Recons. 0.00128 +2024-03-30 08:02:06,742 INFO Train. Iter 262800 : Commit. 0.23158 PPL. 403.69 Recons. 0.00126 +2024-03-30 08:02:38,109 INFO Train. Iter 263000 : Commit. 0.23589 PPL. 404.66 Recons. 0.00131 +2024-03-30 08:03:09,024 INFO Train. Iter 263200 : Commit. 0.23614 PPL. 403.42 Recons. 0.00134 +2024-03-30 08:03:42,124 INFO Train. Iter 263400 : Commit. 0.23321 PPL. 403.51 Recons. 0.00130 +2024-03-30 08:04:13,021 INFO Train. Iter 263600 : Commit. 0.23052 PPL. 403.67 Recons. 0.00126 +2024-03-30 08:04:42,724 INFO Train. Iter 263800 : Commit. 0.23194 PPL. 404.14 Recons. 0.00126 +2024-03-30 08:05:11,305 INFO Train. Iter 264000 : Commit. 0.23570 PPL. 403.77 Recons. 0.00134 +2024-03-30 08:05:38,146 INFO Train. Iter 264200 : Commit. 0.22669 PPL. 403.30 Recons. 0.00120 +2024-03-30 08:06:04,824 INFO Train. Iter 264400 : Commit. 0.22778 PPL. 403.67 Recons. 0.00121 +2024-03-30 08:06:32,200 INFO Train. Iter 264600 : Commit. 0.22979 PPL. 403.61 Recons. 0.00124 +2024-03-30 08:06:58,783 INFO Train. Iter 264800 : Commit. 0.22556 PPL. 403.34 Recons. 0.00119 +2024-03-30 08:07:25,613 INFO Train. Iter 265000 : Commit. 0.23068 PPL. 404.33 Recons. 0.00124 +2024-03-30 08:07:52,141 INFO Train. Iter 265200 : Commit. 0.23494 PPL. 404.14 Recons. 0.00131 +2024-03-30 08:08:18,787 INFO Train. Iter 265400 : Commit. 0.23429 PPL. 403.83 Recons. 0.00131 +2024-03-30 08:08:45,871 INFO Train. Iter 265600 : Commit. 0.22638 PPL. 402.59 Recons. 0.00122 +2024-03-30 08:09:11,503 INFO Train. Iter 265800 : Commit. 0.23377 PPL. 403.95 Recons. 0.00129 +2024-03-30 08:09:38,431 INFO Train. Iter 266000 : Commit. 0.23127 PPL. 403.16 Recons. 0.00128 +2024-03-30 08:10:05,484 INFO Train. Iter 266200 : Commit. 0.23282 PPL. 404.15 Recons. 0.00128 +2024-03-30 08:10:32,768 INFO Train. Iter 266400 : Commit. 0.23047 PPL. 402.93 Recons. 0.00127 +2024-03-30 08:10:59,565 INFO Train. Iter 266600 : Commit. 0.23061 PPL. 403.29 Recons. 0.00125 +2024-03-30 08:11:26,247 INFO Train. Iter 266800 : Commit. 0.23066 PPL. 402.45 Recons. 0.00129 +2024-03-30 08:11:53,613 INFO Train. Iter 267000 : Commit. 0.22911 PPL. 403.08 Recons. 0.00124 +2024-03-30 08:12:21,098 INFO Train. Iter 267200 : Commit. 0.23155 PPL. 403.37 Recons. 0.00127 +2024-03-30 08:12:48,273 INFO Train. Iter 267400 : Commit. 0.23204 PPL. 403.64 Recons. 0.00128 +2024-03-30 08:13:16,562 INFO Train. Iter 267600 : Commit. 0.22979 PPL. 403.50 Recons. 0.00124 +2024-03-30 08:13:43,553 INFO Train. Iter 267800 : Commit. 0.23068 PPL. 403.41 Recons. 0.00126 +2024-03-30 08:14:11,446 INFO Train. Iter 268000 : Commit. 0.22901 PPL. 403.38 Recons. 0.00123 +2024-03-30 08:14:39,980 INFO Train. Iter 268200 : Commit. 0.22469 PPL. 403.16 Recons. 0.00117 +2024-03-30 08:15:07,903 INFO Train. Iter 268400 : Commit. 0.22624 PPL. 402.93 Recons. 0.00121 +2024-03-30 08:15:36,207 INFO Train. Iter 268600 : Commit. 0.23114 PPL. 404.01 Recons. 0.00125 +2024-03-30 08:16:04,398 INFO Train. Iter 268800 : Commit. 0.23485 PPL. 402.95 Recons. 0.00133 +2024-03-30 08:16:31,738 INFO Train. Iter 269000 : Commit. 0.22783 PPL. 403.34 Recons. 0.00121 +2024-03-30 08:16:58,968 INFO Train. Iter 269200 : Commit. 0.22750 PPL. 402.71 Recons. 0.00121 +2024-03-30 08:17:25,929 INFO Train. Iter 269400 : Commit. 0.23807 PPL. 402.73 Recons. 0.00139 +2024-03-30 08:17:52,362 INFO Train. Iter 269600 : Commit. 0.22948 PPL. 402.55 Recons. 0.00126 +2024-03-30 08:18:20,126 INFO Train. Iter 269800 : Commit. 0.23357 PPL. 403.58 Recons. 0.00130 +2024-03-30 08:18:47,572 INFO Train. Iter 270000 : Commit. 0.23307 PPL. 404.07 Recons. 0.00127 +2024-03-30 08:19:15,108 INFO Train. Iter 270200 : Commit. 0.23028 PPL. 402.71 Recons. 0.00126 +2024-03-30 08:19:41,963 INFO Train. Iter 270400 : Commit. 0.23115 PPL. 403.37 Recons. 0.00126 +2024-03-30 08:20:08,999 INFO Train. Iter 270600 : Commit. 0.23250 PPL. 403.31 Recons. 0.00129 +2024-03-30 08:20:36,443 INFO Train. Iter 270800 : Commit. 0.23056 PPL. 403.36 Recons. 0.00126 +2024-03-30 08:21:03,736 INFO Train. Iter 271000 : Commit. 0.22638 PPL. 402.85 Recons. 0.00121 +2024-03-30 08:21:31,275 INFO Train. Iter 271200 : Commit. 0.23156 PPL. 402.88 Recons. 0.00128 +2024-03-30 08:21:58,209 INFO Train. Iter 271400 : Commit. 0.22961 PPL. 403.22 Recons. 0.00123 +2024-03-30 08:22:24,481 INFO Train. Iter 271600 : Commit. 0.23017 PPL. 402.94 Recons. 0.00127 +2024-03-30 08:22:52,800 INFO Train. Iter 271800 : Commit. 0.22556 PPL. 402.57 Recons. 0.00120 +2024-03-30 08:23:20,591 INFO Train. Iter 272000 : Commit. 0.23152 PPL. 402.43 Recons. 0.00130 +2024-03-30 08:23:49,324 INFO Train. Iter 272200 : Commit. 0.22968 PPL. 402.72 Recons. 0.00126 +2024-03-30 08:24:16,686 INFO Train. Iter 272400 : Commit. 0.23063 PPL. 402.92 Recons. 0.00127 +2024-03-30 08:24:44,917 INFO Train. Iter 272600 : Commit. 0.22935 PPL. 402.94 Recons. 0.00126 +2024-03-30 08:25:12,388 INFO Train. Iter 272800 : Commit. 0.23298 PPL. 402.63 Recons. 0.00131 +2024-03-30 08:25:39,792 INFO Train. Iter 273000 : Commit. 0.22835 PPL. 403.14 Recons. 0.00124 +2024-03-30 08:26:09,159 INFO Train. Iter 273200 : Commit. 0.23092 PPL. 403.44 Recons. 0.00128 +2024-03-30 08:26:38,015 INFO Train. Iter 273400 : Commit. 0.22863 PPL. 402.93 Recons. 0.00125 +2024-03-30 08:27:05,603 INFO Train. Iter 273600 : Commit. 0.22712 PPL. 403.03 Recons. 0.00122 +2024-03-30 08:27:37,117 INFO Train. Iter 273800 : Commit. 0.22996 PPL. 402.89 Recons. 0.00128 +2024-03-30 08:28:08,279 INFO Train. Iter 274000 : Commit. 0.23278 PPL. 403.69 Recons. 0.00129 +2024-03-30 08:28:38,843 INFO Train. Iter 274200 : Commit. 0.23280 PPL. 403.11 Recons. 0.00131 +2024-03-30 08:29:09,416 INFO Train. Iter 274400 : Commit. 0.23312 PPL. 402.79 Recons. 0.00132 +2024-03-30 08:29:39,611 INFO Train. Iter 274600 : Commit. 0.23055 PPL. 403.80 Recons. 0.00126 +2024-03-30 08:30:09,545 INFO Train. Iter 274800 : Commit. 0.23277 PPL. 402.64 Recons. 0.00131 +2024-03-30 08:30:39,851 INFO Train. Iter 275000 : Commit. 0.23153 PPL. 403.05 Recons. 0.00129 +2024-03-30 08:31:10,258 INFO Train. Iter 275200 : Commit. 0.22954 PPL. 402.94 Recons. 0.00127 +2024-03-30 08:31:38,688 INFO Train. Iter 275400 : Commit. 0.23110 PPL. 403.02 Recons. 0.00129 +2024-03-30 08:32:08,507 INFO Train. Iter 275600 : Commit. 0.22901 PPL. 403.13 Recons. 0.00124 +2024-03-30 08:32:37,959 INFO Train. Iter 275800 : Commit. 0.22876 PPL. 403.40 Recons. 0.00124 +2024-03-30 08:33:07,815 INFO Train. Iter 276000 : Commit. 0.23014 PPL. 403.06 Recons. 0.00126 +2024-03-30 08:33:37,580 INFO Train. Iter 276200 : Commit. 0.22536 PPL. 402.59 Recons. 0.00120 +2024-03-30 08:34:07,670 INFO Train. Iter 276400 : Commit. 0.23226 PPL. 402.95 Recons. 0.00130 +2024-03-30 08:34:37,288 INFO Train. Iter 276600 : Commit. 0.23045 PPL. 402.65 Recons. 0.00128 +2024-03-30 08:35:06,823 INFO Train. Iter 276800 : Commit. 0.22960 PPL. 403.28 Recons. 0.00125 +2024-03-30 08:35:40,365 INFO Train. Iter 277000 : Commit. 0.23573 PPL. 403.18 Recons. 0.00135 +2024-03-30 08:36:13,680 INFO Train. Iter 277200 : Commit. 0.22765 PPL. 402.01 Recons. 0.00126 +2024-03-30 08:36:44,950 INFO Train. Iter 277400 : Commit. 0.22384 PPL. 402.20 Recons. 0.00119 +2024-03-30 08:37:16,674 INFO Train. Iter 277600 : Commit. 0.23411 PPL. 401.93 Recons. 0.00136 +2024-03-30 08:37:50,423 INFO Train. Iter 277800 : Commit. 0.22889 PPL. 402.52 Recons. 0.00126 +2024-03-30 08:38:23,995 INFO Train. Iter 278000 : Commit. 0.23105 PPL. 403.12 Recons. 0.00127 +2024-03-30 08:38:57,301 INFO Train. Iter 278200 : Commit. 0.22896 PPL. 402.62 Recons. 0.00125 +2024-03-30 08:39:30,815 INFO Train. Iter 278400 : Commit. 0.23139 PPL. 402.68 Recons. 0.00130 +2024-03-30 08:40:03,916 INFO Train. Iter 278600 : Commit. 0.23128 PPL. 403.70 Recons. 0.00126 +2024-03-30 08:40:37,026 INFO Train. Iter 278800 : Commit. 0.23505 PPL. 402.82 Recons. 0.00135 +2024-03-30 08:41:09,128 INFO Train. Iter 279000 : Commit. 0.22550 PPL. 402.00 Recons. 0.00122 +2024-03-30 08:41:36,116 INFO Train. Iter 279200 : Commit. 0.22509 PPL. 402.49 Recons. 0.00120 +2024-03-30 08:42:01,026 INFO Train. Iter 279400 : Commit. 0.23091 PPL. 402.81 Recons. 0.00128 +2024-03-30 08:42:27,477 INFO Train. Iter 279600 : Commit. 0.22605 PPL. 401.91 Recons. 0.00123 +2024-03-30 08:42:53,870 INFO Train. Iter 279800 : Commit. 0.23137 PPL. 402.17 Recons. 0.00131 +2024-03-30 08:43:20,432 INFO Train. Iter 280000 : Commit. 0.23394 PPL. 403.73 Recons. 0.00131 +2024-03-30 08:43:46,679 INFO Train. Iter 280200 : Commit. 0.22943 PPL. 402.57 Recons. 0.00127 +2024-03-30 08:44:12,960 INFO Train. Iter 280400 : Commit. 0.22577 PPL. 402.10 Recons. 0.00123 +2024-03-30 08:44:39,214 INFO Train. Iter 280600 : Commit. 0.23048 PPL. 402.87 Recons. 0.00128 +2024-03-30 08:45:04,549 INFO Train. Iter 280800 : Commit. 0.23551 PPL. 403.98 Recons. 0.00132 +2024-03-30 08:45:30,613 INFO Train. Iter 281000 : Commit. 0.22334 PPL. 401.54 Recons. 0.00120 +2024-03-30 08:45:56,130 INFO Train. Iter 281200 : Commit. 0.22758 PPL. 402.09 Recons. 0.00126 +2024-03-30 08:46:22,435 INFO Train. Iter 281400 : Commit. 0.23318 PPL. 402.93 Recons. 0.00132 +2024-03-30 08:46:48,678 INFO Train. Iter 281600 : Commit. 0.23173 PPL. 402.65 Recons. 0.00130 +2024-03-30 08:47:15,804 INFO Train. Iter 281800 : Commit. 0.22817 PPL. 402.11 Recons. 0.00126 +2024-03-30 08:47:42,517 INFO Train. Iter 282000 : Commit. 0.22851 PPL. 402.49 Recons. 0.00125 +2024-03-30 08:48:08,825 INFO Train. Iter 282200 : Commit. 0.22972 PPL. 402.61 Recons. 0.00128 +2024-03-30 08:48:34,762 INFO Train. Iter 282400 : Commit. 0.23255 PPL. 403.10 Recons. 0.00130 +2024-03-30 08:49:00,313 INFO Train. Iter 282600 : Commit. 0.22776 PPL. 401.73 Recons. 0.00126 +2024-03-30 08:49:26,008 INFO Train. Iter 282800 : Commit. 0.23203 PPL. 402.65 Recons. 0.00131 +2024-03-30 08:49:52,620 INFO Train. Iter 283000 : Commit. 0.23165 PPL. 403.75 Recons. 0.00128 +2024-03-30 08:50:18,224 INFO Train. Iter 283200 : Commit. 0.22885 PPL. 402.62 Recons. 0.00126 +2024-03-30 08:50:44,569 INFO Train. Iter 283400 : Commit. 0.23325 PPL. 402.22 Recons. 0.00133 +2024-03-30 08:51:11,161 INFO Train. Iter 283600 : Commit. 0.22843 PPL. 402.41 Recons. 0.00126 +2024-03-30 08:51:37,948 INFO Train. Iter 283800 : Commit. 0.22803 PPL. 402.04 Recons. 0.00127 +2024-03-30 08:52:05,168 INFO Train. Iter 284000 : Commit. 0.22776 PPL. 403.00 Recons. 0.00123 +2024-03-30 08:52:31,927 INFO Train. Iter 284200 : Commit. 0.22832 PPL. 401.40 Recons. 0.00127 +2024-03-30 08:52:58,875 INFO Train. Iter 284400 : Commit. 0.22964 PPL. 402.82 Recons. 0.00125 +2024-03-30 08:53:25,765 INFO Train. Iter 284600 : Commit. 0.23049 PPL. 403.47 Recons. 0.00127 +2024-03-30 08:53:52,861 INFO Train. Iter 284800 : Commit. 0.23259 PPL. 402.64 Recons. 0.00132 +2024-03-30 08:54:20,923 INFO Train. Iter 285000 : Commit. 0.23194 PPL. 402.92 Recons. 0.00130 +2024-03-30 08:54:47,403 INFO Train. Iter 285200 : Commit. 0.22895 PPL. 402.22 Recons. 0.00125 +2024-03-30 08:55:15,596 INFO Train. Iter 285400 : Commit. 0.23065 PPL. 402.80 Recons. 0.00128 +2024-03-30 08:55:44,017 INFO Train. Iter 285600 : Commit. 0.22743 PPL. 402.21 Recons. 0.00125 +2024-03-30 08:56:12,622 INFO Train. Iter 285800 : Commit. 0.22889 PPL. 402.09 Recons. 0.00127 +2024-03-30 08:56:41,151 INFO Train. Iter 286000 : Commit. 0.23159 PPL. 402.54 Recons. 0.00131 +2024-03-30 08:57:09,825 INFO Train. Iter 286200 : Commit. 0.23200 PPL. 403.16 Recons. 0.00129 +2024-03-30 08:57:38,628 INFO Train. Iter 286400 : Commit. 0.22658 PPL. 402.01 Recons. 0.00123 +2024-03-30 08:58:06,907 INFO Train. Iter 286600 : Commit. 0.22755 PPL. 402.62 Recons. 0.00123 +2024-03-30 08:58:35,631 INFO Train. Iter 286800 : Commit. 0.22855 PPL. 401.87 Recons. 0.00128 +2024-03-30 08:59:03,125 INFO Train. Iter 287000 : Commit. 0.22782 PPL. 402.67 Recons. 0.00125 +2024-03-30 08:59:30,874 INFO Train. Iter 287200 : Commit. 0.22957 PPL. 402.33 Recons. 0.00128 +2024-03-30 08:59:58,539 INFO Train. Iter 287400 : Commit. 0.22640 PPL. 402.17 Recons. 0.00124 +2024-03-30 09:00:25,828 INFO Train. Iter 287600 : Commit. 0.22594 PPL. 401.70 Recons. 0.00123 +2024-03-30 09:00:53,515 INFO Train. Iter 287800 : Commit. 0.22787 PPL. 402.13 Recons. 0.00126 +2024-03-30 09:01:21,326 INFO Train. Iter 288000 : Commit. 0.22720 PPL. 402.25 Recons. 0.00124 +2024-03-30 09:01:48,529 INFO Train. Iter 288200 : Commit. 0.22737 PPL. 401.65 Recons. 0.00126 +2024-03-30 09:02:16,513 INFO Train. Iter 288400 : Commit. 0.22968 PPL. 402.25 Recons. 0.00127 +2024-03-30 09:02:44,458 INFO Train. Iter 288600 : Commit. 0.22856 PPL. 401.85 Recons. 0.00127 +2024-03-30 09:03:12,086 INFO Train. Iter 288800 : Commit. 0.23116 PPL. 402.06 Recons. 0.00130 +2024-03-30 09:03:38,819 INFO Train. Iter 289000 : Commit. 0.23455 PPL. 403.66 Recons. 0.00132 +2024-03-30 09:04:06,774 INFO Train. Iter 289200 : Commit. 0.23095 PPL. 401.58 Recons. 0.00131 +2024-03-30 09:04:35,881 INFO Train. Iter 289400 : Commit. 0.22702 PPL. 401.27 Recons. 0.00126 +2024-03-30 09:05:03,954 INFO Train. Iter 289600 : Commit. 0.23070 PPL. 402.63 Recons. 0.00129 +2024-03-30 09:05:32,742 INFO Train. Iter 289800 : Commit. 0.22724 PPL. 401.53 Recons. 0.00125 +2024-03-30 09:06:01,521 INFO Train. Iter 290000 : Commit. 0.22731 PPL. 403.01 Recons. 0.00121 +2024-03-30 09:06:29,843 INFO Train. Iter 290200 : Commit. 0.22700 PPL. 403.60 Recons. 0.00121 +2024-03-30 09:06:58,584 INFO Train. Iter 290400 : Commit. 0.23030 PPL. 402.35 Recons. 0.00128 +2024-03-30 09:07:27,332 INFO Train. Iter 290600 : Commit. 0.23337 PPL. 402.04 Recons. 0.00133 +2024-03-30 09:07:55,709 INFO Train. Iter 290800 : Commit. 0.23165 PPL. 402.81 Recons. 0.00128 +2024-03-30 09:08:23,327 INFO Train. Iter 291000 : Commit. 0.22580 PPL. 401.61 Recons. 0.00124 +2024-03-30 09:08:51,574 INFO Train. Iter 291200 : Commit. 0.22739 PPL. 402.67 Recons. 0.00123 +2024-03-30 09:09:20,114 INFO Train. Iter 291400 : Commit. 0.23180 PPL. 402.28 Recons. 0.00130 +2024-03-30 09:09:47,637 INFO Train. Iter 291600 : Commit. 0.22452 PPL. 402.02 Recons. 0.00120 +2024-03-30 09:10:15,858 INFO Train. Iter 291800 : Commit. 0.23302 PPL. 401.85 Recons. 0.00134 +2024-03-30 09:10:50,263 INFO Train. Iter 292000 : Commit. 0.22608 PPL. 401.64 Recons. 0.00123 +2024-03-30 09:11:23,999 INFO Train. Iter 292200 : Commit. 0.22573 PPL. 402.47 Recons. 0.00120 +2024-03-30 09:11:58,626 INFO Train. Iter 292400 : Commit. 0.23663 PPL. 402.70 Recons. 0.00137 +2024-03-30 09:12:31,294 INFO Train. Iter 292600 : Commit. 0.22886 PPL. 402.89 Recons. 0.00125 +2024-03-30 09:13:01,697 INFO Train. Iter 292800 : Commit. 0.22938 PPL. 401.93 Recons. 0.00128 +2024-03-30 09:13:33,565 INFO Train. Iter 293000 : Commit. 0.22745 PPL. 402.23 Recons. 0.00125 +2024-03-30 09:14:04,819 INFO Train. Iter 293200 : Commit. 0.22466 PPL. 401.77 Recons. 0.00122 +2024-03-30 09:14:32,593 INFO Train. Iter 293400 : Commit. 0.22973 PPL. 402.79 Recons. 0.00128 +2024-03-30 09:14:58,386 INFO Train. Iter 293600 : Commit. 0.23303 PPL. 402.67 Recons. 0.00133 +2024-03-30 09:15:23,892 INFO Train. Iter 293800 : Commit. 0.22973 PPL. 402.13 Recons. 0.00129 +2024-03-30 09:15:49,716 INFO Train. Iter 294000 : Commit. 0.22768 PPL. 401.95 Recons. 0.00126 +2024-03-30 09:16:15,205 INFO Train. Iter 294200 : Commit. 0.22737 PPL. 402.67 Recons. 0.00124 +2024-03-30 09:16:41,507 INFO Train. Iter 294400 : Commit. 0.22829 PPL. 403.02 Recons. 0.00125 +2024-03-30 09:17:08,299 INFO Train. Iter 294600 : Commit. 0.22491 PPL. 401.35 Recons. 0.00123 +2024-03-30 09:17:33,107 INFO Train. Iter 294800 : Commit. 0.22964 PPL. 401.70 Recons. 0.00129 +2024-03-30 09:17:58,887 INFO Train. Iter 295000 : Commit. 0.23408 PPL. 402.58 Recons. 0.00134 +2024-03-30 09:18:24,576 INFO Train. Iter 295200 : Commit. 0.23018 PPL. 402.70 Recons. 0.00127 +2024-03-30 09:18:50,791 INFO Train. Iter 295400 : Commit. 0.23118 PPL. 401.67 Recons. 0.00132 +2024-03-30 09:19:16,644 INFO Train. Iter 295600 : Commit. 0.22926 PPL. 402.04 Recons. 0.00129 +2024-03-30 09:19:42,367 INFO Train. Iter 295800 : Commit. 0.22716 PPL. 401.65 Recons. 0.00126 +2024-03-30 09:20:08,831 INFO Train. Iter 296000 : Commit. 0.22721 PPL. 402.40 Recons. 0.00124 +2024-03-30 09:20:35,151 INFO Train. Iter 296200 : Commit. 0.22922 PPL. 402.12 Recons. 0.00128 +2024-03-30 09:21:00,061 INFO Train. Iter 296400 : Commit. 0.23067 PPL. 402.03 Recons. 0.00130 +2024-03-30 09:21:25,411 INFO Train. Iter 296600 : Commit. 0.22408 PPL. 401.75 Recons. 0.00122 +2024-03-30 09:21:49,767 INFO Train. Iter 296800 : Commit. 0.22580 PPL. 401.54 Recons. 0.00124 +2024-03-30 09:22:15,935 INFO Train. Iter 297000 : Commit. 0.22334 PPL. 401.47 Recons. 0.00119 +2024-03-30 09:22:41,792 INFO Train. Iter 297200 : Commit. 0.22972 PPL. 402.02 Recons. 0.00129 +2024-03-30 09:23:07,821 INFO Train. Iter 297400 : Commit. 0.22773 PPL. 402.09 Recons. 0.00126 +2024-03-30 09:23:34,488 INFO Train. Iter 297600 : Commit. 0.22746 PPL. 402.08 Recons. 0.00125 +2024-03-30 09:24:01,215 INFO Train. Iter 297800 : Commit. 0.22873 PPL. 401.33 Recons. 0.00128 +2024-03-30 09:24:28,073 INFO Train. Iter 298000 : Commit. 0.22765 PPL. 402.36 Recons. 0.00126 +2024-03-30 09:24:55,013 INFO Train. Iter 298200 : Commit. 0.22468 PPL. 401.32 Recons. 0.00122 +2024-03-30 09:25:21,257 INFO Train. Iter 298400 : Commit. 0.22922 PPL. 401.69 Recons. 0.00129 +2024-03-30 09:25:47,247 INFO Train. Iter 298600 : Commit. 0.22556 PPL. 401.80 Recons. 0.00123 +2024-03-30 09:26:14,411 INFO Train. Iter 298800 : Commit. 0.22548 PPL. 402.03 Recons. 0.00122 +2024-03-30 09:26:41,409 INFO Train. Iter 299000 : Commit. 0.22304 PPL. 401.03 Recons. 0.00120 +2024-03-30 09:27:07,770 INFO Train. Iter 299200 : Commit. 0.22709 PPL. 401.56 Recons. 0.00125 +2024-03-30 09:27:35,587 INFO Train. Iter 299400 : Commit. 0.22939 PPL. 401.51 Recons. 0.00128 +2024-03-30 09:28:03,781 INFO Train. Iter 299600 : Commit. 0.22782 PPL. 401.35 Recons. 0.00127 +2024-03-30 09:28:31,714 INFO Train. Iter 299800 : Commit. 0.22467 PPL. 401.70 Recons. 0.00121 +2024-03-30 09:28:58,527 INFO Train. Iter 300000 : Commit. 0.22797 PPL. 402.10 Recons. 0.00125 diff --git a/ckpt/beatx2_rvqvae/RVQVAE_upper/net_300000.pth b/ckpt/beatx2_rvqvae/RVQVAE_upper/net_300000.pth new file mode 100644 index 0000000000000000000000000000000000000000..f1e873ed757ae8fa47e1607c8c889864e173a675 --- /dev/null +++ b/ckpt/beatx2_rvqvae/RVQVAE_upper/net_300000.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:959d066138b293455a98fb0175b1fe2fcc31da9a1de83c9bfbf093ffea746a0e +size 81794923 diff --git a/ckpt/beatx2_rvqvae/RVQVAE_upper/run.log b/ckpt/beatx2_rvqvae/RVQVAE_upper/run.log new file mode 100644 index 0000000000000000000000000000000000000000..ea5fc2229a4aba103d1ce73547ec2a87e2d12d1a --- /dev/null +++ b/ckpt/beatx2_rvqvae/RVQVAE_upper/run.log @@ -0,0 +1,1546 @@ +2024-03-16 22:22:01,004 INFO { + "batch_size": 256, + "beta": 1.0, + "body_part": "upper", + "code_dim": 512, + "commit": 0.02, + "dataname": "kit", + "depth": 3, + "dilation_growth_rate": 3, + "down_t": 2, + "eval_iter": 1000, + "exp_name": "RVQVAE", + "gamma": 0.05, + "loss_vel": 0.5, + "lr": 0.0002, + "lr_scheduler": [ + 200000 + ], + "mu": 0.99, + "nb_code": 512, + "nb_vis": 20, + "out_dir": "output_beatx2/RVQVAE_upper", + "output_emb_width": 512, + "print_iter": 200, + "quantizer": "ema_reset", + "recons_loss": "l1_smooth", + "results_dir": "visual_results/", + "resume_gpt": null, + "resume_pth": null, + "seed": 123, + "stride_t": 2, + "total_iter": 300000, + "vis_gt": false, + "visual_name": "baseline", + "vq_act": "relu", + "vq_norm": null, + "warm_up_iter": 1000, + "weight_decay": 0.0, + "width": 512, + "window_size": 64 +} +2024-03-16 22:22:01,011 INFO Training on kit, motions are with 21 joints +2024-03-16 22:22:22,467 INFO Warmup. Iter 200 : lr 0.00004 Commit. 0.06728 PPL. 238.77 Recons. 0.30773 +2024-03-16 22:22:43,943 INFO Warmup. Iter 400 : lr 0.00008 Commit. 0.28404 PPL. 188.42 Recons. 0.14487 +2024-03-16 22:23:05,297 INFO Warmup. Iter 600 : lr 0.00012 Commit. 0.63379 PPL. 338.16 Recons. 0.09233 +2024-03-16 22:23:26,692 INFO Warmup. Iter 800 : lr 0.00016 Commit. 1.21107 PPL. 402.01 Recons. 0.06139 +2024-03-16 22:24:09,176 INFO Train. Iter 200 : Commit. 1.62579 PPL. 421.79 Recons. 0.04176 +2024-03-16 22:24:31,179 INFO Train. Iter 400 : Commit. 1.73617 PPL. 423.47 Recons. 0.03526 +2024-03-16 22:24:53,285 INFO Train. Iter 600 : Commit. 1.80605 PPL. 425.27 Recons. 0.03101 +2024-03-16 22:25:15,562 INFO Train. Iter 800 : Commit. 1.83055 PPL. 429.13 Recons. 0.02539 +2024-03-16 22:25:36,206 INFO Train. Iter 1000 : Commit. 1.88383 PPL. 428.03 Recons. 0.02567 +2024-03-16 22:25:58,253 INFO Train. Iter 1200 : Commit. 1.82084 PPL. 431.15 Recons. 0.02119 +2024-03-16 22:26:20,323 INFO Train. Iter 1400 : Commit. 1.88775 PPL. 427.82 Recons. 0.02260 +2024-03-16 22:26:42,460 INFO Train. Iter 1600 : Commit. 1.84252 PPL. 427.97 Recons. 0.02148 +2024-03-16 22:27:04,726 INFO Train. Iter 1800 : Commit. 1.83931 PPL. 429.19 Recons. 0.01979 +2024-03-16 22:27:27,405 INFO Train. Iter 2000 : Commit. 1.78415 PPL. 428.74 Recons. 0.02003 +2024-03-16 22:27:49,413 INFO Train. Iter 2200 : Commit. 1.81204 PPL. 427.72 Recons. 0.02072 +2024-03-16 22:28:11,604 INFO Train. Iter 2400 : Commit. 1.83479 PPL. 428.15 Recons. 0.02117 +2024-03-16 22:28:34,118 INFO Train. Iter 2600 : Commit. 1.73465 PPL. 429.58 Recons. 0.01811 +2024-03-16 22:28:55,061 INFO Train. Iter 2800 : Commit. 1.74408 PPL. 429.18 Recons. 0.01834 +2024-03-16 22:29:17,028 INFO Train. Iter 3000 : Commit. 1.66951 PPL. 429.79 Recons. 0.01615 +2024-03-16 22:29:38,657 INFO Train. Iter 3200 : Commit. 1.74645 PPL. 428.15 Recons. 0.01872 +2024-03-16 22:30:00,398 INFO Train. Iter 3400 : Commit. 1.68809 PPL. 428.99 Recons. 0.01644 +2024-03-16 22:30:22,594 INFO Train. Iter 3600 : Commit. 1.65129 PPL. 429.57 Recons. 0.01689 +2024-03-16 22:30:44,325 INFO Train. Iter 3800 : Commit. 1.67007 PPL. 428.21 Recons. 0.01779 +2024-03-16 22:31:06,195 INFO Train. Iter 4000 : Commit. 1.65058 PPL. 429.26 Recons. 0.01664 +2024-03-16 22:31:28,327 INFO Train. Iter 4200 : Commit. 1.57990 PPL. 428.44 Recons. 0.01493 +2024-03-16 22:31:50,159 INFO Train. Iter 4400 : Commit. 1.59555 PPL. 428.47 Recons. 0.01667 +2024-03-16 22:32:12,102 INFO Train. Iter 4600 : Commit. 1.54425 PPL. 428.39 Recons. 0.01604 +2024-03-16 22:32:32,964 INFO Train. Iter 4800 : Commit. 1.55946 PPL. 428.69 Recons. 0.01625 +2024-03-16 22:32:55,196 INFO Train. Iter 5000 : Commit. 1.51459 PPL. 429.10 Recons. 0.01584 +2024-03-16 22:33:16,848 INFO Train. Iter 5200 : Commit. 1.51071 PPL. 429.23 Recons. 0.01615 +2024-03-16 22:33:38,291 INFO Train. Iter 5400 : Commit. 1.53405 PPL. 428.18 Recons. 0.01565 +2024-03-16 22:34:00,073 INFO Train. Iter 5600 : Commit. 1.42781 PPL. 431.93 Recons. 0.01257 +2024-03-16 22:34:22,100 INFO Train. Iter 5800 : Commit. 1.39522 PPL. 430.59 Recons. 0.01312 +2024-03-16 22:34:44,554 INFO Train. Iter 6000 : Commit. 1.41533 PPL. 431.06 Recons. 0.01367 +2024-03-16 22:35:06,778 INFO Train. Iter 6200 : Commit. 1.47642 PPL. 427.45 Recons. 0.01670 +2024-03-16 22:35:28,528 INFO Train. Iter 6400 : Commit. 1.33733 PPL. 432.87 Recons. 0.01128 +2024-03-16 22:35:50,700 INFO Train. Iter 6600 : Commit. 1.33891 PPL. 431.94 Recons. 0.01217 +2024-03-16 22:36:11,464 INFO Train. Iter 6800 : Commit. 1.39410 PPL. 430.03 Recons. 0.01441 +2024-03-16 22:36:33,297 INFO Train. Iter 7000 : Commit. 1.33388 PPL. 432.89 Recons. 0.01119 +2024-03-16 22:36:55,316 INFO Train. Iter 7200 : Commit. 1.37480 PPL. 430.64 Recons. 0.01379 +2024-03-16 22:37:16,814 INFO Train. Iter 7400 : Commit. 1.34208 PPL. 431.20 Recons. 0.01315 +2024-03-16 22:37:37,611 INFO Train. Iter 7600 : Commit. 1.35375 PPL. 430.64 Recons. 0.01322 +2024-03-16 22:37:58,375 INFO Train. Iter 7800 : Commit. 1.33142 PPL. 431.29 Recons. 0.01320 +2024-03-16 22:38:19,484 INFO Train. Iter 8000 : Commit. 1.36695 PPL. 428.01 Recons. 0.01503 +2024-03-16 22:38:40,265 INFO Train. Iter 8200 : Commit. 1.40111 PPL. 428.84 Recons. 0.01589 +2024-03-16 22:39:00,820 INFO Train. Iter 8400 : Commit. 1.34893 PPL. 430.81 Recons. 0.01355 +2024-03-16 22:39:20,540 INFO Train. Iter 8600 : Commit. 1.29258 PPL. 431.51 Recons. 0.01191 +2024-03-16 22:39:41,772 INFO Train. Iter 8800 : Commit. 1.27403 PPL. 429.69 Recons. 0.01332 +2024-03-16 22:40:02,829 INFO Train. Iter 9000 : Commit. 1.26560 PPL. 431.73 Recons. 0.01206 +2024-03-16 22:40:24,808 INFO Train. Iter 9200 : Commit. 1.23255 PPL. 432.14 Recons. 0.01185 +2024-03-16 22:40:46,763 INFO Train. Iter 9400 : Commit. 1.23568 PPL. 431.91 Recons. 0.01165 +2024-03-16 22:41:08,646 INFO Train. Iter 9600 : Commit. 1.23634 PPL. 430.21 Recons. 0.01250 +2024-03-16 22:41:30,527 INFO Train. Iter 9800 : Commit. 1.22062 PPL. 432.19 Recons. 0.01180 +2024-03-16 22:41:52,634 INFO Train. Iter 10000 : Commit. 1.27515 PPL. 430.01 Recons. 0.01456 +2024-03-16 22:42:14,530 INFO Train. Iter 10200 : Commit. 1.24988 PPL. 431.63 Recons. 0.01239 +2024-03-16 22:42:36,190 INFO Train. Iter 10400 : Commit. 1.25542 PPL. 431.46 Recons. 0.01303 +2024-03-16 22:42:56,758 INFO Train. Iter 10600 : Commit. 1.22280 PPL. 431.43 Recons. 0.01186 +2024-03-16 22:43:18,497 INFO Train. Iter 10800 : Commit. 1.14994 PPL. 434.16 Recons. 0.01003 +2024-03-16 22:43:40,711 INFO Train. Iter 11000 : Commit. 1.15821 PPL. 431.73 Recons. 0.01217 +2024-03-16 22:44:02,485 INFO Train. Iter 11200 : Commit. 1.15549 PPL. 432.78 Recons. 0.01146 +2024-03-16 22:44:24,395 INFO Train. Iter 11400 : Commit. 1.17293 PPL. 431.92 Recons. 0.01231 +2024-03-16 22:44:46,290 INFO Train. Iter 11600 : Commit. 1.24207 PPL. 430.02 Recons. 0.01389 +2024-03-16 22:45:08,066 INFO Train. Iter 11800 : Commit. 1.18575 PPL. 432.78 Recons. 0.01098 +2024-03-16 22:45:29,861 INFO Train. Iter 12000 : Commit. 1.22956 PPL. 432.15 Recons. 0.01352 +2024-03-16 22:45:51,782 INFO Train. Iter 12200 : Commit. 1.22806 PPL. 431.81 Recons. 0.01259 +2024-03-16 22:46:14,057 INFO Train. Iter 12400 : Commit. 1.20161 PPL. 431.39 Recons. 0.01195 +2024-03-16 22:46:35,002 INFO Train. Iter 12600 : Commit. 1.19220 PPL. 432.98 Recons. 0.01138 +2024-03-16 22:46:57,111 INFO Train. Iter 12800 : Commit. 1.13741 PPL. 433.84 Recons. 0.01032 +2024-03-16 22:47:18,737 INFO Train. Iter 13000 : Commit. 1.20114 PPL. 431.05 Recons. 0.01324 +2024-03-16 22:47:40,438 INFO Train. Iter 13200 : Commit. 1.17927 PPL. 430.13 Recons. 0.01387 +2024-03-16 22:48:02,607 INFO Train. Iter 13400 : Commit. 1.15600 PPL. 432.43 Recons. 0.01243 +2024-03-16 22:48:24,960 INFO Train. Iter 13600 : Commit. 1.13145 PPL. 434.22 Recons. 0.01025 +2024-03-16 22:48:47,116 INFO Train. Iter 13800 : Commit. 1.09488 PPL. 434.94 Recons. 0.01003 +2024-03-16 22:49:09,307 INFO Train. Iter 14000 : Commit. 1.07121 PPL. 435.07 Recons. 0.00935 +2024-03-16 22:49:31,391 INFO Train. Iter 14200 : Commit. 1.09643 PPL. 432.09 Recons. 0.01120 +2024-03-16 22:49:52,005 INFO Train. Iter 14400 : Commit. 1.10040 PPL. 433.91 Recons. 0.01123 +2024-03-16 22:50:13,754 INFO Train. Iter 14600 : Commit. 1.11082 PPL. 432.38 Recons. 0.01145 +2024-03-16 22:50:35,418 INFO Train. Iter 14800 : Commit. 1.04361 PPL. 435.23 Recons. 0.00937 +2024-03-16 22:50:57,262 INFO Train. Iter 15000 : Commit. 1.05142 PPL. 436.43 Recons. 0.00883 +2024-03-16 22:51:19,399 INFO Train. Iter 15200 : Commit. 1.13384 PPL. 432.21 Recons. 0.01275 +2024-03-16 22:51:40,629 INFO Train. Iter 15400 : Commit. 1.14655 PPL. 433.31 Recons. 0.01175 +2024-03-16 22:52:01,563 INFO Train. Iter 15600 : Commit. 1.04708 PPL. 435.64 Recons. 0.00944 +2024-03-16 22:52:23,128 INFO Train. Iter 15800 : Commit. 1.12236 PPL. 431.34 Recons. 0.01194 +2024-03-16 22:52:44,161 INFO Train. Iter 16000 : Commit. 1.05442 PPL. 434.25 Recons. 0.01011 +2024-03-16 22:53:05,900 INFO Train. Iter 16200 : Commit. 1.05639 PPL. 435.56 Recons. 0.00955 +2024-03-16 22:53:26,177 INFO Train. Iter 16400 : Commit. 1.02300 PPL. 434.45 Recons. 0.00983 +2024-03-16 22:53:48,322 INFO Train. Iter 16600 : Commit. 0.99811 PPL. 436.99 Recons. 0.00822 +2024-03-16 22:54:09,963 INFO Train. Iter 16800 : Commit. 0.99294 PPL. 436.14 Recons. 0.00872 +2024-03-16 22:54:31,614 INFO Train. Iter 17000 : Commit. 1.01623 PPL. 434.54 Recons. 0.01005 +2024-03-16 22:54:52,938 INFO Train. Iter 17200 : Commit. 1.09692 PPL. 431.88 Recons. 0.01231 +2024-03-16 22:55:14,164 INFO Train. Iter 17400 : Commit. 1.09255 PPL. 432.81 Recons. 0.01166 +2024-03-16 22:55:35,012 INFO Train. Iter 17600 : Commit. 1.05423 PPL. 433.87 Recons. 0.01035 +2024-03-16 22:55:56,201 INFO Train. Iter 17800 : Commit. 1.02642 PPL. 435.35 Recons. 0.00926 +2024-03-16 22:56:17,668 INFO Train. Iter 18000 : Commit. 0.95302 PPL. 436.64 Recons. 0.00807 +2024-03-16 22:56:39,125 INFO Train. Iter 18200 : Commit. 0.95301 PPL. 437.04 Recons. 0.00804 +2024-03-16 22:56:59,310 INFO Train. Iter 18400 : Commit. 0.98989 PPL. 433.24 Recons. 0.01019 +2024-03-16 22:57:20,983 INFO Train. Iter 18600 : Commit. 1.01649 PPL. 433.77 Recons. 0.01040 +2024-03-16 22:57:41,934 INFO Train. Iter 18800 : Commit. 0.98882 PPL. 434.58 Recons. 0.00974 +2024-03-16 22:58:03,021 INFO Train. Iter 19000 : Commit. 1.00174 PPL. 436.31 Recons. 0.00857 +2024-03-16 22:58:24,426 INFO Train. Iter 19200 : Commit. 1.01789 PPL. 431.33 Recons. 0.01161 +2024-03-16 22:58:46,981 INFO Train. Iter 19400 : Commit. 0.98802 PPL. 435.45 Recons. 0.00896 +2024-03-16 22:59:09,271 INFO Train. Iter 19600 : Commit. 0.95419 PPL. 435.10 Recons. 0.00910 +2024-03-16 22:59:31,013 INFO Train. Iter 19800 : Commit. 0.95333 PPL. 435.11 Recons. 0.00897 +2024-03-16 22:59:53,378 INFO Train. Iter 20000 : Commit. 0.90175 PPL. 437.15 Recons. 0.00754 +2024-03-16 23:00:14,415 INFO Train. Iter 20200 : Commit. 0.92287 PPL. 436.35 Recons. 0.00828 +2024-03-16 23:00:36,042 INFO Train. Iter 20400 : Commit. 0.96113 PPL. 433.81 Recons. 0.00976 +2024-03-16 23:00:57,594 INFO Train. Iter 20600 : Commit. 0.91662 PPL. 437.15 Recons. 0.00781 +2024-03-16 23:01:19,075 INFO Train. Iter 20800 : Commit. 0.96006 PPL. 434.42 Recons. 0.00935 +2024-03-16 23:01:40,561 INFO Train. Iter 21000 : Commit. 0.92524 PPL. 436.85 Recons. 0.00762 +2024-03-16 23:02:02,011 INFO Train. Iter 21200 : Commit. 0.95141 PPL. 433.03 Recons. 0.00962 +2024-03-16 23:02:23,452 INFO Train. Iter 21400 : Commit. 0.89219 PPL. 437.49 Recons. 0.00736 +2024-03-16 23:02:44,560 INFO Train. Iter 21600 : Commit. 0.88100 PPL. 437.43 Recons. 0.00728 +2024-03-16 23:03:06,077 INFO Train. Iter 21800 : Commit. 0.96737 PPL. 431.68 Recons. 0.01104 +2024-03-16 23:03:28,387 INFO Train. Iter 22000 : Commit. 0.96822 PPL. 434.69 Recons. 0.00951 +2024-03-16 23:03:49,233 INFO Train. Iter 22200 : Commit. 0.90122 PPL. 436.44 Recons. 0.00780 +2024-03-16 23:04:11,079 INFO Train. Iter 22400 : Commit. 0.88854 PPL. 436.22 Recons. 0.00794 +2024-03-16 23:04:32,236 INFO Train. Iter 22600 : Commit. 0.91686 PPL. 434.15 Recons. 0.00902 +2024-03-16 23:04:53,456 INFO Train. Iter 22800 : Commit. 0.88980 PPL. 435.43 Recons. 0.00821 +2024-03-16 23:05:15,046 INFO Train. Iter 23000 : Commit. 0.88751 PPL. 435.00 Recons. 0.00859 +2024-03-16 23:05:36,562 INFO Train. Iter 23200 : Commit. 0.92019 PPL. 433.96 Recons. 0.00967 +2024-03-16 23:05:58,811 INFO Train. Iter 23400 : Commit. 0.91738 PPL. 433.33 Recons. 0.00921 +2024-03-16 23:06:20,576 INFO Train. Iter 23600 : Commit. 0.83826 PPL. 438.07 Recons. 0.00645 +2024-03-16 23:06:42,089 INFO Train. Iter 23800 : Commit. 0.85509 PPL. 437.26 Recons. 0.00732 +2024-03-16 23:07:03,776 INFO Train. Iter 24000 : Commit. 0.93290 PPL. 431.66 Recons. 0.01074 +2024-03-16 23:07:24,110 INFO Train. Iter 24200 : Commit. 0.88749 PPL. 435.67 Recons. 0.00780 +2024-03-16 23:07:45,651 INFO Train. Iter 24400 : Commit. 0.84281 PPL. 436.73 Recons. 0.00730 +2024-03-16 23:08:07,329 INFO Train. Iter 24600 : Commit. 0.88125 PPL. 432.70 Recons. 0.00933 +2024-03-16 23:08:29,117 INFO Train. Iter 24800 : Commit. 0.87171 PPL. 437.38 Recons. 0.00745 +2024-03-16 23:08:51,058 INFO Train. Iter 25000 : Commit. 0.83881 PPL. 437.98 Recons. 0.00665 +2024-03-16 23:09:12,551 INFO Train. Iter 25200 : Commit. 0.86215 PPL. 434.78 Recons. 0.00884 +2024-03-16 23:09:33,642 INFO Train. Iter 25400 : Commit. 0.90211 PPL. 433.97 Recons. 0.00890 +2024-03-16 23:09:55,233 INFO Train. Iter 25600 : Commit. 0.84504 PPL. 437.15 Recons. 0.00700 +2024-03-16 23:10:16,379 INFO Train. Iter 25800 : Commit. 0.89231 PPL. 433.90 Recons. 0.00888 +2024-03-16 23:10:36,521 INFO Train. Iter 26000 : Commit. 0.85575 PPL. 436.91 Recons. 0.00712 +2024-03-16 23:10:58,553 INFO Train. Iter 26200 : Commit. 0.87460 PPL. 433.56 Recons. 0.00905 +2024-03-16 23:11:20,019 INFO Train. Iter 26400 : Commit. 0.79456 PPL. 438.39 Recons. 0.00595 +2024-03-16 23:11:41,068 INFO Train. Iter 26600 : Commit. 0.87051 PPL. 432.83 Recons. 0.00928 +2024-03-16 23:12:02,615 INFO Train. Iter 26800 : Commit. 0.85722 PPL. 435.39 Recons. 0.00785 +2024-03-16 23:12:23,736 INFO Train. Iter 27000 : Commit. 0.81573 PPL. 437.79 Recons. 0.00652 +2024-03-16 23:12:45,660 INFO Train. Iter 27200 : Commit. 0.83476 PPL. 434.16 Recons. 0.00838 +2024-03-16 23:13:06,865 INFO Train. Iter 27400 : Commit. 0.85203 PPL. 435.08 Recons. 0.00840 +2024-03-16 23:13:28,151 INFO Train. Iter 27600 : Commit. 0.84608 PPL. 435.53 Recons. 0.00795 +2024-03-16 23:13:49,319 INFO Train. Iter 27800 : Commit. 0.85161 PPL. 434.94 Recons. 0.00782 +2024-03-16 23:14:09,565 INFO Train. Iter 28000 : Commit. 0.81387 PPL. 437.54 Recons. 0.00627 +2024-03-16 23:14:31,083 INFO Train. Iter 28200 : Commit. 0.80786 PPL. 434.88 Recons. 0.00777 +2024-03-16 23:14:52,888 INFO Train. Iter 28400 : Commit. 0.84964 PPL. 434.85 Recons. 0.00796 +2024-03-16 23:15:14,272 INFO Train. Iter 28600 : Commit. 0.86319 PPL. 432.39 Recons. 0.00945 +2024-03-16 23:15:35,804 INFO Train. Iter 28800 : Commit. 0.81986 PPL. 435.50 Recons. 0.00697 +2024-03-16 23:15:57,738 INFO Train. Iter 29000 : Commit. 0.84111 PPL. 433.77 Recons. 0.00843 +2024-03-16 23:16:18,999 INFO Train. Iter 29200 : Commit. 0.82927 PPL. 434.45 Recons. 0.00775 +2024-03-16 23:16:40,607 INFO Train. Iter 29400 : Commit. 0.79351 PPL. 436.30 Recons. 0.00684 +2024-03-16 23:17:02,373 INFO Train. Iter 29600 : Commit. 0.80566 PPL. 436.13 Recons. 0.00713 +2024-03-16 23:17:24,367 INFO Train. Iter 29800 : Commit. 0.79077 PPL. 436.72 Recons. 0.00635 +2024-03-16 23:17:44,656 INFO Train. Iter 30000 : Commit. 0.82890 PPL. 432.89 Recons. 0.00846 +2024-03-16 23:18:06,635 INFO Train. Iter 30200 : Commit. 0.81430 PPL. 434.42 Recons. 0.00793 +2024-03-16 23:18:28,564 INFO Train. Iter 30400 : Commit. 0.76847 PPL. 436.07 Recons. 0.00653 +2024-03-16 23:18:50,274 INFO Train. Iter 30600 : Commit. 0.77762 PPL. 433.84 Recons. 0.00733 +2024-03-16 23:19:11,930 INFO Train. Iter 30800 : Commit. 0.78808 PPL. 434.78 Recons. 0.00747 +2024-03-16 23:19:33,253 INFO Train. Iter 31000 : Commit. 0.79315 PPL. 434.37 Recons. 0.00768 +2024-03-16 23:19:54,636 INFO Train. Iter 31200 : Commit. 0.80853 PPL. 435.49 Recons. 0.00700 +2024-03-16 23:20:16,467 INFO Train. Iter 31400 : Commit. 0.74896 PPL. 436.25 Recons. 0.00634 +2024-03-16 23:20:38,071 INFO Train. Iter 31600 : Commit. 0.82623 PPL. 431.06 Recons. 0.00943 +2024-03-16 23:20:58,464 INFO Train. Iter 31800 : Commit. 0.82033 PPL. 434.38 Recons. 0.00795 +2024-03-16 23:21:20,057 INFO Train. Iter 32000 : Commit. 0.86281 PPL. 430.95 Recons. 0.00903 +2024-03-16 23:21:41,727 INFO Train. Iter 32200 : Commit. 0.77525 PPL. 436.10 Recons. 0.00641 +2024-03-16 23:22:02,940 INFO Train. Iter 32400 : Commit. 0.74326 PPL. 437.15 Recons. 0.00589 +2024-03-16 23:22:24,856 INFO Train. Iter 32600 : Commit. 0.72035 PPL. 438.09 Recons. 0.00554 +2024-03-16 23:22:45,823 INFO Train. Iter 32800 : Commit. 0.77890 PPL. 434.38 Recons. 0.00759 +2024-03-16 23:23:08,240 INFO Train. Iter 33000 : Commit. 0.79739 PPL. 434.49 Recons. 0.00743 +2024-03-16 23:23:30,127 INFO Train. Iter 33200 : Commit. 0.75960 PPL. 435.40 Recons. 0.00701 +2024-03-16 23:23:52,071 INFO Train. Iter 33400 : Commit. 0.75659 PPL. 435.71 Recons. 0.00671 +2024-03-16 23:24:13,854 INFO Train. Iter 33600 : Commit. 0.78478 PPL. 433.55 Recons. 0.00776 +2024-03-16 23:24:34,564 INFO Train. Iter 33800 : Commit. 0.73002 PPL. 436.12 Recons. 0.00617 +2024-03-16 23:24:56,392 INFO Train. Iter 34000 : Commit. 0.75204 PPL. 436.02 Recons. 0.00660 +2024-03-16 23:25:17,879 INFO Train. Iter 34200 : Commit. 0.74554 PPL. 434.88 Recons. 0.00716 +2024-03-16 23:25:39,965 INFO Train. Iter 34400 : Commit. 0.76223 PPL. 436.10 Recons. 0.00642 +2024-03-16 23:26:01,875 INFO Train. Iter 34600 : Commit. 0.74355 PPL. 435.12 Recons. 0.00667 +2024-03-16 23:26:22,961 INFO Train. Iter 34800 : Commit. 0.76872 PPL. 433.59 Recons. 0.00799 +2024-03-16 23:26:44,210 INFO Train. Iter 35000 : Commit. 0.79195 PPL. 432.46 Recons. 0.00817 +2024-03-16 23:27:05,597 INFO Train. Iter 35200 : Commit. 0.75727 PPL. 436.39 Recons. 0.00604 +2024-03-16 23:27:26,865 INFO Train. Iter 35400 : Commit. 0.71963 PPL. 435.96 Recons. 0.00597 +2024-03-16 23:27:48,208 INFO Train. Iter 35600 : Commit. 0.76364 PPL. 434.02 Recons. 0.00714 +2024-03-16 23:28:08,565 INFO Train. Iter 35800 : Commit. 0.77268 PPL. 431.64 Recons. 0.00827 +2024-03-16 23:28:29,843 INFO Train. Iter 36000 : Commit. 0.71498 PPL. 436.79 Recons. 0.00533 +2024-03-16 23:28:51,181 INFO Train. Iter 36200 : Commit. 0.77800 PPL. 433.94 Recons. 0.00753 +2024-03-16 23:29:12,554 INFO Train. Iter 36400 : Commit. 0.73290 PPL. 434.96 Recons. 0.00638 +2024-03-16 23:29:33,997 INFO Train. Iter 36600 : Commit. 0.72423 PPL. 434.05 Recons. 0.00693 +2024-03-16 23:29:55,670 INFO Train. Iter 36800 : Commit. 0.70341 PPL. 436.22 Recons. 0.00559 +2024-03-16 23:30:17,475 INFO Train. Iter 37000 : Commit. 0.74045 PPL. 432.91 Recons. 0.00715 +2024-03-16 23:30:39,008 INFO Train. Iter 37200 : Commit. 0.74391 PPL. 433.91 Recons. 0.00698 +2024-03-16 23:31:00,346 INFO Train. Iter 37400 : Commit. 0.71834 PPL. 434.80 Recons. 0.00614 +2024-03-16 23:31:20,638 INFO Train. Iter 37600 : Commit. 0.70064 PPL. 435.05 Recons. 0.00617 +2024-03-16 23:31:41,781 INFO Train. Iter 37800 : Commit. 0.74812 PPL. 435.32 Recons. 0.00642 +2024-03-16 23:32:03,450 INFO Train. Iter 38000 : Commit. 0.71375 PPL. 434.51 Recons. 0.00649 +2024-03-16 23:32:24,614 INFO Train. Iter 38200 : Commit. 0.72275 PPL. 434.54 Recons. 0.00673 +2024-03-16 23:32:46,029 INFO Train. Iter 38400 : Commit. 0.72962 PPL. 434.95 Recons. 0.00632 +2024-03-16 23:33:07,518 INFO Train. Iter 38600 : Commit. 0.72458 PPL. 434.44 Recons. 0.00644 +2024-03-16 23:33:29,611 INFO Train. Iter 38800 : Commit. 0.72395 PPL. 434.55 Recons. 0.00631 +2024-03-16 23:33:52,145 INFO Train. Iter 39000 : Commit. 0.71838 PPL. 435.11 Recons. 0.00592 +2024-03-16 23:34:14,179 INFO Train. Iter 39200 : Commit. 0.74402 PPL. 432.91 Recons. 0.00736 +2024-03-16 23:34:36,643 INFO Train. Iter 39400 : Commit. 0.70965 PPL. 436.31 Recons. 0.00534 +2024-03-16 23:34:57,654 INFO Train. Iter 39600 : Commit. 0.72776 PPL. 433.04 Recons. 0.00706 +2024-03-16 23:35:19,716 INFO Train. Iter 39800 : Commit. 0.71048 PPL. 434.27 Recons. 0.00647 +2024-03-16 23:35:41,820 INFO Train. Iter 40000 : Commit. 0.68031 PPL. 435.25 Recons. 0.00569 +2024-03-16 23:36:03,855 INFO Train. Iter 40200 : Commit. 0.66182 PPL. 436.19 Recons. 0.00526 +2024-03-16 23:36:25,631 INFO Train. Iter 40400 : Commit. 0.72238 PPL. 431.90 Recons. 0.00732 +2024-03-16 23:36:47,363 INFO Train. Iter 40600 : Commit. 0.75067 PPL. 433.44 Recons. 0.00688 +2024-03-16 23:37:09,626 INFO Train. Iter 40800 : Commit. 0.67714 PPL. 434.30 Recons. 0.00582 +2024-03-16 23:37:32,197 INFO Train. Iter 41000 : Commit. 0.69529 PPL. 433.97 Recons. 0.00597 +2024-03-16 23:37:54,032 INFO Train. Iter 41200 : Commit. 0.68859 PPL. 434.67 Recons. 0.00583 +2024-03-16 23:38:16,181 INFO Train. Iter 41400 : Commit. 0.70832 PPL. 434.09 Recons. 0.00623 +2024-03-16 23:38:36,853 INFO Train. Iter 41600 : Commit. 0.68331 PPL. 434.94 Recons. 0.00562 +2024-03-16 23:38:58,591 INFO Train. Iter 41800 : Commit. 0.70820 PPL. 433.86 Recons. 0.00624 +2024-03-16 23:39:20,572 INFO Train. Iter 42000 : Commit. 0.72038 PPL. 433.72 Recons. 0.00650 +2024-03-16 23:39:42,997 INFO Train. Iter 42200 : Commit. 0.67748 PPL. 435.16 Recons. 0.00547 +2024-03-16 23:40:05,181 INFO Train. Iter 42400 : Commit. 0.70547 PPL. 433.44 Recons. 0.00661 +2024-03-16 23:40:26,723 INFO Train. Iter 42600 : Commit. 0.68080 PPL. 434.25 Recons. 0.00602 +2024-03-16 23:40:48,354 INFO Train. Iter 42800 : Commit. 0.69522 PPL. 433.53 Recons. 0.00620 +2024-03-16 23:41:10,341 INFO Train. Iter 43000 : Commit. 0.68438 PPL. 435.01 Recons. 0.00586 +2024-03-16 23:41:32,542 INFO Train. Iter 43200 : Commit. 0.69102 PPL. 435.02 Recons. 0.00577 +2024-03-16 23:41:52,611 INFO Train. Iter 43400 : Commit. 0.68909 PPL. 434.05 Recons. 0.00583 +2024-03-16 23:42:13,849 INFO Train. Iter 43600 : Commit. 0.69667 PPL. 432.19 Recons. 0.00658 +2024-03-16 23:42:35,068 INFO Train. Iter 43800 : Commit. 0.69341 PPL. 434.46 Recons. 0.00618 +2024-03-16 23:42:56,108 INFO Train. Iter 44000 : Commit. 0.69084 PPL. 435.54 Recons. 0.00552 +2024-03-16 23:43:16,950 INFO Train. Iter 44200 : Commit. 0.70685 PPL. 432.86 Recons. 0.00675 +2024-03-16 23:43:38,443 INFO Train. Iter 44400 : Commit. 0.67902 PPL. 432.87 Recons. 0.00646 +2024-03-16 23:43:59,977 INFO Train. Iter 44600 : Commit. 0.68131 PPL. 434.32 Recons. 0.00570 +2024-03-16 23:44:21,680 INFO Train. Iter 44800 : Commit. 0.69932 PPL. 433.24 Recons. 0.00635 +2024-03-16 23:44:42,998 INFO Train. Iter 45000 : Commit. 0.71360 PPL. 431.86 Recons. 0.00685 +2024-03-16 23:45:04,153 INFO Train. Iter 45200 : Commit. 0.70624 PPL. 432.38 Recons. 0.00638 +2024-03-16 23:45:24,711 INFO Train. Iter 45400 : Commit. 0.68657 PPL. 434.91 Recons. 0.00571 +2024-03-16 23:45:46,260 INFO Train. Iter 45600 : Commit. 0.66523 PPL. 434.50 Recons. 0.00552 +2024-03-16 23:46:07,921 INFO Train. Iter 45800 : Commit. 0.63037 PPL. 436.58 Recons. 0.00466 +2024-03-16 23:46:29,197 INFO Train. Iter 46000 : Commit. 0.70281 PPL. 430.59 Recons. 0.00753 +2024-03-16 23:46:50,285 INFO Train. Iter 46200 : Commit. 0.68586 PPL. 433.47 Recons. 0.00591 +2024-03-16 23:47:11,405 INFO Train. Iter 46400 : Commit. 0.66080 PPL. 434.20 Recons. 0.00557 +2024-03-16 23:47:32,512 INFO Train. Iter 46600 : Commit. 0.67238 PPL. 432.29 Recons. 0.00614 +2024-03-16 23:47:53,931 INFO Train. Iter 46800 : Commit. 0.68712 PPL. 432.79 Recons. 0.00671 +2024-03-16 23:48:15,451 INFO Train. Iter 47000 : Commit. 0.69822 PPL. 431.86 Recons. 0.00644 +2024-03-16 23:48:37,038 INFO Train. Iter 47200 : Commit. 0.68480 PPL. 433.66 Recons. 0.00583 +2024-03-16 23:48:58,103 INFO Train. Iter 47400 : Commit. 0.68081 PPL. 431.84 Recons. 0.00658 +2024-03-16 23:49:20,009 INFO Train. Iter 47600 : Commit. 0.67403 PPL. 433.24 Recons. 0.00567 +2024-03-16 23:49:41,898 INFO Train. Iter 47800 : Commit. 0.67943 PPL. 430.30 Recons. 0.00694 +2024-03-16 23:50:04,332 INFO Train. Iter 48000 : Commit. 0.67748 PPL. 434.04 Recons. 0.00575 +2024-03-16 23:50:26,590 INFO Train. Iter 48200 : Commit. 0.65768 PPL. 433.57 Recons. 0.00556 +2024-03-16 23:50:48,706 INFO Train. Iter 48400 : Commit. 0.65198 PPL. 435.60 Recons. 0.00481 +2024-03-16 23:51:10,338 INFO Train. Iter 48600 : Commit. 0.68002 PPL. 430.82 Recons. 0.00653 +2024-03-16 23:51:32,360 INFO Train. Iter 48800 : Commit. 0.65299 PPL. 434.50 Recons. 0.00516 +2024-03-16 23:51:54,014 INFO Train. Iter 49000 : Commit. 0.69573 PPL. 432.07 Recons. 0.00654 +2024-03-16 23:52:14,710 INFO Train. Iter 49200 : Commit. 0.69942 PPL. 432.57 Recons. 0.00634 +2024-03-16 23:52:37,460 INFO Train. Iter 49400 : Commit. 0.66585 PPL. 433.91 Recons. 0.00541 +2024-03-16 23:52:59,898 INFO Train. Iter 49600 : Commit. 0.66698 PPL. 433.13 Recons. 0.00558 +2024-03-16 23:53:21,676 INFO Train. Iter 49800 : Commit. 0.65283 PPL. 433.98 Recons. 0.00540 +2024-03-16 23:53:43,484 INFO Train. Iter 50000 : Commit. 0.67524 PPL. 432.41 Recons. 0.00608 +2024-03-16 23:54:05,332 INFO Train. Iter 50200 : Commit. 0.65851 PPL. 433.58 Recons. 0.00529 +2024-03-16 23:54:27,355 INFO Train. Iter 50400 : Commit. 0.65276 PPL. 432.24 Recons. 0.00605 +2024-03-16 23:54:49,385 INFO Train. Iter 50600 : Commit. 0.64101 PPL. 434.13 Recons. 0.00521 +2024-03-16 23:55:11,455 INFO Train. Iter 50800 : Commit. 0.65535 PPL. 431.85 Recons. 0.00573 +2024-03-16 23:55:33,971 INFO Train. Iter 51000 : Commit. 0.67079 PPL. 431.07 Recons. 0.00661 +2024-03-16 23:55:55,057 INFO Train. Iter 51200 : Commit. 0.67560 PPL. 432.93 Recons. 0.00555 +2024-03-16 23:56:16,738 INFO Train. Iter 51400 : Commit. 0.66588 PPL. 431.64 Recons. 0.00597 +2024-03-16 23:56:37,884 INFO Train. Iter 51600 : Commit. 0.64166 PPL. 433.20 Recons. 0.00533 +2024-03-16 23:56:59,871 INFO Train. Iter 51800 : Commit. 0.64850 PPL. 433.37 Recons. 0.00552 +2024-03-16 23:57:21,697 INFO Train. Iter 52000 : Commit. 0.62595 PPL. 433.59 Recons. 0.00518 +2024-03-16 23:57:43,577 INFO Train. Iter 52200 : Commit. 0.63264 PPL. 435.29 Recons. 0.00471 +2024-03-16 23:58:05,399 INFO Train. Iter 52400 : Commit. 0.63052 PPL. 432.90 Recons. 0.00560 +2024-03-16 23:58:26,853 INFO Train. Iter 52600 : Commit. 0.62073 PPL. 434.48 Recons. 0.00475 +2024-03-16 23:58:48,810 INFO Train. Iter 52800 : Commit. 0.64604 PPL. 432.69 Recons. 0.00585 +2024-03-16 23:59:09,911 INFO Train. Iter 53000 : Commit. 0.64160 PPL. 434.48 Recons. 0.00506 +2024-03-16 23:59:30,216 INFO Train. Iter 53200 : Commit. 0.62935 PPL. 433.70 Recons. 0.00524 +2024-03-16 23:59:51,728 INFO Train. Iter 53400 : Commit. 0.60758 PPL. 434.30 Recons. 0.00467 +2024-03-17 00:00:13,104 INFO Train. Iter 53600 : Commit. 0.62424 PPL. 433.81 Recons. 0.00513 +2024-03-17 00:00:34,469 INFO Train. Iter 53800 : Commit. 0.64355 PPL. 432.82 Recons. 0.00557 +2024-03-17 00:00:55,929 INFO Train. Iter 54000 : Commit. 0.66110 PPL. 430.86 Recons. 0.00634 +2024-03-17 00:01:16,990 INFO Train. Iter 54200 : Commit. 0.63148 PPL. 432.73 Recons. 0.00533 +2024-03-17 00:01:38,525 INFO Train. Iter 54400 : Commit. 0.61005 PPL. 432.47 Recons. 0.00512 +2024-03-17 00:01:59,979 INFO Train. Iter 54600 : Commit. 0.63398 PPL. 432.95 Recons. 0.00544 +2024-03-17 00:02:21,134 INFO Train. Iter 54800 : Commit. 0.61643 PPL. 434.85 Recons. 0.00456 +2024-03-17 00:02:41,755 INFO Train. Iter 55000 : Commit. 0.64676 PPL. 431.04 Recons. 0.00617 +2024-03-17 00:03:02,939 INFO Train. Iter 55200 : Commit. 0.61971 PPL. 434.59 Recons. 0.00468 +2024-03-17 00:03:24,023 INFO Train. Iter 55400 : Commit. 0.62959 PPL. 432.99 Recons. 0.00526 +2024-03-17 00:03:45,074 INFO Train. Iter 55600 : Commit. 0.60890 PPL. 434.67 Recons. 0.00458 +2024-03-17 00:04:05,847 INFO Train. Iter 55800 : Commit. 0.65190 PPL. 431.34 Recons. 0.00632 +2024-03-17 00:04:26,880 INFO Train. Iter 56000 : Commit. 0.63087 PPL. 432.89 Recons. 0.00535 +2024-03-17 00:04:47,835 INFO Train. Iter 56200 : Commit. 0.62578 PPL. 433.49 Recons. 0.00507 +2024-03-17 00:05:08,961 INFO Train. Iter 56400 : Commit. 0.63349 PPL. 432.14 Recons. 0.00549 +2024-03-17 00:05:30,221 INFO Train. Iter 56600 : Commit. 0.62076 PPL. 433.29 Recons. 0.00493 +2024-03-17 00:05:51,684 INFO Train. Iter 56800 : Commit. 0.62635 PPL. 431.46 Recons. 0.00569 +2024-03-17 00:06:12,201 INFO Train. Iter 57000 : Commit. 0.64389 PPL. 433.69 Recons. 0.00496 +2024-03-17 00:06:34,184 INFO Train. Iter 57200 : Commit. 0.59996 PPL. 434.25 Recons. 0.00473 +2024-03-17 00:06:55,331 INFO Train. Iter 57400 : Commit. 0.59782 PPL. 434.65 Recons. 0.00457 +2024-03-17 00:07:17,292 INFO Train. Iter 57600 : Commit. 0.62977 PPL. 431.00 Recons. 0.00579 +2024-03-17 00:07:38,203 INFO Train. Iter 57800 : Commit. 0.64510 PPL. 431.53 Recons. 0.00588 +2024-03-17 00:07:59,801 INFO Train. Iter 58000 : Commit. 0.58645 PPL. 434.87 Recons. 0.00426 +2024-03-17 00:08:21,281 INFO Train. Iter 58200 : Commit. 0.63249 PPL. 430.97 Recons. 0.00589 +2024-03-17 00:08:42,324 INFO Train. Iter 58400 : Commit. 0.63426 PPL. 432.18 Recons. 0.00549 +2024-03-17 00:09:03,354 INFO Train. Iter 58600 : Commit. 0.62406 PPL. 433.70 Recons. 0.00505 +2024-03-17 00:09:24,210 INFO Train. Iter 58800 : Commit. 0.64125 PPL. 432.07 Recons. 0.00570 +2024-03-17 00:09:44,453 INFO Train. Iter 59000 : Commit. 0.63112 PPL. 431.76 Recons. 0.00554 +2024-03-17 00:10:06,314 INFO Train. Iter 59200 : Commit. 0.58844 PPL. 434.34 Recons. 0.00452 +2024-03-17 00:10:27,290 INFO Train. Iter 59400 : Commit. 0.64499 PPL. 429.52 Recons. 0.00626 +2024-03-17 00:10:48,797 INFO Train. Iter 59600 : Commit. 0.64800 PPL. 431.60 Recons. 0.00564 +2024-03-17 00:11:10,233 INFO Train. Iter 59800 : Commit. 0.61667 PPL. 431.22 Recons. 0.00521 +2024-03-17 00:11:31,735 INFO Train. Iter 60000 : Commit. 0.61143 PPL. 434.21 Recons. 0.00445 +2024-03-17 00:11:53,103 INFO Train. Iter 60200 : Commit. 0.60542 PPL. 433.66 Recons. 0.00499 +2024-03-17 00:12:14,317 INFO Train. Iter 60400 : Commit. 0.58046 PPL. 433.47 Recons. 0.00445 +2024-03-17 00:12:35,392 INFO Train. Iter 60600 : Commit. 0.62655 PPL. 430.66 Recons. 0.00580 +2024-03-17 00:12:55,953 INFO Train. Iter 60800 : Commit. 0.61194 PPL. 432.36 Recons. 0.00503 +2024-03-17 00:13:17,518 INFO Train. Iter 61000 : Commit. 0.56753 PPL. 434.88 Recons. 0.00393 +2024-03-17 00:13:39,054 INFO Train. Iter 61200 : Commit. 0.62476 PPL. 430.83 Recons. 0.00604 +2024-03-17 00:14:00,363 INFO Train. Iter 61400 : Commit. 0.61588 PPL. 430.63 Recons. 0.00566 +2024-03-17 00:14:22,571 INFO Train. Iter 61600 : Commit. 0.62238 PPL. 431.60 Recons. 0.00544 +2024-03-17 00:14:43,915 INFO Train. Iter 61800 : Commit. 0.65967 PPL. 429.11 Recons. 0.00658 +2024-03-17 00:15:04,985 INFO Train. Iter 62000 : Commit. 0.60867 PPL. 433.54 Recons. 0.00452 +2024-03-17 00:15:26,225 INFO Train. Iter 62200 : Commit. 0.63124 PPL. 430.47 Recons. 0.00606 +2024-03-17 00:15:47,441 INFO Train. Iter 62400 : Commit. 0.60504 PPL. 432.74 Recons. 0.00507 +2024-03-17 00:16:08,812 INFO Train. Iter 62600 : Commit. 0.61041 PPL. 433.01 Recons. 0.00500 +2024-03-17 00:16:29,151 INFO Train. Iter 62800 : Commit. 0.63430 PPL. 428.87 Recons. 0.00632 +2024-03-17 00:16:50,481 INFO Train. Iter 63000 : Commit. 0.60050 PPL. 434.36 Recons. 0.00427 +2024-03-17 00:17:11,587 INFO Train. Iter 63200 : Commit. 0.61616 PPL. 431.20 Recons. 0.00527 +2024-03-17 00:17:32,730 INFO Train. Iter 63400 : Commit. 0.61992 PPL. 430.80 Recons. 0.00538 +2024-03-17 00:17:53,977 INFO Train. Iter 63600 : Commit. 0.59097 PPL. 434.51 Recons. 0.00413 +2024-03-17 00:18:15,070 INFO Train. Iter 63800 : Commit. 0.61347 PPL. 432.12 Recons. 0.00542 +2024-03-17 00:18:36,244 INFO Train. Iter 64000 : Commit. 0.57218 PPL. 434.21 Recons. 0.00406 +2024-03-17 00:18:57,472 INFO Train. Iter 64200 : Commit. 0.59579 PPL. 430.67 Recons. 0.00519 +2024-03-17 00:19:19,123 INFO Train. Iter 64400 : Commit. 0.60940 PPL. 432.25 Recons. 0.00503 +2024-03-17 00:19:41,211 INFO Train. Iter 64600 : Commit. 0.62382 PPL. 431.30 Recons. 0.00542 +2024-03-17 00:20:02,361 INFO Train. Iter 64800 : Commit. 0.59874 PPL. 432.35 Recons. 0.00486 +2024-03-17 00:20:23,782 INFO Train. Iter 65000 : Commit. 0.63570 PPL. 430.07 Recons. 0.00584 +2024-03-17 00:20:45,480 INFO Train. Iter 65200 : Commit. 0.62607 PPL. 429.39 Recons. 0.00612 +2024-03-17 00:21:06,620 INFO Train. Iter 65400 : Commit. 0.61999 PPL. 431.56 Recons. 0.00528 +2024-03-17 00:21:27,741 INFO Train. Iter 65600 : Commit. 0.60626 PPL. 429.65 Recons. 0.00529 +2024-03-17 00:21:48,988 INFO Train. Iter 65800 : Commit. 0.57521 PPL. 432.93 Recons. 0.00453 +2024-03-17 00:22:10,053 INFO Train. Iter 66000 : Commit. 0.60980 PPL. 430.74 Recons. 0.00563 +2024-03-17 00:22:30,692 INFO Train. Iter 66200 : Commit. 0.61733 PPL. 429.69 Recons. 0.00569 +2024-03-17 00:22:52,205 INFO Train. Iter 66400 : Commit. 0.61624 PPL. 431.49 Recons. 0.00528 +2024-03-17 00:23:12,849 INFO Train. Iter 66600 : Commit. 0.60781 PPL. 431.93 Recons. 0.00516 +2024-03-17 00:23:34,020 INFO Train. Iter 66800 : Commit. 0.60667 PPL. 430.90 Recons. 0.00532 +2024-03-17 00:23:55,023 INFO Train. Iter 67000 : Commit. 0.62530 PPL. 431.10 Recons. 0.00567 +2024-03-17 00:24:16,344 INFO Train. Iter 67200 : Commit. 0.58617 PPL. 433.20 Recons. 0.00462 +2024-03-17 00:24:37,830 INFO Train. Iter 67400 : Commit. 0.59641 PPL. 430.73 Recons. 0.00535 +2024-03-17 00:24:59,205 INFO Train. Iter 67600 : Commit. 0.59759 PPL. 432.45 Recons. 0.00483 +2024-03-17 00:25:20,474 INFO Train. Iter 67800 : Commit. 0.57745 PPL. 433.24 Recons. 0.00434 +2024-03-17 00:25:42,285 INFO Train. Iter 68000 : Commit. 0.60097 PPL. 429.66 Recons. 0.00566 +2024-03-17 00:26:04,129 INFO Train. Iter 68200 : Commit. 0.60210 PPL. 430.83 Recons. 0.00526 +2024-03-17 00:26:25,822 INFO Train. Iter 68400 : Commit. 0.63049 PPL. 428.38 Recons. 0.00626 +2024-03-17 00:26:46,003 INFO Train. Iter 68600 : Commit. 0.61720 PPL. 430.81 Recons. 0.00539 +2024-03-17 00:27:07,512 INFO Train. Iter 68800 : Commit. 0.58371 PPL. 433.58 Recons. 0.00407 +2024-03-17 00:27:29,150 INFO Train. Iter 69000 : Commit. 0.60216 PPL. 429.93 Recons. 0.00540 +2024-03-17 00:27:50,565 INFO Train. Iter 69200 : Commit. 0.57779 PPL. 432.70 Recons. 0.00435 +2024-03-17 00:28:12,582 INFO Train. Iter 69400 : Commit. 0.58667 PPL. 432.51 Recons. 0.00476 +2024-03-17 00:28:34,679 INFO Train. Iter 69600 : Commit. 0.61016 PPL. 430.89 Recons. 0.00521 +2024-03-17 00:28:56,318 INFO Train. Iter 69800 : Commit. 0.60057 PPL. 430.57 Recons. 0.00527 +2024-03-17 00:29:18,065 INFO Train. Iter 70000 : Commit. 0.60455 PPL. 430.27 Recons. 0.00536 +2024-03-17 00:29:40,513 INFO Train. Iter 70200 : Commit. 0.58638 PPL. 431.82 Recons. 0.00478 +2024-03-17 00:30:03,196 INFO Train. Iter 70400 : Commit. 0.59460 PPL. 430.91 Recons. 0.00498 +2024-03-17 00:30:24,060 INFO Train. Iter 70600 : Commit. 0.59008 PPL. 431.56 Recons. 0.00474 +2024-03-17 00:30:45,955 INFO Train. Iter 70800 : Commit. 0.56597 PPL. 433.67 Recons. 0.00403 +2024-03-17 00:31:07,867 INFO Train. Iter 71000 : Commit. 0.60963 PPL. 429.57 Recons. 0.00578 +2024-03-17 00:31:29,629 INFO Train. Iter 71200 : Commit. 0.58144 PPL. 432.03 Recons. 0.00452 +2024-03-17 00:31:50,751 INFO Train. Iter 71400 : Commit. 0.62008 PPL. 428.79 Recons. 0.00574 +2024-03-17 00:32:12,442 INFO Train. Iter 71600 : Commit. 0.60440 PPL. 430.85 Recons. 0.00521 +2024-03-17 00:32:33,785 INFO Train. Iter 71800 : Commit. 0.58015 PPL. 431.70 Recons. 0.00450 +2024-03-17 00:32:55,316 INFO Train. Iter 72000 : Commit. 0.58699 PPL. 430.81 Recons. 0.00517 +2024-03-17 00:33:16,335 INFO Train. Iter 72200 : Commit. 0.59095 PPL. 429.69 Recons. 0.00512 +2024-03-17 00:33:36,807 INFO Train. Iter 72400 : Commit. 0.57517 PPL. 432.44 Recons. 0.00446 +2024-03-17 00:33:58,206 INFO Train. Iter 72600 : Commit. 0.59699 PPL. 431.23 Recons. 0.00507 +2024-03-17 00:34:19,809 INFO Train. Iter 72800 : Commit. 0.55997 PPL. 433.27 Recons. 0.00392 +2024-03-17 00:34:40,604 INFO Train. Iter 73000 : Commit. 0.60317 PPL. 429.20 Recons. 0.00588 +2024-03-17 00:35:02,276 INFO Train. Iter 73200 : Commit. 0.57709 PPL. 432.83 Recons. 0.00438 +2024-03-17 00:35:23,641 INFO Train. Iter 73400 : Commit. 0.57113 PPL. 431.86 Recons. 0.00445 +2024-03-17 00:35:45,360 INFO Train. Iter 73600 : Commit. 0.57366 PPL. 431.30 Recons. 0.00448 +2024-03-17 00:36:07,069 INFO Train. Iter 73800 : Commit. 0.58111 PPL. 431.36 Recons. 0.00501 +2024-03-17 00:36:28,934 INFO Train. Iter 74000 : Commit. 0.57160 PPL. 431.71 Recons. 0.00440 +2024-03-17 00:36:50,986 INFO Train. Iter 74200 : Commit. 0.59139 PPL. 430.21 Recons. 0.00506 +2024-03-17 00:37:11,841 INFO Train. Iter 74400 : Commit. 0.57828 PPL. 430.72 Recons. 0.00473 +2024-03-17 00:37:34,223 INFO Train. Iter 74600 : Commit. 0.57457 PPL. 431.75 Recons. 0.00449 +2024-03-17 00:37:56,821 INFO Train. Iter 74800 : Commit. 0.57632 PPL. 432.09 Recons. 0.00454 +2024-03-17 00:38:19,455 INFO Train. Iter 75000 : Commit. 0.57482 PPL. 430.89 Recons. 0.00480 +2024-03-17 00:38:41,985 INFO Train. Iter 75200 : Commit. 0.60437 PPL. 428.38 Recons. 0.00592 +2024-03-17 00:39:03,766 INFO Train. Iter 75400 : Commit. 0.59667 PPL. 429.30 Recons. 0.00533 +2024-03-17 00:39:25,464 INFO Train. Iter 75600 : Commit. 0.55255 PPL. 433.97 Recons. 0.00360 +2024-03-17 00:39:46,614 INFO Train. Iter 75800 : Commit. 0.56741 PPL. 431.39 Recons. 0.00449 +2024-03-17 00:40:08,263 INFO Train. Iter 76000 : Commit. 0.56850 PPL. 429.76 Recons. 0.00508 +2024-03-17 00:40:29,958 INFO Train. Iter 76200 : Commit. 0.54906 PPL. 433.71 Recons. 0.00372 +2024-03-17 00:40:50,571 INFO Train. Iter 76400 : Commit. 0.60153 PPL. 428.16 Recons. 0.00569 +2024-03-17 00:41:11,680 INFO Train. Iter 76600 : Commit. 0.58744 PPL. 430.31 Recons. 0.00485 +2024-03-17 00:41:32,720 INFO Train. Iter 76800 : Commit. 0.56329 PPL. 432.39 Recons. 0.00407 +2024-03-17 00:41:53,875 INFO Train. Iter 77000 : Commit. 0.57148 PPL. 431.06 Recons. 0.00463 +2024-03-17 00:42:14,926 INFO Train. Iter 77200 : Commit. 0.58287 PPL. 430.15 Recons. 0.00474 +2024-03-17 00:42:35,771 INFO Train. Iter 77400 : Commit. 0.59795 PPL. 429.37 Recons. 0.00521 +2024-03-17 00:42:57,125 INFO Train. Iter 77600 : Commit. 0.57344 PPL. 431.82 Recons. 0.00452 +2024-03-17 00:43:18,528 INFO Train. Iter 77800 : Commit. 0.58253 PPL. 429.61 Recons. 0.00496 +2024-03-17 00:43:40,416 INFO Train. Iter 78000 : Commit. 0.56574 PPL. 430.64 Recons. 0.00460 +2024-03-17 00:44:00,679 INFO Train. Iter 78200 : Commit. 0.55464 PPL. 431.90 Recons. 0.00429 +2024-03-17 00:44:21,962 INFO Train. Iter 78400 : Commit. 0.55895 PPL. 431.72 Recons. 0.00424 +2024-03-17 00:44:42,569 INFO Train. Iter 78600 : Commit. 0.58773 PPL. 429.59 Recons. 0.00499 +2024-03-17 00:45:03,589 INFO Train. Iter 78800 : Commit. 0.56972 PPL. 430.58 Recons. 0.00467 +2024-03-17 00:45:24,499 INFO Train. Iter 79000 : Commit. 0.55919 PPL. 431.73 Recons. 0.00433 +2024-03-17 00:45:46,168 INFO Train. Iter 79200 : Commit. 0.55807 PPL. 431.65 Recons. 0.00430 +2024-03-17 00:46:07,715 INFO Train. Iter 79400 : Commit. 0.56532 PPL. 428.52 Recons. 0.00503 +2024-03-17 00:46:29,258 INFO Train. Iter 79600 : Commit. 0.57092 PPL. 430.65 Recons. 0.00480 +2024-03-17 00:46:50,896 INFO Train. Iter 79800 : Commit. 0.56744 PPL. 431.96 Recons. 0.00419 +2024-03-17 00:47:11,671 INFO Train. Iter 80000 : Commit. 0.57816 PPL. 429.05 Recons. 0.00539 +2024-03-17 00:47:31,777 INFO Train. Iter 80200 : Commit. 0.55988 PPL. 431.53 Recons. 0.00414 +2024-03-17 00:47:53,566 INFO Train. Iter 80400 : Commit. 0.58615 PPL. 428.65 Recons. 0.00561 +2024-03-17 00:48:14,709 INFO Train. Iter 80600 : Commit. 0.56318 PPL. 430.65 Recons. 0.00433 +2024-03-17 00:48:35,772 INFO Train. Iter 80800 : Commit. 0.56417 PPL. 429.11 Recons. 0.00460 +2024-03-17 00:48:56,467 INFO Train. Iter 81000 : Commit. 0.58623 PPL. 429.26 Recons. 0.00514 +2024-03-17 00:49:17,985 INFO Train. Iter 81200 : Commit. 0.54872 PPL. 432.20 Recons. 0.00381 +2024-03-17 00:49:39,460 INFO Train. Iter 81400 : Commit. 0.57884 PPL. 429.02 Recons. 0.00511 +2024-03-17 00:50:01,397 INFO Train. Iter 81600 : Commit. 0.54968 PPL. 432.39 Recons. 0.00390 +2024-03-17 00:50:22,304 INFO Train. Iter 81800 : Commit. 0.55753 PPL. 431.55 Recons. 0.00432 +2024-03-17 00:50:43,373 INFO Train. Iter 82000 : Commit. 0.56226 PPL. 430.12 Recons. 0.00465 +2024-03-17 00:51:03,371 INFO Train. Iter 82200 : Commit. 0.61311 PPL. 426.46 Recons. 0.00614 +2024-03-17 00:51:24,734 INFO Train. Iter 82400 : Commit. 0.57930 PPL. 431.22 Recons. 0.00443 +2024-03-17 00:51:46,064 INFO Train. Iter 82600 : Commit. 0.56111 PPL. 430.78 Recons. 0.00418 +2024-03-17 00:52:07,069 INFO Train. Iter 82800 : Commit. 0.54474 PPL. 431.89 Recons. 0.00388 +2024-03-17 00:52:28,207 INFO Train. Iter 83000 : Commit. 0.58759 PPL. 428.43 Recons. 0.00517 +2024-03-17 00:52:49,467 INFO Train. Iter 83200 : Commit. 0.55811 PPL. 430.77 Recons. 0.00432 +2024-03-17 00:53:10,991 INFO Train. Iter 83400 : Commit. 0.57439 PPL. 430.32 Recons. 0.00472 +2024-03-17 00:53:32,541 INFO Train. Iter 83600 : Commit. 0.54457 PPL. 432.11 Recons. 0.00392 +2024-03-17 00:53:53,939 INFO Train. Iter 83800 : Commit. 0.57523 PPL. 427.22 Recons. 0.00561 +2024-03-17 00:54:13,697 INFO Train. Iter 84000 : Commit. 0.58170 PPL. 428.34 Recons. 0.00497 +2024-03-17 00:54:35,170 INFO Train. Iter 84200 : Commit. 0.55467 PPL. 431.19 Recons. 0.00418 +2024-03-17 00:54:57,036 INFO Train. Iter 84400 : Commit. 0.56413 PPL. 429.81 Recons. 0.00438 +2024-03-17 00:55:18,225 INFO Train. Iter 84600 : Commit. 0.55925 PPL. 430.60 Recons. 0.00440 +2024-03-17 00:55:39,463 INFO Train. Iter 84800 : Commit. 0.53947 PPL. 431.37 Recons. 0.00410 +2024-03-17 00:56:00,665 INFO Train. Iter 85000 : Commit. 0.56615 PPL. 430.00 Recons. 0.00466 +2024-03-17 00:56:22,018 INFO Train. Iter 85200 : Commit. 0.58127 PPL. 429.57 Recons. 0.00468 +2024-03-17 00:56:43,764 INFO Train. Iter 85400 : Commit. 0.54863 PPL. 430.41 Recons. 0.00429 +2024-03-17 00:57:04,652 INFO Train. Iter 85600 : Commit. 0.56152 PPL. 428.87 Recons. 0.00488 +2024-03-17 00:57:25,669 INFO Train. Iter 85800 : Commit. 0.55432 PPL. 430.38 Recons. 0.00424 +2024-03-17 00:57:46,157 INFO Train. Iter 86000 : Commit. 0.55740 PPL. 430.91 Recons. 0.00429 +2024-03-17 00:58:07,614 INFO Train. Iter 86200 : Commit. 0.55091 PPL. 429.06 Recons. 0.00447 +2024-03-17 00:58:29,311 INFO Train. Iter 86400 : Commit. 0.56636 PPL. 429.46 Recons. 0.00454 +2024-03-17 00:58:51,749 INFO Train. Iter 86600 : Commit. 0.56286 PPL. 429.08 Recons. 0.00473 +2024-03-17 00:59:13,573 INFO Train. Iter 86800 : Commit. 0.56126 PPL. 429.91 Recons. 0.00441 +2024-03-17 00:59:35,379 INFO Train. Iter 87000 : Commit. 0.56993 PPL. 428.57 Recons. 0.00491 +2024-03-17 00:59:56,760 INFO Train. Iter 87200 : Commit. 0.53582 PPL. 431.57 Recons. 0.00373 +2024-03-17 01:00:18,233 INFO Train. Iter 87400 : Commit. 0.54509 PPL. 429.40 Recons. 0.00443 +2024-03-17 01:00:39,961 INFO Train. Iter 87600 : Commit. 0.54554 PPL. 430.40 Recons. 0.00415 +2024-03-17 01:01:01,428 INFO Train. Iter 87800 : Commit. 0.57416 PPL. 427.89 Recons. 0.00525 +2024-03-17 01:01:21,368 INFO Train. Iter 88000 : Commit. 0.55334 PPL. 429.53 Recons. 0.00429 +2024-03-17 01:01:42,332 INFO Train. Iter 88200 : Commit. 0.55654 PPL. 429.97 Recons. 0.00437 +2024-03-17 01:02:04,117 INFO Train. Iter 88400 : Commit. 0.55527 PPL. 430.38 Recons. 0.00440 +2024-03-17 01:02:25,923 INFO Train. Iter 88600 : Commit. 0.55469 PPL. 430.28 Recons. 0.00418 +2024-03-17 01:02:47,203 INFO Train. Iter 88800 : Commit. 0.58021 PPL. 425.88 Recons. 0.00530 +2024-03-17 01:03:08,942 INFO Train. Iter 89000 : Commit. 0.53614 PPL. 431.65 Recons. 0.00365 +2024-03-17 01:03:29,903 INFO Train. Iter 89200 : Commit. 0.53697 PPL. 430.86 Recons. 0.00405 +2024-03-17 01:03:51,665 INFO Train. Iter 89400 : Commit. 0.52557 PPL. 430.68 Recons. 0.00397 +2024-03-17 01:04:13,134 INFO Train. Iter 89600 : Commit. 0.57007 PPL. 426.38 Recons. 0.00510 +2024-03-17 01:04:33,487 INFO Train. Iter 89800 : Commit. 0.53671 PPL. 430.65 Recons. 0.00393 +2024-03-17 01:04:54,498 INFO Train. Iter 90000 : Commit. 0.56895 PPL. 427.66 Recons. 0.00521 +2024-03-17 01:05:15,541 INFO Train. Iter 90200 : Commit. 0.56445 PPL. 428.85 Recons. 0.00451 +2024-03-17 01:05:37,621 INFO Train. Iter 90400 : Commit. 0.53167 PPL. 430.66 Recons. 0.00361 +2024-03-17 01:05:58,712 INFO Train. Iter 90600 : Commit. 0.54357 PPL. 428.43 Recons. 0.00447 +2024-03-17 01:06:19,694 INFO Train. Iter 90800 : Commit. 0.55644 PPL. 428.61 Recons. 0.00462 +2024-03-17 01:06:41,125 INFO Train. Iter 91000 : Commit. 0.54439 PPL. 429.98 Recons. 0.00396 +2024-03-17 01:07:02,239 INFO Train. Iter 91200 : Commit. 0.54840 PPL. 428.73 Recons. 0.00444 +2024-03-17 01:07:23,523 INFO Train. Iter 91400 : Commit. 0.53009 PPL. 429.62 Recons. 0.00399 +2024-03-17 01:07:45,104 INFO Train. Iter 91600 : Commit. 0.54538 PPL. 429.74 Recons. 0.00429 +2024-03-17 01:08:06,061 INFO Train. Iter 91800 : Commit. 0.57874 PPL. 425.68 Recons. 0.00570 +2024-03-17 01:08:27,987 INFO Train. Iter 92000 : Commit. 0.53617 PPL. 429.86 Recons. 0.00379 +2024-03-17 01:08:49,357 INFO Train. Iter 92200 : Commit. 0.55323 PPL. 428.56 Recons. 0.00458 +2024-03-17 01:09:11,390 INFO Train. Iter 92400 : Commit. 0.54028 PPL. 429.09 Recons. 0.00436 +2024-03-17 01:09:32,600 INFO Train. Iter 92600 : Commit. 0.56959 PPL. 427.20 Recons. 0.00488 +2024-03-17 01:09:54,667 INFO Train. Iter 92800 : Commit. 0.52456 PPL. 430.55 Recons. 0.00360 +2024-03-17 01:10:16,613 INFO Train. Iter 93000 : Commit. 0.55451 PPL. 427.53 Recons. 0.00477 +2024-03-17 01:10:38,334 INFO Train. Iter 93200 : Commit. 0.55314 PPL. 427.13 Recons. 0.00471 +2024-03-17 01:11:00,397 INFO Train. Iter 93400 : Commit. 0.53792 PPL. 429.33 Recons. 0.00420 +2024-03-17 01:11:22,233 INFO Train. Iter 93600 : Commit. 0.53584 PPL. 430.69 Recons. 0.00374 +2024-03-17 01:11:42,505 INFO Train. Iter 93800 : Commit. 0.52410 PPL. 429.36 Recons. 0.00394 +2024-03-17 01:12:03,528 INFO Train. Iter 94000 : Commit. 0.56208 PPL. 426.92 Recons. 0.00518 +2024-03-17 01:12:24,970 INFO Train. Iter 94200 : Commit. 0.54290 PPL. 429.13 Recons. 0.00403 +2024-03-17 01:12:46,647 INFO Train. Iter 94400 : Commit. 0.54523 PPL. 428.76 Recons. 0.00437 +2024-03-17 01:13:07,521 INFO Train. Iter 94600 : Commit. 0.53726 PPL. 429.77 Recons. 0.00395 +2024-03-17 01:13:28,883 INFO Train. Iter 94800 : Commit. 0.50772 PPL. 430.65 Recons. 0.00362 +2024-03-17 01:13:50,156 INFO Train. Iter 95000 : Commit. 0.54201 PPL. 428.66 Recons. 0.00426 +2024-03-17 01:14:11,433 INFO Train. Iter 95200 : Commit. 0.55408 PPL. 427.51 Recons. 0.00473 +2024-03-17 01:14:33,384 INFO Train. Iter 95400 : Commit. 0.52591 PPL. 429.65 Recons. 0.00383 +2024-03-17 01:14:53,569 INFO Train. Iter 95600 : Commit. 0.51841 PPL. 430.55 Recons. 0.00360 +2024-03-17 01:15:14,773 INFO Train. Iter 95800 : Commit. 0.52858 PPL. 429.35 Recons. 0.00410 +2024-03-17 01:15:35,535 INFO Train. Iter 96000 : Commit. 0.55163 PPL. 426.85 Recons. 0.00481 +2024-03-17 01:15:56,681 INFO Train. Iter 96200 : Commit. 0.53175 PPL. 429.06 Recons. 0.00386 +2024-03-17 01:16:18,161 INFO Train. Iter 96400 : Commit. 0.54734 PPL. 428.01 Recons. 0.00433 +2024-03-17 01:16:38,923 INFO Train. Iter 96600 : Commit. 0.56162 PPL. 426.66 Recons. 0.00515 +2024-03-17 01:17:00,172 INFO Train. Iter 96800 : Commit. 0.53360 PPL. 429.43 Recons. 0.00379 +2024-03-17 01:17:21,780 INFO Train. Iter 97000 : Commit. 0.55489 PPL. 427.15 Recons. 0.00461 +2024-03-17 01:17:42,761 INFO Train. Iter 97200 : Commit. 0.52814 PPL. 429.92 Recons. 0.00376 +2024-03-17 01:18:03,923 INFO Train. Iter 97400 : Commit. 0.52727 PPL. 428.97 Recons. 0.00392 +2024-03-17 01:18:24,052 INFO Train. Iter 97600 : Commit. 0.52624 PPL. 428.55 Recons. 0.00407 +2024-03-17 01:18:45,486 INFO Train. Iter 97800 : Commit. 0.54353 PPL. 427.25 Recons. 0.00482 +2024-03-17 01:19:07,112 INFO Train. Iter 98000 : Commit. 0.55639 PPL. 426.82 Recons. 0.00474 +2024-03-17 01:19:28,288 INFO Train. Iter 98200 : Commit. 0.54428 PPL. 428.11 Recons. 0.00437 +2024-03-17 01:19:49,576 INFO Train. Iter 98400 : Commit. 0.53586 PPL. 428.68 Recons. 0.00407 +2024-03-17 01:20:10,986 INFO Train. Iter 98600 : Commit. 0.52244 PPL. 429.97 Recons. 0.00364 +2024-03-17 01:20:32,756 INFO Train. Iter 98800 : Commit. 0.53932 PPL. 427.61 Recons. 0.00436 +2024-03-17 01:20:53,776 INFO Train. Iter 99000 : Commit. 0.52111 PPL. 429.28 Recons. 0.00370 +2024-03-17 01:21:14,997 INFO Train. Iter 99200 : Commit. 0.53496 PPL. 427.71 Recons. 0.00449 +2024-03-17 01:21:36,388 INFO Train. Iter 99400 : Commit. 0.53549 PPL. 428.49 Recons. 0.00408 +2024-03-17 01:21:56,173 INFO Train. Iter 99600 : Commit. 0.53347 PPL. 428.19 Recons. 0.00420 +2024-03-17 01:22:17,816 INFO Train. Iter 99800 : Commit. 0.53687 PPL. 426.94 Recons. 0.00442 +2024-03-17 01:22:38,838 INFO Train. Iter 100000 : Commit. 0.53856 PPL. 427.81 Recons. 0.00435 +2024-03-17 01:23:00,683 INFO Train. Iter 100200 : Commit. 0.52147 PPL. 428.52 Recons. 0.00369 +2024-03-17 01:23:22,397 INFO Train. Iter 100400 : Commit. 0.53042 PPL. 427.73 Recons. 0.00421 +2024-03-17 01:23:44,096 INFO Train. Iter 100600 : Commit. 0.54224 PPL. 428.74 Recons. 0.00423 +2024-03-17 01:24:06,093 INFO Train. Iter 100800 : Commit. 0.53472 PPL. 426.96 Recons. 0.00429 +2024-03-17 01:24:27,566 INFO Train. Iter 101000 : Commit. 0.54304 PPL. 427.05 Recons. 0.00448 +2024-03-17 01:24:49,744 INFO Train. Iter 101200 : Commit. 0.54175 PPL. 426.88 Recons. 0.00453 +2024-03-17 01:25:10,691 INFO Train. Iter 101400 : Commit. 0.55219 PPL. 426.43 Recons. 0.00480 +2024-03-17 01:25:32,650 INFO Train. Iter 101600 : Commit. 0.54873 PPL. 426.93 Recons. 0.00437 +2024-03-17 01:25:54,833 INFO Train. Iter 101800 : Commit. 0.52275 PPL. 429.38 Recons. 0.00369 +2024-03-17 01:26:16,547 INFO Train. Iter 102000 : Commit. 0.55071 PPL. 425.45 Recons. 0.00488 +2024-03-17 01:26:38,263 INFO Train. Iter 102200 : Commit. 0.55357 PPL. 426.97 Recons. 0.00445 +2024-03-17 01:26:59,821 INFO Train. Iter 102400 : Commit. 0.52981 PPL. 427.53 Recons. 0.00394 +2024-03-17 01:27:21,117 INFO Train. Iter 102600 : Commit. 0.54755 PPL. 426.90 Recons. 0.00452 +2024-03-17 01:27:42,412 INFO Train. Iter 102800 : Commit. 0.53413 PPL. 428.52 Recons. 0.00380 +2024-03-17 01:28:03,732 INFO Train. Iter 103000 : Commit. 0.52067 PPL. 428.62 Recons. 0.00382 +2024-03-17 01:28:25,162 INFO Train. Iter 103200 : Commit. 0.52322 PPL. 428.47 Recons. 0.00389 +2024-03-17 01:28:45,196 INFO Train. Iter 103400 : Commit. 0.52711 PPL. 427.74 Recons. 0.00415 +2024-03-17 01:29:06,529 INFO Train. Iter 103600 : Commit. 0.51725 PPL. 428.30 Recons. 0.00381 +2024-03-17 01:29:27,423 INFO Train. Iter 103800 : Commit. 0.52605 PPL. 425.74 Recons. 0.00432 +2024-03-17 01:29:48,855 INFO Train. Iter 104000 : Commit. 0.52445 PPL. 428.56 Recons. 0.00384 +2024-03-17 01:30:10,267 INFO Train. Iter 104200 : Commit. 0.54445 PPL. 426.41 Recons. 0.00435 +2024-03-17 01:30:31,940 INFO Train. Iter 104400 : Commit. 0.52425 PPL. 427.01 Recons. 0.00388 +2024-03-17 01:30:52,919 INFO Train. Iter 104600 : Commit. 0.52827 PPL. 428.14 Recons. 0.00395 +2024-03-17 01:31:14,019 INFO Train. Iter 104800 : Commit. 0.54132 PPL. 426.12 Recons. 0.00446 +2024-03-17 01:31:35,156 INFO Train. Iter 105000 : Commit. 0.51903 PPL. 429.59 Recons. 0.00350 +2024-03-17 01:31:56,509 INFO Train. Iter 105200 : Commit. 0.51466 PPL. 426.98 Recons. 0.00422 +2024-03-17 01:32:16,635 INFO Train. Iter 105400 : Commit. 0.54843 PPL. 425.73 Recons. 0.00461 +2024-03-17 01:32:38,141 INFO Train. Iter 105600 : Commit. 0.53926 PPL. 427.27 Recons. 0.00419 +2024-03-17 01:32:59,624 INFO Train. Iter 105800 : Commit. 0.53149 PPL. 427.03 Recons. 0.00420 +2024-03-17 01:33:21,388 INFO Train. Iter 106000 : Commit. 0.53219 PPL. 427.77 Recons. 0.00395 +2024-03-17 01:33:43,654 INFO Train. Iter 106200 : Commit. 0.51269 PPL. 428.70 Recons. 0.00371 +2024-03-17 01:34:05,338 INFO Train. Iter 106400 : Commit. 0.50976 PPL. 428.21 Recons. 0.00358 +2024-03-17 01:34:26,859 INFO Train. Iter 106600 : Commit. 0.55126 PPL. 424.23 Recons. 0.00499 +2024-03-17 01:34:48,513 INFO Train. Iter 106800 : Commit. 0.50867 PPL. 429.46 Recons. 0.00330 +2024-03-17 01:35:10,435 INFO Train. Iter 107000 : Commit. 0.52394 PPL. 426.84 Recons. 0.00402 +2024-03-17 01:35:31,626 INFO Train. Iter 107200 : Commit. 0.52751 PPL. 425.40 Recons. 0.00461 +2024-03-17 01:35:53,700 INFO Train. Iter 107400 : Commit. 0.54650 PPL. 425.00 Recons. 0.00459 +2024-03-17 01:36:15,824 INFO Train. Iter 107600 : Commit. 0.51899 PPL. 427.95 Recons. 0.00362 +2024-03-17 01:36:37,771 INFO Train. Iter 107800 : Commit. 0.54944 PPL. 424.74 Recons. 0.00500 +2024-03-17 01:36:59,404 INFO Train. Iter 108000 : Commit. 0.53053 PPL. 428.00 Recons. 0.00383 +2024-03-17 01:37:21,034 INFO Train. Iter 108200 : Commit. 0.55890 PPL. 425.28 Recons. 0.00499 +2024-03-17 01:37:42,653 INFO Train. Iter 108400 : Commit. 0.52843 PPL. 427.62 Recons. 0.00397 +2024-03-17 01:38:03,944 INFO Train. Iter 108600 : Commit. 0.53681 PPL. 427.56 Recons. 0.00429 +2024-03-17 01:38:26,375 INFO Train. Iter 108800 : Commit. 0.52054 PPL. 427.42 Recons. 0.00406 +2024-03-17 01:38:47,691 INFO Train. Iter 109000 : Commit. 0.55024 PPL. 425.47 Recons. 0.00460 +2024-03-17 01:39:07,832 INFO Train. Iter 109200 : Commit. 0.52056 PPL. 429.07 Recons. 0.00357 +2024-03-17 01:39:29,037 INFO Train. Iter 109400 : Commit. 0.52032 PPL. 426.28 Recons. 0.00434 +2024-03-17 01:39:50,373 INFO Train. Iter 109600 : Commit. 0.52014 PPL. 427.94 Recons. 0.00372 +2024-03-17 01:40:11,872 INFO Train. Iter 109800 : Commit. 0.52426 PPL. 426.12 Recons. 0.00441 +2024-03-17 01:40:33,605 INFO Train. Iter 110000 : Commit. 0.52906 PPL. 427.14 Recons. 0.00400 +2024-03-17 01:40:55,541 INFO Train. Iter 110200 : Commit. 0.53671 PPL. 425.77 Recons. 0.00447 +2024-03-17 01:41:17,278 INFO Train. Iter 110400 : Commit. 0.54170 PPL. 425.08 Recons. 0.00464 +2024-03-17 01:41:39,110 INFO Train. Iter 110600 : Commit. 0.52011 PPL. 428.40 Recons. 0.00354 +2024-03-17 01:42:01,027 INFO Train. Iter 110800 : Commit. 0.50491 PPL. 428.61 Recons. 0.00346 +2024-03-17 01:42:22,651 INFO Train. Iter 111000 : Commit. 0.51915 PPL. 425.62 Recons. 0.00432 +2024-03-17 01:42:43,392 INFO Train. Iter 111200 : Commit. 0.53081 PPL. 427.11 Recons. 0.00409 +2024-03-17 01:43:05,477 INFO Train. Iter 111400 : Commit. 0.52688 PPL. 426.66 Recons. 0.00402 +2024-03-17 01:43:27,454 INFO Train. Iter 111600 : Commit. 0.52710 PPL. 425.65 Recons. 0.00424 +2024-03-17 01:43:49,733 INFO Train. Iter 111800 : Commit. 0.52460 PPL. 426.86 Recons. 0.00412 +2024-03-17 01:44:11,628 INFO Train. Iter 112000 : Commit. 0.52557 PPL. 426.91 Recons. 0.00415 +2024-03-17 01:44:34,025 INFO Train. Iter 112200 : Commit. 0.52719 PPL. 427.16 Recons. 0.00399 +2024-03-17 01:44:56,007 INFO Train. Iter 112400 : Commit. 0.52394 PPL. 426.63 Recons. 0.00423 +2024-03-17 01:45:17,940 INFO Train. Iter 112600 : Commit. 0.51941 PPL. 427.92 Recons. 0.00385 +2024-03-17 01:45:39,600 INFO Train. Iter 112800 : Commit. 0.50907 PPL. 427.52 Recons. 0.00360 +2024-03-17 01:46:00,071 INFO Train. Iter 113000 : Commit. 0.51099 PPL. 427.03 Recons. 0.00387 +2024-03-17 01:46:22,273 INFO Train. Iter 113200 : Commit. 0.53169 PPL. 426.08 Recons. 0.00410 +2024-03-17 01:46:44,778 INFO Train. Iter 113400 : Commit. 0.51175 PPL. 425.67 Recons. 0.00404 +2024-03-17 01:47:06,560 INFO Train. Iter 113600 : Commit. 0.53005 PPL. 426.45 Recons. 0.00413 +2024-03-17 01:47:28,252 INFO Train. Iter 113800 : Commit. 0.50376 PPL. 428.28 Recons. 0.00345 +2024-03-17 01:47:49,066 INFO Train. Iter 114000 : Commit. 0.53196 PPL. 424.90 Recons. 0.00450 +2024-03-17 01:48:10,641 INFO Train. Iter 114200 : Commit. 0.50783 PPL. 427.80 Recons. 0.00360 +2024-03-17 01:48:31,460 INFO Train. Iter 114400 : Commit. 0.53040 PPL. 426.27 Recons. 0.00419 +2024-03-17 01:48:52,838 INFO Train. Iter 114600 : Commit. 0.52767 PPL. 426.13 Recons. 0.00411 +2024-03-17 01:49:13,882 INFO Train. Iter 114800 : Commit. 0.51147 PPL. 427.22 Recons. 0.00373 +2024-03-17 01:49:34,524 INFO Train. Iter 115000 : Commit. 0.50561 PPL. 428.36 Recons. 0.00349 +2024-03-17 01:49:55,580 INFO Train. Iter 115200 : Commit. 0.52161 PPL. 426.94 Recons. 0.00405 +2024-03-17 01:50:16,935 INFO Train. Iter 115400 : Commit. 0.51440 PPL. 427.81 Recons. 0.00358 +2024-03-17 01:50:37,993 INFO Train. Iter 115600 : Commit. 0.49803 PPL. 428.46 Recons. 0.00342 +2024-03-17 01:50:59,323 INFO Train. Iter 115800 : Commit. 0.54026 PPL. 422.17 Recons. 0.00487 +2024-03-17 01:51:20,439 INFO Train. Iter 116000 : Commit. 0.55733 PPL. 422.64 Recons. 0.00492 +2024-03-17 01:51:41,999 INFO Train. Iter 116200 : Commit. 0.52084 PPL. 426.24 Recons. 0.00406 +2024-03-17 01:52:03,835 INFO Train. Iter 116400 : Commit. 0.52107 PPL. 426.36 Recons. 0.00390 +2024-03-17 01:52:25,152 INFO Train. Iter 116600 : Commit. 0.52816 PPL. 424.17 Recons. 0.00459 +2024-03-17 01:52:46,658 INFO Train. Iter 116800 : Commit. 0.52808 PPL. 426.88 Recons. 0.00383 +2024-03-17 01:53:06,445 INFO Train. Iter 117000 : Commit. 0.53108 PPL. 424.88 Recons. 0.00463 +2024-03-17 01:53:27,194 INFO Train. Iter 117200 : Commit. 0.53491 PPL. 426.15 Recons. 0.00405 +2024-03-17 01:53:48,550 INFO Train. Iter 117400 : Commit. 0.53066 PPL. 425.56 Recons. 0.00417 +2024-03-17 01:54:10,112 INFO Train. Iter 117600 : Commit. 0.52766 PPL. 426.62 Recons. 0.00416 +2024-03-17 01:54:31,570 INFO Train. Iter 117800 : Commit. 0.50702 PPL. 427.32 Recons. 0.00363 +2024-03-17 01:54:52,966 INFO Train. Iter 118000 : Commit. 0.51397 PPL. 425.82 Recons. 0.00411 +2024-03-17 01:55:14,320 INFO Train. Iter 118200 : Commit. 0.49633 PPL. 427.31 Recons. 0.00366 +2024-03-17 01:55:35,500 INFO Train. Iter 118400 : Commit. 0.50819 PPL. 425.88 Recons. 0.00391 +2024-03-17 01:55:57,047 INFO Train. Iter 118600 : Commit. 0.50178 PPL. 427.33 Recons. 0.00351 +2024-03-17 01:56:17,153 INFO Train. Iter 118800 : Commit. 0.51010 PPL. 426.73 Recons. 0.00380 +2024-03-17 01:56:38,131 INFO Train. Iter 119000 : Commit. 0.51945 PPL. 426.49 Recons. 0.00403 +2024-03-17 01:56:59,373 INFO Train. Iter 119200 : Commit. 0.51952 PPL. 426.50 Recons. 0.00402 +2024-03-17 01:57:20,903 INFO Train. Iter 119400 : Commit. 0.51371 PPL. 426.02 Recons. 0.00388 +2024-03-17 01:57:42,065 INFO Train. Iter 119600 : Commit. 0.52685 PPL. 425.51 Recons. 0.00401 +2024-03-17 01:58:03,243 INFO Train. Iter 119800 : Commit. 0.51529 PPL. 426.41 Recons. 0.00381 +2024-03-17 01:58:24,503 INFO Train. Iter 120000 : Commit. 0.50559 PPL. 426.77 Recons. 0.00353 +2024-03-17 01:58:45,896 INFO Train. Iter 120200 : Commit. 0.51985 PPL. 426.42 Recons. 0.00387 +2024-03-17 01:59:07,162 INFO Train. Iter 120400 : Commit. 0.52613 PPL. 424.84 Recons. 0.00427 +2024-03-17 01:59:27,862 INFO Train. Iter 120600 : Commit. 0.51319 PPL. 426.25 Recons. 0.00381 +2024-03-17 01:59:47,809 INFO Train. Iter 120800 : Commit. 0.51246 PPL. 426.57 Recons. 0.00364 +2024-03-17 02:00:09,231 INFO Train. Iter 121000 : Commit. 0.51484 PPL. 426.24 Recons. 0.00375 +2024-03-17 02:00:30,579 INFO Train. Iter 121200 : Commit. 0.51705 PPL. 425.36 Recons. 0.00398 +2024-03-17 02:00:52,728 INFO Train. Iter 121400 : Commit. 0.50279 PPL. 427.54 Recons. 0.00343 +2024-03-17 02:01:14,136 INFO Train. Iter 121600 : Commit. 0.50330 PPL. 425.19 Recons. 0.00395 +2024-03-17 02:01:35,269 INFO Train. Iter 121800 : Commit. 0.52188 PPL. 425.76 Recons. 0.00410 +2024-03-17 02:01:57,255 INFO Train. Iter 122000 : Commit. 0.51350 PPL. 426.93 Recons. 0.00370 +2024-03-17 02:02:18,351 INFO Train. Iter 122200 : Commit. 0.52080 PPL. 425.37 Recons. 0.00420 +2024-03-17 02:02:39,230 INFO Train. Iter 122400 : Commit. 0.51482 PPL. 425.83 Recons. 0.00416 +2024-03-17 02:03:00,407 INFO Train. Iter 122600 : Commit. 0.51126 PPL. 426.60 Recons. 0.00366 +2024-03-17 02:03:20,764 INFO Train. Iter 122800 : Commit. 0.51109 PPL. 426.27 Recons. 0.00372 +2024-03-17 02:03:42,037 INFO Train. Iter 123000 : Commit. 0.48903 PPL. 427.70 Recons. 0.00326 +2024-03-17 02:04:03,085 INFO Train. Iter 123200 : Commit. 0.53348 PPL. 423.40 Recons. 0.00466 +2024-03-17 02:04:24,183 INFO Train. Iter 123400 : Commit. 0.52649 PPL. 426.16 Recons. 0.00408 +2024-03-17 02:04:45,252 INFO Train. Iter 123600 : Commit. 0.49637 PPL. 427.49 Recons. 0.00326 +2024-03-17 02:05:06,671 INFO Train. Iter 123800 : Commit. 0.52546 PPL. 424.92 Recons. 0.00457 +2024-03-17 02:05:28,239 INFO Train. Iter 124000 : Commit. 0.48825 PPL. 427.77 Recons. 0.00311 +2024-03-17 02:05:50,076 INFO Train. Iter 124200 : Commit. 0.51420 PPL. 423.48 Recons. 0.00431 +2024-03-17 02:06:11,978 INFO Train. Iter 124400 : Commit. 0.51199 PPL. 425.53 Recons. 0.00384 +2024-03-17 02:06:33,280 INFO Train. Iter 124600 : Commit. 0.50121 PPL. 426.80 Recons. 0.00353 +2024-03-17 02:06:55,652 INFO Train. Iter 124800 : Commit. 0.51728 PPL. 425.15 Recons. 0.00420 +2024-03-17 02:07:17,431 INFO Train. Iter 125000 : Commit. 0.53609 PPL. 424.43 Recons. 0.00443 +2024-03-17 02:07:39,501 INFO Train. Iter 125200 : Commit. 0.51512 PPL. 425.70 Recons. 0.00389 +2024-03-17 02:08:01,618 INFO Train. Iter 125400 : Commit. 0.50988 PPL. 426.77 Recons. 0.00374 +2024-03-17 02:08:23,745 INFO Train. Iter 125600 : Commit. 0.50103 PPL. 426.32 Recons. 0.00352 +2024-03-17 02:08:45,269 INFO Train. Iter 125800 : Commit. 0.50704 PPL. 426.59 Recons. 0.00364 +2024-03-17 02:09:06,876 INFO Train. Iter 126000 : Commit. 0.50819 PPL. 425.44 Recons. 0.00392 +2024-03-17 02:09:28,215 INFO Train. Iter 126200 : Commit. 0.51388 PPL. 425.18 Recons. 0.00385 +2024-03-17 02:09:49,680 INFO Train. Iter 126400 : Commit. 0.50442 PPL. 424.33 Recons. 0.00392 +2024-03-17 02:10:09,720 INFO Train. Iter 126600 : Commit. 0.49235 PPL. 426.98 Recons. 0.00336 +2024-03-17 02:10:31,486 INFO Train. Iter 126800 : Commit. 0.53197 PPL. 422.15 Recons. 0.00483 +2024-03-17 02:10:53,690 INFO Train. Iter 127000 : Commit. 0.51515 PPL. 426.03 Recons. 0.00373 +2024-03-17 02:11:15,570 INFO Train. Iter 127200 : Commit. 0.51956 PPL. 424.89 Recons. 0.00406 +2024-03-17 02:11:37,002 INFO Train. Iter 127400 : Commit. 0.50387 PPL. 426.24 Recons. 0.00355 +2024-03-17 02:11:58,260 INFO Train. Iter 127600 : Commit. 0.49774 PPL. 423.88 Recons. 0.00380 +2024-03-17 02:12:19,293 INFO Train. Iter 127800 : Commit. 0.50636 PPL. 426.13 Recons. 0.00346 +2024-03-17 02:12:40,298 INFO Train. Iter 128000 : Commit. 0.49954 PPL. 425.78 Recons. 0.00358 +2024-03-17 02:13:01,144 INFO Train. Iter 128200 : Commit. 0.53309 PPL. 422.07 Recons. 0.00468 +2024-03-17 02:13:22,624 INFO Train. Iter 128400 : Commit. 0.52189 PPL. 424.09 Recons. 0.00399 +2024-03-17 02:13:42,890 INFO Train. Iter 128600 : Commit. 0.50866 PPL. 425.25 Recons. 0.00366 +2024-03-17 02:14:04,375 INFO Train. Iter 128800 : Commit. 0.50396 PPL. 425.60 Recons. 0.00363 +2024-03-17 02:14:25,677 INFO Train. Iter 129000 : Commit. 0.50378 PPL. 425.38 Recons. 0.00370 +2024-03-17 02:14:46,619 INFO Train. Iter 129200 : Commit. 0.49790 PPL. 426.15 Recons. 0.00355 +2024-03-17 02:15:08,554 INFO Train. Iter 129400 : Commit. 0.52486 PPL. 423.42 Recons. 0.00452 +2024-03-17 02:15:30,663 INFO Train. Iter 129600 : Commit. 0.51944 PPL. 425.29 Recons. 0.00391 +2024-03-17 02:15:52,027 INFO Train. Iter 129800 : Commit. 0.49929 PPL. 426.19 Recons. 0.00335 +2024-03-17 02:16:13,031 INFO Train. Iter 130000 : Commit. 0.52425 PPL. 422.66 Recons. 0.00475 +2024-03-17 02:16:34,276 INFO Train. Iter 130200 : Commit. 0.52643 PPL. 423.98 Recons. 0.00429 +2024-03-17 02:16:54,047 INFO Train. Iter 130400 : Commit. 0.51691 PPL. 424.76 Recons. 0.00400 +2024-03-17 02:17:15,505 INFO Train. Iter 130600 : Commit. 0.52521 PPL. 422.95 Recons. 0.00421 +2024-03-17 02:17:36,937 INFO Train. Iter 130800 : Commit. 0.52216 PPL. 424.87 Recons. 0.00411 +2024-03-17 02:17:58,584 INFO Train. Iter 131000 : Commit. 0.52010 PPL. 425.04 Recons. 0.00390 +2024-03-17 02:18:19,862 INFO Train. Iter 131200 : Commit. 0.49725 PPL. 426.14 Recons. 0.00334 +2024-03-17 02:18:41,711 INFO Train. Iter 131400 : Commit. 0.51853 PPL. 423.11 Recons. 0.00451 +2024-03-17 02:19:03,306 INFO Train. Iter 131600 : Commit. 0.51467 PPL. 423.62 Recons. 0.00375 +2024-03-17 02:19:24,906 INFO Train. Iter 131800 : Commit. 0.50704 PPL. 425.38 Recons. 0.00360 +2024-03-17 02:19:46,688 INFO Train. Iter 132000 : Commit. 0.51304 PPL. 424.90 Recons. 0.00392 +2024-03-17 02:20:08,293 INFO Train. Iter 132200 : Commit. 0.50985 PPL. 424.61 Recons. 0.00367 +2024-03-17 02:20:28,950 INFO Train. Iter 132400 : Commit. 0.49269 PPL. 425.94 Recons. 0.00327 +2024-03-17 02:20:50,456 INFO Train. Iter 132600 : Commit. 0.49348 PPL. 425.36 Recons. 0.00345 +2024-03-17 02:21:12,295 INFO Train. Iter 132800 : Commit. 0.49879 PPL. 425.20 Recons. 0.00348 +2024-03-17 02:21:34,060 INFO Train. Iter 133000 : Commit. 0.51033 PPL. 422.44 Recons. 0.00422 +2024-03-17 02:21:55,833 INFO Train. Iter 133200 : Commit. 0.52714 PPL. 422.20 Recons. 0.00445 +2024-03-17 02:22:17,956 INFO Train. Iter 133400 : Commit. 0.50426 PPL. 424.93 Recons. 0.00342 +2024-03-17 02:22:40,519 INFO Train. Iter 133600 : Commit. 0.52194 PPL. 422.06 Recons. 0.00440 +2024-03-17 02:23:02,527 INFO Train. Iter 133800 : Commit. 0.50828 PPL. 426.02 Recons. 0.00335 +2024-03-17 02:23:24,648 INFO Train. Iter 134000 : Commit. 0.52779 PPL. 421.54 Recons. 0.00459 +2024-03-17 02:23:46,881 INFO Train. Iter 134200 : Commit. 0.49923 PPL. 426.37 Recons. 0.00326 +2024-03-17 02:24:07,595 INFO Train. Iter 134400 : Commit. 0.48786 PPL. 425.99 Recons. 0.00325 +2024-03-17 02:24:29,744 INFO Train. Iter 134600 : Commit. 0.49207 PPL. 425.44 Recons. 0.00337 +2024-03-17 02:24:51,549 INFO Train. Iter 134800 : Commit. 0.50244 PPL. 424.41 Recons. 0.00366 +2024-03-17 02:25:13,200 INFO Train. Iter 135000 : Commit. 0.49990 PPL. 425.64 Recons. 0.00346 +2024-03-17 02:25:35,438 INFO Train. Iter 135200 : Commit. 0.50707 PPL. 422.89 Recons. 0.00402 +2024-03-17 02:25:57,486 INFO Train. Iter 135400 : Commit. 0.52367 PPL. 423.22 Recons. 0.00401 +2024-03-17 02:26:20,260 INFO Train. Iter 135600 : Commit. 0.49051 PPL. 426.27 Recons. 0.00309 +2024-03-17 02:26:42,722 INFO Train. Iter 135800 : Commit. 0.52601 PPL. 421.61 Recons. 0.00441 +2024-03-17 02:27:04,985 INFO Train. Iter 136000 : Commit. 0.52286 PPL. 422.20 Recons. 0.00412 +2024-03-17 02:27:25,874 INFO Train. Iter 136200 : Commit. 0.50756 PPL. 423.27 Recons. 0.00369 +2024-03-17 02:27:47,676 INFO Train. Iter 136400 : Commit. 0.52310 PPL. 422.95 Recons. 0.00413 +2024-03-17 02:28:09,634 INFO Train. Iter 136600 : Commit. 0.50198 PPL. 424.04 Recons. 0.00361 +2024-03-17 02:28:31,604 INFO Train. Iter 136800 : Commit. 0.52271 PPL. 421.87 Recons. 0.00437 +2024-03-17 02:28:53,538 INFO Train. Iter 137000 : Commit. 0.49908 PPL. 425.31 Recons. 0.00331 +2024-03-17 02:29:15,249 INFO Train. Iter 137200 : Commit. 0.50122 PPL. 424.44 Recons. 0.00364 +2024-03-17 02:29:36,890 INFO Train. Iter 137400 : Commit. 0.49797 PPL. 424.99 Recons. 0.00342 +2024-03-17 02:29:58,278 INFO Train. Iter 137600 : Commit. 0.50341 PPL. 424.32 Recons. 0.00367 +2024-03-17 02:30:20,305 INFO Train. Iter 137800 : Commit. 0.49742 PPL. 424.06 Recons. 0.00376 +2024-03-17 02:30:41,998 INFO Train. Iter 138000 : Commit. 0.51885 PPL. 423.15 Recons. 0.00417 +2024-03-17 02:31:02,606 INFO Train. Iter 138200 : Commit. 0.50183 PPL. 424.32 Recons. 0.00378 +2024-03-17 02:31:24,426 INFO Train. Iter 138400 : Commit. 0.50385 PPL. 424.80 Recons. 0.00371 +2024-03-17 02:31:45,635 INFO Train. Iter 138600 : Commit. 0.50214 PPL. 424.57 Recons. 0.00364 +2024-03-17 02:32:07,043 INFO Train. Iter 138800 : Commit. 0.50977 PPL. 422.57 Recons. 0.00413 +2024-03-17 02:32:28,347 INFO Train. Iter 139000 : Commit. 0.51565 PPL. 423.33 Recons. 0.00432 +2024-03-17 02:32:50,126 INFO Train. Iter 139200 : Commit. 0.50912 PPL. 425.06 Recons. 0.00351 +2024-03-17 02:33:12,393 INFO Train. Iter 139400 : Commit. 0.50757 PPL. 423.85 Recons. 0.00384 +2024-03-17 02:33:34,250 INFO Train. Iter 139600 : Commit. 0.49344 PPL. 424.51 Recons. 0.00354 +2024-03-17 02:33:55,643 INFO Train. Iter 139800 : Commit. 0.52651 PPL. 422.80 Recons. 0.00449 +2024-03-17 02:34:16,737 INFO Train. Iter 140000 : Commit. 0.50334 PPL. 423.38 Recons. 0.00382 +2024-03-17 02:34:36,908 INFO Train. Iter 140200 : Commit. 0.49998 PPL. 424.42 Recons. 0.00372 +2024-03-17 02:34:58,590 INFO Train. Iter 140400 : Commit. 0.49977 PPL. 423.88 Recons. 0.00359 +2024-03-17 02:35:20,335 INFO Train. Iter 140600 : Commit. 0.48648 PPL. 425.21 Recons. 0.00328 +2024-03-17 02:35:42,249 INFO Train. Iter 140800 : Commit. 0.48492 PPL. 424.76 Recons. 0.00337 +2024-03-17 02:36:04,188 INFO Train. Iter 141000 : Commit. 0.48454 PPL. 425.17 Recons. 0.00322 +2024-03-17 02:36:26,318 INFO Train. Iter 141200 : Commit. 0.52258 PPL. 420.21 Recons. 0.00491 +2024-03-17 02:36:48,147 INFO Train. Iter 141400 : Commit. 0.49212 PPL. 424.99 Recons. 0.00316 +2024-03-17 02:37:10,342 INFO Train. Iter 141600 : Commit. 0.48994 PPL. 424.75 Recons. 0.00334 +2024-03-17 02:37:32,942 INFO Train. Iter 141800 : Commit. 0.50256 PPL. 422.88 Recons. 0.00385 +2024-03-17 02:37:53,826 INFO Train. Iter 142000 : Commit. 0.51581 PPL. 423.43 Recons. 0.00393 +2024-03-17 02:38:15,864 INFO Train. Iter 142200 : Commit. 0.49149 PPL. 424.35 Recons. 0.00334 +2024-03-17 02:38:38,058 INFO Train. Iter 142400 : Commit. 0.51103 PPL. 423.59 Recons. 0.00389 +2024-03-17 02:39:00,278 INFO Train. Iter 142600 : Commit. 0.53148 PPL. 421.45 Recons. 0.00436 +2024-03-17 02:39:21,948 INFO Train. Iter 142800 : Commit. 0.50201 PPL. 424.01 Recons. 0.00350 +2024-03-17 02:39:44,473 INFO Train. Iter 143000 : Commit. 0.52572 PPL. 420.98 Recons. 0.00466 +2024-03-17 02:40:06,019 INFO Train. Iter 143200 : Commit. 0.49659 PPL. 425.29 Recons. 0.00313 +2024-03-17 02:40:27,594 INFO Train. Iter 143400 : Commit. 0.49625 PPL. 424.31 Recons. 0.00350 +2024-03-17 02:40:48,928 INFO Train. Iter 143600 : Commit. 0.51071 PPL. 422.92 Recons. 0.00395 +2024-03-17 02:41:10,149 INFO Train. Iter 143800 : Commit. 0.50053 PPL. 422.62 Recons. 0.00380 +2024-03-17 02:41:31,294 INFO Train. Iter 144000 : Commit. 0.50441 PPL. 424.12 Recons. 0.00369 +2024-03-17 02:41:53,084 INFO Train. Iter 144200 : Commit. 0.49344 PPL. 424.10 Recons. 0.00342 +2024-03-17 02:42:15,619 INFO Train. Iter 144400 : Commit. 0.49355 PPL. 424.16 Recons. 0.00351 +2024-03-17 02:42:37,229 INFO Train. Iter 144600 : Commit. 0.50572 PPL. 421.46 Recons. 0.00412 +2024-03-17 02:42:59,003 INFO Train. Iter 144800 : Commit. 0.49689 PPL. 424.83 Recons. 0.00338 +2024-03-17 02:43:20,396 INFO Train. Iter 145000 : Commit. 0.52373 PPL. 420.80 Recons. 0.00438 +2024-03-17 02:43:41,767 INFO Train. Iter 145200 : Commit. 0.49653 PPL. 424.69 Recons. 0.00334 +2024-03-17 02:44:03,174 INFO Train. Iter 145400 : Commit. 0.50573 PPL. 421.77 Recons. 0.00396 +2024-03-17 02:44:24,517 INFO Train. Iter 145600 : Commit. 0.51038 PPL. 423.22 Recons. 0.00396 +2024-03-17 02:44:45,866 INFO Train. Iter 145800 : Commit. 0.48416 PPL. 424.98 Recons. 0.00311 +2024-03-17 02:45:06,020 INFO Train. Iter 146000 : Commit. 0.51637 PPL. 420.82 Recons. 0.00444 +2024-03-17 02:45:27,613 INFO Train. Iter 146200 : Commit. 0.48765 PPL. 424.11 Recons. 0.00329 +2024-03-17 02:45:49,129 INFO Train. Iter 146400 : Commit. 0.50678 PPL. 423.70 Recons. 0.00360 +2024-03-17 02:46:11,004 INFO Train. Iter 146600 : Commit. 0.48558 PPL. 423.97 Recons. 0.00328 +2024-03-17 02:46:32,354 INFO Train. Iter 146800 : Commit. 0.49676 PPL. 423.19 Recons. 0.00369 +2024-03-17 02:46:53,699 INFO Train. Iter 147000 : Commit. 0.50812 PPL. 422.78 Recons. 0.00399 +2024-03-17 02:47:15,105 INFO Train. Iter 147200 : Commit. 0.49771 PPL. 423.59 Recons. 0.00357 +2024-03-17 02:47:36,779 INFO Train. Iter 147400 : Commit. 0.47509 PPL. 425.41 Recons. 0.00296 +2024-03-17 02:47:58,725 INFO Train. Iter 147600 : Commit. 0.49868 PPL. 422.84 Recons. 0.00377 +2024-03-17 02:48:20,227 INFO Train. Iter 147800 : Commit. 0.48700 PPL. 424.70 Recons. 0.00319 +2024-03-17 02:48:42,146 INFO Train. Iter 148000 : Commit. 0.49943 PPL. 421.39 Recons. 0.00385 +2024-03-17 02:49:03,911 INFO Train. Iter 148200 : Commit. 0.50678 PPL. 422.60 Recons. 0.00390 +2024-03-17 02:49:25,918 INFO Train. Iter 148400 : Commit. 0.48681 PPL. 424.30 Recons. 0.00327 +2024-03-17 02:49:47,889 INFO Train. Iter 148600 : Commit. 0.49917 PPL. 423.06 Recons. 0.00360 +2024-03-17 02:50:09,956 INFO Train. Iter 148800 : Commit. 0.49876 PPL. 423.91 Recons. 0.00345 +2024-03-17 02:50:32,667 INFO Train. Iter 149000 : Commit. 0.49375 PPL. 423.74 Recons. 0.00359 +2024-03-17 02:50:54,848 INFO Train. Iter 149200 : Commit. 0.51932 PPL. 420.46 Recons. 0.00440 +2024-03-17 02:51:16,848 INFO Train. Iter 149400 : Commit. 0.51319 PPL. 423.10 Recons. 0.00374 +2024-03-17 02:51:38,288 INFO Train. Iter 149600 : Commit. 0.48284 PPL. 424.39 Recons. 0.00312 +2024-03-17 02:51:58,825 INFO Train. Iter 149800 : Commit. 0.47546 PPL. 424.54 Recons. 0.00312 +2024-03-17 02:52:21,486 INFO Train. Iter 150000 : Commit. 0.50872 PPL. 421.28 Recons. 0.00400 +2024-03-17 02:52:43,637 INFO Train. Iter 150200 : Commit. 0.50177 PPL. 423.79 Recons. 0.00350 +2024-03-17 02:53:05,564 INFO Train. Iter 150400 : Commit. 0.52277 PPL. 420.37 Recons. 0.00446 +2024-03-17 02:53:27,225 INFO Train. Iter 150600 : Commit. 0.51199 PPL. 421.63 Recons. 0.00397 +2024-03-17 02:53:49,395 INFO Train. Iter 150800 : Commit. 0.49419 PPL. 422.98 Recons. 0.00346 +2024-03-17 02:54:11,576 INFO Train. Iter 151000 : Commit. 0.50614 PPL. 420.96 Recons. 0.00405 +2024-03-17 02:54:33,602 INFO Train. Iter 151200 : Commit. 0.49741 PPL. 424.23 Recons. 0.00334 +2024-03-17 02:54:55,493 INFO Train. Iter 151400 : Commit. 0.48789 PPL. 423.86 Recons. 0.00333 +2024-03-17 02:55:17,148 INFO Train. Iter 151600 : Commit. 0.51276 PPL. 420.68 Recons. 0.00422 +2024-03-17 02:55:38,586 INFO Train. Iter 151800 : Commit. 0.48631 PPL. 424.52 Recons. 0.00316 +2024-03-17 02:56:00,367 INFO Train. Iter 152000 : Commit. 0.50711 PPL. 420.77 Recons. 0.00413 +2024-03-17 02:56:22,494 INFO Train. Iter 152200 : Commit. 0.47131 PPL. 425.05 Recons. 0.00290 +2024-03-17 02:56:44,096 INFO Train. Iter 152400 : Commit. 0.48881 PPL. 422.91 Recons. 0.00349 +2024-03-17 02:57:06,270 INFO Train. Iter 152600 : Commit. 0.48459 PPL. 423.49 Recons. 0.00343 +2024-03-17 02:57:28,256 INFO Train. Iter 152800 : Commit. 0.49264 PPL. 422.58 Recons. 0.00362 +2024-03-17 02:57:50,185 INFO Train. Iter 153000 : Commit. 0.49271 PPL. 423.76 Recons. 0.00343 +2024-03-17 02:58:12,424 INFO Train. Iter 153200 : Commit. 0.50041 PPL. 421.33 Recons. 0.00386 +2024-03-17 02:58:34,255 INFO Train. Iter 153400 : Commit. 0.51108 PPL. 422.56 Recons. 0.00387 +2024-03-17 02:58:55,338 INFO Train. Iter 153600 : Commit. 0.48034 PPL. 423.80 Recons. 0.00316 +2024-03-17 02:59:17,669 INFO Train. Iter 153800 : Commit. 0.48360 PPL. 423.41 Recons. 0.00340 +2024-03-17 02:59:40,264 INFO Train. Iter 154000 : Commit. 0.50797 PPL. 420.99 Recons. 0.00422 +2024-03-17 03:00:02,605 INFO Train. Iter 154200 : Commit. 0.49552 PPL. 423.53 Recons. 0.00341 +2024-03-17 03:00:24,571 INFO Train. Iter 154400 : Commit. 0.47974 PPL. 422.92 Recons. 0.00329 +2024-03-17 03:00:47,155 INFO Train. Iter 154600 : Commit. 0.48521 PPL. 423.14 Recons. 0.00333 +2024-03-17 03:01:09,511 INFO Train. Iter 154800 : Commit. 0.49482 PPL. 422.13 Recons. 0.00382 +2024-03-17 03:01:31,485 INFO Train. Iter 155000 : Commit. 0.50006 PPL. 422.36 Recons. 0.00358 +2024-03-17 03:01:54,152 INFO Train. Iter 155200 : Commit. 0.48927 PPL. 421.56 Recons. 0.00366 +2024-03-17 03:02:16,336 INFO Train. Iter 155400 : Commit. 0.48550 PPL. 423.64 Recons. 0.00319 +2024-03-17 03:02:37,468 INFO Train. Iter 155600 : Commit. 0.50365 PPL. 421.71 Recons. 0.00379 +2024-03-17 03:02:59,258 INFO Train. Iter 155800 : Commit. 0.48225 PPL. 423.04 Recons. 0.00328 +2024-03-17 03:03:21,585 INFO Train. Iter 156000 : Commit. 0.46404 PPL. 424.27 Recons. 0.00301 +2024-03-17 03:03:43,655 INFO Train. Iter 156200 : Commit. 0.50812 PPL. 420.25 Recons. 0.00434 +2024-03-17 03:04:05,542 INFO Train. Iter 156400 : Commit. 0.50947 PPL. 420.33 Recons. 0.00396 +2024-03-17 03:04:27,985 INFO Train. Iter 156600 : Commit. 0.48918 PPL. 423.49 Recons. 0.00321 +2024-03-17 03:04:49,839 INFO Train. Iter 156800 : Commit. 0.48601 PPL. 422.92 Recons. 0.00345 +2024-03-17 03:05:12,139 INFO Train. Iter 157000 : Commit. 0.49388 PPL. 422.93 Recons. 0.00358 +2024-03-17 03:05:34,781 INFO Train. Iter 157200 : Commit. 0.49252 PPL. 422.29 Recons. 0.00347 +2024-03-17 03:05:57,179 INFO Train. Iter 157400 : Commit. 0.48390 PPL. 422.64 Recons. 0.00342 +2024-03-17 03:06:17,742 INFO Train. Iter 157600 : Commit. 0.49862 PPL. 421.62 Recons. 0.00377 +2024-03-17 03:06:39,725 INFO Train. Iter 157800 : Commit. 0.49985 PPL. 421.44 Recons. 0.00368 +2024-03-17 03:07:01,818 INFO Train. Iter 158000 : Commit. 0.49990 PPL. 421.44 Recons. 0.00389 +2024-03-17 03:07:24,228 INFO Train. Iter 158200 : Commit. 0.48725 PPL. 422.81 Recons. 0.00336 +2024-03-17 03:07:46,601 INFO Train. Iter 158400 : Commit. 0.48971 PPL. 421.76 Recons. 0.00368 +2024-03-17 03:08:08,910 INFO Train. Iter 158600 : Commit. 0.50359 PPL. 421.19 Recons. 0.00385 +2024-03-17 03:08:31,563 INFO Train. Iter 158800 : Commit. 0.50195 PPL. 421.71 Recons. 0.00351 +2024-03-17 03:08:53,661 INFO Train. Iter 159000 : Commit. 0.47694 PPL. 423.02 Recons. 0.00309 +2024-03-17 03:09:15,837 INFO Train. Iter 159200 : Commit. 0.49248 PPL. 421.60 Recons. 0.00364 +2024-03-17 03:09:36,473 INFO Train. Iter 159400 : Commit. 0.48375 PPL. 423.30 Recons. 0.00312 +2024-03-17 03:09:58,700 INFO Train. Iter 159600 : Commit. 0.47254 PPL. 423.03 Recons. 0.00328 +2024-03-17 03:10:20,980 INFO Train. Iter 159800 : Commit. 0.49101 PPL. 420.65 Recons. 0.00366 +2024-03-17 03:10:43,324 INFO Train. Iter 160000 : Commit. 0.48193 PPL. 423.20 Recons. 0.00311 +2024-03-17 03:11:05,625 INFO Train. Iter 160200 : Commit. 0.46664 PPL. 423.38 Recons. 0.00296 +2024-03-17 03:11:27,600 INFO Train. Iter 160400 : Commit. 0.48144 PPL. 422.06 Recons. 0.00346 +2024-03-17 03:11:49,371 INFO Train. Iter 160600 : Commit. 0.48501 PPL. 422.86 Recons. 0.00324 +2024-03-17 03:12:11,218 INFO Train. Iter 160800 : Commit. 0.48914 PPL. 422.06 Recons. 0.00346 +2024-03-17 03:12:33,073 INFO Train. Iter 161000 : Commit. 0.48687 PPL. 422.43 Recons. 0.00344 +2024-03-17 03:12:55,123 INFO Train. Iter 161200 : Commit. 0.49095 PPL. 421.02 Recons. 0.00384 +2024-03-17 03:13:16,164 INFO Train. Iter 161400 : Commit. 0.50219 PPL. 422.16 Recons. 0.00366 +2024-03-17 03:13:38,307 INFO Train. Iter 161600 : Commit. 0.49340 PPL. 421.52 Recons. 0.00356 +2024-03-17 03:13:59,977 INFO Train. Iter 161800 : Commit. 0.48652 PPL. 422.43 Recons. 0.00329 +2024-03-17 03:14:21,505 INFO Train. Iter 162000 : Commit. 0.49631 PPL. 420.36 Recons. 0.00402 +2024-03-17 03:14:43,695 INFO Train. Iter 162200 : Commit. 0.49061 PPL. 423.15 Recons. 0.00327 +2024-03-17 03:15:06,904 INFO Train. Iter 162400 : Commit. 0.48466 PPL. 422.18 Recons. 0.00339 +2024-03-17 03:15:28,354 INFO Train. Iter 162600 : Commit. 0.51122 PPL. 419.91 Recons. 0.00402 +2024-03-17 03:15:50,009 INFO Train. Iter 162800 : Commit. 0.51467 PPL. 417.43 Recons. 0.00454 +2024-03-17 03:16:11,203 INFO Train. Iter 163000 : Commit. 0.51819 PPL. 420.48 Recons. 0.00408 +2024-03-17 03:16:33,057 INFO Train. Iter 163200 : Commit. 0.48274 PPL. 422.59 Recons. 0.00313 +2024-03-17 03:16:53,727 INFO Train. Iter 163400 : Commit. 0.48700 PPL. 422.83 Recons. 0.00339 +2024-03-17 03:17:15,471 INFO Train. Iter 163600 : Commit. 0.47569 PPL. 421.89 Recons. 0.00321 +2024-03-17 03:17:36,822 INFO Train. Iter 163800 : Commit. 0.50846 PPL. 418.93 Recons. 0.00441 +2024-03-17 03:17:58,882 INFO Train. Iter 164000 : Commit. 0.48222 PPL. 421.68 Recons. 0.00331 +2024-03-17 03:18:20,730 INFO Train. Iter 164200 : Commit. 0.50218 PPL. 420.77 Recons. 0.00373 +2024-03-17 03:18:41,911 INFO Train. Iter 164400 : Commit. 0.49574 PPL. 421.43 Recons. 0.00372 +2024-03-17 03:19:03,803 INFO Train. Iter 164600 : Commit. 0.49056 PPL. 421.57 Recons. 0.00345 +2024-03-17 03:19:24,843 INFO Train. Iter 164800 : Commit. 0.48442 PPL. 421.83 Recons. 0.00328 +2024-03-17 03:19:46,751 INFO Train. Iter 165000 : Commit. 0.48961 PPL. 420.23 Recons. 0.00378 +2024-03-17 03:20:07,537 INFO Train. Iter 165200 : Commit. 0.49021 PPL. 422.13 Recons. 0.00337 +2024-03-17 03:20:29,577 INFO Train. Iter 165400 : Commit. 0.49437 PPL. 421.56 Recons. 0.00345 +2024-03-17 03:20:51,235 INFO Train. Iter 165600 : Commit. 0.47393 PPL. 421.69 Recons. 0.00323 +2024-03-17 03:21:12,340 INFO Train. Iter 165800 : Commit. 0.49205 PPL. 420.97 Recons. 0.00364 +2024-03-17 03:21:33,582 INFO Train. Iter 166000 : Commit. 0.47046 PPL. 421.54 Recons. 0.00314 +2024-03-17 03:21:55,406 INFO Train. Iter 166200 : Commit. 0.49112 PPL. 421.09 Recons. 0.00356 +2024-03-17 03:22:17,145 INFO Train. Iter 166400 : Commit. 0.50224 PPL. 419.64 Recons. 0.00385 +2024-03-17 03:22:39,248 INFO Train. Iter 166600 : Commit. 0.48640 PPL. 422.01 Recons. 0.00332 +2024-03-17 03:23:00,735 INFO Train. Iter 166800 : Commit. 0.46673 PPL. 423.42 Recons. 0.00290 +2024-03-17 03:23:22,307 INFO Train. Iter 167000 : Commit. 0.47084 PPL. 422.10 Recons. 0.00324 +2024-03-17 03:23:42,617 INFO Train. Iter 167200 : Commit. 0.48843 PPL. 421.23 Recons. 0.00361 +2024-03-17 03:24:04,819 INFO Train. Iter 167400 : Commit. 0.49021 PPL. 421.57 Recons. 0.00341 +2024-03-17 03:24:26,496 INFO Train. Iter 167600 : Commit. 0.48959 PPL. 420.81 Recons. 0.00355 +2024-03-17 03:24:48,214 INFO Train. Iter 167800 : Commit. 0.48667 PPL. 420.68 Recons. 0.00340 +2024-03-17 03:25:09,381 INFO Train. Iter 168000 : Commit. 0.47948 PPL. 421.15 Recons. 0.00350 +2024-03-17 03:25:30,739 INFO Train. Iter 168200 : Commit. 0.49180 PPL. 420.76 Recons. 0.00356 +2024-03-17 03:25:52,412 INFO Train. Iter 168400 : Commit. 0.49146 PPL. 420.11 Recons. 0.00387 +2024-03-17 03:26:13,853 INFO Train. Iter 168600 : Commit. 0.48519 PPL. 421.02 Recons. 0.00355 +2024-03-17 03:26:35,289 INFO Train. Iter 168800 : Commit. 0.48790 PPL. 421.48 Recons. 0.00343 +2024-03-17 03:26:56,910 INFO Train. Iter 169000 : Commit. 0.47781 PPL. 421.62 Recons. 0.00310 +2024-03-17 03:27:16,946 INFO Train. Iter 169200 : Commit. 0.48593 PPL. 421.05 Recons. 0.00349 +2024-03-17 03:27:38,624 INFO Train. Iter 169400 : Commit. 0.47864 PPL. 421.41 Recons. 0.00323 +2024-03-17 03:28:00,298 INFO Train. Iter 169600 : Commit. 0.49054 PPL. 419.33 Recons. 0.00356 +2024-03-17 03:28:21,796 INFO Train. Iter 169800 : Commit. 0.49101 PPL. 421.33 Recons. 0.00327 +2024-03-17 03:28:43,440 INFO Train. Iter 170000 : Commit. 0.47548 PPL. 421.16 Recons. 0.00326 +2024-03-17 03:29:04,975 INFO Train. Iter 170200 : Commit. 0.48220 PPL. 421.45 Recons. 0.00327 +2024-03-17 03:29:26,271 INFO Train. Iter 170400 : Commit. 0.49692 PPL. 419.83 Recons. 0.00376 +2024-03-17 03:29:47,872 INFO Train. Iter 170600 : Commit. 0.48494 PPL. 420.57 Recons. 0.00338 +2024-03-17 03:30:09,698 INFO Train. Iter 170800 : Commit. 0.49181 PPL. 419.76 Recons. 0.00368 +2024-03-17 03:30:30,391 INFO Train. Iter 171000 : Commit. 0.49712 PPL. 420.31 Recons. 0.00349 +2024-03-17 03:30:52,100 INFO Train. Iter 171200 : Commit. 0.49605 PPL. 420.68 Recons. 0.00369 +2024-03-17 03:31:13,558 INFO Train. Iter 171400 : Commit. 0.49003 PPL. 420.39 Recons. 0.00349 +2024-03-17 03:31:35,012 INFO Train. Iter 171600 : Commit. 0.47514 PPL. 422.23 Recons. 0.00323 +2024-03-17 03:31:57,017 INFO Train. Iter 171800 : Commit. 0.48770 PPL. 420.95 Recons. 0.00344 +2024-03-17 03:32:18,269 INFO Train. Iter 172000 : Commit. 0.49599 PPL. 419.56 Recons. 0.00386 +2024-03-17 03:32:39,438 INFO Train. Iter 172200 : Commit. 0.48193 PPL. 421.55 Recons. 0.00310 +2024-03-17 03:33:00,922 INFO Train. Iter 172400 : Commit. 0.47795 PPL. 421.77 Recons. 0.00311 +2024-03-17 03:33:22,691 INFO Train. Iter 172600 : Commit. 0.49118 PPL. 420.88 Recons. 0.00352 +2024-03-17 03:33:44,206 INFO Train. Iter 172800 : Commit. 0.48196 PPL. 420.59 Recons. 0.00325 +2024-03-17 03:34:04,768 INFO Train. Iter 173000 : Commit. 0.47681 PPL. 421.97 Recons. 0.00312 +2024-03-17 03:34:26,468 INFO Train. Iter 173200 : Commit. 0.48433 PPL. 420.19 Recons. 0.00353 +2024-03-17 03:34:48,174 INFO Train. Iter 173400 : Commit. 0.47645 PPL. 421.42 Recons. 0.00307 +2024-03-17 03:35:10,316 INFO Train. Iter 173600 : Commit. 0.48185 PPL. 419.90 Recons. 0.00345 +2024-03-17 03:35:31,849 INFO Train. Iter 173800 : Commit. 0.47169 PPL. 421.99 Recons. 0.00296 +2024-03-17 03:35:54,069 INFO Train. Iter 174000 : Commit. 0.49467 PPL. 418.78 Recons. 0.00387 +2024-03-17 03:36:16,043 INFO Train. Iter 174200 : Commit. 0.49383 PPL. 421.91 Recons. 0.00330 +2024-03-17 03:36:38,406 INFO Train. Iter 174400 : Commit. 0.49968 PPL. 419.14 Recons. 0.00394 +2024-03-17 03:37:00,735 INFO Train. Iter 174600 : Commit. 0.49173 PPL. 420.69 Recons. 0.00353 +2024-03-17 03:37:22,886 INFO Train. Iter 174800 : Commit. 0.47636 PPL. 420.65 Recons. 0.00329 +2024-03-17 03:37:43,673 INFO Train. Iter 175000 : Commit. 0.47603 PPL. 421.10 Recons. 0.00316 +2024-03-17 03:38:05,761 INFO Train. Iter 175200 : Commit. 0.49344 PPL. 420.42 Recons. 0.00364 +2024-03-17 03:38:27,740 INFO Train. Iter 175400 : Commit. 0.49031 PPL. 419.96 Recons. 0.00384 +2024-03-17 03:38:49,629 INFO Train. Iter 175600 : Commit. 0.49325 PPL. 421.28 Recons. 0.00350 +2024-03-17 03:39:11,741 INFO Train. Iter 175800 : Commit. 0.48094 PPL. 418.97 Recons. 0.00353 +2024-03-17 03:39:34,220 INFO Train. Iter 176000 : Commit. 0.48262 PPL. 421.01 Recons. 0.00314 +2024-03-17 03:39:56,311 INFO Train. Iter 176200 : Commit. 0.47615 PPL. 421.48 Recons. 0.00318 +2024-03-17 03:40:18,343 INFO Train. Iter 176400 : Commit. 0.48896 PPL. 420.44 Recons. 0.00356 +2024-03-17 03:40:40,527 INFO Train. Iter 176600 : Commit. 0.48872 PPL. 419.99 Recons. 0.00345 +2024-03-17 03:41:00,970 INFO Train. Iter 176800 : Commit. 0.48942 PPL. 420.34 Recons. 0.00344 +2024-03-17 03:41:23,200 INFO Train. Iter 177000 : Commit. 0.49862 PPL. 418.75 Recons. 0.00381 +2024-03-17 03:41:45,744 INFO Train. Iter 177200 : Commit. 0.47190 PPL. 421.41 Recons. 0.00303 +2024-03-17 03:42:07,674 INFO Train. Iter 177400 : Commit. 0.49656 PPL. 417.84 Recons. 0.00375 +2024-03-17 03:42:29,847 INFO Train. Iter 177600 : Commit. 0.47149 PPL. 422.56 Recons. 0.00285 +2024-03-17 03:42:51,832 INFO Train. Iter 177800 : Commit. 0.48295 PPL. 419.92 Recons. 0.00339 +2024-03-17 03:43:13,765 INFO Train. Iter 178000 : Commit. 0.47911 PPL. 420.35 Recons. 0.00326 +2024-03-17 03:43:36,302 INFO Train. Iter 178200 : Commit. 0.47321 PPL. 420.70 Recons. 0.00311 +2024-03-17 03:43:58,678 INFO Train. Iter 178400 : Commit. 0.48711 PPL. 418.47 Recons. 0.00371 +2024-03-17 03:44:21,083 INFO Train. Iter 178600 : Commit. 0.50759 PPL. 418.01 Recons. 0.00397 +2024-03-17 03:44:42,276 INFO Train. Iter 178800 : Commit. 0.47552 PPL. 420.67 Recons. 0.00312 +2024-03-17 03:45:04,377 INFO Train. Iter 179000 : Commit. 0.47988 PPL. 420.43 Recons. 0.00337 +2024-03-17 03:45:26,284 INFO Train. Iter 179200 : Commit. 0.49529 PPL. 419.07 Recons. 0.00382 +2024-03-17 03:45:48,153 INFO Train. Iter 179400 : Commit. 0.48958 PPL. 419.33 Recons. 0.00354 +2024-03-17 03:46:10,418 INFO Train. Iter 179600 : Commit. 0.48427 PPL. 419.63 Recons. 0.00347 +2024-03-17 03:46:32,812 INFO Train. Iter 179800 : Commit. 0.52187 PPL. 416.30 Recons. 0.00453 +2024-03-17 03:46:54,848 INFO Train. Iter 180000 : Commit. 0.48735 PPL. 419.36 Recons. 0.00352 +2024-03-17 03:47:16,584 INFO Train. Iter 180200 : Commit. 0.46743 PPL. 421.97 Recons. 0.00286 +2024-03-17 03:47:38,923 INFO Train. Iter 180400 : Commit. 0.48037 PPL. 419.28 Recons. 0.00354 +2024-03-17 03:48:01,550 INFO Train. Iter 180600 : Commit. 0.50428 PPL. 418.54 Recons. 0.00385 +2024-03-17 03:48:22,700 INFO Train. Iter 180800 : Commit. 0.47898 PPL. 419.32 Recons. 0.00338 +2024-03-17 03:48:45,421 INFO Train. Iter 181000 : Commit. 0.49868 PPL. 418.90 Recons. 0.00378 +2024-03-17 03:49:07,681 INFO Train. Iter 181200 : Commit. 0.48890 PPL. 419.82 Recons. 0.00360 +2024-03-17 03:49:29,771 INFO Train. Iter 181400 : Commit. 0.46932 PPL. 421.00 Recons. 0.00287 +2024-03-17 03:49:51,988 INFO Train. Iter 181600 : Commit. 0.48361 PPL. 420.07 Recons. 0.00349 +2024-03-17 03:50:14,415 INFO Train. Iter 181800 : Commit. 0.49070 PPL. 420.63 Recons. 0.00334 +2024-03-17 03:50:36,967 INFO Train. Iter 182000 : Commit. 0.48106 PPL. 419.37 Recons. 0.00344 +2024-03-17 03:50:59,088 INFO Train. Iter 182200 : Commit. 0.48497 PPL. 419.01 Recons. 0.00344 +2024-03-17 03:51:20,862 INFO Train. Iter 182400 : Commit. 0.45393 PPL. 420.78 Recons. 0.00273 +2024-03-17 03:51:41,675 INFO Train. Iter 182600 : Commit. 0.47744 PPL. 420.35 Recons. 0.00318 +2024-03-17 03:52:04,656 INFO Train. Iter 182800 : Commit. 0.48644 PPL. 417.71 Recons. 0.00375 +2024-03-17 03:52:27,207 INFO Train. Iter 183000 : Commit. 0.48546 PPL. 420.30 Recons. 0.00321 +2024-03-17 03:52:49,144 INFO Train. Iter 183200 : Commit. 0.47605 PPL. 419.85 Recons. 0.00305 +2024-03-17 03:53:10,914 INFO Train. Iter 183400 : Commit. 0.48808 PPL. 417.93 Recons. 0.00353 +2024-03-17 03:53:32,792 INFO Train. Iter 183600 : Commit. 0.48154 PPL. 418.84 Recons. 0.00334 +2024-03-17 03:53:54,658 INFO Train. Iter 183800 : Commit. 0.49816 PPL. 418.55 Recons. 0.00378 +2024-03-17 03:54:16,918 INFO Train. Iter 184000 : Commit. 0.49333 PPL. 418.00 Recons. 0.00377 +2024-03-17 03:54:39,307 INFO Train. Iter 184200 : Commit. 0.48861 PPL. 420.32 Recons. 0.00326 +2024-03-17 03:55:01,205 INFO Train. Iter 184400 : Commit. 0.48887 PPL. 419.33 Recons. 0.00376 +2024-03-17 03:55:21,939 INFO Train. Iter 184600 : Commit. 0.47499 PPL. 420.92 Recons. 0.00296 +2024-03-17 03:55:44,368 INFO Train. Iter 184800 : Commit. 0.48236 PPL. 418.92 Recons. 0.00340 +2024-03-17 03:56:06,887 INFO Train. Iter 185000 : Commit. 0.48565 PPL. 417.68 Recons. 0.00372 +2024-03-17 03:56:28,806 INFO Train. Iter 185200 : Commit. 0.47883 PPL. 419.70 Recons. 0.00325 +2024-03-17 03:56:50,243 INFO Train. Iter 185400 : Commit. 0.48282 PPL. 420.00 Recons. 0.00321 +2024-03-17 03:57:12,374 INFO Train. Iter 185600 : Commit. 0.47231 PPL. 420.02 Recons. 0.00324 +2024-03-17 03:57:34,242 INFO Train. Iter 185800 : Commit. 0.49260 PPL. 418.73 Recons. 0.00369 +2024-03-17 03:57:55,266 INFO Train. Iter 186000 : Commit. 0.49122 PPL. 418.37 Recons. 0.00372 +2024-03-17 03:58:17,095 INFO Train. Iter 186200 : Commit. 0.49494 PPL. 418.23 Recons. 0.00367 +2024-03-17 03:58:39,856 INFO Train. Iter 186400 : Commit. 0.46340 PPL. 420.60 Recons. 0.00286 +2024-03-17 03:59:00,908 INFO Train. Iter 186600 : Commit. 0.48118 PPL. 418.90 Recons. 0.00335 +2024-03-17 03:59:23,064 INFO Train. Iter 186800 : Commit. 0.48597 PPL. 419.60 Recons. 0.00343 +2024-03-17 03:59:44,216 INFO Train. Iter 187000 : Commit. 0.48021 PPL. 419.38 Recons. 0.00340 +2024-03-17 04:00:06,177 INFO Train. Iter 187200 : Commit. 0.47122 PPL. 420.34 Recons. 0.00299 +2024-03-17 04:00:28,528 INFO Train. Iter 187400 : Commit. 0.49031 PPL. 417.47 Recons. 0.00383 +2024-03-17 04:00:50,203 INFO Train. Iter 187600 : Commit. 0.47070 PPL. 419.69 Recons. 0.00318 +2024-03-17 04:01:11,595 INFO Train. Iter 187800 : Commit. 0.48500 PPL. 418.12 Recons. 0.00328 +2024-03-17 04:01:33,507 INFO Train. Iter 188000 : Commit. 0.48507 PPL. 419.10 Recons. 0.00337 +2024-03-17 04:01:55,044 INFO Train. Iter 188200 : Commit. 0.47695 PPL. 419.57 Recons. 0.00314 +2024-03-17 04:02:15,875 INFO Train. Iter 188400 : Commit. 0.48556 PPL. 418.16 Recons. 0.00369 +2024-03-17 04:02:36,820 INFO Train. Iter 188600 : Commit. 0.49245 PPL. 417.82 Recons. 0.00357 +2024-03-17 04:02:58,437 INFO Train. Iter 188800 : Commit. 0.48325 PPL. 418.38 Recons. 0.00343 +2024-03-17 04:03:19,798 INFO Train. Iter 189000 : Commit. 0.47054 PPL. 419.97 Recons. 0.00294 +2024-03-17 04:03:41,368 INFO Train. Iter 189200 : Commit. 0.49789 PPL. 417.88 Recons. 0.00369 +2024-03-17 04:04:03,085 INFO Train. Iter 189400 : Commit. 0.46972 PPL. 419.84 Recons. 0.00301 +2024-03-17 04:04:23,931 INFO Train. Iter 189600 : Commit. 0.47683 PPL. 420.41 Recons. 0.00309 +2024-03-17 04:04:45,711 INFO Train. Iter 189800 : Commit. 0.47861 PPL. 419.04 Recons. 0.00324 +2024-03-17 04:05:06,832 INFO Train. Iter 190000 : Commit. 0.47325 PPL. 419.06 Recons. 0.00312 +2024-03-17 04:05:28,278 INFO Train. Iter 190200 : Commit. 0.47108 PPL. 419.48 Recons. 0.00307 +2024-03-17 04:05:49,051 INFO Train. Iter 190400 : Commit. 0.49679 PPL. 416.92 Recons. 0.00395 +2024-03-17 04:06:10,503 INFO Train. Iter 190600 : Commit. 0.46260 PPL. 419.46 Recons. 0.00276 +2024-03-17 04:06:32,309 INFO Train. Iter 190800 : Commit. 0.48659 PPL. 416.97 Recons. 0.00355 +2024-03-17 04:06:54,022 INFO Train. Iter 191000 : Commit. 0.46899 PPL. 420.80 Recons. 0.00283 +2024-03-17 04:07:16,135 INFO Train. Iter 191200 : Commit. 0.48659 PPL. 417.26 Recons. 0.00394 +2024-03-17 04:07:38,129 INFO Train. Iter 191400 : Commit. 0.48850 PPL. 420.13 Recons. 0.00320 +2024-03-17 04:08:00,580 INFO Train. Iter 191600 : Commit. 0.46483 PPL. 419.53 Recons. 0.00293 +2024-03-17 04:08:22,039 INFO Train. Iter 191800 : Commit. 0.47424 PPL. 418.49 Recons. 0.00347 +2024-03-17 04:08:43,938 INFO Train. Iter 192000 : Commit. 0.47804 PPL. 419.28 Recons. 0.00317 +2024-03-17 04:09:05,920 INFO Train. Iter 192200 : Commit. 0.48567 PPL. 418.12 Recons. 0.00346 +2024-03-17 04:09:26,572 INFO Train. Iter 192400 : Commit. 0.47118 PPL. 418.48 Recons. 0.00311 +2024-03-17 04:09:48,333 INFO Train. Iter 192600 : Commit. 0.47081 PPL. 419.97 Recons. 0.00300 +2024-03-17 04:10:09,989 INFO Train. Iter 192800 : Commit. 0.51150 PPL. 416.18 Recons. 0.00431 +2024-03-17 04:10:32,411 INFO Train. Iter 193000 : Commit. 0.47906 PPL. 418.53 Recons. 0.00325 +2024-03-17 04:10:54,697 INFO Train. Iter 193200 : Commit. 0.48187 PPL. 417.43 Recons. 0.00348 +2024-03-17 04:11:16,087 INFO Train. Iter 193400 : Commit. 0.45862 PPL. 420.78 Recons. 0.00275 +2024-03-17 04:11:37,424 INFO Train. Iter 193600 : Commit. 0.47323 PPL. 418.17 Recons. 0.00334 +2024-03-17 04:11:58,347 INFO Train. Iter 193800 : Commit. 0.46278 PPL. 419.14 Recons. 0.00297 +2024-03-17 04:12:19,303 INFO Train. Iter 194000 : Commit. 0.48304 PPL. 417.05 Recons. 0.00372 +2024-03-17 04:12:39,540 INFO Train. Iter 194200 : Commit. 0.48111 PPL. 419.79 Recons. 0.00318 +2024-03-17 04:13:01,639 INFO Train. Iter 194400 : Commit. 0.48272 PPL. 417.73 Recons. 0.00341 +2024-03-17 04:13:23,184 INFO Train. Iter 194600 : Commit. 0.49004 PPL. 417.77 Recons. 0.00351 +2024-03-17 04:13:44,408 INFO Train. Iter 194800 : Commit. 0.46897 PPL. 418.93 Recons. 0.00292 +2024-03-17 04:14:06,801 INFO Train. Iter 195000 : Commit. 0.49026 PPL. 416.72 Recons. 0.00387 +2024-03-17 04:14:29,290 INFO Train. Iter 195200 : Commit. 0.48367 PPL. 418.48 Recons. 0.00311 +2024-03-17 04:14:51,478 INFO Train. Iter 195400 : Commit. 0.47886 PPL. 417.71 Recons. 0.00347 +2024-03-17 04:15:13,626 INFO Train. Iter 195600 : Commit. 0.49353 PPL. 417.86 Recons. 0.00339 +2024-03-17 04:15:35,521 INFO Train. Iter 195800 : Commit. 0.47729 PPL. 417.78 Recons. 0.00327 +2024-03-17 04:15:57,363 INFO Train. Iter 196000 : Commit. 0.46986 PPL. 419.19 Recons. 0.00290 +2024-03-17 04:16:18,654 INFO Train. Iter 196200 : Commit. 0.49155 PPL. 416.38 Recons. 0.00385 +2024-03-17 04:16:40,443 INFO Train. Iter 196400 : Commit. 0.47725 PPL. 418.52 Recons. 0.00301 +2024-03-17 04:17:02,040 INFO Train. Iter 196600 : Commit. 0.48988 PPL. 417.36 Recons. 0.00360 +2024-03-17 04:17:23,603 INFO Train. Iter 196800 : Commit. 0.49187 PPL. 416.70 Recons. 0.00370 +2024-03-17 04:17:45,088 INFO Train. Iter 197000 : Commit. 0.47840 PPL. 419.46 Recons. 0.00294 +2024-03-17 04:18:06,868 INFO Train. Iter 197200 : Commit. 0.47739 PPL. 417.83 Recons. 0.00322 +2024-03-17 04:18:28,882 INFO Train. Iter 197400 : Commit. 0.48556 PPL. 417.22 Recons. 0.00365 +2024-03-17 04:18:51,116 INFO Train. Iter 197600 : Commit. 0.48768 PPL. 417.29 Recons. 0.00340 +2024-03-17 04:19:13,073 INFO Train. Iter 197800 : Commit. 0.48616 PPL. 417.21 Recons. 0.00339 +2024-03-17 04:19:34,967 INFO Train. Iter 198000 : Commit. 0.47307 PPL. 418.36 Recons. 0.00303 +2024-03-17 04:19:55,425 INFO Train. Iter 198200 : Commit. 0.48691 PPL. 416.03 Recons. 0.00333 +2024-03-17 04:20:17,584 INFO Train. Iter 198400 : Commit. 0.46056 PPL. 419.01 Recons. 0.00276 +2024-03-17 04:20:39,397 INFO Train. Iter 198600 : Commit. 0.48742 PPL. 416.54 Recons. 0.00346 +2024-03-17 04:21:00,563 INFO Train. Iter 198800 : Commit. 0.47130 PPL. 417.32 Recons. 0.00317 +2024-03-17 04:21:22,181 INFO Train. Iter 199000 : Commit. 0.47692 PPL. 417.74 Recons. 0.00304 +2024-03-17 04:21:43,480 INFO Train. Iter 199200 : Commit. 0.47795 PPL. 416.84 Recons. 0.00322 +2024-03-17 04:22:05,307 INFO Train. Iter 199400 : Commit. 0.49050 PPL. 415.63 Recons. 0.00350 +2024-03-17 04:22:26,732 INFO Train. Iter 199600 : Commit. 0.48718 PPL. 417.01 Recons. 0.00333 +2024-03-17 04:22:48,177 INFO Train. Iter 199800 : Commit. 0.47416 PPL. 417.35 Recons. 0.00307 +2024-03-17 04:23:09,247 INFO Train. Iter 200000 : Commit. 0.48939 PPL. 415.79 Recons. 0.00354 +2024-03-17 04:23:31,360 INFO Train. Iter 200200 : Commit. 0.47142 PPL. 418.69 Recons. 0.00255 +2024-03-17 04:23:53,440 INFO Train. Iter 200400 : Commit. 0.45910 PPL. 418.80 Recons. 0.00231 +2024-03-17 04:24:15,530 INFO Train. Iter 200600 : Commit. 0.46159 PPL. 418.35 Recons. 0.00240 +2024-03-17 04:24:37,838 INFO Train. Iter 200800 : Commit. 0.45406 PPL. 418.48 Recons. 0.00227 +2024-03-17 04:25:00,504 INFO Train. Iter 201000 : Commit. 0.45289 PPL. 418.11 Recons. 0.00227 +2024-03-17 04:25:22,856 INFO Train. Iter 201200 : Commit. 0.45148 PPL. 417.53 Recons. 0.00229 +2024-03-17 04:25:45,121 INFO Train. Iter 201400 : Commit. 0.44902 PPL. 418.08 Recons. 0.00223 +2024-03-17 04:26:06,961 INFO Train. Iter 201600 : Commit. 0.45611 PPL. 418.11 Recons. 0.00232 +2024-03-17 04:26:28,818 INFO Train. Iter 201800 : Commit. 0.46599 PPL. 416.67 Recons. 0.00255 +2024-03-17 04:26:49,287 INFO Train. Iter 202000 : Commit. 0.45126 PPL. 417.59 Recons. 0.00228 +2024-03-17 04:27:11,125 INFO Train. Iter 202200 : Commit. 0.45272 PPL. 416.78 Recons. 0.00234 +2024-03-17 04:27:32,871 INFO Train. Iter 202400 : Commit. 0.45170 PPL. 417.90 Recons. 0.00228 +2024-03-17 04:27:54,818 INFO Train. Iter 202600 : Commit. 0.45595 PPL. 416.65 Recons. 0.00242 +2024-03-17 04:28:17,421 INFO Train. Iter 202800 : Commit. 0.45149 PPL. 417.46 Recons. 0.00231 +2024-03-17 04:28:39,763 INFO Train. Iter 203000 : Commit. 0.45210 PPL. 417.94 Recons. 0.00224 +2024-03-17 04:29:01,654 INFO Train. Iter 203200 : Commit. 0.45147 PPL. 416.21 Recons. 0.00233 +2024-03-17 04:29:23,985 INFO Train. Iter 203400 : Commit. 0.44599 PPL. 417.15 Recons. 0.00221 +2024-03-17 04:29:46,123 INFO Train. Iter 203600 : Commit. 0.44481 PPL. 417.60 Recons. 0.00218 +2024-03-17 04:30:08,550 INFO Train. Iter 203800 : Commit. 0.45074 PPL. 416.22 Recons. 0.00232 +2024-03-17 04:30:29,795 INFO Train. Iter 204000 : Commit. 0.44593 PPL. 417.05 Recons. 0.00220 +2024-03-17 04:30:51,979 INFO Train. Iter 204200 : Commit. 0.45621 PPL. 415.86 Recons. 0.00239 +2024-03-17 04:31:14,420 INFO Train. Iter 204400 : Commit. 0.43894 PPL. 416.23 Recons. 0.00215 +2024-03-17 04:31:37,372 INFO Train. Iter 204600 : Commit. 0.45964 PPL. 416.46 Recons. 0.00246 +2024-03-17 04:31:59,952 INFO Train. Iter 204800 : Commit. 0.44419 PPL. 416.96 Recons. 0.00219 +2024-03-17 04:32:22,460 INFO Train. Iter 205000 : Commit. 0.44378 PPL. 416.33 Recons. 0.00222 +2024-03-17 04:32:43,809 INFO Train. Iter 205200 : Commit. 0.44756 PPL. 416.75 Recons. 0.00225 +2024-03-17 04:33:05,348 INFO Train. Iter 205400 : Commit. 0.44605 PPL. 416.11 Recons. 0.00228 +2024-03-17 04:33:27,472 INFO Train. Iter 205600 : Commit. 0.44841 PPL. 416.93 Recons. 0.00223 +2024-03-17 04:33:47,902 INFO Train. Iter 205800 : Commit. 0.44872 PPL. 416.83 Recons. 0.00226 +2024-03-17 04:34:09,366 INFO Train. Iter 206000 : Commit. 0.44700 PPL. 416.11 Recons. 0.00226 +2024-03-17 04:34:30,648 INFO Train. Iter 206200 : Commit. 0.45848 PPL. 415.69 Recons. 0.00245 +2024-03-17 04:34:52,302 INFO Train. Iter 206400 : Commit. 0.45103 PPL. 416.58 Recons. 0.00228 +2024-03-17 04:35:13,791 INFO Train. Iter 206600 : Commit. 0.44137 PPL. 416.81 Recons. 0.00215 +2024-03-17 04:35:35,886 INFO Train. Iter 206800 : Commit. 0.44544 PPL. 417.07 Recons. 0.00221 +2024-03-17 04:35:58,094 INFO Train. Iter 207000 : Commit. 0.44353 PPL. 417.13 Recons. 0.00216 +2024-03-17 04:36:19,608 INFO Train. Iter 207200 : Commit. 0.44369 PPL. 416.29 Recons. 0.00223 +2024-03-17 04:36:41,170 INFO Train. Iter 207400 : Commit. 0.43953 PPL. 417.08 Recons. 0.00209 +2024-03-17 04:37:02,610 INFO Train. Iter 207600 : Commit. 0.44682 PPL. 416.65 Recons. 0.00227 +2024-03-17 04:37:22,683 INFO Train. Iter 207800 : Commit. 0.44421 PPL. 416.13 Recons. 0.00224 +2024-03-17 04:37:43,761 INFO Train. Iter 208000 : Commit. 0.45640 PPL. 415.84 Recons. 0.00241 +2024-03-17 04:38:05,148 INFO Train. Iter 208200 : Commit. 0.45429 PPL. 416.31 Recons. 0.00237 +2024-03-17 04:38:26,133 INFO Train. Iter 208400 : Commit. 0.44600 PPL. 416.21 Recons. 0.00225 +2024-03-17 04:38:47,306 INFO Train. Iter 208600 : Commit. 0.45123 PPL. 416.64 Recons. 0.00231 +2024-03-17 04:39:08,867 INFO Train. Iter 208800 : Commit. 0.44435 PPL. 416.62 Recons. 0.00219 +2024-03-17 04:39:29,820 INFO Train. Iter 209000 : Commit. 0.44757 PPL. 416.60 Recons. 0.00223 +2024-03-17 04:39:50,657 INFO Train. Iter 209200 : Commit. 0.45287 PPL. 416.01 Recons. 0.00236 +2024-03-17 04:40:11,628 INFO Train. Iter 209400 : Commit. 0.44369 PPL. 416.35 Recons. 0.00221 +2024-03-17 04:40:33,615 INFO Train. Iter 209600 : Commit. 0.44509 PPL. 416.14 Recons. 0.00223 +2024-03-17 04:40:53,738 INFO Train. Iter 209800 : Commit. 0.44435 PPL. 416.57 Recons. 0.00218 +2024-03-17 04:41:15,555 INFO Train. Iter 210000 : Commit. 0.44801 PPL. 415.80 Recons. 0.00229 +2024-03-17 04:41:36,743 INFO Train. Iter 210200 : Commit. 0.44137 PPL. 416.24 Recons. 0.00217 +2024-03-17 04:41:58,547 INFO Train. Iter 210400 : Commit. 0.44366 PPL. 416.81 Recons. 0.00218 +2024-03-17 04:42:20,253 INFO Train. Iter 210600 : Commit. 0.44351 PPL. 416.58 Recons. 0.00216 +2024-03-17 04:42:42,265 INFO Train. Iter 210800 : Commit. 0.44729 PPL. 415.98 Recons. 0.00227 +2024-03-17 04:43:03,455 INFO Train. Iter 211000 : Commit. 0.43990 PPL. 416.09 Recons. 0.00212 +2024-03-17 04:43:24,578 INFO Train. Iter 211200 : Commit. 0.44891 PPL. 416.45 Recons. 0.00226 +2024-03-17 04:43:45,667 INFO Train. Iter 211400 : Commit. 0.44803 PPL. 415.51 Recons. 0.00227 +2024-03-17 04:44:05,773 INFO Train. Iter 211600 : Commit. 0.45450 PPL. 415.97 Recons. 0.00236 +2024-03-17 04:44:27,068 INFO Train. Iter 211800 : Commit. 0.44216 PPL. 416.00 Recons. 0.00216 +2024-03-17 04:44:48,836 INFO Train. Iter 212000 : Commit. 0.44086 PPL. 416.74 Recons. 0.00213 +2024-03-17 04:45:10,168 INFO Train. Iter 212200 : Commit. 0.45059 PPL. 415.28 Recons. 0.00233 +2024-03-17 04:45:31,752 INFO Train. Iter 212400 : Commit. 0.43646 PPL. 415.68 Recons. 0.00211 +2024-03-17 04:45:53,436 INFO Train. Iter 212600 : Commit. 0.44795 PPL. 415.77 Recons. 0.00225 +2024-03-17 04:46:14,818 INFO Train. Iter 212800 : Commit. 0.44318 PPL. 415.70 Recons. 0.00222 +2024-03-17 04:46:36,615 INFO Train. Iter 213000 : Commit. 0.44448 PPL. 415.56 Recons. 0.00220 +2024-03-17 04:46:58,165 INFO Train. Iter 213200 : Commit. 0.44162 PPL. 416.27 Recons. 0.00214 +2024-03-17 04:47:19,406 INFO Train. Iter 213400 : Commit. 0.43482 PPL. 416.12 Recons. 0.00207 +2024-03-17 04:47:39,527 INFO Train. Iter 213600 : Commit. 0.44751 PPL. 415.75 Recons. 0.00226 +2024-03-17 04:48:00,754 INFO Train. Iter 213800 : Commit. 0.44520 PPL. 415.97 Recons. 0.00222 +2024-03-17 04:48:21,788 INFO Train. Iter 214000 : Commit. 0.44717 PPL. 416.03 Recons. 0.00225 +2024-03-17 04:48:42,997 INFO Train. Iter 214200 : Commit. 0.45349 PPL. 415.63 Recons. 0.00238 +2024-03-17 04:49:04,248 INFO Train. Iter 214400 : Commit. 0.44674 PPL. 416.41 Recons. 0.00224 +2024-03-17 04:49:25,933 INFO Train. Iter 214600 : Commit. 0.44114 PPL. 415.94 Recons. 0.00217 +2024-03-17 04:49:47,799 INFO Train. Iter 214800 : Commit. 0.43694 PPL. 415.68 Recons. 0.00213 +2024-03-17 04:50:09,750 INFO Train. Iter 215000 : Commit. 0.44005 PPL. 415.48 Recons. 0.00218 +2024-03-17 04:50:30,763 INFO Train. Iter 215200 : Commit. 0.45186 PPL. 415.68 Recons. 0.00233 +2024-03-17 04:50:51,639 INFO Train. Iter 215400 : Commit. 0.44237 PPL. 415.73 Recons. 0.00222 +2024-03-17 04:51:11,473 INFO Train. Iter 215600 : Commit. 0.43682 PPL. 415.44 Recons. 0.00215 +2024-03-17 04:51:32,803 INFO Train. Iter 215800 : Commit. 0.44657 PPL. 416.00 Recons. 0.00223 +2024-03-17 04:51:54,113 INFO Train. Iter 216000 : Commit. 0.44608 PPL. 415.80 Recons. 0.00223 +2024-03-17 04:52:15,548 INFO Train. Iter 216200 : Commit. 0.44251 PPL. 415.36 Recons. 0.00222 +2024-03-17 04:52:36,676 INFO Train. Iter 216400 : Commit. 0.44667 PPL. 414.21 Recons. 0.00233 +2024-03-17 04:52:57,374 INFO Train. Iter 216600 : Commit. 0.44128 PPL. 416.09 Recons. 0.00215 +2024-03-17 04:53:18,965 INFO Train. Iter 216800 : Commit. 0.44573 PPL. 415.32 Recons. 0.00225 +2024-03-17 04:53:40,247 INFO Train. Iter 217000 : Commit. 0.45796 PPL. 415.31 Recons. 0.00242 +2024-03-17 04:54:01,651 INFO Train. Iter 217200 : Commit. 0.44556 PPL. 415.19 Recons. 0.00225 +2024-03-17 04:54:21,737 INFO Train. Iter 217400 : Commit. 0.43931 PPL. 415.46 Recons. 0.00216 +2024-03-17 04:54:43,416 INFO Train. Iter 217600 : Commit. 0.44045 PPL. 416.04 Recons. 0.00216 +2024-03-17 04:55:05,037 INFO Train. Iter 217800 : Commit. 0.45019 PPL. 416.47 Recons. 0.00226 +2024-03-17 04:55:26,322 INFO Train. Iter 218000 : Commit. 0.44429 PPL. 415.67 Recons. 0.00222 +2024-03-17 04:55:48,113 INFO Train. Iter 218200 : Commit. 0.44959 PPL. 416.19 Recons. 0.00226 +2024-03-17 04:56:09,650 INFO Train. Iter 218400 : Commit. 0.44439 PPL. 416.59 Recons. 0.00219 +2024-03-17 04:56:31,250 INFO Train. Iter 218600 : Commit. 0.45288 PPL. 415.23 Recons. 0.00235 +2024-03-17 04:56:53,165 INFO Train. Iter 218800 : Commit. 0.43618 PPL. 416.19 Recons. 0.00209 +2024-03-17 04:57:14,428 INFO Train. Iter 219000 : Commit. 0.44637 PPL. 415.00 Recons. 0.00228 +2024-03-17 04:57:35,588 INFO Train. Iter 219200 : Commit. 0.44034 PPL. 416.32 Recons. 0.00214 +2024-03-17 04:57:56,054 INFO Train. Iter 219400 : Commit. 0.43843 PPL. 415.63 Recons. 0.00215 +2024-03-17 04:58:17,212 INFO Train. Iter 219600 : Commit. 0.43657 PPL. 415.01 Recons. 0.00214 +2024-03-17 04:58:38,869 INFO Train. Iter 219800 : Commit. 0.43366 PPL. 415.53 Recons. 0.00210 +2024-03-17 04:58:59,941 INFO Train. Iter 220000 : Commit. 0.43453 PPL. 414.73 Recons. 0.00211 +2024-03-17 04:59:21,508 INFO Train. Iter 220200 : Commit. 0.44361 PPL. 416.42 Recons. 0.00220 +2024-03-17 04:59:42,840 INFO Train. Iter 220400 : Commit. 0.43080 PPL. 414.95 Recons. 0.00205 +2024-03-17 05:00:03,992 INFO Train. Iter 220600 : Commit. 0.44112 PPL. 415.07 Recons. 0.00222 +2024-03-17 05:00:24,542 INFO Train. Iter 220800 : Commit. 0.43773 PPL. 416.19 Recons. 0.00211 +2024-03-17 05:00:46,019 INFO Train. Iter 221000 : Commit. 0.43883 PPL. 415.19 Recons. 0.00216 +2024-03-17 05:01:07,502 INFO Train. Iter 221200 : Commit. 0.43368 PPL. 415.05 Recons. 0.00210 +2024-03-17 05:01:28,040 INFO Train. Iter 221400 : Commit. 0.45011 PPL. 415.23 Recons. 0.00232 +2024-03-17 05:01:49,196 INFO Train. Iter 221600 : Commit. 0.44153 PPL. 414.51 Recons. 0.00223 +2024-03-17 05:02:10,223 INFO Train. Iter 221800 : Commit. 0.43941 PPL. 414.27 Recons. 0.00220 +2024-03-17 05:02:31,405 INFO Train. Iter 222000 : Commit. 0.44249 PPL. 415.13 Recons. 0.00221 +2024-03-17 05:02:52,866 INFO Train. Iter 222200 : Commit. 0.44160 PPL. 415.52 Recons. 0.00218 +2024-03-17 05:03:14,396 INFO Train. Iter 222400 : Commit. 0.43129 PPL. 415.27 Recons. 0.00205 +2024-03-17 05:03:35,437 INFO Train. Iter 222600 : Commit. 0.44410 PPL. 414.74 Recons. 0.00223 +2024-03-17 05:03:57,050 INFO Train. Iter 222800 : Commit. 0.43438 PPL. 415.14 Recons. 0.00209 +2024-03-17 05:04:18,321 INFO Train. Iter 223000 : Commit. 0.44599 PPL. 415.41 Recons. 0.00224 +2024-03-17 05:04:38,722 INFO Train. Iter 223200 : Commit. 0.44492 PPL. 415.12 Recons. 0.00225 +2024-03-17 05:05:00,086 INFO Train. Iter 223400 : Commit. 0.43585 PPL. 415.31 Recons. 0.00210 +2024-03-17 05:05:21,750 INFO Train. Iter 223600 : Commit. 0.43813 PPL. 415.32 Recons. 0.00215 +2024-03-17 05:05:43,300 INFO Train. Iter 223800 : Commit. 0.43596 PPL. 414.16 Recons. 0.00216 +2024-03-17 05:06:04,670 INFO Train. Iter 224000 : Commit. 0.43074 PPL. 415.20 Recons. 0.00201 +2024-03-17 05:06:25,653 INFO Train. Iter 224200 : Commit. 0.43456 PPL. 414.78 Recons. 0.00212 +2024-03-17 05:06:46,909 INFO Train. Iter 224400 : Commit. 0.43868 PPL. 414.47 Recons. 0.00219 +2024-03-17 05:07:08,202 INFO Train. Iter 224600 : Commit. 0.43635 PPL. 414.58 Recons. 0.00212 +2024-03-17 05:07:29,617 INFO Train. Iter 224800 : Commit. 0.43514 PPL. 414.95 Recons. 0.00211 +2024-03-17 05:07:51,914 INFO Train. Iter 225000 : Commit. 0.43402 PPL. 414.34 Recons. 0.00211 +2024-03-17 05:08:13,194 INFO Train. Iter 225200 : Commit. 0.44214 PPL. 414.36 Recons. 0.00223 +2024-03-17 05:08:35,203 INFO Train. Iter 225400 : Commit. 0.43817 PPL. 414.78 Recons. 0.00216 +2024-03-17 05:08:57,189 INFO Train. Iter 225600 : Commit. 0.43702 PPL. 414.62 Recons. 0.00215 +2024-03-17 05:09:18,909 INFO Train. Iter 225800 : Commit. 0.43542 PPL. 415.56 Recons. 0.00208 +2024-03-17 05:09:40,970 INFO Train. Iter 226000 : Commit. 0.43779 PPL. 415.00 Recons. 0.00213 +2024-03-17 05:10:01,852 INFO Train. Iter 226200 : Commit. 0.43736 PPL. 414.63 Recons. 0.00214 +2024-03-17 05:10:23,057 INFO Train. Iter 226400 : Commit. 0.43772 PPL. 413.59 Recons. 0.00218 +2024-03-17 05:10:44,475 INFO Train. Iter 226600 : Commit. 0.44555 PPL. 413.97 Recons. 0.00224 +2024-03-17 05:11:05,672 INFO Train. Iter 226800 : Commit. 0.45253 PPL. 414.48 Recons. 0.00235 +2024-03-17 05:11:27,210 INFO Train. Iter 227000 : Commit. 0.44136 PPL. 416.01 Recons. 0.00215 +2024-03-17 05:11:47,299 INFO Train. Iter 227200 : Commit. 0.44111 PPL. 414.43 Recons. 0.00218 +2024-03-17 05:12:08,589 INFO Train. Iter 227400 : Commit. 0.43793 PPL. 413.45 Recons. 0.00219 +2024-03-17 05:12:30,113 INFO Train. Iter 227600 : Commit. 0.44343 PPL. 414.55 Recons. 0.00222 +2024-03-17 05:12:51,629 INFO Train. Iter 227800 : Commit. 0.43203 PPL. 414.36 Recons. 0.00208 +2024-03-17 05:13:12,563 INFO Train. Iter 228000 : Commit. 0.43642 PPL. 414.88 Recons. 0.00210 +2024-03-17 05:13:33,725 INFO Train. Iter 228200 : Commit. 0.43194 PPL. 413.71 Recons. 0.00212 +2024-03-17 05:13:55,222 INFO Train. Iter 228400 : Commit. 0.44659 PPL. 414.04 Recons. 0.00230 +2024-03-17 05:14:16,478 INFO Train. Iter 228600 : Commit. 0.43947 PPL. 414.02 Recons. 0.00218 +2024-03-17 05:14:37,273 INFO Train. Iter 228800 : Commit. 0.43331 PPL. 413.16 Recons. 0.00214 +2024-03-17 05:14:57,528 INFO Train. Iter 229000 : Commit. 0.44042 PPL. 413.69 Recons. 0.00223 +2024-03-17 05:15:18,865 INFO Train. Iter 229200 : Commit. 0.44426 PPL. 412.92 Recons. 0.00235 +2024-03-17 05:15:39,545 INFO Train. Iter 229400 : Commit. 0.44544 PPL. 415.48 Recons. 0.00220 +2024-03-17 05:16:01,062 INFO Train. Iter 229600 : Commit. 0.43273 PPL. 414.16 Recons. 0.00210 +2024-03-17 05:16:22,649 INFO Train. Iter 229800 : Commit. 0.44377 PPL. 414.30 Recons. 0.00225 +2024-03-17 05:16:44,167 INFO Train. Iter 230000 : Commit. 0.43653 PPL. 414.85 Recons. 0.00212 +2024-03-17 05:17:05,505 INFO Train. Iter 230200 : Commit. 0.43910 PPL. 414.06 Recons. 0.00220 +2024-03-17 05:17:27,272 INFO Train. Iter 230400 : Commit. 0.43939 PPL. 414.10 Recons. 0.00221 +2024-03-17 05:17:48,912 INFO Train. Iter 230600 : Commit. 0.42713 PPL. 414.06 Recons. 0.00204 +2024-03-17 05:18:10,069 INFO Train. Iter 230800 : Commit. 0.42997 PPL. 413.95 Recons. 0.00207 +2024-03-17 05:18:30,006 INFO Train. Iter 231000 : Commit. 0.44010 PPL. 413.76 Recons. 0.00222 +2024-03-17 05:18:50,999 INFO Train. Iter 231200 : Commit. 0.44269 PPL. 414.13 Recons. 0.00222 +2024-03-17 05:19:12,322 INFO Train. Iter 231400 : Commit. 0.43403 PPL. 413.66 Recons. 0.00214 +2024-03-17 05:19:34,522 INFO Train. Iter 231600 : Commit. 0.44075 PPL. 413.53 Recons. 0.00223 +2024-03-17 05:19:55,860 INFO Train. Iter 231800 : Commit. 0.44901 PPL. 414.03 Recons. 0.00233 +2024-03-17 05:20:16,810 INFO Train. Iter 232000 : Commit. 0.43629 PPL. 414.06 Recons. 0.00215 +2024-03-17 05:20:38,069 INFO Train. Iter 232200 : Commit. 0.43653 PPL. 414.06 Recons. 0.00215 +2024-03-17 05:20:59,291 INFO Train. Iter 232400 : Commit. 0.44596 PPL. 413.79 Recons. 0.00232 +2024-03-17 05:21:21,203 INFO Train. Iter 232600 : Commit. 0.43456 PPL. 414.32 Recons. 0.00210 +2024-03-17 05:21:42,435 INFO Train. Iter 232800 : Commit. 0.45018 PPL. 413.48 Recons. 0.00235 +2024-03-17 05:22:03,161 INFO Train. Iter 233000 : Commit. 0.43491 PPL. 414.03 Recons. 0.00214 +2024-03-17 05:22:24,654 INFO Train. Iter 233200 : Commit. 0.43748 PPL. 414.05 Recons. 0.00217 +2024-03-17 05:22:45,875 INFO Train. Iter 233400 : Commit. 0.43472 PPL. 413.62 Recons. 0.00213 +2024-03-17 05:23:07,293 INFO Train. Iter 233600 : Commit. 0.43730 PPL. 414.00 Recons. 0.00217 +2024-03-17 05:23:28,225 INFO Train. Iter 233800 : Commit. 0.43080 PPL. 414.09 Recons. 0.00205 +2024-03-17 05:23:49,778 INFO Train. Iter 234000 : Commit. 0.42974 PPL. 414.03 Recons. 0.00209 +2024-03-17 05:24:10,359 INFO Train. Iter 234200 : Commit. 0.43693 PPL. 414.04 Recons. 0.00215 +2024-03-17 05:24:31,465 INFO Train. Iter 234400 : Commit. 0.43900 PPL. 413.93 Recons. 0.00219 +2024-03-17 05:24:53,166 INFO Train. Iter 234600 : Commit. 0.43565 PPL. 413.50 Recons. 0.00215 +2024-03-17 05:25:13,457 INFO Train. Iter 234800 : Commit. 0.43298 PPL. 414.14 Recons. 0.00210 +2024-03-17 05:25:35,068 INFO Train. Iter 235000 : Commit. 0.43574 PPL. 414.65 Recons. 0.00215 +2024-03-17 05:25:56,502 INFO Train. Iter 235200 : Commit. 0.43744 PPL. 414.20 Recons. 0.00216 +2024-03-17 05:26:18,513 INFO Train. Iter 235400 : Commit. 0.42597 PPL. 414.41 Recons. 0.00199 +2024-03-17 05:26:39,689 INFO Train. Iter 235600 : Commit. 0.43282 PPL. 414.86 Recons. 0.00209 +2024-03-17 05:27:00,929 INFO Train. Iter 235800 : Commit. 0.44144 PPL. 413.59 Recons. 0.00225 +2024-03-17 05:27:22,206 INFO Train. Iter 236000 : Commit. 0.44277 PPL. 413.37 Recons. 0.00228 +2024-03-17 05:27:43,423 INFO Train. Iter 236200 : Commit. 0.43671 PPL. 413.73 Recons. 0.00217 +2024-03-17 05:28:04,668 INFO Train. Iter 236400 : Commit. 0.43385 PPL. 413.46 Recons. 0.00215 +2024-03-17 05:28:26,122 INFO Train. Iter 236600 : Commit. 0.44151 PPL. 414.51 Recons. 0.00221 +2024-03-17 05:28:46,079 INFO Train. Iter 236800 : Commit. 0.42281 PPL. 413.72 Recons. 0.00198 +2024-03-17 05:29:07,685 INFO Train. Iter 237000 : Commit. 0.43197 PPL. 414.14 Recons. 0.00209 +2024-03-17 05:29:29,129 INFO Train. Iter 237200 : Commit. 0.44241 PPL. 414.10 Recons. 0.00222 +2024-03-17 05:29:50,334 INFO Train. Iter 237400 : Commit. 0.43334 PPL. 414.28 Recons. 0.00210 +2024-03-17 05:30:11,904 INFO Train. Iter 237600 : Commit. 0.44150 PPL. 412.95 Recons. 0.00228 +2024-03-17 05:30:32,892 INFO Train. Iter 237800 : Commit. 0.44470 PPL. 413.34 Recons. 0.00227 +2024-03-17 05:30:53,882 INFO Train. Iter 238000 : Commit. 0.43735 PPL. 413.45 Recons. 0.00220 +2024-03-17 05:31:14,930 INFO Train. Iter 238200 : Commit. 0.42736 PPL. 413.73 Recons. 0.00202 +2024-03-17 05:31:36,214 INFO Train. Iter 238400 : Commit. 0.44001 PPL. 413.62 Recons. 0.00220 +2024-03-17 05:31:57,689 INFO Train. Iter 238600 : Commit. 0.43707 PPL. 413.47 Recons. 0.00215 +2024-03-17 05:32:18,532 INFO Train. Iter 238800 : Commit. 0.43918 PPL. 413.41 Recons. 0.00220 +2024-03-17 05:32:40,004 INFO Train. Iter 239000 : Commit. 0.43810 PPL. 413.38 Recons. 0.00218 +2024-03-17 05:33:01,359 INFO Train. Iter 239200 : Commit. 0.43912 PPL. 414.24 Recons. 0.00216 +2024-03-17 05:33:22,920 INFO Train. Iter 239400 : Commit. 0.43443 PPL. 413.22 Recons. 0.00214 +2024-03-17 05:33:44,399 INFO Train. Iter 239600 : Commit. 0.44610 PPL. 413.20 Recons. 0.00230 +2024-03-17 05:34:06,000 INFO Train. Iter 239800 : Commit. 0.43485 PPL. 413.23 Recons. 0.00214 +2024-03-17 05:34:27,822 INFO Train. Iter 240000 : Commit. 0.43162 PPL. 413.79 Recons. 0.00207 +2024-03-17 05:34:49,323 INFO Train. Iter 240200 : Commit. 0.43557 PPL. 414.15 Recons. 0.00209 +2024-03-17 05:35:10,361 INFO Train. Iter 240400 : Commit. 0.43969 PPL. 413.34 Recons. 0.00221 +2024-03-17 05:35:30,245 INFO Train. Iter 240600 : Commit. 0.43232 PPL. 413.25 Recons. 0.00209 +2024-03-17 05:35:51,461 INFO Train. Iter 240800 : Commit. 0.43878 PPL. 413.08 Recons. 0.00221 +2024-03-17 05:36:12,813 INFO Train. Iter 241000 : Commit. 0.43277 PPL. 412.82 Recons. 0.00215 +2024-03-17 05:36:33,711 INFO Train. Iter 241200 : Commit. 0.43339 PPL. 413.41 Recons. 0.00211 +2024-03-17 05:36:54,956 INFO Train. Iter 241400 : Commit. 0.43110 PPL. 413.47 Recons. 0.00209 +2024-03-17 05:37:16,203 INFO Train. Iter 241600 : Commit. 0.43269 PPL. 413.15 Recons. 0.00211 +2024-03-17 05:37:37,595 INFO Train. Iter 241800 : Commit. 0.44750 PPL. 413.82 Recons. 0.00229 +2024-03-17 05:37:59,372 INFO Train. Iter 242000 : Commit. 0.44106 PPL. 413.26 Recons. 0.00220 +2024-03-17 05:38:20,567 INFO Train. Iter 242200 : Commit. 0.43807 PPL. 412.96 Recons. 0.00220 +2024-03-17 05:38:42,579 INFO Train. Iter 242400 : Commit. 0.43940 PPL. 414.35 Recons. 0.00214 +2024-03-17 05:39:03,204 INFO Train. Iter 242600 : Commit. 0.43752 PPL. 413.52 Recons. 0.00216 +2024-03-17 05:39:25,312 INFO Train. Iter 242800 : Commit. 0.44547 PPL. 412.02 Recons. 0.00234 +2024-03-17 05:39:47,220 INFO Train. Iter 243000 : Commit. 0.43517 PPL. 413.11 Recons. 0.00214 +2024-03-17 05:40:09,969 INFO Train. Iter 243200 : Commit. 0.43031 PPL. 413.62 Recons. 0.00208 +2024-03-17 05:40:32,167 INFO Train. Iter 243400 : Commit. 0.43261 PPL. 413.37 Recons. 0.00209 +2024-03-17 05:40:54,362 INFO Train. Iter 243600 : Commit. 0.43180 PPL. 413.49 Recons. 0.00207 +2024-03-17 05:41:16,753 INFO Train. Iter 243800 : Commit. 0.42852 PPL. 413.06 Recons. 0.00206 +2024-03-17 05:41:38,858 INFO Train. Iter 244000 : Commit. 0.44166 PPL. 412.31 Recons. 0.00226 +2024-03-17 05:42:01,032 INFO Train. Iter 244200 : Commit. 0.43624 PPL. 413.00 Recons. 0.00216 +2024-03-17 05:42:22,690 INFO Train. Iter 244400 : Commit. 0.43599 PPL. 413.22 Recons. 0.00215 +2024-03-17 05:42:43,278 INFO Train. Iter 244600 : Commit. 0.42812 PPL. 412.81 Recons. 0.00204 +2024-03-17 05:43:05,000 INFO Train. Iter 244800 : Commit. 0.43500 PPL. 412.73 Recons. 0.00215 +2024-03-17 05:43:26,617 INFO Train. Iter 245000 : Commit. 0.43369 PPL. 412.68 Recons. 0.00211 +2024-03-17 05:43:48,267 INFO Train. Iter 245200 : Commit. 0.43637 PPL. 412.46 Recons. 0.00214 +2024-03-17 05:44:10,559 INFO Train. Iter 245400 : Commit. 0.43287 PPL. 412.81 Recons. 0.00209 +2024-03-17 05:44:32,244 INFO Train. Iter 245600 : Commit. 0.43760 PPL. 412.86 Recons. 0.00214 +2024-03-17 05:44:53,842 INFO Train. Iter 245800 : Commit. 0.43225 PPL. 413.21 Recons. 0.00206 +2024-03-17 05:45:15,933 INFO Train. Iter 246000 : Commit. 0.43616 PPL. 412.01 Recons. 0.00215 +2024-03-17 05:45:38,019 INFO Train. Iter 246200 : Commit. 0.43260 PPL. 411.97 Recons. 0.00213 +2024-03-17 05:45:58,717 INFO Train. Iter 246400 : Commit. 0.43922 PPL. 412.48 Recons. 0.00217 +2024-03-17 05:46:20,678 INFO Train. Iter 246600 : Commit. 0.44478 PPL. 411.92 Recons. 0.00230 +2024-03-17 05:46:42,312 INFO Train. Iter 246800 : Commit. 0.42468 PPL. 412.35 Recons. 0.00200 +2024-03-17 05:47:04,209 INFO Train. Iter 247000 : Commit. 0.43089 PPL. 413.15 Recons. 0.00206 +2024-03-17 05:47:25,796 INFO Train. Iter 247200 : Commit. 0.44155 PPL. 413.01 Recons. 0.00219 +2024-03-17 05:47:47,406 INFO Train. Iter 247400 : Commit. 0.43694 PPL. 411.92 Recons. 0.00220 +2024-03-17 05:48:09,158 INFO Train. Iter 247600 : Commit. 0.43389 PPL. 412.72 Recons. 0.00209 +2024-03-17 05:48:31,419 INFO Train. Iter 247800 : Commit. 0.43121 PPL. 412.80 Recons. 0.00207 +2024-03-17 05:48:52,968 INFO Train. Iter 248000 : Commit. 0.43525 PPL. 411.63 Recons. 0.00219 +2024-03-17 05:49:14,192 INFO Train. Iter 248200 : Commit. 0.43432 PPL. 412.79 Recons. 0.00212 +2024-03-17 05:49:35,078 INFO Train. Iter 248400 : Commit. 0.43082 PPL. 412.35 Recons. 0.00209 +2024-03-17 05:49:55,916 INFO Train. Iter 248600 : Commit. 0.43087 PPL. 412.07 Recons. 0.00210 +2024-03-17 05:50:17,535 INFO Train. Iter 248800 : Commit. 0.43887 PPL. 411.41 Recons. 0.00225 +2024-03-17 05:50:39,265 INFO Train. Iter 249000 : Commit. 0.43265 PPL. 412.24 Recons. 0.00210 +2024-03-17 05:51:00,505 INFO Train. Iter 249200 : Commit. 0.43515 PPL. 411.83 Recons. 0.00215 +2024-03-17 05:51:22,919 INFO Train. Iter 249400 : Commit. 0.44300 PPL. 412.25 Recons. 0.00226 +2024-03-17 05:51:44,355 INFO Train. Iter 249600 : Commit. 0.43848 PPL. 412.34 Recons. 0.00217 +2024-03-17 05:52:05,827 INFO Train. Iter 249800 : Commit. 0.43632 PPL. 412.41 Recons. 0.00215 +2024-03-17 05:52:27,038 INFO Train. Iter 250000 : Commit. 0.43204 PPL. 411.37 Recons. 0.00213 +2024-03-17 05:52:49,055 INFO Train. Iter 250200 : Commit. 0.43966 PPL. 411.93 Recons. 0.00222 +2024-03-17 05:53:09,693 INFO Train. Iter 250400 : Commit. 0.43071 PPL. 413.03 Recons. 0.00205 +2024-03-17 05:53:31,760 INFO Train. Iter 250600 : Commit. 0.43338 PPL. 412.26 Recons. 0.00212 +2024-03-17 05:53:53,137 INFO Train. Iter 250800 : Commit. 0.44427 PPL. 411.04 Recons. 0.00235 +2024-03-17 05:54:14,101 INFO Train. Iter 251000 : Commit. 0.43360 PPL. 412.48 Recons. 0.00212 +2024-03-17 05:54:35,077 INFO Train. Iter 251200 : Commit. 0.44139 PPL. 412.44 Recons. 0.00222 +2024-03-17 05:54:56,165 INFO Train. Iter 251400 : Commit. 0.44134 PPL. 410.99 Recons. 0.00227 +2024-03-17 05:55:17,636 INFO Train. Iter 251600 : Commit. 0.43255 PPL. 412.70 Recons. 0.00209 +2024-03-17 05:55:38,659 INFO Train. Iter 251800 : Commit. 0.43134 PPL. 411.69 Recons. 0.00211 +2024-03-17 05:55:59,578 INFO Train. Iter 252000 : Commit. 0.43154 PPL. 412.68 Recons. 0.00209 +2024-03-17 05:56:19,537 INFO Train. Iter 252200 : Commit. 0.42805 PPL. 412.61 Recons. 0.00203 +2024-03-17 05:56:41,308 INFO Train. Iter 252400 : Commit. 0.42822 PPL. 412.74 Recons. 0.00203 +2024-03-17 05:57:03,207 INFO Train. Iter 252600 : Commit. 0.43557 PPL. 411.89 Recons. 0.00215 +2024-03-17 05:57:25,103 INFO Train. Iter 252800 : Commit. 0.43325 PPL. 412.93 Recons. 0.00209 +2024-03-17 05:57:47,191 INFO Train. Iter 253000 : Commit. 0.42679 PPL. 411.52 Recons. 0.00205 +2024-03-17 05:58:08,912 INFO Train. Iter 253200 : Commit. 0.42826 PPL. 412.23 Recons. 0.00204 +2024-03-17 05:58:30,287 INFO Train. Iter 253400 : Commit. 0.43410 PPL. 411.25 Recons. 0.00219 +2024-03-17 05:58:51,681 INFO Train. Iter 253600 : Commit. 0.43672 PPL. 412.30 Recons. 0.00216 +2024-03-17 05:59:13,339 INFO Train. Iter 253800 : Commit. 0.42695 PPL. 411.49 Recons. 0.00206 +2024-03-17 05:59:34,357 INFO Train. Iter 254000 : Commit. 0.43452 PPL. 412.51 Recons. 0.00211 +2024-03-17 05:59:54,688 INFO Train. Iter 254200 : Commit. 0.43352 PPL. 412.36 Recons. 0.00210 +2024-03-17 06:00:15,606 INFO Train. Iter 254400 : Commit. 0.43066 PPL. 412.39 Recons. 0.00207 +2024-03-17 06:00:36,863 INFO Train. Iter 254600 : Commit. 0.43925 PPL. 411.27 Recons. 0.00225 +2024-03-17 06:00:57,777 INFO Train. Iter 254800 : Commit. 0.42533 PPL. 411.47 Recons. 0.00202 +2024-03-17 06:01:18,829 INFO Train. Iter 255000 : Commit. 0.42407 PPL. 411.76 Recons. 0.00200 +2024-03-17 06:01:39,926 INFO Train. Iter 255200 : Commit. 0.43301 PPL. 412.00 Recons. 0.00213 +2024-03-17 06:02:01,509 INFO Train. Iter 255400 : Commit. 0.43782 PPL. 411.85 Recons. 0.00218 +2024-03-17 06:02:22,662 INFO Train. Iter 255600 : Commit. 0.43886 PPL. 412.29 Recons. 0.00220 +2024-03-17 06:02:43,857 INFO Train. Iter 255800 : Commit. 0.44067 PPL. 411.20 Recons. 0.00226 +2024-03-17 06:03:05,303 INFO Train. Iter 256000 : Commit. 0.43472 PPL. 411.89 Recons. 0.00214 +2024-03-17 06:03:25,591 INFO Train. Iter 256200 : Commit. 0.42701 PPL. 412.10 Recons. 0.00205 +2024-03-17 06:03:47,284 INFO Train. Iter 256400 : Commit. 0.43885 PPL. 410.51 Recons. 0.00226 +2024-03-17 06:04:08,268 INFO Train. Iter 256600 : Commit. 0.43193 PPL. 411.21 Recons. 0.00214 +2024-03-17 06:04:29,773 INFO Train. Iter 256800 : Commit. 0.43236 PPL. 412.42 Recons. 0.00209 +2024-03-17 06:04:51,026 INFO Train. Iter 257000 : Commit. 0.43145 PPL. 411.18 Recons. 0.00213 +2024-03-17 06:05:12,028 INFO Train. Iter 257200 : Commit. 0.42675 PPL. 411.77 Recons. 0.00204 +2024-03-17 06:05:32,868 INFO Train. Iter 257400 : Commit. 0.42729 PPL. 411.76 Recons. 0.00208 +2024-03-17 06:05:53,559 INFO Train. Iter 257600 : Commit. 0.42849 PPL. 411.85 Recons. 0.00208 +2024-03-17 06:06:14,935 INFO Train. Iter 257800 : Commit. 0.43129 PPL. 411.66 Recons. 0.00210 +2024-03-17 06:06:35,500 INFO Train. Iter 258000 : Commit. 0.42898 PPL. 411.18 Recons. 0.00208 +2024-03-17 06:06:57,198 INFO Train. Iter 258200 : Commit. 0.43337 PPL. 411.98 Recons. 0.00212 +2024-03-17 06:07:18,598 INFO Train. Iter 258400 : Commit. 0.43354 PPL. 410.97 Recons. 0.00215 +2024-03-17 06:07:39,951 INFO Train. Iter 258600 : Commit. 0.42682 PPL. 411.72 Recons. 0.00204 +2024-03-17 06:08:01,339 INFO Train. Iter 258800 : Commit. 0.42812 PPL. 411.21 Recons. 0.00207 +2024-03-17 06:08:22,686 INFO Train. Iter 259000 : Commit. 0.42199 PPL. 411.39 Recons. 0.00198 +2024-03-17 06:08:44,080 INFO Train. Iter 259200 : Commit. 0.42815 PPL. 411.08 Recons. 0.00211 +2024-03-17 06:09:05,409 INFO Train. Iter 259400 : Commit. 0.43267 PPL. 411.64 Recons. 0.00212 +2024-03-17 06:09:26,557 INFO Train. Iter 259600 : Commit. 0.43432 PPL. 410.95 Recons. 0.00218 +2024-03-17 06:09:48,109 INFO Train. Iter 259800 : Commit. 0.43236 PPL. 411.13 Recons. 0.00214 +2024-03-17 06:10:08,886 INFO Train. Iter 260000 : Commit. 0.42689 PPL. 411.40 Recons. 0.00204 +2024-03-17 06:10:30,371 INFO Train. Iter 260200 : Commit. 0.42818 PPL. 411.11 Recons. 0.00208 +2024-03-17 06:10:51,590 INFO Train. Iter 260400 : Commit. 0.42498 PPL. 411.71 Recons. 0.00204 +2024-03-17 06:11:12,630 INFO Train. Iter 260600 : Commit. 0.43992 PPL. 410.95 Recons. 0.00223 +2024-03-17 06:11:34,612 INFO Train. Iter 260800 : Commit. 0.43311 PPL. 412.26 Recons. 0.00209 +2024-03-17 06:11:56,482 INFO Train. Iter 261000 : Commit. 0.42221 PPL. 411.21 Recons. 0.00202 +2024-03-17 06:12:18,773 INFO Train. Iter 261200 : Commit. 0.42771 PPL. 410.44 Recons. 0.00210 +2024-03-17 06:12:40,805 INFO Train. Iter 261400 : Commit. 0.43513 PPL. 412.43 Recons. 0.00215 +2024-03-17 06:13:02,853 INFO Train. Iter 261600 : Commit. 0.43621 PPL. 411.02 Recons. 0.00220 +2024-03-17 06:13:24,757 INFO Train. Iter 261800 : Commit. 0.42331 PPL. 412.95 Recons. 0.00196 +2024-03-17 06:13:44,971 INFO Train. Iter 262000 : Commit. 0.43020 PPL. 411.06 Recons. 0.00213 +2024-03-17 06:14:06,067 INFO Train. Iter 262200 : Commit. 0.42969 PPL. 411.42 Recons. 0.00211 +2024-03-17 06:14:27,196 INFO Train. Iter 262400 : Commit. 0.42515 PPL. 411.31 Recons. 0.00203 +2024-03-17 06:14:48,524 INFO Train. Iter 262600 : Commit. 0.43039 PPL. 411.28 Recons. 0.00213 +2024-03-17 06:15:10,312 INFO Train. Iter 262800 : Commit. 0.43279 PPL. 410.41 Recons. 0.00217 +2024-03-17 06:15:31,289 INFO Train. Iter 263000 : Commit. 0.43971 PPL. 411.11 Recons. 0.00225 +2024-03-17 06:15:53,470 INFO Train. Iter 263200 : Commit. 0.43234 PPL. 411.63 Recons. 0.00212 +2024-03-17 06:16:15,565 INFO Train. Iter 263400 : Commit. 0.43223 PPL. 410.88 Recons. 0.00214 +2024-03-17 06:16:37,617 INFO Train. Iter 263600 : Commit. 0.42602 PPL. 411.13 Recons. 0.00205 +2024-03-17 06:16:58,508 INFO Train. Iter 263800 : Commit. 0.42790 PPL. 410.70 Recons. 0.00207 +2024-03-17 06:17:21,058 INFO Train. Iter 264000 : Commit. 0.43735 PPL. 410.11 Recons. 0.00227 +2024-03-17 06:17:42,991 INFO Train. Iter 264200 : Commit. 0.43532 PPL. 410.22 Recons. 0.00222 +2024-03-17 06:18:04,405 INFO Train. Iter 264400 : Commit. 0.43631 PPL. 410.58 Recons. 0.00221 +2024-03-17 06:18:25,870 INFO Train. Iter 264600 : Commit. 0.42285 PPL. 410.58 Recons. 0.00205 +2024-03-17 06:18:47,239 INFO Train. Iter 264800 : Commit. 0.43348 PPL. 410.62 Recons. 0.00217 +2024-03-17 06:19:08,552 INFO Train. Iter 265000 : Commit. 0.42942 PPL. 410.84 Recons. 0.00211 +2024-03-17 06:19:29,837 INFO Train. Iter 265200 : Commit. 0.43286 PPL. 411.83 Recons. 0.00211 +2024-03-17 06:19:51,615 INFO Train. Iter 265400 : Commit. 0.42437 PPL. 410.71 Recons. 0.00204 +2024-03-17 06:20:12,968 INFO Train. Iter 265600 : Commit. 0.43582 PPL. 411.28 Recons. 0.00218 +2024-03-17 06:20:33,528 INFO Train. Iter 265800 : Commit. 0.43069 PPL. 410.99 Recons. 0.00214 +2024-03-17 06:20:55,762 INFO Train. Iter 266000 : Commit. 0.43449 PPL. 412.38 Recons. 0.00212 +2024-03-17 06:21:18,210 INFO Train. Iter 266200 : Commit. 0.43305 PPL. 410.36 Recons. 0.00220 +2024-03-17 06:21:40,367 INFO Train. Iter 266400 : Commit. 0.42678 PPL. 410.71 Recons. 0.00207 +2024-03-17 06:22:01,087 INFO Train. Iter 266600 : Commit. 0.43951 PPL. 411.21 Recons. 0.00225 +2024-03-17 06:22:22,213 INFO Train. Iter 266800 : Commit. 0.42923 PPL. 411.22 Recons. 0.00210 +2024-03-17 06:22:43,295 INFO Train. Iter 267000 : Commit. 0.43414 PPL. 411.23 Recons. 0.00217 +2024-03-17 06:23:05,040 INFO Train. Iter 267200 : Commit. 0.41797 PPL. 411.24 Recons. 0.00192 +2024-03-17 06:23:26,019 INFO Train. Iter 267400 : Commit. 0.42408 PPL. 411.34 Recons. 0.00202 +2024-03-17 06:23:47,606 INFO Train. Iter 267600 : Commit. 0.43305 PPL. 411.07 Recons. 0.00218 +2024-03-17 06:24:08,557 INFO Train. Iter 267800 : Commit. 0.43101 PPL. 410.52 Recons. 0.00214 +2024-03-17 06:24:29,288 INFO Train. Iter 268000 : Commit. 0.42820 PPL. 409.95 Recons. 0.00213 +2024-03-17 06:24:50,778 INFO Train. Iter 268200 : Commit. 0.43104 PPL. 411.27 Recons. 0.00209 +2024-03-17 06:25:12,478 INFO Train. Iter 268400 : Commit. 0.43347 PPL. 410.49 Recons. 0.00216 +2024-03-17 06:25:34,289 INFO Train. Iter 268600 : Commit. 0.43009 PPL. 411.61 Recons. 0.00211 +2024-03-17 06:25:56,303 INFO Train. Iter 268800 : Commit. 0.42973 PPL. 411.14 Recons. 0.00210 +2024-03-17 06:26:17,838 INFO Train. Iter 269000 : Commit. 0.42977 PPL. 411.09 Recons. 0.00209 +2024-03-17 06:26:39,796 INFO Train. Iter 269200 : Commit. 0.42379 PPL. 411.35 Recons. 0.00201 +2024-03-17 06:27:01,777 INFO Train. Iter 269400 : Commit. 0.42695 PPL. 411.63 Recons. 0.00203 +2024-03-17 06:27:21,649 INFO Train. Iter 269600 : Commit. 0.42664 PPL. 410.06 Recons. 0.00208 +2024-03-17 06:27:42,902 INFO Train. Iter 269800 : Commit. 0.42916 PPL. 411.46 Recons. 0.00205 +2024-03-17 06:28:03,774 INFO Train. Iter 270000 : Commit. 0.42578 PPL. 411.93 Recons. 0.00200 +2024-03-17 06:28:25,114 INFO Train. Iter 270200 : Commit. 0.43444 PPL. 410.51 Recons. 0.00216 +2024-03-17 06:28:45,433 INFO Train. Iter 270400 : Commit. 0.43339 PPL. 411.82 Recons. 0.00210 +2024-03-17 06:29:07,072 INFO Train. Iter 270600 : Commit. 0.43846 PPL. 410.50 Recons. 0.00225 +2024-03-17 06:29:28,329 INFO Train. Iter 270800 : Commit. 0.42555 PPL. 410.57 Recons. 0.00204 +2024-03-17 06:29:50,163 INFO Train. Iter 271000 : Commit. 0.42550 PPL. 411.07 Recons. 0.00203 +2024-03-17 06:30:11,274 INFO Train. Iter 271200 : Commit. 0.42317 PPL. 410.52 Recons. 0.00205 +2024-03-17 06:30:32,514 INFO Train. Iter 271400 : Commit. 0.43171 PPL. 411.02 Recons. 0.00215 +2024-03-17 06:30:52,526 INFO Train. Iter 271600 : Commit. 0.43259 PPL. 410.66 Recons. 0.00216 +2024-03-17 06:31:13,674 INFO Train. Iter 271800 : Commit. 0.44286 PPL. 410.37 Recons. 0.00232 +2024-03-17 06:31:35,012 INFO Train. Iter 272000 : Commit. 0.43626 PPL. 410.12 Recons. 0.00222 +2024-03-17 06:31:56,359 INFO Train. Iter 272200 : Commit. 0.43260 PPL. 410.94 Recons. 0.00215 +2024-03-17 06:32:17,866 INFO Train. Iter 272400 : Commit. 0.42801 PPL. 410.50 Recons. 0.00211 +2024-03-17 06:32:39,001 INFO Train. Iter 272600 : Commit. 0.43482 PPL. 411.26 Recons. 0.00218 +2024-03-17 06:33:00,342 INFO Train. Iter 272800 : Commit. 0.42871 PPL. 411.41 Recons. 0.00207 +2024-03-17 06:33:21,740 INFO Train. Iter 273000 : Commit. 0.42736 PPL. 411.43 Recons. 0.00205 +2024-03-17 06:33:42,952 INFO Train. Iter 273200 : Commit. 0.43042 PPL. 410.10 Recons. 0.00215 +2024-03-17 06:34:04,507 INFO Train. Iter 273400 : Commit. 0.42396 PPL. 410.44 Recons. 0.00204 +2024-03-17 06:34:25,112 INFO Train. Iter 273600 : Commit. 0.42190 PPL. 410.42 Recons. 0.00200 +2024-03-17 06:34:46,393 INFO Train. Iter 273800 : Commit. 0.42408 PPL. 410.83 Recons. 0.00203 +2024-03-17 06:35:08,100 INFO Train. Iter 274000 : Commit. 0.43228 PPL. 410.87 Recons. 0.00214 +2024-03-17 06:35:30,183 INFO Train. Iter 274200 : Commit. 0.43210 PPL. 411.63 Recons. 0.00209 +2024-03-17 06:35:52,238 INFO Train. Iter 274400 : Commit. 0.42751 PPL. 410.61 Recons. 0.00207 +2024-03-17 06:36:13,878 INFO Train. Iter 274600 : Commit. 0.43451 PPL. 409.94 Recons. 0.00224 +2024-03-17 06:36:35,592 INFO Train. Iter 274800 : Commit. 0.42921 PPL. 411.07 Recons. 0.00209 +2024-03-17 06:36:57,445 INFO Train. Iter 275000 : Commit. 0.42871 PPL. 410.69 Recons. 0.00209 +2024-03-17 06:37:19,398 INFO Train. Iter 275200 : Commit. 0.43639 PPL. 409.75 Recons. 0.00226 +2024-03-17 06:37:40,685 INFO Train. Iter 275400 : Commit. 0.42481 PPL. 411.17 Recons. 0.00203 +2024-03-17 06:38:02,565 INFO Train. Iter 275600 : Commit. 0.43108 PPL. 410.20 Recons. 0.00213 +2024-03-17 06:38:24,861 INFO Train. Iter 275800 : Commit. 0.42579 PPL. 410.32 Recons. 0.00206 +2024-03-17 06:38:47,017 INFO Train. Iter 276000 : Commit. 0.42615 PPL. 410.66 Recons. 0.00204 +2024-03-17 06:39:08,794 INFO Train. Iter 276200 : Commit. 0.42652 PPL. 411.53 Recons. 0.00204 +2024-03-17 06:39:29,854 INFO Train. Iter 276400 : Commit. 0.42932 PPL. 410.65 Recons. 0.00209 +2024-03-17 06:39:51,724 INFO Train. Iter 276600 : Commit. 0.42807 PPL. 409.79 Recons. 0.00210 +2024-03-17 06:40:13,210 INFO Train. Iter 276800 : Commit. 0.43108 PPL. 410.49 Recons. 0.00212 +2024-03-17 06:40:34,574 INFO Train. Iter 277000 : Commit. 0.43821 PPL. 409.23 Recons. 0.00226 +2024-03-17 06:40:55,743 INFO Train. Iter 277200 : Commit. 0.42980 PPL. 410.17 Recons. 0.00212 +2024-03-17 06:41:15,877 INFO Train. Iter 277400 : Commit. 0.43210 PPL. 410.55 Recons. 0.00215 +2024-03-17 06:41:37,469 INFO Train. Iter 277600 : Commit. 0.42657 PPL. 410.68 Recons. 0.00207 +2024-03-17 06:41:59,024 INFO Train. Iter 277800 : Commit. 0.41610 PPL. 410.10 Recons. 0.00193 +2024-03-17 06:42:20,840 INFO Train. Iter 278000 : Commit. 0.41734 PPL. 410.84 Recons. 0.00195 +2024-03-17 06:42:42,624 INFO Train. Iter 278200 : Commit. 0.42760 PPL. 411.48 Recons. 0.00205 +2024-03-17 06:43:03,879 INFO Train. Iter 278400 : Commit. 0.42663 PPL. 409.59 Recons. 0.00210 +2024-03-17 06:43:25,332 INFO Train. Iter 278600 : Commit. 0.42364 PPL. 410.31 Recons. 0.00204 +2024-03-17 06:43:46,728 INFO Train. Iter 278800 : Commit. 0.42438 PPL. 411.27 Recons. 0.00201 +2024-03-17 06:44:07,446 INFO Train. Iter 279000 : Commit. 0.42578 PPL. 410.08 Recons. 0.00207 +2024-03-17 06:44:28,577 INFO Train. Iter 279200 : Commit. 0.42183 PPL. 409.75 Recons. 0.00203 +2024-03-17 06:44:48,945 INFO Train. Iter 279400 : Commit. 0.42728 PPL. 409.72 Recons. 0.00210 +2024-03-17 06:45:10,672 INFO Train. Iter 279600 : Commit. 0.42505 PPL. 410.08 Recons. 0.00206 +2024-03-17 06:45:32,530 INFO Train. Iter 279800 : Commit. 0.42646 PPL. 410.11 Recons. 0.00206 +2024-03-17 06:45:54,246 INFO Train. Iter 280000 : Commit. 0.43943 PPL. 409.76 Recons. 0.00228 +2024-03-17 06:46:16,331 INFO Train. Iter 280200 : Commit. 0.42969 PPL. 409.36 Recons. 0.00215 +2024-03-17 06:46:38,239 INFO Train. Iter 280400 : Commit. 0.42444 PPL. 409.77 Recons. 0.00205 +2024-03-17 06:46:59,523 INFO Train. Iter 280600 : Commit. 0.42604 PPL. 410.41 Recons. 0.00207 +2024-03-17 06:47:20,553 INFO Train. Iter 280800 : Commit. 0.42730 PPL. 410.18 Recons. 0.00209 +2024-03-17 06:47:41,682 INFO Train. Iter 281000 : Commit. 0.42157 PPL. 409.14 Recons. 0.00204 +2024-03-17 06:48:01,571 INFO Train. Iter 281200 : Commit. 0.42526 PPL. 409.53 Recons. 0.00208 +2024-03-17 06:48:22,455 INFO Train. Iter 281400 : Commit. 0.43281 PPL. 409.41 Recons. 0.00220 +2024-03-17 06:48:43,827 INFO Train. Iter 281600 : Commit. 0.43044 PPL. 409.64 Recons. 0.00213 +2024-03-17 06:49:05,585 INFO Train. Iter 281800 : Commit. 0.42624 PPL. 409.44 Recons. 0.00209 +2024-03-17 06:49:27,035 INFO Train. Iter 282000 : Commit. 0.42852 PPL. 409.21 Recons. 0.00213 +2024-03-17 06:49:47,743 INFO Train. Iter 282200 : Commit. 0.42337 PPL. 409.42 Recons. 0.00204 +2024-03-17 06:50:09,088 INFO Train. Iter 282400 : Commit. 0.42488 PPL. 409.17 Recons. 0.00209 +2024-03-17 06:50:30,740 INFO Train. Iter 282600 : Commit. 0.42158 PPL. 409.46 Recons. 0.00204 +2024-03-17 06:50:52,598 INFO Train. Iter 282800 : Commit. 0.42907 PPL. 410.19 Recons. 0.00211 +2024-03-17 06:51:14,385 INFO Train. Iter 283000 : Commit. 0.42698 PPL. 409.16 Recons. 0.00211 +2024-03-17 06:51:34,562 INFO Train. Iter 283200 : Commit. 0.43157 PPL. 408.97 Recons. 0.00219 +2024-03-17 06:51:56,231 INFO Train. Iter 283400 : Commit. 0.43631 PPL. 410.00 Recons. 0.00221 +2024-03-17 06:52:17,447 INFO Train. Iter 283600 : Commit. 0.41938 PPL. 410.31 Recons. 0.00195 +2024-03-17 06:52:38,779 INFO Train. Iter 283800 : Commit. 0.42033 PPL. 410.44 Recons. 0.00198 +2024-03-17 06:53:00,076 INFO Train. Iter 284000 : Commit. 0.42974 PPL. 410.01 Recons. 0.00210 +2024-03-17 06:53:21,697 INFO Train. Iter 284200 : Commit. 0.42376 PPL. 408.85 Recons. 0.00207 +2024-03-17 06:53:43,478 INFO Train. Iter 284400 : Commit. 0.42265 PPL. 409.68 Recons. 0.00203 +2024-03-17 06:54:05,131 INFO Train. Iter 284600 : Commit. 0.42521 PPL. 410.35 Recons. 0.00203 +2024-03-17 06:54:26,458 INFO Train. Iter 284800 : Commit. 0.41774 PPL. 409.61 Recons. 0.00197 +2024-03-17 06:54:47,859 INFO Train. Iter 285000 : Commit. 0.42032 PPL. 409.15 Recons. 0.00202 +2024-03-17 06:55:07,991 INFO Train. Iter 285200 : Commit. 0.42010 PPL. 410.71 Recons. 0.00197 +2024-03-17 06:55:29,909 INFO Train. Iter 285400 : Commit. 0.42424 PPL. 409.58 Recons. 0.00204 +2024-03-17 06:55:51,942 INFO Train. Iter 285600 : Commit. 0.42503 PPL. 409.08 Recons. 0.00208 +2024-03-17 06:56:13,667 INFO Train. Iter 285800 : Commit. 0.42841 PPL. 408.83 Recons. 0.00213 +2024-03-17 06:56:35,348 INFO Train. Iter 286000 : Commit. 0.42632 PPL. 408.16 Recons. 0.00213 +2024-03-17 06:56:56,669 INFO Train. Iter 286200 : Commit. 0.42249 PPL. 410.43 Recons. 0.00201 +2024-03-17 06:57:17,758 INFO Train. Iter 286400 : Commit. 0.42404 PPL. 409.27 Recons. 0.00204 +2024-03-17 06:57:38,667 INFO Train. Iter 286600 : Commit. 0.42566 PPL. 409.37 Recons. 0.00207 +2024-03-17 06:58:00,260 INFO Train. Iter 286800 : Commit. 0.41998 PPL. 409.16 Recons. 0.00201 +2024-03-17 06:58:20,338 INFO Train. Iter 287000 : Commit. 0.42140 PPL. 409.16 Recons. 0.00202 +2024-03-17 06:58:41,830 INFO Train. Iter 287200 : Commit. 0.43033 PPL. 409.72 Recons. 0.00211 +2024-03-17 06:59:02,954 INFO Train. Iter 287400 : Commit. 0.42694 PPL. 409.65 Recons. 0.00208 +2024-03-17 06:59:24,133 INFO Train. Iter 287600 : Commit. 0.42452 PPL. 410.05 Recons. 0.00201 +2024-03-17 06:59:45,724 INFO Train. Iter 287800 : Commit. 0.42439 PPL. 408.77 Recons. 0.00209 +2024-03-17 07:00:07,537 INFO Train. Iter 288000 : Commit. 0.43092 PPL. 409.26 Recons. 0.00215 +2024-03-17 07:00:29,748 INFO Train. Iter 288200 : Commit. 0.43200 PPL. 409.12 Recons. 0.00218 +2024-03-17 07:00:51,898 INFO Train. Iter 288400 : Commit. 0.42146 PPL. 409.28 Recons. 0.00202 +2024-03-17 07:01:14,257 INFO Train. Iter 288600 : Commit. 0.41865 PPL. 408.84 Recons. 0.00200 +2024-03-17 07:01:35,328 INFO Train. Iter 288800 : Commit. 0.42144 PPL. 409.75 Recons. 0.00202 +2024-03-17 07:01:55,815 INFO Train. Iter 289000 : Commit. 0.42638 PPL. 408.49 Recons. 0.00214 +2024-03-17 07:02:16,720 INFO Train. Iter 289200 : Commit. 0.43642 PPL. 408.60 Recons. 0.00225 +2024-03-17 07:02:38,054 INFO Train. Iter 289400 : Commit. 0.41989 PPL. 409.53 Recons. 0.00198 +2024-03-17 07:02:58,975 INFO Train. Iter 289600 : Commit. 0.41927 PPL. 408.30 Recons. 0.00203 +2024-03-17 07:03:20,550 INFO Train. Iter 289800 : Commit. 0.42715 PPL. 409.97 Recons. 0.00208 +2024-03-17 07:03:41,936 INFO Train. Iter 290000 : Commit. 0.42489 PPL. 409.86 Recons. 0.00206 +2024-03-17 07:04:02,858 INFO Train. Iter 290200 : Commit. 0.42706 PPL. 409.70 Recons. 0.00205 +2024-03-17 07:04:24,663 INFO Train. Iter 290400 : Commit. 0.41818 PPL. 409.12 Recons. 0.00198 +2024-03-17 07:04:46,276 INFO Train. Iter 290600 : Commit. 0.42679 PPL. 410.23 Recons. 0.00208 +2024-03-17 07:05:08,145 INFO Train. Iter 290800 : Commit. 0.41755 PPL. 408.39 Recons. 0.00200 +2024-03-17 07:05:28,335 INFO Train. Iter 291000 : Commit. 0.42561 PPL. 409.60 Recons. 0.00206 +2024-03-17 07:05:49,719 INFO Train. Iter 291200 : Commit. 0.41751 PPL. 409.58 Recons. 0.00198 +2024-03-17 07:06:11,053 INFO Train. Iter 291400 : Commit. 0.41208 PPL. 408.72 Recons. 0.00192 +2024-03-17 07:06:31,895 INFO Train. Iter 291600 : Commit. 0.43428 PPL. 409.35 Recons. 0.00220 +2024-03-17 07:06:53,853 INFO Train. Iter 291800 : Commit. 0.42090 PPL. 409.21 Recons. 0.00202 +2024-03-17 07:07:15,385 INFO Train. Iter 292000 : Commit. 0.42985 PPL. 409.23 Recons. 0.00215 +2024-03-17 07:07:36,885 INFO Train. Iter 292200 : Commit. 0.42241 PPL. 409.23 Recons. 0.00203 +2024-03-17 07:07:57,754 INFO Train. Iter 292400 : Commit. 0.42721 PPL. 409.01 Recons. 0.00209 +2024-03-17 07:08:18,924 INFO Train. Iter 292600 : Commit. 0.42590 PPL. 410.10 Recons. 0.00205 +2024-03-17 07:08:39,535 INFO Train. Iter 292800 : Commit. 0.42133 PPL. 410.81 Recons. 0.00195 +2024-03-17 07:09:00,645 INFO Train. Iter 293000 : Commit. 0.42076 PPL. 409.08 Recons. 0.00202 +2024-03-17 07:09:22,068 INFO Train. Iter 293200 : Commit. 0.42033 PPL. 409.02 Recons. 0.00203 +2024-03-17 07:09:43,929 INFO Train. Iter 293400 : Commit. 0.42223 PPL. 409.46 Recons. 0.00204 +2024-03-17 07:10:05,333 INFO Train. Iter 293600 : Commit. 0.42670 PPL. 408.72 Recons. 0.00210 +2024-03-17 07:10:27,375 INFO Train. Iter 293800 : Commit. 0.42116 PPL. 409.20 Recons. 0.00200 +2024-03-17 07:10:49,225 INFO Train. Iter 294000 : Commit. 0.42195 PPL. 409.15 Recons. 0.00201 +2024-03-17 07:11:11,026 INFO Train. Iter 294200 : Commit. 0.42983 PPL. 407.50 Recons. 0.00217 +2024-03-17 07:11:32,649 INFO Train. Iter 294400 : Commit. 0.42694 PPL. 409.27 Recons. 0.00211 +2024-03-17 07:11:54,359 INFO Train. Iter 294600 : Commit. 0.42165 PPL. 408.78 Recons. 0.00203 +2024-03-17 07:12:15,246 INFO Train. Iter 294800 : Commit. 0.41943 PPL. 409.72 Recons. 0.00198 +2024-03-17 07:12:36,738 INFO Train. Iter 295000 : Commit. 0.41928 PPL. 409.47 Recons. 0.00198 +2024-03-17 07:12:58,470 INFO Train. Iter 295200 : Commit. 0.42412 PPL. 409.37 Recons. 0.00204 +2024-03-17 07:13:20,172 INFO Train. Iter 295400 : Commit. 0.42383 PPL. 409.12 Recons. 0.00206 +2024-03-17 07:13:41,347 INFO Train. Iter 295600 : Commit. 0.42694 PPL. 409.43 Recons. 0.00205 +2024-03-17 07:14:02,556 INFO Train. Iter 295800 : Commit. 0.43074 PPL. 409.29 Recons. 0.00214 +2024-03-17 07:14:24,821 INFO Train. Iter 296000 : Commit. 0.42008 PPL. 408.46 Recons. 0.00203 +2024-03-17 07:14:46,071 INFO Train. Iter 296200 : Commit. 0.42948 PPL. 409.43 Recons. 0.00210 +2024-03-17 07:15:07,474 INFO Train. Iter 296400 : Commit. 0.41836 PPL. 408.03 Recons. 0.00202 +2024-03-17 07:15:29,030 INFO Train. Iter 296600 : Commit. 0.42066 PPL. 408.25 Recons. 0.00202 +2024-03-17 07:15:49,886 INFO Train. Iter 296800 : Commit. 0.42555 PPL. 408.54 Recons. 0.00212 +2024-03-17 07:16:11,830 INFO Train. Iter 297000 : Commit. 0.43335 PPL. 408.21 Recons. 0.00223 +2024-03-17 07:16:33,020 INFO Train. Iter 297200 : Commit. 0.42376 PPL. 409.49 Recons. 0.00203 +2024-03-17 07:16:54,482 INFO Train. Iter 297400 : Commit. 0.42319 PPL. 406.98 Recons. 0.00213 +2024-03-17 07:17:15,615 INFO Train. Iter 297600 : Commit. 0.42241 PPL. 409.20 Recons. 0.00204 +2024-03-17 07:17:36,714 INFO Train. Iter 297800 : Commit. 0.42365 PPL. 408.00 Recons. 0.00208 +2024-03-17 07:17:58,349 INFO Train. Iter 298000 : Commit. 0.42023 PPL. 407.78 Recons. 0.00205 +2024-03-17 07:18:19,793 INFO Train. Iter 298200 : Commit. 0.42064 PPL. 409.47 Recons. 0.00201 +2024-03-17 07:18:40,994 INFO Train. Iter 298400 : Commit. 0.42204 PPL. 409.56 Recons. 0.00201 +2024-03-17 07:19:01,693 INFO Train. Iter 298600 : Commit. 0.42349 PPL. 408.36 Recons. 0.00204 +2024-03-17 07:19:22,971 INFO Train. Iter 298800 : Commit. 0.41310 PPL. 408.10 Recons. 0.00194 +2024-03-17 07:19:44,429 INFO Train. Iter 299000 : Commit. 0.42036 PPL. 408.54 Recons. 0.00204 +2024-03-17 07:20:06,057 INFO Train. Iter 299200 : Commit. 0.42235 PPL. 408.98 Recons. 0.00203 +2024-03-17 07:20:27,138 INFO Train. Iter 299400 : Commit. 0.42083 PPL. 409.13 Recons. 0.00199 +2024-03-17 07:20:48,322 INFO Train. Iter 299600 : Commit. 0.42146 PPL. 408.67 Recons. 0.00202 +2024-03-17 07:21:09,748 INFO Train. Iter 299800 : Commit. 0.42003 PPL. 408.48 Recons. 0.00201 +2024-03-17 07:21:31,335 INFO Train. Iter 300000 : Commit. 0.41831 PPL. 409.45 Recons. 0.00197 diff --git a/configs/beat2_rvqvae.yaml b/configs/beat2_rvqvae.yaml new file mode 100644 index 0000000000000000000000000000000000000000..a21d97aa3f70e12ef5a7ec021eb666e784691225 --- /dev/null +++ b/configs/beat2_rvqvae.yaml @@ -0,0 +1,134 @@ +is_train: True +ddp: False +stat: ts +root_path: ./ +out_path: ./outputs/audio2pose/ +project: s2g +data_path: ./datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/ +e_path: weights/AESKConv_240_100.bin +eval_model: motion_representation +e_name: VAESKConv +test_ckpt: ./outputs/audio2pose/custom/0112_001634_emage/last_200.bin +data_path_1: ./datasets/hub/ + +vae_test_len: 32 +vae_test_dim: 330 +vae_test_stride: 20 +vae_length: 240 +vae_codebook_size: 256 +vae_layer: 4 +vae_grow: [1,1,2,1] +variational: False + +# data config +training_speakers: [2] #[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30] #[2] +additional_data: False +cache_path: datasets/beat_cache/beat_smplx_en_emage_2_rvqvae/ +dataset: mix_sep +new_cache: True +use_amass: False +# motion config +ori_joints: beat_smplx_joints +tar_joints: beat_smplx_full +pose_rep: smplxflame_30 +pose_norm: False +pose_fps: 30 +rot6d: True +pre_frames: 4 +pose_dims: 330 +pose_length: 64 +stride: 20 +test_length: 64 +motion_f: 256 +m_pre_encoder: null +m_encoder: null +m_fix_pre: False + +# audio config +audio_rep: onset+amplitude +audio_sr: 16000 +audio_fps: 16000 +audio_norm: False +audio_f: 256 +# a_pre_encoder: tcn_camn +# a_encoder: none +# a_fix_pre: False + +# text config +word_rep: textgrid +word_index_num: 11195 +word_dims: 300 +freeze_wordembed: False +word_f: 256 +t_pre_encoder: fasttext +t_encoder: null +t_fix_pre: False + +# facial config +facial_rep: smplxflame_30 +facial_dims: 100 +facial_norm: False +facial_f: 0 +f_pre_encoder: null +f_encoder: null +f_fix_pre: False + +# speaker config +id_rep: onehot +speaker_f: 0 + +# model config +batch_size: 80 #80 +# warmup_epochs: 1 +# warmup_lr: 1e-6 +lr_base: 4e-4 +model: motion_representation +g_name: VQVAEConvZero +trainer: ae_total +hidden_size: 768 +n_layer: 1 + +rec_weight: 1 +grad_norm: 0.99 +epochs: 200 +test_period: 20 +ll: 3 +lf: 3 +lu: 3 +lh: 3 +cl: 1 +cf: 0 +cu: 1 +ch: 1 + + + +#below is vavae config, copy from QPGESTURE +#Codebook Configs +levels: 1 +downs_t: [3] +strides_t : [2] +emb_width : 512 +l_bins : 512 +l_mu : 0.99 +commit : 0.1 +hvqvae_multipliers : [1] +width: 512 +depth: 3 +m_conv : 1.0 +dilation_growth_rate : 3 +sample_length: 80 +use_bottleneck: True +joint_channel: 6 +# depth: 3 +# width: 128 +# m_conv: 1.0 +# dilation_growth_rate: 1 +# dilation_cycle: None +vel: 1 # 1 -> 0 +acc: 1 # 1 -> 0 +vqvae_reverse_decoder_dilation: True + + +## below is special for emage +rec_pos_weight : 1.0 \ No newline at end of file diff --git a/configs/diffusion_rvqvae_128.yaml b/configs/diffusion_rvqvae_128.yaml new file mode 100644 index 0000000000000000000000000000000000000000..74864848d47c6fd917d9c250d819eb3d0cdbdeee --- /dev/null +++ b/configs/diffusion_rvqvae_128.yaml @@ -0,0 +1,118 @@ +is_train: True +ddp: False +stat: ts +root_path: ./ +out_path: ./outputs/audio2pose/ +project: s2g +data_path: ./datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/ +e_path: weights/AESKConv_240_100.bin +eval_model: motion_representation +e_name: VAESKConv +test_ckpt: ./ckpt/beatx2_cospeech_diffusion/last_500.bin +data_path_1: ./datasets/hub/ +pose_norm: True + + +mean_pose_path: ./mean_std/beatx_2_330_mean.npy +std_pose_path: ./mean_std/beatx_2_330_std.npy + +mean_trans_path: ./mean_std/beatx_2_trans_mean.npy +std_trans_path: ./mean_std/beatx_2_trans_std.npy + + +vqvae_upper_path: ./ckpt/beatx2_rvqvae/RVQVAE_upper/net_300000.pth +vqvae_hands_path: ./ckpt/beatx2_rvqvae/RVQVAE_hands/net_300000.pth +vqvae_lower_path: ./ckpt/beatx2_rvqvae/RVQVAE_lower/net_300000.pth + +vqvae_lower_trans_path: ./ckpt/beatx2_rvqvae/RVQVAE_lower_trans/net_300000.pth +use_trans: True + +decay_epoch: 500 + +vqvae_squeeze_scale: 4 +vqvae_type: rvqvae +vqvae_latent_scale: 5 + +vae_test_len: 32 +vae_test_dim: 330 +vae_test_stride: 20 +vae_length: 240 +vae_codebook_size: 256 +vae_layer: 4 +vae_grow: [1,1,2,1] +variational: False + +# data config +training_speakers: [2] #[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30] +additional_data: False +cache_path: datasets/beat_cache/beat_smplx_en_emage_2_128/ +dataset: beat_sep_lower +new_cache: False + +# motion config +ori_joints: beat_smplx_joints +tar_joints: beat_smplx_full +pose_rep: smplxflame_30 +pose_fps: 30 +rot6d: True +pre_frames: 4 +pose_dims: 330 +pose_length: 128 +stride: 20 +test_length: 128 +motion_f: 256 +m_pre_encoder: null +m_encoder: null +m_fix_pre: False + + +audio_rep: onset+amplitude +audio_sr: 16000 +audio_fps: 16000 +audio_norm: False +audio_f: 256 + + +word_rep: textgrid +word_index_num: 11195 +word_dims: 300 +freeze_wordembed: False +word_f: 256 +t_pre_encoder: fasttext +t_encoder: null +t_fix_pre: False + + +facial_rep: smplxflame_30 +facial_dims: 100 +facial_norm: False +facial_f: 0 +f_pre_encoder: null +f_encoder: null +f_fix_pre: False + + +id_rep: onehot +speaker_f: 0 + + +batch_size: 40 +lr_base: 5e-5 +model: denoiser +g_name: MDM +trainer: diffusion_rvqvae +hidden_size: 768 +n_layer: 1 + +rec_weight: 1 +grad_norm: 0.99 +epochs: 2000 +test_period: 20 +ll: 3 +lf: 3 +lu: 3 +lh: 3 +cl: 1 +cf: 0 +cu: 1 +ch: 1 diff --git a/configs/diffusion_rvqvae_128_hf.yaml b/configs/diffusion_rvqvae_128_hf.yaml new file mode 100644 index 0000000000000000000000000000000000000000..bb22c37705c064b41582ac92d6cdbfded9365706 --- /dev/null +++ b/configs/diffusion_rvqvae_128_hf.yaml @@ -0,0 +1,118 @@ +is_train: True +ddp: False +stat: ts +root_path: ./ +out_path: ./outputs/audio2pose/ +project: s2g +data_path: ./datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/ +e_path: weights/AESKConv_240_100.bin +eval_model: motion_representation +e_name: VAESKConv +test_ckpt: ./ckpt/beatx2_cospeech_diffusion/last_500.bin +data_path_1: ./datasets/hub/ +pose_norm: True + + +mean_pose_path: ./mean_std/beatx_2_330_mean.npy +std_pose_path: ./mean_std/beatx_2_330_std.npy + +mean_trans_path: ./mean_std/beatx_2_trans_mean.npy +std_trans_path: ./mean_std/beatx_2_trans_std.npy + + +vqvae_upper_path: ./ckpt/beatx2_rvqvae/RVQVAE_upper/net_300000.pth +vqvae_hands_path: ./ckpt/beatx2_rvqvae/RVQVAE_hands/net_300000.pth +vqvae_lower_path: ./ckpt/beatx2_rvqvae/RVQVAE_lower/net_300000.pth + +vqvae_lower_trans_path: ./ckpt/beatx2_rvqvae/RVQVAE_lower_trans/net_300000.pth +use_trans: True + +decay_epoch: 500 + +vqvae_squeeze_scale: 4 +vqvae_type: rvqvae +vqvae_latent_scale: 5 + +vae_test_len: 32 +vae_test_dim: 330 +vae_test_stride: 20 +vae_length: 240 +vae_codebook_size: 256 +vae_layer: 4 +vae_grow: [1,1,2,1] +variational: False + +# data config +training_speakers: [2] #[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30] +additional_data: False +cache_path: datasets/beat_cache/web_demo_test/ +dataset: beat_sep_lower_single +new_cache: True + +# motion config +ori_joints: beat_smplx_joints +tar_joints: beat_smplx_full +pose_rep: smplxflame_30 +pose_fps: 30 +rot6d: True +pre_frames: 4 +pose_dims: 330 +pose_length: 128 +stride: 20 +test_length: 128 +motion_f: 256 +m_pre_encoder: null +m_encoder: null +m_fix_pre: False + + +audio_rep: onset+amplitude +audio_sr: 16000 +audio_fps: 16000 +audio_norm: False +audio_f: 256 + + +word_rep: textgrid +word_index_num: 11195 +word_dims: 300 +freeze_wordembed: False +word_f: 256 +t_pre_encoder: fasttext +t_encoder: null +t_fix_pre: False + + +facial_rep: smplxflame_30 +facial_dims: 100 +facial_norm: False +facial_f: 0 +f_pre_encoder: null +f_encoder: null +f_fix_pre: False + + +id_rep: onehot +speaker_f: 0 + + +batch_size: 40 +lr_base: 5e-5 +model: denoiser +g_name: MDM +trainer: diffusion_rvqvae +hidden_size: 768 +n_layer: 1 + +rec_weight: 1 +grad_norm: 0.99 +epochs: 2000 +test_period: 20 +ll: 3 +lf: 3 +lu: 3 +lh: 3 +cl: 1 +cf: 0 +cu: 1 +ch: 1 diff --git a/dataloaders/amass_sep_lower.py b/dataloaders/amass_sep_lower.py new file mode 100644 index 0000000000000000000000000000000000000000..500faca42467490e509ce0354324860d11c406ba --- /dev/null +++ b/dataloaders/amass_sep_lower.py @@ -0,0 +1,713 @@ +import os +import pickle +import math +import shutil +import numpy as np +import lmdb as lmdb +import textgrid as tg +import pandas as pd +import torch +import glob +import json +from termcolor import colored +from loguru import logger +from collections import defaultdict +from torch.utils.data import Dataset +import torch.distributed as dist +#import pyarrow +import pickle +import librosa +import smplx +import glob + +from .build_vocab import Vocab +from .utils.audio_features import Wav2Vec2Model +from .data_tools import joints_list +from .utils import rotation_conversions as rc +from .utils import other_tools + +# ACCAD 120 +# BioMotionLab_NTroje 120 +# CMU 很复杂 +# EKUT 100 +# Eyes_Japan_Dataset 很复杂 +# HumanEva 很复杂 +# KIT 100 +# MPI_HDM05 120 +# MPI_Limits 120 +# MPI_mosh 很复杂 +# SFU 120 +# SSM_synced 很复杂 +# TCD_handMocap 很复杂 +# TotalCapture 60 +# Transitions_mocap 120 + +all_sequences = [ + 'ACCAD', + 'BioMotionLab_NTroje', + 'CMU', + 'EKUT', + 'Eyes_Japan_Dataset', + 'HumanEva', + 'KIT', + 'MPI_HDM05', + 'MPI_Limits', + 'MPI_mosh', + 'SFU', + 'SSM_synced', + 'TCD_handMocap', + 'TotalCapture', + 'Transitions_mocap', +] +amass_test_split = ['Transitions_mocap', 'SSM_synced'] +amass_vald_split = ['HumanEva', 'MPI_HDM05', 'SFU', 'MPI_mosh'] +amass_train_split = ['BioMotionLab_NTroje', 'Eyes_Japan_Dataset', 'TotalCapture', 'KIT', 'ACCAD', 'CMU', 'MPI_Limits', + 'TCD_handMocap', 'EKUT'] + +# 上面这些spilt方式是MOTION CLIP的，但是由于motionx中的framerate处理有问题，我先暂且只挑部分数据集进行训练 +# 这些都是120fps的 +# amass_test_split = ['SFU'] +# amass_vald_split = ['MPI_Limits'] +# amass_train_split = ['BioMotionLab_NTroje', 'MPI_HDM05', 'ACCAD','Transitions_mocap'] + + +amass_splits = { + 'test': amass_test_split, + 'val': amass_vald_split, + 'train': amass_train_split +} +class CustomDataset(Dataset): + def __init__(self, args, loader_type, augmentation=None, kwargs=None, build_cache=True): + self.args = args + self.loader_type = loader_type + + self.rank = dist.get_rank() + self.ori_stride = self.args.stride + self.ori_length = self.args.pose_length + self.alignment = [0,0] # for trinity + + self.ori_joint_list = joints_list[self.args.ori_joints] + self.tar_joint_list = joints_list[self.args.tar_joints] + if 'smplx' in self.args.pose_rep: + self.joint_mask = np.zeros(len(list(self.ori_joint_list.keys()))*3) + self.joints = len(list(self.tar_joint_list.keys())) + for joint_name in self.tar_joint_list: + self.joint_mask[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + else: + self.joints = len(list(self.ori_joint_list.keys()))+1 + self.joint_mask = np.zeros(self.joints*3) + for joint_name in self.tar_joint_list: + if joint_name == "Hips": + self.joint_mask[3:6] = 1 + else: + self.joint_mask[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + # select trainable joints + self.smplx = smplx.create( + self.args.data_path_1+"smplx_models/", + model_type='smplx', + gender='NEUTRAL_2020', + use_face_contour=False, + num_betas=300, + num_expression_coeffs=100, + ext='npz', + use_pca=False, + ).cuda().eval() + + split_rule = pd.read_csv(args.data_path+"train_test_split.csv") + self.selected_file = split_rule.loc[(split_rule['type'] == loader_type) & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + if args.additional_data and loader_type == 'train': + split_b = split_rule.loc[(split_rule['type'] == 'additional') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + #self.selected_file = split_rule.loc[(split_rule['type'] == 'additional') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + self.selected_file = pd.concat([self.selected_file, split_b]) + if self.selected_file.empty: + logger.warning(f"{loader_type} is empty for speaker {self.args.training_speakers}, use train set 0-8 instead") + self.selected_file = split_rule.loc[(split_rule['type'] == 'train') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + self.selected_file = self.selected_file.iloc[0:8] + self.data_dir = args.data_path + + if loader_type == "test": + self.args.multi_length_training = [1.0] + self.max_length = int(args.pose_length * self.args.multi_length_training[-1]) + self.max_audio_pre_len = math.floor(args.pose_length / args.pose_fps * self.args.audio_sr) + if self.max_audio_pre_len > self.args.test_length*self.args.audio_sr: + self.max_audio_pre_len = self.args.test_length*self.args.audio_sr + + if args.word_rep is not None: + with open(f"{args.data_path}weights/vocab.pkl", 'rb') as f: + self.lang_model = pickle.load(f) + + preloaded_dir = self.args.root_path + 'datasets/beat_cache/amass_smplx_en_emage_new/' + loader_type + f"/{args.pose_rep}_cache" + # if args.pose_norm: + # # careful for rotation vectors + # if not os.path.exists(args.data_path+args.mean_pose_path+f"{args.pose_rep.split('_')[0]}/bvh_mean.npy"): + # self.calculate_mean_pose() + # self.mean_pose = np.load(args.data_path+args.mean_pose_path+f"{args.pose_rep.split('_')[0]}/bvh_mean.npy") + # self.std_pose = np.load(args.data_path+args.mean_pose_path+f"{args.pose_rep.split('_')[0]}/bvh_std.npy") + # if args.audio_norm: + # if not os.path.exists(args.data_path+args.mean_pose_path+f"{args.audio_rep.split('_')[0]}/bvh_mean.npy"): + # self.calculate_mean_audio() + # self.mean_audio = np.load(args.data_path+args.mean_pose_path+f"{args.audio_rep.split('_')[0]}/npy_mean.npy") + # self.std_audio = np.load(args.data_path+args.mean_pose_path+f"{args.audio_rep.split('_')[0]}/npy_std.npy") + # if args.facial_norm: + # if not os.path.exists(args.data_path+args.mean_pose_path+f"{args.pose_rep.split('_')[0]}/bvh_mean.npy"): + # self.calculate_mean_face() + # self.mean_facial = np.load(args.data_path+args.mean_pose_path+f"{args.facial_rep}/json_mean.npy") + # self.std_facial = np.load(args.data_path+args.mean_pose_path+f"{args.facial_rep}/json_std.npy") + if self.args.beat_align: + if not os.path.exists(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy"): + self.calculate_mean_velocity(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy") + self.avg_vel = np.load(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy") + + if build_cache and self.rank == 0: + self.build_cache(preloaded_dir) + self.lmdb_env = lmdb.open(preloaded_dir, readonly=True, lock=False) + with self.lmdb_env.begin() as txn: + self.n_samples = txn.stat()["entries"] + + + def calculate_mean_velocity(self, save_path): + self.smplx = smplx.create( + self.args.data_path_1+"smplx_models/", + model_type='smplx', + gender='NEUTRAL_2020', + use_face_contour=False, + num_betas=300, + num_expression_coeffs=100, + ext='npz', + use_pca=False, + ).cuda().eval() + dir_p = self.data_dir + self.args.pose_rep + "/" + all_list = [] + from tqdm import tqdm + for tar in tqdm(os.listdir(dir_p)): + if tar.endswith(".npz"): + m_data = np.load(dir_p+tar, allow_pickle=True) + betas, poses, trans, exps = m_data["betas"], m_data["poses"], m_data["trans"], m_data["expressions"] + n, c = poses.shape[0], poses.shape[1] + betas = betas.reshape(1, 300) + betas = np.tile(betas, (n, 1)) + betas = torch.from_numpy(betas).cuda().float() + poses = torch.from_numpy(poses.reshape(n, c)).cuda().float() + exps = torch.from_numpy(exps.reshape(n, 100)).cuda().float() + trans = torch.from_numpy(trans.reshape(n, 3)).cuda().float() + max_length = 128 + s, r = n//max_length, n%max_length + #print(n, s, r) + all_tensor = [] + for i in range(s): + with torch.no_grad(): + joints = self.smplx( + betas=betas[i*max_length:(i+1)*max_length], + transl=trans[i*max_length:(i+1)*max_length], + expression=exps[i*max_length:(i+1)*max_length], + jaw_pose=poses[i*max_length:(i+1)*max_length, 66:69], + global_orient=poses[i*max_length:(i+1)*max_length,:3], + body_pose=poses[i*max_length:(i+1)*max_length,3:21*3+3], + left_hand_pose=poses[i*max_length:(i+1)*max_length,25*3:40*3], + right_hand_pose=poses[i*max_length:(i+1)*max_length,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[i*max_length:(i+1)*max_length, 69:72], + reye_pose=poses[i*max_length:(i+1)*max_length, 72:75], + )['joints'][:, :55, :].reshape(max_length, 55*3) + all_tensor.append(joints) + if r != 0: + with torch.no_grad(): + joints = self.smplx( + betas=betas[s*max_length:s*max_length+r], + transl=trans[s*max_length:s*max_length+r], + expression=exps[s*max_length:s*max_length+r], + jaw_pose=poses[s*max_length:s*max_length+r, 66:69], + global_orient=poses[s*max_length:s*max_length+r,:3], + body_pose=poses[s*max_length:s*max_length+r,3:21*3+3], + left_hand_pose=poses[s*max_length:s*max_length+r,25*3:40*3], + right_hand_pose=poses[s*max_length:s*max_length+r,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[s*max_length:s*max_length+r, 69:72], + reye_pose=poses[s*max_length:s*max_length+r, 72:75], + )['joints'][:, :55, :].reshape(r, 55*3) + all_tensor.append(joints) + joints = torch.cat(all_tensor, axis=0) + joints = joints.permute(1, 0) + dt = 1/30 + # first steps is forward diff (t+1 - t) / dt + init_vel = (joints[:, 1:2] - joints[:, :1]) / dt + # middle steps are second order (t+1 - t-1) / 2dt + middle_vel = (joints[:, 2:] - joints[:, 0:-2]) / (2 * dt) + # last step is backward diff (t - t-1) / dt + final_vel = (joints[:, -1:] - joints[:, -2:-1]) / dt + #print(joints.shape, init_vel.shape, middle_vel.shape, final_vel.shape) + vel_seq = torch.cat([init_vel, middle_vel, final_vel], dim=1).permute(1, 0).reshape(n, 55, 3) + #print(vel_seq.shape) + #.permute(1, 0).reshape(n, 55, 3) + vel_seq_np = vel_seq.cpu().numpy() + vel_joints_np = np.linalg.norm(vel_seq_np, axis=2) # n * 55 + all_list.append(vel_joints_np) + avg_vel = np.mean(np.concatenate(all_list, axis=0),axis=0) # 55 + np.save(save_path, avg_vel) + + + def build_cache(self, preloaded_dir): + logger.info(f"Audio bit rate: {self.args.audio_fps}") + logger.info("Reading data '{}'...".format(self.data_dir)) + logger.info("Creating the dataset cache...") + if self.args.new_cache: + if os.path.exists(preloaded_dir): + shutil.rmtree(preloaded_dir) + if os.path.exists(preloaded_dir): + logger.info("Found the cache {}".format(preloaded_dir)) + elif self.loader_type == "test": + self.cache_generation( + preloaded_dir, True, + 0, 0, + is_test=True) + else: + self.cache_generation( + preloaded_dir, self.args.disable_filtering, + self.args.clean_first_seconds, self.args.clean_final_seconds, + is_test=False) + + def __len__(self): + return self.n_samples + + + def load_amass(self,data): + ## 这个是用来 + # 修改amass数据里面的朝向，原本在blender里面是Z轴向上，目标是Y轴向上，当时面向目前没改 + + data_dict = {key: data[key] for key in data} + frames = data_dict['poses'].shape[0] + b = data_dict['poses'][...,:3] + b = rc.axis_angle_to_matrix(torch.from_numpy(b)) + rot_matrix = np.array([[1.0, 0.0, 0.0], [0.0 , 0.0, 1.0], [0.0, -1.0, 0.0]]) + c = np.einsum('ij,kjl->kil',rot_matrix,b) + c = rc.matrix_to_axis_angle(torch.from_numpy(c)) + data_dict['poses'][...,:3] = c + + trans_matrix1 = np.array([[1.0, 0.0, 0.0], [0.0 , 0.0, -1.0], [0.0, 1.0, 0.0]]) + data_dict['trans'] = np.einsum("bi,ij->bj",data_dict['trans'],trans_matrix1) + + betas300 = np.zeros(300) + betas300[:16] = data_dict['betas'] + data_dict['betas'] = betas300 + data_dict["expressions"] = np.zeros((frames,100)) + + return data_dict + + def cache_generation(self, out_lmdb_dir, disable_filtering, clean_first_seconds, clean_final_seconds, is_test=False): + # if "wav2vec2" in self.args.audio_rep: + # self.wav2vec_model = Wav2Vec2Model.from_pretrained(f"{self.args.data_path_1}/hub/transformer/wav2vec2-base-960h") + # self.wav2vec_model.feature_extractor._freeze_parameters() + # self.wav2vec_model = self.wav2vec_model.cuda() + # self.wav2vec_model.eval() + + self.n_out_samples = 0 + # create db for samples + if not os.path.exists(out_lmdb_dir): os.makedirs(out_lmdb_dir) + dst_lmdb_env = lmdb.open(out_lmdb_dir, map_size= int(1024 ** 3 * 500))# 500G + n_filtered_out = defaultdict(int) + + + if self.args.use_amass: + amass_dir = '/mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/datasets/AMASS_SMPLX' + for dataset in amass_splits[self.loader_type]: + search_path = os.path.join(amass_dir,dataset, '**', '*.npz') + npz_files = glob.glob(search_path, recursive=True) + for index, file_name in enumerate(npz_files): + f_name = file_name.split('/')[-1] + ext = ".npz" if "smplx" in self.args.pose_rep else ".bvh" + pose_file = file_name + pose_each_file = [] + trans_each_file = [] + trans_v_each_file = [] + shape_each_file = [] + audio_each_file = [] + facial_each_file = [] + word_each_file = [] + emo_each_file = [] + sem_each_file = [] + vid_each_file = [] + id_pose = f_name #1_wayne_0_1_1 + get_foot_contact = True + logger.info(colored(f"# ---- Building cache for Pose {id_pose} ---- #", "blue")) + if "smplx" in self.args.pose_rep: + pose_data = np.load(pose_file, allow_pickle=True) + if len(pose_data.files)==6: + logger.info(colored(f"# ---- state file ---- #", "red")) + continue + assert 30%self.args.pose_fps == 0, 'pose_fps should be an aliquot part of 30' + assert self.args.pose_fps == 30, "should 30" + m_data = np.load(pose_file, allow_pickle=True) + m_data= self.load_amass(m_data) + betas, poses, trans, exps = m_data["betas"], m_data["poses"], m_data["trans"], m_data["expressions"] + mocap_framerate = float(m_data['mocap_frame_rate']) + stride = round(mocap_framerate / self.args.pose_fps) + pose_each_file = poses[::stride] + trans_each_file = trans[::stride] + trans_each_file[:,0] = trans_each_file[:,0] - trans_each_file[0,0] + trans_each_file[:,2] = trans_each_file[:,2] - trans_each_file[0,2] + trans_v_each_file = np.zeros_like(trans_each_file) + trans_v_each_file[1:,0] = trans_each_file[1:,0] - trans_each_file[:-1,0] + trans_v_each_file[0,0] = trans_v_each_file[1,0] + trans_v_each_file[1:,2] = trans_each_file[1:,2] - trans_each_file[:-1,2] + trans_v_each_file[0,2] = trans_v_each_file[1,2] + trans_v_each_file[:,1] = trans_each_file[:,1] + + + shape_each_file = np.repeat(betas.reshape(1, -1), pose_each_file.shape[0], axis=0) + + n, c = poses.shape[0], poses.shape[1] + betas = betas.reshape(1, 300) + betas = np.tile(betas, (n, 1)) + betas = torch.from_numpy(betas).cuda().float() + poses = torch.from_numpy(poses.reshape(n, c)).cuda().float() + exps = torch.from_numpy(exps.reshape(n, 100)).cuda().float() + trans = torch.from_numpy(trans.reshape(n, 3)).cuda().float() + + if get_foot_contact: + max_length = 128 + s, r = n//max_length, n%max_length + #print(n, s, r) + all_tensor = [] + for i in range(s): + with torch.no_grad(): + joints = self.smplx( + betas=betas[i*max_length:(i+1)*max_length], + transl=trans[i*max_length:(i+1)*max_length], + expression=exps[i*max_length:(i+1)*max_length], + jaw_pose=poses[i*max_length:(i+1)*max_length, 66:69], + global_orient=poses[i*max_length:(i+1)*max_length,:3], + body_pose=poses[i*max_length:(i+1)*max_length,3:21*3+3], + left_hand_pose=poses[i*max_length:(i+1)*max_length,25*3:40*3], + right_hand_pose=poses[i*max_length:(i+1)*max_length,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[i*max_length:(i+1)*max_length, 69:72], + reye_pose=poses[i*max_length:(i+1)*max_length, 72:75], + )['joints'][:, (7,8,10,11), :].reshape(max_length, 4, 3).cpu() + all_tensor.append(joints) + if r != 0: + with torch.no_grad(): + joints = self.smplx( + betas=betas[s*max_length:s*max_length+r], + transl=trans[s*max_length:s*max_length+r], + expression=exps[s*max_length:s*max_length+r], + jaw_pose=poses[s*max_length:s*max_length+r, 66:69], + global_orient=poses[s*max_length:s*max_length+r,:3], + body_pose=poses[s*max_length:s*max_length+r,3:21*3+3], + left_hand_pose=poses[s*max_length:s*max_length+r,25*3:40*3], + right_hand_pose=poses[s*max_length:s*max_length+r,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[s*max_length:s*max_length+r, 69:72], + reye_pose=poses[s*max_length:s*max_length+r, 72:75], + )['joints'][:, (7,8,10,11), :].reshape(r, 4, 3).cpu() + all_tensor.append(joints) + joints = torch.cat(all_tensor, axis=0) # all, 4, 3 + # print(joints.shape) + feetv = torch.zeros(joints.shape[1], joints.shape[0]) + joints = joints.permute(1, 0, 2) + #print(joints.shape, feetv.shape) + feetv[:, :-1] = (joints[:, 1:] - joints[:, :-1]).norm(dim=-1) + #print(feetv.shape) + contacts = (feetv < 0.01).numpy().astype(float) + # print(contacts.shape, contacts) + contacts = contacts.transpose(1, 0)[::stride] + pose_each_file = pose_each_file * self.joint_mask + pose_each_file = pose_each_file[:, self.joint_mask.astype(bool)] + pose_each_file = np.concatenate([pose_each_file, contacts], axis=1) + # print(pose_each_file.shape) + else: + pose_each_file = pose_each_file * self.joint_mask + pose_each_file = pose_each_file[:, self.joint_mask.astype(bool)] + + # print(pose_each_file.shape) + + + if self.args.id_rep is not None: + vid_each_file = np.repeat(np.array(int(100)-1).reshape(1, 1), pose_each_file.shape[0], axis=0) + + filtered_result = self._sample_from_clip( + dst_lmdb_env, + audio_each_file, pose_each_file, trans_each_file, trans_v_each_file,shape_each_file, facial_each_file, word_each_file, + vid_each_file, emo_each_file, sem_each_file, + disable_filtering, clean_first_seconds, clean_final_seconds, is_test, + ) + for type in filtered_result.keys(): + n_filtered_out[type] += filtered_result[type] + + + + with dst_lmdb_env.begin() as txn: + logger.info(colored(f"no. of samples: {txn.stat()['entries']}", "cyan")) + n_total_filtered = 0 + for type, n_filtered in n_filtered_out.items(): + logger.info("{}: {}".format(type, n_filtered)) + n_total_filtered += n_filtered + logger.info(colored("no. of excluded samples: {} ({:.1f}%)".format( + n_total_filtered, 100 * n_total_filtered / (txn.stat()["entries"] + n_total_filtered)), "cyan")) + dst_lmdb_env.sync() + dst_lmdb_env.close() + + def _sample_from_clip( + self, dst_lmdb_env, audio_each_file, pose_each_file, trans_each_file, trans_v_each_file,shape_each_file, facial_each_file, word_each_file, + vid_each_file, emo_each_file, sem_each_file, + disable_filtering, clean_first_seconds, clean_final_seconds, is_test, + ): + """ + for data cleaning, we ignore the data for first and final n s + for test, we return all data + """ + # audio_start = int(self.alignment[0] * self.args.audio_fps) + # pose_start = int(self.alignment[1] * self.args.pose_fps) + #logger.info(f"before: {audio_each_file.shape} {pose_each_file.shape}") + # audio_each_file = audio_each_file[audio_start:] + # pose_each_file = pose_each_file[pose_start:] + # trans_each_file = + #logger.info(f"after alignment: {audio_each_file.shape} {pose_each_file.shape}") + #print(pose_each_file.shape) + round_seconds_skeleton = pose_each_file.shape[0] // self.args.pose_fps # assume 1500 frames / 15 fps = 100 s + #print(round_seconds_skeleton) + if audio_each_file != []: + if self.args.audio_rep != "wave16k": + round_seconds_audio = len(audio_each_file) // self.args.audio_fps # assume 16,000,00 / 16,000 = 100 s + elif self.args.audio_rep == "mfcc": + round_seconds_audio = audio_each_file.shape[0] // self.args.audio_fps + else: + round_seconds_audio = audio_each_file.shape[0] // self.args.audio_sr + if facial_each_file != []: + round_seconds_facial = facial_each_file.shape[0] // self.args.pose_fps + logger.info(f"audio: {round_seconds_audio}s, pose: {round_seconds_skeleton}s, facial: {round_seconds_facial}s") + round_seconds_skeleton = min(round_seconds_audio, round_seconds_skeleton, round_seconds_facial) + max_round = max(round_seconds_audio, round_seconds_skeleton, round_seconds_facial) + if round_seconds_skeleton != max_round: + logger.warning(f"reduce to {round_seconds_skeleton}s, ignore {max_round-round_seconds_skeleton}s") + else: + logger.info(f"pose: {round_seconds_skeleton}s, audio: {round_seconds_audio}s") + round_seconds_skeleton = min(round_seconds_audio, round_seconds_skeleton) + max_round = max(round_seconds_audio, round_seconds_skeleton) + if round_seconds_skeleton != max_round: + logger.warning(f"reduce to {round_seconds_skeleton}s, ignore {max_round-round_seconds_skeleton}s") + + clip_s_t, clip_e_t = clean_first_seconds, round_seconds_skeleton - clean_final_seconds # assume [10, 90]s + clip_s_f_audio, clip_e_f_audio = self.args.audio_fps * clip_s_t, clip_e_t * self.args.audio_fps # [160,000,90*160,000] + clip_s_f_pose, clip_e_f_pose = clip_s_t * self.args.pose_fps, clip_e_t * self.args.pose_fps # [150,90*15] + + + for ratio in self.args.multi_length_training: + if is_test:# stride = length for test + cut_length = clip_e_f_pose - clip_s_f_pose + self.args.stride = cut_length + self.max_length = cut_length + else: + self.args.stride = int(ratio*self.ori_stride) + cut_length = int(self.ori_length*ratio) + + num_subdivision = math.floor((clip_e_f_pose - clip_s_f_pose - cut_length) / self.args.stride) + 1 + logger.info(f"pose from frame {clip_s_f_pose} to {clip_e_f_pose}, length {cut_length}") + logger.info(f"{num_subdivision} clips is expected with stride {self.args.stride}") + + if audio_each_file != []: + audio_short_length = math.floor(cut_length / self.args.pose_fps * self.args.audio_fps) + """ + for audio sr = 16000, fps = 15, pose_length = 34, + audio short length = 36266.7 -> 36266 + this error is fine. + """ + logger.info(f"audio from frame {clip_s_f_audio} to {clip_e_f_audio}, length {audio_short_length}") + + n_filtered_out = defaultdict(int) + sample_pose_list = [] + sample_audio_list = [] + sample_facial_list = [] + sample_shape_list = [] + sample_word_list = [] + sample_emo_list = [] + sample_sem_list = [] + sample_vid_list = [] + sample_trans_list = [] + sample_trans_v_list = [] + + for i in range(num_subdivision): # cut into around 2s chip, (self npose) + start_idx = clip_s_f_pose + i * self.args.stride + fin_idx = start_idx + cut_length + sample_pose = pose_each_file[start_idx:fin_idx] + + sample_trans = trans_each_file[start_idx:fin_idx] + sample_trans_v = trans_v_each_file[start_idx:fin_idx] + sample_shape = shape_each_file[start_idx:fin_idx] + # print(sample_pose.shape) + if self.args.audio_rep is not None and audio_each_file != []: + audio_start = clip_s_f_audio + math.floor(i * self.args.stride * self.args.audio_fps / self.args.pose_fps) + audio_end = audio_start + audio_short_length + sample_audio = audio_each_file[audio_start:audio_end] + else: + sample_audio = np.array([-1]) + sample_facial = facial_each_file[start_idx:fin_idx] if self.args.facial_rep is not None else np.array([-1]) + sample_word = word_each_file[start_idx:fin_idx] if self.args.word_rep is not None else np.array([-1]) + sample_emo = emo_each_file[start_idx:fin_idx] if self.args.emo_rep is not None else np.array([-1]) + sample_sem = sem_each_file[start_idx:fin_idx] if self.args.sem_rep is not None else np.array([-1]) + sample_vid = vid_each_file[start_idx:fin_idx] if self.args.id_rep is not None else np.array([-1]) + + if sample_pose.any() != None: + # filtering motion skeleton data + sample_pose, filtering_message = MotionPreprocessor(sample_pose).get() + is_correct_motion = (sample_pose != []) + if is_correct_motion or disable_filtering: + sample_pose_list.append(sample_pose) + sample_audio_list.append(sample_audio) + sample_facial_list.append(sample_facial) + sample_shape_list.append(sample_shape) + sample_word_list.append(sample_word) + sample_vid_list.append(sample_vid) + sample_emo_list.append(sample_emo) + sample_sem_list.append(sample_sem) + sample_trans_list.append(sample_trans) + sample_trans_v_list.append(sample_trans_v) + else: + n_filtered_out[filtering_message] += 1 + + if len(sample_pose_list) > 0: + with dst_lmdb_env.begin(write=True) as txn: + for pose, audio, facial, shape, word, vid, emo, sem, trans,trans_v in zip( + sample_pose_list, + sample_audio_list, + sample_facial_list, + sample_shape_list, + sample_word_list, + sample_vid_list, + sample_emo_list, + sample_sem_list, + sample_trans_list, + sample_trans_v_list,): + k = "{:005}".format(self.n_out_samples).encode("ascii") + v = [pose, audio, facial, shape, word, emo, sem, vid, trans,trans_v] + v = pickle.dumps(v,5) + txn.put(k, v) + self.n_out_samples += 1 + return n_filtered_out + + def __getitem__(self, idx): + with self.lmdb_env.begin(write=False) as txn: + key = "{:005}".format(idx).encode("ascii") + sample = txn.get(key) + sample = pickle.loads(sample) + tar_pose, in_audio, in_facial, in_shape, in_word, emo, sem, vid, trans,trans_v = sample + #print(in_shape) + #vid = torch.from_numpy(vid).int() + emo = torch.from_numpy(emo).int() + sem = torch.from_numpy(sem).float() + in_audio = np.zeros([68266,2]) + in_audio = torch.from_numpy(in_audio).float() + in_word = np.zeros([128]) + in_facial = np.zeros([128,100]) + in_word = torch.from_numpy(in_word).float() if self.args.word_cache else torch.from_numpy(in_word).int() + if self.loader_type == "test": + tar_pose = torch.from_numpy(tar_pose).float() + trans = torch.from_numpy(trans).float() + trans_v = torch.from_numpy(trans_v).float() + in_facial = torch.from_numpy(in_facial).float() + vid = torch.from_numpy(vid).float() + in_shape = torch.from_numpy(in_shape).float() + else: + in_shape = torch.from_numpy(in_shape).reshape((in_shape.shape[0], -1)).float() + trans = torch.from_numpy(trans).reshape((trans.shape[0], -1)).float() + trans_v = torch.from_numpy(trans_v).reshape((trans_v.shape[0], -1)).float() + vid = torch.from_numpy(vid).reshape((vid.shape[0], -1)).float() + tar_pose = torch.from_numpy(tar_pose).reshape((tar_pose.shape[0], -1)).float() + in_facial = torch.from_numpy(in_facial).reshape((in_facial.shape[0], -1)).float() + return {"pose":tar_pose, "audio":in_audio, "facial":in_facial, "beta": in_shape, "word":in_word, "id":vid, "emo":emo, "sem":sem, "trans":trans,"trans_v":trans_v} + + +class MotionPreprocessor: + def __init__(self, skeletons): + self.skeletons = skeletons + #self.mean_pose = mean_pose + self.filtering_message = "PASS" + + def get(self): + assert (self.skeletons is not None) + + # filtering + if self.skeletons != []: + if self.check_pose_diff(): + self.skeletons = [] + self.filtering_message = "pose" + # elif self.check_spine_angle(): + # self.skeletons = [] + # self.filtering_message = "spine angle" + # elif self.check_static_motion(): + # self.skeletons = [] + # self.filtering_message = "motion" + + # if self.skeletons != []: + # self.skeletons = self.skeletons.tolist() + # for i, frame in enumerate(self.skeletons): + # assert not np.isnan(self.skeletons[i]).any() # missing joints + + return self.skeletons, self.filtering_message + + def check_static_motion(self, verbose=True): + def get_variance(skeleton, joint_idx): + wrist_pos = skeleton[:, joint_idx] + variance = np.sum(np.var(wrist_pos, axis=0)) + return variance + + left_arm_var = get_variance(self.skeletons, 6) + right_arm_var = get_variance(self.skeletons, 9) + + th = 0.0014 # exclude 13110 + # th = 0.002 # exclude 16905 + if left_arm_var < th and right_arm_var < th: + if verbose: + print("skip - check_static_motion left var {}, right var {}".format(left_arm_var, right_arm_var)) + return True + else: + if verbose: + print("pass - check_static_motion left var {}, right var {}".format(left_arm_var, right_arm_var)) + return False + + + def check_pose_diff(self, verbose=False): +# diff = np.abs(self.skeletons - self.mean_pose) # 186*1 +# diff = np.mean(diff) + +# # th = 0.017 +# th = 0.02 #0.02 # exclude 3594 +# if diff < th: +# if verbose: +# print("skip - check_pose_diff {:.5f}".format(diff)) +# return True +# # th = 3.5 #0.02 # exclude 3594 +# # if 3.5 < diff < 5: +# # if verbose: +# # print("skip - check_pose_diff {:.5f}".format(diff)) +# # return True +# else: +# if verbose: +# print("pass - check_pose_diff {:.5f}".format(diff)) + return False + + + def check_spine_angle(self, verbose=True): + def angle_between(v1, v2): + v1_u = v1 / np.linalg.norm(v1) + v2_u = v2 / np.linalg.norm(v2) + return np.arccos(np.clip(np.dot(v1_u, v2_u), -1.0, 1.0)) + + angles = [] + for i in range(self.skeletons.shape[0]): + spine_vec = self.skeletons[i, 1] - self.skeletons[i, 0] + angle = angle_between(spine_vec, [0, -1, 0]) + angles.append(angle) + + if np.rad2deg(max(angles)) > 30 or np.rad2deg(np.mean(angles)) > 20: # exclude 4495 + # if np.rad2deg(max(angles)) > 20: # exclude 8270 + if verbose: + print("skip - check_spine_angle {:.5f}, {:.5f}".format(max(angles), np.mean(angles))) + return True + else: + if verbose: + print("pass - check_spine_angle {:.5f}".format(max(angles))) + return False \ No newline at end of file diff --git a/dataloaders/beat_sep.py b/dataloaders/beat_sep.py new file mode 100644 index 0000000000000000000000000000000000000000..b04615d8c1a1ad370a88e09ab68e78d533fc4d0d --- /dev/null +++ b/dataloaders/beat_sep.py @@ -0,0 +1,772 @@ +import os +import pickle +import math +import shutil +import numpy as np +import lmdb as lmdb +import textgrid as tg +import pandas as pd +import torch +import glob +import json +from termcolor import colored +from loguru import logger +from collections import defaultdict +from torch.utils.data import Dataset +import torch.distributed as dist +#import pyarrow +import pickle +import librosa +import smplx + +from .build_vocab import Vocab +from .utils.audio_features import Wav2Vec2Model +from .data_tools import joints_list +from .utils import rotation_conversions as rc +from .utils import other_tools + +class CustomDataset(Dataset): + def __init__(self, args, loader_type, augmentation=None, kwargs=None, build_cache=True): + self.args = args + self.loader_type = loader_type + + self.rank = dist.get_rank() + self.ori_stride = self.args.stride + self.ori_length = self.args.pose_length + self.alignment = [0,0] # for trinity + + self.ori_joint_list = joints_list[self.args.ori_joints] + self.tar_joint_list = joints_list[self.args.tar_joints] + if 'smplx' in self.args.pose_rep: + self.joint_mask = np.zeros(len(list(self.ori_joint_list.keys()))*3) + self.joints = len(list(self.tar_joint_list.keys())) + for joint_name in self.tar_joint_list: + self.joint_mask[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + else: + self.joints = len(list(self.ori_joint_list.keys()))+1 + self.joint_mask = np.zeros(self.joints*3) + for joint_name in self.tar_joint_list: + if joint_name == "Hips": + self.joint_mask[3:6] = 1 + else: + self.joint_mask[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + # select trainable joints + + split_rule = pd.read_csv(args.data_path+"train_test_split.csv") + self.selected_file = split_rule.loc[(split_rule['type'] == loader_type) & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + if args.additional_data and loader_type == 'train': + split_b = split_rule.loc[(split_rule['type'] == 'additional') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + #self.selected_file = split_rule.loc[(split_rule['type'] == 'additional') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + self.selected_file = pd.concat([self.selected_file, split_b]) + if self.selected_file.empty: + logger.warning(f"{loader_type} is empty for speaker {self.args.training_speakers}, use train set 0-8 instead") + self.selected_file = split_rule.loc[(split_rule['type'] == 'train') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + self.selected_file = self.selected_file.iloc[0:8] + self.data_dir = args.data_path + + if loader_type == "test": + self.args.multi_length_training = [1.0] + self.max_length = int(args.pose_length * self.args.multi_length_training[-1]) + self.max_audio_pre_len = math.floor(args.pose_length / args.pose_fps * self.args.audio_sr) + if self.max_audio_pre_len > self.args.test_length*self.args.audio_sr: + self.max_audio_pre_len = self.args.test_length*self.args.audio_sr + + if args.word_rep is not None: + with open(f"{args.data_path}weights/vocab.pkl", 'rb') as f: + self.lang_model = pickle.load(f) + + preloaded_dir = self.args.root_path + self.args.cache_path + loader_type + f"/{args.pose_rep}_cache" + # if args.pose_norm: + # # careful for rotation vectors + # if not os.path.exists(args.data_path+args.mean_pose_path+f"{args.pose_rep.split('_')[0]}/bvh_mean.npy"): + # self.calculate_mean_pose() + # self.mean_pose = np.load(args.data_path+args.mean_pose_path+f"{args.pose_rep.split('_')[0]}/bvh_mean.npy") + # self.std_pose = np.load(args.data_path+args.mean_pose_path+f"{args.pose_rep.split('_')[0]}/bvh_std.npy") + # if args.audio_norm: + # if not os.path.exists(args.data_path+args.mean_pose_path+f"{args.audio_rep.split('_')[0]}/bvh_mean.npy"): + # self.calculate_mean_audio() + # self.mean_audio = np.load(args.data_path+args.mean_pose_path+f"{args.audio_rep.split('_')[0]}/npy_mean.npy") + # self.std_audio = np.load(args.data_path+args.mean_pose_path+f"{args.audio_rep.split('_')[0]}/npy_std.npy") + # if args.facial_norm: + # if not os.path.exists(args.data_path+args.mean_pose_path+f"{args.pose_rep.split('_')[0]}/bvh_mean.npy"): + # self.calculate_mean_face() + # self.mean_facial = np.load(args.data_path+args.mean_pose_path+f"{args.facial_rep}/json_mean.npy") + # self.std_facial = np.load(args.data_path+args.mean_pose_path+f"{args.facial_rep}/json_std.npy") + if self.args.beat_align: + if not os.path.exists(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy"): + self.calculate_mean_velocity(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy") + self.avg_vel = np.load(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy") + + if build_cache and self.rank == 0: + self.build_cache(preloaded_dir) + self.lmdb_env = lmdb.open(preloaded_dir, readonly=True, lock=False) + with self.lmdb_env.begin() as txn: + self.n_samples = txn.stat()["entries"] + + + def calculate_mean_velocity(self, save_path): + self.smplx = smplx.create( + self.args.data_path_1+"smplx_models/", + model_type='smplx', + gender='NEUTRAL_2020', + use_face_contour=False, + num_betas=300, + num_expression_coeffs=100, + ext='npz', + use_pca=False, + ).cuda().eval() + dir_p = self.data_dir + self.args.pose_rep + "/" + all_list = [] + from tqdm import tqdm + for tar in tqdm(os.listdir(dir_p)): + if tar.endswith(".npz"): + m_data = np.load(dir_p+tar, allow_pickle=True) + betas, poses, trans, exps = m_data["betas"], m_data["poses"], m_data["trans"], m_data["expressions"] + n, c = poses.shape[0], poses.shape[1] + betas = betas.reshape(1, 300) + betas = np.tile(betas, (n, 1)) + betas = torch.from_numpy(betas).cuda().float() + poses = torch.from_numpy(poses.reshape(n, c)).cuda().float() + exps = torch.from_numpy(exps.reshape(n, 100)).cuda().float() + trans = torch.from_numpy(trans.reshape(n, 3)).cuda().float() + max_length = 128 + s, r = n//max_length, n%max_length + #print(n, s, r) + all_tensor = [] + for i in range(s): + with torch.no_grad(): + joints = self.smplx( + betas=betas[i*max_length:(i+1)*max_length], + transl=trans[i*max_length:(i+1)*max_length], + expression=exps[i*max_length:(i+1)*max_length], + jaw_pose=poses[i*max_length:(i+1)*max_length, 66:69], + global_orient=poses[i*max_length:(i+1)*max_length,:3], + body_pose=poses[i*max_length:(i+1)*max_length,3:21*3+3], + left_hand_pose=poses[i*max_length:(i+1)*max_length,25*3:40*3], + right_hand_pose=poses[i*max_length:(i+1)*max_length,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[i*max_length:(i+1)*max_length, 69:72], + reye_pose=poses[i*max_length:(i+1)*max_length, 72:75], + )['joints'][:, :55, :].reshape(max_length, 55*3) + all_tensor.append(joints) + if r != 0: + with torch.no_grad(): + joints = self.smplx( + betas=betas[s*max_length:s*max_length+r], + transl=trans[s*max_length:s*max_length+r], + expression=exps[s*max_length:s*max_length+r], + jaw_pose=poses[s*max_length:s*max_length+r, 66:69], + global_orient=poses[s*max_length:s*max_length+r,:3], + body_pose=poses[s*max_length:s*max_length+r,3:21*3+3], + left_hand_pose=poses[s*max_length:s*max_length+r,25*3:40*3], + right_hand_pose=poses[s*max_length:s*max_length+r,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[s*max_length:s*max_length+r, 69:72], + reye_pose=poses[s*max_length:s*max_length+r, 72:75], + )['joints'][:, :55, :].reshape(r, 55*3) + all_tensor.append(joints) + joints = torch.cat(all_tensor, axis=0) + joints = joints.permute(1, 0) + dt = 1/30 + # first steps is forward diff (t+1 - t) / dt + init_vel = (joints[:, 1:2] - joints[:, :1]) / dt + # middle steps are second order (t+1 - t-1) / 2dt + middle_vel = (joints[:, 2:] - joints[:, 0:-2]) / (2 * dt) + # last step is backward diff (t - t-1) / dt + final_vel = (joints[:, -1:] - joints[:, -2:-1]) / dt + #print(joints.shape, init_vel.shape, middle_vel.shape, final_vel.shape) + vel_seq = torch.cat([init_vel, middle_vel, final_vel], dim=1).permute(1, 0).reshape(n, 55, 3) + #print(vel_seq.shape) + #.permute(1, 0).reshape(n, 55, 3) + vel_seq_np = vel_seq.cpu().numpy() + vel_joints_np = np.linalg.norm(vel_seq_np, axis=2) # n * 55 + all_list.append(vel_joints_np) + avg_vel = np.mean(np.concatenate(all_list, axis=0),axis=0) # 55 + np.save(save_path, avg_vel) + + + def build_cache(self, preloaded_dir): + logger.info(f"Audio bit rate: {self.args.audio_fps}") + logger.info("Reading data '{}'...".format(self.data_dir)) + logger.info("Creating the dataset cache...") + if self.args.new_cache: + if os.path.exists(preloaded_dir): + shutil.rmtree(preloaded_dir) + if os.path.exists(preloaded_dir): + logger.info("Found the cache {}".format(preloaded_dir)) + elif self.loader_type == "test": + self.cache_generation( + preloaded_dir, True, + 0, 0, + is_test=True) + else: + self.cache_generation( + preloaded_dir, self.args.disable_filtering, + self.args.clean_first_seconds, self.args.clean_final_seconds, + is_test=False) + + def __len__(self): + return self.n_samples + + + def cache_generation(self, out_lmdb_dir, disable_filtering, clean_first_seconds, clean_final_seconds, is_test=False): + # if "wav2vec2" in self.args.audio_rep: + # self.wav2vec_model = Wav2Vec2Model.from_pretrained(f"{self.args.data_path_1}/hub/transformer/wav2vec2-base-960h") + # self.wav2vec_model.feature_extractor._freeze_parameters() + # self.wav2vec_model = self.wav2vec_model.cuda() + # self.wav2vec_model.eval() + + self.n_out_samples = 0 + # create db for samples + if not os.path.exists(out_lmdb_dir): os.makedirs(out_lmdb_dir) + dst_lmdb_env = lmdb.open(out_lmdb_dir, map_size= int(1024 ** 3 * 50))# 50G + n_filtered_out = defaultdict(int) + + for index, file_name in self.selected_file.iterrows(): + f_name = file_name["id"] + ext = ".npz" if "smplx" in self.args.pose_rep else ".bvh" + pose_file = self.data_dir + self.args.pose_rep + "/" + f_name + ext + pose_each_file = [] + trans_each_file = [] + shape_each_file = [] + audio_each_file = [] + facial_each_file = [] + word_each_file = [] + emo_each_file = [] + sem_each_file = [] + vid_each_file = [] + id_pose = f_name #1_wayne_0_1_1 + + logger.info(colored(f"# ---- Building cache for Pose {id_pose} ---- #", "blue")) + if "smplx" in self.args.pose_rep: + pose_data = np.load(pose_file, allow_pickle=True) + assert 30%self.args.pose_fps == 0, 'pose_fps should be an aliquot part of 30' + stride = int(30/self.args.pose_fps) + pose_each_file = pose_data["poses"][::stride] * self.joint_mask + pose_each_file = pose_each_file[:, self.joint_mask.astype(bool)] + # print(pose_each_file.shape) + trans_each_file = pose_data["trans"][::stride] + shape_each_file = np.repeat(pose_data["betas"].reshape(1, 300), pose_each_file.shape[0], axis=0) + if self.args.facial_rep is not None: + logger.info(f"# ---- Building cache for Facial {id_pose} and Pose {id_pose} ---- #") + facial_each_file = pose_data["expressions"][::stride] + if self.args.facial_norm: + facial_each_file = (facial_each_file - self.mean_facial) / self.std_facial + + else: + assert 120%self.args.pose_fps == 0, 'pose_fps should be an aliquot part of 120' + stride = int(120/self.args.pose_fps) + with open(pose_file, "r") as pose_data: + for j, line in enumerate(pose_data.readlines()): + if j < 431: continue + if j%stride != 0:continue + data = np.fromstring(line, dtype=float, sep=" ") + rot_data = rc.euler_angles_to_matrix(torch.from_numpy(np.deg2rad(data)).reshape(-1, self.joints,3), "XYZ") + rot_data = rc.matrix_to_axis_angle(rot_data).reshape(-1, self.joints*3) + rot_data = rot_data.numpy() * self.joint_mask + + pose_each_file.append(rot_data) + trans_each_file.append(data[:3]) + + pose_each_file = np.array(pose_each_file) + # print(pose_each_file.shape) + trans_each_file = np.array(trans_each_file) + shape_each_file = np.repeat(np.array(-1).reshape(1, 1), pose_each_file.shape[0], axis=0) + if self.args.facial_rep is not None: + logger.info(f"# ---- Building cache for Facial {id_pose} and Pose {id_pose} ---- #") + facial_file = pose_file.replace(self.args.pose_rep, self.args.facial_rep).replace("bvh", "json") + assert 60%self.args.pose_fps == 0, 'pose_fps should be an aliquot part of 120' + stride = int(60/self.args.pose_fps) + if not os.path.exists(facial_file): + logger.warning(f"# ---- file not found for Facial {id_pose}, skip all files with the same id ---- #") + self.selected_file = self.selected_file.drop(self.selected_file[self.selected_file['id'] == id_pose].index) + continue + with open(facial_file, 'r') as facial_data_file: + facial_data = json.load(facial_data_file) + for j, frame_data in enumerate(facial_data['frames']): + if j%stride != 0:continue + facial_each_file.append(frame_data['weights']) + facial_each_file = np.array(facial_each_file) + if self.args.facial_norm: + facial_each_file = (facial_each_file - self.mean_facial) / self.std_facial + + if self.args.id_rep is not None: + vid_each_file = np.repeat(np.array(int(f_name.split("_")[0])-1).reshape(1, 1), pose_each_file.shape[0], axis=0) + + if self.args.audio_rep is not None: + logger.info(f"# ---- Building cache for Audio {id_pose} and Pose {id_pose} ---- #") + audio_file = pose_file.replace(self.args.pose_rep, 'wave16k').replace(ext, ".wav") + if not os.path.exists(audio_file): + logger.warning(f"# ---- file not found for Audio {id_pose}, skip all files with the same id ---- #") + self.selected_file = self.selected_file.drop(self.selected_file[self.selected_file['id'] == id_pose].index) + continue + audio_each_file, sr = librosa.load(audio_file) + audio_each_file = librosa.resample(audio_each_file, orig_sr=sr, target_sr=self.args.audio_sr) + if self.args.audio_rep == "onset+amplitude": + from numpy.lib import stride_tricks + frame_length = 1024 + # hop_length = 512 + shape = (audio_each_file.shape[-1] - frame_length + 1, frame_length) + strides = (audio_each_file.strides[-1], audio_each_file.strides[-1]) + rolling_view = stride_tricks.as_strided(audio_each_file, shape=shape, strides=strides) + amplitude_envelope = np.max(np.abs(rolling_view), axis=1) + # pad the last frame_length-1 samples + amplitude_envelope = np.pad(amplitude_envelope, (0, frame_length-1), mode='constant', constant_values=amplitude_envelope[-1]) + audio_onset_f = librosa.onset.onset_detect(y=audio_each_file, sr=self.args.audio_sr, units='frames') + onset_array = np.zeros(len(audio_each_file), dtype=float) + onset_array[audio_onset_f] = 1.0 + # print(amplitude_envelope.shape, audio_each_file.shape, onset_array.shape) + audio_each_file = np.concatenate([amplitude_envelope.reshape(-1, 1), onset_array.reshape(-1, 1)], axis=1) + elif self.args.audio_rep == "mfcc": + audio_each_file = librosa.feature.melspectrogram(y=audio_each_file, sr=self.args.audio_sr, n_mels=128, hop_length=int(self.args.audio_sr/self.args.audio_fps)) + audio_each_file = audio_each_file.transpose(1, 0) + # print(audio_each_file.shape, pose_each_file.shape) + if self.args.audio_norm and self.args.audio_rep == "wave16k": + audio_each_file = (audio_each_file - self.mean_audio) / self.std_audio + # print(audio_each_file.shape) + time_offset = 0 + if self.args.word_rep is not None: + logger.info(f"# ---- Building cache for Word {id_pose} and Pose {id_pose} ---- #") + word_file = f"{self.data_dir}{self.args.word_rep}/{id_pose}.TextGrid" + if not os.path.exists(word_file): + logger.warning(f"# ---- file not found for Word {id_pose}, skip all files with the same id ---- #") + self.selected_file = self.selected_file.drop(self.selected_file[self.selected_file['id'] == id_pose].index) + continue + tgrid = tg.TextGrid.fromFile(word_file) + if self.args.t_pre_encoder == "bert": + from transformers import AutoTokenizer, BertModel + tokenizer = AutoTokenizer.from_pretrained(self.args.data_path_1 + "hub/bert-base-uncased", local_files_only=True) + model = BertModel.from_pretrained(self.args.data_path_1 + "hub/bert-base-uncased", local_files_only=True).eval() + list_word = [] + all_hidden = [] + max_len = 400 + last = 0 + word_token_mapping = [] + first = True + for i, word in enumerate(tgrid[0]): + last = i + if (i%max_len != 0) or (i==0): + if word.mark == "": + list_word.append(".") + else: + list_word.append(word.mark) + else: + max_counter = max_len + str_word = ' '.join(map(str, list_word)) + if first: + global_len = 0 + end = -1 + offset_word = [] + for k, wordvalue in enumerate(list_word): + start = end+1 + end = start+len(wordvalue) + offset_word.append((start, end)) + #print(offset_word) + token_scan = tokenizer.encode_plus(str_word, return_offsets_mapping=True)['offset_mapping'] + #print(token_scan) + for start, end in offset_word: + sub_mapping = [] + for i, (start_t, end_t) in enumerate(token_scan[1:-1]): + if int(start) <= int(start_t) and int(end_t) <= int(end): + #print(i+global_len) + sub_mapping.append(i+global_len) + word_token_mapping.append(sub_mapping) + #print(len(word_token_mapping)) + global_len = word_token_mapping[-1][-1] + 1 + list_word = [] + if word.mark == "": + list_word.append(".") + else: + list_word.append(word.mark) + + with torch.no_grad(): + inputs = tokenizer(str_word, return_tensors="pt") + outputs = model(**inputs) + last_hidden_states = outputs.last_hidden_state.reshape(-1, 768).cpu().numpy()[1:-1, :] + all_hidden.append(last_hidden_states) + + #list_word = list_word[:10] + if list_word == []: + pass + else: + if first: + global_len = 0 + str_word = ' '.join(map(str, list_word)) + end = -1 + offset_word = [] + for k, wordvalue in enumerate(list_word): + start = end+1 + end = start+len(wordvalue) + offset_word.append((start, end)) + #print(offset_word) + token_scan = tokenizer.encode_plus(str_word, return_offsets_mapping=True)['offset_mapping'] + #print(token_scan) + for start, end in offset_word: + sub_mapping = [] + for i, (start_t, end_t) in enumerate(token_scan[1:-1]): + if int(start) <= int(start_t) and int(end_t) <= int(end): + sub_mapping.append(i+global_len) + #print(sub_mapping) + word_token_mapping.append(sub_mapping) + #print(len(word_token_mapping)) + with torch.no_grad(): + inputs = tokenizer(str_word, return_tensors="pt") + outputs = model(**inputs) + last_hidden_states = outputs.last_hidden_state.reshape(-1, 768).cpu().numpy()[1:-1, :] + all_hidden.append(last_hidden_states) + last_hidden_states = np.concatenate(all_hidden, axis=0) + + for i in range(pose_each_file.shape[0]): + found_flag = False + current_time = i/self.args.pose_fps + time_offset + j_last = 0 + for j, word in enumerate(tgrid[0]): + word_n, word_s, word_e = word.mark, word.minTime, word.maxTime + if word_s<=current_time and current_time<=word_e: + if self.args.word_cache and self.args.t_pre_encoder == 'bert': + mapping_index = word_token_mapping[j] + #print(mapping_index, word_s, word_e) + s_t = np.linspace(word_s, word_e, len(mapping_index)+1) + #print(s_t) + for tt, t_sep in enumerate(s_t[1:]): + if current_time <= t_sep: + #if len(mapping_index) > 1: print(mapping_index[tt]) + word_each_file.append(last_hidden_states[mapping_index[tt]]) + break + else: + if word_n == " ": + word_each_file.append(self.lang_model.PAD_token) + else: + word_each_file.append(self.lang_model.get_word_index(word_n)) + found_flag = True + j_last = j + break + else: continue + if not found_flag: + if self.args.word_cache and self.args.t_pre_encoder == 'bert': + word_each_file.append(last_hidden_states[j_last]) + else: + word_each_file.append(self.lang_model.UNK_token) + word_each_file = np.array(word_each_file) + #print(word_each_file.shape) + + if self.args.emo_rep is not None: + logger.info(f"# ---- Building cache for Emo {id_pose} and Pose {id_pose} ---- #") + rtype, start = int(id_pose.split('_')[3]), int(id_pose.split('_')[3]) + if rtype == 0 or rtype == 2 or rtype == 4 or rtype == 6: + if start >= 1 and start <= 64: + score = 0 + elif start >= 65 and start <= 72: + score = 1 + elif start >= 73 and start <= 80: + score = 2 + elif start >= 81 and start <= 86: + score = 3 + elif start >= 87 and start <= 94: + score = 4 + elif start >= 95 and start <= 102: + score = 5 + elif start >= 103 and start <= 110: + score = 6 + elif start >= 111 and start <= 118: + score = 7 + else: pass + else: + # you may denote as unknown in the future + score = 0 + emo_each_file = np.repeat(np.array(score).reshape(1, 1), pose_each_file.shape[0], axis=0) + #print(emo_each_file) + + if self.args.sem_rep is not None: + logger.info(f"# ---- Building cache for Sem {id_pose} and Pose {id_pose} ---- #") + sem_file = f"{self.data_dir}{self.args.sem_rep}/{id_pose}.txt" + sem_all = pd.read_csv(sem_file, + sep='\t', + names=["name", "start_time", "end_time", "duration", "score", "keywords"]) + # we adopt motion-level semantic score here. + for i in range(pose_each_file.shape[0]): + found_flag = False + for j, (start, end, score) in enumerate(zip(sem_all['start_time'],sem_all['end_time'], sem_all['score'])): + current_time = i/self.args.pose_fps + time_offset + if start<=current_time and current_time<=end: + sem_each_file.append(score) + found_flag=True + break + else: continue + if not found_flag: sem_each_file.append(0.) + sem_each_file = np.array(sem_each_file) + #print(sem_each_file) + + filtered_result = self._sample_from_clip( + dst_lmdb_env, + audio_each_file, pose_each_file, trans_each_file, shape_each_file, facial_each_file, word_each_file, + vid_each_file, emo_each_file, sem_each_file, + disable_filtering, clean_first_seconds, clean_final_seconds, is_test, + ) + for type in filtered_result.keys(): + n_filtered_out[type] += filtered_result[type] + + with dst_lmdb_env.begin() as txn: + logger.info(colored(f"no. of samples: {txn.stat()['entries']}", "cyan")) + n_total_filtered = 0 + for type, n_filtered in n_filtered_out.items(): + logger.info("{}: {}".format(type, n_filtered)) + n_total_filtered += n_filtered + logger.info(colored("no. of excluded samples: {} ({:.1f}%)".format( + n_total_filtered, 100 * n_total_filtered / (txn.stat()["entries"] + n_total_filtered)), "cyan")) + dst_lmdb_env.sync() + dst_lmdb_env.close() + + def _sample_from_clip( + self, dst_lmdb_env, audio_each_file, pose_each_file, trans_each_file, shape_each_file, facial_each_file, word_each_file, + vid_each_file, emo_each_file, sem_each_file, + disable_filtering, clean_first_seconds, clean_final_seconds, is_test, + ): + """ + for data cleaning, we ignore the data for first and final n s + for test, we return all data + """ + # audio_start = int(self.alignment[0] * self.args.audio_fps) + # pose_start = int(self.alignment[1] * self.args.pose_fps) + #logger.info(f"before: {audio_each_file.shape} {pose_each_file.shape}") + # audio_each_file = audio_each_file[audio_start:] + # pose_each_file = pose_each_file[pose_start:] + # trans_each_file = + #logger.info(f"after alignment: {audio_each_file.shape} {pose_each_file.shape}") + #print(pose_each_file.shape) + round_seconds_skeleton = pose_each_file.shape[0] // self.args.pose_fps # assume 1500 frames / 15 fps = 100 s + #print(round_seconds_skeleton) + if audio_each_file != []: + if self.args.audio_rep != "wave16k": + round_seconds_audio = len(audio_each_file) // self.args.audio_fps # assume 16,000,00 / 16,000 = 100 s + elif self.args.audio_rep == "mfcc": + round_seconds_audio = audio_each_file.shape[0] // self.args.audio_fps + else: + round_seconds_audio = audio_each_file.shape[0] // self.args.audio_sr + if facial_each_file != []: + round_seconds_facial = facial_each_file.shape[0] // self.args.pose_fps + logger.info(f"audio: {round_seconds_audio}s, pose: {round_seconds_skeleton}s, facial: {round_seconds_facial}s") + round_seconds_skeleton = min(round_seconds_audio, round_seconds_skeleton, round_seconds_facial) + max_round = max(round_seconds_audio, round_seconds_skeleton, round_seconds_facial) + if round_seconds_skeleton != max_round: + logger.warning(f"reduce to {round_seconds_skeleton}s, ignore {max_round-round_seconds_skeleton}s") + else: + logger.info(f"pose: {round_seconds_skeleton}s, audio: {round_seconds_audio}s") + round_seconds_skeleton = min(round_seconds_audio, round_seconds_skeleton) + max_round = max(round_seconds_audio, round_seconds_skeleton) + if round_seconds_skeleton != max_round: + logger.warning(f"reduce to {round_seconds_skeleton}s, ignore {max_round-round_seconds_skeleton}s") + + clip_s_t, clip_e_t = clean_first_seconds, round_seconds_skeleton - clean_final_seconds # assume [10, 90]s + clip_s_f_audio, clip_e_f_audio = self.args.audio_fps * clip_s_t, clip_e_t * self.args.audio_fps # [160,000,90*160,000] + clip_s_f_pose, clip_e_f_pose = clip_s_t * self.args.pose_fps, clip_e_t * self.args.pose_fps # [150,90*15] + + + for ratio in self.args.multi_length_training: + if is_test:# stride = length for test + cut_length = clip_e_f_pose - clip_s_f_pose + self.args.stride = cut_length + self.max_length = cut_length + else: + self.args.stride = int(ratio*self.ori_stride) + cut_length = int(self.ori_length*ratio) + + num_subdivision = math.floor((clip_e_f_pose - clip_s_f_pose - cut_length) / self.args.stride) + 1 + logger.info(f"pose from frame {clip_s_f_pose} to {clip_e_f_pose}, length {cut_length}") + logger.info(f"{num_subdivision} clips is expected with stride {self.args.stride}") + + if audio_each_file != []: + audio_short_length = math.floor(cut_length / self.args.pose_fps * self.args.audio_fps) + """ + for audio sr = 16000, fps = 15, pose_length = 34, + audio short length = 36266.7 -> 36266 + this error is fine. + """ + logger.info(f"audio from frame {clip_s_f_audio} to {clip_e_f_audio}, length {audio_short_length}") + + n_filtered_out = defaultdict(int) + sample_pose_list = [] + sample_audio_list = [] + sample_facial_list = [] + sample_shape_list = [] + sample_word_list = [] + sample_emo_list = [] + sample_sem_list = [] + sample_vid_list = [] + sample_trans_list = [] + + for i in range(num_subdivision): # cut into around 2s chip, (self npose) + start_idx = clip_s_f_pose + i * self.args.stride + fin_idx = start_idx + cut_length + sample_pose = pose_each_file[start_idx:fin_idx] + sample_trans = trans_each_file[start_idx:fin_idx] + sample_shape = shape_each_file[start_idx:fin_idx] + # print(sample_pose.shape) + if self.args.audio_rep is not None: + audio_start = clip_s_f_audio + math.floor(i * self.args.stride * self.args.audio_fps / self.args.pose_fps) + audio_end = audio_start + audio_short_length + sample_audio = audio_each_file[audio_start:audio_end] + else: + sample_audio = np.array([-1]) + sample_facial = facial_each_file[start_idx:fin_idx] if self.args.facial_rep is not None else np.array([-1]) + sample_word = word_each_file[start_idx:fin_idx] if self.args.word_rep is not None else np.array([-1]) + sample_emo = emo_each_file[start_idx:fin_idx] if self.args.emo_rep is not None else np.array([-1]) + sample_sem = sem_each_file[start_idx:fin_idx] if self.args.sem_rep is not None else np.array([-1]) + sample_vid = vid_each_file[start_idx:fin_idx] if self.args.id_rep is not None else np.array([-1]) + + if sample_pose.any() != None: + # filtering motion skeleton data + sample_pose, filtering_message = MotionPreprocessor(sample_pose).get() + is_correct_motion = (sample_pose != []) + if is_correct_motion or disable_filtering: + sample_pose_list.append(sample_pose) + sample_audio_list.append(sample_audio) + sample_facial_list.append(sample_facial) + sample_shape_list.append(sample_shape) + sample_word_list.append(sample_word) + sample_vid_list.append(sample_vid) + sample_emo_list.append(sample_emo) + sample_sem_list.append(sample_sem) + sample_trans_list.append(sample_trans) + else: + n_filtered_out[filtering_message] += 1 + + if len(sample_pose_list) > 0: + with dst_lmdb_env.begin(write=True) as txn: + for pose, audio, facial, shape, word, vid, emo, sem, trans in zip( + sample_pose_list, + sample_audio_list, + sample_facial_list, + sample_shape_list, + sample_word_list, + sample_vid_list, + sample_emo_list, + sample_sem_list, + sample_trans_list,): + k = "{:005}".format(self.n_out_samples).encode("ascii") + v = [pose, audio, facial, shape, word, emo, sem, vid, trans] + v = pickle.dumps(v,5) + txn.put(k, v) + self.n_out_samples += 1 + return n_filtered_out + + def __getitem__(self, idx): + with self.lmdb_env.begin(write=False) as txn: + key = "{:005}".format(idx).encode("ascii") + sample = txn.get(key) + sample = pickle.loads(sample) + tar_pose, in_audio, in_facial, in_shape, in_word, emo, sem, vid, trans = sample + #print(in_shape) + #vid = torch.from_numpy(vid).int() + emo = torch.from_numpy(emo).int() + sem = torch.from_numpy(sem).float() + in_audio = torch.from_numpy(in_audio).float() + in_word = torch.from_numpy(in_word).float() if self.args.word_cache else torch.from_numpy(in_word).int() + if self.loader_type == "test": + tar_pose = torch.from_numpy(tar_pose).float() + trans = torch.from_numpy(trans).float() + in_facial = torch.from_numpy(in_facial).float() + vid = torch.from_numpy(vid).float() + in_shape = torch.from_numpy(in_shape).float() + else: + in_shape = torch.from_numpy(in_shape).reshape((in_shape.shape[0], -1)).float() + trans = torch.from_numpy(trans).reshape((trans.shape[0], -1)).float() + vid = torch.from_numpy(vid).reshape((vid.shape[0], -1)).float() + tar_pose = torch.from_numpy(tar_pose).reshape((tar_pose.shape[0], -1)).float() + in_facial = torch.from_numpy(in_facial).reshape((in_facial.shape[0], -1)).float() + return {"pose":tar_pose, "audio":in_audio, "facial":in_facial, "beta": in_shape, "word":in_word, "id":vid, "emo":emo, "sem":sem, "trans":trans} + + +class MotionPreprocessor: + def __init__(self, skeletons): + self.skeletons = skeletons + #self.mean_pose = mean_pose + self.filtering_message = "PASS" + + def get(self): + assert (self.skeletons is not None) + + # filtering + if self.skeletons != []: + if self.check_pose_diff(): + self.skeletons = [] + self.filtering_message = "pose" + # elif self.check_spine_angle(): + # self.skeletons = [] + # self.filtering_message = "spine angle" + # elif self.check_static_motion(): + # self.skeletons = [] + # self.filtering_message = "motion" + + # if self.skeletons != []: + # self.skeletons = self.skeletons.tolist() + # for i, frame in enumerate(self.skeletons): + # assert not np.isnan(self.skeletons[i]).any() # missing joints + + return self.skeletons, self.filtering_message + + def check_static_motion(self, verbose=True): + def get_variance(skeleton, joint_idx): + wrist_pos = skeleton[:, joint_idx] + variance = np.sum(np.var(wrist_pos, axis=0)) + return variance + + left_arm_var = get_variance(self.skeletons, 6) + right_arm_var = get_variance(self.skeletons, 9) + + th = 0.0014 # exclude 13110 + # th = 0.002 # exclude 16905 + if left_arm_var < th and right_arm_var < th: + if verbose: + print("skip - check_static_motion left var {}, right var {}".format(left_arm_var, right_arm_var)) + return True + else: + if verbose: + print("pass - check_static_motion left var {}, right var {}".format(left_arm_var, right_arm_var)) + return False + + + def check_pose_diff(self, verbose=False): +# diff = np.abs(self.skeletons - self.mean_pose) # 186*1 +# diff = np.mean(diff) + +# # th = 0.017 +# th = 0.02 #0.02 # exclude 3594 +# if diff < th: +# if verbose: +# print("skip - check_pose_diff {:.5f}".format(diff)) +# return True +# # th = 3.5 #0.02 # exclude 3594 +# # if 3.5 < diff < 5: +# # if verbose: +# # print("skip - check_pose_diff {:.5f}".format(diff)) +# # return True +# else: +# if verbose: +# print("pass - check_pose_diff {:.5f}".format(diff)) + return False + + + def check_spine_angle(self, verbose=True): + def angle_between(v1, v2): + v1_u = v1 / np.linalg.norm(v1) + v2_u = v2 / np.linalg.norm(v2) + return np.arccos(np.clip(np.dot(v1_u, v2_u), -1.0, 1.0)) + + angles = [] + for i in range(self.skeletons.shape[0]): + spine_vec = self.skeletons[i, 1] - self.skeletons[i, 0] + angle = angle_between(spine_vec, [0, -1, 0]) + angles.append(angle) + + if np.rad2deg(max(angles)) > 30 or np.rad2deg(np.mean(angles)) > 20: # exclude 4495 + # if np.rad2deg(max(angles)) > 20: # exclude 8270 + if verbose: + print("skip - check_spine_angle {:.5f}, {:.5f}".format(max(angles), np.mean(angles))) + return True + else: + if verbose: + print("pass - check_spine_angle {:.5f}".format(max(angles))) + return False \ No newline at end of file diff --git a/dataloaders/beat_sep_lower.py b/dataloaders/beat_sep_lower.py new file mode 100644 index 0000000000000000000000000000000000000000..1f78fd2aafaa63154861b7af5ecbed41a0f51113 --- /dev/null +++ b/dataloaders/beat_sep_lower.py @@ -0,0 +1,876 @@ +import os +import pickle +import math +import shutil +import numpy as np +import lmdb as lmdb +import textgrid as tg +import pandas as pd +import torch +import glob +import json +from termcolor import colored +from loguru import logger +from collections import defaultdict +from torch.utils.data import Dataset +import torch.distributed as dist +#import pyarrow +import pickle +import librosa +import smplx + +from .build_vocab import Vocab +from .utils.audio_features import Wav2Vec2Model +from .data_tools import joints_list +from .utils import rotation_conversions as rc +from .utils import other_tools + +class CustomDataset(Dataset): + def __init__(self, args, loader_type, augmentation=None, kwargs=None, build_cache=True): + self.args = args + self.loader_type = loader_type + + self.rank = dist.get_rank() + self.ori_stride = self.args.stride + self.ori_length = self.args.pose_length + self.alignment = [0,0] # for trinity + + self.ori_joint_list = joints_list[self.args.ori_joints] + self.tar_joint_list = joints_list[self.args.tar_joints] + if 'smplx' in self.args.pose_rep: + self.joint_mask = np.zeros(len(list(self.ori_joint_list.keys()))*3) + self.joints = len(list(self.tar_joint_list.keys())) + for joint_name in self.tar_joint_list: + self.joint_mask[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + else: + self.joints = len(list(self.ori_joint_list.keys()))+1 + self.joint_mask = np.zeros(self.joints*3) + for joint_name in self.tar_joint_list: + if joint_name == "Hips": + self.joint_mask[3:6] = 1 + else: + self.joint_mask[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + # select trainable joints + self.smplx = smplx.create( + self.args.data_path_1+"smplx_models/", + model_type='smplx', + gender='NEUTRAL_2020', + use_face_contour=False, + num_betas=300, + num_expression_coeffs=100, + ext='npz', + use_pca=False, + ).cuda().eval() + + split_rule = pd.read_csv(args.data_path+"train_test_split.csv") + self.selected_file = split_rule.loc[(split_rule['type'] == loader_type) & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + if args.additional_data and loader_type == 'train': + split_b = split_rule.loc[(split_rule['type'] == 'additional') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + #self.selected_file = split_rule.loc[(split_rule['type'] == 'additional') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + self.selected_file = pd.concat([self.selected_file, split_b]) + if self.selected_file.empty: + logger.warning(f"{loader_type} is empty for speaker {self.args.training_speakers}, use train set 0-8 instead") + self.selected_file = split_rule.loc[(split_rule['type'] == 'train') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + self.selected_file = self.selected_file.iloc[0:8] + self.data_dir = args.data_path + + if loader_type == "test": + self.args.multi_length_training = [1.0] + self.max_length = int(args.pose_length * self.args.multi_length_training[-1]) + self.max_audio_pre_len = math.floor(args.pose_length / args.pose_fps * self.args.audio_sr) + if self.max_audio_pre_len > self.args.test_length*self.args.audio_sr: + self.max_audio_pre_len = self.args.test_length*self.args.audio_sr + + if args.word_rep is not None: + with open(f"{args.data_path}weights/vocab.pkl", 'rb') as f: + self.lang_model = pickle.load(f) + + preloaded_dir = self.args.root_path + self.args.cache_path + loader_type + f"/{args.pose_rep}_cache" + # if args.pose_norm: + # # careful for rotation vectors + # if not os.path.exists(args.data_path+args.mean_pose_path+f"{args.pose_rep.split('_')[0]}/bvh_mean.npy"): + # self.calculate_mean_pose() + # self.mean_pose = np.load(args.data_path+args.mean_pose_path+f"{args.pose_rep.split('_')[0]}/bvh_mean.npy") + # self.std_pose = np.load(args.data_path+args.mean_pose_path+f"{args.pose_rep.split('_')[0]}/bvh_std.npy") + # if args.audio_norm: + # if not os.path.exists(args.data_path+args.mean_pose_path+f"{args.audio_rep.split('_')[0]}/bvh_mean.npy"): + # self.calculate_mean_audio() + # self.mean_audio = np.load(args.data_path+args.mean_pose_path+f"{args.audio_rep.split('_')[0]}/npy_mean.npy") + # self.std_audio = np.load(args.data_path+args.mean_pose_path+f"{args.audio_rep.split('_')[0]}/npy_std.npy") + # if args.facial_norm: + # if not os.path.exists(args.data_path+args.mean_pose_path+f"{args.pose_rep.split('_')[0]}/bvh_mean.npy"): + # self.calculate_mean_face() + # self.mean_facial = np.load(args.data_path+args.mean_pose_path+f"{args.facial_rep}/json_mean.npy") + # self.std_facial = np.load(args.data_path+args.mean_pose_path+f"{args.facial_rep}/json_std.npy") + if self.args.beat_align: + if not os.path.exists(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy"): + self.calculate_mean_velocity(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy") + self.avg_vel = np.load(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy") + + if build_cache and self.rank == 0: + self.build_cache(preloaded_dir) + self.lmdb_env = lmdb.open(preloaded_dir, readonly=True, lock=False) + with self.lmdb_env.begin() as txn: + self.n_samples = txn.stat()["entries"] + + + def calculate_mean_velocity(self, save_path): + self.smplx = smplx.create( + self.args.data_path_1+"smplx_models/", + model_type='smplx', + gender='NEUTRAL_2020', + use_face_contour=False, + num_betas=300, + num_expression_coeffs=100, + ext='npz', + use_pca=False, + ).cuda().eval() + dir_p = self.data_dir + self.args.pose_rep + "/" + all_list = [] + from tqdm import tqdm + for tar in tqdm(os.listdir(dir_p)): + if tar.endswith(".npz"): + m_data = np.load(dir_p+tar, allow_pickle=True) + betas, poses, trans, exps = m_data["betas"], m_data["poses"], m_data["trans"], m_data["expressions"] + n, c = poses.shape[0], poses.shape[1] + betas = betas.reshape(1, 300) + betas = np.tile(betas, (n, 1)) + betas = torch.from_numpy(betas).cuda().float() + poses = torch.from_numpy(poses.reshape(n, c)).cuda().float() + exps = torch.from_numpy(exps.reshape(n, 100)).cuda().float() + trans = torch.from_numpy(trans.reshape(n, 3)).cuda().float() + max_length = 128 + s, r = n//max_length, n%max_length + #print(n, s, r) + all_tensor = [] + for i in range(s): + with torch.no_grad(): + joints = self.smplx( + betas=betas[i*max_length:(i+1)*max_length], + transl=trans[i*max_length:(i+1)*max_length], + expression=exps[i*max_length:(i+1)*max_length], + jaw_pose=poses[i*max_length:(i+1)*max_length, 66:69], + global_orient=poses[i*max_length:(i+1)*max_length,:3], + body_pose=poses[i*max_length:(i+1)*max_length,3:21*3+3], + left_hand_pose=poses[i*max_length:(i+1)*max_length,25*3:40*3], + right_hand_pose=poses[i*max_length:(i+1)*max_length,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[i*max_length:(i+1)*max_length, 69:72], + reye_pose=poses[i*max_length:(i+1)*max_length, 72:75], + )['joints'][:, :55, :].reshape(max_length, 55*3) + all_tensor.append(joints) + if r != 0: + with torch.no_grad(): + joints = self.smplx( + betas=betas[s*max_length:s*max_length+r], + transl=trans[s*max_length:s*max_length+r], + expression=exps[s*max_length:s*max_length+r], + jaw_pose=poses[s*max_length:s*max_length+r, 66:69], + global_orient=poses[s*max_length:s*max_length+r,:3], + body_pose=poses[s*max_length:s*max_length+r,3:21*3+3], + left_hand_pose=poses[s*max_length:s*max_length+r,25*3:40*3], + right_hand_pose=poses[s*max_length:s*max_length+r,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[s*max_length:s*max_length+r, 69:72], + reye_pose=poses[s*max_length:s*max_length+r, 72:75], + )['joints'][:, :55, :].reshape(r, 55*3) + all_tensor.append(joints) + joints = torch.cat(all_tensor, axis=0) + joints = joints.permute(1, 0) + dt = 1/30 + # first steps is forward diff (t+1 - t) / dt + init_vel = (joints[:, 1:2] - joints[:, :1]) / dt + # middle steps are second order (t+1 - t-1) / 2dt + middle_vel = (joints[:, 2:] - joints[:, 0:-2]) / (2 * dt) + # last step is backward diff (t - t-1) / dt + final_vel = (joints[:, -1:] - joints[:, -2:-1]) / dt + #print(joints.shape, init_vel.shape, middle_vel.shape, final_vel.shape) + vel_seq = torch.cat([init_vel, middle_vel, final_vel], dim=1).permute(1, 0).reshape(n, 55, 3) + #print(vel_seq.shape) + #.permute(1, 0).reshape(n, 55, 3) + vel_seq_np = vel_seq.cpu().numpy() + vel_joints_np = np.linalg.norm(vel_seq_np, axis=2) # n * 55 + all_list.append(vel_joints_np) + avg_vel = np.mean(np.concatenate(all_list, axis=0),axis=0) # 55 + np.save(save_path, avg_vel) + + + def build_cache(self, preloaded_dir): + logger.info(f"Audio bit rate: {self.args.audio_fps}") + logger.info("Reading data '{}'...".format(self.data_dir)) + logger.info("Creating the dataset cache...") + if self.args.new_cache: + if os.path.exists(preloaded_dir): + shutil.rmtree(preloaded_dir) + if os.path.exists(preloaded_dir): + logger.info("Found the cache {}".format(preloaded_dir)) + elif self.loader_type == "test": + self.cache_generation( + preloaded_dir, True, + 0, 0, + is_test=True) + else: + self.cache_generation( + preloaded_dir, self.args.disable_filtering, + self.args.clean_first_seconds, self.args.clean_final_seconds, + is_test=False) + + def __len__(self): + return self.n_samples + + + def cache_generation(self, out_lmdb_dir, disable_filtering, clean_first_seconds, clean_final_seconds, is_test=False): + # if "wav2vec2" in self.args.audio_rep: + # self.wav2vec_model = Wav2Vec2Model.from_pretrained(f"{self.args.data_path_1}/hub/transformer/wav2vec2-base-960h") + # self.wav2vec_model.feature_extractor._freeze_parameters() + # self.wav2vec_model = self.wav2vec_model.cuda() + # self.wav2vec_model.eval() + + self.n_out_samples = 0 + # create db for samples + if not os.path.exists(out_lmdb_dir): os.makedirs(out_lmdb_dir) + dst_lmdb_env = lmdb.open(out_lmdb_dir, map_size= int(1024 ** 3 * 500))# 500G + n_filtered_out = defaultdict(int) + + for index, file_name in self.selected_file.iterrows(): + f_name = file_name["id"] + ext = ".npz" if "smplx" in self.args.pose_rep else ".bvh" + pose_file = self.data_dir + self.args.pose_rep + "/" + f_name + ext + pose_each_file = [] + trans_each_file = [] + trans_v_each_file = [] + shape_each_file = [] + audio_each_file = [] + facial_each_file = [] + word_each_file = [] + emo_each_file = [] + sem_each_file = [] + vid_each_file = [] + id_pose = f_name #1_wayne_0_1_1 + + logger.info(colored(f"# ---- Building cache for Pose {id_pose} ---- #", "blue")) + if "smplx" in self.args.pose_rep: + pose_data = np.load(pose_file, allow_pickle=True) + assert 30%self.args.pose_fps == 0, 'pose_fps should be an aliquot part of 30' + stride = int(30/self.args.pose_fps) + pose_each_file = pose_data["poses"][::stride] + trans_each_file = pose_data["trans"][::stride] + trans_each_file[:,0] = trans_each_file[:,0] - trans_each_file[0,0] + trans_each_file[:,2] = trans_each_file[:,2] - trans_each_file[0,2] + trans_v_each_file = np.zeros_like(trans_each_file) + trans_v_each_file[1:,0] = trans_each_file[1:,0] - trans_each_file[:-1,0] + trans_v_each_file[0,0] = trans_v_each_file[1,0] + trans_v_each_file[1:,2] = trans_each_file[1:,2] - trans_each_file[:-1,2] + trans_v_each_file[0,2] = trans_v_each_file[1,2] + trans_v_each_file[:,1] = trans_each_file[:,1] + shape_each_file = np.repeat(pose_data["betas"].reshape(1, 300), pose_each_file.shape[0], axis=0) + + assert self.args.pose_fps == 30, "should 30" + m_data = np.load(pose_file, allow_pickle=True) + betas, poses, trans, exps = m_data["betas"], m_data["poses"], m_data["trans"], m_data["expressions"] + n, c = poses.shape[0], poses.shape[1] + betas = betas.reshape(1, 300) + betas = np.tile(betas, (n, 1)) + betas = torch.from_numpy(betas).cuda().float() + poses = torch.from_numpy(poses.reshape(n, c)).cuda().float() + exps = torch.from_numpy(exps.reshape(n, 100)).cuda().float() + trans = torch.from_numpy(trans.reshape(n, 3)).cuda().float() + max_length = 128 # 为什么这里需要一个max_length + s, r = n//max_length, n%max_length + #print(n, s, r) + all_tensor = [] + for i in range(s): + with torch.no_grad(): + joints = self.smplx( + betas=betas[i*max_length:(i+1)*max_length], + transl=trans[i*max_length:(i+1)*max_length], + expression=exps[i*max_length:(i+1)*max_length], + jaw_pose=poses[i*max_length:(i+1)*max_length, 66:69], + global_orient=poses[i*max_length:(i+1)*max_length,:3], + body_pose=poses[i*max_length:(i+1)*max_length,3:21*3+3], + left_hand_pose=poses[i*max_length:(i+1)*max_length,25*3:40*3], + right_hand_pose=poses[i*max_length:(i+1)*max_length,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[i*max_length:(i+1)*max_length, 69:72], + reye_pose=poses[i*max_length:(i+1)*max_length, 72:75], + )['joints'][:, (7,8,10,11), :].reshape(max_length, 4, 3).cpu() + all_tensor.append(joints) + if r != 0: + with torch.no_grad(): + joints = self.smplx( + betas=betas[s*max_length:s*max_length+r], + transl=trans[s*max_length:s*max_length+r], + expression=exps[s*max_length:s*max_length+r], + jaw_pose=poses[s*max_length:s*max_length+r, 66:69], + global_orient=poses[s*max_length:s*max_length+r,:3], + body_pose=poses[s*max_length:s*max_length+r,3:21*3+3], + left_hand_pose=poses[s*max_length:s*max_length+r,25*3:40*3], + right_hand_pose=poses[s*max_length:s*max_length+r,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[s*max_length:s*max_length+r, 69:72], + reye_pose=poses[s*max_length:s*max_length+r, 72:75], + )['joints'][:, (7,8,10,11), :].reshape(r, 4, 3).cpu() + all_tensor.append(joints) + joints = torch.cat(all_tensor, axis=0) # all, 4, 3 + # print(joints.shape) + feetv = torch.zeros(joints.shape[1], joints.shape[0]) + joints = joints.permute(1, 0, 2) + #print(joints.shape, feetv.shape) + feetv[:, :-1] = (joints[:, 1:] - joints[:, :-1]).norm(dim=-1) + #print(feetv.shape) + contacts = (feetv < 0.01).numpy().astype(float) + # print(contacts.shape, contacts) + contacts = contacts.transpose(1, 0) + pose_each_file = pose_each_file * self.joint_mask + pose_each_file = pose_each_file[:, self.joint_mask.astype(bool)] + pose_each_file = np.concatenate([pose_each_file, contacts], axis=1) + # print(pose_each_file.shape) + + + if self.args.facial_rep is not None: + logger.info(f"# ---- Building cache for Facial {id_pose} and Pose {id_pose} ---- #") + facial_each_file = pose_data["expressions"][::stride] + if self.args.facial_norm: + facial_each_file = (facial_each_file - self.mean_facial) / self.std_facial + + else: + assert 120%self.args.pose_fps == 0, 'pose_fps should be an aliquot part of 120' + stride = int(120/self.args.pose_fps) + with open(pose_file, "r") as pose_data: + for j, line in enumerate(pose_data.readlines()): + if j < 431: continue + if j%stride != 0:continue + data = np.fromstring(line, dtype=float, sep=" ") + rot_data = rc.euler_angles_to_matrix(torch.from_numpy(np.deg2rad(data)).reshape(-1, self.joints,3), "XYZ") + rot_data = rc.matrix_to_axis_angle(rot_data).reshape(-1, self.joints*3) + rot_data = rot_data.numpy() * self.joint_mask + + pose_each_file.append(rot_data) + trans_each_file.append(data[:3]) + + pose_each_file = np.array(pose_each_file) + # print(pose_each_file.shape) + trans_each_file = np.array(trans_each_file) + shape_each_file = np.repeat(np.array(-1).reshape(1, 1), pose_each_file.shape[0], axis=0) + if self.args.facial_rep is not None: + logger.info(f"# ---- Building cache for Facial {id_pose} and Pose {id_pose} ---- #") + facial_file = pose_file.replace(self.args.pose_rep, self.args.facial_rep).replace("bvh", "json") + assert 60%self.args.pose_fps == 0, 'pose_fps should be an aliquot part of 120' + stride = int(60/self.args.pose_fps) + if not os.path.exists(facial_file): + logger.warning(f"# ---- file not found for Facial {id_pose}, skip all files with the same id ---- #") + self.selected_file = self.selected_file.drop(self.selected_file[self.selected_file['id'] == id_pose].index) + continue + with open(facial_file, 'r') as facial_data_file: + facial_data = json.load(facial_data_file) + for j, frame_data in enumerate(facial_data['frames']): + if j%stride != 0:continue + facial_each_file.append(frame_data['weights']) + facial_each_file = np.array(facial_each_file) + if self.args.facial_norm: + facial_each_file = (facial_each_file - self.mean_facial) / self.std_facial + + if self.args.id_rep is not None: + vid_each_file = np.repeat(np.array(int(f_name.split("_")[0])-1).reshape(1, 1), pose_each_file.shape[0], axis=0) + + if self.args.audio_rep is not None: + logger.info(f"# ---- Building cache for Audio {id_pose} and Pose {id_pose} ---- #") + audio_file = pose_file.replace(self.args.pose_rep, 'wave16k').replace(ext, ".wav") + if not os.path.exists(audio_file): + logger.warning(f"# ---- file not found for Audio {id_pose}, skip all files with the same id ---- #") + self.selected_file = self.selected_file.drop(self.selected_file[self.selected_file['id'] == id_pose].index) + continue + audio_save_path = audio_file.replace("wave16k", "onset_amplitude").replace(".wav", ".npy") + if self.args.audio_rep == "onset+amplitude" and os.path.exists(audio_save_path): + audio_each_file = np.load(audio_save_path) + logger.warning(f"# ---- file found cache for Audio {id_pose} ---- #") + elif self.args.audio_rep == "onset+amplitude": + audio_each_file, sr = librosa.load(audio_file) + audio_each_file = librosa.resample(audio_each_file, orig_sr=sr, target_sr=self.args.audio_sr) + from numpy.lib import stride_tricks + frame_length = 1024 + # hop_length = 512 + shape = (audio_each_file.shape[-1] - frame_length + 1, frame_length) + strides = (audio_each_file.strides[-1], audio_each_file.strides[-1]) + rolling_view = stride_tricks.as_strided(audio_each_file, shape=shape, strides=strides) + amplitude_envelope = np.max(np.abs(rolling_view), axis=1) + # pad the last frame_length-1 samples + amplitude_envelope = np.pad(amplitude_envelope, (0, frame_length-1), mode='constant', constant_values=amplitude_envelope[-1]) + audio_onset_f = librosa.onset.onset_detect(y=audio_each_file, sr=self.args.audio_sr, units='frames') + onset_array = np.zeros(len(audio_each_file), dtype=float) + onset_array[audio_onset_f] = 1.0 + # print(amplitude_envelope.shape, audio_each_file.shape, onset_array.shape) + audio_each_file = np.concatenate([amplitude_envelope.reshape(-1, 1), onset_array.reshape(-1, 1)], axis=1) + audio_save_path = audio_file.replace("wave16k", "onset_amplitude").replace(".wav", ".npy") + np.save(audio_save_path, audio_each_file) + + elif self.args.audio_rep == "mfcc": + audio_each_file = librosa.feature.melspectrogram(y=audio_each_file, sr=self.args.audio_sr, n_mels=128, hop_length=int(self.args.audio_sr/self.args.audio_fps)) + audio_each_file = audio_each_file.transpose(1, 0) + # print(audio_each_file.shape, pose_each_file.shape) + if self.args.audio_norm and self.args.audio_rep == "wave16k": + audio_each_file = (audio_each_file - self.mean_audio) / self.std_audio + + time_offset = 0 + if self.args.word_rep is not None: + logger.info(f"# ---- Building cache for Word {id_pose} and Pose {id_pose} ---- #") + word_file = f"{self.data_dir}{self.args.word_rep}/{id_pose}.TextGrid" + if not os.path.exists(word_file): + logger.warning(f"# ---- file not found for Word {id_pose}, skip all files with the same id ---- #") + self.selected_file = self.selected_file.drop(self.selected_file[self.selected_file['id'] == id_pose].index) + continue + word_save_path = f"{self.data_dir}{self.args.t_pre_encoder}/{id_pose}.npy" + if os.path.exists(word_save_path): + word_each_file = np.load(word_save_path) + logger.warning(f"# ---- file found cache for Word {id_pose} ---- #") + else: + tgrid = tg.TextGrid.fromFile(word_file) + if self.args.t_pre_encoder == "bert": + from transformers import AutoTokenizer, BertModel + tokenizer = AutoTokenizer.from_pretrained(self.args.data_path_1 + "hub/bert-base-uncased", local_files_only=True) + model = BertModel.from_pretrained(self.args.data_path_1 + "hub/bert-base-uncased", local_files_only=True).eval() + list_word = [] + all_hidden = [] + max_len = 400 + last = 0 + word_token_mapping = [] + first = True + for i, word in enumerate(tgrid[0]): + last = i + if (i%max_len != 0) or (i==0): + if word.mark == "": + list_word.append(".") + else: + list_word.append(word.mark) + else: + max_counter = max_len + str_word = ' '.join(map(str, list_word)) + if first: + global_len = 0 + end = -1 + offset_word = [] + for k, wordvalue in enumerate(list_word): + start = end+1 + end = start+len(wordvalue) + offset_word.append((start, end)) + #print(offset_word) + token_scan = tokenizer.encode_plus(str_word, return_offsets_mapping=True)['offset_mapping'] + #print(token_scan) + for start, end in offset_word: + sub_mapping = [] + for i, (start_t, end_t) in enumerate(token_scan[1:-1]): + if int(start) <= int(start_t) and int(end_t) <= int(end): + #print(i+global_len) + sub_mapping.append(i+global_len) + word_token_mapping.append(sub_mapping) + #print(len(word_token_mapping)) + global_len = word_token_mapping[-1][-1] + 1 + list_word = [] + if word.mark == "": + list_word.append(".") + else: + list_word.append(word.mark) + + with torch.no_grad(): + inputs = tokenizer(str_word, return_tensors="pt") + outputs = model(**inputs) + last_hidden_states = outputs.last_hidden_state.reshape(-1, 768).cpu().numpy()[1:-1, :] + all_hidden.append(last_hidden_states) + + #list_word = list_word[:10] + if list_word == []: + pass + else: + if first: + global_len = 0 + str_word = ' '.join(map(str, list_word)) + end = -1 + offset_word = [] + for k, wordvalue in enumerate(list_word): + start = end+1 + end = start+len(wordvalue) + offset_word.append((start, end)) + #print(offset_word) + token_scan = tokenizer.encode_plus(str_word, return_offsets_mapping=True)['offset_mapping'] + #print(token_scan) + for start, end in offset_word: + sub_mapping = [] + for i, (start_t, end_t) in enumerate(token_scan[1:-1]): + if int(start) <= int(start_t) and int(end_t) <= int(end): + sub_mapping.append(i+global_len) + #print(sub_mapping) + word_token_mapping.append(sub_mapping) + #print(len(word_token_mapping)) + with torch.no_grad(): + inputs = tokenizer(str_word, return_tensors="pt") + outputs = model(**inputs) + last_hidden_states = outputs.last_hidden_state.reshape(-1, 768).cpu().numpy()[1:-1, :] + all_hidden.append(last_hidden_states) + last_hidden_states = np.concatenate(all_hidden, axis=0) + + for i in range(pose_each_file.shape[0]): + found_flag = False + current_time = i/self.args.pose_fps + time_offset + j_last = 0 + for j, word in enumerate(tgrid[0]): + word_n, word_s, word_e = word.mark, word.minTime, word.maxTime + if word_s<=current_time and current_time<=word_e: + if self.args.word_cache and self.args.t_pre_encoder == 'bert': + mapping_index = word_token_mapping[j] + #print(mapping_index, word_s, word_e) + s_t = np.linspace(word_s, word_e, len(mapping_index)+1) + #print(s_t) + for tt, t_sep in enumerate(s_t[1:]): + if current_time <= t_sep: + #if len(mapping_index) > 1: print(mapping_index[tt]) + word_each_file.append(last_hidden_states[mapping_index[tt]]) + break + else: + if word_n == " ": + word_each_file.append(self.lang_model.PAD_token) + else: + word_each_file.append(self.lang_model.get_word_index(word_n)) + found_flag = True + j_last = j + break + else: continue + if not found_flag: + if self.args.word_cache and self.args.t_pre_encoder == 'bert': + word_each_file.append(last_hidden_states[j_last]) + else: + word_each_file.append(self.lang_model.UNK_token) + word_each_file = np.array(word_each_file) + word_save_path = f"{self.data_dir}{self.args.t_pre_encoder}/{id_pose}.npy" + np.save(word_save_path, word_each_file) + #print(word_each_file.shape) + #print(word_each_file.shape) + + if self.args.emo_rep is not None: + logger.info(f"# ---- Building cache for Emo {id_pose} and Pose {id_pose} ---- #") + rtype, start = int(id_pose.split('_')[3]), int(id_pose.split('_')[3]) + if rtype == 0 or rtype == 2 or rtype == 4 or rtype == 6: + if start >= 1 and start <= 64: + score = 0 + elif start >= 65 and start <= 72: + score = 1 + elif start >= 73 and start <= 80: + score = 2 + elif start >= 81 and start <= 86: + score = 3 + elif start >= 87 and start <= 94: + score = 4 + elif start >= 95 and start <= 102: + score = 5 + elif start >= 103 and start <= 110: + score = 6 + elif start >= 111 and start <= 118: + score = 7 + else: pass + else: + # you may denote as unknown in the future + score = 0 + emo_each_file = np.repeat(np.array(score).reshape(1, 1), pose_each_file.shape[0], axis=0) + #print(emo_each_file) + + if self.args.sem_rep is not None: + logger.info(f"# ---- Building cache for Sem {id_pose} and Pose {id_pose} ---- #") + sem_file = f"{self.data_dir}{self.args.sem_rep}/{id_pose}.txt" + sem_all = pd.read_csv(sem_file, + sep='\t', + names=["name", "start_time", "end_time", "duration", "score", "keywords"]) + # we adopt motion-level semantic score here. + for i in range(pose_each_file.shape[0]): + found_flag = False + for j, (start, end, score) in enumerate(zip(sem_all['start_time'],sem_all['end_time'], sem_all['score'])): + current_time = i/self.args.pose_fps + time_offset + if start<=current_time and current_time<=end: + sem_each_file.append(score) + found_flag=True + break + else: continue + if not found_flag: sem_each_file.append(0.) + sem_each_file = np.array(sem_each_file) + #print(sem_each_file) + + filtered_result = self._sample_from_clip( + dst_lmdb_env, + audio_each_file, pose_each_file, trans_each_file, trans_v_each_file,shape_each_file, facial_each_file, word_each_file, + vid_each_file, emo_each_file, sem_each_file, + disable_filtering, clean_first_seconds, clean_final_seconds, is_test, + ) + for type in filtered_result.keys(): + n_filtered_out[type] += filtered_result[type] + + with dst_lmdb_env.begin() as txn: + logger.info(colored(f"no. of samples: {txn.stat()['entries']}", "cyan")) + n_total_filtered = 0 + for type, n_filtered in n_filtered_out.items(): + logger.info("{}: {}".format(type, n_filtered)) + n_total_filtered += n_filtered + logger.info(colored("no. of excluded samples: {} ({:.1f}%)".format( + n_total_filtered, 100 * n_total_filtered / (txn.stat()["entries"] + n_total_filtered)), "cyan")) + dst_lmdb_env.sync() + dst_lmdb_env.close() + + def _sample_from_clip( + self, dst_lmdb_env, audio_each_file, pose_each_file, trans_each_file, trans_v_each_file,shape_each_file, facial_each_file, word_each_file, + vid_each_file, emo_each_file, sem_each_file, + disable_filtering, clean_first_seconds, clean_final_seconds, is_test, + ): + """ + for data cleaning, we ignore the data for first and final n s + for test, we return all data + """ + # audio_start = int(self.alignment[0] * self.args.audio_fps) + # pose_start = int(self.alignment[1] * self.args.pose_fps) + #logger.info(f"before: {audio_each_file.shape} {pose_each_file.shape}") + # audio_each_file = audio_each_file[audio_start:] + # pose_each_file = pose_each_file[pose_start:] + # trans_each_file = + #logger.info(f"after alignment: {audio_each_file.shape} {pose_each_file.shape}") + #print(pose_each_file.shape) + round_seconds_skeleton = pose_each_file.shape[0] // self.args.pose_fps # assume 1500 frames / 15 fps = 100 s + #print(round_seconds_skeleton) + if audio_each_file != []: + if self.args.audio_rep != "wave16k": + round_seconds_audio = len(audio_each_file) // self.args.audio_fps # assume 16,000,00 / 16,000 = 100 s + elif self.args.audio_rep == "mfcc": + round_seconds_audio = audio_each_file.shape[0] // self.args.audio_fps + else: + round_seconds_audio = audio_each_file.shape[0] // self.args.audio_sr + if facial_each_file != []: + round_seconds_facial = facial_each_file.shape[0] // self.args.pose_fps + logger.info(f"audio: {round_seconds_audio}s, pose: {round_seconds_skeleton}s, facial: {round_seconds_facial}s") + round_seconds_skeleton = min(round_seconds_audio, round_seconds_skeleton, round_seconds_facial) + max_round = max(round_seconds_audio, round_seconds_skeleton, round_seconds_facial) + if round_seconds_skeleton != max_round: + logger.warning(f"reduce to {round_seconds_skeleton}s, ignore {max_round-round_seconds_skeleton}s") + else: + logger.info(f"pose: {round_seconds_skeleton}s, audio: {round_seconds_audio}s") + round_seconds_skeleton = min(round_seconds_audio, round_seconds_skeleton) + max_round = max(round_seconds_audio, round_seconds_skeleton) + if round_seconds_skeleton != max_round: + logger.warning(f"reduce to {round_seconds_skeleton}s, ignore {max_round-round_seconds_skeleton}s") + + clip_s_t, clip_e_t = clean_first_seconds, round_seconds_skeleton - clean_final_seconds # assume [10, 90]s + clip_s_f_audio, clip_e_f_audio = self.args.audio_fps * clip_s_t, clip_e_t * self.args.audio_fps # [160,000,90*160,000] + clip_s_f_pose, clip_e_f_pose = clip_s_t * self.args.pose_fps, clip_e_t * self.args.pose_fps # [150,90*15] + + + for ratio in self.args.multi_length_training: + if is_test:# stride = length for test + cut_length = clip_e_f_pose - clip_s_f_pose + self.args.stride = cut_length + self.max_length = cut_length + else: + self.args.stride = int(ratio*self.ori_stride) + cut_length = int(self.ori_length*ratio) + + num_subdivision = math.floor((clip_e_f_pose - clip_s_f_pose - cut_length) / self.args.stride) + 1 + logger.info(f"pose from frame {clip_s_f_pose} to {clip_e_f_pose}, length {cut_length}") + logger.info(f"{num_subdivision} clips is expected with stride {self.args.stride}") + + if audio_each_file != []: + audio_short_length = math.floor(cut_length / self.args.pose_fps * self.args.audio_fps) + """ + for audio sr = 16000, fps = 15, pose_length = 34, + audio short length = 36266.7 -> 36266 + this error is fine. + """ + logger.info(f"audio from frame {clip_s_f_audio} to {clip_e_f_audio}, length {audio_short_length}") + + n_filtered_out = defaultdict(int) + sample_pose_list = [] + sample_audio_list = [] + sample_facial_list = [] + sample_shape_list = [] + sample_word_list = [] + sample_emo_list = [] + sample_sem_list = [] + sample_vid_list = [] + sample_trans_list = [] + sample_trans_v_list = [] + + for i in range(num_subdivision): # cut into around 2s chip, (self npose) + start_idx = clip_s_f_pose + i * self.args.stride + fin_idx = start_idx + cut_length + sample_pose = pose_each_file[start_idx:fin_idx] + + sample_trans = trans_each_file[start_idx:fin_idx] + sample_trans_v = trans_v_each_file[start_idx:fin_idx] + sample_shape = shape_each_file[start_idx:fin_idx] + # print(sample_pose.shape) + if self.args.audio_rep is not None: + audio_start = clip_s_f_audio + math.floor(i * self.args.stride * self.args.audio_fps / self.args.pose_fps) + audio_end = audio_start + audio_short_length + sample_audio = audio_each_file[audio_start:audio_end] + else: + sample_audio = np.array([-1]) + sample_facial = facial_each_file[start_idx:fin_idx] if self.args.facial_rep is not None else np.array([-1]) + sample_word = word_each_file[start_idx:fin_idx] if self.args.word_rep is not None else np.array([-1]) + sample_emo = emo_each_file[start_idx:fin_idx] if self.args.emo_rep is not None else np.array([-1]) + sample_sem = sem_each_file[start_idx:fin_idx] if self.args.sem_rep is not None else np.array([-1]) + sample_vid = vid_each_file[start_idx:fin_idx] if self.args.id_rep is not None else np.array([-1]) + + if sample_pose.any() != None: + # filtering motion skeleton data + sample_pose, filtering_message = MotionPreprocessor(sample_pose).get() + is_correct_motion = (sample_pose != []) + if is_correct_motion or disable_filtering: + sample_pose_list.append(sample_pose) + sample_audio_list.append(sample_audio) + sample_facial_list.append(sample_facial) + sample_shape_list.append(sample_shape) + sample_word_list.append(sample_word) + sample_vid_list.append(sample_vid) + sample_emo_list.append(sample_emo) + sample_sem_list.append(sample_sem) + sample_trans_list.append(sample_trans) + sample_trans_v_list.append(sample_trans_v) + else: + n_filtered_out[filtering_message] += 1 + + if len(sample_pose_list) > 0: + with dst_lmdb_env.begin(write=True) as txn: + for pose, audio, facial, shape, word, vid, emo, sem, trans,trans_v in zip( + sample_pose_list, + sample_audio_list, + sample_facial_list, + sample_shape_list, + sample_word_list, + sample_vid_list, + sample_emo_list, + sample_sem_list, + sample_trans_list, + sample_trans_v_list,): + k = "{:005}".format(self.n_out_samples).encode("ascii") + v = [pose, audio, facial, shape, word, emo, sem, vid, trans,trans_v] + v = pickle.dumps(v,5) + txn.put(k, v) + self.n_out_samples += 1 + return n_filtered_out + + def __getitem__(self, idx): + with self.lmdb_env.begin(write=False) as txn: + key = "{:005}".format(idx).encode("ascii") + sample = txn.get(key) + sample = pickle.loads(sample) + tar_pose, in_audio, in_facial, in_shape, in_word, emo, sem, vid, trans,trans_v = sample + #print(in_shape) + #vid = torch.from_numpy(vid).int() + emo = torch.from_numpy(emo).int() + sem = torch.from_numpy(sem).float() + in_audio = torch.from_numpy(in_audio).float() + in_word = torch.from_numpy(in_word).float() if self.args.word_cache else torch.from_numpy(in_word).int() + if self.loader_type == "test": + tar_pose = torch.from_numpy(tar_pose).float() + trans = torch.from_numpy(trans).float() + trans_v = torch.from_numpy(trans_v).float() + in_facial = torch.from_numpy(in_facial).float() + vid = torch.from_numpy(vid).float() + in_shape = torch.from_numpy(in_shape).float() + else: + in_shape = torch.from_numpy(in_shape).reshape((in_shape.shape[0], -1)).float() + trans = torch.from_numpy(trans).reshape((trans.shape[0], -1)).float() + trans_v = torch.from_numpy(trans_v).reshape((trans_v.shape[0], -1)).float() + vid = torch.from_numpy(vid).reshape((vid.shape[0], -1)).float() + tar_pose = torch.from_numpy(tar_pose).reshape((tar_pose.shape[0], -1)).float() + in_facial = torch.from_numpy(in_facial).reshape((in_facial.shape[0], -1)).float() + return {"pose":tar_pose, "audio":in_audio, "facial":in_facial, "beta": in_shape, "word":in_word, "id":vid, "emo":emo, "sem":sem, "trans":trans,"trans_v":trans_v} + + +class MotionPreprocessor: + def __init__(self, skeletons): + self.skeletons = skeletons + #self.mean_pose = mean_pose + self.filtering_message = "PASS" + + def get(self): + assert (self.skeletons is not None) + + # filtering + if self.skeletons != []: + if self.check_pose_diff(): + self.skeletons = [] + self.filtering_message = "pose" + # elif self.check_spine_angle(): + # self.skeletons = [] + # self.filtering_message = "spine angle" + # elif self.check_static_motion(): + # self.skeletons = [] + # self.filtering_message = "motion" + + # if self.skeletons != []: + # self.skeletons = self.skeletons.tolist() + # for i, frame in enumerate(self.skeletons): + # assert not np.isnan(self.skeletons[i]).any() # missing joints + + return self.skeletons, self.filtering_message + + def check_static_motion(self, verbose=True): + def get_variance(skeleton, joint_idx): + wrist_pos = skeleton[:, joint_idx] + variance = np.sum(np.var(wrist_pos, axis=0)) + return variance + + left_arm_var = get_variance(self.skeletons, 6) + right_arm_var = get_variance(self.skeletons, 9) + + th = 0.0014 # exclude 13110 + # th = 0.002 # exclude 16905 + if left_arm_var < th and right_arm_var < th: + if verbose: + print("skip - check_static_motion left var {}, right var {}".format(left_arm_var, right_arm_var)) + return True + else: + if verbose: + print("pass - check_static_motion left var {}, right var {}".format(left_arm_var, right_arm_var)) + return False + + + def check_pose_diff(self, verbose=False): +# diff = np.abs(self.skeletons - self.mean_pose) # 186*1 +# diff = np.mean(diff) + +# # th = 0.017 +# th = 0.02 #0.02 # exclude 3594 +# if diff < th: +# if verbose: +# print("skip - check_pose_diff {:.5f}".format(diff)) +# return True +# # th = 3.5 #0.02 # exclude 3594 +# # if 3.5 < diff < 5: +# # if verbose: +# # print("skip - check_pose_diff {:.5f}".format(diff)) +# # return True +# else: +# if verbose: +# print("pass - check_pose_diff {:.5f}".format(diff)) + return False + + + def check_spine_angle(self, verbose=True): + def angle_between(v1, v2): + v1_u = v1 / np.linalg.norm(v1) + v2_u = v2 / np.linalg.norm(v2) + return np.arccos(np.clip(np.dot(v1_u, v2_u), -1.0, 1.0)) + + angles = [] + for i in range(self.skeletons.shape[0]): + spine_vec = self.skeletons[i, 1] - self.skeletons[i, 0] + angle = angle_between(spine_vec, [0, -1, 0]) + angles.append(angle) + + if np.rad2deg(max(angles)) > 30 or np.rad2deg(np.mean(angles)) > 20: # exclude 4495 + # if np.rad2deg(max(angles)) > 20: # exclude 8270 + if verbose: + print("skip - check_spine_angle {:.5f}, {:.5f}".format(max(angles), np.mean(angles))) + return True + else: + if verbose: + print("pass - check_spine_angle {:.5f}".format(max(angles))) + return False \ No newline at end of file diff --git a/dataloaders/beat_sep_lower_single.py b/dataloaders/beat_sep_lower_single.py new file mode 100644 index 0000000000000000000000000000000000000000..89908372000c1edda213bff30740d4aeca80d2ba --- /dev/null +++ b/dataloaders/beat_sep_lower_single.py @@ -0,0 +1,730 @@ +import os +import pickle +import math +import shutil +import numpy as np +import lmdb as lmdb +import textgrid as tg +import pandas as pd +import torch +import glob +import json +from termcolor import colored +from loguru import logger +from collections import defaultdict +from torch.utils.data import Dataset +import torch.distributed as dist +#import pyarrow +import pickle +import librosa +import smplx + +from .build_vocab import Vocab +from .utils.audio_features import Wav2Vec2Model +from .data_tools import joints_list +from .utils import rotation_conversions as rc +from .utils import other_tools + +class CustomDataset(Dataset): + def __init__(self, args, loader_type, augmentation=None, kwargs=None, build_cache=True): + + self.audio_file_path = args.audio_file_path + self.textgrid_file_path = args.textgrid_file_path + self.default_pose_file = "./demo/examples/2_scott_0_1_1.npz" + + self.args = args + self.loader_type = loader_type + + self.rank = 0 + self.ori_stride = self.args.stride + self.ori_length = self.args.pose_length + self.alignment = [0,0] # for trinity + + self.ori_joint_list = joints_list[self.args.ori_joints] + self.tar_joint_list = joints_list[self.args.tar_joints] + if 'smplx' in self.args.pose_rep: + self.joint_mask = np.zeros(len(list(self.ori_joint_list.keys()))*3) + self.joints = len(list(self.tar_joint_list.keys())) + for joint_name in self.tar_joint_list: + self.joint_mask[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + else: + self.joints = len(list(self.ori_joint_list.keys()))+1 + self.joint_mask = np.zeros(self.joints*3) + for joint_name in self.tar_joint_list: + if joint_name == "Hips": + self.joint_mask[3:6] = 1 + else: + self.joint_mask[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + # select trainable joints + self.smplx = smplx.create( + self.args.data_path_1+"smplx_models/", + model_type='smplx', + gender='NEUTRAL_2020', + use_face_contour=False, + num_betas=300, + num_expression_coeffs=100, + ext='npz', + use_pca=False, + ).cuda().eval() + + split_rule = pd.read_csv(args.data_path+"train_test_split.csv") + self.selected_file = split_rule.loc[(split_rule['type'] == loader_type) & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + if args.additional_data and loader_type == 'train': + split_b = split_rule.loc[(split_rule['type'] == 'additional') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + #self.selected_file = split_rule.loc[(split_rule['type'] == 'additional') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + self.selected_file = pd.concat([self.selected_file, split_b]) + if self.selected_file.empty: + logger.warning(f"{loader_type} is empty for speaker {self.args.training_speakers}, use train set 0-8 instead") + self.selected_file = split_rule.loc[(split_rule['type'] == 'train') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + self.selected_file = self.selected_file.iloc[0:8] + self.data_dir = args.data_path + + if loader_type == "test": + self.args.multi_length_training = [1.0] + self.max_length = int(args.pose_length * self.args.multi_length_training[-1]) + self.max_audio_pre_len = math.floor(args.pose_length / args.pose_fps * self.args.audio_sr) + if self.max_audio_pre_len > self.args.test_length*self.args.audio_sr: + self.max_audio_pre_len = self.args.test_length*self.args.audio_sr + + if args.word_rep is not None: + with open(f"{args.data_path}weights/vocab.pkl", 'rb') as f: + self.lang_model = pickle.load(f) + + preloaded_dir = self.args.tmp_dir+'/' + loader_type + f"/{args.pose_rep}_cache" + + if self.args.beat_align: + if not os.path.exists(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy"): + self.calculate_mean_velocity(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy") + self.avg_vel = np.load(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy") + + if build_cache and self.rank == 0: + self.build_cache(preloaded_dir) + self.lmdb_env = lmdb.open(preloaded_dir, readonly=True, lock=False) + with self.lmdb_env.begin() as txn: + self.n_samples = txn.stat()["entries"] + + + + + def calculate_mean_velocity(self, save_path): + self.smplx = smplx.create( + self.args.data_path_1+"smplx_models/", + model_type='smplx', + gender='NEUTRAL_2020', + use_face_contour=False, + num_betas=300, + num_expression_coeffs=100, + ext='npz', + use_pca=False, + ).cuda().eval() + dir_p = self.data_dir + self.args.pose_rep + "/" + all_list = [] + from tqdm import tqdm + for tar in tqdm(os.listdir(dir_p)): + if tar.endswith(".npz"): + m_data = np.load(dir_p+tar, allow_pickle=True) + betas, poses, trans, exps = m_data["betas"], m_data["poses"], m_data["trans"], m_data["expressions"] + n, c = poses.shape[0], poses.shape[1] + betas = betas.reshape(1, 300) + betas = np.tile(betas, (n, 1)) + betas = torch.from_numpy(betas).cuda().float() + poses = torch.from_numpy(poses.reshape(n, c)).cuda().float() + exps = torch.from_numpy(exps.reshape(n, 100)).cuda().float() + trans = torch.from_numpy(trans.reshape(n, 3)).cuda().float() + max_length = 128 + s, r = n//max_length, n%max_length + #print(n, s, r) + all_tensor = [] + for i in range(s): + with torch.no_grad(): + joints = self.smplx( + betas=betas[i*max_length:(i+1)*max_length], + transl=trans[i*max_length:(i+1)*max_length], + expression=exps[i*max_length:(i+1)*max_length], + jaw_pose=poses[i*max_length:(i+1)*max_length, 66:69], + global_orient=poses[i*max_length:(i+1)*max_length,:3], + body_pose=poses[i*max_length:(i+1)*max_length,3:21*3+3], + left_hand_pose=poses[i*max_length:(i+1)*max_length,25*3:40*3], + right_hand_pose=poses[i*max_length:(i+1)*max_length,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[i*max_length:(i+1)*max_length, 69:72], + reye_pose=poses[i*max_length:(i+1)*max_length, 72:75], + )['joints'][:, :55, :].reshape(max_length, 55*3) + all_tensor.append(joints) + if r != 0: + with torch.no_grad(): + joints = self.smplx( + betas=betas[s*max_length:s*max_length+r], + transl=trans[s*max_length:s*max_length+r], + expression=exps[s*max_length:s*max_length+r], + jaw_pose=poses[s*max_length:s*max_length+r, 66:69], + global_orient=poses[s*max_length:s*max_length+r,:3], + body_pose=poses[s*max_length:s*max_length+r,3:21*3+3], + left_hand_pose=poses[s*max_length:s*max_length+r,25*3:40*3], + right_hand_pose=poses[s*max_length:s*max_length+r,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[s*max_length:s*max_length+r, 69:72], + reye_pose=poses[s*max_length:s*max_length+r, 72:75], + )['joints'][:, :55, :].reshape(r, 55*3) + all_tensor.append(joints) + joints = torch.cat(all_tensor, axis=0) + joints = joints.permute(1, 0) + dt = 1/30 + # first steps is forward diff (t+1 - t) / dt + init_vel = (joints[:, 1:2] - joints[:, :1]) / dt + # middle steps are second order (t+1 - t-1) / 2dt + middle_vel = (joints[:, 2:] - joints[:, 0:-2]) / (2 * dt) + # last step is backward diff (t - t-1) / dt + final_vel = (joints[:, -1:] - joints[:, -2:-1]) / dt + #print(joints.shape, init_vel.shape, middle_vel.shape, final_vel.shape) + vel_seq = torch.cat([init_vel, middle_vel, final_vel], dim=1).permute(1, 0).reshape(n, 55, 3) + #print(vel_seq.shape) + #.permute(1, 0).reshape(n, 55, 3) + vel_seq_np = vel_seq.cpu().numpy() + vel_joints_np = np.linalg.norm(vel_seq_np, axis=2) # n * 55 + all_list.append(vel_joints_np) + avg_vel = np.mean(np.concatenate(all_list, axis=0),axis=0) # 55 + np.save(save_path, avg_vel) + + + def build_cache(self, preloaded_dir): + logger.info(f"Audio bit rate: {self.args.audio_fps}") + logger.info("Reading data '{}'...".format(self.data_dir)) + logger.info("Creating the dataset cache...") + if self.args.new_cache: + if os.path.exists(preloaded_dir): + shutil.rmtree(preloaded_dir) + if os.path.exists(preloaded_dir): + logger.info("Found the cache {}".format(preloaded_dir)) + elif self.loader_type == "test": + self.cache_generation( + preloaded_dir, True, + 0, 0, + is_test=True) + else: + self.cache_generation( + preloaded_dir, self.args.disable_filtering, + self.args.clean_first_seconds, self.args.clean_final_seconds, + is_test=False) + + def __len__(self): + return self.n_samples + + + def cache_generation(self, out_lmdb_dir, disable_filtering, clean_first_seconds, clean_final_seconds, is_test=False): + # if "wav2vec2" in self.args.audio_rep: + # self.wav2vec_model = Wav2Vec2Model.from_pretrained(f"{self.args.data_path_1}/hub/transformer/wav2vec2-base-960h") + # self.wav2vec_model.feature_extractor._freeze_parameters() + # self.wav2vec_model = self.wav2vec_model.cuda() + # self.wav2vec_model.eval() + + self.n_out_samples = 0 + # create db for samples + if not os.path.exists(out_lmdb_dir): os.makedirs(out_lmdb_dir) + dst_lmdb_env = lmdb.open(out_lmdb_dir, map_size= int(1024 ** 3 * 500))# 500G + n_filtered_out = defaultdict(int) + + + #f_name = file_name["id"] + ext = ".npz" if "smplx" in self.args.pose_rep else ".bvh" + pose_file = self.default_pose_file + pose_each_file = [] + trans_each_file = [] + trans_v_each_file = [] + shape_each_file = [] + audio_each_file = [] + facial_each_file = [] + word_each_file = [] + emo_each_file = [] + sem_each_file = [] + vid_each_file = [] + id_pose = "tmp" #1_wayne_0_1_1 + + logger.info(colored(f"# ---- Building cache for Pose {id_pose} ---- #", "blue")) + if "smplx" in self.args.pose_rep: + pose_data = np.load(pose_file, allow_pickle=True) + assert 30%self.args.pose_fps == 0, 'pose_fps should be an aliquot part of 30' + stride = int(30/self.args.pose_fps) + pose_each_file = pose_data["poses"][::stride] + trans_each_file = pose_data["trans"][::stride] + trans_each_file[:,0] = trans_each_file[:,0] - trans_each_file[0,0] + trans_each_file[:,2] = trans_each_file[:,2] - trans_each_file[0,2] + trans_v_each_file = np.zeros_like(trans_each_file) + trans_v_each_file[1:,0] = trans_each_file[1:,0] - trans_each_file[:-1,0] + trans_v_each_file[0,0] = trans_v_each_file[1,0] + trans_v_each_file[1:,2] = trans_each_file[1:,2] - trans_each_file[:-1,2] + trans_v_each_file[0,2] = trans_v_each_file[1,2] + trans_v_each_file[:,1] = trans_each_file[:,1] + shape_each_file = np.repeat(pose_data["betas"].reshape(1, 300), pose_each_file.shape[0], axis=0) + + assert self.args.pose_fps == 30, "should 30" + m_data = np.load(pose_file, allow_pickle=True) + betas, poses, trans, exps = m_data["betas"], m_data["poses"], m_data["trans"], m_data["expressions"] + n, c = poses.shape[0], poses.shape[1] + betas = betas.reshape(1, 300) + betas = np.tile(betas, (n, 1)) + betas = torch.from_numpy(betas).cuda().float() + poses = torch.from_numpy(poses.reshape(n, c)).cuda().float() + exps = torch.from_numpy(exps.reshape(n, 100)).cuda().float() + trans = torch.from_numpy(trans.reshape(n, 3)).cuda().float() + max_length = 128 # 为什么这里需要一个max_length + s, r = n//max_length, n%max_length + #print(n, s, r) + all_tensor = [] + for i in range(s): + with torch.no_grad(): + joints = self.smplx( + betas=betas[i*max_length:(i+1)*max_length], + transl=trans[i*max_length:(i+1)*max_length], + expression=exps[i*max_length:(i+1)*max_length], + jaw_pose=poses[i*max_length:(i+1)*max_length, 66:69], + global_orient=poses[i*max_length:(i+1)*max_length,:3], + body_pose=poses[i*max_length:(i+1)*max_length,3:21*3+3], + left_hand_pose=poses[i*max_length:(i+1)*max_length,25*3:40*3], + right_hand_pose=poses[i*max_length:(i+1)*max_length,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[i*max_length:(i+1)*max_length, 69:72], + reye_pose=poses[i*max_length:(i+1)*max_length, 72:75], + )['joints'][:, (7,8,10,11), :].reshape(max_length, 4, 3).cpu() + all_tensor.append(joints) + if r != 0: + with torch.no_grad(): + joints = self.smplx( + betas=betas[s*max_length:s*max_length+r], + transl=trans[s*max_length:s*max_length+r], + expression=exps[s*max_length:s*max_length+r], + jaw_pose=poses[s*max_length:s*max_length+r, 66:69], + global_orient=poses[s*max_length:s*max_length+r,:3], + body_pose=poses[s*max_length:s*max_length+r,3:21*3+3], + left_hand_pose=poses[s*max_length:s*max_length+r,25*3:40*3], + right_hand_pose=poses[s*max_length:s*max_length+r,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[s*max_length:s*max_length+r, 69:72], + reye_pose=poses[s*max_length:s*max_length+r, 72:75], + )['joints'][:, (7,8,10,11), :].reshape(r, 4, 3).cpu() + all_tensor.append(joints) + joints = torch.cat(all_tensor, axis=0) # all, 4, 3 + # print(joints.shape) + feetv = torch.zeros(joints.shape[1], joints.shape[0]) + joints = joints.permute(1, 0, 2) + #print(joints.shape, feetv.shape) + feetv[:, :-1] = (joints[:, 1:] - joints[:, :-1]).norm(dim=-1) + #print(feetv.shape) + contacts = (feetv < 0.01).numpy().astype(float) + # print(contacts.shape, contacts) + contacts = contacts.transpose(1, 0) + pose_each_file = pose_each_file * self.joint_mask + pose_each_file = pose_each_file[:, self.joint_mask.astype(bool)] + pose_each_file = np.concatenate([pose_each_file, contacts], axis=1) + # print(pose_each_file.shape) + + + if self.args.facial_rep is not None: + logger.info(f"# ---- Building cache for Facial {id_pose} and Pose {id_pose} ---- #") + facial_each_file = pose_data["expressions"][::stride] + if self.args.facial_norm: + facial_each_file = (facial_each_file - self.mean_facial) / self.std_facial + + if self.args.id_rep is not None: + vid_each_file = np.repeat(np.array(int(999)-1).reshape(1, 1), pose_each_file.shape[0], axis=0) + + if self.args.audio_rep is not None: + logger.info(f"# ---- Building cache for Audio {id_pose} and Pose {id_pose} ---- #") + audio_file = self.audio_file_path + if not os.path.exists(audio_file): + logger.warning(f"# ---- file not found for Audio {id_pose}, skip all files with the same id ---- #") + self.selected_file = self.selected_file.drop(self.selected_file[self.selected_file['id'] == id_pose].index) + + audio_save_path = audio_file.replace("wave16k", "onset_amplitude").replace(".wav", ".npy") + + if self.args.audio_rep == "onset+amplitude": + audio_each_file, sr = librosa.load(audio_file) + audio_each_file = librosa.resample(audio_each_file, orig_sr=sr, target_sr=self.args.audio_sr) + from numpy.lib import stride_tricks + frame_length = 1024 + # hop_length = 512 + shape = (audio_each_file.shape[-1] - frame_length + 1, frame_length) + strides = (audio_each_file.strides[-1], audio_each_file.strides[-1]) + rolling_view = stride_tricks.as_strided(audio_each_file, shape=shape, strides=strides) + amplitude_envelope = np.max(np.abs(rolling_view), axis=1) + # pad the last frame_length-1 samples + amplitude_envelope = np.pad(amplitude_envelope, (0, frame_length-1), mode='constant', constant_values=amplitude_envelope[-1]) + audio_onset_f = librosa.onset.onset_detect(y=audio_each_file, sr=self.args.audio_sr, units='frames') + onset_array = np.zeros(len(audio_each_file), dtype=float) + onset_array[audio_onset_f] = 1.0 + # print(amplitude_envelope.shape, audio_each_file.shape, onset_array.shape) + audio_each_file = np.concatenate([amplitude_envelope.reshape(-1, 1), onset_array.reshape(-1, 1)], axis=1) + + + elif self.args.audio_rep == "mfcc": + audio_each_file = librosa.feature.melspectrogram(y=audio_each_file, sr=self.args.audio_sr, n_mels=128, hop_length=int(self.args.audio_sr/self.args.audio_fps)) + audio_each_file = audio_each_file.transpose(1, 0) + # print(audio_each_file.shape, pose_each_file.shape) + if self.args.audio_norm and self.args.audio_rep == "wave16k": + audio_each_file = (audio_each_file - self.mean_audio) / self.std_audio + + time_offset = 0 + if self.args.word_rep is not None: + logger.info(f"# ---- Building cache for Word {id_pose} and Pose {id_pose} ---- #") + word_file = self.textgrid_file_path + if not os.path.exists(word_file): + logger.warning(f"# ---- file not found for Word {id_pose}, skip all files with the same id ---- #") + self.selected_file = self.selected_file.drop(self.selected_file[self.selected_file['id'] == id_pose].index) + word_save_path = f"{self.data_dir}{self.args.t_pre_encoder}/{id_pose}.npy" + + tgrid = tg.TextGrid.fromFile(word_file) + + for i in range(pose_each_file.shape[0]): + found_flag = False + current_time = i/self.args.pose_fps + time_offset + j_last = 0 + for j, word in enumerate(tgrid[0]): + word_n, word_s, word_e = word.mark, word.minTime, word.maxTime + if word_s<=current_time and current_time<=word_e: + if word_n == " ": + word_each_file.append(self.lang_model.PAD_token) + else: + word_each_file.append(self.lang_model.get_word_index(word_n)) + found_flag = True + j_last = j + break + else: continue + if not found_flag: + word_each_file.append(self.lang_model.UNK_token) + word_each_file = np.array(word_each_file) + + + + if self.args.emo_rep is not None: + logger.info(f"# ---- Building cache for Emo {id_pose} and Pose {id_pose} ---- #") + rtype, start = int(id_pose.split('_')[3]), int(id_pose.split('_')[3]) + if rtype == 0 or rtype == 2 or rtype == 4 or rtype == 6: + if start >= 1 and start <= 64: + score = 0 + elif start >= 65 and start <= 72: + score = 1 + elif start >= 73 and start <= 80: + score = 2 + elif start >= 81 and start <= 86: + score = 3 + elif start >= 87 and start <= 94: + score = 4 + elif start >= 95 and start <= 102: + score = 5 + elif start >= 103 and start <= 110: + score = 6 + elif start >= 111 and start <= 118: + score = 7 + else: pass + else: + # you may denote as unknown in the future + score = 0 + emo_each_file = np.repeat(np.array(score).reshape(1, 1), pose_each_file.shape[0], axis=0) + #print(emo_each_file) + + if self.args.sem_rep is not None: + logger.info(f"# ---- Building cache for Sem {id_pose} and Pose {id_pose} ---- #") + sem_file = f"{self.data_dir}{self.args.sem_rep}/{id_pose}.txt" + sem_all = pd.read_csv(sem_file, + sep='\t', + names=["name", "start_time", "end_time", "duration", "score", "keywords"]) + # we adopt motion-level semantic score here. + for i in range(pose_each_file.shape[0]): + found_flag = False + for j, (start, end, score) in enumerate(zip(sem_all['start_time'],sem_all['end_time'], sem_all['score'])): + current_time = i/self.args.pose_fps + time_offset + if start<=current_time and current_time<=end: + sem_each_file.append(score) + found_flag=True + break + else: continue + if not found_flag: sem_each_file.append(0.) + sem_each_file = np.array(sem_each_file) + #print(sem_each_file) + + filtered_result = self._sample_from_clip( + dst_lmdb_env, + audio_each_file, pose_each_file, trans_each_file, trans_v_each_file,shape_each_file, facial_each_file, word_each_file, + vid_each_file, emo_each_file, sem_each_file, + disable_filtering, clean_first_seconds, clean_final_seconds, is_test, + ) + for type in filtered_result.keys(): + n_filtered_out[type] += filtered_result[type] + + + + +#### ---------for_end------------ #### + with dst_lmdb_env.begin() as txn: + logger.info(colored(f"no. of samples: {txn.stat()['entries']}", "cyan")) + n_total_filtered = 0 + for type, n_filtered in n_filtered_out.items(): + logger.info("{}: {}".format(type, n_filtered)) + n_total_filtered += n_filtered + logger.info(colored("no. of excluded samples: {} ({:.1f}%)".format( + n_total_filtered, 100 * n_total_filtered / (txn.stat()["entries"] + n_total_filtered)), "cyan")) + dst_lmdb_env.sync() + dst_lmdb_env.close() + + def _sample_from_clip( + self, dst_lmdb_env, audio_each_file, pose_each_file, trans_each_file, trans_v_each_file,shape_each_file, facial_each_file, word_each_file, + vid_each_file, emo_each_file, sem_each_file, + disable_filtering, clean_first_seconds, clean_final_seconds, is_test, + ): + """ + for data cleaning, we ignore the data for first and final n s + for test, we return all data + """ + # audio_start = int(self.alignment[0] * self.args.audio_fps) + # pose_start = int(self.alignment[1] * self.args.pose_fps) + #logger.info(f"before: {audio_each_file.shape} {pose_each_file.shape}") + # audio_each_file = audio_each_file[audio_start:] + # pose_each_file = pose_each_file[pose_start:] + # trans_each_file = + #logger.info(f"after alignment: {audio_each_file.shape} {pose_each_file.shape}") + #print(pose_each_file.shape) + round_seconds_skeleton = pose_each_file.shape[0] // self.args.pose_fps # assume 1500 frames / 15 fps = 100 s + #print(round_seconds_skeleton) + if audio_each_file is not None: + if self.args.audio_rep != "wave16k": + round_seconds_audio = len(audio_each_file) // self.args.audio_fps # assume 16,000,00 / 16,000 = 100 s + elif self.args.audio_rep == "mfcc": + round_seconds_audio = audio_each_file.shape[0] // self.args.audio_fps + else: + round_seconds_audio = audio_each_file.shape[0] // self.args.audio_sr + if facial_each_file is not None: + round_seconds_facial = facial_each_file.shape[0] // self.args.pose_fps + logger.info(f"audio: {round_seconds_audio}s, pose: {round_seconds_skeleton}s, facial: {round_seconds_facial}s") + round_seconds_skeleton = min(round_seconds_audio, round_seconds_skeleton, round_seconds_facial) + max_round = max(round_seconds_audio, round_seconds_skeleton, round_seconds_facial) + if round_seconds_skeleton != max_round: + logger.warning(f"reduce to {round_seconds_skeleton}s, ignore {max_round-round_seconds_skeleton}s") + else: + logger.info(f"pose: {round_seconds_skeleton}s, audio: {round_seconds_audio}s") + round_seconds_skeleton = min(round_seconds_audio, round_seconds_skeleton) + max_round = max(round_seconds_audio, round_seconds_skeleton) + if round_seconds_skeleton != max_round: + logger.warning(f"reduce to {round_seconds_skeleton}s, ignore {max_round-round_seconds_skeleton}s") + + clip_s_t, clip_e_t = clean_first_seconds, round_seconds_skeleton - clean_final_seconds # assume [10, 90]s + clip_s_f_audio, clip_e_f_audio = self.args.audio_fps * clip_s_t, clip_e_t * self.args.audio_fps # [160,000,90*160,000] + clip_s_f_pose, clip_e_f_pose = clip_s_t * self.args.pose_fps, clip_e_t * self.args.pose_fps # [150,90*15] + + + for ratio in self.args.multi_length_training: + if is_test:# stride = length for test + cut_length = clip_e_f_pose - clip_s_f_pose + self.args.stride = cut_length + self.max_length = cut_length + else: + self.args.stride = int(ratio*self.ori_stride) + cut_length = int(self.ori_length*ratio) + + num_subdivision = math.floor((clip_e_f_pose - clip_s_f_pose - cut_length) / self.args.stride) + 1 + logger.info(f"pose from frame {clip_s_f_pose} to {clip_e_f_pose}, length {cut_length}") + logger.info(f"{num_subdivision} clips is expected with stride {self.args.stride}") + + if audio_each_file is not None: + audio_short_length = math.floor(cut_length / self.args.pose_fps * self.args.audio_fps) + """ + for audio sr = 16000, fps = 15, pose_length = 34, + audio short length = 36266.7 -> 36266 + this error is fine. + """ + logger.info(f"audio from frame {clip_s_f_audio} to {clip_e_f_audio}, length {audio_short_length}") + + n_filtered_out = defaultdict(int) + sample_pose_list = [] + sample_audio_list = [] + sample_facial_list = [] + sample_shape_list = [] + sample_word_list = [] + sample_emo_list = [] + sample_sem_list = [] + sample_vid_list = [] + sample_trans_list = [] + sample_trans_v_list = [] + + for i in range(num_subdivision): # cut into around 2s chip, (self npose) + start_idx = clip_s_f_pose + i * self.args.stride + fin_idx = start_idx + cut_length + sample_pose = pose_each_file[start_idx:fin_idx] + + sample_trans = trans_each_file[start_idx:fin_idx] + sample_trans_v = trans_v_each_file[start_idx:fin_idx] + sample_shape = shape_each_file[start_idx:fin_idx] + # print(sample_pose.shape) + if self.args.audio_rep is not None: + audio_start = clip_s_f_audio + math.floor(i * self.args.stride * self.args.audio_fps / self.args.pose_fps) + audio_end = audio_start + audio_short_length + sample_audio = audio_each_file[audio_start:audio_end] + else: + sample_audio = np.array([-1]) + sample_facial = facial_each_file[start_idx:fin_idx] if self.args.facial_rep is not None else np.array([-1]) + sample_word = word_each_file[start_idx:fin_idx] if self.args.word_rep is not None else np.array([-1]) + sample_emo = emo_each_file[start_idx:fin_idx] if self.args.emo_rep is not None else np.array([-1]) + sample_sem = sem_each_file[start_idx:fin_idx] if self.args.sem_rep is not None else np.array([-1]) + sample_vid = vid_each_file[start_idx:fin_idx] if self.args.id_rep is not None else np.array([-1]) + + if sample_pose.any() != None: + # filtering motion skeleton data + sample_pose, filtering_message = MotionPreprocessor(sample_pose).get() + is_correct_motion = (sample_pose is not None) + if is_correct_motion or disable_filtering: + sample_pose_list.append(sample_pose) + sample_audio_list.append(sample_audio) + sample_facial_list.append(sample_facial) + sample_shape_list.append(sample_shape) + sample_word_list.append(sample_word) + sample_vid_list.append(sample_vid) + sample_emo_list.append(sample_emo) + sample_sem_list.append(sample_sem) + sample_trans_list.append(sample_trans) + sample_trans_v_list.append(sample_trans_v) + else: + n_filtered_out[filtering_message] += 1 + + if len(sample_pose_list) > 0: + with dst_lmdb_env.begin(write=True) as txn: + for pose, audio, facial, shape, word, vid, emo, sem, trans,trans_v in zip( + sample_pose_list, + sample_audio_list, + sample_facial_list, + sample_shape_list, + sample_word_list, + sample_vid_list, + sample_emo_list, + sample_sem_list, + sample_trans_list, + sample_trans_v_list,): + k = "{:005}".format(self.n_out_samples).encode("ascii") + v = [pose, audio, facial, shape, word, emo, sem, vid, trans,trans_v] + v = pickle.dumps(v,5) + txn.put(k, v) + self.n_out_samples += 1 + return n_filtered_out + + def __getitem__(self, idx): + with self.lmdb_env.begin(write=False) as txn: + key = "{:005}".format(idx).encode("ascii") + sample = txn.get(key) + sample = pickle.loads(sample) + tar_pose, in_audio, in_facial, in_shape, in_word, emo, sem, vid, trans,trans_v = sample + #print(in_shape) + #vid = torch.from_numpy(vid).int() + emo = torch.from_numpy(emo).int() + sem = torch.from_numpy(sem).float() + in_audio = torch.from_numpy(in_audio).float() + in_word = torch.from_numpy(in_word).float() if self.args.word_cache else torch.from_numpy(in_word).int() + if self.loader_type == "test": + tar_pose = torch.from_numpy(tar_pose).float() + trans = torch.from_numpy(trans).float() + trans_v = torch.from_numpy(trans_v).float() + in_facial = torch.from_numpy(in_facial).float() + vid = torch.from_numpy(vid).float() + in_shape = torch.from_numpy(in_shape).float() + else: + in_shape = torch.from_numpy(in_shape).reshape((in_shape.shape[0], -1)).float() + trans = torch.from_numpy(trans).reshape((trans.shape[0], -1)).float() + trans_v = torch.from_numpy(trans_v).reshape((trans_v.shape[0], -1)).float() + vid = torch.from_numpy(vid).reshape((vid.shape[0], -1)).float() + tar_pose = torch.from_numpy(tar_pose).reshape((tar_pose.shape[0], -1)).float() + in_facial = torch.from_numpy(in_facial).reshape((in_facial.shape[0], -1)).float() + return {"pose":tar_pose, "audio":in_audio, "facial":in_facial, "beta": in_shape, "word":in_word, "id":vid, "emo":emo, "sem":sem, "trans":trans,"trans_v":trans_v} + + +class MotionPreprocessor: + def __init__(self, skeletons): + self.skeletons = skeletons + #self.mean_pose = mean_pose + self.filtering_message = "PASS" + + def get(self): + assert (self.skeletons is not None) + + # filtering + if self.skeletons is not None: + if self.check_pose_diff(): + self.skeletons = [] + self.filtering_message = "pose" + # elif self.check_spine_angle(): + # self.skeletons = [] + # self.filtering_message = "spine angle" + # elif self.check_static_motion(): + # self.skeletons = [] + # self.filtering_message = "motion" + + # if self.skeletons is not None: + # self.skeletons = self.skeletons.tolist() + # for i, frame in enumerate(self.skeletons): + # assert not np.isnan(self.skeletons[i]).any() # missing joints + + return self.skeletons, self.filtering_message + + def check_static_motion(self, verbose=True): + def get_variance(skeleton, joint_idx): + wrist_pos = skeleton[:, joint_idx] + variance = np.sum(np.var(wrist_pos, axis=0)) + return variance + + left_arm_var = get_variance(self.skeletons, 6) + right_arm_var = get_variance(self.skeletons, 9) + + th = 0.0014 # exclude 13110 + # th = 0.002 # exclude 16905 + if left_arm_var < th and right_arm_var < th: + if verbose: + print("skip - check_static_motion left var {}, right var {}".format(left_arm_var, right_arm_var)) + return True + else: + if verbose: + print("pass - check_static_motion left var {}, right var {}".format(left_arm_var, right_arm_var)) + return False + + + def check_pose_diff(self, verbose=False): +# diff = np.abs(self.skeletons - self.mean_pose) # 186*1 +# diff = np.mean(diff) + +# # th = 0.017 +# th = 0.02 #0.02 # exclude 3594 +# if diff < th: +# if verbose: +# print("skip - check_pose_diff {:.5f}".format(diff)) +# return True +# # th = 3.5 #0.02 # exclude 3594 +# # if 3.5 < diff < 5: +# # if verbose: +# # print("skip - check_pose_diff {:.5f}".format(diff)) +# # return True +# else: +# if verbose: +# print("pass - check_pose_diff {:.5f}".format(diff)) + return False + + + def check_spine_angle(self, verbose=True): + def angle_between(v1, v2): + v1_u = v1 / np.linalg.norm(v1) + v2_u = v2 / np.linalg.norm(v2) + return np.arccos(np.clip(np.dot(v1_u, v2_u), -1.0, 1.0)) + + angles = [] + for i in range(self.skeletons.shape[0]): + spine_vec = self.skeletons[i, 1] - self.skeletons[i, 0] + angle = angle_between(spine_vec, [0, -1, 0]) + angles.append(angle) + + if np.rad2deg(max(angles)) > 30 or np.rad2deg(np.mean(angles)) > 20: # exclude 4495 + # if np.rad2deg(max(angles)) > 20: # exclude 8270 + if verbose: + print("skip - check_spine_angle {:.5f}, {:.5f}".format(max(angles), np.mean(angles))) + return True + else: + if verbose: + print("pass - check_spine_angle {:.5f}".format(max(angles))) + return False \ No newline at end of file diff --git a/dataloaders/beat_smplx2020.py b/dataloaders/beat_smplx2020.py new file mode 100644 index 0000000000000000000000000000000000000000..3674244faa73e645e98f65981eac586671fa5a07 --- /dev/null +++ b/dataloaders/beat_smplx2020.py @@ -0,0 +1,763 @@ +import os +import pickle +import math +import shutil +import numpy as np +import lmdb as lmdb +import textgrid as tg +import pandas as pd +import torch +import glob +import json +from termcolor import colored +from loguru import logger +from collections import defaultdict +from torch.utils.data import Dataset +import torch.distributed as dist +import pyarrow +import librosa +import smplx + +from .build_vocab import Vocab +from .utils.audio_features import Wav2Vec2Model +from .data_tools import joints_list +from .utils import rotation_conversions as rc +from .utils import other_tools + +class CustomDataset(Dataset): + def __init__(self, args, loader_type, augmentation=None, kwargs=None, build_cache=True): + self.args = args + self.loader_type = loader_type + + self.rank = dist.get_rank() + self.ori_stride = self.args.stride + self.ori_length = self.args.pose_length + self.alignment = [0,0] # for trinity + + self.ori_joint_list = joints_list[self.args.ori_joints] + self.tar_joint_list = joints_list[self.args.tar_joints] + if 'smplx' in self.args.pose_rep: + self.joint_mask = np.zeros(len(list(self.ori_joint_list.keys()))*3) + self.joints = len(list(self.ori_joint_list.keys())) + for joint_name in self.tar_joint_list: + self.joint_mask[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + else: + self.joints = len(list(self.ori_joint_list.keys()))+1 + self.joint_mask = np.zeros(self.joints*3) + for joint_name in self.tar_joint_list: + if joint_name == "Hips": + self.joint_mask[3:6] = 1 + else: + self.joint_mask[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + # select trainable joints + + split_rule = pd.read_csv(args.data_path+"train_test_split.csv") + self.selected_file = split_rule.loc[(split_rule['type'] == loader_type) & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + if args.additional_data and loader_type == 'train': + split_b = split_rule.loc[(split_rule['type'] == 'additional') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + #self.selected_file = split_rule.loc[(split_rule['type'] == 'additional') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + self.selected_file = pd.concat([self.selected_file, split_b]) + if self.selected_file.empty: + logger.warning(f"{loader_type} is empty for speaker {self.args.training_speakers}, use train set 0-8 instead") + self.selected_file = split_rule.loc[(split_rule['type'] == 'train') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + self.selected_file = self.selected_file.iloc[0:8] + self.data_dir = args.data_path + + if loader_type == "test": + self.args.multi_length_training = [1.0] + self.max_length = int(args.pose_length * self.args.multi_length_training[-1]) + self.max_audio_pre_len = math.floor(args.pose_length / args.pose_fps * self.args.audio_sr) + if self.max_audio_pre_len > self.args.test_length*self.args.audio_sr: + self.max_audio_pre_len = self.args.test_length*self.args.audio_sr + + if args.word_rep is not None: + with open(f"{args.data_path}weights/vocab.pkl", 'rb') as f: + self.lang_model = pickle.load(f) + + preloaded_dir = self.args.root_path + self.args.cache_path + loader_type + f"/{args.pose_rep}_cache" + # if args.pose_norm: + # # careful for rotation vectors + # if not os.path.exists(args.data_path+args.mean_pose_path+f"{args.pose_rep.split('_')[0]}/bvh_mean.npy"): + # self.calculate_mean_pose() + # self.mean_pose = np.load(args.data_path+args.mean_pose_path+f"{args.pose_rep.split('_')[0]}/bvh_mean.npy") + # self.std_pose = np.load(args.data_path+args.mean_pose_path+f"{args.pose_rep.split('_')[0]}/bvh_std.npy") + # if args.audio_norm: + # if not os.path.exists(args.data_path+args.mean_pose_path+f"{args.audio_rep.split('_')[0]}/bvh_mean.npy"): + # self.calculate_mean_audio() + # self.mean_audio = np.load(args.data_path+args.mean_pose_path+f"{args.audio_rep.split('_')[0]}/npy_mean.npy") + # self.std_audio = np.load(args.data_path+args.mean_pose_path+f"{args.audio_rep.split('_')[0]}/npy_std.npy") + # if args.facial_norm: + # if not os.path.exists(args.data_path+args.mean_pose_path+f"{args.pose_rep.split('_')[0]}/bvh_mean.npy"): + # self.calculate_mean_face() + # self.mean_facial = np.load(args.data_path+args.mean_pose_path+f"{args.facial_rep}/json_mean.npy") + # self.std_facial = np.load(args.data_path+args.mean_pose_path+f"{args.facial_rep}/json_std.npy") + if self.args.beat_align: + if not os.path.exists(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy"): + self.calculate_mean_velocity(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy") + self.avg_vel = np.load(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy") + + if build_cache and self.rank == 0: + self.build_cache(preloaded_dir) + self.lmdb_env = lmdb.open(preloaded_dir, readonly=True, lock=False) + with self.lmdb_env.begin() as txn: + self.n_samples = txn.stat()["entries"] + + + def calculate_mean_velocity(self, save_path): + self.smplx = smplx.create( + self.args.data_path_1+"smplx_models/", + model_type='smplx', + gender='NEUTRAL_2020', + use_face_contour=False, + num_betas=300, + num_expression_coeffs=100, + ext='npz', + use_pca=False, + ).cuda().eval() + dir_p = self.data_dir + self.args.pose_rep + "/" + all_list = [] + from tqdm import tqdm + for tar in tqdm(os.listdir(dir_p)): + if tar.endswith(".npz"): + m_data = np.load(dir_p+tar, allow_pickle=True) + betas, poses, trans, exps = m_data["betas"], m_data["poses"], m_data["trans"], m_data["expressions"] + n, c = poses.shape[0], poses.shape[1] + betas = betas.reshape(1, 300) + betas = np.tile(betas, (n, 1)) + betas = torch.from_numpy(betas).cuda().float() + poses = torch.from_numpy(poses.reshape(n, c)).cuda().float() + exps = torch.from_numpy(exps.reshape(n, 100)).cuda().float() + trans = torch.from_numpy(trans.reshape(n, 3)).cuda().float() + max_length = 128 + s, r = n//max_length, n%max_length + #print(n, s, r) + all_tensor = [] + for i in range(s): + with torch.no_grad(): + joints = self.smplx( + betas=betas[i*max_length:(i+1)*max_length], + transl=trans[i*max_length:(i+1)*max_length], + expression=exps[i*max_length:(i+1)*max_length], + jaw_pose=poses[i*max_length:(i+1)*max_length, 66:69], + global_orient=poses[i*max_length:(i+1)*max_length,:3], + body_pose=poses[i*max_length:(i+1)*max_length,3:21*3+3], + left_hand_pose=poses[i*max_length:(i+1)*max_length,25*3:40*3], + right_hand_pose=poses[i*max_length:(i+1)*max_length,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[i*max_length:(i+1)*max_length, 69:72], + reye_pose=poses[i*max_length:(i+1)*max_length, 72:75], + )['joints'][:, :55, :].reshape(max_length, 55*3) + all_tensor.append(joints) + if r != 0: + with torch.no_grad(): + joints = self.smplx( + betas=betas[s*max_length:s*max_length+r], + transl=trans[s*max_length:s*max_length+r], + expression=exps[s*max_length:s*max_length+r], + jaw_pose=poses[s*max_length:s*max_length+r, 66:69], + global_orient=poses[s*max_length:s*max_length+r,:3], + body_pose=poses[s*max_length:s*max_length+r,3:21*3+3], + left_hand_pose=poses[s*max_length:s*max_length+r,25*3:40*3], + right_hand_pose=poses[s*max_length:s*max_length+r,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[s*max_length:s*max_length+r, 69:72], + reye_pose=poses[s*max_length:s*max_length+r, 72:75], + )['joints'][:, :55, :].reshape(r, 55*3) + all_tensor.append(joints) + joints = torch.cat(all_tensor, axis=0) + joints = joints.permute(1, 0) + dt = 1/30 + # first steps is forward diff (t+1 - t) / dt + init_vel = (joints[:, 1:2] - joints[:, :1]) / dt + # middle steps are second order (t+1 - t-1) / 2dt + middle_vel = (joints[:, 2:] - joints[:, 0:-2]) / (2 * dt) + # last step is backward diff (t - t-1) / dt + final_vel = (joints[:, -1:] - joints[:, -2:-1]) / dt + #print(joints.shape, init_vel.shape, middle_vel.shape, final_vel.shape) + vel_seq = torch.cat([init_vel, middle_vel, final_vel], dim=1).permute(1, 0).reshape(n, 55, 3) + #print(vel_seq.shape) + #.permute(1, 0).reshape(n, 55, 3) + vel_seq_np = vel_seq.cpu().numpy() + vel_joints_np = np.linalg.norm(vel_seq_np, axis=2) # n * 55 + all_list.append(vel_joints_np) + avg_vel = np.mean(np.concatenate(all_list, axis=0),axis=0) # 55 + np.save(save_path, avg_vel) + + + def build_cache(self, preloaded_dir): + logger.info(f"Audio bit rate: {self.args.audio_fps}") + logger.info("Reading data '{}'...".format(self.data_dir)) + logger.info("Creating the dataset cache...") + if self.args.new_cache: + if os.path.exists(preloaded_dir): + shutil.rmtree(preloaded_dir) + if os.path.exists(preloaded_dir): + logger.info("Found the cache {}".format(preloaded_dir)) + elif self.loader_type == "test": + self.cache_generation( + preloaded_dir, True, + 0, 0, + is_test=True) + else: + self.cache_generation( + preloaded_dir, self.args.disable_filtering, + self.args.clean_first_seconds, self.args.clean_final_seconds, + is_test=False) + + def __len__(self): + return self.n_samples + + + def cache_generation(self, out_lmdb_dir, disable_filtering, clean_first_seconds, clean_final_seconds, is_test=False): + # if "wav2vec2" in self.args.audio_rep: + # self.wav2vec_model = Wav2Vec2Model.from_pretrained(f"{self.args.data_path_1}/hub/transformer/wav2vec2-base-960h") + # self.wav2vec_model.feature_extractor._freeze_parameters() + # self.wav2vec_model = self.wav2vec_model.cuda() + # self.wav2vec_model.eval() + + self.n_out_samples = 0 + # create db for samples + if not os.path.exists(out_lmdb_dir): os.makedirs(out_lmdb_dir) + dst_lmdb_env = lmdb.open(out_lmdb_dir, map_size= int(1024 ** 3 * 50))# 50G + n_filtered_out = defaultdict(int) + + for index, file_name in self.selected_file.iterrows(): + f_name = file_name["id"] + ext = ".npz" if "smplx" in self.args.pose_rep else ".bvh" + pose_file = self.data_dir + self.args.pose_rep + "/" + f_name + ext + pose_each_file = [] + trans_each_file = [] + shape_each_file = [] + audio_each_file = [] + facial_each_file = [] + word_each_file = [] + emo_each_file = [] + sem_each_file = [] + vid_each_file = [] + id_pose = f_name #1_wayne_0_1_1 + + logger.info(colored(f"# ---- Building cache for Pose {id_pose} ---- #", "blue")) + if "smplx" in self.args.pose_rep: + pose_data = np.load(pose_file, allow_pickle=True) + assert 30%self.args.pose_fps == 0, 'pose_fps should be an aliquot part of 30' + stride = int(30/self.args.pose_fps) + pose_each_file = pose_data["poses"][::stride] * self.joint_mask + trans_each_file = pose_data["trans"][::stride] + shape_each_file = np.repeat(pose_data["betas"].reshape(1, 300), pose_each_file.shape[0], axis=0) + if self.args.facial_rep is not None: + logger.info(f"# ---- Building cache for Facial {id_pose} and Pose {id_pose} ---- #") + facial_each_file = pose_data["expressions"][::stride] + if self.args.facial_norm: + facial_each_file = (facial_each_file - self.mean_facial) / self.std_facial + + else: + assert 120%self.args.pose_fps == 0, 'pose_fps should be an aliquot part of 120' + stride = int(120/self.args.pose_fps) + with open(pose_file, "r") as pose_data: + for j, line in enumerate(pose_data.readlines()): + if j < 431: continue + if j%stride != 0:continue + data = np.fromstring(line, dtype=float, sep=" ") + rot_data = rc.euler_angles_to_matrix(torch.from_numpy(np.deg2rad(data)).reshape(-1, self.joints,3), "XYZ") + rot_data = rc.matrix_to_axis_angle(rot_data).reshape(-1, self.joints*3) + rot_data = rot_data.numpy() * self.joint_mask + + pose_each_file.append(rot_data) + trans_each_file.append(data[:3]) + + pose_each_file = np.array(pose_each_file) + # print(pose_each_file.shape) + trans_each_file = np.array(trans_each_file) + shape_each_file = np.repeat(np.array(-1).reshape(1, 1), pose_each_file.shape[0], axis=0) + if self.args.facial_rep is not None: + logger.info(f"# ---- Building cache for Facial {id_pose} and Pose {id_pose} ---- #") + facial_file = pose_file.replace(self.args.pose_rep, self.args.facial_rep).replace("bvh", "json") + assert 60%self.args.pose_fps == 0, 'pose_fps should be an aliquot part of 120' + stride = int(60/self.args.pose_fps) + if not os.path.exists(facial_file): + logger.warning(f"# ---- file not found for Facial {id_pose}, skip all files with the same id ---- #") + self.selected_file = self.selected_file.drop(self.selected_file[self.selected_file['id'] == id_pose].index) + continue + with open(facial_file, 'r') as facial_data_file: + facial_data = json.load(facial_data_file) + for j, frame_data in enumerate(facial_data['frames']): + if j%stride != 0:continue + facial_each_file.append(frame_data['weights']) + facial_each_file = np.array(facial_each_file) + if self.args.facial_norm: + facial_each_file = (facial_each_file - self.mean_facial) / self.std_facial + + if self.args.id_rep is not None: + vid_each_file = np.repeat(np.array(int(f_name.split("_")[0])-1).reshape(1, 1), pose_each_file.shape[0], axis=0) + + if self.args.audio_rep is not None: + logger.info(f"# ---- Building cache for Audio {id_pose} and Pose {id_pose} ---- #") + audio_file = pose_file.replace(self.args.pose_rep, 'wave16k').replace(ext, ".wav") + if not os.path.exists(audio_file): + logger.warning(f"# ---- file not found for Audio {id_pose}, skip all files with the same id ---- #") + self.selected_file = self.selected_file.drop(self.selected_file[self.selected_file['id'] == id_pose].index) + continue + audio_each_file, sr = librosa.load(audio_file) + audio_each_file = librosa.resample(audio_each_file, orig_sr=sr, target_sr=self.args.audio_sr) + if self.args.audio_rep == "onset+amplitude": + from numpy.lib import stride_tricks + frame_length = 1024 + # hop_length = 512 + shape = (audio_each_file.shape[-1] - frame_length + 1, frame_length) + strides = (audio_each_file.strides[-1], audio_each_file.strides[-1]) + rolling_view = stride_tricks.as_strided(audio_each_file, shape=shape, strides=strides) + amplitude_envelope = np.max(np.abs(rolling_view), axis=1) + # pad the last frame_length-1 samples + amplitude_envelope = np.pad(amplitude_envelope, (0, frame_length-1), mode='constant', constant_values=amplitude_envelope[-1]) + audio_onset_f = librosa.onset.onset_detect(y=audio_each_file, sr=self.args.audio_sr, units='frames') + onset_array = np.zeros(len(audio_each_file), dtype=float) + onset_array[audio_onset_f] = 1.0 + # print(amplitude_envelope.shape, audio_each_file.shape, onset_array.shape) + audio_each_file = np.concatenate([amplitude_envelope.reshape(-1, 1), onset_array.reshape(-1, 1)], axis=1) + elif self.args.audio_rep == "mfcc": + audio_each_file = librosa.feature.mfcc(audio_each_file, sr=self.args.audio_sr, n_mfcc=13, hop_length=int(self.args.audio_sr/self.args.audio_fps)) + + if self.args.audio_norm and self.args.audio_rep == "wave16k": + audio_each_file = (audio_each_file - self.mean_audio) / self.std_audio + + time_offset = 0 + if self.args.word_rep is not None: + logger.info(f"# ---- Building cache for Word {id_pose} and Pose {id_pose} ---- #") + word_file = f"{self.data_dir}{self.args.word_rep}/{id_pose}.TextGrid" + if not os.path.exists(word_file): + logger.warning(f"# ---- file not found for Word {id_pose}, skip all files with the same id ---- #") + self.selected_file = self.selected_file.drop(self.selected_file[self.selected_file['id'] == id_pose].index) + continue + tgrid = tg.TextGrid.fromFile(word_file) + if self.args.t_pre_encoder == "bert": + from transformers import AutoTokenizer, BertModel + tokenizer = AutoTokenizer.from_pretrained(self.args.data_path_1 + "hub/bert-base-uncased", local_files_only=True) + model = BertModel.from_pretrained(self.args.data_path_1 + "hub/bert-base-uncased", local_files_only=True).eval() + list_word = [] + all_hidden = [] + max_len = 400 + last = 0 + word_token_mapping = [] + first = True + for i, word in enumerate(tgrid[0]): + last = i + if (i%max_len != 0) or (i==0): + if word.mark == "": + list_word.append(".") + else: + list_word.append(word.mark) + else: + max_counter = max_len + str_word = ' '.join(map(str, list_word)) + if first: + global_len = 0 + end = -1 + offset_word = [] + for k, wordvalue in enumerate(list_word): + start = end+1 + end = start+len(wordvalue) + offset_word.append((start, end)) + #print(offset_word) + token_scan = tokenizer.encode_plus(str_word, return_offsets_mapping=True)['offset_mapping'] + #print(token_scan) + for start, end in offset_word: + sub_mapping = [] + for i, (start_t, end_t) in enumerate(token_scan[1:-1]): + if int(start) <= int(start_t) and int(end_t) <= int(end): + #print(i+global_len) + sub_mapping.append(i+global_len) + word_token_mapping.append(sub_mapping) + #print(len(word_token_mapping)) + global_len = word_token_mapping[-1][-1] + 1 + list_word = [] + if word.mark == "": + list_word.append(".") + else: + list_word.append(word.mark) + + with torch.no_grad(): + inputs = tokenizer(str_word, return_tensors="pt") + outputs = model(**inputs) + last_hidden_states = outputs.last_hidden_state.reshape(-1, 768).cpu().numpy()[1:-1, :] + all_hidden.append(last_hidden_states) + + #list_word = list_word[:10] + if list_word == []: + pass + else: + if first: + global_len = 0 + str_word = ' '.join(map(str, list_word)) + end = -1 + offset_word = [] + for k, wordvalue in enumerate(list_word): + start = end+1 + end = start+len(wordvalue) + offset_word.append((start, end)) + #print(offset_word) + token_scan = tokenizer.encode_plus(str_word, return_offsets_mapping=True)['offset_mapping'] + #print(token_scan) + for start, end in offset_word: + sub_mapping = [] + for i, (start_t, end_t) in enumerate(token_scan[1:-1]): + if int(start) <= int(start_t) and int(end_t) <= int(end): + sub_mapping.append(i+global_len) + #print(sub_mapping) + word_token_mapping.append(sub_mapping) + #print(len(word_token_mapping)) + with torch.no_grad(): + inputs = tokenizer(str_word, return_tensors="pt") + outputs = model(**inputs) + last_hidden_states = outputs.last_hidden_state.reshape(-1, 768).cpu().numpy()[1:-1, :] + all_hidden.append(last_hidden_states) + last_hidden_states = np.concatenate(all_hidden, axis=0) + + for i in range(pose_each_file.shape[0]): + found_flag = False + current_time = i/self.args.pose_fps + time_offset + j_last = 0 + for j, word in enumerate(tgrid[0]): + word_n, word_s, word_e = word.mark, word.minTime, word.maxTime + if word_s<=current_time and current_time<=word_e: + if self.args.word_cache and self.args.t_pre_encoder == 'bert': + mapping_index = word_token_mapping[j] + #print(mapping_index, word_s, word_e) + s_t = np.linspace(word_s, word_e, len(mapping_index)+1) + #print(s_t) + for tt, t_sep in enumerate(s_t[1:]): + if current_time <= t_sep: + #if len(mapping_index) > 1: print(mapping_index[tt]) + word_each_file.append(last_hidden_states[mapping_index[tt]]) + break + else: + if word_n == " ": + word_each_file.append(self.lang_model.PAD_token) + else: + word_each_file.append(self.lang_model.get_word_index(word_n)) + found_flag = True + j_last = j + break + else: continue + if not found_flag: + if self.args.word_cache and self.args.t_pre_encoder == 'bert': + word_each_file.append(last_hidden_states[j_last]) + else: + word_each_file.append(self.lang_model.UNK_token) + word_each_file = np.array(word_each_file) + #print(word_each_file.shape) + + if self.args.emo_rep is not None: + logger.info(f"# ---- Building cache for Emo {id_pose} and Pose {id_pose} ---- #") + rtype, start = int(id_pose.split('_')[3]), int(id_pose.split('_')[3]) + if rtype == 0 or rtype == 2 or rtype == 4 or rtype == 6: + if start >= 1 and start <= 64: + score = 0 + elif start >= 65 and start <= 72: + score = 1 + elif start >= 73 and start <= 80: + score = 2 + elif start >= 81 and start <= 86: + score = 3 + elif start >= 87 and start <= 94: + score = 4 + elif start >= 95 and start <= 102: + score = 5 + elif start >= 103 and start <= 110: + score = 6 + elif start >= 111 and start <= 118: + score = 7 + else: pass + else: + # you may denote as unknown in the future + score = 0 + emo_each_file = np.repeat(np.array(score).reshape(1, 1), pose_each_file.shape[0], axis=0) + #print(emo_each_file) + + if self.args.sem_rep is not None: + logger.info(f"# ---- Building cache for Sem {id_pose} and Pose {id_pose} ---- #") + sem_file = f"{self.data_dir}{self.args.sem_rep}/{id_pose}.txt" + sem_all = pd.read_csv(sem_file, + sep='\t', + names=["name", "start_time", "end_time", "duration", "score", "keywords"]) + # we adopt motion-level semantic score here. + for i in range(pose_each_file.shape[0]): + found_flag = False + for j, (start, end, score) in enumerate(zip(sem_all['start_time'],sem_all['end_time'], sem_all['score'])): + current_time = i/self.args.pose_fps + time_offset + if start<=current_time and current_time<=end: + sem_each_file.append(score) + found_flag=True + break + else: continue + if not found_flag: sem_each_file.append(0.) + sem_each_file = np.array(sem_each_file) + #print(sem_each_file) + + filtered_result = self._sample_from_clip( + dst_lmdb_env, + audio_each_file, pose_each_file, trans_each_file, shape_each_file, facial_each_file, word_each_file, + vid_each_file, emo_each_file, sem_each_file, + disable_filtering, clean_first_seconds, clean_final_seconds, is_test, + ) + for type in filtered_result.keys(): + n_filtered_out[type] += filtered_result[type] + + with dst_lmdb_env.begin() as txn: + logger.info(colored(f"no. of samples: {txn.stat()['entries']}", "cyan")) + n_total_filtered = 0 + for type, n_filtered in n_filtered_out.items(): + logger.info("{}: {}".format(type, n_filtered)) + n_total_filtered += n_filtered + logger.info(colored("no. of excluded samples: {} ({:.1f}%)".format( + n_total_filtered, 100 * n_total_filtered / (txn.stat()["entries"] + n_total_filtered)), "cyan")) + dst_lmdb_env.sync() + dst_lmdb_env.close() + + def _sample_from_clip( + self, dst_lmdb_env, audio_each_file, pose_each_file, trans_each_file, shape_each_file, facial_each_file, word_each_file, + vid_each_file, emo_each_file, sem_each_file, + disable_filtering, clean_first_seconds, clean_final_seconds, is_test, + ): + """ + for data cleaning, we ignore the data for first and final n s + for test, we return all data + """ + # audio_start = int(self.alignment[0] * self.args.audio_fps) + # pose_start = int(self.alignment[1] * self.args.pose_fps) + #logger.info(f"before: {audio_each_file.shape} {pose_each_file.shape}") + # audio_each_file = audio_each_file[audio_start:] + # pose_each_file = pose_each_file[pose_start:] + # trans_each_file = + #logger.info(f"after alignment: {audio_each_file.shape} {pose_each_file.shape}") + #print(pose_each_file.shape) + round_seconds_skeleton = pose_each_file.shape[0] // self.args.pose_fps # assume 1500 frames / 15 fps = 100 s + #print(round_seconds_skeleton) + if audio_each_file != []: + round_seconds_audio = len(audio_each_file) // self.args.audio_fps # assume 16,000,00 / 16,000 = 100 s + if facial_each_file != []: + round_seconds_facial = facial_each_file.shape[0] // self.args.pose_fps + logger.info(f"audio: {round_seconds_audio}s, pose: {round_seconds_skeleton}s, facial: {round_seconds_facial}s") + round_seconds_skeleton = min(round_seconds_audio, round_seconds_skeleton, round_seconds_facial) + max_round = max(round_seconds_audio, round_seconds_skeleton, round_seconds_facial) + if round_seconds_skeleton != max_round: + logger.warning(f"reduce to {round_seconds_skeleton}s, ignore {max_round-round_seconds_skeleton}s") + else: + logger.info(f"pose: {round_seconds_skeleton}s, audio: {round_seconds_audio}s") + round_seconds_skeleton = min(round_seconds_audio, round_seconds_skeleton) + max_round = max(round_seconds_audio, round_seconds_skeleton) + if round_seconds_skeleton != max_round: + logger.warning(f"reduce to {round_seconds_skeleton}s, ignore {max_round-round_seconds_skeleton}s") + + clip_s_t, clip_e_t = clean_first_seconds, round_seconds_skeleton - clean_final_seconds # assume [10, 90]s + clip_s_f_audio, clip_e_f_audio = self.args.audio_fps * clip_s_t, clip_e_t * self.args.audio_fps # [160,000,90*160,000] + clip_s_f_pose, clip_e_f_pose = clip_s_t * self.args.pose_fps, clip_e_t * self.args.pose_fps # [150,90*15] + + + for ratio in self.args.multi_length_training: + if is_test:# stride = length for test + cut_length = clip_e_f_pose - clip_s_f_pose + self.args.stride = cut_length + self.max_length = cut_length + else: + self.args.stride = int(ratio*self.ori_stride) + cut_length = int(self.ori_length*ratio) + + num_subdivision = math.floor((clip_e_f_pose - clip_s_f_pose - cut_length) / self.args.stride) + 1 + logger.info(f"pose from frame {clip_s_f_pose} to {clip_e_f_pose}, length {cut_length}") + logger.info(f"{num_subdivision} clips is expected with stride {self.args.stride}") + + if audio_each_file != []: + audio_short_length = math.floor(cut_length / self.args.pose_fps * self.args.audio_fps) + """ + for audio sr = 16000, fps = 15, pose_length = 34, + audio short length = 36266.7 -> 36266 + this error is fine. + """ + logger.info(f"audio from frame {clip_s_f_audio} to {clip_e_f_audio}, length {audio_short_length}") + + n_filtered_out = defaultdict(int) + sample_pose_list = [] + sample_audio_list = [] + sample_facial_list = [] + sample_shape_list = [] + sample_word_list = [] + sample_emo_list = [] + sample_sem_list = [] + sample_vid_list = [] + sample_trans_list = [] + + for i in range(num_subdivision): # cut into around 2s chip, (self npose) + start_idx = clip_s_f_pose + i * self.args.stride + fin_idx = start_idx + cut_length + sample_pose = pose_each_file[start_idx:fin_idx] + sample_trans = trans_each_file[start_idx:fin_idx] + sample_shape = shape_each_file[start_idx:fin_idx] + # print(sample_pose.shape) + if self.args.audio_rep is not None: + audio_start = clip_s_f_audio + math.floor(i * self.args.stride * self.args.audio_fps / self.args.pose_fps) + audio_end = audio_start + audio_short_length + sample_audio = audio_each_file[audio_start:audio_end] + else: + sample_audio = np.array([-1]) + sample_facial = facial_each_file[start_idx:fin_idx] if self.args.facial_rep is not None else np.array([-1]) + sample_word = word_each_file[start_idx:fin_idx] if self.args.word_rep is not None else np.array([-1]) + sample_emo = emo_each_file[start_idx:fin_idx] if self.args.emo_rep is not None else np.array([-1]) + sample_sem = sem_each_file[start_idx:fin_idx] if self.args.sem_rep is not None else np.array([-1]) + sample_vid = vid_each_file[start_idx:fin_idx] if self.args.id_rep is not None else np.array([-1]) + + if sample_pose.any() != None: + # filtering motion skeleton data + sample_pose, filtering_message = MotionPreprocessor(sample_pose).get() + is_correct_motion = (sample_pose != []) + if is_correct_motion or disable_filtering: + sample_pose_list.append(sample_pose) + sample_audio_list.append(sample_audio) + sample_facial_list.append(sample_facial) + sample_shape_list.append(sample_shape) + sample_word_list.append(sample_word) + sample_vid_list.append(sample_vid) + sample_emo_list.append(sample_emo) + sample_sem_list.append(sample_sem) + sample_trans_list.append(sample_trans) + else: + n_filtered_out[filtering_message] += 1 + + if len(sample_pose_list) > 0: + with dst_lmdb_env.begin(write=True) as txn: + for pose, audio, facial, shape, word, vid, emo, sem, trans in zip( + sample_pose_list, + sample_audio_list, + sample_facial_list, + sample_shape_list, + sample_word_list, + sample_vid_list, + sample_emo_list, + sample_sem_list, + sample_trans_list,): + k = "{:005}".format(self.n_out_samples).encode("ascii") + v = [pose, audio, facial, shape, word, emo, sem, vid, trans] + v = pyarrow.serialize(v).to_buffer() + txn.put(k, v) + self.n_out_samples += 1 + return n_filtered_out + + def __getitem__(self, idx): + with self.lmdb_env.begin(write=False) as txn: + key = "{:005}".format(idx).encode("ascii") + sample = txn.get(key) + sample = pyarrow.deserialize(sample) + tar_pose, in_audio, in_facial, in_shape, in_word, emo, sem, vid, trans = sample + #print(in_shape) + #vid = torch.from_numpy(vid).int() + emo = torch.from_numpy(emo).int() + sem = torch.from_numpy(sem).float() + in_audio = torch.from_numpy(in_audio).float() + in_word = torch.from_numpy(in_word).float() if self.args.word_cache else torch.from_numpy(in_word).int() + if self.loader_type == "test": + tar_pose = torch.from_numpy(tar_pose).float() + trans = torch.from_numpy(trans).float() + in_facial = torch.from_numpy(in_facial).float() + vid = torch.from_numpy(vid).float() + in_shape = torch.from_numpy(in_shape).float() + else: + in_shape = torch.from_numpy(in_shape).reshape((in_shape.shape[0], -1)).float() + trans = torch.from_numpy(trans).reshape((trans.shape[0], -1)).float() + vid = torch.from_numpy(vid).reshape((vid.shape[0], -1)).float() + tar_pose = torch.from_numpy(tar_pose).reshape((tar_pose.shape[0], -1)).float() + in_facial = torch.from_numpy(in_facial).reshape((in_facial.shape[0], -1)).float() + return {"pose":tar_pose, "audio":in_audio, "facial":in_facial, "beta": in_shape, "word":in_word, "id":vid, "emo":emo, "sem":sem, "trans":trans} + + +class MotionPreprocessor: + def __init__(self, skeletons): + self.skeletons = skeletons + #self.mean_pose = mean_pose + self.filtering_message = "PASS" + + def get(self): + assert (self.skeletons is not None) + + # filtering + if self.skeletons != []: + if self.check_pose_diff(): + self.skeletons = [] + self.filtering_message = "pose" + # elif self.check_spine_angle(): + # self.skeletons = [] + # self.filtering_message = "spine angle" + # elif self.check_static_motion(): + # self.skeletons = [] + # self.filtering_message = "motion" + + # if self.skeletons != []: + # self.skeletons = self.skeletons.tolist() + # for i, frame in enumerate(self.skeletons): + # assert not np.isnan(self.skeletons[i]).any() # missing joints + + return self.skeletons, self.filtering_message + + def check_static_motion(self, verbose=True): + def get_variance(skeleton, joint_idx): + wrist_pos = skeleton[:, joint_idx] + variance = np.sum(np.var(wrist_pos, axis=0)) + return variance + + left_arm_var = get_variance(self.skeletons, 6) + right_arm_var = get_variance(self.skeletons, 9) + + th = 0.0014 # exclude 13110 + # th = 0.002 # exclude 16905 + if left_arm_var < th and right_arm_var < th: + if verbose: + print("skip - check_static_motion left var {}, right var {}".format(left_arm_var, right_arm_var)) + return True + else: + if verbose: + print("pass - check_static_motion left var {}, right var {}".format(left_arm_var, right_arm_var)) + return False + + + def check_pose_diff(self, verbose=False): +# diff = np.abs(self.skeletons - self.mean_pose) # 186*1 +# diff = np.mean(diff) + +# # th = 0.017 +# th = 0.02 #0.02 # exclude 3594 +# if diff < th: +# if verbose: +# print("skip - check_pose_diff {:.5f}".format(diff)) +# return True +# # th = 3.5 #0.02 # exclude 3594 +# # if 3.5 < diff < 5: +# # if verbose: +# # print("skip - check_pose_diff {:.5f}".format(diff)) +# # return True +# else: +# if verbose: +# print("pass - check_pose_diff {:.5f}".format(diff)) + return False + + + def check_spine_angle(self, verbose=True): + def angle_between(v1, v2): + v1_u = v1 / np.linalg.norm(v1) + v2_u = v2 / np.linalg.norm(v2) + return np.arccos(np.clip(np.dot(v1_u, v2_u), -1.0, 1.0)) + + angles = [] + for i in range(self.skeletons.shape[0]): + spine_vec = self.skeletons[i, 1] - self.skeletons[i, 0] + angle = angle_between(spine_vec, [0, -1, 0]) + angles.append(angle) + + if np.rad2deg(max(angles)) > 30 or np.rad2deg(np.mean(angles)) > 20: # exclude 4495 + # if np.rad2deg(max(angles)) > 20: # exclude 8270 + if verbose: + print("skip - check_spine_angle {:.5f}, {:.5f}".format(max(angles), np.mean(angles))) + return True + else: + if verbose: + print("pass - check_spine_angle {:.5f}".format(max(angles))) + return False \ No newline at end of file diff --git a/dataloaders/build_vocab.py b/dataloaders/build_vocab.py new file mode 100644 index 0000000000000000000000000000000000000000..fa1ca7af2a372f4ffc966160012edd60ba10c168 --- /dev/null +++ b/dataloaders/build_vocab.py @@ -0,0 +1,199 @@ +import numpy as np +import glob +import os +import pickle +import lmdb +#import pyarrow +import fasttext +from loguru import logger +from scipy import linalg + + +class Vocab: + PAD_token = 0 + SOS_token = 1 + EOS_token = 2 + UNK_token = 3 + + def __init__(self, name, insert_default_tokens=True): + self.name = name + self.trimmed = False + self.word_embedding_weights = None + self.reset_dictionary(insert_default_tokens) + + def reset_dictionary(self, insert_default_tokens=True): + self.word2index = {} + self.word2count = {} + if insert_default_tokens: + self.index2word = {self.PAD_token: "", self.SOS_token: "", + self.EOS_token: "", self.UNK_token: ""} + else: + self.index2word = {self.UNK_token: ""} + self.n_words = len(self.index2word) # count default tokens + + def index_word(self, word): + if word not in self.word2index: + self.word2index[word] = self.n_words + self.word2count[word] = 1 + self.index2word[self.n_words] = word + self.n_words += 1 + else: + self.word2count[word] += 1 + + def add_vocab(self, other_vocab): + for word, _ in other_vocab.word2count.items(): + self.index_word(word) + + # remove words below a certain count threshold + def trim(self, min_count): + if self.trimmed: + return + self.trimmed = True + + keep_words = [] + + for k, v in self.word2count.items(): + if v >= min_count: + keep_words.append(k) + + print(' word trimming, kept %s / %s = %.4f' % ( + len(keep_words), len(self.word2index), len(keep_words) / len(self.word2index) + )) + + # reinitialize dictionary + self.reset_dictionary() + for word in keep_words: + self.index_word(word) + + def get_word_index(self, word): + if word in self.word2index: + return self.word2index[word] + else: + return self.UNK_token + + def load_word_vectors(self, pretrained_path, embedding_dim=300): + print(" loading word vectors from '{}'...".format(pretrained_path)) + + # initialize embeddings to random values for special words + init_sd = 1 / np.sqrt(embedding_dim) + weights = np.random.normal(0, scale=init_sd, size=[self.n_words, embedding_dim]) + weights = weights.astype(np.float32) + + # read word vectors + word_model = fasttext.load_model(pretrained_path) + for word, id in self.word2index.items(): + vec = word_model.get_word_vector(word) + weights[id] = vec + self.word_embedding_weights = weights + + def __get_embedding_weight(self, pretrained_path, embedding_dim=300): + """ function modified from http://ronny.rest/blog/post_2017_08_04_glove/ """ + print("Loading word embedding '{}'...".format(pretrained_path)) + cache_path = pretrained_path + weights = None + + # use cached file if it exists + if os.path.exists(cache_path): # + with open(cache_path, 'rb') as f: + print(' using cached result from {}'.format(cache_path)) + weights = pickle.load(f) + if weights.shape != (self.n_words, embedding_dim): + logging.warning(' failed to load word embedding weights. reinitializing...') + weights = None + + if weights is None: + # initialize embeddings to random values for special and OOV words + init_sd = 1 / np.sqrt(embedding_dim) + weights = np.random.normal(0, scale=init_sd, size=[self.n_words, embedding_dim]) + weights = weights.astype(np.float32) + + with open(pretrained_path, encoding="utf-8", mode="r") as textFile: + num_embedded_words = 0 + for line_raw in textFile: + # extract the word, and embeddings vector + line = line_raw.split() + try: + word, vector = (line[0], np.array(line[1:], dtype=np.float32)) + # if word == 'love': # debugging + # print(word, vector) + + # if it is in our vocab, then update the corresponding weights + id = self.word2index.get(word, None) + if id is not None: + weights[id] = vector + num_embedded_words += 1 + except ValueError: + print(' parsing error at {}...'.format(line_raw[:50])) + continue + print(' {} / {} word vectors are found in the embedding'.format(num_embedded_words, len(self.word2index))) + + with open(cache_path, 'wb') as f: + pickle.dump(weights, f) + return weights + + +def build_vocab(name, data_path, cache_path, word_vec_path=None, feat_dim=None): + print(' building a language model...') + #if not os.path.exists(cache_path): + lang_model = Vocab(name) + print(' indexing words from {}'.format(data_path)) + index_words_from_textgrid(lang_model, data_path) + + if word_vec_path is not None: + lang_model.load_word_vectors(word_vec_path, feat_dim) + else: + print(' loaded from {}'.format(cache_path)) + with open(cache_path, 'rb') as f: + lang_model = pickle.load(f) + if word_vec_path is None: + lang_model.word_embedding_weights = None + elif lang_model.word_embedding_weights.shape[0] != lang_model.n_words: + logging.warning(' failed to load word embedding weights. check this') + assert False + + with open(cache_path, 'wb') as f: + pickle.dump(lang_model, f) + + + return lang_model + + +def index_words(lang_model, data_path): + #index words form text + with open(data_path, "r") as f: + for line in f.readlines(): + line = line.replace(",", " ") + line = line.replace(".", " ") + line = line.replace("?", " ") + line = line.replace("!", " ") + for word in line.split(): + lang_model.index_word(word) + print(' indexed %d words' % lang_model.n_words) + +def index_words_from_textgrid(lang_model, data_path): + import textgrid as tg + from tqdm import tqdm + #trainvaltest=os.listdir(data_path) + # for loadtype in trainvaltest: + # if "." in loadtype: continue #ignore .ipynb_checkpoints + texts = os.listdir(data_path+"/textgrid/") + #print(texts) + for textfile in tqdm(texts): + tgrid = tg.TextGrid.fromFile(data_path+"/textgrid/"+textfile) + for word in tgrid[0]: + word_n, word_s, word_e = word.mark, word.minTime, word.maxTime + word_n = word_n.replace(",", " ") + word_n = word_n.replace(".", " ") + word_n = word_n.replace("?", " ") + word_n = word_n.replace("!", " ") + #print(word_n) + lang_model.index_word(word_n) + print(' indexed %d words' % lang_model.n_words) + print(lang_model.word2index, lang_model.word2count) + +if __name__ == "__main__": + # 11195 for all, 5793 for 4 speakers + # build_vocab("beat_english_15_141", "/home/ma-user/work/datasets/beat_cache/beat_english_15_141/", "/home/ma-user/work/datasets/beat_cache/beat_english_15_141/vocab.pkl", "/home/ma-user/work/datasets/cc.en.300.bin", 300) + build_vocab("beat_chinese_v1.0.0", "/data/datasets/beat_chinese_v1.0.0/", "/data/datasets/beat_chinese_v1.0.0/weights/vocab.pkl", "/home/ma-user/work/cc.zh.300.bin", 300) + + \ No newline at end of file diff --git a/dataloaders/data_tools.py b/dataloaders/data_tools.py new file mode 100644 index 0000000000000000000000000000000000000000..a1e17a5c30d07c425238e4f94154d0c4f445f72d --- /dev/null +++ b/dataloaders/data_tools.py @@ -0,0 +1,1756 @@ +import numpy as np +import glob +import os +import pickle +import lmdb +#import pyarrow +import fasttext +from loguru import logger +from scipy import linalg +from .pymo.parsers import BVHParser +from .pymo.viz_tools import * +from .pymo.preprocessing import * + + + + +# pose version fpsxx_trinity/japanese_joints(_xxx) +joints_list = { + "trinity_joints":{ + 'Hips': [6,6], + 'Spine': [3,9], + 'Spine1': [3,12], + 'Spine2': [3,15], + 'Spine3': [3,18], + 'Neck': [3,21], + 'Neck1': [3,24], + 'Head': [3,27], + 'RShoulder': [3,30], + 'RArm': [3,33], + 'RArm1': [3,36], + 'RHand': [3,39], + 'RHandT1': [3,42], + 'RHandT2': [3,45], + 'RHandT3': [3,48], + 'RHandI1': [3,51], + 'RHandI2': [3,54], + 'RHandI3': [3,57], + 'RHandM1': [3,60], + 'RHandM2': [3,63], + 'RHandM3': [3,66], + 'RHandR1': [3,69], + 'RHandR2': [3,72], + 'RHandR3': [3,75], + 'RHandP1': [3,78], + 'RHandP2': [3,81], + 'RHandP3': [3,84], + 'LShoulder': [3,87], + 'LArm': [3,90], + 'LArm1': [3,93], + 'LHand': [3,96], + 'LHandT1': [3,99], + 'LHandT2': [3,102], + 'LHandT3': [3,105], + 'LHandI1': [3,108], + 'LHandI2': [3,111], + 'LHandI3': [3,114], + 'LHandM1': [3,117], + 'LHandM2': [3,120], + 'LHandM3': [3,123], + 'LHandR1': [3,126], + 'LHandR2': [3,129], + 'LHandR3': [3,132], + 'LHandP1': [3,135], + 'LHandP2': [3,138], + 'LHandP3': [3,141], + 'RUpLeg': [3,144], + 'RLeg': [3,147], + 'RFoot': [3,150], + 'RFootF': [3,153], + 'RToeBase': [3,156], + 'LUpLeg': [3,159], + 'LLeg': [3,162], + 'LFoot': [3,165], + 'LFootF': [3,168], + 'LToeBase': [3,171],}, + "trinity_joints_123":{ + 'Spine': 3 , + 'Neck': 3 , + 'Neck1': 3 , + 'RShoulder': 3 , + 'RArm': 3 , + 'RArm1': 3 , + 'RHand': 3 , + 'RHandT1': 3 , + 'RHandT2': 3 , + 'RHandT3': 3 , + 'RHandI1': 3 , + 'RHandI2': 3 , + 'RHandI3': 3 , + 'RHandM1': 3 , + 'RHandM2': 3 , + 'RHandM3': 3 , + 'RHandR1': 3 , + 'RHandR2': 3 , + 'RHandR3': 3 , + 'RHandP1': 3 , + 'RHandP2': 3 , + 'RHandP3': 3 , + 'LShoulder': 3 , + 'LArm': 3 , + 'LArm1': 3 , + 'LHand': 3 , + 'LHandT1': 3 , + 'LHandT2': 3 , + 'LHandT3': 3 , + 'LHandI1': 3 , + 'LHandI2': 3 , + 'LHandI3': 3 , + 'LHandM1': 3 , + 'LHandM2': 3 , + 'LHandM3': 3 , + 'LHandR1': 3 , + 'LHandR2': 3 , + 'LHandR3': 3 , + 'LHandP1': 3 , + 'LHandP2': 3 , + 'LHandP3': 3 ,}, + "trinity_joints_168":{ + 'Hips': 3 , + 'Spine': 3 , + 'Spine1': 3 , + 'Spine2': 3 , + 'Spine3': 3 , + 'Neck': 3 , + 'Neck1': 3 , + 'Head': 3 , + 'RShoulder': 3 , + 'RArm': 3 , + 'RArm1': 3 , + 'RHand': 3 , + 'RHandT1': 3 , + 'RHandT2': 3 , + 'RHandT3': 3 , + 'RHandI1': 3 , + 'RHandI2': 3 , + 'RHandI3': 3 , + 'RHandM1': 3 , + 'RHandM2': 3 , + 'RHandM3': 3 , + 'RHandR1': 3 , + 'RHandR2': 3 , + 'RHandR3': 3 , + 'RHandP1': 3 , + 'RHandP2': 3 , + 'RHandP3': 3 , + 'LShoulder': 3 , + 'LArm': 3 , + 'LArm1': 3 , + 'LHand': 3 , + 'LHandT1': 3 , + 'LHandT2': 3 , + 'LHandT3': 3 , + 'LHandI1': 3 , + 'LHandI2': 3 , + 'LHandI3': 3 , + 'LHandM1': 3 , + 'LHandM2': 3 , + 'LHandM3': 3 , + 'LHandR1': 3 , + 'LHandR2': 3 , + 'LHandR3': 3 , + 'LHandP1': 3 , + 'LHandP2': 3 , + 'LHandP3': 3 , + 'RUpLeg': 3 , + 'RLeg': 3 , + 'RFoot': 3 , + 'RFootF': 3 , + 'RToeBase': 3 , + 'LUpLeg': 3 , + 'LLeg': 3 , + 'LFoot': 3 , + 'LFootF': 3 , + 'LToeBase': 3 ,}, + "trinity_joints_138":{ + "Hips": 3 , + 'Spine': 3 , + 'Spine1': 3 , + 'Spine2': 3 , + 'Spine3': 3 , + 'Neck': 3 , + 'Neck1': 3 , + 'Head': 3 , + 'RShoulder': 3 , + 'RArm': 3 , + 'RArm1': 3 , + 'RHand': 3 , + 'RHandT1': 3 , + 'RHandT2': 3 , + 'RHandT3': 3 , + 'RHandI1': 3 , + 'RHandI2': 3 , + 'RHandI3': 3 , + 'RHandM1': 3 , + 'RHandM2': 3 , + 'RHandM3': 3 , + 'RHandR1': 3 , + 'RHandR2': 3 , + 'RHandR3': 3 , + 'RHandP1': 3 , + 'RHandP2': 3 , + 'RHandP3': 3 , + 'LShoulder': 3 , + 'LArm': 3 , + 'LArm1': 3 , + 'LHand': 3 , + 'LHandT1': 3 , + 'LHandT2': 3 , + 'LHandT3': 3 , + 'LHandI1': 3 , + 'LHandI2': 3 , + 'LHandI3': 3 , + 'LHandM1': 3 , + 'LHandM2': 3 , + 'LHandM3': 3 , + 'LHandR1': 3 , + 'LHandR2': 3 , + 'LHandR3': 3 , + 'LHandP1': 3 , + 'LHandP2': 3 , + 'LHandP3': 3 ,}, + "beat_smplx_joints": { + 'pelvis': [3,3], + 'left_hip': [3,6], + 'right_hip': [3,9], + 'spine1': [3,12], + 'left_knee': [3,15], + 'right_knee': [3,18], + 'spine2': [3,21], + 'left_ankle': [3,24], + 'right_ankle': [3,27], + + 'spine3': [3,30], + 'left_foot': [3,33], + 'right_foot': [3,36], + 'neck': [3,39], + 'left_collar': [3,42], + 'right_collar': [3,45], + 'head': [3,48], + 'left_shoulder': [3,51], + + 'right_shoulder': [3,54], + 'left_elbow': [3,57], + 'right_elbow': [3,60], + 'left_wrist': [3,63], + 'right_wrist': [3,66], + + 'jaw': [3,69], + 'left_eye_smplhf': [3,72], + 'right_eye_smplhf': [3,75], + 'left_index1': [3,78], + 'left_index2': [3,81], + + 'left_index3': [3,84], + 'left_middle1': [3,87], + 'left_middle2': [3,90], + 'left_middle3': [3,93], + 'left_pinky1': [3,96], + + 'left_pinky2': [3,99], + 'left_pinky3': [3,102], + 'left_ring1': [3,105], + 'left_ring2': [3,108], + + 'left_ring3': [3,111], + 'left_thumb1': [3,114], + 'left_thumb2': [3,117], + 'left_thumb3': [3,120], + 'right_index1': [3,123], + 'right_index2': [3,126], + 'right_index3': [3,129], + 'right_middle1': [3,132], + + 'right_middle2': [3,135], + 'right_middle3': [3,138], + 'right_pinky1': [3,141], + 'right_pinky2': [3,144], + 'right_pinky3': [3,147], + + 'right_ring1': [3,150], + 'right_ring2': [3,153], + 'right_ring3': [3,156], + 'right_thumb1': [3,159], + 'right_thumb2': [3,162], + 'right_thumb3': [3,165], + +# 'nose': [3,168], +# 'right_eye': [3,171], +# 'left_eye': [3,174], +# 'right_ear': [3,177], + +# 'left_ear': [3,180], +# 'left_big_toe': [3,183], +# 'left_small_toe': [3,186], +# 'left_heel': [3,189], + +# 'right_big_toe': [3,192], +# 'right_small_toe': [3,195], +# 'right_heel': [3,198], +# 'left_thumb': [3,201], +# 'left_index': [3,204], +# 'left_middle': [3,207], + +# 'left_ring': [3,210], +# 'left_pinky': [3,213], +# 'right_thumb': [3,216], +# 'right_index': [3,219], +# 'right_middle': [3,222], +# 'right_ring': [3,225], + +# 'right_pinky': [3,228], +# 'right_eye_brow1': [3,231], +# 'right_eye_brow2': [3,234], +# 'right_eye_brow3': [3,237], + +# 'right_eye_brow4': [3,240], +# 'right_eye_brow5': [3,243], +# 'left_eye_brow5': [3,246], +# 'left_eye_brow4': [3,249], + +# 'left_eye_brow3': [3,252], +# 'left_eye_brow2': [3,255], +# 'left_eye_brow1': [3,258], +# 'nose1': [3,261], +# 'nose2': [3,264], +# 'nose3': [3,267], + +# 'nose4': [3,270], +# 'right_nose_2': [3,273], +# 'right_nose_1': [3,276], +# 'nose_middle': [3,279], +# 'left_nose_1': [3,282], +# 'left_nose_2': [3,285], + +# 'right_eye1': [3,288], +# 'right_eye2': [3,291], +# 'right_eye3': [3,294], +# 'right_eye4': [3,297], + +# 'right_eye5': [3,300], +# 'right_eye6': [3,303], +# 'left_eye4': [3,306], +# 'left_eye3': [3,309], + +# 'left_eye2': [3,312], +# 'left_eye1': [3,315], +# 'left_eye6': [3,318], +# 'left_eye5': [3,321], +# 'right_mouth_1': [3,324], +# 'right_mouth_2': [3,327], +# 'right_mouth_3': [3,330], +# 'mouth_top': [3,333], +# 'left_mouth_3': [3,336], +# 'left_mouth_2': [3,339], +# 'left_mouth_1': [3,342], +# 'left_mouth_5': [3,345], +# 'left_mouth_4': [3,348], +# 'mouth_bottom': [3,351], +# 'right_mouth_4': [3,354], +# 'right_mouth_5': [3,357], +# 'right_lip_1': [3,360], +# 'right_lip_2': [3,363], +# 'lip_top': [3,366], +# 'left_lip_2': [3,369], + +# 'left_lip_1': [3,372], +# 'left_lip_3': [3,375], +# 'lip_bottom': [3,378], +# 'right_lip_3': [3,381], +# 'right_contour_1': [3,384], +# 'right_contour_2': [3,387], +# 'right_contour_3': [3,390], +# 'right_contour_4': [3,393], +# 'right_contour_5': [3,396], +# 'right_contour_6': [3,399], +# 'right_contour_7': [3,402], +# 'right_contour_8': [3,405], +# 'contour_middle': [3,408], +# 'left_contour_8': [3,411], +# 'left_contour_7': [3,414], +# 'left_contour_6': [3,417], +# 'left_contour_5': [3,420], +# 'left_contour_4': [3,423], +# 'left_contour_3': [3,426], +# 'left_contour_2': [3,429], +# 'left_contour_1': [3,432], + }, + + "beat_smplx_no_eyes": { + "pelvis":3, + "left_hip":3, + "right_hip":3, + "spine1":3, + "left_knee":3, + "right_knee":3, + "spine2":3, + "left_ankle":3, + "right_ankle":3, + "spine3":3, + "left_foot":3, + "right_foot":3, + "neck":3, + "left_collar":3, + "right_collar":3, + "head":3, + "left_shoulder":3, + "right_shoulder":3, + "left_elbow":3, + "right_elbow":3, + "left_wrist":3, + "right_wrist":3, + "jaw":3, + # "left_eye_smplhf":3, + # "right_eye_smplhf":3, + "left_index1":3, + "left_index2":3, + "left_index3":3, + "left_middle1":3, + "left_middle2":3, + "left_middle3":3, + "left_pinky1":3, + "left_pinky2":3, + "left_pinky3":3, + "left_ring1":3, + "left_ring2":3, + "left_ring3":3, + "left_thumb1":3, + "left_thumb2":3, + "left_thumb3":3, + "right_index1":3, + "right_index2":3, + "right_index3":3, + "right_middle1":3, + "right_middle2":3, + "right_middle3":3, + "right_pinky1":3, + "right_pinky2":3, + "right_pinky3":3, + "right_ring1":3, + "right_ring2":3, + "right_ring3":3, + "right_thumb1":3, + "right_thumb2":3, + "right_thumb3":3, + }, + + "beat_smplx_full": { + "pelvis":3, + "left_hip":3, + "right_hip":3, + "spine1":3, + "left_knee":3, + "right_knee":3, + "spine2":3, + "left_ankle":3, + "right_ankle":3, + "spine3":3, + "left_foot":3, + "right_foot":3, + "neck":3, + "left_collar":3, + "right_collar":3, + "head":3, + "left_shoulder":3, + "right_shoulder":3, + "left_elbow":3, + "right_elbow":3, + "left_wrist":3, + "right_wrist":3, + "jaw":3, + "left_eye_smplhf":3, + "right_eye_smplhf":3, + "left_index1":3, + "left_index2":3, + "left_index3":3, + "left_middle1":3, + "left_middle2":3, + "left_middle3":3, + "left_pinky1":3, + "left_pinky2":3, + "left_pinky3":3, + "left_ring1":3, + "left_ring2":3, + "left_ring3":3, + "left_thumb1":3, + "left_thumb2":3, + "left_thumb3":3, + "right_index1":3, + "right_index2":3, + "right_index3":3, + "right_middle1":3, + "right_middle2":3, + "right_middle3":3, + "right_pinky1":3, + "right_pinky2":3, + "right_pinky3":3, + "right_ring1":3, + "right_ring2":3, + "right_ring3":3, + "right_thumb1":3, + "right_thumb2":3, + "right_thumb3":3, + }, + + "beat_smplx_upall": { + # "pelvis":3, + # "left_hip":3, + # "right_hip":3, + "spine1":3, + # "left_knee":3, + # "right_knee":3, + "spine2":3, + # "left_ankle":3, + # "right_ankle":3, + "spine3":3, + # "left_foot":3, + # "right_foot":3, + "neck":3, + "left_collar":3, + "right_collar":3, + "head":3, + "left_shoulder":3, + "right_shoulder":3, + "left_elbow":3, + "right_elbow":3, + "left_wrist":3, + "right_wrist":3, + # "jaw":3, + # "left_eye_smplhf":3, + # "right_eye_smplhf":3, + "left_index1":3, + "left_index2":3, + "left_index3":3, + "left_middle1":3, + "left_middle2":3, + "left_middle3":3, + "left_pinky1":3, + "left_pinky2":3, + "left_pinky3":3, + "left_ring1":3, + "left_ring2":3, + "left_ring3":3, + "left_thumb1":3, + "left_thumb2":3, + "left_thumb3":3, + "right_index1":3, + "right_index2":3, + "right_index3":3, + "right_middle1":3, + "right_middle2":3, + "right_middle3":3, + "right_pinky1":3, + "right_pinky2":3, + "right_pinky3":3, + "right_ring1":3, + "right_ring2":3, + "right_ring3":3, + "right_thumb1":3, + "right_thumb2":3, + "right_thumb3":3, + }, + + "beat_smplx_upper": { + #"pelvis":3, + # "left_hip":3, + # "right_hip":3, + "spine1":3, + # "left_knee":3, + # "right_knee":3, + "spine2":3, + # "left_ankle":3, + # "right_ankle":3, + "spine3":3, + # "left_foot":3, + # "right_foot":3, + "neck":3, + "left_collar":3, + "right_collar":3, + "head":3, + "left_shoulder":3, + "right_shoulder":3, + "left_elbow":3, + "right_elbow":3, + "left_wrist":3, + "right_wrist":3, + # "jaw":3, + # "left_eye_smplhf":3, + # "right_eye_smplhf":3, + # "left_index1":3, + # "left_index2":3, + # "left_index3":3, + # "left_middle1":3, + # "left_middle2":3, + # "left_middle3":3, + # "left_pinky1":3, + # "left_pinky2":3, + # "left_pinky3":3, + # "left_ring1":3, + # "left_ring2":3, + # "left_ring3":3, + # "left_thumb1":3, + # "left_thumb2":3, + # "left_thumb3":3, + # "right_index1":3, + # "right_index2":3, + # "right_index3":3, + # "right_middle1":3, + # "right_middle2":3, + # "right_middle3":3, + # "right_pinky1":3, + # "right_pinky2":3, + # "right_pinky3":3, + # "right_ring1":3, + # "right_ring2":3, + # "right_ring3":3, + # "right_thumb1":3, + # "right_thumb2":3, + # "right_thumb3":3, + }, + + "beat_smplx_hands": { + #"pelvis":3, + # "left_hip":3, + # "right_hip":3, + # "spine1":3, + # "left_knee":3, + # "right_knee":3, + # "spine2":3, + # "left_ankle":3, + # "right_ankle":3, + # "spine3":3, + # "left_foot":3, + # "right_foot":3, + # "neck":3, + # "left_collar":3, + # "right_collar":3, + # "head":3, + # "left_shoulder":3, + # "right_shoulder":3, + # "left_elbow":3, + # "right_elbow":3, + # "left_wrist":3, + # "right_wrist":3, + # "jaw":3, + # "left_eye_smplhf":3, + # "right_eye_smplhf":3, + "left_index1":3, + "left_index2":3, + "left_index3":3, + "left_middle1":3, + "left_middle2":3, + "left_middle3":3, + "left_pinky1":3, + "left_pinky2":3, + "left_pinky3":3, + "left_ring1":3, + "left_ring2":3, + "left_ring3":3, + "left_thumb1":3, + "left_thumb2":3, + "left_thumb3":3, + "right_index1":3, + "right_index2":3, + "right_index3":3, + "right_middle1":3, + "right_middle2":3, + "right_middle3":3, + "right_pinky1":3, + "right_pinky2":3, + "right_pinky3":3, + "right_ring1":3, + "right_ring2":3, + "right_ring3":3, + "right_thumb1":3, + "right_thumb2":3, + "right_thumb3":3, + }, + + "beat_smplx_lower": { + "pelvis":3, + "left_hip":3, + "right_hip":3, + # "spine1":3, + "left_knee":3, + "right_knee":3, + # "spine2":3, + "left_ankle":3, + "right_ankle":3, + # "spine3":3, + "left_foot":3, + "right_foot":3, + # "neck":3, + # "left_collar":3, + # "right_collar":3, + # "head":3, + # "left_shoulder":3, + # "right_shoulder":3, + # "left_elbow":3, + # "right_elbow":3, + # "left_wrist":3, + # "right_wrist":3, + # "jaw":3, + # "left_eye_smplhf":3, + # "right_eye_smplhf":3, + # "left_index1":3, + # "left_index2":3, + # "left_index3":3, + # "left_middle1":3, + # "left_middle2":3, + # "left_middle3":3, + # "left_pinky1":3, + # "left_pinky2":3, + # "left_pinky3":3, + # "left_ring1":3, + # "left_ring2":3, + # "left_ring3":3, + # "left_thumb1":3, + # "left_thumb2":3, + # "left_thumb3":3, + # "right_index1":3, + # "right_index2":3, + # "right_index3":3, + # "right_middle1":3, + # "right_middle2":3, + # "right_middle3":3, + # "right_pinky1":3, + # "right_pinky2":3, + # "right_pinky3":3, + # "right_ring1":3, + # "right_ring2":3, + # "right_ring3":3, + # "right_thumb1":3, + # "right_thumb2":3, + # "right_thumb3":3, + }, + + "beat_smplx_face": { + # "pelvis":3, + # "left_hip":3, + # "right_hip":3, + # # "spine1":3, + # "left_knee":3, + # "right_knee":3, + # # "spine2":3, + # "left_ankle":3, + # "right_ankle":3, + # # "spine3":3, + # "left_foot":3, + # "right_foot":3, + # "neck":3, + # "left_collar":3, + # "right_collar":3, + # "head":3, + # "left_shoulder":3, + # "right_shoulder":3, + # "left_elbow":3, + # "right_elbow":3, + # "left_wrist":3, + # "right_wrist":3, + "jaw":3, + # "left_eye_smplhf":3, + # "right_eye_smplhf":3, + # "left_index1":3, + # "left_index2":3, + # "left_index3":3, + # "left_middle1":3, + # "left_middle2":3, + # "left_middle3":3, + # "left_pinky1":3, + # "left_pinky2":3, + # "left_pinky3":3, + # "left_ring1":3, + # "left_ring2":3, + # "left_ring3":3, + # "left_thumb1":3, + # "left_thumb2":3, + # "left_thumb3":3, + # "right_index1":3, + # "right_index2":3, + # "right_index3":3, + # "right_middle1":3, + # "right_middle2":3, + # "right_middle3":3, + # "right_pinky1":3, + # "right_pinky2":3, + # "right_pinky3":3, + # "right_ring1":3, + # "right_ring2":3, + # "right_ring3":3, + # "right_thumb1":3, + # "right_thumb2":3, + # "right_thumb3":3, + }, + + "beat_joints": { + 'Hips': [6,6], + 'Spine': [3,9], + 'Spine1': [3,12], + 'Spine2': [3,15], + 'Spine3': [3,18], + 'Neck': [3,21], + 'Neck1': [3,24], + 'Head': [3,27], + 'HeadEnd': [3,30], + + 'RShoulder': [3,33], + 'RArm': [3,36], + 'RArm1': [3,39], + 'RHand': [3,42], + 'RHandM1': [3,45], + 'RHandM2': [3,48], + 'RHandM3': [3,51], + 'RHandM4': [3,54], + + 'RHandR': [3,57], + 'RHandR1': [3,60], + 'RHandR2': [3,63], + 'RHandR3': [3,66], + 'RHandR4': [3,69], + + 'RHandP': [3,72], + 'RHandP1': [3,75], + 'RHandP2': [3,78], + 'RHandP3': [3,81], + 'RHandP4': [3,84], + + 'RHandI': [3,87], + 'RHandI1': [3,90], + 'RHandI2': [3,93], + 'RHandI3': [3,96], + 'RHandI4': [3,99], + + 'RHandT1': [3,102], + 'RHandT2': [3,105], + 'RHandT3': [3,108], + 'RHandT4': [3,111], + + 'LShoulder': [3,114], + 'LArm': [3,117], + 'LArm1': [3,120], + 'LHand': [3,123], + 'LHandM1': [3,126], + 'LHandM2': [3,129], + 'LHandM3': [3,132], + 'LHandM4': [3,135], + + 'LHandR': [3,138], + 'LHandR1': [3,141], + 'LHandR2': [3,144], + 'LHandR3': [3,147], + 'LHandR4': [3,150], + + 'LHandP': [3,153], + 'LHandP1': [3,156], + 'LHandP2': [3,159], + 'LHandP3': [3,162], + 'LHandP4': [3,165], + + 'LHandI': [3,168], + 'LHandI1': [3,171], + 'LHandI2': [3,174], + 'LHandI3': [3,177], + 'LHandI4': [3,180], + + 'LHandT1': [3,183], + 'LHandT2': [3,186], + 'LHandT3': [3,189], + 'LHandT4': [3,192], + + 'RUpLeg': [3,195], + 'RLeg': [3,198], + 'RFoot': [3,201], + 'RFootF': [3,204], + 'RToeBase': [3,207], + 'RToeBaseEnd': [3,210], + + 'LUpLeg': [3,213], + 'LLeg': [3,216], + 'LFoot': [3,219], + 'LFootF': [3,222], + 'LToeBase': [3,225], + 'LToeBaseEnd': [3,228],}, + + "beat_full":{ + 'Hips': 3, + 'Spine': 3 , + 'Spine1': 3 , + 'Spine2': 3 , + 'Spine3': 3 , + 'Neck': 3 , + 'Neck1': 3 , + 'Head' : 3, + 'HeadEnd' : 3, + 'RShoulder': 3 , + 'RArm': 3 , + 'RArm1': 3 , + 'RHand': 3 , + 'RHandM1': 3 , + 'RHandM2': 3 , + 'RHandM3': 3 , + 'RHandM4': 3 , + 'RHandR': 3 , + 'RHandR1': 3 , + 'RHandR2': 3 , + 'RHandR3': 3 , + 'RHandR4': 3 , + 'RHandP': 3 , + 'RHandP1': 3 , + 'RHandP2': 3 , + 'RHandP3': 3 , + 'RHandP4': 3 , + 'RHandI': 3 , + 'RHandI1': 3 , + 'RHandI2': 3 , + 'RHandI3': 3 , + 'RHandI4': 3 , + 'RHandT1': 3 , + 'RHandT2': 3 , + 'RHandT3': 3 , + 'RHandT4': 3 , + 'LShoulder': 3 , + 'LArm': 3 , + 'LArm1': 3 , + 'LHand': 3 , + 'LHandM1': 3 , + 'LHandM2': 3 , + 'LHandM3': 3 , + 'LHandM4': 3 , + 'LHandR': 3 , + 'LHandR1': 3 , + 'LHandR2': 3 , + 'LHandR3': 3 , + 'LHandR4': 3 , + 'LHandP': 3 , + 'LHandP1': 3 , + 'LHandP2': 3 , + 'LHandP3': 3 , + 'LHandP4': 3 , + 'LHandI': 3 , + 'LHandI1': 3 , + 'LHandI2': 3 , + 'LHandI3': 3 , + 'LHandI4': 3 , + 'LHandT1': 3 , + 'LHandT2': 3 , + 'LHandT3': 3 , + 'LHandT4': 3 , + 'RUpLeg': 3, + 'RLeg': 3, + 'RFoot': 3, + 'RFootF': 3, + 'RToeBase': 3, + 'RToeBaseEnd': 3, + 'LUpLeg': 3, + 'LLeg': 3, + 'LFoot': 3, + 'LFootF': 3, + 'LToeBase': 3, + 'LToeBaseEnd': 3, + }, + + "japanese_joints":{ + 'Hips': [6,6], + 'Spine': [6,12], + 'Spine1': [6,18], + 'Spine2': [6,24], + 'Spine3': [6,30], + 'Neck': [6,36], + 'Neck1': [6,42], + 'Head': [6,48], + 'RShoulder': [6,54], + 'RArm': [6,60], + 'RArm1': [6,66], + 'RHand': [6,72], + 'RHandM1': [6,78], + 'RHandM2': [6,84], + 'RHandM3': [6,90], + 'RHandR': [6,96], + 'RHandR1': [6,102], + 'RHandR2': [6,108], + 'RHandR3': [6,114], + 'RHandP': [6,120], + 'RHandP1': [6,126], + 'RHandP2': [6,132], + 'RHandP3': [6,138], + 'RHandI': [6,144], + 'RHandI1': [6,150], + 'RHandI2': [6,156], + 'RHandI3': [6,162], + 'RHandT1': [6,168], + 'RHandT2': [6,174], + 'RHandT3': [6,180], + 'LShoulder': [6,186], + 'LArm': [6,192], + 'LArm1': [6,198], + 'LHand': [6,204], + 'LHandM1': [6,210], + 'LHandM2': [6,216], + 'LHandM3': [6,222], + 'LHandR': [6,228], + 'LHandR1': [6,234], + 'LHandR2': [6,240], + 'LHandR3': [6,246], + 'LHandP': [6,252], + 'LHandP1': [6,258], + 'LHandP2': [6,264], + 'LHandP3': [6,270], + 'LHandI': [6,276], + 'LHandI1': [6,282], + 'LHandI2': [6,288], + 'LHandI3': [6,294], + 'LHandT1': [6,300], + 'LHandT2': [6,306], + 'LHandT3': [6,312], + 'RUpLeg': [6,318], + 'RLeg': [6,324], + 'RFoot': [6,330], + 'RFootF': [6,336], + 'RToeBase': [6,342], + 'LUpLeg': [6,348], + 'LLeg': [6,354], + 'LFoot': [6,360], + 'LFootF': [6,366], + 'LToeBase': [6,372],}, + + "yostar":{ + 'Hips': [6,6], + 'Spine': [3,9], + 'Spine1': [3,12], + 'Bone040': [3,15], + 'Bone041': [3,18], + + 'Bone034': [3,21], + 'Bone035': [3,24], + 'Bone036': [3,27], + 'Bone037': [3,30], + 'Bone038': [3,33], + 'Bone039': [3,36], + + 'RibbonL1': [3,39], + 'RibbonL1_end': [3,42], + + 'Chest': [3,45], + 'L_eri': [3,48], + 'R_eri': [3,51], + 'Neck': [3,54], + 'Head': [3,57], + 'Head_end': [3,60], + + 'RBackHair_1': [3,63], + 'RBackHair_2': [3,66], + 'RBackHair_3': [3,69], + 'RBackHair_4': [3,72], + 'RBackHair_end': [3,75], + + 'RFrontHair': [3,78], + 'CFrontHair_1': [3,81], + 'CFrontHair_2': [3,84], + 'CFrontHair_3': [3,87], + 'CFrontHair_emd': [3,90], + + 'LFrontHair_1': [3,93], + 'LFrontHair_2': [3,96], + 'LFrontHair_3': [3,99], + + 'LBackHair_1': [3,102], + 'LBackHair_2': [3,105], + 'LBackHair_3': [3,108], + 'LBackHair_4': [3,111], + 'LBackHair_end': [3,114], + + 'LSideHair_1': [3,117], + 'LSideHair_2': [3,120], + 'LSideHair_3': [3,123], + 'LSideHair_4': [3,126], + 'LSideHair_5': [3,129], + 'LSideHair_6': [3,132], + 'LSideHair_7': [3,135], + 'LSideHair_end': [3,138], + + 'CBackHair_1': [3,141], + 'CBackHair_2': [3,144], + 'CBackHair_3': [3,147], + 'CBackHair_4': [3,150], + 'CBackHair_end': [3,153], + + 'RSideHair_1': [3,156], + 'RSideHair_2': [3,159], + 'RSideHair_3': [3,162], + 'RSideHair_4': [3,165], + + 'RibbonR_1': [3,168], + 'RibbonR_2': [3,171], + 'RibbonR_3': [3,174], + + 'RibbonL_1': [3,177], + 'RibbonL_2': [3,180], + 'RibbonL_3': [3,183], + + 'LeftEye': [3,186], + 'LeftEye_end': [3,189], + 'RightEye': [3,192], + 'RightEye_end': [3,195], + + 'LeftShoulder': [3,198], + 'LeftArm': [3,201], + 'LeftForearm': [3,204], + 'LeftHand': [3,207], + 'LeftHandThumb1': [3,210], + 'LeftHandThumb2': [3,213], + 'LeftHandThumb3': [3,216], + 'LeftHandThumb_end': [3,219], + + 'LeftHandIndex1': [3,222], + 'LeftHandIndex2': [3,225], + 'LeftHandIndex3': [3,228], + 'LeftHandIndex_end': [3,231], + + 'LeftHandMiddle1': [3,234], + 'LeftHandMiddle2': [3,237], + 'LeftHandMiddle3': [3,240], + 'LeftHandMiddle_end': [3,243], + + 'LeftHandRing1': [3,246], + 'LeftHandRing2': [3,249], + 'LeftHandRing3': [3,252], + 'LeftHandRing_end': [3,255], + + 'LeftHandPinky1': [3,258], + 'LeftHandPinky2': [3,261], + 'LeftHandPinky3': [3,264], + 'LeftHandPinky_end': [3,267], + + 'RightShoulder': [3,270], + 'RightArm': [3,273], + 'RightForearm': [3,276], + 'RightHand': [3,279], + 'RightHandThumb1': [3,282], + 'RightHandThumb2': [3,285], + 'RightHandThumb3': [3,288], + 'RightHandThumb_end': [3,291], + + 'RightHandIndex1': [3,294], + 'RightHandIndex2': [3,297], + 'RightHandIndex3': [3,300], + 'RightHandIndex_end': [3,303], + + 'RightHandMiddle1': [3,306], + 'RightHandMiddle2': [3,309], + 'RightHandMiddle3': [3,312], + 'RightHandMiddle_end': [3,315], + + 'RightHandRing1': [3,318], + 'RightHandRing2': [3,321], + 'RightHandRing3': [3,324], + 'RightHandRing_end': [3,327], + + 'RightHandPinky1': [3,330], + 'RightHandPinky2': [3,333], + 'RightHandPinky3': [3,336], + 'RightHandPinky_end': [3,339], + + 'RibbonR1': [3,342], + 'RibbonR1_end': [3,345], + 'RibbonR2': [3,348], + 'RibbonR2_end': [3,351], + 'RibbonL2': [3,354], + 'RibbonL2_end': [3,357], + + 'LeftUpLeg': [3,360], + 'LeftLeg': [3,363], + 'LeftFoot': [3,366], + 'LeftToe': [3,369], + 'LeftToe_end': [3,372], + + 'RightUpLeg': [3,375], + 'RightLEg': [3,378], + 'RightFoot': [3,381], + 'RightToe': [3,384], + 'RightToe_end': [3,387], + + 'bone_skirtF00': [3, 390], + 'bone_skirtF01': [3, 393], + 'bone_skirtF02': [3, 396], + 'bone_skirtF03': [3, 399], + 'Bone020': [3, 402], + 'Bone026': [3, 405], + + 'bone_skirtF_R_00': [3, 408], + 'bone_skirtF_R_01': [3, 411], + 'bone_skirtF_R_02': [3, 414], + 'bone_skirtF_R_03': [3, 417], + 'Bone019': [3, 420], + 'Bone028': [3, 423], + + 'bone_skirtR00': [3, 426], + 'bone_skirtR01': [3, 429], + 'bone_skirtR02': [3, 432], + 'bone_skirtR03': [3, 435], + 'Bone018': [3, 438], + 'Bone029': [3, 441], + + 'bone_skirtF_L_00': [3, 444], + 'bone_skirtF_L_01': [3, 447], + 'bone_skirtF_L_02': [3, 450], + 'bone_skirtF_L_03': [3, 453], + 'Bone021': [3, 456], + 'Bone027': [3, 459], + + 'bone_skirtL00': [3, 462], + 'bone_skirtL01': [3, 465], + 'bone_skirtL02': [3, 468], + 'bone_skirtL03': [3, 471], + 'Bone022': [3, 474], + 'Bone033': [3, 477], + + 'bone_skirtB_L_00': [3, 480], + 'bone_skirtB_L_01': [3, 483], + 'bone_skirtB_L_02': [3, 486], + 'bone_skirtB_L_03': [3, 489], + 'Bone023': [3, 492], + 'Bone032': [3, 495], + + 'bone_skirtB00': [3, 498], + 'bone_skirtB01': [3, 501], + 'bone_skirtB02': [3, 504], + 'bone_skirtB03': [3, 507], + 'Bone024': [3, 510], + 'Bone031': [3, 513], + + 'bone_skirtB_R_00': [3, 516], + 'bone_skirtB_R_01': [3, 519], + 'bone_skirtB_R_02': [3, 521], + 'bone_skirtB_R_03': [3, 524], + 'Bone025': [3, 527], + 'Bone030': [3, 530], + }, + + "yostar_fullbody_213":{ + 'Hips': 3 , + 'Spine': 3 , + 'Spine1': 3 , + 'Chest': 3 , + 'L_eri': 3 , + 'R_eri': 3 , + 'Neck': 3 , + 'Head': 3 , + 'Head_end': 3 , + + 'LeftEye': 3, + 'LeftEye_end': 3, + 'RightEye': 3, + 'RightEye_end': 3, + + 'LeftShoulder': 3, + 'LeftArm': 3, + 'LeftForearm': 3, + 'LeftHand': 3, + 'LeftHandThumb1': 3, + 'LeftHandThumb2': 3, + 'LeftHandThumb3': 3, + 'LeftHandThumb_end': 3, + + 'LeftHandIndex1': 3, + 'LeftHandIndex2': 3, + 'LeftHandIndex3': 3, + 'LeftHandIndex_end': 3, + + 'LeftHandMiddle1': 3, + 'LeftHandMiddle2': 3, + 'LeftHandMiddle3': 3, + 'LeftHandMiddle_end': 3, + + 'LeftHandRing1': 3, + 'LeftHandRing2': 3, + 'LeftHandRing3': 3, + 'LeftHandRing_end': 3, + + 'LeftHandPinky1': 3, + 'LeftHandPinky2': 3, + 'LeftHandPinky3': 3, + 'LeftHandPinky_end':3, + + 'RightShoulder': 3, + 'RightArm': 3, + 'RightForearm': 3, + 'RightHand': 3, + 'RightHandThumb1': 3, + 'RightHandThumb2': 3, + 'RightHandThumb3': 3, + 'RightHandThumb_end': 3, + + 'RightHandIndex1': 3, + 'RightHandIndex2': 3, + 'RightHandIndex3': 3, + 'RightHandIndex_end': 3, + + 'RightHandMiddle1': 3, + 'RightHandMiddle2': 3, + 'RightHandMiddle3': 3, + 'RightHandMiddle_end': 3, + + 'RightHandRing1': 3, + 'RightHandRing2': 3, + 'RightHandRing3': 3, + 'RightHandRing_end': 3, + + 'RightHandPinky1': 3, + 'RightHandPinky2': 3, + 'RightHandPinky3': 3, + 'RightHandPinky_end': 3, + + 'LeftUpLeg': 3, + 'LeftLeg': 3, + 'LeftFoot': 3, + 'LeftToe': 3, + 'LeftToe_end': 3, + + 'RightUpLeg': 3, + 'RightLEg': 3, + 'RightFoot': 3, + 'RightToe': 3, + 'RightToe_end': 3, + }, + "yostar_mainbody_48": { + #'Hips': 3 , + 'Spine': 3 , + 'Spine1': 3 , + 'Chest': 3 , + 'L_eri': 3 , + 'R_eri': 3 , + 'Neck': 3 , + 'Head': 3 , + 'Head_end': 3 , + + 'LeftShoulder': 3, + 'LeftArm': 3, + 'LeftForearm': 3, + 'LeftHand': 3, + + 'RightShoulder': 3, + 'RightArm': 3, + 'RightForearm': 3, + 'RightHand': 3, + }, + "yostar_mainbody_69": { + 'Hips': 3 , + 'Spine': 3 , + 'Spine1': 3 , + 'Chest': 3 , + 'L_eri': 3 , + 'R_eri': 3 , + 'Neck': 3 , + 'Head': 3 , + 'Head_end': 3 , + + 'LeftShoulder': 3, + 'LeftArm': 3, + 'LeftForearm': 3, + 'LeftHand': 3, + + 'RightShoulder': 3, + 'RightArm': 3, + 'RightForearm': 3, + 'RightHand': 3, + + 'LeftUpLeg': 3, + 'LeftLeg': 3, + 'LeftFoot': 3, + + 'RightUpLeg': 3, + 'RightLEg': 3, + 'RightFoot': 3, + }, + + "yostar_upbody_168": { + #'Hips': 3 , + 'Spine': 3 , + 'Spine1': 3 , + 'Chest': 3 , + 'L_eri': 3 , + 'R_eri': 3 , + 'Neck': 3 , + 'Head': 3 , + 'Head_end': 3 , + + 'LeftShoulder': 3, + 'LeftArm': 3, + 'LeftForearm': 3, + 'LeftHand': 3, + 'LeftHandThumb1': 3, + 'LeftHandThumb2': 3, + 'LeftHandThumb3': 3, + 'LeftHandThumb_end': 3, + + 'LeftHandIndex1': 3, + 'LeftHandIndex2': 3, + 'LeftHandIndex3': 3, + 'LeftHandIndex_end': 3, + + 'LeftHandMiddle1': 3, + 'LeftHandMiddle2': 3, + 'LeftHandMiddle3': 3, + 'LeftHandMiddle_end': 3, + + 'LeftHandRing1': 3, + 'LeftHandRing2': 3, + 'LeftHandRing3': 3, + 'LeftHandRing_end': 3, + + 'LeftHandPinky1': 3, + 'LeftHandPinky2': 3, + 'LeftHandPinky3': 3, + 'LeftHandPinky_end':3, + + 'RightShoulder': 3, + 'RightArm': 3, + 'RightForearm': 3, + 'RightHand': 3, + 'RightHandThumb1': 3, + 'RightHandThumb2': 3, + 'RightHandThumb3': 3, + 'RightHandThumb_end': 3, + + 'RightHandIndex1': 3, + 'RightHandIndex2': 3, + 'RightHandIndex3': 3, + 'RightHandIndex_end': 3, + + 'RightHandMiddle1': 3, + 'RightHandMiddle2': 3, + 'RightHandMiddle3': 3, + 'RightHandMiddle_end': 3, + + 'RightHandRing1': 3, + 'RightHandRing2': 3, + 'RightHandRing3': 3, + 'RightHandRing_end': 3, + + 'RightHandPinky1': 3, + 'RightHandPinky2': 3, + 'RightHandPinky3': 3, + 'RightHandPinky_end': 3, + }, + "spine_neck_141":{ + 'Spine': 3 , + 'Neck': 3 , + 'Neck1': 3 , + 'RShoulder': 3 , + 'RArm': 3 , + 'RArm1': 3 , + 'RHand': 3 , + 'RHandM1': 3 , + 'RHandM2': 3 , + 'RHandM3': 3 , + 'RHandR': 3 , + 'RHandR1': 3 , + 'RHandR2': 3 , + 'RHandR3': 3 , + 'RHandP': 3 , + 'RHandP1': 3 , + 'RHandP2': 3 , + 'RHandP3': 3 , + 'RHandI': 3 , + 'RHandI1': 3 , + 'RHandI2': 3 , + 'RHandI3': 3 , + 'RHandT1': 3 , + 'RHandT2': 3 , + 'RHandT3': 3 , + 'LShoulder': 3 , + 'LArm': 3 , + 'LArm1': 3 , + 'LHand': 3 , + 'LHandM1': 3 , + 'LHandM2': 3 , + 'LHandM3': 3 , + 'LHandR': 3 , + 'LHandR1': 3 , + 'LHandR2': 3 , + 'LHandR3': 3 , + 'LHandP': 3 , + 'LHandP1': 3 , + 'LHandP2': 3 , + 'LHandP3': 3 , + 'LHandI': 3 , + 'LHandI1': 3 , + 'LHandI2': 3 , + 'LHandI3': 3 , + 'LHandT1': 3 , + 'LHandT2': 3 , + 'LHandT3': 3 ,}, +} + + +class FIDCalculator(object): + ''' + todo + ''' + def __init__(self): + self.gt_rot = None # pandas dataframe for n frames * joints * 6 + self.gt_pos = None # n frames * (joints + 13) * 3 + self.op_rot = None # pandas dataframe for n frames * joints * 6 + self.op_pos = None # n frames * (joints + 13) * 3 + + + def load(self, path, load_type, save_pos=False): + ''' + select gt or op for load_type + ''' + parser = BVHParser() + parsed_data = parser.parse(path) + if load_type == 'gt': + self.gt_rot = parsed_data.values + elif load_type == 'op': + self.op_rot = parsed_data.values + else: print('error, select gt or op for load_type') + + if save_pos: + mp = MocapParameterizer('position') + positions = mp.fit_transform([parsed_data]) + if load_type == 'gt': + self.gt_pos = positions[0].values + elif load_type == 'op': + self.op_pos = positions[0].values + else: print('error, select gt or op for load_type') + + + def _joint_selector(self, selected_joints, ori_data): + selected_data = pd.DataFrame(columns=[]) + + for joint_name in selected_joints: + selected_data[joint_name] = ori_data[joint_name] + return selected_data.to_numpy() + + + def cal_vol(self, dtype): + if dtype == 'pos': + gt = self.gt_pos + op = self.op_pos + else: + gt = self.gt_rot + op = self.op_rot + + gt_v = gt.to_numpy()[1:, :] - gt.to_numpy()[0:-1, :] + op_v = op.to_numpy()[1:, :] - op.to_numpy()[0:-1, :] + if dtype == 'pos': + self.gt_vol_pos = pd.DataFrame(gt_v, columns = gt.columns.tolist()) + self.op_vol_pos = pd.DataFrame(op_v, columns = gt.columns.tolist()) + else: + self.gt_vol_rot = pd.DataFrame(gt_v, columns = gt.columns.tolist()) + self.op_vol_rot = pd.DataFrame(op_v, columns = gt.columns.tolist()) + + + @staticmethod + def frechet_distance(samples_A, samples_B): + A_mu = np.mean(samples_A, axis=0) + A_sigma = np.cov(samples_A, rowvar=False) + B_mu = np.mean(samples_B, axis=0) + B_sigma = np.cov(samples_B, rowvar=False) + try: + frechet_dist = FIDCalculator.calculate_frechet_distance(A_mu, A_sigma, B_mu, B_sigma) + except ValueError: + frechet_dist = 1e+10 + return frechet_dist + + + @staticmethod + def calculate_frechet_distance(mu1, sigma1, mu2, sigma2, eps=1e-6): + """ from https://github.com/mseitzer/pytorch-fid/blob/master/fid_score.py """ + """Numpy implementation of the Frechet Distance. + The Frechet distance between two multivariate Gaussians X_1 ~ N(mu_1, C_1) + and X_2 ~ N(mu_2, C_2) is + d^2 = ||mu_1 - mu_2||^2 + Tr(C_1 + C_2 - 2*sqrt(C_1*C_2)). + Stable version by Dougal J. Sutherland. + Params: + -- mu1 : Numpy array containing the activations of a layer of the + inception net (like returned by the function 'get_predictions') + for generated samples. + -- mu2 : The sample mean over activations, precalculated on an + representative data set. + -- sigma1: The covariance matrix over activations for generated samples. + -- sigma2: The covariance matrix over activations, precalculated on an + representative data set. + Returns: + -- : The Frechet Distance. + """ + + mu1 = np.atleast_1d(mu1) + mu2 = np.atleast_1d(mu2) + #print(mu1[0], mu2[0]) + sigma1 = np.atleast_2d(sigma1) + sigma2 = np.atleast_2d(sigma2) + #print(sigma1[0], sigma2[0]) + assert mu1.shape == mu2.shape, \ + 'Training and test mean vectors have different lengths' + assert sigma1.shape == sigma2.shape, \ + 'Training and test covariances have different dimensions' + + diff = mu1 - mu2 + + # Product might be almost singular + covmean, _ = linalg.sqrtm(sigma1.dot(sigma2), disp=False) + #print(diff, covmean[0]) + if not np.isfinite(covmean).all(): + msg = ('fid calculation produces singular product; ' + 'adding %s to diagonal of cov estimates') % eps + print(msg) + offset = np.eye(sigma1.shape[0]) * eps + covmean = linalg.sqrtm((sigma1 + offset).dot(sigma2 + offset)) + + # Numerical error might give slight imaginary component + if np.iscomplexobj(covmean): + if not np.allclose(np.diagonal(covmean).imag, 0, atol=1e-3): + m = np.max(np.abs(covmean.imag)) + raise ValueError('Imaginary component {}'.format(m)) + covmean = covmean.real + + tr_covmean = np.trace(covmean) + + return (diff.dot(diff) + np.trace(sigma1) + + np.trace(sigma2) - 2 * tr_covmean) + + + def calculate_fid(self, cal_type, joint_type, high_level_opt): + + if cal_type == 'pos': + if self.gt_pos.shape != self.op_pos.shape: + min_val = min(self.gt_pos.shape[0],self.op_pos.shape[0]) + gt = self.gt_pos[:min_val] + op = self.op_pos[:min_val] + else: + gt = self.gt_pos + op = self.op_pos + full_body = gt.columns.tolist() + elif cal_type == 'rot': + if self.gt_rot.shape != self.op_rot.shape: + min_val = min(self.gt_rot.shape[0],self.op_rot.shape[0]) + gt = self.gt_rot[:min_val] + op = self.op_rot[:min_val] + else: + gt = self.gt_rot + op = self.op_rot + full_body_with_offset = gt.columns.tolist() + full_body = [o for o in full_body_with_offset if ('position' not in o)] + elif cal_type == 'pos_vol': + assert self.gt_vol_pos.shape == self.op_vol_pos.shape + gt = self.gt_vol_pos + op = self.op_vol_pos + full_body_with_offset = gt.columns.tolist() + full_body = gt.columns.tolist() + elif cal_type == 'rot_vol': + assert self.gt_vol_rot.shape == self.op_vol_rot.shape + gt = self.gt_vol_rot + op = self.op_vol_rot + full_body_with_offset = gt.columns.tolist() + full_body = [o for o in full_body_with_offset if ('position' not in o)] + #print(f'full_body contains {len(full_body)//3} joints') + + if joint_type == 'full_upper_body': + selected_body = [o for o in full_body if ('Leg' not in o) and ('Foot' not in o) and ('Toe' not in o)] + elif joint_type == 'upper_body': + selected_body = [o for o in full_body if ('Hand' not in o) and ('Leg' not in o) and ('Foot' not in o) and ('Toe' not in o)] + elif joint_type == 'fingers': + selected_body = [o for o in full_body if ('Hand' in o)] + elif joint_type == 'indivdual': + pass + else: print('error, plz select correct joint type') + #print(f'calculate fid for {len(selected_body)//3} joints') + + gt = self._joint_selector(selected_body, gt) + op = self._joint_selector(selected_body, op) + + if high_level_opt == 'fid': + fid = FIDCalculator.frechet_distance(gt, op) + return fid + elif high_level_opt == 'var': + var_gt = gt.var() + var_op = op.var() + return var_gt, var_op + elif high_level_opt == 'mean': + mean_gt = gt.mean() + mean_op = op.mean() + return mean_gt, mean_op + else: return 0 + + +def result2target_vis(pose_version, res_bvhlist, save_path, demo_name, verbose=True): + if "trinity" in pose_version: + ori_list = joints_list[pose_version[6:-4]] + target_list = joints_list[pose_version[6:]] + file_content_length = 336 + elif "beat" in pose_version or "spine_neck_141" in pose_version: + ori_list = joints_list["beat_joints"] + target_list = joints_list["spine_neck_141"] + file_content_length = 431 + elif "yostar" in pose_version: + ori_list = joints_list["yostar"] + target_list = joints_list[pose_version] + file_content_length = 1056 + else: + ori_list = joints_list["japanese_joints"] + target_list = joints_list[pose_version] + file_content_length = 366 + + bvh_files_dirs = sorted(glob.glob(f'{res_bvhlist}*.bvh'), key=str) + #test_seq_list = os.list_dir(demo_name).sort() + + counter = 0 + if not os.path.exists(save_path): + os.makedirs(save_path) + for i, bvh_file_dir in enumerate(bvh_files_dirs): + short_name = bvh_file_dir.split("/")[-1][11:] + #print(short_name) + wirte_file = open(os.path.join(save_path, f'res_{short_name}'),'w+') + with open(f"{demo_name}{short_name}",'r') as pose_data_pre: + pose_data_pre_file = pose_data_pre.readlines() + for j, line in enumerate(pose_data_pre_file[0:file_content_length]): + wirte_file.write(line) + offset_data = pose_data_pre_file[file_content_length] + offset_data = np.fromstring(offset_data, dtype=float, sep=' ') + wirte_file.close() + + wirte_file = open(os.path.join(save_path, f'res_{short_name}'),'r') + ori_lines = wirte_file.readlines() + with open(bvh_file_dir, 'r') as pose_data: + pose_data_file = pose_data.readlines() + ori_lines[file_content_length-2] = 'Frames: ' + str(len(pose_data_file)-1) + '\n' + wirte_file.close() + + wirte_file = open(os.path.join(save_path, f'res_{short_name}'),'w+') + wirte_file.writelines(i for i in ori_lines[:file_content_length]) + wirte_file.close() + + with open(os.path.join(save_path, f'res_{short_name}'),'a+') as wirte_file: + with open(bvh_file_dir, 'r') as pose_data: + data_each_file = [] + pose_data_file = pose_data.readlines() + for j, line in enumerate(pose_data_file): + if not j: + pass + else: + data = np.fromstring(line, dtype=float, sep=' ') + data_rotation = offset_data.copy() + for iii, (k, v) in enumerate(target_list.items()): # here is 147 rotations by 3 + #print(data_rotation[ori_list[k][1]-v:ori_list[k][1]], data[iii*3:iii*3+3]) + data_rotation[ori_list[k][1]-v:ori_list[k][1]] = data[iii*3:iii*3+3] + data_each_file.append(data_rotation) + + for line_data in data_each_file: + line_data = np.array2string(line_data, max_line_width=np.inf, precision=6, suppress_small=False, separator=' ') + wirte_file.write(line_data[1:-2]+'\n') + + counter += 1 + if verbose: + logger.info('data_shape:', data_rotation.shape, 'process:', counter, '/', len(bvh_files_dirs)) \ No newline at end of file diff --git a/dataloaders/mix_sep.py b/dataloaders/mix_sep.py new file mode 100644 index 0000000000000000000000000000000000000000..4e5fbe024806909c6615d261767919e285c64365 --- /dev/null +++ b/dataloaders/mix_sep.py @@ -0,0 +1,637 @@ +import os +import pickle +import math +import shutil +import numpy as np +import lmdb as lmdb +import textgrid as tg +import pandas as pd +import torch +import glob +import json +from termcolor import colored +from loguru import logger +from collections import defaultdict +from torch.utils.data import Dataset +import torch.distributed as dist +#import pyarrow +import pickle +import librosa +import smplx +import glob + +from .build_vocab import Vocab +from .utils.audio_features import Wav2Vec2Model +from .data_tools import joints_list +from .utils import rotation_conversions as rc +from .utils import other_tools + +# ACCAD 120 +# BioMotionLab_NTroje 120 +# CMU 很复杂 +# EKUT 100 +# Eyes_Japan_Dataset 很复杂 +# HumanEva 很复杂 +# KIT 100 +# MPI_HDM05 120 +# MPI_Limits 120 +# MPI_mosh 很复杂 +# SFU 120 +# SSM_synced 很复杂 +# TCD_handMocap 很复杂 +# TotalCapture 60 +# Transitions_mocap 120 + +all_sequences = [ + 'ACCAD', + 'BioMotionLab_NTroje', + 'CMU', + 'EKUT', + 'Eyes_Japan_Dataset', + 'HumanEva', + 'KIT', + 'MPI_HDM05', + 'MPI_Limits', + 'MPI_mosh', + 'SFU', + 'SSM_synced', + 'TCD_handMocap', + 'TotalCapture', + 'Transitions_mocap', +] +amass_test_split = ['Transitions_mocap', 'SSM_synced'] +amass_vald_split = ['HumanEva', 'MPI_HDM05', 'SFU', 'MPI_mosh'] +amass_train_split = ['BioMotionLab_NTroje', 'Eyes_Japan_Dataset', 'TotalCapture', 'KIT', 'ACCAD', 'CMU', 'MPI_Limits', + 'TCD_handMocap', 'EKUT'] + +# 上面这些spilt方式是MOTION CLIP的，但是由于motionx中的framerate处理有问题，我先暂且只挑部分数据集进行训练 +# 这些都是120fps的 +# amass_test_split = ['SFU'] +# amass_vald_split = ['MPI_Limits'] +# amass_train_split = ['BioMotionLab_NTroje', 'MPI_HDM05', 'ACCAD','Transitions_mocap'] + + +amass_splits = { + 'test': amass_test_split, + 'val': amass_vald_split, + 'train': amass_train_split +} +# assert len(amass_splits['train'] + amass_splits['test'] + amass_splits['vald']) == len(all_sequences) == 15 + +class CustomDataset(Dataset): + def __init__(self, args, loader_type, augmentation=None, kwargs=None, build_cache=True): + self.args = args + self.loader_type = loader_type + + self.rank = 0 + self.ori_stride = self.args.stride + self.ori_length = self.args.pose_length + self.alignment = [0,0] # for trinity + + self.ori_joint_list = joints_list[self.args.ori_joints] + self.tar_joint_list = joints_list[self.args.tar_joints] + if 'smplx' in self.args.pose_rep: + self.joint_mask = np.zeros(len(list(self.ori_joint_list.keys()))*3) + self.joints = len(list(self.tar_joint_list.keys())) + for joint_name in self.tar_joint_list: + self.joint_mask[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + else: + self.joints = len(list(self.ori_joint_list.keys()))+1 + self.joint_mask = np.zeros(self.joints*3) + for joint_name in self.tar_joint_list: + if joint_name == "Hips": + self.joint_mask[3:6] = 1 + else: + self.joint_mask[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + # select trainable joints + + split_rule = pd.read_csv(args.data_path+"train_test_split.csv") + self.selected_file = split_rule.loc[(split_rule['type'] == loader_type) & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + if args.additional_data and loader_type == 'train': + split_b = split_rule.loc[(split_rule['type'] == 'additional') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + #self.selected_file = split_rule.loc[(split_rule['type'] == 'additional') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + self.selected_file = pd.concat([self.selected_file, split_b]) + if self.selected_file.empty: + logger.warning(f"{loader_type} is empty for speaker {self.args.training_speakers}, use train set 0-8 instead") + self.selected_file = split_rule.loc[(split_rule['type'] == 'train') & (split_rule['id'].str.split("_").str[0].astype(int).isin(self.args.training_speakers))] + self.selected_file = self.selected_file.iloc[0:8] + self.data_dir = args.data_path + self.use_amass = args.use_amass + self.beatx_during_time = 0 + self.amass_during_time = 0 + + if loader_type == "test": + self.args.multi_length_training = [1.0] + self.max_length = int(args.pose_length * self.args.multi_length_training[-1]) + self.max_audio_pre_len = math.floor(args.pose_length / args.pose_fps * self.args.audio_sr) + if self.max_audio_pre_len > self.args.test_length*self.args.audio_sr: + self.max_audio_pre_len = self.args.test_length*self.args.audio_sr + preloaded_dir = self.args.root_path + self.args.cache_path + loader_type + f"/{args.pose_rep}_cache" + + if self.args.beat_align: + if not os.path.exists(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy"): + self.calculate_mean_velocity(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy") + self.avg_vel = np.load(args.data_path+f"weights/mean_vel_{args.pose_rep}.npy") + + if build_cache and self.rank == 0: + self.build_cache(preloaded_dir) + self.lmdb_env = lmdb.open(preloaded_dir, readonly=True, lock=False) + with self.lmdb_env.begin() as txn: + self.n_samples = txn.stat()["entries"] + + self.norm = True + self.mean = np.load('./mean_std/beatx_2_330_mean.npy') + self.std = np.load('./mean_std/beatx_2_330_std.npy') + + self.trans_mean = np.load('./mean_std/beatx_2_trans_mean.npy') + self.trans_std = np.load('./mean_std/beatx_2_trans_std.npy') + + def load_amass(self,data): + ## 这个是用来 + # 修改amass数据里面的朝向，原本在blender里面是Z轴向上，目标是Y轴向上，当时面向目前没改 + + data_dict = {key: data[key] for key in data} + frames = data_dict['poses'].shape[0] + b = data_dict['poses'][...,:3] + b = rc.axis_angle_to_matrix(torch.from_numpy(b)) + rot_matrix = np.array([[1.0, 0.0, 0.0], [0.0 , 0.0, 1.0], [0.0, -1.0, 0.0]]) + c = np.einsum('ij,kjl->kil',rot_matrix,b) + c = rc.matrix_to_axis_angle(torch.from_numpy(c)) + data_dict['poses'][...,:3] = c + + trans_matrix1 = np.array([[1.0, 0.0, 0.0], [0.0 , 0.0, -1.0], [0.0, 1.0, 0.0]]) + data_dict['trans'] = np.einsum("bi,ij->bj",data_dict['trans'],trans_matrix1) + + betas300 = np.zeros(300) + betas300[:16] = data_dict['betas'] + data_dict['betas'] = betas300 + data_dict["expressions"] = np.zeros((frames,100)) + + return data_dict + + + def calculate_mean_velocity(self, save_path): + self.smplx = smplx.create( + self.args.data_path_1+"smplx_models/", + model_type='smplx', + gender='NEUTRAL_2020', + use_face_contour=False, + num_betas=300, + num_expression_coeffs=100, + ext='npz', + use_pca=False, + ).cuda().eval() + dir_p = self.data_dir + self.args.pose_rep + "/" + all_list = [] + from tqdm import tqdm + for tar in tqdm(os.listdir(dir_p)): + if tar.endswith(".npz"): + m_data = np.load(dir_p+tar, allow_pickle=True) + betas, poses, trans, exps = m_data["betas"], m_data["poses"], m_data["trans"], m_data["expressions"] + n, c = poses.shape[0], poses.shape[1] + betas = betas.reshape(1, 300) + betas = np.tile(betas, (n, 1)) + betas = torch.from_numpy(betas).cuda().float() + poses = torch.from_numpy(poses.reshape(n, c)).cuda().float() + exps = torch.from_numpy(exps.reshape(n, 100)).cuda().float() + trans = torch.from_numpy(trans.reshape(n, 3)).cuda().float() + max_length = 128 + s, r = n//max_length, n%max_length + #print(n, s, r) + all_tensor = [] + for i in range(s): + with torch.no_grad(): + joints = self.smplx( + betas=betas[i*max_length:(i+1)*max_length], + transl=trans[i*max_length:(i+1)*max_length], + expression=exps[i*max_length:(i+1)*max_length], + jaw_pose=poses[i*max_length:(i+1)*max_length, 66:69], + global_orient=poses[i*max_length:(i+1)*max_length,:3], + body_pose=poses[i*max_length:(i+1)*max_length,3:21*3+3], + left_hand_pose=poses[i*max_length:(i+1)*max_length,25*3:40*3], + right_hand_pose=poses[i*max_length:(i+1)*max_length,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[i*max_length:(i+1)*max_length, 69:72], + reye_pose=poses[i*max_length:(i+1)*max_length, 72:75], + )['joints'][:, :55, :].reshape(max_length, 55*3) + all_tensor.append(joints) + if r != 0: + with torch.no_grad(): + joints = self.smplx( + betas=betas[s*max_length:s*max_length+r], + transl=trans[s*max_length:s*max_length+r], + expression=exps[s*max_length:s*max_length+r], + jaw_pose=poses[s*max_length:s*max_length+r, 66:69], + global_orient=poses[s*max_length:s*max_length+r,:3], + body_pose=poses[s*max_length:s*max_length+r,3:21*3+3], + left_hand_pose=poses[s*max_length:s*max_length+r,25*3:40*3], + right_hand_pose=poses[s*max_length:s*max_length+r,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=poses[s*max_length:s*max_length+r, 69:72], + reye_pose=poses[s*max_length:s*max_length+r, 72:75], + )['joints'][:, :55, :].reshape(r, 55*3) + all_tensor.append(joints) + joints = torch.cat(all_tensor, axis=0) + joints = joints.permute(1, 0) + dt = 1/30 + # first steps is forward diff (t+1 - t) / dt + init_vel = (joints[:, 1:2] - joints[:, :1]) / dt + # middle steps are second order (t+1 - t-1) / 2dt + middle_vel = (joints[:, 2:] - joints[:, 0:-2]) / (2 * dt) + # last step is backward diff (t - t-1) / dt + final_vel = (joints[:, -1:] - joints[:, -2:-1]) / dt + #print(joints.shape, init_vel.shape, middle_vel.shape, final_vel.shape) + vel_seq = torch.cat([init_vel, middle_vel, final_vel], dim=1).permute(1, 0).reshape(n, 55, 3) + #print(vel_seq.shape) + #.permute(1, 0).reshape(n, 55, 3) + vel_seq_np = vel_seq.cpu().numpy() + vel_joints_np = np.linalg.norm(vel_seq_np, axis=2) # n * 55 + all_list.append(vel_joints_np) + avg_vel = np.mean(np.concatenate(all_list, axis=0),axis=0) # 55 + np.save(save_path, avg_vel) + + + def build_cache(self, preloaded_dir): + logger.info(f"Audio bit rate: {self.args.audio_fps}") + logger.info("Reading data '{}'...".format(self.data_dir)) + logger.info("Creating the dataset cache...") + if self.args.new_cache: + if os.path.exists(preloaded_dir): + shutil.rmtree(preloaded_dir) + if os.path.exists(preloaded_dir): + logger.info("Found the cache {}".format(preloaded_dir)) + elif self.loader_type == "test": + self.cache_generation( + preloaded_dir, True, + 0, 0, + is_test=True) + else: + self.cache_generation( + preloaded_dir, self.args.disable_filtering, + self.args.clean_first_seconds, self.args.clean_final_seconds, + is_test=False) + logger.info(f"BEATX during time is {self.beatx_during_time}s !") + logger.info(f"AMASS during time is {self.amass_during_time}s !") + + ## 对于BEATX train ,val ,test: 69800s ,7695s, 18672s ,总计 26.7h + ## + + def __len__(self): + return self.n_samples + + + def cache_generation(self, out_lmdb_dir, disable_filtering, clean_first_seconds, clean_final_seconds, is_test=False): + # if "wav2vec2" in self.args.audio_rep: + # self.wav2vec_model = Wav2Vec2Model.from_pretrained(f"{self.args.data_path_1}/hub/transformer/wav2vec2-base-960h") + # self.wav2vec_model.feature_extractor._freeze_parameters() + # self.wav2vec_model = self.wav2vec_model.cuda() + # self.wav2vec_model.eval() + + self.n_out_samples = 0 + # create db for samples + if not os.path.exists(out_lmdb_dir): os.makedirs(out_lmdb_dir) + dst_lmdb_env = lmdb.open(out_lmdb_dir, map_size= int(1024 ** 3 * 50))# 50G + n_filtered_out = defaultdict(int) + + for index, file_name in self.selected_file.iterrows(): + f_name = file_name["id"] + ext = ".npz" if "smplx" in self.args.pose_rep else ".bvh" + pose_file = self.data_dir + self.args.pose_rep + "/" + f_name + ext + pose_each_file = [] + trans_each_file = [] + trans_v_each_file = [] + shape_each_file = [] + audio_each_file = [] + facial_each_file = [] + word_each_file = [] + emo_each_file = [] + sem_each_file = [] + vid_each_file = [] + id_pose = f_name #1_wayne_0_1_1 + + logger.info(colored(f"# ---- Building cache for Pose {id_pose} ---- #", "blue")) + if "smplx" in self.args.pose_rep: + pose_data = np.load(pose_file, allow_pickle=True) + assert 30%self.args.pose_fps == 0, 'pose_fps should be an aliquot part of 30' + stride = int(30/self.args.pose_fps) + pose_each_file = pose_data["poses"][::stride] * self.joint_mask + pose_each_file = pose_each_file[:, self.joint_mask.astype(bool)] + # print(pose_each_file.shape) + self.beatx_during_time += pose_each_file.shape[0]/30 + trans_each_file = pose_data["trans"][::stride] + trans_each_file[:,0] = trans_each_file[:,0] - trans_each_file[0,0] + trans_each_file[:,2] = trans_each_file[:,2] - trans_each_file[0,2] + trans_v_each_file = np.zeros_like(trans_each_file) + trans_v_each_file[1:,0] = trans_each_file[1:,0] - trans_each_file[:-1,0] + trans_v_each_file[0,0] = trans_v_each_file[1,0] + trans_v_each_file[1:,2] = trans_each_file[1:,2] - trans_each_file[:-1,2] + trans_v_each_file[0,2] = trans_v_each_file[1,2] + trans_v_each_file[:,1] = trans_each_file[:,1] + + + shape_each_file = np.repeat(pose_data["betas"].reshape(1, 300), pose_each_file.shape[0], axis=0) + if self.args.facial_rep is not None: + logger.info(f"# ---- Building cache for Facial {id_pose} and Pose {id_pose} ---- #") + facial_each_file = pose_data["expressions"][::stride] + if self.args.facial_norm: + facial_each_file = (facial_each_file - self.mean_facial) / self.std_facial + + if self.args.id_rep is not None: + vid_each_file = np.repeat(np.array(int(f_name.split("_")[0])-1).reshape(1, 1), pose_each_file.shape[0], axis=0) + + filtered_result = self._sample_from_clip( + dst_lmdb_env, + pose_each_file, trans_each_file,trans_v_each_file, shape_each_file, + vid_each_file, + disable_filtering, clean_first_seconds, clean_final_seconds, is_test, + ) + for type in filtered_result.keys(): + n_filtered_out[type] += filtered_result[type] + + if self.args.use_amass: + amass_dir = '/mnt/fu09a/chenbohong/PantoMatrix/scripts/EMAGE_2024/datasets/AMASS_SMPLX' + for dataset in amass_splits[self.loader_type]: + search_path = os.path.join(amass_dir,dataset, '**', '*.npz') + npz_files = glob.glob(search_path, recursive=True) + for index, file_name in enumerate(npz_files): + f_name = file_name.split('/')[-1] + ext = ".npz" if "smplx" in self.args.pose_rep else ".bvh" + pose_file = file_name + pose_each_file = [] + trans_each_file = [] + trans_v_each_file = [] + shape_each_file = [] + audio_each_file = [] + facial_each_file = [] + word_each_file = [] + emo_each_file = [] + sem_each_file = [] + vid_each_file = [] + id_pose = f_name #1_wayne_0_1_1 + + logger.info(colored(f"# ---- Building cache for Pose {id_pose} ---- #", "blue")) + if "smplx" in self.args.pose_rep: + pose_data = np.load(pose_file, allow_pickle=True) + if len(pose_data.files)==6: + logger.info(colored(f"# ---- state file ---- #", "red")) + continue + assert 30%self.args.pose_fps == 0, 'pose_fps should be an aliquot part of 30' + pose_each_file = self.load_amass(pose_data) + fps = pose_data['mocap_frame_rate'] + stride =round(fps/30) + pose_each_file = pose_data["poses"][::stride] * self.joint_mask + pose_each_file = pose_each_file[:, self.joint_mask.astype(bool)] + trans_each_file = pose_data["trans"][::stride] + + + trans_each_file[:,0] = trans_each_file[:,0] - trans_each_file[0,0] + trans_each_file[:,2] = trans_each_file[:,2] - trans_each_file[0,2] + trans_v_each_file = np.zeros_like(trans_each_file) + trans_v_each_file[1:,0] = trans_each_file[1:,0] - trans_each_file[:-1,0] + trans_v_each_file[0,0] = trans_v_each_file[1,0] + trans_v_each_file[1:,2] = trans_each_file[1:,2] - trans_each_file[:-1,2] + trans_v_each_file[0,2] = trans_v_each_file[1,2] + trans_v_each_file[:,1] = trans_each_file[:,1] + + + + shape_each_file = np.repeat(pose_data["betas"].reshape(1, -1), pose_each_file.shape[0], axis=0) + + if self.args.id_rep is not None: + vid_each_file = np.repeat(np.array(int(100)-1).reshape(1, 1), pose_each_file.shape[0], axis=0) + + filtered_result = self._sample_from_clip( + dst_lmdb_env, + pose_each_file, trans_each_file,trans_v_each_file, shape_each_file, + vid_each_file, + disable_filtering, clean_first_seconds, clean_final_seconds, is_test, + ) + for type in filtered_result.keys(): + n_filtered_out[type] += filtered_result[type] + + + with dst_lmdb_env.begin() as txn: + logger.info(colored(f"no. of samples: {txn.stat()['entries']}", "cyan")) + n_total_filtered = 0 + for type, n_filtered in n_filtered_out.items(): + logger.info("{}: {}".format(type, n_filtered)) + n_total_filtered += n_filtered + logger.info(colored("no. of excluded samples: {} ({:.1f}%)".format( + n_total_filtered, 100 * n_total_filtered / (txn.stat()["entries"] + n_total_filtered)), "cyan")) + dst_lmdb_env.sync() + dst_lmdb_env.close() + + def _sample_from_clip( + self, dst_lmdb_env, pose_each_file, trans_each_file, trans_v_each_file,shape_each_file, + vid_each_file, + disable_filtering, clean_first_seconds, clean_final_seconds, is_test, + ): + """ + for data cleaning, we ignore the data for first and final n s + for test, we return all data + """ + # audio_start = int(self.alignment[0] * self.args.audio_fps) + # pose_start = int(self.alignment[1] * self.args.pose_fps) + #logger.info(f"before: {audio_each_file.shape} {pose_each_file.shape}") + # audio_each_file = audio_each_file[audio_start:] + # pose_each_file = pose_each_file[pose_start:] + # trans_each_file = + #logger.info(f"after alignment: {audio_each_file.shape} {pose_each_file.shape}") + #print(pose_each_file.shape) + round_seconds_skeleton = pose_each_file.shape[0] // self.args.pose_fps # assume 1500 frames / 15 fps = 100 s + #print(round_seconds_skeleton) + + clip_s_t, clip_e_t = clean_first_seconds, round_seconds_skeleton - clean_final_seconds # assume [10, 90]s + clip_s_f_audio, clip_e_f_audio = self.args.audio_fps * clip_s_t, clip_e_t * self.args.audio_fps # [160,000,90*160,000] + clip_s_f_pose, clip_e_f_pose = clip_s_t * self.args.pose_fps, clip_e_t * self.args.pose_fps # [150,90*15] + + + for ratio in self.args.multi_length_training: + if is_test:# stride = length for test + cut_length = clip_e_f_pose - clip_s_f_pose + self.args.stride = cut_length + self.max_length = cut_length + else: + self.args.stride = int(ratio*self.ori_stride) + cut_length = int(self.ori_length*ratio) + + num_subdivision = math.floor((clip_e_f_pose - clip_s_f_pose - cut_length) / self.args.stride) + 1 + logger.info(f"pose from frame {clip_s_f_pose} to {clip_e_f_pose}, length {cut_length}") + logger.info(f"{num_subdivision} clips is expected with stride {self.args.stride}") + + + n_filtered_out = defaultdict(int) + sample_pose_list = [] + sample_audio_list = [] + sample_shape_list = [] + sample_vid_list = [] + sample_trans_list = [] + sample_trans_v_list = [] + + for i in range(num_subdivision): # cut into around 2s chip, (self npose) + start_idx = clip_s_f_pose + i * self.args.stride + fin_idx = start_idx + cut_length + sample_pose = pose_each_file[start_idx:fin_idx] + sample_trans = trans_each_file[start_idx:fin_idx] + sample_trans_v = trans_v_each_file[start_idx:fin_idx] + sample_shape = shape_each_file[start_idx:fin_idx] + # print(sample_pose.shape) + + + sample_vid = vid_each_file[start_idx:fin_idx] if self.args.id_rep is not None else np.array([-1]) + + if sample_pose.any() != None: + # filtering motion skeleton data + sample_pose, filtering_message = MotionPreprocessor(sample_pose).get() + is_correct_motion = (sample_pose is not None) + if is_correct_motion or disable_filtering: + sample_pose_list.append(sample_pose) + + sample_shape_list.append(sample_shape) + + sample_vid_list.append(sample_vid) + + + sample_trans_list.append(sample_trans) + sample_trans_v_list.append(sample_trans_v) + else: + n_filtered_out[filtering_message] += 1 + + if len(sample_pose_list) > 0: + with dst_lmdb_env.begin(write=True) as txn: + for pose, shape, vid, trans,trans_v in zip( + sample_pose_list, + sample_shape_list, + sample_vid_list, + sample_trans_list, + sample_trans_v_list, + ): + k = "{:005}".format(self.n_out_samples).encode("ascii") + v = [pose , shape, vid, trans,trans_v] + v = pickle.dumps(v,5) + txn.put(k, v) + self.n_out_samples += 1 + return n_filtered_out + + def __getitem__(self, idx): + with self.lmdb_env.begin(write=False) as txn: + key = "{:005}".format(idx).encode("ascii") + sample = txn.get(key) + sample = pickle.loads(sample) + tar_pose, in_shape, vid, trans,trans_v = sample + tar_pose = torch.from_numpy(tar_pose).float() + tar_pose = rc.axis_angle_to_matrix(tar_pose.reshape(-1, 55, 3)) + tar_pose = rc.matrix_to_rotation_6d(tar_pose).reshape(-1, 55*6) + + if self.norm: + tar_pose = (tar_pose - self.mean) / self.std + trans_v = (trans_v-self.trans_mean)/self.trans_std + + if self.loader_type == "test": + tar_pose = tar_pose.float() + trans = torch.from_numpy(trans).float() + trans_v = torch.from_numpy(trans_v).float() + vid = torch.from_numpy(vid).float() + in_shape = torch.from_numpy(in_shape).float() + else: + in_shape = torch.from_numpy(in_shape).reshape((in_shape.shape[0], -1)).float() + trans = torch.from_numpy(trans).reshape((trans.shape[0], -1)).float() + trans_v = torch.from_numpy(trans_v).reshape((trans_v.shape[0], -1)).float() + vid = torch.from_numpy(vid).reshape((vid.shape[0], -1)).float() + tar_pose = tar_pose.reshape((tar_pose.shape[0], -1)).float() + tar_pose = torch.cat([tar_pose, trans_v], dim=1) + return tar_pose + +class MotionPreprocessor: + def __init__(self, skeletons): + self.skeletons = skeletons + #self.mean_pose = mean_pose + self.filtering_message = "PASS" + + def get(self): + assert (self.skeletons is not None) + + # filtering + if self.skeletons is not None: + if self.check_pose_diff(): + self.skeletons = [] + self.filtering_message = "pose" + # elif self.check_spine_angle(): + # self.skeletons = [] + # self.filtering_message = "spine angle" + # elif self.check_static_motion(): + # self.skeletons = [] + # self.filtering_message = "motion" + + # if self.skeletons != []: + # self.skeletons = self.skeletons.tolist() + # for i, frame in enumerate(self.skeletons): + # assert not np.isnan(self.skeletons[i]).any() # missing joints + + return self.skeletons, self.filtering_message + + def check_static_motion(self, verbose=True): + def get_variance(skeleton, joint_idx): + wrist_pos = skeleton[:, joint_idx] + variance = np.sum(np.var(wrist_pos, axis=0)) + return variance + + left_arm_var = get_variance(self.skeletons, 6) + right_arm_var = get_variance(self.skeletons, 9) + + th = 0.0014 # exclude 13110 + # th = 0.002 # exclude 16905 + if left_arm_var < th and right_arm_var < th: + if verbose: + print("skip - check_static_motion left var {}, right var {}".format(left_arm_var, right_arm_var)) + return True + else: + if verbose: + print("pass - check_static_motion left var {}, right var {}".format(left_arm_var, right_arm_var)) + return False + + + def check_pose_diff(self, verbose=False): +# diff = np.abs(self.skeletons - self.mean_pose) # 186*1 +# diff = np.mean(diff) + +# # th = 0.017 +# th = 0.02 #0.02 # exclude 3594 +# if diff < th: +# if verbose: +# print("skip - check_pose_diff {:.5f}".format(diff)) +# return True +# # th = 3.5 #0.02 # exclude 3594 +# # if 3.5 < diff < 5: +# # if verbose: +# # print("skip - check_pose_diff {:.5f}".format(diff)) +# # return True +# else: +# if verbose: +# print("pass - check_pose_diff {:.5f}".format(diff)) + return False + + + def check_spine_angle(self, verbose=True): + def angle_between(v1, v2): + v1_u = v1 / np.linalg.norm(v1) + v2_u = v2 / np.linalg.norm(v2) + return np.arccos(np.clip(np.dot(v1_u, v2_u), -1.0, 1.0)) + + angles = [] + for i in range(self.skeletons.shape[0]): + spine_vec = self.skeletons[i, 1] - self.skeletons[i, 0] + angle = angle_between(spine_vec, [0, -1, 0]) + angles.append(angle) + + if np.rad2deg(max(angles)) > 30 or np.rad2deg(np.mean(angles)) > 20: # exclude 4495 + # if np.rad2deg(max(angles)) > 20: # exclude 8270 + if verbose: + print("skip - check_spine_angle {:.5f}, {:.5f}".format(max(angles), np.mean(angles))) + return True + else: + if verbose: + print("pass - check_spine_angle {:.5f}".format(max(angles))) + return False \ No newline at end of file diff --git a/dataloaders/pymo/Quaternions.py b/dataloaders/pymo/Quaternions.py new file mode 100644 index 0000000000000000000000000000000000000000..d4b754871310a264e2bd2675479db9a79d24358e --- /dev/null +++ b/dataloaders/pymo/Quaternions.py @@ -0,0 +1,468 @@ +import numpy as np + +class Quaternions: + """ + Quaternions is a wrapper around a numpy ndarray + that allows it to act as if it were an narray of + a quaternion data type. + + Therefore addition, subtraction, multiplication, + division, negation, absolute, are all defined + in terms of quaternion operations such as quaternion + multiplication. + + This allows for much neater code and many routines + which conceptually do the same thing to be written + in the same way for point data and for rotation data. + + The Quaternions class has been desgined such that it + should support broadcasting and slicing in all of the + usual ways. + """ + + def __init__(self, qs): + if isinstance(qs, np.ndarray): + + if len(qs.shape) == 1: qs = np.array([qs]) + self.qs = qs + return + + if isinstance(qs, Quaternions): + self.qs = qs.qs + return + + raise TypeError('Quaternions must be constructed from iterable, numpy array, or Quaternions, not %s' % type(qs)) + + def __str__(self): return "Quaternions("+ str(self.qs) + ")" + def __repr__(self): return "Quaternions("+ repr(self.qs) + ")" + + """ Helper Methods for Broadcasting and Data extraction """ + + @classmethod + def _broadcast(cls, sqs, oqs, scalar=False): + + if isinstance(oqs, float): return sqs, oqs * np.ones(sqs.shape[:-1]) + + ss = np.array(sqs.shape) if not scalar else np.array(sqs.shape[:-1]) + os = np.array(oqs.shape) + + if len(ss) != len(os): + raise TypeError('Quaternions cannot broadcast together shapes %s and %s' % (sqs.shape, oqs.shape)) + + if np.all(ss == os): return sqs, oqs + + if not np.all((ss == os) | (os == np.ones(len(os))) | (ss == np.ones(len(ss)))): + raise TypeError('Quaternions cannot broadcast together shapes %s and %s' % (sqs.shape, oqs.shape)) + + sqsn, oqsn = sqs.copy(), oqs.copy() + + for a in np.where(ss == 1)[0]: sqsn = sqsn.repeat(os[a], axis=a) + for a in np.where(os == 1)[0]: oqsn = oqsn.repeat(ss[a], axis=a) + + return sqsn, oqsn + + """ Adding Quaterions is just Defined as Multiplication """ + + def __add__(self, other): return self * other + def __sub__(self, other): return self / other + + """ Quaterion Multiplication """ + + def __mul__(self, other): + """ + Quaternion multiplication has three main methods. + + When multiplying a Quaternions array by Quaternions + normal quaternion multiplication is performed. + + When multiplying a Quaternions array by a vector + array of the same shape, where the last axis is 3, + it is assumed to be a Quaternion by 3D-Vector + multiplication and the 3D-Vectors are rotated + in space by the Quaternions. + + When multipplying a Quaternions array by a scalar + or vector of different shape it is assumed to be + a Quaternions by Scalars multiplication and the + Quaternions are scaled using Slerp and the identity + quaternions. + """ + + """ If Quaternions type do Quaternions * Quaternions """ + if isinstance(other, Quaternions): + + sqs, oqs = Quaternions._broadcast(self.qs, other.qs) + + q0 = sqs[...,0]; q1 = sqs[...,1]; + q2 = sqs[...,2]; q3 = sqs[...,3]; + r0 = oqs[...,0]; r1 = oqs[...,1]; + r2 = oqs[...,2]; r3 = oqs[...,3]; + + qs = np.empty(sqs.shape) + qs[...,0] = r0 * q0 - r1 * q1 - r2 * q2 - r3 * q3 + qs[...,1] = r0 * q1 + r1 * q0 - r2 * q3 + r3 * q2 + qs[...,2] = r0 * q2 + r1 * q3 + r2 * q0 - r3 * q1 + qs[...,3] = r0 * q3 - r1 * q2 + r2 * q1 + r3 * q0 + + return Quaternions(qs) + + """ If array type do Quaternions * Vectors """ + if isinstance(other, np.ndarray) and other.shape[-1] == 3: + vs = Quaternions(np.concatenate([np.zeros(other.shape[:-1] + (1,)), other], axis=-1)) + return (self * (vs * -self)).imaginaries + + """ If float do Quaternions * Scalars """ + if isinstance(other, np.ndarray) or isinstance(other, float): + return Quaternions.slerp(Quaternions.id_like(self), self, other) + + raise TypeError('Cannot multiply/add Quaternions with type %s' % str(type(other))) + + def __div__(self, other): + """ + When a Quaternion type is supplied, division is defined + as multiplication by the inverse of that Quaternion. + + When a scalar or vector is supplied it is defined + as multiplicaion of one over the supplied value. + Essentially a scaling. + """ + + if isinstance(other, Quaternions): return self * (-other) + if isinstance(other, np.ndarray): return self * (1.0 / other) + if isinstance(other, float): return self * (1.0 / other) + raise TypeError('Cannot divide/subtract Quaternions with type %s' + str(type(other))) + + def __eq__(self, other): return self.qs == other.qs + def __ne__(self, other): return self.qs != other.qs + + def __neg__(self): + """ Invert Quaternions """ + return Quaternions(self.qs * np.array([[1, -1, -1, -1]])) + + def __abs__(self): + """ Unify Quaternions To Single Pole """ + qabs = self.normalized().copy() + top = np.sum(( qabs.qs) * np.array([1,0,0,0]), axis=-1) + bot = np.sum((-qabs.qs) * np.array([1,0,0,0]), axis=-1) + qabs.qs[top < bot] = -qabs.qs[top < bot] + return qabs + + def __iter__(self): return iter(self.qs) + def __len__(self): return len(self.qs) + + def __getitem__(self, k): return Quaternions(self.qs[k]) + def __setitem__(self, k, v): self.qs[k] = v.qs + + @property + def lengths(self): + return np.sum(self.qs**2.0, axis=-1)**0.5 + + @property + def reals(self): + return self.qs[...,0] + + @property + def imaginaries(self): + return self.qs[...,1:4] + + @property + def shape(self): return self.qs.shape[:-1] + + def repeat(self, n, **kwargs): + return Quaternions(self.qs.repeat(n, **kwargs)) + + def normalized(self): + return Quaternions(self.qs / self.lengths[...,np.newaxis]) + + def log(self): + norm = abs(self.normalized()) + imgs = norm.imaginaries + lens = np.sqrt(np.sum(imgs**2, axis=-1)) + lens = np.arctan2(lens, norm.reals) / (lens + 1e-10) + return imgs * lens[...,np.newaxis] + + def constrained(self, axis): + + rl = self.reals + im = np.sum(axis * self.imaginaries, axis=-1) + + t1 = -2 * np.arctan2(rl, im) + np.pi + t2 = -2 * np.arctan2(rl, im) - np.pi + + top = Quaternions.exp(axis[np.newaxis] * (t1[:,np.newaxis] / 2.0)) + bot = Quaternions.exp(axis[np.newaxis] * (t2[:,np.newaxis] / 2.0)) + img = self.dot(top) > self.dot(bot) + + ret = top.copy() + ret[ img] = top[ img] + ret[~img] = bot[~img] + return ret + + def constrained_x(self): return self.constrained(np.array([1,0,0])) + def constrained_y(self): return self.constrained(np.array([0,1,0])) + def constrained_z(self): return self.constrained(np.array([0,0,1])) + + def dot(self, q): return np.sum(self.qs * q.qs, axis=-1) + + def copy(self): return Quaternions(np.copy(self.qs)) + + def reshape(self, s): + self.qs.reshape(s) + return self + + def interpolate(self, ws): + return Quaternions.exp(np.average(abs(self).log, axis=0, weights=ws)) + + def euler(self, order='xyz'): + + q = self.normalized().qs + q0 = q[...,0] + q1 = q[...,1] + q2 = q[...,2] + q3 = q[...,3] + es = np.zeros(self.shape + (3,)) + + if order == 'xyz': + es[...,0] = np.arctan2(2 * (q0 * q1 + q2 * q3), 1 - 2 * (q1 * q1 + q2 * q2)) + es[...,1] = np.arcsin((2 * (q0 * q2 - q3 * q1)).clip(-1,1)) + es[...,2] = np.arctan2(2 * (q0 * q3 + q1 * q2), 1 - 2 * (q2 * q2 + q3 * q3)) + elif order == 'yzx': + es[...,0] = np.arctan2(2 * (q1 * q0 - q2 * q3), -q1 * q1 + q2 * q2 - q3 * q3 + q0 * q0) + es[...,1] = np.arctan2(2 * (q2 * q0 - q1 * q3), q1 * q1 - q2 * q2 - q3 * q3 + q0 * q0) + es[...,2] = np.arcsin((2 * (q1 * q2 + q3 * q0)).clip(-1,1)) + else: + raise NotImplementedError('Cannot convert from ordering %s' % order) + + """ + + # These conversion don't appear to work correctly for Maya. + # http://bediyap.com/programming/convert-quaternion-to-euler-rotations/ + + if order == 'xyz': + es[...,0] = np.arctan2(2 * (q0 * q3 - q1 * q2), q0 * q0 + q1 * q1 - q2 * q2 - q3 * q3) + es[...,1] = np.arcsin((2 * (q1 * q3 + q0 * q2)).clip(-1,1)) + es[...,2] = np.arctan2(2 * (q0 * q1 - q2 * q3), q0 * q0 - q1 * q1 - q2 * q2 + q3 * q3) + elif order == 'yzx': + es[...,0] = np.arctan2(2 * (q0 * q1 - q2 * q3), q0 * q0 - q1 * q1 + q2 * q2 - q3 * q3) + es[...,1] = np.arcsin((2 * (q1 * q2 + q0 * q3)).clip(-1,1)) + es[...,2] = np.arctan2(2 * (q0 * q2 - q1 * q3), q0 * q0 + q1 * q1 - q2 * q2 - q3 * q3) + elif order == 'zxy': + es[...,0] = np.arctan2(2 * (q0 * q2 - q1 * q3), q0 * q0 - q1 * q1 - q2 * q2 + q3 * q3) + es[...,1] = np.arcsin((2 * (q0 * q1 + q2 * q3)).clip(-1,1)) + es[...,2] = np.arctan2(2 * (q0 * q3 - q1 * q2), q0 * q0 - q1 * q1 + q2 * q2 - q3 * q3) + elif order == 'xzy': + es[...,0] = np.arctan2(2 * (q0 * q2 + q1 * q3), q0 * q0 + q1 * q1 - q2 * q2 - q3 * q3) + es[...,1] = np.arcsin((2 * (q0 * q3 - q1 * q2)).clip(-1,1)) + es[...,2] = np.arctan2(2 * (q0 * q1 + q2 * q3), q0 * q0 - q1 * q1 + q2 * q2 - q3 * q3) + elif order == 'yxz': + es[...,0] = np.arctan2(2 * (q1 * q2 + q0 * q3), q0 * q0 - q1 * q1 + q2 * q2 - q3 * q3) + es[...,1] = np.arcsin((2 * (q0 * q1 - q2 * q3)).clip(-1,1)) + es[...,2] = np.arctan2(2 * (q1 * q3 + q0 * q2), q0 * q0 - q1 * q1 - q2 * q2 + q3 * q3) + elif order == 'zyx': + es[...,0] = np.arctan2(2 * (q0 * q1 + q2 * q3), q0 * q0 - q1 * q1 - q2 * q2 + q3 * q3) + es[...,1] = np.arcsin((2 * (q0 * q2 - q1 * q3)).clip(-1,1)) + es[...,2] = np.arctan2(2 * (q0 * q3 + q1 * q2), q0 * q0 + q1 * q1 - q2 * q2 - q3 * q3) + else: + raise KeyError('Unknown ordering %s' % order) + + """ + + # https://github.com/ehsan/ogre/blob/master/OgreMain/src/OgreMatrix3.cpp + # Use this class and convert from matrix + + return es + + + def average(self): + + if len(self.shape) == 1: + + import numpy.core.umath_tests as ut + system = ut.matrix_multiply(self.qs[:,:,np.newaxis], self.qs[:,np.newaxis,:]).sum(axis=0) + w, v = np.linalg.eigh(system) + qiT_dot_qref = (self.qs[:,:,np.newaxis] * v[np.newaxis,:,:]).sum(axis=1) + return Quaternions(v[:,np.argmin((1.-qiT_dot_qref**2).sum(axis=0))]) + + else: + + raise NotImplementedError('Cannot average multi-dimensionsal Quaternions') + + def angle_axis(self): + + norm = self.normalized() + s = np.sqrt(1 - (norm.reals**2.0)) + s[s == 0] = 0.001 + + angles = 2.0 * np.arccos(norm.reals) + axis = norm.imaginaries / s[...,np.newaxis] + + return angles, axis + + + def transforms(self): + + qw = self.qs[...,0] + qx = self.qs[...,1] + qy = self.qs[...,2] + qz = self.qs[...,3] + + x2 = qx + qx; y2 = qy + qy; z2 = qz + qz; + xx = qx * x2; yy = qy * y2; wx = qw * x2; + xy = qx * y2; yz = qy * z2; wy = qw * y2; + xz = qx * z2; zz = qz * z2; wz = qw * z2; + + m = np.empty(self.shape + (3,3)) + m[...,0,0] = 1.0 - (yy + zz) + m[...,0,1] = xy - wz + m[...,0,2] = xz + wy + m[...,1,0] = xy + wz + m[...,1,1] = 1.0 - (xx + zz) + m[...,1,2] = yz - wx + m[...,2,0] = xz - wy + m[...,2,1] = yz + wx + m[...,2,2] = 1.0 - (xx + yy) + + return m + + def ravel(self): + return self.qs.ravel() + + @classmethod + def id(cls, n): + + if isinstance(n, tuple): + qs = np.zeros(n + (4,)) + qs[...,0] = 1.0 + return Quaternions(qs) + + if isinstance(n, int) or isinstance(n, long): + qs = np.zeros((n,4)) + qs[:,0] = 1.0 + return Quaternions(qs) + + raise TypeError('Cannot Construct Quaternion from %s type' % str(type(n))) + + @classmethod + def id_like(cls, a): + qs = np.zeros(a.shape + (4,)) + qs[...,0] = 1.0 + return Quaternions(qs) + + @classmethod + def exp(cls, ws): + + ts = np.sum(ws**2.0, axis=-1)**0.5 + ts[ts == 0] = 0.001 + ls = np.sin(ts) / ts + + qs = np.empty(ws.shape[:-1] + (4,)) + qs[...,0] = np.cos(ts) + qs[...,1] = ws[...,0] * ls + qs[...,2] = ws[...,1] * ls + qs[...,3] = ws[...,2] * ls + + return Quaternions(qs).normalized() + + @classmethod + def slerp(cls, q0s, q1s, a): + + fst, snd = cls._broadcast(q0s.qs, q1s.qs) + fst, a = cls._broadcast(fst, a, scalar=True) + snd, a = cls._broadcast(snd, a, scalar=True) + + len = np.sum(fst * snd, axis=-1) + + neg = len < 0.0 + len[neg] = -len[neg] + snd[neg] = -snd[neg] + + amount0 = np.zeros(a.shape) + amount1 = np.zeros(a.shape) + + linear = (1.0 - len) < 0.01 + omegas = np.arccos(len[~linear]) + sinoms = np.sin(omegas) + + amount0[ linear] = 1.0 - a[linear] + amount1[ linear] = a[linear] + amount0[~linear] = np.sin((1.0 - a[~linear]) * omegas) / sinoms + amount1[~linear] = np.sin( a[~linear] * omegas) / sinoms + + return Quaternions( + amount0[...,np.newaxis] * fst + + amount1[...,np.newaxis] * snd) + + @classmethod + def between(cls, v0s, v1s): + a = np.cross(v0s, v1s) + w = np.sqrt((v0s**2).sum(axis=-1) * (v1s**2).sum(axis=-1)) + (v0s * v1s).sum(axis=-1) + return Quaternions(np.concatenate([w[...,np.newaxis], a], axis=-1)).normalized() + + @classmethod + def from_angle_axis(cls, angles, axis): + axis = axis / (np.sqrt(np.sum(axis**2, axis=-1)) + 1e-10)[...,np.newaxis] + sines = np.sin(angles / 2.0)[...,np.newaxis] + cosines = np.cos(angles / 2.0)[...,np.newaxis] + return Quaternions(np.concatenate([cosines, axis * sines], axis=-1)) + + @classmethod + def from_euler(cls, es, order='xyz', world=False): + + axis = { + 'x' : np.array([1,0,0]), + 'y' : np.array([0,1,0]), + 'z' : np.array([0,0,1]), + } + + q0s = Quaternions.from_angle_axis(es[...,0], axis[order[0]]) + q1s = Quaternions.from_angle_axis(es[...,1], axis[order[1]]) + q2s = Quaternions.from_angle_axis(es[...,2], axis[order[2]]) + + return (q2s * (q1s * q0s)) if world else (q0s * (q1s * q2s)) + + @classmethod + def from_transforms(cls, ts): + + d0, d1, d2 = ts[...,0,0], ts[...,1,1], ts[...,2,2] + + q0 = ( d0 + d1 + d2 + 1.0) / 4.0 + q1 = ( d0 - d1 - d2 + 1.0) / 4.0 + q2 = (-d0 + d1 - d2 + 1.0) / 4.0 + q3 = (-d0 - d1 + d2 + 1.0) / 4.0 + + q0 = np.sqrt(q0.clip(0,None)) + q1 = np.sqrt(q1.clip(0,None)) + q2 = np.sqrt(q2.clip(0,None)) + q3 = np.sqrt(q3.clip(0,None)) + + c0 = (q0 >= q1) & (q0 >= q2) & (q0 >= q3) + c1 = (q1 >= q0) & (q1 >= q2) & (q1 >= q3) + c2 = (q2 >= q0) & (q2 >= q1) & (q2 >= q3) + c3 = (q3 >= q0) & (q3 >= q1) & (q3 >= q2) + + q1[c0] *= np.sign(ts[c0,2,1] - ts[c0,1,2]) + q2[c0] *= np.sign(ts[c0,0,2] - ts[c0,2,0]) + q3[c0] *= np.sign(ts[c0,1,0] - ts[c0,0,1]) + + q0[c1] *= np.sign(ts[c1,2,1] - ts[c1,1,2]) + q2[c1] *= np.sign(ts[c1,1,0] + ts[c1,0,1]) + q3[c1] *= np.sign(ts[c1,0,2] + ts[c1,2,0]) + + q0[c2] *= np.sign(ts[c2,0,2] - ts[c2,2,0]) + q1[c2] *= np.sign(ts[c2,1,0] + ts[c2,0,1]) + q3[c2] *= np.sign(ts[c2,2,1] + ts[c2,1,2]) + + q0[c3] *= np.sign(ts[c3,1,0] - ts[c3,0,1]) + q1[c3] *= np.sign(ts[c3,2,0] + ts[c3,0,2]) + q2[c3] *= np.sign(ts[c3,2,1] + ts[c3,1,2]) + + qs = np.empty(ts.shape[:-2] + (4,)) + qs[...,0] = q0 + qs[...,1] = q1 + qs[...,2] = q2 + qs[...,3] = q3 + + return cls(qs) + + + \ No newline at end of file diff --git a/dataloaders/pymo/__init__.py b/dataloaders/pymo/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/dataloaders/pymo/data.py b/dataloaders/pymo/data.py new file mode 100644 index 0000000000000000000000000000000000000000..7be4f0a819aa041218b8a3d78e700017253d277c --- /dev/null +++ b/dataloaders/pymo/data.py @@ -0,0 +1,53 @@ +import numpy as np + +class Joint(): + def __init__(self, name, parent=None, children=None): + self.name = name + self.parent = parent + self.children = children + +class MocapData(): + def __init__(self): + self.skeleton = {} + self.values = None + self.channel_names = [] + self.framerate = 0.0 + self.root_name = '' + + def traverse(self, j=None): + stack = [self.root_name] + while stack: + joint = stack.pop() + yield joint + for c in self.skeleton[joint]['children']: + stack.append(c) + + def clone(self): + import copy + new_data = MocapData() + new_data.skeleton = copy.copy(self.skeleton) + new_data.values = copy.copy(self.values) + new_data.channel_names = copy.copy(self.channel_names) + new_data.root_name = copy.copy(self.root_name) + new_data.framerate = copy.copy(self.framerate) + return new_data + + def get_all_channels(self): + '''Returns all of the channels parsed from the file as a 2D numpy array''' + + frames = [f[1] for f in self.values] + return np.asarray([[channel[2] for channel in frame] for frame in frames]) + + def get_skeleton_tree(self): + tree = [] + root_key = [j for j in self.skeleton if self.skeleton[j]['parent']==None][0] + + root_joint = Joint(root_key) + + def get_empty_channels(self): + #TODO + pass + + def get_constant_channels(self): + #TODO + pass diff --git a/dataloaders/pymo/features.py b/dataloaders/pymo/features.py new file mode 100644 index 0000000000000000000000000000000000000000..fec29ed5758f79b61f296e01e9b077cba573f495 --- /dev/null +++ b/dataloaders/pymo/features.py @@ -0,0 +1,43 @@ +''' +A set of mocap feature extraction functions + +Created by Omid Alemi | Nov 17 2017 + +''' +import numpy as np +import pandas as pd +import peakutils +import matplotlib.pyplot as plt + +def get_foot_contact_idxs(signal, t=0.02, min_dist=120): + up_idxs = peakutils.indexes(signal, thres=t/max(signal), min_dist=min_dist) + down_idxs = peakutils.indexes(-signal, thres=t/min(signal), min_dist=min_dist) + + return [up_idxs, down_idxs] + + +def create_foot_contact_signal(mocap_track, col_name, start=1, t=0.02, min_dist=120): + signal = mocap_track.values[col_name].values + idxs = get_foot_contact_idxs(signal, t, min_dist) + + step_signal = [] + + c = start + for f in range(len(signal)): + if f in idxs[1]: + c = 0 + elif f in idxs[0]: + c = 1 + + step_signal.append(c) + + return step_signal + +def plot_foot_up_down(mocap_track, col_name, t=0.02, min_dist=120): + + signal = mocap_track.values[col_name].values + idxs = get_foot_contact_idxs(signal, t, min_dist) + + plt.plot(mocap_track.values.index, signal) + plt.plot(mocap_track.values.index[idxs[0]], signal[idxs[0]], 'ro') + plt.plot(mocap_track.values.index[idxs[1]], signal[idxs[1]], 'go') diff --git a/dataloaders/pymo/mocapplayer/data-template.js b/dataloaders/pymo/mocapplayer/data-template.js new file mode 100644 index 0000000000000000000000000000000000000000..68a51392fb7d2458487eae2a00a3ed03c1e7153a --- /dev/null +++ b/dataloaders/pymo/mocapplayer/data-template.js @@ -0,0 +1,3 @@ +var dataBuffer = `$$DATA$$`; + +start(dataBuffer); \ No newline at end of file diff --git a/dataloaders/pymo/mocapplayer/js/skeletonFactory.js b/dataloaders/pymo/mocapplayer/js/skeletonFactory.js new file mode 100644 index 0000000000000000000000000000000000000000..e1d072b7df2fb40772e93f2dee595e467744e36b --- /dev/null +++ b/dataloaders/pymo/mocapplayer/js/skeletonFactory.js @@ -0,0 +1,233 @@ +bm_v = new THREE.MeshPhongMaterial({ + color: 0x08519c, + emissive: 0x08306b, + specular: 0x08519c, + shininess: 10, + side: THREE.DoubleSide +}); + +jm_v = new THREE.MeshPhongMaterial({ + color: 0x08306b, + emissive: 0x000000, + specular: 0x111111, + shininess: 90, + side: THREE.DoubleSide +}); + +bm_a = new THREE.MeshPhongMaterial({ + color: 0x980043, + emissive: 0x67001f, + specular: 0x6a51a3, + shininess: 10, + side: THREE.DoubleSide +}); + +jm_a = new THREE.MeshPhongMaterial({ + color: 0x67001f, + emissive: 0x000000, + specular: 0x111111, + shininess: 90, + side: THREE.DoubleSide +}); + +bm_b = new THREE.MeshPhongMaterial({ + color: 0x3f007d, + emissive: 0x3f007d, + specular: 0x807dba, + shininess: 2, + side: THREE.DoubleSide +}); + +jm_b = new THREE.MeshPhongMaterial({ + color: 0x3f007d, + emissive: 0x000000, + specular: 0x807dba, + shininess: 90, + side: THREE.DoubleSide +}); + +//------------------ + + +jointmaterial = new THREE.MeshLambertMaterial({ + color: 0xc57206, + emissive: 0x271c18, + side: THREE.DoubleSide, + // shading: THREE.FlatShading, + wireframe: false, + shininess: 90, +}); + +bonematerial = new THREE.MeshPhongMaterial({ + color: 0xbd9a6d, + emissive: 0x271c18, + side: THREE.DoubleSide, + // shading: THREE.FlatShading, + wireframe: false +}); + +jointmaterial2 = new THREE.MeshPhongMaterial({ + color: 0x1562a2, + emissive: 0x000000, + specular: 0x111111, + shininess: 30, + side: THREE.DoubleSide +}); + +bonematerial2 = new THREE.MeshPhongMaterial({ + color: 0x552211, + emissive: 0x882211, + // emissive: 0x000000, + specular: 0x111111, + shininess: 30, + side: THREE.DoubleSide +}); + +bonematerial3 = new THREE.MeshPhongMaterial({ + color: 0x176793, + emissive: 0x000000, + specular: 0x111111, + shininess: 90, + side: THREE.DoubleSide +}); + + + +jointmaterial4 = new THREE.MeshPhongMaterial({ + color: 0xFF8A00, + emissive: 0x000000, + specular: 0x111111, + shininess: 90, + side: THREE.DoubleSide +}); + + +bonematerial4 = new THREE.MeshPhongMaterial({ + color: 0x53633D, + emissive: 0x000000, + specular: 0xFFC450, + shininess: 90, + side: THREE.DoubleSide +}); + + + +bonematerial44 = new THREE.MeshPhongMaterial({ + color: 0x582A72, + emissive: 0x000000, + specular: 0xFFC450, + shininess: 90, + side: THREE.DoubleSide +}); + +jointmaterial5 = new THREE.MeshPhongMaterial({ + color: 0xAA5533, + emissive: 0x000000, + specular: 0x111111, + shininess: 30, + side: THREE.DoubleSide +}); + +bonematerial5 = new THREE.MeshPhongMaterial({ + color: 0x552211, + emissive: 0x772211, + specular: 0x111111, + shininess: 30, + side: THREE.DoubleSide +}); + + +markermaterial = new THREE.MeshPhongMaterial({ + color: 0xc57206, + emissive: 0x271c18, + side: THREE.DoubleSide, + // shading: THREE.FlatShading, + wireframe: false, + shininess: 20, +}); + +markermaterial2 = new THREE.MeshPhongMaterial({ + color: 0x1562a2, + emissive: 0x271c18, + side: THREE.DoubleSide, + // shading: THREE.FlatShading, + wireframe: false, + shininess: 20, +}); + +markermaterial3 = new THREE.MeshPhongMaterial({ + color: 0x555555, + emissive: 0x999999, + side: THREE.DoubleSide, + // shading: THREE.FlatShading, + wireframe: false, + shininess: 20, +}); + + +var makeMarkerGeometry_Sphere10 = function(markerName, scale) { + return new THREE.SphereGeometry(10, 60, 60); +}; + +var makeMarkerGeometry_Sphere3 = function(markerName, scale) { + return new THREE.SphereGeometry(3, 60, 60); +}; + +var makeMarkerGeometry_SphereX = function(markerName, scale) { + return new THREE.SphereGeometry(5, 60, 60); +}; + +var makeJointGeometry_SphereX = function(X) { + return function(jointName, scale) { + return new THREE.SphereGeometry(X, 60, 60); + }; +}; + + +var makeJointGeometry_Sphere1 = function(jointName, scale) { + return new THREE.SphereGeometry(2 / scale, 60, 60); +}; + +var makeJointGeometry_Sphere2 = function(jointName, scale) { + return new THREE.SphereGeometry(1 / scale, 60, 60); +}; + +var makeJointGeometry_Dode = function(jointName, scale) { + return new THREE.DodecahedronGeometry(1 / scale, 0); +}; + +var makeBoneGeometry_Cylinder1 = function(joint1Name, joint2Name, length, scale) { + return new THREE.CylinderGeometry(1.5 / scale, 0.7 / scale, length, 40); +}; + +var makeBoneGeometry_Cylinder2 = function(joint1Name, joint2Name, length, scale) { + // if (joint1Name.includes("LeftHip")) + // length = 400; + return new THREE.CylinderGeometry(1.5 / scale, 0.2 / scale, length, 40); +}; + +var makeBoneGeometry_Cylinder3 = function(joint1Name, joint2Name, length, scale) { + var c1 = new THREE.CylinderGeometry(1.5 / scale, 0.2 / scale, length / 1, 20); + var c2 = new THREE.CylinderGeometry(0.2 / scale, 1.5 / scale, length / 1, 40); + + var material = new THREE.MeshPhongMaterial({ + color: 0xF7FE2E + }); + var mmesh = new THREE.Mesh(c1, material); + mmesh.updateMatrix(); + c2.merge(mmesh.geometry, mmesh.matrix); + return c2; +}; + +var makeBoneGeometry_Box1 = function(joint1Name, joint2Name, length, scale) { + return new THREE.BoxGeometry(1 / scale, length, 1 / scale, 40); +}; + + +var makeJointGeometry_Empty = function(jointName, scale) { + return new THREE.SphereGeometry(0.001, 60, 60); +}; + +var makeBoneGeometry_Empty = function(joint1Name, joint2Name, length, scale) { + return new THREE.CylinderGeometry(0.001, 0.001, 0.001, 40); +}; diff --git a/dataloaders/pymo/mocapplayer/libs/jquery.min.js b/dataloaders/pymo/mocapplayer/libs/jquery.min.js new file mode 100644 index 0000000000000000000000000000000000000000..b8c4187de18dd413ad3029839ce0773549e92a14 --- /dev/null +++ b/dataloaders/pymo/mocapplayer/libs/jquery.min.js @@ -0,0 +1,4 @@ +/*! jQuery v2.2.3 | (c) jQuery Foundation | jquery.org/license */ +!function(a,b){"object"==typeof module&&"object"==typeof module.exports?module.exports=a.document?b(a,!0):function(a){if(!a.document)throw new Error("jQuery requires a window with a document");return b(a)}:b(a)}("undefined"!=typeof window?window:this,function(a,b){var c=[],d=a.document,e=c.slice,f=c.concat,g=c.push,h=c.indexOf,i={},j=i.toString,k=i.hasOwnProperty,l={},m="2.2.3",n=function(a,b){return new n.fn.init(a,b)},o=/^[\s\uFEFF\xA0]+|[\s\uFEFF\xA0]+$/g,p=/^-ms-/,q=/-([\da-z])/gi,r=function(a,b){return b.toUpperCase()};n.fn=n.prototype={jquery:m,constructor:n,selector:"",length:0,toArray:function(){return e.call(this)},get:function(a){return null!=a?0>a?this[a+this.length]:this[a]:e.call(this)},pushStack:function(a){var b=n.merge(this.constructor(),a);return b.prevObject=this,b.context=this.context,b},each:function(a){return n.each(this,a)},map:function(a){return this.pushStack(n.map(this,function(b,c){return a.call(b,c,b)}))},slice:function(){return this.pushStack(e.apply(this,arguments))},first:function(){return this.eq(0)},last:function(){return this.eq(-1)},eq:function(a){var b=this.length,c=+a+(0>a?b:0);return this.pushStack(c>=0&&b>c?[this[c]]:[])},end:function(){return this.prevObject||this.constructor()},push:g,sort:c.sort,splice:c.splice},n.extend=n.fn.extend=function(){var a,b,c,d,e,f,g=arguments[0]||{},h=1,i=arguments.length,j=!1;for("boolean"==typeof g&&(j=g,g=arguments[h]||{},h++),"object"==typeof g||n.isFunction(g)||(g={}),h===i&&(g=this,h--);i>h;h++)if(null!=(a=arguments[h]))for(b in a)c=g[b],d=a[b],g!==d&&(j&&d&&(n.isPlainObject(d)||(e=n.isArray(d)))?(e?(e=!1,f=c&&n.isArray(c)?c:[]):f=c&&n.isPlainObject(c)?c:{},g[b]=n.extend(j,f,d)):void 0!==d&&(g[b]=d));return g},n.extend({expando:"jQuery"+(m+Math.random()).replace(/\D/g,""),isReady:!0,error:function(a){throw new Error(a)},noop:function(){},isFunction:function(a){return"function"===n.type(a)},isArray:Array.isArray,isWindow:function(a){return null!=a&&a===a.window},isNumeric:function(a){var b=a&&a.toString();return!n.isArray(a)&&b-parseFloat(b)+1>=0},isPlainObject:function(a){var b;if("object"!==n.type(a)||a.nodeType||n.isWindow(a))return!1;if(a.constructor&&!k.call(a,"constructor")&&!k.call(a.constructor.prototype||{},"isPrototypeOf"))return!1;for(b in a);return void 0===b||k.call(a,b)},isEmptyObject:function(a){var b;for(b in a)return!1;return!0},type:function(a){return null==a?a+"":"object"==typeof a||"function"==typeof a?i[j.call(a)]||"object":typeof a},globalEval:function(a){var b,c=eval;a=n.trim(a),a&&(1===a.indexOf("use strict")?(b=d.createElement("script"),b.text=a,d.head.appendChild(b).parentNode.removeChild(b)):c(a))},camelCase:function(a){return a.replace(p,"ms-").replace(q,r)},nodeName:function(a,b){return a.nodeName&&a.nodeName.toLowerCase()===b.toLowerCase()},each:function(a,b){var c,d=0;if(s(a)){for(c=a.length;c>d;d++)if(b.call(a[d],d,a[d])===!1)break}else for(d in a)if(b.call(a[d],d,a[d])===!1)break;return a},trim:function(a){return null==a?"":(a+"").replace(o,"")},makeArray:function(a,b){var c=b||[];return null!=a&&(s(Object(a))?n.merge(c,"string"==typeof a?[a]:a):g.call(c,a)),c},inArray:function(a,b,c){return null==b?-1:h.call(b,a,c)},merge:function(a,b){for(var c=+b.length,d=0,e=a.length;c>d;d++)a[e++]=b[d];return a.length=e,a},grep:function(a,b,c){for(var d,e=[],f=0,g=a.length,h=!c;g>f;f++)d=!b(a[f],f),d!==h&&e.push(a[f]);return e},map:function(a,b,c){var d,e,g=0,h=[];if(s(a))for(d=a.length;d>g;g++)e=b(a[g],g,c),null!=e&&h.push(e);else for(g in a)e=b(a[g],g,c),null!=e&&h.push(e);return f.apply([],h)},guid:1,proxy:function(a,b){var c,d,f;return"string"==typeof b&&(c=a[b],b=a,a=c),n.isFunction(a)?(d=e.call(arguments,2),f=function(){return a.apply(b||this,d.concat(e.call(arguments)))},f.guid=a.guid=a.guid||n.guid++,f):void 0},now:Date.now,support:l}),"function"==typeof Symbol&&(n.fn[Symbol.iterator]=c[Symbol.iterator]),n.each("Boolean Number String Function Array Date RegExp Object Error Symbol".split(" "),function(a,b){i["[object "+b+"]"]=b.toLowerCase()});function s(a){var b=!!a&&"length"in a&&a.length,c=n.type(a);return"function"===c||n.isWindow(a)?!1:"array"===c||0===b||"number"==typeof b&&b>0&&b-1 in a}var t=function(a){var b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u="sizzle"+1*new Date,v=a.document,w=0,x=0,y=ga(),z=ga(),A=ga(),B=function(a,b){return a===b&&(l=!0),0},C=1<<31,D={}.hasOwnProperty,E=[],F=E.pop,G=E.push,H=E.push,I=E.slice,J=function(a,b){for(var c=0,d=a.length;d>c;c++)if(a[c]===b)return c;return-1},K="checked|selected|async|autofocus|autoplay|controls|defer|disabled|hidden|ismap|loop|multiple|open|readonly|required|scoped",L="[\\x20\\t\\r\\n\\f]",M="(?:\\\\.|[\\w-]|[^\\x00-\\xa0])+",N="\\["+L+"*("+M+")(?:"+L+"*([*^$|!~]?=)"+L+"*(?:'((?:\\\\.|[^\\\\'])*)'|\"((?:\\\\.|[^\\\\\"])*)\"|("+M+"))|)"+L+"*\\]",O=":("+M+")(?:\\((('((?:\\\\.|[^\\\\'])*)'|\"((?:\\\\.|[^\\\\\"])*)\")|((?:\\\\.|[^\\\\()[\\]]|"+N+")*)|.*)\\)|)",P=new RegExp(L+"+","g"),Q=new RegExp("^"+L+"+|((?:^|[^\\\\])(?:\\\\.)*)"+L+"+$","g"),R=new RegExp("^"+L+"*,"+L+"*"),S=new RegExp("^"+L+"*([>+~]|"+L+")"+L+"*"),T=new RegExp("="+L+"*([^\\]'\"]*?)"+L+"*\\]","g"),U=new RegExp(O),V=new RegExp("^"+M+"$"),W={ID:new RegExp("^#("+M+")"),CLASS:new RegExp("^\\.("+M+")"),TAG:new RegExp("^("+M+"|[*])"),ATTR:new RegExp("^"+N),PSEUDO:new RegExp("^"+O),CHILD:new RegExp("^:(only|first|last|nth|nth-last)-(child|of-type)(?:\\("+L+"*(even|odd|(([+-]|)(\\d*)n|)"+L+"*(?:([+-]|)"+L+"*(\\d+)|))"+L+"*\\)|)","i"),bool:new RegExp("^(?:"+K+")$","i"),needsContext:new RegExp("^"+L+"*[>+~]|:(even|odd|eq|gt|lt|nth|first|last)(?:\\("+L+"*((?:-\\d)?\\d*)"+L+"*\\)|)(?=[^-]|$)","i")},X=/^(?:input|select|textarea|button)$/i,Y=/^h\d$/i,Z=/^[^{]+\{\s*\[native \w/,$=/^(?:#([\w-]+)|(\w+)|\.([\w-]+))$/,_=/[+~]/,aa=/'|\\/g,ba=new RegExp("\\\\([\\da-f]{1,6}"+L+"?|("+L+")|.)","ig"),ca=function(a,b,c){var d="0x"+b-65536;return d!==d||c?b:0>d?String.fromCharCode(d+65536):String.fromCharCode(d>>10|55296,1023&d|56320)},da=function(){m()};try{H.apply(E=I.call(v.childNodes),v.childNodes),E[v.childNodes.length].nodeType}catch(ea){H={apply:E.length?function(a,b){G.apply(a,I.call(b))}:function(a,b){var c=a.length,d=0;while(a[c++]=b[d++]);a.length=c-1}}}function fa(a,b,d,e){var f,h,j,k,l,o,r,s,w=b&&b.ownerDocument,x=b?b.nodeType:9;if(d=d||[],"string"!=typeof a||!a||1!==x&&9!==x&&11!==x)return d;if(!e&&((b?b.ownerDocument||b:v)!==n&&m(b),b=b||n,p)){if(11!==x&&(o=$.exec(a)))if(f=o[1]){if(9===x){if(!(j=b.getElementById(f)))return d;if(j.id===f)return d.push(j),d}else if(w&&(j=w.getElementById(f))&&t(b,j)&&j.id===f)return d.push(j),d}else{if(o[2])return H.apply(d,b.getElementsByTagName(a)),d;if((f=o[3])&&c.getElementsByClassName&&b.getElementsByClassName)return H.apply(d,b.getElementsByClassName(f)),d}if(c.qsa&&!A[a+" "]&&(!q||!q.test(a))){if(1!==x)w=b,s=a;else if("object"!==b.nodeName.toLowerCase()){(k=b.getAttribute("id"))?k=k.replace(aa,"\\$&"):b.setAttribute("id",k=u),r=g(a),h=r.length,l=V.test(k)?"#"+k:"[id='"+k+"']";while(h--)r[h]=l+" "+qa(r[h]);s=r.join(","),w=_.test(a)&&oa(b.parentNode)||b}if(s)try{return H.apply(d,w.querySelectorAll(s)),d}catch(y){}finally{k===u&&b.removeAttribute("id")}}}return i(a.replace(Q,"$1"),b,d,e)}function ga(){var a=[];function b(c,e){return a.push(c+" ")>d.cacheLength&&delete b[a.shift()],b[c+" "]=e}return b}function ha(a){return a[u]=!0,a}function ia(a){var b=n.createElement("div");try{return!!a(b)}catch(c){return!1}finally{b.parentNode&&b.parentNode.removeChild(b),b=null}}function ja(a,b){var c=a.split("|"),e=c.length;while(e--)d.attrHandle[c[e]]=b}function ka(a,b){var c=b&&a,d=c&&1===a.nodeType&&1===b.nodeType&&(~b.sourceIndex||C)-(~a.sourceIndex||C);if(d)return d;if(c)while(c=c.nextSibling)if(c===b)return-1;return a?1:-1}function la(a){return function(b){var c=b.nodeName.toLowerCase();return"input"===c&&b.type===a}}function ma(a){return function(b){var c=b.nodeName.toLowerCase();return("input"===c||"button"===c)&&b.type===a}}function na(a){return ha(function(b){return b=+b,ha(function(c,d){var e,f=a([],c.length,b),g=f.length;while(g--)c[e=f[g]]&&(c[e]=!(d[e]=c[e]))})})}function oa(a){return a&&"undefined"!=typeof a.getElementsByTagName&&a}c=fa.support={},f=fa.isXML=function(a){var b=a&&(a.ownerDocument||a).documentElement;return b?"HTML"!==b.nodeName:!1},m=fa.setDocument=function(a){var b,e,g=a?a.ownerDocument||a:v;return g!==n&&9===g.nodeType&&g.documentElement?(n=g,o=n.documentElement,p=!f(n),(e=n.defaultView)&&e.top!==e&&(e.addEventListener?e.addEventListener("unload",da,!1):e.attachEvent&&e.attachEvent("onunload",da)),c.attributes=ia(function(a){return a.className="i",!a.getAttribute("className")}),c.getElementsByTagName=ia(function(a){return a.appendChild(n.createComment("")),!a.getElementsByTagName("*").length}),c.getElementsByClassName=Z.test(n.getElementsByClassName),c.getById=ia(function(a){return o.appendChild(a).id=u,!n.getElementsByName||!n.getElementsByName(u).length}),c.getById?(d.find.ID=function(a,b){if("undefined"!=typeof b.getElementById&&p){var c=b.getElementById(a);return c?[c]:[]}},d.filter.ID=function(a){var b=a.replace(ba,ca);return function(a){return a.getAttribute("id")===b}}):(delete d.find.ID,d.filter.ID=function(a){var b=a.replace(ba,ca);return function(a){var c="undefined"!=typeof a.getAttributeNode&&a.getAttributeNode("id");return c&&c.value===b}}),d.find.TAG=c.getElementsByTagName?function(a,b){return"undefined"!=typeof b.getElementsByTagName?b.getElementsByTagName(a):c.qsa?b.querySelectorAll(a):void 0}:function(a,b){var c,d=[],e=0,f=b.getElementsByTagName(a);if("*"===a){while(c=f[e++])1===c.nodeType&&d.push(c);return d}return f},d.find.CLASS=c.getElementsByClassName&&function(a,b){return"undefined"!=typeof b.getElementsByClassName&&p?b.getElementsByClassName(a):void 0},r=[],q=[],(c.qsa=Z.test(n.querySelectorAll))&&(ia(function(a){o.appendChild(a).innerHTML="",a.querySelectorAll("[msallowcapture^='']").length&&q.push("[*^$]="+L+"*(?:''|\"\")"),a.querySelectorAll("[selected]").length||q.push("\\["+L+"*(?:value|"+K+")"),a.querySelectorAll("[id~="+u+"-]").length||q.push("~="),a.querySelectorAll(":checked").length||q.push(":checked"),a.querySelectorAll("a#"+u+"+*").length||q.push(".#.+[+~]")}),ia(function(a){var b=n.createElement("input");b.setAttribute("type","hidden"),a.appendChild(b).setAttribute("name","D"),a.querySelectorAll("[name=d]").length&&q.push("name"+L+"*[*^$|!~]?="),a.querySelectorAll(":enabled").length||q.push(":enabled",":disabled"),a.querySelectorAll("*,:x"),q.push(",.*:")})),(c.matchesSelector=Z.test(s=o.matches||o.webkitMatchesSelector||o.mozMatchesSelector||o.oMatchesSelector||o.msMatchesSelector))&&ia(function(a){c.disconnectedMatch=s.call(a,"div"),s.call(a,"[s!='']:x"),r.push("!=",O)}),q=q.length&&new RegExp(q.join("|")),r=r.length&&new RegExp(r.join("|")),b=Z.test(o.compareDocumentPosition),t=b||Z.test(o.contains)?function(a,b){var c=9===a.nodeType?a.documentElement:a,d=b&&b.parentNode;return a===d||!(!d||1!==d.nodeType||!(c.contains?c.contains(d):a.compareDocumentPosition&&16&a.compareDocumentPosition(d)))}:function(a,b){if(b)while(b=b.parentNode)if(b===a)return!0;return!1},B=b?function(a,b){if(a===b)return l=!0,0;var d=!a.compareDocumentPosition-!b.compareDocumentPosition;return d?d:(d=(a.ownerDocument||a)===(b.ownerDocument||b)?a.compareDocumentPosition(b):1,1&d||!c.sortDetached&&b.compareDocumentPosition(a)===d?a===n||a.ownerDocument===v&&t(v,a)?-1:b===n||b.ownerDocument===v&&t(v,b)?1:k?J(k,a)-J(k,b):0:4&d?-1:1)}:function(a,b){if(a===b)return l=!0,0;var c,d=0,e=a.parentNode,f=b.parentNode,g=[a],h=[b];if(!e||!f)return a===n?-1:b===n?1:e?-1:f?1:k?J(k,a)-J(k,b):0;if(e===f)return ka(a,b);c=a;while(c=c.parentNode)g.unshift(c);c=b;while(c=c.parentNode)h.unshift(c);while(g[d]===h[d])d++;return d?ka(g[d],h[d]):g[d]===v?-1:h[d]===v?1:0},n):n},fa.matches=function(a,b){return fa(a,null,null,b)},fa.matchesSelector=function(a,b){if((a.ownerDocument||a)!==n&&m(a),b=b.replace(T,"='$1']"),c.matchesSelector&&p&&!A[b+" "]&&(!r||!r.test(b))&&(!q||!q.test(b)))try{var d=s.call(a,b);if(d||c.disconnectedMatch||a.document&&11!==a.document.nodeType)return d}catch(e){}return fa(b,n,null,[a]).length>0},fa.contains=function(a,b){return(a.ownerDocument||a)!==n&&m(a),t(a,b)},fa.attr=function(a,b){(a.ownerDocument||a)!==n&&m(a);var e=d.attrHandle[b.toLowerCase()],f=e&&D.call(d.attrHandle,b.toLowerCase())?e(a,b,!p):void 0;return void 0!==f?f:c.attributes||!p?a.getAttribute(b):(f=a.getAttributeNode(b))&&f.specified?f.value:null},fa.error=function(a){throw new Error("Syntax error, unrecognized expression: "+a)},fa.uniqueSort=function(a){var b,d=[],e=0,f=0;if(l=!c.detectDuplicates,k=!c.sortStable&&a.slice(0),a.sort(B),l){while(b=a[f++])b===a[f]&&(e=d.push(f));while(e--)a.splice(d[e],1)}return k=null,a},e=fa.getText=function(a){var b,c="",d=0,f=a.nodeType;if(f){if(1===f||9===f||11===f){if("string"==typeof a.textContent)return a.textContent;for(a=a.firstChild;a;a=a.nextSibling)c+=e(a)}else if(3===f||4===f)return a.nodeValue}else while(b=a[d++])c+=e(b);return c},d=fa.selectors={cacheLength:50,createPseudo:ha,match:W,attrHandle:{},find:{},relative:{">":{dir:"parentNode",first:!0}," ":{dir:"parentNode"},"+":{dir:"previousSibling",first:!0},"~":{dir:"previousSibling"}},preFilter:{ATTR:function(a){return a[1]=a[1].replace(ba,ca),a[3]=(a[3]||a[4]||a[5]||"").replace(ba,ca),"~="===a[2]&&(a[3]=" "+a[3]+" "),a.slice(0,4)},CHILD:function(a){return a[1]=a[1].toLowerCase(),"nth"===a[1].slice(0,3)?(a[3]||fa.error(a[0]),a[4]=+(a[4]?a[5]+(a[6]||1):2*("even"===a[3]||"odd"===a[3])),a[5]=+(a[7]+a[8]||"odd"===a[3])):a[3]&&fa.error(a[0]),a},PSEUDO:function(a){var b,c=!a[6]&&a[2];return W.CHILD.test(a[0])?null:(a[3]?a[2]=a[4]||a[5]||"":c&&U.test(c)&&(b=g(c,!0))&&(b=c.indexOf(")",c.length-b)-c.length)&&(a[0]=a[0].slice(0,b),a[2]=c.slice(0,b)),a.slice(0,3))}},filter:{TAG:function(a){var b=a.replace(ba,ca).toLowerCase();return"*"===a?function(){return!0}:function(a){return a.nodeName&&a.nodeName.toLowerCase()===b}},CLASS:function(a){var b=y[a+" "];return b||(b=new RegExp("(^|"+L+")"+a+"("+L+"|$)"))&&y(a,function(a){return b.test("string"==typeof a.className&&a.className||"undefined"!=typeof a.getAttribute&&a.getAttribute("class")||"")})},ATTR:function(a,b,c){return function(d){var e=fa.attr(d,a);return null==e?"!="===b:b?(e+="","="===b?e===c:"!="===b?e!==c:"^="===b?c&&0===e.indexOf(c):"*="===b?c&&e.indexOf(c)>-1:"$="===b?c&&e.slice(-c.length)===c:"~="===b?(" "+e.replace(P," ")+" ").indexOf(c)>-1:"|="===b?e===c||e.slice(0,c.length+1)===c+"-":!1):!0}},CHILD:function(a,b,c,d,e){var f="nth"!==a.slice(0,3),g="last"!==a.slice(-4),h="of-type"===b;return 1===d&&0===e?function(a){return!!a.parentNode}:function(b,c,i){var j,k,l,m,n,o,p=f!==g?"nextSibling":"previousSibling",q=b.parentNode,r=h&&b.nodeName.toLowerCase(),s=!i&&!h,t=!1;if(q){if(f){while(p){m=b;while(m=m[p])if(h?m.nodeName.toLowerCase()===r:1===m.nodeType)return!1;o=p="only"===a&&!o&&"nextSibling"}return!0}if(o=[g?q.firstChild:q.lastChild],g&&s){m=q,l=m[u]||(m[u]={}),k=l[m.uniqueID]||(l[m.uniqueID]={}),j=k[a]||[],n=j[0]===w&&j[1],t=n&&j[2],m=n&&q.childNodes[n];while(m=++n&&m&&m[p]||(t=n=0)||o.pop())if(1===m.nodeType&&++t&&m===b){k[a]=[w,n,t];break}}else if(s&&(m=b,l=m[u]||(m[u]={}),k=l[m.uniqueID]||(l[m.uniqueID]={}),j=k[a]||[],n=j[0]===w&&j[1],t=n),t===!1)while(m=++n&&m&&m[p]||(t=n=0)||o.pop())if((h?m.nodeName.toLowerCase()===r:1===m.nodeType)&&++t&&(s&&(l=m[u]||(m[u]={}),k=l[m.uniqueID]||(l[m.uniqueID]={}),k[a]=[w,t]),m===b))break;return t-=e,t===d||t%d===0&&t/d>=0}}},PSEUDO:function(a,b){var c,e=d.pseudos[a]||d.setFilters[a.toLowerCase()]||fa.error("unsupported pseudo: "+a);return e[u]?e(b):e.length>1?(c=[a,a,"",b],d.setFilters.hasOwnProperty(a.toLowerCase())?ha(function(a,c){var d,f=e(a,b),g=f.length;while(g--)d=J(a,f[g]),a[d]=!(c[d]=f[g])}):function(a){return e(a,0,c)}):e}},pseudos:{not:ha(function(a){var b=[],c=[],d=h(a.replace(Q,"$1"));return d[u]?ha(function(a,b,c,e){var f,g=d(a,null,e,[]),h=a.length;while(h--)(f=g[h])&&(a[h]=!(b[h]=f))}):function(a,e,f){return b[0]=a,d(b,null,f,c),b[0]=null,!c.pop()}}),has:ha(function(a){return function(b){return fa(a,b).length>0}}),contains:ha(function(a){return a=a.replace(ba,ca),function(b){return(b.textContent||b.innerText||e(b)).indexOf(a)>-1}}),lang:ha(function(a){return V.test(a||"")||fa.error("unsupported lang: "+a),a=a.replace(ba,ca).toLowerCase(),function(b){var c;do if(c=p?b.lang:b.getAttribute("xml:lang")||b.getAttribute("lang"))return c=c.toLowerCase(),c===a||0===c.indexOf(a+"-");while((b=b.parentNode)&&1===b.nodeType);return!1}}),target:function(b){var c=a.location&&a.location.hash;return c&&c.slice(1)===b.id},root:function(a){return a===o},focus:function(a){return a===n.activeElement&&(!n.hasFocus||n.hasFocus())&&!!(a.type||a.href||~a.tabIndex)},enabled:function(a){return a.disabled===!1},disabled:function(a){return a.disabled===!0},checked:function(a){var b=a.nodeName.toLowerCase();return"input"===b&&!!a.checked||"option"===b&&!!a.selected},selected:function(a){return a.parentNode&&a.parentNode.selectedIndex,a.selected===!0},empty:function(a){for(a=a.firstChild;a;a=a.nextSibling)if(a.nodeType<6)return!1;return!0},parent:function(a){return!d.pseudos.empty(a)},header:function(a){return Y.test(a.nodeName)},input:function(a){return X.test(a.nodeName)},button:function(a){var b=a.nodeName.toLowerCase();return"input"===b&&"button"===a.type||"button"===b},text:function(a){var b;return"input"===a.nodeName.toLowerCase()&&"text"===a.type&&(null==(b=a.getAttribute("type"))||"text"===b.toLowerCase())},first:na(function(){return[0]}),last:na(function(a,b){return[b-1]}),eq:na(function(a,b,c){return[0>c?c+b:c]}),even:na(function(a,b){for(var c=0;b>c;c+=2)a.push(c);return a}),odd:na(function(a,b){for(var c=1;b>c;c+=2)a.push(c);return a}),lt:na(function(a,b,c){for(var d=0>c?c+b:c;--d>=0;)a.push(d);return a}),gt:na(function(a,b,c){for(var d=0>c?c+b:c;++db;b++)d+=a[b].value;return d}function ra(a,b,c){var d=b.dir,e=c&&"parentNode"===d,f=x++;return b.first?function(b,c,f){while(b=b[d])if(1===b.nodeType||e)return a(b,c,f)}:function(b,c,g){var h,i,j,k=[w,f];if(g){while(b=b[d])if((1===b.nodeType||e)&&a(b,c,g))return!0}else while(b=b[d])if(1===b.nodeType||e){if(j=b[u]||(b[u]={}),i=j[b.uniqueID]||(j[b.uniqueID]={}),(h=i[d])&&h[0]===w&&h[1]===f)return k[2]=h[2];if(i[d]=k,k[2]=a(b,c,g))return!0}}}function sa(a){return a.length>1?function(b,c,d){var e=a.length;while(e--)if(!a[e](b,c,d))return!1;return!0}:a[0]}function ta(a,b,c){for(var d=0,e=b.length;e>d;d++)fa(a,b[d],c);return c}function ua(a,b,c,d,e){for(var f,g=[],h=0,i=a.length,j=null!=b;i>h;h++)(f=a[h])&&(c&&!c(f,d,e)||(g.push(f),j&&b.push(h)));return g}function va(a,b,c,d,e,f){return d&&!d[u]&&(d=va(d)),e&&!e[u]&&(e=va(e,f)),ha(function(f,g,h,i){var j,k,l,m=[],n=[],o=g.length,p=f||ta(b||"*",h.nodeType?[h]:h,[]),q=!a||!f&&b?p:ua(p,m,a,h,i),r=c?e||(f?a:o||d)?[]:g:q;if(c&&c(q,r,h,i),d){j=ua(r,n),d(j,[],h,i),k=j.length;while(k--)(l=j[k])&&(r[n[k]]=!(q[n[k]]=l))}if(f){if(e||a){if(e){j=[],k=r.length;while(k--)(l=r[k])&&j.push(q[k]=l);e(null,r=[],j,i)}k=r.length;while(k--)(l=r[k])&&(j=e?J(f,l):m[k])>-1&&(f[j]=!(g[j]=l))}}else r=ua(r===g?r.splice(o,r.length):r),e?e(null,g,r,i):H.apply(g,r)})}function wa(a){for(var b,c,e,f=a.length,g=d.relative[a[0].type],h=g||d.relative[" "],i=g?1:0,k=ra(function(a){return a===b},h,!0),l=ra(function(a){return J(b,a)>-1},h,!0),m=[function(a,c,d){var e=!g&&(d||c!==j)||((b=c).nodeType?k(a,c,d):l(a,c,d));return b=null,e}];f>i;i++)if(c=d.relative[a[i].type])m=[ra(sa(m),c)];else{if(c=d.filter[a[i].type].apply(null,a[i].matches),c[u]){for(e=++i;f>e;e++)if(d.relative[a[e].type])break;return va(i>1&&sa(m),i>1&&qa(a.slice(0,i-1).concat({value:" "===a[i-2].type?"*":""})).replace(Q,"$1"),c,e>i&&wa(a.slice(i,e)),f>e&&wa(a=a.slice(e)),f>e&&qa(a))}m.push(c)}return sa(m)}function xa(a,b){var c=b.length>0,e=a.length>0,f=function(f,g,h,i,k){var l,o,q,r=0,s="0",t=f&&[],u=[],v=j,x=f||e&&d.find.TAG("*",k),y=w+=null==v?1:Math.random()||.1,z=x.length;for(k&&(j=g===n||g||k);s!==z&&null!=(l=x[s]);s++){if(e&&l){o=0,g||l.ownerDocument===n||(m(l),h=!p);while(q=a[o++])if(q(l,g||n,h)){i.push(l);break}k&&(w=y)}c&&((l=!q&&l)&&r--,f&&t.push(l))}if(r+=s,c&&s!==r){o=0;while(q=b[o++])q(t,u,g,h);if(f){if(r>0)while(s--)t[s]||u[s]||(u[s]=F.call(i));u=ua(u)}H.apply(i,u),k&&!f&&u.length>0&&r+b.length>1&&fa.uniqueSort(i)}return k&&(w=y,j=v),t};return c?ha(f):f}return h=fa.compile=function(a,b){var c,d=[],e=[],f=A[a+" "];if(!f){b||(b=g(a)),c=b.length;while(c--)f=wa(b[c]),f[u]?d.push(f):e.push(f);f=A(a,xa(e,d)),f.selector=a}return f},i=fa.select=function(a,b,e,f){var i,j,k,l,m,n="function"==typeof a&&a,o=!f&&g(a=n.selector||a);if(e=e||[],1===o.length){if(j=o[0]=o[0].slice(0),j.length>2&&"ID"===(k=j[0]).type&&c.getById&&9===b.nodeType&&p&&d.relative[j[1].type]){if(b=(d.find.ID(k.matches[0].replace(ba,ca),b)||[])[0],!b)return e;n&&(b=b.parentNode),a=a.slice(j.shift().value.length)}i=W.needsContext.test(a)?0:j.length;while(i--){if(k=j[i],d.relative[l=k.type])break;if((m=d.find[l])&&(f=m(k.matches[0].replace(ba,ca),_.test(j[0].type)&&oa(b.parentNode)||b))){if(j.splice(i,1),a=f.length&&qa(j),!a)return H.apply(e,f),e;break}}}return(n||h(a,o))(f,b,!p,e,!b||_.test(a)&&oa(b.parentNode)||b),e},c.sortStable=u.split("").sort(B).join("")===u,c.detectDuplicates=!!l,m(),c.sortDetached=ia(function(a){return 1&a.compareDocumentPosition(n.createElement("div"))}),ia(function(a){return a.innerHTML="","#"===a.firstChild.getAttribute("href")})||ja("type|href|height|width",function(a,b,c){return c?void 0:a.getAttribute(b,"type"===b.toLowerCase()?1:2)}),c.attributes&&ia(function(a){return a.innerHTML="",a.firstChild.setAttribute("value",""),""===a.firstChild.getAttribute("value")})||ja("value",function(a,b,c){return c||"input"!==a.nodeName.toLowerCase()?void 0:a.defaultValue}),ia(function(a){return null==a.getAttribute("disabled")})||ja(K,function(a,b,c){var d;return c?void 0:a[b]===!0?b.toLowerCase():(d=a.getAttributeNode(b))&&d.specified?d.value:null}),fa}(a);n.find=t,n.expr=t.selectors,n.expr[":"]=n.expr.pseudos,n.uniqueSort=n.unique=t.uniqueSort,n.text=t.getText,n.isXMLDoc=t.isXML,n.contains=t.contains;var u=function(a,b,c){var d=[],e=void 0!==c;while((a=a[b])&&9!==a.nodeType)if(1===a.nodeType){if(e&&n(a).is(c))break;d.push(a)}return d},v=function(a,b){for(var c=[];a;a=a.nextSibling)1===a.nodeType&&a!==b&&c.push(a);return c},w=n.expr.match.needsContext,x=/^<([\w-]+)\s*\/?>(?:<\/\1>|)$/,y=/^.[^:#\[\.,]*$/;function z(a,b,c){if(n.isFunction(b))return n.grep(a,function(a,d){return!!b.call(a,d,a)!==c});if(b.nodeType)return n.grep(a,function(a){return a===b!==c});if("string"==typeof b){if(y.test(b))return n.filter(b,a,c);b=n.filter(b,a)}return n.grep(a,function(a){return h.call(b,a)>-1!==c})}n.filter=function(a,b,c){var d=b[0];return c&&(a=":not("+a+")"),1===b.length&&1===d.nodeType?n.find.matchesSelector(d,a)?[d]:[]:n.find.matches(a,n.grep(b,function(a){return 1===a.nodeType}))},n.fn.extend({find:function(a){var b,c=this.length,d=[],e=this;if("string"!=typeof a)return this.pushStack(n(a).filter(function(){for(b=0;c>b;b++)if(n.contains(e[b],this))return!0}));for(b=0;c>b;b++)n.find(a,e[b],d);return d=this.pushStack(c>1?n.unique(d):d),d.selector=this.selector?this.selector+" "+a:a,d},filter:function(a){return this.pushStack(z(this,a||[],!1))},not:function(a){return this.pushStack(z(this,a||[],!0))},is:function(a){return!!z(this,"string"==typeof a&&w.test(a)?n(a):a||[],!1).length}});var A,B=/^(?:\s*(<[\w\W]+>)[^>]*|#([\w-]*))$/,C=n.fn.init=function(a,b,c){var e,f;if(!a)return this;if(c=c||A,"string"==typeof a){if(e="<"===a[0]&&">"===a[a.length-1]&&a.length>=3?[null,a,null]:B.exec(a),!e||!e[1]&&b)return!b||b.jquery?(b||c).find(a):this.constructor(b).find(a);if(e[1]){if(b=b instanceof n?b[0]:b,n.merge(this,n.parseHTML(e[1],b&&b.nodeType?b.ownerDocument||b:d,!0)),x.test(e[1])&&n.isPlainObject(b))for(e in b)n.isFunction(this[e])?this[e](b[e]):this.attr(e,b[e]);return this}return f=d.getElementById(e[2]),f&&f.parentNode&&(this.length=1,this[0]=f),this.context=d,this.selector=a,this}return a.nodeType?(this.context=this[0]=a,this.length=1,this):n.isFunction(a)?void 0!==c.ready?c.ready(a):a(n):(void 0!==a.selector&&(this.selector=a.selector,this.context=a.context),n.makeArray(a,this))};C.prototype=n.fn,A=n(d);var D=/^(?:parents|prev(?:Until|All))/,E={children:!0,contents:!0,next:!0,prev:!0};n.fn.extend({has:function(a){var b=n(a,this),c=b.length;return this.filter(function(){for(var a=0;c>a;a++)if(n.contains(this,b[a]))return!0})},closest:function(a,b){for(var c,d=0,e=this.length,f=[],g=w.test(a)||"string"!=typeof a?n(a,b||this.context):0;e>d;d++)for(c=this[d];c&&c!==b;c=c.parentNode)if(c.nodeType<11&&(g?g.index(c)>-1:1===c.nodeType&&n.find.matchesSelector(c,a))){f.push(c);break}return this.pushStack(f.length>1?n.uniqueSort(f):f)},index:function(a){return a?"string"==typeof a?h.call(n(a),this[0]):h.call(this,a.jquery?a[0]:a):this[0]&&this[0].parentNode?this.first().prevAll().length:-1},add:function(a,b){return this.pushStack(n.uniqueSort(n.merge(this.get(),n(a,b))))},addBack:function(a){return this.add(null==a?this.prevObject:this.prevObject.filter(a))}});function F(a,b){while((a=a[b])&&1!==a.nodeType);return a}n.each({parent:function(a){var b=a.parentNode;return b&&11!==b.nodeType?b:null},parents:function(a){return u(a,"parentNode")},parentsUntil:function(a,b,c){return u(a,"parentNode",c)},next:function(a){return F(a,"nextSibling")},prev:function(a){return F(a,"previousSibling")},nextAll:function(a){return u(a,"nextSibling")},prevAll:function(a){return u(a,"previousSibling")},nextUntil:function(a,b,c){return u(a,"nextSibling",c)},prevUntil:function(a,b,c){return u(a,"previousSibling",c)},siblings:function(a){return v((a.parentNode||{}).firstChild,a)},children:function(a){return v(a.firstChild)},contents:function(a){return a.contentDocument||n.merge([],a.childNodes)}},function(a,b){n.fn[a]=function(c,d){var e=n.map(this,b,c);return"Until"!==a.slice(-5)&&(d=c),d&&"string"==typeof d&&(e=n.filter(d,e)),this.length>1&&(E[a]||n.uniqueSort(e),D.test(a)&&e.reverse()),this.pushStack(e)}});var G=/\S+/g;function H(a){var b={};return n.each(a.match(G)||[],function(a,c){b[c]=!0}),b}n.Callbacks=function(a){a="string"==typeof a?H(a):n.extend({},a);var b,c,d,e,f=[],g=[],h=-1,i=function(){for(e=a.once,d=b=!0;g.length;h=-1){c=g.shift();while(++h-1)f.splice(c,1),h>=c&&h--}),this},has:function(a){return a?n.inArray(a,f)>-1:f.length>0},empty:function(){return f&&(f=[]),this},disable:function(){return e=g=[],f=c="",this},disabled:function(){return!f},lock:function(){return e=g=[],c||(f=c=""),this},locked:function(){return!!e},fireWith:function(a,c){return e||(c=c||[],c=[a,c.slice?c.slice():c],g.push(c),b||i()),this},fire:function(){return j.fireWith(this,arguments),this},fired:function(){return!!d}};return j},n.extend({Deferred:function(a){var b=[["resolve","done",n.Callbacks("once memory"),"resolved"],["reject","fail",n.Callbacks("once memory"),"rejected"],["notify","progress",n.Callbacks("memory")]],c="pending",d={state:function(){return c},always:function(){return e.done(arguments).fail(arguments),this},then:function(){var a=arguments;return n.Deferred(function(c){n.each(b,function(b,f){var g=n.isFunction(a[b])&&a[b];e[f[1]](function(){var a=g&&g.apply(this,arguments);a&&n.isFunction(a.promise)?a.promise().progress(c.notify).done(c.resolve).fail(c.reject):c[f[0]+"With"](this===d?c.promise():this,g?[a]:arguments)})}),a=null}).promise()},promise:function(a){return null!=a?n.extend(a,d):d}},e={};return d.pipe=d.then,n.each(b,function(a,f){var g=f[2],h=f[3];d[f[1]]=g.add,h&&g.add(function(){c=h},b[1^a][2].disable,b[2][2].lock),e[f[0]]=function(){return e[f[0]+"With"](this===e?d:this,arguments),this},e[f[0]+"With"]=g.fireWith}),d.promise(e),a&&a.call(e,e),e},when:function(a){var b=0,c=e.call(arguments),d=c.length,f=1!==d||a&&n.isFunction(a.promise)?d:0,g=1===f?a:n.Deferred(),h=function(a,b,c){return function(d){b[a]=this,c[a]=arguments.length>1?e.call(arguments):d,c===i?g.notifyWith(b,c):--f||g.resolveWith(b,c)}},i,j,k;if(d>1)for(i=new Array(d),j=new Array(d),k=new Array(d);d>b;b++)c[b]&&n.isFunction(c[b].promise)?c[b].promise().progress(h(b,j,i)).done(h(b,k,c)).fail(g.reject):--f;return f||g.resolveWith(k,c),g.promise()}});var I;n.fn.ready=function(a){return n.ready.promise().done(a),this},n.extend({isReady:!1,readyWait:1,holdReady:function(a){a?n.readyWait++:n.ready(!0)},ready:function(a){(a===!0?--n.readyWait:n.isReady)||(n.isReady=!0,a!==!0&&--n.readyWait>0||(I.resolveWith(d,[n]),n.fn.triggerHandler&&(n(d).triggerHandler("ready"),n(d).off("ready"))))}});function J(){d.removeEventListener("DOMContentLoaded",J),a.removeEventListener("load",J),n.ready()}n.ready.promise=function(b){return I||(I=n.Deferred(),"complete"===d.readyState||"loading"!==d.readyState&&!d.documentElement.doScroll?a.setTimeout(n.ready):(d.addEventListener("DOMContentLoaded",J),a.addEventListener("load",J))),I.promise(b)},n.ready.promise();var K=function(a,b,c,d,e,f,g){var h=0,i=a.length,j=null==c;if("object"===n.type(c)){e=!0;for(h in c)K(a,b,h,c[h],!0,f,g)}else if(void 0!==d&&(e=!0,n.isFunction(d)||(g=!0),j&&(g?(b.call(a,d),b=null):(j=b,b=function(a,b,c){return j.call(n(a),c)})),b))for(;i>h;h++)b(a[h],c,g?d:d.call(a[h],h,b(a[h],c)));return e?a:j?b.call(a):i?b(a[0],c):f},L=function(a){return 1===a.nodeType||9===a.nodeType||!+a.nodeType};function M(){this.expando=n.expando+M.uid++}M.uid=1,M.prototype={register:function(a,b){var c=b||{};return a.nodeType?a[this.expando]=c:Object.defineProperty(a,this.expando,{value:c,writable:!0,configurable:!0}),a[this.expando]},cache:function(a){if(!L(a))return{};var b=a[this.expando];return b||(b={},L(a)&&(a.nodeType?a[this.expando]=b:Object.defineProperty(a,this.expando,{value:b,configurable:!0}))),b},set:function(a,b,c){var d,e=this.cache(a);if("string"==typeof b)e[b]=c;else for(d in b)e[d]=b[d];return e},get:function(a,b){return void 0===b?this.cache(a):a[this.expando]&&a[this.expando][b]},access:function(a,b,c){var d;return void 0===b||b&&"string"==typeof b&&void 0===c?(d=this.get(a,b),void 0!==d?d:this.get(a,n.camelCase(b))):(this.set(a,b,c),void 0!==c?c:b)},remove:function(a,b){var c,d,e,f=a[this.expando];if(void 0!==f){if(void 0===b)this.register(a);else{n.isArray(b)?d=b.concat(b.map(n.camelCase)):(e=n.camelCase(b),b in f?d=[b,e]:(d=e,d=d in f?[d]:d.match(G)||[])),c=d.length;while(c--)delete f[d[c]]}(void 0===b||n.isEmptyObject(f))&&(a.nodeType?a[this.expando]=void 0:delete a[this.expando])}},hasData:function(a){var b=a[this.expando];return void 0!==b&&!n.isEmptyObject(b)}};var N=new M,O=new M,P=/^(?:\{[\w\W]*\}|\[[\w\W]*\])$/,Q=/[A-Z]/g;function R(a,b,c){var d;if(void 0===c&&1===a.nodeType)if(d="data-"+b.replace(Q,"-$&").toLowerCase(),c=a.getAttribute(d),"string"==typeof c){try{c="true"===c?!0:"false"===c?!1:"null"===c?null:+c+""===c?+c:P.test(c)?n.parseJSON(c):c; +}catch(e){}O.set(a,b,c)}else c=void 0;return c}n.extend({hasData:function(a){return O.hasData(a)||N.hasData(a)},data:function(a,b,c){return O.access(a,b,c)},removeData:function(a,b){O.remove(a,b)},_data:function(a,b,c){return N.access(a,b,c)},_removeData:function(a,b){N.remove(a,b)}}),n.fn.extend({data:function(a,b){var c,d,e,f=this[0],g=f&&f.attributes;if(void 0===a){if(this.length&&(e=O.get(f),1===f.nodeType&&!N.get(f,"hasDataAttrs"))){c=g.length;while(c--)g[c]&&(d=g[c].name,0===d.indexOf("data-")&&(d=n.camelCase(d.slice(5)),R(f,d,e[d])));N.set(f,"hasDataAttrs",!0)}return e}return"object"==typeof a?this.each(function(){O.set(this,a)}):K(this,function(b){var c,d;if(f&&void 0===b){if(c=O.get(f,a)||O.get(f,a.replace(Q,"-$&").toLowerCase()),void 0!==c)return c;if(d=n.camelCase(a),c=O.get(f,d),void 0!==c)return c;if(c=R(f,d,void 0),void 0!==c)return c}else d=n.camelCase(a),this.each(function(){var c=O.get(this,d);O.set(this,d,b),a.indexOf("-")>-1&&void 0!==c&&O.set(this,a,b)})},null,b,arguments.length>1,null,!0)},removeData:function(a){return this.each(function(){O.remove(this,a)})}}),n.extend({queue:function(a,b,c){var d;return a?(b=(b||"fx")+"queue",d=N.get(a,b),c&&(!d||n.isArray(c)?d=N.access(a,b,n.makeArray(c)):d.push(c)),d||[]):void 0},dequeue:function(a,b){b=b||"fx";var c=n.queue(a,b),d=c.length,e=c.shift(),f=n._queueHooks(a,b),g=function(){n.dequeue(a,b)};"inprogress"===e&&(e=c.shift(),d--),e&&("fx"===b&&c.unshift("inprogress"),delete f.stop,e.call(a,g,f)),!d&&f&&f.empty.fire()},_queueHooks:function(a,b){var c=b+"queueHooks";return N.get(a,c)||N.access(a,c,{empty:n.Callbacks("once memory").add(function(){N.remove(a,[b+"queue",c])})})}}),n.fn.extend({queue:function(a,b){var c=2;return"string"!=typeof a&&(b=a,a="fx",c--),arguments.length",""],thead:[1,"","
"],col:[2,"","
"],tr:[2,"","
"],td:[3,"","
"],_default:[0,"",""]};$.optgroup=$.option,$.tbody=$.tfoot=$.colgroup=$.caption=$.thead,$.th=$.td;function _(a,b){var c="undefined"!=typeof a.getElementsByTagName?a.getElementsByTagName(b||"*"):"undefined"!=typeof a.querySelectorAll?a.querySelectorAll(b||"*"):[];return void 0===b||b&&n.nodeName(a,b)?n.merge([a],c):c}function aa(a,b){for(var c=0,d=a.length;d>c;c++)N.set(a[c],"globalEval",!b||N.get(b[c],"globalEval"))}var ba=/<|&#?\w+;/;function ca(a,b,c,d,e){for(var f,g,h,i,j,k,l=b.createDocumentFragment(),m=[],o=0,p=a.length;p>o;o++)if(f=a[o],f||0===f)if("object"===n.type(f))n.merge(m,f.nodeType?[f]:f);else if(ba.test(f)){g=g||l.appendChild(b.createElement("div")),h=(Y.exec(f)||["",""])[1].toLowerCase(),i=$[h]||$._default,g.innerHTML=i[1]+n.htmlPrefilter(f)+i[2],k=i[0];while(k--)g=g.lastChild;n.merge(m,g.childNodes),g=l.firstChild,g.textContent=""}else m.push(b.createTextNode(f));l.textContent="",o=0;while(f=m[o++])if(d&&n.inArray(f,d)>-1)e&&e.push(f);else if(j=n.contains(f.ownerDocument,f),g=_(l.appendChild(f),"script"),j&&aa(g),c){k=0;while(f=g[k++])Z.test(f.type||"")&&c.push(f)}return l}!function(){var a=d.createDocumentFragment(),b=a.appendChild(d.createElement("div")),c=d.createElement("input");c.setAttribute("type","radio"),c.setAttribute("checked","checked"),c.setAttribute("name","t"),b.appendChild(c),l.checkClone=b.cloneNode(!0).cloneNode(!0).lastChild.checked,b.innerHTML="",l.noCloneChecked=!!b.cloneNode(!0).lastChild.defaultValue}();var da=/^key/,ea=/^(?:mouse|pointer|contextmenu|drag|drop)|click/,fa=/^([^.]*)(?:\.(.+)|)/;function ga(){return!0}function ha(){return!1}function ia(){try{return d.activeElement}catch(a){}}function ja(a,b,c,d,e,f){var g,h;if("object"==typeof b){"string"!=typeof c&&(d=d||c,c=void 0);for(h in b)ja(a,h,c,d,b[h],f);return a}if(null==d&&null==e?(e=c,d=c=void 0):null==e&&("string"==typeof c?(e=d,d=void 0):(e=d,d=c,c=void 0)),e===!1)e=ha;else if(!e)return a;return 1===f&&(g=e,e=function(a){return n().off(a),g.apply(this,arguments)},e.guid=g.guid||(g.guid=n.guid++)),a.each(function(){n.event.add(this,b,e,d,c)})}n.event={global:{},add:function(a,b,c,d,e){var f,g,h,i,j,k,l,m,o,p,q,r=N.get(a);if(r){c.handler&&(f=c,c=f.handler,e=f.selector),c.guid||(c.guid=n.guid++),(i=r.events)||(i=r.events={}),(g=r.handle)||(g=r.handle=function(b){return"undefined"!=typeof n&&n.event.triggered!==b.type?n.event.dispatch.apply(a,arguments):void 0}),b=(b||"").match(G)||[""],j=b.length;while(j--)h=fa.exec(b[j])||[],o=q=h[1],p=(h[2]||"").split(".").sort(),o&&(l=n.event.special[o]||{},o=(e?l.delegateType:l.bindType)||o,l=n.event.special[o]||{},k=n.extend({type:o,origType:q,data:d,handler:c,guid:c.guid,selector:e,needsContext:e&&n.expr.match.needsContext.test(e),namespace:p.join(".")},f),(m=i[o])||(m=i[o]=[],m.delegateCount=0,l.setup&&l.setup.call(a,d,p,g)!==!1||a.addEventListener&&a.addEventListener(o,g)),l.add&&(l.add.call(a,k),k.handler.guid||(k.handler.guid=c.guid)),e?m.splice(m.delegateCount++,0,k):m.push(k),n.event.global[o]=!0)}},remove:function(a,b,c,d,e){var f,g,h,i,j,k,l,m,o,p,q,r=N.hasData(a)&&N.get(a);if(r&&(i=r.events)){b=(b||"").match(G)||[""],j=b.length;while(j--)if(h=fa.exec(b[j])||[],o=q=h[1],p=(h[2]||"").split(".").sort(),o){l=n.event.special[o]||{},o=(d?l.delegateType:l.bindType)||o,m=i[o]||[],h=h[2]&&new RegExp("(^|\\.)"+p.join("\\.(?:.*\\.|)")+"(\\.|$)"),g=f=m.length;while(f--)k=m[f],!e&&q!==k.origType||c&&c.guid!==k.guid||h&&!h.test(k.namespace)||d&&d!==k.selector&&("**"!==d||!k.selector)||(m.splice(f,1),k.selector&&m.delegateCount--,l.remove&&l.remove.call(a,k));g&&!m.length&&(l.teardown&&l.teardown.call(a,p,r.handle)!==!1||n.removeEvent(a,o,r.handle),delete i[o])}else for(o in i)n.event.remove(a,o+b[j],c,d,!0);n.isEmptyObject(i)&&N.remove(a,"handle events")}},dispatch:function(a){a=n.event.fix(a);var b,c,d,f,g,h=[],i=e.call(arguments),j=(N.get(this,"events")||{})[a.type]||[],k=n.event.special[a.type]||{};if(i[0]=a,a.delegateTarget=this,!k.preDispatch||k.preDispatch.call(this,a)!==!1){h=n.event.handlers.call(this,a,j),b=0;while((f=h[b++])&&!a.isPropagationStopped()){a.currentTarget=f.elem,c=0;while((g=f.handlers[c++])&&!a.isImmediatePropagationStopped())a.rnamespace&&!a.rnamespace.test(g.namespace)||(a.handleObj=g,a.data=g.data,d=((n.event.special[g.origType]||{}).handle||g.handler).apply(f.elem,i),void 0!==d&&(a.result=d)===!1&&(a.preventDefault(),a.stopPropagation()))}return k.postDispatch&&k.postDispatch.call(this,a),a.result}},handlers:function(a,b){var c,d,e,f,g=[],h=b.delegateCount,i=a.target;if(h&&i.nodeType&&("click"!==a.type||isNaN(a.button)||a.button<1))for(;i!==this;i=i.parentNode||this)if(1===i.nodeType&&(i.disabled!==!0||"click"!==a.type)){for(d=[],c=0;h>c;c++)f=b[c],e=f.selector+" ",void 0===d[e]&&(d[e]=f.needsContext?n(e,this).index(i)>-1:n.find(e,this,null,[i]).length),d[e]&&d.push(f);d.length&&g.push({elem:i,handlers:d})}return h]*)\/>/gi,la=/\s*$/g;function pa(a,b){return n.nodeName(a,"table")&&n.nodeName(11!==b.nodeType?b:b.firstChild,"tr")?a.getElementsByTagName("tbody")[0]||a.appendChild(a.ownerDocument.createElement("tbody")):a}function qa(a){return a.type=(null!==a.getAttribute("type"))+"/"+a.type,a}function ra(a){var b=na.exec(a.type);return b?a.type=b[1]:a.removeAttribute("type"),a}function sa(a,b){var c,d,e,f,g,h,i,j;if(1===b.nodeType){if(N.hasData(a)&&(f=N.access(a),g=N.set(b,f),j=f.events)){delete g.handle,g.events={};for(e in j)for(c=0,d=j[e].length;d>c;c++)n.event.add(b,e,j[e][c])}O.hasData(a)&&(h=O.access(a),i=n.extend({},h),O.set(b,i))}}function ta(a,b){var c=b.nodeName.toLowerCase();"input"===c&&X.test(a.type)?b.checked=a.checked:"input"!==c&&"textarea"!==c||(b.defaultValue=a.defaultValue)}function ua(a,b,c,d){b=f.apply([],b);var e,g,h,i,j,k,m=0,o=a.length,p=o-1,q=b[0],r=n.isFunction(q);if(r||o>1&&"string"==typeof q&&!l.checkClone&&ma.test(q))return a.each(function(e){var f=a.eq(e);r&&(b[0]=q.call(this,e,f.html())),ua(f,b,c,d)});if(o&&(e=ca(b,a[0].ownerDocument,!1,a,d),g=e.firstChild,1===e.childNodes.length&&(e=g),g||d)){for(h=n.map(_(e,"script"),qa),i=h.length;o>m;m++)j=e,m!==p&&(j=n.clone(j,!0,!0),i&&n.merge(h,_(j,"script"))),c.call(a[m],j,m);if(i)for(k=h[h.length-1].ownerDocument,n.map(h,ra),m=0;i>m;m++)j=h[m],Z.test(j.type||"")&&!N.access(j,"globalEval")&&n.contains(k,j)&&(j.src?n._evalUrl&&n._evalUrl(j.src):n.globalEval(j.textContent.replace(oa,"")))}return a}function va(a,b,c){for(var d,e=b?n.filter(b,a):a,f=0;null!=(d=e[f]);f++)c||1!==d.nodeType||n.cleanData(_(d)),d.parentNode&&(c&&n.contains(d.ownerDocument,d)&&aa(_(d,"script")),d.parentNode.removeChild(d));return a}n.extend({htmlPrefilter:function(a){return a.replace(ka,"<$1>")},clone:function(a,b,c){var d,e,f,g,h=a.cloneNode(!0),i=n.contains(a.ownerDocument,a);if(!(l.noCloneChecked||1!==a.nodeType&&11!==a.nodeType||n.isXMLDoc(a)))for(g=_(h),f=_(a),d=0,e=f.length;e>d;d++)ta(f[d],g[d]);if(b)if(c)for(f=f||_(a),g=g||_(h),d=0,e=f.length;e>d;d++)sa(f[d],g[d]);else sa(a,h);return g=_(h,"script"),g.length>0&&aa(g,!i&&_(a,"script")),h},cleanData:function(a){for(var b,c,d,e=n.event.special,f=0;void 0!==(c=a[f]);f++)if(L(c)){if(b=c[N.expando]){if(b.events)for(d in b.events)e[d]?n.event.remove(c,d):n.removeEvent(c,d,b.handle);c[N.expando]=void 0}c[O.expando]&&(c[O.expando]=void 0)}}}),n.fn.extend({domManip:ua,detach:function(a){return va(this,a,!0)},remove:function(a){return va(this,a)},text:function(a){return K(this,function(a){return void 0===a?n.text(this):this.empty().each(function(){1!==this.nodeType&&11!==this.nodeType&&9!==this.nodeType||(this.textContent=a)})},null,a,arguments.length)},append:function(){return ua(this,arguments,function(a){if(1===this.nodeType||11===this.nodeType||9===this.nodeType){var b=pa(this,a);b.appendChild(a)}})},prepend:function(){return ua(this,arguments,function(a){if(1===this.nodeType||11===this.nodeType||9===this.nodeType){var b=pa(this,a);b.insertBefore(a,b.firstChild)}})},before:function(){return ua(this,arguments,function(a){this.parentNode&&this.parentNode.insertBefore(a,this)})},after:function(){return ua(this,arguments,function(a){this.parentNode&&this.parentNode.insertBefore(a,this.nextSibling)})},empty:function(){for(var a,b=0;null!=(a=this[b]);b++)1===a.nodeType&&(n.cleanData(_(a,!1)),a.textContent="");return this},clone:function(a,b){return a=null==a?!1:a,b=null==b?a:b,this.map(function(){return n.clone(this,a,b)})},html:function(a){return K(this,function(a){var b=this[0]||{},c=0,d=this.length;if(void 0===a&&1===b.nodeType)return b.innerHTML;if("string"==typeof a&&!la.test(a)&&!$[(Y.exec(a)||["",""])[1].toLowerCase()]){a=n.htmlPrefilter(a);try{for(;d>c;c++)b=this[c]||{},1===b.nodeType&&(n.cleanData(_(b,!1)),b.innerHTML=a);b=0}catch(e){}}b&&this.empty().append(a)},null,a,arguments.length)},replaceWith:function(){var a=[];return ua(this,arguments,function(b){var c=this.parentNode;n.inArray(this,a)<0&&(n.cleanData(_(this)),c&&c.replaceChild(b,this))},a)}}),n.each({appendTo:"append",prependTo:"prepend",insertBefore:"before",insertAfter:"after",replaceAll:"replaceWith"},function(a,b){n.fn[a]=function(a){for(var c,d=[],e=n(a),f=e.length-1,h=0;f>=h;h++)c=h===f?this:this.clone(!0),n(e[h])[b](c),g.apply(d,c.get());return this.pushStack(d)}});var wa,xa={HTML:"block",BODY:"block"};function ya(a,b){var c=n(b.createElement(a)).appendTo(b.body),d=n.css(c[0],"display");return c.detach(),d}function za(a){var b=d,c=xa[a];return c||(c=ya(a,b),"none"!==c&&c||(wa=(wa||n("' + link = 'New Window'%url + return IPython.display.HTML(iframe+link) + +def nb_play_mocap(mocap, mf, meta=None, frame_time=1/30, scale=1, camera_z=500, base_url=None): + data_template = 'var dataBuffer = `$$DATA$$`;' + data_template += 'var metadata = $$META$$;' + data_template += 'start(dataBuffer, metadata, $$CZ$$, $$SCALE$$, $$FRAMETIME$$);' + dir_path = os.path.dirname(os.path.realpath(__file__)) + + + if base_url is None: + base_url = os.path.join(dir_path, 'mocapplayer/playBuffer.html') + + # print(dir_path) + + if mf == 'bvh': + pass + elif mf == 'pos': + cols = list(mocap.values.columns) + for c in cols: + if 'rotation' in c: + cols.remove(c) + + data_csv = mocap.values.to_csv(index=False, columns=cols) + + if meta is not None: + lines = [','.join(item) for item in meta.astype('str')] + meta_csv = '[' + ','.join('[%s]'%l for l in lines) +']' + else: + meta_csv = '[]' + + data_assigned = data_template.replace('$$DATA$$', data_csv) + data_assigned = data_assigned.replace('$$META$$', meta_csv) + data_assigned = data_assigned.replace('$$CZ$$', str(camera_z)) + data_assigned = data_assigned.replace('$$SCALE$$', str(scale)) + data_assigned = data_assigned.replace('$$FRAMETIME$$', str(frame_time)) + + else: + return + + + + with open(os.path.join(dir_path, 'mocapplayer/data.js'), 'w') as oFile: + oFile.write(data_assigned) + + url = '%s?&cz=200&order=xzyi&frame_time=%f&scale=%f'%(base_url, frame_time, scale) + iframe = '' + link = 'New Window'%url + return IPython.display.HTML(iframe+link) diff --git a/dataloaders/pymo/writers.py b/dataloaders/pymo/writers.py new file mode 100644 index 0000000000000000000000000000000000000000..834ef639bb3c86e7ca94a0c6de2fa868a48c3ff9 --- /dev/null +++ b/dataloaders/pymo/writers.py @@ -0,0 +1,55 @@ +import numpy as np +import pandas as pd + +class BVHWriter(): + def __init__(self): + pass + + def write(self, X, ofile): + + # Writing the skeleton info + ofile.write('HIERARCHY\n') + + self.motions_ = [] + self._printJoint(X, X.root_name, 0, ofile) + + # Writing the motion header + ofile.write('MOTION\n') + ofile.write('Frames: %d\n'%X.values.shape[0]) + ofile.write('Frame Time: %f\n'%X.framerate) + + # Writing the data + self.motions_ = np.asarray(self.motions_).T + lines = [" ".join(item) for item in self.motions_.astype(str)] + ofile.write("".join("%s\n"%l for l in lines)) + + def _printJoint(self, X, joint, tab, ofile): + + if X.skeleton[joint]['parent'] == None: + ofile.write('ROOT %s\n'%joint) + elif len(X.skeleton[joint]['children']) > 0: + ofile.write('%sJOINT %s\n'%('\t'*(tab), joint)) + else: + ofile.write('%sEnd site\n'%('\t'*(tab))) + + ofile.write('%s{\n'%('\t'*(tab))) + + ofile.write('%sOFFSET %3.5f %3.5f %3.5f\n'%('\t'*(tab+1), + X.skeleton[joint]['offsets'][0], + X.skeleton[joint]['offsets'][1], + X.skeleton[joint]['offsets'][2])) + channels = X.skeleton[joint]['channels'] + n_channels = len(channels) + + if n_channels > 0: + for ch in channels: + self.motions_.append(np.asarray(X.values['%s_%s'%(joint, ch)].values)) + + if len(X.skeleton[joint]['children']) > 0: + ch_str = ''.join(' %s'*n_channels%tuple(channels)) + ofile.write('%sCHANNELS %d%s\n' %('\t'*(tab+1), n_channels, ch_str)) + + for c in X.skeleton[joint]['children']: + self._printJoint(X, c, tab+1, ofile) + + ofile.write('%s}\n'%('\t'*(tab))) diff --git a/dataloaders/utils/audio_features.py b/dataloaders/utils/audio_features.py new file mode 100644 index 0000000000000000000000000000000000000000..596c40ee03cf21c6ec159e3c3542b3086ef38e73 --- /dev/null +++ b/dataloaders/utils/audio_features.py @@ -0,0 +1,209 @@ +"""modified from https://github.com/yesheng-THU/GFGE/blob/main/data_processing/audio_features.py""" +import numpy as np +import librosa +import math +import os +import scipy.io.wavfile as wav +import torch +import torch.nn as nn +import torch.nn.functional as F +import copy +from tqdm import tqdm +from transformers import Wav2Vec2Model, Wav2Vec2Config +from transformers.modeling_outputs import BaseModelOutput +from typing import Optional, Tuple +_CONFIG_FOR_DOC = "Wav2Vec2Config" + +# the implementation of Wav2Vec2Model is borrowed from https://huggingface.co/transformers/_modules/transformers/models/wav2vec2/modeling_wav2vec2.html#Wav2Vec2Model +# initialize our encoder with the pre-trained wav2vec 2.0 weights. +def _compute_mask_indices( + shape: Tuple[int, int], + mask_prob: float, + mask_length: int, + attention_mask: Optional[torch.Tensor] = None, + min_masks: int = 0, +) -> np.ndarray: + bsz, all_sz = shape + mask = np.full((bsz, all_sz), False) + + all_num_mask = int( + mask_prob * all_sz / float(mask_length) + + np.random.rand() + ) + all_num_mask = max(min_masks, all_num_mask) + mask_idcs = [] + padding_mask = attention_mask.ne(1) if attention_mask is not None else None + for i in range(bsz): + if padding_mask is not None: + sz = all_sz - padding_mask[i].long().sum().item() + num_mask = int( + mask_prob * sz / float(mask_length) + + np.random.rand() + ) + num_mask = max(min_masks, num_mask) + else: + sz = all_sz + num_mask = all_num_mask + + lengths = np.full(num_mask, mask_length) + + if sum(lengths) == 0: + lengths[0] = min(mask_length, sz - 1) + + min_len = min(lengths) + if sz - min_len <= num_mask: + min_len = sz - num_mask - 1 + + mask_idc = np.random.choice(sz - min_len, num_mask, replace=False) + mask_idc = np.asarray([mask_idc[j] + offset for j in range(len(mask_idc)) for offset in range(lengths[j])]) + mask_idcs.append(np.unique(mask_idc[mask_idc < sz])) + + min_len = min([len(m) for m in mask_idcs]) + for i, mask_idc in enumerate(mask_idcs): + if len(mask_idc) > min_len: + mask_idc = np.random.choice(mask_idc, min_len, replace=False) + mask[i, mask_idc] = True + return mask + +# linear interpolation layer +def linear_interpolation(features, input_fps, output_fps, output_len=None): + features = features.transpose(1, 2) + seq_len = features.shape[2] / float(input_fps) + if output_len is None: + output_len = int(seq_len * output_fps) + output_features = F.interpolate(features,size=output_len,align_corners=True,mode='linear') + return output_features.transpose(1, 2) + +class Wav2Vec2Model(Wav2Vec2Model): + def __init__(self, config): + super().__init__(config) + self.audio_fps = 15 #args.audio_fps + #input_values 16K hz, 49fps, 20ms overlap, 25ms recepion field + def forward( + self, + input_values, + dataset="beat", + attention_mask=None, + output_attentions=None, + output_hidden_states=None, + return_dict=None, + frame_num=None + ): + #print(input_values.shape) + self.config.output_attentions = True + output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions + output_hidden_states = ( + output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states + ) + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + + hidden_states = self.feature_extractor(input_values) + hidden_states = hidden_states.transpose(1, 2) + #print(hidden_states.shape) + if dataset == "beat": + hidden_states = linear_interpolation(hidden_states, 49, self.audio_fps, output_len=frame_num) + #print(hidden_states.shape) + if attention_mask is not None: + output_lengths = self._get_feat_extract_output_lengths(attention_mask.sum(-1)) + attention_mask = torch.zeros( + hidden_states.shape[:2], dtype=hidden_states.dtype, device=hidden_states.device + ) + attention_mask[ + (torch.arange(attention_mask.shape[0], device=hidden_states.device), output_lengths - 1) + ] = 1 + attention_mask = attention_mask.flip([-1]).cumsum(-1).flip([-1]).bool() + + hidden_states = self.feature_projection(hidden_states)[0] + #print(hidden_states.shape) + if self.config.apply_spec_augment and self.training: + batch_size, sequence_length, hidden_size = hidden_states.size() + if self.config.mask_time_prob > 0: + mask_time_indices = _compute_mask_indices( + (batch_size, sequence_length), + self.config.mask_time_prob, + self.config.mask_time_length, + attention_mask=attention_mask, + min_masks=2, + ) + hidden_states[torch.from_numpy(mask_time_indices)] = self.masked_spec_embed.to(hidden_states.dtype) + if self.config.mask_feature_prob > 0: + mask_feature_indices = _compute_mask_indices( + (batch_size, hidden_size), + self.config.mask_feature_prob, + self.config.mask_feature_length, + ) + mask_feature_indices = torch.from_numpy(mask_feature_indices).to(hidden_states.device) + hidden_states[mask_feature_indices[:, None].expand(-1, sequence_length, -1)] = 0 + encoder_outputs = self.encoder( + hidden_states, + attention_mask=attention_mask, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + ) + hidden_states = encoder_outputs[0] + #print(encoder_outputs.shape) + if not return_dict: + return (hidden_states,) + encoder_outputs[1:] + + return BaseModelOutput( + last_hidden_state=hidden_states, + hidden_states=encoder_outputs.hidden_states, + attentions=encoder_outputs.attentions, + ) + + +def extract_wav2vec2(file_folder, destpath, fps, inference_length=16000*20): + wav2vec_model = Wav2Vec2Model.from_pretrained("/home/ma-user/work/datasets/hub/transformer/wav2vec2-base-960h") + wav2vec_model.feature_extractor._freeze_parameters() + wav2vec_model = wav2vec_model.cuda() + wav2vec_model.eval() + audio_mean = np.load("/home/ma-user/work/datasets/beat_cache/beat_english_15_141/train/wave16k/npy_mean.npy") + audio_std = np.load("/home/ma-user/work/datasets/beat_cache/beat_english_15_141/train/wave16k/npy_std.npy") + if not os.path.exists(destpath): os.mkdir(destpath) + with torch.no_grad(): + for file_name in tqdm(os.listdir(file_folder)): + if "mean" in file_name or "std" in file_name or "pynb" in file_name: continue + audio_np = np.load(file_folder+file_name) + audio_np = (audio_np-audio_mean)/audio_std + audio_torch = torch.from_numpy(audio_np).cuda() + audio_torch = audio_torch.reshape(1, -1) + #print(audio_torch.shape, audio_np.shape) + + if audio_torch.shape[1] > inference_length: + num_div = audio_torch.shape[1] // inference_length + remain = audio_torch.shape[1] % inference_length + for i in range(num_div): + audio_feat = wav2vec_model(audio_torch[:, i*inference_length:(i+1)*inference_length]).last_hidden_state.cpu().numpy().reshape(-1, 768) + if i == 0: + audio_feat_all = audio_feat + else: + audio_feat_all = np.concatenate((audio_feat_all, audio_feat), 0) + if remain > 1600: #0.25s + audio_feat = wav2vec_model(audio_torch[:, num_div*inference_length:num_div*inference_length+remain]).last_hidden_state.cpu().numpy().reshape(-1, 768) + audio_feat_all = np.concatenate((audio_feat_all, audio_feat), 0) + else: + audio_feat_all = wav2vec_model(audio_torch).last_hidden_state.cpu().numpy().reshape(-1, 768) + #print(audio_feat_all.shape, audio_np.shape[0]/16000*15, torch.cuda.memory_cached() / 1E9) + np.save(destpath+file_name, audio_feat_all) + +def extract_melspec(file, destpath, fps, n_mels=128): + fs,X = wav.read(file) + X = X.astype(float)/math.pow(2,15) + target_sr = 48000 + X_48k = librosa.resample(X, orig_sr=fs, target_sr=target_sr, res_type="kaiser_best") + n_fft=int(target_sr*0.13) + hop_len=int(target_sr/fps) + C = librosa.feature.melspectrogram(y=X_48k, sr=target_sr, n_fft=n_fft, hop_length=hop_len, n_mels=n_mels, fmin=0.0, fmax=8000) + #C2 = librosa.feature.melspectrogram(y=X, sr=fs, n_fft=1024, hop_length=512) + #print(C.shape, C2.shape) + C = np.log(C) + np.save(destpath,np.transpose(C)) + + +if __name__ == "__main__": + #calculate mean and build cache for data. + target_fps = 15 + ori_data_path = f"/home/ma-user/work/datasets/beat_cache/beat_english_{target_fps}_141/" + for data_type in ["train", "val", "test"]: + extract_wav2vec2(ori_data_path+data_type+"/wave16k/", ori_data_path+data_type+f"/wav2vec2_{target_fps}/", target_fps) \ No newline at end of file diff --git a/dataloaders/utils/other_tools.py b/dataloaders/utils/other_tools.py new file mode 100644 index 0000000000000000000000000000000000000000..08ce836b36ef1148f7d0eabaed131ac10f287020 --- /dev/null +++ b/dataloaders/utils/other_tools.py @@ -0,0 +1,676 @@ +import os +import numpy as np +import random +import torch +import shutil +import csv +import pprint +import pandas as pd +from loguru import logger +from collections import OrderedDict +import matplotlib.pyplot as plt +import pickle +import time + +import numpy as np + +def adjust_array(x, k): + len_x = len(x) + len_k = len(k) + + # If x is shorter than k, pad with zeros + if len_x < len_k: + return np.pad(x, (0, len_k - len_x), 'constant') + + # If x is longer than k, truncate x + elif len_x > len_k: + return x[:len_k] + + # If both are of same length + else: + return x + +def onset_to_frame(onset_times, audio_length, fps): + # Calculate total number of frames for the given audio length + total_frames = int(audio_length * fps) + + # Create an array of zeros of shape (total_frames,) + frame_array = np.zeros(total_frames, dtype=np.int32) + + # For each onset time, calculate the frame number and set it to 1 + for onset in onset_times: + frame_num = int(onset * fps) + # Check if the frame number is within the array bounds + if 0 <= frame_num < total_frames: + frame_array[frame_num] = 1 + + return frame_array + +def smooth_animations(animation1, animation2, blend_frames): + """ + Smoothly transition between two animation clips using linear interpolation. + + Parameters: + - animation1: The first animation clip, a numpy array of shape [n, k]. + - animation2: The second animation clip, a numpy array of shape [n, k]. + - blend_frames: Number of frames over which to blend the two animations. + + Returns: + - A smoothly blended animation clip of shape [2n, k]. + """ + + # Ensure blend_frames doesn't exceed the length of either animation + blend_frames = min(blend_frames, len(animation1), len(animation2)) + + # Extract overlapping sections + overlap_a1 = animation1[-blend_frames:-blend_frames+1, :] + overlap_a2 = animation2[blend_frames-1:blend_frames, :] + + # Create blend weights for linear interpolation + alpha = np.linspace(0, 1, 2 * blend_frames).reshape(-1, 1) + + # Linearly interpolate between overlapping sections + blended_overlap = overlap_a1 * (1 - alpha) + overlap_a2 * alpha + + # Extend the animations to form the result with 2n frames + if blend_frames == len(animation1) and blend_frames == len(animation2): + result = blended_overlap + else: + before_blend = animation1[:-blend_frames] + after_blend = animation2[blend_frames:] + result = np.vstack((before_blend, blended_overlap, after_blend)) + return result + + +def interpolate_sequence(quaternions): + bs, n, j, _ = quaternions.shape + new_n = 2 * n + new_quaternions = torch.zeros((bs, new_n, j, 4), device=quaternions.device, dtype=quaternions.dtype) + + for i in range(n): + q1 = quaternions[:, i, :, :] + new_quaternions[:, 2*i, :, :] = q1 + + if i < n - 1: + q2 = quaternions[:, i + 1, :, :] + new_quaternions[:, 2*i + 1, :, :] = slerp(q1, q2, 0.5) + else: + # For the last point, duplicate the value + new_quaternions[:, 2*i + 1, :, :] = q1 + + return new_quaternions + +def quaternion_multiply(q1, q2): + w1, x1, y1, z1 = q1 + w2, x2, y2, z2 = q2 + w = w1 * w2 - x1 * x2 - y1 * y2 - z1 * z2 + x = w1 * x2 + x1 * w2 + y1 * z2 - z1 * y2 + y = w1 * y2 + y1 * w2 + z1 * x2 - x1 * z2 + z = w1 * z2 + z1 * w2 + x1 * y2 - y1 * x2 + return w, x, y, z + +def quaternion_conjugate(q): + w, x, y, z = q + return (w, -x, -y, -z) + +def slerp(q1, q2, t): + dot = torch.sum(q1 * q2, dim=-1, keepdim=True) + + flip = (dot < 0).float() + q2 = (1 - flip * 2) * q2 + dot = dot * (1 - flip * 2) + + DOT_THRESHOLD = 0.9995 + mask = (dot > DOT_THRESHOLD).float() + + theta_0 = torch.acos(dot) + theta = theta_0 * t + + q3 = q2 - q1 * dot + q3 = q3 / torch.norm(q3, dim=-1, keepdim=True) + + interpolated = (torch.cos(theta) * q1 + torch.sin(theta) * q3) + + return mask * (q1 + t * (q2 - q1)) + (1 - mask) * interpolated + +def estimate_linear_velocity(data_seq, dt): + ''' + Given some batched data sequences of T timesteps in the shape (B, T, ...), estimates + the velocity for the middle T-2 steps using a second order central difference scheme. + The first and last frames are with forward and backward first-order + differences, respectively + - h : step size + ''' + # first steps is forward diff (t+1 - t) / dt + init_vel = (data_seq[:, 1:2] - data_seq[:, :1]) / dt + # middle steps are second order (t+1 - t-1) / 2dt + middle_vel = (data_seq[:, 2:] - data_seq[:, 0:-2]) / (2 * dt) + # last step is backward diff (t - t-1) / dt + final_vel = (data_seq[:, -1:] - data_seq[:, -2:-1]) / dt + + vel_seq = torch.cat([init_vel, middle_vel, final_vel], dim=1) + return vel_seq + + +def estimate_angular_velocity(rot_seq, dt): + ''' + Given a batch of sequences of T rotation matrices, estimates angular velocity at T-2 steps. + Input sequence should be of shape (B, T, ..., 3, 3) + ''' + # see https://en.wikipedia.org/wiki/Angular_velocity#Calculation_from_the_orientation_matrix + dRdt = estimate_linear_velocity(rot_seq, dt) + R = rot_seq + RT = R.transpose(-1, -2) + # compute skew-symmetric angular velocity tensor + w_mat = torch.matmul(dRdt, RT) + # pull out angular velocity vector by averaging symmetric entries + w_x = (-w_mat[..., 1, 2] + w_mat[..., 2, 1]) / 2.0 + w_y = (w_mat[..., 0, 2] - w_mat[..., 2, 0]) / 2.0 + w_z = (-w_mat[..., 0, 1] + w_mat[..., 1, 0]) / 2.0 + w = torch.stack([w_x, w_y, w_z], axis=-1) + return w + +import matplotlib.image as mpimg +from io import BytesIO + +def image_from_bytes(image_bytes): + return mpimg.imread(BytesIO(image_bytes), format='PNG') + + + +def process_frame(i, vertices_all, vertices1_all, faces, output_dir, use_matplotlib, filenames, camera_params, camera_params1): + import matplotlib + matplotlib.use('Agg') + import matplotlib.pyplot as plt + import trimesh + import pyvirtualdisplay as Display + + vertices = vertices_all[i] + vertices1 = vertices1_all[i] + filename = f"{output_dir}frame_{i}.png" + filenames.append(filename) + if i%100 == 0: + print('processed', i, 'frames') + #time_s = time.time() + #print(vertices.shape) + if use_matplotlib: + fig = plt.figure(figsize=(20, 10)) + ax = fig.add_subplot(121, projection="3d") + fig.subplots_adjust(left=0, right=1, bottom=0, top=1) + #ax.view_init(elev=0, azim=90) + x = vertices[:, 0] + y = vertices[:, 1] + z = vertices[:, 2] + ax.scatter(x, y, z, s=0.5) + ax.set_xlim([-1.0, 1.0]) + ax.set_ylim([-0.5, 1.5])#heigth + ax.set_zlim([-0, 2])#depth + ax.set_box_aspect((1,1,1)) + else: + mesh = trimesh.Trimesh(vertices, faces) + scene = mesh.scene() + scene.camera.fov = camera_params['fov'] + scene.camera.resolution = camera_params['resolution'] + scene.camera.z_near = camera_params['z_near'] + scene.camera.z_far = camera_params['z_far'] + scene.graph[scene.camera.name] = camera_params['transform'] + fig, ax =plt.subplots(1,2, figsize=(16, 6)) + image = scene.save_image(resolution=[640, 480], visible=False) + im0 = ax[0].imshow(image_from_bytes(image)) + ax[0].axis('off') + + if use_matplotlib: + ax2 = fig.add_subplot(122, projection="3d") + ax2.set_box_aspect((1,1,1)) + fig.subplots_adjust(left=0, right=1, bottom=0, top=1) + x1 = vertices1[:, 0] + y1 = vertices1[:, 1] + z1 = vertices1[:, 2] + ax2.scatter(x1, y1, z1, s=0.5) + ax2.set_xlim([-1.0, 1.0]) + ax2.set_ylim([-0.5, 1.5])#heigth + ax2.set_zlim([-0, 2]) + plt.savefig(filename, bbox_inches='tight') + plt.close(fig) + else: + mesh1 = trimesh.Trimesh(vertices1, faces) + scene1 = mesh1.scene() + scene1.camera.fov = camera_params1['fov'] + scene1.camera.resolution = camera_params1['resolution'] + scene1.camera.z_near = camera_params1['z_near'] + scene1.camera.z_far = camera_params1['z_far'] + scene1.graph[scene1.camera.name] = camera_params1['transform'] + image1 = scene1.save_image(resolution=[640, 480], visible=False) + im1 = ax[1].imshow(image_from_bytes(image1)) + ax[1].axis('off') + plt.savefig(filename, bbox_inches='tight') + plt.close(fig) + +def generate_images(frames, vertices_all, vertices1_all, faces, output_dir, use_matplotlib, filenames): + import multiprocessing + import trimesh + num_cores = multiprocessing.cpu_count() # This will get the number of cores on your machine. + mesh = trimesh.Trimesh(vertices_all[0], faces) + scene = mesh.scene() + camera_params = { + 'fov': scene.camera.fov, + 'resolution': scene.camera.resolution, + 'focal': scene.camera.focal, + 'z_near': scene.camera.z_near, + "z_far": scene.camera.z_far, + 'transform': scene.graph[scene.camera.name][0] + } + mesh1 = trimesh.Trimesh(vertices1_all[0], faces) + scene1 = mesh1.scene() + camera_params1 = { + 'fov': scene1.camera.fov, + 'resolution': scene1.camera.resolution, + 'focal': scene1.camera.focal, + 'z_near': scene1.camera.z_near, + "z_far": scene1.camera.z_far, + 'transform': scene1.graph[scene1.camera.name][0] + } + # Use a Pool to manage the processes + # print(num_cores) + progress = multiprocessing.Value('i', 0) + lock = multiprocessing.Lock() + with multiprocessing.Pool(num_cores) as pool: + pool.starmap(process_frame, [(i, vertices_all, vertices1_all, faces, output_dir, use_matplotlib, filenames, camera_params, camera_params1) for i in range(frames)]) + +def render_one_sequence( + res_npz_path, + gt_npz_path, + output_dir, + audio_path, + model_folder="/data/datasets/smplx_models/", + model_type='smplx', + gender='NEUTRAL_2020', + ext='npz', + num_betas=300, + num_expression_coeffs=100, + use_face_contour=False, + use_matplotlib=False, + args=None): + import smplx + import matplotlib.pyplot as plt + import imageio + from tqdm import tqdm + import os + import numpy as np + import torch + import moviepy.editor as mp + import librosa + + model = smplx.create(model_folder, model_type=model_type, + gender=gender, use_face_contour=use_face_contour, + num_betas=num_betas, + num_expression_coeffs=num_expression_coeffs, + ext=ext, use_pca=False).cuda() + + #data_npz = np.load(f"{output_dir}{res_npz_path}.npz") + data_np_body = np.load(res_npz_path, allow_pickle=True) + gt_np_body = np.load(gt_npz_path, allow_pickle=True) + + if not os.path.exists(output_dir): os.makedirs(output_dir) + filenames = [] + if not use_matplotlib: + import trimesh + #import pyrender + from pyvirtualdisplay import Display + display = Display(visible=0, size=(640, 480)) + display.start() + faces = np.load(f"{model_folder}/smplx/SMPLX_NEUTRAL_2020.npz", allow_pickle=True)["f"] + seconds = 1 + #data_npz["jaw_pose"].shape[0] + n = data_np_body["poses"].shape[0] + beta = torch.from_numpy(data_np_body["betas"]).to(torch.float32).unsqueeze(0).cuda() + beta = beta.repeat(n, 1) + expression = torch.from_numpy(data_np_body["expressions"][:n]).to(torch.float32).cuda() + jaw_pose = torch.from_numpy(data_np_body["poses"][:n, 66:69]).to(torch.float32).cuda() + pose = torch.from_numpy(data_np_body["poses"][:n]).to(torch.float32).cuda() + transl = torch.from_numpy(data_np_body["trans"][:n]).to(torch.float32).cuda() + # print(beta.shape, expression.shape, jaw_pose.shape, pose.shape, transl.shape, pose[:,:3].shape) + output = model(betas=beta, transl=transl, expression=expression, jaw_pose=jaw_pose, + global_orient=pose[:,:3], body_pose=pose[:,3:21*3+3], left_hand_pose=pose[:,25*3:40*3], right_hand_pose=pose[:,40*3:55*3], + leye_pose=pose[:, 69:72], + reye_pose=pose[:, 72:75], + return_verts=True) + vertices_all = output["vertices"].cpu().detach().numpy() + + beta1 = torch.from_numpy(gt_np_body["betas"]).to(torch.float32).unsqueeze(0).cuda() + expression1 = torch.from_numpy(gt_np_body["expressions"][:n]).to(torch.float32).cuda() + jaw_pose1 = torch.from_numpy(gt_np_body["poses"][:n,66:69]).to(torch.float32).cuda() + pose1 = torch.from_numpy(gt_np_body["poses"][:n]).to(torch.float32).cuda() + transl1 = torch.from_numpy(gt_np_body["trans"][:n]).to(torch.float32).cuda() + output1 = model(betas=beta1, transl=transl1, expression=expression1, jaw_pose=jaw_pose1, global_orient=pose1[:,:3], body_pose=pose1[:,3:21*3+3], left_hand_pose=pose1[:,25*3:40*3], right_hand_pose=pose1[:,40*3:55*3], + leye_pose=pose1[:, 69:72], + reye_pose=pose1[:, 72:75],return_verts=True) + vertices1_all = output1["vertices"].cpu().detach().numpy() + if args.debug: + seconds = 1 + else: + seconds = vertices_all.shape[0]//30 + # camera_settings = None + time_s = time.time() + generate_images(int(seconds*30), vertices_all, vertices1_all, faces, output_dir, use_matplotlib, filenames) + filenames = [f"{output_dir}frame_{i}.png" for i in range(int(seconds*30))] + # print(time.time()-time_s) + # for i in tqdm(range(seconds*30)): + # vertices = vertices_all[i] + # vertices1 = vertices1_all[i] + # filename = f"{output_dir}frame_{i}.png" + # filenames.append(filename) + # #time_s = time.time() + # #print(vertices.shape) + # if use_matplotlib: + # fig = plt.figure(figsize=(20, 10)) + # ax = fig.add_subplot(121, projection="3d") + # fig.subplots_adjust(left=0, right=1, bottom=0, top=1) + # #ax.view_init(elev=0, azim=90) + # x = vertices[:, 0] + # y = vertices[:, 1] + # z = vertices[:, 2] + # ax.scatter(x, y, z, s=0.5) + # ax.set_xlim([-1.0, 1.0]) + # ax.set_ylim([-0.5, 1.5])#heigth + # ax.set_zlim([-0, 2])#depth + # ax.set_box_aspect((1,1,1)) + # else: + # mesh = trimesh.Trimesh(vertices, faces) + # if i == 0: + # scene = mesh.scene() + # camera_params = { + # 'fov': scene.camera.fov, + # 'resolution': scene.camera.resolution, + # 'focal': scene.camera.focal, + # 'z_near': scene.camera.z_near, + # "z_far": scene.camera.z_far, + # 'transform': scene.graph[scene.camera.name][0] + # } + # else: + # scene = mesh.scene() + # scene.camera.fov = camera_params['fov'] + # scene.camera.resolution = camera_params['resolution'] + # scene.camera.z_near = camera_params['z_near'] + # scene.camera.z_far = camera_params['z_far'] + # scene.graph[scene.camera.name] = camera_params['transform'] + # fig, ax =plt.subplots(1,2, figsize=(16, 6)) + # image = scene.save_image(resolution=[640, 480], visible=False) + # #print((time.time()-time_s)) + # im0 = ax[0].imshow(image_from_bytes(image)) + # ax[0].axis('off') + + # # beta1 = torch.from_numpy(gt_np_body["betas"]).to(torch.float32).unsqueeze(0) + # # expression1 = torch.from_numpy(gt_np_body["expressions"][i]).to(torch.float32).unsqueeze(0) + # # jaw_pose1 = torch.from_numpy(gt_np_body["poses"][i][66:69]).to(torch.float32).unsqueeze(0) + # # pose1 = torch.from_numpy(gt_np_body["poses"][i]).to(torch.float32).unsqueeze(0) + # # transl1 = torch.from_numpy(gt_np_body["trans"][i]).to(torch.float32).unsqueeze(0) + # # #print(beta.shape, expression.shape, jaw_pose.shape, pose.shape, transl.shape)global_orient=pose[0:1,:3], + # # output1 = model(betas=beta1, transl=transl1, expression=expression1, jaw_pose=jaw_pose1, global_orient=pose1[0:1,:3], body_pose=pose1[0:1,3:21*3+3], left_hand_pose=pose1[0:1,25*3:40*3], right_hand_pose=pose1[0:1,40*3:55*3], return_verts=True) + # # vertices1 = output1["vertices"].cpu().detach().numpy()[0] + + # if use_matplotlib: + # ax2 = fig.add_subplot(122, projection="3d") + # ax2.set_box_aspect((1,1,1)) + # fig.subplots_adjust(left=0, right=1, bottom=0, top=1) + # #ax2.view_init(elev=0, azim=90) + # x1 = vertices1[:, 0] + # y1 = vertices1[:, 1] + # z1 = vertices1[:, 2] + # ax2.scatter(x1, y1, z1, s=0.5) + # ax2.set_xlim([-1.0, 1.0]) + # ax2.set_ylim([-0.5, 1.5])#heigth + # ax2.set_zlim([-0, 2]) + # plt.savefig(filename, bbox_inches='tight') + # plt.close(fig) + # else: + # mesh1 = trimesh.Trimesh(vertices1, faces) + # if i == 0: + # scene1 = mesh1.scene() + # camera_params1 = { + # 'fov': scene1.camera.fov, + # 'resolution': scene1.camera.resolution, + # 'focal': scene1.camera.focal, + # 'z_near': scene1.camera.z_near, + # "z_far": scene1.camera.z_far, + # 'transform': scene1.graph[scene1.camera.name][0] + # } + # else: + # scene1 = mesh1.scene() + # scene1.camera.fov = camera_params1['fov'] + # scene1.camera.resolution = camera_params1['resolution'] + # scene1.camera.z_near = camera_params1['z_near'] + # scene1.camera.z_far = camera_params1['z_far'] + # scene1.graph[scene1.camera.name] = camera_params1['transform'] + # image1 = scene1.save_image(resolution=[640, 480], visible=False) + # im1 = ax[1].imshow(image_from_bytes(image1)) + # ax[1].axis('off') + # plt.savefig(filename, bbox_inches='tight') + # plt.close(fig) + + # display.stop() + # print(filenames) + images = [imageio.imread(filename) for filename in filenames] + imageio.mimsave(f"{output_dir}raw_{res_npz_path.split('/')[-1][:-4]}.mp4", images, fps=30) + for filename in filenames: + os.remove(filename) + + video = mp.VideoFileClip(f"{output_dir}raw_{res_npz_path.split('/')[-1][:-4]}.mp4") + # audio, sr = librosa.load(audio_path) + # audio = audio[:seconds*sr] + # print(audio.shape, seconds, sr) + # import soundfile as sf + # sf.write(f"{output_dir}{res_npz_path.split('/')[-1][:-4]}.wav", audio, 16000, 'PCM_24') + # audio_tmp = librosa.output.write_wav(f"{output_dir}{res_npz_path.split('/')[-1][:-4]}.wav", audio, sr=16000) + audio = mp.AudioFileClip(audio_path) + if audio.duration > video.duration: + audio = audio.subclip(0, video.duration) + final_clip = video.set_audio(audio) + final_clip.write_videofile(f"{output_dir}{res_npz_path.split('/')[-1][4:-4]}.mp4") + os.remove(f"{output_dir}raw_{res_npz_path.split('/')[-1][:-4]}.mp4") + +def print_exp_info(args): + logger.info(pprint.pformat(vars(args))) + logger.info(f"# ------------ {args.name} ----------- #") + logger.info("PyTorch version: {}".format(torch.__version__)) + logger.info("CUDA version: {}".format(torch.version.cuda)) + logger.info("{} GPUs".format(torch.cuda.device_count())) + logger.info(f"Random Seed: {args.random_seed}") + +def args2csv(args, get_head=False, list4print=[]): + for k, v in args.items(): + if isinstance(args[k], dict): + args2csv(args[k], get_head, list4print) + else: list4print.append(k) if get_head else list4print.append(v) + return list4print + +class EpochTracker: + def __init__(self, metric_names, metric_directions): + assert len(metric_names) == len(metric_directions), "Metric names and directions should have the same length" + + + self.metric_names = metric_names + self.states = ['train', 'val', 'test'] + self.types = ['last', 'best'] + + + self.values = {name: {state: {type_: {'value': np.inf if not is_higher_better else -np.inf, 'epoch': 0} + for type_ in self.types} + for state in self.states} + for name, is_higher_better in zip(metric_names, metric_directions)} + + self.loss_meters = {name: {state: AverageMeter(f"{name}_{state}") + for state in self.states} + for name in metric_names} + + + self.is_higher_better = {name: direction for name, direction in zip(metric_names, metric_directions)} + self.train_history = {name: [] for name in metric_names} + self.val_history = {name: [] for name in metric_names} + + + def update_meter(self, name, state, value): + self.loss_meters[name][state].update(value) + + + def update_values(self, name, state, epoch): + value_avg = self.loss_meters[name][state].avg + new_best = False + + + if ((value_avg < self.values[name][state]['best']['value'] and not self.is_higher_better[name]) or + (value_avg > self.values[name][state]['best']['value'] and self.is_higher_better[name])): + self.values[name][state]['best']['value'] = value_avg + self.values[name][state]['best']['epoch'] = epoch + new_best = True + self.values[name][state]['last']['value'] = value_avg + self.values[name][state]['last']['epoch'] = epoch + return new_best + + + def get(self, name, state, type_): + return self.values[name][state][type_] + + + def reset(self): + for name in self.metric_names: + for state in self.states: + self.loss_meters[name][state].reset() + + + def flatten_values(self): + flat_dict = {} + for name in self.metric_names: + for state in self.states: + for type_ in self.types: + value_key = f"{name}_{state}_{type_}" + epoch_key = f"{name}_{state}_{type_}_epoch" + flat_dict[value_key] = self.values[name][state][type_]['value'] + flat_dict[epoch_key] = self.values[name][state][type_]['epoch'] + return flat_dict + + def update_and_plot(self, name, epoch, save_path): + new_best_train = self.update_values(name, 'train', epoch) + new_best_val = self.update_values(name, 'val', epoch) + + + self.train_history[name].append(self.loss_meters[name]['train'].avg) + self.val_history[name].append(self.loss_meters[name]['val'].avg) + + + train_values = self.train_history[name] + val_values = self.val_history[name] + epochs = list(range(1, len(train_values) + 1)) + + + plt.figure(figsize=(10, 6)) + plt.plot(epochs, train_values, label='Train') + plt.plot(epochs, val_values, label='Val') + plt.title(f'Train vs Val {name} over epochs') + plt.xlabel('Epochs') + plt.ylabel(name) + plt.legend() + plt.savefig(save_path) + plt.close() + + + return new_best_train, new_best_val + + + + +def record_trial(args, tracker): + """ + 1. record notes, score, env_name, experments_path, + """ + csv_path = args.out_path + "custom/" +args.csv_name+".csv" + all_print_dict = vars(args) + all_print_dict.update(tracker.flatten_values()) + if not os.path.exists(csv_path): + pd.DataFrame([all_print_dict]).to_csv(csv_path, index=False) + else: + df_existing = pd.read_csv(csv_path) + df_new = pd.DataFrame([all_print_dict]) + df_aligned = df_existing.append(df_new).fillna("") + df_aligned.to_csv(csv_path, index=False) + + +def set_random_seed(args): + os.environ['PYTHONHASHSEED'] = str(args.random_seed) + random.seed(args.random_seed) + np.random.seed(args.random_seed) + torch.manual_seed(args.random_seed) + torch.cuda.manual_seed_all(args.random_seed) + torch.cuda.manual_seed(args.random_seed) + torch.backends.cudnn.deterministic = args.deterministic #args.CUDNN_DETERMINISTIC + torch.backends.cudnn.benchmark = args.benchmark + torch.backends.cudnn.enabled = args.cudnn_enabled + + +def save_checkpoints(save_path, model, opt=None, epoch=None, lrs=None): + if lrs is not None: + states = { 'model_state': model.state_dict(), + 'epoch': epoch + 1, + 'opt_state': opt.state_dict(), + 'lrs':lrs.state_dict(),} + elif opt is not None: + states = { 'model_state': model.state_dict(), + 'epoch': epoch + 1, + 'opt_state': opt.state_dict(),} + else: + states = { 'model_state': model.state_dict(),} + torch.save(states, save_path) + + +def load_checkpoints(model, save_path, load_name='model'): + states = torch.load(save_path) + new_weights = OrderedDict() + flag=False + for k, v in states['model_state'].items(): + #print(k) + if "module" not in k: + break + else: + new_weights[k[7:]]=v + flag=True + if flag: + try: + model.load_state_dict(new_weights) + except: + #print(states['model_state']) + model.load_state_dict(states['model_state']) + else: + model.load_state_dict(states['model_state']) + logger.info(f"load self-pretrained checkpoints for {load_name}") + + +def model_complexity(model, args): + from ptflops import get_model_complexity_info + flops, params = get_model_complexity_info(model, (args.T_GLOBAL._DIM, args.TRAIN.CROP, args.TRAIN), + as_strings=False, print_per_layer_stat=False) + logging.info('{:<30} {:<8} BFlops'.format('Computational complexity: ', flops / 1e9)) + logging.info('{:<30} {:<8} MParams'.format('Number of parameters: ', params / 1e6)) + + +class AverageMeter(object): + """Computes and stores the average and current value""" + def __init__(self, name, fmt=':f'): + self.name = name + self.fmt = fmt + self.reset() + + def reset(self): + self.val = 0 + self.avg = 0 + self.sum = 0 + self.count = 0 + + def update(self, val, n=1): + self.val = val + self.sum += val * n + self.count += n + self.avg = self.sum / self.count + + def __str__(self): + fmtstr = '{name} {val' + self.fmt + '} ({avg' + self.fmt + '})' + return fmtstr.format(**self.__dict__) \ No newline at end of file diff --git a/dataloaders/utils/rotation_conversions.py b/dataloaders/utils/rotation_conversions.py new file mode 100644 index 0000000000000000000000000000000000000000..f2bfaa1b2247622bff35d3f9b15e8eb84064aa53 --- /dev/null +++ b/dataloaders/utils/rotation_conversions.py @@ -0,0 +1,550 @@ +# Copyright (c) Facebook, Inc. and its affiliates. All rights reserved. + +import functools +from typing import Optional + +import torch +import torch.nn.functional as F + + +""" +The transformation matrices returned from the functions in this file assume +the points on which the transformation will be applied are column vectors. +i.e. the R matrix is structured as + + R = [ + [Rxx, Rxy, Rxz], + [Ryx, Ryy, Ryz], + [Rzx, Rzy, Rzz], + ] # (3, 3) + +This matrix can be applied to column vectors by post multiplication +by the points e.g. + + points = [[0], [1], [2]] # (3 x 1) xyz coordinates of a point + transformed_points = R * points + +To apply the same matrix to points which are row vectors, the R matrix +can be transposed and pre multiplied by the points: + +e.g. + points = [[0, 1, 2]] # (1 x 3) xyz coordinates of a point + transformed_points = points * R.transpose(1, 0) +""" + + +def quaternion_to_matrix(quaternions): + """ + Convert rotations given as quaternions to rotation matrices. + + Args: + quaternions: quaternions with real part first, + as tensor of shape (..., 4). + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + r, i, j, k = torch.unbind(quaternions, -1) + two_s = 2.0 / (quaternions * quaternions).sum(-1) + + o = torch.stack( + ( + 1 - two_s * (j * j + k * k), + two_s * (i * j - k * r), + two_s * (i * k + j * r), + two_s * (i * j + k * r), + 1 - two_s * (i * i + k * k), + two_s * (j * k - i * r), + two_s * (i * k - j * r), + two_s * (j * k + i * r), + 1 - two_s * (i * i + j * j), + ), + -1, + ) + return o.reshape(quaternions.shape[:-1] + (3, 3)) + + +def _copysign(a, b): + """ + Return a tensor where each element has the absolute value taken from the, + corresponding element of a, with sign taken from the corresponding + element of b. This is like the standard copysign floating-point operation, + but is not careful about negative 0 and NaN. + + Args: + a: source tensor. + b: tensor whose signs will be used, of the same shape as a. + + Returns: + Tensor of the same shape as a with the signs of b. + """ + signs_differ = (a < 0) != (b < 0) + return torch.where(signs_differ, -a, a) + + +def _sqrt_positive_part(x): + """ + Returns torch.sqrt(torch.max(0, x)) + but with a zero subgradient where x is 0. + """ + ret = torch.zeros_like(x) + positive_mask = x > 0 + ret[positive_mask] = torch.sqrt(x[positive_mask]) + return ret + + +def matrix_to_quaternion(matrix): + """ + Convert rotations given as rotation matrices to quaternions. + + Args: + matrix: Rotation matrices as tensor of shape (..., 3, 3). + + Returns: + quaternions with real part first, as tensor of shape (..., 4). + """ + if matrix.size(-1) != 3 or matrix.size(-2) != 3: + raise ValueError(f"Invalid rotation matrix shape f{matrix.shape}.") + m00 = matrix[..., 0, 0] + m11 = matrix[..., 1, 1] + m22 = matrix[..., 2, 2] + o0 = 0.5 * _sqrt_positive_part(1 + m00 + m11 + m22) + x = 0.5 * _sqrt_positive_part(1 + m00 - m11 - m22) + y = 0.5 * _sqrt_positive_part(1 - m00 + m11 - m22) + z = 0.5 * _sqrt_positive_part(1 - m00 - m11 + m22) + o1 = _copysign(x, matrix[..., 2, 1] - matrix[..., 1, 2]) + o2 = _copysign(y, matrix[..., 0, 2] - matrix[..., 2, 0]) + o3 = _copysign(z, matrix[..., 1, 0] - matrix[..., 0, 1]) + return torch.stack((o0, o1, o2, o3), -1) + + +def _axis_angle_rotation(axis: str, angle): + """ + Return the rotation matrices for one of the rotations about an axis + of which Euler angles describe, for each value of the angle given. + + Args: + axis: Axis label "X" or "Y or "Z". + angle: any shape tensor of Euler angles in radians + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + + cos = torch.cos(angle) + sin = torch.sin(angle) + one = torch.ones_like(angle) + zero = torch.zeros_like(angle) + + if axis == "X": + R_flat = (one, zero, zero, zero, cos, -sin, zero, sin, cos) + if axis == "Y": + R_flat = (cos, zero, sin, zero, one, zero, -sin, zero, cos) + if axis == "Z": + R_flat = (cos, -sin, zero, sin, cos, zero, zero, zero, one) + + return torch.stack(R_flat, -1).reshape(angle.shape + (3, 3)) + + +def euler_angles_to_matrix(euler_angles, convention: str): + """ + Convert rotations given as Euler angles in radians to rotation matrices. + + Args: + euler_angles: Euler angles in radians as tensor of shape (..., 3). + convention: Convention string of three uppercase letters from + {"X", "Y", and "Z"}. + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + if euler_angles.dim() == 0 or euler_angles.shape[-1] != 3: + raise ValueError("Invalid input euler angles.") + if len(convention) != 3: + raise ValueError("Convention must have 3 letters.") + if convention[1] in (convention[0], convention[2]): + raise ValueError(f"Invalid convention {convention}.") + for letter in convention: + if letter not in ("X", "Y", "Z"): + raise ValueError(f"Invalid letter {letter} in convention string.") + matrices = map(_axis_angle_rotation, convention, torch.unbind(euler_angles, -1)) + return functools.reduce(torch.matmul, matrices) + + +def _angle_from_tan( + axis: str, other_axis: str, data, horizontal: bool, tait_bryan: bool +): + """ + Extract the first or third Euler angle from the two members of + the matrix which are positive constant times its sine and cosine. + + Args: + axis: Axis label "X" or "Y or "Z" for the angle we are finding. + other_axis: Axis label "X" or "Y or "Z" for the middle axis in the + convention. + data: Rotation matrices as tensor of shape (..., 3, 3). + horizontal: Whether we are looking for the angle for the third axis, + which means the relevant entries are in the same row of the + rotation matrix. If not, they are in the same column. + tait_bryan: Whether the first and third axes in the convention differ. + + Returns: + Euler Angles in radians for each matrix in data as a tensor + of shape (...). + """ + + i1, i2 = {"X": (2, 1), "Y": (0, 2), "Z": (1, 0)}[axis] + if horizontal: + i2, i1 = i1, i2 + even = (axis + other_axis) in ["XY", "YZ", "ZX"] + if horizontal == even: + return torch.atan2(data[..., i1], data[..., i2]) + if tait_bryan: + return torch.atan2(-data[..., i2], data[..., i1]) + return torch.atan2(data[..., i2], -data[..., i1]) + + +def _index_from_letter(letter: str): + if letter == "X": + return 0 + if letter == "Y": + return 1 + if letter == "Z": + return 2 + + +def matrix_to_euler_angles(matrix, convention: str): + """ + Convert rotations given as rotation matrices to Euler angles in radians. + + Args: + matrix: Rotation matrices as tensor of shape (..., 3, 3). + convention: Convention string of three uppercase letters. + + Returns: + Euler angles in radians as tensor of shape (..., 3). + """ + if len(convention) != 3: + raise ValueError("Convention must have 3 letters.") + if convention[1] in (convention[0], convention[2]): + raise ValueError(f"Invalid convention {convention}.") + for letter in convention: + if letter not in ("X", "Y", "Z"): + raise ValueError(f"Invalid letter {letter} in convention string.") + if matrix.size(-1) != 3 or matrix.size(-2) != 3: + raise ValueError(f"Invalid rotation matrix shape f{matrix.shape}.") + i0 = _index_from_letter(convention[0]) + i2 = _index_from_letter(convention[2]) + tait_bryan = i0 != i2 + if tait_bryan: + central_angle = torch.asin( + matrix[..., i0, i2] * (-1.0 if i0 - i2 in [-1, 2] else 1.0) + ) + else: + central_angle = torch.acos(matrix[..., i0, i0]) + + o = ( + _angle_from_tan( + convention[0], convention[1], matrix[..., i2], False, tait_bryan + ), + central_angle, + _angle_from_tan( + convention[2], convention[1], matrix[..., i0, :], True, tait_bryan + ), + ) + return torch.stack(o, -1) + + +def random_quaternions( + n: int, dtype: Optional[torch.dtype] = None, device=None, requires_grad=False +): + """ + Generate random quaternions representing rotations, + i.e. versors with nonnegative real part. + + Args: + n: Number of quaternions in a batch to return. + dtype: Type to return. + device: Desired device of returned tensor. Default: + uses the current device for the default tensor type. + requires_grad: Whether the resulting tensor should have the gradient + flag set. + + Returns: + Quaternions as tensor of shape (N, 4). + """ + o = torch.randn((n, 4), dtype=dtype, device=device, requires_grad=requires_grad) + s = (o * o).sum(1) + o = o / _copysign(torch.sqrt(s), o[:, 0])[:, None] + return o + + +def random_rotations( + n: int, dtype: Optional[torch.dtype] = None, device=None, requires_grad=False +): + """ + Generate random rotations as 3x3 rotation matrices. + + Args: + n: Number of rotation matrices in a batch to return. + dtype: Type to return. + device: Device of returned tensor. Default: if None, + uses the current device for the default tensor type. + requires_grad: Whether the resulting tensor should have the gradient + flag set. + + Returns: + Rotation matrices as tensor of shape (n, 3, 3). + """ + quaternions = random_quaternions( + n, dtype=dtype, device=device, requires_grad=requires_grad + ) + return quaternion_to_matrix(quaternions) + + +def random_rotation( + dtype: Optional[torch.dtype] = None, device=None, requires_grad=False +): + """ + Generate a single random 3x3 rotation matrix. + + Args: + dtype: Type to return + device: Device of returned tensor. Default: if None, + uses the current device for the default tensor type + requires_grad: Whether the resulting tensor should have the gradient + flag set + + Returns: + Rotation matrix as tensor of shape (3, 3). + """ + return random_rotations(1, dtype, device, requires_grad)[0] + + +def standardize_quaternion(quaternions): + """ + Convert a unit quaternion to a standard form: one in which the real + part is non negative. + + Args: + quaternions: Quaternions with real part first, + as tensor of shape (..., 4). + + Returns: + Standardized quaternions as tensor of shape (..., 4). + """ + return torch.where(quaternions[..., 0:1] < 0, -quaternions, quaternions) + + +def quaternion_raw_multiply(a, b): + """ + Multiply two quaternions. + Usual torch rules for broadcasting apply. + + Args: + a: Quaternions as tensor of shape (..., 4), real part first. + b: Quaternions as tensor of shape (..., 4), real part first. + + Returns: + The product of a and b, a tensor of quaternions shape (..., 4). + """ + aw, ax, ay, az = torch.unbind(a, -1) + bw, bx, by, bz = torch.unbind(b, -1) + ow = aw * bw - ax * bx - ay * by - az * bz + ox = aw * bx + ax * bw + ay * bz - az * by + oy = aw * by - ax * bz + ay * bw + az * bx + oz = aw * bz + ax * by - ay * bx + az * bw + return torch.stack((ow, ox, oy, oz), -1) + + +def quaternion_multiply(a, b): + """ + Multiply two quaternions representing rotations, returning the quaternion + representing their composition, i.e. the versor with nonnegative real part. + Usual torch rules for broadcasting apply. + + Args: + a: Quaternions as tensor of shape (..., 4), real part first. + b: Quaternions as tensor of shape (..., 4), real part first. + + Returns: + The product of a and b, a tensor of quaternions of shape (..., 4). + """ + ab = quaternion_raw_multiply(a, b) + return standardize_quaternion(ab) + + +def quaternion_invert(quaternion): + """ + Given a quaternion representing rotation, get the quaternion representing + its inverse. + + Args: + quaternion: Quaternions as tensor of shape (..., 4), with real part + first, which must be versors (unit quaternions). + + Returns: + The inverse, a tensor of quaternions of shape (..., 4). + """ + + return quaternion * quaternion.new_tensor([1, -1, -1, -1]) + + +def quaternion_apply(quaternion, point): + """ + Apply the rotation given by a quaternion to a 3D point. + Usual torch rules for broadcasting apply. + + Args: + quaternion: Tensor of quaternions, real part first, of shape (..., 4). + point: Tensor of 3D points of shape (..., 3). + + Returns: + Tensor of rotated points of shape (..., 3). + """ + if point.size(-1) != 3: + raise ValueError(f"Points are not in 3D, f{point.shape}.") + real_parts = point.new_zeros(point.shape[:-1] + (1,)) + point_as_quaternion = torch.cat((real_parts, point), -1) + out = quaternion_raw_multiply( + quaternion_raw_multiply(quaternion, point_as_quaternion), + quaternion_invert(quaternion), + ) + return out[..., 1:] + + +def axis_angle_to_matrix(axis_angle): + """ + Convert rotations given as axis/angle to rotation matrices. + + Args: + axis_angle: Rotations given as a vector in axis angle form, + as a tensor of shape (..., 3), where the magnitude is + the angle turned anticlockwise in radians around the + vector's direction. + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + return quaternion_to_matrix(axis_angle_to_quaternion(axis_angle)) + + +def matrix_to_axis_angle(matrix): + """ + Convert rotations given as rotation matrices to axis/angle. + + Args: + matrix: Rotation matrices as tensor of shape (..., 3, 3). + + Returns: + Rotations given as a vector in axis angle form, as a tensor + of shape (..., 3), where the magnitude is the angle + turned anticlockwise in radians around the vector's + direction. + """ + return quaternion_to_axis_angle(matrix_to_quaternion(matrix)) + + +def axis_angle_to_quaternion(axis_angle): + """ + Convert rotations given as axis/angle to quaternions. + + Args: + axis_angle: Rotations given as a vector in axis angle form, + as a tensor of shape (..., 3), where the magnitude is + the angle turned anticlockwise in radians around the + vector's direction. + + Returns: + quaternions with real part first, as tensor of shape (..., 4). + """ + angles = torch.norm(axis_angle, p=2, dim=-1, keepdim=True) + half_angles = 0.5 * angles + eps = 1e-6 + small_angles = angles.abs() < eps + sin_half_angles_over_angles = torch.empty_like(angles) + sin_half_angles_over_angles[~small_angles] = ( + torch.sin(half_angles[~small_angles]) / angles[~small_angles] + ) + # for x small, sin(x/2) is about x/2 - (x/2)^3/6 + # so sin(x/2)/x is about 1/2 - (x*x)/48 + sin_half_angles_over_angles[small_angles] = ( + 0.5 - (angles[small_angles] * angles[small_angles]) / 48 + ) + quaternions = torch.cat( + [torch.cos(half_angles), axis_angle * sin_half_angles_over_angles], dim=-1 + ) + return quaternions + + +def quaternion_to_axis_angle(quaternions): + """ + Convert rotations given as quaternions to axis/angle. + + Args: + quaternions: quaternions with real part first, + as tensor of shape (..., 4). + + Returns: + Rotations given as a vector in axis angle form, as a tensor + of shape (..., 3), where the magnitude is the angle + turned anticlockwise in radians around the vector's + direction. + """ + norms = torch.norm(quaternions[..., 1:], p=2, dim=-1, keepdim=True) + half_angles = torch.atan2(norms, quaternions[..., :1]) + angles = 2 * half_angles + eps = 1e-6 + small_angles = angles.abs() < eps + sin_half_angles_over_angles = torch.empty_like(angles) + sin_half_angles_over_angles[~small_angles] = ( + torch.sin(half_angles[~small_angles]) / angles[~small_angles] + ) + # for x small, sin(x/2) is about x/2 - (x/2)^3/6 + # so sin(x/2)/x is about 1/2 - (x*x)/48 + sin_half_angles_over_angles[small_angles] = ( + 0.5 - (angles[small_angles] * angles[small_angles]) / 48 + ) + return quaternions[..., 1:] / sin_half_angles_over_angles + + +def rotation_6d_to_matrix(d6: torch.Tensor) -> torch.Tensor: + """ + Converts 6D rotation representation by Zhou et al. [1] to rotation matrix + using Gram--Schmidt orthogonalisation per Section B of [1]. + Args: + d6: 6D rotation representation, of size (*, 6) + + Returns: + batch of rotation matrices of size (*, 3, 3) + + [1] Zhou, Y., Barnes, C., Lu, J., Yang, J., & Li, H. + On the Continuity of Rotation Representations in Neural Networks. + IEEE Conference on Computer Vision and Pattern Recognition, 2019. + Retrieved from http://arxiv.org/abs/1812.07035 + """ + + a1, a2 = d6[..., :3], d6[..., 3:] + b1 = F.normalize(a1, dim=-1) + b2 = a2 - (b1 * a2).sum(-1, keepdim=True) * b1 + b2 = F.normalize(b2, dim=-1) + b3 = torch.cross(b1, b2, dim=-1) + return torch.stack((b1, b2, b3), dim=-2) + + +def matrix_to_rotation_6d(matrix: torch.Tensor) -> torch.Tensor: + """ + Converts rotation matrices to 6D rotation representation by Zhou et al. [1] + by dropping the last row. Note that 6D representation is not unique. + Args: + matrix: batch of rotation matrices of size (*, 3, 3) + + Returns: + 6D rotation representation, of size (*, 6) + + [1] Zhou, Y., Barnes, C., Lu, J., Yang, J., & Li, H. + On the Continuity of Rotation Representations in Neural Networks. + IEEE Conference on Computer Vision and Pattern Recognition, 2019. + Retrieved from http://arxiv.org/abs/1812.07035 + """ + return matrix[..., :2, :].clone().reshape(*matrix.size()[:-2], 6) diff --git a/datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/train_test_split.csv b/datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/train_test_split.csv new file mode 100644 index 0000000000000000000000000000000000000000..24ef34b4543baf0ca8d5f65644c1c3010f24e554 --- /dev/null +++ b/datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/train_test_split.csv @@ -0,0 +1,1621 @@ +id,type +10_kieks_0_103_103,test +10_kieks_0_104_104,train +10_kieks_0_10_10,val +10_kieks_0_111_111,test +10_kieks_0_112_112,train +10_kieks_0_11_11,val +10_kieks_0_12_12,val +10_kieks_0_13_13,val +10_kieks_0_14_14,val +10_kieks_0_15_15,val +10_kieks_0_16_16,val +10_kieks_0_1_1,test +10_kieks_0_2_2,test +10_kieks_0_3_3,test +10_kieks_0_4_4,test +10_kieks_0_5_5,test +10_kieks_0_65_65,test +10_kieks_0_66_66,train +10_kieks_0_6_6,test +10_kieks_0_73_73,test +10_kieks_0_74_74,train +10_kieks_0_7_7,test +10_kieks_0_81_81,test +10_kieks_0_82_82,train +10_kieks_0_87_87,test +10_kieks_0_88_88,train +10_kieks_0_8_8,test +10_kieks_0_95_95,test +10_kieks_0_96_96,train +10_kieks_0_9_9,val +10_kieks_1_1_1,additional +10_kieks_1_2_2,additional +10_kieks_1_3_3,additional +11_nidal_0_100_100,train +11_nidal_0_101_101,train +11_nidal_0_102_102,train +11_nidal_0_103_103,test +11_nidal_0_104_104,train +11_nidal_0_105_105,train +11_nidal_0_106_106,train +11_nidal_0_107_107,train +11_nidal_0_108_108,train +11_nidal_0_109_109,train +11_nidal_0_10_10,val +11_nidal_0_110_110,train +11_nidal_0_111_111,test +11_nidal_0_112_112,train +11_nidal_0_113_113,train +11_nidal_0_114_114,train +11_nidal_0_115_115,train +11_nidal_0_116_116,train +11_nidal_0_117_117,train +11_nidal_0_118_118,train +11_nidal_0_11_11,val +11_nidal_0_12_12,val +11_nidal_0_13_13,val +11_nidal_0_14_14,val +11_nidal_0_15_15,val +11_nidal_0_16_16,val +11_nidal_0_17_17,train +11_nidal_0_18_18,train +11_nidal_0_19_19,train +11_nidal_0_1_1,test +11_nidal_0_20_20,train +11_nidal_0_21_21,train +11_nidal_0_22_22,train +11_nidal_0_23_23,train +11_nidal_0_24_24,train +11_nidal_0_25_25,train +11_nidal_0_26_26,train +11_nidal_0_27_27,train +11_nidal_0_28_28,train +11_nidal_0_29_29,train +11_nidal_0_2_2,test +11_nidal_0_30_30,train +11_nidal_0_31_31,train +11_nidal_0_32_32,train +11_nidal_0_33_33,train +11_nidal_0_34_34,train +11_nidal_0_35_35,train +11_nidal_0_36_36,train +11_nidal_0_37_37,train +11_nidal_0_38_38,train +11_nidal_0_39_39,train +11_nidal_0_3_3,test +11_nidal_0_40_40,train +11_nidal_0_41_41,train +11_nidal_0_42_42,train +11_nidal_0_43_43,train +11_nidal_0_44_44,train +11_nidal_0_45_45,train +11_nidal_0_46_46,train +11_nidal_0_47_47,train +11_nidal_0_48_48,train +11_nidal_0_49_49,train +11_nidal_0_4_4,test +11_nidal_0_50_50,train +11_nidal_0_51_51,train +11_nidal_0_52_52,train +11_nidal_0_53_53,train +11_nidal_0_54_54,train +11_nidal_0_55_55,train +11_nidal_0_56_56,train +11_nidal_0_57_57,train +11_nidal_0_58_58,train +11_nidal_0_59_59,train +11_nidal_0_5_5,test +11_nidal_0_60_60,train +11_nidal_0_61_61,train +11_nidal_0_62_62,train +11_nidal_0_63_63,train +11_nidal_0_64_64,train +11_nidal_0_65_65,test +11_nidal_0_66_66,train +11_nidal_0_67_67,train +11_nidal_0_68_68,train +11_nidal_0_69_69,train +11_nidal_0_6_6,test +11_nidal_0_70_70,train +11_nidal_0_71_71,train +11_nidal_0_72_72,train +11_nidal_0_73_73,test +11_nidal_0_74_74,train +11_nidal_0_75_75,train +11_nidal_0_76_76,train +11_nidal_0_77_77,train +11_nidal_0_78_78,train +11_nidal_0_79_79,train +11_nidal_0_7_7,test +11_nidal_0_80_80,train +11_nidal_0_81_81,test +11_nidal_0_82_82,train +11_nidal_0_83_83,train +11_nidal_0_84_84,train +11_nidal_0_85_85,train +11_nidal_0_86_86,train +11_nidal_0_87_87,test +11_nidal_0_88_88,train +11_nidal_0_89_89,train +11_nidal_0_8_8,test +11_nidal_0_90_90,train +11_nidal_0_91_91,train +11_nidal_0_92_92,train +11_nidal_0_93_93,train +11_nidal_0_94_94,train +11_nidal_0_95_95,test +11_nidal_0_96_96,train +11_nidal_0_97_97,train +11_nidal_0_98_98,train +11_nidal_0_99_99,train +11_nidal_0_9_9,val +11_nidal_1_10_10,additional +11_nidal_1_11_11,additional +11_nidal_1_12_12,additional +11_nidal_1_1_1,additional +11_nidal_1_2_2,additional +11_nidal_1_3_3,additional +11_nidal_1_4_4,additional +11_nidal_1_5_5,additional +11_nidal_1_6_6,additional +11_nidal_1_7_7,additional +11_nidal_1_8_8,additional +11_nidal_1_9_9,additional +12_zhao_0_103_103,test +12_zhao_0_104_104,train +12_zhao_0_10_10,val +12_zhao_0_111_111,test +12_zhao_0_112_112,train +12_zhao_0_11_11,train +12_zhao_0_12_12,train +12_zhao_0_13_13,train +12_zhao_0_14_14,train +12_zhao_0_15_15,train +12_zhao_0_16_16,train +12_zhao_0_1_1,test +12_zhao_0_2_2,test +12_zhao_0_3_3,train +12_zhao_0_4_4,train +12_zhao_0_5_5,train +12_zhao_0_65_65,test +12_zhao_0_66_66,train +12_zhao_0_6_6,train +12_zhao_0_73_73,test +12_zhao_0_74_74,train +12_zhao_0_7_7,train +12_zhao_0_81_81,test +12_zhao_0_82_82,train +12_zhao_0_87_87,test +12_zhao_0_88_88,train +12_zhao_0_8_8,train +12_zhao_0_95_95,test +12_zhao_0_96_96,train +12_zhao_0_9_9,val +12_zhao_1_1_1,additional +12_zhao_1_2_2,additional +12_zhao_1_3_3,additional +13_lu_0_103_103,test +13_lu_0_104_104,train +13_lu_0_10_10,val +13_lu_0_111_111,test +13_lu_0_112_112,train +13_lu_0_11_11,train +13_lu_0_12_12,train +13_lu_0_13_13,train +13_lu_0_14_14,train +13_lu_0_15_15,train +13_lu_0_16_16,train +13_lu_0_1_1,test +13_lu_0_2_2,test +13_lu_0_3_3,train +13_lu_0_4_4,train +13_lu_0_5_5,train +13_lu_0_65_65,test +13_lu_0_66_66,train +13_lu_0_6_6,train +13_lu_0_73_73,test +13_lu_0_74_74,train +13_lu_0_7_7,train +13_lu_0_81_81,test +13_lu_0_82_82,train +13_lu_0_87_87,test +13_lu_0_88_88,train +13_lu_0_8_8,train +13_lu_0_95_95,test +13_lu_0_96_96,train +13_lu_0_9_9,val +13_lu_1_1_1,additional +13_lu_1_2_2,additional +13_lu_1_3_3,additional +15_carlos_0_103_103,test +15_carlos_0_104_104,train +15_carlos_0_10_10,val +15_carlos_0_111_111,test +15_carlos_0_112_112,train +15_carlos_0_11_11,train +15_carlos_0_13_13,train +15_carlos_0_14_14,train +15_carlos_0_15_15,train +15_carlos_0_16_16,train +15_carlos_0_1_1,test +15_carlos_0_2_2,test +15_carlos_0_3_3,train +15_carlos_0_4_4,train +15_carlos_0_5_5,train +15_carlos_0_65_65,test +15_carlos_0_66_66,train +15_carlos_0_73_73,test +15_carlos_0_74_74,train +15_carlos_0_7_7,train +15_carlos_0_81_81,test +15_carlos_0_82_82,train +15_carlos_0_87_87,test +15_carlos_0_88_88,train +15_carlos_0_8_8,train +15_carlos_0_95_95,test +15_carlos_0_96_96,train +15_carlos_0_9_9,val +15_carlos_1_1_1,additional +15_carlos_1_2_2,additional +15_carlos_1_3_3,additional +16_jorge_0_103_103,test +16_jorge_0_104_104,train +16_jorge_0_10_10,val +16_jorge_0_111_111,test +16_jorge_0_112_112,train +16_jorge_0_11_11,train +16_jorge_0_12_12,train +16_jorge_0_13_13,train +16_jorge_0_14_14,train +16_jorge_0_15_15,train +16_jorge_0_16_16,train +16_jorge_0_1_1,test +16_jorge_0_2_2,test +16_jorge_0_3_3,train +16_jorge_0_4_4,train +16_jorge_0_5_5,train +16_jorge_0_65_65,test +16_jorge_0_66_66,train +16_jorge_0_6_6,train +16_jorge_0_73_73,test +16_jorge_0_74_74,train +16_jorge_0_7_7,train +16_jorge_0_81_81,test +16_jorge_0_82_82,train +16_jorge_0_87_87,test +16_jorge_0_88_88,train +16_jorge_0_8_8,train +16_jorge_0_95_95,test +16_jorge_0_96_96,train +16_jorge_0_9_9,val +16_jorge_1_1_1,additional +16_jorge_1_2_2,additional +16_jorge_1_3_3,additional +17_itoi_0_103_103,test +17_itoi_0_104_104,train +17_itoi_0_10_10,val +17_itoi_0_111_111,test +17_itoi_0_112_112,train +17_itoi_0_11_11,train +17_itoi_0_12_12,train +17_itoi_0_13_13,train +17_itoi_0_14_14,train +17_itoi_0_15_15,train +17_itoi_0_16_16,train +17_itoi_0_1_1,test +17_itoi_0_2_2,test +17_itoi_0_3_3,train +17_itoi_0_4_4,train +17_itoi_0_5_5,train +17_itoi_0_65_65,test +17_itoi_0_66_66,train +17_itoi_0_6_6,train +17_itoi_0_73_73,test +17_itoi_0_74_74,train +17_itoi_0_7_7,train +17_itoi_0_81_81,test +17_itoi_0_82_82,train +17_itoi_0_87_87,test +17_itoi_0_88_88,train +17_itoi_0_8_8,train +17_itoi_0_95_95,test +17_itoi_0_96_96,train +17_itoi_0_9_9,val +17_itoi_1_1_1,additional +17_itoi_1_2_2,additional +17_itoi_1_3_3,additional +18_daiki_0_103_103,test +18_daiki_0_104_104,train +18_daiki_0_10_10,val +18_daiki_0_111_111,test +18_daiki_0_112_112,train +18_daiki_0_11_11,train +18_daiki_0_12_12,train +18_daiki_0_13_13,train +18_daiki_0_14_14,train +18_daiki_0_15_15,train +18_daiki_0_16_16,train +18_daiki_0_1_1,test +18_daiki_0_2_2,test +18_daiki_0_3_3,train +18_daiki_0_4_4,train +18_daiki_0_5_5,train +18_daiki_0_65_65,test +18_daiki_0_66_66,train +18_daiki_0_6_6,train +18_daiki_0_73_73,test +18_daiki_0_74_74,train +18_daiki_0_7_7,train +18_daiki_0_81_81,test +18_daiki_0_82_82,train +18_daiki_0_87_87,test +18_daiki_0_88_88,train +18_daiki_0_8_8,train +18_daiki_0_95_95,test +18_daiki_0_96_96,train +18_daiki_0_9_9,val +18_daiki_1_1_1,additional +18_daiki_1_2_2,additional +18_daiki_1_3_3,additional +1_wayne_0_100_100,train +1_wayne_0_101_101,train +1_wayne_0_102_102,train +1_wayne_0_103_103,test +1_wayne_0_104_104,train +1_wayne_0_105_105,train +1_wayne_0_106_106,train +1_wayne_0_107_107,train +1_wayne_0_108_108,train +1_wayne_0_109_109,train +1_wayne_0_10_10,val +1_wayne_0_110_110,train +1_wayne_0_111_111,test +1_wayne_0_112_112,train +1_wayne_0_113_113,train +1_wayne_0_114_114,train +1_wayne_0_115_115,train +1_wayne_0_116_116,train +1_wayne_0_117_117,train +1_wayne_0_118_118,train +1_wayne_0_11_11,val +1_wayne_0_12_12,val +1_wayne_0_13_13,val +1_wayne_0_14_14,val +1_wayne_0_15_15,val +1_wayne_0_16_16,val +1_wayne_0_17_17,train +1_wayne_0_18_18,train +1_wayne_0_19_19,train +1_wayne_0_1_1,test +1_wayne_0_20_20,train +1_wayne_0_21_21,train +1_wayne_0_22_22,train +1_wayne_0_23_23,train +1_wayne_0_24_24,train +1_wayne_0_25_25,train +1_wayne_0_26_26,train +1_wayne_0_27_27,train +1_wayne_0_28_28,train +1_wayne_0_29_29,train +1_wayne_0_2_2,test +1_wayne_0_30_30,train +1_wayne_0_31_31,train +1_wayne_0_32_32,train +1_wayne_0_33_33,train +1_wayne_0_34_34,train +1_wayne_0_35_35,train +1_wayne_0_36_36,train +1_wayne_0_37_37,train +1_wayne_0_38_38,train +1_wayne_0_39_39,train +1_wayne_0_3_3,test +1_wayne_0_40_40,train +1_wayne_0_41_41,train +1_wayne_0_42_42,train +1_wayne_0_43_43,train +1_wayne_0_44_44,train +1_wayne_0_45_45,train +1_wayne_0_46_46,train +1_wayne_0_47_47,train +1_wayne_0_48_48,train +1_wayne_0_49_49,train +1_wayne_0_4_4,test +1_wayne_0_50_50,train +1_wayne_0_51_51,train +1_wayne_0_52_52,train +1_wayne_0_53_53,train +1_wayne_0_54_54,train +1_wayne_0_55_55,train +1_wayne_0_56_56,train +1_wayne_0_57_57,train +1_wayne_0_58_58,train +1_wayne_0_59_59,train +1_wayne_0_5_5,test +1_wayne_0_60_60,train +1_wayne_0_61_61,train +1_wayne_0_62_62,train +1_wayne_0_63_63,train +1_wayne_0_64_64,train +1_wayne_0_65_65,test +1_wayne_0_66_66,train +1_wayne_0_67_67,train +1_wayne_0_68_68,train +1_wayne_0_69_69,train +1_wayne_0_6_6,test +1_wayne_0_70_70,train +1_wayne_0_71_71,train +1_wayne_0_72_72,train +1_wayne_0_73_73,test +1_wayne_0_74_74,train +1_wayne_0_75_75,train +1_wayne_0_76_76,train +1_wayne_0_77_77,train +1_wayne_0_78_78,train +1_wayne_0_79_79,train +1_wayne_0_7_7,test +1_wayne_0_80_80,train +1_wayne_0_81_81,test +1_wayne_0_82_82,train +1_wayne_0_83_83,train +1_wayne_0_84_84,train +1_wayne_0_85_85,train +1_wayne_0_86_86,train +1_wayne_0_87_87,test +1_wayne_0_88_88,train +1_wayne_0_89_89,train +1_wayne_0_8_8,test +1_wayne_0_90_90,train +1_wayne_0_91_91,train +1_wayne_0_92_92,train +1_wayne_0_93_93,train +1_wayne_0_94_94,train +1_wayne_0_95_95,test +1_wayne_0_96_96,train +1_wayne_0_97_97,train +1_wayne_0_98_98,train +1_wayne_0_99_99,train +1_wayne_0_9_9,val +1_wayne_1_10_10,additional +1_wayne_1_11_11,additional +1_wayne_1_12_12,additional +1_wayne_1_1_1,additional +1_wayne_1_2_2,additional +1_wayne_1_3_3,additional +1_wayne_1_4_4,additional +1_wayne_1_5_5,additional +1_wayne_1_6_6,additional +1_wayne_1_7_7,additional +1_wayne_1_8_8,additional +1_wayne_1_9_9,additional +20_li_0_103_103,test +20_li_0_104_104,train +20_li_0_10_10,val +20_li_0_111_111,test +20_li_0_112_112,train +20_li_0_11_11,train +20_li_0_12_12,train +20_li_0_13_13,train +20_li_0_14_14,train +20_li_0_15_15,train +20_li_0_16_16,train +20_li_0_1_1,test +20_li_0_2_2,test +20_li_0_3_3,train +20_li_0_4_4,train +20_li_0_5_5,train +20_li_0_65_65,test +20_li_0_66_66,train +20_li_0_6_6,train +20_li_0_73_73,test +20_li_0_74_74,train +20_li_0_7_7,train +20_li_0_81_81,test +20_li_0_82_82,train +20_li_0_87_87,test +20_li_0_88_88,train +20_li_0_8_8,train +20_li_0_95_95,test +20_li_0_96_96,train +20_li_0_9_9,val +20_li_1_1_1,additional +20_li_1_2_2,additional +20_li_1_3_3,additional +21_ayana_0_100_100,train +21_ayana_0_101_101,train +21_ayana_0_102_102,train +21_ayana_0_103_103,test +21_ayana_0_104_104,train +21_ayana_0_105_105,train +21_ayana_0_106_106,train +21_ayana_0_107_107,train +21_ayana_0_108_108,train +21_ayana_0_109_109,train +21_ayana_0_10_10,val +21_ayana_0_110_110,train +21_ayana_0_111_111,test +21_ayana_0_112_112,train +21_ayana_0_113_113,train +21_ayana_0_114_114,train +21_ayana_0_115_115,train +21_ayana_0_116_116,train +21_ayana_0_117_117,train +21_ayana_0_118_118,train +21_ayana_0_11_11,val +21_ayana_0_12_12,val +21_ayana_0_13_13,val +21_ayana_0_14_14,val +21_ayana_0_15_15,val +21_ayana_0_16_16,val +21_ayana_0_17_17,train +21_ayana_0_18_18,train +21_ayana_0_19_19,train +21_ayana_0_20_20,train +21_ayana_0_21_21,train +21_ayana_0_22_22,train +21_ayana_0_23_23,train +21_ayana_0_24_24,train +21_ayana_0_25_25,train +21_ayana_0_26_26,train +21_ayana_0_27_27,train +21_ayana_0_28_28,train +21_ayana_0_29_29,train +21_ayana_0_30_30,train +21_ayana_0_31_31,train +21_ayana_0_32_32,train +21_ayana_0_33_33,train +21_ayana_0_34_34,train +21_ayana_0_35_35,train +21_ayana_0_36_36,train +21_ayana_0_37_37,train +21_ayana_0_38_38,train +21_ayana_0_39_39,train +21_ayana_0_40_40,train +21_ayana_0_41_41,train +21_ayana_0_42_42,train +21_ayana_0_43_43,train +21_ayana_0_44_44,train +21_ayana_0_45_45,train +21_ayana_0_46_46,train +21_ayana_0_47_47,train +21_ayana_0_48_48,train +21_ayana_0_49_49,train +21_ayana_0_50_50,train +21_ayana_0_51_51,train +21_ayana_0_52_52,train +21_ayana_0_53_53,train +21_ayana_0_54_54,train +21_ayana_0_55_55,train +21_ayana_0_56_56,train +21_ayana_0_57_57,train +21_ayana_0_58_58,train +21_ayana_0_59_59,train +21_ayana_0_60_60,train +21_ayana_0_61_61,train +21_ayana_0_62_62,train +21_ayana_0_63_63,train +21_ayana_0_64_64,train +21_ayana_0_65_65,test +21_ayana_0_66_66,train +21_ayana_0_67_67,train +21_ayana_0_68_68,train +21_ayana_0_69_69,train +21_ayana_0_70_70,train +21_ayana_0_71_71,train +21_ayana_0_72_72,train +21_ayana_0_73_73,test +21_ayana_0_74_74,train +21_ayana_0_75_75,train +21_ayana_0_76_76,train +21_ayana_0_77_77,train +21_ayana_0_78_78,train +21_ayana_0_79_79,train +21_ayana_0_80_80,train +21_ayana_0_81_81,test +21_ayana_0_82_82,train +21_ayana_0_83_83,train +21_ayana_0_84_84,train +21_ayana_0_85_85,train +21_ayana_0_86_86,train +21_ayana_0_87_87,test +21_ayana_0_88_88,train +21_ayana_0_89_89,train +21_ayana_0_90_90,train +21_ayana_0_91_91,train +21_ayana_0_92_92,train +21_ayana_0_93_93,train +21_ayana_0_94_94,train +21_ayana_0_95_95,test +21_ayana_0_96_96,train +21_ayana_0_97_97,train +21_ayana_0_98_98,train +21_ayana_0_99_99,train +21_ayana_0_9_9,val +21_ayana_1_10_10,additional +21_ayana_1_11_11,additional +21_ayana_1_12_12,additional +21_ayana_1_1_1,additional +21_ayana_1_2_2,additional +21_ayana_1_3_3,additional +21_ayana_1_4_4,additional +21_ayana_1_5_5,additional +21_ayana_1_6_6,additional +21_ayana_1_7_7,additional +21_ayana_1_8_8,additional +21_ayana_1_9_9,additional +22_luqi_0_103_103,test +22_luqi_0_104_104,train +22_luqi_0_111_111,test +22_luqi_0_112_112,train +22_luqi_0_1_1,test +22_luqi_0_2_2,test +22_luqi_0_3_3,train +22_luqi_0_4_4,train +22_luqi_0_5_5,train +22_luqi_0_65_65,test +22_luqi_0_66_66,train +22_luqi_0_6_6,train +22_luqi_0_73_73,test +22_luqi_0_74_74,train +22_luqi_0_7_7,train +22_luqi_0_81_81,test +22_luqi_0_82_82,train +22_luqi_0_87_87,test +22_luqi_0_88_88,train +22_luqi_0_8_8,train +22_luqi_0_95_95,test +22_luqi_0_96_96,train +22_luqi_1_1_1,additional +22_luqi_1_2_2,additional +22_luqi_1_3_3,additional +23_hailing_0_103_103,test +23_hailing_0_104_104,train +23_hailing_0_10_10,val +23_hailing_0_111_111,test +23_hailing_0_112_112,train +23_hailing_0_11_11,train +23_hailing_0_12_12,train +23_hailing_0_13_13,train +23_hailing_0_14_14,train +23_hailing_0_15_15,train +23_hailing_0_16_16,train +23_hailing_0_1_1,test +23_hailing_0_2_2,test +23_hailing_0_3_3,train +23_hailing_0_4_4,train +23_hailing_0_5_5,train +23_hailing_0_65_65,test +23_hailing_0_66_66,train +23_hailing_0_6_6,train +23_hailing_0_7_7,train +23_hailing_0_81_81,test +23_hailing_0_82_82,train +23_hailing_0_87_87,test +23_hailing_0_88_88,train +23_hailing_0_8_8,train +23_hailing_0_95_95,test +23_hailing_0_96_96,train +23_hailing_0_9_9,val +23_hailing_1_1_1,additional +23_hailing_1_2_2,additional +23_hailing_1_3_3,additional +24_kexin_0_103_103,test +24_kexin_0_104_104,train +24_kexin_0_10_10,val +24_kexin_0_111_111,test +24_kexin_0_112_112,train +24_kexin_0_11_11,train +24_kexin_0_12_12,train +24_kexin_0_13_13,train +24_kexin_0_14_14,train +24_kexin_0_15_15,train +24_kexin_0_16_16,train +24_kexin_0_1_1,test +24_kexin_0_2_2,test +24_kexin_0_3_3,train +24_kexin_0_4_4,train +24_kexin_0_5_5,train +24_kexin_0_65_65,test +24_kexin_0_66_66,train +24_kexin_0_6_6,train +24_kexin_0_73_73,test +24_kexin_0_74_74,train +24_kexin_0_7_7,train +24_kexin_0_81_81,test +24_kexin_0_82_82,train +24_kexin_0_87_87,test +24_kexin_0_88_88,train +24_kexin_0_8_8,train +24_kexin_0_95_95,test +24_kexin_0_96_96,train +24_kexin_0_9_9,val +24_kexin_1_1_1,additional +24_kexin_1_2_2,additional +24_kexin_1_3_3,additional +25_goto_0_103_103,test +25_goto_0_104_104,train +25_goto_0_10_10,val +25_goto_0_111_111,test +25_goto_0_112_112,train +25_goto_0_11_11,train +25_goto_0_12_12,train +25_goto_0_13_13,train +25_goto_0_14_14,train +25_goto_0_15_15,train +25_goto_0_16_16,train +25_goto_0_2_2,test +25_goto_0_3_3,train +25_goto_0_4_4,train +25_goto_0_65_65,test +25_goto_0_66_66,train +25_goto_0_6_6,train +25_goto_0_73_73,test +25_goto_0_74_74,train +25_goto_0_7_7,train +25_goto_0_81_81,test +25_goto_0_82_82,train +25_goto_0_87_87,test +25_goto_0_88_88,train +25_goto_0_8_8,train +25_goto_0_95_95,test +25_goto_0_96_96,train +25_goto_0_9_9,val +25_goto_1_1_1,additional +25_goto_1_2_2,additional +25_goto_1_3_3,additional +27_yingqing_0_103_103,test +27_yingqing_0_104_104,train +27_yingqing_0_10_10,val +27_yingqing_0_111_111,test +27_yingqing_0_112_112,train +27_yingqing_0_11_11,train +27_yingqing_0_12_12,train +27_yingqing_0_13_13,train +27_yingqing_0_14_14,train +27_yingqing_0_15_15,train +27_yingqing_0_16_16,train +27_yingqing_0_1_1,test +27_yingqing_0_2_2,test +27_yingqing_0_3_3,train +27_yingqing_0_4_4,train +27_yingqing_0_5_5,train +27_yingqing_0_65_65,test +27_yingqing_0_66_66,train +27_yingqing_0_6_6,train +27_yingqing_0_73_73,test +27_yingqing_0_74_74,train +27_yingqing_0_7_7,train +27_yingqing_0_81_81,test +27_yingqing_0_82_82,train +27_yingqing_0_87_87,test +27_yingqing_0_88_88,train +27_yingqing_0_8_8,train +27_yingqing_0_95_95,test +27_yingqing_0_96_96,train +27_yingqing_0_9_9,val +27_yingqing_1_1_1,additional +27_yingqing_1_2_2,additional +27_yingqing_1_3_3,additional +28_tiffnay_0_103_103,test +28_tiffnay_0_104_104,train +28_tiffnay_0_10_10,val +28_tiffnay_0_111_111,test +28_tiffnay_0_112_112,train +28_tiffnay_0_11_11,train +28_tiffnay_0_12_12,train +28_tiffnay_0_13_13,train +28_tiffnay_0_14_14,train +28_tiffnay_0_15_15,train +28_tiffnay_0_16_16,train +28_tiffnay_0_1_1,test +28_tiffnay_0_2_2,test +28_tiffnay_0_3_3,train +28_tiffnay_0_4_4,train +28_tiffnay_0_5_5,train +28_tiffnay_0_65_65,test +28_tiffnay_0_66_66,train +28_tiffnay_0_6_6,train +28_tiffnay_0_73_73,test +28_tiffnay_0_74_74,train +28_tiffnay_0_7_7,train +28_tiffnay_0_81_81,test +28_tiffnay_0_82_82,train +28_tiffnay_0_87_87,test +28_tiffnay_0_88_88,train +28_tiffnay_0_8_8,train +28_tiffnay_0_95_95,test +28_tiffnay_0_96_96,train +28_tiffnay_0_9_9,val +28_tiffnay_1_1_1,additional +28_tiffnay_1_2_2,additional +28_tiffnay_1_3_3,additional +2_scott_0_100_100,train +2_scott_0_101_101,train +2_scott_0_102_102,train +2_scott_0_103_103,test +2_scott_0_104_104,train +2_scott_0_105_105,train +2_scott_0_106_106,train +2_scott_0_107_107,train +2_scott_0_108_108,train +2_scott_0_109_109,train +2_scott_0_10_10,val +2_scott_0_110_110,train +2_scott_0_111_111,test +2_scott_0_112_112,train +2_scott_0_113_113,train +2_scott_0_114_114,train +2_scott_0_115_115,train +2_scott_0_116_116,train +2_scott_0_117_117,train +2_scott_0_118_118,train +2_scott_0_11_11,val +2_scott_0_12_12,val +2_scott_0_13_13,val +2_scott_0_14_14,val +2_scott_0_15_15,val +2_scott_0_16_16,val +2_scott_0_17_17,train +2_scott_0_18_18,train +2_scott_0_19_19,train +2_scott_0_1_1,test +2_scott_0_20_20,train +2_scott_0_21_21,train +2_scott_0_22_22,train +2_scott_0_23_23,train +2_scott_0_24_24,train +2_scott_0_25_25,train +2_scott_0_26_26,train +2_scott_0_27_27,train +2_scott_0_28_28,train +2_scott_0_29_29,train +2_scott_0_2_2,test +2_scott_0_30_30,train +2_scott_0_31_31,train +2_scott_0_32_32,train +2_scott_0_33_33,train +2_scott_0_34_34,train +2_scott_0_35_35,train +2_scott_0_36_36,train +2_scott_0_37_37,train +2_scott_0_38_38,train +2_scott_0_39_39,train +2_scott_0_3_3,test +2_scott_0_40_40,train +2_scott_0_41_41,train +2_scott_0_42_42,train +2_scott_0_43_43,train +2_scott_0_44_44,train +2_scott_0_45_45,train +2_scott_0_46_46,train +2_scott_0_47_47,train +2_scott_0_48_48,train +2_scott_0_49_49,train +2_scott_0_4_4,test +2_scott_0_50_50,train +2_scott_0_51_51,train +2_scott_0_52_52,train +2_scott_0_53_53,train +2_scott_0_54_54,train +2_scott_0_55_55,train +2_scott_0_56_56,train +2_scott_0_57_57,train +2_scott_0_58_58,train +2_scott_0_59_59,train +2_scott_0_5_5,test +2_scott_0_60_60,train +2_scott_0_61_61,train +2_scott_0_62_62,train +2_scott_0_63_63,train +2_scott_0_64_64,train +2_scott_0_65_65,test +2_scott_0_66_66,train +2_scott_0_67_67,train +2_scott_0_68_68,train +2_scott_0_69_69,train +2_scott_0_6_6,test +2_scott_0_70_70,train +2_scott_0_71_71,train +2_scott_0_72_72,train +2_scott_0_73_73,test +2_scott_0_74_74,train +2_scott_0_75_75,train +2_scott_0_76_76,train +2_scott_0_77_77,train +2_scott_0_78_78,train +2_scott_0_79_79,train +2_scott_0_7_7,test +2_scott_0_80_80,train +2_scott_0_81_81,test +2_scott_0_82_82,train +2_scott_0_83_83,train +2_scott_0_84_84,train +2_scott_0_85_85,train +2_scott_0_86_86,train +2_scott_0_87_87,test +2_scott_0_88_88,train +2_scott_0_89_89,train +2_scott_0_8_8,test +2_scott_0_90_90,train +2_scott_0_91_91,train +2_scott_0_92_92,train +2_scott_0_93_93,train +2_scott_0_94_94,train +2_scott_0_95_95,test +2_scott_0_96_96,train +2_scott_0_97_97,train +2_scott_0_98_98,train +2_scott_0_99_99,train +2_scott_0_9_9,val +2_scott_1_10_10,additional +2_scott_1_11_11,additional +2_scott_1_12_12,additional +2_scott_1_1_1,additional +2_scott_1_2_2,additional +2_scott_1_3_3,additional +2_scott_1_4_4,additional +2_scott_1_5_5,additional +2_scott_1_6_6,additional +2_scott_1_7_7,additional +2_scott_1_8_8,additional +2_scott_1_9_9,additional +30_katya_0_103_103,test +30_katya_0_104_104,train +30_katya_0_10_10,val +30_katya_0_111_111,test +30_katya_0_112_112,train +30_katya_0_11_11,train +30_katya_0_12_12,train +30_katya_0_13_13,train +30_katya_0_14_14,train +30_katya_0_15_15,train +30_katya_0_16_16,train +30_katya_0_1_1,test +30_katya_0_2_2,test +30_katya_0_3_3,train +30_katya_0_4_4,train +30_katya_0_5_5,train +30_katya_0_65_65,test +30_katya_0_66_66,train +30_katya_0_6_6,train +30_katya_0_73_73,test +30_katya_0_74_74,train +30_katya_0_7_7,train +30_katya_0_81_81,test +30_katya_0_82_82,train +30_katya_0_87_87,test +30_katya_0_88_88,train +30_katya_0_8_8,train +30_katya_0_95_95,test +30_katya_0_96_96,train +30_katya_0_9_9,val +30_katya_1_1_1,additional +30_katya_1_2_2,additional +30_katya_1_3_3,additional +3_solomon_0_100_100,train +3_solomon_0_101_101,train +3_solomon_0_102_102,train +3_solomon_0_103_103,test +3_solomon_0_104_104,train +3_solomon_0_105_105,train +3_solomon_0_106_106,train +3_solomon_0_107_107,train +3_solomon_0_108_108,train +3_solomon_0_109_109,train +3_solomon_0_10_10,val +3_solomon_0_110_110,train +3_solomon_0_111_111,test +3_solomon_0_112_112,train +3_solomon_0_113_113,train +3_solomon_0_114_114,train +3_solomon_0_115_115,train +3_solomon_0_116_116,train +3_solomon_0_117_117,train +3_solomon_0_118_118,train +3_solomon_0_11_11,val +3_solomon_0_12_12,val +3_solomon_0_13_13,val +3_solomon_0_14_14,val +3_solomon_0_15_15,val +3_solomon_0_16_16,val +3_solomon_0_17_17,train +3_solomon_0_18_18,train +3_solomon_0_19_19,train +3_solomon_0_1_1,test +3_solomon_0_20_20,train +3_solomon_0_21_21,train +3_solomon_0_22_22,train +3_solomon_0_23_23,train +3_solomon_0_24_24,train +3_solomon_0_25_25,train +3_solomon_0_26_26,train +3_solomon_0_27_27,train +3_solomon_0_28_28,train +3_solomon_0_29_29,train +3_solomon_0_2_2,test +3_solomon_0_30_30,train +3_solomon_0_31_31,train +3_solomon_0_32_32,train +3_solomon_0_33_33,train +3_solomon_0_34_34,train +3_solomon_0_35_35,train +3_solomon_0_36_36,train +3_solomon_0_37_37,train +3_solomon_0_38_38,train +3_solomon_0_39_39,train +3_solomon_0_3_3,test +3_solomon_0_40_40,train +3_solomon_0_41_41,train +3_solomon_0_42_42,train +3_solomon_0_43_43,train +3_solomon_0_44_44,train +3_solomon_0_45_45,train +3_solomon_0_46_46,train +3_solomon_0_47_47,train +3_solomon_0_48_48,train +3_solomon_0_49_49,train +3_solomon_0_4_4,test +3_solomon_0_50_50,train +3_solomon_0_51_51,train +3_solomon_0_52_52,train +3_solomon_0_53_53,train +3_solomon_0_54_54,train +3_solomon_0_55_55,train +3_solomon_0_56_56,train +3_solomon_0_57_57,train +3_solomon_0_58_58,train +3_solomon_0_59_59,train +3_solomon_0_5_5,test +3_solomon_0_60_60,train +3_solomon_0_61_61,train +3_solomon_0_62_62,train +3_solomon_0_63_63,train +3_solomon_0_64_64,train +3_solomon_0_65_65,test +3_solomon_0_66_66,train +3_solomon_0_67_67,train +3_solomon_0_68_68,train +3_solomon_0_69_69,train +3_solomon_0_6_6,test +3_solomon_0_70_70,train +3_solomon_0_71_71,train +3_solomon_0_72_72,train +3_solomon_0_73_73,test +3_solomon_0_74_74,train +3_solomon_0_75_75,train +3_solomon_0_76_76,train +3_solomon_0_77_77,train +3_solomon_0_78_78,train +3_solomon_0_79_79,train +3_solomon_0_7_7,test +3_solomon_0_80_80,train +3_solomon_0_81_81,test +3_solomon_0_82_82,train +3_solomon_0_83_83,train +3_solomon_0_84_84,train +3_solomon_0_85_85,train +3_solomon_0_86_86,train +3_solomon_0_87_87,test +3_solomon_0_88_88,train +3_solomon_0_89_89,train +3_solomon_0_8_8,test +3_solomon_0_90_90,train +3_solomon_0_91_91,train +3_solomon_0_92_92,train +3_solomon_0_93_93,train +3_solomon_0_94_94,train +3_solomon_0_95_95,test +3_solomon_0_96_96,train +3_solomon_0_97_97,train +3_solomon_0_98_98,train +3_solomon_0_99_99,train +3_solomon_0_9_9,val +3_solomon_1_10_10,additional +3_solomon_1_11_11,additional +3_solomon_1_12_12,additional +3_solomon_1_1_1,additional +3_solomon_1_2_2,additional +3_solomon_1_3_3,additional +3_solomon_1_4_4,additional +3_solomon_1_5_5,additional +3_solomon_1_6_6,additional +3_solomon_1_7_7,additional +3_solomon_1_8_8,additional +3_solomon_1_9_9,additional +4_lawrence_0_100_100,train +4_lawrence_0_101_101,train +4_lawrence_0_102_102,train +4_lawrence_0_103_103,test +4_lawrence_0_104_104,train +4_lawrence_0_105_105,train +4_lawrence_0_106_106,train +4_lawrence_0_107_107,train +4_lawrence_0_108_108,train +4_lawrence_0_109_109,train +4_lawrence_0_10_10,val +4_lawrence_0_110_110,train +4_lawrence_0_111_111,test +4_lawrence_0_112_112,train +4_lawrence_0_113_113,train +4_lawrence_0_114_114,train +4_lawrence_0_115_115,train +4_lawrence_0_116_116,train +4_lawrence_0_117_117,train +4_lawrence_0_118_118,train +4_lawrence_0_11_11,val +4_lawrence_0_12_12,val +4_lawrence_0_13_13,val +4_lawrence_0_14_14,val +4_lawrence_0_15_15,val +4_lawrence_0_16_16,val +4_lawrence_0_17_17,train +4_lawrence_0_18_18,train +4_lawrence_0_19_19,train +4_lawrence_0_1_1,test +4_lawrence_0_20_20,train +4_lawrence_0_21_21,train +4_lawrence_0_22_22,train +4_lawrence_0_23_23,train +4_lawrence_0_24_24,train +4_lawrence_0_25_25,train +4_lawrence_0_26_26,train +4_lawrence_0_27_27,train +4_lawrence_0_28_28,train +4_lawrence_0_29_29,train +4_lawrence_0_2_2,test +4_lawrence_0_30_30,train +4_lawrence_0_31_31,train +4_lawrence_0_32_32,train +4_lawrence_0_33_33,train +4_lawrence_0_34_34,train +4_lawrence_0_35_35,train +4_lawrence_0_36_36,train +4_lawrence_0_37_37,train +4_lawrence_0_38_38,train +4_lawrence_0_39_39,train +4_lawrence_0_3_3,test +4_lawrence_0_40_40,train +4_lawrence_0_41_41,train +4_lawrence_0_42_42,train +4_lawrence_0_43_43,train +4_lawrence_0_44_44,train +4_lawrence_0_45_45,train +4_lawrence_0_46_46,train +4_lawrence_0_47_47,train +4_lawrence_0_48_48,train +4_lawrence_0_49_49,train +4_lawrence_0_4_4,test +4_lawrence_0_50_50,train +4_lawrence_0_51_51,train +4_lawrence_0_52_52,train +4_lawrence_0_53_53,train +4_lawrence_0_54_54,train +4_lawrence_0_55_55,train +4_lawrence_0_56_56,train +4_lawrence_0_57_57,train +4_lawrence_0_58_58,train +4_lawrence_0_59_59,train +4_lawrence_0_5_5,test +4_lawrence_0_60_60,train +4_lawrence_0_61_61,train +4_lawrence_0_62_62,train +4_lawrence_0_63_63,train +4_lawrence_0_64_64,train +4_lawrence_0_65_65,test +4_lawrence_0_66_66,train +4_lawrence_0_67_67,train +4_lawrence_0_68_68,train +4_lawrence_0_69_69,train +4_lawrence_0_6_6,test +4_lawrence_0_70_70,train +4_lawrence_0_71_71,train +4_lawrence_0_72_72,train +4_lawrence_0_73_73,test +4_lawrence_0_74_74,train +4_lawrence_0_75_75,train +4_lawrence_0_76_76,train +4_lawrence_0_77_77,train +4_lawrence_0_78_78,train +4_lawrence_0_79_79,train +4_lawrence_0_7_7,test +4_lawrence_0_80_80,train +4_lawrence_0_81_81,test +4_lawrence_0_82_82,train +4_lawrence_0_83_83,train +4_lawrence_0_84_84,train +4_lawrence_0_85_85,train +4_lawrence_0_86_86,train +4_lawrence_0_87_87,test +4_lawrence_0_88_88,train +4_lawrence_0_89_89,train +4_lawrence_0_8_8,test +4_lawrence_0_90_90,train +4_lawrence_0_91_91,train +4_lawrence_0_92_92,train +4_lawrence_0_93_93,train +4_lawrence_0_94_94,train +4_lawrence_0_95_95,test +4_lawrence_0_96_96,train +4_lawrence_0_97_97,train +4_lawrence_0_98_98,train +4_lawrence_0_99_99,train +4_lawrence_0_9_9,val +4_lawrence_1_10_10,additional +4_lawrence_1_11_11,additional +4_lawrence_1_12_12,additional +4_lawrence_1_1_1,additional +4_lawrence_1_2_2,additional +4_lawrence_1_3_3,additional +4_lawrence_1_4_4,additional +4_lawrence_1_5_5,additional +4_lawrence_1_6_6,additional +4_lawrence_1_7_7,additional +4_lawrence_1_8_8,additional +4_lawrence_1_9_9,additional +5_stewart_0_103_103,test +5_stewart_0_104_104,train +5_stewart_0_10_10,val +5_stewart_0_111_111,test +5_stewart_0_112_112,train +5_stewart_0_11_11,val +5_stewart_0_12_12,val +5_stewart_0_13_13,val +5_stewart_0_14_14,val +5_stewart_0_15_15,val +5_stewart_0_16_16,val +5_stewart_0_1_1,test +5_stewart_0_2_2,test +5_stewart_0_3_3,test +5_stewart_0_4_4,test +5_stewart_0_5_5,test +5_stewart_0_65_65,test +5_stewart_0_66_66,train +5_stewart_0_6_6,test +5_stewart_0_73_73,test +5_stewart_0_74_74,train +5_stewart_0_7_7,test +5_stewart_0_81_81,test +5_stewart_0_82_82,train +5_stewart_0_87_87,test +5_stewart_0_88_88,train +5_stewart_0_8_8,test +5_stewart_0_95_95,test +5_stewart_0_96_96,train +5_stewart_0_9_9,val +5_stewart_1_1_1,additional +5_stewart_1_2_2,additional +5_stewart_1_3_3,additional +6_carla_0_100_100,train +6_carla_0_101_101,train +6_carla_0_102_102,train +6_carla_0_103_103,test +6_carla_0_104_104,train +6_carla_0_105_105,train +6_carla_0_106_106,train +6_carla_0_107_107,train +6_carla_0_108_108,train +6_carla_0_109_109,train +6_carla_0_110_110,train +6_carla_0_111_111,test +6_carla_0_112_112,train +6_carla_0_113_113,train +6_carla_0_114_114,train +6_carla_0_115_115,train +6_carla_0_116_116,train +6_carla_0_117_117,train +6_carla_0_118_118,train +6_carla_0_49_49,train +6_carla_0_50_50,train +6_carla_0_51_51,train +6_carla_0_52_52,train +6_carla_0_53_53,train +6_carla_0_54_54,train +6_carla_0_55_55,train +6_carla_0_56_56,train +6_carla_0_57_57,train +6_carla_0_58_58,train +6_carla_0_59_59,train +6_carla_0_60_60,train +6_carla_0_61_61,train +6_carla_0_62_62,train +6_carla_0_63_63,train +6_carla_0_64_64,train +6_carla_0_65_65,test +6_carla_0_66_66,train +6_carla_0_67_67,train +6_carla_0_68_68,train +6_carla_0_69_69,train +6_carla_0_70_70,train +6_carla_0_71_71,train +6_carla_0_72_72,train +6_carla_0_73_73,test +6_carla_0_74_74,train +6_carla_0_75_75,train +6_carla_0_76_76,train +6_carla_0_77_77,train +6_carla_0_78_78,train +6_carla_0_79_79,train +6_carla_0_80_80,train +6_carla_0_81_81,test +6_carla_0_82_82,train +6_carla_0_83_83,train +6_carla_0_84_84,train +6_carla_0_85_85,train +6_carla_0_86_86,train +6_carla_0_87_87,test +6_carla_0_88_88,train +6_carla_0_89_89,train +6_carla_0_90_90,train +6_carla_0_91_91,train +6_carla_0_92_92,train +6_carla_0_93_93,train +6_carla_0_94_94,train +6_carla_0_95_95,test +6_carla_0_96_96,train +6_carla_0_97_97,train +6_carla_0_98_98,train +6_carla_0_99_99,train +6_carla_1_10_10,additional +6_carla_1_11_11,additional +6_carla_1_12_12,additional +6_carla_1_1_1,additional +6_carla_1_2_2,additional +6_carla_1_3_3,additional +6_carla_1_4_4,additional +6_carla_1_5_5,additional +6_carla_1_6_6,additional +6_carla_1_7_7,additional +6_carla_1_8_8,additional +6_carla_1_9_9,additional +7_sophie_0_100_100,train +7_sophie_0_101_101,train +7_sophie_0_102_102,train +7_sophie_0_103_103,test +7_sophie_0_104_104,train +7_sophie_0_105_105,train +7_sophie_0_106_106,train +7_sophie_0_107_107,train +7_sophie_0_108_108,train +7_sophie_0_109_109,train +7_sophie_0_10_10,val +7_sophie_0_110_110,train +7_sophie_0_111_111,test +7_sophie_0_112_112,train +7_sophie_0_113_113,train +7_sophie_0_114_114,train +7_sophie_0_115_115,train +7_sophie_0_116_116,train +7_sophie_0_117_117,train +7_sophie_0_118_118,train +7_sophie_0_11_11,val +7_sophie_0_12_12,val +7_sophie_0_13_13,val +7_sophie_0_14_14,val +7_sophie_0_15_15,val +7_sophie_0_16_16,val +7_sophie_0_17_17,train +7_sophie_0_18_18,train +7_sophie_0_19_19,train +7_sophie_0_1_1,test +7_sophie_0_20_20,train +7_sophie_0_21_21,train +7_sophie_0_22_22,train +7_sophie_0_23_23,train +7_sophie_0_24_24,train +7_sophie_0_25_25,train +7_sophie_0_26_26,train +7_sophie_0_27_27,train +7_sophie_0_28_28,train +7_sophie_0_29_29,train +7_sophie_0_2_2,test +7_sophie_0_30_30,train +7_sophie_0_31_31,train +7_sophie_0_32_32,train +7_sophie_0_33_33,train +7_sophie_0_34_34,train +7_sophie_0_35_35,train +7_sophie_0_36_36,train +7_sophie_0_37_37,train +7_sophie_0_38_38,train +7_sophie_0_39_39,train +7_sophie_0_3_3,test +7_sophie_0_40_40,train +7_sophie_0_41_41,train +7_sophie_0_42_42,train +7_sophie_0_43_43,train +7_sophie_0_44_44,train +7_sophie_0_45_45,train +7_sophie_0_46_46,train +7_sophie_0_47_47,train +7_sophie_0_48_48,train +7_sophie_0_49_49,train +7_sophie_0_4_4,test +7_sophie_0_50_50,train +7_sophie_0_51_51,train +7_sophie_0_52_52,train +7_sophie_0_53_53,train +7_sophie_0_54_54,train +7_sophie_0_55_55,train +7_sophie_0_56_56,train +7_sophie_0_57_57,train +7_sophie_0_58_58,train +7_sophie_0_59_59,train +7_sophie_0_5_5,test +7_sophie_0_60_60,train +7_sophie_0_61_61,train +7_sophie_0_62_62,train +7_sophie_0_63_63,train +7_sophie_0_64_64,train +7_sophie_0_65_65,test +7_sophie_0_66_66,train +7_sophie_0_67_67,train +7_sophie_0_68_68,train +7_sophie_0_69_69,train +7_sophie_0_6_6,test +7_sophie_0_70_70,train +7_sophie_0_71_71,train +7_sophie_0_72_72,train +7_sophie_0_73_73,test +7_sophie_0_74_74,train +7_sophie_0_75_75,train +7_sophie_0_76_76,train +7_sophie_0_77_77,train +7_sophie_0_78_78,train +7_sophie_0_79_79,train +7_sophie_0_7_7,test +7_sophie_0_80_80,train +7_sophie_0_81_81,test +7_sophie_0_82_82,train +7_sophie_0_83_83,train +7_sophie_0_84_84,train +7_sophie_0_85_85,train +7_sophie_0_86_86,train +7_sophie_0_87_87,test +7_sophie_0_88_88,train +7_sophie_0_89_89,train +7_sophie_0_8_8,test +7_sophie_0_90_90,train +7_sophie_0_91_91,train +7_sophie_0_92_92,train +7_sophie_0_93_93,train +7_sophie_0_94_94,train +7_sophie_0_95_95,test +7_sophie_0_96_96,train +7_sophie_0_97_97,train +7_sophie_0_98_98,train +7_sophie_0_99_99,train +7_sophie_0_9_9,val +7_sophie_1_10_10,additional +7_sophie_1_11_11,additional +7_sophie_1_12_12,additional +7_sophie_1_1_1,additional +7_sophie_1_2_2,additional +7_sophie_1_3_3,additional +7_sophie_1_4_4,additional +7_sophie_1_5_5,additional +7_sophie_1_6_6,additional +7_sophie_1_7_7,additional +7_sophie_1_8_8,additional +7_sophie_1_9_9,additional +9_miranda_0_100_100,train +9_miranda_0_101_101,train +9_miranda_0_102_102,train +9_miranda_0_103_103,test +9_miranda_0_104_104,train +9_miranda_0_105_105,train +9_miranda_0_106_106,train +9_miranda_0_107_107,train +9_miranda_0_108_108,train +9_miranda_0_109_109,train +9_miranda_0_10_10,val +9_miranda_0_110_110,train +9_miranda_0_111_111,test +9_miranda_0_112_112,train +9_miranda_0_113_113,train +9_miranda_0_114_114,train +9_miranda_0_115_115,train +9_miranda_0_116_116,train +9_miranda_0_117_117,train +9_miranda_0_118_118,train +9_miranda_0_11_11,val +9_miranda_0_12_12,val +9_miranda_0_13_13,val +9_miranda_0_14_14,val +9_miranda_0_15_15,val +9_miranda_0_16_16,val +9_miranda_0_17_17,train +9_miranda_0_18_18,train +9_miranda_0_19_19,train +9_miranda_0_20_20,train +9_miranda_0_21_21,train +9_miranda_0_22_22,train +9_miranda_0_23_23,train +9_miranda_0_24_24,train +9_miranda_0_25_25,train +9_miranda_0_26_26,train +9_miranda_0_27_27,train +9_miranda_0_28_28,train +9_miranda_0_29_29,train +9_miranda_0_30_30,train +9_miranda_0_31_31,train +9_miranda_0_32_32,train +9_miranda_0_33_33,train +9_miranda_0_34_34,train +9_miranda_0_35_35,train +9_miranda_0_36_36,train +9_miranda_0_37_37,train +9_miranda_0_38_38,train +9_miranda_0_39_39,train +9_miranda_0_40_40,train +9_miranda_0_41_41,train +9_miranda_0_42_42,train +9_miranda_0_43_43,train +9_miranda_0_44_44,train +9_miranda_0_45_45,train +9_miranda_0_46_46,train +9_miranda_0_47_47,train +9_miranda_0_48_48,train +9_miranda_0_49_49,train +9_miranda_0_50_50,train +9_miranda_0_51_51,train +9_miranda_0_52_52,train +9_miranda_0_53_53,train +9_miranda_0_54_54,train +9_miranda_0_55_55,train +9_miranda_0_56_56,train +9_miranda_0_57_57,train +9_miranda_0_58_58,train +9_miranda_0_59_59,train +9_miranda_0_60_60,train +9_miranda_0_61_61,train +9_miranda_0_62_62,train +9_miranda_0_63_63,train +9_miranda_0_64_64,train +9_miranda_0_65_65,test +9_miranda_0_66_66,train +9_miranda_0_67_67,train +9_miranda_0_68_68,train +9_miranda_0_69_69,train +9_miranda_0_70_70,train +9_miranda_0_71_71,train +9_miranda_0_72_72,train +9_miranda_0_73_73,test +9_miranda_0_74_74,train +9_miranda_0_75_75,train +9_miranda_0_76_76,train +9_miranda_0_77_77,train +9_miranda_0_78_78,train +9_miranda_0_79_79,train +9_miranda_0_80_80,train +9_miranda_0_81_81,test +9_miranda_0_82_82,train +9_miranda_0_83_83,train +9_miranda_0_84_84,train +9_miranda_0_85_85,train +9_miranda_0_86_86,train +9_miranda_0_87_87,test +9_miranda_0_88_88,train +9_miranda_0_89_89,train +9_miranda_0_90_90,train +9_miranda_0_91_91,train +9_miranda_0_92_92,train +9_miranda_0_93_93,train +9_miranda_0_94_94,train +9_miranda_0_95_95,test +9_miranda_0_96_96,train +9_miranda_0_97_97,train +9_miranda_0_98_98,train +9_miranda_0_99_99,train +9_miranda_0_9_9,val +9_miranda_1_10_10,additional +9_miranda_1_11_11,additional +9_miranda_1_12_12,additional +9_miranda_1_1_1,additional +9_miranda_1_2_2,additional +9_miranda_1_3_3,additional +9_miranda_1_4_4,additional +9_miranda_1_5_5,additional +9_miranda_1_6_6,additional +9_miranda_1_7_7,additional +9_miranda_1_8_8,additional +9_miranda_1_9_9,additional \ No newline at end of file diff --git a/datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/weights/mean_vel_smplxflame_30.npy b/datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/weights/mean_vel_smplxflame_30.npy new file mode 100644 index 0000000000000000000000000000000000000000..0e4a6657c0f467876250b1258f2523756edbce57 --- /dev/null +++ b/datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/weights/mean_vel_smplxflame_30.npy @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d338491ce7d0da939cbb0ea326486d0b9c8654b3d33c7f6d32ca01e9a49ebf79 +size 348 diff --git a/datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/weights/vocab.pkl b/datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/weights/vocab.pkl new file mode 100644 index 0000000000000000000000000000000000000000..3daf14aa7f32d7823b19fd765ff5739bb9a1bd32 --- /dev/null +++ b/datasets/BEAT_SMPL/beat_v2.0.0/beat_english_v2.0.0/weights/vocab.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:54fbcea7b19e0ee9b5c5836c85087a682d3a9513041091ce3e95d83eed0b2acd +size 13821361 diff --git a/datasets/hub/pretrained_vq/face_vertex_1layer_790.bin b/datasets/hub/pretrained_vq/face_vertex_1layer_790.bin new file mode 100644 index 0000000000000000000000000000000000000000..fadc9adb218332a3eb34c0728eb4017be319a9b6 --- /dev/null +++ b/datasets/hub/pretrained_vq/face_vertex_1layer_790.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ce9240dcf08b1629855708c00d0aef116dccfdd35a0f6e2eb1d4aba430ae4ac9 +size 8935204 diff --git a/datasets/hub/smplx_models/smplx/SMPLX_NEUTRAL_2020.npz b/datasets/hub/smplx_models/smplx/SMPLX_NEUTRAL_2020.npz new file mode 100644 index 0000000000000000000000000000000000000000..998c299e1ea055a6091e87edb92774c888940cd7 --- /dev/null +++ b/datasets/hub/smplx_models/smplx/SMPLX_NEUTRAL_2020.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bdf06146e27d92022fe5dadad3b9203373f6879eca8e4d8235359ee3ec6a5a74 +size 167264530 diff --git a/demo/examples/2_scott_0_1_1.npz b/demo/examples/2_scott_0_1_1.npz new file mode 100644 index 0000000000000000000000000000000000000000..4d96616f0a09bd6df0971b9547d2e8376b078fcf --- /dev/null +++ b/demo/examples/2_scott_0_1_1.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:37b112fd59fcabb09270d6ca3c74e7459cc5b9729564bcacf1f75609f3999592 +size 2831524 diff --git a/demo/examples/2_scott_0_1_1.wav b/demo/examples/2_scott_0_1_1.wav new file mode 100644 index 0000000000000000000000000000000000000000..de87e147fcb09f5b7b52a2cda8b04a73fff00609 --- /dev/null +++ b/demo/examples/2_scott_0_1_1.wav @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b877ea7f3d19f26f0c252e00a798409e0eb6d2c63fa6d81f89ce5fda381e9354 +size 4102276 diff --git a/demo/examples/2_scott_0_2_2.wav b/demo/examples/2_scott_0_2_2.wav new file mode 100644 index 0000000000000000000000000000000000000000..7beef3a2e2a915bb377ee89cbb99d9621f0bac9b --- /dev/null +++ b/demo/examples/2_scott_0_2_2.wav @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fce4f54acf0dc67f75f7a726d30aef22358354c846da1e639ee6b283d621c95d +size 1984044 diff --git a/demo/examples/2_scott_0_3_3.wav b/demo/examples/2_scott_0_3_3.wav new file mode 100644 index 0000000000000000000000000000000000000000..c53883e47f0b7d5f7947fefaa9224e5cb126b83e --- /dev/null +++ b/demo/examples/2_scott_0_3_3.wav @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5f75277749bbd98d4059a0f9811720813b69c701b6be32432189d738b3d036a7 +size 2176044 diff --git a/demo/examples/2_scott_0_4_4.wav b/demo/examples/2_scott_0_4_4.wav new file mode 100644 index 0000000000000000000000000000000000000000..4baafea26febb0a2abcce94075dc7273a370d878 --- /dev/null +++ b/demo/examples/2_scott_0_4_4.wav @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:691e496ee2129dbce66a5fcfd87b7aa00d1f49caa61788cad9fab5bbca5cccf7 +size 2144044 diff --git a/demo/examples/2_scott_0_5_5.wav b/demo/examples/2_scott_0_5_5.wav new file mode 100644 index 0000000000000000000000000000000000000000..ec9f153ec0a867eb655381f67fdaa35ed9d465da --- /dev/null +++ b/demo/examples/2_scott_0_5_5.wav @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5de5920fe7427294c4f1912b054b858ff4cc75cc5015fcb63c6b25c9edc7179c +size 2464044 diff --git a/demo/install_mfa.sh b/demo/install_mfa.sh new file mode 100644 index 0000000000000000000000000000000000000000..34477a491e6b7d00f0fd8bbdd8f89caae94a4a5d --- /dev/null +++ b/demo/install_mfa.sh @@ -0,0 +1,3 @@ +conda install -c conda-forge kalpy +mfa model download acoustic english_us_arpa +mfa model download dictionary english_us_arpa \ No newline at end of file diff --git a/diffusion/fp16_util.py b/diffusion/fp16_util.py new file mode 100644 index 0000000000000000000000000000000000000000..1ccb93e4843b6257c3151b763356ef501f1acec8 --- /dev/null +++ b/diffusion/fp16_util.py @@ -0,0 +1,236 @@ +""" +Helpers to train with 16-bit precision. +""" + +import numpy as np +import torch as th +import torch.nn as nn +from torch._utils import _flatten_dense_tensors, _unflatten_dense_tensors + +from diffusion import logger + +INITIAL_LOG_LOSS_SCALE = 20.0 + + +def convert_module_to_f16(l): + """ + Convert primitive modules to float16. + """ + if isinstance(l, (nn.Conv1d, nn.Conv2d, nn.Conv3d)): + l.weight.data = l.weight.data.half() + if l.bias is not None: + l.bias.data = l.bias.data.half() + + +def convert_module_to_f32(l): + """ + Convert primitive modules to float32, undoing convert_module_to_f16(). + """ + if isinstance(l, (nn.Conv1d, nn.Conv2d, nn.Conv3d)): + l.weight.data = l.weight.data.float() + if l.bias is not None: + l.bias.data = l.bias.data.float() + + +def make_master_params(param_groups_and_shapes): + """ + Copy model parameters into a (differently-shaped) list of full-precision + parameters. + """ + master_params = [] + for param_group, shape in param_groups_and_shapes: + master_param = nn.Parameter( + _flatten_dense_tensors( + [param.detach().float() for (_, param) in param_group] + ).view(shape) + ) + master_param.requires_grad = True + master_params.append(master_param) + return master_params + + +def model_grads_to_master_grads(param_groups_and_shapes, master_params): + """ + Copy the gradients from the model parameters into the master parameters + from make_master_params(). + """ + for master_param, (param_group, shape) in zip( + master_params, param_groups_and_shapes + ): + master_param.grad = _flatten_dense_tensors( + [param_grad_or_zeros(param) for (_, param) in param_group] + ).view(shape) + + +def master_params_to_model_params(param_groups_and_shapes, master_params): + """ + Copy the master parameter data back into the model parameters. + """ + # Without copying to a list, if a generator is passed, this will + # silently not copy any parameters. + for master_param, (param_group, _) in zip(master_params, param_groups_and_shapes): + for (_, param), unflat_master_param in zip( + param_group, unflatten_master_params(param_group, master_param.view(-1)) + ): + param.detach().copy_(unflat_master_param) + + +def unflatten_master_params(param_group, master_param): + return _unflatten_dense_tensors(master_param, [param for (_, param) in param_group]) + + +def get_param_groups_and_shapes(named_model_params): + named_model_params = list(named_model_params) + scalar_vector_named_params = ( + [(n, p) for (n, p) in named_model_params if p.ndim <= 1], + (-1), + ) + matrix_named_params = ( + [(n, p) for (n, p) in named_model_params if p.ndim > 1], + (1, -1), + ) + return [scalar_vector_named_params, matrix_named_params] + + +def master_params_to_state_dict( + model, param_groups_and_shapes, master_params, use_fp16 +): + if use_fp16: + state_dict = model.state_dict() + for master_param, (param_group, _) in zip( + master_params, param_groups_and_shapes + ): + for (name, _), unflat_master_param in zip( + param_group, unflatten_master_params(param_group, master_param.view(-1)) + ): + assert name in state_dict + state_dict[name] = unflat_master_param + else: + state_dict = model.state_dict() + for i, (name, _value) in enumerate(model.named_parameters()): + assert name in state_dict + state_dict[name] = master_params[i] + return state_dict + + +def state_dict_to_master_params(model, state_dict, use_fp16): + if use_fp16: + named_model_params = [ + (name, state_dict[name]) for name, _ in model.named_parameters() + ] + param_groups_and_shapes = get_param_groups_and_shapes(named_model_params) + master_params = make_master_params(param_groups_and_shapes) + else: + master_params = [state_dict[name] for name, _ in model.named_parameters()] + return master_params + + +def zero_master_grads(master_params): + for param in master_params: + param.grad = None + + +def zero_grad(model_params): + for param in model_params: + # Taken from https://pytorch.org/docs/stable/_modules/torch/optim/optimizer.html#Optimizer.add_param_group + if param.grad is not None: + param.grad.detach_() + param.grad.zero_() + + +def param_grad_or_zeros(param): + if param.grad is not None: + return param.grad.data.detach() + else: + return th.zeros_like(param) + + +class MixedPrecisionTrainer: + def __init__( + self, + *, + model, + use_fp16=False, + fp16_scale_growth=1e-3, + initial_lg_loss_scale=INITIAL_LOG_LOSS_SCALE, + ): + self.model = model + self.use_fp16 = use_fp16 + self.fp16_scale_growth = fp16_scale_growth + + self.model_params = list(self.model.parameters()) + self.master_params = self.model_params + self.param_groups_and_shapes = None + self.lg_loss_scale = initial_lg_loss_scale + + if self.use_fp16: + self.param_groups_and_shapes = get_param_groups_and_shapes( + self.model.named_parameters() + ) + self.master_params = make_master_params(self.param_groups_and_shapes) + self.model.convert_to_fp16() + + def zero_grad(self): + zero_grad(self.model_params) + + def backward(self, loss: th.Tensor): + if self.use_fp16: + loss_scale = 2 ** self.lg_loss_scale + (loss * loss_scale).backward() + else: + loss.backward() + + def optimize(self, opt: th.optim.Optimizer): + if self.use_fp16: + return self._optimize_fp16(opt) + else: + return self._optimize_normal(opt) + + def _optimize_fp16(self, opt: th.optim.Optimizer): + logger.logkv_mean("lg_loss_scale", self.lg_loss_scale) + model_grads_to_master_grads(self.param_groups_and_shapes, self.master_params) + grad_norm, param_norm = self._compute_norms(grad_scale=2 ** self.lg_loss_scale) + if check_overflow(grad_norm): + self.lg_loss_scale -= 1 + logger.log(f"Found NaN, decreased lg_loss_scale to {self.lg_loss_scale}") + zero_master_grads(self.master_params) + return False + + logger.logkv_mean("grad_norm", grad_norm) + logger.logkv_mean("param_norm", param_norm) + + self.master_params[0].grad.mul_(1.0 / (2 ** self.lg_loss_scale)) + opt.step() + zero_master_grads(self.master_params) + master_params_to_model_params(self.param_groups_and_shapes, self.master_params) + self.lg_loss_scale += self.fp16_scale_growth + return True + + def _optimize_normal(self, opt: th.optim.Optimizer): + grad_norm, param_norm = self._compute_norms() + logger.logkv_mean("grad_norm", grad_norm) + logger.logkv_mean("param_norm", param_norm) + opt.step() + return True + + def _compute_norms(self, grad_scale=1.0): + grad_norm = 0.0 + param_norm = 0.0 + for p in self.master_params: + with th.no_grad(): + param_norm += th.norm(p, p=2, dtype=th.float32).item() ** 2 + if p.grad is not None: + grad_norm += th.norm(p.grad, p=2, dtype=th.float32).item() ** 2 + return np.sqrt(grad_norm) / grad_scale, np.sqrt(param_norm) + + def master_params_to_state_dict(self, master_params): + return master_params_to_state_dict( + self.model, self.param_groups_and_shapes, master_params, self.use_fp16 + ) + + def state_dict_to_master_params(self, state_dict): + return state_dict_to_master_params(self.model, state_dict, self.use_fp16) + + +def check_overflow(value): + return (value == float("inf")) or (value == -float("inf")) or (value != value) diff --git a/diffusion/gaussian_diffusion.py b/diffusion/gaussian_diffusion.py new file mode 100644 index 0000000000000000000000000000000000000000..e66adf9aee70197b4f2b9a3d6dff5b60c2ee83ff --- /dev/null +++ b/diffusion/gaussian_diffusion.py @@ -0,0 +1,1619 @@ +# This code is based on https://github.com/openai/guided-diffusion +""" +This code started out as a PyTorch port of Ho et al's diffusion models: +https://github.com/hojonathanho/diffusion/blob/1e0dceb3b3495bbe19116a5e1b3596cd0706c543/diffusion_tf/diffusion_utils_2.py + +Docstrings have been added, as well as DDIM sampling and a new collection of beta schedules. +""" + +import enum +import math +import pdb + +import numpy as np +import torch +import torch as th +from copy import deepcopy +from diffusion.nn import mean_flat, sum_flat +from diffusion.losses import normal_kl, discretized_gaussian_log_likelihood + +def get_named_beta_schedule(schedule_name, num_diffusion_timesteps, scale_betas=1.): + """ + Get a pre-defined beta schedule for the given name. + + The beta schedule library consists of beta schedules which remain similar + in the limit of num_diffusion_timesteps. + Beta schedules may be added, but should not be removed or changed once + they are committed to maintain backwards compatibility. + """ + if schedule_name == "linear": + # Linear schedule from Ho et al, extended to work for any number of + # diffusion steps. + scale = scale_betas * 1000 / num_diffusion_timesteps + beta_start = scale * 0.0001 + beta_end = scale * 0.02 + return np.linspace( + beta_start, beta_end, num_diffusion_timesteps, dtype=np.float64 + ) + elif schedule_name == "cosine": + return betas_for_alpha_bar( + num_diffusion_timesteps, + lambda t: math.cos((t + 0.008) / 1.008 * math.pi / 2) ** 2, + ) + else: + raise NotImplementedError(f"unknown beta schedule: {schedule_name}") + + +def betas_for_alpha_bar(num_diffusion_timesteps, alpha_bar, max_beta=0.999): + """ + Create a beta schedule that discretizes the given alpha_t_bar function, + which defines the cumulative product of (1-beta) over time from t = [0,1]. + + :param num_diffusion_timesteps: the number of betas to produce. + :param alpha_bar: a lambda that takes an argument t from 0 to 1 and + produces the cumulative product of (1-beta) up to that + part of the diffusion process. + :param max_beta: the maximum beta to use; use values lower than 1 to + prevent singularities. + """ + betas = [] + for i in range(num_diffusion_timesteps): + t1 = i / num_diffusion_timesteps + t2 = (i + 1) / num_diffusion_timesteps + betas.append(min(1 - alpha_bar(t2) / alpha_bar(t1), max_beta)) + return np.array(betas) + + +class ModelMeanType(enum.Enum): + """ + Which type of output the model predicts. + """ + + PREVIOUS_X = enum.auto() # the model predicts x_{t-1} + START_X = enum.auto() # the model predicts x_0 + EPSILON = enum.auto() # the model predicts epsilon + + +class ModelVarType(enum.Enum): + """ + What is used as the model's output variance. + + The LEARNED_RANGE option has been added to allow the model to predict + values between FIXED_SMALL and FIXED_LARGE, making its job easier. + """ + + LEARNED = enum.auto() + FIXED_SMALL = enum.auto() + FIXED_LARGE = enum.auto() + LEARNED_RANGE = enum.auto() + + +class LossType(enum.Enum): + MSE = enum.auto() # use raw MSE loss (and KL when learning variances) + RESCALED_MSE = ( + enum.auto() + ) # use raw MSE loss (with RESCALED_KL when learning variances) + KL = enum.auto() # use the variational lower-bound + RESCALED_KL = enum.auto() # like KL, but rescale to estimate the full VLB + + def is_vb(self): + return self == LossType.KL or self == LossType.RESCALED_KL + + +class GaussianDiffusion: + """ + Utilities for training and sampling diffusion models. + + Ported directly from here, and then adapted over time to further experimentation. + https://github.com/hojonathanho/diffusion/blob/1e0dceb3b3495bbe19116a5e1b3596cd0706c543/diffusion_tf/diffusion_utils_2.py#L42 + + :param betas: a 1-D numpy array of betas for each diffusion timestep, + starting at T and going to 1. + :param model_mean_type: a ModelMeanType determining what the model outputs. + :param model_var_type: a ModelVarType determining how variance is output. + :param loss_type: a LossType determining the loss function to use. + :param rescale_timesteps: if True, pass floating point timesteps into the + model so that they are always scaled like in the + original paper (0 to 1000). + """ + + def __init__( + self, + *, + betas, + model_mean_type, + model_var_type, + loss_type, + rescale_timesteps=False, + lambda_rcxyz=0., + lambda_vel=0., + lambda_pose=1., + lambda_orient=1., + lambda_loc=1., + data_rep='rot6d', + lambda_root_vel=0., + lambda_vel_rcxyz=0., + lambda_fc=0., + ): + self.model_mean_type = model_mean_type + self.model_var_type = model_var_type + self.loss_type = loss_type + self.rescale_timesteps = rescale_timesteps + self.data_rep = data_rep + + if data_rep != 'rot_vel' and lambda_pose != 1.: + raise ValueError('lambda_pose is relevant only when training on velocities!') + self.lambda_pose = lambda_pose + self.lambda_orient = lambda_orient + self.lambda_loc = lambda_loc + + self.lambda_rcxyz = lambda_rcxyz + self.lambda_vel = lambda_vel + self.lambda_root_vel = lambda_root_vel + self.lambda_vel_rcxyz = lambda_vel_rcxyz + self.lambda_fc = lambda_fc + + if self.lambda_rcxyz > 0. or self.lambda_vel > 0. or self.lambda_root_vel > 0. or \ + self.lambda_vel_rcxyz > 0. or self.lambda_fc > 0.: + assert self.loss_type == LossType.MSE, 'Geometric losses are supported by MSE loss type only!' + + # Use float64 for accuracy. + betas = np.array(betas, dtype=np.float64) + self.betas = betas + assert len(betas.shape) == 1, "betas must be 1-D" + assert (betas > 0).all() and (betas <= 1).all() + + self.num_timesteps = int(betas.shape[0]) + + alphas = 1.0 - betas + self.alphas_cumprod = np.cumprod(alphas, axis=0) + self.alphas_cumprod_prev = np.append(1.0, self.alphas_cumprod[:-1]) + self.alphas_cumprod_next = np.append(self.alphas_cumprod[1:], 0.0) + assert self.alphas_cumprod_prev.shape == (self.num_timesteps,) + + # calculations for diffusion q(x_t | x_{t-1}) and others + self.sqrt_alphas_cumprod = np.sqrt(self.alphas_cumprod) + self.sqrt_one_minus_alphas_cumprod = np.sqrt(1.0 - self.alphas_cumprod) + self.log_one_minus_alphas_cumprod = np.log(1.0 - self.alphas_cumprod) + self.sqrt_recip_alphas_cumprod = np.sqrt(1.0 / self.alphas_cumprod) + self.sqrt_recipm1_alphas_cumprod = np.sqrt(1.0 / self.alphas_cumprod - 1) + + # calculations for posterior q(x_{t-1} | x_t, x_0) + self.posterior_variance = ( + betas * (1.0 - self.alphas_cumprod_prev) / (1.0 - self.alphas_cumprod) + ) + # log calculation clipped because the posterior variance is 0 at the + # beginning of the diffusion chain. + self.posterior_log_variance_clipped = np.log( + np.append(self.posterior_variance[1], self.posterior_variance[1:]) + ) + self.posterior_mean_coef1 = ( + betas * np.sqrt(self.alphas_cumprod_prev) / (1.0 - self.alphas_cumprod) + ) + self.posterior_mean_coef2 = ( + (1.0 - self.alphas_cumprod_prev) + * np.sqrt(alphas) + / (1.0 - self.alphas_cumprod) + ) + + self.l2_loss = lambda a, b: (a - b) ** 2 # th.nn.MSELoss(reduction='none') # must be None for handling mask later on. + self.smooth_l1_loss = th.nn.SmoothL1Loss(reduction='none') + + def masked_l2(self, a, b, mask): + # assuming a.shape == b.shape == bs, J, Jdim, seqlen + # assuming mask.shape == bs, 1, 1, seqlen + # loss = self.l2_loss(a, b) # 20221217 + loss = self.smooth_l1_loss(a, b) + loss = sum_flat(loss * mask.float()) # gives \sigma_euclidean over unmasked elements + n_entries = a.shape[1] * a.shape[2] + non_zero_elements = sum_flat(mask) * n_entries + # print('mask', mask.shape) + # print('non_zero_elements', non_zero_elements) + # print('loss', loss) + mse_loss_val = loss / non_zero_elements + # print('mse_loss_val', mse_loss_val) + return mse_loss_val + + + def q_mean_variance(self, x_start, t): + """ + Get the distribution q(x_t | x_0). + + :param x_start: the [N x C x ...] tensor of noiseless inputs. + :param t: the number of diffusion steps (minus 1). Here, 0 means one step. + :return: A tuple (mean, variance, log_variance), all of x_start's shape. + """ + mean = ( + _extract_into_tensor(self.sqrt_alphas_cumprod, t, x_start.shape) * x_start + ) + variance = _extract_into_tensor(1.0 - self.alphas_cumprod, t, x_start.shape) + log_variance = _extract_into_tensor( + self.log_one_minus_alphas_cumprod, t, x_start.shape + ) + return mean, variance, log_variance + + def q_sample(self, x_start, t, noise=None): + """ + Diffuse the dataset for a given number of diffusion steps. + + In other words, sample from q(x_t | x_0). + + :param x_start: the initial dataset batch. + :param t: the number of diffusion steps (minus 1). Here, 0 means one step. + :param noise: if specified, the split-out normal noise. + :return: A noisy version of x_start. + """ + if noise is None: + noise = th.randn_like(x_start) + assert noise.shape == x_start.shape + return ( + _extract_into_tensor(self.sqrt_alphas_cumprod, t, x_start.shape) * x_start + + _extract_into_tensor(self.sqrt_one_minus_alphas_cumprod, t, x_start.shape) + * noise + ) + + def q_posterior_mean_variance(self, x_start, x_t, t): + """ + Compute the mean and variance of the diffusion posterior: + + q(x_{t-1} | x_t, x_0) + + """ + assert x_start.shape == x_t.shape + posterior_mean = ( + _extract_into_tensor(self.posterior_mean_coef1, t, x_t.shape) * x_start + + _extract_into_tensor(self.posterior_mean_coef2, t, x_t.shape) * x_t + ) + posterior_variance = _extract_into_tensor(self.posterior_variance, t, x_t.shape) + posterior_log_variance_clipped = _extract_into_tensor( + self.posterior_log_variance_clipped, t, x_t.shape + ) + assert ( + posterior_mean.shape[0] + == posterior_variance.shape[0] + == posterior_log_variance_clipped.shape[0] + == x_start.shape[0] + ) + return posterior_mean, posterior_variance, posterior_log_variance_clipped + + def p_mean_variance( + self, model, x, t, clip_denoised=True, denoised_fn=None, model_kwargs=None + ): + """ + Apply the model to get p(x_{t-1} | x_t), as well as a prediction of + the initial x, x_0. + + :param model: the model, which takes a signal and a batch of timesteps + as input. + :param x: the [N x C x ...] tensor at time t. + :param t: a 1-D Tensor of timesteps. + :param clip_denoised: if True, clip the denoised signal into [-1, 1]. + :param denoised_fn: if not None, a function which applies to the + x_start prediction before it is used to sample. Applies before + clip_denoised. + :param model_kwargs: if not None, a dict of extra keyword arguments to + pass to the model. This can be used for conditioning. + :return: a dict with the following keys: + - 'mean': the model mean output. + - 'variance': the model variance output. + - 'log_variance': the log of 'variance'. + - 'pred_xstart': the prediction for x_0. + """ + if model_kwargs is None: + model_kwargs = {} + + B, C = x.shape[:2] + assert t.shape == (B,) + model_output = model(x, self._scale_timesteps(t), **model_kwargs) + #Classifier-free guidence + ''' + model_output_uncond = model(x, self._scale_timesteps(t), **model_kwargs,uncond_info=True) + cfg_scale=torch.ones(1, device='cuda') * 3.0 + model_output_body=model_output_body_uncond + (cfg_scale.view(-1, 1, 1, 1) * (model_output_body - model_output_body_uncond)) + model_output_hand=model_output_hand_uncond + (cfg_scale.view(-1, 1, 1, 1) * (model_output_hand - model_output_hand_uncond)) + ''' + + if 'inpainting_mask' in model_kwargs['y'].keys() and 'inpainted_motion' in model_kwargs['y'].keys(): + inpainting_mask, inpainted_motion = model_kwargs['y']['inpainting_mask'], model_kwargs['y']['inpainted_motion'] + assert self.model_mean_type == ModelMeanType.START_X, 'This feature supports only X_start pred for mow!' + assert model_output.shape == inpainting_mask.shape == inpainted_motion.shape + model_output = (model_output * ~inpainting_mask) + (inpainted_motion * inpainting_mask) + # print('model_output', model_output.shape, model_output) + # print('inpainting_mask', inpainting_mask.shape, inpainting_mask[0,0,0,:]) + # print('inpainted_motion', inpainted_motion.shape, inpainted_motion) + + if self.model_var_type in [ModelVarType.LEARNED, ModelVarType.LEARNED_RANGE]: + assert model_output.shape == (B, C * 2, *x.shape[2:]) + model_output, model_var_values = th.split(model_output, C, dim=1) + if self.model_var_type == ModelVarType.LEARNED: + model_log_variance = model_var_values + model_variance = th.exp(model_log_variance) + else: + min_log = _extract_into_tensor( + self.posterior_log_variance_clipped, t, x.shape + ) + max_log = _extract_into_tensor(np.log(self.betas), t, x.shape) + # The model_var_values is [-1, 1] for [min_var, max_var]. + frac = (model_var_values + 1) / 2 + model_log_variance = frac * max_log + (1 - frac) * min_log + model_variance = th.exp(model_log_variance) + else: + model_variance, model_log_variance = { + # for fixedlarge, we set the initial (log-)variance like so + # to get a better decoder log likelihood. + ModelVarType.FIXED_LARGE: ( + np.append(self.posterior_variance[1], self.betas[1:]), + np.log(np.append(self.posterior_variance[1], self.betas[1:])), + ), + ModelVarType.FIXED_SMALL: ( + self.posterior_variance, + self.posterior_log_variance_clipped, + ), + }[self.model_var_type] + # print('model_variance', model_variance) + # print('model_log_variance',model_log_variance) + # print('self.posterior_variance', self.posterior_variance) + # print('self.posterior_log_variance_clipped', self.posterior_log_variance_clipped) + # print('self.model_var_type', self.model_var_type) + + + model_variance = _extract_into_tensor(model_variance, t, x.shape) + model_log_variance = _extract_into_tensor(model_log_variance, t, x.shape) + + def process_xstart(x): + if denoised_fn is not None: + x = denoised_fn(x) + if clip_denoised: + # print('clip_denoised', clip_denoised) + return x.clamp(-1, 1) + return x + + if self.model_mean_type == ModelMeanType.PREVIOUS_X: + pred_xstart = process_xstart( + self._predict_xstart_from_xprev(x_t=x, t=t, xprev=model_output) + ) + model_mean = model_output + elif self.model_mean_type in [ModelMeanType.START_X, ModelMeanType.EPSILON]: # THIS IS US! + if self.model_mean_type == ModelMeanType.START_X: + pred_xstart = process_xstart(model_output) + else: + pred_xstart = process_xstart( + self._predict_xstart_from_eps(x_t=x, t=t, eps=model_output) + ) + model_mean, _, _ = self.q_posterior_mean_variance( + x_start=pred_xstart, x_t=x, t=t + ) + else: + raise NotImplementedError(self.model_mean_type) + + assert ( + model_mean.shape == model_log_variance.shape == pred_xstart.shape == x.shape + ) + return { + "mean": model_mean, + "variance": model_variance, + "log_variance": model_log_variance, + "pred_xstart": pred_xstart, + } + + def _predict_xstart_from_eps(self, x_t, t, eps): + assert x_t.shape == eps.shape + return ( + _extract_into_tensor(self.sqrt_recip_alphas_cumprod, t, x_t.shape) * x_t + - _extract_into_tensor(self.sqrt_recipm1_alphas_cumprod, t, x_t.shape) * eps + ) + + def _predict_xstart_from_xprev(self, x_t, t, xprev): + assert x_t.shape == xprev.shape + return ( # (xprev - coef2*x_t) / coef1 + _extract_into_tensor(1.0 / self.posterior_mean_coef1, t, x_t.shape) * xprev + - _extract_into_tensor( + self.posterior_mean_coef2 / self.posterior_mean_coef1, t, x_t.shape + ) + * x_t + ) + + def _predict_eps_from_xstart(self, x_t, t, pred_xstart): + return ( + _extract_into_tensor(self.sqrt_recip_alphas_cumprod, t, x_t.shape) * x_t + - pred_xstart + ) / _extract_into_tensor(self.sqrt_recipm1_alphas_cumprod, t, x_t.shape) + + def _scale_timesteps(self, t): + if self.rescale_timesteps: + return t.float() * (1000.0 / self.num_timesteps) + return t + + def condition_mean(self, cond_fn, p_mean_var, x, t, model_kwargs=None): + """ + Compute the mean for the previous step, given a function cond_fn that + computes the gradient of a conditional log probability with respect to + x. In particular, cond_fn computes grad(log(p(y|x))), and we want to + condition on y. + + This uses the conditioning strategy from Sohl-Dickstein et al. (2015). + """ + gradient = cond_fn(x, self._scale_timesteps(t), **model_kwargs) + new_mean = ( + p_mean_var["mean"].float() + p_mean_var["variance"] * gradient.float() + ) + return new_mean + + def condition_mean_with_grad(self, cond_fn, p_mean_var, x, t, model_kwargs=None): + """ + Compute the mean for the previous step, given a function cond_fn that + computes the gradient of a conditional log probability with respect to + x. In particular, cond_fn computes grad(log(p(y|x))), and we want to + condition on y. + + This uses the conditioning strategy from Sohl-Dickstein et al. (2015). + """ + gradient = cond_fn(x, t, p_mean_var, **model_kwargs) + new_mean = ( + p_mean_var["mean"].float() + p_mean_var["variance"] * gradient.float() + ) + return new_mean + + def condition_score(self, cond_fn, p_mean_var, x, t, model_kwargs=None): + """ + Compute what the p_mean_variance output would have been, should the + model's score function be conditioned by cond_fn. + + See condition_mean() for details on cond_fn. + + Unlike condition_mean(), this instead uses the conditioning strategy + from Song et al (2020). + """ + alpha_bar = _extract_into_tensor(self.alphas_cumprod, t, x.shape) + + eps = self._predict_eps_from_xstart(x, t, p_mean_var["pred_xstart"]) + eps = eps - (1 - alpha_bar).sqrt() * cond_fn( + x, self._scale_timesteps(t), **model_kwargs + ) + + out = p_mean_var.copy() + out["pred_xstart"] = self._predict_xstart_from_eps(x, t, eps) + out["mean"], _, _ = self.q_posterior_mean_variance( + x_start=out["pred_xstart"], x_t=x, t=t + ) + return out + + def condition_score_with_grad(self, cond_fn, p_mean_var, x, t, model_kwargs=None): + """ + Compute what the p_mean_variance output would have been, should the + model's score function be conditioned by cond_fn. + + See condition_mean() for details on cond_fn. + + Unlike condition_mean(), this instead uses the conditioning strategy + from Song et al (2020). + """ + alpha_bar = _extract_into_tensor(self.alphas_cumprod, t, x.shape) + + eps = self._predict_eps_from_xstart(x, t, p_mean_var["pred_xstart"]) + eps = eps - (1 - alpha_bar).sqrt() * cond_fn( + x, t, p_mean_var, **model_kwargs + ) + + out = p_mean_var.copy() + out["pred_xstart"] = self._predict_xstart_from_eps(x, t, eps) + out["mean"], _, _ = self.q_posterior_mean_variance( + x_start=out["pred_xstart"], x_t=x, t=t + ) + return out + + def p_sample( + self, + model, + x, + t, + clip_denoised=True, + denoised_fn=None, + cond_fn=None, + model_kwargs=None, + const_noise=False, + ): + """ + Sample x_{t-1} from the model at the given timestep. + + :param model: the model to sample from. + :param x: the current tensor at x_{t-1}. + :param t: the value of t, starting at 0 for the first diffusion step. + :param clip_denoised: if True, clip the x_start prediction to [-1, 1]. + :param denoised_fn: if not None, a function which applies to the + x_start prediction before it is used to sample. + :param cond_fn: if not None, this is a gradient function that acts + similarly to the model. + :param model_kwargs: if not None, a dict of extra keyword arguments to + pass to the model. This can be used for conditioning. + :return: a dict containing the following keys: + - 'sample': a random sample from the model. + - 'pred_xstart': a prediction of x_0. + """ + out = self.p_mean_variance( + model, + x, + t, + clip_denoised=clip_denoised, + denoised_fn=denoised_fn, + model_kwargs=model_kwargs, + ) # 'mean' （1, 135, 1, 240）, 'variance', 'log_variance', 'pred_xstart' + noise = th.randn_like(x) + # print('const_noise', const_noise) + if const_noise: + noise = noise[[0]].repeat(x.shape[0], 1, 1, 1) + + nonzero_mask = ( + (t != 0).float().view(-1, *([1] * (len(x.shape) - 1))) + ) # no noise when t == 0 + if cond_fn is not None: + out["mean"] = self.condition_mean( + cond_fn, out, x, t, model_kwargs=model_kwargs + ) + # print('mean', out["mean"].shape, out["mean"]) + # print('log_variance', out["log_variance"].shape, out["log_variance"]) + # print('nonzero_mask', nonzero_mask.shape, nonzero_mask) + sample = out["mean"] + nonzero_mask * th.exp(0.5 * out["log_variance"]) * noise + return {"sample": sample, "pred_xstart": out["pred_xstart"]} + + def p_sample_with_grad( + self, + model, + x, + t, + clip_denoised=True, + denoised_fn=None, + cond_fn=None, + model_kwargs=None, + ): + """ + Sample x_{t-1} from the model at the given timestep. + + :param model: the model to sample from. + :param x: the current tensor at x_{t-1}. + :param t: the value of t, starting at 0 for the first diffusion step. + :param clip_denoised: if True, clip the x_start prediction to [-1, 1]. + :param denoised_fn: if not None, a function which applies to the + x_start prediction before it is used to sample. + :param cond_fn: if not None, this is a gradient function that acts + similarly to the model. + :param model_kwargs: if not None, a dict of extra keyword arguments to + pass to the model. This can be used for conditioning. + :return: a dict containing the following keys: + - 'sample': a random sample from the model. + - 'pred_xstart': a prediction of x_0. + """ + with th.enable_grad(): + x = x.detach().requires_grad_() + out = self.p_mean_variance( + model, + x, + t, + clip_denoised=clip_denoised, + denoised_fn=denoised_fn, + model_kwargs=model_kwargs, + ) + noise = th.randn_like(x) + nonzero_mask = ( + (t != 0).float().view(-1, *([1] * (len(x.shape) - 1))) + ) # no noise when t == 0 + if cond_fn is not None: + out["mean"] = self.condition_mean_with_grad( + cond_fn, out, x, t, model_kwargs=model_kwargs + ) + sample = out["mean"] + nonzero_mask * th.exp(0.5 * out["log_variance"]) * noise + return {"sample": sample, "pred_xstart": out["pred_xstart"].detach()} + + def p_sample_loop( + self, + model, + shape, + noise=None, + clip_denoised=True, + denoised_fn=None, + cond_fn=None, + model_kwargs=None, + device=None, + progress=False, + skip_timesteps=0, + init_image=None, + randomize_class=False, + cond_fn_with_grad=False, + dump_steps=None, + const_noise=False, + ): + """ + Generate samples from the model. + + :param model: the model module. + :param shape: the shape of the samples, (N, C, H, W). + :param noise: if specified, the noise from the encoder to sample. + Should be of the same shape as `shape`. + :param clip_denoised: if True, clip x_start predictions to [-1, 1]. + :param denoised_fn: if not None, a function which applies to the + x_start prediction before it is used to sample. + :param cond_fn: if not None, this is a gradient function that acts + similarly to the model. + :param model_kwargs: if not None, a dict of extra keyword arguments to + pass to the model. This can be used for conditioning. + :param device: if specified, the device to create the samples on. + If not specified, use a model parameter's device. + :param progress: if True, show a tqdm progress bar. + :param const_noise: If True, will noise all samples with the same noise throughout sampling + :return: a non-differentiable batch of samples. + """ + final = None + if dump_steps is not None: + dump = [] + + for i, sample in enumerate(self.p_sample_loop_progressive( + model, + shape, + noise=noise, + clip_denoised=clip_denoised, + denoised_fn=denoised_fn, + cond_fn=cond_fn, + model_kwargs=model_kwargs, + device=device, + progress=progress, + skip_timesteps=skip_timesteps, + init_image=init_image, + randomize_class=randomize_class, + cond_fn_with_grad=cond_fn_with_grad, + const_noise=const_noise, + )): + if dump_steps is not None and i in dump_steps: + dump.append(deepcopy(sample["sample"])) + final = sample + if dump_steps is not None: + return dump + return final["sample"] + + def p_sample_loop_progressive( + self, + model, + shape, + noise=None, + clip_denoised=True, + denoised_fn=None, + cond_fn=None, + model_kwargs=None, + device=None, + progress=False, + skip_timesteps=0, + init_image=None, + randomize_class=False, + cond_fn_with_grad=False, + const_noise=False, + ): + """ + Generate samples from the model and yield intermediate samples from + each timestep of diffusion. + + Arguments are the same as p_sample_loop(). + Returns a generator over dicts, where each dict is the return value of + p_sample(). + """ + if device is None: + device = next(model.parameters()).device + assert isinstance(shape, (tuple, list)) + if noise is not None: + img = noise + else: + img = th.randn(*shape, device=device) + + if skip_timesteps and init_image is None: + init_image = th.zeros_like(img) + + indices = list(range(self.num_timesteps - skip_timesteps))[::-1] + + if init_image is not None: + my_t = th.ones([shape[0]], device=device, dtype=th.long) * indices[0] + img = self.q_sample(init_image, my_t, img) + + if progress: + # Lazy import so that we don't depend on tqdm. + from tqdm.auto import tqdm + + indices = tqdm(indices) + + for i in indices: + t = th.tensor([i] * shape[0], device=device) + if randomize_class and 'y' in model_kwargs: + model_kwargs['y'] = th.randint(low=0, high=model.num_classes, + size=model_kwargs['y'].shape, + device=model_kwargs['y'].device) + with th.no_grad(): + sample_fn = self.p_sample_with_grad if cond_fn_with_grad else self.p_sample + out = sample_fn( + model, + img, + t, + clip_denoised=clip_denoised, + denoised_fn=denoised_fn, + cond_fn=cond_fn, + model_kwargs=model_kwargs, + const_noise=const_noise, + ) + yield out + img = out["sample"] + + def ddim_sample( + self, + model, + x, + t, + clip_denoised=True, + denoised_fn=None, + cond_fn=None, + model_kwargs=None, + eta=0.0, + ): + """ + Sample x_{t-1} from the model using DDIM. + + Same usage as p_sample(). + """ + out_orig = self.p_mean_variance( + model, + x, + t, + clip_denoised=clip_denoised, + denoised_fn=denoised_fn, + model_kwargs=model_kwargs, + ) + if cond_fn is not None: + out = self.condition_score(cond_fn, out_orig, x, t, model_kwargs=model_kwargs) + else: + out = out_orig + + # Usually our model outputs epsilon, but we re-derive it + # in case we used x_start or x_prev prediction. + eps = self._predict_eps_from_xstart(x, t, out["pred_xstart"]) + + alpha_bar = _extract_into_tensor(self.alphas_cumprod, t, x.shape) + alpha_bar_prev = _extract_into_tensor(self.alphas_cumprod_prev, t, x.shape) + sigma = ( + eta + * th.sqrt((1 - alpha_bar_prev) / (1 - alpha_bar)) + * th.sqrt(1 - alpha_bar / alpha_bar_prev) + ) + # Equation 12. + noise = th.randn_like(x) + mean_pred = ( + out["pred_xstart"] * th.sqrt(alpha_bar_prev) + + th.sqrt(1 - alpha_bar_prev - sigma ** 2) * eps + ) + nonzero_mask = ( + (t != 0).float().view(-1, *([1] * (len(x.shape) - 1))) + ) # no noise when t == 0 + sample = mean_pred + nonzero_mask * sigma * noise + return {"sample": sample, "pred_xstart": out_orig["pred_xstart"]} + + def ddim_sample_with_grad( + self, + model, + x, + t, + clip_denoised=True, + denoised_fn=None, + cond_fn=None, + model_kwargs=None, + eta=0.0, + ): + """ + Sample x_{t-1} from the model using DDIM. + + Same usage as p_sample(). + """ + with th.enable_grad(): + x = x.detach().requires_grad_() + out_orig = self.p_mean_variance( + model, + x, + t, + clip_denoised=clip_denoised, + denoised_fn=denoised_fn, + model_kwargs=model_kwargs, + ) + if cond_fn is not None: + out = self.condition_score_with_grad(cond_fn, out_orig, x, t, + model_kwargs=model_kwargs) + else: + out = out_orig + + out["pred_xstart"] = out["pred_xstart"].detach() + + # Usually our model outputs epsilon, but we re-derive it + # in case we used x_start or x_prev prediction. + eps = self._predict_eps_from_xstart(x, t, out["pred_xstart"]) + + alpha_bar = _extract_into_tensor(self.alphas_cumprod, t, x.shape) + alpha_bar_prev = _extract_into_tensor(self.alphas_cumprod_prev, t, x.shape) + sigma = ( + eta + * th.sqrt((1 - alpha_bar_prev) / (1 - alpha_bar)) + * th.sqrt(1 - alpha_bar / alpha_bar_prev) + ) + # Equation 12. + noise = th.randn_like(x) + mean_pred = ( + out["pred_xstart"] * th.sqrt(alpha_bar_prev) + + th.sqrt(1 - alpha_bar_prev - sigma ** 2) * eps + ) + nonzero_mask = ( + (t != 0).float().view(-1, *([1] * (len(x.shape) - 1))) + ) # no noise when t == 0 + sample = mean_pred + nonzero_mask * sigma * noise + return {"sample": sample, "pred_xstart": out_orig["pred_xstart"].detach()} + + def ddim_reverse_sample( + self, + model, + x, + t, + clip_denoised=True, + denoised_fn=None, + model_kwargs=None, + eta=0.0, + ): + """ + Sample x_{t+1} from the model using DDIM reverse ODE. + """ + assert eta == 0.0, "Reverse ODE only for deterministic path" + out = self.p_mean_variance( + model, + x, + t, + clip_denoised=clip_denoised, + denoised_fn=denoised_fn, + model_kwargs=model_kwargs, + ) + # Usually our model outputs epsilon, but we re-derive it + # in case we used x_start or x_prev prediction. + eps = ( + _extract_into_tensor(self.sqrt_recip_alphas_cumprod, t, x.shape) * x + - out["pred_xstart"] + ) / _extract_into_tensor(self.sqrt_recipm1_alphas_cumprod, t, x.shape) + alpha_bar_next = _extract_into_tensor(self.alphas_cumprod_next, t, x.shape) + + # Equation 12. reversed + mean_pred = ( + out["pred_xstart"] * th.sqrt(alpha_bar_next) + + th.sqrt(1 - alpha_bar_next) * eps + ) + + return {"sample": mean_pred, "pred_xstart": out["pred_xstart"]} + + def ddim_sample_loop( + self, + model, + shape, + noise=None, + clip_denoised=True, + denoised_fn=None, + cond_fn=None, + model_kwargs=None, + device=None, + progress=False, + eta=0.0, + skip_timesteps=0, + init_image=None, + randomize_class=False, + cond_fn_with_grad=False, + dump_steps=None, + const_noise=False, + ): + """ + Generate samples from the model using DDIM. + + Same usage as p_sample_loop(). + """ + if dump_steps is not None: + raise NotImplementedError() + if const_noise == True: + raise NotImplementedError() + + final = None + for sample in self.ddim_sample_loop_progressive( + model, + shape, + noise=noise, + clip_denoised=clip_denoised, + denoised_fn=denoised_fn, + cond_fn=cond_fn, + model_kwargs=model_kwargs, + device=device, + progress=progress, + eta=eta, + skip_timesteps=skip_timesteps, + init_image=init_image, + randomize_class=randomize_class, + cond_fn_with_grad=cond_fn_with_grad, + ): + final = sample + return final["sample"] + + def ddim_sample_loop_progressive( + self, + model, + shape, + noise=None, + clip_denoised=True, + denoised_fn=None, + cond_fn=None, + model_kwargs=None, + device=None, + progress=False, + eta=0.0, + skip_timesteps=0, + init_image=None, + randomize_class=False, + cond_fn_with_grad=False, + ): + """ + Use DDIM to sample from the model and yield intermediate samples from + each timestep of DDIM. + + Same usage as p_sample_loop_progressive(). + """ + if device is None: + device = next(model.parameters()).device + assert isinstance(shape, (tuple, list)) + if noise is not None: + img = noise + else: + img = th.randn(*shape, device=device) + + if skip_timesteps and init_image is None: + init_image = th.zeros_like(img) + + indices = list(range(self.num_timesteps - skip_timesteps))[::-1] + + if init_image is not None: + my_t = th.ones([shape[0]], device=device, dtype=th.long) * indices[0] + img = self.q_sample(init_image, my_t, img) + + if progress: + # Lazy import so that we don't depend on tqdm. + from tqdm.auto import tqdm + + indices = tqdm(indices) + + for i in indices: + t = th.tensor([i] * shape[0], device=device) + if randomize_class and 'y' in model_kwargs: + model_kwargs['y'] = th.randint(low=0, high=model.num_classes, + size=model_kwargs['y'].shape, + device=model_kwargs['y'].device) + with th.no_grad(): + sample_fn = self.ddim_sample_with_grad if cond_fn_with_grad else self.ddim_sample + out = sample_fn( + model, + img, + t, + clip_denoised=clip_denoised, + denoised_fn=denoised_fn, + cond_fn=cond_fn, + model_kwargs=model_kwargs, + eta=eta, + ) + yield out + img = out["sample"] + + def plms_sample( + self, + model, + x, + t, + clip_denoised=True, + denoised_fn=None, + cond_fn=None, + model_kwargs=None, + cond_fn_with_grad=False, + order=2, + old_out=None, + ): + """ + Sample x_{t-1} from the model using Pseudo Linear Multistep. + + Same usage as p_sample(). + """ + if not int(order) or not 1 <= order <= 4: + raise ValueError('order is invalid (should be int from 1-4).') + + def get_model_output(x, t): + with th.set_grad_enabled(cond_fn_with_grad and cond_fn is not None): + x = x.detach().requires_grad_() if cond_fn_with_grad else x + out_orig = self.p_mean_variance( + model, + x, + t, + clip_denoised=clip_denoised, + denoised_fn=denoised_fn, + model_kwargs=model_kwargs, + ) + if cond_fn is not None: + if cond_fn_with_grad: + out = self.condition_score_with_grad(cond_fn, out_orig, x, t, model_kwargs=model_kwargs) + x = x.detach() + else: + out = self.condition_score(cond_fn, out_orig, x, t, model_kwargs=model_kwargs) + else: + out = out_orig + + # Usually our model outputs epsilon, but we re-derive it + # in case we used x_start or x_prev prediction. + eps = self._predict_eps_from_xstart(x, t, out["pred_xstart"]) + return eps, out, out_orig + + alpha_bar = _extract_into_tensor(self.alphas_cumprod, t, x.shape) + alpha_bar_prev = _extract_into_tensor(self.alphas_cumprod_prev, t, x.shape) + eps, out, out_orig = get_model_output(x, t) + + if order > 1 and old_out is None: + # Pseudo Improved Euler + old_eps = [eps] + mean_pred = out["pred_xstart"] * th.sqrt(alpha_bar_prev) + th.sqrt(1 - alpha_bar_prev) * eps + eps_2, _, _ = get_model_output(mean_pred, t - 1) + eps_prime = (eps + eps_2) / 2 + pred_prime = self._predict_xstart_from_eps(x, t, eps_prime) + mean_pred = pred_prime * th.sqrt(alpha_bar_prev) + th.sqrt(1 - alpha_bar_prev) * eps_prime + else: + # Pseudo Linear Multistep (Adams-Bashforth) + old_eps = old_out["old_eps"] + old_eps.append(eps) + cur_order = min(order, len(old_eps)) + if cur_order == 1: + eps_prime = old_eps[-1] + elif cur_order == 2: + eps_prime = (3 * old_eps[-1] - old_eps[-2]) / 2 + elif cur_order == 3: + eps_prime = (23 * old_eps[-1] - 16 * old_eps[-2] + 5 * old_eps[-3]) / 12 + elif cur_order == 4: + eps_prime = (55 * old_eps[-1] - 59 * old_eps[-2] + 37 * old_eps[-3] - 9 * old_eps[-4]) / 24 + else: + raise RuntimeError('cur_order is invalid.') + pred_prime = self._predict_xstart_from_eps(x, t, eps_prime) + mean_pred = pred_prime * th.sqrt(alpha_bar_prev) + th.sqrt(1 - alpha_bar_prev) * eps_prime + + if len(old_eps) >= order: + old_eps.pop(0) + + nonzero_mask = (t != 0).float().view(-1, *([1] * (len(x.shape) - 1))) + sample = mean_pred * nonzero_mask + out["pred_xstart"] * (1 - nonzero_mask) + + return {"sample": sample, "pred_xstart": out_orig["pred_xstart"], "old_eps": old_eps} + + def plms_sample_loop( + self, + model, + shape, + noise=None, + clip_denoised=True, + denoised_fn=None, + cond_fn=None, + model_kwargs=None, + device=None, + progress=False, + skip_timesteps=0, + init_image=None, + randomize_class=False, + cond_fn_with_grad=False, + order=2, + ): + """ + Generate samples from the model using Pseudo Linear Multistep. + + Same usage as p_sample_loop(). + """ + final = None + for sample in self.plms_sample_loop_progressive( + model, + shape, + noise=noise, + clip_denoised=clip_denoised, + denoised_fn=denoised_fn, + cond_fn=cond_fn, + model_kwargs=model_kwargs, + device=device, + progress=progress, + skip_timesteps=skip_timesteps, + init_image=init_image, + randomize_class=randomize_class, + cond_fn_with_grad=cond_fn_with_grad, + order=order, + ): + final = sample + return final["sample"] + + def plms_sample_loop_progressive( + self, + model, + shape, + noise=None, + clip_denoised=True, + denoised_fn=None, + cond_fn=None, + model_kwargs=None, + device=None, + progress=False, + skip_timesteps=0, + init_image=None, + randomize_class=False, + cond_fn_with_grad=False, + order=2, + ): + """ + Use PLMS to sample from the model and yield intermediate samples from each + timestep of PLMS. + + Same usage as p_sample_loop_progressive(). + """ + if device is None: + device = next(model.parameters()).device + assert isinstance(shape, (tuple, list)) + if noise is not None: + img = noise + else: + img = th.randn(*shape, device=device) + + if skip_timesteps and init_image is None: + init_image = th.zeros_like(img) + + indices = list(range(self.num_timesteps - skip_timesteps))[::-1] + + if init_image is not None: + my_t = th.ones([shape[0]], device=device, dtype=th.long) * indices[0] + img = self.q_sample(init_image, my_t, img) + + if progress: + # Lazy import so that we don't depend on tqdm. + from tqdm.auto import tqdm + + indices = tqdm(indices) + + old_out = None + + for i in indices: + t = th.tensor([i] * shape[0], device=device) + if randomize_class and 'y' in model_kwargs: + model_kwargs['y'] = th.randint(low=0, high=model.num_classes, + size=model_kwargs['y'].shape, + device=model_kwargs['y'].device) + with th.no_grad(): + out = self.plms_sample( + model, + img, + t, + clip_denoised=clip_denoised, + denoised_fn=denoised_fn, + cond_fn=cond_fn, + model_kwargs=model_kwargs, + cond_fn_with_grad=cond_fn_with_grad, + order=order, + old_out=old_out, + ) + yield out + old_out = out + img = out["sample"] + + def _vb_terms_bpd( + self, model, x_start, x_t, t, clip_denoised=True, model_kwargs=None + ): + """ + Get a term for the variational lower-bound. + + The resulting units are bits (rather than nats, as one might expect). + This allows for comparison to other papers. + + :return: a dict with the following keys: + - 'output': a shape [N] tensor of NLLs or KLs. + - 'pred_xstart': the x_0 predictions. + """ + true_mean, _, true_log_variance_clipped = self.q_posterior_mean_variance( + x_start=x_start, x_t=x_t, t=t + ) + out = self.p_mean_variance( + model, x_t, t, clip_denoised=clip_denoised, model_kwargs=model_kwargs + ) + kl = normal_kl( + true_mean, true_log_variance_clipped, out["mean"], out["log_variance"] + ) + kl = mean_flat(kl) / np.log(2.0) + + decoder_nll = -discretized_gaussian_log_likelihood( + x_start, means=out["mean"], log_scales=0.5 * out["log_variance"] + ) + assert decoder_nll.shape == x_start.shape + decoder_nll = mean_flat(decoder_nll) / np.log(2.0) + + # At the first timestep return the decoder NLL, + # otherwise return KL(q(x_{t-1}|x_t,x_0) || p(x_{t-1}|x_t)) + output = th.where((t == 0), decoder_nll, kl) + return {"output": output, "pred_xstart": out["pred_xstart"]} + + def training_losses(self, model, x_start, t, model_kwargs=None, noise=None, dataset=None): + """ + Compute training losses for a single timestep. + + :param model: the model to evaluate loss on. + :param x_start: the [N x C x ...] tensor of inputs. + :param t: a batch of timestep indices. + :param model_kwargs: if not None, a dict of extra keyword arguments to + pass to the model. This can be used for conditioning. + :param noise: if specified, the specific Gaussian noise to try to remove. + :return: a dict with the key "loss" containing a tensor of shape [N]. + Some mean or variance settings may also have other keys. + """ + + # enc = model.model._modules['module'] + enc = model.model + mask = model_kwargs['y']['mask'] + # get_xyz = lambda sample: enc.rot2xyz(sample, mask=None, pose_rep=enc.pose_rep, translation=enc.translation, + # glob=enc.glob, + # # jointstype='vertices', # 3.4 iter/sec # USED ALSO IN MotionCLIP + # jointstype='smpl', # 3.4 iter/sec + # vertstrans=False) + + if model_kwargs is None: + model_kwargs = {} + if noise is None: + noise = th.randn_like(x_start) + x_t = self.q_sample(x_start, t, noise=noise) # torch.Size([64, 251, 1, 196]), add noisy + + terms = {} + + if self.loss_type == LossType.KL or self.loss_type == LossType.RESCALED_KL: # LossType.MSE + terms["loss"] = self._vb_terms_bpd( + model=model, + x_start=x_start, + x_t=x_t, + t=t, + clip_denoised=False, + model_kwargs=model_kwargs, + )["output"] + if self.loss_type == LossType.RESCALED_KL: + terms["loss"] *= self.num_timesteps + elif self.loss_type == LossType.MSE or self.loss_type == LossType.RESCALED_MSE: + model_output = model(x_t, self._scale_timesteps(t), **model_kwargs) + + if self.model_var_type in [ # ModelVarType.FIXED_SMALL: 2 + ModelVarType.LEARNED, + ModelVarType.LEARNED_RANGE, + ]: + B, C = x_t.shape[:2] + assert model_output.shape == (B, C * 2, *x_t.shape[2:]) + model_output, model_var_values = th.split(model_output, C, dim=1) + # Learn the variance using the variational bound, but don't let + # it affect our mean prediction. + frozen_out = th.cat([model_output.detach(), model_var_values], dim=1) + terms["vb"] = self._vb_terms_bpd( + model=lambda *args, r=frozen_out: r, + x_start=x_start, + x_t=x_t, + t=t, + clip_denoised=False, + )["output"] + if self.loss_type == LossType.RESCALED_MSE: + # Divide by 1000 for equivalence with initial implementation. + # Without a factor of 1/1000, the VB term hurts the MSE term. + terms["vb"] *= self.num_timesteps / 1000.0 + + target = { + ModelMeanType.PREVIOUS_X: self.q_posterior_mean_variance( + x_start=x_start, x_t=x_t, t=t + )[0], + ModelMeanType.START_X: x_start, + ModelMeanType.EPSILON: noise, + }[self.model_mean_type] # ModelMeanType.START_X: 2 + assert model_output.shape == target.shape == x_start.shape # [bs, njoints, nfeats, nframes] + + # pdb.set_trace() # target (2, 135, 1, 240) + + terms["rot_mse"] = self.masked_l2(target, model_output, mask) # mean_flat(rot_mse) # [64, 251, 1, 196], -, [64, 1, 1, 196] + + target_xyz, model_output_xyz = None, None + + if self.lambda_rcxyz > 0.: # 0.0 + target_xyz = get_xyz(target) # [bs, nvertices(vertices)/njoints(smpl), 3, nframes] + model_output_xyz = get_xyz(model_output) # [bs, nvertices, 3, nframes] + terms["rcxyz_mse"] = self.masked_l2(target_xyz, model_output_xyz, mask) # mean_flat((target_xyz - model_output_xyz) ** 2) + + if self.lambda_vel_rcxyz > 0.: # 0.0 + if self.data_rep == 'rot6d' and dataset.dataname in ['humanact12', 'uestc']: + target_xyz = get_xyz(target) if target_xyz is None else target_xyz + model_output_xyz = get_xyz(model_output) if model_output_xyz is None else model_output_xyz + target_xyz_vel = (target_xyz[:, :, :, 1:] - target_xyz[:, :, :, :-1]) + model_output_xyz_vel = (model_output_xyz[:, :, :, 1:] - model_output_xyz[:, :, :, :-1]) + terms["vel_xyz_mse"] = self.masked_l2(target_xyz_vel, model_output_xyz_vel, mask[:, :, :, 1:]) + + if self.lambda_fc > 0.: # 0.0 + torch.autograd.set_detect_anomaly(True) + if self.data_rep == 'rot6d' and dataset.dataname in ['humanact12', 'uestc']: + target_xyz = get_xyz(target) if target_xyz is None else target_xyz + model_output_xyz = get_xyz(model_output) if model_output_xyz is None else model_output_xyz + # 'L_Ankle', # 7, 'R_Ankle', # 8 , 'L_Foot', # 10, 'R_Foot', # 11 + l_ankle_idx, r_ankle_idx, l_foot_idx, r_foot_idx = 7, 8, 10, 11 + relevant_joints = [l_ankle_idx, l_foot_idx, r_ankle_idx, r_foot_idx] + gt_joint_xyz = target_xyz[:, relevant_joints, :, :] # [BatchSize, 4, 3, Frames] + gt_joint_vel = torch.linalg.norm(gt_joint_xyz[:, :, :, 1:] - gt_joint_xyz[:, :, :, :-1], axis=2) # [BatchSize, 4, Frames] + fc_mask = torch.unsqueeze((gt_joint_vel <= 0.01), dim=2).repeat(1, 1, 3, 1) + pred_joint_xyz = model_output_xyz[:, relevant_joints, :, :] # [BatchSize, 4, 3, Frames] + pred_vel = pred_joint_xyz[:, :, :, 1:] - pred_joint_xyz[:, :, :, :-1] + pred_vel[~fc_mask] = 0 + terms["fc"] = self.masked_l2(pred_vel, + torch.zeros(pred_vel.shape, device=pred_vel.device), + mask[:, :, :, 1:]) + if self.lambda_vel > 0.: # 0.0 + target_vel = (target[..., 1:] - target[..., :-1]) + model_output_vel = (model_output[..., 1:] - model_output[..., :-1]) + terms["vel_mse"] = self.masked_l2(target_vel[:, :-1, :, :], # Remove last joint, is the root location! + model_output_vel[:, :-1, :, :], + mask[:, :, :, 1:]) # mean_flat((target_vel - model_output_vel) ** 2) + + terms["loss"] = terms["rot_mse"] + terms.get('vb', 0.) +\ + (self.lambda_vel * terms.get('vel_mse', 0.)) +\ + (self.lambda_rcxyz * terms.get('rcxyz_mse', 0.)) + \ + (self.lambda_fc * terms.get('fc', 0.)) + + else: + raise NotImplementedError(self.loss_type) + + return terms + + def fc_loss_rot_repr(self, gt_xyz, pred_xyz, mask): + def to_np_cpu(x): + return x.detach().cpu().numpy() + """ + pose_xyz: SMPL batch tensor of shape: [BatchSize, 24, 3, Frames] + """ + # 'L_Ankle', # 7, 'R_Ankle', # 8 , 'L_Foot', # 10, 'R_Foot', # 11 + + l_ankle_idx, r_ankle_idx = 7, 8 + l_foot_idx, r_foot_idx = 10, 11 + """ Contact calculated by 'Kfir Method' Commented code)""" + # contact_signal = torch.zeros((pose_xyz.shape[0], pose_xyz.shape[3], 2), device=pose_xyz.device) # [BatchSize, Frames, 2] + # left_xyz = 0.5 * (pose_xyz[:, l_ankle_idx, :, :] + pose_xyz[:, l_foot_idx, :, :]) # [BatchSize, 3, Frames] + # right_xyz = 0.5 * (pose_xyz[:, r_ankle_idx, :, :] + pose_xyz[:, r_foot_idx, :, :]) + # left_z, right_z = left_xyz[:, 2, :], right_xyz[:, 2, :] # [BatchSize, Frames] + # left_velocity = torch.linalg.norm(left_xyz[:, :, 2:] - left_xyz[:, :, :-2], axis=1) # [BatchSize, Frames] + # right_velocity = torch.linalg.norm(left_xyz[:, :, 2:] - left_xyz[:, :, :-2], axis=1) + # + # left_z_mask = left_z <= torch.mean(torch.sort(left_z)[0][:, :left_z.shape[1] // 5], axis=-1) + # left_z_mask = torch.stack([left_z_mask, left_z_mask], dim=-1) # [BatchSize, Frames, 2] + # left_z_mask[:, :, 1] = False # Blank right side + # contact_signal[left_z_mask] = 0.4 + # + # right_z_mask = right_z <= torch.mean(torch.sort(right_z)[0][:, :right_z.shape[1] // 5], axis=-1) + # right_z_mask = torch.stack([right_z_mask, right_z_mask], dim=-1) # [BatchSize, Frames, 2] + # right_z_mask[:, :, 0] = False # Blank left side + # contact_signal[right_z_mask] = 0.4 + # contact_signal[left_z <= (torch.mean(torch.sort(left_z)[:left_z.shape[0] // 5]) + 20), 0] = 1 + # contact_signal[right_z <= (torch.mean(torch.sort(right_z)[:right_z.shape[0] // 5]) + 20), 1] = 1 + + # plt.plot(to_np_cpu(left_z[0]), label='left_z') + # plt.plot(to_np_cpu(left_velocity[0]), label='left_velocity') + # plt.plot(to_np_cpu(contact_signal[0, :, 0]), label='left_fc') + # plt.grid() + # plt.legend() + # plt.show() + # plt.plot(to_np_cpu(right_z[0]), label='right_z') + # plt.plot(to_np_cpu(right_velocity[0]), label='right_velocity') + # plt.plot(to_np_cpu(contact_signal[0, :, 1]), label='right_fc') + # plt.grid() + # plt.legend() + # plt.show() + + gt_joint_xyz = gt_xyz[:, [l_ankle_idx, l_foot_idx, r_ankle_idx, r_foot_idx], :, :] # [BatchSize, 4, 3, Frames] + gt_joint_vel = torch.linalg.norm(gt_joint_xyz[:, :, :, 1:] - gt_joint_xyz[:, :, :, :-1], axis=2) # [BatchSize, 4, Frames] + fc_mask = (gt_joint_vel <= 0.01) + pred_joint_xyz = pred_xyz[:, [l_ankle_idx, l_foot_idx, r_ankle_idx, r_foot_idx], :, :] # [BatchSize, 4, 3, Frames] + pred_joint_vel = torch.linalg.norm(pred_joint_xyz[:, :, :, 1:] - pred_joint_xyz[:, :, :, :-1], axis=2) # [BatchSize, 4, Frames] + pred_joint_vel[~fc_mask] = 0 # Blank non-contact velocities frames. [BS,4,FRAMES] + pred_joint_vel = torch.unsqueeze(pred_joint_vel, dim=2) + + """DEBUG CODE""" + # print(f'mask: {mask.shape}') + # print(f'pred_joint_vel: {pred_joint_vel.shape}') + # plt.title(f'Joint: {joint_idx}') + # plt.plot(to_np_cpu(gt_joint_vel[0]), label='velocity') + # plt.plot(to_np_cpu(fc_mask[0]), label='fc') + # plt.grid() + # plt.legend() + # plt.show() + return self.masked_l2(pred_joint_vel, torch.zeros(pred_joint_vel.shape, device=pred_joint_vel.device), + mask[:, :, :, 1:]) + # TODO - NOT USED YET, JUST COMMITING TO NOT DELETE THIS AND KEEP INITIAL IMPLEMENTATION, NOT DONE! + def foot_contact_loss_humanml3d(self, target, model_output): + # root_rot_velocity (B, seq_len, 1) + # root_linear_velocity (B, seq_len, 2) + # root_y (B, seq_len, 1) + # ric_data (B, seq_len, (joint_num - 1)*3) , XYZ + # rot_data (B, seq_len, (joint_num - 1)*6) , 6D + # local_velocity (B, seq_len, joint_num*3) , XYZ + # foot contact (B, seq_len, 4) , + + target_fc = target[:, -4:, :, :] + root_rot_velocity = target[:, :1, :, :] + root_linear_velocity = target[:, 1:3, :, :] + root_y = target[:, 3:4, :, :] + ric_data = target[:, 4:67, :, :] # 4+(3*21)=67 + rot_data = target[:, 67:193, :, :] # 67+(6*21)=193 + local_velocity = target[:, 193:259, :, :] # 193+(3*22)=259 + contact = target[:, 259:, :, :] # 193+(3*22)=259 + contact_mask_gt = contact > 0.5 # contact mask order for indexes are fid_l [7, 10], fid_r [8, 11] + vel_lf_7 = local_velocity[:, 7 * 3:8 * 3, :, :] + vel_rf_8 = local_velocity[:, 8 * 3:9 * 3, :, :] + vel_lf_10 = local_velocity[:, 10 * 3:11 * 3, :, :] + vel_rf_11 = local_velocity[:, 11 * 3:12 * 3, :, :] + + calc_vel_lf_7 = ric_data[:, 6 * 3:7 * 3, :, 1:] - ric_data[:, 6 * 3:7 * 3, :, :-1] + calc_vel_rf_8 = ric_data[:, 7 * 3:8 * 3, :, 1:] - ric_data[:, 7 * 3:8 * 3, :, :-1] + calc_vel_lf_10 = ric_data[:, 9 * 3:10 * 3, :, 1:] - ric_data[:, 9 * 3:10 * 3, :, :-1] + calc_vel_rf_11 = ric_data[:, 10 * 3:11 * 3, :, 1:] - ric_data[:, 10 * 3:11 * 3, :, :-1] + + # vel_foots = torch.stack([vel_lf_7, vel_lf_10, vel_rf_8, vel_rf_11], dim=1) + for chosen_vel_foot_calc, chosen_vel_foot, joint_idx, contact_mask_idx in zip( + [calc_vel_lf_7, calc_vel_rf_8, calc_vel_lf_10, calc_vel_rf_11], + [vel_lf_7, vel_lf_10, vel_rf_8, vel_rf_11], + [7, 10, 8, 11], + [0, 1, 2, 3]): + tmp_mask_gt = contact_mask_gt[:, contact_mask_idx, :, :].cpu().detach().numpy().reshape(-1).astype(int) + chosen_vel_norm = np.linalg.norm(chosen_vel_foot.cpu().detach().numpy().reshape((3, -1)), axis=0) + chosen_vel_calc_norm = np.linalg.norm(chosen_vel_foot_calc.cpu().detach().numpy().reshape((3, -1)), + axis=0) + + print(tmp_mask_gt.shape) + print(chosen_vel_foot.shape) + print(chosen_vel_calc_norm.shape) + import matplotlib.pyplot as plt + plt.plot(tmp_mask_gt, label='FC mask') + plt.plot(chosen_vel_norm, label='Vel. XYZ norm (from vector)') + plt.plot(chosen_vel_calc_norm, label='Vel. XYZ norm (calculated diff XYZ)') + + plt.title(f'FC idx {contact_mask_idx}, Joint Index {joint_idx}') + plt.legend() + plt.show() + # print(vel_foots.shape) + return 0 + # TODO - NOT USED YET, JUST COMMITING TO NOT DELETE THIS AND KEEP INITIAL IMPLEMENTATION, NOT DONE! + def velocity_consistency_loss_humanml3d(self, target, model_output): + # root_rot_velocity (B, seq_len, 1) + # root_linear_velocity (B, seq_len, 2) + # root_y (B, seq_len, 1) + # ric_data (B, seq_len, (joint_num - 1)*3) , XYZ + # rot_data (B, seq_len, (joint_num - 1)*6) , 6D + # local_velocity (B, seq_len, joint_num*3) , XYZ + # foot contact (B, seq_len, 4) , + + target_fc = target[:, -4:, :, :] + root_rot_velocity = target[:, :1, :, :] + root_linear_velocity = target[:, 1:3, :, :] + root_y = target[:, 3:4, :, :] + ric_data = target[:, 4:67, :, :] # 4+(3*21)=67 + rot_data = target[:, 67:193, :, :] # 67+(6*21)=193 + local_velocity = target[:, 193:259, :, :] # 193+(3*22)=259 + contact = target[:, 259:, :, :] # 193+(3*22)=259 + + calc_vel_from_xyz = ric_data[:, :, :, 1:] - ric_data[:, :, :, :-1] + velocity_from_vector = local_velocity[:, 3:, :, 1:] # Slicing out root + r_rot_quat, r_pos = motion_process.recover_root_rot_pos(target.permute(0, 2, 3, 1).type(th.FloatTensor)) + print(f'r_rot_quat: {r_rot_quat.shape}') + print(f'calc_vel_from_xyz: {calc_vel_from_xyz.shape}') + calc_vel_from_xyz = calc_vel_from_xyz.permute(0, 2, 3, 1) + calc_vel_from_xyz = calc_vel_from_xyz.reshape((1, 1, -1, 21, 3)).type(th.FloatTensor) + r_rot_quat_adapted = r_rot_quat[..., :-1, None, :].repeat((1,1,1,21,1)).to(calc_vel_from_xyz.device) + print(f'calc_vel_from_xyz: {calc_vel_from_xyz.shape} , {calc_vel_from_xyz.device}') + print(f'r_rot_quat_adapted: {r_rot_quat_adapted.shape}, {r_rot_quat_adapted.device}') + + calc_vel_from_xyz = motion_process.qrot(r_rot_quat_adapted, calc_vel_from_xyz) + calc_vel_from_xyz = calc_vel_from_xyz.reshape((1, 1, -1, 21 * 3)) + calc_vel_from_xyz = calc_vel_from_xyz.permute(0, 3, 1, 2) + print(f'calc_vel_from_xyz: {calc_vel_from_xyz.shape} , {calc_vel_from_xyz.device}') + + import matplotlib.pyplot as plt + for i in range(21): + plt.plot(np.linalg.norm(calc_vel_from_xyz[:,i*3:(i+1)*3,:,:].cpu().detach().numpy().reshape((3, -1)), axis=0), label='Calc Vel') + plt.plot(np.linalg.norm(velocity_from_vector[:,i*3:(i+1)*3,:,:].cpu().detach().numpy().reshape((3, -1)), axis=0), label='Vector Vel') + plt.title(f'Joint idx: {i}') + plt.legend() + plt.show() + print(calc_vel_from_xyz.shape) + print(velocity_from_vector.shape) + diff = calc_vel_from_xyz-velocity_from_vector + print(np.linalg.norm(diff.cpu().detach().numpy().reshape((63, -1)), axis=0)) + + return 0 + + + def _prior_bpd(self, x_start): + """ + Get the prior KL term for the variational lower-bound, measured in + bits-per-dim. + + This term can't be optimized, as it only depends on the encoder. + + :param x_start: the [N x C x ...] tensor of inputs. + :return: a batch of [N] KL values (in bits), one per batch element. + """ + batch_size = x_start.shape[0] + t = th.tensor([self.num_timesteps - 1] * batch_size, device=x_start.device) + qt_mean, _, qt_log_variance = self.q_mean_variance(x_start, t) + kl_prior = normal_kl( + mean1=qt_mean, logvar1=qt_log_variance, mean2=0.0, logvar2=0.0 + ) + return mean_flat(kl_prior) / np.log(2.0) + + def calc_bpd_loop(self, model, x_start, clip_denoised=True, model_kwargs=None): + """ + Compute the entire variational lower-bound, measured in bits-per-dim, + as well as other related quantities. + + :param model: the model to evaluate loss on. + :param x_start: the [N x C x ...] tensor of inputs. + :param clip_denoised: if True, clip denoised samples. + :param model_kwargs: if not None, a dict of extra keyword arguments to + pass to the model. This can be used for conditioning. + + :return: a dict containing the following keys: + - total_bpd: the total variational lower-bound, per batch element. + - prior_bpd: the prior term in the lower-bound. + - vb: an [N x T] tensor of terms in the lower-bound. + - xstart_mse: an [N x T] tensor of x_0 MSEs for each timestep. + - mse: an [N x T] tensor of epsilon MSEs for each timestep. + """ + device = x_start.device + batch_size = x_start.shape[0] + + vb = [] + xstart_mse = [] + mse = [] + for t in list(range(self.num_timesteps))[::-1]: + t_batch = th.tensor([t] * batch_size, device=device) + noise = th.randn_like(x_start) + x_t = self.q_sample(x_start=x_start, t=t_batch, noise=noise) + # Calculate VLB term at the current timestep + with th.no_grad(): + out = self._vb_terms_bpd( + model, + x_start=x_start, + x_t=x_t, + t=t_batch, + clip_denoised=clip_denoised, + model_kwargs=model_kwargs, + ) + vb.append(out["output"]) + xstart_mse.append(mean_flat((out["pred_xstart"] - x_start) ** 2)) + eps = self._predict_eps_from_xstart(x_t, t_batch, out["pred_xstart"]) + mse.append(mean_flat((eps - noise) ** 2)) + + vb = th.stack(vb, dim=1) + xstart_mse = th.stack(xstart_mse, dim=1) + mse = th.stack(mse, dim=1) + + prior_bpd = self._prior_bpd(x_start) + total_bpd = vb.sum(dim=1) + prior_bpd + return { + "total_bpd": total_bpd, + "prior_bpd": prior_bpd, + "vb": vb, + "xstart_mse": xstart_mse, + "mse": mse, + } + + +def _extract_into_tensor(arr, timesteps, broadcast_shape): + """ + Extract values from a 1-D numpy array for a batch of indices. + + :param arr: the 1-D numpy array. + :param timesteps: a tensor of indices into the array to extract. + :param broadcast_shape: a larger shape of K dimensions with the batch + dimension equal to the length of timesteps. + :return: a tensor of shape [batch_size, 1, ...] where the shape has K dims. + """ + res = th.from_numpy(arr).to(device=timesteps.device)[timesteps].float() + while len(res.shape) < len(broadcast_shape): + res = res[..., None] + return res.expand(broadcast_shape) diff --git a/diffusion/inpainting_gaussian_diffusion.py b/diffusion/inpainting_gaussian_diffusion.py new file mode 100644 index 0000000000000000000000000000000000000000..1721fcc2acf99ddba8d316afcb372ab7a6805d9b --- /dev/null +++ b/diffusion/inpainting_gaussian_diffusion.py @@ -0,0 +1,66 @@ +from diffusion.respace import SpacedDiffusion +from .gaussian_diffusion import _extract_into_tensor +import torch as th + +class InpaintingGaussianDiffusion(SpacedDiffusion): + def q_sample(self, x_start, t, noise=None, model_kwargs=None): + """ + overrides q_sample to use the inpainting mask + + same usage as in GaussianDiffusion + """ + if noise is None: + noise = th.randn_like(x_start) + assert noise.shape == x_start.shape + + bs, feat, _, frames = noise.shape + inpainting_mask = th.zeros_like(noise).to(noise.device) + inpainting_mask[:,:10] = 1 #just inpainting root trajectory, for training + noise *= 1. - inpainting_mask + return ( + _extract_into_tensor(self.sqrt_alphas_cumprod, t, x_start.shape) * x_start + + _extract_into_tensor(self.sqrt_one_minus_alphas_cumprod, t, x_start.shape) + * noise + ) + + def p_sample( + self, + model, + x, + t, + clip_denoised=True, + denoised_fn=None, + cond_fn=None, + model_kwargs=None, + const_noise=False, + ): + """ + overrides p_sample to use the inpainting mask + + same usage as in GaussianDiffusion + """ + out = self.p_mean_variance( + model, + x, + t, + clip_denoised=clip_denoised, + denoised_fn=denoised_fn, + model_kwargs=model_kwargs, + ) + noise = th.randn_like(x) + if const_noise: + noise = noise[[0]].repeat(x.shape[0], 1, 1, 1) + + inpainting_mask = th.zeros_like(noise).to(noise.device) + inpainting_mask[:,:10] = 1 #just inpainting root trajectory, for inference + noise *= 1. - inpainting_mask + + nonzero_mask = ( + (t != 0).float().view(-1, *([1] * (len(x.shape) - 1))) + ) # no noise when t == 0 + if cond_fn is not None: + out["mean"] = self.condition_mean( + cond_fn, out, x, t, model_kwargs=model_kwargs + ) + sample = out["mean"] + nonzero_mask * th.exp(0.5 * out["log_variance"]) * noise + return {"sample": sample, "pred_xstart": out["pred_xstart"]} \ No newline at end of file diff --git a/diffusion/logger.py b/diffusion/logger.py new file mode 100644 index 0000000000000000000000000000000000000000..b1d856dcfea6b56a2ee8d37b286887430dbfac30 --- /dev/null +++ b/diffusion/logger.py @@ -0,0 +1,495 @@ +""" +Logger copied from OpenAI baselines to avoid extra RL-based dependencies: +https://github.com/openai/baselines/blob/ea25b9e8b234e6ee1bca43083f8f3cf974143998/baselines/logger.py +""" + +import os +import sys +import shutil +import os.path as osp +import json +import time +import datetime +import tempfile +import warnings +from collections import defaultdict +from contextlib import contextmanager + +DEBUG = 10 +INFO = 20 +WARN = 30 +ERROR = 40 + +DISABLED = 50 + + +class KVWriter(object): + def writekvs(self, kvs): + raise NotImplementedError + + +class SeqWriter(object): + def writeseq(self, seq): + raise NotImplementedError + + +class HumanOutputFormat(KVWriter, SeqWriter): + def __init__(self, filename_or_file): + if isinstance(filename_or_file, str): + self.file = open(filename_or_file, "wt") + self.own_file = True + else: + assert hasattr(filename_or_file, "read"), ( + "expected file or str, got %s" % filename_or_file + ) + self.file = filename_or_file + self.own_file = False + + def writekvs(self, kvs): + # Create strings for printing + key2str = {} + for (key, val) in sorted(kvs.items()): + if hasattr(val, "__float__"): + valstr = "%-8.3g" % val + else: + valstr = str(val) + key2str[self._truncate(key)] = self._truncate(valstr) + + # Find max widths + if len(key2str) == 0: + print("WARNING: tried to write empty key-value dict") + return + else: + keywidth = max(map(len, key2str.keys())) + valwidth = max(map(len, key2str.values())) + + # Write out the data + dashes = "-" * (keywidth + valwidth + 7) + lines = [dashes] + for (key, val) in sorted(key2str.items(), key=lambda kv: kv[0].lower()): + lines.append( + "| %s%s | %s%s |" + % (key, " " * (keywidth - len(key)), val, " " * (valwidth - len(val))) + ) + lines.append(dashes) + self.file.write("\n".join(lines) + "\n") + + # Flush the output to the file + self.file.flush() + + def _truncate(self, s): + maxlen = 30 + return s[: maxlen - 3] + "..." if len(s) > maxlen else s + + def writeseq(self, seq): + seq = list(seq) + for (i, elem) in enumerate(seq): + self.file.write(elem) + if i < len(seq) - 1: # add space unless this is the last one + self.file.write(" ") + self.file.write("\n") + self.file.flush() + + def close(self): + if self.own_file: + self.file.close() + + +class JSONOutputFormat(KVWriter): + def __init__(self, filename): + self.file = open(filename, "wt") + + def writekvs(self, kvs): + for k, v in sorted(kvs.items()): + if hasattr(v, "dtype"): + kvs[k] = float(v) + self.file.write(json.dumps(kvs) + "\n") + self.file.flush() + + def close(self): + self.file.close() + + +class CSVOutputFormat(KVWriter): + def __init__(self, filename): + self.file = open(filename, "w+t") + self.keys = [] + self.sep = "," + + def writekvs(self, kvs): + # Add our current row to the history + extra_keys = list(kvs.keys() - self.keys) + extra_keys.sort() + if extra_keys: + self.keys.extend(extra_keys) + self.file.seek(0) + lines = self.file.readlines() + self.file.seek(0) + for (i, k) in enumerate(self.keys): + if i > 0: + self.file.write(",") + self.file.write(k) + self.file.write("\n") + for line in lines[1:]: + self.file.write(line[:-1]) + self.file.write(self.sep * len(extra_keys)) + self.file.write("\n") + for (i, k) in enumerate(self.keys): + if i > 0: + self.file.write(",") + v = kvs.get(k) + if v is not None: + self.file.write(str(v)) + self.file.write("\n") + self.file.flush() + + def close(self): + self.file.close() + + +class TensorBoardOutputFormat(KVWriter): + """ + Dumps key/value pairs into TensorBoard's numeric format. + """ + + def __init__(self, dir): + os.makedirs(dir, exist_ok=True) + self.dir = dir + self.step = 1 + prefix = "events" + path = osp.join(osp.abspath(dir), prefix) + import tensorflow as tf + from tensorflow.python import pywrap_tensorflow + from tensorflow.core.util import event_pb2 + from tensorflow.python.util import compat + + self.tf = tf + self.event_pb2 = event_pb2 + self.pywrap_tensorflow = pywrap_tensorflow + self.writer = pywrap_tensorflow.EventsWriter(compat.as_bytes(path)) + + def writekvs(self, kvs): + def summary_val(k, v): + kwargs = {"tag": k, "simple_value": float(v)} + return self.tf.Summary.Value(**kwargs) + + summary = self.tf.Summary(value=[summary_val(k, v) for k, v in kvs.items()]) + event = self.event_pb2.Event(wall_time=time.time(), summary=summary) + event.step = ( + self.step + ) # is there any reason why you'd want to specify the step? + self.writer.WriteEvent(event) + self.writer.Flush() + self.step += 1 + + def close(self): + if self.writer: + self.writer.Close() + self.writer = None + + +def make_output_format(format, ev_dir, log_suffix=""): + os.makedirs(ev_dir, exist_ok=True) + if format == "stdout": + return HumanOutputFormat(sys.stdout) + elif format == "log": + return HumanOutputFormat(osp.join(ev_dir, "log%s.txt" % log_suffix)) + elif format == "json": + return JSONOutputFormat(osp.join(ev_dir, "progress%s.json" % log_suffix)) + elif format == "csv": + return CSVOutputFormat(osp.join(ev_dir, "progress%s.csv" % log_suffix)) + elif format == "tensorboard": + return TensorBoardOutputFormat(osp.join(ev_dir, "tb%s" % log_suffix)) + else: + raise ValueError("Unknown format specified: %s" % (format,)) + + +# ================================================================ +# API +# ================================================================ + + +def logkv(key, val): + """ + Log a value of some diagnostic + Call this once for each diagnostic quantity, each iteration + If called many times, last value will be used. + """ + get_current().logkv(key, val) + + +def logkv_mean(key, val): + """ + The same as logkv(), but if called many times, values averaged. + """ + get_current().logkv_mean(key, val) + + +def logkvs(d): + """ + Log a dictionary of key-value pairs + """ + for (k, v) in d.items(): + logkv(k, v) + + +def dumpkvs(): + """ + Write all of the diagnostics from the current iteration + """ + return get_current().dumpkvs() + + +def getkvs(): + return get_current().name2val + + +def log(*args, level=INFO): + """ + Write the sequence of args, with no separators, to the console and output files (if you've configured an output file). + """ + get_current().log(*args, level=level) + + +def debug(*args): + log(*args, level=DEBUG) + + +def info(*args): + log(*args, level=INFO) + + +def warn(*args): + log(*args, level=WARN) + + +def error(*args): + log(*args, level=ERROR) + + +def set_level(level): + """ + Set logging threshold on current logger. + """ + get_current().set_level(level) + + +def set_comm(comm): + get_current().set_comm(comm) + + +def get_dir(): + """ + Get directory that log files are being written to. + will be None if there is no output directory (i.e., if you didn't call start) + """ + return get_current().get_dir() + + +record_tabular = logkv +dump_tabular = dumpkvs + + +@contextmanager +def profile_kv(scopename): + logkey = "wait_" + scopename + tstart = time.time() + try: + yield + finally: + get_current().name2val[logkey] += time.time() - tstart + + +def profile(n): + """ + Usage: + @profile("my_func") + def my_func(): code + """ + + def decorator_with_name(func): + def func_wrapper(*args, **kwargs): + with profile_kv(n): + return func(*args, **kwargs) + + return func_wrapper + + return decorator_with_name + + +# ================================================================ +# Backend +# ================================================================ + + +def get_current(): + if Logger.CURRENT is None: + _configure_default_logger() + + return Logger.CURRENT + + +class Logger(object): + DEFAULT = None # A logger with no output files. (See right below class definition) + # So that you can still log to the terminal without setting up any output files + CURRENT = None # Current logger being used by the free functions above + + def __init__(self, dir, output_formats, comm=None): + self.name2val = defaultdict(float) # values this iteration + self.name2cnt = defaultdict(int) + self.level = INFO + self.dir = dir + self.output_formats = output_formats + self.comm = comm + + # Logging API, forwarded + # ---------------------------------------- + def logkv(self, key, val): + self.name2val[key] = val + + def logkv_mean(self, key, val): + oldval, cnt = self.name2val[key], self.name2cnt[key] + self.name2val[key] = oldval * cnt / (cnt + 1) + val / (cnt + 1) + self.name2cnt[key] = cnt + 1 + + def dumpkvs(self): + if self.comm is None: + d = self.name2val + else: + d = mpi_weighted_mean( + self.comm, + { + name: (val, self.name2cnt.get(name, 1)) + for (name, val) in self.name2val.items() + }, + ) + if self.comm.rank != 0: + d["dummy"] = 1 # so we don't get a warning about empty dict + out = d.copy() # Return the dict for unit testing purposes + for fmt in self.output_formats: + if isinstance(fmt, KVWriter): + fmt.writekvs(d) + self.name2val.clear() + self.name2cnt.clear() + return out + + def log(self, *args, level=INFO): + if self.level <= level: + self._do_log(args) + + # Configuration + # ---------------------------------------- + def set_level(self, level): + self.level = level + + def set_comm(self, comm): + self.comm = comm + + def get_dir(self): + return self.dir + + def close(self): + for fmt in self.output_formats: + fmt.close() + + # Misc + # ---------------------------------------- + def _do_log(self, args): + for fmt in self.output_formats: + if isinstance(fmt, SeqWriter): + fmt.writeseq(map(str, args)) + + +def get_rank_without_mpi_import(): + # check environment variables here instead of importing mpi4py + # to avoid calling MPI_Init() when this module is imported + for varname in ["PMI_RANK", "OMPI_COMM_WORLD_RANK"]: + if varname in os.environ: + return int(os.environ[varname]) + return 0 + + +def mpi_weighted_mean(comm, local_name2valcount): + """ + Copied from: https://github.com/openai/baselines/blob/ea25b9e8b234e6ee1bca43083f8f3cf974143998/baselines/common/mpi_util.py#L110 + Perform a weighted average over dicts that are each on a different node + Input: local_name2valcount: dict mapping key -> (value, count) + Returns: key -> mean + """ + all_name2valcount = comm.gather(local_name2valcount) + if comm.rank == 0: + name2sum = defaultdict(float) + name2count = defaultdict(float) + for n2vc in all_name2valcount: + for (name, (val, count)) in n2vc.items(): + try: + val = float(val) + except ValueError: + if comm.rank == 0: + warnings.warn( + "WARNING: tried to compute mean on non-float {}={}".format( + name, val + ) + ) + else: + name2sum[name] += val * count + name2count[name] += count + return {name: name2sum[name] / name2count[name] for name in name2sum} + else: + return {} + + +def configure(dir=None, format_strs=None, comm=None, log_suffix=""): + """ + If comm is provided, average all numerical stats across that comm + """ + if dir is None: + dir = os.getenv("OPENAI_LOGDIR") + if dir is None: + dir = osp.join( + tempfile.gettempdir(), + datetime.datetime.now().strftime("openai-%Y-%m-%d-%H-%M-%S-%f"), + ) + assert isinstance(dir, str) + dir = os.path.expanduser(dir) + os.makedirs(os.path.expanduser(dir), exist_ok=True) + + rank = get_rank_without_mpi_import() + if rank > 0: + log_suffix = log_suffix + "-rank%03i" % rank + + if format_strs is None: + if rank == 0: + format_strs = os.getenv("OPENAI_LOG_FORMAT", "stdout,log,csv").split(",") + else: + format_strs = os.getenv("OPENAI_LOG_FORMAT_MPI", "log").split(",") + format_strs = filter(None, format_strs) + output_formats = [make_output_format(f, dir, log_suffix) for f in format_strs] + + Logger.CURRENT = Logger(dir=dir, output_formats=output_formats, comm=comm) + if output_formats: + log("Logging to %s" % dir) + + +def _configure_default_logger(): + configure() + Logger.DEFAULT = Logger.CURRENT + + +def reset(): + if Logger.CURRENT is not Logger.DEFAULT: + Logger.CURRENT.close() + Logger.CURRENT = Logger.DEFAULT + log("Reset logger") + + +@contextmanager +def scoped_configure(dir=None, format_strs=None, comm=None): + prevlogger = Logger.CURRENT + configure(dir=dir, format_strs=format_strs, comm=comm) + try: + yield + finally: + Logger.CURRENT.close() + Logger.CURRENT = prevlogger + diff --git a/diffusion/losses.py b/diffusion/losses.py new file mode 100644 index 0000000000000000000000000000000000000000..e3fded1953584eaaf183f3d2399be545a5003e0a --- /dev/null +++ b/diffusion/losses.py @@ -0,0 +1,77 @@ +# This code is based on https://github.com/openai/guided-diffusion +""" +Helpers for various likelihood-based losses. These are ported from the original +Ho et al. diffusion models codebase: +https://github.com/hojonathanho/diffusion/blob/1e0dceb3b3495bbe19116a5e1b3596cd0706c543/diffusion_tf/utils.py +""" + +import numpy as np +import torch as th + + +def normal_kl(mean1, logvar1, mean2, logvar2): + """ + Compute the KL divergence between two gaussians. + + Shapes are automatically broadcasted, so batches can be compared to + scalars, among other use cases. + """ + tensor = None + for obj in (mean1, logvar1, mean2, logvar2): + if isinstance(obj, th.Tensor): + tensor = obj + break + assert tensor is not None, "at least one argument must be a Tensor" + + # Force variances to be Tensors. Broadcasting helps convert scalars to + # Tensors, but it does not work for th.exp(). + logvar1, logvar2 = [ + x if isinstance(x, th.Tensor) else th.tensor(x).to(tensor) + for x in (logvar1, logvar2) + ] + + return 0.5 * ( + -1.0 + + logvar2 + - logvar1 + + th.exp(logvar1 - logvar2) + + ((mean1 - mean2) ** 2) * th.exp(-logvar2) + ) + + +def approx_standard_normal_cdf(x): + """ + A fast approximation of the cumulative distribution function of the + standard normal. + """ + return 0.5 * (1.0 + th.tanh(np.sqrt(2.0 / np.pi) * (x + 0.044715 * th.pow(x, 3)))) + + +def discretized_gaussian_log_likelihood(x, *, means, log_scales): + """ + Compute the log-likelihood of a Gaussian distribution discretizing to a + given image. + + :param x: the target images. It is assumed that this was uint8 values, + rescaled to the range [-1, 1]. + :param means: the Gaussian mean Tensor. + :param log_scales: the Gaussian log stddev Tensor. + :return: a tensor like x of log probabilities (in nats). + """ + assert x.shape == means.shape == log_scales.shape + centered_x = x - means + inv_stdv = th.exp(-log_scales) + plus_in = inv_stdv * (centered_x + 1.0 / 255.0) + cdf_plus = approx_standard_normal_cdf(plus_in) + min_in = inv_stdv * (centered_x - 1.0 / 255.0) + cdf_min = approx_standard_normal_cdf(min_in) + log_cdf_plus = th.log(cdf_plus.clamp(min=1e-12)) + log_one_minus_cdf_min = th.log((1.0 - cdf_min).clamp(min=1e-12)) + cdf_delta = cdf_plus - cdf_min + log_probs = th.where( + x < -0.999, + log_cdf_plus, + th.where(x > 0.999, log_one_minus_cdf_min, th.log(cdf_delta.clamp(min=1e-12))), + ) + assert log_probs.shape == x.shape + return log_probs diff --git a/diffusion/model_util.py b/diffusion/model_util.py new file mode 100644 index 0000000000000000000000000000000000000000..6fe9a76f88fea4ce68188a46a3ebf859e290c4a9 --- /dev/null +++ b/diffusion/model_util.py @@ -0,0 +1,51 @@ +import pdb + +from diffusion import gaussian_diffusion as gd +from diffusion.respace import SpacedDiffusion, space_timesteps + + + +def create_gaussian_diffusion(DiffusionClass=SpacedDiffusion,use_ddim=False): + noise_schedule = 'cosine' + sigma_small = True + lambda_vel = 0.0 + lambda_rcxyz = 0.0 + lambda_fc = 0.0 + + # default params + predict_xstart = True # we always predict x_start (a.k.a. x0), that's our deal! + steps = 1000 + scale_beta = 1. # no scaling + timestep_respacing =None + if use_ddim: + timestep_respacing = 'ddim50' # can be used for ddim sampling, we don't use it. + learn_sigma = False + rescale_timesteps = False + + betas = gd.get_named_beta_schedule(noise_schedule, steps, scale_beta) + loss_type = gd.LossType.MSE + + if not timestep_respacing: + timestep_respacing = [steps] + + return DiffusionClass( + use_timesteps=space_timesteps(steps, timestep_respacing), + betas=betas, + model_mean_type=( + gd.ModelMeanType.EPSILON if not predict_xstart else gd.ModelMeanType.START_X + ), + model_var_type=( + ( + gd.ModelVarType.FIXED_LARGE + if not sigma_small + else gd.ModelVarType.FIXED_SMALL + ) + if not learn_sigma + else gd.ModelVarType.LEARNED_RANGE + ), + loss_type=loss_type, + rescale_timesteps=rescale_timesteps, + lambda_vel=lambda_vel, + lambda_rcxyz=lambda_rcxyz, + lambda_fc=lambda_fc, + ) \ No newline at end of file diff --git a/diffusion/nn.py b/diffusion/nn.py new file mode 100644 index 0000000000000000000000000000000000000000..41c18e7dd3d8cae1e719638e87c27f718f6a94e6 --- /dev/null +++ b/diffusion/nn.py @@ -0,0 +1,197 @@ +# This code is based on https://github.com/openai/guided-diffusion +""" +Various utilities for neural networks. +""" + +import math + +import torch as th +import torch.nn as nn + + +# PyTorch 1.7 has SiLU, but we support PyTorch 1.5. +class SiLU(nn.Module): + def forward(self, x): + return x * th.sigmoid(x) + + +class GroupNorm32(nn.GroupNorm): + def forward(self, x): + return super().forward(x.float()).type(x.dtype) + + +def conv_nd(dims, *args, **kwargs): + """ + Create a 1D, 2D, or 3D convolution module. + """ + if dims == 1: + return nn.Conv1d(*args, **kwargs) + elif dims == 2: + return nn.Conv2d(*args, **kwargs) + elif dims == 3: + return nn.Conv3d(*args, **kwargs) + raise ValueError(f"unsupported dimensions: {dims}") + + +def linear(*args, **kwargs): + """ + Create a linear module. + """ + return nn.Linear(*args, **kwargs) + + +def avg_pool_nd(dims, *args, **kwargs): + """ + Create a 1D, 2D, or 3D average pooling module. + """ + if dims == 1: + return nn.AvgPool1d(*args, **kwargs) + elif dims == 2: + return nn.AvgPool2d(*args, **kwargs) + elif dims == 3: + return nn.AvgPool3d(*args, **kwargs) + raise ValueError(f"unsupported dimensions: {dims}") + + +def update_ema(target_params, source_params, rate=0.99): + """ + Update target parameters to be closer to those of source parameters using + an exponential moving average. + + :param target_params: the target parameter sequence. + :param source_params: the source parameter sequence. + :param rate: the EMA rate (closer to 1 means slower). + """ + for targ, src in zip(target_params, source_params): + targ.detach().mul_(rate).add_(src, alpha=1 - rate) + + +def zero_module(module): + """ + Zero out the parameters of a module and return it. + """ + for p in module.parameters(): + p.detach().zero_() + return module + + +def scale_module(module, scale): + """ + Scale the parameters of a module and return it. + """ + for p in module.parameters(): + p.detach().mul_(scale) + return module + + +def mean_flat(tensor): + """ + Take the mean over all non-batch dimensions. + """ + return tensor.mean(dim=list(range(1, len(tensor.shape)))) + +def sum_flat(tensor): + """ + Take the sum over all non-batch dimensions. + """ + return tensor.sum(dim=list(range(1, len(tensor.shape)))) + + +def normalization(channels): + """ + Make a standard normalization layer. + + :param channels: number of input channels. + :return: an nn.Module for normalization. + """ + return GroupNorm32(32, channels) + + +def timestep_embedding(timesteps, dim, max_period=10000): + """ + Create sinusoidal timestep embeddings. + + :param timesteps: a 1-D Tensor of N indices, one per batch element. + These may be fractional. + :param dim: the dimension of the output. + :param max_period: controls the minimum frequency of the embeddings. + :return: an [N x dim] Tensor of positional embeddings. + """ + half = dim // 2 + freqs = th.exp( + -math.log(max_period) * th.arange(start=0, end=half, dtype=th.float32) / half + ).to(device=timesteps.device) + args = timesteps[:, None].float() * freqs[None] + embedding = th.cat([th.cos(args), th.sin(args)], dim=-1) + if dim % 2: + embedding = th.cat([embedding, th.zeros_like(embedding[:, :1])], dim=-1) + return embedding + + +def checkpoint(func, inputs, params, flag): + """ + Evaluate a function without caching intermediate activations, allowing for + reduced memory at the expense of extra compute in the backward pass. + :param func: the function to evaluate. + :param inputs: the argument sequence to pass to `func`. + :param params: a sequence of parameters `func` depends on but does not + explicitly take as arguments. + :param flag: if False, disable gradient checkpointing. + """ + if flag: + args = tuple(inputs) + tuple(params) + return CheckpointFunction.apply(func, len(inputs), *args) + else: + return func(*inputs) + + +class CheckpointFunction(th.autograd.Function): + @staticmethod + @th.cuda.amp.custom_fwd + def forward(ctx, run_function, length, *args): + ctx.run_function = run_function + ctx.input_length = length + ctx.save_for_backward(*args) + with th.no_grad(): + output_tensors = ctx.run_function(*args[:length]) + return output_tensors + + @staticmethod + @th.cuda.amp.custom_bwd + def backward(ctx, *output_grads): + args = list(ctx.saved_tensors) + + # Filter for inputs that require grad. If none, exit early. + input_indices = [i for (i, x) in enumerate(args) if x.requires_grad] + if not input_indices: + return (None, None) + tuple(None for _ in args) + + with th.enable_grad(): + for i in input_indices: + if i < ctx.input_length: + # Not sure why the OAI code does this little + # dance. It might not be necessary. + args[i] = args[i].detach().requires_grad_() + args[i] = args[i].view_as(args[i]) + output_tensors = ctx.run_function(*args[:ctx.input_length]) + + if isinstance(output_tensors, th.Tensor): + output_tensors = [output_tensors] + + # Filter for outputs that require grad. If none, exit early. + out_and_grads = [(o, g) for (o, g) in zip(output_tensors, output_grads) if o.requires_grad] + if not out_and_grads: + return (None, None) + tuple(None for _ in args) + + # Compute gradients on the filtered tensors. + computed_grads = th.autograd.grad( + [o for (o, g) in out_and_grads], + [args[i] for i in input_indices], + [g for (o, g) in out_and_grads] + ) + + # Reassemble the complete gradient tuple. + input_grads = [None for _ in args] + for (i, g) in zip(input_indices, computed_grads): + input_grads[i] = g + return (None, None) + tuple(input_grads) diff --git a/diffusion/resample.py b/diffusion/resample.py new file mode 100644 index 0000000000000000000000000000000000000000..c82eccdcd47c468d41e7cbe02de6a731f2c9bf81 --- /dev/null +++ b/diffusion/resample.py @@ -0,0 +1,154 @@ +from abc import ABC, abstractmethod + +import numpy as np +import torch as th +import torch.distributed as dist + + +def create_named_schedule_sampler(name, diffusion): + """ + Create a ScheduleSampler from a library of pre-defined samplers. + + :param name: the name of the sampler. + :param diffusion: the diffusion object to sample for. + """ + if name == "uniform": + return UniformSampler(diffusion) + elif name == "loss-second-moment": + return LossSecondMomentResampler(diffusion) + else: + raise NotImplementedError(f"unknown schedule sampler: {name}") + + +class ScheduleSampler(ABC): + """ + A distribution over timesteps in the diffusion process, intended to reduce + variance of the objective. + + By default, samplers perform unbiased importance sampling, in which the + objective's mean is unchanged. + However, subclasses may override sample() to change how the resampled + terms are reweighted, allowing for actual changes in the objective. + """ + + @abstractmethod + def weights(self): + """ + Get a numpy array of weights, one per diffusion step. + + The weights needn't be normalized, but must be positive. + """ + + def sample(self, batch_size, device): + """ + Importance-sample timesteps for a batch. + + :param batch_size: the number of timesteps. + :param device: the torch device to save to. + :return: a tuple (timesteps, weights): + - timesteps: a tensor of timestep indices. + - weights: a tensor of weights to scale the resulting losses. + """ + w = self.weights() + p = w / np.sum(w) + indices_np = np.random.choice(len(p), size=(batch_size,), p=p) + indices = th.from_numpy(indices_np).long().to(device) + weights_np = 1 / (len(p) * p[indices_np]) + weights = th.from_numpy(weights_np).float().to(device) + return indices, weights + + +class UniformSampler(ScheduleSampler): + def __init__(self, diffusion): + self.diffusion = diffusion + self._weights = np.ones([diffusion.num_timesteps]) + + def weights(self): + return self._weights + + +class LossAwareSampler(ScheduleSampler): + def update_with_local_losses(self, local_ts, local_losses): + """ + Update the reweighting using losses from a model. + + Call this method from each rank with a batch of timesteps and the + corresponding losses for each of those timesteps. + This method will perform synchronization to make sure all of the ranks + maintain the exact same reweighting. + + :param local_ts: an integer Tensor of timesteps. + :param local_losses: a 1D Tensor of losses. + """ + batch_sizes = [ + th.tensor([0], dtype=th.int32, device=local_ts.device) + for _ in range(dist.get_world_size()) + ] + dist.all_gather( + batch_sizes, + th.tensor([len(local_ts)], dtype=th.int32, device=local_ts.device), + ) + + # Pad all_gather batches to be the maximum batch size. + batch_sizes = [x.item() for x in batch_sizes] + max_bs = max(batch_sizes) + + timestep_batches = [th.zeros(max_bs).to(local_ts) for bs in batch_sizes] + loss_batches = [th.zeros(max_bs).to(local_losses) for bs in batch_sizes] + dist.all_gather(timestep_batches, local_ts) + dist.all_gather(loss_batches, local_losses) + timesteps = [ + x.item() for y, bs in zip(timestep_batches, batch_sizes) for x in y[:bs] + ] + losses = [x.item() for y, bs in zip(loss_batches, batch_sizes) for x in y[:bs]] + self.update_with_all_losses(timesteps, losses) + + @abstractmethod + def update_with_all_losses(self, ts, losses): + """ + Update the reweighting using losses from a model. + + Sub-classes should override this method to update the reweighting + using losses from the model. + + This method directly updates the reweighting without synchronizing + between workers. It is called by update_with_local_losses from all + ranks with identical arguments. Thus, it should have deterministic + behavior to maintain state across workers. + + :param ts: a list of int timesteps. + :param losses: a list of float losses, one per timestep. + """ + + +class LossSecondMomentResampler(LossAwareSampler): + def __init__(self, diffusion, history_per_term=10, uniform_prob=0.001): + self.diffusion = diffusion + self.history_per_term = history_per_term + self.uniform_prob = uniform_prob + self._loss_history = np.zeros( + [diffusion.num_timesteps, history_per_term], dtype=np.float64 + ) + self._loss_counts = np.zeros([diffusion.num_timesteps], dtype=np.int) + + def weights(self): + if not self._warmed_up(): + return np.ones([self.diffusion.num_timesteps], dtype=np.float64) + weights = np.sqrt(np.mean(self._loss_history ** 2, axis=-1)) + weights /= np.sum(weights) + weights *= 1 - self.uniform_prob + weights += self.uniform_prob / len(weights) + return weights + + def update_with_all_losses(self, ts, losses): + for t, loss in zip(ts, losses): + if self._loss_counts[t] == self.history_per_term: + # Shift out the oldest loss term. + self._loss_history[t, :-1] = self._loss_history[t, 1:] + self._loss_history[t, -1] = loss + else: + self._loss_history[t, self._loss_counts[t]] = loss + self._loss_counts[t] += 1 + + def _warmed_up(self): + return (self._loss_counts == self.history_per_term).all() diff --git a/diffusion/respace.py b/diffusion/respace.py new file mode 100644 index 0000000000000000000000000000000000000000..13a3c0667029b75aa82202ef709fc7cb2fb337f4 --- /dev/null +++ b/diffusion/respace.py @@ -0,0 +1,129 @@ +# This code is based on https://github.com/openai/guided-diffusion +import numpy as np +import torch as th + +from .gaussian_diffusion import GaussianDiffusion + + +def space_timesteps(num_timesteps, section_counts): + """ + Create a list of timesteps to use from an original diffusion process, + given the number of timesteps we want to take from equally-sized portions + of the original process. + + For example, if there's 300 timesteps and the section counts are [10,15,20] + then the first 100 timesteps are strided to be 10 timesteps, the second 100 + are strided to be 15 timesteps, and the final 100 are strided to be 20. + + If the stride is a string starting with "ddim", then the fixed striding + from the DDIM paper is used, and only one section is allowed. + + :param num_timesteps: the number of diffusion steps in the original + process to divide up. + :param section_counts: either a list of numbers, or a string containing + comma-separated numbers, indicating the step count + per section. As a special case, use "ddimN" where N + is a number of steps to use the striding from the + DDIM paper. + :return: a set of diffusion steps from the original process to use. + """ + if isinstance(section_counts, str): + if section_counts.startswith("ddim"): + desired_count = int(section_counts[len("ddim") :]) + for i in range(1, num_timesteps): + if len(range(0, num_timesteps, i)) == desired_count: + return set(range(0, num_timesteps, i)) + raise ValueError( + f"cannot create exactly {num_timesteps} steps with an integer stride" + ) + section_counts = [int(x) for x in section_counts.split(",")] + size_per = num_timesteps // len(section_counts) + extra = num_timesteps % len(section_counts) + start_idx = 0 + all_steps = [] + for i, section_count in enumerate(section_counts): + size = size_per + (1 if i < extra else 0) + if size < section_count: + raise ValueError( + f"cannot divide section of {size} steps into {section_count}" + ) + if section_count <= 1: + frac_stride = 1 + else: + frac_stride = (size - 1) / (section_count - 1) + cur_idx = 0.0 + taken_steps = [] + for _ in range(section_count): + taken_steps.append(start_idx + round(cur_idx)) + cur_idx += frac_stride + all_steps += taken_steps + start_idx += size + return set(all_steps) + + +class SpacedDiffusion(GaussianDiffusion): + """ + A diffusion process which can skip steps in a base diffusion process. + + :param use_timesteps: a collection (sequence or set) of timesteps from the + original diffusion process to retain. + :param kwargs: the kwargs to create the base diffusion process. + """ + + def __init__(self, use_timesteps, **kwargs): + self.use_timesteps = set(use_timesteps) + self.timestep_map = [] + self.original_num_steps = len(kwargs["betas"]) + + base_diffusion = GaussianDiffusion(**kwargs) # pylint: disable=missing-kwoa + last_alpha_cumprod = 1.0 + new_betas = [] + for i, alpha_cumprod in enumerate(base_diffusion.alphas_cumprod): + if i in self.use_timesteps: + new_betas.append(1 - alpha_cumprod / last_alpha_cumprod) + last_alpha_cumprod = alpha_cumprod + self.timestep_map.append(i) + kwargs["betas"] = np.array(new_betas) + super().__init__(**kwargs) + + def p_mean_variance( + self, model, *args, **kwargs + ): # pylint: disable=signature-differs + return super().p_mean_variance(self._wrap_model(model), *args, **kwargs) + + def training_losses( + self, model, *args, **kwargs + ): # pylint: disable=signature-differs + return super().training_losses(self._wrap_model(model), *args, **kwargs) + + def condition_mean(self, cond_fn, *args, **kwargs): + return super().condition_mean(self._wrap_model(cond_fn), *args, **kwargs) + + def condition_score(self, cond_fn, *args, **kwargs): + return super().condition_score(self._wrap_model(cond_fn), *args, **kwargs) + + def _wrap_model(self, model): + if isinstance(model, _WrappedModel): + return model + return _WrappedModel( + model, self.timestep_map, self.rescale_timesteps, self.original_num_steps + ) + + def _scale_timesteps(self, t): + # Scaling is done by the wrapped model. + return t + + +class _WrappedModel: + def __init__(self, model, timestep_map, rescale_timesteps, original_num_steps): + self.model = model + self.timestep_map = timestep_map + self.rescale_timesteps = rescale_timesteps + self.original_num_steps = original_num_steps + + def __call__(self, x, ts, **kwargs): + map_tensor = th.tensor(self.timestep_map, device=ts.device, dtype=ts.dtype) + new_ts = map_tensor[ts] + if self.rescale_timesteps: + new_ts = new_ts.float() * (1000.0 / self.original_num_steps) + return self.model(x, new_ts, **kwargs) diff --git a/diffusion_rvqvae_trainer.py b/diffusion_rvqvae_trainer.py new file mode 100644 index 0000000000000000000000000000000000000000..259b073a18dca2513bd56579578a0d74430f780a --- /dev/null +++ b/diffusion_rvqvae_trainer.py @@ -0,0 +1,736 @@ +import train +import os +import time +import csv +import sys +import warnings +import random +import torch +import torch.nn as nn +import torch.nn.functional as F +from torch.utils.tensorboard import SummaryWriter +from torch.nn.parallel import DistributedDataParallel as DDP +import numpy as np +import time +import pprint +from loguru import logger +from utils import rotation_conversions as rc +import smplx +from utils import config, logger_tools, other_tools, metric, data_transfer +from dataloaders import data_tools +from optimizers.optim_factory import create_optimizer +from optimizers.scheduler_factory import create_scheduler +from optimizers.loss_factory import get_loss_func +from dataloaders.data_tools import joints_list +import librosa +from diffusion.model_util import create_gaussian_diffusion +from diffusion.resample import create_named_schedule_sampler +from models.vq.model import RVQVAE +import pickle +from models.motionclip import get_model +import clip + + +class CustomTrainer(train.BaseTrainer): + ''' + Multi-Modal AutoEncoder + ''' + def __init__(self, args): + super().__init__(args) + self.args = args + self.joints = self.train_data.joints + self.ori_joint_list = joints_list[self.args.ori_joints] + self.tar_joint_list_face = joints_list["beat_smplx_face"] + self.tar_joint_list_upper = joints_list["beat_smplx_upper"] + self.tar_joint_list_hands = joints_list["beat_smplx_hands"] + self.tar_joint_list_lower = joints_list["beat_smplx_lower"] + + self.joint_mask_face = np.zeros(len(list(self.ori_joint_list.keys()))*3) + self.joints = 55 + for joint_name in self.tar_joint_list_face: + self.joint_mask_face[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + self.joint_mask_upper = np.zeros(len(list(self.ori_joint_list.keys()))*3) + for joint_name in self.tar_joint_list_upper: + self.joint_mask_upper[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + self.joint_mask_hands = np.zeros(len(list(self.ori_joint_list.keys()))*3) + for joint_name in self.tar_joint_list_hands: + self.joint_mask_hands[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + self.joint_mask_lower = np.zeros(len(list(self.ori_joint_list.keys()))*3) + for joint_name in self.tar_joint_list_lower: + self.joint_mask_lower[self.ori_joint_list[joint_name][1] - self.ori_joint_list[joint_name][0]:self.ori_joint_list[joint_name][1]] = 1 + + self.tracker = other_tools.EpochTracker(["fid", "l1div", "bc", "rec", "trans", "vel", "transv", 'dis', 'gen', 'acc', 'transa', 'exp', 'lvd', 'mse', "cls", "rec_face", "latent", "cls_full", "cls_self", "cls_word", "latent_word","latent_self","predict_x0_loss"], [False,True,True, False, False, False, False, False, False, False, False, False, False, False, False, False, False,False, False, False,False,False,False]) + + vq_model_module = __import__(f"models.motion_representation", fromlist=["something"]) + self.args.vae_layer = 2 + self.args.vae_length = 256 + self.args.vae_test_dim = 106 + self.vq_model_face = getattr(vq_model_module, "VQVAEConvZero")(self.args).to(self.rank) + other_tools.load_checkpoints(self.vq_model_face, "./datasets/hub/pretrained_vq/face_vertex_1layer_790.bin", args.e_name) + + + vq_type = self.args.vqvae_type + if vq_type=="vqvae": + + self.args.vae_layer = 4 + self.args.vae_test_dim = 78 + self.vq_model_upper = getattr(vq_model_module, "VQVAEConvZero")(self.args).to(self.rank) + other_tools.load_checkpoints(self.vq_model_upper, args.vqvae_upper_path, args.e_name) + self.args.vae_test_dim = 180 + self.vq_model_hands = getattr(vq_model_module, "VQVAEConvZero")(self.args).to(self.rank) + other_tools.load_checkpoints(self.vq_model_hands, args.vqvae_hands_path, args.e_name) + self.args.vae_test_dim = 54 + self.args.vae_layer = 4 + self.vq_model_lower = getattr(vq_model_module, "VQVAEConvZero")(self.args).to(self.rank) + other_tools.load_checkpoints(self.vq_model_lower, args.vqvae_lower_path, args.e_name) + + elif vq_type=="rvqvae": + + args.num_quantizers = 6 + args.shared_codebook = False + args.quantize_dropout_prob = 0.2 + args.mu = 0.99 + + args.nb_code = 512 + args.code_dim = 512 + args.code_dim = 512 + args.down_t = 2 + args.stride_t = 2 + args.width = 512 + args.depth = 3 + args.dilation_growth_rate = 3 + args.vq_act = "relu" + args.vq_norm = None + + dim_pose = 78 + args.body_part = "upper" + self.vq_model_upper = RVQVAE(args, + dim_pose, + args.nb_code, + args.code_dim, + args.code_dim, + args.down_t, + args.stride_t, + args.width, + args.depth, + args.dilation_growth_rate, + args.vq_act, + args.vq_norm) + + dim_pose = 180 + args.body_part = "hands" + self.vq_model_hands = RVQVAE(args, + dim_pose, + args.nb_code, + args.code_dim, + args.code_dim, + args.down_t, + args.stride_t, + args.width, + args.depth, + args.dilation_growth_rate, + args.vq_act, + args.vq_norm) + + dim_pose = 54 + if args.use_trans: + dim_pose = 57 + self.args.vqvae_lower_path = self.args.vqvae_lower_trans_path + args.body_part = "lower" + self.vq_model_lower = RVQVAE(args, + dim_pose, + args.nb_code, + args.code_dim, + args.code_dim, + args.down_t, + args.stride_t, + args.width, + args.depth, + args.dilation_growth_rate, + args.vq_act, + args.vq_norm) + + self.vq_model_upper.load_state_dict(torch.load(self.args.vqvae_upper_path)['net']) + self.vq_model_hands.load_state_dict(torch.load(self.args.vqvae_hands_path)['net']) + self.vq_model_lower.load_state_dict(torch.load(self.args.vqvae_lower_path)['net']) + + self.vqvae_latent_scale = self.args.vqvae_latent_scale + + self.vq_model_upper.eval().to(self.rank) + self.vq_model_hands.eval().to(self.rank) + self.vq_model_lower.eval().to(self.rank) + + + + + + self.args.vae_test_dim = 61 + self.args.vae_layer = 4 + self.args.vae_test_dim = 330 + self.args.vae_layer = 4 + self.args.vae_length = 240 + + + self.vq_model_face.eval() + self.vq_model_upper.eval() + self.vq_model_hands.eval() + self.vq_model_lower.eval() + + self.cls_loss = nn.NLLLoss().to(self.rank) + self.reclatent_loss = nn.MSELoss().to(self.rank) + self.vel_loss = torch.nn.L1Loss(reduction='mean').to(self.rank) + self.rec_loss = get_loss_func("GeodesicLoss").to(self.rank) + self.log_softmax = nn.LogSoftmax(dim=2).to(self.rank) + + self.diffusion = create_gaussian_diffusion() + self.schedule_sampler_type = 'uniform' + self.schedule_sampler = create_named_schedule_sampler(self.schedule_sampler_type, self.diffusion) + self.mean = np.load(args.mean_pose_path) + self.std = np.load(args.std_pose_path) + + self.use_trans = args.use_trans + if self.use_trans: + self.trans_mean = np.load(args.mean_trans_path) + self.trans_std = np.load(args.std_trans_path) + self.trans_mean = torch.from_numpy(self.trans_mean).cuda() + self.trans_std = torch.from_numpy(self.trans_std).cuda() + + + joints = [3,6,9,12,13,14,15,16,17,18,19,20,21] + upper_body_mask = [] + for i in joints: + upper_body_mask.extend([i*6, i*6+1, i*6+2, i*6+3, i*6+4, i*6+5]) + + joints = list(range(25,55)) + hands_body_mask = [] + for i in joints: + hands_body_mask.extend([i*6, i*6+1, i*6+2, i*6+3, i*6+4, i*6+5]) + + joints = [0,1,2,4,5,7,8,10,11] + lower_body_mask = [] + for i in joints: + lower_body_mask.extend([i*6, i*6+1, i*6+2, i*6+3, i*6+4, i*6+5]) + + self.mean_upper = self.mean[upper_body_mask] + self.mean_hands = self.mean[hands_body_mask] + self.mean_lower = self.mean[lower_body_mask] + self.std_upper = self.std[upper_body_mask] + self.std_hands = self.std[hands_body_mask] + self.std_lower = self.std[lower_body_mask] + + self.mean_upper = torch.from_numpy(self.mean_upper).cuda() + self.mean_hands = torch.from_numpy(self.mean_hands).cuda() + self.mean_lower = torch.from_numpy(self.mean_lower).cuda() + self.std_upper = torch.from_numpy(self.std_upper).cuda() + self.std_hands = torch.from_numpy(self.std_hands).cuda() + self.std_lower = torch.from_numpy(self.std_lower).cuda() + + + def inverse_selection(self, filtered_t, selection_array, n): + original_shape_t = np.zeros((n, selection_array.size)) + selected_indices = np.where(selection_array == 1)[0] + for i in range(n): + original_shape_t[i, selected_indices] = filtered_t[i] + return original_shape_t + + def inverse_selection_tensor(self, filtered_t, selection_array, n): + selection_array = torch.from_numpy(selection_array).cuda() + original_shape_t = torch.zeros((n, 165)).cuda() + selected_indices = torch.where(selection_array == 1)[0] + for i in range(n): + original_shape_t[i, selected_indices] = filtered_t[i] + return original_shape_t + + def _load_data(self, dict_data): + tar_pose_raw = dict_data["pose"] + tar_pose = tar_pose_raw[:, :, :165].to(self.rank) + tar_contact = tar_pose_raw[:, :, 165:169].to(self.rank) + tar_trans = dict_data["trans"].to(self.rank) + tar_trans_v = dict_data["trans_v"].to(self.rank) + tar_exps = dict_data["facial"].to(self.rank) + in_audio = dict_data["audio"].to(self.rank) + in_word = dict_data["word"].to(self.rank) + tar_beta = dict_data["beta"].to(self.rank) + tar_id = dict_data["id"].to(self.rank).long() + bs, n, j = tar_pose.shape[0], tar_pose.shape[1], self.joints + + tar_pose_jaw = tar_pose[:, :, 66:69] + tar_pose_jaw = rc.axis_angle_to_matrix(tar_pose_jaw.reshape(bs, n, 1, 3)) + tar_pose_jaw = rc.matrix_to_rotation_6d(tar_pose_jaw).reshape(bs, n, 1*6) + tar_pose_face = torch.cat([tar_pose_jaw, tar_exps], dim=2) + + tar_pose_hands = tar_pose[:, :, 25*3:55*3] + tar_pose_hands = rc.axis_angle_to_matrix(tar_pose_hands.reshape(bs, n, 30, 3)) + tar_pose_hands = rc.matrix_to_rotation_6d(tar_pose_hands).reshape(bs, n, 30*6) + + tar_pose_upper = tar_pose[:, :, self.joint_mask_upper.astype(bool)] + tar_pose_upper = rc.axis_angle_to_matrix(tar_pose_upper.reshape(bs, n, 13, 3)) + tar_pose_upper = rc.matrix_to_rotation_6d(tar_pose_upper).reshape(bs, n, 13*6) + + tar_pose_leg = tar_pose[:, :, self.joint_mask_lower.astype(bool)] + tar_pose_leg = rc.axis_angle_to_matrix(tar_pose_leg.reshape(bs, n, 9, 3)) + tar_pose_leg = rc.matrix_to_rotation_6d(tar_pose_leg).reshape(bs, n, 9*6) + + tar_pose_lower = tar_pose_leg + + + tar4dis = torch.cat([tar_pose_jaw, tar_pose_upper, tar_pose_hands, tar_pose_leg], dim=2) + + + if self.args.pose_norm: + tar_pose_upper = (tar_pose_upper - self.mean_upper) / self.std_upper + tar_pose_hands = (tar_pose_hands - self.mean_hands) / self.std_hands + tar_pose_lower = (tar_pose_lower - self.mean_lower) / self.std_lower + + if self.use_trans: + tar_trans_v = (tar_trans_v - self.trans_mean)/self.trans_std + tar_pose_lower = torch.cat([tar_pose_lower,tar_trans_v], dim=-1) + + latent_face_top = self.vq_model_face.map2latent(tar_pose_face) # bs*n/4 + latent_upper_top = self.vq_model_upper.map2latent(tar_pose_upper) + latent_hands_top = self.vq_model_hands.map2latent(tar_pose_hands) + latent_lower_top = self.vq_model_lower.map2latent(tar_pose_lower) + + latent_in = torch.cat([latent_upper_top, latent_hands_top, latent_lower_top], dim=2)/self.args.vqvae_latent_scale + + + tar_pose_6d = rc.axis_angle_to_matrix(tar_pose.reshape(bs, n, 55, 3)) + tar_pose_6d = rc.matrix_to_rotation_6d(tar_pose_6d).reshape(bs, n, 55*6) + latent_all = torch.cat([tar_pose_6d, tar_trans, tar_contact], dim=-1) + style_feature = None + if self.args.use_motionclip: + motionclip_feat = tar_pose_6d[...,:22*6] + batch = {} + bs,seq,feat = motionclip_feat.shape + batch['x']=motionclip_feat.permute(0,2,1).contiguous() + batch['y']=torch.zeros(bs).int().cuda() + batch['mask']=torch.ones([bs,seq]).bool().cuda() + style_feature = self.motionclip.encoder(batch)['mu'].detach().float() + + + + # print(tar_index_value_upper_top.shape, index_in.shape) + return { + "tar_pose_jaw": tar_pose_jaw, + "tar_pose_face": tar_pose_face, + "tar_pose_upper": tar_pose_upper, + "tar_pose_lower": tar_pose_lower, + "tar_pose_hands": tar_pose_hands, + 'tar_pose_leg': tar_pose_leg, + "in_audio": in_audio, + "in_word": in_word, + "tar_trans": tar_trans, + "tar_exps": tar_exps, + "tar_beta": tar_beta, + "tar_pose": tar_pose, + "tar4dis": tar4dis, + "latent_face_top": latent_face_top, + "latent_upper_top": latent_upper_top, + "latent_hands_top": latent_hands_top, + "latent_lower_top": latent_lower_top, + "latent_in": latent_in, + "tar_id": tar_id, + "latent_all": latent_all, + "tar_pose_6d": tar_pose_6d, + "tar_contact": tar_contact, + "style_feature":style_feature, + } + + def _g_training(self, loaded_data, use_adv, mode="train", epoch=0): + bs, n, j = loaded_data["tar_pose"].shape[0], loaded_data["tar_pose"].shape[1], self.joints + + cond_ = {'y':{}} + cond_['y']['audio'] = loaded_data['in_audio'] + cond_['y']['word'] = loaded_data['in_word'] + cond_['y']['id'] = loaded_data['tar_id'] + cond_['y']['seed'] = loaded_data['latent_in'][:,:self.args.pre_frames] + cond_['y']['mask'] = (torch.zeros([self.args.batch_size, 1, 1, self.args.pose_length//self.args.vqvae_squeeze_scale]) < 1).cuda() + cond_['y']['style_feature'] = loaded_data['style_feature'] + x0 = loaded_data['latent_in'] + x0 = x0.permute(0, 2, 1).unsqueeze(2) + t, weights = self.schedule_sampler.sample(x0.shape[0], x0.device) + g_loss_final = self.diffusion.training_losses(self.model,x0,t,model_kwargs = cond_)["loss"].mean() + self.tracker.update_meter("predict_x0_loss", "train", g_loss_final.item()) + + if mode == 'train': + return g_loss_final + + + def _g_test(self, loaded_data): + + sample_fn = self.diffusion.p_sample_loop + + mode = 'test' + bs, n, j = loaded_data["tar_pose"].shape[0], loaded_data["tar_pose"].shape[1], self.joints + tar_pose = loaded_data["tar_pose"] + tar_beta = loaded_data["tar_beta"] + tar_exps = loaded_data["tar_exps"] + tar_contact = loaded_data["tar_contact"] + tar_trans = loaded_data["tar_trans"] + in_word = loaded_data["in_word"] + in_audio = loaded_data["in_audio"] + in_x0 = loaded_data['latent_in'] + in_seed = loaded_data['latent_in'] + + remain = n%8 + if remain != 0: + tar_pose = tar_pose[:, :-remain, :] + tar_beta = tar_beta[:, :-remain, :] + tar_trans = tar_trans[:, :-remain, :] + in_word = in_word[:, :-remain] + tar_exps = tar_exps[:, :-remain, :] + tar_contact = tar_contact[:, :-remain, :] + in_x0 = in_x0[:, :in_x0.shape[1]-(remain//self.args.vqvae_squeeze_scale), :] + in_seed = in_seed[:, :in_x0.shape[1]-(remain//self.args.vqvae_squeeze_scale), :] + n = n - remain + + tar_pose_jaw = tar_pose[:, :, 66:69] + tar_pose_jaw = rc.axis_angle_to_matrix(tar_pose_jaw.reshape(bs, n, 1, 3)) + tar_pose_jaw = rc.matrix_to_rotation_6d(tar_pose_jaw).reshape(bs, n, 1*6) + tar_pose_face = torch.cat([tar_pose_jaw, tar_exps], dim=2) + + tar_pose_hands = tar_pose[:, :, 25*3:55*3] + tar_pose_hands = rc.axis_angle_to_matrix(tar_pose_hands.reshape(bs, n, 30, 3)) + tar_pose_hands = rc.matrix_to_rotation_6d(tar_pose_hands).reshape(bs, n, 30*6) + + tar_pose_upper = tar_pose[:, :, self.joint_mask_upper.astype(bool)] + tar_pose_upper = rc.axis_angle_to_matrix(tar_pose_upper.reshape(bs, n, 13, 3)) + tar_pose_upper = rc.matrix_to_rotation_6d(tar_pose_upper).reshape(bs, n, 13*6) + + tar_pose_leg = tar_pose[:, :, self.joint_mask_lower.astype(bool)] + tar_pose_leg = rc.axis_angle_to_matrix(tar_pose_leg.reshape(bs, n, 9, 3)) + tar_pose_leg = rc.matrix_to_rotation_6d(tar_pose_leg).reshape(bs, n, 9*6) + tar_pose_lower = torch.cat([tar_pose_leg, tar_trans, tar_contact], dim=2) + + tar_pose_6d = rc.axis_angle_to_matrix(tar_pose.reshape(bs, n, 55, 3)) + tar_pose_6d = rc.matrix_to_rotation_6d(tar_pose_6d).reshape(bs, n, 55*6) + latent_all = torch.cat([tar_pose_6d, tar_trans, tar_contact], dim=-1) + + rec_all_face = [] + rec_all_upper = [] + rec_all_lower = [] + rec_all_hands = [] + vqvae_squeeze_scale = self.args.vqvae_squeeze_scale + roundt = (n - self.args.pre_frames * vqvae_squeeze_scale) // (self.args.pose_length - self.args.pre_frames * vqvae_squeeze_scale) + remain = (n - self.args.pre_frames * vqvae_squeeze_scale) % (self.args.pose_length - self.args.pre_frames * vqvae_squeeze_scale) + round_l = self.args.pose_length - self.args.pre_frames * vqvae_squeeze_scale + + + for i in range(0, roundt): + in_word_tmp = in_word[:, i*(round_l):(i+1)*(round_l)+self.args.pre_frames * vqvae_squeeze_scale] + + in_audio_tmp = in_audio[:, i*(16000//30*round_l):(i+1)*(16000//30*round_l)+16000//30*self.args.pre_frames * vqvae_squeeze_scale] + in_id_tmp = loaded_data['tar_id'][:, i*(round_l):(i+1)*(round_l)+self.args.pre_frames] + in_seed_tmp = in_seed[:, i*(round_l)//vqvae_squeeze_scale:(i+1)*(round_l)//vqvae_squeeze_scale+self.args.pre_frames] + in_x0_tmp = in_x0[:, i*(round_l)//vqvae_squeeze_scale:(i+1)*(round_l)//vqvae_squeeze_scale+self.args.pre_frames] + mask_val = torch.ones(bs, self.args.pose_length, self.args.pose_dims+3+4).float().cuda() + mask_val[:, :self.args.pre_frames, :] = 0.0 + if i == 0: + in_seed_tmp = in_seed_tmp[:, :self.args.pre_frames, :] + else: + in_seed_tmp = last_sample[:, -self.args.pre_frames:, :] + + cond_ = {'y':{}} + cond_['y']['audio'] = in_audio_tmp + cond_['y']['word'] = in_word_tmp + cond_['y']['id'] = in_id_tmp + cond_['y']['seed'] =in_seed_tmp + cond_['y']['mask'] = (torch.zeros([self.args.batch_size, 1, 1, self.args.pose_length]) < 1).cuda() + + + + cond_['y']['style_feature'] = torch.zeros([bs, 512]).cuda() + + shape_ = (bs, 1536, 1, 32) + sample = sample_fn( + self.model, + shape_, + clip_denoised=False, + model_kwargs=cond_, + skip_timesteps=0, + init_image=None, + progress=True, + dump_steps=None, + noise=None, + const_noise=False, + ) + sample = sample.squeeze().permute(1,0).unsqueeze(0) + + last_sample = sample.clone() + + rec_latent_upper = sample[...,:512] + rec_latent_hands = sample[...,512:1024] + rec_latent_lower = sample[...,1024:1536] + + + + if i == 0: + rec_all_upper.append(rec_latent_upper) + rec_all_hands.append(rec_latent_hands) + rec_all_lower.append(rec_latent_lower) + else: + rec_all_upper.append(rec_latent_upper[:, self.args.pre_frames:]) + rec_all_hands.append(rec_latent_hands[:, self.args.pre_frames:]) + rec_all_lower.append(rec_latent_lower[:, self.args.pre_frames:]) + + rec_all_upper = torch.cat(rec_all_upper, dim=1) * self.vqvae_latent_scale + rec_all_hands = torch.cat(rec_all_hands, dim=1) * self.vqvae_latent_scale + rec_all_lower = torch.cat(rec_all_lower, dim=1) * self.vqvae_latent_scale + + rec_upper = self.vq_model_upper.latent2origin(rec_all_upper)[0] + rec_hands = self.vq_model_hands.latent2origin(rec_all_hands)[0] + rec_lower = self.vq_model_lower.latent2origin(rec_all_lower)[0] + + + if self.use_trans: + rec_trans_v = rec_lower[...,-3:] + rec_trans_v = rec_trans_v * self.trans_std + self.trans_mean + rec_trans = torch.zeros_like(rec_trans_v) + rec_trans = torch.cumsum(rec_trans_v, dim=-2) + rec_trans[...,1]=rec_trans_v[...,1] + rec_lower = rec_lower[...,:-3] + + if self.args.pose_norm: + rec_upper = rec_upper * self.std_upper + self.mean_upper + rec_hands = rec_hands * self.std_hands + self.mean_hands + rec_lower = rec_lower * self.std_lower + self.mean_lower + + + + + n = n - remain + tar_pose = tar_pose[:, :n, :] + tar_exps = tar_exps[:, :n, :] + tar_trans = tar_trans[:, :n, :] + tar_beta = tar_beta[:, :n, :] + + + rec_exps = tar_exps + #rec_pose_jaw = rec_face[:, :, :6] + rec_pose_legs = rec_lower[:, :, :54] + bs, n = rec_pose_legs.shape[0], rec_pose_legs.shape[1] + rec_pose_upper = rec_upper.reshape(bs, n, 13, 6) + rec_pose_upper = rc.rotation_6d_to_matrix(rec_pose_upper)# + rec_pose_upper = rc.matrix_to_axis_angle(rec_pose_upper).reshape(bs*n, 13*3) + rec_pose_upper_recover = self.inverse_selection_tensor(rec_pose_upper, self.joint_mask_upper, bs*n) + rec_pose_lower = rec_pose_legs.reshape(bs, n, 9, 6) + rec_pose_lower = rc.rotation_6d_to_matrix(rec_pose_lower) + rec_lower2global = rc.matrix_to_rotation_6d(rec_pose_lower.clone()).reshape(bs, n, 9*6) + rec_pose_lower = rc.matrix_to_axis_angle(rec_pose_lower).reshape(bs*n, 9*3) + rec_pose_lower_recover = self.inverse_selection_tensor(rec_pose_lower, self.joint_mask_lower, bs*n) + rec_pose_hands = rec_hands.reshape(bs, n, 30, 6) + rec_pose_hands = rc.rotation_6d_to_matrix(rec_pose_hands) + rec_pose_hands = rc.matrix_to_axis_angle(rec_pose_hands).reshape(bs*n, 30*3) + rec_pose_hands_recover = self.inverse_selection_tensor(rec_pose_hands, self.joint_mask_hands, bs*n) + rec_pose = rec_pose_upper_recover + rec_pose_lower_recover + rec_pose_hands_recover + rec_pose[:, 66:69] = tar_pose.reshape(bs*n, 55*3)[:, 66:69] + + rec_pose = rc.axis_angle_to_matrix(rec_pose.reshape(bs*n, j, 3)) + rec_pose = rc.matrix_to_rotation_6d(rec_pose).reshape(bs, n, j*6) + tar_pose = rc.axis_angle_to_matrix(tar_pose.reshape(bs*n, j, 3)) + tar_pose = rc.matrix_to_rotation_6d(tar_pose).reshape(bs, n, j*6) + + return { + 'rec_pose': rec_pose, + 'rec_trans': rec_trans, + 'tar_pose': tar_pose, + 'tar_exps': tar_exps, + 'tar_beta': tar_beta, + 'tar_trans': tar_trans, + 'rec_exps': rec_exps, + } + + def train(self, epoch): + + use_adv = bool(epoch>=self.args.no_adv_epoch) + self.model.train() + t_start = time.time() + self.tracker.reset() + for its, batch_data in enumerate(self.train_loader): + loaded_data = self._load_data(batch_data) + t_data = time.time() - t_start + + self.opt.zero_grad() + g_loss_final = 0 + g_loss_final += self._g_training(loaded_data, use_adv, 'train', epoch) + + g_loss_final.backward() + if self.args.grad_norm != 0: + torch.nn.utils.clip_grad_norm_(self.model.parameters(), self.args.grad_norm) + self.opt.step() + + mem_cost = torch.cuda.memory_cached() / 1E9 + lr_g = self.opt.param_groups[0]['lr'] + + t_train = time.time() - t_start - t_data + t_start = time.time() + if its % self.args.log_period == 0: + self.train_recording(epoch, its, t_data, t_train, mem_cost, lr_g) + if self.args.debug: + if its == 1: break + self.opt_s.step(epoch) + + + + + + def test(self, epoch): + + results_save_path = self.checkpoint_path + f"/{epoch}/" + if os.path.exists(results_save_path): + return 0 + os.makedirs(results_save_path) + start_time = time.time() + total_length = 0 + test_seq_list = self.test_data.selected_file + align = 0 + latent_out = [] + latent_ori = [] + l2_all = 0 + lvel = 0 + self.model.eval() + self.smplx.eval() + self.eval_copy.eval() + with torch.no_grad(): + for its, batch_data in enumerate(self.test_loader): + loaded_data = self._load_data(batch_data) + net_out = self._g_test(loaded_data) + tar_pose = net_out['tar_pose'] + rec_pose = net_out['rec_pose'] + tar_exps = net_out['tar_exps'] + tar_beta = net_out['tar_beta'] + rec_trans = net_out['rec_trans'] + tar_trans = net_out['tar_trans'] + rec_exps = net_out['rec_exps'] + bs, n, j = tar_pose.shape[0], tar_pose.shape[1], self.joints + if (30/self.args.pose_fps) != 1: + assert 30%self.args.pose_fps == 0 + n *= int(30/self.args.pose_fps) + tar_pose = torch.nn.functional.interpolate(tar_pose.permute(0, 2, 1), scale_factor=30/self.args.pose_fps, mode='linear').permute(0,2,1) + rec_pose = torch.nn.functional.interpolate(rec_pose.permute(0, 2, 1), scale_factor=30/self.args.pose_fps, mode='linear').permute(0,2,1) + + + rec_pose = rc.rotation_6d_to_matrix(rec_pose.reshape(bs*n, j, 6)) + rec_pose = rc.matrix_to_rotation_6d(rec_pose).reshape(bs, n, j*6) + tar_pose = rc.rotation_6d_to_matrix(tar_pose.reshape(bs*n, j, 6)) + tar_pose = rc.matrix_to_rotation_6d(tar_pose).reshape(bs, n, j*6) + remain = n%self.args.vae_test_len + latent_out.append(self.eval_copy.map2latent(rec_pose[:, :n-remain]).reshape(-1, self.args.vae_length).detach().cpu().numpy()) # bs * n/8 * 240 + latent_ori.append(self.eval_copy.map2latent(tar_pose[:, :n-remain]).reshape(-1, self.args.vae_length).detach().cpu().numpy()) + + rec_pose = rc.rotation_6d_to_matrix(rec_pose.reshape(bs*n, j, 6)) + rec_pose = rc.matrix_to_axis_angle(rec_pose).reshape(bs*n, j*3) + tar_pose = rc.rotation_6d_to_matrix(tar_pose.reshape(bs*n, j, 6)) + tar_pose = rc.matrix_to_axis_angle(tar_pose).reshape(bs*n, j*3) + + vertices_rec = self.smplx( + betas=tar_beta.reshape(bs*n, 300), + transl=rec_trans.reshape(bs*n, 3)-rec_trans.reshape(bs*n, 3), + expression=tar_exps.reshape(bs*n, 100)-tar_exps.reshape(bs*n, 100), + jaw_pose=rec_pose[:, 66:69], + global_orient=rec_pose[:,:3], + body_pose=rec_pose[:,3:21*3+3], + left_hand_pose=rec_pose[:,25*3:40*3], + right_hand_pose=rec_pose[:,40*3:55*3], + return_joints=True, + leye_pose=rec_pose[:, 69:72], + reye_pose=rec_pose[:, 72:75], + ) + + vertices_rec_face = self.smplx( + betas=tar_beta.reshape(bs*n, 300), + transl=rec_trans.reshape(bs*n, 3)-rec_trans.reshape(bs*n, 3), + expression=rec_exps.reshape(bs*n, 100), + jaw_pose=rec_pose[:, 66:69], + global_orient=rec_pose[:,:3]-rec_pose[:,:3], + body_pose=rec_pose[:,3:21*3+3]-rec_pose[:,3:21*3+3], + left_hand_pose=rec_pose[:,25*3:40*3]-rec_pose[:,25*3:40*3], + right_hand_pose=rec_pose[:,40*3:55*3]-rec_pose[:,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=rec_pose[:, 69:72]-rec_pose[:, 69:72], + reye_pose=rec_pose[:, 72:75]-rec_pose[:, 72:75], + ) + vertices_tar_face = self.smplx( + betas=tar_beta.reshape(bs*n, 300), + transl=tar_trans.reshape(bs*n, 3)-tar_trans.reshape(bs*n, 3), + expression=tar_exps.reshape(bs*n, 100), + jaw_pose=tar_pose[:, 66:69], + global_orient=tar_pose[:,:3]-tar_pose[:,:3], + body_pose=tar_pose[:,3:21*3+3]-tar_pose[:,3:21*3+3], + left_hand_pose=tar_pose[:,25*3:40*3]-tar_pose[:,25*3:40*3], + right_hand_pose=tar_pose[:,40*3:55*3]-tar_pose[:,40*3:55*3], + return_verts=True, + return_joints=True, + leye_pose=tar_pose[:, 69:72]-tar_pose[:, 69:72], + reye_pose=tar_pose[:, 72:75]-tar_pose[:, 72:75], + ) + joints_rec = vertices_rec["joints"].detach().cpu().numpy().reshape(1, n, 127*3)[0, :n, :55*3] + # joints_tar = vertices_tar["joints"].detach().cpu().numpy().reshape(1, n, 127*3)[0, :n, :55*3] + facial_rec = vertices_rec_face['vertices'].reshape(1, n, -1)[0, :n] + facial_tar = vertices_tar_face['vertices'].reshape(1, n, -1)[0, :n] + face_vel_loss = self.vel_loss(facial_rec[1:, :] - facial_tar[:-1, :], facial_tar[1:, :] - facial_tar[:-1, :]) + l2 = self.reclatent_loss(facial_rec, facial_tar) + l2_all += l2.item() * n + lvel += face_vel_loss.item() * n + + _ = self.l1_calculator.run(joints_rec) + if self.alignmenter is not None: + in_audio_eval, sr = librosa.load(self.args.data_path+"wave16k/"+test_seq_list.iloc[its]['id']+".wav") + in_audio_eval = librosa.resample(in_audio_eval, orig_sr=sr, target_sr=self.args.audio_sr) + a_offset = int(self.align_mask * (self.args.audio_sr / self.args.pose_fps)) + onset_bt = self.alignmenter.load_audio(in_audio_eval[:int(self.args.audio_sr / self.args.pose_fps*n)], a_offset, len(in_audio_eval)-a_offset, True) + beat_vel = self.alignmenter.load_pose(joints_rec, self.align_mask, n-self.align_mask, 30, True) + align += (self.alignmenter.calculate_align(onset_bt, beat_vel, 30) * (n-2*self.align_mask)) + + tar_pose_np = tar_pose.detach().cpu().numpy() + rec_pose_np = rec_pose.detach().cpu().numpy() + rec_trans_np = rec_trans.detach().cpu().numpy().reshape(bs*n, 3) + rec_exp_np = rec_exps.detach().cpu().numpy().reshape(bs*n, 100) + tar_exp_np = tar_exps.detach().cpu().numpy().reshape(bs*n, 100) + tar_trans_np = tar_trans.detach().cpu().numpy().reshape(bs*n, 3) + gt_npz = np.load(self.args.data_path+self.args.pose_rep +"/"+test_seq_list.iloc[its]['id']+".npz", allow_pickle=True) + np.savez(results_save_path+"gt_"+test_seq_list.iloc[its]['id']+'.npz', + betas=gt_npz["betas"], + poses=tar_pose_np, + expressions=tar_exp_np, + trans=tar_trans_np, + model='smplx2020', + gender='neutral', + mocap_frame_rate = 30 , + ) + np.savez(results_save_path+"res_"+test_seq_list.iloc[its]['id']+'.npz', + betas=gt_npz["betas"], + poses=rec_pose_np, + expressions=rec_exp_np, + trans=rec_trans_np, + model='smplx2020', + gender='neutral', + mocap_frame_rate = 30, + ) + total_length += n + + logger.info(f"l2 loss: {l2_all/total_length}") + logger.info(f"lvel loss: {lvel/total_length}") + + latent_out_all = np.concatenate(latent_out, axis=0) + latent_ori_all = np.concatenate(latent_ori, axis=0) + fid = data_tools.FIDCalculator.frechet_distance(latent_out_all, latent_ori_all) + logger.info(f"fid score: {fid}") + self.test_recording("fid", fid, epoch) + + align_avg = align/(total_length-2*len(self.test_loader)*self.align_mask) + logger.info(f"align score: {align_avg}") + self.test_recording("bc", align_avg, epoch) + + l1div = self.l1_calculator.avg() + logger.info(f"l1div score: {l1div}") + self.test_recording("l1div", l1div, epoch) + + #data_tools.result2target_vis(self.args.pose_version, results_save_path, results_save_path, self.test_demo, False) + end_time = time.time() - start_time + logger.info(f"total inference time: {int(end_time)} s for {int(total_length/self.args.pose_fps)} s motion") + + + + diff --git a/mean_std/beatx_2_330_mean.npy b/mean_std/beatx_2_330_mean.npy new file mode 100644 index 0000000000000000000000000000000000000000..0aee1651c67d50d686e91c1a696a6972087b13d2 --- /dev/null +++ b/mean_std/beatx_2_330_mean.npy @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6102c3f4d49fea07edcde67924967c12cdfe4bfbf768ce688a2b703f871b8401 +size 1448 diff --git a/mean_std/beatx_2_330_std.npy b/mean_std/beatx_2_330_std.npy new file mode 100644 index 0000000000000000000000000000000000000000..8fa5cbe4dc5e2b987011fc052978b02494d37a07 --- /dev/null +++ b/mean_std/beatx_2_330_std.npy @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:896296af04cd228db8d7482872bf0d555190788fd02b06f5a7ac1c16097b8386 +size 1448 diff --git a/mean_std/beatx_2_trans_mean.npy b/mean_std/beatx_2_trans_mean.npy new file mode 100644 index 0000000000000000000000000000000000000000..74e17ee7a1ad9ab9412a3c372f365f561c333977 --- /dev/null +++ b/mean_std/beatx_2_trans_mean.npy @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bc26022f3cc62c47439f7d3f4d2217cee51ac7f2c7b257f646c965916cf1fafb +size 152 diff --git a/mean_std/beatx_2_trans_std.npy b/mean_std/beatx_2_trans_std.npy new file mode 100644 index 0000000000000000000000000000000000000000..0960fd580d67de28634cc918f317e85c2c18d993 --- /dev/null +++ b/mean_std/beatx_2_trans_std.npy @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:592c00d106e18e5dfdc75f195944a5492cba21ea477d2d4bfdb5569c19de5a79 +size 152 diff --git a/models/denoiser.py b/models/denoiser.py new file mode 100644 index 0000000000000000000000000000000000000000..1849b16bf41c3658bd9d3ba1e310ba8f9c53c61b --- /dev/null +++ b/models/denoiser.py @@ -0,0 +1,367 @@ +import pdb + +import numpy as np +import torch +import torch.nn as nn +import torch.nn.functional as F +from .utils.layer import BasicBlock +from einops import rearrange +import pickle +from .timm_transformer.transformer import Block as mytimmBlock + +class MDM(nn.Module): + def __init__(self, args): + super().__init__() + + + njoints=768 + nfeats=1 + latent_dim=512 + ff_size=1024 + num_layers=8 + num_heads=4 + dropout=0.1 + ablation=None + activation="gelu" + legacy=False + data_rep='rot6d' + dataset='amass' + audio_feat_dim = 64 + emb_trans_dec=False + audio_rep='' + n_seed=8 + cond_mode='' + kargs={} + + if args.vqvae_type == 'rvqvae': + njoints = 1536 + elif args.vqvae_type == 'novqvae': + njoints = 312 + self.args= args + self.legacy = legacy + self.njoints = njoints + self.nfeats = nfeats + self.data_rep = data_rep + + self.latent_dim = latent_dim + + self.ff_size = ff_size + self.num_layers = num_layers + self.num_heads = num_heads + self.dropout = dropout + + self.ablation = ablation + self.activation = activation + self.action_emb = kargs.get('action_emb', None) + + self.input_feats = self.njoints * self.nfeats + + self.cond_mask_prob = kargs.get('cond_mask_prob', 0.3) + self.use_motionclip = args.use_motionclip + + if args.audio_rep == 'onset+amplitude': + self.WavEncoder = WavEncoder(args.audio_f,audio_in=2) + self.audio_feat_dim = args.audio_f + + self.text_encoder_body = nn.Linear(300, args.audio_f) + + with open(f"{args.data_path}weights/vocab.pkl", 'rb') as f: + self.lang_model = pickle.load(f) + pre_trained_embedding = self.lang_model.word_embedding_weights + self.text_pre_encoder_body = nn.Embedding.from_pretrained(torch.FloatTensor(pre_trained_embedding),freeze=args.t_fix_pre) + + + + + self.sequence_pos_encoder = PositionalEncoding(self.latent_dim, self.dropout) + self.emb_trans_dec = emb_trans_dec + + self.cond_mode = cond_mode + self.num_head = 8 + + self.mytimmblocks = nn.ModuleList([ + mytimmBlock(dim=self.latent_dim,num_heads=self.num_heads,mlp_ratio=self.ff_size//self.latent_dim,drop_path=self.dropout) #hidden是对应于输入x的维度，attn_heads应该是12，这里写1是为了方便调试流程 + for _ in range(self.num_layers)]) + + self.embed_timestep = TimestepEmbedder(self.latent_dim, self.sequence_pos_encoder) + self.n_seed = n_seed + + + self.style_dim = 64 + self.embed_style = nn.Linear(6, self.style_dim) + self.embed_text = nn.Linear(self.input_feats*4, self.latent_dim) + + + + self.output_process = OutputProcess(self.data_rep, self.input_feats, self.latent_dim, self.njoints, + self.nfeats) + + + self.rel_pos = SinusoidalEmbeddings(self.latent_dim // self.num_head) + self.input_process = InputProcess(self.data_rep, self.input_feats , self.latent_dim) + self.input_process2 = nn.Linear(self.latent_dim * 2 + self.audio_feat_dim, self.latent_dim) + if self.use_motionclip: + self.input_process3 = nn.Linear(self.latent_dim + 512, self.latent_dim) + + self.mix_audio_text = nn.Linear(args.audio_f+args.word_f,256) + + + + def mask_cond(self, cond, force_mask=False): + bs, d = cond.shape + if force_mask: + return torch.zeros_like(cond) + elif self.training and self.cond_mask_prob > 0.: + mask = torch.bernoulli(torch.ones(bs, device=cond.device) * self.cond_mask_prob).view(bs, 1) # 1-> use null_cond, 0-> use real cond + return cond * (1. - mask) + + else: + return cond + + def mask_cond_audio(self, cond, force_mask=False): + bs, d = cond.shape + if force_mask: + return torch.zeros_like(cond) + elif self.training and self.cond_mask_prob_audio > 0.: + mask = torch.bernoulli(torch.ones(bs, device=cond.device) * self.cond_mask_prob_audio).view(bs, 1) # 1-> use null_cond, 0-> use real cond + return cond * (1. - mask) + else: + return cond + + + def forward(self, x, timesteps, y=None,uncond_info=False): + """ + x: [batch_size, njoints, nfeats, max_frames], denoted x_t in the paper + timesteps: [batch_size] (int) + seed: [batch_size, njoints, nfeats] + """ + _,_,_,noise_length = x.shape + y = y.copy() + + bs, njoints, nfeats, nframes = x.shape # 300 ,1141, 1, 88 + emb_t = self.embed_timestep(timesteps) # [1, bs, d], (1, 2, 256) + + force_mask = y.get('uncond', False) # False + #force_mask=uncond_info + + if self.n_seed != 0: + embed_text = self.embed_text(y['seed'].reshape(bs, -1)) # (bs, 256-64) + emb_seed = embed_text + + audio_feat = self.WavEncoder(y['audio']).permute(1, 0, 2) + text_feat = self.text_pre_encoder_body(y['word']) + text_feat = self.text_encoder_body(text_feat).permute(1, 0, 2) + + at_feat = torch.cat([audio_feat,text_feat],dim=2) + at_feat = self.mix_audio_text(at_feat) + at_feat = F.avg_pool1d(at_feat.permute(1,2,0), self.args.vqvae_squeeze_scale).permute(2,0,1) + + # This part is test for timm transformer blocks + x = x.reshape(bs, njoints * nfeats, 1, nframes) # [300, 1141, 1, 88] -> [300, 1141, 1, 88] + # self-attention + x_ = self.input_process(x) # [300, 1141, 1, 88] -> [88, 300, 256] + + # local-cross-attention + + xseq = torch.cat((x_, at_feat), axis=2) # [88, 300, 256], [88, 300, 64] -> [88, 300, 320] + # all frames + embed_style_2 = (emb_seed + emb_t).repeat(nframes, 1, 1) # [300, 256] ,[1, 300, 256] -> [88, 300, 256] + xseq = torch.cat((embed_style_2, xseq), axis=2) # -> [88, 300, 576] + xseq = self.input_process2(xseq) #[88, 300, 576] -> [88, 300, 256] + + if self.use_motionclip: + xseq = torch.cat((xseq, self.mask_cond(y['style_feature'],force_mask).unsqueeze(0).repeat(nframes, 1, 1)), axis = 2) + xseq = self.input_process3(xseq) + + + # 下面10行都是位置编码,感觉加了会好一点点，不知道是不是错觉 + xseq = xseq.permute(1, 0, 2) # [88, 300, 256] -> [300, 88, 256] + xseq = xseq.view(bs, nframes, self.num_head, -1) # [300, 88, 256] -> [300, 88, 8, 32] + xseq = xseq.permute(0, 2, 1, 3) # [300, 88, 8, 32] -> [300, 8, 88, 32] + xseq = xseq.reshape(bs * self.num_head, nframes, -1) # [300, 8, 88, 32] -> [2400, 88, 32] + pos_emb = self.rel_pos(xseq) # (88, 32) + xseq, _ = apply_rotary_pos_emb(xseq, xseq, pos_emb) # [2400, 88, 32] + xseq_rpe = xseq.reshape(bs, self.num_head, nframes, -1) # [300, 8, 88, 32] + xseq = xseq_rpe.permute(0, 2, 1, 3) # [300, 8, 88, 32] -> [300, 88, 8, 32] + xseq = xseq.view(bs, nframes, -1) # [300, 88, 8, 32] -> [300, 88, 256] + + for block in self.mytimmblocks: + xseq = block(xseq) + + xseq = xseq.permute(1, 0, 2) # [300, 88, 256] -> [88 ,300, 256] + output = xseq + + + output = self.output_process(output) # [88, 300, 256] -> [300, 1141, 1, 88] + return output[...,:noise_length] + + + @staticmethod + def apply_rotary(x, sinusoidal_pos): + sin, cos = sinusoidal_pos + x1, x2 = x[..., 0::2], x[..., 1::2] + # 如果是旋转query key的话，下面这个直接cat就行，因为要进行矩阵乘法，最终会在这个维度求和。（只要保持query和key的最后一个dim的每一个位置对应上就可以） + # torch.cat([x1 * cos - x2 * sin, x2 * cos + x1 * sin], dim=-1) + # 如果是旋转value的话，下面这个stack后再flatten才可以，因为训练好的模型最后一个dim是两两之间交替的。 + return torch.stack([x1 * cos - x2 * sin, x2 * cos + x1 * sin], dim=-1).flatten(-2, -1) + + + +class PositionalEncoding(nn.Module): + def __init__(self, d_model, dropout=0.1, max_len=5000): + super(PositionalEncoding, self).__init__() + self.dropout = nn.Dropout(p=dropout) + + pe = torch.zeros(max_len, d_model) # (5000, 128) + position = torch.arange(0, max_len, dtype=torch.float).unsqueeze(1) # (5000, 1) + div_term = torch.exp(torch.arange(0, d_model, 2).float() * (-np.log(10000.0) / d_model)) + pe[:, 0::2] = torch.sin(position * div_term) + pe[:, 1::2] = torch.cos(position * div_term) + pe = pe.unsqueeze(0).transpose(0, 1) + + self.register_buffer('pe', pe) + + def forward(self, x): + # not used in the final model + x = x + self.pe[:x.shape[0], :] + return self.dropout(x) + + + +class TimestepEmbedder(nn.Module): + def __init__(self, latent_dim, sequence_pos_encoder): + super().__init__() + self.latent_dim = latent_dim + self.sequence_pos_encoder = sequence_pos_encoder + + time_embed_dim = self.latent_dim + self.time_embed = nn.Sequential( + nn.Linear(self.latent_dim, time_embed_dim), + nn.SiLU(), + nn.Linear(time_embed_dim, time_embed_dim), + ) + + def forward(self, timesteps): + return self.time_embed(self.sequence_pos_encoder.pe[timesteps]).permute(1, 0, 2) + + +class InputProcess(nn.Module): + def __init__(self, data_rep, input_feats, latent_dim): + super().__init__() + self.data_rep = data_rep + self.input_feats = input_feats + self.latent_dim = latent_dim + self.poseEmbedding = nn.Linear(self.input_feats, self.latent_dim) + if self.data_rep == 'rot_vel': + self.velEmbedding = nn.Linear(self.input_feats, self.latent_dim) + + def forward(self, x): + bs, njoints, nfeats, nframes = x.shape + x = x.permute((3, 0, 1, 2)).reshape(nframes, bs, njoints*nfeats) + + if self.data_rep in ['rot6d', 'xyz', 'hml_vec']: + x = self.poseEmbedding(x) # [seqlen, bs, d] + return x + elif self.data_rep == 'rot_vel': + first_pose = x[[0]] # [1, bs, 150] + first_pose = self.poseEmbedding(first_pose) # [1, bs, d] + vel = x[1:] # [seqlen-1, bs, 150] + vel = self.velEmbedding(vel) # [seqlen-1, bs, d] + return torch.cat((first_pose, vel), axis=0) # [seqlen, bs, d] + else: + raise ValueError + + +class OutputProcess(nn.Module): + def __init__(self, data_rep, input_feats, latent_dim, njoints, nfeats): + super().__init__() + self.data_rep = data_rep + self.input_feats = input_feats + self.latent_dim = latent_dim + self.njoints = njoints + self.nfeats = nfeats + self.poseFinal = nn.Linear(self.latent_dim, self.input_feats) + if self.data_rep == 'rot_vel': + self.velFinal = nn.Linear(self.latent_dim, self.input_feats) + + def forward(self, output): + nframes, bs, d = output.shape + if self.data_rep in ['rot6d', 'xyz', 'hml_vec']: + output = self.poseFinal(output) # [88, 300, 256] -> [88, 300, 1141] + elif self.data_rep == 'rot_vel': + first_pose = output[[0]] # [1, bs, d] + first_pose = self.poseFinal(first_pose) # [1, bs, 150] + vel = output[1:] # [seqlen-1, bs, d] + vel = self.velFinal(vel) # [seqlen-1, bs, 150] + output = torch.cat((first_pose, vel), axis=0) # [seqlen, bs, 150] + else: + raise ValueError + output = output.reshape(nframes, bs, self.njoints, self.nfeats) + output = output.permute(1, 2, 3, 0) # [bs, njoints, nfeats, nframes] + return output + + +class WavEncoder(nn.Module): + def __init__(self, out_dim, audio_in=1): + super().__init__() + self.out_dim = out_dim + self.feat_extractor = nn.Sequential( + BasicBlock(audio_in, out_dim//4, 15, 5, first_dilation=1700, downsample=True), + BasicBlock(out_dim//4, out_dim//4, 15, 6, first_dilation=0, downsample=True), + BasicBlock(out_dim//4, out_dim//4, 15, 1, first_dilation=7, ), + BasicBlock(out_dim//4, out_dim//2, 15, 6, first_dilation=0, downsample=True), + BasicBlock(out_dim//2, out_dim//2, 15, 1, first_dilation=7), + BasicBlock(out_dim//2, out_dim, 15, 3, first_dilation=0,downsample=True), + ) + def forward(self, wav_data): + if wav_data.dim() == 2: + wav_data = wav_data.unsqueeze(1) + else: + wav_data = wav_data.transpose(1, 2) + out = self.feat_extractor(wav_data) + return out.transpose(1, 2) + +class SinusoidalEmbeddings(nn.Module): + def __init__(self, dim): + super().__init__() + inv_freq = 1. / (10000 ** (torch.arange(0, dim, 2).float() / dim)) + self.register_buffer('inv_freq', inv_freq) + + def forward(self, x): + n = x.shape[-2] + t = torch.arange(n, device = x.device).type_as(self.inv_freq) + freqs = torch.einsum('i , j -> i j', t, self.inv_freq) + return torch.cat((freqs, freqs), dim=-1) + +def rotate_half(x): + x = rearrange(x, 'b ... (r d) -> b (...) r d', r = 2) + x1, x2 = x.unbind(dim = -2) + return torch.cat((-x2, x1), dim = -1) + +def apply_rotary_pos_emb(q, k, freqs): + q, k = map(lambda t: (t * freqs.cos()) + (rotate_half(t) * freqs.sin()), (q, k)) + return q, k + +if __name__ == '__main__': + ''' + cd ./main/model + python mdm.py + ''' + n_frames = 240 + + n_seed = 8 + + model = MDM(modeltype='', njoints=1140, nfeats=1, cond_mode = 'cross_local_attention5_style1', action_emb='tensor', audio_rep='mfcc', + arch='mytrans_enc', latent_dim=256, n_seed=n_seed, cond_mask_prob=0.1) + + x = torch.randn(2, 1140, 1, 88) + t = torch.tensor([12, 85]) + + model_kwargs_ = {'y': {}} + model_kwargs_['y']['mask'] = (torch.zeros([1, 1, 1, n_frames]) < 1) # [..., n_seed:] + model_kwargs_['y']['audio'] = torch.randn(2, 88, 13).permute(1, 0, 2) # [n_seed:, ...] + model_kwargs_['y']['style'] = torch.randn(2, 6) + model_kwargs_['y']['mask_local'] = torch.ones(2, 88).bool() + model_kwargs_['y']['seed'] = x[..., 0:n_seed] + y = model(x, t, model_kwargs_['y']) + print(y.shape) diff --git a/models/emage.py b/models/emage.py new file mode 100644 index 0000000000000000000000000000000000000000..fa4851ec619ad0c4ba9cdd9f7cd797fb004c7437 --- /dev/null +++ b/models/emage.py @@ -0,0 +1,248 @@ +import torch +import torch.nn as nn +import os +import math +import pickle +import numpy as np +import torch.nn.functional as F +from torch.nn.utils import weight_norm +from .utils.build_vocab import Vocab +from torch.nn import TransformerEncoder, TransformerEncoderLayer +from .utils.layer import BasicBlock, TextEncoderTCN +from .utils.wav2vec import Wav2Vec2Model +from .utils.layer import ResBlock, init_weight +from loguru import logger +from .motion_encoder import * +import copy + +class WavEncoder(nn.Module): + def __init__(self, out_dim, audio_in=1): + super().__init__() + self.out_dim = out_dim + self.feat_extractor = nn.Sequential( + BasicBlock(audio_in, out_dim//4, 15, 5, first_dilation=1600, downsample=True), + BasicBlock(out_dim//4, out_dim//4, 15, 6, first_dilation=0, downsample=True), + BasicBlock(out_dim//4, out_dim//4, 15, 1, first_dilation=7, ), + BasicBlock(out_dim//4, out_dim//2, 15, 6, first_dilation=0, downsample=True), + BasicBlock(out_dim//2, out_dim//2, 15, 1, first_dilation=7), + BasicBlock(out_dim//2, out_dim, 15, 3, first_dilation=0,downsample=True), + ) + def forward(self, wav_data): + if wav_data.dim() == 2: + wav_data = wav_data.unsqueeze(1) + else: + wav_data = wav_data.transpose(1, 2) + out = self.feat_extractor(wav_data) + return out.transpose(1, 2) + + +class MLP(nn.Module): + def __init__(self, in_dim, hidden_size, out_dim): + super().__init__() + self.mlp = nn.Sequential( + nn.Linear(in_dim, hidden_size), + nn.LeakyReLU(0.2, True), + nn.Linear(hidden_size, out_dim) + ) + def forward(self, inputs): + out = self.mlp(inputs) + return out + + +class PeriodicPositionalEncoding(nn.Module): + def __init__(self, d_model, dropout=0.1, period=15, max_seq_len=60): # 12s + super(PeriodicPositionalEncoding, self).__init__() + self.dropout = nn.Dropout(p=dropout) + pe = torch.zeros(period, d_model) + position = torch.arange(0, period, dtype=torch.float).unsqueeze(1) + div_term = torch.exp(torch.arange(0, d_model, 2).float() * (-math.log(10000.0) / d_model)) + pe[:, 0::2] = torch.sin(position * div_term) + pe[:, 1::2] = torch.cos(position * div_term) + pe = pe.unsqueeze(0) # (1, period, d_model) + repeat_num = (max_seq_len//period) + 1 + pe = pe.repeat(1, repeat_num, 1) # (1, repeat_num, period, d_model) + self.register_buffer('pe', pe) + def forward(self, x): + # print(self.pe.shape, x.shape) + x = x + self.pe[:, :x.size(1), :] + return self.dropout(x) + + +class MAGE_Transformer(nn.Module): + def __init__(self, args): + super(MAGE_Transformer, self).__init__() + self.args = args + with open(f"{args.data_path}weights/vocab.pkl", 'rb') as f: + self.lang_model = pickle.load(f) + pre_trained_embedding = self.lang_model.word_embedding_weights + self.text_pre_encoder_face = nn.Embedding.from_pretrained(torch.FloatTensor(pre_trained_embedding),freeze=args.t_fix_pre) + self.text_encoder_face = nn.Linear(300, args.audio_f) + self.text_encoder_face = nn.Linear(300, args.audio_f) + self.text_pre_encoder_body = nn.Embedding.from_pretrained(torch.FloatTensor(pre_trained_embedding),freeze=args.t_fix_pre) + self.text_encoder_body = nn.Linear(300, args.audio_f) + self.text_encoder_body = nn.Linear(300, args.audio_f) + + self.audio_pre_encoder_face = WavEncoder(args.audio_f, audio_in=2) + self.audio_pre_encoder_body = WavEncoder(args.audio_f, audio_in=2) + + self.at_attn_face = nn.Linear(args.audio_f*2, args.audio_f*2) + self.at_attn_body = nn.Linear(args.audio_f*2, args.audio_f*2) + + args_top = copy.deepcopy(self.args) + args_top.vae_layer = 3 + args_top.vae_length = args.motion_f + args_top.vae_test_dim = args.pose_dims+3+4 + self.motion_encoder = VQEncoderV6(args_top) # masked motion to latent bs t 333 to bs t 256 + + # face decoder + self.feature2face = nn.Linear(args.audio_f*2, args.hidden_size) + self.face2latent = nn.Linear(args.hidden_size, args.vae_codebook_size) + self.transformer_de_layer = nn.TransformerDecoderLayer( + d_model=self.args.hidden_size, + nhead=4, + dim_feedforward=self.args.hidden_size*2, + batch_first=True + ) + self.face_decoder = nn.TransformerDecoder(self.transformer_de_layer, num_layers=4) + self.position_embeddings = PeriodicPositionalEncoding(self.args.hidden_size, period=self.args.pose_length, max_seq_len=self.args.pose_length) + + # motion decoder + self.transformer_en_layer = nn.TransformerEncoderLayer( + d_model=self.args.hidden_size, + nhead=4, + dim_feedforward=self.args.hidden_size*2, + batch_first=True + ) + self.motion_self_encoder = nn.TransformerEncoder(self.transformer_en_layer, num_layers=1) + self.audio_feature2motion = nn.Linear(args.audio_f, args.hidden_size) + self.feature2motion = nn.Linear(args.motion_f, args.hidden_size) + + self.bodyhints_face = MLP(args.motion_f, args.hidden_size, args.motion_f) + self.bodyhints_body = MLP(args.motion_f, args.hidden_size, args.motion_f) + self.motion2latent_upper = MLP(args.hidden_size, args.hidden_size, self.args.hidden_size) + self.motion2latent_hands = MLP(args.hidden_size, args.hidden_size, self.args.hidden_size) + self.motion2latent_lower = MLP(args.hidden_size, args.hidden_size, self.args.hidden_size) + self.wordhints_decoder = nn.TransformerDecoder(self.transformer_de_layer, num_layers=8) + + self.upper_decoder = nn.TransformerDecoder(self.transformer_de_layer, num_layers=1) + self.hands_decoder = nn.TransformerDecoder(self.transformer_de_layer, num_layers=1) + self.lower_decoder = nn.TransformerDecoder(self.transformer_de_layer, num_layers=1) + + self.face_classifier = MLP(self.args.vae_codebook_size, args.hidden_size, self.args.vae_codebook_size) + self.upper_classifier = MLP(self.args.vae_codebook_size, args.hidden_size, self.args.vae_codebook_size) + self.hands_classifier = MLP(self.args.vae_codebook_size, args.hidden_size, self.args.vae_codebook_size) + self.lower_classifier = MLP(self.args.vae_codebook_size, args.hidden_size, self.args.vae_codebook_size) + + self.mask_embeddings = nn.Parameter(torch.zeros(1, 1, self.args.pose_dims+3+4)) + self.motion_down_upper = nn.Linear(args.hidden_size, self.args.vae_codebook_size) + self.motion_down_hands = nn.Linear(args.hidden_size, self.args.vae_codebook_size) + self.motion_down_lower = nn.Linear(args.hidden_size, self.args.vae_codebook_size) + self.motion_down_upper = nn.Linear(args.hidden_size, self.args.vae_codebook_size) + self.motion_down_hands = nn.Linear(args.hidden_size, self.args.vae_codebook_size) + self.motion_down_lower = nn.Linear(args.hidden_size, self.args.vae_codebook_size) + self._reset_parameters() + + self.spearker_encoder_body = nn.Embedding(25, args.hidden_size) + self.spearker_encoder_face = nn.Embedding(25, args.hidden_size) + + def _reset_parameters(self): + nn.init.normal_(self.mask_embeddings, 0, self.args.hidden_size ** -0.5) + + def forward(self, in_audio=None, in_word=None, mask=None, is_test=None, in_motion=None, use_attentions=True, use_word=True, in_id = None): + in_word_face = self.text_pre_encoder_face(in_word) + in_word_face = self.text_encoder_face(in_word_face) + in_word_body = self.text_pre_encoder_body(in_word) + in_word_body = self.text_encoder_body(in_word_body) + bs, t, c = in_word_face.shape + in_audio_face = self.audio_pre_encoder_face(in_audio) + in_audio_body = self.audio_pre_encoder_body(in_audio) + if in_audio_face.shape[1] != in_motion.shape[1]: + diff_length = in_motion.shape[1]- in_audio_face.shape[1] + if diff_length < 0: + in_audio_face = in_audio_face[:, :diff_length, :] + in_audio_body = in_audio_body[:, :diff_length, :] + else: + in_audio_face = torch.cat((in_audio_face, in_audio_face[:,-diff_length:]),1) + in_audio_body = torch.cat((in_audio_body, in_audio_body[:,-diff_length:]),1) + + if use_attentions: + alpha_at_face = torch.cat([in_word_face, in_audio_face], dim=-1).reshape(bs, t, c*2) + alpha_at_face = self.at_attn_face(alpha_at_face).reshape(bs, t, c, 2) + alpha_at_face = alpha_at_face.softmax(dim=-1) + fusion_face = in_word_face * alpha_at_face[:,:,:,1] + in_audio_face * alpha_at_face[:,:,:,0] + alpha_at_body = torch.cat([in_word_body, in_audio_body], dim=-1).reshape(bs, t, c*2) + alpha_at_body = self.at_attn_body(alpha_at_body).reshape(bs, t, c, 2) + alpha_at_body = alpha_at_body.softmax(dim=-1) + fusion_body = in_word_body * alpha_at_body[:,:,:,1] + in_audio_body * alpha_at_body[:,:,:,0] + else: + fusion_face = in_word_face + in_audio_face + fusion_body = in_word_body + in_audio_body + + masked_embeddings = self.mask_embeddings.expand_as(in_motion) + masked_motion = torch.where(mask == 1, masked_embeddings, in_motion) # bs, t, 256 + body_hint = self.motion_encoder(masked_motion) # bs t 256 + speaker_embedding_face = self.spearker_encoder_face(in_id).squeeze(2) + speaker_embedding_body = self.spearker_encoder_body(in_id).squeeze(2) + + # decode face + use_body_hints = True + if use_body_hints: + body_hint_face = self.bodyhints_face(body_hint) + fusion_face = torch.cat([fusion_face, body_hint_face], dim=2) + a2g_face = self.feature2face(fusion_face) + face_embeddings = speaker_embedding_face + face_embeddings = self.position_embeddings(face_embeddings) + decoded_face = self.face_decoder(tgt=face_embeddings, memory=a2g_face) + face_latent = self.face2latent(decoded_face) + cls_face = self.face_classifier(face_latent) + + # motion spatial encoder + body_hint_body = self.bodyhints_body(body_hint) + motion_embeddings = self.feature2motion(body_hint_body) + motion_embeddings = speaker_embedding_body + motion_embeddings + motion_embeddings = self.position_embeddings(motion_embeddings) + + # bi-directional self-attention + motion_refined_embeddings = self.motion_self_encoder(motion_embeddings) + + # audio to gesture cross-modal attention + if use_word : + a2g_motion = self.audio_feature2motion(fusion_body) + motion_refined_embeddings_in = motion_refined_embeddings + speaker_embedding_body + motion_refined_embeddings_in = self.position_embeddings(motion_refined_embeddings) + word_hints = self.wordhints_decoder(tgt=motion_refined_embeddings_in, memory=a2g_motion) + motion_refined_embeddings = motion_refined_embeddings + word_hints + + # feedforward + upper_latent = self.motion2latent_upper(motion_refined_embeddings) + hands_latent = self.motion2latent_hands(motion_refined_embeddings) + lower_latent = self.motion2latent_lower(motion_refined_embeddings) + + upper_latent_in = upper_latent + speaker_embedding_body + upper_latent_in = self.position_embeddings(upper_latent_in) + hands_latent_in = hands_latent + speaker_embedding_body + hands_latent_in = self.position_embeddings(hands_latent_in) + lower_latent_in = lower_latent + speaker_embedding_body + lower_latent_in = self.position_embeddings(lower_latent_in) + + # transformer decoder + motion_upper = self.upper_decoder(tgt=upper_latent_in, memory=hands_latent+lower_latent) + motion_hands = self.hands_decoder(tgt=hands_latent_in, memory=upper_latent+lower_latent) + motion_lower = self.lower_decoder(tgt=lower_latent_in, memory=upper_latent+hands_latent) + upper_latent = self.motion_down_upper(motion_upper+upper_latent) + hands_latent = self.motion_down_hands(motion_hands+hands_latent) + lower_latent = self.motion_down_lower(motion_lower+lower_latent) + cls_lower = self.lower_classifier(lower_latent) + cls_upper = self.upper_classifier(upper_latent) + cls_hands = self.hands_classifier(hands_latent) + + return { + "rec_face":face_latent, + "rec_upper":upper_latent, + "rec_lower":lower_latent, + "rec_hands":hands_latent, + "cls_face":cls_face, + "cls_upper":cls_upper, + "cls_lower":cls_lower, + "cls_hands":cls_hands, + } \ No newline at end of file diff --git a/models/motion_encoder.py b/models/motion_encoder.py new file mode 100644 index 0000000000000000000000000000000000000000..afa8513ea66ea0446230d796aff277ed142b8801 --- /dev/null +++ b/models/motion_encoder.py @@ -0,0 +1,789 @@ +import random +import math +import numpy as np +import torch +import torch.nn as nn +import torch.nn.functional as F +import smplx + +# ----------- 1 full conv-based encoder------------- # +""" +from tm2t +TM2T: Stochastical and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and Texts +https://github.com/EricGuo5513/TM2T +""" +from .quantizer import * +from .utils.layer import ResBlock, init_weight + +class SCFormer(nn.Module): + def __init__(self, args): + super(VQEncoderV3, self).__init__() + + + n_down = args.vae_layer + channels = [args.vae_length] + for i in range(n_down-1): + channels.append(args.vae_length) + + input_size = args.vae_test_dim + assert len(channels) == n_down + layers = [ + nn.Conv1d(input_size, channels[0], 4, 2, 1), + nn.LeakyReLU(0.2, inplace=True), + ResBlock(channels[0]), + ] + + for i in range(1, n_down): + layers += [ + nn.Conv1d(channels[i-1], channels[i], 4, 2, 1), + nn.LeakyReLU(0.2, inplace=True), + ResBlock(channels[i]), + ] + self.main = nn.Sequential(*layers) + # self.out_net = nn.Linear(output_size, output_size) + self.main.apply(init_weight) + # self.out_net.apply(init_weight) + def forward(self, inputs): # bs t n + ''' + face 51 or 106 + hand 30*(15) + upper body + lower body + global 1*3 + max length around 180 --> 450 + ''' + bs, t, n = inputs.shape + inputs = inputs.reshape(bs*t, n) + inputs = self.spatial_transformer_encoder(inputs) # bs*t c + cs = inputs.shape[1] + inputs = inputs.reshape(bs, t, cs).permute(0, 2, 1).reshape(bs*cs, t) + inputs = self.temporal_cnn_encoder(inputs) # bs*c t + ct = inputs.shape[1] + outputs = inputs.reshape(bs, cs, ct).permute(0, 2, 1) # bs ct cs + return outputs + +class VQEncoderV3(nn.Module): + def __init__(self, args): + super(VQEncoderV3, self).__init__() + n_down = args.vae_layer + channels = [args.vae_length] + for i in range(n_down-1): + channels.append(args.vae_length) + + input_size = args.vae_test_dim + assert len(channels) == n_down + layers = [ + nn.Conv1d(input_size, channels[0], 4, 2, 1), + nn.LeakyReLU(0.2, inplace=True), + ResBlock(channels[0]), + ] + + for i in range(1, n_down): + layers += [ + nn.Conv1d(channels[i-1], channels[i], 4, 2, 1), + nn.LeakyReLU(0.2, inplace=True), + ResBlock(channels[i]), + ] + self.main = nn.Sequential(*layers) + # self.out_net = nn.Linear(output_size, output_size) + self.main.apply(init_weight) + # self.out_net.apply(init_weight) + def forward(self, inputs): + inputs = inputs.permute(0, 2, 1) + outputs = self.main(inputs).permute(0, 2, 1) + return outputs + +class VQEncoderV6(nn.Module): + def __init__(self, args): + super(VQEncoderV6, self).__init__() + n_down = args.vae_layer + channels = [args.vae_length] + for i in range(n_down-1): + channels.append(args.vae_length) + + input_size = args.vae_test_dim + assert len(channels) == n_down + layers = [ + nn.Conv1d(input_size, channels[0], 3, 1, 1), + nn.LeakyReLU(0.2, inplace=True), + ResBlock(channels[0]), + ] + + for i in range(1, n_down): + layers += [ + nn.Conv1d(channels[i-1], channels[i], 3, 1, 1), + nn.LeakyReLU(0.2, inplace=True), + ResBlock(channels[i]), + ] + self.main = nn.Sequential(*layers) + # self.out_net = nn.Linear(output_size, output_size) + self.main.apply(init_weight) + # self.out_net.apply(init_weight) + def forward(self, inputs): + inputs = inputs.permute(0, 2, 1) + outputs = self.main(inputs).permute(0, 2, 1) + return outputs + +class VQEncoderV4(nn.Module): + def __init__(self, args): + super(VQEncoderV4, self).__init__() + n_down = args.vae_layer + channels = [args.vae_length] + for i in range(n_down-1): + channels.append(args.vae_length) + + input_size = args.vae_test_dim + assert len(channels) == n_down + layers = [ + nn.Conv1d(input_size, channels[0], 4, 2, 1), + nn.LeakyReLU(0.2, inplace=True), + ResBlock(channels[0]), + ] + + for i in range(1, n_down): + layers += [ + nn.Conv1d(channels[i-1], channels[i], 3, 1, 1), + nn.LeakyReLU(0.2, inplace=True), + ResBlock(channels[i]), + ] + self.main = nn.Sequential(*layers) + # self.out_net = nn.Linear(output_size, output_size) + self.main.apply(init_weight) + # self.out_net.apply(init_weight) + def forward(self, inputs): + inputs = inputs.permute(0, 2, 1) + outputs = self.main(inputs).permute(0, 2, 1) + # print(outputs.shape) + return outputs + +class VQEncoderV5(nn.Module): + def __init__(self, args): + super(VQEncoderV5, self).__init__() + n_down = args.vae_layer + channels = [args.vae_length] + for i in range(n_down-1): + channels.append(args.vae_length) + + input_size = args.vae_test_dim + assert len(channels) == n_down + layers = [ + nn.Conv1d(input_size, channels[0], 3, 1, 1), + nn.LeakyReLU(0.2, inplace=True), + ResBlock(channels[0]), + ] + + for i in range(1, n_down): + layers += [ + nn.Conv1d(channels[i-1], channels[i], 3, 1, 1), + nn.LeakyReLU(0.2, inplace=True), + ResBlock(channels[i]), + ] + self.main = nn.Sequential(*layers) + # self.out_net = nn.Linear(output_size, output_size) + self.main.apply(init_weight) + # self.out_net.apply(init_weight) + def forward(self, inputs): + inputs = inputs.permute(0, 2, 1) + outputs = self.main(inputs).permute(0, 2, 1) + # print(outputs.shape) + return outputs + +class VQDecoderV4(nn.Module): + def __init__(self, args): + super(VQDecoderV4, self).__init__() + n_up = args.vae_layer + channels = [] + for i in range(n_up-1): + channels.append(args.vae_length) + channels.append(args.vae_length) + channels.append(args.vae_test_dim) + input_size = args.vae_length + n_resblk = 2 + assert len(channels) == n_up + 1 + if input_size == channels[0]: + layers = [] + else: + layers = [nn.Conv1d(input_size, channels[0], kernel_size=3, stride=1, padding=1)] + + for i in range(n_resblk): + layers += [ResBlock(channels[0])] + # channels = channels + for i in range(n_up): + up_factor = 2 if i < n_up - 1 else 1 + layers += [ + nn.Upsample(scale_factor=up_factor, mode='nearest'), + nn.Conv1d(channels[i], channels[i+1], kernel_size=3, stride=1, padding=1), + nn.LeakyReLU(0.2, inplace=True) + ] + layers += [nn.Conv1d(channels[-1], channels[-1], kernel_size=3, stride=1, padding=1)] + self.main = nn.Sequential(*layers) + self.main.apply(init_weight) + + def forward(self, inputs): + inputs = inputs.permute(0, 2, 1) + outputs = self.main(inputs).permute(0, 2, 1) + return outputs + +class VQDecoderV5(nn.Module): + def __init__(self, args): + super(VQDecoderV5, self).__init__() + n_up = args.vae_layer + channels = [] + for i in range(n_up-1): + channels.append(args.vae_length) + channels.append(args.vae_length) + channels.append(args.vae_test_dim) + input_size = args.vae_length + n_resblk = 2 + assert len(channels) == n_up + 1 + if input_size == channels[0]: + layers = [] + else: + layers = [nn.Conv1d(input_size, channels[0], kernel_size=3, stride=1, padding=1)] + + for i in range(n_resblk): + layers += [ResBlock(channels[0])] + # channels = channels + for i in range(n_up): + up_factor = 2 if i < n_up - 1 else 1 + layers += [ + #nn.Upsample(scale_factor=up_factor, mode='nearest'), + nn.Conv1d(channels[i], channels[i+1], kernel_size=3, stride=1, padding=1), + nn.LeakyReLU(0.2, inplace=True) + ] + layers += [nn.Conv1d(channels[-1], channels[-1], kernel_size=3, stride=1, padding=1)] + self.main = nn.Sequential(*layers) + self.main.apply(init_weight) + + def forward(self, inputs): + inputs = inputs.permute(0, 2, 1) + outputs = self.main(inputs).permute(0, 2, 1) + return outputs + +class VQDecoderV7(nn.Module): + def __init__(self, args): + super(VQDecoderV7, self).__init__() + n_up = args.vae_layer + channels = [] + for i in range(n_up-1): + channels.append(args.vae_length) + channels.append(args.vae_length) + channels.append(args.vae_test_dim+4) + input_size = args.vae_length + n_resblk = 2 + assert len(channels) == n_up + 1 + if input_size == channels[0]: + layers = [] + else: + layers = [nn.Conv1d(input_size, channels[0], kernel_size=3, stride=1, padding=1)] + + for i in range(n_resblk): + layers += [ResBlock(channels[0])] + # channels = channels + for i in range(n_up): + up_factor = 2 if i < n_up - 1 else 1 + layers += [ + #nn.Upsample(scale_factor=up_factor, mode='nearest'), + nn.Conv1d(channels[i], channels[i+1], kernel_size=3, stride=1, padding=1), + nn.LeakyReLU(0.2, inplace=True) + ] + layers += [nn.Conv1d(channels[-1], channels[-1], kernel_size=3, stride=1, padding=1)] + self.main = nn.Sequential(*layers) + self.main.apply(init_weight) + + def forward(self, inputs): + inputs = inputs.permute(0, 2, 1) + outputs = self.main(inputs).permute(0, 2, 1) + return outputs + +class VQDecoderV3(nn.Module): + def __init__(self, args): + super(VQDecoderV3, self).__init__() + n_up = args.vae_layer + channels = [] + for i in range(n_up-1): + channels.append(args.vae_length) + channels.append(args.vae_length) + channels.append(args.vae_test_dim) + input_size = args.vae_length + n_resblk = 2 + assert len(channels) == n_up + 1 + if input_size == channels[0]: + layers = [] + else: + layers = [nn.Conv1d(input_size, channels[0], kernel_size=3, stride=1, padding=1)] + + for i in range(n_resblk): + layers += [ResBlock(channels[0])] + # channels = channels + for i in range(n_up): + layers += [ + nn.Upsample(scale_factor=2, mode='nearest'), + nn.Conv1d(channels[i], channels[i+1], kernel_size=3, stride=1, padding=1), + nn.LeakyReLU(0.2, inplace=True) + ] + layers += [nn.Conv1d(channels[-1], channels[-1], kernel_size=3, stride=1, padding=1)] + self.main = nn.Sequential(*layers) + self.main.apply(init_weight) + + def forward(self, inputs): + inputs = inputs.permute(0, 2, 1) + outputs = self.main(inputs).permute(0, 2, 1) + return outputs + +class VQDecoderV6(nn.Module): + def __init__(self, args): + super(VQDecoderV6, self).__init__() + n_up = args.vae_layer + channels = [] + for i in range(n_up-1): + channels.append(args.vae_length) + channels.append(args.vae_length) + channels.append(args.vae_test_dim) + input_size = args.vae_length * 2 + n_resblk = 2 + assert len(channels) == n_up + 1 + if input_size == channels[0]: + layers = [] + else: + layers = [nn.Conv1d(input_size, channels[0], kernel_size=3, stride=1, padding=1)] + + for i in range(n_resblk): + layers += [ResBlock(channels[0])] + # channels = channels + for i in range(n_up): + layers += [ + # nn.Upsample(scale_factor=2, mode='nearest'), + nn.Conv1d(channels[i], channels[i+1], kernel_size=3, stride=1, padding=1), + nn.LeakyReLU(0.2, inplace=True) + ] + layers += [nn.Conv1d(channels[-1], channels[-1], kernel_size=3, stride=1, padding=1)] + self.main = nn.Sequential(*layers) + self.main.apply(init_weight) + + def forward(self, inputs): + inputs = inputs.permute(0, 2, 1) + outputs = self.main(inputs).permute(0, 2, 1) + return outputs + + +# -----------2 conv+mlp based fix-length input ae ------------- # +from .utils.layer import reparameterize, ConvNormRelu, BasicBlock +""" +from Trimodal, +encoder: + bs, n, c_in --conv--> bs, n/k, c_out_0 --mlp--> bs, c_out_1, only support fixed length +decoder: + bs, c_out_1 --mlp--> bs, n/k*c_out_0 --> bs, n/k, c_out_0 --deconv--> bs, n, c_in +""" +class PoseEncoderConv(nn.Module): + def __init__(self, length, dim, feature_length=32): + super().__init__() + self.base = feature_length + self.net = nn.Sequential( + ConvNormRelu(dim, self.base, batchnorm=True), #32 + ConvNormRelu(self.base, self.base*2, batchnorm=True), #30 + ConvNormRelu(self.base*2, self.base*2, True, batchnorm=True), #14 + nn.Conv1d(self.base*2, self.base, 3) + ) + self.out_net = nn.Sequential( + nn.Linear(12*self.base, self.base*4), # for 34 frames + nn.BatchNorm1d(self.base*4), + nn.LeakyReLU(True), + nn.Linear(self.base*4, self.base*2), + nn.BatchNorm1d(self.base*2), + nn.LeakyReLU(True), + nn.Linear(self.base*2, self.base), + ) + self.fc_mu = nn.Linear(self.base, self.base) + self.fc_logvar = nn.Linear(self.base, self.base) + + def forward(self, poses, variational_encoding=None): + poses = poses.transpose(1, 2) # to (bs, dim, seq) + out = self.net(poses) + out = out.flatten(1) + out = self.out_net(out) + mu = self.fc_mu(out) + logvar = self.fc_logvar(out) + if variational_encoding: + z = reparameterize(mu, logvar) + else: + z = mu + return z, mu, logvar + + +class PoseDecoderFC(nn.Module): + def __init__(self, gen_length, pose_dim, use_pre_poses=False): + super().__init__() + self.gen_length = gen_length + self.pose_dim = pose_dim + self.use_pre_poses = use_pre_poses + + in_size = 32 + if use_pre_poses: + self.pre_pose_net = nn.Sequential( + nn.Linear(pose_dim * 4, 32), + nn.BatchNorm1d(32), + nn.ReLU(), + nn.Linear(32, 32), + ) + in_size += 32 + + self.net = nn.Sequential( + nn.Linear(in_size, 128), + nn.BatchNorm1d(128), + nn.ReLU(), + nn.Linear(128, 128), + nn.BatchNorm1d(128), + nn.ReLU(), + nn.Linear(128, 256), + nn.BatchNorm1d(256), + nn.ReLU(), + nn.Linear(256, 512), + nn.BatchNorm1d(512), + nn.ReLU(), + nn.Linear(512, gen_length * pose_dim), + ) + + def forward(self, latent_code, pre_poses=None): + if self.use_pre_poses: + pre_pose_feat = self.pre_pose_net(pre_poses.reshape(pre_poses.shape[0], -1)) + feat = torch.cat((pre_pose_feat, latent_code), dim=1) + else: + feat = latent_code + output = self.net(feat) + output = output.view(-1, self.gen_length, self.pose_dim) + return output + + +class PoseDecoderConv(nn.Module): + def __init__(self, length, dim, use_pre_poses=False, feature_length=32): + super().__init__() + self.use_pre_poses = use_pre_poses + self.feat_size = feature_length + + if use_pre_poses: + self.pre_pose_net = nn.Sequential( + nn.Linear(dim * 4, 32), + nn.BatchNorm1d(32), + nn.ReLU(), + nn.Linear(32, 32), + ) + self.feat_size += 32 + + if length == 64: + self.pre_net = nn.Sequential( + nn.Linear(self.feat_size, self.feat_size), + nn.BatchNorm1d(self.feat_size), + nn.LeakyReLU(True), + nn.Linear(self.feat_size, self.feat_size//8*64), + ) + elif length == 34: + self.pre_net = nn.Sequential( + nn.Linear(self.feat_size, self.feat_size*2), + nn.BatchNorm1d(self.feat_size*2), + nn.LeakyReLU(True), + nn.Linear(self.feat_size*2, self.feat_size//8*34), + ) + elif length == 32: + self.pre_net = nn.Sequential( + nn.Linear(self.feat_size, self.feat_size*2), + nn.BatchNorm1d(self.feat_size*2), + nn.LeakyReLU(True), + nn.Linear(self.feat_size*2, self.feat_size//8*32), + ) + else: + assert False + self.decoder_size = self.feat_size//8 + self.net = nn.Sequential( + nn.ConvTranspose1d(self.decoder_size, self.feat_size, 3), + nn.BatchNorm1d(self.feat_size), + nn.LeakyReLU(0.2, True), + + nn.ConvTranspose1d(self.feat_size, self.feat_size, 3), + nn.BatchNorm1d(self.feat_size), + nn.LeakyReLU(0.2, True), + nn.Conv1d(self.feat_size, self.feat_size*2, 3), + nn.Conv1d(self.feat_size*2, dim, 3), + ) + + def forward(self, feat, pre_poses=None): + if self.use_pre_poses: + pre_pose_feat = self.pre_pose_net(pre_poses.reshape(pre_poses.shape[0], -1)) + feat = torch.cat((pre_pose_feat, feat), dim=1) + #print(feat.shape) + out = self.pre_net(feat) + #print(out.shape) + out = out.view(feat.shape[0], self.decoder_size, -1) + #print(out.shape) + out = self.net(out) + out = out.transpose(1, 2) + return out + +''' +Our CaMN Modification +''' +class PoseEncoderConvResNet(nn.Module): + def __init__(self, length, dim, feature_length=32): + super().__init__() + self.base = feature_length + self.conv1=BasicBlock(dim, self.base, reduce_first = 1, downsample = False, first_dilation=1) #34 + self.conv2=BasicBlock(self.base, self.base*2, downsample = False, first_dilation=1,) #34 + self.conv3=BasicBlock(self.base*2, self.base*2, first_dilation=1, downsample = True, stride=2)#17 + self.conv4=BasicBlock(self.base*2, self.base, first_dilation=1, downsample = False) + + self.out_net = nn.Sequential( + # nn.Linear(864, 256), # for 64 frames + nn.Linear(17*self.base, self.base*4), # for 34 frames + nn.BatchNorm1d(self.base*4), + nn.LeakyReLU(True), + nn.Linear(self.base*4, self.base*2), + nn.BatchNorm1d(self.base*2), + nn.LeakyReLU(True), + nn.Linear(self.base*2, self.base), + ) + + self.fc_mu = nn.Linear(self.base, self.base) + self.fc_logvar = nn.Linear(self.base, self.base) + + def forward(self, poses, variational_encoding=None): + poses = poses.transpose(1, 2) # to (bs, dim, seq) + out1 = self.conv1(poses) + out2 = self.conv2(out1) + out3 = self.conv3(out2) + out = self.conv4(out3) + out = out.flatten(1) + out = self.out_net(out) + mu = self.fc_mu(out) + logvar = self.fc_logvar(out) + if variational_encoding: + z = reparameterize(mu, logvar) + else: + z = mu + return z, mu, logvar + + +# -----------3 lstm ------------- # +''' +bs, n, c_int --> bs, n, c_out or bs, 1 (hidden), c_out +''' +class AELSTM(nn.Module): + def __init__(self, args): + super().__init__() + self.motion_emb = nn.Linear(args.vae_test_dim, args.vae_length) + self.lstm = nn.LSTM(args.vae_length, hidden_size=args.vae_length, num_layers=4, batch_first=True, + bidirectional=True, dropout=0.3) + self.out = nn.Sequential( + nn.Linear(args.vae_length, args.vae_length//2), + nn.LeakyReLU(0.2, True), + nn.Linear(args.vae_length//2, args.vae_test_dim) + ) + self.hidden_size = args.vae_length + + def forward(self, inputs): + poses = self.motion_emb(inputs) + out, _ = self.lstm(poses) + out = out[:, :, :self.hidden_size] + out[:, :, self.hidden_size:] + out_poses = self.out(out) + return { + "poses_feat":out, + "rec_pose": out_poses, + } + +class PoseDecoderLSTM(nn.Module): + """ + input bs*n*64 + """ + def __init__(self,pose_dim, feature_length): + super().__init__() + self.pose_dim = pose_dim + self.base = feature_length + self.hidden_size = 256 + self.lstm_d = nn.LSTM(self.base, hidden_size=self.hidden_size, num_layers=4, batch_first=True, + bidirectional=True, dropout=0.3) + self.out_d = nn.Sequential( + nn.Linear(self.hidden_size, self.hidden_size // 2), + nn.LeakyReLU(True), + nn.Linear(self.hidden_size // 2, self.pose_dim) + ) + + def forward(self, latent_code): + output, _ = self.lstm_d(latent_code) + output = output[:, :, :self.hidden_size] + output[:, :, self.hidden_size:] # sum bidirectional outputs + #print("outd:", output.shape) + output = self.out_d(output.reshape(-1, output.shape[2])) + output = output.view(latent_code.shape[0], latent_code.shape[1], -1) + #print("resotuput:", output.shape) + return output + +# ---------------4 transformer --------------- # +class PositionalEncoding(nn.Module): + def __init__(self, d_model, dropout=0.1, max_len=5000): + super(PositionalEncoding, self).__init__() + self.dropout = nn.Dropout(p=dropout) + pe = torch.zeros(max_len, d_model) + position = torch.arange(0, max_len, dtype=torch.float).unsqueeze(1) + div_term = torch.exp(torch.arange(0, d_model, 2).float() * (-np.log(10000.0) / d_model)) + pe[:, 0::2] = torch.sin(position * div_term) + pe[:, 1::2] = torch.cos(position * div_term) + pe = pe.unsqueeze(0)#.transpose(0, 1) + + self.register_buffer('pe', pe) + + def forward(self, x): + #print(self.pe.shape, x.shape) + x = x + self.pe[:, :x.shape[1]] + return self.dropout(x) + +class Encoder_TRANSFORMER(nn.Module): + def __init__(self, args): + super().__init__() + self.skelEmbedding = nn.Linear(args.vae_test_dim, args.vae_length) + self.sequence_pos_encoder = PositionalEncoding(args.vae_length, 0.3) + seqTransEncoderLayer = nn.TransformerEncoderLayer(d_model=args.vae_length, + nhead=4, + dim_feedforward=1025, + dropout=0.3, + activation="gelu", + batch_first=True + ) + self.seqTransEncoder = nn.TransformerEncoder(seqTransEncoderLayer, + num_layers=4) + def _generate_square_subsequent_mask(self, sz): + mask = (torch.triu(torch.ones(sz, sz)) == 1).transpose(0, 1) + mask = mask.float().masked_fill(mask == 0, float('-inf')).masked_fill(mask == 1, float(0.0)) + return mask + + def forward(self, inputs): + x = self.skelEmbedding(inputs) #bs * n * 128 + #print(x.shape) + xseq = self.sequence_pos_encoder(x) + device = xseq.device + #mask = self._generate_square_subsequent_mask(xseq.size(1)).to(device) + final = self.seqTransEncoder(xseq) + #print(final.shape) + mu = final[:, 0:1, :] + logvar = final[:, 1:2, :] + return final, mu, logvar + +class Decoder_TRANSFORMER(nn.Module): + def __init__(self, args): + super().__init__() + self.vae_test_len = args.vae_test_len + self.vae_length = args.vae_length + self.sequence_pos_encoder = PositionalEncoding(args.vae_length, 0.3) + seqTransDecoderLayer = nn.TransformerDecoderLayer(d_model=args.vae_length, + nhead=4, + dim_feedforward=1024, + dropout=0.3, + activation="gelu", + batch_first=True) + self.seqTransDecoder = nn.TransformerDecoder(seqTransDecoderLayer, + num_layers=4) + self.finallayer = nn.Linear(args.vae_length, args.vae_test_dim) + + def forward(self, inputs): + timequeries = torch.zeros(inputs.shape[0], self.vae_test_len, self.vae_length, device=inputs.device) + timequeries = self.sequence_pos_encoder(timequeries) + output = self.seqTransDecoder(tgt=timequeries, memory=inputs) + output = self.finallayer(output) + return output + +# --------- 5 skcnn --------------- # +''' +from NeMF, +NeMF: Neural Motion Fields for Kinematic Animation +''' +from .utils.skeleton import ResidualBlock, SkeletonResidual, residual_ratio, SkeletonConv, SkeletonPool, find_neighbor, build_edge_topology +class LocalEncoder(nn.Module): + def __init__(self, args, topology): + super(LocalEncoder, self).__init__() + args.channel_base = 6 + args.activation = "tanh" + args.use_residual_blocks=True + args.z_dim=1024 + args.temporal_scale=8 + args.kernel_size=4 + args.num_layers=args.vae_layer + args.skeleton_dist=2 + args.extra_conv=0 + # check how to reflect in 1d + args.padding_mode="constant" + args.skeleton_pool="mean" + args.upsampling="linear" + + + self.topologies = [topology] + self.channel_base = [args.channel_base] + + self.channel_list = [] + self.edge_num = [len(topology)] + self.pooling_list = [] + self.layers = nn.ModuleList() + self.args = args + # self.convs = [] + + kernel_size = args.kernel_size + kernel_even = False if kernel_size % 2 else True + padding = (kernel_size - 1) // 2 + bias = True + self.grow = args.vae_grow + for i in range(args.num_layers): + self.channel_base.append(self.channel_base[-1]*self.grow[i]) + + for i in range(args.num_layers): + seq = [] + neighbour_list = find_neighbor(self.topologies[i], args.skeleton_dist) + in_channels = self.channel_base[i] * self.edge_num[i] + out_channels = self.channel_base[i + 1] * self.edge_num[i] + if i == 0: + self.channel_list.append(in_channels) + self.channel_list.append(out_channels) + last_pool = True if i == args.num_layers - 1 else False + + # (T, J, D) => (T, J', D) + pool = SkeletonPool(edges=self.topologies[i], pooling_mode=args.skeleton_pool, + channels_per_edge=out_channels // len(neighbour_list), last_pool=last_pool) + + if args.use_residual_blocks: + # (T, J, D) => (T/2, J', 2D) + seq.append(SkeletonResidual(self.topologies[i], neighbour_list, joint_num=self.edge_num[i], in_channels=in_channels, out_channels=out_channels, + kernel_size=kernel_size, stride=2, padding=padding, padding_mode=args.padding_mode, bias=bias, + extra_conv=args.extra_conv, pooling_mode=args.skeleton_pool, activation=args.activation, last_pool=last_pool)) + else: + for _ in range(args.extra_conv): + # (T, J, D) => (T, J, D) + seq.append(SkeletonConv(neighbour_list, in_channels=in_channels, out_channels=in_channels, + joint_num=self.edge_num[i], kernel_size=kernel_size - 1 if kernel_even else kernel_size, + stride=1, + padding=padding, padding_mode=args.padding_mode, bias=bias)) + seq.append(nn.PReLU() if args.activation == 'relu' else nn.Tanh()) + # (T, J, D) => (T/2, J, 2D) + seq.append(SkeletonConv(neighbour_list, in_channels=in_channels, out_channels=out_channels, + joint_num=self.edge_num[i], kernel_size=kernel_size, stride=2, + padding=padding, padding_mode=args.padding_mode, bias=bias, add_offset=False, + in_offset_channel=3 * self.channel_base[i] // self.channel_base[0])) + # self.convs.append(seq[-1]) + + seq.append(pool) + seq.append(nn.PReLU() if args.activation == 'relu' else nn.Tanh()) + self.layers.append(nn.Sequential(*seq)) + + self.topologies.append(pool.new_edges) + self.pooling_list.append(pool.pooling_list) + self.edge_num.append(len(self.topologies[-1])) + + # in_features = self.channel_base[-1] * len(self.pooling_list[-1]) + # in_features *= int(args.temporal_scale / 2) + # self.reduce = nn.Linear(in_features, args.z_dim) + # self.mu = nn.Linear(in_features, args.z_dim) + # self.logvar = nn.Linear(in_features, args.z_dim) + + def forward(self, input): + #bs, n, c = input.shape[0], input.shape[1], input.shape[2] + output = input.permute(0, 2, 1)#input.reshape(bs, n, -1, 6) + for layer in self.layers: + output = layer(output) + #output = output.view(output.shape[0], -1) + output = output.permute(0, 2, 1) + return output \ No newline at end of file diff --git a/models/motion_representation.py b/models/motion_representation.py new file mode 100644 index 0000000000000000000000000000000000000000..b5d93b49931d45ae0b8bf5013c76b08445e13eff --- /dev/null +++ b/models/motion_representation.py @@ -0,0 +1,431 @@ +import random +import math +import numpy as np +import torch +import torch.nn as nn +import torch.nn.functional as F +import smplx +import copy +from .motion_encoder import * + +# ----------- AE, VAE ------------- # +class VAEConvZero(nn.Module): + def __init__(self, args): + super(VAEConvZero, self).__init__() + self.encoder = VQEncoderV5(args) + # self.quantizer = Quantizer(args.vae_codebook_size, args.vae_length, args.vae_quantizer_lambda) + self.decoder = VQDecoderV5(args) + + def forward(self, inputs): + pre_latent = self.encoder(inputs) + # print(pre_latent.shape) + # embedding_loss, vq_latent, _, perplexity = self.quantizer(pre_latent) + rec_pose = self.decoder(pre_latent) + return { + # "poses_feat":vq_latent, + # "embedding_loss":embedding_loss, + # "perplexity":perplexity, + "rec_pose": rec_pose + } + +class VAEConv(nn.Module): + def __init__(self, args): + super(VAEConv, self).__init__() + self.encoder = VQEncoderV3(args) + self.decoder = VQDecoderV3(args) + self.fc_mu = nn.Linear(args.vae_length, args.vae_length) + self.fc_logvar = nn.Linear(args.vae_length, args.vae_length) + self.variational = args.variational + + def forward(self, inputs): + pre_latent = self.encoder(inputs) + mu, logvar = None, None + if self.variational: + mu = self.fc_mu(pre_latent) + logvar = self.fc_logvar(pre_latent) + pre_latent = reparameterize(mu, logvar) + rec_pose = self.decoder(pre_latent) + return { + "poses_feat":pre_latent, + "rec_pose": rec_pose, + "pose_mu": mu, + "pose_logvar": logvar, + } + + def map2latent(self, inputs): + pre_latent = self.encoder(inputs) + if self.variational: + mu = self.fc_mu(pre_latent) + logvar = self.fc_logvar(pre_latent) + pre_latent = reparameterize(mu, logvar) + return pre_latent + + def decode(self, pre_latent): + rec_pose = self.decoder(pre_latent) + return rec_pose + +class VAESKConv(VAEConv): + def __init__(self, args): + super(VAESKConv, self).__init__(args) + smpl_fname = args.data_path_1+'smplx_models/smplx/SMPLX_NEUTRAL_2020.npz' + smpl_data = np.load(smpl_fname, encoding='latin1') + parents = smpl_data['kintree_table'][0].astype(np.int32) + edges = build_edge_topology(parents) + self.encoder = LocalEncoder(args, edges) + self.decoder = VQDecoderV3(args) + +class VAEConvMLP(VAEConv): + def __init__(self, args): + super(VAEConvMLP, self).__init__(args) + self.encoder = PoseEncoderConv(args.vae_test_len, args.vae_test_dim, feature_length=args.vae_length) + self.decoder = PoseDecoderConv(args.vae_test_len, args.vae_test_dim, feature_length=args.vae_length) + +class VAELSTM(VAEConv): + def __init__(self, args): + super(VAELSTM, self).__init__(args) + pose_dim = args.vae_test_dim + feature_length = args.vae_length + self.encoder = PoseEncoderLSTM_Resnet(pose_dim, feature_length=feature_length) + self.decoder = PoseDecoderLSTM(pose_dim, feature_length=feature_length) + +class VAETransformer(VAEConv): + def __init__(self, args): + super(VAETransformer, self).__init__(args) + self.encoder = Encoder_TRANSFORMER(args) + self.decoder = Decoder_TRANSFORMER(args) + +# ----------- VQVAE --------------- # +class VQVAEConv(nn.Module): + def __init__(self, args): + super(VQVAEConv, self).__init__() + self.encoder = VQEncoderV3(args) + self.quantizer = Quantizer(args.vae_codebook_size, args.vae_length, args.vae_quantizer_lambda) + self.decoder = VQDecoderV3(args) + + def forward(self, inputs): + pre_latent = self.encoder(inputs) + # print(pre_latent.shape) + embedding_loss, vq_latent, _, perplexity = self.quantizer(pre_latent) + rec_pose = self.decoder(vq_latent) + return { + "poses_feat":vq_latent, + "embedding_loss":embedding_loss, + "perplexity":perplexity, + "rec_pose": rec_pose + } + + def map2index(self, inputs): + pre_latent = self.encoder(inputs) + index = self.quantizer.map2index(pre_latent) + return index + + def map2latent(self, inputs): + pre_latent = self.encoder(inputs) + index = self.quantizer.map2index(pre_latent) + z_q = self.quantizer.get_codebook_entry(index) + return z_q + + def decode(self, index): + z_q = self.quantizer.get_codebook_entry(index) + rec_pose = self.decoder(z_q) + return rec_pose + +class VQVAESKConv(VQVAEConv): + def __init__(self, args): + super(VQVAESKConv, self).__init__(args) + smpl_fname = args.data_path_1+'smplx_models/smplx/SMPLX_NEUTRAL_2020.npz' + smpl_data = np.load(smpl_fname, encoding='latin1') + parents = smpl_data['kintree_table'][0].astype(np.int32) + edges = build_edge_topology(parents) + self.encoder = LocalEncoder(args, edges) + + +class VQVAEConvStride(nn.Module): + def __init__(self, args): + super(VQVAEConvStride, self).__init__() + self.encoder = VQEncoderV4(args) + self.quantizer = Quantizer(args.vae_codebook_size, args.vae_length, args.vae_quantizer_lambda) + self.decoder = VQDecoderV4(args) + + def forward(self, inputs): + pre_latent = self.encoder(inputs) + # print(pre_latent.shape) + embedding_loss, vq_latent, _, perplexity = self.quantizer(pre_latent) + rec_pose = self.decoder(vq_latent) + return { + "poses_feat":vq_latent, + "embedding_loss":embedding_loss, + "perplexity":perplexity, + "rec_pose": rec_pose + } + + def map2index(self, inputs): + pre_latent = self.encoder(inputs) + index = self.quantizer.map2index(pre_latent) + return index + + def map2latent(self, inputs): + pre_latent = self.encoder(inputs) + index = self.quantizer.map2index(pre_latent) + z_q = self.quantizer.get_codebook_entry(index) + return z_q + + def decode(self, index): + z_q = self.quantizer.get_codebook_entry(index) + rec_pose = self.decoder(z_q) + return rec_pose + +class VQVAEConvZero(nn.Module): + def __init__(self, args): + super(VQVAEConvZero, self).__init__() + self.encoder = VQEncoderV5(args) + self.quantizer = Quantizer(args.vae_codebook_size, args.vae_length, args.vae_quantizer_lambda) + self.decoder = VQDecoderV5(args) + + def forward(self, inputs): + pre_latent = self.encoder(inputs) + # print(pre_latent.shape) + embedding_loss, vq_latent, _, perplexity = self.quantizer(pre_latent) + rec_pose = self.decoder(vq_latent) + return { + "poses_feat":vq_latent, + "embedding_loss":embedding_loss, + "perplexity":perplexity, + "rec_pose": rec_pose + } + + def map2index(self, inputs): + pre_latent = self.encoder(inputs) + index = self.quantizer.map2index(pre_latent) + return index + + def map2latent(self, inputs): + pre_latent = self.encoder(inputs) + index = self.quantizer.map2index(pre_latent) + z_q = self.quantizer.get_codebook_entry(index) + return z_q + + def decode(self, index): + z_q = self.quantizer.get_codebook_entry(index) + rec_pose = self.decoder(z_q) + return rec_pose + + +class VAEConvZero(nn.Module): + def __init__(self, args): + super(VAEConvZero, self).__init__() + self.encoder = VQEncoderV5(args) + # self.quantizer = Quantizer(args.vae_codebook_size, args.vae_length, args.vae_quantizer_lambda) + self.decoder = VQDecoderV5(args) + + def forward(self, inputs): + pre_latent = self.encoder(inputs) + # print(pre_latent.shape) + # embedding_loss, vq_latent, _, perplexity = self.quantizer(pre_latent) + rec_pose = self.decoder(pre_latent) + return { + # "poses_feat":vq_latent, + # "embedding_loss":embedding_loss, + # "perplexity":perplexity, + "rec_pose": rec_pose + } + + # def map2index(self, inputs): + # pre_latent = self.encoder(inputs) + # index = self.quantizer.map2index(pre_latent) + # return index + + # def map2latent(self, inputs): + # pre_latent = self.encoder(inputs) + # index = self.quantizer.map2index(pre_latent) + # z_q = self.quantizer.get_codebook_entry(index) + # return z_q + + # def decode(self, index): + # z_q = self.quantizer.get_codebook_entry(index) + # rec_pose = self.decoder(z_q) + # return rec_pose + + +class VQVAEConvZero3(nn.Module): + def __init__(self, args): + super(VQVAEConvZero3, self).__init__() + self.encoder = VQEncoderV5(args) + self.quantizer = Quantizer(args.vae_codebook_size, args.vae_length, args.vae_quantizer_lambda) + self.decoder = VQDecoderV5(args) + + def forward(self, inputs): + pre_latent = self.encoder(inputs) + # print(pre_latent.shape) + embedding_loss, vq_latent, _, perplexity = self.quantizer(pre_latent) + rec_pose = self.decoder(vq_latent) + return { + "poses_feat":vq_latent, + "embedding_loss":embedding_loss, + "perplexity":perplexity, + "rec_pose": rec_pose + } + + def map2index(self, inputs): + pre_latent = self.encoder(inputs) + index = self.quantizer.map2index(pre_latent) + return index + + def map2latent(self, inputs): + pre_latent = self.encoder(inputs) + index = self.quantizer.map2index(pre_latent) + z_q = self.quantizer.get_codebook_entry(index) + return z_q + + def decode(self, index): + z_q = self.quantizer.get_codebook_entry(index) + rec_pose = self.decoder(z_q) + return rec_pose + +class VQVAEConvZero2(nn.Module): + def __init__(self, args): + super(VQVAEConvZero2, self).__init__() + self.encoder = VQEncoderV5(args) + self.quantizer = Quantizer(args.vae_codebook_size, args.vae_length, args.vae_quantizer_lambda) + self.decoder = VQDecoderV7(args) + + def forward(self, inputs): + pre_latent = self.encoder(inputs) + # print(pre_latent.shape) + embedding_loss, vq_latent, _, perplexity = self.quantizer(pre_latent) + rec_pose = self.decoder(vq_latent) + return { + "poses_feat":vq_latent, + "embedding_loss":embedding_loss, + "perplexity":perplexity, + "rec_pose": rec_pose + } + + def map2index(self, inputs): + pre_latent = self.encoder(inputs) + index = self.quantizer.map2index(pre_latent) + return index + + def map2latent(self, inputs): + pre_latent = self.encoder(inputs) + index = self.quantizer.map2index(pre_latent) + z_q = self.quantizer.get_codebook_entry(index) + return z_q + + def decode(self, index): + z_q = self.quantizer.get_codebook_entry(index) + rec_pose = self.decoder(z_q) + return rec_pose + +class VQVAE2(nn.Module): + def __init__(self, args): + super(VQVAE2, self).__init__() + # Bottom-level encoder and decoder + args_bottom = copy.deepcopy(args) + args_bottom.vae_layer = 2 + self.bottom_encoder = VQEncoderV6(args_bottom) + self.bottom_quantizer = Quantizer(args.vae_codebook_size, args.vae_length, args.vae_quantizer_lambda) + args_bottom.vae_test_dim = args.vae_test_dim + self.bottom_decoder = VQDecoderV6(args_bottom) + + # Top-level encoder and decoder + args_top = copy.deepcopy(args) + args_top.vae_layer = 3 + args_top.vae_test_dim = args.vae_length + self.top_encoder = VQEncoderV3(args_top) # Adjust according to the top level's design + self.quantize_conv_t = nn.Conv1d(args.vae_length+args.vae_length, args.vae_length, 1) + self.top_quantizer = Quantizer(args.vae_codebook_size, args.vae_length, args.vae_quantizer_lambda) + # self.upsample_t_up = nn.Upsample(scale_factor=2, mode='nearest') + layers = [ + nn.Upsample(scale_factor=2, mode='nearest'), + nn.Conv1d(args.vae_length, args.vae_length, kernel_size=3, stride=1, padding=1), + nn.LeakyReLU(0.2, inplace=True), + nn.Upsample(scale_factor=2, mode='nearest'), + nn.Conv1d(args.vae_length, args.vae_length, kernel_size=3, stride=1, padding=1), + nn.LeakyReLU(0.2, inplace=True), + nn.Upsample(scale_factor=2, mode='nearest'), + nn.Conv1d(args.vae_length, args.vae_length, kernel_size=3, stride=1, padding=1), + nn.LeakyReLU(0.2, inplace=True) + ] + self.upsample_t= nn.Sequential(*layers) + self.top_decoder = VQDecoderV3(args_top) # Adjust to handle top level features appropriately + + def forward(self, inputs): + # Bottom-level processing + enc_b = self.bottom_encoder(inputs) + enc_t = self.top_encoder(enc_b) + #print(enc_b.shape, enc_t.shape) + top_embedding_loss, quant_t, _, top_perplexity = self.top_quantizer(enc_t) + #print(quant_t.shape) + dec_t = self.top_decoder(quant_t) + #print(dec_t.shape) + enc_b = torch.cat([dec_t, enc_b], dim=2).permute(0,2,1) + #print(enc_b.shape) + quant_b = self.quantize_conv_t(enc_b).permute(0,2,1) + #print("5",quant_b.shape) + bottom_embedding_loss, quant_b, _, bottom_perplexity = self.bottom_quantizer(quant_b) + #print("6",quant_b.shape) + upsample_t = self.upsample_t(quant_t.permute(0,2,1)).permute(0,2,1) + #print("7",upsample_t.shape) + quant = torch.cat([upsample_t, quant_b], 2) + rec_pose = self.bottom_decoder(quant) + # print(quant_t.shape, quant_b.shape, rec_pose.shape) + return { + "poses_feat_top": quant_t, + "pose_feat_bottom": quant_b, + "embedding_loss":top_embedding_loss+bottom_embedding_loss, + #"perplexity":perplexity, + "rec_pose": rec_pose + } + + def map2index(self, inputs): + enc_b = self.bottom_encoder(inputs) + enc_t = self.top_encoder(enc_b) + + _, quant_t, _, _ = self.top_quantizer(enc_t) + top_index = self.top_quantizer.map2index(enc_t) + dec_t = self.top_decoder(quant_t) + + enc_b = torch.cat([dec_t, enc_b], dim=2).permute(0,2,1) + #print(enc_b.shape) + quant_b = self.quantize_conv_t(enc_b).permute(0,2,1) + # quant_b = self.quantize_conv_t(enc_b) + bottom_index = self.bottom_quantizer.map2index(quant_b) + return top_index, bottom_index + + def get_top_laent(self, top_index): + z_q_top = self.top_quantizer.get_codebook_entry(top_index) + return z_q_top + + def map2latent(self, inputs): + enc_b = self.bottom_encoder(inputs) + enc_t = self.top_encoder(enc_b) + + _, quant_t, _, _ = self.top_quantizer(enc_t) + top_index = self.top_quantizer.map2index(enc_t) + dec_t = self.top_decoder(quant_t) + + enc_b = torch.cat([dec_t, enc_b], dim=2).permute(0,2,1) + #print(enc_b.shape) + quant_b = self.quantize_conv_t(enc_b).permute(0,2,1) + # quant_b = self.quantize_conv_t(enc_b) + bottom_index = self.bottom_quantizer.map2index(quant_b) + z_q_top = self.top_quantizer.get_codebook_entry(top_index) + z_q_bottom = self.bottom_quantizer.get_codebook_entry(bottom_index) + return z_q_top, z_q_bottom + + def map2latent_top(self, inputs): + enc_b = self.bottom_encoder(inputs) + enc_t = self.top_encoder(enc_b) + top_index = self.top_quantizer.map2index(enc_t) + z_q_top = self.top_quantizer.get_codebook_entry(top_index) + return z_q_top + + def decode(self, top_index, bottom_index): + quant_t = self.top_quantizer.get_codebook_entry(top_index) + quant_b = self.bottom_quantizer.get_codebook_entry(bottom_index) + upsample_t = self.upsample_t(quant_t.permute(0,2,1)).permute(0,2,1) + #print("7",upsample_t.shape) + quant = torch.cat([upsample_t, quant_b], 2) + rec_pose = self.bottom_decoder(quant) + return rec_pose \ No newline at end of file diff --git a/models/motionclip.py b/models/motionclip.py new file mode 100644 index 0000000000000000000000000000000000000000..4d00f4882e08ce518de3b5a341c49e4259204cee --- /dev/null +++ b/models/motionclip.py @@ -0,0 +1,272 @@ +import torch +import torch.nn as nn +import numpy as np +import clip + +class PositionalEncoding(nn.Module): + def __init__(self, d_model, dropout=0.1, max_len=5000): + super(PositionalEncoding, self).__init__() + self.dropout = nn.Dropout(p=dropout) + + pe = torch.zeros(max_len, d_model) + position = torch.arange(0, max_len, dtype=torch.float).unsqueeze(1) + div_term = torch.exp(torch.arange(0, d_model, 2).float() * (-np.log(10000.0) / d_model)) + pe[:, 0::2] = torch.sin(position * div_term) + pe[:, 1::2] = torch.cos(position * div_term) + pe = pe.unsqueeze(0).transpose(0, 1) + + self.register_buffer('pe', pe) + + def forward(self, x): + # not used in the final model + x = x + self.pe[:x.shape[0], :] + return self.dropout(x) + + +class Encoder_TRANSFORMER(nn.Module): + def __init__(self, modeltype, njoints, nfeats, num_frames, num_classes, translation, pose_rep, glob, glob_rot, + latent_dim=256, ff_size=1024, num_layers=4, num_heads=4, dropout=0.1, + ablation=None, activation="gelu", **kargs): + super().__init__() + + self.modeltype = modeltype + self.njoints = njoints + self.nfeats = nfeats + self.num_frames = num_frames + self.num_classes = num_classes + + self.pose_rep = pose_rep + self.glob = glob + self.glob_rot = glob_rot + self.translation = translation + + self.latent_dim = latent_dim + + self.ff_size = ff_size + self.num_layers = num_layers + self.num_heads = num_heads + self.dropout = dropout + + self.ablation = ablation + self.activation = activation + + self.input_feats = self.njoints*self.nfeats + + self.muQuery = nn.Parameter(torch.randn(1, self.latent_dim)) + self.sigmaQuery = nn.Parameter(torch.randn(1, self.latent_dim)) + self.skelEmbedding = nn.Linear(self.input_feats, self.latent_dim) + + self.sequence_pos_encoder = PositionalEncoding(self.latent_dim, self.dropout) + + seqTransEncoderLayer = nn.TransformerEncoderLayer(d_model=self.latent_dim, + nhead=self.num_heads, + dim_feedforward=self.ff_size, + dropout=self.dropout, + activation=self.activation) + self.seqTransEncoder = nn.TransformerEncoder(seqTransEncoderLayer, + num_layers=self.num_layers) + + def forward(self, batch): + x, y, mask = batch["x"], batch["y"], batch["mask"] + bs, nfeats, nframes = x.shape + x = x.permute((2, 0, 1)).reshape(nframes, bs, nfeats) + + # embedding of the skeleton + x = self.skelEmbedding(x) + + # Blank Y to 0's , no classes in our model, only learned token + y = y - y + xseq = torch.cat((self.muQuery[y][None], self.sigmaQuery[y][None], x), axis=0) + + # add positional encoding + xseq = self.sequence_pos_encoder(xseq) + + # create a bigger mask, to allow attend to mu and sigma + muandsigmaMask = torch.ones((bs, 2), dtype=bool, device=x.device) + + maskseq = torch.cat((muandsigmaMask, mask), axis=1) + + final = self.seqTransEncoder(xseq, src_key_padding_mask=~maskseq) + mu = final[0] + logvar = final[1] + + return {"mu": mu} + + +class Decoder_TRANSFORMER(nn.Module): + def __init__(self, modeltype, njoints, nfeats, num_frames, num_classes, translation, pose_rep, glob, glob_rot, + latent_dim=256, ff_size=1024, num_layers=4, num_heads=4, dropout=0.1, activation="gelu", + ablation=None, **kargs): + super().__init__() + + self.modeltype = modeltype + self.njoints = njoints + self.nfeats = nfeats + self.num_frames = num_frames + self.num_classes = num_classes + + self.pose_rep = pose_rep + self.glob = glob + self.glob_rot = glob_rot + self.translation = translation + + self.latent_dim = latent_dim + + self.ff_size = ff_size + self.num_layers = num_layers + self.num_heads = num_heads + self.dropout = dropout + + self.ablation = ablation + + self.activation = activation + + self.input_feats = self.njoints*self.nfeats + + # only for ablation / not used in the final model + if self.ablation == "zandtime": + self.ztimelinear = nn.Linear(self.latent_dim + self.num_classes, self.latent_dim) + + self.actionBiases = nn.Parameter(torch.randn(1, self.latent_dim)) + + # only for ablation / not used in the final model + if self.ablation == "time_encoding": + self.sequence_pos_encoder = TimeEncoding(self.dropout) + else: + self.sequence_pos_encoder = PositionalEncoding(self.latent_dim, self.dropout) + + seqTransDecoderLayer = nn.TransformerDecoderLayer(d_model=self.latent_dim, + nhead=self.num_heads, + dim_feedforward=self.ff_size, + dropout=self.dropout, + activation=activation) + self.seqTransDecoder = nn.TransformerDecoder(seqTransDecoderLayer, + num_layers=self.num_layers) + + self.finallayer = nn.Linear(self.latent_dim, self.input_feats) + + def forward(self, batch, use_text_emb=False): + z, y, mask, lengths = batch["z"], batch["y"], batch["mask"], batch["lengths"] + if use_text_emb: + z = batch["clip_text_emb"] + latent_dim = z.shape[1] + bs, nframes = mask.shape + njoints, nfeats = self.njoints, self.nfeats + + # only for ablation / not used in the final model + if self.ablation == "zandtime": + yoh = F.one_hot(y, self.num_classes) + z = torch.cat((z, yoh), axis=1) + z = self.ztimelinear(z) + z = z[None] # sequence of size 1 + else: + # only for ablation / not used in the final model + if self.ablation == "concat_bias": + # sequence of size 2 + z = torch.stack((z, self.actionBiases[y]), axis=0) + else: + z = z[None] # sequence of size 1 # + + timequeries = torch.zeros(nframes, bs, latent_dim, device=z.device) + + # only for ablation / not used in the final model + if self.ablation == "time_encoding": + timequeries = self.sequence_pos_encoder(timequeries, mask, lengths) + else: + timequeries = self.sequence_pos_encoder(timequeries) + + output = self.seqTransDecoder(tgt=timequeries, memory=z, + tgt_key_padding_mask=~mask) + + output = self.finallayer(output).reshape(nframes, bs, njoints, nfeats) + + # zero for padded area + output[~mask.T] = 0 + output = output.permute(1, 2, 3, 0) + + if use_text_emb: + batch["txt_output"] = output + else: + batch["output"] = output + return batch + + + +class MOTIONCLIP(nn.Module): + def __init__(self, encoder, decoder, device, lambdas, latent_dim, outputxyz, + pose_rep, glob, glob_rot, translation, jointstype, vertstrans, clip_lambdas={}, **kwargs): + super().__init__() + + self.encoder = encoder + self.decoder = decoder + + self.outputxyz = outputxyz + + self.lambdas = lambdas + self.clip_lambdas = clip_lambdas + + self.latent_dim = latent_dim + self.pose_rep = pose_rep + self.glob = glob + self.glob_rot = glob_rot + self.device = device + self.translation = translation + self.jointstype = jointstype + self.vertstrans = vertstrans + + self.clip_model = kwargs['clip_model'] + self.clip_training = kwargs.get('clip_training', False) + if self.clip_training and self.clip_model: + self.clip_model.training = True + else: + if self.clip_model: + assert self.clip_model.training == False # make sure clip is frozen + + + def forward(self, batch): + + # encode + batch.update(self.encoder(batch)) + batch["z"] = batch["mu"] + # decode + batch.update(self.decoder(batch)) + return batch + + + + +def get_gen_model(parameters, clip_model): + encoder = Encoder_TRANSFORMER(**parameters) + decoder = Decoder_TRANSFORMER(**parameters) + parameters["outputxyz"] = "rcxyz" in parameters["lambdas"] + return MOTIONCLIP(encoder, decoder, clip_model=clip_model, **parameters).to(parameters["device"]) + + +def get_model(parameters): + + # clip_model, preprocess = clip.load("ViT-B/32", device=device) # Must set jit=False for training + clip_model, clip_preprocess = clip.load("ViT-B/32", device=parameters['device'], jit=False) # Must set jit=False for training + clip.model.convert_weights(clip_model) # Actually this line is unnecessary since clip by default already on float16 + + for domain in parameters.get('clip_training', '').split('_'): + clip_num_layers = parameters.get('clip_layers', 12) + if domain == 'text': + clip_model.initialize_parameters() + clip_model.transformer.resblocks = clip_model.transformer.resblocks[:clip_num_layers] + if domain == 'image': + clip_model.initialize_parameters() + clip_model.visual.transformer = clip_model.transformer.resblocks[:clip_num_layers] + + # NO Clip Training ,Freeze CLIP weights + if parameters.get('clip_training', '') == '': + clip_model.eval() + for p in clip_model.parameters(): + p.requires_grad = False + + model = get_gen_model(parameters, clip_model) + return model + + + + + diff --git a/models/myvqvae.py b/models/myvqvae.py new file mode 100644 index 0000000000000000000000000000000000000000..40a46ed506a735f1fab869cddf68ad6a83fb35db --- /dev/null +++ b/models/myvqvae.py @@ -0,0 +1,847 @@ +import pdb + +import sys +[sys.path.append(i) for i in ['.', '..']] +sys.path.append("./models/qp_vqvae") +sys.path.append("./models/qp_vqvae/utils") +import numpy as np +import torch as t +import torch.nn as nn + +from .qp_vqvae.encdec import Encoder, Decoder, assert_shape +from .qp_vqvae.bottleneck import NoBottleneck, Bottleneck +from .qp_vqvae.utils.logger import average_metrics + +from .qp_vqvae.utils.torch_utils import parse_args +import torch.nn.functional as F +args = parse_args() +mydevice = t.device('cuda:' + args.gpu) + +def dont_update(params): + for param in params: + param.requires_grad = False + +def update(params): + for param in params: + param.requires_grad = True + +def calculate_strides(strides, downs): + return [stride ** down for stride, down in zip(strides, downs)] + +# def _loss_fn(loss_fn, x_target, x_pred, hps): +# if loss_fn == 'l1': +# return t.mean(t.abs(x_pred - x_target)) / hps.bandwidth['l1'] +# elif loss_fn == 'l2': +# return t.mean((x_pred - x_target) ** 2) / hps.bandwidth['l2'] +# elif loss_fn == 'linf': +# residual = ((x_pred - x_target) ** 2).reshape(x_target.shape[0], -1) +# values, _ = t.topk(residual, hps.linf_k, dim=1) +# return t.mean(values) / hps.bandwidth['l2'] +# elif loss_fn == 'lmix': +# loss = 0.0 +# if hps.lmix_l1: +# loss += hps.lmix_l1 * _loss_fn('l1', x_target, x_pred, hps) +# if hps.lmix_l2: +# loss += hps.lmix_l2 * _loss_fn('l2', x_target, x_pred, hps) +# if hps.lmix_linf: +# loss += hps.lmix_linf * _loss_fn('linf', x_target, x_pred, hps) +# return loss +# else: +# assert False, f"Unknown loss_fn {loss_fn}" +def _loss_fn(x_target, x_pred): + smooth_l1_loss = nn.SmoothL1Loss(reduction='none') + return smooth_l1_loss(x_pred,x_target).mean() + #return t.mean(t.abs(x_pred - x_target)) + +class VQVAE(nn.Module): + def __init__(self, hps, input_dim=72): + super().__init__() + self.hps = hps + input_dim=hps.pose_dims + input_shape = (hps.sample_length, input_dim) + levels = hps.levels + downs_t = hps.downs_t + strides_t = hps.strides_t + emb_width = hps.emb_width + l_bins = hps.l_bins + mu = hps.l_mu + commit = hps.commit + #root_weight = hps.root_weight + # spectral = hps.spectral + # multispectral = hps.multispectral + multipliers = hps.hvqvae_multipliers + use_bottleneck = hps.use_bottleneck + if use_bottleneck: + print('We use bottleneck!') + else: + print('We do not use bottleneck!') + if not hasattr(hps, 'dilation_cycle'): + hps.dilation_cycle = None + block_kwargs = dict(width=hps.width, depth=hps.depth, m_conv=hps.m_conv, \ + dilation_growth_rate=hps.dilation_growth_rate, \ + dilation_cycle=hps.dilation_cycle, \ + reverse_decoder_dilation=hps.vqvae_reverse_decoder_dilation) + + self.sample_length = input_shape[0] + x_shape, x_channels = input_shape[:-1], input_shape[-1] + self.x_shape = x_shape + + self.downsamples = calculate_strides(strides_t, downs_t) + self.hop_lengths = np.cumprod(self.downsamples) + self.z_shapes = z_shapes = [(x_shape[0] // self.hop_lengths[level],) for level in range(levels)] + self.levels = levels + + if multipliers is None: + self.multipliers = [1] * levels + else: + assert len(multipliers) == levels, "Invalid number of multipliers" + self.multipliers = multipliers + def _block_kwargs(level): + this_block_kwargs = dict(block_kwargs) + this_block_kwargs["width"] *= self.multipliers[level] + this_block_kwargs["depth"] *= self.multipliers[level] + return this_block_kwargs + + encoder = lambda level: Encoder(x_channels, emb_width, level + 1, + downs_t[:level+1], strides_t[:level+1], **_block_kwargs(level)) # different from supplemental + decoder = lambda level: Decoder(x_channels, emb_width, level + 1, + downs_t[:level+1], strides_t[:level+1], **_block_kwargs(level)) + self.encoders = nn.ModuleList() + self.decoders = nn.ModuleList() + for level in range(levels): + self.encoders.append(encoder(level)) + self.decoders.append(decoder(level)) + + if use_bottleneck: + self.bottleneck = Bottleneck(l_bins, emb_width, mu, levels) # 512, 512, 0.99, 1 + else: + self.bottleneck = NoBottleneck(levels) + + self.downs_t = downs_t + self.strides_t = strides_t + self.l_bins = l_bins + self.commit = commit + #self.root_weight = root_weight + self.reg = hps.reg if hasattr(hps, 'reg') else 0 + self.acc = hps.acc if hasattr(hps, 'acc') else 0 + self.vel = hps.vel if hasattr(hps, 'vel') else 0 + if self.reg == 0: + print('No motion regularization!') + # self.spectral = spectral + # self.multispectral = multispectral + + def preprocess(self, x): + # x: NTC [-1,1] -> NCT [-1,1] + assert len(x.shape) == 3 + x = x.permute(0,2,1).float() + return x + + def postprocess(self, x): + # x: NTC [-1,1] <- NCT [-1,1] + x = x.permute(0,2,1) + return x + + def _decode(self, zs, start_level=0, end_level=None): + # Decode + if end_level is None: + end_level = self.levels + assert len(zs) == end_level - start_level + xs_quantised = self.bottleneck.decode(zs, start_level=start_level, end_level=end_level) + assert len(xs_quantised) == end_level - start_level + + # Use only lowest level + decoder, x_quantised = self.decoders[start_level], xs_quantised[0:1] + x_out = decoder(x_quantised, all_levels=False) + x_out = self.postprocess(x_out) + return x_out + + def decode(self, zs, start_level=0, end_level=None, bs_chunks=1): + z_chunks = [t.chunk(z, bs_chunks, dim=0) for z in zs] + x_outs = [] + for i in range(bs_chunks): + zs_i = [z_chunk[i] for z_chunk in z_chunks] + x_out = self._decode(zs_i, start_level=start_level, end_level=end_level) + x_outs.append(x_out) + return t.cat(x_outs, dim=0) + + def _encode(self, x, start_level=0, end_level=None): + # Encode + if end_level is None: + end_level = self.levels + x_in = self.preprocess(x) + xs = [] + for level in range(self.levels): + encoder = self.encoders[level] + x_out = encoder(x_in) + xs.append(x_out[-1]) + zs = self.bottleneck.encode(xs) + return zs[start_level:end_level] + + def encode(self, x, start_level=0, end_level=None, bs_chunks=1): + x_chunks = t.chunk(x, bs_chunks, dim=0) + zs_list = [] + for x_i in x_chunks: + zs_i = self._encode(x_i, start_level=start_level, end_level=end_level) + zs_list.append(zs_i) + zs = [t.cat(zs_level_list, dim=0) for zs_level_list in zip(*zs_list)] + return zs + + def sample(self, n_samples): + zs = [t.randint(0, self.l_bins, size=(n_samples, *z_shape), device=mydevice) for z_shape in self.z_shapes] + return self.decode(zs) + + def forward(self, x): # ([256, 80, 282]) + metrics = {} + + N = x.shape[0] + + # Encode/Decode + x_in = self.preprocess(x) # ([256, 282, 80]) + xs = [] + for level in range(self.levels): + encoder = self.encoders[level] + x_out = encoder(x_in) + xs.append(x_out[-1]) + # xs[0]: (32, 512, 30) + zs, xs_quantised, commit_losses, quantiser_metrics = self.bottleneck(xs) #xs[0].shape=([256, 512, 5]) + ''' + zs[0]: (32, 30) + xs_quantised[0]: (32, 512, 30) + commit_losses[0]: 0.0009 + quantiser_metrics[0]: + fit 0.4646 + pn 0.0791 + entropy 5.9596 + used_curr 512 + usage 512 + dk 0.0006 + ''' + x_outs = [] + for level in range(self.levels): + decoder = self.decoders[level] + x_out = decoder(xs_quantised[level:level+1], all_levels=False) + assert_shape(x_out, x_in.shape) + x_outs.append(x_out) + # x_outs[0]: (32, 45, 240) + + # Loss + # def _spectral_loss(x_target, x_out, self.hps): + # if hps.use_nonrelative_specloss: + # sl = spectral_loss(x_target, x_out, self.hps) / hps.bandwidth['spec'] + # else: + # sl = spectral_convergence(x_target, x_out, self.hps) + # sl = t.mean(sl) + # return sl + + # def _multispectral_loss(x_target, x_out, self.hps): + # sl = multispectral_loss(x_target, x_out, self.hps) / hps.bandwidth['spec'] + # sl = t.mean(sl) + # return sl + + recons_loss = t.zeros(()).cuda() + regularization = t.zeros(()).cuda() + velocity_loss = t.zeros(()).cuda() + acceleration_loss = t.zeros(()).cuda() + # spec_loss = t.zeros(()).to(x.device) + # multispec_loss = t.zeros(()).to(x.device) + # x_target = audio_postprocess(x.float(), self.hps) + x_target = x.float() + + for level in reversed(range(self.levels)): + x_out = self.postprocess(x_outs[level]) # (32, 240, 45) + # x_out = audio_postprocess(x_out, self.hps) + + # scale_factor = t.ones(self.hps.pose_dims).to(x_target.device) + # scale_factor[:3]=self.root_weight + # x_target = x_target * scale_factor + # x_out = x_out * scale_factor + + # this_recons_loss = _loss_fn(loss_fn, x_target, x_out, hps) + this_recons_loss = _loss_fn(x_target, x_out) + # this_spec_loss = _spectral_loss(x_target, x_out, hps) + # this_multispec_loss = _multispectral_loss(x_target, x_out, hps) + metrics[f'recons_loss_l{level + 1}'] = this_recons_loss + # metrics[f'spectral_loss_l{level + 1}'] = this_spec_loss + # metrics[f'multispectral_loss_l{level + 1}'] = this_multispec_loss + recons_loss += this_recons_loss + # spec_loss += this_spec_loss + # multispec_loss += this_multispec_loss + regularization += t.mean((x_out[:, 2:] + x_out[:, :-2] - 2 * x_out[:, 1:-1])**2) + + velocity_loss += _loss_fn( x_out[:, 1:] - x_out[:, :-1], x_target[:, 1:] - x_target[:, :-1]) + acceleration_loss += _loss_fn(x_out[:, 2:] + x_out[:, :-2] - 2 * x_out[:, 1:-1], x_target[:, 2:] + x_target[:, :-2] - 2 * x_target[:, 1:-1]) + # if not hasattr(self.) + commit_loss = sum(commit_losses) + # loss = recons_loss + self.spectral * spec_loss + self.multispectral * multispec_loss + self.commit * commit_loss + # pdb.set_trace() + loss = recons_loss + commit_loss * self.commit + self.reg * regularization + self.vel * velocity_loss + self.acc * acceleration_loss + ''' x:-0.8474 ~ 1.1465 + 0.2080 + 5.5e-5 * 0.02 + 0.0011 + 0.0163 * 1 + 0.0274 * 1 + ''' + encodings = F.one_hot(zs[0].reshape(-1), self.hps.l_bins).float() + avg_probs = t.mean(encodings, dim=0) + perplexity = t.exp(-t.sum(avg_probs * t.log(avg_probs + 1e-10))) + with t.no_grad(): + # sc = t.mean(spectral_convergence(x_target, x_out, hps)) + # l2_loss = _loss_fn("l2", x_target, x_out, hps) + l1_loss = _loss_fn(x_target, x_out) + + # linf_loss = _loss_fn("linf", x_target, x_out, hps) + + quantiser_metrics = average_metrics(quantiser_metrics) + + metrics.update(dict( + loss = loss, + recons_loss=recons_loss, + # spectral_loss=spec_loss, + # multispectral_loss=multispec_loss, + # spectral_convergence=sc, + # l2_loss=l2_loss, + l1_loss=l1_loss, + # linf_loss=linf_loss, + commit_loss=commit_loss, + regularization=regularization, + velocity_loss=velocity_loss, + acceleration_loss=acceleration_loss, + perplexity=perplexity, + **quantiser_metrics)) + + for key, val in metrics.items(): + metrics[key] = val.detach() + + return { + # "poses_feat":vq_latent, + # "embedding_loss":embedding_loss, + # "perplexity":perplexity, + "rec_pose": x_out, + "loss": loss, + "metrics": metrics, + "embedding_loss": commit_loss * self.commit, + } + + +class VQVAE_Encoder(nn.Module): + def __init__(self, hps, input_dim=72): + super().__init__() + self.hps = hps + input_dim=hps.pose_dims + input_shape = (hps.sample_length, input_dim) + levels = hps.levels + downs_t = hps.downs_t + strides_t = hps.strides_t + emb_width = hps.emb_width + l_bins = hps.l_bins + mu = hps.l_mu + commit = hps.commit + # spectral = hps.spectral + # multispectral = hps.multispectral + multipliers = hps.hvqvae_multipliers + use_bottleneck = hps.use_bottleneck + if use_bottleneck: + print('We use bottleneck!') + else: + print('We do not use bottleneck!') + if not hasattr(hps, 'dilation_cycle'): + hps.dilation_cycle = None + block_kwargs = dict(width=hps.width, depth=hps.depth, m_conv=hps.m_conv, \ + dilation_growth_rate=hps.dilation_growth_rate, \ + dilation_cycle=hps.dilation_cycle, \ + reverse_decoder_dilation=hps.vqvae_reverse_decoder_dilation) + + self.sample_length = input_shape[0] + x_shape, x_channels = input_shape[:-1], input_shape[-1] + self.x_shape = x_shape + + self.downsamples = calculate_strides(strides_t, downs_t) + self.hop_lengths = np.cumprod(self.downsamples) + self.z_shapes = z_shapes = [(x_shape[0] // self.hop_lengths[level],) for level in range(levels)] + self.levels = levels + + if multipliers is None: + self.multipliers = [1] * levels + else: + assert len(multipliers) == levels, "Invalid number of multipliers" + self.multipliers = multipliers + def _block_kwargs(level): + this_block_kwargs = dict(block_kwargs) + this_block_kwargs["width"] *= self.multipliers[level] + this_block_kwargs["depth"] *= self.multipliers[level] + return this_block_kwargs + + encoder = lambda level: Encoder(x_channels, emb_width, level + 1, + downs_t[:level+1], strides_t[:level+1], **_block_kwargs(level)) # different from supplemental + decoder = lambda level: Decoder(x_channels, emb_width, level + 1, + downs_t[:level+1], strides_t[:level+1], **_block_kwargs(level)) + self.encoders = nn.ModuleList() + self.decoders = nn.ModuleList() + for level in range(levels): + self.encoders.append(encoder(level)) + self.decoders.append(decoder(level)) + + if use_bottleneck: + self.bottleneck = Bottleneck(l_bins, emb_width, mu, levels) # 512, 512, 0.99, 1 + else: + self.bottleneck = NoBottleneck(levels) + + self.downs_t = downs_t + self.strides_t = strides_t + self.l_bins = l_bins + self.commit = commit + self.reg = hps.reg if hasattr(hps, 'reg') else 0 + self.acc = hps.acc if hasattr(hps, 'acc') else 0 + self.vel = hps.vel if hasattr(hps, 'vel') else 0 + if self.reg == 0: + print('No motion regularization!') + # self.spectral = spectral + # self.multispectral = multispectral + + def preprocess(self, x): + # x: NTC [-1,1] -> NCT [-1,1] + assert len(x.shape) == 3 + x = x.permute(0,2,1).float() + return x + + def postprocess(self, x): + # x: NTC [-1,1] <- NCT [-1,1] + x = x.permute(0,2,1) + return x + + + def sample(self, n_samples): + zs = [t.randint(0, self.l_bins, size=(n_samples, *z_shape), device=mydevice) for z_shape in self.z_shapes] + return self.decode(zs) + + def forward(self, x): # ([256, 80, 282]) + metrics = {} + + N = x.shape[0] + + # Encode/Decode + x_in = self.preprocess(x) + xs = [] + for level in range(self.levels): + encoder = self.encoders[level] + x_out = encoder(x_in) + xs.append(x_out[-1]) + # xs[0]: (32, 512, 30) + zs, xs_quantised, commit_losses, quantiser_metrics = self.bottleneck(xs) #xs[0].shape=([256, 512, 5]) + return zs[0],xs[0] , xs_quantised[0] + + +class VQVAE_Decoder(nn.Module): + def __init__(self, hps, input_dim=72): + super().__init__() + self.hps = hps + input_dim=hps.pose_dims + input_shape = (hps.sample_length, input_dim) + levels = hps.levels + downs_t = hps.downs_t + strides_t = hps.strides_t + emb_width = hps.emb_width + l_bins = hps.l_bins + mu = hps.l_mu + commit = hps.commit + # spectral = hps.spectral + # multispectral = hps.multispectral + multipliers = hps.hvqvae_multipliers + use_bottleneck = hps.use_bottleneck + if use_bottleneck: + print('We use bottleneck!') + else: + print('We do not use bottleneck!') + if not hasattr(hps, 'dilation_cycle'): + hps.dilation_cycle = None + block_kwargs = dict(width=hps.width, depth=hps.depth, m_conv=hps.m_conv, \ + dilation_growth_rate=hps.dilation_growth_rate, \ + dilation_cycle=hps.dilation_cycle, \ + reverse_decoder_dilation=hps.vqvae_reverse_decoder_dilation) + + self.sample_length = input_shape[0] + x_shape, x_channels = input_shape[:-1], input_shape[-1] + self.x_shape = x_shape + + self.downsamples = calculate_strides(strides_t, downs_t) + self.hop_lengths = np.cumprod(self.downsamples) + self.z_shapes = z_shapes = [(x_shape[0] // self.hop_lengths[level],) for level in range(levels)] + self.levels = levels + + if multipliers is None: + self.multipliers = [1] * levels + else: + assert len(multipliers) == levels, "Invalid number of multipliers" + self.multipliers = multipliers + def _block_kwargs(level): + this_block_kwargs = dict(block_kwargs) + this_block_kwargs["width"] *= self.multipliers[level] + this_block_kwargs["depth"] *= self.multipliers[level] + return this_block_kwargs + + encoder = lambda level: Encoder(x_channels, emb_width, level + 1, + downs_t[:level+1], strides_t[:level+1], **_block_kwargs(level)) # different from supplemental + decoder = lambda level: Decoder(x_channels, emb_width, level + 1, + downs_t[:level+1], strides_t[:level+1], **_block_kwargs(level)) + self.encoders = nn.ModuleList() + self.decoders = nn.ModuleList() + for level in range(levels): + self.encoders.append(encoder(level)) + self.decoders.append(decoder(level)) + + if use_bottleneck: + self.bottleneck = Bottleneck(l_bins, emb_width, mu, levels) # 512, 512, 0.99, 1 + else: + self.bottleneck = NoBottleneck(levels) + + self.downs_t = downs_t + self.strides_t = strides_t + self.l_bins = l_bins + self.commit = commit + self.reg = hps.reg if hasattr(hps, 'reg') else 0 + self.acc = hps.acc if hasattr(hps, 'acc') else 0 + self.vel = hps.vel if hasattr(hps, 'vel') else 0 + if self.reg == 0: + print('No motion regularization!') + # self.spectral = spectral + # self.multispectral = multispectral + + def preprocess(self, x): + # x: NTC [-1,1] -> NCT [-1,1] + assert len(x.shape) == 3 + x = x.permute(0,2,1).float() + return x + + def postprocess(self, x): + # x: NTC [-1,1] <- NCT [-1,1] + x = x.permute(0,2,1) + return x + + + def forward(self, xs): # ([256, 80, 282]) + xs=[xs] + zs, xs_quantised, commit_losses, quantiser_metrics = self.bottleneck(xs) + x_outs = [] + for level in range(self.levels): + decoder = self.decoders[level] + x_out = decoder(xs_quantised[level:level+1], all_levels=False) + x_outs.append(x_out) + + + for level in reversed(range(self.levels)): + x_out = self.postprocess(x_outs[level]) # (32, 240, 45) + + return x_out + + + + +class Residual_VQVAE(nn.Module): + def __init__(self, hps, input_dim=72): + super().__init__() + self.hps = hps + input_dim=hps.pose_dims + input_shape = (hps.sample_length, input_dim) + levels = hps.levels + downs_t = hps.downs_t + strides_t = hps.strides_t + emb_width = hps.emb_width + l_bins = hps.l_bins + mu = hps.l_mu + commit = hps.commit + root_weight = hps.root_weight + # spectral = hps.spectral + # multispectral = hps.multispectral + multipliers = hps.hvqvae_multipliers + use_bottleneck = hps.use_bottleneck + if use_bottleneck: + print('We use bottleneck!') + else: + print('We do not use bottleneck!') + if not hasattr(hps, 'dilation_cycle'): + hps.dilation_cycle = None + block_kwargs = dict(width=hps.width, depth=hps.depth, m_conv=hps.m_conv, \ + dilation_growth_rate=hps.dilation_growth_rate, \ + dilation_cycle=hps.dilation_cycle, \ + reverse_decoder_dilation=hps.vqvae_reverse_decoder_dilation) + + self.sample_length = input_shape[0] + x_shape, x_channels = input_shape[:-1], input_shape[-1] + self.x_shape = x_shape + + self.downsamples = calculate_strides(strides_t, downs_t) + self.hop_lengths = np.cumprod(self.downsamples) + self.z_shapes = z_shapes = [(x_shape[0] // self.hop_lengths[level],) for level in range(levels)] + self.levels = levels + + if multipliers is None: + self.multipliers = [1] * levels + else: + assert len(multipliers) == levels, "Invalid number of multipliers" + self.multipliers = multipliers + def _block_kwargs(level): + this_block_kwargs = dict(block_kwargs) + this_block_kwargs["width"] *= self.multipliers[level] + this_block_kwargs["depth"] *= self.multipliers[level] + return this_block_kwargs + + encoder = lambda level: Encoder(x_channels, emb_width, level + 1, + downs_t[:level+1], strides_t[:level+1], **_block_kwargs(level)) # different from supplemental + decoder = lambda level: Decoder(x_channels, emb_width, level + 1, + downs_t[:level+1], strides_t[:level+1], **_block_kwargs(level)) + self.encoders = nn.ModuleList() + self.decoders = nn.ModuleList() + for level in range(levels): + self.encoders.append(encoder(level)) + self.decoders.append(decoder(level)) + + if use_bottleneck: + self.bottleneck = Bottleneck(l_bins, emb_width, mu, levels) # 512, 512, 0.99, 1 + else: + self.bottleneck = NoBottleneck(levels) + + self.downs_t = downs_t + self.strides_t = strides_t + self.l_bins = l_bins + self.commit = commit + self.root_weight = root_weight + self.reg = hps.reg if hasattr(hps, 'reg') else 0 + self.acc = hps.acc if hasattr(hps, 'acc') else 0 + self.vel = hps.vel if hasattr(hps, 'vel') else 0 + if self.reg == 0: + print('No motion regularization!') + # self.spectral = spectral + # self.multispectral = multispectral + + def preprocess(self, x): + # x: NTC [-1,1] -> NCT [-1,1] + assert len(x.shape) == 3 + x = x.permute(0,2,1).float() + return x + + def postprocess(self, x): + # x: NTC [-1,1] <- NCT [-1,1] + x = x.permute(0,2,1) + return x + + def _decode(self, zs, start_level=0, end_level=None): + # Decode + if end_level is None: + end_level = self.levels + assert len(zs) == end_level - start_level + xs_quantised = self.bottleneck.decode(zs, start_level=start_level, end_level=end_level) + assert len(xs_quantised) == end_level - start_level + + # Use only lowest level + decoder, x_quantised = self.decoders[start_level], xs_quantised[0:1] + x_out = decoder(x_quantised, all_levels=False) + x_out = self.postprocess(x_out) + return x_out + + def decode(self, zs, start_level=0, end_level=None, bs_chunks=1): + z_chunks = [t.chunk(z, bs_chunks, dim=0) for z in zs] + x_outs = [] + for i in range(bs_chunks): + zs_i = [z_chunk[i] for z_chunk in z_chunks] + x_out = self._decode(zs_i, start_level=start_level, end_level=end_level) + x_outs.append(x_out) + return t.cat(x_outs, dim=0) + + def _encode(self, x, start_level=0, end_level=None): + # Encode + if end_level is None: + end_level = self.levels + x_in = self.preprocess(x) + xs = [] + for level in range(self.levels): + encoder = self.encoders[level] + x_out = encoder(x_in) + xs.append(x_out[-1]) + zs = self.bottleneck.encode(xs) + return zs[start_level:end_level] + + def encode(self, x, start_level=0, end_level=None, bs_chunks=1): + x_chunks = t.chunk(x, bs_chunks, dim=0) + zs_list = [] + for x_i in x_chunks: + zs_i = self._encode(x_i, start_level=start_level, end_level=end_level) + zs_list.append(zs_i) + zs = [t.cat(zs_level_list, dim=0) for zs_level_list in zip(*zs_list)] + return zs + + def sample(self, n_samples): + zs = [t.randint(0, self.l_bins, size=(n_samples, *z_shape), device=mydevice) for z_shape in self.z_shapes] + return self.decode(zs) + + def forward(self, x): # ([256, 80, 282]) + metrics = {} + + N = x.shape[0] + + # Encode/Decode + x_in = self.preprocess(x) # ([256, 282, 80]) + xs = [] + for level in range(self.levels): + encoder = self.encoders[level] + x_out = encoder(x_in) + xs.append(x_out[-1]) + # xs[0]: (32, 512, 30) + zs, xs_quantised, commit_losses, quantiser_metrics = self.bottleneck(xs) #xs[0].shape=([256, 512, 5]) + ''' + zs[0]: (32, 30) + xs_quantised[0]: (32, 512, 30) + commit_losses[0]: 0.0009 + quantiser_metrics[0]: + fit 0.4646 + pn 0.0791 + entropy 5.9596 + used_curr 512 + usage 512 + dk 0.0006 + ''' + x_outs = [] + for level in range(self.levels): + decoder = self.decoders[level] + x_out = decoder(xs_quantised[level:level+1], all_levels=False) + assert_shape(x_out, x_in.shape) + x_outs.append(x_out) + # x_outs[0]: (32, 45, 240) + + # Loss + # def _spectral_loss(x_target, x_out, self.hps): + # if hps.use_nonrelative_specloss: + # sl = spectral_loss(x_target, x_out, self.hps) / hps.bandwidth['spec'] + # else: + # sl = spectral_convergence(x_target, x_out, self.hps) + # sl = t.mean(sl) + # return sl + + # def _multispectral_loss(x_target, x_out, self.hps): + # sl = multispectral_loss(x_target, x_out, self.hps) / hps.bandwidth['spec'] + # sl = t.mean(sl) + # return sl + + recons_loss = t.zeros(()).cuda() + regularization = t.zeros(()).cuda() + velocity_loss = t.zeros(()).cuda() + acceleration_loss = t.zeros(()).cuda() + # spec_loss = t.zeros(()).to(x.device) + # multispec_loss = t.zeros(()).to(x.device) + # x_target = audio_postprocess(x.float(), self.hps) + x_target = x.float() + + for level in reversed(range(self.levels)): + x_out = self.postprocess(x_outs[level]) # (32, 240, 45) + # x_out = audio_postprocess(x_out, self.hps) + + scale_factor = t.ones(self.hps.pose_dims).to(x_target.device) + scale_factor[:3]=self.root_weight + x_target = x_target * scale_factor + x_out = x_out * scale_factor + + # this_recons_loss = _loss_fn(loss_fn, x_target, x_out, hps) + this_recons_loss = _loss_fn(x_target, x_out) + # this_spec_loss = _spectral_loss(x_target, x_out, hps) + # this_multispec_loss = _multispectral_loss(x_target, x_out, hps) + metrics[f'recons_loss_l{level + 1}'] = this_recons_loss + # metrics[f'spectral_loss_l{level + 1}'] = this_spec_loss + # metrics[f'multispectral_loss_l{level + 1}'] = this_multispec_loss + recons_loss += this_recons_loss + # spec_loss += this_spec_loss + # multispec_loss += this_multispec_loss + regularization += t.mean((x_out[:, 2:] + x_out[:, :-2] - 2 * x_out[:, 1:-1])**2) + + velocity_loss += _loss_fn( x_out[:, 1:] - x_out[:, :-1], x_target[:, 1:] - x_target[:, :-1]) + acceleration_loss += _loss_fn(x_out[:, 2:] + x_out[:, :-2] - 2 * x_out[:, 1:-1], x_target[:, 2:] + x_target[:, :-2] - 2 * x_target[:, 1:-1]) + # if not hasattr(self.) + commit_loss = sum(commit_losses) + # loss = recons_loss + self.spectral * spec_loss + self.multispectral * multispec_loss + self.commit * commit_loss + # pdb.set_trace() + loss = recons_loss + commit_loss * self.commit + self.reg * regularization + self.vel * velocity_loss + self.acc * acceleration_loss + ''' x:-0.8474 ~ 1.1465 + 0.2080 + 5.5e-5 * 0.02 + 0.0011 + 0.0163 * 1 + 0.0274 * 1 + ''' + encodings = F.one_hot(zs[0].reshape(-1), self.hps.l_bins).float() + avg_probs = t.mean(encodings, dim=0) + perplexity = t.exp(-t.sum(avg_probs * t.log(avg_probs + 1e-10))) + with t.no_grad(): + # sc = t.mean(spectral_convergence(x_target, x_out, hps)) + # l2_loss = _loss_fn("l2", x_target, x_out, hps) + l1_loss = _loss_fn(x_target, x_out) + + # linf_loss = _loss_fn("linf", x_target, x_out, hps) + + quantiser_metrics = average_metrics(quantiser_metrics) + + metrics.update(dict( + loss = loss, + recons_loss=recons_loss, + # spectral_loss=spec_loss, + # multispectral_loss=multispec_loss, + # spectral_convergence=sc, + # l2_loss=l2_loss, + l1_loss=l1_loss, + # linf_loss=linf_loss, + commit_loss=commit_loss, + regularization=regularization, + velocity_loss=velocity_loss, + acceleration_loss=acceleration_loss, + perplexity=perplexity, + **quantiser_metrics)) + + for key, val in metrics.items(): + metrics[key] = val.detach() + + return { + # "poses_feat":vq_latent, + # "embedding_loss":embedding_loss, + # "perplexity":perplexity, + "rec_pose": x_out, + "loss": loss, + "metrics": metrics, + "embedding_loss": commit_loss * self.commit, + } + + + + + + + + + + + + + +if __name__ == '__main__': + ''' + cd codebook/ + python vqvae.py --config=./codebook.yml --train --no_cuda 2 --gpu 2 + ''' + import yaml + from pprint import pprint + from easydict import EasyDict + + with open(args.config) as f: + config = yaml.safe_load(f) + + for k, v in vars(args).items(): + config[k] = v + pprint(config) + + config = EasyDict(config) + + x = t.rand(32, 40, 15 * 9).to(mydevice) + model = VQVAE(config.VQVAE, 15 * 9) # n_joints * n_chanels + + model = nn.DataParallel(model, device_ids=[eval(i) for i in config.no_cuda]) + model = model.to(mydevice) + model = model.train() + output, loss, metrics = model(x) + pdb.set_trace() diff --git a/models/qp_vqvae/__init__.py b/models/qp_vqvae/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/models/qp_vqvae/bottleneck.py b/models/qp_vqvae/bottleneck.py new file mode 100644 index 0000000000000000000000000000000000000000..dd98a717fed0ca3d5c4c28dd23f617556fa99438 --- /dev/null +++ b/models/qp_vqvae/bottleneck.py @@ -0,0 +1,331 @@ +import pdb + +import numpy as np +import torch as t +import torch.nn as nn +import torch.nn.functional as F +import utils.dist_adapter as dist +import sys +[sys.path.append(i) for i in ['.', '..']] +from utils.torch_utils import parse_args + + +args = parse_args() +mydevice = t.device('cuda:' + args.gpu) + +class BottleneckBlock(nn.Module): + def __init__(self, k_bins, emb_width, mu): + super().__init__() + self.k_bins = k_bins + self.emb_width = emb_width + self.mu = mu + self.reset_k() + self.threshold = 1.0 + + def reset_k(self): + self.init = False + self.k_sum = None + self.k_elem = None + self.register_buffer('k', t.zeros(self.k_bins, self.emb_width).cuda()) + + def _tile(self, x): + d, ew = x.shape # 960, 512 + if d < self.k_bins: + n_repeats = (self.k_bins + d - 1) // d + std = 0.01 / np.sqrt(ew) + x = x.repeat(n_repeats, 1) + x = x + t.randn_like(x) * std + return x + + def init_k(self, x): + mu, emb_width, k_bins = self.mu, self.emb_width, self.k_bins # mu=0.99, emb_width=512, k_bins=512 + self.init = True + # init k_w using random vectors from x + y = self._tile(x) + _k_rand = y[t.randperm(y.shape[0])][:k_bins] # (512, 512), a random permutation of integers from 0 to n - 1 + # dist.broadcast(_k_rand, 0) + self.k = _k_rand + assert self.k.shape == (k_bins, emb_width) + self.k_sum = self.k + self.k_elem = t.ones(k_bins, device=self.k.device) + + def restore_k(self, num_tokens=None, threshold=1.0): + mu, emb_width, k_bins = self.mu, self.emb_width, self.k_bins + self.init = True + assert self.k.shape == (k_bins, emb_width) + self.k_sum = self.k.clone() + self.k_elem = t.ones(k_bins, device=self.k.device) + if num_tokens is not None: + expected_usage = num_tokens / k_bins + self.k_elem.data.mul_(expected_usage) + self.k_sum.data.mul_(expected_usage) + self.threshold = threshold + + def update_k(self, x, x_l): # (960, 512), (960) + mu, emb_width, k_bins = self.mu, self.emb_width, self.k_bins # mu=0.99, emb_width=512, k_bins=512 + with t.no_grad(): + # Calculate new centres + x_l_onehot = t.zeros(k_bins, x.shape[0], device=x.device) # (512(k_bins), 960(N * L)) + x_l_onehot.scatter_(0, x_l.view(1, x.shape[0]), 1) # (1, 190) -> (512, 960), find which axis + + _k_sum = t.matmul(x_l_onehot, x) #(512(k_bins), 512(w)) + _k_elem = x_l_onehot.sum(dim=-1) # (512(k_bins)) + y = self._tile(x) # (960, 512) + _k_rand = y[t.randperm(y.shape[0])][:k_bins] # (512, 512) + + # dist.broadcast(_k_rand, 0) + # dist.all_reduce(_k_sum) + # dist.all_reduce(_k_elem) + + # Update centres + old_k = self.k + self.k_sum = mu * self.k_sum + (1. - mu) * _k_sum # w, k_bins + self.k_elem = mu * self.k_elem + (1. - mu) * _k_elem # k_bins + usage = (self.k_elem.view(k_bins, 1) >= self.threshold).float() + self.k = usage * (self.k_sum.view(k_bins, emb_width) / self.k_elem.view(k_bins, 1)) \ + + (1 - usage) * _k_rand + _k_prob = _k_elem / t.sum(_k_elem) # x_l_onehot.mean(dim=-1) # prob of each bin + entropy = -t.sum(_k_prob * t.log(_k_prob + 1e-8)) # entropy ie how diverse + used_curr = (_k_elem >= self.threshold).sum() + usage = t.sum(usage) + dk = t.norm(self.k - old_k) / np.sqrt(np.prod(old_k.shape)) + return dict(entropy=entropy, + used_curr=used_curr, + usage=usage, + dk=dk) + + def preprocess(self, x): + # NCT -> NTC -> [NT, C] + x = x.permute(0, 2, 1).contiguous() + x = x.view(-1, x.shape[-1]) # x_en = (N * L, w), k_j = (w, k_bins) + + if x.shape[-1] == self.emb_width: + prenorm = t.norm(x - t.mean(x)) / np.sqrt(np.prod(x.shape)) # np.sqrt - product of array elements over a given axis + elif x.shape[-1] == 2 * self.emb_width: + x1, x2 = x[...,:self.emb_width], x[...,self.emb_width:] + prenorm = (t.norm(x1 - t.mean(x1)) / np.sqrt(np.prod(x1.shape))) + (t.norm(x2 - t.mean(x2)) / np.sqrt(np.prod(x2.shape))) + + # Normalise + x = x1 + x2 + else: + assert False, f"Expected {x.shape[-1]} to be (1 or 2) * {self.emb_width}" + return x, prenorm + + def postprocess(self, x_l, x_d, x_shape): + # [NT, C] -> NTC -> NCT + N, T = x_shape + x_d = x_d.view(N, T, -1).permute(0, 2, 1).contiguous() + x_l = x_l.view(N, T) + return x_l, x_d + + def quantise(self, x): + # Calculate latent code x_l + k_w = self.k.t() # (512, 512) + distance = t.sum(x ** 2, dim=-1, keepdim=True) - 2 * t.matmul(x, k_w) + t.sum(k_w ** 2, dim=0, keepdim=True) # (960(N * L), 512(b)) + min_distance, x_l = t.min(distance, dim=-1) # (960), (960) + fit = t.mean(min_distance) + return x_l, fit + + def dequantise(self, x_l): + x = F.embedding(x_l, self.k) # self.k: (512, 512) weighted array + return x + + def encode(self, x): + N, width, T = x.shape + + # Preprocess. + x, prenorm = self.preprocess(x) + + # Quantise + x_l, fit = self.quantise(x) + + # Postprocess. + x_l = x_l.view(N, T) + return x_l + + def decode(self, x_l): + N, T = x_l.shape + width = self.emb_width + + # Dequantise + x_d = self.dequantise(x_l) + + # Postprocess + x_d = x_d.view(N, T, width).permute(0, 2, 1).contiguous() + return x_d + + def forward(self, x, update_k=True): + N, width, T = x.shape # 32, 512, 30 + + # Preprocess + x, prenorm = self.preprocess(x) # (960, 512), 0.2888 + + # Init k if not inited + if update_k and not self.init: + self.init_k(x) + + # Quantise and dequantise through bottleneck + x_l, fit = self.quantise(x) # (960), 34.1081 + x_d = self.dequantise(x_l) # (960, 512) + + # Update embeddings + if update_k: + update_metrics = self.update_k(x, x_l) + else: + update_metrics = {} + + # Loss + commit_loss = t.norm(x_d.detach() - x) ** 2 / np.prod(x.shape) # L2 loss -> L1 loss + + # Passthrough + x_d = x + (x_d - x).detach() + + # Postprocess + x_l, x_d = self.postprocess(x_l, x_d, (N,T)) + return x_l, x_d, commit_loss, dict(fit=fit, + pn=prenorm, + **update_metrics) + + +class Bottleneck(nn.Module): + def __init__(self, l_bins, emb_width, mu, levels): + super().__init__() + self.levels = levels + level_block = lambda level: BottleneckBlock(l_bins, emb_width, mu) + self.level_blocks = nn.ModuleList() + for level in range(self.levels): + self.level_blocks.append(level_block(level)) + + def encode(self, xs): + zs = [level_block.encode(x) for (level_block, x) in zip(self.level_blocks, xs)] + return zs + + def decode(self, zs, start_level=0, end_level=None): + if end_level is None: + end_level = self.levels + xs_quantised = [level_block.decode(z) for (level_block, z) in zip(self.level_blocks[start_level:end_level], zs)] + return xs_quantised + + def forward(self, xs): + zs, xs_quantised, commit_losses, metrics = [], [], [], [] + for level in range(self.levels): + level_block = self.level_blocks[level] + x = xs[level] # (32, 512, 30) + z, x_quantised, commit_loss, metric = level_block(x, update_k=self.training) + ''' + z: (32, 30) + x_quantised: (32, 512, 30) + commit_loss: 0.0666 + metric: same as models/vqvae.py + ''' + zs.append(z) + if not self.training: + # Be extra paranoid and make sure the encoder weights can't + # change from straight-through estimator + x_quantised = x_quantised.detach() + xs_quantised.append(x_quantised) + commit_losses.append(commit_loss) + if self.training: + metrics.append(metric) + return zs, xs_quantised, commit_losses, metrics + + + + + +class Residual_Bottleneck(nn.Module): + def __init__(self, l_bins, emb_width, mu, levels): + super().__init__() + self.levels = levels + self.residuals = 4 + level_block = lambda level: BottleneckBlock(l_bins, emb_width, mu) + self.level_blocks = nn.ModuleList() + for level in range(self.levels): + self.level_blocks.append(level_block(level)) + + for residual in range(self.residuals): + self.residual_blocks.append(level_block(residual)) + + def encode(self, xs): + zs = [level_block.encode(x) for (level_block, x) in zip(self.level_blocks, xs)] + return zs + + def decode(self, zs, start_level=0, end_level=None): + if end_level is None: + end_level = self.levels + xs_quantised = [level_block.decode(z) for (level_block, z) in zip(self.level_blocks[start_level:end_level], zs)] + return xs_quantised + + def forward(self, xs): + zs, xs_quantised, commit_losses, metrics = [], [], [], [] + for level in range(self.levels): + level_block = self.level_blocks[level] + x = xs[level] # (32, 512, 30) + + residual = x + quantized_out = 0. + + for residual_num in range(self.residuals): + residual_block = self.residual_blocks[residual_num] + z, x_quantised, commit_loss, metric = residual_block(x, update_k=self.training) + residual = residual - x_quantised.detach() + quantized_out = quantized_out + x_quantised + + z, x_quantised, commit_loss, metric = level_block(x, update_k=self.training) + ''' + z: (32, 30) + x_quantised: (32, 512, 30) + commit_loss: 0.0666 + metric: same as models/vqvae.py + ''' + zs.append(z) + if not self.training: + # Be extra paranoid and make sure the encoder weights can't + # change from straight-through estimator + x_quantised = x_quantised.detach() + xs_quantised.append(x_quantised) + commit_losses.append(commit_loss) + if self.training: + metrics.append(metric) + return zs, xs_quantised, commit_losses, metrics + + + + +class NoBottleneckBlock(nn.Module): + def restore_k(self): + pass + +class NoBottleneck(nn.Module): + def __init__(self, levels): + super().__init__() + self.level_blocks = nn.ModuleList() + self.levels = levels + for level in range(levels): + self.level_blocks.append(NoBottleneckBlock()) + + def encode(self, xs): + return xs + + def decode(self, zs, start_level=0, end_level=None): + if end_level is None: + end_level = self.levels + return zs + + def forward(self, xs): + zero = t.zeros(()).cuda() + commit_losses = [zero for _ in range(self.levels)] + metrics = [dict(entropy=zero, usage=zero, used_curr=zero, pn=zero, dk=zero) for _ in range(self.levels)] + return xs, xs, commit_losses, metrics + +if __name__ == '__main__': + ''' + python -m models.bottleneck --config configs/sep_vqvae.yaml --train --no_cuda 2 --gpu 2 + ''' + # x = [t.rand(32, 512, 30)] + # bottleneck = Bottleneck(512, 512, 0.99, 1).to(mydevice) + # zs, xs_quantised, commit_losses, quantiser_metrics = bottleneck(x) + + x = t.rand(32, 512, 30) + model = BottleneckBlock(k_bins=512, emb_width=512, mu=0.99) + zs, xs_quantised, commit_losses, quantiser_metrics = model(x) diff --git a/models/qp_vqvae/codebook.yml b/models/qp_vqvae/codebook.yml new file mode 100644 index 0000000000000000000000000000000000000000..d0532bb0531cd9fe6bc57dfca84e17e2e5f47dbf --- /dev/null +++ b/models/qp_vqvae/codebook.yml @@ -0,0 +1,71 @@ +VQVAE: + #Codebook Configs + levels: 1 + downs_t: [3,] # 3 -> 1 + strides_t : [2,] # 2 -> 3 + emb_width : 512 + l_bins : 512 + l_mu : 0.99 + commit : 0.02 + hvqvae_multipliers : [1,] + width: 512 + depth: 3 + m_conv : 1.0 + dilation_growth_rate : 3 + sample_length: 30 + use_bottleneck: True + joint_channel: 9 + # depth: 3 + # width: 128 + # m_conv: 1.0 + # dilation_growth_rate: 1 + # dilation_cycle: None + vel: 1 # 1 -> 0 + acc: 1 # 1 -> 0 + vqvae_reverse_decoder_dilation: True + +# BEAT +train_data_path: "../dataset/BEAT/speaker_10_state_0/lmdb/lmdb_train" # speaker_1_state_0 +val_data_path: "../dataset/BEAT/speaker_10_state_0/lmdb/lmdb_valid" + +# 60 fps + rotation lmdb +data_mean: [0.96776, 0.03511, -0.00725, -0.03507, 0.96983, -0.00267, 0.00705, 0.00363, 0.99770, 0.99930, 0.01376, 0.00255, -0.01374, 0.99938, -0.00253, -0.00263, 0.00236, 0.99987, 0.99216, 0.01860, -0.00709, -0.01882, 0.98965, -0.02039, 0.00531, 0.02222, 0.99601, 0.99612, -0.00993, -0.00998, 0.00991, 0.99691, -0.00743, 0.00991, 0.00820, 0.99890, 0.97768, 0.03605, 0.00750, -0.03840, 0.98164, 0.01540, -0.00722, -0.01537, 0.98648, 0.97946, 0.01763, 0.03590, -0.01997, 0.97667, 0.02363, -0.03636, -0.02139, 0.97316, 0.99365, 0.00565, 0.00280, -0.00546, 0.99544, -0.00042, -0.00264, 0.00288, 0.99802, 0.96583, -0.00000, 0.03884, -0.01122, 0.93705, 0.25010, -0.03542, -0.24257, 0.94337, 0.42426, -0.00898, 0.03954, 0.00077, 0.78974, -0.20711, -0.03733, 0.36827, 0.39807, 0.77377, -0.03384, -0.00000, 0.02925, 0.01928, -0.84132, 0.03604, 0.63766, 0.00386, 0.92148, -0.00060, 0.02677, 0.00000, 0.96406, 0.08089, -0.02854, -0.07688, 0.91554, 0.97416, 0.00000, 0.01114, 0.00433, 0.96865, -0.19804, -0.01083, 0.19909, 0.95347, 0.40921, -0.00936, -0.05575, -0.00286, 0.82451, 0.12718, 0.04597, -0.30625, 0.40531, 0.79050, 0.02049, 0.00000, 0.01737, -0.00978, 0.82851, 0.00535, -0.66105, -0.02535, 0.91757, -0.00631, 0.02182, -0.00000, 0.95817, -0.09851, -0.01829, 0.09236, 0.91430] +data_std: [0.02841, 0.23917, 0.06431, 0.23880, 0.02782, 0.01986, 0.06575, 0.01430, 0.00336, 0.00108, 0.03157, 0.01468, 0.03155, 0.00106, 0.00621, 0.01473, 0.00612, 0.00018, 0.01512, 0.11668, 0.03705, 0.11614, 0.01536, 0.07804, 0.03890, 0.07677, 0.00516, 0.00416, 0.07577, 0.04228, 0.07578, 0.00407, 0.01605, 0.04228, 0.01571, 0.00136, 0.02789, 0.15938, 0.12886, 0.15728, 0.02378, 0.09677, 0.13076, 0.09421, 0.02422, 0.03618, 0.12665, 0.14733, 0.12803, 0.07007, 0.15448, 0.14572, 0.15622, 0.07448, 0.02085, 0.09297, 0.05959, 0.09228, 0.01465, 0.01851, 0.06068, 0.01429, 0.00739, 0.20466, 0.00001, 0.15420, 0.04271, 0.19514, 0.13914, 0.14860, 0.15817, 0.05350, 0.16375, 0.29095, 0.84077, 0.46422, 0.24234, 0.24331, 0.75914, 0.31183, 0.17603, 0.28407, 0.56519, 0.00013, 0.23171, 0.44004, 0.20889, 0.51453, 0.28072, 0.49853, 0.18833, 0.07924, 0.32926, 0.00007, 0.16923, 0.18816, 0.33851, 0.17250, 0.10350, 0.10469, 0.00000, 0.19984, 0.04231, 0.09819, 0.10514, 0.19528, 0.11157, 0.02373, 0.15367, 0.26760, 0.85681, 0.43448, 0.24673, 0.23313, 0.78615, 0.30631, 0.16722, 0.28200, 0.54330, 0.00016, 0.22882, 0.44907, 0.24319, 0.49285, 0.25625, 0.50377, 0.20667, 0.07828, 0.32970, 0.00034, 0.19347, 0.18646, 0.33914, 0.17255, 0.10199] + +n_poses: 240 # 30 -> 40 -> 240 +n_codes: 30 +motion_resampling_framerate: 60 # 20 -> 60 +subdivision_stride: 32 # 10 -> 30 +batch_size: 256 +loader_workers: 2 +epochs: 500 # 500 -> 10 +save_per_epochs: 25 # 20 -> 1 +model_save_path: "./output/train_codebook" +name: "codebook" + +lr: 0.00003 # 0.00003 -> +betas: [0.5, 0.999] +milestones: [100, 200] +gamma: 0.1 + +end2end: + lr: 0.0002 + epochs: 100 + betas: [0.99, 0.999] + model_save_path: "./output/train_end2end" + save_per_epochs: 10 + name: "end2end" + +PAE: + epochs: 100 + save_per_epochs: 10 + n_poses: 240 + subdivision_stride: 1 + model_save_path: "./output/train_PAE" + figs_save_path: "./output/train_PAE/figs" + name: "PAE" + +beat_data_to_lmdb: + path: "../dataset/BEAT/speaker_10_state_0" + lmdb_name: "lmdb" + mode: "rotation" diff --git a/models/qp_vqvae/encdec.py b/models/qp_vqvae/encdec.py new file mode 100644 index 0000000000000000000000000000000000000000..5db19825dff8b170a56fed48f653df4e3d4fb1be --- /dev/null +++ b/models/qp_vqvae/encdec.py @@ -0,0 +1,136 @@ +import pdb + +import torch as t +import torch.nn as nn +from resnet import Resnet, Resnet1D +from utils.torch_utils import assert_shape + +class EncoderConvBlock(nn.Module): + def __init__(self, input_emb_width, output_emb_width, down_t, + stride_t, width, depth, m_conv, + dilation_growth_rate=1, dilation_cycle=None, zero_out=False, + res_scale=False): + super().__init__() + blocks = [] + filter_t, pad_t = stride_t * 2, stride_t // 2 + if down_t > 0: + for i in range(down_t): + block = nn.Sequential( + # nn.Conv1d(input_emb_width if i == 0 else width, width, filter_t, stride_t, pad_t, padding_mode='replicate'), + nn.Conv1d(input_emb_width if i == 0 else width, width, filter_t, stride_t, pad_t), + Resnet1D(width, depth, m_conv, dilation_growth_rate, dilation_cycle, zero_out, res_scale), + ) + blocks.append(block) + block = nn.Conv1d(width, output_emb_width, 3, 1, 1) + # block = nn.Conv1d(width, output_emb_width, 3, 1, 1, padding_mode='replicate') + blocks.append(block) + self.model = nn.Sequential(*blocks) + + def forward(self, x): + return self.model(x) + +class DecoderConvBock(nn.Module): + def __init__(self, input_emb_width, output_emb_width, down_t, + stride_t, width, depth, m_conv, dilation_growth_rate=1, dilation_cycle=None, zero_out=False, res_scale=False, reverse_decoder_dilation=False, checkpoint_res=False): + super().__init__() + blocks = [] + if down_t > 0: + filter_t, pad_t = stride_t * 2, stride_t // 2 + block = nn.Conv1d(output_emb_width, width, 3, 1, 1) + # block = nn.Conv1d(output_emb_width, width, 3, 1, 1, padding_mode='replicate') + blocks.append(block) + for i in range(down_t): + block = nn.Sequential( + Resnet1D(width, depth, m_conv, dilation_growth_rate, dilation_cycle, zero_out=zero_out, res_scale=res_scale, reverse_dilation=reverse_decoder_dilation, checkpoint_res=checkpoint_res), + nn.ConvTranspose1d(width, input_emb_width if i == (down_t - 1) else width, filter_t, stride_t, pad_t) + ) + blocks.append(block) + self.model = nn.Sequential(*blocks) + + def forward(self, x): + return self.model(x) + +class Encoder(nn.Module): + def __init__(self, input_emb_width, output_emb_width, levels, downs_t, + strides_t, **block_kwargs): + super().__init__() + self.input_emb_width = input_emb_width + self.output_emb_width = output_emb_width + self.levels = levels + self.downs_t = downs_t + self.strides_t = strides_t + + block_kwargs_copy = dict(**block_kwargs) + if 'reverse_decoder_dilation' in block_kwargs_copy: + del block_kwargs_copy['reverse_decoder_dilation'] + level_block = lambda level, down_t, stride_t: EncoderConvBlock(input_emb_width if level == 0 else output_emb_width, + output_emb_width, + down_t, stride_t, + **block_kwargs_copy) + self.level_blocks = nn.ModuleList() + iterator = zip(list(range(self.levels)), downs_t, strides_t) + for level, down_t, stride_t in iterator: + self.level_blocks.append(level_block(level, down_t, stride_t)) + + def forward(self, x): + N, T = x.shape[0], x.shape[-1] + emb = self.input_emb_width + assert_shape(x, (N, emb, T)) + xs = [] + + # 64, 32, ... + iterator = zip(list(range(self.levels)), self.downs_t, self.strides_t) + for level, down_t, stride_t in iterator: + level_block = self.level_blocks[level] + x = level_block(x) + emb, T = self.output_emb_width, T // (stride_t ** down_t) + assert_shape(x, (N, emb, T)) + xs.append(x) + + return xs + +class Decoder(nn.Module): + def __init__(self, input_emb_width, output_emb_width, levels, downs_t, + strides_t, **block_kwargs): + super().__init__() + self.input_emb_width = input_emb_width + self.output_emb_width = output_emb_width + self.levels = levels + + self.downs_t = downs_t + + self.strides_t = strides_t + + level_block = lambda level, down_t, stride_t: DecoderConvBock(output_emb_width, + output_emb_width, + down_t, stride_t, + **block_kwargs) + self.level_blocks = nn.ModuleList() + iterator = zip(list(range(self.levels)), downs_t, strides_t) + for level, down_t, stride_t in iterator: + self.level_blocks.append(level_block(level, down_t, stride_t)) + + self.out = nn.Conv1d(output_emb_width, input_emb_width, 3, 1, 1) + # self.out = nn.Conv1d(output_emb_width, input_emb_width, 3, 1, 1, padding_mode='replicate') + def forward(self, xs, all_levels=True): + if all_levels: + assert len(xs) == self.levels + else: + assert len(xs) == 1 + x = xs[-1] + N, T = x.shape[0], x.shape[-1] + emb = self.output_emb_width + assert_shape(x, (N, emb, T)) + + # 32, 64 ... + iterator = reversed(list(zip(list(range(self.levels)), self.downs_t, self.strides_t))) + for level, down_t, stride_t in iterator: + level_block = self.level_blocks[level] + x = level_block(x) + emb, T = self.output_emb_width, T * (stride_t ** down_t) + assert_shape(x, (N, emb, T)) + if level != 0 and all_levels: + x = x + xs[level - 1] + + x = self.out(x) + return x diff --git a/models/qp_vqvae/resnet.py b/models/qp_vqvae/resnet.py new file mode 100644 index 0000000000000000000000000000000000000000..a4c97ccaeddcaa1e4994d5a7b837dead8453338f --- /dev/null +++ b/models/qp_vqvae/resnet.py @@ -0,0 +1,78 @@ +import math +import torch.nn as nn +import utils.dist_adapter as dist +from utils.checkpoint import checkpoint + +class ResConvBlock(nn.Module): + def __init__(self, n_in, n_state): + super().__init__() + self.model = nn.Sequential( + nn.ReLU(), + nn.Conv2d(n_in, n_state, 3, 1, 1), + nn.ReLU(), + nn.Conv2d(n_state, n_in, 1, 1, 0), + ) + + def forward(self, x): + return x + self.model(x) + +class Resnet(nn.Module): + def __init__(self, n_in, n_depth, m_conv=1.0): + super().__init__() + self.model = nn.Sequential(*[ResConvBlock(n_in, int(m_conv * n_in)) for _ in range(n_depth)]) + + def forward(self, x): + return self.model(x) + +class ResConv1DBlock(nn.Module): + def __init__(self, n_in, n_state, dilation=1, zero_out=False, res_scale=1.0): + super().__init__() + padding = dilation + self.model = nn.Sequential( + nn.LeakyReLU(0.2), + #nn.ReLU(), + nn.Conv1d(n_in, n_state, 3, 1, padding, dilation), + # nn.Conv1d(n_in, n_state, 3, 1, padding, dilation, padding_mode='replicate'), + nn.LeakyReLU(0.2), + #nn.ReLU(), + nn.Conv1d(n_state, n_in, 1, 1, 0,), + # nn.Conv1d(n_state, n_in, 1, 1, 0, padding_mode='replicate'), + ) + if zero_out: + out = self.model[-1] + nn.init.zeros_(out.weight) + nn.init.zeros_(out.bias) + self.res_scale = res_scale + def forward(self, x): + return x + self.res_scale * self.model(x) + +class Resnet1D(nn.Module): + def __init__(self, n_in, n_depth, m_conv=1.0, dilation_growth_rate=1, dilation_cycle=None, zero_out=False, res_scale=False, reverse_dilation=False, checkpoint_res=False): + super().__init__() + def _get_depth(depth): + if dilation_cycle is None: + return depth + else: + return depth % dilation_cycle + blocks = [ResConv1DBlock(n_in, int(m_conv * n_in), + dilation=dilation_growth_rate ** _get_depth(depth), + zero_out=zero_out, + res_scale=1.0 if not res_scale else 1.0 / math.sqrt(n_depth)) + for depth in range(n_depth)] + if reverse_dilation: + blocks = blocks[::-1] + self.checkpoint_res = checkpoint_res + if self.checkpoint_res == 1: + if dist.get_rank() == 0: + print("Checkpointing convs") + self.blocks = nn.ModuleList(blocks) + else: + self.model = nn.Sequential(*blocks) + + def forward(self, x): + if self.checkpoint_res == 1: + for block in self.blocks: + x = checkpoint(block, (x, ), block.parameters(), True) + return x + else: + return self.model(x) diff --git a/models/qp_vqvae/simpleVqvae.py b/models/qp_vqvae/simpleVqvae.py new file mode 100644 index 0000000000000000000000000000000000000000..7ef61981ff212d14682601562f590c74b0640cc5 --- /dev/null +++ b/models/qp_vqvae/simpleVqvae.py @@ -0,0 +1,226 @@ +import pdb + +import torch +import torch.nn as nn +import torch.nn.functional as F + + +class ConvNorm(torch.nn.Module): + def __init__(self, in_channels, out_channels, kernel_size=1, stride=1, + padding=None, dilation=1, bias=True, w_init_gain='linear'): + super(ConvNorm, self).__init__() + if padding is None: + assert(kernel_size % 2 == 1) + padding = int(dilation * (kernel_size - 1) / 2) + + self.conv = torch.nn.Conv1d(in_channels, out_channels, + kernel_size=kernel_size, stride=stride, + padding=padding, dilation=dilation, + bias=bias) + + torch.nn.init.xavier_uniform_( + self.conv.weight, gain=torch.nn.init.calculate_gain(w_init_gain)) + + def forward(self, signal): + conv_signal = self.conv(signal) + return conv_signal + + +class LinearNorm(torch.nn.Module): + def __init__(self, in_dim, out_dim, bias=True, w_init_gain='linear'): + super(LinearNorm, self).__init__() + self.linear_layer = torch.nn.Linear(in_dim, out_dim, bias=bias) + + torch.nn.init.xavier_uniform_( + self.linear_layer.weight, + gain=torch.nn.init.calculate_gain(w_init_gain)) + + def forward(self, x): + return self.linear_layer(x) + + +class Decoder(nn.Module): + def __init__(self, dim_pre=64, dim_out=45): + super(Decoder, self).__init__() + convolutions = [] + for i in range(3): + conv_layer = nn.Sequential( + ConvNorm(dim_pre, + dim_pre, + kernel_size=5, stride=1, + padding=2, + dilation=1, w_init_gain='relu'), + nn.BatchNorm1d(dim_pre)) + convolutions.append(conv_layer) + self.convolutions = nn.ModuleList(convolutions) + self.linear_projection = LinearNorm(dim_pre, dim_out) + + def forward(self, z, target=None): + z = F.interpolate(z.transpose(1, 2), scale_factor=2) + for conv in self.convolutions: + z = F.relu(conv(z)) + z = z.transpose(1, 2) # (b, 240, 64) + decoder_output = self.linear_projection(z) + if target is None: + return decoder_output + else: + loss = F.l1_loss(decoder_output, target) + return loss, decoder_output + + +class Encoder(nn.Module): + ''' + reference from: https://github.com/bshall/VectorQuantizedCPC/blob/master/model.py + ''' + + def __init__(self, in_channels, channels, n_embeddings, z_dim, c_dim): + super(Encoder, self).__init__() + self.conv = nn.Conv1d(in_channels, channels, 4, 2, 1, bias=False) # T // 2 + # self.conv = nn.Conv1d(in_channels, channels, 3, 1, 1, bias=False) # T + self.encoder = nn.Sequential( + nn.LayerNorm(channels), + nn.ReLU(True), + nn.Linear(channels, channels, bias=False), + nn.LayerNorm(channels), + nn.ReLU(True), + nn.Linear(channels, channels, bias=False), + nn.LayerNorm(channels), + nn.ReLU(True), + nn.Linear(channels, channels, bias=False), + nn.LayerNorm(channels), + nn.ReLU(True), + nn.Linear(channels, channels, bias=False), + nn.LayerNorm(channels), + nn.ReLU(True), + nn.Linear(channels, z_dim), + ) + self.codebook = VQEmbeddingEMA(n_embeddings, z_dim) + self.rnn = nn.LSTM(z_dim, c_dim, batch_first=True) + + def encode(self, mel): + z = self.conv(mel) + z_beforeVQ = self.encoder(z.transpose(1, 2)) + z, r, indices = self.codebook.encode(z_beforeVQ) + c, _ = self.rnn(z) + return z, c, z_beforeVQ, indices + + def forward(self, mels): + z = self.conv(mels.float()) # (bz, 80, 128) -> (bz, 512, 128/2) + z_beforeVQ = self.encoder(z.transpose(1, 2)) # (bz, 512, 128/2) -> (bz, 128/2, 512) -> (bz, 128/2, 64) + z, r, loss, perplexity = self.codebook(z_beforeVQ) # z: (bz, 128/2, 64) + z, r, indices = self.codebook.encode(z_beforeVQ) + c, _ = self.rnn(z) # (64, 128/2, 64) -> (64, 128/2, 256) + return z, c, z_beforeVQ, loss, perplexity + + +class VQEmbeddingEMA(nn.Module): + ''' + reference from: https://github.com/bshall/VectorQuantizedCPC/blob/master/model.py + ''' + + def __init__(self, n_embeddings, embedding_dim, commitment_cost=2, decay=0.9999, epsilon=1e-7): + super(VQEmbeddingEMA, self).__init__() + self.commitment_cost = commitment_cost + self.decay = decay + self.epsilon = epsilon + + init_bound = 1 / 512 + embedding = torch.Tensor(n_embeddings, embedding_dim) + embedding.uniform_(-init_bound, init_bound) + self.register_buffer("embedding", embedding) # only change during forward + self.register_buffer("ema_count", torch.zeros(n_embeddings)) + self.register_buffer("ema_weight", self.embedding.clone()) + + def encode(self, x): + M, D = self.embedding.size() + x_flat = x.detach().reshape(-1, D) + + distances = torch.addmm(torch.sum(self.embedding ** 2, dim=1) + + torch.sum(x_flat ** 2, dim=1, keepdim=True), + x_flat, self.embedding.t(), + alpha=-2.0, beta=1.0) + + indices = torch.argmin(distances.float(), dim=-1) + quantized = F.embedding(indices, self.embedding) + quantized = quantized.view_as(x) + residual = x - quantized + return quantized, residual, indices.view(x.size(0), x.size(1)) + + def forward(self, x): + M, D = self.embedding.size() + x_flat = x.detach().reshape(-1, D) + + distances = torch.addmm(torch.sum(self.embedding ** 2, dim=1) + + torch.sum(x_flat ** 2, dim=1, keepdim=True), + x_flat, self.embedding.t(), + alpha=-2.0, beta=1.0) # calculate the distance between each ele in embedding and x + + indices = torch.argmin(distances.float(), dim=-1) + encodings = F.one_hot(indices, M).float() + quantized = F.embedding(indices, self.embedding) + quantized = quantized.view_as(x) + + if self.training: # EMA based codebook learning + self.ema_count = self.decay * self.ema_count + (1 - self.decay) * torch.sum(encodings, dim=0) + + n = torch.sum(self.ema_count) + self.ema_count = (self.ema_count + self.epsilon) / (n + M * self.epsilon) * n + + dw = torch.matmul(encodings.t(), x_flat) + self.ema_weight = self.decay * self.ema_weight + (1 - self.decay) * dw + + self.embedding = self.ema_weight / self.ema_count.unsqueeze(-1) + + e_latent_loss = F.mse_loss(x, quantized.detach()) + loss = self.commitment_cost * e_latent_loss + + residual = x - quantized + + quantized = x + (quantized - x).detach() + + avg_probs = torch.mean(encodings, dim=0) + perplexity = torch.exp(-torch.sum(avg_probs * torch.log(avg_probs + 1e-10))) + + return quantized, residual, loss, perplexity + + +class simpleVQVAE(nn.Module): + def __init__(self): + super(simpleVQVAE, self).__init__() + self.encoder = Encoder(in_channels=15 * 3, channels=512, n_embeddings=512, z_dim=64, c_dim=256) + self.decoder = Decoder(dim_pre=64, dim_out=45) + + def encode(self,x): + z, _, _, indices = self.encoder.encode(x.transpose(1, 2)) + return [indices] + + def forward(self, x): + z, c, z_beforeVQ, loss_vq, perplexity = self.encoder(x.transpose(1, 2)) + + loss_recon, output = self.decoder(z, x) + return output, loss_vq + loss_recon, perplexity + + +if __name__ == '__main__': + ''' + cd codebook/ + python -m models.simpleVqvae + ''' + # model = Encoder(in_channels=80, channels=512, n_embeddings=512, z_dim=64, c_dim=256) + # x = torch.rand(2, 80, 128) + # z, c, z_beforeVQ, loss, perplexity = model(x) + ''' + z: (2, 64, 64) + c: (2, 64, 256) + z_beforeVQ: (2, 64, 64) + loss + perplexity + ''' + + model = Encoder(in_channels=15 * 3, channels=512, n_embeddings=512, z_dim=64, c_dim=256) + model2 = Decoder(dim_pre=64, dim_out=45) + x = torch.rand(2, 240, 15 * 3) + z, c, z_beforeVQ, loss, perplexity = model(x.transpose(1, 2)) + pdb.set_trace() + loss, output = model2(z, x) + diff --git a/models/qp_vqvae/utils/checkpoint.py b/models/qp_vqvae/utils/checkpoint.py new file mode 100644 index 0000000000000000000000000000000000000000..cfbcccb34d415a50f6bc7ef86334215d54779c18 --- /dev/null +++ b/models/qp_vqvae/utils/checkpoint.py @@ -0,0 +1,32 @@ +# Simple gradient checkpointing. Works with distributed data parallel +import torch as t + +def checkpoint(func, inputs, params, flag): + if flag: + args = inputs + tuple(params) + return CheckpointFunction.apply(func, len(inputs), *args) + else: + return func(*inputs) + +class CheckpointFunction(t.autograd.Function): + @staticmethod + def forward(ctx, run_function, length, *args): + ctx.run_function = run_function + ctx.input_tensors = list(args[:length]) + ctx.input_params = list(args[length:]) + with t.no_grad(): + output_tensors = ctx.run_function(*ctx.input_tensors) + return output_tensors + + @staticmethod + def backward(ctx, *output_grads): + for i in range(len(ctx.input_tensors)): + temp = ctx.input_tensors[i] + ctx.input_tensors[i] = temp.detach() + ctx.input_tensors[i].requires_grad = temp.requires_grad + with t.enable_grad(): + output_tensors = ctx.run_function(*ctx.input_tensors) + input_grads = t.autograd.grad(output_tensors, ctx.input_tensors + ctx.input_params, output_grads, allow_unused=True) + del ctx.input_tensors + del output_tensors + return (None, None) + input_grads diff --git a/models/qp_vqvae/utils/dist_adapter.py b/models/qp_vqvae/utils/dist_adapter.py new file mode 100644 index 0000000000000000000000000000000000000000..b67af947b3ac4f7e37e843b37d0499fc1ea5e7ef --- /dev/null +++ b/models/qp_vqvae/utils/dist_adapter.py @@ -0,0 +1,86 @@ +import torch.distributed as dist +from enum import Enum + +class ReduceOp(Enum): + SUM = 0, + PRODUCT = 1, + MIN = 2, + MAX = 3 + + def ToDistOp(self): + return { + self.SUM: dist.ReduceOp.SUM, + self.PRODUCT: dist.ReduceOp.PRODUCT, + self.MIN: dist.ReduceOp.MIN, + self.MAX: dist.ReduceOp.MAX + }[self] + +def is_available(): + return dist.is_available() + +def get_rank(): + if is_available(): + return _get_rank() + else: + return 0 + +def get_world_size(): + if is_available(): + return _get_world_size() + else: + return 1 + +def barrier(): + if is_available(): + return _barrier() + #else: do nothing + +def all_gather(tensor_list, tensor): + if is_available(): + return _all_gather(tensor_list, tensor) + else: + tensor_list[0] = tensor + +def all_reduce(tensor, op=ReduceOp.SUM): + if is_available(): + return _all_reduce(tensor, op) + #else: do nothing + +def reduce(tensor, dst, op=ReduceOp.SUM): + if is_available(): + return _reduce(tensor, dst, op) + #else: do nothing + +def broadcast(tensor, src): + if is_available(): + return _broadcast(tensor, src) + #else: do nothing + +def init_process_group(backend, init_method): + if is_available(): + return _init_process_group(backend, init_method) + #else: do nothing + +def _get_rank(): + return dist.get_rank() + +def _barrier(): + return dist.barrier() + +def _get_world_size(): + return dist.get_world_size() + +def _all_gather(tensor_list, tensor): + return dist.all_gather(tensor_list, tensor) + +def _all_reduce(tensor, op): + return dist.all_reduce(tensor, op.ToDistOp()) + +def _reduce(tensor, dst, op): + return dist.reduce(tensor, dst, op.ToDistOp()) + +def _broadcast(tensor, src): + return dist.broadcast(tensor, src) + +def _init_process_group(backend, init_method): + return dist.init_process_group(backend, init_method) \ No newline at end of file diff --git a/models/qp_vqvae/utils/logger.py b/models/qp_vqvae/utils/logger.py new file mode 100644 index 0000000000000000000000000000000000000000..dcf902f3666ccc4b838392c14188400acc2ef8dd --- /dev/null +++ b/models/qp_vqvae/utils/logger.py @@ -0,0 +1,155 @@ +import torch as t +import torch.distributed as dist +from tqdm import tqdm +from datetime import date +import os +import sys + +import sys +sys.path.append(os.path.dirname(os.path.dirname(sys.path[0]))) +from utils.torch_utils import parse_args + + +args = parse_args() +mydevice = t.device('cuda:' + args.gpu) + +def def_tqdm(x): + return tqdm(x, leave=True, file=sys.stdout, bar_format="{n_fmt}/{total_fmt} [{elapsed}<{remaining}, {rate_fmt}{postfix}]") + +def get_range(x): + if dist.get_rank() == 0: + return def_tqdm(x) + else: + return x + +def init_logging(hps, local_rank, rank): + logdir = f"{hps.local_logdir}/{hps.name}" + if local_rank == 0: + if not os.path.exists(logdir): + os.makedirs(logdir) + with open(logdir + 'argv.txt', 'w') as f: + f.write(hps.argv + '\n') + print("Logging to", logdir) + logger = Logger(logdir, rank) + metrics = Metrics() + logger.add_text('hps', str(hps)) + return logger, metrics + +def get_name(hps): + name = "" + for key, value in hps.items(): + name += f"{key}_{value}_" + return name + +def average_metrics(_metrics): + metrics = {} + for _metric in _metrics: + for key, val in _metric.items(): + if key not in metrics: + metrics[key] = [] + metrics[key].append(val) + return {key: sum(vals)//len(vals) for key, vals in metrics.items()} + +class Metrics: + def __init__(self): + self.sum = {} + self.n = {} + + def update(self, tag, val, batch): + # v is average value over batch + # store total value and total batch, returns dist average + sum = t.tensor(val * batch).float().to(mydevice) + n = t.tensor(batch).float().to(mydevice) + dist.all_reduce(sum) + dist.all_reduce(n) + sum = sum.item() + n = n.item() + self.sum[tag] = self.sum.get(tag, 0.0) + sum + self.n[tag] = self.n.get(tag, 0.0) + n + return sum / n + + def avg(self, tag): + if tag in self.sum: + return self.sum[tag] / self.n[tag] + else: + return 0.0 + + def reset(self): + self.sum = {} + self.n = {} + +class Logger: + def __init__(self, logdir, rank): + if rank == 0: + from tensorboardX import SummaryWriter + self.sw = SummaryWriter(f"{logdir}/logs") + self.iters = 0 + self.rank = rank + self.works = [] + self.logdir = logdir + + def step(self): + self.iters += 1 + + def flush(self): + if self.rank == 0: + self.sw.flush() + + def add_text(self, tag, text): + if self.rank == 0: + self.sw.add_text(tag, text, self.iters) + + def add_audios(self, tag, auds, sample_rate=22050, max_len=None, max_log=8): + if self.rank == 0: + for i in range(min(len(auds), max_log)): + if max_len: + self.sw.add_audio(f"{i}/{tag}", auds[i][:max_len * sample_rate], self.iters, sample_rate) + else: + self.sw.add_audio(f"{i}/{tag}", auds[i], self.iters, sample_rate) + + def add_audio(self, tag, aud, sample_rate=22050): + if self.rank == 0: + self.sw.add_audio(tag, aud, self.iters, sample_rate) + + def add_images(self, tag, img, dataformats="NHWC"): + if self.rank == 0: + self.sw.add_images(tag, img, self.iters, dataformats=dataformats) + + def add_image(self, tag, img): + if self.rank == 0: + self.sw.add_image(tag, img, self.iters) + + def add_scalar(self, tag, val): + if self.rank == 0: + self.sw.add_scalar(tag, val, self.iters) + + def get_range(self, loader): + if self.rank == 0: + self.trange = def_tqdm(loader) + else: + self.trange = loader + return enumerate(self.trange) + + def close_range(self): + if self.rank == 0: + self.trange.close() + + def set_postfix(self, *args, **kwargs): + if self.rank == 0: + self.trange.set_postfix(*args, **kwargs) + + # For logging summaries of varies graph ops + def add_reduce_scalar(self, tag, layer, val): + if self.iters % 100 == 0: + with t.no_grad(): + val = val.float().norm()/float(val.numel()) + work = dist.reduce(val, 0, async_op=True) + self.works.append((tag, layer, val, work)) + + def finish_reduce(self): + for tag, layer, val, work in self.works: + work.wait() + if self.rank == 0: + val = val.item()/dist.get_world_size() + self.lw[layer].add_scalar(tag, val, self.iters) + self.works = [] diff --git a/models/qp_vqvae/utils/parse_args.py b/models/qp_vqvae/utils/parse_args.py new file mode 100644 index 0000000000000000000000000000000000000000..1a4bd75b26ec795fff46ced5766e5e8581b0b486 --- /dev/null +++ b/models/qp_vqvae/utils/parse_args.py @@ -0,0 +1 @@ +import configargparse diff --git a/models/qp_vqvae/utils/torch_utils.py b/models/qp_vqvae/utils/torch_utils.py new file mode 100644 index 0000000000000000000000000000000000000000..846a26faece198400833120f82f7aba5861ebe4c --- /dev/null +++ b/models/qp_vqvae/utils/torch_utils.py @@ -0,0 +1,50 @@ +import gc +import torch as t + +def freeze_model(model): + model.eval() + for params in model.parameters(): + params.requires_grad = False + + +def unfreeze_model(model): + model.train() + for params in model.parameters(): + params.requires_grad = True + +def zero_grad(model): + for p in model.parameters(): + if p.requires_grad and p.grad is not None: + p.grad = None + +def empty_cache(): + gc.collect() + t.cuda.empty_cache() + +def assert_shape(x, exp_shape): + assert x.shape == exp_shape, f"Expected {exp_shape} got {x.shape}" + +def count_parameters(model): + return sum(p.numel() for p in model.parameters() if p.requires_grad) + +def count_state(model): + return sum(s.numel() for s in model.state_dict().values()) + +import argparse + +def parse_args(): + parser = argparse.ArgumentParser(description='Codebook') + parser.add_argument('--config', default='./configs/codebook.yml') + parser.add_argument('--gpu', type=str, default='2') + parser.add_argument('--no_cuda', type=list, default=['2']) + parser.add_argument('--prefix', type=str, required=False, default='knn_pred_wavvq') + parser.add_argument('--save_path', type=str, required=False, default="./Speech2GestureMatching/output/") + parser.add_argument('--code_path', type=str, required=False) + parser.add_argument('--VQVAE_model_path', type=str, required=False) + parser.add_argument('--BEAT_path', type=str, default="../dataset/orig_BEAT/speakers/") + parser.add_argument('--save_dir', type=str, default="../dataset/BEAT") + parser.add_argument('--step', type=str, default="1") + parser.add_argument('--stage', type=str, default="train") + args = parser.parse_args() + return args + diff --git a/models/quantizer.py b/models/quantizer.py new file mode 100644 index 0000000000000000000000000000000000000000..896c973ea25513e27feccc564d85d5dd361a4dc5 --- /dev/null +++ b/models/quantizer.py @@ -0,0 +1,159 @@ +import torch +import torch.nn as nn +import torch.nn.functional as F + + +class Quantizer(nn.Module): + def __init__(self, n_e, e_dim, beta): + super(Quantizer, self).__init__() + + self.e_dim = e_dim + self.n_e = n_e + self.beta = beta + + self.embedding = nn.Embedding(self.n_e, self.e_dim) + self.embedding.weight.data.uniform_(-1.0 / self.n_e, 1.0 / self.n_e) + + def forward(self, z): + """ + Inputs the output of the encoder network z and maps it to a discrete + one-hot vectort that is the index of the closest embedding vector e_j + z (continuous) -> z_q (discrete) + :param z (B, seq_len, channel): + :return z_q: + """ + assert z.shape[-1] == self.e_dim + z_flattened = z.contiguous().view(-1, self.e_dim) + + # B x V + d = torch.sum(z_flattened ** 2, dim=1, keepdim=True) + \ + torch.sum(self.embedding.weight**2, dim=1) - 2 * \ + torch.matmul(z_flattened, self.embedding.weight.t()) + # B x 1 + min_encoding_indices = torch.argmin(d, dim=1) + z_q = self.embedding(min_encoding_indices).view(z.shape) + + # compute loss for embedding + loss = torch.mean((z_q - z.detach())**2) + self.beta * \ + torch.mean((z_q.detach() - z)**2) + + # preserve gradients + z_q = z + (z_q - z).detach() + + min_encodings = F.one_hot(min_encoding_indices, self.n_e).type(z.dtype) + e_mean = torch.mean(min_encodings, dim=0) + perplexity = torch.exp(-torch.sum(e_mean*torch.log(e_mean + 1e-10))) + return loss, z_q, min_encoding_indices, perplexity + + def map2index(self, z): + """ + Inputs the output of the encoder network z and maps it to a discrete + one-hot vectort that is the index of the closest embedding vector e_j + z (continuous) -> z_q (discrete) + :param z (B, seq_len, channel): + :return z_q: + """ + assert z.shape[-1] == self.e_dim + #print(z.shape) + z_flattened = z.contiguous().view(-1, self.e_dim) + #print(z_flattened.shape) + + # B x V + d = torch.sum(z_flattened ** 2, dim=1, keepdim=True) + \ + torch.sum(self.embedding.weight**2, dim=1) - 2 * \ + torch.matmul(z_flattened, self.embedding.weight.t()) + # B x 1 + min_encoding_indices = torch.argmin(d, dim=1) + return min_encoding_indices.reshape(z.shape[0], -1) + + def get_codebook_entry(self, indices): + """ + + :param indices(B, seq_len): + :return z_q(B, seq_len, e_dim): + """ + index_flattened = indices.view(-1) + z_q = self.embedding(index_flattened) + z_q = z_q.view(indices.shape + (self.e_dim, )).contiguous() + return z_q + + +class EmbeddingEMA(nn.Module): + def __init__(self, num_tokens, codebook_dim, decay=0.99, eps=1e-5): + super(EmbeddingEMA, self).__init__() + self.decay = decay + self.eps = eps + weight = torch.randn(num_tokens, codebook_dim) + self.weight = nn.Parameter(weight, requires_grad=False) + self.cluster_size = nn.Parameter(torch.zeros(num_tokens), requires_grad=False) + self.embed_avg = nn.Parameter(weight.clone(), requires_grad=False) + self.update = True + + def forward(self, embed_id): + return F.embedding(embed_id, self.weight) + + def cluster_size_ema_update(self, new_cluster_size): + self.cluster_size.data.mul_(self.decay).add_(new_cluster_size, alpha=1 - self.decay) + + def embed_avg_ema_update(self, new_emb_avg): + self.embed_avg.data.mul_(self.decay).add(new_emb_avg, alpha=1 - self.decay) + + def weight_update(self, num_tokens): + n = self.cluster_size.sum() + smoothed_cluster_size = ( + (self.cluster_size + self.eps) / (n + num_tokens*self.eps) * n + ) + embed_normalized = self.embed_avg / smoothed_cluster_size.unsqueeze(1) + self.weight.data.copy_(embed_normalized) + + +class EMAVectorQuantizer(nn.Module): + def __init__(self, n_embed, embedding_dim, beta, decay=0.99, eps=1e-5): + super(EMAVectorQuantizer, self).__init__() + self.codebook_dim = embedding_dim + self.num_tokens = n_embed + self.beta = beta + self.embedding = EmbeddingEMA(self.num_tokens, self.codebook_dim, decay, eps) + + def forward(self, z): + z_flattened = z.view(-1, self.codebook_dim) + + d = torch.sum(z_flattened ** 2, dim=1, keepdim=True) + \ + torch.sum(self.embedding.weight ** 2, dim=1) - 2 * \ + torch.matmul(z_flattened, self.embedding.weight.t()) + + min_encoding_indices = torch.argmin(d, dim=1) + z_q = self.embedding(min_encoding_indices).view(z.shape) + + min_encodings = F.one_hot(min_encoding_indices, self.num_tokens).type(z.dtype) + e_mean = torch.mean(min_encodings, dim=0) + perplexity = torch.exp(-torch.sum(e_mean * torch.log(e_mean + 1e-10))) + + if self.training and self.embedding.update: + encoding_sum = min_encodings.sum(0) + embed_sum = min_encodings.transpose(0, 1)@z_flattened + + self.embedding.cluster_size_ema_update(encoding_sum) + self.embedding.embed_avg_ema_update(embed_sum) + self.embedding.weight_update(self.num_tokens) + + loss = self.beta * F.mse_loss(z_q.detach(), z) + + z_q = z + (z_q - z).detach() + return loss, z_q, min_encoding_indices, perplexity + + +# class GumbelQuantizer(nn.Module): +# def __init__(self, num_hiddens, embedding_dim, n_embed, straight_through=True, +# kl_weight=5e-4, temp_init=1.0): +# super(GumbelQuantizer, self).__init__() +# +# self.embedding_dim = embedding_dim +# self.n_embed = n_embed +# +# self.straight_through = straight_through +# self.temperature = temp_init +# self.kl_weight = kl_weight +# +# self.proj = nn.Linear(num_hiddens, n_embed) +# self.embed = nn.Embedding(n_embed, embedding_dim) diff --git a/models/timm_transformer/__init__.py b/models/timm_transformer/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..6bcdd2c8926fc01175bc0ca94e1b8bdd20186f84 --- /dev/null +++ b/models/timm_transformer/__init__.py @@ -0,0 +1 @@ +#from .config import use_fused_attn \ No newline at end of file diff --git a/models/timm_transformer/config.py b/models/timm_transformer/config.py new file mode 100644 index 0000000000000000000000000000000000000000..47d5d0a341f8968e801c803c6f439370b5511e04 --- /dev/null +++ b/models/timm_transformer/config.py @@ -0,0 +1,149 @@ +""" Model / Layer Config singleton state +""" +import os +import warnings +from typing import Any, Optional + +import torch + +__all__ = [ + 'is_exportable', 'is_scriptable', 'is_no_jit', 'use_fused_attn', + 'set_exportable', 'set_scriptable', 'set_no_jit', 'set_layer_config', 'set_fused_attn' +] + +# Set to True if prefer to have layers with no jit optimization (includes activations) +_NO_JIT = False + +# Set to True if prefer to have activation layers with no jit optimization +# NOTE not currently used as no difference between no_jit and no_activation jit as only layers obeying +# the jit flags so far are activations. This will change as more layers are updated and/or added. +_NO_ACTIVATION_JIT = False + +# Set to True if exporting a model with Same padding via ONNX +_EXPORTABLE = False + +# Set to True if wanting to use torch.jit.script on a model +_SCRIPTABLE = False + + +# use torch.scaled_dot_product_attention where possible +_HAS_FUSED_ATTN = hasattr(torch.nn.functional, 'scaled_dot_product_attention') +if 'TIMM_FUSED_ATTN' in os.environ: + _USE_FUSED_ATTN = int(os.environ['TIMM_FUSED_ATTN']) +else: + _USE_FUSED_ATTN = 1 # 0 == off, 1 == on (for tested use), 2 == on (for experimental use) + + +def is_no_jit(): + return _NO_JIT + + +class set_no_jit: + def __init__(self, mode: bool) -> None: + global _NO_JIT + self.prev = _NO_JIT + _NO_JIT = mode + + def __enter__(self) -> None: + pass + + def __exit__(self, *args: Any) -> bool: + global _NO_JIT + _NO_JIT = self.prev + return False + + +def is_exportable(): + return _EXPORTABLE + + +class set_exportable: + def __init__(self, mode: bool) -> None: + global _EXPORTABLE + self.prev = _EXPORTABLE + _EXPORTABLE = mode + + def __enter__(self) -> None: + pass + + def __exit__(self, *args: Any) -> bool: + global _EXPORTABLE + _EXPORTABLE = self.prev + return False + + +def is_scriptable(): + return _SCRIPTABLE + + +class set_scriptable: + def __init__(self, mode: bool) -> None: + global _SCRIPTABLE + self.prev = _SCRIPTABLE + _SCRIPTABLE = mode + + def __enter__(self) -> None: + pass + + def __exit__(self, *args: Any) -> bool: + global _SCRIPTABLE + _SCRIPTABLE = self.prev + return False + + +class set_layer_config: + """ Layer config context manager that allows setting all layer config flags at once. + If a flag arg is None, it will not change the current value. + """ + def __init__( + self, + scriptable: Optional[bool] = None, + exportable: Optional[bool] = None, + no_jit: Optional[bool] = None, + no_activation_jit: Optional[bool] = None): + global _SCRIPTABLE + global _EXPORTABLE + global _NO_JIT + global _NO_ACTIVATION_JIT + self.prev = _SCRIPTABLE, _EXPORTABLE, _NO_JIT, _NO_ACTIVATION_JIT + if scriptable is not None: + _SCRIPTABLE = scriptable + if exportable is not None: + _EXPORTABLE = exportable + if no_jit is not None: + _NO_JIT = no_jit + if no_activation_jit is not None: + _NO_ACTIVATION_JIT = no_activation_jit + + def __enter__(self) -> None: + pass + + def __exit__(self, *args: Any) -> bool: + global _SCRIPTABLE + global _EXPORTABLE + global _NO_JIT + global _NO_ACTIVATION_JIT + _SCRIPTABLE, _EXPORTABLE, _NO_JIT, _NO_ACTIVATION_JIT = self.prev + return False + + +def use_fused_attn(experimental: bool = False) -> bool: + # NOTE: ONNX export cannot handle F.scaled_dot_product_attention as of pytorch 2.0 + if not _HAS_FUSED_ATTN or _EXPORTABLE: + return False + if experimental: + return _USE_FUSED_ATTN > 1 + return _USE_FUSED_ATTN > 0 + + +def set_fused_attn(enable: bool = True, experimental: bool = False): + global _USE_FUSED_ATTN + if not _HAS_FUSED_ATTN: + warnings.warn('This version of pytorch does not have F.scaled_dot_product_attention, fused_attn flag ignored.') + return + if experimental and enable: + _USE_FUSED_ATTN = 2 + elif enable: + _USE_FUSED_ATTN = 1 + else: + _USE_FUSED_ATTN = 0 diff --git a/models/timm_transformer/helpers.py b/models/timm_transformer/helpers.py new file mode 100644 index 0000000000000000000000000000000000000000..94cf2ece48a6e5e55c74b70b60c899f26af345c6 --- /dev/null +++ b/models/timm_transformer/helpers.py @@ -0,0 +1,17 @@ +from itertools import repeat +import collections.abc + +# From PyTorch internals +def _ntuple(n): + def parse(x): + if isinstance(x, collections.abc.Iterable) and not isinstance(x, str): + return tuple(x) + return tuple(repeat(x, n)) + return parse + + +to_1tuple = _ntuple(1) +to_2tuple = _ntuple(2) +to_3tuple = _ntuple(3) +to_4tuple = _ntuple(4) +to_ntuple = _ntuple diff --git a/models/timm_transformer/layers.py b/models/timm_transformer/layers.py new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/models/timm_transformer/transformer.py b/models/timm_transformer/transformer.py new file mode 100644 index 0000000000000000000000000000000000000000..ea4c6caa555ed0e78ce7becb7aa6f018d3a5acfc --- /dev/null +++ b/models/timm_transformer/transformer.py @@ -0,0 +1,199 @@ +import logging +import math +from collections import OrderedDict +from functools import partial +from typing import Callable, List, Optional, Sequence, Tuple, Union + +import torch +import torch.nn as nn +import torch.nn.functional as F +import torch.utils.checkpoint +from torch.jit import Final + +from .config import use_fused_attn +from .helpers import to_2tuple +__all__ = ['VisionTransformer'] # model_registry will add each entrypoint fn to this + + +_logger = logging.getLogger(__name__) + + +def drop_path(x, drop_prob: float = 0., training: bool = False, scale_by_keep: bool = True): + """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks). + + This is the same as the DropConnect impl I created for EfficientNet, etc networks, however, + the original name is misleading as 'Drop Connect' is a different form of dropout in a separate paper... + See discussion: https://github.com/tensorflow/tpu/issues/494#issuecomment-532968956 ... I've opted for + changing the layer and argument names to 'drop path' rather than mix DropConnect as a layer name and use + 'survival rate' as the argument. + + """ + if drop_prob == 0. or not training: + return x + keep_prob = 1 - drop_prob + shape = (x.shape[0],) + (1,) * (x.ndim - 1) # work with diff dim tensors, not just 2D ConvNets + random_tensor = x.new_empty(shape).bernoulli_(keep_prob) + if keep_prob > 0.0 and scale_by_keep: + random_tensor.div_(keep_prob) + return x * random_tensor + + + +class DropPath(nn.Module): + """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks). + """ + def __init__(self, drop_prob: float = 0., scale_by_keep: bool = True): + super(DropPath, self).__init__() + self.drop_prob = drop_prob + self.scale_by_keep = scale_by_keep + + def forward(self, x): + return drop_path(x, self.drop_prob, self.training, self.scale_by_keep) + + def extra_repr(self): + return f'drop_prob={round(self.drop_prob,3):0.3f}' + +class Attention(nn.Module): + fused_attn: Final[bool] + + def __init__( + self, + dim, + num_heads=8, + qkv_bias=False, + qk_norm=False, + attn_drop=0., + proj_drop=0., + norm_layer=nn.LayerNorm, + ): + super().__init__() + assert dim % num_heads == 0, 'dim should be divisible by num_heads' + self.num_heads = num_heads + self.head_dim = dim // num_heads + self.scale = self.head_dim ** -0.5 + self.fused_attn = use_fused_attn() + + self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias) + self.q_norm = norm_layer(self.head_dim) if qk_norm else nn.Identity() + self.k_norm = norm_layer(self.head_dim) if qk_norm else nn.Identity() + self.attn_drop = nn.Dropout(attn_drop) + self.proj = nn.Linear(dim, dim) + self.proj_drop = nn.Dropout(proj_drop) + + def forward(self, x): + B, N, C = x.shape + qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, self.head_dim).permute(2, 0, 3, 1, 4) + q, k, v = qkv.unbind(0) + q, k = self.q_norm(q), self.k_norm(k) + + if self.fused_attn: + x = F.scaled_dot_product_attention( + q, k, v, + dropout_p=self.attn_drop.p, + ) + else: + q = q * self.scale + attn = q @ k.transpose(-2, -1) + attn = attn.softmax(dim=-1) + attn = self.attn_drop(attn) + x = attn @ v + + x = x.transpose(1, 2).reshape(B, N, C) + x = self.proj(x) + x = self.proj_drop(x) + return x + + +class LayerScale(nn.Module): + def __init__(self, dim, init_values=1e-5, inplace=False): + super().__init__() + self.inplace = inplace + self.gamma = nn.Parameter(init_values * torch.ones(dim)) + + def forward(self, x): + return x.mul_(self.gamma) if self.inplace else x * self.gamma + + +class Mlp(nn.Module): + """ MLP as used in Vision Transformer, MLP-Mixer and related networks + """ + def __init__( + self, + in_features, + hidden_features=None, + out_features=None, + act_layer=nn.GELU, + norm_layer=None, + bias=True, + drop=0., + use_conv=False, + ): + super().__init__() + out_features = out_features or in_features + hidden_features = hidden_features or in_features + bias = to_2tuple(bias) + drop_probs = to_2tuple(drop) + linear_layer = partial(nn.Conv2d, kernel_size=1) if use_conv else nn.Linear + + self.fc1 = linear_layer(in_features, hidden_features, bias=bias[0]) + self.act = act_layer() + self.drop1 = nn.Dropout(drop_probs[0]) + self.norm = norm_layer(hidden_features) if norm_layer is not None else nn.Identity() + self.fc2 = linear_layer(hidden_features, out_features, bias=bias[1]) + self.drop2 = nn.Dropout(drop_probs[1]) + + def forward(self, x): + x = self.fc1(x) + x = self.act(x) + x = self.drop1(x) + x = self.fc2(x) + x = self.drop2(x) + return x + + +class Block(nn.Module): + + def __init__( + self, + dim, + num_heads, + mlp_ratio=4., + qkv_bias=False, + qk_norm=False, + proj_drop=0., + attn_drop=0., + init_values=None, + drop_path=0., + act_layer=nn.GELU, + norm_layer=nn.LayerNorm, + mlp_layer=Mlp, + ): + super().__init__() + self.norm1 = norm_layer(dim) + self.attn = Attention( + dim, + num_heads=num_heads, + qkv_bias=qkv_bias, + qk_norm=qk_norm, + attn_drop=attn_drop, + proj_drop=proj_drop, + norm_layer=norm_layer, + ) + self.ls1 = LayerScale(dim, init_values=init_values) if init_values else nn.Identity() + self.drop_path1 = DropPath(drop_path) if drop_path > 0. else nn.Identity() + + self.norm2 = norm_layer(dim) + self.mlp = mlp_layer( + in_features=dim, + hidden_features=int(dim * mlp_ratio), + act_layer=act_layer, + drop=proj_drop, + ) + self.ls2 = LayerScale(dim, init_values=init_values) if init_values else nn.Identity() + self.drop_path2 = DropPath(drop_path) if drop_path > 0. else nn.Identity() + + def forward(self, x): + x = x + self.drop_path1(self.ls1(self.attn(self.norm1(x)))) + x = x + self.drop_path2(self.ls2(self.mlp(self.norm2(x)))) + return x + diff --git a/models/utils/__init__.py b/models/utils/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/models/utils/audio_utils.py b/models/utils/audio_utils.py new file mode 100644 index 0000000000000000000000000000000000000000..39f428af596b2be187a78cc2abf36c458d35ccba --- /dev/null +++ b/models/utils/audio_utils.py @@ -0,0 +1,148 @@ +import numpy as np +import torch as t +import models.utils.dist_adapter as dist +import soundfile +import librosa +from models.utils.dist_utils import print_once + +class DefaultSTFTValues: + def __init__(self, hps): + self.sr = hps.sr + self.n_fft = 2048 + self.hop_length = 256 + self.window_size = 6 * self.hop_length + +class STFTValues: + def __init__(self, hps, n_fft, hop_length, window_size): + self.sr = hps.sr + self.n_fft = n_fft + self.hop_length = hop_length + self.window_size = window_size + +def calculate_bandwidth(dataset, hps, duration=600): + hps = DefaultSTFTValues(hps) + n_samples = int(dataset.sr * duration) + l1, total, total_sq, n_seen, idx = 0.0, 0.0, 0.0, 0.0, dist.get_rank() + spec_norm_total, spec_nelem = 0.0, 0.0 + while n_seen < n_samples: + x = dataset[idx] + if isinstance(x, (tuple, list)): + x, y = x + samples = x.astype(np.float64) + stft = librosa.core.stft(np.mean(samples, axis=1), hps.n_fft, hop_length=hps.hop_length, win_length=hps.window_size) + spec = np.absolute(stft) + spec_norm_total += np.linalg.norm(spec) + spec_nelem += 1 + n_seen += int(np.prod(samples.shape)) + l1 += np.sum(np.abs(samples)) + total += np.sum(samples) + total_sq += np.sum(samples ** 2) + idx += max(16, dist.get_world_size()) + + if dist.is_available(): + from jukebox.utils.dist_utils import allreduce + n_seen = allreduce(n_seen) + total = allreduce(total) + total_sq = allreduce(total_sq) + l1 = allreduce(l1) + spec_nelem = allreduce(spec_nelem) + spec_norm_total = allreduce(spec_norm_total) + + mean = total / n_seen + bandwidth = dict(l2 = total_sq / n_seen - mean ** 2, + l1 = l1 / n_seen, + spec = spec_norm_total / spec_nelem) + print_once(bandwidth) + return bandwidth + +def audio_preprocess(x, hps): + # Extra layer in case we want to experiment with different preprocessing + # For two channel, blend randomly into mono (standard is .5 left, .5 right) + + # x: NTC + # x = x.float() + # if x.shape[-1]==2: + # if hps.aug_blend: + # mix=t.rand((x.shape[0],1), device=x.device) #np.random.rand() + # else: + # mix = 0.5 + # x=(mix*x[:,:,0]+(1-mix)*x[:,:,1]) + # elif x.shape[-1]==1: + # x=x[:,:,0] + # else: + # assert False, f'Expected channels {hps.channels}. Got unknown {x.shape[-1]} channels' + + # # x: NT -> NTC + # x = x.unsqueeze(2) + return x + +def audio_postprocess(x, hps): + return x + +def stft(sig, hps): + return t.stft(sig, hps.n_fft, hps.hop_length, win_length=hps.window_size, window=t.hann_window(hps.window_size, device=sig.device)) + +def spec(x, hps): + return t.norm(stft(x, hps), p=2, dim=-1) + +def norm(x): + return (x.view(x.shape[0], -1) ** 2).sum(dim=-1).sqrt() + +def squeeze(x): + if len(x.shape) == 3: + assert x.shape[-1] in [1,2] + x = t.mean(x, -1) + if len(x.shape) != 2: + raise ValueError(f'Unknown input shape {x.shape}') + return x + +def spectral_loss(x_in, x_out, hps): + hps = DefaultSTFTValues(hps) + spec_in = spec(squeeze(x_in.float()), hps) + spec_out = spec(squeeze(x_out.float()), hps) + return norm(spec_in - spec_out) + +def multispectral_loss(x_in, x_out, hps): + losses = [] + assert len(hps.multispec_loss_n_fft) == len(hps.multispec_loss_hop_length) == len(hps.multispec_loss_window_size) + args = [hps.multispec_loss_n_fft, + hps.multispec_loss_hop_length, + hps.multispec_loss_window_size] + for n_fft, hop_length, window_size in zip(*args): + hps = STFTValues(hps, n_fft, hop_length, window_size) + spec_in = spec(squeeze(x_in.float()), hps) + spec_out = spec(squeeze(x_out.float()), hps) + losses.append(norm(spec_in - spec_out)) + return sum(losses) / len(losses) + +def spectral_convergence(x_in, x_out, hps, epsilon=2e-3): + hps = DefaultSTFTValues(hps) + spec_in = spec(squeeze(x_in.float()), hps) + spec_out = spec(squeeze(x_out.float()), hps) + + gt_norm = norm(spec_in) + residual_norm = norm(spec_in - spec_out) + mask = (gt_norm > epsilon).float() + return (residual_norm * mask) / t.clamp(gt_norm, min=epsilon) + +def log_magnitude_loss(x_in, x_out, hps, epsilon=1e-4): + hps = DefaultSTFTValues(hps) + spec_in = t.log(spec(squeeze(x_in.float()), hps) + epsilon) + spec_out = t.log(spec(squeeze(x_out.float()), hps) + epsilon) + return t.mean(t.abs(spec_in - spec_out)) + +def load_audio(file, sr, offset, duration, mono=False): + # Librosa loads more filetypes than soundfile + x, _ = librosa.load(file, sr=sr, mono=mono, offset=offset/sr, duration=duration/sr) + if len(x.shape) == 1: + x = x.reshape((1, -1)) + return x + + +def save_wav(fname, aud, sr): + # clip before saving? + aud = t.clamp(aud, -1, 1).cpu().numpy() + for i in list(range(aud.shape[0])): + soundfile.write(f'{fname}/item_{i}.wav', aud[i], samplerate=sr, format='wav') + + diff --git a/models/utils/build_vocab.py b/models/utils/build_vocab.py new file mode 100644 index 0000000000000000000000000000000000000000..79b623ff52c05d0213a7fd24094edf0a7b3c9334 --- /dev/null +++ b/models/utils/build_vocab.py @@ -0,0 +1,144 @@ +import numpy as np +import glob +import os +import pickle +import lmdb +import fasttext +from loguru import logger +from scipy import linalg + + +class Vocab: + PAD_token = 0 + SOS_token = 1 + EOS_token = 2 + UNK_token = 3 + + def __init__(self, name, insert_default_tokens=True): + self.name = name + self.trimmed = False + self.word_embedding_weights = None + self.reset_dictionary(insert_default_tokens) + + def reset_dictionary(self, insert_default_tokens=True): + self.word2index = {} + self.word2count = {} + if insert_default_tokens: + self.index2word = {self.PAD_token: "", self.SOS_token: "", + self.EOS_token: "", self.UNK_token: ""} + else: + self.index2word = {self.UNK_token: ""} + self.n_words = len(self.index2word) # count default tokens + + def index_word(self, word): + if word not in self.word2index: + self.word2index[word] = self.n_words + self.word2count[word] = 1 + self.index2word[self.n_words] = word + self.n_words += 1 + else: + self.word2count[word] += 1 + + def add_vocab(self, other_vocab): + for word, _ in other_vocab.word2count.items(): + self.index_word(word) + + # remove words below a certain count threshold + def trim(self, min_count): + if self.trimmed: + return + self.trimmed = True + + keep_words = [] + + for k, v in self.word2count.items(): + if v >= min_count: + keep_words.append(k) + + print(' word trimming, kept %s / %s = %.4f' % ( + len(keep_words), len(self.word2index), len(keep_words) / len(self.word2index) + )) + + # reinitialize dictionary + self.reset_dictionary() + for word in keep_words: + self.index_word(word) + + def get_word_index(self, word): + if word in self.word2index: + return self.word2index[word] + else: + return self.UNK_token + + def load_word_vectors(self, pretrained_path, embedding_dim=300): + print(" loading word vectors from '{}'...".format(pretrained_path)) + + # initialize embeddings to random values for special words + init_sd = 1 / np.sqrt(embedding_dim) + weights = np.random.normal(0, scale=init_sd, size=[self.n_words, embedding_dim]) + weights = weights.astype(np.float32) + + # read word vectors + word_model = fasttext.load_model(pretrained_path) + for word, id in self.word2index.items(): + vec = word_model.get_word_vector(word) + weights[id] = vec + self.word_embedding_weights = weights + +def build_vocab(name, data_path, cache_path, word_vec_path=None, feat_dim=None): + print(' building a language model...') + lang_model = Vocab(name) + print(' indexing words from {}'.format(data_path)) + index_words_from_textgrid(lang_model, data_path) + + if word_vec_path is not None: + lang_model.load_word_vectors(word_vec_path, feat_dim) + else: + print(' loaded from {}'.format(cache_path)) + with open(cache_path, 'rb') as f: + lang_model = pickle.load(f) + if word_vec_path is None: + lang_model.word_embedding_weights = None + elif lang_model.word_embedding_weights.shape[0] != lang_model.n_words: + logging.warning(' failed to load word embedding weights. check this') + assert False + + with open(cache_path, 'wb') as f: + pickle.dump(lang_model, f) + + return lang_model + +def index_words(lang_model, data_path): + #index words form text + with open(data_path, "r") as f: + for line in f.readlines(): + line = line.replace(",", " ") + line = line.replace(".", " ") + line = line.replace("?", " ") + line = line.replace("!", " ") + for word in line.split(): + lang_model.index_word(word) + print(' indexed %d words' % lang_model.n_words) + +def index_words_from_textgrid(lang_model, data_path): + import textgrid as tg + trainvaltest=os.listdir(data_path) + for loadtype in trainvaltest: + if "." in loadtype: continue #ignore .ipynb_checkpoints + texts = os.listdir(data_path+loadtype+"/text/") + for textfile in texts: + tgrid = tg.TextGrid.fromFile(data_path+loadtype+"/text/"+textfile) + for word in tgrid[0]: + word_n, word_s, word_e = word.mark, word.minTime, word.maxTime + word_n = word_n.replace(",", " ") + word_n = word_n.replace(".", " ") + word_n = word_n.replace("?", " ") + word_n = word_n.replace("!", " ") + #print(word_n) + lang_model.index_word(word_n) + print(' indexed %d words' % lang_model.n_words) + +if __name__ == "__main__": + #11195 for all, 5793 for 4 speakers + build_vocab("beat_english_15_141", "/home/ma-user/work/datasets/beat_cache/beat_english_15_141/", "/home/ma-user/work/datasets/beat_cache/beat_english_15_141/vocab.pkl", "/home/ma-user/work/datasets/cc.en.300.bin", 300) + \ No newline at end of file diff --git a/models/utils/fk.py b/models/utils/fk.py new file mode 100644 index 0000000000000000000000000000000000000000..d6ae32341c1ccab559772e053fecf6ded608dc1f --- /dev/null +++ b/models/utils/fk.py @@ -0,0 +1,149 @@ +"""Based on Daniel Holden code from: + A Deep Learning Framework for Character Motion Synthesis and Editing + (http://www.ipab.inf.ed.ac.uk/cgvu/motionsynthesis.pdf) +""" + +import os + +import numpy as np +import torch +import torch.nn as nn +from .rotations import euler_angles_to_matrix, quaternion_to_matrix, rotation_6d_to_matrix + + +class ForwardKinematicsLayer(nn.Module): + """ Forward Kinematics Layer Class """ + + def __init__(self, args=None, parents=None, positions=None, device=None): + super().__init__() + self.b_idxs = None + if device is None: + self.device = torch.device('cuda' if torch.cuda.is_available() else 'cpu') + else: + self.device = device + if parents is None and positions is None: + # Load SMPL skeleton (their joint order is different from the one we use for bvh export) + smpl_fname = os.path.join(args.smpl.smpl_body_model, args.data.gender, 'model.npz') + smpl_data = np.load(smpl_fname, encoding='latin1') + self.parents = torch.from_numpy(smpl_data['kintree_table'][0].astype(np.int32)).to(self.device) + self.parents = self.parents.long() + self.positions = torch.from_numpy(smpl_data['J'].astype(np.float32)).to(self.device) + self.positions[1:] -= self.positions[self.parents[1:]] + else: + self.parents = torch.from_numpy(parents).to(self.device) + self.parents = self.parents.long() + self.positions = torch.from_numpy(positions).to(self.device) + self.positions = self.positions.float() + self.positions[0] = 0 + + def rotate(self, t0s, t1s): + return torch.matmul(t0s, t1s) + + def identity_rotation(self, rotations): + diagonal = torch.diag(torch.tensor([1.0, 1.0, 1.0, 1.0])).to(self.device) + diagonal = torch.reshape( + diagonal, torch.Size([1] * len(rotations.shape[:2]) + [4, 4])) + ts = diagonal.repeat(rotations.shape[:2] + torch.Size([1, 1])) + return ts + + def make_fast_rotation_matrices(self, positions, rotations): + if len(rotations.shape) == 4 and rotations.shape[-2:] == torch.Size([3, 3]): + rot_matrices = rotations + elif rotations.shape[-1] == 3: + rot_matrices = euler_angles_to_matrix(rotations, convention='XYZ') + elif rotations.shape[-1] == 4: + rot_matrices = quaternion_to_matrix(rotations) + elif rotations.shape[-1] == 6: + rot_matrices = rotation_6d_to_matrix(rotations) + else: + raise NotImplementedError(f'Unimplemented rotation representation in FK layer, shape of {rotations.shape}') + + rot_matrices = torch.cat([rot_matrices, positions[..., None]], dim=-1) + zeros = torch.zeros(rot_matrices.shape[:-2] + torch.Size([1, 3])).to(self.device) + ones = torch.ones(rot_matrices.shape[:-2] + torch.Size([1, 1])).to(self.device) + zerosones = torch.cat([zeros, ones], dim=-1) + rot_matrices = torch.cat([rot_matrices, zerosones], dim=-2) + return rot_matrices + + def rotate_global(self, parents, positions, rotations): + locals = self.make_fast_rotation_matrices(positions, rotations) + globals = self.identity_rotation(rotations) + + globals = torch.cat([locals[:, 0:1], globals[:, 1:]], dim=1) + b_size = positions.shape[0] + if self.b_idxs is None: + self.b_idxs = torch.LongTensor(np.arange(b_size)).to(self.device) + elif self.b_idxs.shape[-1] != b_size: + self.b_idxs = torch.LongTensor(np.arange(b_size)).to(self.device) + + for i in range(1, positions.shape[1]): + globals[:, i] = self.rotate( + globals[self.b_idxs, parents[i]], locals[:, i]) + + return globals + + def get_tpose_joints(self, offsets, parents): + num_joints = len(parents) + joints = [offsets[:, 0]] + for j in range(1, len(parents)): + joints.append(joints[parents[j]] + offsets[:, j]) + + return torch.stack(joints, dim=1) + + def canonical_to_local(self, canonical_xform, global_orient=None): + """ + Args: + canonical_xform: (B, J, 3, 3) + global_orient: (B, 3, 3) + + Returns: + local_xform: (B, J, 3, 3) + """ + local_xform = torch.zeros_like(canonical_xform) + + if global_orient is None: + global_xform = canonical_xform + else: + global_xform = torch.matmul(global_orient.unsqueeze(1), canonical_xform) + for i in range(global_xform.shape[1]): + if i == 0: + local_xform[:, i] = global_xform[:, i] + else: + local_xform[:, i] = torch.bmm(torch.linalg.inv(global_xform[:, self.parents[i]]), global_xform[:, i]) + + return local_xform + + def global_to_local(self, global_xform): + """ + Args: + global_xform: (B, J, 3, 3) + + Returns: + local_xform: (B, J, 3, 3) + """ + local_xform = torch.zeros_like(global_xform) + + for i in range(global_xform.shape[1]): + if i == 0: + local_xform[:, i] = global_xform[:, i] + else: + local_xform[:, i] = torch.bmm(torch.linalg.inv(global_xform[:, self.parents[i]]), global_xform[:, i]) + + return local_xform + + def forward(self, rotations, positions=None): + """ + Args: + rotations (B, J, D) + + Returns: + The global position of each joint after FK (B, J, 3) + """ + # Get the full transform with rotations for skinning + b_size = rotations.shape[0] + if positions is None: + positions = self.positions.repeat(b_size, 1, 1) + transforms = self.rotate_global(self.parents, positions, rotations) + coordinates = transforms[:, :, :3, 3] / transforms[:, :, 3:, 3] + + return coordinates, transforms diff --git a/models/utils/layer.py b/models/utils/layer.py new file mode 100644 index 0000000000000000000000000000000000000000..86f8013512086280656ee10225952642abe7b11e --- /dev/null +++ b/models/utils/layer.py @@ -0,0 +1,217 @@ +import random +import math +import numpy as np +import torch +import torch.nn as nn +from torch.nn.utils import weight_norm +import torch.nn.functional as F + +from .build_vocab import Vocab + +class Chomp1d(nn.Module): + def __init__(self, chomp_size): + super(Chomp1d, self).__init__() + self.chomp_size = chomp_size + + def forward(self, x): + return x[:, :, :-self.chomp_size].contiguous() + + +class TemporalBlock(nn.Module): + def __init__(self, n_inputs, n_outputs, kernel_size, stride, dilation, padding, dropout=0.2): + super(TemporalBlock, self).__init__() + self.conv1 = weight_norm(nn.Conv1d(n_inputs, n_outputs, kernel_size, + stride=stride, padding=padding, dilation=dilation)) + self.chomp1 = Chomp1d(padding) + self.relu1 = nn.ReLU() + self.dropout1 = nn.Dropout(dropout) + + self.conv2 = weight_norm(nn.Conv1d(n_outputs, n_outputs, kernel_size, + stride=stride, padding=padding, dilation=dilation)) + self.chomp2 = Chomp1d(padding) + self.relu2 = nn.ReLU() + self.dropout2 = nn.Dropout(dropout) + + self.net = nn.Sequential(self.conv1, self.chomp1, self.relu1, self.dropout1, + self.conv2, self.chomp2, self.relu2, self.dropout2) + self.downsample = nn.Conv1d(n_inputs, n_outputs, 1) if n_inputs != n_outputs else None + self.relu = nn.ReLU() + self.init_weights() + + def init_weights(self): + self.conv1.weight.data.normal_(0, 0.01) + self.conv2.weight.data.normal_(0, 0.01) + if self.downsample is not None: + self.downsample.weight.data.normal_(0, 0.01) + + def forward(self, x): + out = self.net(x) + res = x if self.downsample is None else self.downsample(x) + return self.relu(out + res) + + +class TemporalConvNet(nn.Module): + def __init__(self, num_inputs, num_channels, kernel_size=2, dropout=0.2): + super(TemporalConvNet, self).__init__() + layers = [] + num_levels = len(num_channels) + for i in range(num_levels): + dilation_size = 2 ** i + in_channels = num_inputs if i == 0 else num_channels[i-1] + out_channels = num_channels[i] + layers += [TemporalBlock(in_channels, out_channels, kernel_size, stride=1, dilation=dilation_size, + padding=(kernel_size-1) * dilation_size, dropout=dropout)] + + self.network = nn.Sequential(*layers) + + def forward(self, x): + return self.network(x) + + +class TextEncoderTCN(nn.Module): + """ based on https://github.com/locuslab/TCN/blob/master/TCN/word_cnn/model.py """ + def __init__(self, args, n_words=11195, embed_size=300, pre_trained_embedding=None, + kernel_size=2, dropout=0.3, emb_dropout=0.1, word_cache=False): + super(TextEncoderTCN, self).__init__() +# if word_cache: +# self.embedding = None +# else: +# if pre_trained_embedding is not None: # use pre-trained embedding (fasttext) +# #print(pre_trained_embedding.shape) +# assert pre_trained_embedding.shape[0] == n_words +# assert pre_trained_embedding.shape[1] == embed_size +# self.embedding = nn.Embedding.from_pretrained(torch.FloatTensor(pre_trained_embedding), +# freeze=args.freeze_wordembed) +# else: +# self.embedding = nn.Embedding(n_words, embed_size) + + num_channels = [args.hidden_size] #* args.n_layer + self.tcn = TemporalConvNet(embed_size, num_channels, kernel_size, dropout=dropout) + self.decoder = nn.Linear(num_channels[-1], args.word_f) + self.drop = nn.Dropout(emb_dropout) + #self.emb_dropout = emb_dropout + self.init_weights() + + def init_weights(self): + self.decoder.bias.data.fill_(0) + self.decoder.weight.data.normal_(0, 0.01) + + def forward(self, input): + #print(input.shape) +# if self.embedding is None: +# emb = self.drop(input) +# else: +# emb = self.drop(self.embedding(input)) + y = self.tcn(input.transpose(1, 2)).transpose(1, 2) + y = self.decoder(y) + return y, torch.max(y, dim=1)[0] + + + + + + + + + +def reparameterize(mu, logvar): + std = torch.exp(0.5 * logvar) + eps = torch.randn_like(std) + return mu + eps * std + +def ConvNormRelu(in_channels, out_channels, downsample=False, padding=0, batchnorm=True): + if not downsample: + k = 3 + s = 1 + else: + k = 4 + s = 2 + conv_block = nn.Conv1d(in_channels, out_channels, kernel_size=k, stride=s, padding=padding) + norm_block = nn.BatchNorm1d(out_channels) + if batchnorm: + net = nn.Sequential( + conv_block, + norm_block, + nn.LeakyReLU(0.2, True) + ) + else: + net = nn.Sequential( + conv_block, + nn.LeakyReLU(0.2, True) + ) + return net + +class BasicBlock(nn.Module): + """ based on timm: https://github.com/rwightman/pytorch-image-models """ + def __init__(self, inplanes, planes, ker_size, stride=1, downsample=None, cardinality=1, base_width=64, + reduce_first=1, dilation=1, first_dilation=None, act_layer=nn.LeakyReLU, norm_layer=nn.BatchNorm1d, attn_layer=None, aa_layer=None, drop_block=None, drop_path=None): + super(BasicBlock, self).__init__() + + self.conv1 = nn.Conv1d( + inplanes, planes, kernel_size=ker_size, stride=stride, padding=first_dilation, + dilation=dilation, bias=True) + self.bn1 = norm_layer(planes) + self.act1 = act_layer(inplace=True) + self.conv2 = nn.Conv1d( + planes, planes, kernel_size=ker_size, padding=ker_size//2, dilation=dilation, bias=True) + self.bn2 = norm_layer(planes) + self.act2 = act_layer(inplace=True) + if downsample is not None: + self.downsample = nn.Sequential( + nn.Conv1d(inplanes, planes, stride=stride, kernel_size=ker_size, padding=first_dilation, dilation=dilation, bias=True), + norm_layer(planes), + ) + else: self.downsample=None + self.stride = stride + self.dilation = dilation + self.drop_block = drop_block + self.drop_path = drop_path + + def zero_init_last_bn(self): + nn.init.zeros_(self.bn2.weight) + + def forward(self, x): + shortcut = x + x = self.conv1(x) + x = self.bn1(x) + x = self.act1(x) + x = self.conv2(x) + x = self.bn2(x) + if self.downsample is not None: + shortcut = self.downsample(shortcut) + x += shortcut + x = self.act2(x) + return x + +def init_weight(m): + if isinstance(m, nn.Conv1d) or isinstance(m, nn.Linear) or isinstance(m, nn.ConvTranspose1d): + nn.init.xavier_normal_(m.weight) + # m.bias.data.fill_(0.01) + if m.bias is not None: + nn.init.constant_(m.bias, 0) + +def init_weight_skcnn(m): + if isinstance(m, nn.Conv1d) or isinstance(m, nn.Linear) or isinstance(m, nn.ConvTranspose1d): + nn.init.kaiming_uniform_(m.weight, a=math.sqrt(5)) + # m.bias.data.fill_(0.01) + if m.bias is not None: + #nn.init.constant_(m.bias, 0) + fan_in, _ = nn.init._calculate_fan_in_and_fan_out(m.weight) + bound = 1 / math.sqrt(fan_in) + nn.init.uniform_(m.bias, -bound, bound) + +class ResBlock(nn.Module): + def __init__(self, channel): + super(ResBlock, self).__init__() + self.model = nn.Sequential( + nn.Conv1d(channel, channel, kernel_size=3, stride=1, padding=1), + nn.LeakyReLU(0.2, inplace=True), + nn.Conv1d(channel, channel, kernel_size=3, stride=1, padding=1), + ) + + def forward(self, x): + residual = x + out = self.model(x) + out += residual + return out + \ No newline at end of file diff --git a/models/utils/rotation_conversions.py b/models/utils/rotation_conversions.py new file mode 100644 index 0000000000000000000000000000000000000000..f2bfaa1b2247622bff35d3f9b15e8eb84064aa53 --- /dev/null +++ b/models/utils/rotation_conversions.py @@ -0,0 +1,550 @@ +# Copyright (c) Facebook, Inc. and its affiliates. All rights reserved. + +import functools +from typing import Optional + +import torch +import torch.nn.functional as F + + +""" +The transformation matrices returned from the functions in this file assume +the points on which the transformation will be applied are column vectors. +i.e. the R matrix is structured as + + R = [ + [Rxx, Rxy, Rxz], + [Ryx, Ryy, Ryz], + [Rzx, Rzy, Rzz], + ] # (3, 3) + +This matrix can be applied to column vectors by post multiplication +by the points e.g. + + points = [[0], [1], [2]] # (3 x 1) xyz coordinates of a point + transformed_points = R * points + +To apply the same matrix to points which are row vectors, the R matrix +can be transposed and pre multiplied by the points: + +e.g. + points = [[0, 1, 2]] # (1 x 3) xyz coordinates of a point + transformed_points = points * R.transpose(1, 0) +""" + + +def quaternion_to_matrix(quaternions): + """ + Convert rotations given as quaternions to rotation matrices. + + Args: + quaternions: quaternions with real part first, + as tensor of shape (..., 4). + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + r, i, j, k = torch.unbind(quaternions, -1) + two_s = 2.0 / (quaternions * quaternions).sum(-1) + + o = torch.stack( + ( + 1 - two_s * (j * j + k * k), + two_s * (i * j - k * r), + two_s * (i * k + j * r), + two_s * (i * j + k * r), + 1 - two_s * (i * i + k * k), + two_s * (j * k - i * r), + two_s * (i * k - j * r), + two_s * (j * k + i * r), + 1 - two_s * (i * i + j * j), + ), + -1, + ) + return o.reshape(quaternions.shape[:-1] + (3, 3)) + + +def _copysign(a, b): + """ + Return a tensor where each element has the absolute value taken from the, + corresponding element of a, with sign taken from the corresponding + element of b. This is like the standard copysign floating-point operation, + but is not careful about negative 0 and NaN. + + Args: + a: source tensor. + b: tensor whose signs will be used, of the same shape as a. + + Returns: + Tensor of the same shape as a with the signs of b. + """ + signs_differ = (a < 0) != (b < 0) + return torch.where(signs_differ, -a, a) + + +def _sqrt_positive_part(x): + """ + Returns torch.sqrt(torch.max(0, x)) + but with a zero subgradient where x is 0. + """ + ret = torch.zeros_like(x) + positive_mask = x > 0 + ret[positive_mask] = torch.sqrt(x[positive_mask]) + return ret + + +def matrix_to_quaternion(matrix): + """ + Convert rotations given as rotation matrices to quaternions. + + Args: + matrix: Rotation matrices as tensor of shape (..., 3, 3). + + Returns: + quaternions with real part first, as tensor of shape (..., 4). + """ + if matrix.size(-1) != 3 or matrix.size(-2) != 3: + raise ValueError(f"Invalid rotation matrix shape f{matrix.shape}.") + m00 = matrix[..., 0, 0] + m11 = matrix[..., 1, 1] + m22 = matrix[..., 2, 2] + o0 = 0.5 * _sqrt_positive_part(1 + m00 + m11 + m22) + x = 0.5 * _sqrt_positive_part(1 + m00 - m11 - m22) + y = 0.5 * _sqrt_positive_part(1 - m00 + m11 - m22) + z = 0.5 * _sqrt_positive_part(1 - m00 - m11 + m22) + o1 = _copysign(x, matrix[..., 2, 1] - matrix[..., 1, 2]) + o2 = _copysign(y, matrix[..., 0, 2] - matrix[..., 2, 0]) + o3 = _copysign(z, matrix[..., 1, 0] - matrix[..., 0, 1]) + return torch.stack((o0, o1, o2, o3), -1) + + +def _axis_angle_rotation(axis: str, angle): + """ + Return the rotation matrices for one of the rotations about an axis + of which Euler angles describe, for each value of the angle given. + + Args: + axis: Axis label "X" or "Y or "Z". + angle: any shape tensor of Euler angles in radians + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + + cos = torch.cos(angle) + sin = torch.sin(angle) + one = torch.ones_like(angle) + zero = torch.zeros_like(angle) + + if axis == "X": + R_flat = (one, zero, zero, zero, cos, -sin, zero, sin, cos) + if axis == "Y": + R_flat = (cos, zero, sin, zero, one, zero, -sin, zero, cos) + if axis == "Z": + R_flat = (cos, -sin, zero, sin, cos, zero, zero, zero, one) + + return torch.stack(R_flat, -1).reshape(angle.shape + (3, 3)) + + +def euler_angles_to_matrix(euler_angles, convention: str): + """ + Convert rotations given as Euler angles in radians to rotation matrices. + + Args: + euler_angles: Euler angles in radians as tensor of shape (..., 3). + convention: Convention string of three uppercase letters from + {"X", "Y", and "Z"}. + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + if euler_angles.dim() == 0 or euler_angles.shape[-1] != 3: + raise ValueError("Invalid input euler angles.") + if len(convention) != 3: + raise ValueError("Convention must have 3 letters.") + if convention[1] in (convention[0], convention[2]): + raise ValueError(f"Invalid convention {convention}.") + for letter in convention: + if letter not in ("X", "Y", "Z"): + raise ValueError(f"Invalid letter {letter} in convention string.") + matrices = map(_axis_angle_rotation, convention, torch.unbind(euler_angles, -1)) + return functools.reduce(torch.matmul, matrices) + + +def _angle_from_tan( + axis: str, other_axis: str, data, horizontal: bool, tait_bryan: bool +): + """ + Extract the first or third Euler angle from the two members of + the matrix which are positive constant times its sine and cosine. + + Args: + axis: Axis label "X" or "Y or "Z" for the angle we are finding. + other_axis: Axis label "X" or "Y or "Z" for the middle axis in the + convention. + data: Rotation matrices as tensor of shape (..., 3, 3). + horizontal: Whether we are looking for the angle for the third axis, + which means the relevant entries are in the same row of the + rotation matrix. If not, they are in the same column. + tait_bryan: Whether the first and third axes in the convention differ. + + Returns: + Euler Angles in radians for each matrix in data as a tensor + of shape (...). + """ + + i1, i2 = {"X": (2, 1), "Y": (0, 2), "Z": (1, 0)}[axis] + if horizontal: + i2, i1 = i1, i2 + even = (axis + other_axis) in ["XY", "YZ", "ZX"] + if horizontal == even: + return torch.atan2(data[..., i1], data[..., i2]) + if tait_bryan: + return torch.atan2(-data[..., i2], data[..., i1]) + return torch.atan2(data[..., i2], -data[..., i1]) + + +def _index_from_letter(letter: str): + if letter == "X": + return 0 + if letter == "Y": + return 1 + if letter == "Z": + return 2 + + +def matrix_to_euler_angles(matrix, convention: str): + """ + Convert rotations given as rotation matrices to Euler angles in radians. + + Args: + matrix: Rotation matrices as tensor of shape (..., 3, 3). + convention: Convention string of three uppercase letters. + + Returns: + Euler angles in radians as tensor of shape (..., 3). + """ + if len(convention) != 3: + raise ValueError("Convention must have 3 letters.") + if convention[1] in (convention[0], convention[2]): + raise ValueError(f"Invalid convention {convention}.") + for letter in convention: + if letter not in ("X", "Y", "Z"): + raise ValueError(f"Invalid letter {letter} in convention string.") + if matrix.size(-1) != 3 or matrix.size(-2) != 3: + raise ValueError(f"Invalid rotation matrix shape f{matrix.shape}.") + i0 = _index_from_letter(convention[0]) + i2 = _index_from_letter(convention[2]) + tait_bryan = i0 != i2 + if tait_bryan: + central_angle = torch.asin( + matrix[..., i0, i2] * (-1.0 if i0 - i2 in [-1, 2] else 1.0) + ) + else: + central_angle = torch.acos(matrix[..., i0, i0]) + + o = ( + _angle_from_tan( + convention[0], convention[1], matrix[..., i2], False, tait_bryan + ), + central_angle, + _angle_from_tan( + convention[2], convention[1], matrix[..., i0, :], True, tait_bryan + ), + ) + return torch.stack(o, -1) + + +def random_quaternions( + n: int, dtype: Optional[torch.dtype] = None, device=None, requires_grad=False +): + """ + Generate random quaternions representing rotations, + i.e. versors with nonnegative real part. + + Args: + n: Number of quaternions in a batch to return. + dtype: Type to return. + device: Desired device of returned tensor. Default: + uses the current device for the default tensor type. + requires_grad: Whether the resulting tensor should have the gradient + flag set. + + Returns: + Quaternions as tensor of shape (N, 4). + """ + o = torch.randn((n, 4), dtype=dtype, device=device, requires_grad=requires_grad) + s = (o * o).sum(1) + o = o / _copysign(torch.sqrt(s), o[:, 0])[:, None] + return o + + +def random_rotations( + n: int, dtype: Optional[torch.dtype] = None, device=None, requires_grad=False +): + """ + Generate random rotations as 3x3 rotation matrices. + + Args: + n: Number of rotation matrices in a batch to return. + dtype: Type to return. + device: Device of returned tensor. Default: if None, + uses the current device for the default tensor type. + requires_grad: Whether the resulting tensor should have the gradient + flag set. + + Returns: + Rotation matrices as tensor of shape (n, 3, 3). + """ + quaternions = random_quaternions( + n, dtype=dtype, device=device, requires_grad=requires_grad + ) + return quaternion_to_matrix(quaternions) + + +def random_rotation( + dtype: Optional[torch.dtype] = None, device=None, requires_grad=False +): + """ + Generate a single random 3x3 rotation matrix. + + Args: + dtype: Type to return + device: Device of returned tensor. Default: if None, + uses the current device for the default tensor type + requires_grad: Whether the resulting tensor should have the gradient + flag set + + Returns: + Rotation matrix as tensor of shape (3, 3). + """ + return random_rotations(1, dtype, device, requires_grad)[0] + + +def standardize_quaternion(quaternions): + """ + Convert a unit quaternion to a standard form: one in which the real + part is non negative. + + Args: + quaternions: Quaternions with real part first, + as tensor of shape (..., 4). + + Returns: + Standardized quaternions as tensor of shape (..., 4). + """ + return torch.where(quaternions[..., 0:1] < 0, -quaternions, quaternions) + + +def quaternion_raw_multiply(a, b): + """ + Multiply two quaternions. + Usual torch rules for broadcasting apply. + + Args: + a: Quaternions as tensor of shape (..., 4), real part first. + b: Quaternions as tensor of shape (..., 4), real part first. + + Returns: + The product of a and b, a tensor of quaternions shape (..., 4). + """ + aw, ax, ay, az = torch.unbind(a, -1) + bw, bx, by, bz = torch.unbind(b, -1) + ow = aw * bw - ax * bx - ay * by - az * bz + ox = aw * bx + ax * bw + ay * bz - az * by + oy = aw * by - ax * bz + ay * bw + az * bx + oz = aw * bz + ax * by - ay * bx + az * bw + return torch.stack((ow, ox, oy, oz), -1) + + +def quaternion_multiply(a, b): + """ + Multiply two quaternions representing rotations, returning the quaternion + representing their composition, i.e. the versor with nonnegative real part. + Usual torch rules for broadcasting apply. + + Args: + a: Quaternions as tensor of shape (..., 4), real part first. + b: Quaternions as tensor of shape (..., 4), real part first. + + Returns: + The product of a and b, a tensor of quaternions of shape (..., 4). + """ + ab = quaternion_raw_multiply(a, b) + return standardize_quaternion(ab) + + +def quaternion_invert(quaternion): + """ + Given a quaternion representing rotation, get the quaternion representing + its inverse. + + Args: + quaternion: Quaternions as tensor of shape (..., 4), with real part + first, which must be versors (unit quaternions). + + Returns: + The inverse, a tensor of quaternions of shape (..., 4). + """ + + return quaternion * quaternion.new_tensor([1, -1, -1, -1]) + + +def quaternion_apply(quaternion, point): + """ + Apply the rotation given by a quaternion to a 3D point. + Usual torch rules for broadcasting apply. + + Args: + quaternion: Tensor of quaternions, real part first, of shape (..., 4). + point: Tensor of 3D points of shape (..., 3). + + Returns: + Tensor of rotated points of shape (..., 3). + """ + if point.size(-1) != 3: + raise ValueError(f"Points are not in 3D, f{point.shape}.") + real_parts = point.new_zeros(point.shape[:-1] + (1,)) + point_as_quaternion = torch.cat((real_parts, point), -1) + out = quaternion_raw_multiply( + quaternion_raw_multiply(quaternion, point_as_quaternion), + quaternion_invert(quaternion), + ) + return out[..., 1:] + + +def axis_angle_to_matrix(axis_angle): + """ + Convert rotations given as axis/angle to rotation matrices. + + Args: + axis_angle: Rotations given as a vector in axis angle form, + as a tensor of shape (..., 3), where the magnitude is + the angle turned anticlockwise in radians around the + vector's direction. + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + return quaternion_to_matrix(axis_angle_to_quaternion(axis_angle)) + + +def matrix_to_axis_angle(matrix): + """ + Convert rotations given as rotation matrices to axis/angle. + + Args: + matrix: Rotation matrices as tensor of shape (..., 3, 3). + + Returns: + Rotations given as a vector in axis angle form, as a tensor + of shape (..., 3), where the magnitude is the angle + turned anticlockwise in radians around the vector's + direction. + """ + return quaternion_to_axis_angle(matrix_to_quaternion(matrix)) + + +def axis_angle_to_quaternion(axis_angle): + """ + Convert rotations given as axis/angle to quaternions. + + Args: + axis_angle: Rotations given as a vector in axis angle form, + as a tensor of shape (..., 3), where the magnitude is + the angle turned anticlockwise in radians around the + vector's direction. + + Returns: + quaternions with real part first, as tensor of shape (..., 4). + """ + angles = torch.norm(axis_angle, p=2, dim=-1, keepdim=True) + half_angles = 0.5 * angles + eps = 1e-6 + small_angles = angles.abs() < eps + sin_half_angles_over_angles = torch.empty_like(angles) + sin_half_angles_over_angles[~small_angles] = ( + torch.sin(half_angles[~small_angles]) / angles[~small_angles] + ) + # for x small, sin(x/2) is about x/2 - (x/2)^3/6 + # so sin(x/2)/x is about 1/2 - (x*x)/48 + sin_half_angles_over_angles[small_angles] = ( + 0.5 - (angles[small_angles] * angles[small_angles]) / 48 + ) + quaternions = torch.cat( + [torch.cos(half_angles), axis_angle * sin_half_angles_over_angles], dim=-1 + ) + return quaternions + + +def quaternion_to_axis_angle(quaternions): + """ + Convert rotations given as quaternions to axis/angle. + + Args: + quaternions: quaternions with real part first, + as tensor of shape (..., 4). + + Returns: + Rotations given as a vector in axis angle form, as a tensor + of shape (..., 3), where the magnitude is the angle + turned anticlockwise in radians around the vector's + direction. + """ + norms = torch.norm(quaternions[..., 1:], p=2, dim=-1, keepdim=True) + half_angles = torch.atan2(norms, quaternions[..., :1]) + angles = 2 * half_angles + eps = 1e-6 + small_angles = angles.abs() < eps + sin_half_angles_over_angles = torch.empty_like(angles) + sin_half_angles_over_angles[~small_angles] = ( + torch.sin(half_angles[~small_angles]) / angles[~small_angles] + ) + # for x small, sin(x/2) is about x/2 - (x/2)^3/6 + # so sin(x/2)/x is about 1/2 - (x*x)/48 + sin_half_angles_over_angles[small_angles] = ( + 0.5 - (angles[small_angles] * angles[small_angles]) / 48 + ) + return quaternions[..., 1:] / sin_half_angles_over_angles + + +def rotation_6d_to_matrix(d6: torch.Tensor) -> torch.Tensor: + """ + Converts 6D rotation representation by Zhou et al. [1] to rotation matrix + using Gram--Schmidt orthogonalisation per Section B of [1]. + Args: + d6: 6D rotation representation, of size (*, 6) + + Returns: + batch of rotation matrices of size (*, 3, 3) + + [1] Zhou, Y., Barnes, C., Lu, J., Yang, J., & Li, H. + On the Continuity of Rotation Representations in Neural Networks. + IEEE Conference on Computer Vision and Pattern Recognition, 2019. + Retrieved from http://arxiv.org/abs/1812.07035 + """ + + a1, a2 = d6[..., :3], d6[..., 3:] + b1 = F.normalize(a1, dim=-1) + b2 = a2 - (b1 * a2).sum(-1, keepdim=True) * b1 + b2 = F.normalize(b2, dim=-1) + b3 = torch.cross(b1, b2, dim=-1) + return torch.stack((b1, b2, b3), dim=-2) + + +def matrix_to_rotation_6d(matrix: torch.Tensor) -> torch.Tensor: + """ + Converts rotation matrices to 6D rotation representation by Zhou et al. [1] + by dropping the last row. Note that 6D representation is not unique. + Args: + matrix: batch of rotation matrices of size (*, 3, 3) + + Returns: + 6D rotation representation, of size (*, 6) + + [1] Zhou, Y., Barnes, C., Lu, J., Yang, J., & Li, H. + On the Continuity of Rotation Representations in Neural Networks. + IEEE Conference on Computer Vision and Pattern Recognition, 2019. + Retrieved from http://arxiv.org/abs/1812.07035 + """ + return matrix[..., :2, :].clone().reshape(*matrix.size()[:-2], 6) diff --git a/models/utils/rotations.py b/models/utils/rotations.py new file mode 100644 index 0000000000000000000000000000000000000000..55729b2724c9c34234bddb63a826aa1f9a4321b9 --- /dev/null +++ b/models/utils/rotations.py @@ -0,0 +1,587 @@ +# Copyright (c) Meta Platforms, Inc. and affiliates. +# All rights reserved. +# +# This source code is licensed under the BSD-style license found in the +# LICENSE file in the root directory of this source tree. + +from typing import Optional, Union + +import torch +import torch.nn.functional as F + +Device = Union[str, torch.device] + + +""" +The transformation matrices returned from the functions in this file assume +the points on which the transformation will be applied are column vectors. +i.e. the R matrix is structured as + + R = [ + [Rxx, Rxy, Rxz], + [Ryx, Ryy, Ryz], + [Rzx, Rzy, Rzz], + ] # (3, 3) + +This matrix can be applied to column vectors by post multiplication +by the points e.g. + + points = [[0], [1], [2]] # (3 x 1) xyz coordinates of a point + transformed_points = R * points + +To apply the same matrix to points which are row vectors, the R matrix +can be transposed and pre multiplied by the points: + +e.g. + points = [[0, 1, 2]] # (1 x 3) xyz coordinates of a point + transformed_points = points * R.transpose(1, 0) +""" + + +def quaternion_to_matrix(quaternions: torch.Tensor) -> torch.Tensor: + """ + Convert rotations given as quaternions to rotation matrices. + + Args: + quaternions: quaternions with real part first, + as tensor of shape (..., 4). + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + r, i, j, k = torch.unbind(quaternions, -1) + two_s = 2.0 / (quaternions * quaternions).sum(-1) + + o = torch.stack( + ( + 1 - two_s * (j * j + k * k), + two_s * (i * j - k * r), + two_s * (i * k + j * r), + two_s * (i * j + k * r), + 1 - two_s * (i * i + k * k), + two_s * (j * k - i * r), + two_s * (i * k - j * r), + two_s * (j * k + i * r), + 1 - two_s * (i * i + j * j), + ), + -1, + ) + return o.reshape(quaternions.shape[:-1] + (3, 3)) + + +def _copysign(a: torch.Tensor, b: torch.Tensor) -> torch.Tensor: + """ + Return a tensor where each element has the absolute value taken from the, + corresponding element of a, with sign taken from the corresponding + element of b. This is like the standard copysign floating-point operation, + but is not careful about negative 0 and NaN. + + Args: + a: source tensor. + b: tensor whose signs will be used, of the same shape as a. + + Returns: + Tensor of the same shape as a with the signs of b. + """ + signs_differ = (a < 0) != (b < 0) + return torch.where(signs_differ, -a, a) + + +def _sqrt_positive_part(x: torch.Tensor) -> torch.Tensor: + """ + Returns torch.sqrt(torch.max(0, x)) + but with a zero subgradient where x is 0. + """ + ret = torch.zeros_like(x) + positive_mask = x > 0 + ret[positive_mask] = torch.sqrt(x[positive_mask]) + return ret + + +def matrix_to_quaternion(matrix: torch.Tensor) -> torch.Tensor: + """ + Convert rotations given as rotation matrices to quaternions. + + Args: + matrix: Rotation matrices as tensor of shape (..., 3, 3). + + Returns: + quaternions with real part first, as tensor of shape (..., 4). + """ + if matrix.size(-1) != 3 or matrix.size(-2) != 3: + raise ValueError(f"Invalid rotation matrix shape {matrix.shape}.") + + batch_dim = matrix.shape[:-2] + m00, m01, m02, m10, m11, m12, m20, m21, m22 = torch.unbind( + matrix.reshape(batch_dim + (9,)), dim=-1 + ) + + q_abs = _sqrt_positive_part( + torch.stack( + [ + 1.0 + m00 + m11 + m22, + 1.0 + m00 - m11 - m22, + 1.0 - m00 + m11 - m22, + 1.0 - m00 - m11 + m22, + ], + dim=-1, + ) + ) + + # we produce the desired quaternion multiplied by each of r, i, j, k + quat_by_rijk = torch.stack( + [ + torch.stack([q_abs[..., 0] ** 2, m21 - m12, m02 - m20, m10 - m01], dim=-1), + torch.stack([m21 - m12, q_abs[..., 1] ** 2, m10 + m01, m02 + m20], dim=-1), + torch.stack([m02 - m20, m10 + m01, q_abs[..., 2] ** 2, m12 + m21], dim=-1), + torch.stack([m10 - m01, m20 + m02, m21 + m12, q_abs[..., 3] ** 2], dim=-1), + ], + dim=-2, + ) + + # We floor here at 0.1 but the exact level is not important; if q_abs is small, + # the candidate won't be picked. + flr = torch.tensor(0.1).to(dtype=q_abs.dtype, device=q_abs.device) + quat_candidates = quat_by_rijk / (2.0 * q_abs[..., None].max(flr)) + + # if not for numerical problems, quat_candidates[i] should be same (up to a sign), + # forall i; we pick the best-conditioned one (with the largest denominator) + + return quat_candidates[ + F.one_hot(q_abs.argmax(dim=-1), num_classes=4) > 0.5, : # pyre-ignore[16] + ].reshape(batch_dim + (4,)) + + +def _axis_angle_rotation(axis: str, angle: torch.Tensor) -> torch.Tensor: + """ + Return the rotation matrices for one of the rotations about an axis + of which Euler angles describe, for each value of the angle given. + + Args: + axis: Axis label "X" or "Y or "Z". + angle: any shape tensor of Euler angles in radians + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + + cos = torch.cos(angle) + sin = torch.sin(angle) + one = torch.ones_like(angle) + zero = torch.zeros_like(angle) + + if axis == "X": + R_flat = (one, zero, zero, zero, cos, -sin, zero, sin, cos) + elif axis == "Y": + R_flat = (cos, zero, sin, zero, one, zero, -sin, zero, cos) + elif axis == "Z": + R_flat = (cos, -sin, zero, sin, cos, zero, zero, zero, one) + else: + raise ValueError("letter must be either X, Y or Z.") + + return torch.stack(R_flat, -1).reshape(angle.shape + (3, 3)) + + +def euler_angles_to_matrix(euler_angles: torch.Tensor, convention: str) -> torch.Tensor: + """ + Convert rotations given as Euler angles in radians to rotation matrices. + + Args: + euler_angles: Euler angles in radians as tensor of shape (..., 3). + convention: Convention string of three uppercase letters from + {"X", "Y", and "Z"}. + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + if euler_angles.dim() == 0 or euler_angles.shape[-1] != 3: + raise ValueError("Invalid input euler angles.") + if len(convention) != 3: + raise ValueError("Convention must have 3 letters.") + if convention[1] in (convention[0], convention[2]): + raise ValueError(f"Invalid convention {convention}.") + for letter in convention: + if letter not in ("X", "Y", "Z"): + raise ValueError(f"Invalid letter {letter} in convention string.") + matrices = [ + _axis_angle_rotation(c, e) + for c, e in zip(convention, torch.unbind(euler_angles, -1)) + ] + # return functools.reduce(torch.matmul, matrices) + return torch.matmul(torch.matmul(matrices[0], matrices[1]), matrices[2]) + + +def _angle_from_tan( + axis: str, other_axis: str, data, horizontal: bool, tait_bryan: bool +) -> torch.Tensor: + """ + Extract the first or third Euler angle from the two members of + the matrix which are positive constant times its sine and cosine. + + Args: + axis: Axis label "X" or "Y or "Z" for the angle we are finding. + other_axis: Axis label "X" or "Y or "Z" for the middle axis in the + convention. + data: Rotation matrices as tensor of shape (..., 3, 3). + horizontal: Whether we are looking for the angle for the third axis, + which means the relevant entries are in the same row of the + rotation matrix. If not, they are in the same column. + tait_bryan: Whether the first and third axes in the convention differ. + + Returns: + Euler Angles in radians for each matrix in data as a tensor + of shape (...). + """ + + i1, i2 = {"X": (2, 1), "Y": (0, 2), "Z": (1, 0)}[axis] + if horizontal: + i2, i1 = i1, i2 + even = (axis + other_axis) in ["XY", "YZ", "ZX"] + if horizontal == even: + return torch.atan2(data[..., i1], data[..., i2]) + if tait_bryan: + return torch.atan2(-data[..., i2], data[..., i1]) + return torch.atan2(data[..., i2], -data[..., i1]) + + +def _index_from_letter(letter: str) -> int: + if letter == "X": + return 0 + if letter == "Y": + return 1 + if letter == "Z": + return 2 + raise ValueError("letter must be either X, Y or Z.") + + +def matrix_to_euler_angles(matrix: torch.Tensor, convention: str) -> torch.Tensor: + """ + Convert rotations given as rotation matrices to Euler angles in radians. + + Args: + matrix: Rotation matrices as tensor of shape (..., 3, 3). + convention: Convention string of three uppercase letters. + + Returns: + Euler angles in radians as tensor of shape (..., 3). + """ + if len(convention) != 3: + raise ValueError("Convention must have 3 letters.") + if convention[1] in (convention[0], convention[2]): + raise ValueError(f"Invalid convention {convention}.") + for letter in convention: + if letter not in ("X", "Y", "Z"): + raise ValueError(f"Invalid letter {letter} in convention string.") + if matrix.size(-1) != 3 or matrix.size(-2) != 3: + raise ValueError(f"Invalid rotation matrix shape {matrix.shape}.") + i0 = _index_from_letter(convention[0]) + i2 = _index_from_letter(convention[2]) + tait_bryan = i0 != i2 + if tait_bryan: + central_angle = torch.asin( + matrix[..., i0, i2] * (-1.0 if i0 - i2 in [-1, 2] else 1.0) + ) + else: + central_angle = torch.acos(matrix[..., i0, i0]) + + o = ( + _angle_from_tan( + convention[0], convention[1], matrix[..., i2], False, tait_bryan + ), + central_angle, + _angle_from_tan( + convention[2], convention[1], matrix[..., i0, :], True, tait_bryan + ), + ) + return torch.stack(o, -1) + + +def random_quaternions( + n: int, dtype: Optional[torch.dtype] = None, device: Optional[Device] = None +) -> torch.Tensor: + """ + Generate random quaternions representing rotations, + i.e. versors with nonnegative real part. + + Args: + n: Number of quaternions in a batch to return. + dtype: Type to return. + device: Desired device of returned tensor. Default: + uses the current device for the default tensor type. + + Returns: + Quaternions as tensor of shape (N, 4). + """ + if isinstance(device, str): + device = torch.device(device) + o = torch.randn((n, 4), dtype=dtype, device=device) + s = (o * o).sum(1) + o = o / _copysign(torch.sqrt(s), o[:, 0])[:, None] + return o + + +def random_rotations( + n: int, dtype: Optional[torch.dtype] = None, device: Optional[Device] = None +) -> torch.Tensor: + """ + Generate random rotations as 3x3 rotation matrices. + + Args: + n: Number of rotation matrices in a batch to return. + dtype: Type to return. + device: Device of returned tensor. Default: if None, + uses the current device for the default tensor type. + + Returns: + Rotation matrices as tensor of shape (n, 3, 3). + """ + quaternions = random_quaternions(n, dtype=dtype, device=device) + return quaternion_to_matrix(quaternions) + + +def random_rotation( + dtype: Optional[torch.dtype] = None, device: Optional[Device] = None +) -> torch.Tensor: + """ + Generate a single random 3x3 rotation matrix. + + Args: + dtype: Type to return + device: Device of returned tensor. Default: if None, + uses the current device for the default tensor type + + Returns: + Rotation matrix as tensor of shape (3, 3). + """ + return random_rotations(1, dtype, device)[0] + + +def standardize_quaternion(quaternions: torch.Tensor) -> torch.Tensor: + """ + Convert a unit quaternion to a standard form: one in which the real + part is non negative. + + Args: + quaternions: Quaternions with real part first, + as tensor of shape (..., 4). + + Returns: + Standardized quaternions as tensor of shape (..., 4). + """ + return torch.where(quaternions[..., 0:1] < 0, -quaternions, quaternions) + + +def quaternion_raw_multiply(a: torch.Tensor, b: torch.Tensor) -> torch.Tensor: + """ + Multiply two quaternions. + Usual torch rules for broadcasting apply. + + Args: + a: Quaternions as tensor of shape (..., 4), real part first. + b: Quaternions as tensor of shape (..., 4), real part first. + + Returns: + The product of a and b, a tensor of quaternions shape (..., 4). + """ + aw, ax, ay, az = torch.unbind(a, -1) + bw, bx, by, bz = torch.unbind(b, -1) + ow = aw * bw - ax * bx - ay * by - az * bz + ox = aw * bx + ax * bw + ay * bz - az * by + oy = aw * by - ax * bz + ay * bw + az * bx + oz = aw * bz + ax * by - ay * bx + az * bw + return torch.stack((ow, ox, oy, oz), -1) + + +def quaternion_multiply(a: torch.Tensor, b: torch.Tensor) -> torch.Tensor: + """ + Multiply two quaternions representing rotations, returning the quaternion + representing their composition, i.e. the versor with nonnegative real part. + Usual torch rules for broadcasting apply. + + Args: + a: Quaternions as tensor of shape (..., 4), real part first. + b: Quaternions as tensor of shape (..., 4), real part first. + + Returns: + The product of a and b, a tensor of quaternions of shape (..., 4). + """ + ab = quaternion_raw_multiply(a, b) + return standardize_quaternion(ab) + + +def quaternion_invert(quaternion: torch.Tensor) -> torch.Tensor: + """ + Given a quaternion representing rotation, get the quaternion representing + its inverse. + + Args: + quaternion: Quaternions as tensor of shape (..., 4), with real part + first, which must be versors (unit quaternions). + + Returns: + The inverse, a tensor of quaternions of shape (..., 4). + """ + + scaling = torch.tensor([1, -1, -1, -1], device=quaternion.device) + return quaternion * scaling + + +def quaternion_apply(quaternion: torch.Tensor, point: torch.Tensor) -> torch.Tensor: + """ + Apply the rotation given by a quaternion to a 3D point. + Usual torch rules for broadcasting apply. + + Args: + quaternion: Tensor of quaternions, real part first, of shape (..., 4). + point: Tensor of 3D points of shape (..., 3). + + Returns: + Tensor of rotated points of shape (..., 3). + """ + if point.size(-1) != 3: + raise ValueError(f"Points are not in 3D, {point.shape}.") + real_parts = point.new_zeros(point.shape[:-1] + (1,)) + point_as_quaternion = torch.cat((real_parts, point), -1) + out = quaternion_raw_multiply( + quaternion_raw_multiply(quaternion, point_as_quaternion), + quaternion_invert(quaternion), + ) + return out[..., 1:] + + +def axis_angle_to_matrix(axis_angle: torch.Tensor) -> torch.Tensor: + """ + Convert rotations given as axis/angle to rotation matrices. + + Args: + axis_angle: Rotations given as a vector in axis angle form, + as a tensor of shape (..., 3), where the magnitude is + the angle turned anticlockwise in radians around the + vector's direction. + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + return quaternion_to_matrix(axis_angle_to_quaternion(axis_angle)) + + +def matrix_to_axis_angle(matrix: torch.Tensor) -> torch.Tensor: + """ + Convert rotations given as rotation matrices to axis/angle. + + Args: + matrix: Rotation matrices as tensor of shape (..., 3, 3). + + Returns: + Rotations given as a vector in axis angle form, as a tensor + of shape (..., 3), where the magnitude is the angle + turned anticlockwise in radians around the vector's + direction. + """ + return quaternion_to_axis_angle(matrix_to_quaternion(matrix)) + + +def axis_angle_to_quaternion(axis_angle: torch.Tensor) -> torch.Tensor: + """ + Convert rotations given as axis/angle to quaternions. + + Args: + axis_angle: Rotations given as a vector in axis angle form, + as a tensor of shape (..., 3), where the magnitude is + the angle turned anticlockwise in radians around the + vector's direction. + + Returns: + quaternions with real part first, as tensor of shape (..., 4). + """ + angles = torch.norm(axis_angle, p=2, dim=-1, keepdim=True) + half_angles = angles * 0.5 + eps = 1e-6 + small_angles = angles.abs() < eps + sin_half_angles_over_angles = torch.empty_like(angles) + sin_half_angles_over_angles[~small_angles] = ( + torch.sin(half_angles[~small_angles]) / angles[~small_angles] + ) + # for x small, sin(x/2) is about x/2 - (x/2)^3/6 + # so sin(x/2)/x is about 1/2 - (x*x)/48 + sin_half_angles_over_angles[small_angles] = ( + 0.5 - (angles[small_angles] * angles[small_angles]) / 48 + ) + quaternions = torch.cat( + [torch.cos(half_angles), axis_angle * sin_half_angles_over_angles], dim=-1 + ) + return quaternions + + +def quaternion_to_axis_angle(quaternions: torch.Tensor) -> torch.Tensor: + """ + Convert rotations given as quaternions to axis/angle. + + Args: + quaternions: quaternions with real part first, + as tensor of shape (..., 4). + + Returns: + Rotations given as a vector in axis angle form, as a tensor + of shape (..., 3), where the magnitude is the angle + turned anticlockwise in radians around the vector's + direction. + """ + norms = torch.norm(quaternions[..., 1:], p=2, dim=-1, keepdim=True) + half_angles = torch.atan2(norms, quaternions[..., :1]) + angles = 2 * half_angles + eps = 1e-6 + small_angles = angles.abs() < eps + sin_half_angles_over_angles = torch.empty_like(angles) + sin_half_angles_over_angles[~small_angles] = ( + torch.sin(half_angles[~small_angles]) / angles[~small_angles] + ) + # for x small, sin(x/2) is about x/2 - (x/2)^3/6 + # so sin(x/2)/x is about 1/2 - (x*x)/48 + sin_half_angles_over_angles[small_angles] = ( + 0.5 - (angles[small_angles] * angles[small_angles]) / 48 + ) + return quaternions[..., 1:] / sin_half_angles_over_angles + + +def rotation_6d_to_matrix(d6: torch.Tensor) -> torch.Tensor: + """ + Converts 6D rotation representation by Zhou et al. [1] to rotation matrix + using Gram--Schmidt orthogonalization per Section B of [1]. + Args: + d6: 6D rotation representation, of size (*, 6) + + Returns: + batch of rotation matrices of size (*, 3, 3) + + [1] Zhou, Y., Barnes, C., Lu, J., Yang, J., & Li, H. + On the Continuity of Rotation Representations in Neural Networks. + IEEE Conference on Computer Vision and Pattern Recognition, 2019. + Retrieved from http://arxiv.org/abs/1812.07035 + """ + + a1, a2 = d6[..., :3], d6[..., 3:] + b1 = F.normalize(a1, dim=-1) + b2 = a2 - (b1 * a2).sum(-1, keepdim=True) * b1 + b2 = F.normalize(b2, dim=-1) + b3 = torch.cross(b1, b2, dim=-1) + return torch.stack((b1, b2, b3), dim=-2) + + +def matrix_to_rotation_6d(matrix: torch.Tensor) -> torch.Tensor: + """ + Converts rotation matrices to 6D rotation representation by Zhou et al. [1] + by dropping the last row. Note that 6D representation is not unique. + Args: + matrix: batch of rotation matrices of size (*, 3, 3) + + Returns: + 6D rotation representation, of size (*, 6) + + [1] Zhou, Y., Barnes, C., Lu, J., Yang, J., & Li, H. + On the Continuity of Rotation Representations in Neural Networks. + IEEE Conference on Computer Vision and Pattern Recognition, 2019. + Retrieved from http://arxiv.org/abs/1812.07035 + """ + batch_dim = matrix.size()[:-2] + return matrix[..., :2, :].clone().reshape(batch_dim + (6,)) diff --git a/models/utils/skeleton.py b/models/utils/skeleton.py new file mode 100644 index 0000000000000000000000000000000000000000..123656b7516aec1b424f9f87d384837eb820ccc9 --- /dev/null +++ b/models/utils/skeleton.py @@ -0,0 +1,636 @@ +import math + +import numpy as np +import torch +import torch.nn as nn +import torch.nn.functional as F + + +class SkeletonConv(nn.Module): + def __init__(self, neighbour_list, in_channels, out_channels, kernel_size, joint_num, stride=1, padding=0, + bias=True, padding_mode='zeros', add_offset=False, in_offset_channel=0): + self.in_channels_per_joint = in_channels // joint_num + self.out_channels_per_joint = out_channels // joint_num + if in_channels % joint_num != 0 or out_channels % joint_num != 0: + raise Exception('BAD') + super(SkeletonConv, self).__init__() + + if padding_mode == 'zeros': + padding_mode = 'constant' + if padding_mode == 'reflection': + padding_mode = 'reflect' + + self.expanded_neighbour_list = [] + self.expanded_neighbour_list_offset = [] + self.neighbour_list = neighbour_list + self.add_offset = add_offset + self.joint_num = joint_num + + self.stride = stride + self.dilation = 1 + self.groups = 1 + self.padding = padding + self.padding_mode = padding_mode + self._padding_repeated_twice = (padding, padding) + + for neighbour in neighbour_list: + expanded = [] + for k in neighbour: + for i in range(self.in_channels_per_joint): + expanded.append(k * self.in_channels_per_joint + i) + self.expanded_neighbour_list.append(expanded) + + if self.add_offset: + self.offset_enc = SkeletonLinear(neighbour_list, in_offset_channel * len(neighbour_list), out_channels) + + for neighbour in neighbour_list: + expanded = [] + for k in neighbour: + for i in range(add_offset): + expanded.append(k * in_offset_channel + i) + self.expanded_neighbour_list_offset.append(expanded) + + self.weight = torch.zeros(out_channels, in_channels, kernel_size) + if bias: + self.bias = torch.zeros(out_channels) + else: + self.register_parameter('bias', None) + + self.mask = torch.zeros_like(self.weight) + for i, neighbour in enumerate(self.expanded_neighbour_list): + self.mask[self.out_channels_per_joint * i: self.out_channels_per_joint * (i + 1), neighbour, ...] = 1 + self.mask = nn.Parameter(self.mask, requires_grad=False) + + self.description = 'SkeletonConv(in_channels_per_armature={}, out_channels_per_armature={}, kernel_size={}, ' \ + 'joint_num={}, stride={}, padding={}, bias={})'.format( + in_channels // joint_num, out_channels // joint_num, kernel_size, joint_num, stride, padding, bias + ) + + self.reset_parameters() + + def reset_parameters(self): + for i, neighbour in enumerate(self.expanded_neighbour_list): + """ Use temporary variable to avoid assign to copy of slice, which might lead to unexpected result """ + tmp = torch.zeros_like(self.weight[self.out_channels_per_joint * i: self.out_channels_per_joint * (i + 1), + neighbour, ...]) + nn.init.kaiming_uniform_(tmp, a=math.sqrt(5)) + self.weight[self.out_channels_per_joint * i: self.out_channels_per_joint * (i + 1), + neighbour, ...] = tmp + if self.bias is not None: + fan_in, _ = nn.init._calculate_fan_in_and_fan_out( + self.weight[self.out_channels_per_joint * i: self.out_channels_per_joint * (i + 1), neighbour, ...]) + bound = 1 / math.sqrt(fan_in) + tmp = torch.zeros_like( + self.bias[self.out_channels_per_joint * i: self.out_channels_per_joint * (i + 1)]) + nn.init.uniform_(tmp, -bound, bound) + self.bias[self.out_channels_per_joint * i: self.out_channels_per_joint * (i + 1)] = tmp + + self.weight = nn.Parameter(self.weight) + if self.bias is not None: + self.bias = nn.Parameter(self.bias) + + def set_offset(self, offset): + if not self.add_offset: + raise Exception('Wrong Combination of Parameters') + self.offset = offset.reshape(offset.shape[0], -1) + + def forward(self, input): + # print('SkeletonConv') + weight_masked = self.weight * self.mask + #print(f'input: {input.size()}') + res = F.conv1d(F.pad(input, self._padding_repeated_twice, mode=self.padding_mode), + weight_masked, self.bias, self.stride, + 0, self.dilation, self.groups) + + if self.add_offset: + offset_res = self.offset_enc(self.offset) + offset_res = offset_res.reshape(offset_res.shape + (1, )) + res += offset_res / 100 + #print(f'res: {res.size()}') + return res + + +class SkeletonLinear(nn.Module): + def __init__(self, neighbour_list, in_channels, out_channels, extra_dim1=False): + super(SkeletonLinear, self).__init__() + self.neighbour_list = neighbour_list + self.in_channels = in_channels + self.out_channels = out_channels + self.in_channels_per_joint = in_channels // len(neighbour_list) + self.out_channels_per_joint = out_channels // len(neighbour_list) + self.extra_dim1 = extra_dim1 + self.expanded_neighbour_list = [] + + for neighbour in neighbour_list: + expanded = [] + for k in neighbour: + for i in range(self.in_channels_per_joint): + expanded.append(k * self.in_channels_per_joint + i) + self.expanded_neighbour_list.append(expanded) + + self.weight = torch.zeros(out_channels, in_channels) + self.mask = torch.zeros(out_channels, in_channels) + self.bias = nn.Parameter(torch.Tensor(out_channels)) + + self.reset_parameters() + + def reset_parameters(self): + for i, neighbour in enumerate(self.expanded_neighbour_list): + tmp = torch.zeros_like( + self.weight[i*self.out_channels_per_joint: (i + 1)*self.out_channels_per_joint, neighbour] + ) + self.mask[i*self.out_channels_per_joint: (i + 1)*self.out_channels_per_joint, neighbour] = 1 + nn.init.kaiming_uniform_(tmp, a=math.sqrt(5)) + self.weight[i*self.out_channels_per_joint: (i + 1)*self.out_channels_per_joint, neighbour] = tmp + + fan_in, _ = nn.init._calculate_fan_in_and_fan_out(self.weight) + bound = 1 / math.sqrt(fan_in) + nn.init.uniform_(self.bias, -bound, bound) + + self.weight = nn.Parameter(self.weight) + self.mask = nn.Parameter(self.mask, requires_grad=False) + + def forward(self, input): + input = input.reshape(input.shape[0], -1) + weight_masked = self.weight * self.mask + res = F.linear(input, weight_masked, self.bias) + if self.extra_dim1: + res = res.reshape(res.shape + (1,)) + return res + + +class SkeletonPool(nn.Module): + def __init__(self, edges, pooling_mode, channels_per_edge, last_pool=False): + super(SkeletonPool, self).__init__() + + if pooling_mode != 'mean': + raise Exception('Unimplemented pooling mode in matrix_implementation') + + self.channels_per_edge = channels_per_edge + self.pooling_mode = pooling_mode + self.edge_num = len(edges) + # self.edge_num = len(edges) + 1 + self.seq_list = [] + self.pooling_list = [] + self.new_edges = [] + degree = [0] * 100 # each element represents the degree of the corresponding joint + + for edge in edges: + degree[edge[0]] += 1 + degree[edge[1]] += 1 + + # seq_list contains multiple sub-lists where each sub-list is an edge chain from the joint whose degree > 2 to the end effectors or joints whose degree > 2. + def find_seq(j, seq): + nonlocal self, degree, edges + + if degree[j] > 2 and j != 0: + self.seq_list.append(seq) + seq = [] + + if degree[j] == 1: + self.seq_list.append(seq) + return + + for idx, edge in enumerate(edges): + if edge[0] == j: + find_seq(edge[1], seq + [idx]) + + find_seq(0, []) + # print(f'self.seq_list: {self.seq_list}') + + for seq in self.seq_list: + if last_pool: + self.pooling_list.append(seq) + continue + if len(seq) % 2 == 1: + self.pooling_list.append([seq[0]]) + self.new_edges.append(edges[seq[0]]) + seq = seq[1:] + for i in range(0, len(seq), 2): + self.pooling_list.append([seq[i], seq[i + 1]]) + self.new_edges.append([edges[seq[i]][0], edges[seq[i + 1]][1]]) + # print(f'self.pooling_list: {self.pooling_list}') + # print(f'self.new_egdes: {self.new_edges}') + + # add global position + # self.pooling_list.append([self.edge_num - 1]) + + self.description = 'SkeletonPool(in_edge_num={}, out_edge_num={})'.format( + len(edges), len(self.pooling_list) + ) + + self.weight = torch.zeros(len(self.pooling_list) * channels_per_edge, self.edge_num * channels_per_edge) + + for i, pair in enumerate(self.pooling_list): + for j in pair: + for c in range(channels_per_edge): + self.weight[i * channels_per_edge + c, j * channels_per_edge + c] = 1.0 / len(pair) + + self.weight = nn.Parameter(self.weight, requires_grad=False) + + def forward(self, input: torch.Tensor): + # print('SkeletonPool') + # print(f'input: {input.size()}') + # print(f'self.weight: {self.weight.size()}') + return torch.matmul(self.weight, input) + + +class SkeletonUnpool(nn.Module): + def __init__(self, pooling_list, channels_per_edge): + super(SkeletonUnpool, self).__init__() + self.pooling_list = pooling_list + self.input_edge_num = len(pooling_list) + self.output_edge_num = 0 + self.channels_per_edge = channels_per_edge + for t in self.pooling_list: + self.output_edge_num += len(t) + + self.description = 'SkeletonUnpool(in_edge_num={}, out_edge_num={})'.format( + self.input_edge_num, self.output_edge_num, + ) + + self.weight = torch.zeros(self.output_edge_num * channels_per_edge, self.input_edge_num * channels_per_edge) + + for i, pair in enumerate(self.pooling_list): + for j in pair: + for c in range(channels_per_edge): + self.weight[j * channels_per_edge + c, i * channels_per_edge + c] = 1 + + self.weight = nn.Parameter(self.weight) + self.weight.requires_grad_(False) + + def forward(self, input: torch.Tensor): + # print('SkeletonUnpool') + # print(f'input: {input.size()}') + # print(f'self.weight: {self.weight.size()}') + return torch.matmul(self.weight, input) + + +""" +Helper functions for skeleton operation +""" + + +def dfs(x, fa, vis, dist): + vis[x] = 1 + for y in range(len(fa)): + if (fa[y] == x or fa[x] == y) and vis[y] == 0: + dist[y] = dist[x] + 1 + dfs(y, fa, vis, dist) + + +""" +def find_neighbor_joint(fa, threshold): + neighbor_list = [[]] + for x in range(1, len(fa)): + vis = [0 for _ in range(len(fa))] + dist = [0 for _ in range(len(fa))] + dist[0] = 10000 + dfs(x, fa, vis, dist) + neighbor = [] + for j in range(1, len(fa)): + if dist[j] <= threshold: + neighbor.append(j) + neighbor_list.append(neighbor) + + neighbor = [0] + for i, x in enumerate(neighbor_list): + if i == 0: continue + if 1 in x: + neighbor.append(i) + neighbor_list[i] = [0] + neighbor_list[i] + neighbor_list[0] = neighbor + return neighbor_list + + +def build_edge_topology(topology, offset): + # get all edges (pa, child, offset) + edges = [] + joint_num = len(topology) + for i in range(1, joint_num): + edges.append((topology[i], i, offset[i])) + return edges +""" + + +def build_edge_topology(topology): + # get all edges (pa, child) + edges = [] + joint_num = len(topology) + edges.append((0, joint_num)) # add an edge between the root joint and a virtual joint + for i in range(1, joint_num): + edges.append((topology[i], i)) + return edges + + +def build_joint_topology(edges, origin_names): + parent = [] + offset = [] + names = [] + edge2joint = [] + joint_from_edge = [] # -1 means virtual joint + joint_cnt = 0 + out_degree = [0] * (len(edges) + 10) + for edge in edges: + out_degree[edge[0]] += 1 + + # add root joint + joint_from_edge.append(-1) + parent.append(0) + offset.append(np.array([0, 0, 0])) + names.append(origin_names[0]) + joint_cnt += 1 + + def make_topology(edge_idx, pa): + nonlocal edges, parent, offset, names, edge2joint, joint_from_edge, joint_cnt + edge = edges[edge_idx] + if out_degree[edge[0]] > 1: + parent.append(pa) + offset.append(np.array([0, 0, 0])) + names.append(origin_names[edge[1]] + '_virtual') + edge2joint.append(-1) + pa = joint_cnt + joint_cnt += 1 + + parent.append(pa) + offset.append(edge[2]) + names.append(origin_names[edge[1]]) + edge2joint.append(edge_idx) + pa = joint_cnt + joint_cnt += 1 + + for idx, e in enumerate(edges): + if e[0] == edge[1]: + make_topology(idx, pa) + + for idx, e in enumerate(edges): + if e[0] == 0: + make_topology(idx, 0) + + return parent, offset, names, edge2joint + + +def calc_edge_mat(edges): + edge_num = len(edges) + # edge_mat[i][j] = distance between edge(i) and edge(j) + edge_mat = [[100000] * edge_num for _ in range(edge_num)] + for i in range(edge_num): + edge_mat[i][i] = 0 + + # initialize edge_mat with direct neighbor + for i, a in enumerate(edges): + for j, b in enumerate(edges): + link = 0 + for x in range(2): + for y in range(2): + if a[x] == b[y]: + link = 1 + if link: + edge_mat[i][j] = 1 + + # calculate all the pairs distance + for k in range(edge_num): + for i in range(edge_num): + for j in range(edge_num): + edge_mat[i][j] = min(edge_mat[i][j], edge_mat[i][k] + edge_mat[k][j]) + return edge_mat + + +def find_neighbor(edges, d): + """ + Args: + edges: The list contains N elements, each element represents (parent, child). + d: Distance between edges (the distance of the same edge is 0 and the distance of adjacent edges is 1). + + Returns: + The list contains N elements, each element is a list of edge indices whose distance <= d. + """ + edge_mat = calc_edge_mat(edges) + neighbor_list = [] + edge_num = len(edge_mat) + for i in range(edge_num): + neighbor = [] + for j in range(edge_num): + if edge_mat[i][j] <= d: + neighbor.append(j) + neighbor_list.append(neighbor) + + # # add neighbor for global part + # global_part_neighbor = neighbor_list[0].copy() + # """ + # Line #373 is buggy. Thanks @crissallan!! + # See issue #30 (https://github.com/DeepMotionEditing/deep-motion-editing/issues/30) + # However, fixing this bug will make it unable to load the pretrained model and + # affect the reproducibility of quantitative error reported in the paper. + # It is not a fatal bug so we didn't touch it and we are looking for possible solutions. + # """ + # for i in global_part_neighbor: + # neighbor_list[i].append(edge_num) + # neighbor_list.append(global_part_neighbor) + + return neighbor_list + + +def calc_node_depth(topology): + def dfs(node, topology): + if topology[node] < 0: + return 0 + return 1 + dfs(topology[node], topology) + depth = [] + for i in range(len(topology)): + depth.append(dfs(i, topology)) + + return depth + + +def residual_ratio(k): + return 1 / (k + 1) + + +class Affine(nn.Module): + def __init__(self, num_parameters, scale=True, bias=True, scale_init=1.0): + super(Affine, self).__init__() + if scale: + self.scale = nn.Parameter(torch.ones(num_parameters) * scale_init) + else: + self.register_parameter('scale', None) + + if bias: + self.bias = nn.Parameter(torch.zeros(num_parameters)) + else: + self.register_parameter('bias', None) + + def forward(self, input): + output = input + if self.scale is not None: + scale = self.scale.unsqueeze(0) + while scale.dim() < input.dim(): + scale = scale.unsqueeze(2) + output = output.mul(scale) + + if self.bias is not None: + bias = self.bias.unsqueeze(0) + while bias.dim() < input.dim(): + bias = bias.unsqueeze(2) + output += bias + + return output + + +class BatchStatistics(nn.Module): + def __init__(self, affine=-1): + super(BatchStatistics, self).__init__() + self.affine = nn.Sequential() if affine == -1 else Affine(affine) + self.loss = 0 + + def clear_loss(self): + self.loss = 0 + + def compute_loss(self, input): + input_flat = input.view(input.size(1), input.numel() // input.size(1)) + mu = input_flat.mean(1) + logvar = (input_flat.pow(2).mean(1) - mu.pow(2)).sqrt().log() + + self.loss = mu.pow(2).mean() + logvar.pow(2).mean() + + def forward(self, input): + self.compute_loss(input) + return self.affine(input) + + +class ResidualBlock(nn.Module): + def __init__(self, in_channels, out_channels, kernel_size, stride, padding, residual_ratio, activation, batch_statistics=False, last_layer=False): + super(ResidualBlock, self).__init__() + + self.residual_ratio = residual_ratio + self.shortcut_ratio = 1 - residual_ratio + + residual = [] + residual.append(nn.Conv1d(in_channels, out_channels, kernel_size, stride, padding)) + if batch_statistics: + residual.append(BatchStatistics(out_channels)) + if not last_layer: + residual.append(nn.PReLU() if activation == 'relu' else nn.Tanh()) + self.residual = nn.Sequential(*residual) + + self.shortcut = nn.Sequential( + nn.AvgPool1d(kernel_size=2) if stride == 2 else nn.Sequential(), + nn.Conv1d(in_channels, out_channels, kernel_size=1, stride=1, padding=0), + BatchStatistics(out_channels) if (in_channels != out_channels and batch_statistics is True) else nn.Sequential() + ) + + def forward(self, input): + return self.residual(input).mul(self.residual_ratio) + self.shortcut(input).mul(self.shortcut_ratio) + + +class ResidualBlockTranspose(nn.Module): + def __init__(self, in_channels, out_channels, kernel_size, stride, padding, residual_ratio, activation): + super(ResidualBlockTranspose, self).__init__() + + self.residual_ratio = residual_ratio + self.shortcut_ratio = 1 - residual_ratio + + self.residual = nn.Sequential( + nn.ConvTranspose1d(in_channels, out_channels, kernel_size, stride, padding), + nn.PReLU() if activation == 'relu' else nn.Tanh() + ) + + self.shortcut = nn.Sequential( + nn.Upsample(scale_factor=2, mode='linear', align_corners=False) if stride == 2 else nn.Sequential(), + nn.Conv1d(in_channels, out_channels, kernel_size=1, stride=1, padding=0) + ) + + def forward(self, input): + return self.residual(input).mul(self.residual_ratio) + self.shortcut(input).mul(self.shortcut_ratio) + + +class SkeletonResidual(nn.Module): + def __init__(self, topology, neighbour_list, joint_num, in_channels, out_channels, kernel_size, stride, padding, padding_mode, bias, extra_conv, pooling_mode, activation, last_pool): + super(SkeletonResidual, self).__init__() + + kernel_even = False if kernel_size % 2 else True + + seq = [] + for _ in range(extra_conv): + # (T, J, D) => (T, J, D) + seq.append(SkeletonConv(neighbour_list, in_channels=in_channels, out_channels=in_channels, + joint_num=joint_num, kernel_size=kernel_size - 1 if kernel_even else kernel_size, + stride=1, + padding=padding, padding_mode=padding_mode, bias=bias)) + seq.append(nn.PReLU() if activation == 'relu' else nn.Tanh()) + # (T, J, D) => (T/2, J, 2D) + seq.append(SkeletonConv(neighbour_list, in_channels=in_channels, out_channels=out_channels, + joint_num=joint_num, kernel_size=kernel_size, stride=stride, + padding=padding, padding_mode=padding_mode, bias=bias, add_offset=False)) + seq.append(nn.GroupNorm(10, out_channels)) # FIXME: REMEMBER TO CHANGE BACK !!! + self.residual = nn.Sequential(*seq) + + # (T, J, D) => (T/2, J, 2D) + self.shortcut = SkeletonConv(neighbour_list, in_channels=in_channels, out_channels=out_channels, + joint_num=joint_num, kernel_size=1, stride=stride, padding=0, + bias=True, add_offset=False) + + seq = [] + # (T/2, J, 2D) => (T/2, J', 2D) + pool = SkeletonPool(edges=topology, pooling_mode=pooling_mode, + channels_per_edge=out_channels // len(neighbour_list), last_pool=last_pool) + if len(pool.pooling_list) != pool.edge_num: + seq.append(pool) + seq.append(nn.PReLU() if activation == 'relu' else nn.Tanh()) + self.common = nn.Sequential(*seq) + + def forward(self, input): + output = self.residual(input) + self.shortcut(input) + + return self.common(output) + + +class SkeletonResidualTranspose(nn.Module): + def __init__(self, neighbour_list, joint_num, in_channels, out_channels, kernel_size, padding, padding_mode, bias, extra_conv, pooling_list, upsampling, activation, last_layer): + super(SkeletonResidualTranspose, self).__init__() + + kernel_even = False if kernel_size % 2 else True + + seq = [] + # (T, J, D) => (2T, J, D) + if upsampling is not None: + seq.append(nn.Upsample(scale_factor=2, mode=upsampling, align_corners=False)) + # (2T, J, D) => (2T, J', D) + unpool = SkeletonUnpool(pooling_list, in_channels // len(neighbour_list)) + if unpool.input_edge_num != unpool.output_edge_num: + seq.append(unpool) + self.common = nn.Sequential(*seq) + + seq = [] + for _ in range(extra_conv): + # (2T, J', D) => (2T, J', D) + seq.append(SkeletonConv(neighbour_list, in_channels=in_channels, out_channels=in_channels, + joint_num=joint_num, kernel_size=kernel_size - 1 if kernel_even else kernel_size, + stride=1, + padding=padding, padding_mode=padding_mode, bias=bias)) + seq.append(nn.PReLU() if activation == 'relu' else nn.Tanh()) + # (2T, J', D) => (2T, J', D/2) + seq.append(SkeletonConv(neighbour_list, in_channels=in_channels, out_channels=out_channels, + joint_num=joint_num, kernel_size=kernel_size - 1 if kernel_even else kernel_size, + stride=1, + padding=padding, padding_mode=padding_mode, bias=bias, add_offset=False)) + self.residual = nn.Sequential(*seq) + + # (2T, J', D) => (2T, J', D/2) + self.shortcut = SkeletonConv(neighbour_list, in_channels=in_channels, out_channels=out_channels, + joint_num=joint_num, kernel_size=1, stride=1, padding=0, + bias=True, add_offset=False) + + if activation == 'relu': + self.activation = nn.PReLU() if not last_layer else None + else: + self.activation = nn.Tanh() if not last_layer else None + + def forward(self, input): + output = self.common(input) + output = self.residual(output) + self.shortcut(output) + + if self.activation is not None: + return self.activation(output) + else: + return output \ No newline at end of file diff --git a/models/utils/wav2vec.py b/models/utils/wav2vec.py new file mode 100644 index 0000000000000000000000000000000000000000..ca23fe1d5a03834986885ed776cbf83c29e391ea --- /dev/null +++ b/models/utils/wav2vec.py @@ -0,0 +1,150 @@ +import torch +import torch.nn as nn +import torch.nn.functional as F +import numpy as np +import copy +import math +from transformers import Wav2Vec2Model,Wav2Vec2Config +from transformers.modeling_outputs import BaseModelOutput +from typing import Optional, Tuple +_CONFIG_FOR_DOC = "Wav2Vec2Config" + +# the implementation of Wav2Vec2Model is borrowed from https://huggingface.co/transformers/_modules/transformers/models/wav2vec2/modeling_wav2vec2.html#Wav2Vec2Model +# initialize our encoder with the pre-trained wav2vec 2.0 weights. +def _compute_mask_indices( + shape: Tuple[int, int], + mask_prob: float, + mask_length: int, + attention_mask: Optional[torch.Tensor] = None, + min_masks: int = 0, +) -> np.ndarray: + bsz, all_sz = shape + mask = np.full((bsz, all_sz), False) + + all_num_mask = int( + mask_prob * all_sz / float(mask_length) + + np.random.rand() + ) + all_num_mask = max(min_masks, all_num_mask) + mask_idcs = [] + padding_mask = attention_mask.ne(1) if attention_mask is not None else None + for i in range(bsz): + if padding_mask is not None: + sz = all_sz - padding_mask[i].long().sum().item() + num_mask = int( + mask_prob * sz / float(mask_length) + + np.random.rand() + ) + num_mask = max(min_masks, num_mask) + else: + sz = all_sz + num_mask = all_num_mask + + lengths = np.full(num_mask, mask_length) + + if sum(lengths) == 0: + lengths[0] = min(mask_length, sz - 1) + + min_len = min(lengths) + if sz - min_len <= num_mask: + min_len = sz - num_mask - 1 + + mask_idc = np.random.choice(sz - min_len, num_mask, replace=False) + mask_idc = np.asarray([mask_idc[j] + offset for j in range(len(mask_idc)) for offset in range(lengths[j])]) + mask_idcs.append(np.unique(mask_idc[mask_idc < sz])) + + min_len = min([len(m) for m in mask_idcs]) + for i, mask_idc in enumerate(mask_idcs): + if len(mask_idc) > min_len: + mask_idc = np.random.choice(mask_idc, min_len, replace=False) + mask[i, mask_idc] = True + return mask + +# linear interpolation layer +def linear_interpolation(features, input_fps, output_fps, output_len=None): + features = features.transpose(1, 2) + seq_len = features.shape[2] / float(input_fps) + if output_len is None: + output_len = int(seq_len * output_fps) + output_features = F.interpolate(features,size=output_len,align_corners=True,mode='linear') + return output_features.transpose(1, 2) + +class Wav2Vec2Model(Wav2Vec2Model): + def __init__(self, config): + super().__init__(config) + self.args = config + self.args.audio_fps = 15 #args.audio_fps + #input_values 16K hz, 49fps, 20ms overlap, 25ms recepion field + def forward( + self, + input_values, + dataset="beat", + attention_mask=None, + output_attentions=None, + output_hidden_states=None, + return_dict=None, + frame_num=None + ): + #print(input_values.shape) + self.config.output_attentions = True + output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions + output_hidden_states = ( + output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states + ) + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + + hidden_states = self.feature_extractor(input_values) + hidden_states = hidden_states.transpose(1, 2) + #print(hidden_states.shape) + if dataset == "beat": + hidden_states = linear_interpolation(hidden_states, 49, self.args.audio_fps, output_len=frame_num) + #print(hidden_states.shape) + if attention_mask is not None: + output_lengths = self._get_feat_extract_output_lengths(attention_mask.sum(-1)) + attention_mask = torch.zeros( + hidden_states.shape[:2], dtype=hidden_states.dtype, device=hidden_states.device + ) + attention_mask[ + (torch.arange(attention_mask.shape[0], device=hidden_states.device), output_lengths - 1) + ] = 1 + attention_mask = attention_mask.flip([-1]).cumsum(-1).flip([-1]).bool() + + hidden_states = self.feature_projection(hidden_states)[0] + #print(hidden_states.shape) + if self.config.apply_spec_augment and self.training: + batch_size, sequence_length, hidden_size = hidden_states.size() + if self.config.mask_time_prob > 0: + mask_time_indices = _compute_mask_indices( + (batch_size, sequence_length), + self.config.mask_time_prob, + self.config.mask_time_length, + attention_mask=attention_mask, + min_masks=2, + ) + hidden_states[torch.from_numpy(mask_time_indices)] = self.masked_spec_embed.to(hidden_states.dtype) + if self.config.mask_feature_prob > 0: + mask_feature_indices = _compute_mask_indices( + (batch_size, hidden_size), + self.config.mask_feature_prob, + self.config.mask_feature_length, + ) + mask_feature_indices = torch.from_numpy(mask_feature_indices).to(hidden_states.device) + hidden_states[mask_feature_indices[:, None].expand(-1, sequence_length, -1)] = 0 + encoder_outputs = self.encoder( + hidden_states, + attention_mask=attention_mask, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + ) + hidden_states = encoder_outputs[0] + #print(encoder_outputs.shape) + if not return_dict: + return (hidden_states,) + encoder_outputs[1:] + + return hidden_states +# BaseModelOutput( +# last_hidden_state=hidden_states, +# hidden_states=encoder_outputs.hidden_states, +# attentions=encoder_outputs.attentions, +# ) \ No newline at end of file diff --git a/models/vq/__init__.py b/models/vq/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/models/vq/encdec.py b/models/vq/encdec.py new file mode 100644 index 0000000000000000000000000000000000000000..0ee18696f1051439ebb6d73a746ea811a5e17973 --- /dev/null +++ b/models/vq/encdec.py @@ -0,0 +1,68 @@ +import torch.nn as nn +from models.vq.resnet import Resnet1D + + +class Encoder(nn.Module): + def __init__(self, + input_emb_width=3, + output_emb_width=512, + down_t=2, + stride_t=2, + width=512, + depth=3, + dilation_growth_rate=3, + activation='relu', + norm=None): + super().__init__() + + blocks = [] + filter_t, pad_t = stride_t * 2, stride_t // 2 + blocks.append(nn.Conv1d(input_emb_width, width, 3, 1, 1)) + blocks.append(nn.ReLU()) + + for i in range(down_t): + input_dim = width + block = nn.Sequential( + nn.Conv1d(input_dim, width, filter_t, stride_t, pad_t), + Resnet1D(width, depth, dilation_growth_rate, activation=activation, norm=norm), + ) + blocks.append(block) + blocks.append(nn.Conv1d(width, output_emb_width, 3, 1, 1)) + self.model = nn.Sequential(*blocks) + + def forward(self, x): + return self.model(x) + + +class Decoder(nn.Module): + def __init__(self, + input_emb_width=3, + output_emb_width=512, + down_t=2, + stride_t=2, + width=512, + depth=3, + dilation_growth_rate=3, + activation='relu', + norm=None): + super().__init__() + blocks = [] + + blocks.append(nn.Conv1d(output_emb_width, width, 3, 1, 1)) + blocks.append(nn.ReLU()) + for i in range(down_t): + out_dim = width + block = nn.Sequential( + Resnet1D(width, depth, dilation_growth_rate, reverse_dilation=True, activation=activation, norm=norm), + nn.Upsample(scale_factor=2, mode='nearest'), + nn.Conv1d(width, out_dim, 3, 1, 1) + ) + blocks.append(block) + blocks.append(nn.Conv1d(width, width, 3, 1, 1)) + blocks.append(nn.ReLU()) + blocks.append(nn.Conv1d(width, input_emb_width, 3, 1, 1)) + self.model = nn.Sequential(*blocks) + + def forward(self, x): + x = self.model(x) + return x.permute(0, 2, 1) \ No newline at end of file diff --git a/models/vq/model.py b/models/vq/model.py new file mode 100644 index 0000000000000000000000000000000000000000..404e32ac84fabf8a974e9b1dc4242a74c5429772 --- /dev/null +++ b/models/vq/model.py @@ -0,0 +1,146 @@ +import random + +import torch.nn as nn +from models.vq.encdec import Encoder, Decoder +from models.vq.residual_vq import ResidualVQ + +class RVQVAE(nn.Module): + def __init__(self, + args, + input_width=263, + nb_code=1024, + code_dim=512, + output_emb_width=512, + down_t=3, + stride_t=2, + width=512, + depth=3, + dilation_growth_rate=3, + activation='relu', + norm=None): + + super().__init__() + assert output_emb_width == code_dim + self.code_dim = code_dim + self.num_code = nb_code + # self.quant = args.quantizer + self.encoder = Encoder(input_width, output_emb_width, down_t, stride_t, width, depth, + dilation_growth_rate, activation=activation, norm=norm) + self.decoder = Decoder(input_width, output_emb_width, down_t, stride_t, width, depth, + dilation_growth_rate, activation=activation, norm=norm) + rvqvae_config = { + 'num_quantizers': args.num_quantizers, + 'shared_codebook': args.shared_codebook, + 'quantize_dropout_prob': args.quantize_dropout_prob, + 'quantize_dropout_cutoff_index': 0, + 'nb_code': nb_code, + 'code_dim':code_dim, + 'args': args, + } + self.quantizer = ResidualVQ(**rvqvae_config) + + def preprocess(self, x): + # (bs, T, Jx3) -> (bs, Jx3, T) + x = x.permute(0, 2, 1).float() + return x + + def postprocess(self, x): + # (bs, Jx3, T) -> (bs, T, Jx3) + x = x.permute(0, 2, 1) + return x + + def encode(self, x): + N, T, _ = x.shape + x_in = self.preprocess(x) + x_encoder = self.encoder(x_in) + # print(x_encoder.shape) + code_idx, all_codes = self.quantizer.quantize(x_encoder, return_latent=True) + # print(code_idx.shape) + # code_idx = code_idx.view(N, -1) + # (N, T, Q) + # print() + return code_idx, all_codes + + def forward(self, x): + x_in = self.preprocess(x) + # Encode + x_encoder = self.encoder(x_in) + + ## quantization + # x_quantized, code_idx, commit_loss, perplexity = self.quantizer(x_encoder, sample_codebook_temp=0.5, + # force_dropout_index=0) #TODO hardcode + x_quantized, code_idx, commit_loss, perplexity = self.quantizer(x_encoder, sample_codebook_temp=0.5) + + # print(code_idx[0, :, 1]) + ## decoder + x_out = self.decoder(x_quantized) + # x_out = self.postprocess(x_decoder) + return { + 'rec_pose': x_out, + 'commit_loss': commit_loss, + 'perplexity': perplexity, + } + + + def forward_decoder(self, x): + x_d = self.quantizer.get_codes_from_indices(x) + # x_d = x_d.view(1, -1, self.code_dim).permute(0, 2, 1).contiguous() + x = x_d.sum(dim=0).permute(0, 2, 1) + + # decoder + x_out = self.decoder(x) + # x_out = self.postprocess(x_decoder) + return x_out + + def map2latent(self,x): + x_in = self.preprocess(x) + # Encode + x_encoder = self.encoder(x_in) + x_encoder = x_encoder.permute(0,2,1) + return x_encoder + + def latent2origin(self,x): + x = x.permute(0,2,1) + x_quantized, code_idx, commit_loss, perplexity = self.quantizer(x, sample_codebook_temp=0.5) + # print(code_idx[0, :, 1]) + ## decoder + x_out = self.decoder(x_quantized) + # x_out = self.postprocess(x_decoder) + return x_out, commit_loss, perplexity + + +class LengthEstimator(nn.Module): + def __init__(self, input_size, output_size): + super(LengthEstimator, self).__init__() + nd = 512 + self.output = nn.Sequential( + nn.Linear(input_size, nd), + nn.LayerNorm(nd), + nn.LeakyReLU(0.2, inplace=True), + + nn.Dropout(0.2), + nn.Linear(nd, nd // 2), + nn.LayerNorm(nd // 2), + nn.LeakyReLU(0.2, inplace=True), + + nn.Dropout(0.2), + nn.Linear(nd // 2, nd // 4), + nn.LayerNorm(nd // 4), + nn.LeakyReLU(0.2, inplace=True), + + nn.Linear(nd // 4, output_size) + ) + + self.output.apply(self.__init_weights) + + def __init_weights(self, module): + if isinstance(module, (nn.Linear, nn.Embedding)): + module.weight.data.normal_(mean=0.0, std=0.02) + if isinstance(module, nn.Linear) and module.bias is not None: + module.bias.data.zero_() + elif isinstance(module, nn.LayerNorm): + module.bias.data.zero_() + module.weight.data.fill_(1.0) + + def forward(self, text_emb): + return self.output(text_emb) \ No newline at end of file diff --git a/models/vq/quantizer.py b/models/vq/quantizer.py new file mode 100644 index 0000000000000000000000000000000000000000..23ebff7dad1e1342e442a17342678f240cb31123 --- /dev/null +++ b/models/vq/quantizer.py @@ -0,0 +1,180 @@ +import numpy as np +import torch +import torch.nn as nn +import torch.nn.functional as F +from einops import rearrange, repeat, reduce, pack, unpack + +# from vector_quantize_pytorch import ResidualVQ + +#Borrow from vector_quantize_pytorch + +def log(t, eps = 1e-20): + return torch.log(t.clamp(min = eps)) + +def gumbel_noise(t): + noise = torch.zeros_like(t).uniform_(0, 1) + return -log(-log(noise)) + +def gumbel_sample( + logits, + temperature = 1., + stochastic = False, + dim = -1, + training = True +): + + if training and stochastic and temperature > 0: + sampling_logits = (logits / temperature) + gumbel_noise(logits) + else: + sampling_logits = logits + + ind = sampling_logits.argmax(dim = dim) + + return ind + +class QuantizeEMAReset(nn.Module): + def __init__(self, nb_code, code_dim, args): + super(QuantizeEMAReset, self).__init__() + self.nb_code = nb_code + self.code_dim = code_dim + self.mu = args.mu ##TO_DO + self.reset_codebook() + + def reset_codebook(self): + self.init = False + self.code_sum = None + self.code_count = None + self.register_buffer('codebook', torch.zeros(self.nb_code, self.code_dim, requires_grad=False).cuda()) + + def _tile(self, x): + nb_code_x, code_dim = x.shape + if nb_code_x < self.nb_code: + n_repeats = (self.nb_code + nb_code_x - 1) // nb_code_x + std = 0.01 / np.sqrt(code_dim) + out = x.repeat(n_repeats, 1) + out = out + torch.randn_like(out) * std + else: + out = x + return out + + def init_codebook(self, x): + out = self._tile(x) + self.codebook = out[:self.nb_code] + self.code_sum = self.codebook.clone() + self.code_count = torch.ones(self.nb_code, device=self.codebook.device) + self.init = True + + def quantize(self, x, sample_codebook_temp=0.): + # N X C -> C X N + k_w = self.codebook.t() + # x: NT X C + # NT X N + distance = torch.sum(x ** 2, dim=-1, keepdim=True) - \ + 2 * torch.matmul(x, k_w) + \ + torch.sum(k_w ** 2, dim=0, keepdim=True) # (N * L, b) + + # code_idx = torch.argmin(distance, dim=-1) + + code_idx = gumbel_sample(-distance, dim = -1, temperature = sample_codebook_temp, stochastic=True, training = self.training) + + return code_idx + + def dequantize(self, code_idx): + x = F.embedding(code_idx, self.codebook) + return x + + def get_codebook_entry(self, indices): + return self.dequantize(indices).permute(0, 2, 1) + + @torch.no_grad() + def compute_perplexity(self, code_idx): + # Calculate new centres + code_onehot = torch.zeros(self.nb_code, code_idx.shape[0], device=code_idx.device) # nb_code, N * L + code_onehot.scatter_(0, code_idx.view(1, code_idx.shape[0]), 1) + + code_count = code_onehot.sum(dim=-1) # nb_code + prob = code_count / torch.sum(code_count) + perplexity = torch.exp(-torch.sum(prob * torch.log(prob + 1e-7))) + return perplexity + + @torch.no_grad() + def update_codebook(self, x, code_idx): + code_onehot = torch.zeros(self.nb_code, x.shape[0], device=x.device) # nb_code, N * L + code_onehot.scatter_(0, code_idx.view(1, x.shape[0]), 1) + + code_sum = torch.matmul(code_onehot, x) # nb_code, c + code_count = code_onehot.sum(dim=-1) # nb_code + + out = self._tile(x) + code_rand = out[:self.nb_code] + + # Update centres + self.code_sum = self.mu * self.code_sum + (1. - self.mu) * code_sum + self.code_count = self.mu * self.code_count + (1. - self.mu) * code_count + + usage = (self.code_count.view(self.nb_code, 1) >= 1.0).float() + code_update = self.code_sum.view(self.nb_code, self.code_dim) / self.code_count.view(self.nb_code, 1) + self.codebook = usage * code_update + (1-usage) * code_rand + + + prob = code_count / torch.sum(code_count) + perplexity = torch.exp(-torch.sum(prob * torch.log(prob + 1e-7))) + + return perplexity + + def preprocess(self, x): + # NCT -> NTC -> [NT, C] + # x = x.permute(0, 2, 1).contiguous() + # x = x.view(-1, x.shape[-1]) + x = rearrange(x, 'n c t -> (n t) c') + return x + + def forward(self, x, return_idx=False, temperature=0.): + N, width, T = x.shape + + x = self.preprocess(x) + if self.training and not self.init: + self.init_codebook(x) + + code_idx = self.quantize(x, temperature) + x_d = self.dequantize(code_idx) + + if self.training: + perplexity = self.update_codebook(x, code_idx) + else: + perplexity = self.compute_perplexity(code_idx) + + commit_loss = F.mse_loss(x, x_d.detach()) # It's right. the t2m-gpt paper is wrong on embed loss and commitment loss. + + # Passthrough + x_d = x + (x_d - x).detach() + + # Postprocess + x_d = x_d.view(N, T, -1).permute(0, 2, 1).contiguous() + code_idx = code_idx.view(N, T).contiguous() + # print(code_idx[0]) + if return_idx: + return x_d, code_idx, commit_loss, perplexity + return x_d, commit_loss, perplexity + +class QuantizeEMA(QuantizeEMAReset): + @torch.no_grad() + def update_codebook(self, x, code_idx): + code_onehot = torch.zeros(self.nb_code, x.shape[0], device=x.device) # nb_code, N * L + code_onehot.scatter_(0, code_idx.view(1, x.shape[0]), 1) + + code_sum = torch.matmul(code_onehot, x) # nb_code, c + code_count = code_onehot.sum(dim=-1) # nb_code + + # Update centres + self.code_sum = self.mu * self.code_sum + (1. - self.mu) * code_sum + self.code_count = self.mu * self.code_count + (1. - self.mu) * code_count + + usage = (self.code_count.view(self.nb_code, 1) >= 1.0).float() + code_update = self.code_sum.view(self.nb_code, self.code_dim) / self.code_count.view(self.nb_code, 1) + self.codebook = usage * code_update + (1-usage) * self.codebook + + prob = code_count / torch.sum(code_count) + perplexity = torch.exp(-torch.sum(prob * torch.log(prob + 1e-7))) + + return perplexity diff --git a/models/vq/residual_vq.py b/models/vq/residual_vq.py new file mode 100644 index 0000000000000000000000000000000000000000..6478260124d174ad35f539a612859ce749983fb8 --- /dev/null +++ b/models/vq/residual_vq.py @@ -0,0 +1,194 @@ +import random +from math import ceil +from functools import partial +from itertools import zip_longest +from random import randrange + +import torch +from torch import nn +import torch.nn.functional as F +# from vector_quantize_pytorch.vector_quantize_pytorch import VectorQuantize +from models.vq.quantizer import QuantizeEMAReset, QuantizeEMA + +from einops import rearrange, repeat, pack, unpack + +# helper functions + +def exists(val): + return val is not None + +def default(val, d): + return val if exists(val) else d + +def round_up_multiple(num, mult): + return ceil(num / mult) * mult + +# main class + +class ResidualVQ(nn.Module): + """ Follows Algorithm 1. in https://arxiv.org/pdf/2107.03312.pdf """ + def __init__( + self, + num_quantizers, + shared_codebook=False, + quantize_dropout_prob=0.5, + quantize_dropout_cutoff_index=0, + **kwargs + ): + super().__init__() + + self.num_quantizers = num_quantizers + + # self.layers = nn.ModuleList([VectorQuantize(accept_image_fmap = accept_image_fmap, **kwargs) for _ in range(num_quantizers)]) + if shared_codebook: + layer = QuantizeEMAReset(**kwargs) + self.layers = nn.ModuleList([layer for _ in range(num_quantizers)]) + else: + self.layers = nn.ModuleList([QuantizeEMAReset(**kwargs) for _ in range(num_quantizers)]) + # self.layers = nn.ModuleList([QuantizeEMA(**kwargs) for _ in range(num_quantizers)]) + + # self.quantize_dropout = quantize_dropout and num_quantizers > 1 + + assert quantize_dropout_cutoff_index >= 0 and quantize_dropout_prob >= 0 + + self.quantize_dropout_cutoff_index = quantize_dropout_cutoff_index + self.quantize_dropout_prob = quantize_dropout_prob + + + @property + def codebooks(self): + codebooks = [layer.codebook for layer in self.layers] + codebooks = torch.stack(codebooks, dim = 0) + return codebooks # 'q c d' + + def get_codes_from_indices(self, indices): #indices shape 'b n q' # dequantize + + batch, quantize_dim = indices.shape[0], indices.shape[-1] + + # because of quantize dropout, one can pass in indices that are coarse + # and the network should be able to reconstruct + + if quantize_dim < self.num_quantizers: + indices = F.pad(indices, (0, self.num_quantizers - quantize_dim), value = -1) + + # get ready for gathering + + codebooks = repeat(self.codebooks, 'q c d -> q b c d', b = batch) + gather_indices = repeat(indices, 'b n q -> q b n d', d = codebooks.shape[-1]) + + # take care of quantizer dropout + + mask = gather_indices == -1. + gather_indices = gather_indices.masked_fill(mask, 0) # have it fetch a dummy code to be masked out later + + # print(gather_indices.max(), gather_indices.min()) + all_codes = codebooks.gather(2, gather_indices) # gather all codes + + # mask out any codes that were dropout-ed + + all_codes = all_codes.masked_fill(mask, 0.) + + return all_codes # 'q b n d' + + def get_codebook_entry(self, indices): #indices shape 'b n q' + all_codes = self.get_codes_from_indices(indices) #'q b n d' + latent = torch.sum(all_codes, dim=0) #'b n d' + latent = latent.permute(0, 2, 1) + return latent + + def forward(self, x, return_all_codes = False, sample_codebook_temp = None, force_dropout_index=-1): + # debug check + # print(self.codebooks[:,0,0].detach().cpu().numpy()) + num_quant, quant_dropout_prob, device = self.num_quantizers, self.quantize_dropout_prob, x.device + + quantized_out = 0. + residual = x + + all_losses = [] + all_indices = [] + all_perplexity = [] + + + should_quantize_dropout = self.training and random.random() < self.quantize_dropout_prob + + start_drop_quantize_index = num_quant + # To ensure the first-k layers learn things as much as possible, we randomly dropout the last q - k layers + if should_quantize_dropout: + start_drop_quantize_index = randrange(self.quantize_dropout_cutoff_index, num_quant) # keep quant layers <= quantize_dropout_cutoff_index, TODO vary in batch + null_indices_shape = [x.shape[0], x.shape[-1]] # 'b*n' + null_indices = torch.full(null_indices_shape, -1., device = device, dtype = torch.long) + # null_loss = 0. + + if force_dropout_index >= 0: + should_quantize_dropout = True + start_drop_quantize_index = force_dropout_index + null_indices_shape = [x.shape[0], x.shape[-1]] # 'b*n' + null_indices = torch.full(null_indices_shape, -1., device=device, dtype=torch.long) + + # print(force_dropout_index) + # go through the layers + + for quantizer_index, layer in enumerate(self.layers): + + if should_quantize_dropout and quantizer_index > start_drop_quantize_index: + all_indices.append(null_indices) + # all_losses.append(null_loss) + continue + + # layer_indices = None + # if return_loss: + # layer_indices = indices[..., quantizer_index] #gt indices + + # quantized, *rest = layer(residual, indices = layer_indices, sample_codebook_temp = sample_codebook_temp) #single quantizer TODO + quantized, *rest = layer(residual, return_idx=True, temperature=sample_codebook_temp) #single quantizer + + # print(quantized.shape, residual.shape) + residual -= quantized.detach() + quantized_out += quantized + + embed_indices, loss, perplexity = rest + all_indices.append(embed_indices) + all_losses.append(loss) + all_perplexity.append(perplexity) + + + # stack all losses and indices + all_indices = torch.stack(all_indices, dim=-1) + all_losses = sum(all_losses)/len(all_losses) + all_perplexity = sum(all_perplexity)/len(all_perplexity) + + ret = (quantized_out, all_indices, all_losses, all_perplexity) + + if return_all_codes: + # whether to return all codes from all codebooks across layers + all_codes = self.get_codes_from_indices(all_indices) + + # will return all codes in shape (quantizer, batch, sequence length, codebook dimension) + ret = (*ret, all_codes) + + return ret + + def quantize(self, x, return_latent=False): + all_indices = [] + quantized_out = 0. + residual = x + all_codes = [] + for quantizer_index, layer in enumerate(self.layers): + + quantized, *rest = layer(residual, return_idx=True) #single quantizer + + residual = residual - quantized.detach() + quantized_out = quantized_out + quantized + + embed_indices, loss, perplexity = rest + all_indices.append(embed_indices) + # print(quantizer_index, embed_indices[0]) + # print(quantizer_index, quantized[0]) + # break + all_codes.append(quantized) + + code_idx = torch.stack(all_indices, dim=-1) + all_codes = torch.stack(all_codes, dim=0) + if return_latent: + return code_idx, all_codes + return code_idx \ No newline at end of file diff --git a/models/vq/resnet.py b/models/vq/resnet.py new file mode 100644 index 0000000000000000000000000000000000000000..88d230d31055a6e66b0c9b75b29fff8935b8a1fd --- /dev/null +++ b/models/vq/resnet.py @@ -0,0 +1,84 @@ +import torch.nn as nn +import torch + +class nonlinearity(nn.Module): + def __init(self): + super().__init__() + + def forward(self, x): + return x * torch.sigmoid(x) + + +class ResConv1DBlock(nn.Module): + def __init__(self, n_in, n_state, dilation=1, activation='silu', norm=None, dropout=0.2): + super(ResConv1DBlock, self).__init__() + + padding = dilation + self.norm = norm + + if norm == "LN": + self.norm1 = nn.LayerNorm(n_in) + self.norm2 = nn.LayerNorm(n_in) + elif norm == "GN": + self.norm1 = nn.GroupNorm(num_groups=32, num_channels=n_in, eps=1e-6, affine=True) + self.norm2 = nn.GroupNorm(num_groups=32, num_channels=n_in, eps=1e-6, affine=True) + elif norm == "BN": + self.norm1 = nn.BatchNorm1d(num_features=n_in, eps=1e-6, affine=True) + self.norm2 = nn.BatchNorm1d(num_features=n_in, eps=1e-6, affine=True) + else: + self.norm1 = nn.Identity() + self.norm2 = nn.Identity() + + if activation == "relu": + self.activation1 = nn.ReLU() + self.activation2 = nn.ReLU() + + elif activation == "silu": + self.activation1 = nonlinearity() + self.activation2 = nonlinearity() + + elif activation == "gelu": + self.activation1 = nn.GELU() + self.activation2 = nn.GELU() + + self.conv1 = nn.Conv1d(n_in, n_state, 3, 1, padding, dilation) + self.conv2 = nn.Conv1d(n_state, n_in, 1, 1, 0, ) + self.dropout = nn.Dropout(dropout) + + def forward(self, x): + x_orig = x + if self.norm == "LN": + x = self.norm1(x.transpose(-2, -1)) + x = self.activation1(x.transpose(-2, -1)) + else: + x = self.norm1(x) + x = self.activation1(x) + + x = self.conv1(x) + + if self.norm == "LN": + x = self.norm2(x.transpose(-2, -1)) + x = self.activation2(x.transpose(-2, -1)) + else: + x = self.norm2(x) + x = self.activation2(x) + + x = self.conv2(x) + x = self.dropout(x) + x = x + x_orig + return x + + +class Resnet1D(nn.Module): + def __init__(self, n_in, n_depth, dilation_growth_rate=1, reverse_dilation=True, activation='relu', norm=None): + super().__init__() + + blocks = [ResConv1DBlock(n_in, n_in, dilation=dilation_growth_rate ** depth, activation=activation, norm=norm) + for depth in range(n_depth)] + if reverse_dilation: + blocks = blocks[::-1] + + self.model = nn.Sequential(*blocks) + + def forward(self, x): + return self.model(x) \ No newline at end of file diff --git a/optimizers/__init__.py b/optimizers/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/optimizers/loss_factory.py b/optimizers/loss_factory.py new file mode 100644 index 0000000000000000000000000000000000000000..193728094efce456ddad37103206dc0e65a6df3f --- /dev/null +++ b/optimizers/loss_factory.py @@ -0,0 +1,118 @@ +# Copyright (c) HuaWei, Inc. and its affiliates. +# liu.haiyang@huawei.com + +import torch.nn as nn +import torch.nn.functional as F +import torch +import numpy as np + + +class GeodesicLoss(nn.Module): + def __init__(self): + super(GeodesicLoss, self).__init__() + + def compute_geodesic_distance(self, m1, m2): + """ Compute the geodesic distance between two rotation matrices. + + Args: + m1, m2: Two rotation matrices with the shape (batch x 3 x 3). + + Returns: + The minimal angular difference between two rotation matrices in radian form [0, pi]. + """ + m1 = m1.reshape(-1, 3, 3) + m2 = m2.reshape(-1, 3, 3) + batch = m1.shape[0] + m = torch.bmm(m1, m2.transpose(1, 2)) # batch*3*3 + + cos = (m[:, 0, 0] + m[:, 1, 1] + m[:, 2, 2] - 1) / 2 + cos = torch.clamp(cos, min=-1 + 1E-6, max=1-1E-6) + + theta = torch.acos(cos) + + return theta + + def __call__(self, m1, m2, reduction='mean'): + loss = self.compute_geodesic_distance(m1, m2) + + if reduction == 'mean': + return loss.mean() + elif reduction == 'none': + return loss + else: + raise RuntimeError(f'unsupported reduction: {reduction}') + + +class BCE_Loss(nn.Module): + def __init__(self, args=None): + super(BCE_Loss, self).__init__() + + def forward(self, fake_outputs, real_target): + final_loss = F.cross_entropy(fake_outputs, real_target, reduce="mean") + return final_loss + +class weight_Loss(nn.Module): + def __init__(self, args=None): + super(weight_Loss, self).__init__() + def forward(self, weight_f): + weight_loss_div = torch.mean(weight_f[:, :, 0]*weight_f[:, :, 1]) + weight_loss_gap = torch.mean(-torch.log(torch.max(weight_f[:, :, 0], dim=1)[0] - torch.min(weight_f[:, :, 0], dim=1)[0])) + return weight_loss_div, weight_loss_gap + + +class HuberLoss(nn.Module): + def __init__(self, beta=0.1, reduction="mean"): + super(HuberLoss, self).__init__() + self.beta = beta + self.reduction = reduction + + def forward(self, outputs, targets): + final_loss = F.smooth_l1_loss(outputs / self.beta, targets / self.beta, reduction=self.reduction) * self.beta + return final_loss + + +class KLDLoss(nn.Module): + def __init__(self, beta=0.1): + super(KLDLoss, self).__init__() + self.beta = beta + + def forward(self, outputs, targets): + final_loss = F.smooth_l1_loss((outputs / self.beta, targets / self.beta) * self.beta) + return final_loss + + +class REGLoss(nn.Module): + def __init__(self, beta=0.1): + super(REGLoss, self).__init__() + self.beta = beta + + def forward(self, outputs, targets): + final_loss = F.smooth_l1_loss((outputs / self.beta, targets / self.beta) * self.beta) + return final_loss + + +class L2Loss(nn.Module): + def __init__(self): + super(L2Loss, self).__init__() + + def forward(self, outputs, targets): + final_loss = F.l2_loss(outputs, targets) + return final_loss + +LOSS_FUNC_LUT = { + "bce_loss": BCE_Loss, + "l2_loss": L2Loss, + "huber_loss": HuberLoss, + "kl_loss": KLDLoss, + "id_loss": REGLoss, + "GeodesicLoss": GeodesicLoss, + "weight_Loss": weight_Loss, + } + + +def get_loss_func(loss_name, **kwargs): + loss_func_class = LOSS_FUNC_LUT.get(loss_name) + loss_func = loss_func_class(**kwargs) + return loss_func + + diff --git a/optimizers/optim_factory.py b/optimizers/optim_factory.py new file mode 100644 index 0000000000000000000000000000000000000000..5627baeb643328a06b1e5caa64dec1b219d93d22 --- /dev/null +++ b/optimizers/optim_factory.py @@ -0,0 +1,176 @@ +""" Optimizer Factory w/ Custom Weight Decay +Hacked together by / Copyright 2020 Ross Wightman +""" +from typing import Optional + +import torch +import torch.nn as nn +import torch.optim as optim + +from .timm.adafactor import Adafactor +from .timm.adahessian import Adahessian +from .timm.adamp import AdamP +from .timm.lookahead import Lookahead +from .timm.nadam import Nadam +from .timm.novograd import NovoGrad +from .timm.nvnovograd import NvNovoGrad +from .timm.radam import RAdam +from .timm.rmsprop_tf import RMSpropTF +from .timm.sgdp import SGDP +from .timm.adabelief import AdaBelief + +try: + from apex.optimizers import FusedNovoGrad, FusedAdam, FusedLAMB, FusedSGD + has_apex = True +except ImportError: + has_apex = False + + +def add_weight_decay(model, weight_decay=1e-5, skip_list=()): + decay = [] + no_decay = [] + for name, param in model.named_parameters(): + if not param.requires_grad: + continue # frozen weights + if len(param.shape) == 1 or name.endswith(".bias") or name in skip_list: + no_decay.append(param) + else: + decay.append(param) + return [ + {"params": no_decay, "weight_decay": 0.}, + {"params": decay, "weight_decay": weight_decay}] + + +def optimizer_kwargs(args, lr_weight): + """ args/argparse to kwargs helper + Convert optimizer args in argparse args or args like object to keyword args for updated create fn. + """ + kwargs = dict( + optimizer_name=args.opt, + learning_rate=args.lr_base,#*args.batch_size/128*lr_weight, + weight_decay=args.weight_decay, + momentum=args.momentum) + if getattr(args, "opt_eps", None) is not None: + kwargs["eps"] = args.opt_eps + if getattr(args, "opt_betas", None) is not None: + kwargs["betas"] = args.opt_betas + if getattr(args, "opt_args", None) is not None: + kwargs.update(args.opt_args) + return kwargs + + +def create_optimizer(args, model, filter_bias_and_bn=True, lr_weight=1): + """ Legacy optimizer factory for backwards compatibility. + NOTE: Use create_optimizer_v2 for new code. + """ + return create_optimizer_v2( + model, + **optimizer_kwargs(args, lr_weight), + filter_bias_and_bn=filter_bias_and_bn, + ) + + +def create_optimizer_v2( + model: nn.Module, + optimizer_name: str = "sgd", + learning_rate: Optional[float] = None, + weight_decay: float = 0., + momentum: float = 0.9, + filter_bias_and_bn: bool = True, + **kwargs): + """ Create an optimizer. + + TODO currently the model is passed in and all parameters are selected for optimization. + For more general use an interface that allows selection of parameters to optimize and lr groups, one of: + * a filter fn interface that further breaks params into groups in a weight_decay compatible fashion + * expose the parameters interface and leave it up to caller + + Args: + model (nn.Module): model containing parameters to optimize + optimizer_name: name of optimizer to create + learning_rate: initial learning rate + weight_decay: weight decay to apply in optimizer + momentum: momentum for momentum based optimizers (others may use betas via kwargs) + filter_bias_and_bn: filter out bias, bn and other 1d params from weight decay + **kwargs: extra optimizer specific kwargs to pass through + + Returns: + Optimizer + """ + opt_lower = optimizer_name.lower() + if weight_decay and filter_bias_and_bn: + skip = {} + if hasattr(model, "no_weight_decay"): + skip = model.no_weight_decay() + parameters = add_weight_decay(model, weight_decay, skip) + weight_decay = 0. + else: + parameters = model.parameters() + if "fused" in opt_lower: + assert has_apex and torch.cuda.is_available(), "APEX and CUDA required for fused optimizers" + + opt_args = dict(lr=learning_rate, weight_decay=weight_decay, **kwargs) + opt_split = opt_lower.split("_") + opt_lower = opt_split[-1] + if opt_lower == "sgd" or opt_lower == "nesterov": + opt_args.pop("eps", None) + optimizer = optim.SGD(parameters, momentum=momentum, nesterov=True, **opt_args) + elif opt_lower == "momentum": + opt_args.pop("eps", None) + optimizer = optim.SGD(parameters, momentum=momentum, nesterov=False, **opt_args) + + elif opt_lower == "adam": + optimizer = optim.Adam(parameters, **opt_args) + elif opt_lower == "adabelief": + optimizer = AdaBelief(parameters, rectify=False, **opt_args) + elif opt_lower == "adamw": + optimizer = optim.AdamW(parameters, lr=learning_rate, weight_decay=weight_decay) + + elif opt_lower == "nadam": + optimizer = Nadam(parameters, **opt_args) + elif opt_lower == "radam": + optimizer = RAdam(parameters, **opt_args) + elif opt_lower == "adamp": + optimizer = AdamP(parameters, wd_ratio=0.01, nesterov=True, **opt_args) + elif opt_lower == "sgdp": + optimizer = SGDP(parameters, momentum=momentum, nesterov=True, **opt_args) + elif opt_lower == "adadelta": + optimizer = optim.Adadelta(parameters, **opt_args) + elif opt_lower == "adafactor": + if not learning_rate: + opt_args["lr"] = None + optimizer = Adafactor(parameters, **opt_args) + elif opt_lower == "adahessian": + optimizer = Adahessian(parameters, **opt_args) + elif opt_lower == "rmsprop": + optimizer = optim.RMSprop(parameters, alpha=0.9, momentum=momentum, **opt_args) + elif opt_lower == "rmsproptf": + optimizer = RMSpropTF(parameters, alpha=0.9, momentum=momentum, **opt_args) + elif opt_lower == "novograd": + optimizer = NovoGrad(parameters, **opt_args) + elif opt_lower == "nvnovograd": + optimizer = NvNovoGrad(parameters, **opt_args) + elif opt_lower == "fusedsgd": + opt_args.pop("eps", None) + optimizer = FusedSGD(parameters, momentum=momentum, nesterov=True, **opt_args) + elif opt_lower == "fusedmomentum": + opt_args.pop("eps", None) + optimizer = FusedSGD(parameters, momentum=momentum, nesterov=False, **opt_args) + elif opt_lower == "fusedadam": + optimizer = FusedAdam(parameters, adam_w_mode=False, **opt_args) + elif opt_lower == "fusedadamw": + optimizer = FusedAdam(parameters, adam_w_mode=True, **opt_args) + elif opt_lower == "fusedlamb": + optimizer = FusedLAMB(parameters, **opt_args) + elif opt_lower == "fusednovograd": + opt_args.setdefault("betas", (0.95, 0.98)) + optimizer = FusedNovoGrad(parameters, **opt_args) + else: + assert False and "Invalid optimizer" + raise ValueError + + if len(opt_split) > 1: + if opt_split[0] == "lookahead": + optimizer = Lookahead(optimizer) + + return optimizer diff --git a/optimizers/scheduler_factory.py b/optimizers/scheduler_factory.py new file mode 100644 index 0000000000000000000000000000000000000000..8cff09723d2c2128317220bb2d316ddf7b035b76 --- /dev/null +++ b/optimizers/scheduler_factory.py @@ -0,0 +1,103 @@ +""" Scheduler Factory +Hacked together by / Copyright 2020 Ross Wightman +""" +from .timm.cosine_lr import CosineLRScheduler +from .timm.tanh_lr import TanhLRScheduler +from .timm.step_lr import StepLRScheduler +from .timm.plateau_lr import PlateauLRScheduler +import torch + +def create_scheduler(args, optimizer, **kwargs): + num_epochs = args.epochs + + if getattr(args, 'lr_noise', None) is not None: + lr_noise = getattr(args, 'lr_noise') + if isinstance(lr_noise, (list, tuple)): + noise_range = [n * num_epochs for n in lr_noise] + if len(noise_range) == 1: + noise_range = noise_range[0] + else: + noise_range = lr_noise * num_epochs + else: + noise_range = None + + lr_scheduler = None + if args.lr_policy == 'cosine': + lr_scheduler = CosineLRScheduler( + optimizer, + t_initial=num_epochs, + t_mul=getattr(args, 'lr_cycle_mul', 1.), + lr_min=args.lr_min, + decay_rate=args.decay_rate, + warmup_lr_init=args.warmup_lr, + warmup_t=args.warmup_epochs, + cycle_limit=getattr(args, 'lr_cycle_limit', 1), + t_in_epochs=True, + noise_range_t=noise_range, + noise_pct=getattr(args, 'lr_noise_pct', 0.67), + noise_std=getattr(args, 'lr_noise_std', 1.), + noise_seed=getattr(args, 'seed', 42), + ) + num_epochs = lr_scheduler.get_cycle_length() + args.COOLDOWN_EPOCHS + elif args.lr_policy == 'tanh': + lr_scheduler = TanhLRScheduler( + optimizer, + t_initial=num_epochs, + t_mul=getattr(args, 'lr_cycle_mul', 1.), + lr_min=args.min_lr, + warmup_lr_init=args.warmup_lr, + warmup_t=args.warmup_epochs, + cycle_limit=getattr(args, 'lr_cycle_limit', 1), + t_in_epochs=True, + noise_range_t=noise_range, + noise_pct=getattr(args, 'lr_noise_pct', 0.67), + noise_std=getattr(args, 'lr_noise_std', 1.), + noise_seed=getattr(args, 'seed', 42), + ) + num_epochs = lr_scheduler.get_cycle_length() + args.COOLDOWN_EPOCHS + elif args.lr_policy == 'step': + lr_scheduler = StepLRScheduler( + optimizer, + decay_t=args.decay_epochs - getattr(kwargs, 'init_epoch', 0), # for D + decay_rate=args.decay_rate, + warmup_lr_init=args.warmup_lr, + warmup_t=args.warmup_epochs, + noise_range_t=noise_range, + noise_pct=getattr(args, 'lr_noise_pct', 0.67), + noise_std=getattr(args, 'lr_noise_std', 1.), + noise_seed=getattr(args, 'seed', 42), + ) + elif args.lr_policy == 'plateau': + mode = 'min' if 'loss' in getattr(args, 'eval_metric', '') else 'max' + lr_scheduler = PlateauLRScheduler( + optimizer, + decay_rate=args.decay_rate, + patience_t=args.patience_epochs, + lr_min=args.min_lr, + mode=mode, + warmup_lr_init=args.warmup_lr, + warmup_t=args.warmup_epochs, + cooldown_t=0, + noise_range_t=noise_range, + noise_pct=getattr(args, 'lr_noise_pct', 0.67), + noise_std=getattr(args, 'lr_noise_std', 1.), + noise_seed=getattr(args, 'seed', 42), + ) + elif args.lr_policy == "onecyclelr": + lr_scheduler = torch.optim.lr_scheduler.OneCycleLR( + optimizer, + max_lr=args.LR, + total_steps=kwargs["total_steps"], + pct_start=args.PCT_START, + div_factor=args.DIV_FACTOR_ONECOS, + final_div_factor=args.FIN_DACTOR_ONCCOS, + ) + elif args.lr_policy == "cosinerestart": + lr_scheduler = torch.optim.lr_scheduler.CosineAnnealingWarmRestarts( + optimizer, + T_0 = kwargs["total_steps"], + T_mult=2, + eta_min = 1e-6, + last_epoch=-1, + ) + return lr_scheduler \ No newline at end of file diff --git a/optimizers/timm/__init__.py b/optimizers/timm/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..612eaaebb068a160663278c01db7c544a67907f3 --- /dev/null +++ b/optimizers/timm/__init__.py @@ -0,0 +1,16 @@ +from .adamp import AdamP +from .adamw import AdamW +from .adafactor import Adafactor +from .adahessian import Adahessian +from .lookahead import Lookahead +from .nadam import Nadam +from .novograd import NovoGrad +from .nvnovograd import NvNovoGrad +from .radam import RAdam +from .rmsprop_tf import RMSpropTF +from .sgdp import SGDP +from .adabelief import AdaBelief +from .cosine_lr import CosineLRScheduler +from .plateau_lr import PlateauLRScheduler +from .step_lr import StepLRScheduler +from .tanh_lr import TanhLRScheduler \ No newline at end of file diff --git a/optimizers/timm/adabelief.py b/optimizers/timm/adabelief.py new file mode 100644 index 0000000000000000000000000000000000000000..a26d7b27ac85ce65a02bc2e938058b685d914a65 --- /dev/null +++ b/optimizers/timm/adabelief.py @@ -0,0 +1,205 @@ +import math +import torch +from torch.optim.optimizer import Optimizer + + +class AdaBelief(Optimizer): + r"""Implements AdaBelief algorithm. Modified from Adam in PyTorch + + Arguments: + params (iterable): iterable of parameters to optimize or dicts defining + parameter groups + lr (float, optional): learning rate (default: 1e-3) + betas (Tuple[float, float], optional): coefficients used for computing + running averages of gradient and its square (default: (0.9, 0.999)) + eps (float, optional): term added to the denominator to improve + numerical stability (default: 1e-16) + weight_decay (float, optional): weight decay (L2 penalty) (default: 0) + amsgrad (boolean, optional): whether to use the AMSGrad variant of this + algorithm from the paper `On the Convergence of Adam and Beyond`_ + (default: False) + weight_decouple (boolean, optional): ( default: True) If set as True, then + the optimizer uses decoupled weight decay as in AdamW + fixed_decay (boolean, optional): (default: False) This is used when weight_decouple + is set as True. + When fixed_decay == True, the weight decay is performed as + $W_{new} = W_{old} - W_{old} \times decay$. + When fixed_decay == False, the weight decay is performed as + $W_{new} = W_{old} - W_{old} \times decay \times lr$. Note that in this case, the + weight decay ratio decreases with learning rate (lr). + rectify (boolean, optional): (default: True) If set as True, then perform the rectified + update similar to RAdam + degenerated_to_sgd (boolean, optional) (default:True) If set as True, then perform SGD update + when variance of gradient is high + reference: AdaBelief Optimizer, adapting stepsizes by the belief in observed gradients, NeurIPS 2020 + + For a complete table of recommended hyperparameters, see https://github.com/juntang-zhuang/Adabelief-Optimizer' + For example train/args for EfficientNet see these gists + - link to train_scipt: https://gist.github.com/juntang-zhuang/0a501dd51c02278d952cf159bc233037 + - link to args.yaml: https://gist.github.com/juntang-zhuang/517ce3c27022b908bb93f78e4f786dc3 + """ + + def __init__(self, params, lr=1e-3, betas=(0.9, 0.999), eps=1e-16, + weight_decay=0, amsgrad=False, weight_decouple=True, fixed_decay=False, rectify=True, + degenerated_to_sgd=True): + + if not 0.0 <= lr: + raise ValueError("Invalid learning rate: {}".format(lr)) + if not 0.0 <= eps: + raise ValueError("Invalid epsilon value: {}".format(eps)) + if not 0.0 <= betas[0] < 1.0: + raise ValueError("Invalid beta parameter at index 0: {}".format(betas[0])) + if not 0.0 <= betas[1] < 1.0: + raise ValueError("Invalid beta parameter at index 1: {}".format(betas[1])) + + self.degenerated_to_sgd = degenerated_to_sgd + if isinstance(params, (list, tuple)) and len(params) > 0 and isinstance(params[0], dict): + for param in params: + if 'betas' in param and (param['betas'][0] != betas[0] or param['betas'][1] != betas[1]): + param['buffer'] = [[None, None, None] for _ in range(10)] + + defaults = dict(lr=lr, betas=betas, eps=eps, + weight_decay=weight_decay, amsgrad=amsgrad, buffer=[[None, None, None] for _ in range(10)]) + super(AdaBelief, self).__init__(params, defaults) + + self.degenerated_to_sgd = degenerated_to_sgd + self.weight_decouple = weight_decouple + self.rectify = rectify + self.fixed_decay = fixed_decay + + def __setstate__(self, state): + super(AdaBelief, self).__setstate__(state) + for group in self.param_groups: + group.setdefault('amsgrad', False) + + def reset(self): + for group in self.param_groups: + for p in group['params']: + state = self.state[p] + amsgrad = group['amsgrad'] + + # State initialization + state['step'] = 0 + # Exponential moving average of gradient values + state['exp_avg'] = torch.zeros_like(p.data) + + # Exponential moving average of squared gradient values + state['exp_avg_var'] = torch.zeros_like(p.data) + if amsgrad: + # Maintains max of all exp. moving avg. of sq. grad. values + state['max_exp_avg_var'] = torch.zeros_like(p.data) + + def step(self, closure=None): + """Performs a single optimization step. + Arguments: + closure (callable, optional): A closure that reevaluates the model + and returns the loss. + """ + loss = None + if closure is not None: + loss = closure() + + for group in self.param_groups: + for p in group['params']: + if p.grad is None: + continue + + # cast data type + half_precision = False + if p.data.dtype == torch.float16: + half_precision = True + p.data = p.data.float() + p.grad = p.grad.float() + + grad = p.grad.data + if grad.is_sparse: + raise RuntimeError( + 'AdaBelief does not support sparse gradients, please consider SparseAdam instead') + amsgrad = group['amsgrad'] + + state = self.state[p] + + beta1, beta2 = group['betas'] + + # State initialization + if len(state) == 0: + state['step'] = 0 + # Exponential moving average of gradient values + state['exp_avg'] = torch.zeros_like(p.data) + # Exponential moving average of squared gradient values + state['exp_avg_var'] = torch.zeros_like(p.data) + if amsgrad: + # Maintains max of all exp. moving avg. of sq. grad. values + state['max_exp_avg_var'] = torch.zeros_like(p.data) + + # perform weight decay, check if decoupled weight decay + if self.weight_decouple: + if not self.fixed_decay: + p.data.mul_(1.0 - group['lr'] * group['weight_decay']) + else: + p.data.mul_(1.0 - group['weight_decay']) + else: + if group['weight_decay'] != 0: + grad.add_(p.data, alpha=group['weight_decay']) + + # get current state variable + exp_avg, exp_avg_var = state['exp_avg'], state['exp_avg_var'] + + state['step'] += 1 + bias_correction1 = 1 - beta1 ** state['step'] + bias_correction2 = 1 - beta2 ** state['step'] + + # Update first and second moment running average + exp_avg.mul_(beta1).add_(grad, alpha=1 - beta1) + grad_residual = grad - exp_avg + exp_avg_var.mul_(beta2).addcmul_( grad_residual, grad_residual, value=1 - beta2) + + if amsgrad: + max_exp_avg_var = state['max_exp_avg_var'] + # Maintains the maximum of all 2nd moment running avg. till now + torch.max(max_exp_avg_var, exp_avg_var.add_(group['eps']), out=max_exp_avg_var) + + # Use the max. for normalizing running avg. of gradient + denom = (max_exp_avg_var.sqrt() / math.sqrt(bias_correction2)).add_(group['eps']) + else: + denom = (exp_avg_var.add_(group['eps']).sqrt() / math.sqrt(bias_correction2)).add_(group['eps']) + + # update + if not self.rectify: + # Default update + step_size = group['lr'] / bias_correction1 + p.data.addcdiv_( exp_avg, denom, value=-step_size) + + else: # Rectified update, forked from RAdam + buffered = group['buffer'][int(state['step'] % 10)] + if state['step'] == buffered[0]: + N_sma, step_size = buffered[1], buffered[2] + else: + buffered[0] = state['step'] + beta2_t = beta2 ** state['step'] + N_sma_max = 2 / (1 - beta2) - 1 + N_sma = N_sma_max - 2 * state['step'] * beta2_t / (1 - beta2_t) + buffered[1] = N_sma + + # more conservative since it's an approximated value + if N_sma >= 5: + step_size = math.sqrt( + (1 - beta2_t) * (N_sma - 4) / (N_sma_max - 4) * (N_sma - 2) / N_sma * N_sma_max / ( + N_sma_max - 2)) / (1 - beta1 ** state['step']) + elif self.degenerated_to_sgd: + step_size = 1.0 / (1 - beta1 ** state['step']) + else: + step_size = -1 + buffered[2] = step_size + + if N_sma >= 5: + denom = exp_avg_var.sqrt().add_(group['eps']) + p.data.addcdiv_(exp_avg, denom, value=-step_size * group['lr']) + elif step_size > 0: + p.data.add_( exp_avg, alpha=-step_size * group['lr']) + + if half_precision: + p.data = p.data.half() + p.grad = p.grad.half() + + return loss diff --git a/optimizers/timm/adafactor.py b/optimizers/timm/adafactor.py new file mode 100644 index 0000000000000000000000000000000000000000..088ce3acd82e2be1b393afafa05f48435e538a1a --- /dev/null +++ b/optimizers/timm/adafactor.py @@ -0,0 +1,174 @@ +""" Adafactor Optimizer + +Lifted from https://github.com/pytorch/fairseq/blob/master/fairseq/optim/adafactor.py + +Original header/copyright below. + +""" +# Copyright (c) Facebook, Inc. and its affiliates. +# +# This source code is licensed under the MIT license found in the +# LICENSE file in the root directory of this source tree. +import torch +import math + + +class Adafactor(torch.optim.Optimizer): + """Implements Adafactor algorithm. + This implementation is based on: `Adafactor: Adaptive Learning Rates with Sublinear Memory Cost` + (see https://arxiv.org/abs/1804.04235) + + Note that this optimizer internally adjusts the learning rate depending on the + *scale_parameter*, *relative_step* and *warmup_init* options. + + To use a manual (external) learning rate schedule you should set `scale_parameter=False` and + `relative_step=False`. + + Arguments: + params (iterable): iterable of parameters to optimize or dicts defining parameter groups + lr (float, optional): external learning rate (default: None) + eps (tuple[float, float]): regularization constants for square gradient + and parameter scale respectively (default: (1e-30, 1e-3)) + clip_threshold (float): threshold of root mean square of final gradient update (default: 1.0) + decay_rate (float): coefficient used to compute running averages of square gradient (default: -0.8) + beta1 (float): coefficient used for computing running averages of gradient (default: None) + weight_decay (float, optional): weight decay (L2 penalty) (default: 0) + scale_parameter (bool): if True, learning rate is scaled by root mean square of parameter (default: True) + relative_step (bool): if True, time-dependent learning rate is computed + instead of external learning rate (default: True) + warmup_init (bool): time-dependent learning rate computation depends on + whether warm-up initialization is being used (default: False) + """ + + def __init__(self, params, lr=None, eps=1e-30, eps_scale=1e-3, clip_threshold=1.0, + decay_rate=-0.8, betas=None, weight_decay=0.0, scale_parameter=True, warmup_init=False): + relative_step = lr is None + if warmup_init and not relative_step: + raise ValueError('warmup_init requires relative_step=True') + + beta1 = None if betas is None else betas[0] # make it compat with standard betas arg + defaults = dict(lr=lr, eps=eps, eps_scale=eps_scale, clip_threshold=clip_threshold, decay_rate=decay_rate, + beta1=beta1, weight_decay=weight_decay, scale_parameter=scale_parameter, + relative_step=relative_step, warmup_init=warmup_init) + super(Adafactor, self).__init__(params, defaults) + + @staticmethod + def _get_lr(param_group, param_state): + if param_group['relative_step']: + min_step = 1e-6 * param_state['step'] if param_group['warmup_init'] else 1e-2 + lr_t = min(min_step, 1.0 / math.sqrt(param_state['step'])) + param_scale = 1.0 + if param_group['scale_parameter']: + param_scale = max(param_group['eps_scale'], param_state['RMS']) + param_group['lr'] = lr_t * param_scale + return param_group['lr'] + + @staticmethod + def _get_options(param_group, param_shape): + factored = len(param_shape) >= 2 + use_first_moment = param_group['beta1'] is not None + return factored, use_first_moment + + @staticmethod + def _rms(tensor): + return tensor.norm(2) / (tensor.numel() ** 0.5) + + def _approx_sq_grad(self, exp_avg_sq_row, exp_avg_sq_col): + r_factor = (exp_avg_sq_row / exp_avg_sq_row.mean(dim=-1, keepdim=True)).rsqrt_().unsqueeze(-1) + c_factor = exp_avg_sq_col.unsqueeze(-2).rsqrt() + return torch.mul(r_factor, c_factor) + + def step(self, closure=None): + """Performs a single optimization step. + Arguments: + closure (callable, optional): A closure that reevaluates the model and returns the loss. + """ + loss = None + if closure is not None: + loss = closure() + + for group in self.param_groups: + for p in group['params']: + if p.grad is None: + continue + grad = p.grad.data + if grad.dtype in {torch.float16, torch.bfloat16}: + grad = grad.float() + if grad.is_sparse: + raise RuntimeError('Adafactor does not support sparse gradients.') + + state = self.state[p] + grad_shape = grad.shape + + factored, use_first_moment = self._get_options(group, grad_shape) + # State Initialization + if len(state) == 0: + state['step'] = 0 + + if use_first_moment: + # Exponential moving average of gradient values + state['exp_avg'] = torch.zeros_like(grad) + if factored: + state['exp_avg_sq_row'] = torch.zeros(grad_shape[:-1]).to(grad) + state['exp_avg_sq_col'] = torch.zeros(grad_shape[:-2] + grad_shape[-1:]).to(grad) + else: + state['exp_avg_sq'] = torch.zeros_like(grad) + + state['RMS'] = 0 + else: + if use_first_moment: + state['exp_avg'] = state['exp_avg'].to(grad) + if factored: + state['exp_avg_sq_row'] = state['exp_avg_sq_row'].to(grad) + state['exp_avg_sq_col'] = state['exp_avg_sq_col'].to(grad) + else: + state['exp_avg_sq'] = state['exp_avg_sq'].to(grad) + + p_data_fp32 = p.data + if p.data.dtype in {torch.float16, torch.bfloat16}: + p_data_fp32 = p_data_fp32.float() + + state['step'] += 1 + state['RMS'] = self._rms(p_data_fp32) + lr_t = self._get_lr(group, state) + + beta2t = 1.0 - math.pow(state['step'], group['decay_rate']) + update = grad ** 2 + group['eps'] + if factored: + exp_avg_sq_row = state['exp_avg_sq_row'] + exp_avg_sq_col = state['exp_avg_sq_col'] + + exp_avg_sq_row.mul_(beta2t).add_(1.0 - beta2t, update.mean(dim=-1)) + exp_avg_sq_col.mul_(beta2t).add_(1.0 - beta2t, update.mean(dim=-2)) + #exp_avg_sq_row.mul_(beta2t).add_(update.mean(dim=-1), alpha=1.0 - beta2t) # pytorch 1.6+ + #exp_avg_sq_col.mul_(beta2t).add_(update.mean(dim=-2), alpha=1.0 - beta2t) + + # Approximation of exponential moving average of square of gradient + update = self._approx_sq_grad(exp_avg_sq_row, exp_avg_sq_col) + update.mul_(grad) + else: + exp_avg_sq = state['exp_avg_sq'] + + exp_avg_sq.mul_(beta2t).add_(1.0 - beta2t, update) + #exp_avg_sq.mul_(beta2t).add_(update, alpha=1.0 - beta2t) # pytorch 1.6+ + update = exp_avg_sq.rsqrt().mul_(grad) + + update.div_((self._rms(update) / group['clip_threshold']).clamp_(min=1.0)) + update.mul_(lr_t) + + if use_first_moment: + exp_avg = state['exp_avg'] + exp_avg.mul_(group["beta1"]).add_(1 - group["beta1"], update) + #exp_avg.mul_(group['beta1']).add_(update, alpha=1 - group['beta1']) # pytorch 1.6+ + update = exp_avg + + if group['weight_decay'] != 0: + p_data_fp32.add_(-group["weight_decay"] * lr_t, p_data_fp32) + #p_data_fp32.add_(p_data_fp32, alpha=-group['weight_decay'] * lr_t) # pytorch 1.6+ + + p_data_fp32.add_(-update) + + if p.data.dtype in {torch.float16, torch.bfloat16}: + p.data.copy_(p_data_fp32) + + return loss \ No newline at end of file diff --git a/optimizers/timm/adahessian.py b/optimizers/timm/adahessian.py new file mode 100644 index 0000000000000000000000000000000000000000..985c67ca686a65f61f5c5b1a7db3e5bba815a19b --- /dev/null +++ b/optimizers/timm/adahessian.py @@ -0,0 +1,156 @@ +""" AdaHessian Optimizer + +Lifted from https://github.com/davda54/ada-hessian/blob/master/ada_hessian.py +Originally licensed MIT, Copyright 2020, David Samuel +""" +import torch + + +class Adahessian(torch.optim.Optimizer): + """ + Implements the AdaHessian algorithm from "ADAHESSIAN: An Adaptive Second OrderOptimizer for Machine Learning" + + Arguments: + params (iterable): iterable of parameters to optimize or dicts defining parameter groups + lr (float, optional): learning rate (default: 0.1) + betas ((float, float), optional): coefficients used for computing running averages of gradient and the + squared hessian trace (default: (0.9, 0.999)) + eps (float, optional): term added to the denominator to improve numerical stability (default: 1e-8) + weight_decay (float, optional): weight decay (L2 penalty) (default: 0.0) + hessian_power (float, optional): exponent of the hessian trace (default: 1.0) + update_each (int, optional): compute the hessian trace approximation only after *this* number of steps + (to save time) (default: 1) + n_samples (int, optional): how many times to sample `z` for the approximation of the hessian trace (default: 1) + """ + + def __init__(self, params, lr=0.1, betas=(0.9, 0.999), eps=1e-8, weight_decay=0.0, + hessian_power=1.0, update_each=1, n_samples=1, avg_conv_kernel=False): + if not 0.0 <= lr: + raise ValueError(f"Invalid learning rate: {lr}") + if not 0.0 <= eps: + raise ValueError(f"Invalid epsilon value: {eps}") + if not 0.0 <= betas[0] < 1.0: + raise ValueError(f"Invalid beta parameter at index 0: {betas[0]}") + if not 0.0 <= betas[1] < 1.0: + raise ValueError(f"Invalid beta parameter at index 1: {betas[1]}") + if not 0.0 <= hessian_power <= 1.0: + raise ValueError(f"Invalid Hessian power value: {hessian_power}") + + self.n_samples = n_samples + self.update_each = update_each + self.avg_conv_kernel = avg_conv_kernel + + # use a separate generator that deterministically generates the same `z`s across all GPUs in case of distributed training + self.seed = 2147483647 + self.generator = torch.Generator().manual_seed(self.seed) + + defaults = dict(lr=lr, betas=betas, eps=eps, weight_decay=weight_decay, hessian_power=hessian_power) + super(Adahessian, self).__init__(params, defaults) + + for p in self.get_params(): + p.hess = 0.0 + self.state[p]["hessian step"] = 0 + + @property + def is_second_order(self): + return True + + def get_params(self): + """ + Gets all parameters in all param_groups with gradients + """ + + return (p for group in self.param_groups for p in group['params'] if p.requires_grad) + + def zero_hessian(self): + """ + Zeros out the accumalated hessian traces. + """ + + for p in self.get_params(): + if not isinstance(p.hess, float) and self.state[p]["hessian step"] % self.update_each == 0: + p.hess.zero_() + + @torch.no_grad() + def set_hessian(self): + """ + Computes the Hutchinson approximation of the hessian trace and accumulates it for each trainable parameter. + """ + + params = [] + for p in filter(lambda p: p.grad is not None, self.get_params()): + if self.state[p]["hessian step"] % self.update_each == 0: # compute the trace only each `update_each` step + params.append(p) + self.state[p]["hessian step"] += 1 + + if len(params) == 0: + return + + if self.generator.device != params[0].device: # hackish way of casting the generator to the right device + self.generator = torch.Generator(params[0].device).manual_seed(self.seed) + + grads = [p.grad for p in params] + + for i in range(self.n_samples): + # Rademacher distribution {-1.0, 1.0} + zs = [torch.randint(0, 2, p.size(), generator=self.generator, device=p.device) * 2.0 - 1.0 for p in params] + h_zs = torch.autograd.grad( + grads, params, grad_outputs=zs, only_inputs=True, retain_graph=i < self.n_samples - 1) + for h_z, z, p in zip(h_zs, zs, params): + p.hess += h_z * z / self.n_samples # approximate the expected values of z*(H@z) + + @torch.no_grad() + def step(self, closure=None): + """ + Performs a single optimization step. + Arguments: + closure (callable, optional) -- a closure that reevaluates the model and returns the loss (default: None) + """ + + loss = None + if closure is not None: + loss = closure() + + self.zero_hessian() + self.set_hessian() + + for group in self.param_groups: + for p in group['params']: + if p.grad is None or p.hess is None: + continue + + if self.avg_conv_kernel and p.dim() == 4: + p.hess = torch.abs(p.hess).mean(dim=[2, 3], keepdim=True).expand_as(p.hess).clone() + + # Perform correct stepweight decay as in AdamW + p.mul_(1 - group['lr'] * group['weight_decay']) + + state = self.state[p] + + # State initialization + if len(state) == 1: + state['step'] = 0 + # Exponential moving average of gradient values + state['exp_avg'] = torch.zeros_like(p) + # Exponential moving average of Hessian diagonal square values + state['exp_hessian_diag_sq'] = torch.zeros_like(p) + + exp_avg, exp_hessian_diag_sq = state['exp_avg'], state['exp_hessian_diag_sq'] + beta1, beta2 = group['betas'] + state['step'] += 1 + + # Decay the first and second moment running average coefficient + exp_avg.mul_(beta1).add_(p.grad, alpha=1 - beta1) + exp_hessian_diag_sq.mul_(beta2).addcmul_(p.hess, p.hess, value=1 - beta2) + + bias_correction1 = 1 - beta1 ** state['step'] + bias_correction2 = 1 - beta2 ** state['step'] + + k = group['hessian_power'] + denom = (exp_hessian_diag_sq / bias_correction2).pow_(k / 2).add_(group['eps']) + + # make update + step_size = group['lr'] / bias_correction1 + p.addcdiv_(exp_avg, denom, value=-step_size) + + return loss diff --git a/optimizers/timm/adamp.py b/optimizers/timm/adamp.py new file mode 100644 index 0000000000000000000000000000000000000000..468c3e865e0ceb6fb2bf22f9388237a783314f07 --- /dev/null +++ b/optimizers/timm/adamp.py @@ -0,0 +1,107 @@ +""" +AdamP Optimizer Implementation copied from https://github.com/clovaai/AdamP/blob/master/adamp/adamp.py + +Paper: `Slowing Down the Weight Norm Increase in Momentum-based Optimizers` - https://arxiv.org/abs/2006.08217 +Code: https://github.com/clovaai/AdamP + +Copyright (c) 2020-present NAVER Corp. +MIT license +""" + +import torch +import torch.nn as nn +from torch.optim.optimizer import Optimizer, required +import math + +class AdamP(Optimizer): + def __init__(self, params, lr=1e-3, betas=(0.9, 0.999), eps=1e-8, + weight_decay=0, delta=0.1, wd_ratio=0.1, nesterov=False): + defaults = dict(lr=lr, betas=betas, eps=eps, weight_decay=weight_decay, + delta=delta, wd_ratio=wd_ratio, nesterov=nesterov) + super(AdamP, self).__init__(params, defaults) + + def _channel_view(self, x): + return x.view(x.size(0), -1) + + def _layer_view(self, x): + return x.view(1, -1) + + def _cosine_similarity(self, x, y, eps, view_func): + x = view_func(x) + y = view_func(y) + + x_norm = x.norm(dim=1).add_(eps) + y_norm = y.norm(dim=1).add_(eps) + dot = (x * y).sum(dim=1) + + return dot.abs() / x_norm / y_norm + + def _projection(self, p, grad, perturb, delta, wd_ratio, eps): + wd = 1 + expand_size = [-1] + [1] * (len(p.shape) - 1) + for view_func in [self._channel_view, self._layer_view]: + + cosine_sim = self._cosine_similarity(grad, p.data, eps, view_func) + + if cosine_sim.max() < delta / math.sqrt(view_func(p.data).size(1)): + p_n = p.data / view_func(p.data).norm(dim=1).view(expand_size).add_(eps) + perturb -= p_n * view_func(p_n * perturb).sum(dim=1).view(expand_size) + wd = wd_ratio + + return perturb, wd + + return perturb, wd + + def step(self, closure=None): + loss = None + if closure is not None: + loss = closure() + + for group in self.param_groups: + for p in group['params']: + if p.grad is None: + continue + + grad = p.grad.data + beta1, beta2 = group['betas'] + nesterov = group['nesterov'] + + state = self.state[p] + + # State initialization + if len(state) == 0: + state['step'] = 0 + state['exp_avg'] = torch.zeros_like(p.data) + state['exp_avg_sq'] = torch.zeros_like(p.data) + + # Adam + exp_avg, exp_avg_sq = state['exp_avg'], state['exp_avg_sq'] + + state['step'] += 1 + bias_correction1 = 1 - beta1 ** state['step'] + bias_correction2 = 1 - beta2 ** state['step'] + + exp_avg.mul_(beta1).add_(1 - beta1, grad) + exp_avg_sq.mul_(beta2).addcmul_(1 - beta2, grad, grad) + + denom = (exp_avg_sq.sqrt() / math.sqrt(bias_correction2)).add_(group['eps']) + step_size = group['lr'] / bias_correction1 + + if nesterov: + perturb = (beta1 * exp_avg + (1 - beta1) * grad) / denom + else: + perturb = exp_avg / denom + + # Projection + wd_ratio = 1 + if len(p.shape) > 1: + perturb, wd_ratio = self._projection(p, grad, perturb, group['delta'], group['wd_ratio'], group['eps']) + + # Weight decay + if group['weight_decay'] > 0: + p.data.mul_(1 - group['lr'] * group['weight_decay'] * wd_ratio) + + # Step + p.data.add_(-step_size, perturb) + + return loss diff --git a/optimizers/timm/adamw.py b/optimizers/timm/adamw.py new file mode 100644 index 0000000000000000000000000000000000000000..66f9a959de586356a29ace2f9c57d3fee8d1057a --- /dev/null +++ b/optimizers/timm/adamw.py @@ -0,0 +1,117 @@ +""" AdamW Optimizer +Impl copied from PyTorch master +""" +import math +import torch +from torch.optim.optimizer import Optimizer + + +class AdamW(Optimizer): + r"""Implements AdamW algorithm. + + The original Adam algorithm was proposed in `Adam: A Method for Stochastic Optimization`_. + The AdamW variant was proposed in `Decoupled Weight Decay Regularization`_. + + Arguments: + params (iterable): iterable of parameters to optimize or dicts defining + parameter groups + lr (float, optional): learning rate (default: 1e-3) + betas (Tuple[float, float], optional): coefficients used for computing + running averages of gradient and its square (default: (0.9, 0.999)) + eps (float, optional): term added to the denominator to improve + numerical stability (default: 1e-8) + weight_decay (float, optional): weight decay coefficient (default: 1e-2) + amsgrad (boolean, optional): whether to use the AMSGrad variant of this + algorithm from the paper `On the Convergence of Adam and Beyond`_ + (default: False) + + .. _Adam\: A Method for Stochastic Optimization: + https://arxiv.org/abs/1412.6980 + .. _Decoupled Weight Decay Regularization: + https://arxiv.org/abs/1711.05101 + .. _On the Convergence of Adam and Beyond: + https://openreview.net/forum?id=ryQu7f-RZ + """ + + def __init__(self, params, lr=1e-3, betas=(0.9, 0.999), eps=1e-8, + weight_decay=1e-2, amsgrad=False): + if not 0.0 <= lr: + raise ValueError("Invalid learning rate: {}".format(lr)) + if not 0.0 <= eps: + raise ValueError("Invalid epsilon value: {}".format(eps)) + if not 0.0 <= betas[0] < 1.0: + raise ValueError("Invalid beta parameter at index 0: {}".format(betas[0])) + if not 0.0 <= betas[1] < 1.0: + raise ValueError("Invalid beta parameter at index 1: {}".format(betas[1])) + defaults = dict(lr=lr, betas=betas, eps=eps, + weight_decay=weight_decay, amsgrad=amsgrad) + super(AdamW, self).__init__(params, defaults) + + def __setstate__(self, state): + super(AdamW, self).__setstate__(state) + for group in self.param_groups: + group.setdefault('amsgrad', False) + + def step(self, closure=None): + """Performs a single optimization step. + + Arguments: + closure (callable, optional): A closure that reevaluates the model + and returns the loss. + """ + loss = None + if closure is not None: + loss = closure() + + for group in self.param_groups: + for p in group['params']: + if p.grad is None: + continue + + # Perform stepweight decay + p.data.mul_(1 - group['lr'] * group['weight_decay']) + + # Perform optimization step + grad = p.grad.data + if grad.is_sparse: + raise RuntimeError('Adam does not support sparse gradients, please consider SparseAdam instead') + amsgrad = group['amsgrad'] + + state = self.state[p] + + # State initialization + if len(state) == 0: + state['step'] = 0 + # Exponential moving average of gradient values + state['exp_avg'] = torch.zeros_like(p.data) + # Exponential moving average of squared gradient values + state['exp_avg_sq'] = torch.zeros_like(p.data) + if amsgrad: + # Maintains max of all exp. moving avg. of sq. grad. values + state['max_exp_avg_sq'] = torch.zeros_like(p.data) + + exp_avg, exp_avg_sq = state['exp_avg'], state['exp_avg_sq'] + if amsgrad: + max_exp_avg_sq = state['max_exp_avg_sq'] + beta1, beta2 = group['betas'] + + state['step'] += 1 + bias_correction1 = 1 - beta1 ** state['step'] + bias_correction2 = 1 - beta2 ** state['step'] + + # Decay the first and second moment running average coefficient + exp_avg.mul_(beta1).add_(1 - beta1, grad) + exp_avg_sq.mul_(beta2).addcmul_(1 - beta2, grad, grad) + if amsgrad: + # Maintains the maximum of all 2nd moment running avg. till now + torch.max(max_exp_avg_sq, exp_avg_sq, out=max_exp_avg_sq) + # Use the max. for normalizing running avg. of gradient + denom = (max_exp_avg_sq.sqrt() / math.sqrt(bias_correction2)).add_(group['eps']) + else: + denom = (exp_avg_sq.sqrt() / math.sqrt(bias_correction2)).add_(group['eps']) + + step_size = group['lr'] / bias_correction1 + + p.data.addcdiv_(-step_size, exp_avg, denom) + + return loss diff --git a/optimizers/timm/cosine_lr.py b/optimizers/timm/cosine_lr.py new file mode 100644 index 0000000000000000000000000000000000000000..1532f092b5cc8c0af5125967cfb84b32ce03ca4a --- /dev/null +++ b/optimizers/timm/cosine_lr.py @@ -0,0 +1,116 @@ +""" Cosine Scheduler + +Cosine LR schedule with warmup, cycle/restarts, noise. + +Hacked together by / Copyright 2020 Ross Wightman +""" +import logging +import math +import numpy as np +import torch + +from .scheduler import Scheduler + + +_logger = logging.getLogger(__name__) + + +class CosineLRScheduler(Scheduler): + """ + Cosine decay with restarts. + This is described in the paper https://arxiv.org/abs/1608.03983. + + Inspiration from + https://github.com/allenai/allennlp/blob/master/allennlp/training/learning_rate_schedulers/cosine.py + """ + + def __init__(self, + optimizer: torch.optim.Optimizer, + t_initial: int, + t_mul: float = 1., + lr_min: float = 0., + decay_rate: float = 1., + warmup_t=0, + warmup_lr_init=0, + warmup_prefix=False, + cycle_limit=0, + t_in_epochs=True, + noise_range_t=None, + noise_pct=0.67, + noise_std=1.0, + noise_seed=42, + initialize=True) -> None: + super().__init__( + optimizer, param_group_field="lr", + noise_range_t=noise_range_t, noise_pct=noise_pct, noise_std=noise_std, noise_seed=noise_seed, + initialize=initialize) + + assert t_initial > 0 + assert lr_min >= 0 + if t_initial == 1 and t_mul == 1 and decay_rate == 1: + _logger.warning("Cosine annealing scheduler will have no effect on the learning " + "rate since t_initial = t_mul = eta_mul = 1.") + self.t_initial = t_initial + self.t_mul = t_mul + self.lr_min = lr_min + self.decay_rate = decay_rate + self.cycle_limit = cycle_limit + self.warmup_t = warmup_t + self.warmup_lr_init = warmup_lr_init + self.warmup_prefix = warmup_prefix + self.t_in_epochs = t_in_epochs + if self.warmup_t: + self.warmup_steps = [(v - warmup_lr_init) / self.warmup_t for v in self.base_values] + super().update_groups(self.warmup_lr_init) + else: + self.warmup_steps = [1 for _ in self.base_values] + + def _get_lr(self, t): + if t < self.warmup_t: + lrs = [self.warmup_lr_init + t * s for s in self.warmup_steps] + else: + if self.warmup_prefix: + t = t - self.warmup_t + + if self.t_mul != 1: + i = math.floor(math.log(1 - t / self.t_initial * (1 - self.t_mul), self.t_mul)) + t_i = self.t_mul ** i * self.t_initial + t_curr = t - (1 - self.t_mul ** i) / (1 - self.t_mul) * self.t_initial + else: + i = t // self.t_initial + t_i = self.t_initial + t_curr = t - (self.t_initial * i) + + gamma = self.decay_rate ** i + lr_min = self.lr_min * gamma + lr_max_values = [v * gamma for v in self.base_values] + + if self.cycle_limit == 0 or (self.cycle_limit > 0 and i < self.cycle_limit): + lrs = [ + lr_min + 0.5 * (lr_max - lr_min) * (1 + math.cos(math.pi * t_curr / t_i)) for lr_max in lr_max_values + ] + else: + lrs = [self.lr_min for _ in self.base_values] + + return lrs + + def get_epoch_values(self, epoch: int): + if self.t_in_epochs: + return self._get_lr(epoch) + else: + return None + + def get_update_values(self, num_updates: int): + if not self.t_in_epochs: + return self._get_lr(num_updates) + else: + return None + + def get_cycle_length(self, cycles=0): + if not cycles: + cycles = self.cycle_limit + cycles = max(1, cycles) + if self.t_mul == 1.0: + return self.t_initial * cycles + else: + return int(math.floor(-self.t_initial * (self.t_mul ** cycles - 1) / (1 - self.t_mul))) diff --git a/optimizers/timm/lookahead.py b/optimizers/timm/lookahead.py new file mode 100644 index 0000000000000000000000000000000000000000..6b5b7f38ec8cb6594e3986b66223fa2881daeca3 --- /dev/null +++ b/optimizers/timm/lookahead.py @@ -0,0 +1,92 @@ +""" Lookahead Optimizer Wrapper. +Implementation modified from: https://github.com/alphadl/lookahead.pytorch +Paper: `Lookahead Optimizer: k steps forward, 1 step back` - https://arxiv.org/abs/1907.08610 + +Hacked together by / Copyright 2020 Ross Wightman +""" +import torch +from torch.optim.optimizer import Optimizer +from collections import defaultdict + + +class Lookahead(Optimizer): + def __init__(self, base_optimizer, alpha=0.5, k=6): + if not 0.0 <= alpha <= 1.0: + raise ValueError(f'Invalid slow update rate: {alpha}') + if not 1 <= k: + raise ValueError(f'Invalid lookahead steps: {k}') + defaults = dict(lookahead_alpha=alpha, lookahead_k=k, lookahead_step=0) + self.base_optimizer = base_optimizer + self.param_groups = self.base_optimizer.param_groups + self.defaults = base_optimizer.defaults + self.defaults.update(defaults) + self.state = defaultdict(dict) + # manually add our defaults to the param groups + for name, default in defaults.items(): + for group in self.param_groups: + group.setdefault(name, default) + + def update_slow(self, group): + for fast_p in group["params"]: + if fast_p.grad is None: + continue + param_state = self.state[fast_p] + if 'slow_buffer' not in param_state: + param_state['slow_buffer'] = torch.empty_like(fast_p.data) + param_state['slow_buffer'].copy_(fast_p.data) + slow = param_state['slow_buffer'] + slow.add_(group['lookahead_alpha'], fast_p.data - slow) + fast_p.data.copy_(slow) + + def sync_lookahead(self): + for group in self.param_groups: + self.update_slow(group) + + def step(self, closure=None): + #assert id(self.param_groups) == id(self.base_optimizer.param_groups) + loss = self.base_optimizer.step(closure) + for group in self.param_groups: + group['lookahead_step'] += 1 + if group['lookahead_step'] % group['lookahead_k'] == 0: + self.update_slow(group) + return loss + + def state_dict(self): + fast_state_dict = self.base_optimizer.state_dict() + slow_state = { + (id(k) if isinstance(k, torch.Tensor) else k): v + for k, v in self.state.items() + } + fast_state = fast_state_dict['state'] + param_groups = fast_state_dict['param_groups'] + return { + 'state': fast_state, + 'slow_state': slow_state, + 'param_groups': param_groups, + } + + def load_state_dict(self, state_dict): + fast_state_dict = { + 'state': state_dict['state'], + 'param_groups': state_dict['param_groups'], + } + self.base_optimizer.load_state_dict(fast_state_dict) + + # We want to restore the slow state, but share param_groups reference + # with base_optimizer. This is a bit redundant but least code + slow_state_new = False + if 'slow_state' not in state_dict: + print('Loading state_dict from optimizer without Lookahead applied.') + state_dict['slow_state'] = defaultdict(dict) + slow_state_new = True + slow_state_dict = { + 'state': state_dict['slow_state'], + 'param_groups': state_dict['param_groups'], # this is pointless but saves code + } + super(Lookahead, self).load_state_dict(slow_state_dict) + self.param_groups = self.base_optimizer.param_groups # make both ref same container + if slow_state_new: + # reapply defaults to catch missing lookahead specific ones + for name, default in self.defaults.items(): + for group in self.param_groups: + group.setdefault(name, default) diff --git a/optimizers/timm/nadam.py b/optimizers/timm/nadam.py new file mode 100644 index 0000000000000000000000000000000000000000..d994d1b83485c9b068de73f5f3cf2efb1e5bec39 --- /dev/null +++ b/optimizers/timm/nadam.py @@ -0,0 +1,88 @@ +import torch +from torch.optim import Optimizer + + +class Nadam(Optimizer): + """Implements Nadam algorithm (a variant of Adam based on Nesterov momentum). + + It has been proposed in `Incorporating Nesterov Momentum into Adam`__. + + Arguments: + params (iterable): iterable of parameters to optimize or dicts defining + parameter groups + lr (float, optional): learning rate (default: 2e-3) + betas (Tuple[float, float], optional): coefficients used for computing + running averages of gradient and its square + eps (float, optional): term added to the denominator to improve + numerical stability (default: 1e-8) + weight_decay (float, optional): weight decay (L2 penalty) (default: 0) + schedule_decay (float, optional): momentum schedule decay (default: 4e-3) + + __ http://cs229.stanford.edu/proj2015/054_report.pdf + __ http://www.cs.toronto.edu/~fritz/absps/momentum.pdf + + Originally taken from: https://github.com/pytorch/pytorch/pull/1408 + NOTE: Has potential issues but does work well on some problems. + """ + + def __init__(self, params, lr=2e-3, betas=(0.9, 0.999), eps=1e-8, + weight_decay=0, schedule_decay=4e-3): + defaults = dict(lr=lr, betas=betas, eps=eps, + weight_decay=weight_decay, schedule_decay=schedule_decay) + super(Nadam, self).__init__(params, defaults) + + def step(self, closure=None): + """Performs a single optimization step. + + Arguments: + closure (callable, optional): A closure that reevaluates the model + and returns the loss. + """ + loss = None + if closure is not None: + loss = closure() + + for group in self.param_groups: + for p in group['params']: + if p.grad is None: + continue + grad = p.grad.data + state = self.state[p] + + # State initialization + if len(state) == 0: + state['step'] = 0 + state['m_schedule'] = 1. + state['exp_avg'] = grad.new().resize_as_(grad).zero_() + state['exp_avg_sq'] = grad.new().resize_as_(grad).zero_() + + # Warming momentum schedule + m_schedule = state['m_schedule'] + schedule_decay = group['schedule_decay'] + exp_avg, exp_avg_sq = state['exp_avg'], state['exp_avg_sq'] + beta1, beta2 = group['betas'] + eps = group['eps'] + state['step'] += 1 + t = state['step'] + + if group['weight_decay'] != 0: + grad = grad.add(group['weight_decay'], p.data) + + momentum_cache_t = beta1 * \ + (1. - 0.5 * (0.96 ** (t * schedule_decay))) + momentum_cache_t_1 = beta1 * \ + (1. - 0.5 * (0.96 ** ((t + 1) * schedule_decay))) + m_schedule_new = m_schedule * momentum_cache_t + m_schedule_next = m_schedule * momentum_cache_t * momentum_cache_t_1 + state['m_schedule'] = m_schedule_new + + # Decay the first and second moment running average coefficient + exp_avg.mul_(beta1).add_(1. - beta1, grad) + exp_avg_sq.mul_(beta2).addcmul_(1. - beta2, grad, grad) + exp_avg_sq_prime = exp_avg_sq / (1. - beta2 ** t) + denom = exp_avg_sq_prime.sqrt_().add_(eps) + + p.data.addcdiv_(-group['lr'] * (1. - momentum_cache_t) / (1. - m_schedule_new), grad, denom) + p.data.addcdiv_(-group['lr'] * momentum_cache_t_1 / (1. - m_schedule_next), exp_avg, denom) + + return loss diff --git a/optimizers/timm/novograd.py b/optimizers/timm/novograd.py new file mode 100644 index 0000000000000000000000000000000000000000..4137c6aa9406360d29f5f7234ebbdef294404d0e --- /dev/null +++ b/optimizers/timm/novograd.py @@ -0,0 +1,77 @@ +"""NovoGrad Optimizer. +Original impl by Masashi Kimura (Convergence Lab): https://github.com/convergence-lab/novograd +Paper: `Stochastic Gradient Methods with Layer-wise Adaptive Moments for Training of Deep Networks` + - https://arxiv.org/abs/1905.11286 +""" + +import torch +from torch.optim.optimizer import Optimizer +import math + + +class NovoGrad(Optimizer): + def __init__(self, params, grad_averaging=False, lr=0.1, betas=(0.95, 0.98), eps=1e-8, weight_decay=0): + defaults = dict(lr=lr, betas=betas, eps=eps, weight_decay=weight_decay) + super(NovoGrad, self).__init__(params, defaults) + self._lr = lr + self._beta1 = betas[0] + self._beta2 = betas[1] + self._eps = eps + self._wd = weight_decay + self._grad_averaging = grad_averaging + + self._momentum_initialized = False + + def step(self, closure=None): + loss = None + if closure is not None: + loss = closure() + + if not self._momentum_initialized: + for group in self.param_groups: + for p in group['params']: + if p.grad is None: + continue + state = self.state[p] + grad = p.grad.data + if grad.is_sparse: + raise RuntimeError('NovoGrad does not support sparse gradients') + + v = torch.norm(grad)**2 + m = grad/(torch.sqrt(v) + self._eps) + self._wd * p.data + state['step'] = 0 + state['v'] = v + state['m'] = m + state['grad_ema'] = None + self._momentum_initialized = True + + for group in self.param_groups: + for p in group['params']: + if p.grad is None: + continue + state = self.state[p] + state['step'] += 1 + + step, v, m = state['step'], state['v'], state['m'] + grad_ema = state['grad_ema'] + + grad = p.grad.data + g2 = torch.norm(grad)**2 + grad_ema = g2 if grad_ema is None else grad_ema * \ + self._beta2 + g2 * (1. - self._beta2) + grad *= 1.0 / (torch.sqrt(grad_ema) + self._eps) + + if self._grad_averaging: + grad *= (1. - self._beta1) + + g2 = torch.norm(grad)**2 + v = self._beta2*v + (1. - self._beta2)*g2 + m = self._beta1*m + (grad / (torch.sqrt(v) + self._eps) + self._wd * p.data) + bias_correction1 = 1 - self._beta1 ** step + bias_correction2 = 1 - self._beta2 ** step + step_size = group['lr'] * math.sqrt(bias_correction2) / bias_correction1 + + state['v'], state['m'] = v, m + state['grad_ema'] = grad_ema + p.data.add_(-step_size, m) + return loss diff --git a/optimizers/timm/nvnovograd.py b/optimizers/timm/nvnovograd.py new file mode 100644 index 0000000000000000000000000000000000000000..323312d2fc36d028124f7a7ec604d248e71503cd --- /dev/null +++ b/optimizers/timm/nvnovograd.py @@ -0,0 +1,118 @@ +""" Nvidia NovoGrad Optimizer. +Original impl by Nvidia from Jasper example: + - https://github.com/NVIDIA/DeepLearningExamples/blob/master/PyTorch/SpeechRecognition/Jasper +Paper: `Stochastic Gradient Methods with Layer-wise Adaptive Moments for Training of Deep Networks` + - https://arxiv.org/abs/1905.11286 +""" + +import torch +from torch.optim.optimizer import Optimizer +import math + + +class NvNovoGrad(Optimizer): + """ + Implements Novograd algorithm. + + Args: + params (iterable): iterable of parameters to optimize or dicts defining + parameter groups + lr (float, optional): learning rate (default: 1e-3) + betas (Tuple[float, float], optional): coefficients used for computing + running averages of gradient and its square (default: (0.95, 0.98)) + eps (float, optional): term added to the denominator to improve + numerical stability (default: 1e-8) + weight_decay (float, optional): weight decay (L2 penalty) (default: 0) + grad_averaging: gradient averaging + amsgrad (boolean, optional): whether to use the AMSGrad variant of this + algorithm from the paper `On the Convergence of Adam and Beyond`_ + (default: False) + """ + + def __init__(self, params, lr=1e-3, betas=(0.95, 0.98), eps=1e-8, + weight_decay=0, grad_averaging=False, amsgrad=False): + if not 0.0 <= lr: + raise ValueError("Invalid learning rate: {}".format(lr)) + if not 0.0 <= eps: + raise ValueError("Invalid epsilon value: {}".format(eps)) + if not 0.0 <= betas[0] < 1.0: + raise ValueError("Invalid beta parameter at index 0: {}".format(betas[0])) + if not 0.0 <= betas[1] < 1.0: + raise ValueError("Invalid beta parameter at index 1: {}".format(betas[1])) + defaults = dict(lr=lr, betas=betas, eps=eps, + weight_decay=weight_decay, + grad_averaging=grad_averaging, + amsgrad=amsgrad) + + super(NvNovoGrad, self).__init__(params, defaults) + + def __setstate__(self, state): + super(NvNovoGrad, self).__setstate__(state) + for group in self.param_groups: + group.setdefault('amsgrad', False) + + def step(self, closure=None): + """Performs a single optimization step. + + Arguments: + closure (callable, optional): A closure that reevaluates the model + and returns the loss. + """ + loss = None + if closure is not None: + loss = closure() + + for group in self.param_groups: + for p in group['params']: + if p.grad is None: + continue + grad = p.grad.data + if grad.is_sparse: + raise RuntimeError('Sparse gradients are not supported.') + amsgrad = group['amsgrad'] + + state = self.state[p] + + # State initialization + if len(state) == 0: + state['step'] = 0 + # Exponential moving average of gradient values + state['exp_avg'] = torch.zeros_like(p.data) + # Exponential moving average of squared gradient values + state['exp_avg_sq'] = torch.zeros([]).to(state['exp_avg'].device) + if amsgrad: + # Maintains max of all exp. moving avg. of sq. grad. values + state['max_exp_avg_sq'] = torch.zeros([]).to(state['exp_avg'].device) + + exp_avg, exp_avg_sq = state['exp_avg'], state['exp_avg_sq'] + if amsgrad: + max_exp_avg_sq = state['max_exp_avg_sq'] + beta1, beta2 = group['betas'] + + state['step'] += 1 + + norm = torch.sum(torch.pow(grad, 2)) + + if exp_avg_sq == 0: + exp_avg_sq.copy_(norm) + else: + exp_avg_sq.mul_(beta2).add_(1 - beta2, norm) + + if amsgrad: + # Maintains the maximum of all 2nd moment running avg. till now + torch.max(max_exp_avg_sq, exp_avg_sq, out=max_exp_avg_sq) + # Use the max. for normalizing running avg. of gradient + denom = max_exp_avg_sq.sqrt().add_(group['eps']) + else: + denom = exp_avg_sq.sqrt().add_(group['eps']) + + grad.div_(denom) + if group['weight_decay'] != 0: + grad.add_(group['weight_decay'], p.data) + if group['grad_averaging']: + grad.mul_(1 - beta1) + exp_avg.mul_(beta1).add_(grad) + + p.data.add_(-group['lr'], exp_avg) + + return loss diff --git a/optimizers/timm/plateau_lr.py b/optimizers/timm/plateau_lr.py new file mode 100644 index 0000000000000000000000000000000000000000..4f2cacb65a1bf23d10aa6fd296f74579571043cf --- /dev/null +++ b/optimizers/timm/plateau_lr.py @@ -0,0 +1,113 @@ +""" Plateau Scheduler + +Adapts PyTorch plateau scheduler and allows application of noise, warmup. + +Hacked together by / Copyright 2020 Ross Wightman +""" +import torch + +from .scheduler import Scheduler + + +class PlateauLRScheduler(Scheduler): + """Decay the LR by a factor every time the validation loss plateaus.""" + + def __init__(self, + optimizer, + decay_rate=0.1, + patience_t=10, + verbose=True, + threshold=1e-4, + cooldown_t=0, + warmup_t=0, + warmup_lr_init=0, + lr_min=0, + mode='max', + noise_range_t=None, + noise_type='normal', + noise_pct=0.67, + noise_std=1.0, + noise_seed=None, + initialize=True, + ): + super().__init__(optimizer, 'lr', initialize=initialize) + + self.lr_scheduler = torch.optim.lr_scheduler.ReduceLROnPlateau( + self.optimizer, + patience=patience_t, + factor=decay_rate, + verbose=verbose, + threshold=threshold, + cooldown=cooldown_t, + mode=mode, + min_lr=lr_min + ) + + self.noise_range = noise_range_t + self.noise_pct = noise_pct + self.noise_type = noise_type + self.noise_std = noise_std + self.noise_seed = noise_seed if noise_seed is not None else 42 + self.warmup_t = warmup_t + self.warmup_lr_init = warmup_lr_init + if self.warmup_t: + self.warmup_steps = [(v - warmup_lr_init) / self.warmup_t for v in self.base_values] + super().update_groups(self.warmup_lr_init) + else: + self.warmup_steps = [1 for _ in self.base_values] + self.restore_lr = None + + def state_dict(self): + return { + 'best': self.lr_scheduler.best, + 'last_epoch': self.lr_scheduler.last_epoch, + } + + def load_state_dict(self, state_dict): + self.lr_scheduler.best = state_dict['best'] + if 'last_epoch' in state_dict: + self.lr_scheduler.last_epoch = state_dict['last_epoch'] + + # override the base class step fn completely + def step(self, epoch, metric=None): + if epoch <= self.warmup_t: + lrs = [self.warmup_lr_init + epoch * s for s in self.warmup_steps] + super().update_groups(lrs) + else: + if self.restore_lr is not None: + # restore actual LR from before our last noise perturbation before stepping base + for i, param_group in enumerate(self.optimizer.param_groups): + param_group['lr'] = self.restore_lr[i] + self.restore_lr = None + + self.lr_scheduler.step(metric, epoch) # step the base scheduler + + if self.noise_range is not None: + if isinstance(self.noise_range, (list, tuple)): + apply_noise = self.noise_range[0] <= epoch < self.noise_range[1] + else: + apply_noise = epoch >= self.noise_range + if apply_noise: + self._apply_noise(epoch) + + def _apply_noise(self, epoch): + g = torch.Generator() + g.manual_seed(self.noise_seed + epoch) + if self.noise_type == 'normal': + while True: + # resample if noise out of percent limit, brute force but shouldn't spin much + noise = torch.randn(1, generator=g).item() + if abs(noise) < self.noise_pct: + break + else: + noise = 2 * (torch.rand(1, generator=g).item() - 0.5) * self.noise_pct + + # apply the noise on top of previous LR, cache the old value so we can restore for normal + # stepping of base scheduler + restore_lr = [] + for i, param_group in enumerate(self.optimizer.param_groups): + old_lr = float(param_group['lr']) + restore_lr.append(old_lr) + new_lr = old_lr + old_lr * noise + param_group['lr'] = new_lr + self.restore_lr = restore_lr diff --git a/optimizers/timm/radam.py b/optimizers/timm/radam.py new file mode 100644 index 0000000000000000000000000000000000000000..9987a334460286b1a6c8ec6d57ee023596a74219 --- /dev/null +++ b/optimizers/timm/radam.py @@ -0,0 +1,152 @@ +"""RAdam Optimizer. +Implementation lifted from: https://github.com/LiyuanLucasLiu/RAdam +Paper: `On the Variance of the Adaptive Learning Rate and Beyond` - https://arxiv.org/abs/1908.03265 +""" +import math +import torch +from torch.optim.optimizer import Optimizer, required + + +class RAdam(Optimizer): + + def __init__(self, params, lr=1e-3, betas=(0.9, 0.999), eps=1e-8, weight_decay=0): + defaults = dict(lr=lr, betas=betas, eps=eps, weight_decay=weight_decay) + self.buffer = [[None, None, None] for ind in range(10)] + super(RAdam, self).__init__(params, defaults) + + def __setstate__(self, state): + super(RAdam, self).__setstate__(state) + + def step(self, closure=None): + + loss = None + if closure is not None: + loss = closure() + + for group in self.param_groups: + + for p in group['params']: + if p.grad is None: + continue + grad = p.grad.data.float() + if grad.is_sparse: + raise RuntimeError('RAdam does not support sparse gradients') + + p_data_fp32 = p.data.float() + + state = self.state[p] + + if len(state) == 0: + state['step'] = 0 + state['exp_avg'] = torch.zeros_like(p_data_fp32) + state['exp_avg_sq'] = torch.zeros_like(p_data_fp32) + else: + state['exp_avg'] = state['exp_avg'].type_as(p_data_fp32) + state['exp_avg_sq'] = state['exp_avg_sq'].type_as(p_data_fp32) + + exp_avg, exp_avg_sq = state['exp_avg'], state['exp_avg_sq'] + beta1, beta2 = group['betas'] + + exp_avg_sq.mul_(beta2).addcmul_(1 - beta2, grad, grad) + exp_avg.mul_(beta1).add_(1 - beta1, grad) + + state['step'] += 1 + buffered = self.buffer[int(state['step'] % 10)] + if state['step'] == buffered[0]: + N_sma, step_size = buffered[1], buffered[2] + else: + buffered[0] = state['step'] + beta2_t = beta2 ** state['step'] + N_sma_max = 2 / (1 - beta2) - 1 + N_sma = N_sma_max - 2 * state['step'] * beta2_t / (1 - beta2_t) + buffered[1] = N_sma + + # more conservative since it's an approximated value + if N_sma >= 5: + step_size = group['lr'] * math.sqrt( + (1 - beta2_t) * (N_sma - 4) / (N_sma_max - 4) * (N_sma - 2) / N_sma * N_sma_max / ( + N_sma_max - 2)) / (1 - beta1 ** state['step']) + else: + step_size = group['lr'] / (1 - beta1 ** state['step']) + buffered[2] = step_size + + if group['weight_decay'] != 0: + p_data_fp32.add_(-group['weight_decay'] * group['lr'], p_data_fp32) + + # more conservative since it's an approximated value + if N_sma >= 5: + denom = exp_avg_sq.sqrt().add_(group['eps']) + p_data_fp32.addcdiv_(-step_size, exp_avg, denom) + else: + p_data_fp32.add_(-step_size, exp_avg) + + p.data.copy_(p_data_fp32) + + return loss + + +class PlainRAdam(Optimizer): + + def __init__(self, params, lr=1e-3, betas=(0.9, 0.999), eps=1e-8, weight_decay=0): + defaults = dict(lr=lr, betas=betas, eps=eps, weight_decay=weight_decay) + + super(PlainRAdam, self).__init__(params, defaults) + + def __setstate__(self, state): + super(PlainRAdam, self).__setstate__(state) + + def step(self, closure=None): + + loss = None + if closure is not None: + loss = closure() + + for group in self.param_groups: + + for p in group['params']: + if p.grad is None: + continue + grad = p.grad.data.float() + if grad.is_sparse: + raise RuntimeError('RAdam does not support sparse gradients') + + p_data_fp32 = p.data.float() + + state = self.state[p] + + if len(state) == 0: + state['step'] = 0 + state['exp_avg'] = torch.zeros_like(p_data_fp32) + state['exp_avg_sq'] = torch.zeros_like(p_data_fp32) + else: + state['exp_avg'] = state['exp_avg'].type_as(p_data_fp32) + state['exp_avg_sq'] = state['exp_avg_sq'].type_as(p_data_fp32) + + exp_avg, exp_avg_sq = state['exp_avg'], state['exp_avg_sq'] + beta1, beta2 = group['betas'] + + exp_avg_sq.mul_(beta2).addcmul_(1 - beta2, grad, grad) + exp_avg.mul_(beta1).add_(1 - beta1, grad) + + state['step'] += 1 + beta2_t = beta2 ** state['step'] + N_sma_max = 2 / (1 - beta2) - 1 + N_sma = N_sma_max - 2 * state['step'] * beta2_t / (1 - beta2_t) + + if group['weight_decay'] != 0: + p_data_fp32.add_(-group['weight_decay'] * group['lr'], p_data_fp32) + + # more conservative since it's an approximated value + if N_sma >= 5: + step_size = group['lr'] * math.sqrt( + (1 - beta2_t) * (N_sma - 4) / (N_sma_max - 4) * (N_sma - 2) / N_sma * N_sma_max / ( + N_sma_max - 2)) / (1 - beta1 ** state['step']) + denom = exp_avg_sq.sqrt().add_(group['eps']) + p_data_fp32.addcdiv_(-step_size, exp_avg, denom) + else: + step_size = group['lr'] / (1 - beta1 ** state['step']) + p_data_fp32.add_(-step_size, exp_avg) + + p.data.copy_(p_data_fp32) + + return loss diff --git a/optimizers/timm/rmsprop_tf.py b/optimizers/timm/rmsprop_tf.py new file mode 100644 index 0000000000000000000000000000000000000000..5115555cd26040e3af297a6e79e7bd5e4d202623 --- /dev/null +++ b/optimizers/timm/rmsprop_tf.py @@ -0,0 +1,136 @@ +""" RMSProp modified to behave like Tensorflow impl + +Originally cut & paste from PyTorch RMSProp +https://github.com/pytorch/pytorch/blob/063946d2b3f3f1e953a2a3b54e0b34f1393de295/torch/optim/rmsprop.py +Licensed under BSD-Clause 3 (ish), https://github.com/pytorch/pytorch/blob/master/LICENSE + +Modifications Copyright 2020 Ross Wightman +""" + +import torch +from torch.optim import Optimizer + + +class RMSpropTF(Optimizer): + """Implements RMSprop algorithm (TensorFlow style epsilon) + + NOTE: This is a direct cut-and-paste of PyTorch RMSprop with eps applied before sqrt + and a few other modifications to closer match Tensorflow for matching hyper-params. + + Noteworthy changes include: + 1. Epsilon applied inside square-root + 2. square_avg initialized to ones + 3. LR scaling of update accumulated in momentum buffer + + Proposed by G. Hinton in his + `course `_. + + The centered version first appears in `Generating Sequences + With Recurrent Neural Networks `_. + + Arguments: + params (iterable): iterable of parameters to optimize or dicts defining + parameter groups + lr (float, optional): learning rate (default: 1e-2) + momentum (float, optional): momentum factor (default: 0) + alpha (float, optional): smoothing (decay) constant (default: 0.9) + eps (float, optional): term added to the denominator to improve + numerical stability (default: 1e-10) + centered (bool, optional) : if ``True``, compute the centered RMSProp, + the gradient is normalized by an estimation of its variance + weight_decay (float, optional): weight decay (L2 penalty) (default: 0) + decoupled_decay (bool, optional): decoupled weight decay as per https://arxiv.org/abs/1711.05101 + lr_in_momentum (bool, optional): learning rate scaling is included in the momentum buffer + update as per defaults in Tensorflow + + """ + + def __init__(self, params, lr=1e-2, alpha=0.9, eps=1e-10, weight_decay=0, momentum=0., centered=False, + decoupled_decay=False, lr_in_momentum=True): + if not 0.0 <= lr: + raise ValueError("Invalid learning rate: {}".format(lr)) + if not 0.0 <= eps: + raise ValueError("Invalid epsilon value: {}".format(eps)) + if not 0.0 <= momentum: + raise ValueError("Invalid momentum value: {}".format(momentum)) + if not 0.0 <= weight_decay: + raise ValueError("Invalid weight_decay value: {}".format(weight_decay)) + if not 0.0 <= alpha: + raise ValueError("Invalid alpha value: {}".format(alpha)) + + defaults = dict(lr=lr, momentum=momentum, alpha=alpha, eps=eps, centered=centered, weight_decay=weight_decay, + decoupled_decay=decoupled_decay, lr_in_momentum=lr_in_momentum) + super(RMSpropTF, self).__init__(params, defaults) + + def __setstate__(self, state): + super(RMSpropTF, self).__setstate__(state) + for group in self.param_groups: + group.setdefault('momentum', 0) + group.setdefault('centered', False) + + def step(self, closure=None): + """Performs a single optimization step. + + Arguments: + closure (callable, optional): A closure that reevaluates the model + and returns the loss. + """ + loss = None + if closure is not None: + loss = closure() + + for group in self.param_groups: + for p in group['params']: + if p.grad is None: + continue + grad = p.grad.data + if grad.is_sparse: + raise RuntimeError('RMSprop does not support sparse gradients') + state = self.state[p] + + # State initialization + if len(state) == 0: + state['step'] = 0 + state['square_avg'] = torch.ones_like(p.data) # PyTorch inits to zero + if group['momentum'] > 0: + state['momentum_buffer'] = torch.zeros_like(p.data) + if group['centered']: + state['grad_avg'] = torch.zeros_like(p.data) + + square_avg = state['square_avg'] + one_minus_alpha = 1. - group['alpha'] + + state['step'] += 1 + + if group['weight_decay'] != 0: + if 'decoupled_decay' in group and group['decoupled_decay']: + p.data.add_(-group['weight_decay'], p.data) + else: + grad = grad.add(group['weight_decay'], p.data) + + # Tensorflow order of ops for updating squared avg + square_avg.add_(one_minus_alpha, grad.pow(2) - square_avg) + # square_avg.mul_(alpha).addcmul_(1 - alpha, grad, grad) # PyTorch original + + if group['centered']: + grad_avg = state['grad_avg'] + grad_avg.add_(one_minus_alpha, grad - grad_avg) + # grad_avg.mul_(alpha).add_(1 - alpha, grad) # PyTorch original + avg = square_avg.addcmul(-1, grad_avg, grad_avg).add(group['eps']).sqrt_() # eps moved in sqrt + else: + avg = square_avg.add(group['eps']).sqrt_() # eps moved in sqrt + + if group['momentum'] > 0: + buf = state['momentum_buffer'] + # Tensorflow accumulates the LR scaling in the momentum buffer + if 'lr_in_momentum' in group and group['lr_in_momentum']: + buf.mul_(group['momentum']).addcdiv_(group['lr'], grad, avg) + p.data.add_(-buf) + else: + # PyTorch scales the param update by LR + buf.mul_(group['momentum']).addcdiv_(grad, avg) + p.data.add_(-group['lr'], buf) + else: + p.data.addcdiv_(-group['lr'], grad, avg) + + return loss diff --git a/optimizers/timm/scheduler.py b/optimizers/timm/scheduler.py new file mode 100644 index 0000000000000000000000000000000000000000..21d51509c87a0783c6b61986c574a3ed5366e165 --- /dev/null +++ b/optimizers/timm/scheduler.py @@ -0,0 +1,105 @@ +from typing import Dict, Any + +import torch + + +class Scheduler: + """ Parameter Scheduler Base Class + A scheduler base class that can be used to schedule any optimizer parameter groups. + + Unlike the builtin PyTorch schedulers, this is intended to be consistently called + * At the END of each epoch, before incrementing the epoch count, to calculate next epoch's value + * At the END of each optimizer update, after incrementing the update count, to calculate next update's value + + The schedulers built on this should try to remain as stateless as possible (for simplicity). + + This family of schedulers is attempting to avoid the confusion of the meaning of 'last_epoch' + and -1 values for special behaviour. All epoch and update counts must be tracked in the training + code and explicitly passed in to the schedulers on the corresponding step or step_update call. + + Based on ideas from: + * https://github.com/pytorch/fairseq/tree/master/fairseq/optim/lr_scheduler + * https://github.com/allenai/allennlp/tree/master/allennlp/training/learning_rate_schedulers + """ + + def __init__(self, + optimizer: torch.optim.Optimizer, + param_group_field: str, + noise_range_t=None, + noise_type='normal', + noise_pct=0.67, + noise_std=1.0, + noise_seed=None, + initialize: bool = True) -> None: + self.optimizer = optimizer + self.param_group_field = param_group_field + self._initial_param_group_field = f"initial_{param_group_field}" + if initialize: + for i, group in enumerate(self.optimizer.param_groups): + if param_group_field not in group: + raise KeyError(f"{param_group_field} missing from param_groups[{i}]") + group.setdefault(self._initial_param_group_field, group[param_group_field]) + else: + for i, group in enumerate(self.optimizer.param_groups): + if self._initial_param_group_field not in group: + raise KeyError(f"{self._initial_param_group_field} missing from param_groups[{i}]") + self.base_values = [group[self._initial_param_group_field] for group in self.optimizer.param_groups] + self.metric = None # any point to having this for all? + self.noise_range_t = noise_range_t + self.noise_pct = noise_pct + self.noise_type = noise_type + self.noise_std = noise_std + self.noise_seed = noise_seed if noise_seed is not None else 42 + self.update_groups(self.base_values) + + def state_dict(self) -> Dict[str, Any]: + return {key: value for key, value in self.__dict__.items() if key != 'optimizer'} + + def load_state_dict(self, state_dict: Dict[str, Any]) -> None: + self.__dict__.update(state_dict) + + def get_epoch_values(self, epoch: int): + return None + + def get_update_values(self, num_updates: int): + return None + + def step(self, epoch: int, metric: float = None) -> None: + self.metric = metric + values = self.get_epoch_values(epoch) + if values is not None: + values = self._add_noise(values, epoch) + self.update_groups(values) + + def step_update(self, num_updates: int, metric: float = None): + self.metric = metric + values = self.get_update_values(num_updates) + if values is not None: + values = self._add_noise(values, num_updates) + self.update_groups(values) + + def update_groups(self, values): + if not isinstance(values, (list, tuple)): + values = [values] * len(self.optimizer.param_groups) + for param_group, value in zip(self.optimizer.param_groups, values): + param_group[self.param_group_field] = value + + def _add_noise(self, lrs, t): + if self.noise_range_t is not None: + if isinstance(self.noise_range_t, (list, tuple)): + apply_noise = self.noise_range_t[0] <= t < self.noise_range_t[1] + else: + apply_noise = t >= self.noise_range_t + if apply_noise: + g = torch.Generator() + g.manual_seed(self.noise_seed + t) + if self.noise_type == 'normal': + while True: + # resample if noise out of percent limit, brute force but shouldn't spin much + noise = torch.randn(1, generator=g).item() + if abs(noise) < self.noise_pct: + break + else: + noise = 2 * (torch.rand(1, generator=g).item() - 0.5) * self.noise_pct + lrs = [v + v * noise for v in lrs] + return lrs diff --git a/optimizers/timm/sgdp.py b/optimizers/timm/sgdp.py new file mode 100644 index 0000000000000000000000000000000000000000..f4a94aa332d7030a70e888342eb6cc4623d69836 --- /dev/null +++ b/optimizers/timm/sgdp.py @@ -0,0 +1,96 @@ +""" +SGDP Optimizer Implementation copied from https://github.com/clovaai/AdamP/blob/master/adamp/sgdp.py + +Paper: `Slowing Down the Weight Norm Increase in Momentum-based Optimizers` - https://arxiv.org/abs/2006.08217 +Code: https://github.com/clovaai/AdamP + +Copyright (c) 2020-present NAVER Corp. +MIT license +""" + +import torch +import torch.nn as nn +from torch.optim.optimizer import Optimizer, required +import math + +class SGDP(Optimizer): + def __init__(self, params, lr=required, momentum=0, dampening=0, + weight_decay=0, nesterov=False, eps=1e-8, delta=0.1, wd_ratio=0.1): + defaults = dict(lr=lr, momentum=momentum, dampening=dampening, weight_decay=weight_decay, + nesterov=nesterov, eps=eps, delta=delta, wd_ratio=wd_ratio) + super(SGDP, self).__init__(params, defaults) + + def _channel_view(self, x): + return x.view(x.size(0), -1) + + def _layer_view(self, x): + return x.view(1, -1) + + def _cosine_similarity(self, x, y, eps, view_func): + x = view_func(x) + y = view_func(y) + + x_norm = x.norm(dim=1).add_(eps) + y_norm = y.norm(dim=1).add_(eps) + dot = (x * y).sum(dim=1) + + return dot.abs() / x_norm / y_norm + + def _projection(self, p, grad, perturb, delta, wd_ratio, eps): + wd = 1 + expand_size = [-1] + [1] * (len(p.shape) - 1) + for view_func in [self._channel_view, self._layer_view]: + + cosine_sim = self._cosine_similarity(grad, p.data, eps, view_func) + + if cosine_sim.max() < delta / math.sqrt(view_func(p.data).size(1)): + p_n = p.data / view_func(p.data).norm(dim=1).view(expand_size).add_(eps) + perturb -= p_n * view_func(p_n * perturb).sum(dim=1).view(expand_size) + wd = wd_ratio + + return perturb, wd + + return perturb, wd + + def step(self, closure=None): + loss = None + if closure is not None: + loss = closure() + + for group in self.param_groups: + weight_decay = group['weight_decay'] + momentum = group['momentum'] + dampening = group['dampening'] + nesterov = group['nesterov'] + + for p in group['params']: + if p.grad is None: + continue + grad = p.grad.data + state = self.state[p] + + # State initialization + if len(state) == 0: + state['momentum'] = torch.zeros_like(p.data) + + # SGD + buf = state['momentum'] + buf.mul_(momentum).add_(1 - dampening, grad) + if nesterov: + d_p = grad + momentum * buf + else: + d_p = buf + + # Projection + wd_ratio = 1 + if len(p.shape) > 1: + d_p, wd_ratio = self._projection(p, grad, d_p, group['delta'], group['wd_ratio'], group['eps']) + + # Weight decay + if weight_decay != 0: + p.data.mul_(1 - group['lr'] * group['weight_decay'] * wd_ratio / (1-momentum)) + + # Step + p.data.add_(-group['lr'], d_p) + + return loss diff --git a/optimizers/timm/step_lr.py b/optimizers/timm/step_lr.py new file mode 100644 index 0000000000000000000000000000000000000000..f797e1a8cf35999531dd5f1ccbbe09a9d0cf30a9 --- /dev/null +++ b/optimizers/timm/step_lr.py @@ -0,0 +1,63 @@ +""" Step Scheduler + +Basic step LR schedule with warmup, noise. + +Hacked together by / Copyright 2020 Ross Wightman +""" +import math +import torch + +from .scheduler import Scheduler + + +class StepLRScheduler(Scheduler): + """ + """ + + def __init__(self, + optimizer: torch.optim.Optimizer, + decay_t: float, + decay_rate: float = 1., + warmup_t=0, + warmup_lr_init=0, + t_in_epochs=True, + noise_range_t=None, + noise_pct=0.67, + noise_std=1.0, + noise_seed=42, + initialize=True, + ) -> None: + super().__init__( + optimizer, param_group_field="lr", + noise_range_t=noise_range_t, noise_pct=noise_pct, noise_std=noise_std, noise_seed=noise_seed, + initialize=initialize) + + self.decay_t = decay_t + self.decay_rate = decay_rate + self.warmup_t = warmup_t + self.warmup_lr_init = warmup_lr_init + self.t_in_epochs = t_in_epochs + if self.warmup_t: + self.warmup_steps = [(v - warmup_lr_init) / self.warmup_t for v in self.base_values] + super().update_groups(self.warmup_lr_init) + else: + self.warmup_steps = [1 for _ in self.base_values] + + def _get_lr(self, t): + if t < self.warmup_t: + lrs = [self.warmup_lr_init + t * s for s in self.warmup_steps] + else: + lrs = [v * (self.decay_rate ** (t // self.decay_t)) for v in self.base_values] + return lrs + + def get_epoch_values(self, epoch: int): + if self.t_in_epochs: + return self._get_lr(epoch) + else: + return None + + def get_update_values(self, num_updates: int): + if not self.t_in_epochs: + return self._get_lr(num_updates) + else: + return None diff --git a/optimizers/timm/tanh_lr.py b/optimizers/timm/tanh_lr.py new file mode 100644 index 0000000000000000000000000000000000000000..8cc338bb1df7a564d9207b32ab0f59cdf1ef4c59 --- /dev/null +++ b/optimizers/timm/tanh_lr.py @@ -0,0 +1,120 @@ +""" TanH Scheduler + +TanH schedule with warmup, cycle/restarts, noise. + +Hacked together by / Copyright 2020 Ross Wightman +""" +import logging +import math +import numpy as np +import torch + +from .scheduler import Scheduler + + +_logger = logging.getLogger(__name__) + + +class TanhLRScheduler(Scheduler): + """ + Hyberbolic-Tangent decay with restarts. + This is described in the paper https://arxiv.org/abs/1806.01593 + """ + + def __init__(self, + optimizer: torch.optim.Optimizer, + t_initial: int, + lb: float = -6., + ub: float = 4., + t_mul: float = 1., + lr_min: float = 0., + decay_rate: float = 1., + warmup_t=0, + warmup_lr_init=0, + warmup_prefix=False, + cycle_limit=0, + t_in_epochs=True, + noise_range_t=None, + noise_pct=0.67, + noise_std=1.0, + noise_seed=42, + initialize=True) -> None: + super().__init__( + optimizer, param_group_field="lr", + noise_range_t=noise_range_t, noise_pct=noise_pct, noise_std=noise_std, noise_seed=noise_seed, + initialize=initialize) + + assert t_initial > 0 + assert lr_min >= 0 + assert lb < ub + assert cycle_limit >= 0 + assert warmup_t >= 0 + assert warmup_lr_init >= 0 + self.lb = lb + self.ub = ub + self.t_initial = t_initial + self.t_mul = t_mul + self.lr_min = lr_min + self.decay_rate = decay_rate + self.cycle_limit = cycle_limit + self.warmup_t = warmup_t + self.warmup_lr_init = warmup_lr_init + self.warmup_prefix = warmup_prefix + self.t_in_epochs = t_in_epochs + if self.warmup_t: + t_v = self.base_values if self.warmup_prefix else self._get_lr(self.warmup_t) + self.warmup_steps = [(v - warmup_lr_init) / self.warmup_t for v in t_v] + super().update_groups(self.warmup_lr_init) + else: + self.warmup_steps = [1 for _ in self.base_values] + + def _get_lr(self, t): + if t < self.warmup_t: + lrs = [self.warmup_lr_init + t * s for s in self.warmup_steps] + else: + if self.warmup_prefix: + t = t - self.warmup_t + + if self.t_mul != 1: + i = math.floor(math.log(1 - t / self.t_initial * (1 - self.t_mul), self.t_mul)) + t_i = self.t_mul ** i * self.t_initial + t_curr = t - (1 - self.t_mul ** i) / (1 - self.t_mul) * self.t_initial + else: + i = t // self.t_initial + t_i = self.t_initial + t_curr = t - (self.t_initial * i) + + if self.cycle_limit == 0 or (self.cycle_limit > 0 and i < self.cycle_limit): + gamma = self.decay_rate ** i + lr_min = self.lr_min * gamma + lr_max_values = [v * gamma for v in self.base_values] + + tr = t_curr / t_i + lrs = [ + lr_min + 0.5 * (lr_max - lr_min) * (1 - math.tanh(self.lb * (1. - tr) + self.ub * tr)) + for lr_max in lr_max_values + ] + else: + lrs = [self.lr_min * (self.decay_rate ** self.cycle_limit) for _ in self.base_values] + return lrs + + def get_epoch_values(self, epoch: int): + if self.t_in_epochs: + return self._get_lr(epoch) + else: + return None + + def get_update_values(self, num_updates: int): + if not self.t_in_epochs: + return self._get_lr(num_updates) + else: + return None + + def get_cycle_length(self, cycles=0): + if not cycles: + cycles = self.cycle_limit + cycles = max(1, cycles) + if self.t_mul == 1.0: + return self.t_initial * cycles + else: + return int(math.floor(-self.t_initial * (self.t_mul ** cycles - 1) / (1 - self.t_mul))) diff --git a/requirements.txt b/requirements.txt new file mode 100644 index 0000000000000000000000000000000000000000..02884db4078a237e42b40d4f1570679d97a69191 --- /dev/null +++ b/requirements.txt @@ -0,0 +1,35 @@ +torch +loguru +tqdm +smplx +torchvision +soundfile +lmdb +librosa +PyYAML +wandb +tensorboard +pynvml +matplotlib +configargparse +pandas +opencv-python +fasttest +IPython +h5py +einops +git+https://github.com/openai/CLIP.git +textgrid +termcolor +transformers +gdown + +pyrender +trimesh +imageio +moviepy +spaces + +pip install git+https://github.com/RobinWitch/Montreal-Forced-Aligner.git +pip install pgvector +pip install Bio \ No newline at end of file diff --git a/rvq_beatx_train.py b/rvq_beatx_train.py new file mode 100644 index 0000000000000000000000000000000000000000..c817c294b0aff1e64b59a7f054a4f4aa26d72d0c --- /dev/null +++ b/rvq_beatx_train.py @@ -0,0 +1,367 @@ +import pynvml + +def get_gpt_id(): + pynvml.nvmlInit() + gpu_indices = [] + device_count = pynvml.nvmlDeviceGetCount() + for i in range(device_count): + handle = pynvml.nvmlDeviceGetHandleByIndex(i) + memory_info = pynvml.nvmlDeviceGetMemoryInfo(handle) + perf_state = pynvml.nvmlDeviceGetPowerState(handle) + #if perf_state == 8 and memory_info.used < 2000 * 1024 * 1024: + if perf_state == 8 : + gpu_indices.append(i) + assert len(gpu_indices) > 0, "There is no GPU with performance state P8 and low memory usage" + pynvml.nvmlShutdown() + print(f"usalbe gpu ids: {gpu_indices} , now we use {gpu_indices[0]}") + return str(gpu_indices[0]) +dev = get_gpt_id() +import os +os.environ["CUDA_VISIBLE_DEVICES"] = dev +import json + +import torch +import torch.nn as nn +import torch.optim as optim +from torch.utils.tensorboard import SummaryWriter +import logging +import sys + + + +import warnings +warnings.filterwarnings('ignore') +from models.vq.model import RVQVAE + +def get_logger(out_dir): + logger = logging.getLogger('Exp') + logger.setLevel(logging.INFO) + formatter = logging.Formatter("%(asctime)s %(levelname)s %(message)s") + + file_path = os.path.join(out_dir, "run.log") + file_hdlr = logging.FileHandler(file_path) + file_hdlr.setFormatter(formatter) + + strm_hdlr = logging.StreamHandler(sys.stdout) + strm_hdlr.setFormatter(formatter) + + logger.addHandler(file_hdlr) + logger.addHandler(strm_hdlr) + return logger + + +class ReConsLoss(nn.Module): + def __init__(self, recons_loss, nb_joints): + super(ReConsLoss, self).__init__() + + if recons_loss == 'l1': + self.Loss = torch.nn.L1Loss() + elif recons_loss == 'l2' : + self.Loss = torch.nn.MSELoss() + elif recons_loss == 'l1_smooth' : + self.Loss = torch.nn.SmoothL1Loss() + + # 4 global motion associated to root + # 12 local motion (3 local xyz, 3 vel xyz, 6 rot6d) + # 3 global vel xyz + # 4 foot contact + self.nb_joints = nb_joints + self.motion_dim = (nb_joints - 1) * 12 + 4 + 3 + 4 + + def forward(self, motion_pred, motion_gt) : + loss = self.Loss(motion_pred[..., : self.motion_dim], motion_gt[..., :self.motion_dim]) + return loss + + def forward_vel(self, motion_pred, motion_gt) : + loss = self.Loss(motion_pred[..., 4 : (self.nb_joints - 1) * 3 + 4], motion_gt[..., 4 : (self.nb_joints - 1) * 3 + 4]) + return loss + + def my_forward(self,motion_pred,motion_gt,mask) : + loss = self.Loss(motion_pred[..., mask], motion_gt[..., mask]) + return loss + + + +import argparse + +def get_args_parser(): + parser = argparse.ArgumentParser(description='Optimal Transport AutoEncoder training for AIST', + add_help=True, + formatter_class=argparse.ArgumentDefaultsHelpFormatter) + + ## dataloader + parser.add_argument('--dataname', type=str, default='kit', help='dataset directory') + parser.add_argument('--batch-size', default=128, type=int, help='batch size') + parser.add_argument('--window-size', type=int, default=64, help='training motion length') + parser.add_argument('--body_part',type=str,default='whole') + ## optimization + parser.add_argument('--total-iter', default=200000, type=int, help='number of total iterations to run') + parser.add_argument('--warm-up-iter', default=1000, type=int, help='number of total iterations for warmup') + parser.add_argument('--lr', default=2e-4, type=float, help='max learning rate') + parser.add_argument('--lr-scheduler', default=[50000, 400000], nargs="+", type=int, help="learning rate schedule (iterations)") + parser.add_argument('--gamma', default=0.05, type=float, help="learning rate decay") + + parser.add_argument('--weight-decay', default=0.0, type=float, help='weight decay') + parser.add_argument("--commit", type=float, default=0.02, help="hyper-parameter for the commitment loss") + parser.add_argument('--loss-vel', type=float, default=0.1, help='hyper-parameter for the velocity loss') + parser.add_argument('--recons-loss', type=str, default='l2', help='reconstruction loss') + + ## vqvae arch + parser.add_argument("--code-dim", type=int, default=512, help="embedding dimension") + parser.add_argument("--nb-code", type=int, default=512, help="nb of embedding") + parser.add_argument("--mu", type=float, default=0.99, help="exponential moving average to update the codebook") + parser.add_argument("--down-t", type=int, default=2, help="downsampling rate") + parser.add_argument("--stride-t", type=int, default=2, help="stride size") + parser.add_argument("--width", type=int, default=512, help="width of the network") + parser.add_argument("--depth", type=int, default=3, help="depth of the network") + parser.add_argument("--dilation-growth-rate", type=int, default=3, help="dilation growth rate") + parser.add_argument("--output-emb-width", type=int, default=512, help="output embedding width") + parser.add_argument('--vq-act', type=str, default='relu', choices = ['relu', 'silu', 'gelu'], help='dataset directory') + parser.add_argument('--vq-norm', type=str, default=None, help='dataset directory') + + ## quantizer + parser.add_argument("--quantizer", type=str, default='ema_reset', choices = ['ema', 'orig', 'ema_reset', 'reset'], help="eps for optimal transport") + parser.add_argument('--beta', type=float, default=1.0, help='commitment loss in standard VQ') + + ## resume + parser.add_argument("--resume-pth", type=str, default=None, help='resume pth for VQ') + parser.add_argument("--resume-gpt", type=str, default=None, help='resume pth for GPT') + + + ## output directory + parser.add_argument('--out-dir', type=str, default='output_vqfinal/', help='output directory') + parser.add_argument('--results-dir', type=str, default='visual_results/', help='output directory') + parser.add_argument('--visual-name', type=str, default='baseline', help='output directory') + parser.add_argument('--exp-name', type=str, default='exp_debug', help='name of the experiment, will create a file inside out-dir') + ## other + parser.add_argument('--print-iter', default=200, type=int, help='print frequency') + parser.add_argument('--eval-iter', default=1000, type=int, help='evaluation frequency') + parser.add_argument('--seed', default=123, type=int, help='seed for initializing training.') + + parser.add_argument('--vis-gt', action='store_true', help='whether visualize GT motions') + parser.add_argument('--nb-vis', default=20, type=int, help='nb of visualizations') + + + return parser.parse_args() + +def update_lr_warm_up(optimizer, nb_iter, warm_up_iter, lr): + + current_lr = lr * (nb_iter + 1) / (warm_up_iter + 1) + for param_group in optimizer.param_groups: + param_group["lr"] = current_lr + + return optimizer, current_lr + +##### ---- Exp dirs ---- ##### +args = get_args_parser() +torch.manual_seed(args.seed) + +args.out_dir = os.path.join(args.out_dir, f'{args.exp_name}_{args.body_part}') +os.makedirs(args.out_dir, exist_ok = True) + +##### ---- Logger ---- ##### +logger = get_logger(args.out_dir) +writer = SummaryWriter(args.out_dir) +logger.info(json.dumps(vars(args), indent=4, sort_keys=True)) + + +if args.dataname == 'kit' : + dataset_opt_path = 'checkpoints/kit/Comp_v6_KLD005/opt.txt' + args.nb_joints = 21 + +elif args.dataname == 't2m': + dataset_opt_path = 'checkpoints/t2m/Comp_v6_KLD005/opt.txt' + args.nb_joints = 22 + +elif args.dataname == 'h3d623': + dataset_opt_path = 'checkpoints/t2m/Comp_v6_KLD005/opt.txt' + args.nb_joints = 52 + + +##### ---- Dataloader ---- ##### +from dataloaders.mix_sep import CustomDataset +from utils.config import parse_args + +dataset_args = parse_args("configs/beat2_rvqvae.yaml") +build_cache = not os.path.exists(dataset_args.cache_path) + +trainSet = CustomDataset(dataset_args,"train",build_cache = build_cache) +train_loader = torch.utils.data.DataLoader(trainSet, + args.batch_size, + shuffle=True, + #sampler=sampler, + num_workers=8, + #collate_fn=collate_fn, + drop_last = True) + + +def cycle(iterable): + while True: + for x in iterable: + yield x + +train_loader_iter = cycle(train_loader) + + + +if args.body_part in "upper": + joints = [3,6,9,12,13,14,15,16,17,18,19,20,21] + upper_body_mask = [] + for i in joints: + upper_body_mask.extend([i*6, i*6+1, i*6+2, i*6+3, i*6+4, i*6+5]) + mask = upper_body_mask + rec_mask = list(range(len(mask))) + + +elif args.body_part in "hands": + + joints = list(range(25,55)) + hands_body_mask = [] + for i in joints: + hands_body_mask.extend([i*6, i*6+1, i*6+2, i*6+3, i*6+4, i*6+5]) + mask = hands_body_mask + rec_mask = list(range(len(mask))) + + +elif args.body_part in "lower": + joints = [0,1,2,4,5,7,8,10,11] + lower_body_mask = [] + for i in joints: + lower_body_mask.extend([i*6, i*6+1, i*6+2, i*6+3, i*6+4, i*6+5]) + mask = lower_body_mask + rec_mask = list(range(len(mask))) + +elif args.body_part in "lower_trans": + joints = [0,1,2,4,5,7,8,10,11] + lower_body_mask = [] + for i in joints: + lower_body_mask.extend([i*6, i*6+1, i*6+2, i*6+3, i*6+4, i*6+5]) + lower_body_mask.extend([330,331,332]) + mask = lower_body_mask + rec_mask = list(range(len(mask))) + + + + +##### ---- Network ---- ##### +if args.body_part in "upper": + dim_pose = 78 +elif args.body_part in "hands": + dim_pose = 180 +elif args.body_part in "lower": + dim_pose = 54 +elif args.body_part in "lower_trans": + dim_pose = 57 +elif args.body_part in "whole": + dim_pose = 312 + + +args.num_quantizers = 6 +args.shared_codebook = False +args.quantize_dropout_prob = 0.2 +net = RVQVAE(args, + dim_pose, + args.nb_code, + args.code_dim, + args.code_dim, + args.down_t, + args.stride_t, + args.width, + args.depth, + args.dilation_growth_rate, + args.vq_act, + args.vq_norm) + + +if args.resume_pth : + logger.info('loading checkpoint from {}'.format(args.resume_pth)) + ckpt = torch.load(args.resume_pth, map_location='cpu') + net.load_state_dict(ckpt['net'], strict=True) +net.train() +net.cuda() + +##### ---- Optimizer & Scheduler ---- ##### +optimizer = optim.AdamW(net.parameters(), lr=args.lr, betas=(0.9, 0.99), weight_decay=args.weight_decay) +scheduler = torch.optim.lr_scheduler.MultiStepLR(optimizer, milestones=args.lr_scheduler, gamma=args.gamma) + + +Loss = ReConsLoss(args.recons_loss, args.nb_joints) + +##### ------ warm-up ------- ##### +avg_recons, avg_perplexity, avg_commit = 0., 0., 0. + +for nb_iter in range(1, args.warm_up_iter): + + optimizer, current_lr = update_lr_warm_up(optimizer, nb_iter, args.warm_up_iter, args.lr) + + gt_motion = next(train_loader_iter) + gt_motion = gt_motion[...,mask].cuda().float() # (bs, 64, dim) + + pred_motion, loss_commit, perplexity = net(gt_motion).values() + loss_motion = Loss.my_forward(pred_motion, gt_motion,rec_mask) + loss_vel = 0#Loss.my_forward(pred_motion, gt_motion,vel_mask) + + loss = loss_motion + args.commit * loss_commit + args.loss_vel * loss_vel + + optimizer.zero_grad() + loss.backward() + optimizer.step() + + avg_recons += loss_motion.item() + avg_perplexity += perplexity.item() + avg_commit += loss_commit.item() + + if nb_iter % args.print_iter == 0 : + avg_recons /= args.print_iter + avg_perplexity /= args.print_iter + avg_commit /= args.print_iter + + logger.info(f"Warmup. Iter {nb_iter} : lr {current_lr:.5f} \t Commit. {avg_commit:.5f} \t PPL. {avg_perplexity:.2f} \t Recons. {avg_recons:.5f}") + + avg_recons, avg_perplexity, avg_commit = 0., 0., 0. + +##### ---- Training ---- ##### +avg_recons, avg_perplexity, avg_commit = 0., 0., 0. +#best_fid, best_iter, best_div, best_top1, best_top2, best_top3, best_matching, writer, logger = eval_trans.evaluation_vqvae(args.out_dir, val_loader, net, logger, writer, 0, best_fid=1000, best_iter=0, best_div=100, best_top1=0, best_top2=0, best_top3=0, best_matching=100, eval_wrapper=eval_wrapper) +args.eval_iter = args.eval_iter * 10 +for nb_iter in range(1, args.total_iter + 1): + + gt_motion = next(train_loader_iter) + gt_motion = gt_motion[...,mask].cuda().float() # bs, nb_joints, joints_dim, seq_len + + pred_motion, loss_commit, perplexity = net(gt_motion) + loss_motion = Loss.my_forward(pred_motion, gt_motion,rec_mask) + loss_vel = 0 + + loss = loss_motion + args.commit * loss_commit + args.loss_vel * loss_vel + + optimizer.zero_grad() + loss.backward() + optimizer.step() + scheduler.step() + + avg_recons += loss_motion.item() + avg_perplexity += perplexity.item() + avg_commit += loss_commit.item() + + if nb_iter % args.print_iter == 0 : + avg_recons /= args.print_iter + avg_perplexity /= args.print_iter + avg_commit /= args.print_iter + + writer.add_scalar('./Train/L1', avg_recons, nb_iter) + writer.add_scalar('./Train/PPL', avg_perplexity, nb_iter) + writer.add_scalar('./Train/Commit', avg_commit, nb_iter) + + logger.info(f"Train. Iter {nb_iter} : \t Commit. {avg_commit:.5f} \t PPL. {avg_perplexity:.2f} \t Recons. {avg_recons:.5f}") + + avg_recons, avg_perplexity, avg_commit = 0., 0., 0., + + # if nb_iter % args.eval_iter==0 : + # best_fid, best_iter, best_div, best_top1, best_top2, best_top3, best_matching, writer, logger = eval_trans.evaluation_vqvae(args.out_dir, val_loader, net, logger, writer, nb_iter, best_fid, best_iter, best_div, best_top1, best_top2, best_top3, best_matching, eval_wrapper=eval_wrapper) + # eval_trans.my_evaluation_vqvae(args.out_dir, val_loader, net, logger, writer) + if nb_iter % args.eval_iter==0 : + torch.save({'net' : net.state_dict()}, os.path.join(args.out_dir, f'net_{nb_iter}.pth')) + #net.load_state_dict('/mnt/fu06/chenbohong/T2M-GPT/output/VQVAE/net_last.pth') + +# run command diff --git a/system_utils.py b/system_utils.py new file mode 100644 index 0000000000000000000000000000000000000000..ed2de71ffa059fe5c87465cb17a0d780f8a33210 --- /dev/null +++ b/system_utils.py @@ -0,0 +1,18 @@ +import pynvml + +def get_gpt_id(): + # return "1" + pynvml.nvmlInit() + gpu_indices = [] + device_count = pynvml.nvmlDeviceGetCount() + for i in range(device_count): + handle = pynvml.nvmlDeviceGetHandleByIndex(i) + memory_info = pynvml.nvmlDeviceGetMemoryInfo(handle) + perf_state = pynvml.nvmlDeviceGetPowerState(handle) + #if perf_state == 8 and memory_info.used < 2000 * 1024 * 1024: + if perf_state == 8: + gpu_indices.append(i) + assert len(gpu_indices) > 0, "There is no GPU with performance state P8 and low memory usage" + pynvml.nvmlShutdown() + print(f"usalbe gpu ids: {gpu_indices} , now we use {gpu_indices[-1]}") + return str(gpu_indices[-1]) diff --git a/test.py b/test.py new file mode 100644 index 0000000000000000000000000000000000000000..3ef9ccdeaa14d4545e2004aa3549da7439c225c7 --- /dev/null +++ b/test.py @@ -0,0 +1,252 @@ +from system_utils import get_gpt_id +dev = get_gpt_id() +import os +os.environ["CUDA_VISIBLE_DEVICES"] = dev +import signal +import time +import csv +import sys +import warnings +import random +import torch +import torch.nn as nn +import torch.nn.functional as F +import torch.distributed as dist +from torch.nn.parallel import DistributedDataParallel as DDP +import torch.multiprocessing as mp +import numpy as np +import time +import pprint +from loguru import logger +import smplx +from torch.utils.tensorboard import SummaryWriter +import wandb +import matplotlib.pyplot as plt +from utils import config, logger_tools, other_tools, metric +from dataloaders import data_tools +from dataloaders.build_vocab import Vocab +from optimizers.optim_factory import create_optimizer +from optimizers.scheduler_factory import create_scheduler +from optimizers.loss_factory import get_loss_func +import socket + +class BaseTrainer(object): + def __init__(self, args): + self.args = args + self.rank = dist.get_rank() + self.checkpoint_path = args.out_path + "custom/" + args.name + args.notes + "/" #wandb.run.dir #args.cache_path+args.out_path+"/"+args.name + if self.rank==0: + if self.args.stat == "ts": + self.writer = SummaryWriter(log_dir=args.out_path + "custom/" + args.name + args.notes + "/") + else: + wandb.init(project=args.project, entity="liu1997", dir=args.out_path, name=args.name[12:] + args.notes) + wandb.config.update(args) + self.writer = None + #self.test_demo = args.data_path + args.test_data_path + "bvh_full/" + # self.train_data = __import__(f"dataloaders.{args.dataset}", fromlist=["something"]).CustomDataset(args, "train") + # self.train_loader = torch.utils.data.DataLoader( + # self.train_data, + # batch_size=args.batch_size, + # shuffle=False if args.ddp else True, + # num_workers=args.loader_workers, + # drop_last=True, + # sampler=torch.utils.data.distributed.DistributedSampler(self.train_data) if args.ddp else None, + # ) + # self.train_length = len(self.train_loader) + # logger.info(f"Init train dataloader success") + + # self.val_data = __import__(f"dataloaders.{args.dataset}", fromlist=["something"]).CustomDataset(args, "val") + # self.val_loader = torch.utils.data.DataLoader( + # self.val_data, + # batch_size=args.batch_size, + # shuffle=False, + # num_workers=args.loader_workers, + # drop_last=False, + # sampler=torch.utils.data.distributed.DistributedSampler(self.val_data) if args.ddp else None, + # ) + # logger.info(f"Init val dataloader success") + if self.rank == 0: + self.test_data = __import__(f"dataloaders.{args.dataset}", fromlist=["something"]).CustomDataset(args, "test") + self.test_loader = torch.utils.data.DataLoader( + self.test_data, + batch_size=1, + shuffle=False, + num_workers=args.loader_workers, + drop_last=False, + ) + logger.info(f"Init test dataloader success") + model_module = __import__(f"models.{args.model}", fromlist=["something"]) + + if args.ddp: + self.model = getattr(model_module, args.g_name)(args).to(self.rank) + process_group = torch.distributed.new_group() + self.model = torch.nn.SyncBatchNorm.convert_sync_batchnorm(self.model, process_group) + self.model = DDP(self.model, device_ids=[self.rank], output_device=self.rank, + broadcast_buffers=False, find_unused_parameters=False) + else: + self.model = torch.nn.DataParallel(getattr(model_module, args.g_name)(args), args.gpus).cuda() + + if self.rank == 0: + logger.info(self.model) + logger.info(f"init {args.g_name} success") + if args.stat == "wandb": + wandb.watch(self.model) + + # if args.d_name is not None: + # if args.ddp: + # self.d_model = getattr(model_module, args.d_name)(args).to(self.rank) + # self.d_model = torch.nn.SyncBatchNorm.convert_sync_batchnorm(self.d_model, process_group) + # self.d_model = DDP(self.d_model, device_ids=[self.rank], output_device=self.rank, + # broadcast_buffers=False, find_unused_parameters=False) + # else: + # self.d_model = torch.nn.DataParallel(getattr(model_module, args.d_name)(args), args.gpus).cuda() + # if self.rank == 0: + # logger.info(self.d_model) + # logger.info(f"init {args.d_name} success") + # if args.stat == "wandb": + # wandb.watch(self.d_model) + # self.opt_d = create_optimizer(args, self.d_model, lr_weight=args.d_lr_weight) + # self.opt_d_s = create_scheduler(args, self.opt_d) + + if args.e_name is not None: + """ + bugs on DDP training using eval_model, using additional eval_copy for evaluation + """ + eval_model_module = __import__(f"models.{args.eval_model}", fromlist=["something"]) + # eval copy is for single card evaluation + if self.args.ddp: + self.eval_model = getattr(eval_model_module, args.e_name)(args).to(self.rank) + self.eval_copy = getattr(eval_model_module, args.e_name)(args).to(self.rank) + else: + self.eval_model = getattr(eval_model_module, args.e_name)(args) + self.eval_copy = getattr(eval_model_module, args.e_name)(args).to(self.rank) + + #if self.rank == 0: + other_tools.load_checkpoints(self.eval_copy, args.data_path+args.e_path, args.e_name) + other_tools.load_checkpoints(self.eval_model, args.data_path+args.e_path, args.e_name) + if self.args.ddp: + self.eval_model = torch.nn.SyncBatchNorm.convert_sync_batchnorm(self.eval_model, process_group) + self.eval_model = DDP(self.eval_model, device_ids=[self.rank], output_device=self.rank, + broadcast_buffers=False, find_unused_parameters=False) + self.eval_model.eval() + self.eval_copy.eval() + if self.rank == 0: + logger.info(self.eval_model) + logger.info(f"init {args.e_name} success") + if args.stat == "wandb": + wandb.watch(self.eval_model) + self.smplx = smplx.create( + self.args.data_path_1+"smplx_models/", + model_type='smplx', + gender='NEUTRAL_2020', + use_face_contour=False, + num_betas=300, + num_expression_coeffs=100, + ext='npz', + use_pca=False, + ).to(self.rank).eval() + self.alignmenter = metric.alignment(0.3, 7, self.train_data.avg_vel, upper_body=[3,6,9,12,13,14,15,16,17,18,19,20,21]) if self.rank == 0 else None + self.align_mask = 60 + self.l1_calculator = metric.L1div() if self.rank == 0 else None + + def train_recording(self, epoch, its, t_data, t_train, mem_cost, lr_g, lr_d=None): + pstr = "[%03d][%03d/%03d] "%(epoch, its, self.train_length) + for name, states in self.tracker.loss_meters.items(): + metric = states['train'] + if metric.count > 0: + pstr += "{}: {:.3f}\t".format(name, metric.avg) + self.writer.add_scalar(f"train/{name}", metric.avg, epoch*self.train_length+its) if self.args.stat == "ts" else wandb.log({name: metric.avg}, step=epoch*self.train_length+its) + pstr += "glr: {:.1e}\t".format(lr_g) + self.writer.add_scalar("lr/glr", lr_g, epoch*self.train_length+its) if self.args.stat == "ts" else wandb.log({'glr': lr_g}, step=epoch*self.train_length+its) + if lr_d is not None: + pstr += "dlr: {:.1e}\t".format(lr_d) + self.writer.add_scalar("lr/dlr", lr_d, epoch*self.train_length+its) if self.args.stat == "ts" else wandb.log({'dlr': lr_d}, step=epoch*self.train_length+its) + pstr += "dtime: %04d\t"%(t_data*1000) + pstr += "ntime: %04d\t"%(t_train*1000) + pstr += "mem: {:.2f} ".format(mem_cost*len(self.args.gpus)) + logger.info(pstr) + + def val_recording(self, epoch): + pstr_curr = "Curr info >>>> " + pstr_best = "Best info >>>> " + for name, states in self.tracker.loss_meters.items(): + metric = states['val'] + if metric.count > 0: + pstr_curr += "{}: {:.3f} \t".format(name, metric.avg) + if epoch != 0: + if self.args.stat == "ts": + self.writer.add_scalars(f"val/{name}", {name+"_val":metric.avg, name+"_train":states['train'].avg}, epoch*self.train_length) + else: + wandb.log({name+"_val": metric.avg, name+"_train":states['train'].avg}, step=epoch*self.train_length) + new_best_train, new_best_val = self.tracker.update_and_plot(name, epoch, self.checkpoint_path+f"{name}_{self.args.name+self.args.notes}.png") + if new_best_val: + other_tools.save_checkpoints(os.path.join(self.checkpoint_path, f"{name}.bin"), self.model, opt=None, epoch=None, lrs=None) + for k, v in self.tracker.values.items(): + metric = v['val']['best'] + if self.tracker.loss_meters[k]['val'].count > 0: + pstr_best += "{}: {:.3f}({:03d})\t".format(k, metric['value'], metric['epoch']) + logger.info(pstr_curr) + logger.info(pstr_best) + + def test_recording(self, dict_name, value, epoch): + self.tracker.update_meter(dict_name, "test", value) + _ = self.tracker.update_values(dict_name, 'test', epoch) + +@logger.catch +def main_worker(rank, world_size, args): + #os.environ['TRANSFORMERS_CACHE'] = args.data_path_1 + "hub/" + if not sys.warnoptions: + warnings.simplefilter("ignore") + dist.init_process_group(backend="nccl", rank=rank, world_size=world_size) + + logger_tools.set_args_and_logger(args, rank) + other_tools.set_random_seed(args) + other_tools.print_exp_info(args) + + # return one intance of trainer + trainer = __import__(f"{args.trainer}_trainer", fromlist=["something"]).CustomTrainer(args) if args.trainer != "base" else BaseTrainer(args) + other_tools.load_checkpoints(trainer.model, args.test_ckpt, args.g_name) + trainer.test(999) + + + + + + +def is_port_in_use(port): + with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as s: + return s.connect_ex(('127.0.0.1', port)) == 0 + +def find_available_port(start_port, end_port): + for port in range(start_port, end_port + 1): + if not is_port_in_use(port): + return port + return None + +if __name__ == "__main__": + os.environ["MASTER_ADDR"]='127.0.0.1' + # 设置初始的端口号 + start_port = 21575 + end_port = 21699 + os.environ["MASTER_PORT"]=f'16{dev}75' + # 检测初始指定的端口是否被占用 + master_port = int(os.environ.get("MASTER_PORT", start_port)) + if is_port_in_use(master_port): + new_port = find_available_port(start_port, end_port) + if new_port is not None: + os.environ["MASTER_PORT"] = str(new_port) + print(f"Port {master_port} is in use. Switched to port {new_port}") + else: + print("No available ports in the range.") + #os.environ["MASTER_PORT"]=f'16{dev}75' + #os.environ["TORCH_DISTRIBUTED_DEBUG"] = "DETAIL" + args = config.parse_args() + if args.ddp: + mp.set_start_method("spawn", force=True) + mp.spawn( + main_worker, + args=(len(args.gpus), args,), + nprocs=len(args.gpus), + ) + else: + main_worker(0, 1, args) \ No newline at end of file diff --git a/tmp.py b/tmp.py new file mode 100644 index 0000000000000000000000000000000000000000..91e5bb45cca254a74aca73dce3ae70f4ff092c55 --- /dev/null +++ b/tmp.py @@ -0,0 +1,22 @@ +import torch +from transformers import pipeline +import librosa +import os +os.environ["http_proxy"] = "http://10.76.5.191:7890" +os.environ["https_proxy"] = "http://10.76.5.191:7890" +device = "cuda:0" if torch.cuda.is_available() else "cpu" + +pipe = pipeline( + "automatic-speech-recognition", + model="openai/whisper-tiny.en", + chunk_length_s=30, + device=device, +) + +audio,sr = librosa.load("/mnt/data3/cbh/SynTalker/demo/test3/1_wayne_0_2_2.wav",sr=None) +sample = audio + +prediction = pipe(sample.copy(), batch_size=8)["text"] + +# # we can also return timestamps for the predictions +# prediction = pipe(sample.copy(), batch_size=8, return_timestamps=True)["chunks"] diff --git a/train.py b/train.py new file mode 100644 index 0000000000000000000000000000000000000000..e63ad12003038b7c6cf3b51a787269c46a92b199 --- /dev/null +++ b/train.py @@ -0,0 +1,317 @@ +from system_utils import get_gpt_id +dev = get_gpt_id() +import os +os.environ["CUDA_VISIBLE_DEVICES"] = dev +import signal +import time +import csv +import sys +import warnings +import random +import torch +import torch.nn as nn +import torch.nn.functional as F +import torch.distributed as dist +from torch.nn.parallel import DistributedDataParallel as DDP +import torch.multiprocessing as mp +import numpy as np +import time +import pprint +from loguru import logger +import smplx +from torch.utils.tensorboard import SummaryWriter +import wandb +import matplotlib.pyplot as plt +from utils import config, logger_tools, other_tools, metric +from dataloaders import data_tools +from dataloaders.build_vocab import Vocab +from optimizers.optim_factory import create_optimizer +from optimizers.scheduler_factory import create_scheduler +from optimizers.loss_factory import get_loss_func + + +class BaseTrainer(object): + def __init__(self, args): + self.args = args + self.rank = 0 + self.checkpoint_path = args.out_path + "custom/" + args.name + args.notes + "/" #wandb.run.dir #args.cache_path+args.out_path+"/"+args.name + if self.rank==0: + if self.args.stat == "ts": + self.writer = SummaryWriter(log_dir=args.out_path + "custom/" + args.name + args.notes + "/") + else: + wandb.init(project=args.project, entity="liu1997", dir=args.out_path, name=args.name[12:] + args.notes) + wandb.config.update(args) + self.writer = None + #self.test_demo = args.data_path + args.test_data_path + "bvh_full/" + self.train_data_beatx = __import__(f"dataloaders.{args.dataset}", fromlist=["something"]).CustomDataset(args, "train") + self.train_data = self.train_data_beatx + if args.use_amass: + self.train_data_amass = __import__(f"dataloaders.amass_sep_lower", fromlist=["something"]).CustomDataset(args, "train") + beatx_len = len(self.train_data_beatx) + smplx_len = len(self.train_data_amass) + self.train_data = torch.utils.data.ConcatDataset([self.train_data_beatx,self.train_data_amass,]) + + self.train_loader = torch.utils.data.DataLoader( + self.train_data, + batch_size=args.batch_size, + shuffle=False if args.ddp else True, + num_workers=args.loader_workers, + drop_last=True, + sampler=torch.utils.data.distributed.DistributedSampler(self.train_data) if args.ddp else None, + ) + self.train_length = len(self.train_loader) + logger.info(f"Init train dataloader success") + + # self.val_data = __import__(f"dataloaders.{args.dataset}", fromlist=["something"]).CustomDataset(args, "val") + # self.val_loader = torch.utils.data.DataLoader( + # self.val_data, + # batch_size=args.batch_size, + # shuffle=False, + # num_workers=args.loader_workers, + # drop_last=False, + # sampler=torch.utils.data.distributed.DistributedSampler(self.val_data) if args.ddp else None, + # ) + logger.info(f"Init val dataloader success") + if self.rank == 0: + self.test_data = __import__(f"dataloaders.{args.dataset}", fromlist=["something"]).CustomDataset(args, "test") + self.test_loader = torch.utils.data.DataLoader( + self.test_data, + batch_size=1, + shuffle=False, + num_workers=args.loader_workers, + drop_last=False, + ) + logger.info(f"Init test dataloader success") + model_module = __import__(f"models.{args.model}", fromlist=["something"]) + + if args.ddp: + self.model = getattr(model_module, args.g_name)(args).to(self.rank) + process_group = torch.distributed.new_group() + self.model = torch.nn.SyncBatchNorm.convert_sync_batchnorm(self.model, process_group) + self.model = DDP(self.model, device_ids=[self.rank], output_device=self.rank, + broadcast_buffers=False, find_unused_parameters=False) + else: + self.model = torch.nn.DataParallel(getattr(model_module, args.g_name)(args), args.gpus).cuda() + + if self.rank == 0: + logger.info(self.model) + logger.info(f"init {args.g_name} success") + if args.stat == "wandb": + wandb.watch(self.model) + + if args.d_name is not None: + if args.ddp: + self.d_model = getattr(model_module, args.d_name)(args).to(self.rank) + self.d_model = torch.nn.SyncBatchNorm.convert_sync_batchnorm(self.d_model, process_group) + self.d_model = DDP(self.d_model, device_ids=[self.rank], output_device=self.rank, + broadcast_buffers=False, find_unused_parameters=False) + else: + self.d_model = torch.nn.DataParallel(getattr(model_module, args.d_name)(args), args.gpus).cuda() + if self.rank == 0: + logger.info(self.d_model) + logger.info(f"init {args.d_name} success") + if args.stat == "wandb": + wandb.watch(self.d_model) + self.opt_d = create_optimizer(args, self.d_model, lr_weight=args.d_lr_weight) + self.opt_d_s = create_scheduler(args, self.opt_d) + + if args.e_name is not None: + """ + bugs on DDP training using eval_model, using additional eval_copy for evaluation + """ + eval_model_module = __import__(f"models.{args.eval_model}", fromlist=["something"]) + # eval copy is for single card evaluation + if self.args.ddp: + self.eval_model = getattr(eval_model_module, args.e_name)(args).to(self.rank) + self.eval_copy = getattr(eval_model_module, args.e_name)(args).to(self.rank) + else: + self.eval_model = getattr(eval_model_module, args.e_name)(args) + self.eval_copy = getattr(eval_model_module, args.e_name)(args).to(self.rank) + + #if self.rank == 0: + other_tools.load_checkpoints(self.eval_copy, args.data_path+args.e_path, args.e_name) + other_tools.load_checkpoints(self.eval_model, args.data_path+args.e_path, args.e_name) + if self.args.ddp: + self.eval_model = torch.nn.SyncBatchNorm.convert_sync_batchnorm(self.eval_model, process_group) + self.eval_model = DDP(self.eval_model, device_ids=[self.rank], output_device=self.rank, + broadcast_buffers=False, find_unused_parameters=False) + self.eval_model.eval() + self.eval_copy.eval() + if self.rank == 0: + logger.info(self.eval_model) + logger.info(f"init {args.e_name} success") + if args.stat == "wandb": + wandb.watch(self.eval_model) + self.opt = create_optimizer(args, self.model) + self.opt_s = create_scheduler(args, self.opt) + self.smplx = smplx.create( + self.args.data_path_1+"smplx_models/", + model_type='smplx', + gender='NEUTRAL_2020', + use_face_contour=False, + num_betas=300, + num_expression_coeffs=100, + ext='npz', + use_pca=False, + ).to(self.rank).eval() + self.alignmenter = metric.alignment(0.3, 7, self.train_data.avg_vel, upper_body=[3,6,9,12,13,14,15,16,17,18,19,20,21]) if self.rank == 0 else None + self.align_mask = 60 + self.l1_calculator = metric.L1div() if self.rank == 0 else None + + + def inverse_selection(self, filtered_t, selection_array, n): + original_shape_t = np.zeros((n, selection_array.size)) + selected_indices = np.where(selection_array == 1)[0] + for i in range(n): + original_shape_t[i, selected_indices] = filtered_t[i] + return original_shape_t + + # def inverse_selection_6d(self, filtered_t, selection_array, n): + # original_shape_t = np.zeros((n, selection_array.size)) + # selected_indices = np.where(selection_array == 1)[0] + # new_selected_indices = np.zeros((n, selected_indices.size*2)) + # new_selected_indices[:, ::2] = selected_indices + # new_selected_indices[:, 1::2] = selected_indices + # selected_indices = new_selected_indices.astype(np.bool) + # for i in range(n): + # original_shape_t[i, selected_indices] = filtered_t[i] + # return original_shape_t + + def inverse_selection_tensor(self, filtered_t, selection_array, n): + selection_array = torch.from_numpy(selection_array).cuda() + selected_indices = torch.where(selection_array == 1)[0] + if len(filtered_t.shape) == 2: + original_shape_t = torch.zeros((n, 165)).cuda() + for i in range(n): + original_shape_t[i, selected_indices] = filtered_t[i] + elif len(filtered_t.shape) == 3: + bs, n, _ = filtered_t.shape + original_shape_t = torch.zeros((bs, n, 165), device='cuda') + expanded_indices = selected_indices.unsqueeze(0).unsqueeze(0).expand(bs, n, -1) + original_shape_t.scatter_(2, expanded_indices, filtered_t) + return original_shape_t + + def inverse_selection_tensor_6d(self, filtered_t, selection_array, n): + new_selected_array = np.zeros((330)) + new_selected_array[::2] = selection_array + new_selected_array[1::2] = selection_array + selection_array = new_selected_array + selection_array = torch.from_numpy(selection_array).cuda() + selected_indices = torch.where(selection_array == 1)[0] + if len(filtered_t.shape) == 2: + original_shape_t = torch.zeros((n, 330)).cuda() + for i in range(n): + original_shape_t[i, selected_indices] = filtered_t[i] + elif len(filtered_t.shape) == 3: + bs, n, _ = filtered_t.shape + original_shape_t = torch.zeros((bs, n, 330), device='cuda') + expanded_indices = selected_indices.unsqueeze(0).unsqueeze(0).expand(bs, n, -1) + original_shape_t.scatter_(2, expanded_indices, filtered_t) + return original_shape_t + + def train_recording(self, epoch, its, t_data, t_train, mem_cost, lr_g, lr_d=None): + pstr = "[%03d][%03d/%03d] "%(epoch, its, self.train_length) + for name, states in self.tracker.loss_meters.items(): + metric = states['train'] + if metric.count > 0: + pstr += "{}: {:.3f}\t".format(name, metric.avg) + self.writer.add_scalar(f"train/{name}", metric.avg, epoch*self.train_length+its) if self.args.stat == "ts" else wandb.log({name: metric.avg}, step=epoch*self.train_length+its) + pstr += "glr: {:.1e}\t".format(lr_g) + self.writer.add_scalar("lr/glr", lr_g, epoch*self.train_length+its) if self.args.stat == "ts" else wandb.log({'glr': lr_g}, step=epoch*self.train_length+its) + if lr_d is not None: + pstr += "dlr: {:.1e}\t".format(lr_d) + self.writer.add_scalar("lr/dlr", lr_d, epoch*self.train_length+its) if self.args.stat == "ts" else wandb.log({'dlr': lr_d}, step=epoch*self.train_length+its) + pstr += "dtime: %04d\t"%(t_data*1000) + pstr += "ntime: %04d\t"%(t_train*1000) + pstr += "mem: {:.2f} ".format(mem_cost*len(self.args.gpus)) + logger.info(pstr) + + def val_recording(self, epoch): + pstr_curr = "Curr info >>>> " + pstr_best = "Best info >>>> " + for name, states in self.tracker.loss_meters.items(): + metric = states['val'] + if metric.count > 0: + pstr_curr += "{}: {:.3f} \t".format(name, metric.avg) + if epoch != 0: + if self.args.stat == "ts": + self.writer.add_scalars(f"val/{name}", {name+"_val":metric.avg, name+"_train":states['train'].avg}, epoch*self.train_length) + else: + wandb.log({name+"_val": metric.avg, name+"_train":states['train'].avg}, step=epoch*self.train_length) + new_best_train, new_best_val = self.tracker.update_and_plot(name, epoch, self.checkpoint_path+f"{name}_{self.args.name+self.args.notes}.png") + if new_best_val: + other_tools.save_checkpoints(os.path.join(self.checkpoint_path, f"{name}.bin"), self.model, opt=None, epoch=None, lrs=None) + for k, v in self.tracker.values.items(): + metric = v['val']['best'] + if self.tracker.loss_meters[k]['val'].count > 0: + pstr_best += "{}: {:.3f}({:03d})\t".format(k, metric['value'], metric['epoch']) + logger.info(pstr_curr) + logger.info(pstr_best) + + def test_recording(self, dict_name, value, epoch): + self.tracker.update_meter(dict_name, "test", value) + _ = self.tracker.update_values(dict_name, 'test', epoch) + +@logger.catch +def main_worker(rank, world_size, args): + #os.environ['TRANSFORMERS_CACHE'] = args.data_path_1 + "hub/" + if not sys.warnoptions: + warnings.simplefilter("ignore") + dist.init_process_group(backend="nccl", rank=rank, world_size=world_size) + + logger_tools.set_args_and_logger(args, rank) + other_tools.set_random_seed(args) + other_tools.print_exp_info(args) + + # return one intance of trainer + trainer = __import__(f"{args.trainer}_trainer", fromlist=["something"]).CustomTrainer(args) if args.trainer != "base" else BaseTrainer(args) + logger.info("Training from scratch ...") + start_time = time.time() + for epoch in range(args.epochs+1): + if args.ddp: trainer.val_loader.sampler.set_epoch(epoch) + # trainer.val(epoch) + # if (epoch) % args.test_period == 1: trainer.val(epoch) + epoch_time = time.time()-start_time + if trainer.rank == 0: logger.info("Time info >>>> elapsed: %.2f mins\t"%(epoch_time/60)+"remain: %.2f mins"%((args.epochs/(epoch+1e-7)-1)*epoch_time/60)) + if epoch != args.epochs: + if args.ddp: trainer.train_loader.sampler.set_epoch(epoch) + trainer.tracker.reset() + trainer.train(epoch) + if args.debug: + other_tools.save_checkpoints(os.path.join(trainer.checkpoint_path, f"last_{epoch}.bin"), trainer.model, opt=None, epoch=None, lrs=None) + other_tools.load_checkpoints(trainer.model, os.path.join(trainer.checkpoint_path, f"last_{epoch}.bin"), args.g_name) + #other_tools.load_checkpoints(trainer.model, "/home/s24273/datasets/hub/pretrained_vq/last_140.bin", args.g_name) + trainer.test(epoch) + if (epoch) % args.test_period == 0 and epoch !=0: + if rank == 0: + other_tools.save_checkpoints(os.path.join(trainer.checkpoint_path, f"last_{epoch}.bin"), trainer.model, opt=None, epoch=None, lrs=None) + #trainer.test(epoch) + + if rank == 0: + for k, v in trainer.tracker.values.items(): + if trainer.tracker.loss_meters[k]['val'].count > 0: + other_tools.load_checkpoints(trainer.model, os.path.join(trainer.checkpoint_path, f"{k}.bin"), args.g_name) + logger.info(f"inference on ckpt {k}_val_{v['val']['best']['epoch']}:") + trainer.test(v['valb']['est']['epoch']) + other_tools.record_trial(args, trainer.tracker) + wandb.log({"fid_test": trainer.tracker["fid"]["test"]["best"]}) + if args.stat == "ts": + trainer.writer.close() + else: + wandb.finish() + + +if __name__ == "__main__": + os.environ["MASTER_ADDR"]='127.0.0.1' + os.environ["MASTER_PORT"]=f'9{dev}75' + #os.environ["TORCH_DISTRIBUTED_DEBUG"] = "DETAIL" + args = config.parse_args() + if args.ddp: + mp.set_start_method("spawn", force=True) + mp.spawn( + main_worker, + args=(len(args.gpus), args,), + nprocs=len(args.gpus), + ) + else: + main_worker(0, 1, args) \ No newline at end of file diff --git a/utils/config.py b/utils/config.py new file mode 100644 index 0000000000000000000000000000000000000000..15561ec297d4745cb99fa7fcb0a8e7280d1122a4 --- /dev/null +++ b/utils/config.py @@ -0,0 +1,316 @@ +import configargparse +import time +import json +import yaml +import os + +def str2bool(v): + """ from https://stackoverflow.com/a/43357954/1361529 """ + if isinstance(v, bool): + return v + if v.lower() in ('yes', 'true', 't', 'y', '1'): + return True + elif v.lower() in ('no', 'false', 'f', 'n', '0'): + return False + else: + raise configargparse.ArgumentTypeError('Boolean value expected.') + + +def parse_args(config_path=None): + """ + requirement for config + 1. command > yaml > default + 2. avoid re-definition + 3. lowercase letters is better + 4. hierarchical is not necessary + """ + parser = configargparse.ArgParser() + parser.add("-c", "--config", required=True, is_config_file=True) + parser.add("--project", default="audio2pose", type=str) # wandb project name + parser.add("--stat", default="ts", type=str) + parser.add("--csv_name", default="a2g_0", type=str) # local device id + parser.add("--notes", default="", type=str) + parser.add("--trainer", default="camn", type=str) + + parser.add("--l", default=4, type=int) + # ------------- path and save name ---------------- # + parser.add("--is_train", default=True, type=str2bool) + parser.add("--debug", default=False, type=str2bool) + # different between environments + parser.add("--root_path", default="/home/ma-user/work/") + parser.add("--cache_path", default="/outputs/audio2pose/", type=str) + parser.add("--out_path", default="/outputs/audio2pose/", type=str) + parser.add("--data_path", default="/outputs/audio2pose/", type=str) + parser.add("--train_data_path", default="/datasets/trinity/train/", type=str) + parser.add("--val_data_path", default="/datasets/trinity/val/", type=str) + parser.add("--test_data_path", default="/datasets/trinity/test/", type=str) + parser.add("--mean_pose_path", default="/datasets/trinity/train/", type=str) + parser.add("--std_pose_path", default="/datasets/trinity/train/", type=str) + parser.add("--mean_trans_path", default="", type=str) + parser.add("--std_trans_path", default="", type=str) + # for pretrian weights + parser.add("--data_path_1", default="../../datasets/checkpoints/", type=str) + parser.add("--vqvae_upper_path", default="", type=str) + parser.add("--vqvae_hands_path", default="", type=str) + parser.add("--vqvae_lower_path", default="", type=str) + parser.add("--vqvae_lower_trans_path", default="", type=str) + parser.add("--use_trans", default=False, type=str2bool) + parser.add("--use_motionclip", default=False, type=str2bool) + + parser.add("--vqvae_latent_scale",default=1.0,type=float) + parser.add("--vqvae_squeeze_scale", default="1", type=int) + parser.add("--vqvae_type", default="vqvae", type=str) + + + # ------------------- evaluation ----------------------- # + parser.add("--test_ckpt", default="/datasets/beat_cache/beat_4english_15_141/last.bin") + parser.add("--eval_model", default="vae", type=str) + parser.add("--e_name", default=None, type=str) #HalfEmbeddingNet + parser.add("--e_path", default="/datasets/beat/generated_data/self_vae_128.bin") + parser.add("--variational", default=False, type=str2bool) + parser.add("--vae_length", default=256, type=int) + parser.add("--vae_test_dim", default=141, type=int) + parser.add("--vae_test_len", default=34, type=int) + parser.add("--vae_test_stride", default=10, type=int) + #parser.add("--vae_pose_length", default=34, type=int) + parser.add("--test_period", default=20, type=int) + parser.add("--vae_codebook_size", default=1024, type=int) + parser.add("--vae_quantizer_lambda", default=1., type=float) + + parser.add("--vae_layer", default=2, type=int) + parser.add("--vae_grow", default=[1,1,2,1], type=int, nargs="*") + parser.add("--lf", default=0., type=float) + parser.add("--ll", default=0., type=float) + parser.add("--lu", default=0., type=float) + parser.add("--lh", default=0., type=float) + parser.add("--cf", default=0., type=float) + parser.add("--cl", default=0., type=float) + parser.add("--cu", default=0., type=float) + parser.add("--ch", default=0., type=float) + + + # --------------- data ---------------------------- # + parser.add("--use_amass", default=False, type=str2bool) + parser.add("--additional_data", default=False, type=str2bool) + parser.add("--train_trans", default=True, type=str2bool) + parser.add("--dataset", default="beat", type=str) + parser.add("--rot6d", default=True, type=str2bool) + parser.add("--ori_joints", default="spine_neck_141", type=str) + parser.add("--tar_joints", default="spine_neck_141", type=str) + parser.add("--training_speakers", default=[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30], type=int, nargs="*") + #parser.add("--pose_version", default="spine_neck_141", type=str) + parser.add("--new_cache", default=True, type=str2bool) + parser.add("--beat_align", default=True, type=str2bool) + parser.add("--cache_only", default=False, type=str2bool) + parser.add("--word_cache", default=False, type=str2bool) + parser.add("--use_aug", default=False, type=str2bool) + parser.add("--disable_filtering", default=False, type=str2bool) + parser.add("--clean_first_seconds", default=0, type=int) + parser.add("--clean_final_seconds", default=0, type=int) + + parser.add("--audio_rep", default=None, type=str) + parser.add("--audio_sr", default=16000, type=int) + parser.add("--word_rep", default=None, type=str) + parser.add("--emo_rep", default=None, type=str) + parser.add("--sem_rep", default=None, type=str) + parser.add("--facial_rep", default=None, type=str) + parser.add("--pose_rep", default="bvhrot", type=str) + parser.add("--id_rep", default="onehot", type=str) + parser.add("--speaker_id", default="onehot", type=str) + + parser.add("--a_pre_encoder", default=None, type=str) + parser.add("--a_encoder", default=None, type=str) + parser.add("--a_fix_pre", default=False, type=str2bool) + parser.add("--t_pre_encoder", default=None, type=str) + parser.add("--t_encoder", default=None, type=str) + parser.add("--t_fix_pre", default=False, type=str2bool) + parser.add("--m_pre_encoder", default=None, type=str) + parser.add("--m_encoder", default=None, type=str) + parser.add("--m_fix_pre", default=False, type=str2bool) + parser.add("--f_pre_encoder", default=None, type=str) + parser.add("--f_encoder", default=None, type=str) + parser.add("--f_fix_pre", default=False, type=str2bool) + parser.add("--m_decoder", default=None, type=str) + parser.add("--decode_fusion", default=None, type=str) + parser.add("--atmr", default=0.0, type=float) + parser.add("--ttmr", default=0., type=float) + parser.add("--mtmr", default=0., type=float) + parser.add("--ftmr", default=0., type=float) + parser.add("--asmr", default=0., type=float) + parser.add("--tsmr", default=0., type=float) + parser.add("--msmr", default=0., type=float) + parser.add("--fsmr", default=0., type=float) +# parser.add("--m_encoder", default=None, type=str) +# parser.add("--m_pre_fix", default=None, type=str) +# parser.add("--id_rep", default=None, type=str) + + parser.add("--freeze_wordembed", default=True, type=str2bool) + parser.add("--audio_fps", default=16000, type=int) + parser.add("--facial_fps", default=15, type=int) + parser.add("--pose_fps", default=15, type=int) + + parser.add("--audio_dims", default=1, type=int) + parser.add("--facial_dims", default=39, type=int) + parser.add("--pose_dims", default=123, type=int) + parser.add("--word_index_num", default=5793, type=int) + parser.add("--word_dims", default=300, type=int) + parser.add("--speaker_dims", default=4, type=int) + parser.add("--emotion_dims", default=8, type=int) + + parser.add("--audio_norm", default=False, type=str2bool) + parser.add("--facial_norm", default=False, type=str2bool) + parser.add("--pose_norm", default=False, type=str2bool) + + parser.add("--pose_length", default=34, type=int) + parser.add("--pre_frames", default=4, type=int) + parser.add("--stride", default=10, type=int) + parser.add("--pre_type", default="zero", type=str) + + parser.add("--audio_f", default=0, type=int) + parser.add("--motion_f", default=0, type=int) + parser.add("--facial_f", default=0, type=int) + parser.add("--speaker_f", default=0, type=int) + parser.add("--word_f", default=0, type=int) + parser.add("--emotion_f", default=0, type=int) + parser.add("--aud_prob", default=1.0, type=float) + parser.add("--pos_prob", default=1.0, type=float) + parser.add("--txt_prob", default=1.0, type=float) + parser.add("--fac_prob", default=1.0, type=float) + parser.add("--multi_length_training", default=[1.0], type=float, nargs="*") + # --------------- model ---------------------------- # + parser.add("--pretrain", default=False, type=str2bool) + parser.add("--model", default="camn", type=str) + parser.add("--g_name", default="CaMN", type=str) + parser.add("--d_name", default=None, type=str) #ConvDiscriminator + parser.add("--dropout_prob", default=0.3, type=float) + parser.add("--n_layer", default=4, type=int) + parser.add("--hidden_size", default=300, type=int) + #parser.add("--period", default=34, type=int) + parser.add("--test_length", default=34, type=int) + # Self-designed "Multi-Stage", "Seprate", or "Original" + parser.add("--finger_net", default="original", type=str) + parser.add("--pos_encoding_type", default="sin", type=str) + parser.add("--queue_size", default=1024, type=int) + + # --------------- training ------------------------- # + parser.add("--epochs", default=120, type=int) + parser.add("--epoch_stage", default=0, type=int) + parser.add("--grad_norm", default=0, type=float) + parser.add("--no_adv_epoch", default=999, type=int) + parser.add("--batch_size", default=128, type=int) + parser.add("--opt", default="adam", type=str) + parser.add("--lr_base", default=0.00025, type=float) + parser.add("--opt_betas", default=[0.5, 0.999], type=float, nargs="*") + parser.add("--weight_decay", default=0., type=float) + # for warmup and cosine + parser.add("--lr_min", default=1e-7, type=float) + parser.add("--warmup_lr", default=5e-4, type=float) + parser.add("--warmup_epochs", default=0, type=int) + parser.add("--decay_epochs", default=9999, type=int) + parser.add("--decay_rate", default=0.1, type=float) + parser.add("--lr_policy", default="step", type=str) + # for sgd + parser.add("--momentum", default=0.8, type=float) + parser.add("--nesterov", default=True, type=str2bool) + parser.add("--amsgrad", default=False, type=str2bool) + parser.add("--d_lr_weight", default=0.2, type=float) + parser.add("--rec_weight", default=500, type=float) + parser.add("--adv_weight", default=20.0, type=float) + parser.add("--fid_weight", default=0.0, type=float) + parser.add("--vel_weight", default=0.0, type=float) + parser.add("--acc_weight", default=0.0, type=float) + parser.add("--kld_weight", default=0.0, type=float) + parser.add("--kld_aud_weight", default=0.0, type=float) + parser.add("--kld_fac_weight", default=0.0, type=float) + parser.add("--ali_weight", default=0.0, type=float) + parser.add("--ita_weight", default=0.0, type=float) + parser.add("--iwa_weight", default=0.0, type=float) + parser.add("--wei_weight", default=0.0, type=float) + parser.add("--gap_weight", default=0.0, type=float) + parser.add("--atcont", default=0.0, type=float) + parser.add("--fusion_mode", default="sum", type=str) + + parser.add("--div_reg_weight", default=0.0, type=float) + parser.add("--rec_aud_weight", default=0.0, type=float) + parser.add("--rec_ver_weight", default=0.0, type=float) + parser.add("--rec_pos_weight", default=0.0, type=float) + parser.add("--rec_fac_weight", default=0.0, type=float) + parser.add("--rec_txt_weight", default=0.0, type=float) +# parser.add("--gan_noise_size", default=0, type=int) + # --------------- ha2g -------------------------- # + parser.add("--n_pre_poses", default=4, type=int) + parser.add("--n_poses", default=34, type=int) + parser.add("--input_context", default="both", type=str) + parser.add("--loss_contrastive_pos_weight", default=0.2, type=float) + parser.add("--loss_contrastive_neg_weight", default=0.005, type=float) + parser.add("--loss_physical_weight", default=0.0, type=float) + parser.add("--loss_reg_weight", default=0.05, type=float) + parser.add("--loss_regression_weight", default=70.0, type=float) + parser.add("--loss_gan_weight", default=5.0, type=float) + parser.add("--loss_kld_weight", default=0.1, type=float) + parser.add("--z_type", default="speaker", type=str) + # --------------- device -------------------------- # + parser.add("--random_seed", default=2021, type=int) + parser.add("--deterministic", default=True, type=str2bool) + parser.add("--benchmark", default=True, type=str2bool) + parser.add("--cudnn_enabled", default=True, type=str2bool) + # mix precision + parser.add("--apex", default=False, type=str2bool) + parser.add("--gpus", default=[0], type=int, nargs="*") + parser.add("--loader_workers", default=0, type=int) + parser.add("--ddp", default=False, type=str2bool) + parser.add("--sparse", default=1, type=int) + #parser.add("--world_size") + + # --------------- vqvae -------------------------- # + parser.add("--levels", default=1, type=int) + parser.add("--downs_t", default=[3] ,type=int, nargs="*") + parser.add("--strides_t", default=[2], type=int,nargs="*") + parser.add("--emb_width", default=512, type=int) + parser.add("--l_bins", default=512, type=int) + parser.add("--l_mu", default=0.99, type=float) + parser.add("--commit", default=0.02, type=float) + parser.add("--hvqvae_multipliers", default=[1], type=int,nargs="*") + parser.add("--width", default=512, type=int) + parser.add("--depth", default=3, type=int) + parser.add("--m_conv", default=1.0, type=float) + parser.add("--dilation_growth_rate", default=3, type=int) + parser.add("--sample_length", default=34, type=int) + parser.add("--use_bottleneck", default=True, type=str2bool) + parser.add("--joint_channel", default=3, type=int) + parser.add("--vel", default=1, type=int) + parser.add("--acc", default=1, type=int) + parser.add("--vqvae_reverse_decoder_dilation", default=True, type=str2bool) + parser.add("--vqvae_ckpt",type=str) + parser.add("--root_weight",default=1.0,type=float) + + # --------------- render -------------------------- # + parser.add("--render_video_fps", default=30, type=int) + parser.add("--render_video_width", default=1920, type=int) + parser.add("--render_video_height", default=720, type=int) + cpu_cores = os.cpu_count() if os.cpu_count() is not None else 1 + default_concurrent = max(1, cpu_cores // 2) + parser.add("--render_concurrent_num", default=default_concurrent, type=int) + parser.add("--render_tmp_img_filetype", default="bmp", type=str) + + # logging + parser.add("--log_period", default=10, type=int) + + if config_path: + args = parser.parse_args(["--config", config_path]) + else: + args = parser.parse_args() + + idc = 0 + for i, char in enumerate(args.config): + if char == "/": idc = i + args.name = args.config[idc+1:-5] + + is_train = args.is_train + + if is_train: + time_local = time.localtime() + name_expend = "%02d%02d_%02d%02d%02d_"%(time_local[1], time_local[2],time_local[3], time_local[4], time_local[5]) + args.name = name_expend + args.name + + return args \ No newline at end of file diff --git a/utils/data_transfer.py b/utils/data_transfer.py new file mode 100644 index 0000000000000000000000000000000000000000..025110cac44f42d581c4414bd3d0c6c7b21f33e0 --- /dev/null +++ b/utils/data_transfer.py @@ -0,0 +1,202 @@ +import os +import logging +import random +import h5py +import numpy as np +import pickle +import math +import numbers +import torch +import torch.nn as nn +import torch.nn.functional as F +from torch.optim.lr_scheduler import StepLR +from torch.distributions import Normal + + +def _index_from_letter(letter: str) -> int: + if letter == "X": + return 0 + if letter == "Y": + return 1 + if letter == "Z": + return 2 + raise ValueError("letter must be either X, Y or Z.") + + +def _angle_from_tan( + axis: str, other_axis: str, data, horizontal: bool, tait_bryan: bool +) -> torch.Tensor: + """ + Extract the first or third Euler angle from the two members of + the matrix which are positive constant times its sine and cosine. + + Args: + axis: Axis label "X" or "Y or "Z" for the angle we are finding. + other_axis: Axis label "X" or "Y or "Z" for the middle axis in the + convention. + data: Rotation matrices as tensor of shape (..., 3, 3). + horizontal: Whether we are looking for the angle for the third axis, + which means the relevant entries are in the same row of the + rotation matrix. If not, they are in the same column. + tait_bryan: Whether the first and third axes in the convention differ. + + Returns: + Euler Angles in radians for each matrix in data as a tensor + of shape (...). + """ + + i1, i2 = {"X": (2, 1), "Y": (0, 2), "Z": (1, 0)}[axis] + if horizontal: + i2, i1 = i1, i2 + even = (axis + other_axis) in ["XY", "YZ", "ZX"] + if horizontal == even: + return torch.atan2(data[..., i1], data[..., i2]) + if tait_bryan: + return torch.atan2(-data[..., i2], data[..., i1]) + return torch.atan2(data[..., i2], -data[..., i1]) + + +def _axis_angle_rotation(axis: str, angle: torch.Tensor) -> torch.Tensor: + """ + Return the rotation matrices for one of the rotations about an axis + of which Euler angles describe, for each value of the angle given. + + Args: + axis: Axis label "X" or "Y or "Z". + angle: any shape tensor of Euler angles in radians + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + + cos = torch.cos(angle) + sin = torch.sin(angle) + one = torch.ones_like(angle) + zero = torch.zeros_like(angle) + + if axis == "X": + R_flat = (one, zero, zero, zero, cos, -sin, zero, sin, cos) + elif axis == "Y": + R_flat = (cos, zero, sin, zero, one, zero, -sin, zero, cos) + elif axis == "Z": + R_flat = (cos, -sin, zero, sin, cos, zero, zero, zero, one) + else: + raise ValueError("letter must be either X, Y or Z.") + + return torch.stack(R_flat, -1).reshape(angle.shape + (3, 3)) + +def euler_angles_to_matrix(euler_angles: torch.Tensor, convention: str) -> torch.Tensor: + """ + Convert rotations given as Euler angles in radians to rotation matrices. + + Args: + euler_angles: Euler angles in radians as tensor of shape (..., 3). + convention: Convention string of three uppercase letters from + {"X", "Y", and "Z"}. + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + if euler_angles.dim() == 0 or euler_angles.shape[-1] != 3: + raise ValueError("Invalid input euler angles.") + if len(convention) != 3: + raise ValueError("Convention must have 3 letters.") + if convention[1] in (convention[0], convention[2]): + raise ValueError(f"Invalid convention {convention}.") + for letter in convention: + if letter not in ("X", "Y", "Z"): + raise ValueError(f"Invalid letter {letter} in convention string.") + matrices = [ + _axis_angle_rotation(c, e) + for c, e in zip(convention, torch.unbind(euler_angles, -1)) + ] + # return functools.reduce(torch.matmul, matrices) + return torch.matmul(torch.matmul(matrices[0], matrices[1]), matrices[2]) + + +def matrix_to_rotation_6d(matrix: torch.Tensor) -> torch.Tensor: + """ + Converts rotation matrices to 6D rotation representation by Zhou et al. [1] + by dropping the last row. Note that 6D representation is not unique. + Args: + matrix: batch of rotation matrices of size (*, 3, 3) + Returns: + 6D rotation representation, of size (*, 6) + [1] Zhou, Y., Barnes, C., Lu, J., Yang, J., & Li, H. + On the Continuity of Rotation Representations in Neural Networks. + IEEE Conference on Computer Vision and Pattern Recognition, 2019. + Retrieved from http://arxiv.org/abs/1812.07035 + """ + return matrix[..., :2, :].clone().reshape(*matrix.size()[:-2], 6) + + +def rotation_6d_to_matrix(d6: torch.Tensor) -> torch.Tensor: + """ + Args: + d6: 6D rotation representation, of size (*, 6) + Returns: + batch of rotation matrices of size (*, 3, 3) + """ + a1, a2 = d6[..., :3], d6[..., 3:] + b1 = F.normalize(a1, dim=-1) + b2 = a2 - (b1 * a2).sum(-1, keepdim=True) * b1 + b2 = F.normalize(b2, dim=-1) + b3 = torch.cross(b1, b2, dim=-1) + return torch.stack((b1, b2, b3), dim=-2) + + +def matrix_to_euler_angles(matrix: torch.Tensor, convention: str) -> torch.Tensor: + """ + Convert rotations given as rotation matrices to Euler angles in radians. + + Args: + matrix: Rotation matrices as tensor of shape (..., 3, 3). + convention: Convention string of three uppercase letters. + + Returns: + Euler angles in radians as tensor of shape (..., 3). + """ + if len(convention) != 3: + raise ValueError("Convention must have 3 letters.") + if convention[1] in (convention[0], convention[2]): + raise ValueError(f"Invalid convention {convention}.") + for letter in convention: + if letter not in ("X", "Y", "Z"): + raise ValueError(f"Invalid letter {letter} in convention string.") + if matrix.size(-1) != 3 or matrix.size(-2) != 3: + raise ValueError(f"Invalid rotation matrix shape {matrix.shape}.") + i0 = _index_from_letter(convention[0]) + i2 = _index_from_letter(convention[2]) + tait_bryan = i0 != i2 + if tait_bryan: + central_angle = torch.asin( + matrix[..., i0, i2] * (-1.0 if i0 - i2 in [-1, 2] else 1.0) + ) + else: + central_angle = torch.acos(matrix[..., i0, i0]) + + o = ( + _angle_from_tan( + convention[0], convention[1], matrix[..., i2], False, tait_bryan + ), + central_angle, + _angle_from_tan( + convention[2], convention[1], matrix[..., i0, :], True, tait_bryan + ), + ) + return torch.stack(o, -1) + + +def so3_relative_angle(m1, m2): + m1 = m1.reshape(-1, 3, 3) + m2 = m2.reshape(-1, 3, 3) + #print(m2.shape) + m = torch.bmm(m1, m2.transpose(1, 2)) # batch*3*3 + #print(m.shape) + cos = (m[:, 0, 0] + m[:, 1, 1] + m[:, 2, 2] - 1) / 2 + #print(cos.shape) + cos = torch.clamp(cos, min=-1 + 1E-6, max=1-1E-6) + #print(cos.shape) + theta = torch.acos(cos) + #print(theta.shape) + return torch.mean(theta) diff --git a/utils/fast_render.py b/utils/fast_render.py new file mode 100644 index 0000000000000000000000000000000000000000..d85520694154856e2ad7c8bd21051b0c29ed75cc --- /dev/null +++ b/utils/fast_render.py @@ -0,0 +1,267 @@ +import os +import time +import numpy as np +import pyrender +import trimesh +import queue +import imageio +import threading +import multiprocessing +import utils.media +import glob + +def deg_to_rad(degrees): + return degrees * np.pi / 180 + +def create_pose_camera(angle_deg): + angle_rad = deg_to_rad(angle_deg) + return np.array([ + [1.0, 0.0, 0.0, 0.0], + [0.0, np.cos(angle_rad), -np.sin(angle_rad), 1.0], + [0.0, np.sin(angle_rad), np.cos(angle_rad), 5.0], + [0.0, 0.0, 0.0, 1.0] + ]) + +def create_pose_light(angle_deg): + angle_rad = deg_to_rad(angle_deg) + return np.array([ + [1.0, 0.0, 0.0, 0.0], + [0.0, np.cos(angle_rad), -np.sin(angle_rad), 0.0], + [0.0, np.sin(angle_rad), np.cos(angle_rad), 3.0], + [0.0, 0.0, 0.0, 1.0] + ]) + +def create_scene_with_mesh(vertices, faces, uniform_color, pose_camera, pose_light): + trimesh_mesh = trimesh.Trimesh(vertices=vertices, faces=faces, vertex_colors=uniform_color) + mesh = pyrender.Mesh.from_trimesh(trimesh_mesh, smooth=True) + scene = pyrender.Scene() + scene.add(mesh) + camera = pyrender.OrthographicCamera(xmag=1.0, ymag=1.0) + scene.add(camera, pose=pose_camera) + light = pyrender.DirectionalLight(color=[1.0, 1.0, 1.0], intensity=4.0) + scene.add(light, pose=pose_light) + return scene + +def do_render_one_frame(renderer, frame_idx, vertices, vertices1, faces): + if frame_idx % 100 == 0: + print('processed', frame_idx, 'frames') + + uniform_color = [220, 220, 220, 255] + pose_camera = create_pose_camera(angle_deg=-2) + pose_light = create_pose_light(angle_deg=-30) + + figs = [] + for vtx in [vertices, vertices1]: + # print(vtx.shape) + scene = create_scene_with_mesh(vtx, faces, uniform_color, pose_camera, pose_light) + fig, _ = renderer.render(scene) + figs.append(fig) + + return figs[0], figs[1] + +def do_render_one_frame_no_gt(renderer, frame_idx, vertices, faces): + if frame_idx % 100 == 0: + print('processed', frame_idx, 'frames') + + uniform_color = [220, 220, 220, 255] + pose_camera = create_pose_camera(angle_deg=-2) + pose_light = create_pose_light(angle_deg=-30) + + figs = [] + # for vtx in [vertices]: + # print(vtx.shape) + # print(vertices.shape) + scene = create_scene_with_mesh(vertices, faces, uniform_color, pose_camera, pose_light) + fig, _ = renderer.render(scene) + figs.append(fig) + + return figs[0] + +def write_images_from_queue(fig_queue, output_dir, img_filetype): + while True: + e = fig_queue.get() + if e is None: + break + fid, fig1, fig2 = e + filename = os.path.join(output_dir, f"frame_{fid}.{img_filetype}") + merged_fig = np.hstack((fig1, fig2)) + try: + imageio.imwrite(filename, merged_fig) + except Exception as ex: + print(f"Error writing image {filename}: {ex}") + raise ex + +def write_images_from_queue_no_gt(fig_queue, output_dir, img_filetype): + while True: + e = fig_queue.get() + if e is None: + break + fid, fig1, fig2 = e + filename = os.path.join(output_dir, f"frame_{fid}.{img_filetype}") + merged_fig = fig1 #np.hstack((fig1)) + try: + imageio.imwrite(filename, merged_fig) + except Exception as ex: + print(f"Error writing image {filename}: {ex}") + raise ex + + +def render_frames_and_enqueue(fids, frame_vertex_pairs, faces, render_width, render_height, fig_queue): + fig_resolution = (render_width // 2, render_height) + renderer = pyrender.OffscreenRenderer(*fig_resolution) + + for idx, fid in enumerate(fids): + fig1, fig2 = do_render_one_frame(renderer, fid, frame_vertex_pairs[idx][0], frame_vertex_pairs[idx][1], faces) + fig_queue.put((fid, fig1, fig2)) + + renderer.delete() + +def render_frames_and_enqueue_no_gt(fids, frame_vertex_pairs, faces, render_width, render_height, fig_queue): + fig_resolution = (render_width // 2, render_height) + renderer = pyrender.OffscreenRenderer(*fig_resolution) + + for idx, fid in enumerate(fids): + fig1 = do_render_one_frame_no_gt(renderer, fid, frame_vertex_pairs[idx][0], faces) + fig_queue.put((fid, fig1)) + + renderer.delete() + +def sub_process_process_frame(subprocess_index, render_video_width, render_video_height, render_tmp_img_filetype, fids, frame_vertex_pairs, faces, output_dir): + begin_ts = time.time() + print(f"subprocess_index={subprocess_index} begin_ts={begin_ts}") + + fig_queue = queue.Queue() + render_frames_and_enqueue(fids, frame_vertex_pairs, faces, render_video_width, render_video_height, fig_queue) + fig_queue.put(None) + render_end_ts = time.time() + + image_writer_thread = threading.Thread(target=write_images_from_queue, args=(fig_queue, output_dir, render_tmp_img_filetype)) + image_writer_thread.start() + image_writer_thread.join() + + write_end_ts = time.time() + print( + f"subprocess_index={subprocess_index} " + f"render={render_end_ts - begin_ts:.2f} " + f"all={write_end_ts - begin_ts:.2f} " + f"begin_ts={begin_ts:.2f} " + f"render_end_ts={render_end_ts:.2f} " + f"write_end_ts={write_end_ts:.2f}" + ) + +def sub_process_process_frame_no_gt(subprocess_index, render_video_width, render_video_height, render_tmp_img_filetype, fids, frame_vertex_pairs, faces, output_dir): + begin_ts = time.time() + print(f"subprocess_index={subprocess_index} begin_ts={begin_ts}") + + fig_queue = queue.Queue() + render_frames_and_enqueue(fids, frame_vertex_pairs, faces, render_video_width, render_video_height, fig_queue) + fig_queue.put(None) + render_end_ts = time.time() + + image_writer_thread = threading.Thread(target=write_images_from_queue_no_gt, args=(fig_queue, output_dir, render_tmp_img_filetype)) + image_writer_thread.start() + image_writer_thread.join() + + write_end_ts = time.time() + print( + f"subprocess_index={subprocess_index} " + f"render={render_end_ts - begin_ts:.2f} " + f"all={write_end_ts - begin_ts:.2f} " + f"begin_ts={begin_ts:.2f} " + f"render_end_ts={render_end_ts:.2f} " + f"write_end_ts={write_end_ts:.2f}" + ) + +def distribute_frames(frames, render_video_fps, render_concurent_nums, vertices_all, vertices1_all): + sample_interval = max(1, int(30 // render_video_fps)) + subproc_frame_ids = [[] for _ in range(render_concurent_nums)] + subproc_vertices = [[] for _ in range(render_concurent_nums)] + sampled_frame_id = 0 + + for i in range(frames): + if i % sample_interval != 0: + continue + subprocess_index = sampled_frame_id % render_concurent_nums + subproc_frame_ids[subprocess_index].append(sampled_frame_id) + subproc_vertices[subprocess_index].append((vertices_all[i], vertices1_all[i])) + sampled_frame_id += 1 + + return subproc_frame_ids, subproc_vertices + +def distribute_frames_no_gt(frames, render_video_fps, render_concurent_nums, vertices_all): + sample_interval = max(1, int(30 // render_video_fps)) + subproc_frame_ids = [[] for _ in range(render_concurent_nums)] + subproc_vertices = [[] for _ in range(render_concurent_nums)] + sampled_frame_id = 0 + + for i in range(frames): + if i % sample_interval != 0: + continue + subprocess_index = sampled_frame_id % render_concurent_nums + subproc_frame_ids[subprocess_index].append(sampled_frame_id) + subproc_vertices[subprocess_index].append((vertices_all[i], vertices_all[i])) + sampled_frame_id += 1 + + return subproc_frame_ids, subproc_vertices + +def generate_silent_videos(render_video_fps, + render_video_width, + render_video_height, + render_concurent_nums, + render_tmp_img_filetype, + frames, + vertices_all, + vertices1_all, + faces, + output_dir): + + subproc_frame_ids, subproc_vertices = distribute_frames(frames, render_video_fps, render_concurent_nums, vertices_all, vertices1_all) + + print(f"generate_silent_videos concurrentNum={render_concurent_nums} time={time.time()}") + with multiprocessing.Pool(render_concurent_nums) as pool: + pool.starmap( + sub_process_process_frame, + [ + (subprocess_index, render_video_width, render_video_height, render_tmp_img_filetype, subproc_frame_ids[subprocess_index], subproc_vertices[subprocess_index], faces, output_dir) + for subprocess_index in range(render_concurent_nums) + ] + ) + + output_file = os.path.join(output_dir, "silence_video.mp4") + utils.media.convert_img_to_mp4(os.path.join(output_dir, f"frame_%d.{render_tmp_img_filetype}"), output_file, render_video_fps) + filenames = glob.glob(os.path.join(output_dir, f"*.{render_tmp_img_filetype}")) + for filename in filenames: + os.remove(filename) + + return output_file + +def generate_silent_videos_no_gt(render_video_fps, + render_video_width, + render_video_height, + render_concurent_nums, + render_tmp_img_filetype, + frames, + vertices_all, + faces, + output_dir): + + subproc_frame_ids, subproc_vertices = distribute_frames_no_gt(frames, render_video_fps, render_concurent_nums, vertices_all) + + print(f"generate_silent_videos concurrentNum={render_concurent_nums} time={time.time()}") + #sub_process_process_frame_no_gt(0, render_video_width, render_video_height, render_tmp_img_filetype, subproc_frame_ids[0], subproc_vertices[0], faces, output_dir) + with multiprocessing.Pool(render_concurent_nums) as pool: + pool.starmap( + sub_process_process_frame_no_gt, + [ + (subprocess_index, render_video_width, render_video_height, render_tmp_img_filetype, subproc_frame_ids[subprocess_index], subproc_vertices[subprocess_index], faces, output_dir) + for subprocess_index in range(render_concurent_nums) + ] + ) + + output_file = os.path.join(output_dir, "silence_video.mp4") + utils.media.convert_img_to_mp4(os.path.join(output_dir, f"frame_%d.{render_tmp_img_filetype}"), output_file, render_video_fps) + filenames = glob.glob(os.path.join(output_dir, f"*.{render_tmp_img_filetype}")) + for filename in filenames: + os.remove(filename) + + return output_file \ No newline at end of file diff --git a/utils/logger_tools.py b/utils/logger_tools.py new file mode 100644 index 0000000000000000000000000000000000000000..9e62355cb3763cfd34dc82b0d50c3b65a115064e --- /dev/null +++ b/utils/logger_tools.py @@ -0,0 +1,59 @@ +import os +import inspect +import sys +import yaml +#import wandb +from loguru import logger + +def setup_logger(save_dir, distributed_rank=0, filename="log.txt", mode="a"): + """setup logger for training and testing. + Args: + save_dir(str): location to save log file + distributed_rank(int): device rank when multi-gpu environment + filename (string): log save name. + mode(str): log file write mode, `append` or `override`. default is `a`. + + Return: + logger instance. + """ + loguru_format = ( + "{time: MM-DD HH:mm:ss} | " + #"{level: <8} | " + #"{name}:{line} - {message}" + "{message}" + ) + + logger.remove() + save_file = os.path.join(save_dir, filename) + if mode == "o" and os.path.exists(save_file): + os.remove(save_file) + # only keep logger in rank0 process + if distributed_rank == 0: + logger.add( + sys.stderr, + format=loguru_format, + level="INFO", + enqueue=True, + ) + logger.add(save_file, + format=loguru_format, + ) + + +def set_args_and_logger(args, rank): + """ + set logger file and print args + """ + args_name_dir = args.out_path + "custom/" + args.name + args.notes + "/" + if rank == 0: + if not os.path.exists(args_name_dir): os.makedirs(args_name_dir) + args_name = args_name_dir + "/" + args.name +".yaml" + if os.path.exists(args_name): + s_add = 10 + logger.warning(f"Already exist args, add {s_add} to ran_seed to continue training") + args.random_seed += s_add + else: + with open(args_name, "w+") as f: + yaml.dump(args.__dict__, f, default_flow_style=True) + #json.dump(args.__dict__, f) + setup_logger(args_name_dir, rank, filename=f"{args.name}.txt") \ No newline at end of file diff --git a/utils/media.py b/utils/media.py new file mode 100644 index 0000000000000000000000000000000000000000..2bd21e079a9e48f97f1511bd289d39f4aeccc40e --- /dev/null +++ b/utils/media.py @@ -0,0 +1,39 @@ +import numpy as np +import subprocess + +def add_audio_to_video(silent_video_path, audio_path, output_video_path): + command = [ + 'ffmpeg', + '-y', + '-i', silent_video_path, + '-i', audio_path, + '-map', '0:v', + '-map', '1:a', + '-c:v', 'copy', + '-shortest', + output_video_path + ] + + try: + subprocess.run(command, check=True) + print(f"Video with audio generated successfully: {output_video_path}") + except subprocess.CalledProcessError as e: + print(f"Error occurred: {e}") + + +def convert_img_to_mp4(input_pattern, output_file, framerate=30): + command = [ + 'ffmpeg', + '-framerate', str(framerate), + '-i', input_pattern, + '-c:v', 'libx264', + '-pix_fmt', 'yuv420p', + output_file, + '-y' + ] + + try: + subprocess.run(command, check=True) + print(f"Video conversion successful. Output file: {output_file}") + except subprocess.CalledProcessError as e: + print(f"Error during video conversion: {e}") diff --git a/utils/metric.py b/utils/metric.py new file mode 100644 index 0000000000000000000000000000000000000000..53930062137b7ee82adb21ce226f572be77176e5 --- /dev/null +++ b/utils/metric.py @@ -0,0 +1,242 @@ +import librosa +import glob +import os +import numpy as np +import matplotlib.pyplot as plt +import librosa.display +from matplotlib.pyplot import figure +import math +from scipy.signal import argrelextrema + + +class L1div(object): + def __init__(self): + self.counter = 0 + self.sum = 0 + def run(self, results): + self.counter += results.shape[0] + mean = np.mean(results, 0) + for i in range(results.shape[0]): + results[i, :] = abs(results[i, :] - mean) + sum_l1 = np.sum(results) + self.sum += sum_l1 + def avg(self): + return self.sum/self.counter + def reset(self): + self.counter = 0 + self.sum = 0 + + +class SRGR(object): + def __init__(self, threshold=0.1, joints=47): + self.threshold = threshold + self.pose_dimes = joints + self.counter = 0 + self.sum = 0 + + def run(self, results, targets, semantic): + results = results.reshape(-1, self.pose_dimes, 3) + targets = targets.reshape(-1, self.pose_dimes, 3) + semantic = semantic.reshape(-1) + diff = np.sum(abs(results-targets),2) + success = np.where(diffself.threshold) + #print(vel.shape) + #t_end = 80 + #vel[::2, :] -= 0.000001 + #print(vel[t_start:t_end, i], vel[t_start:t_end, i].shape) + beat_vel = argrelextrema(vel[t_start:t_end, i], np.less, order=self.order) # n*47 + #print(beat_vel, t_start, t_end) + beat_vel_list = [] + for j in beat_vel[0]: + if j in vel_mask[0]: + beat_vel_list.append(j) + beat_vel = np.array(beat_vel_list) + beat_vel_all.append(beat_vel) + #print(beat_vel_all) + return beat_vel_all #beat_right_arm, beat_right_shoulder, beat_right_wrist, beat_left_arm, beat_left_shoulder, beat_left_wrist + + + def load_data(self, wave, pose, t_start, t_end, pose_fps): + onset_raw, onset_bt, onset_bt_rms = self.load_audio(wave, t_start, t_end) + beat_right_arm, beat_right_shoulder, beat_right_wrist, beat_left_arm, beat_left_shoulder, beat_left_wrist = self.load_pose(pose, t_start, t_end, pose_fps) + return onset_raw, onset_bt, onset_bt_rms, beat_right_arm, beat_right_shoulder, beat_right_wrist, beat_left_arm, beat_left_shoulder, beat_left_wrist + + def eval_random_pose(self, wave, pose, t_start, t_end, pose_fps, num_random=60): + onset_raw, onset_bt, onset_bt_rms = self.load_audio(wave, t_start, t_end) + dur = t_end - t_start + for i in range(num_random): + beat_right_arm, beat_right_shoulder, beat_right_wrist, beat_left_arm, beat_left_shoulder, beat_left_wrist = self.load_pose(pose, i, i+dur, pose_fps) + dis_all_b2a= self.calculate_align(onset_raw, onset_bt, onset_bt_rms, beat_right_arm, beat_right_shoulder, beat_right_wrist, beat_left_arm, beat_left_shoulder, beat_left_wrist) + print(f"{i}s: ",dis_all_b2a) + + + @staticmethod + def plot_onsets(audio, sr, onset_times_1, onset_times_2): + import librosa + import librosa.display + import matplotlib.pyplot as plt + # Plot audio waveform + fig, axarr = plt.subplots(2, 1, figsize=(10, 10), sharex=True) + + # Plot audio waveform in both subplots + librosa.display.waveshow(audio, sr=sr, alpha=0.7, ax=axarr[0]) + librosa.display.waveshow(audio, sr=sr, alpha=0.7, ax=axarr[1]) + + # Plot onsets from first method on the first subplot + for onset in onset_times_1: + axarr[0].axvline(onset, color='r', linestyle='--', alpha=0.9, label='Onset Method 1') + axarr[0].legend() + axarr[0].set(title='Onset Method 1', xlabel='', ylabel='Amplitude') + + # Plot onsets from second method on the second subplot + for onset in onset_times_2: + axarr[1].axvline(onset, color='b', linestyle='-', alpha=0.7, label='Onset Method 2') + axarr[1].legend() + axarr[1].set(title='Onset Method 2', xlabel='Time (s)', ylabel='Amplitude') + + + # Add legend (eliminate duplicate labels) + handles, labels = plt.gca().get_legend_handles_labels() + by_label = dict(zip(labels, handles)) + plt.legend(by_label.values(), by_label.keys()) + + # Show plot + plt.title("Audio waveform with Onsets") + plt.savefig("./onset.png", dpi=500) + + def audio_beat_vis(self, onset_raw, onset_bt, onset_bt_rms): + figure(figsize=(24, 6), dpi=80) + fig, ax = plt.subplots(nrows=4, sharex=True) + librosa.display.specshow(librosa.amplitude_to_db(self.S, ref=np.max), + y_axis='log', x_axis='time', ax=ax[0]) + ax[0].label_outer() + ax[1].plot(self.times, self.oenv, label='Onset strength') + ax[1].vlines(librosa.frames_to_time(onset_raw), 0, self.oenv.max(), label='Raw onsets', color='r') + ax[1].legend() + ax[1].label_outer() + + ax[2].plot(self.times, self.oenv, label='Onset strength') + ax[2].vlines(librosa.frames_to_time(onset_bt), 0, self.oenv.max(), label='Backtracked', color='r') + ax[2].legend() + ax[2].label_outer() + + ax[3].plot(self.times, self.rms[0], label='RMS') + ax[3].vlines(librosa.frames_to_time(onset_bt_rms), 0, self.oenv.max(), label='Backtracked (RMS)', color='r') + ax[3].legend() + fig.savefig("./onset.png", dpi=500) + + @staticmethod + def motion_frames2time(vel, offset, pose_fps): + time_vel = vel/pose_fps + offset + return time_vel + + @staticmethod + def GAHR(a, b, sigma): + dis_all_a2b = 0 + dis_all_b2a = 0 + for b_each in b: + l2_min = np.inf + for a_each in a: + l2_dis = abs(a_each - b_each) + if l2_dis < l2_min: + l2_min = l2_dis + dis_all_b2a += math.exp(-(l2_min**2)/(2*sigma**2)) + dis_all_b2a /= len(b) + return dis_all_b2a + + @staticmethod + def fix_directed_GAHR(a, b, sigma): + a = alignment.motion_frames2time(a, 0, 30) + b = alignment.motion_frames2time(b, 0, 30) + t = len(a)/30 + a = [0] + a + [t] + b = [0] + b + [t] + dis_a2b = alignment.GAHR(a, b, sigma) + return dis_a2b + + def calculate_align(self, onset_bt_rms, beat_vel, pose_fps=30): + audio_bt = onset_bt_rms + avg_dis_all_b2a_list = [] + for its, beat_vel_each in enumerate(beat_vel): + if its not in self.upper_body: + continue + #print(beat_vel_each) + #print(audio_bt.shape, beat_vel_each.shape) + pose_bt = self.motion_frames2time(beat_vel_each, 0, pose_fps) + #print(pose_bt) + avg_dis_all_b2a_list.append(self.GAHR(pose_bt, audio_bt, self.sigma)) + # avg_dis_all_b2a = max(avg_dis_all_b2a_list) + avg_dis_all_b2a = sum(avg_dis_all_b2a_list)/len(avg_dis_all_b2a_list) #max(avg_dis_all_b2a_list) + #print(avg_dis_all_b2a, sum(avg_dis_all_b2a_list)/47) + return avg_dis_all_b2a \ No newline at end of file diff --git a/utils/other_tools.py b/utils/other_tools.py new file mode 100644 index 0000000000000000000000000000000000000000..a07cd967e613d7a6d7d4abc7f804665dafa6afd7 --- /dev/null +++ b/utils/other_tools.py @@ -0,0 +1,820 @@ +import os +import numpy as np +import random +import torch +import shutil +import csv +import pprint +import pandas as pd +from loguru import logger +from collections import OrderedDict +import matplotlib.pyplot as plt +import pickle +import time +import hashlib +from scipy.spatial.transform import Rotation as R +from scipy.spatial.transform import Slerp +import cv2 + + +def resize_motion_sequence_tensor(sequence, target_frames): + """ + Resize a batch of 8-frame motion sequences to a specified number of frames using interpolation. + + :param sequence: A (bs, 8, 165) tensor representing a batch of 8-frame motion sequences + :param target_frames: An integer representing the desired number of frames in the output sequences + :return: A (bs, target_frames, 165) tensor representing the resized motion sequences + """ + bs, _, _ = sequence.shape + + # Create a time vector for the original and target sequences + original_time = torch.linspace(0, 1, 8, device=sequence.device).view(1, -1, 1) + target_time = torch.linspace(0, 1, target_frames, device=sequence.device).view(1, -1, 1) + + # Permute the dimensions to (bs, 165, 8) for interpolation + sequence = sequence.permute(0, 2, 1) + + # Interpolate each joint's motion to the target number of frames + resized_sequence = torch.nn.functional.interpolate(sequence, size=target_frames, mode='linear', align_corners=True) + + # Permute the dimensions back to (bs, target_frames, 165) + resized_sequence = resized_sequence.permute(0, 2, 1) + + return resized_sequence + +def adjust_speed_according_to_ratio_tensor(chunks): + """ + Adjust the playback speed within a batch of 32-frame chunks according to random intervals. + + :param chunks: A (bs, 32, 165) tensor representing a batch of motion chunks + :return: A (bs, 32, 165) tensor representing the motion chunks after speed adjustment + """ + bs, _, _ = chunks.shape + + # Step 1: Divide the chunk into 4 equal intervals of 8 frames + equal_intervals = torch.chunk(chunks, 4, dim=1) + + # Step 2: Randomly sample 3 points within the chunk to determine new intervals + success = 0 + all_success = [] + #sample_points = torch.sort(torch.randint(1, 32, (bs, 3), device=chunks.device), dim=1).values + # new_intervals_boundaries = torch.cat([torch.zeros((bs, 1), device=chunks.device, dtype=torch.long), sample_points, 32*torch.ones((bs, 1), device=chunks.device, dtype=torch.long)], dim=1) + while success != 1: + sample_points = sorted(random.sample(range(1, 32), 3)) + new_intervals_boundaries = [0] + sample_points + [32] + new_intervals = [chunks[0][new_intervals_boundaries[i]:new_intervals_boundaries[i+1]] for i in range(4)] + speed_ratios = [8 / len(new_interval) for new_interval in new_intervals] + # if any of the speed ratios is greater than 3 or less than 0.33, resample + if all([0.33 <= speed_ratio <= 3 for speed_ratio in speed_ratios]): + success += 1 + all_success.append(new_intervals_boundaries) + new_intervals_boundaries = torch.from_numpy(np.array(all_success)) + # print(new_intervals_boundaries) + all_shapes = new_intervals_boundaries[:, 1:] - new_intervals_boundaries[:, :-1] + # Step 4: Adjust the speed of each new interval + adjusted_intervals = [] + # print(equal_intervals[0].shape) + for i in range(4): + adjusted_interval = resize_motion_sequence_tensor(equal_intervals[i], all_shapes[0, i]) + adjusted_intervals.append(adjusted_interval) + + # Step 5: Concatenate the adjusted intervals + adjusted_chunk = torch.cat(adjusted_intervals, dim=1) + + return adjusted_chunk + +def compute_exact_iou(bbox1, bbox2): + x1 = max(bbox1[0], bbox2[0]) + y1 = max(bbox1[1], bbox2[1]) + x2 = min(bbox1[0] + bbox1[2], bbox2[0] + bbox2[2]) + y2 = min(bbox1[1] + bbox1[3], bbox2[1] + bbox2[3]) + + intersection_area = max(0, x2 - x1) * max(0, y2 - y1) + bbox1_area = bbox1[2] * bbox1[3] + bbox2_area = bbox2[2] * bbox2[3] + union_area = bbox1_area + bbox2_area - intersection_area + + if union_area == 0: + return 0 + + return intersection_area / union_area + +def compute_iou(mask1, mask2): + # Compute the intersection + intersection = np.logical_and(mask1, mask2).sum() + + # Compute the union + union = np.logical_or(mask1, mask2).sum() + + # Compute the IoU + iou = intersection / union + + return iou + +def blankblending(all_frames, x, n): + return all_frames[x:x+n+1] + +def load_video_as_numpy_array(video_path): + cap = cv2.VideoCapture(video_path) + + # Using list comprehension to read frames and store in a list + frames = [frame for ret, frame in iter(lambda: cap.read(), (False, None)) if ret] + + cap.release() + + return np.array(frames) + +def synthesize_intermediate_frames_bidirectional(all_frames, x, n): + frame1 = all_frames[x] + frame2 = all_frames[x + n] + + # Convert the frames to grayscale + gray1 = cv2.cvtColor(frame1, cv2.COLOR_BGR2GRAY) + gray2 = cv2.cvtColor(frame2, cv2.COLOR_BGR2GRAY) + + # Calculate the forward and backward optical flow + forward_flow = cv2.calcOpticalFlowFarneback(gray1, gray2, None, 0.5, 3, 15, 3, 5, 1.2, 0) + backward_flow = cv2.calcOpticalFlowFarneback(gray2, gray1, None, 0.5, 3, 15, 3, 5, 1.2, 0) + + synthesized_frames = [] + for i in range(1, n): # For each intermediate frame between x and x + n + alpha = i / n # Interpolation factor + + # Compute the intermediate forward and backward flow + intermediate_forward_flow = forward_flow * alpha + intermediate_backward_flow = backward_flow * (1 - alpha) + + # Warp the frames based on the intermediate flow + h, w = frame1.shape[:2] + flow_map = np.column_stack((np.repeat(np.arange(h), w), np.tile(np.arange(w), h))) + forward_displacement = flow_map + intermediate_forward_flow.reshape(-1, 2) + backward_displacement = flow_map - intermediate_backward_flow.reshape(-1, 2) + + # Use cv2.remap for efficient warping + remap_x_forward, remap_y_forward = np.clip(forward_displacement[:, 1], 0, w - 1), np.clip(forward_displacement[:, 0], 0, h - 1) + remap_x_backward, remap_y_backward = np.clip(backward_displacement[:, 1], 0, w - 1), np.clip(backward_displacement[:, 0], 0, h - 1) + + warped_forward = cv2.remap(frame1, remap_x_forward.reshape(h, w).astype(np.float32), remap_y_forward.reshape(h, w).astype(np.float32), interpolation=cv2.INTER_LINEAR) + warped_backward = cv2.remap(frame2, remap_x_backward.reshape(h, w).astype(np.float32), remap_y_backward.reshape(h, w).astype(np.float32), interpolation=cv2.INTER_LINEAR) + + # Blend the warped frames to generate the intermediate frame + intermediate_frame = cv2.addWeighted(warped_forward, 1 - alpha, warped_backward, alpha, 0) + synthesized_frames.append(intermediate_frame) + + return synthesized_frames # Return n-2 synthesized intermediate frames + + +def linear_interpolate_frames(all_frames, x, n): + frame1 = all_frames[x] + frame2 = all_frames[x + n] + + synthesized_frames = [] + for i in range(1, n): # For each intermediate frame between x and x + n + alpha = i / (n) # Correct interpolation factor + inter_frame = cv2.addWeighted(frame1, 1 - alpha, frame2, alpha, 0) + synthesized_frames.append(inter_frame) + return synthesized_frames[:-1] + +def warp_frame(src_frame, flow): + h, w = flow.shape[:2] + flow_map = np.column_stack((np.repeat(np.arange(h), w), np.tile(np.arange(w), h))) + displacement = flow_map + flow.reshape(-1, 2) + + # Extract x and y coordinates of the displacement + x_coords = np.clip(displacement[:, 1], 0, w - 1).reshape(h, w).astype(np.float32) + y_coords = np.clip(displacement[:, 0], 0, h - 1).reshape(h, w).astype(np.float32) + + # Use cv2.remap for efficient warping + warped_frame = cv2.remap(src_frame, x_coords, y_coords, interpolation=cv2.INTER_LINEAR) + + return warped_frame + +def synthesize_intermediate_frames(all_frames, x, n): + # Calculate Optical Flow between the first and last frame + frame1 = cv2.cvtColor(all_frames[x], cv2.COLOR_BGR2GRAY) + frame2 = cv2.cvtColor(all_frames[x + n], cv2.COLOR_BGR2GRAY) + flow = cv2.calcOpticalFlowFarneback(frame1, frame2, None, 0.5, 3, 15, 3, 5, 1.2, 0) + + synthesized_frames = [] + for i in range(1, n): # For each intermediate frame + alpha = i / (n) # Interpolation factor + intermediate_flow = flow * alpha # Interpolate the flow + intermediate_frame = warp_frame(all_frames[x], intermediate_flow) # Warp the first frame + synthesized_frames.append(intermediate_frame) + + return synthesized_frames + + +def map2color(s): + m = hashlib.md5() + m.update(s.encode('utf-8')) + color_code = m.hexdigest()[:6] + return '#' + color_code + +def euclidean_distance(a, b): + return np.sqrt(np.sum((a - b)**2)) + +def adjust_array(x, k): + len_x = len(x) + len_k = len(k) + + # If x is shorter than k, pad with zeros + if len_x < len_k: + return np.pad(x, (0, len_k - len_x), 'constant') + + # If x is longer than k, truncate x + elif len_x > len_k: + return x[:len_k] + + # If both are of same length + else: + return x + +def onset_to_frame(onset_times, audio_length, fps): + # Calculate total number of frames for the given audio length + total_frames = int(audio_length * fps) + + # Create an array of zeros of shape (total_frames,) + frame_array = np.zeros(total_frames, dtype=np.int32) + + # For each onset time, calculate the frame number and set it to 1 + for onset in onset_times: + frame_num = int(onset * fps) + # Check if the frame number is within the array bounds + if 0 <= frame_num < total_frames: + frame_array[frame_num] = 1 + + return frame_array + +# def np_slerp(q1, q2, t): +# dot_product = np.sum(q1 * q2, axis=-1) +# q2_flip = np.where(dot_product[:, None] < 0, -q2, q2) # Flip quaternions where dot_product is negative +# dot_product = np.abs(dot_product) + +# angle = np.arccos(np.clip(dot_product, -1, 1)) +# sin_angle = np.sin(angle) + +# t1 = np.sin((1.0 - t) * angle) / sin_angle +# t2 = np.sin(t * angle) / sin_angle + +# return t1 * q1 + t2 * q2_flip + + +def smooth_rotvec_animations(animation1, animation2, blend_frames): + """ + Smoothly transition between two animation clips using SLERP. + + Parameters: + - animation1: The first animation clip, a numpy array of shape [n, k]. + - animation2: The second animation clip, a numpy array of shape [n, k]. + - blend_frames: Number of frames over which to blend the two animations. + + Returns: + - A smoothly blended animation clip of shape [2n, k]. + """ + + # Ensure blend_frames doesn't exceed the length of either animation + n1, k1 = animation1.shape + n2, k2 = animation2.shape + animation1 = animation1.reshape(n1, k1//3, 3) + animation2 = animation2.reshape(n2, k2//3, 3) + blend_frames = min(blend_frames, len(animation1), len(animation2)) + all_int = [] + for i in range(k1//3): + # Convert rotation vectors to quaternion for the overlapping part + q = R.from_rotvec(np.concatenate([animation1[0:1, i], animation2[-2:-1, i]], axis=0))#.as_quat() + # q2 = R.from_rotvec()#.as_quat() + times = [0, blend_frames * 2 - 1] + slerp = Slerp(times, q) + interpolated = slerp(np.arange(blend_frames * 2)) + interpolated_rotvecs = interpolated.as_rotvec() + all_int.append(interpolated_rotvecs) + interpolated_rotvecs = np.concatenate(all_int, axis=1) + # result = np.vstack((animation1[:-blend_frames], interpolated_rotvecs, animation2[blend_frames:])) + result = interpolated_rotvecs.reshape(2*n1, k1) + return result + +def smooth_animations(animation1, animation2, blend_frames): + """ + Smoothly transition between two animation clips using linear interpolation. + + Parameters: + - animation1: The first animation clip, a numpy array of shape [n, k]. + - animation2: The second animation clip, a numpy array of shape [n, k]. + - blend_frames: Number of frames over which to blend the two animations. + + Returns: + - A smoothly blended animation clip of shape [2n, k]. + """ + + # Ensure blend_frames doesn't exceed the length of either animation + blend_frames = min(blend_frames, len(animation1), len(animation2)) + + # Extract overlapping sections + overlap_a1 = animation1[-blend_frames:-blend_frames+1, :] + overlap_a2 = animation2[blend_frames-1:blend_frames, :] + + # Create blend weights for linear interpolation + alpha = np.linspace(0, 1, 2 * blend_frames).reshape(-1, 1) + + # Linearly interpolate between overlapping sections + blended_overlap = overlap_a1 * (1 - alpha) + overlap_a2 * alpha + + # Extend the animations to form the result with 2n frames + if blend_frames == len(animation1) and blend_frames == len(animation2): + result = blended_overlap + else: + before_blend = animation1[:-blend_frames] + after_blend = animation2[blend_frames:] + result = np.vstack((before_blend, blended_overlap, after_blend)) + return result + +def interpolate_sequence(quaternions): + bs, n, j, _ = quaternions.shape + new_n = 2 * n + new_quaternions = torch.zeros((bs, new_n, j, 4), device=quaternions.device, dtype=quaternions.dtype) + + for i in range(n): + q1 = quaternions[:, i, :, :] + new_quaternions[:, 2*i, :, :] = q1 + + if i < n - 1: + q2 = quaternions[:, i + 1, :, :] + new_quaternions[:, 2*i + 1, :, :] = slerp(q1, q2, 0.5) + else: + # For the last point, duplicate the value + new_quaternions[:, 2*i + 1, :, :] = q1 + + return new_quaternions + +def quaternion_multiply(q1, q2): + w1, x1, y1, z1 = q1 + w2, x2, y2, z2 = q2 + w = w1 * w2 - x1 * x2 - y1 * y2 - z1 * z2 + x = w1 * x2 + x1 * w2 + y1 * z2 - z1 * y2 + y = w1 * y2 + y1 * w2 + z1 * x2 - x1 * z2 + z = w1 * z2 + z1 * w2 + x1 * y2 - y1 * x2 + return w, x, y, z + +def quaternion_conjugate(q): + w, x, y, z = q + return (w, -x, -y, -z) + +def slerp(q1, q2, t): + dot = torch.sum(q1 * q2, dim=-1, keepdim=True) + + flip = (dot < 0).float() + q2 = (1 - flip * 2) * q2 + dot = dot * (1 - flip * 2) + + DOT_THRESHOLD = 0.9995 + mask = (dot > DOT_THRESHOLD).float() + + theta_0 = torch.acos(dot) + theta = theta_0 * t + + q3 = q2 - q1 * dot + q3 = q3 / torch.norm(q3, dim=-1, keepdim=True) + + interpolated = (torch.cos(theta) * q1 + torch.sin(theta) * q3) + + return mask * (q1 + t * (q2 - q1)) + (1 - mask) * interpolated + +def estimate_linear_velocity(data_seq, dt): + ''' + Given some batched data sequences of T timesteps in the shape (B, T, ...), estimates + the velocity for the middle T-2 steps using a second order central difference scheme. + The first and last frames are with forward and backward first-order + differences, respectively + - h : step size + ''' + # first steps is forward diff (t+1 - t) / dt + init_vel = (data_seq[:, 1:2] - data_seq[:, :1]) / dt + # middle steps are second order (t+1 - t-1) / 2dt + middle_vel = (data_seq[:, 2:] - data_seq[:, 0:-2]) / (2 * dt) + # last step is backward diff (t - t-1) / dt + final_vel = (data_seq[:, -1:] - data_seq[:, -2:-1]) / dt + + vel_seq = torch.cat([init_vel, middle_vel, final_vel], dim=1) + return vel_seq + +def velocity2position(data_seq, dt, init_pos): + res_trans = [] + for i in range(data_seq.shape[1]): + if i == 0: + res_trans.append(init_pos.unsqueeze(1)) + else: + res = data_seq[:, i-1:i] * dt + res_trans[-1] + res_trans.append(res) + return torch.cat(res_trans, dim=1) + +def estimate_angular_velocity(rot_seq, dt): + ''' + Given a batch of sequences of T rotation matrices, estimates angular velocity at T-2 steps. + Input sequence should be of shape (B, T, ..., 3, 3) + ''' + # see https://en.wikipedia.org/wiki/Angular_velocity#Calculation_from_the_orientation_matrix + dRdt = estimate_linear_velocity(rot_seq, dt) + R = rot_seq + RT = R.transpose(-1, -2) + # compute skew-symmetric angular velocity tensor + w_mat = torch.matmul(dRdt, RT) + # pull out angular velocity vector by averaging symmetric entries + w_x = (-w_mat[..., 1, 2] + w_mat[..., 2, 1]) / 2.0 + w_y = (w_mat[..., 0, 2] - w_mat[..., 2, 0]) / 2.0 + w_z = (-w_mat[..., 0, 1] + w_mat[..., 1, 0]) / 2.0 + w = torch.stack([w_x, w_y, w_z], axis=-1) + return w + +def image_from_bytes(image_bytes): + import matplotlib.image as mpimg + from io import BytesIO + return mpimg.imread(BytesIO(image_bytes), format='PNG') + +def process_frame(i, vertices_all, vertices1_all, faces, output_dir, use_matplotlib, filenames, camera_params, camera_params1): + import matplotlib + matplotlib.use('Agg') + import matplotlib.pyplot as plt + import trimesh + import pyvirtualdisplay as Display + + vertices = vertices_all[i] + vertices1 = vertices1_all[i] + filename = f"{output_dir}frame_{i}.png" + filenames.append(filename) + if i%100 == 0: + print('processed', i, 'frames') + #time_s = time.time() + #print(vertices.shape) + if use_matplotlib: + fig = plt.figure(figsize=(20, 10)) + ax = fig.add_subplot(121, projection="3d") + fig.subplots_adjust(left=0, right=1, bottom=0, top=1) + #ax.view_init(elev=0, azim=90) + x = vertices[:, 0] + y = vertices[:, 1] + z = vertices[:, 2] + ax.scatter(x, y, z, s=0.5) + ax.set_xlim([-1.0, 1.0]) + ax.set_ylim([-0.5, 1.5])#heigth + ax.set_zlim([-0, 2])#depth + ax.set_box_aspect((1,1,1)) + else: + mesh = trimesh.Trimesh(vertices, faces) + scene = mesh.scene() + scene.camera.fov = camera_params['fov'] + scene.camera.resolution = camera_params['resolution'] + scene.camera.z_near = camera_params['z_near'] + scene.camera.z_far = camera_params['z_far'] + scene.graph[scene.camera.name] = camera_params['transform'] + fig, ax =plt.subplots(1,2, figsize=(16, 6)) + image = scene.save_image(resolution=[640, 480], visible=False) + im0 = ax[0].imshow(image_from_bytes(image)) + ax[0].axis('off') + + if use_matplotlib: + ax2 = fig.add_subplot(122, projection="3d") + ax2.set_box_aspect((1,1,1)) + fig.subplots_adjust(left=0, right=1, bottom=0, top=1) + x1 = vertices1[:, 0] + y1 = vertices1[:, 1] + z1 = vertices1[:, 2] + ax2.scatter(x1, y1, z1, s=0.5) + ax2.set_xlim([-1.0, 1.0]) + ax2.set_ylim([-0.5, 1.5])#heigth + ax2.set_zlim([-0, 2]) + plt.savefig(filename, bbox_inches='tight') + plt.close(fig) + else: + mesh1 = trimesh.Trimesh(vertices1, faces) + scene1 = mesh1.scene() + scene1.camera.fov = camera_params1['fov'] + scene1.camera.resolution = camera_params1['resolution'] + scene1.camera.z_near = camera_params1['z_near'] + scene1.camera.z_far = camera_params1['z_far'] + scene1.graph[scene1.camera.name] = camera_params1['transform'] + image1 = scene1.save_image(resolution=[640, 480], visible=False) + im1 = ax[1].imshow(image_from_bytes(image1)) + ax[1].axis('off') + plt.savefig(filename, bbox_inches='tight') + plt.close(fig) + +def generate_images(frames, vertices_all, vertices1_all, faces, output_dir, use_matplotlib, filenames): + import multiprocessing + import trimesh + num_cores = multiprocessing.cpu_count() # This will get the number of cores on your machine. + mesh = trimesh.Trimesh(vertices_all[0], faces) + scene = mesh.scene() + camera_params = { + 'fov': scene.camera.fov, + 'resolution': scene.camera.resolution, + 'focal': scene.camera.focal, + 'z_near': scene.camera.z_near, + "z_far": scene.camera.z_far, + 'transform': scene.graph[scene.camera.name][0] + } + mesh1 = trimesh.Trimesh(vertices1_all[0], faces) + scene1 = mesh1.scene() + camera_params1 = { + 'fov': scene1.camera.fov, + 'resolution': scene1.camera.resolution, + 'focal': scene1.camera.focal, + 'z_near': scene1.camera.z_near, + "z_far": scene1.camera.z_far, + 'transform': scene1.graph[scene1.camera.name][0] + } + # Use a Pool to manage the processes + # print(num_cores) + progress = multiprocessing.Value('i', 0) + lock = multiprocessing.Lock() + with multiprocessing.Pool(num_cores) as pool: + pool.starmap(process_frame, [(i, vertices_all, vertices1_all, faces, output_dir, use_matplotlib, filenames, camera_params, camera_params1) for i in range(frames)]) + +def render_one_sequence( + res_npz_path, + gt_npz_path, + output_dir, + audio_path, + model_folder="/data/datasets/smplx_models/", + model_type='smplx', + gender='NEUTRAL_2020', + ext='npz', + num_betas=300, + num_expression_coeffs=100, + use_face_contour=False, + use_matplotlib=False, + args=None): + import smplx + import matplotlib.pyplot as plt + import imageio + from tqdm import tqdm + import os + import numpy as np + import torch + import moviepy.editor as mp + import librosa + + model = smplx.create(model_folder, model_type=model_type, + gender=gender, use_face_contour=use_face_contour, + num_betas=num_betas, + num_expression_coeffs=num_expression_coeffs, + ext=ext, use_pca=False).cuda() + + #data_npz = np.load(f"{output_dir}{res_npz_path}.npz") + data_np_body = np.load(res_npz_path, allow_pickle=True) + gt_np_body = np.load(gt_npz_path, allow_pickle=True) + + if not os.path.exists(output_dir): os.makedirs(output_dir) + filenames = [] + if not use_matplotlib: + import trimesh + #import pyrender + from pyvirtualdisplay import Display + display = Display(visible=0, size=(640, 480)) + display.start() + faces = np.load(f"{model_folder}/smplx/SMPLX_NEUTRAL_2020.npz", allow_pickle=True)["f"] + seconds = 1 + + n = data_np_body["poses"].shape[0] + beta = torch.from_numpy(data_np_body["betas"]).to(torch.float32).unsqueeze(0).cuda() + beta = beta.repeat(n, 1) + expression = torch.from_numpy(data_np_body["expressions"][:n]).to(torch.float32).cuda() + jaw_pose = torch.from_numpy(data_np_body["poses"][:n, 66:69]).to(torch.float32).cuda() + pose = torch.from_numpy(data_np_body["poses"][:n]).to(torch.float32).cuda() + transl = torch.from_numpy(data_np_body["trans"][:n]).to(torch.float32).cuda() + + output = model(betas=beta, transl=transl, expression=expression, jaw_pose=jaw_pose, + global_orient=pose[:,:3], body_pose=pose[:,3:21*3+3], left_hand_pose=pose[:,25*3:40*3], right_hand_pose=pose[:,40*3:55*3], + leye_pose=pose[:, 69:72], + reye_pose=pose[:, 72:75], + return_verts=True) + vertices_all = output["vertices"].cpu().detach().numpy() + + beta1 = torch.from_numpy(gt_np_body["betas"]).to(torch.float32).unsqueeze(0).cuda() + expression1 = torch.from_numpy(gt_np_body["expressions"][:n]).to(torch.float32).cuda() + jaw_pose1 = torch.from_numpy(gt_np_body["poses"][:n,66:69]).to(torch.float32).cuda() + pose1 = torch.from_numpy(gt_np_body["poses"][:n]).to(torch.float32).cuda() + transl1 = torch.from_numpy(gt_np_body["trans"][:n]).to(torch.float32).cuda() + output1 = model(betas=beta1, transl=transl1, expression=expression1, jaw_pose=jaw_pose1, global_orient=pose1[:,:3], body_pose=pose1[:,3:21*3+3], left_hand_pose=pose1[:,25*3:40*3], right_hand_pose=pose1[:,40*3:55*3], + leye_pose=pose1[:, 69:72], + reye_pose=pose1[:, 72:75],return_verts=True) + vertices1_all = output1["vertices"].cpu().detach().numpy() + if args.debug: + seconds = 1 + else: + seconds = vertices_all.shape[0]//30 + + time_s = time.time() + generate_images(int(seconds*30), vertices_all, vertices1_all, faces, output_dir, use_matplotlib, filenames) + filenames = [f"{output_dir}frame_{i}.png" for i in range(int(seconds*30))] + + images = [imageio.imread(filename) for filename in filenames] + imageio.mimsave(f"{output_dir}raw_{res_npz_path.split('/')[-1][:-4]}.mp4", images, fps=30) + for filename in filenames: + os.remove(filename) + + video = mp.VideoFileClip(f"{output_dir}raw_{res_npz_path.split('/')[-1][:-4]}.mp4") + audio = mp.AudioFileClip(audio_path) + if audio.duration > video.duration: + audio = audio.subclip(0, video.duration) + final_clip = video.set_audio(audio) + final_clip.write_videofile(f"{output_dir}{res_npz_path.split('/')[-1][4:-4]}.mp4") + os.remove(f"{output_dir}raw_{res_npz_path.split('/')[-1][:-4]}.mp4") + +def print_exp_info(args): + logger.info(pprint.pformat(vars(args))) + logger.info(f"# ------------ {args.name} ----------- #") + logger.info("PyTorch version: {}".format(torch.__version__)) + logger.info("CUDA version: {}".format(torch.version.cuda)) + logger.info("{} GPUs".format(torch.cuda.device_count())) + logger.info(f"Random Seed: {args.random_seed}") + +def args2csv(args, get_head=False, list4print=[]): + for k, v in args.items(): + if isinstance(args[k], dict): + args2csv(args[k], get_head, list4print) + else: list4print.append(k) if get_head else list4print.append(v) + return list4print + +class EpochTracker: + def __init__(self, metric_names, metric_directions): + assert len(metric_names) == len(metric_directions), "Metric names and directions should have the same length" + + + self.metric_names = metric_names + self.states = ['train', 'val', 'test'] + self.types = ['last', 'best'] + + + self.values = {name: {state: {type_: {'value': np.inf if not is_higher_better else -np.inf, 'epoch': 0} + for type_ in self.types} + for state in self.states} + for name, is_higher_better in zip(metric_names, metric_directions)} + + self.loss_meters = {name: {state: AverageMeter(f"{name}_{state}") + for state in self.states} + for name in metric_names} + + + self.is_higher_better = {name: direction for name, direction in zip(metric_names, metric_directions)} + self.train_history = {name: [] for name in metric_names} + self.val_history = {name: [] for name in metric_names} + + + def update_meter(self, name, state, value): + self.loss_meters[name][state].update(value) + + + def update_values(self, name, state, epoch): + value_avg = self.loss_meters[name][state].avg + new_best = False + + + if ((value_avg < self.values[name][state]['best']['value'] and not self.is_higher_better[name]) or + (value_avg > self.values[name][state]['best']['value'] and self.is_higher_better[name])): + self.values[name][state]['best']['value'] = value_avg + self.values[name][state]['best']['epoch'] = epoch + new_best = True + self.values[name][state]['last']['value'] = value_avg + self.values[name][state]['last']['epoch'] = epoch + return new_best + + + def get(self, name, state, type_): + return self.values[name][state][type_] + + + def reset(self): + for name in self.metric_names: + for state in self.states: + self.loss_meters[name][state].reset() + + + def flatten_values(self): + flat_dict = {} + for name in self.metric_names: + for state in self.states: + for type_ in self.types: + value_key = f"{name}_{state}_{type_}" + epoch_key = f"{name}_{state}_{type_}_epoch" + flat_dict[value_key] = self.values[name][state][type_]['value'] + flat_dict[epoch_key] = self.values[name][state][type_]['epoch'] + return flat_dict + + def update_and_plot(self, name, epoch, save_path): + new_best_train = self.update_values(name, 'train', epoch) + new_best_val = self.update_values(name, 'val', epoch) + + + self.train_history[name].append(self.loss_meters[name]['train'].avg) + self.val_history[name].append(self.loss_meters[name]['val'].avg) + + + train_values = self.train_history[name] + val_values = self.val_history[name] + epochs = list(range(1, len(train_values) + 1)) + + + plt.figure(figsize=(10, 6)) + plt.plot(epochs, train_values, label='Train') + plt.plot(epochs, val_values, label='Val') + plt.title(f'Train vs Val {name} over epochs') + plt.xlabel('Epochs') + plt.ylabel(name) + plt.legend() + plt.savefig(save_path) + plt.close() + + + return new_best_train, new_best_val + +def record_trial(args, tracker): + """ + 1. record notes, score, env_name, experments_path, + """ + csv_path = args.out_path + "custom/" +args.csv_name+".csv" + all_print_dict = vars(args) + all_print_dict.update(tracker.flatten_values()) + if not os.path.exists(csv_path): + pd.DataFrame([all_print_dict]).to_csv(csv_path, index=False) + else: + df_existing = pd.read_csv(csv_path) + df_new = pd.DataFrame([all_print_dict]) + df_aligned = df_existing.append(df_new).fillna("") + df_aligned.to_csv(csv_path, index=False) + +def set_random_seed(args): + os.environ['PYTHONHASHSEED'] = str(args.random_seed) + random.seed(args.random_seed) + np.random.seed(args.random_seed) + torch.manual_seed(args.random_seed) + torch.cuda.manual_seed_all(args.random_seed) + torch.cuda.manual_seed(args.random_seed) + torch.backends.cudnn.deterministic = False #default: False + torch.backends.cudnn.benchmark = True #default: False + torch.backends.cudnn.enabled = args.cudnn_enabled #default: True + +def save_checkpoints(save_path, model, opt=None, epoch=None, lrs=None): + if lrs is not None: + states = { 'model_state': model.state_dict(), + 'epoch': epoch + 1, + 'opt_state': opt.state_dict(), + 'lrs':lrs.state_dict(),} + elif opt is not None: + states = { 'model_state': model.state_dict(), + 'epoch': epoch + 1, + 'opt_state': opt.state_dict(),} + else: + states = { 'model_state': model.state_dict(),} + torch.save(states, save_path) + +def load_checkpoints(model, save_path, load_name='model'): + states = torch.load(save_path) + new_weights = OrderedDict() + flag=False + for k, v in states['model_state'].items(): + #print(k) + if "module" not in k: + break + else: + new_weights[k[7:]]=v + flag=True + if flag: + try: + model.load_state_dict(new_weights) + except: + #print(states['model_state']) + model.load_state_dict(states['model_state']) + else: + model.load_state_dict(states['model_state']) + logger.info(f"load self-pretrained checkpoints for {load_name}") + +def model_complexity(model, args): + from ptflops import get_model_complexity_info + flops, params = get_model_complexity_info(model, (args.T_GLOBAL._DIM, args.TRAIN.CROP, args.TRAIN), + as_strings=False, print_per_layer_stat=False) + logging.info('{:<30} {:<8} BFlops'.format('Computational complexity: ', flops / 1e9)) + logging.info('{:<30} {:<8} MParams'.format('Number of parameters: ', params / 1e6)) + +class AverageMeter(object): + """Computes and stores the average and current value""" + def __init__(self, name, fmt=':f'): + self.name = name + self.fmt = fmt + self.reset() + + def reset(self): + self.val = 0 + self.avg = 0 + self.sum = 0 + self.count = 0 + + def update(self, val, n=1): + self.val = val + self.sum += val * n + self.count += n + self.avg = self.sum / self.count + + def __str__(self): + fmtstr = '{name} {val' + self.fmt + '} ({avg' + self.fmt + '})' + return fmtstr.format(**self.__dict__) \ No newline at end of file diff --git a/utils/other_tools_hf.py b/utils/other_tools_hf.py new file mode 100644 index 0000000000000000000000000000000000000000..9613f2320104fd9b24d3524ca020014be967c6da --- /dev/null +++ b/utils/other_tools_hf.py @@ -0,0 +1,956 @@ +import os +import numpy as np +import random +import torch +import shutil +import csv +import pprint +import pandas as pd +from loguru import logger +from collections import OrderedDict +import matplotlib.pyplot as plt +import pickle +import time +import hashlib +from scipy.spatial.transform import Rotation as R +from scipy.spatial.transform import Slerp +import cv2 +import utils.media +import utils.fast_render + +def write_wav_names_to_csv(folder_path, csv_path): + """ + Traverse a folder and write the base names of all .wav files to a CSV file. + + :param folder_path: Path to the folder to traverse. + :param csv_path: Path to the CSV file to write. + """ + # Open the CSV file for writing + with open(csv_path, mode='w', newline='') as file: + writer = csv.writer(file) + # Write the header + writer.writerow(['id', 'type']) + + # Walk through the folder + for root, dirs, files in os.walk(folder_path): + for file in files: + # Check if the file ends with .wav + if file.endswith('.wav'): + # Extract the base name without the extension + base_name = os.path.splitext(file)[0] + # Write the base name and type to the CSV + writer.writerow([base_name, 'test']) + +def resize_motion_sequence_tensor(sequence, target_frames): + """ + Resize a batch of 8-frame motion sequences to a specified number of frames using interpolation. + + :param sequence: A (bs, 8, 165) tensor representing a batch of 8-frame motion sequences + :param target_frames: An integer representing the desired number of frames in the output sequences + :return: A (bs, target_frames, 165) tensor representing the resized motion sequences + """ + bs, _, _ = sequence.shape + + # Create a time vector for the original and target sequences + original_time = torch.linspace(0, 1, 8, device=sequence.device).view(1, -1, 1) + target_time = torch.linspace(0, 1, target_frames, device=sequence.device).view(1, -1, 1) + + # Permute the dimensions to (bs, 165, 8) for interpolation + sequence = sequence.permute(0, 2, 1) + + # Interpolate each joint's motion to the target number of frames + resized_sequence = torch.nn.functional.interpolate(sequence, size=target_frames, mode='linear', align_corners=True) + + # Permute the dimensions back to (bs, target_frames, 165) + resized_sequence = resized_sequence.permute(0, 2, 1) + + return resized_sequence + +def adjust_speed_according_to_ratio_tensor(chunks): + """ + Adjust the playback speed within a batch of 32-frame chunks according to random intervals. + + :param chunks: A (bs, 32, 165) tensor representing a batch of motion chunks + :return: A (bs, 32, 165) tensor representing the motion chunks after speed adjustment + """ + bs, _, _ = chunks.shape + + # Step 1: Divide the chunk into 4 equal intervals of 8 frames + equal_intervals = torch.chunk(chunks, 4, dim=1) + + # Step 2: Randomly sample 3 points within the chunk to determine new intervals + success = 0 + all_success = [] + #sample_points = torch.sort(torch.randint(1, 32, (bs, 3), device=chunks.device), dim=1).values + # new_intervals_boundaries = torch.cat([torch.zeros((bs, 1), device=chunks.device, dtype=torch.long), sample_points, 32*torch.ones((bs, 1), device=chunks.device, dtype=torch.long)], dim=1) + while success != 1: + sample_points = sorted(random.sample(range(1, 32), 3)) + new_intervals_boundaries = [0] + sample_points + [32] + new_intervals = [chunks[0][new_intervals_boundaries[i]:new_intervals_boundaries[i+1]] for i in range(4)] + speed_ratios = [8 / len(new_interval) for new_interval in new_intervals] + # if any of the speed ratios is greater than 3 or less than 0.33, resample + if all([0.33 <= speed_ratio <= 3 for speed_ratio in speed_ratios]): + success += 1 + all_success.append(new_intervals_boundaries) + new_intervals_boundaries = torch.from_numpy(np.array(all_success)) + # print(new_intervals_boundaries) + all_shapes = new_intervals_boundaries[:, 1:] - new_intervals_boundaries[:, :-1] + # Step 4: Adjust the speed of each new interval + adjusted_intervals = [] + # print(equal_intervals[0].shape) + for i in range(4): + adjusted_interval = resize_motion_sequence_tensor(equal_intervals[i], all_shapes[0, i]) + adjusted_intervals.append(adjusted_interval) + + # Step 5: Concatenate the adjusted intervals + adjusted_chunk = torch.cat(adjusted_intervals, dim=1) + + return adjusted_chunk + +def compute_exact_iou(bbox1, bbox2): + x1 = max(bbox1[0], bbox2[0]) + y1 = max(bbox1[1], bbox2[1]) + x2 = min(bbox1[0] + bbox1[2], bbox2[0] + bbox2[2]) + y2 = min(bbox1[1] + bbox1[3], bbox2[1] + bbox2[3]) + + intersection_area = max(0, x2 - x1) * max(0, y2 - y1) + bbox1_area = bbox1[2] * bbox1[3] + bbox2_area = bbox2[2] * bbox2[3] + union_area = bbox1_area + bbox2_area - intersection_area + + if union_area == 0: + return 0 + + return intersection_area / union_area + +def compute_iou(mask1, mask2): + # Compute the intersection + intersection = np.logical_and(mask1, mask2).sum() + + # Compute the union + union = np.logical_or(mask1, mask2).sum() + + # Compute the IoU + iou = intersection / union + + return iou + +def blankblending(all_frames, x, n): + return all_frames[x:x+n+1] + + +def load_video_as_numpy_array(video_path): + cap = cv2.VideoCapture(video_path) + + # Using list comprehension to read frames and store in a list + frames = [frame for ret, frame in iter(lambda: cap.read(), (False, None)) if ret] + + cap.release() + + return np.array(frames) + +def synthesize_intermediate_frames_bidirectional(all_frames, x, n): + frame1 = all_frames[x] + frame2 = all_frames[x + n] + + # Convert the frames to grayscale + gray1 = cv2.cvtColor(frame1, cv2.COLOR_BGR2GRAY) + gray2 = cv2.cvtColor(frame2, cv2.COLOR_BGR2GRAY) + + # Calculate the forward and backward optical flow + forward_flow = cv2.calcOpticalFlowFarneback(gray1, gray2, None, 0.5, 3, 15, 3, 5, 1.2, 0) + backward_flow = cv2.calcOpticalFlowFarneback(gray2, gray1, None, 0.5, 3, 15, 3, 5, 1.2, 0) + + synthesized_frames = [] + for i in range(1, n): # For each intermediate frame between x and x + n + alpha = i / n # Interpolation factor + + # Compute the intermediate forward and backward flow + intermediate_forward_flow = forward_flow * alpha + intermediate_backward_flow = backward_flow * (1 - alpha) + + # Warp the frames based on the intermediate flow + h, w = frame1.shape[:2] + flow_map = np.column_stack((np.repeat(np.arange(h), w), np.tile(np.arange(w), h))) + forward_displacement = flow_map + intermediate_forward_flow.reshape(-1, 2) + backward_displacement = flow_map - intermediate_backward_flow.reshape(-1, 2) + + # Use cv2.remap for efficient warping + remap_x_forward, remap_y_forward = np.clip(forward_displacement[:, 1], 0, w - 1), np.clip(forward_displacement[:, 0], 0, h - 1) + remap_x_backward, remap_y_backward = np.clip(backward_displacement[:, 1], 0, w - 1), np.clip(backward_displacement[:, 0], 0, h - 1) + + warped_forward = cv2.remap(frame1, remap_x_forward.reshape(h, w).astype(np.float32), remap_y_forward.reshape(h, w).astype(np.float32), interpolation=cv2.INTER_LINEAR) + warped_backward = cv2.remap(frame2, remap_x_backward.reshape(h, w).astype(np.float32), remap_y_backward.reshape(h, w).astype(np.float32), interpolation=cv2.INTER_LINEAR) + + # Blend the warped frames to generate the intermediate frame + intermediate_frame = cv2.addWeighted(warped_forward, 1 - alpha, warped_backward, alpha, 0) + synthesized_frames.append(intermediate_frame) + + return synthesized_frames # Return n-2 synthesized intermediate frames + + +def linear_interpolate_frames(all_frames, x, n): + frame1 = all_frames[x] + frame2 = all_frames[x + n] + + synthesized_frames = [] + for i in range(1, n): # For each intermediate frame between x and x + n + alpha = i / (n) # Correct interpolation factor + inter_frame = cv2.addWeighted(frame1, 1 - alpha, frame2, alpha, 0) + synthesized_frames.append(inter_frame) + return synthesized_frames[:-1] + +def warp_frame(src_frame, flow): + h, w = flow.shape[:2] + flow_map = np.column_stack((np.repeat(np.arange(h), w), np.tile(np.arange(w), h))) + displacement = flow_map + flow.reshape(-1, 2) + + # Extract x and y coordinates of the displacement + x_coords = np.clip(displacement[:, 1], 0, w - 1).reshape(h, w).astype(np.float32) + y_coords = np.clip(displacement[:, 0], 0, h - 1).reshape(h, w).astype(np.float32) + + # Use cv2.remap for efficient warping + warped_frame = cv2.remap(src_frame, x_coords, y_coords, interpolation=cv2.INTER_LINEAR) + + return warped_frame + +def synthesize_intermediate_frames(all_frames, x, n): + # Calculate Optical Flow between the first and last frame + frame1 = cv2.cvtColor(all_frames[x], cv2.COLOR_BGR2GRAY) + frame2 = cv2.cvtColor(all_frames[x + n], cv2.COLOR_BGR2GRAY) + flow = cv2.calcOpticalFlowFarneback(frame1, frame2, None, 0.5, 3, 15, 3, 5, 1.2, 0) + + synthesized_frames = [] + for i in range(1, n): # For each intermediate frame + alpha = i / (n) # Interpolation factor + intermediate_flow = flow * alpha # Interpolate the flow + intermediate_frame = warp_frame(all_frames[x], intermediate_flow) # Warp the first frame + synthesized_frames.append(intermediate_frame) + + return synthesized_frames + + +def map2color(s): + m = hashlib.md5() + m.update(s.encode('utf-8')) + color_code = m.hexdigest()[:6] + return '#' + color_code + +def euclidean_distance(a, b): + return np.sqrt(np.sum((a - b)**2)) + +def adjust_array(x, k): + len_x = len(x) + len_k = len(k) + + # If x is shorter than k, pad with zeros + if len_x < len_k: + return np.pad(x, (0, len_k - len_x), 'constant') + + # If x is longer than k, truncate x + elif len_x > len_k: + return x[:len_k] + + # If both are of same length + else: + return x + +def onset_to_frame(onset_times, audio_length, fps): + # Calculate total number of frames for the given audio length + total_frames = int(audio_length * fps) + + # Create an array of zeros of shape (total_frames,) + frame_array = np.zeros(total_frames, dtype=np.int32) + + # For each onset time, calculate the frame number and set it to 1 + for onset in onset_times: + frame_num = int(onset * fps) + # Check if the frame number is within the array bounds + if 0 <= frame_num < total_frames: + frame_array[frame_num] = 1 + + return frame_array + +# def np_slerp(q1, q2, t): +# dot_product = np.sum(q1 * q2, axis=-1) +# q2_flip = np.where(dot_product[:, None] < 0, -q2, q2) # Flip quaternions where dot_product is negative +# dot_product = np.abs(dot_product) + +# angle = np.arccos(np.clip(dot_product, -1, 1)) +# sin_angle = np.sin(angle) + +# t1 = np.sin((1.0 - t) * angle) / sin_angle +# t2 = np.sin(t * angle) / sin_angle + +# return t1 * q1 + t2 * q2_flip + + +def smooth_rotvec_animations(animation1, animation2, blend_frames): + """ + Smoothly transition between two animation clips using SLERP. + + Parameters: + - animation1: The first animation clip, a numpy array of shape [n, k]. + - animation2: The second animation clip, a numpy array of shape [n, k]. + - blend_frames: Number of frames over which to blend the two animations. + + Returns: + - A smoothly blended animation clip of shape [2n, k]. + """ + + # Ensure blend_frames doesn't exceed the length of either animation + n1, k1 = animation1.shape + n2, k2 = animation2.shape + animation1 = animation1.reshape(n1, k1//3, 3) + animation2 = animation2.reshape(n2, k2//3, 3) + blend_frames = min(blend_frames, len(animation1), len(animation2)) + all_int = [] + for i in range(k1//3): + # Convert rotation vectors to quaternion for the overlapping part + q = R.from_rotvec(np.concatenate([animation1[0:1, i], animation2[-2:-1, i]], axis=0))#.as_quat() + # q2 = R.from_rotvec()#.as_quat() + times = [0, blend_frames * 2 - 1] + slerp = Slerp(times, q) + interpolated = slerp(np.arange(blend_frames * 2)) + interpolated_rotvecs = interpolated.as_rotvec() + all_int.append(interpolated_rotvecs) + interpolated_rotvecs = np.concatenate(all_int, axis=1) + # result = np.vstack((animation1[:-blend_frames], interpolated_rotvecs, animation2[blend_frames:])) + result = interpolated_rotvecs.reshape(2*n1, k1) + return result + +def smooth_animations(animation1, animation2, blend_frames): + """ + Smoothly transition between two animation clips using linear interpolation. + + Parameters: + - animation1: The first animation clip, a numpy array of shape [n, k]. + - animation2: The second animation clip, a numpy array of shape [n, k]. + - blend_frames: Number of frames over which to blend the two animations. + + Returns: + - A smoothly blended animation clip of shape [2n, k]. + """ + + # Ensure blend_frames doesn't exceed the length of either animation + blend_frames = min(blend_frames, len(animation1), len(animation2)) + + # Extract overlapping sections + overlap_a1 = animation1[-blend_frames:-blend_frames+1, :] + overlap_a2 = animation2[blend_frames-1:blend_frames, :] + + # Create blend weights for linear interpolation + alpha = np.linspace(0, 1, 2 * blend_frames).reshape(-1, 1) + + # Linearly interpolate between overlapping sections + blended_overlap = overlap_a1 * (1 - alpha) + overlap_a2 * alpha + + # Extend the animations to form the result with 2n frames + if blend_frames == len(animation1) and blend_frames == len(animation2): + result = blended_overlap + else: + before_blend = animation1[:-blend_frames] + after_blend = animation2[blend_frames:] + result = np.vstack((before_blend, blended_overlap, after_blend)) + return result + +def interpolate_sequence(quaternions): + bs, n, j, _ = quaternions.shape + new_n = 2 * n + new_quaternions = torch.zeros((bs, new_n, j, 4), device=quaternions.device, dtype=quaternions.dtype) + + for i in range(n): + q1 = quaternions[:, i, :, :] + new_quaternions[:, 2*i, :, :] = q1 + + if i < n - 1: + q2 = quaternions[:, i + 1, :, :] + new_quaternions[:, 2*i + 1, :, :] = slerp(q1, q2, 0.5) + else: + # For the last point, duplicate the value + new_quaternions[:, 2*i + 1, :, :] = q1 + + return new_quaternions + +def quaternion_multiply(q1, q2): + w1, x1, y1, z1 = q1 + w2, x2, y2, z2 = q2 + w = w1 * w2 - x1 * x2 - y1 * y2 - z1 * z2 + x = w1 * x2 + x1 * w2 + y1 * z2 - z1 * y2 + y = w1 * y2 + y1 * w2 + z1 * x2 - x1 * z2 + z = w1 * z2 + z1 * w2 + x1 * y2 - y1 * x2 + return w, x, y, z + +def quaternion_conjugate(q): + w, x, y, z = q + return (w, -x, -y, -z) + +def slerp(q1, q2, t): + dot = torch.sum(q1 * q2, dim=-1, keepdim=True) + + flip = (dot < 0).float() + q2 = (1 - flip * 2) * q2 + dot = dot * (1 - flip * 2) + + DOT_THRESHOLD = 0.9995 + mask = (dot > DOT_THRESHOLD).float() + + theta_0 = torch.acos(dot) + theta = theta_0 * t + + q3 = q2 - q1 * dot + q3 = q3 / torch.norm(q3, dim=-1, keepdim=True) + + interpolated = (torch.cos(theta) * q1 + torch.sin(theta) * q3) + + return mask * (q1 + t * (q2 - q1)) + (1 - mask) * interpolated + +def estimate_linear_velocity(data_seq, dt): + ''' + Given some batched data sequences of T timesteps in the shape (B, T, ...), estimates + the velocity for the middle T-2 steps using a second order central difference scheme. + The first and last frames are with forward and backward first-order + differences, respectively + - h : step size + ''' + # first steps is forward diff (t+1 - t) / dt + init_vel = (data_seq[:, 1:2] - data_seq[:, :1]) / dt + # middle steps are second order (t+1 - t-1) / 2dt + middle_vel = (data_seq[:, 2:] - data_seq[:, 0:-2]) / (2 * dt) + # last step is backward diff (t - t-1) / dt + final_vel = (data_seq[:, -1:] - data_seq[:, -2:-1]) / dt + + vel_seq = torch.cat([init_vel, middle_vel, final_vel], dim=1) + return vel_seq + +def velocity2position(data_seq, dt, init_pos): + res_trans = [] + for i in range(data_seq.shape[1]): + if i == 0: + res_trans.append(init_pos.unsqueeze(1)) + else: + res = data_seq[:, i-1:i] * dt + res_trans[-1] + res_trans.append(res) + return torch.cat(res_trans, dim=1) + +def estimate_angular_velocity(rot_seq, dt): + ''' + Given a batch of sequences of T rotation matrices, estimates angular velocity at T-2 steps. + Input sequence should be of shape (B, T, ..., 3, 3) + ''' + # see https://en.wikipedia.org/wiki/Angular_velocity#Calculation_from_the_orientation_matrix + dRdt = estimate_linear_velocity(rot_seq, dt) + R = rot_seq + RT = R.transpose(-1, -2) + # compute skew-symmetric angular velocity tensor + w_mat = torch.matmul(dRdt, RT) + # pull out angular velocity vector by averaging symmetric entries + w_x = (-w_mat[..., 1, 2] + w_mat[..., 2, 1]) / 2.0 + w_y = (w_mat[..., 0, 2] - w_mat[..., 2, 0]) / 2.0 + w_z = (-w_mat[..., 0, 1] + w_mat[..., 1, 0]) / 2.0 + w = torch.stack([w_x, w_y, w_z], axis=-1) + return w + +def image_from_bytes(image_bytes): + import matplotlib.image as mpimg + from io import BytesIO + return mpimg.imread(BytesIO(image_bytes), format='PNG') + +def process_frame(i, vertices_all, vertices1_all, faces, output_dir, filenames): + import matplotlib + matplotlib.use('Agg') + import matplotlib.pyplot as plt + import trimesh + import pyrender + + def deg_to_rad(degrees): + return degrees * np.pi / 180 + + uniform_color = [220, 220, 220, 255] + resolution = (1000, 1000) + figsize = (10, 10) + + fig, axs = plt.subplots( + nrows=1, + ncols=2, + figsize=(figsize[0] * 2, figsize[1] * 1) + ) + axs = axs.flatten() + + vertices = vertices_all[i] + vertices1 = vertices1_all[i] + filename = f"{output_dir}frame_{i}.png" + filenames.append(filename) + if i%100 == 0: + print('processed', i, 'frames') + #time_s = time.time() + #print(vertices.shape) + angle_rad = deg_to_rad(-2) + pose_camera = np.array([ + [1.0, 0.0, 0.0, 0.0], + [0.0, np.cos(angle_rad), -np.sin(angle_rad), 1.0], + [0.0, np.sin(angle_rad), np.cos(angle_rad), 5.0], + [0.0, 0.0, 0.0, 1.0] + ]) + angle_rad = deg_to_rad(-30) + pose_light = np.array([ + [1.0, 0.0, 0.0, 0.0], + [0.0, np.cos(angle_rad), -np.sin(angle_rad), 0.0], + [0.0, np.sin(angle_rad), np.cos(angle_rad), 3.0], + [0.0, 0.0, 0.0, 1.0] + ]) + + for vtx_idx, vtx in enumerate([vertices, vertices1]): + trimesh_mesh = trimesh.Trimesh( + vertices=vtx, + faces=faces, + vertex_colors=uniform_color + ) + mesh = pyrender.Mesh.from_trimesh( + trimesh_mesh, smooth=True + ) + scene = pyrender.Scene() + scene.add(mesh) + camera = pyrender.OrthographicCamera(xmag=1.0, ymag=1.0) + scene.add(camera, pose=pose_camera) + light = pyrender.DirectionalLight(color=[1.0, 1.0, 1.0], intensity=4.0) + scene.add(light, pose=pose_light) + renderer = pyrender.OffscreenRenderer(*resolution) + color, _ = renderer.render(scene) + axs[vtx_idx].imshow(color) + axs[vtx_idx].axis('off') + renderer.delete() + + plt.savefig(filename, bbox_inches='tight') + plt.close(fig) + +def generate_images(frames, vertices_all, vertices1_all, faces, output_dir, filenames): + import multiprocessing + # import trimesh + num_cores = multiprocessing.cpu_count() - 1 # This will get the number of cores on your machine. + # mesh = trimesh.Trimesh(vertices_all[0], faces) + # scene = mesh.scene() + # fov = scene.camera.fov.copy() + # fov[0] = 80.0 + # fov[1] = 60.0 + # camera_params = { + # 'fov': fov, + # 'resolution': scene.camera.resolution, + # 'focal': scene.camera.focal, + # 'z_near': scene.camera.z_near, + # "z_far": scene.camera.z_far, + # 'transform': scene.graph[scene.camera.name][0] + # } + # mesh1 = trimesh.Trimesh(vertices1_all[0], faces) + # scene1 = mesh1.scene() + # camera_params1 = { + # 'fov': fov, + # 'resolution': scene1.camera.resolution, + # 'focal': scene1.camera.focal, + # 'z_near': scene1.camera.z_near, + # "z_far": scene1.camera.z_far, + # 'transform': scene1.graph[scene1.camera.name][0] + # } + # Use a Pool to manage the processes + # print(num_cores) + # for i in range(frames): + # process_frame(i, vertices_all, vertices1_all, faces, output_dir, use_matplotlib, filenames, camera_params, camera_params1) + for i in range(frames): + process_frame(i*3, vertices_all, vertices1_all, faces, output_dir, filenames) + + # progress = multiprocessing.Value('i', 0) + # lock = multiprocessing.Lock() + # with multiprocessing.Pool(num_cores) as pool: + # # pool.starmap(process_frame, [(i, vertices_all, vertices1_all, faces, output_dir, use_matplotlib, filenames, camera_params, camera_params1) for i in range(frames)]) + # pool.starmap( + # process_frame, + # [ + # (i, vertices_all, vertices1_all, faces, output_dir, filenames) + # for i in range(frames) + # ] + # ) + + # progress = multiprocessing.Value('i', 0) + # lock = multiprocessing.Lock() + # with multiprocessing.Pool(num_cores) as pool: + # # pool.starmap(process_frame, [(i, vertices_all, vertices1_all, faces, output_dir, use_matplotlib, filenames, camera_params, camera_params1) for i in range(frames)]) + # pool.starmap( + # process_frame, + # [ + # (i, vertices_all, vertices1_all, faces, output_dir, filenames) + # for i in range(frames) + # ] + # ) + +def render_one_sequence( + res_npz_path, + gt_npz_path, + output_dir, + audio_path, + model_folder="/data/datasets/smplx_models/", + model_type='smplx', + gender='NEUTRAL_2020', + ext='npz', + num_betas=300, + num_expression_coeffs=100, + use_face_contour=False, + use_matplotlib=False, + args=None): + import smplx + import matplotlib.pyplot as plt + import imageio + from tqdm import tqdm + import os + import numpy as np + import torch + import moviepy.editor as mp + import librosa + + model = smplx.create(model_folder, model_type=model_type, + gender=gender, use_face_contour=use_face_contour, + num_betas=num_betas, + num_expression_coeffs=num_expression_coeffs, + ext=ext, use_pca=False).cuda() + + #data_npz = np.load(f"{output_dir}{res_npz_path}.npz") + data_np_body = np.load(res_npz_path, allow_pickle=True) + gt_np_body = np.load(gt_npz_path, allow_pickle=True) + + if not os.path.exists(output_dir): os.makedirs(output_dir) + # if not use_matplotlib: + # import trimesh + #import pyrender + from pyvirtualdisplay import Display + #''' + #display = Display(visible=0, size=(1000, 1000)) + #display.start() + faces = np.load(f"{model_folder}/smplx/SMPLX_NEUTRAL_2020.npz", allow_pickle=True)["f"] + seconds = 1 + #data_npz["jaw_pose"].shape[0] + n = data_np_body["poses"].shape[0] + beta = torch.from_numpy(data_np_body["betas"]).to(torch.float32).unsqueeze(0).cuda() + beta = beta.repeat(n, 1) + expression = torch.from_numpy(data_np_body["expressions"][:n]).to(torch.float32).cuda() + jaw_pose = torch.from_numpy(data_np_body["poses"][:n, 66:69]).to(torch.float32).cuda() + pose = torch.from_numpy(data_np_body["poses"][:n]).to(torch.float32).cuda() + transl = torch.from_numpy(data_np_body["trans"][:n]).to(torch.float32).cuda() + # print(beta.shape, expression.shape, jaw_pose.shape, pose.shape, transl.shape, pose[:,:3].shape) + output = model(betas=beta, transl=transl, expression=expression, jaw_pose=jaw_pose, + global_orient=pose[:,:3], body_pose=pose[:,3:21*3+3], left_hand_pose=pose[:,25*3:40*3], right_hand_pose=pose[:,40*3:55*3], + leye_pose=pose[:, 69:72], + reye_pose=pose[:, 72:75], + return_verts=True) + vertices_all = output["vertices"].cpu().detach().numpy() + + beta1 = torch.from_numpy(gt_np_body["betas"]).to(torch.float32).unsqueeze(0).cuda() + expression1 = torch.from_numpy(gt_np_body["expressions"][:n]).to(torch.float32).cuda() + jaw_pose1 = torch.from_numpy(gt_np_body["poses"][:n,66:69]).to(torch.float32).cuda() + pose1 = torch.from_numpy(gt_np_body["poses"][:n]).to(torch.float32).cuda() + transl1 = torch.from_numpy(gt_np_body["trans"][:n]).to(torch.float32).cuda() + output1 = model(betas=beta1, transl=transl1, expression=expression1, jaw_pose=jaw_pose1, global_orient=pose1[:,:3], body_pose=pose1[:,3:21*3+3], left_hand_pose=pose1[:,25*3:40*3], right_hand_pose=pose1[:,40*3:55*3], + leye_pose=pose1[:, 69:72], + reye_pose=pose1[:, 72:75],return_verts=True) + vertices1_all = output1["vertices"].cpu().detach().numpy() + if args.debug: + seconds = 1 + else: + seconds = vertices_all.shape[0]//30 + silent_video_file_path = utils.fast_render.generate_silent_videos(args.render_video_fps, + args.render_video_width, + args.render_video_height, + args.render_concurrent_num, + args.render_tmp_img_filetype, + int(seconds*args.render_video_fps), + vertices_all, + vertices1_all, + faces, + output_dir) + base_filename_without_ext = os.path.splitext(os.path.basename(res_npz_path))[0] + final_clip = os.path.join(output_dir, f"{base_filename_without_ext}.mp4") + utils.media.add_audio_to_video(silent_video_file_path, audio_path, final_clip) + os.remove(silent_video_file_path) + return final_clip + +def render_one_sequence_no_gt( + res_npz_path, + output_dir, + audio_path, + model_folder="/data/datasets/smplx_models/", + model_type='smplx', + gender='NEUTRAL_2020', + ext='npz', + num_betas=300, + num_expression_coeffs=100, + use_face_contour=False, + use_matplotlib=False, + args=None): + import smplx + import matplotlib.pyplot as plt + import imageio + from tqdm import tqdm + import os + import numpy as np + import torch + import moviepy.editor as mp + import librosa + + model = smplx.create(model_folder, model_type=model_type, + gender=gender, use_face_contour=use_face_contour, + num_betas=num_betas, + num_expression_coeffs=num_expression_coeffs, + ext=ext, use_pca=False).cuda() + + #data_npz = np.load(f"{output_dir}{res_npz_path}.npz") + data_np_body = np.load(res_npz_path, allow_pickle=True) + # gt_np_body = np.load(gt_npz_path, allow_pickle=True) + + if not os.path.exists(output_dir): os.makedirs(output_dir) + # if not use_matplotlib: + # import trimesh + #import pyrender + #''' + #display = Display(visible=0, size=(1000, 1000)) + #display.start() + faces = np.load(f"{model_folder}/smplx/SMPLX_NEUTRAL_2020.npz", allow_pickle=True)["f"] + seconds = 1 + #data_npz["jaw_pose"].shape[0] + n = data_np_body["poses"].shape[0] + beta = torch.from_numpy(data_np_body["betas"]).to(torch.float32).unsqueeze(0).cuda() + beta = beta.repeat(n, 1) + expression = torch.from_numpy(data_np_body["expressions"][:n]).to(torch.float32).cuda() + jaw_pose = torch.from_numpy(data_np_body["poses"][:n, 66:69]).to(torch.float32).cuda() + pose = torch.from_numpy(data_np_body["poses"][:n]).to(torch.float32).cuda() + transl = torch.from_numpy(data_np_body["trans"][:n]).to(torch.float32).cuda() + # print(beta.shape, expression.shape, jaw_pose.shape, pose.shape, transl.shape, pose[:,:3].shape) + output = model(betas=beta, transl=transl, expression=expression, jaw_pose=jaw_pose, + global_orient=pose[:,:3], body_pose=pose[:,3:21*3+3], left_hand_pose=pose[:,25*3:40*3], right_hand_pose=pose[:,40*3:55*3], + leye_pose=pose[:, 69:72], + reye_pose=pose[:, 72:75], + return_verts=True) + vertices_all = output["vertices"].cpu().detach().numpy() + + # beta1 = torch.from_numpy(gt_np_body["betas"]).to(torch.float32).unsqueeze(0).cuda() + # expression1 = torch.from_numpy(gt_np_body["expressions"][:n]).to(torch.float32).cuda() + # jaw_pose1 = torch.from_numpy(gt_np_body["poses"][:n,66:69]).to(torch.float32).cuda() + # pose1 = torch.from_numpy(gt_np_body["poses"][:n]).to(torch.float32).cuda() + # transl1 = torch.from_numpy(gt_np_body["trans"][:n]).to(torch.float32).cuda() + # output1 = model(betas=beta1, transl=transl1, expression=expression1, jaw_pose=jaw_pose1, global_orient=pose1[:,:3], body_pose=pose1[:,3:21*3+3], left_hand_pose=pose1[:,25*3:40*3], right_hand_pose=pose1[:,40*3:55*3], + # leye_pose=pose1[:, 69:72], + # reye_pose=pose1[:, 72:75],return_verts=True) + # vertices1_all = output1["vertices"].cpu().detach().numpy() + if args.debug: + seconds = 1 + else: + seconds = vertices_all.shape[0]//30 + silent_video_file_path = utils.fast_render.generate_silent_videos_no_gt(args.render_video_fps, + args.render_video_width, + args.render_video_height, + args.render_concurrent_num, + args.render_tmp_img_filetype, + int(seconds*args.render_video_fps), + vertices_all, + faces, + output_dir) + base_filename_without_ext = os.path.splitext(os.path.basename(res_npz_path))[0] + final_clip = os.path.join(output_dir, f"{base_filename_without_ext}.mp4") + utils.media.add_audio_to_video(silent_video_file_path, audio_path, final_clip) + os.remove(silent_video_file_path) + return final_clip + +def print_exp_info(args): + logger.info(pprint.pformat(vars(args))) + logger.info(f"# ------------ {args.name} ----------- #") + logger.info("PyTorch version: {}".format(torch.__version__)) + logger.info("CUDA version: {}".format(torch.version.cuda)) + logger.info("{} GPUs".format(torch.cuda.device_count())) + logger.info(f"Random Seed: {args.random_seed}") + +def args2csv(args, get_head=False, list4print=[]): + for k, v in args.items(): + if isinstance(args[k], dict): + args2csv(args[k], get_head, list4print) + else: list4print.append(k) if get_head else list4print.append(v) + return list4print + +class EpochTracker: + def __init__(self, metric_names, metric_directions): + assert len(metric_names) == len(metric_directions), "Metric names and directions should have the same length" + + + self.metric_names = metric_names + self.states = ['train', 'val', 'test'] + self.types = ['last', 'best'] + + + self.values = {name: {state: {type_: {'value': np.inf if not is_higher_better else -np.inf, 'epoch': 0} + for type_ in self.types} + for state in self.states} + for name, is_higher_better in zip(metric_names, metric_directions)} + + self.loss_meters = {name: {state: AverageMeter(f"{name}_{state}") + for state in self.states} + for name in metric_names} + + + self.is_higher_better = {name: direction for name, direction in zip(metric_names, metric_directions)} + self.train_history = {name: [] for name in metric_names} + self.val_history = {name: [] for name in metric_names} + + + def update_meter(self, name, state, value): + self.loss_meters[name][state].update(value) + + + def update_values(self, name, state, epoch): + value_avg = self.loss_meters[name][state].avg + new_best = False + + + if ((value_avg < self.values[name][state]['best']['value'] and not self.is_higher_better[name]) or + (value_avg > self.values[name][state]['best']['value'] and self.is_higher_better[name])): + self.values[name][state]['best']['value'] = value_avg + self.values[name][state]['best']['epoch'] = epoch + new_best = True + self.values[name][state]['last']['value'] = value_avg + self.values[name][state]['last']['epoch'] = epoch + return new_best + + + def get(self, name, state, type_): + return self.values[name][state][type_] + + + def reset(self): + for name in self.metric_names: + for state in self.states: + self.loss_meters[name][state].reset() + + + def flatten_values(self): + flat_dict = {} + for name in self.metric_names: + for state in self.states: + for type_ in self.types: + value_key = f"{name}_{state}_{type_}" + epoch_key = f"{name}_{state}_{type_}_epoch" + flat_dict[value_key] = self.values[name][state][type_]['value'] + flat_dict[epoch_key] = self.values[name][state][type_]['epoch'] + return flat_dict + + def update_and_plot(self, name, epoch, save_path): + new_best_train = self.update_values(name, 'train', epoch) + new_best_val = self.update_values(name, 'val', epoch) + + + self.train_history[name].append(self.loss_meters[name]['train'].avg) + self.val_history[name].append(self.loss_meters[name]['val'].avg) + + + train_values = self.train_history[name] + val_values = self.val_history[name] + epochs = list(range(1, len(train_values) + 1)) + + + plt.figure(figsize=(10, 6)) + plt.plot(epochs, train_values, label='Train') + plt.plot(epochs, val_values, label='Val') + plt.title(f'Train vs Val {name} over epochs') + plt.xlabel('Epochs') + plt.ylabel(name) + plt.legend() + plt.savefig(save_path) + plt.close() + + + return new_best_train, new_best_val + +def record_trial(args, tracker): + """ + 1. record notes, score, env_name, experments_path, + """ + csv_path = args.out_path + "custom/" +args.csv_name+".csv" + all_print_dict = vars(args) + all_print_dict.update(tracker.flatten_values()) + if not os.path.exists(csv_path): + pd.DataFrame([all_print_dict]).to_csv(csv_path, index=False) + else: + df_existing = pd.read_csv(csv_path) + df_new = pd.DataFrame([all_print_dict]) + df_aligned = df_existing.append(df_new).fillna("") + df_aligned.to_csv(csv_path, index=False) + +def set_random_seed(args): + os.environ['PYTHONHASHSEED'] = str(args.random_seed) + random.seed(args.random_seed) + np.random.seed(args.random_seed) + torch.manual_seed(args.random_seed) + torch.cuda.manual_seed_all(args.random_seed) + torch.cuda.manual_seed(args.random_seed) + torch.backends.cudnn.deterministic = args.deterministic #args.CUDNN_DETERMINISTIC + torch.backends.cudnn.benchmark = args.benchmark + torch.backends.cudnn.enabled = args.cudnn_enabled + +def save_checkpoints(save_path, model, opt=None, epoch=None, lrs=None): + if lrs is not None: + states = { 'model_state': model.state_dict(), + 'epoch': epoch + 1, + 'opt_state': opt.state_dict(), + 'lrs':lrs.state_dict(),} + elif opt is not None: + states = { 'model_state': model.state_dict(), + 'epoch': epoch + 1, + 'opt_state': opt.state_dict(),} + else: + states = { 'model_state': model.state_dict(),} + torch.save(states, save_path) + +def load_checkpoints(model, save_path, load_name='model'): + states = torch.load(save_path) + new_weights = OrderedDict() + flag=False + for k, v in states['model_state'].items(): + #print(k) + if "module" not in k: + break + else: + new_weights[k[7:]]=v + flag=True + if flag: + try: + model.load_state_dict(new_weights) + except: + #print(states['model_state']) + model.load_state_dict(states['model_state']) + else: + model.load_state_dict(states['model_state']) + logger.info(f"load self-pretrained checkpoints for {load_name}") + +def model_complexity(model, args): + from ptflops import get_model_complexity_info + flops, params = get_model_complexity_info(model, (args.T_GLOBAL._DIM, args.TRAIN.CROP, args.TRAIN), + as_strings=False, print_per_layer_stat=False) + logging.info('{:<30} {:<8} BFlops'.format('Computational complexity: ', flops / 1e9)) + logging.info('{:<30} {:<8} MParams'.format('Number of parameters: ', params / 1e6)) + +class AverageMeter(object): + """Computes and stores the average and current value""" + def __init__(self, name, fmt=':f'): + self.name = name + self.fmt = fmt + self.reset() + + def reset(self): + self.val = 0 + self.avg = 0 + self.sum = 0 + self.count = 0 + + def update(self, val, n=1): + self.val = val + self.sum += val * n + self.count += n + self.avg = self.sum / self.count + + def __str__(self): + fmtstr = '{name} {val' + self.fmt + '} ({avg' + self.fmt + '})' + return fmtstr.format(**self.__dict__) \ No newline at end of file diff --git a/utils/rotation_conversions.py b/utils/rotation_conversions.py new file mode 100644 index 0000000000000000000000000000000000000000..f2bfaa1b2247622bff35d3f9b15e8eb84064aa53 --- /dev/null +++ b/utils/rotation_conversions.py @@ -0,0 +1,550 @@ +# Copyright (c) Facebook, Inc. and its affiliates. All rights reserved. + +import functools +from typing import Optional + +import torch +import torch.nn.functional as F + + +""" +The transformation matrices returned from the functions in this file assume +the points on which the transformation will be applied are column vectors. +i.e. the R matrix is structured as + + R = [ + [Rxx, Rxy, Rxz], + [Ryx, Ryy, Ryz], + [Rzx, Rzy, Rzz], + ] # (3, 3) + +This matrix can be applied to column vectors by post multiplication +by the points e.g. + + points = [[0], [1], [2]] # (3 x 1) xyz coordinates of a point + transformed_points = R * points + +To apply the same matrix to points which are row vectors, the R matrix +can be transposed and pre multiplied by the points: + +e.g. + points = [[0, 1, 2]] # (1 x 3) xyz coordinates of a point + transformed_points = points * R.transpose(1, 0) +""" + + +def quaternion_to_matrix(quaternions): + """ + Convert rotations given as quaternions to rotation matrices. + + Args: + quaternions: quaternions with real part first, + as tensor of shape (..., 4). + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + r, i, j, k = torch.unbind(quaternions, -1) + two_s = 2.0 / (quaternions * quaternions).sum(-1) + + o = torch.stack( + ( + 1 - two_s * (j * j + k * k), + two_s * (i * j - k * r), + two_s * (i * k + j * r), + two_s * (i * j + k * r), + 1 - two_s * (i * i + k * k), + two_s * (j * k - i * r), + two_s * (i * k - j * r), + two_s * (j * k + i * r), + 1 - two_s * (i * i + j * j), + ), + -1, + ) + return o.reshape(quaternions.shape[:-1] + (3, 3)) + + +def _copysign(a, b): + """ + Return a tensor where each element has the absolute value taken from the, + corresponding element of a, with sign taken from the corresponding + element of b. This is like the standard copysign floating-point operation, + but is not careful about negative 0 and NaN. + + Args: + a: source tensor. + b: tensor whose signs will be used, of the same shape as a. + + Returns: + Tensor of the same shape as a with the signs of b. + """ + signs_differ = (a < 0) != (b < 0) + return torch.where(signs_differ, -a, a) + + +def _sqrt_positive_part(x): + """ + Returns torch.sqrt(torch.max(0, x)) + but with a zero subgradient where x is 0. + """ + ret = torch.zeros_like(x) + positive_mask = x > 0 + ret[positive_mask] = torch.sqrt(x[positive_mask]) + return ret + + +def matrix_to_quaternion(matrix): + """ + Convert rotations given as rotation matrices to quaternions. + + Args: + matrix: Rotation matrices as tensor of shape (..., 3, 3). + + Returns: + quaternions with real part first, as tensor of shape (..., 4). + """ + if matrix.size(-1) != 3 or matrix.size(-2) != 3: + raise ValueError(f"Invalid rotation matrix shape f{matrix.shape}.") + m00 = matrix[..., 0, 0] + m11 = matrix[..., 1, 1] + m22 = matrix[..., 2, 2] + o0 = 0.5 * _sqrt_positive_part(1 + m00 + m11 + m22) + x = 0.5 * _sqrt_positive_part(1 + m00 - m11 - m22) + y = 0.5 * _sqrt_positive_part(1 - m00 + m11 - m22) + z = 0.5 * _sqrt_positive_part(1 - m00 - m11 + m22) + o1 = _copysign(x, matrix[..., 2, 1] - matrix[..., 1, 2]) + o2 = _copysign(y, matrix[..., 0, 2] - matrix[..., 2, 0]) + o3 = _copysign(z, matrix[..., 1, 0] - matrix[..., 0, 1]) + return torch.stack((o0, o1, o2, o3), -1) + + +def _axis_angle_rotation(axis: str, angle): + """ + Return the rotation matrices for one of the rotations about an axis + of which Euler angles describe, for each value of the angle given. + + Args: + axis: Axis label "X" or "Y or "Z". + angle: any shape tensor of Euler angles in radians + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + + cos = torch.cos(angle) + sin = torch.sin(angle) + one = torch.ones_like(angle) + zero = torch.zeros_like(angle) + + if axis == "X": + R_flat = (one, zero, zero, zero, cos, -sin, zero, sin, cos) + if axis == "Y": + R_flat = (cos, zero, sin, zero, one, zero, -sin, zero, cos) + if axis == "Z": + R_flat = (cos, -sin, zero, sin, cos, zero, zero, zero, one) + + return torch.stack(R_flat, -1).reshape(angle.shape + (3, 3)) + + +def euler_angles_to_matrix(euler_angles, convention: str): + """ + Convert rotations given as Euler angles in radians to rotation matrices. + + Args: + euler_angles: Euler angles in radians as tensor of shape (..., 3). + convention: Convention string of three uppercase letters from + {"X", "Y", and "Z"}. + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + if euler_angles.dim() == 0 or euler_angles.shape[-1] != 3: + raise ValueError("Invalid input euler angles.") + if len(convention) != 3: + raise ValueError("Convention must have 3 letters.") + if convention[1] in (convention[0], convention[2]): + raise ValueError(f"Invalid convention {convention}.") + for letter in convention: + if letter not in ("X", "Y", "Z"): + raise ValueError(f"Invalid letter {letter} in convention string.") + matrices = map(_axis_angle_rotation, convention, torch.unbind(euler_angles, -1)) + return functools.reduce(torch.matmul, matrices) + + +def _angle_from_tan( + axis: str, other_axis: str, data, horizontal: bool, tait_bryan: bool +): + """ + Extract the first or third Euler angle from the two members of + the matrix which are positive constant times its sine and cosine. + + Args: + axis: Axis label "X" or "Y or "Z" for the angle we are finding. + other_axis: Axis label "X" or "Y or "Z" for the middle axis in the + convention. + data: Rotation matrices as tensor of shape (..., 3, 3). + horizontal: Whether we are looking for the angle for the third axis, + which means the relevant entries are in the same row of the + rotation matrix. If not, they are in the same column. + tait_bryan: Whether the first and third axes in the convention differ. + + Returns: + Euler Angles in radians for each matrix in data as a tensor + of shape (...). + """ + + i1, i2 = {"X": (2, 1), "Y": (0, 2), "Z": (1, 0)}[axis] + if horizontal: + i2, i1 = i1, i2 + even = (axis + other_axis) in ["XY", "YZ", "ZX"] + if horizontal == even: + return torch.atan2(data[..., i1], data[..., i2]) + if tait_bryan: + return torch.atan2(-data[..., i2], data[..., i1]) + return torch.atan2(data[..., i2], -data[..., i1]) + + +def _index_from_letter(letter: str): + if letter == "X": + return 0 + if letter == "Y": + return 1 + if letter == "Z": + return 2 + + +def matrix_to_euler_angles(matrix, convention: str): + """ + Convert rotations given as rotation matrices to Euler angles in radians. + + Args: + matrix: Rotation matrices as tensor of shape (..., 3, 3). + convention: Convention string of three uppercase letters. + + Returns: + Euler angles in radians as tensor of shape (..., 3). + """ + if len(convention) != 3: + raise ValueError("Convention must have 3 letters.") + if convention[1] in (convention[0], convention[2]): + raise ValueError(f"Invalid convention {convention}.") + for letter in convention: + if letter not in ("X", "Y", "Z"): + raise ValueError(f"Invalid letter {letter} in convention string.") + if matrix.size(-1) != 3 or matrix.size(-2) != 3: + raise ValueError(f"Invalid rotation matrix shape f{matrix.shape}.") + i0 = _index_from_letter(convention[0]) + i2 = _index_from_letter(convention[2]) + tait_bryan = i0 != i2 + if tait_bryan: + central_angle = torch.asin( + matrix[..., i0, i2] * (-1.0 if i0 - i2 in [-1, 2] else 1.0) + ) + else: + central_angle = torch.acos(matrix[..., i0, i0]) + + o = ( + _angle_from_tan( + convention[0], convention[1], matrix[..., i2], False, tait_bryan + ), + central_angle, + _angle_from_tan( + convention[2], convention[1], matrix[..., i0, :], True, tait_bryan + ), + ) + return torch.stack(o, -1) + + +def random_quaternions( + n: int, dtype: Optional[torch.dtype] = None, device=None, requires_grad=False +): + """ + Generate random quaternions representing rotations, + i.e. versors with nonnegative real part. + + Args: + n: Number of quaternions in a batch to return. + dtype: Type to return. + device: Desired device of returned tensor. Default: + uses the current device for the default tensor type. + requires_grad: Whether the resulting tensor should have the gradient + flag set. + + Returns: + Quaternions as tensor of shape (N, 4). + """ + o = torch.randn((n, 4), dtype=dtype, device=device, requires_grad=requires_grad) + s = (o * o).sum(1) + o = o / _copysign(torch.sqrt(s), o[:, 0])[:, None] + return o + + +def random_rotations( + n: int, dtype: Optional[torch.dtype] = None, device=None, requires_grad=False +): + """ + Generate random rotations as 3x3 rotation matrices. + + Args: + n: Number of rotation matrices in a batch to return. + dtype: Type to return. + device: Device of returned tensor. Default: if None, + uses the current device for the default tensor type. + requires_grad: Whether the resulting tensor should have the gradient + flag set. + + Returns: + Rotation matrices as tensor of shape (n, 3, 3). + """ + quaternions = random_quaternions( + n, dtype=dtype, device=device, requires_grad=requires_grad + ) + return quaternion_to_matrix(quaternions) + + +def random_rotation( + dtype: Optional[torch.dtype] = None, device=None, requires_grad=False +): + """ + Generate a single random 3x3 rotation matrix. + + Args: + dtype: Type to return + device: Device of returned tensor. Default: if None, + uses the current device for the default tensor type + requires_grad: Whether the resulting tensor should have the gradient + flag set + + Returns: + Rotation matrix as tensor of shape (3, 3). + """ + return random_rotations(1, dtype, device, requires_grad)[0] + + +def standardize_quaternion(quaternions): + """ + Convert a unit quaternion to a standard form: one in which the real + part is non negative. + + Args: + quaternions: Quaternions with real part first, + as tensor of shape (..., 4). + + Returns: + Standardized quaternions as tensor of shape (..., 4). + """ + return torch.where(quaternions[..., 0:1] < 0, -quaternions, quaternions) + + +def quaternion_raw_multiply(a, b): + """ + Multiply two quaternions. + Usual torch rules for broadcasting apply. + + Args: + a: Quaternions as tensor of shape (..., 4), real part first. + b: Quaternions as tensor of shape (..., 4), real part first. + + Returns: + The product of a and b, a tensor of quaternions shape (..., 4). + """ + aw, ax, ay, az = torch.unbind(a, -1) + bw, bx, by, bz = torch.unbind(b, -1) + ow = aw * bw - ax * bx - ay * by - az * bz + ox = aw * bx + ax * bw + ay * bz - az * by + oy = aw * by - ax * bz + ay * bw + az * bx + oz = aw * bz + ax * by - ay * bx + az * bw + return torch.stack((ow, ox, oy, oz), -1) + + +def quaternion_multiply(a, b): + """ + Multiply two quaternions representing rotations, returning the quaternion + representing their composition, i.e. the versor with nonnegative real part. + Usual torch rules for broadcasting apply. + + Args: + a: Quaternions as tensor of shape (..., 4), real part first. + b: Quaternions as tensor of shape (..., 4), real part first. + + Returns: + The product of a and b, a tensor of quaternions of shape (..., 4). + """ + ab = quaternion_raw_multiply(a, b) + return standardize_quaternion(ab) + + +def quaternion_invert(quaternion): + """ + Given a quaternion representing rotation, get the quaternion representing + its inverse. + + Args: + quaternion: Quaternions as tensor of shape (..., 4), with real part + first, which must be versors (unit quaternions). + + Returns: + The inverse, a tensor of quaternions of shape (..., 4). + """ + + return quaternion * quaternion.new_tensor([1, -1, -1, -1]) + + +def quaternion_apply(quaternion, point): + """ + Apply the rotation given by a quaternion to a 3D point. + Usual torch rules for broadcasting apply. + + Args: + quaternion: Tensor of quaternions, real part first, of shape (..., 4). + point: Tensor of 3D points of shape (..., 3). + + Returns: + Tensor of rotated points of shape (..., 3). + """ + if point.size(-1) != 3: + raise ValueError(f"Points are not in 3D, f{point.shape}.") + real_parts = point.new_zeros(point.shape[:-1] + (1,)) + point_as_quaternion = torch.cat((real_parts, point), -1) + out = quaternion_raw_multiply( + quaternion_raw_multiply(quaternion, point_as_quaternion), + quaternion_invert(quaternion), + ) + return out[..., 1:] + + +def axis_angle_to_matrix(axis_angle): + """ + Convert rotations given as axis/angle to rotation matrices. + + Args: + axis_angle: Rotations given as a vector in axis angle form, + as a tensor of shape (..., 3), where the magnitude is + the angle turned anticlockwise in radians around the + vector's direction. + + Returns: + Rotation matrices as tensor of shape (..., 3, 3). + """ + return quaternion_to_matrix(axis_angle_to_quaternion(axis_angle)) + + +def matrix_to_axis_angle(matrix): + """ + Convert rotations given as rotation matrices to axis/angle. + + Args: + matrix: Rotation matrices as tensor of shape (..., 3, 3). + + Returns: + Rotations given as a vector in axis angle form, as a tensor + of shape (..., 3), where the magnitude is the angle + turned anticlockwise in radians around the vector's + direction. + """ + return quaternion_to_axis_angle(matrix_to_quaternion(matrix)) + + +def axis_angle_to_quaternion(axis_angle): + """ + Convert rotations given as axis/angle to quaternions. + + Args: + axis_angle: Rotations given as a vector in axis angle form, + as a tensor of shape (..., 3), where the magnitude is + the angle turned anticlockwise in radians around the + vector's direction. + + Returns: + quaternions with real part first, as tensor of shape (..., 4). + """ + angles = torch.norm(axis_angle, p=2, dim=-1, keepdim=True) + half_angles = 0.5 * angles + eps = 1e-6 + small_angles = angles.abs() < eps + sin_half_angles_over_angles = torch.empty_like(angles) + sin_half_angles_over_angles[~small_angles] = ( + torch.sin(half_angles[~small_angles]) / angles[~small_angles] + ) + # for x small, sin(x/2) is about x/2 - (x/2)^3/6 + # so sin(x/2)/x is about 1/2 - (x*x)/48 + sin_half_angles_over_angles[small_angles] = ( + 0.5 - (angles[small_angles] * angles[small_angles]) / 48 + ) + quaternions = torch.cat( + [torch.cos(half_angles), axis_angle * sin_half_angles_over_angles], dim=-1 + ) + return quaternions + + +def quaternion_to_axis_angle(quaternions): + """ + Convert rotations given as quaternions to axis/angle. + + Args: + quaternions: quaternions with real part first, + as tensor of shape (..., 4). + + Returns: + Rotations given as a vector in axis angle form, as a tensor + of shape (..., 3), where the magnitude is the angle + turned anticlockwise in radians around the vector's + direction. + """ + norms = torch.norm(quaternions[..., 1:], p=2, dim=-1, keepdim=True) + half_angles = torch.atan2(norms, quaternions[..., :1]) + angles = 2 * half_angles + eps = 1e-6 + small_angles = angles.abs() < eps + sin_half_angles_over_angles = torch.empty_like(angles) + sin_half_angles_over_angles[~small_angles] = ( + torch.sin(half_angles[~small_angles]) / angles[~small_angles] + ) + # for x small, sin(x/2) is about x/2 - (x/2)^3/6 + # so sin(x/2)/x is about 1/2 - (x*x)/48 + sin_half_angles_over_angles[small_angles] = ( + 0.5 - (angles[small_angles] * angles[small_angles]) / 48 + ) + return quaternions[..., 1:] / sin_half_angles_over_angles + + +def rotation_6d_to_matrix(d6: torch.Tensor) -> torch.Tensor: + """ + Converts 6D rotation representation by Zhou et al. [1] to rotation matrix + using Gram--Schmidt orthogonalisation per Section B of [1]. + Args: + d6: 6D rotation representation, of size (*, 6) + + Returns: + batch of rotation matrices of size (*, 3, 3) + + [1] Zhou, Y., Barnes, C., Lu, J., Yang, J., & Li, H. + On the Continuity of Rotation Representations in Neural Networks. + IEEE Conference on Computer Vision and Pattern Recognition, 2019. + Retrieved from http://arxiv.org/abs/1812.07035 + """ + + a1, a2 = d6[..., :3], d6[..., 3:] + b1 = F.normalize(a1, dim=-1) + b2 = a2 - (b1 * a2).sum(-1, keepdim=True) * b1 + b2 = F.normalize(b2, dim=-1) + b3 = torch.cross(b1, b2, dim=-1) + return torch.stack((b1, b2, b3), dim=-2) + + +def matrix_to_rotation_6d(matrix: torch.Tensor) -> torch.Tensor: + """ + Converts rotation matrices to 6D rotation representation by Zhou et al. [1] + by dropping the last row. Note that 6D representation is not unique. + Args: + matrix: batch of rotation matrices of size (*, 3, 3) + + Returns: + 6D rotation representation, of size (*, 6) + + [1] Zhou, Y., Barnes, C., Lu, J., Yang, J., & Li, H. + On the Continuity of Rotation Representations in Neural Networks. + IEEE Conference on Computer Vision and Pattern Recognition, 2019. + Retrieved from http://arxiv.org/abs/1812.07035 + """ + return matrix[..., :2, :].clone().reshape(*matrix.size()[:-2], 6)