Namespace(aa='rand-m9-mstd0.5-inc1', auto_resume=True, batch_size=512, cfg_path='iFormer_l.yaml', clip_grad=None, color_jitter=0.4, crop_pct=None, cutmix=1.0, cutmix_minmax=None, data_path='imagenet', data_set='IMNET', device='cuda', disable_eval=False, dist_backend='nccl', dist_eval=True, dist_on_itp=False, dist_url='env://', distillation_alpha=0.5, distillation_tau=1.0, distillation_type='none', distributed=True, drop_path=0.1, enable_wandb=False, epochs=300, eval=False, eval_data_path=None, finetune='', gpu=0, head_init_scale=1.0, imagenet_default_mean_and_std=True, input_size=224, layer_decay=1.0, layer_scale_init_value=0, local_rank=-1, log_dir=None, lr=0.008, min_lr=1e-06, mixup=0.8, mixup_mode='batch', mixup_prob=1.0, mixup_switch_prob=0.5, model='iFormer_l', model_ema=True, model_ema_decay=0.9999, model_ema_eval=True, model_ema_force_cpu=False, model_key='model|module', model_prefix='', momentum=0.9, nb_classes=1000, num_workers=16, opt='adamw', opt_betas=None, opt_eps=1e-08, output_dir='', pin_mem=True, project='iFormer', rank=0, recount=1, remode='pixel', reprob=0.25, resplit=False, resume='', save_ckpt=True, save_ckpt_freq=1, save_ckpt_num=3, seed=0, smoothing=0.1, start_epoch=0, teacher_model='regnety_160', teacher_path='regnety_160-a5fe301d.pth', train_interpolation='bicubic', update_freq=1, use_amp=False, wandb_ckpt=False, warmup_epochs=20, warmup_steps=-1, weight_decay=0.05, weight_decay_end=None, world_size=16)
Transform = 
RandomResizedCropAndInterpolation(size=(224, 224), scale=(0.08, 1.0), ratio=(0.75, 1.3333), interpolation=PIL.Image.BICUBIC)
RandomHorizontalFlip(p=0.5)
<timm.data.auto_augment.RandAugment object at 0x7f73da1fcdf0>
ToTensor()
Normalize(mean=tensor([0.4850, 0.4560, 0.4060]), std=tensor([0.2290, 0.2240, 0.2250]))
<timm.data.random_erasing.RandomErasing object at 0x7f73da1dca90>
---------------------------
reading from datapath imagenet
Number of the class = 1000
Transform = 
Resize(size=256, interpolation=bicubic, max_size=None, antialias=warn)
CenterCrop(size=(224, 224))
ToTensor()
Normalize(mean=(0.485, 0.456, 0.406), std=(0.229, 0.224, 0.225))
---------------------------
reading from datapath imagenet
Number of the class = 1000
Sampler_train = <torch.utils.data.distributed.DistributedSampler object at 0x7f73da1dcdf0>
Mixup is activated!
Using EMA with decay = 0.99990000
Model = iFormer(
  (downsample_layers): ModuleList(
    (0): Sequential(
      (0): Conv2d_BN(
        (c): Conv2d(3, 24, kernel_size=(5, 5), stride=(2, 2), padding=(2, 2), bias=False)
        (bn): BatchNorm2d(24, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      )
      (1): GELU(approximate='none')
      (2): EdgeResidual(
        (conv_exp_bn1): Conv2d_BN(
          (c): Conv2d(24, 96, kernel_size=(5, 5), stride=(2, 2), padding=(2, 2), bias=False)
          (bn): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        )
        (act): GELU(approximate='none')
        (conv_pwl_bn2): Conv2d_BN(
          (c): Conv2d(96, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
          (bn): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        )
      )
    )
    (1): Sequential(
      (0): Conv2d_BN(
        (c): Conv2d(48, 96, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
        (bn): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      )
    )
    (2): Sequential(
      (0): Conv2d_BN(
        (c): Conv2d(96, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
        (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      )
    )
    (3): Sequential(
      (0): Conv2d_BN(
        (c): Conv2d(256, 384, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
        (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      )
    )
  )
  (stages): ModuleList(
    (0): Sequential(
      (0): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(48, 48, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=48, bias=False)
                (bn): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(48, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(192, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): Identity()
          )
        )
      )
      (1): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(48, 48, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=48, bias=False)
                (bn): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(48, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(192, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
    )
    (1): Sequential(
      (0): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(96, 96, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=96, bias=False)
                (bn): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(96, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (1): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(96, 96, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=96, bias=False)
                (bn): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(96, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
    )
    (2): Sequential(
      (0): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(256, 256, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=256, bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (1): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(256, 256, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=256, bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (2): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(256, 256, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=256, bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (3): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(256, 256, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=256, bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (4): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(256, 256, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=256, bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (5): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(256, 256, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=256, bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (6): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(256, 256, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=256, bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (7): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(256, 256, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=256, bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (8): BasicBlock(
        (block): RepCPE(
          (cpe): Residual(
            (m): Conv2d_BN(
              (c): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=256, bias=False)
              (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
            (drop_path): Identity()
          )
        )
      )
      (9): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual(
            (m): SHMA(
              (q): Conv2d_BN(
                (c): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (k): Conv2d_BN(
                (c): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (gate_act): Sigmoid()
              (attn_drop): Dropout(p=0.0, inplace=False)
              (proj): Conv2d_BN(
                (c): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (v_gate): Conv2d_BN(
                (c): Conv2d(256, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (10): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(256, 768, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(768, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (11): BasicBlock(
        (block): RepCPE(
          (cpe): Residual(
            (m): Conv2d_BN(
              (c): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=256, bias=False)
              (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
            (drop_path): Identity()
          )
        )
      )
      (12): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual(
            (m): SHMA(
              (q): Conv2d_BN(
                (c): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (k): Conv2d_BN(
                (c): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (gate_act): Sigmoid()
              (attn_drop): Dropout(p=0.0, inplace=False)
              (proj): Conv2d_BN(
                (c): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (v_gate): Conv2d_BN(
                (c): Conv2d(256, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (13): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(256, 768, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(768, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (14): BasicBlock(
        (block): RepCPE(
          (cpe): Residual(
            (m): Conv2d_BN(
              (c): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=256, bias=False)
              (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
            (drop_path): Identity()
          )
        )
      )
      (15): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual(
            (m): SHMA(
              (q): Conv2d_BN(
                (c): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (k): Conv2d_BN(
                (c): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (gate_act): Sigmoid()
              (attn_drop): Dropout(p=0.0, inplace=False)
              (proj): Conv2d_BN(
                (c): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (v_gate): Conv2d_BN(
                (c): Conv2d(256, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (16): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(256, 768, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(768, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (17): BasicBlock(
        (block): RepCPE(
          (cpe): Residual(
            (m): Conv2d_BN(
              (c): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=256, bias=False)
              (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
            (drop_path): Identity()
          )
        )
      )
      (18): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual(
            (m): SHMA(
              (q): Conv2d_BN(
                (c): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (k): Conv2d_BN(
                (c): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (gate_act): Sigmoid()
              (attn_drop): Dropout(p=0.0, inplace=False)
              (proj): Conv2d_BN(
                (c): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (v_gate): Conv2d_BN(
                (c): Conv2d(256, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (19): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(256, 768, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(768, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (20): BasicBlock(
        (block): RepCPE(
          (cpe): Residual(
            (m): Conv2d_BN(
              (c): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=256, bias=False)
              (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
            (drop_path): Identity()
          )
        )
      )
      (21): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual(
            (m): SHMA(
              (q): Conv2d_BN(
                (c): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (k): Conv2d_BN(
                (c): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (gate_act): Sigmoid()
              (attn_drop): Dropout(p=0.0, inplace=False)
              (proj): Conv2d_BN(
                (c): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (v_gate): Conv2d_BN(
                (c): Conv2d(256, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (22): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(256, 768, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(768, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (23): BasicBlock(
        (block): RepCPE(
          (cpe): Residual(
            (m): Conv2d_BN(
              (c): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=256, bias=False)
              (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
            (drop_path): Identity()
          )
        )
      )
      (24): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual(
            (m): SHMA(
              (q): Conv2d_BN(
                (c): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (k): Conv2d_BN(
                (c): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (gate_act): Sigmoid()
              (attn_drop): Dropout(p=0.0, inplace=False)
              (proj): Conv2d_BN(
                (c): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (v_gate): Conv2d_BN(
                (c): Conv2d(256, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (25): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(256, 768, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(768, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (26): BasicBlock(
        (block): RepCPE(
          (cpe): Residual(
            (m): Conv2d_BN(
              (c): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=256, bias=False)
              (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
            (drop_path): Identity()
          )
        )
      )
      (27): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual(
            (m): SHMA(
              (q): Conv2d_BN(
                (c): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (k): Conv2d_BN(
                (c): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (gate_act): Sigmoid()
              (attn_drop): Dropout(p=0.0, inplace=False)
              (proj): Conv2d_BN(
                (c): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (v_gate): Conv2d_BN(
                (c): Conv2d(256, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (28): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(256, 768, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(768, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (29): BasicBlock(
        (block): RepCPE(
          (cpe): Residual(
            (m): Conv2d_BN(
              (c): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=256, bias=False)
              (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
            (drop_path): Identity()
          )
        )
      )
      (30): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual(
            (m): SHMA(
              (q): Conv2d_BN(
                (c): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (k): Conv2d_BN(
                (c): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (gate_act): Sigmoid()
              (attn_drop): Dropout(p=0.0, inplace=False)
              (proj): Conv2d_BN(
                (c): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (v_gate): Conv2d_BN(
                (c): Conv2d(256, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (31): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(256, 768, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(768, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (32): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(256, 256, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=256, bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
    )
    (3): Sequential(
      (0): BasicBlock(
        (block): RepCPE(
          (cpe): Residual(
            (m): Conv2d_BN(
              (c): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=384, bias=False)
              (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
            (drop_path): Identity()
          )
        )
      )
      (1): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual(
            (m): SHMA(
              (q): Conv2d_BN(
                (c): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (k): Conv2d_BN(
                (c): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (gate_act): Sigmoid()
              (attn_drop): Dropout(p=0.0, inplace=False)
              (proj): Conv2d_BN(
                (c): Conv2d(384, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (v_gate): Conv2d_BN(
                (c): Conv2d(384, 768, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (2): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 1152, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1152, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(1152, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (3): BasicBlock(
        (block): RepCPE(
          (cpe): Residual(
            (m): Conv2d_BN(
              (c): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=384, bias=False)
              (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
            (drop_path): Identity()
          )
        )
      )
      (4): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual(
            (m): SHMA(
              (q): Conv2d_BN(
                (c): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (k): Conv2d_BN(
                (c): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (gate_act): Sigmoid()
              (attn_drop): Dropout(p=0.0, inplace=False)
              (proj): Conv2d_BN(
                (c): Conv2d(384, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (v_gate): Conv2d_BN(
                (c): Conv2d(384, 768, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (5): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 1152, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1152, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(1152, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
    )
  )
  (classifier): Classfier(
    (classifier): BN_Linear(
      (bn): BatchNorm1d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      (l): Linear(in_features=384, out_features=1000, bias=True)
    )
  )
)
number of params: 14779808
LR = 0.00800000
Batch size = 8192
Update frequent = 1
Number of training examples = 1281167
Number of training training per epoch = 156
Param groups = {
  "decay": {
    "weight_decay": 0.05,
    "params": [
      "downsample_layers.0.0.c.weight",
      "downsample_layers.0.2.conv_exp_bn1.c.weight",
      "downsample_layers.0.2.conv_pwl_bn2.c.weight",
      "downsample_layers.1.0.c.weight",
      "downsample_layers.2.0.c.weight",
      "downsample_layers.3.0.c.weight",
      "stages.0.0.block.token_channel_mixer.m.0.c.weight",
      "stages.0.0.block.token_channel_mixer.m.1.c.weight",
      "stages.0.0.block.token_channel_mixer.m.3.c.weight",
      "stages.0.1.block.token_channel_mixer.m.0.c.weight",
      "stages.0.1.block.token_channel_mixer.m.1.c.weight",
      "stages.0.1.block.token_channel_mixer.m.3.c.weight",
      "stages.1.0.block.token_channel_mixer.m.0.c.weight",
      "stages.1.0.block.token_channel_mixer.m.1.c.weight",
      "stages.1.0.block.token_channel_mixer.m.3.c.weight",
      "stages.1.1.block.token_channel_mixer.m.0.c.weight",
      "stages.1.1.block.token_channel_mixer.m.1.c.weight",
      "stages.1.1.block.token_channel_mixer.m.3.c.weight",
      "stages.2.0.block.token_channel_mixer.m.0.c.weight",
      "stages.2.0.block.token_channel_mixer.m.1.c.weight",
      "stages.2.0.block.token_channel_mixer.m.3.c.weight",
      "stages.2.1.block.token_channel_mixer.m.0.c.weight",
      "stages.2.1.block.token_channel_mixer.m.1.c.weight",
      "stages.2.1.block.token_channel_mixer.m.3.c.weight",
      "stages.2.2.block.token_channel_mixer.m.0.c.weight",
      "stages.2.2.block.token_channel_mixer.m.1.c.weight",
      "stages.2.2.block.token_channel_mixer.m.3.c.weight",
      "stages.2.3.block.token_channel_mixer.m.0.c.weight",
      "stages.2.3.block.token_channel_mixer.m.1.c.weight",
      "stages.2.3.block.token_channel_mixer.m.3.c.weight",
      "stages.2.4.block.token_channel_mixer.m.0.c.weight",
      "stages.2.4.block.token_channel_mixer.m.1.c.weight",
      "stages.2.4.block.token_channel_mixer.m.3.c.weight",
      "stages.2.5.block.token_channel_mixer.m.0.c.weight",
      "stages.2.5.block.token_channel_mixer.m.1.c.weight",
      "stages.2.5.block.token_channel_mixer.m.3.c.weight",
      "stages.2.6.block.token_channel_mixer.m.0.c.weight",
      "stages.2.6.block.token_channel_mixer.m.1.c.weight",
      "stages.2.6.block.token_channel_mixer.m.3.c.weight",
      "stages.2.7.block.token_channel_mixer.m.0.c.weight",
      "stages.2.7.block.token_channel_mixer.m.1.c.weight",
      "stages.2.7.block.token_channel_mixer.m.3.c.weight",
      "stages.2.8.block.cpe.m.c.weight",
      "stages.2.9.block.token_channel_mixer.m.q.c.weight",
      "stages.2.9.block.token_channel_mixer.m.k.c.weight",
      "stages.2.9.block.token_channel_mixer.m.proj.c.weight",
      "stages.2.9.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.2.10.block.channel_mixer.m.0.c.weight",
      "stages.2.10.block.channel_mixer.m.2.c.weight",
      "stages.2.11.block.cpe.m.c.weight",
      "stages.2.12.block.token_channel_mixer.m.q.c.weight",
      "stages.2.12.block.token_channel_mixer.m.k.c.weight",
      "stages.2.12.block.token_channel_mixer.m.proj.c.weight",
      "stages.2.12.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.2.13.block.channel_mixer.m.0.c.weight",
      "stages.2.13.block.channel_mixer.m.2.c.weight",
      "stages.2.14.block.cpe.m.c.weight",
      "stages.2.15.block.token_channel_mixer.m.q.c.weight",
      "stages.2.15.block.token_channel_mixer.m.k.c.weight",
      "stages.2.15.block.token_channel_mixer.m.proj.c.weight",
      "stages.2.15.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.2.16.block.channel_mixer.m.0.c.weight",
      "stages.2.16.block.channel_mixer.m.2.c.weight",
      "stages.2.17.block.cpe.m.c.weight",
      "stages.2.18.block.token_channel_mixer.m.q.c.weight",
      "stages.2.18.block.token_channel_mixer.m.k.c.weight",
      "stages.2.18.block.token_channel_mixer.m.proj.c.weight",
      "stages.2.18.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.2.19.block.channel_mixer.m.0.c.weight",
      "stages.2.19.block.channel_mixer.m.2.c.weight",
      "stages.2.20.block.cpe.m.c.weight",
      "stages.2.21.block.token_channel_mixer.m.q.c.weight",
      "stages.2.21.block.token_channel_mixer.m.k.c.weight",
      "stages.2.21.block.token_channel_mixer.m.proj.c.weight",
      "stages.2.21.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.2.22.block.channel_mixer.m.0.c.weight",
      "stages.2.22.block.channel_mixer.m.2.c.weight",
      "stages.2.23.block.cpe.m.c.weight",
      "stages.2.24.block.token_channel_mixer.m.q.c.weight",
      "stages.2.24.block.token_channel_mixer.m.k.c.weight",
      "stages.2.24.block.token_channel_mixer.m.proj.c.weight",
      "stages.2.24.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.2.25.block.channel_mixer.m.0.c.weight",
      "stages.2.25.block.channel_mixer.m.2.c.weight",
      "stages.2.26.block.cpe.m.c.weight",
      "stages.2.27.block.token_channel_mixer.m.q.c.weight",
      "stages.2.27.block.token_channel_mixer.m.k.c.weight",
      "stages.2.27.block.token_channel_mixer.m.proj.c.weight",
      "stages.2.27.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.2.28.block.channel_mixer.m.0.c.weight",
      "stages.2.28.block.channel_mixer.m.2.c.weight",
      "stages.2.29.block.cpe.m.c.weight",
      "stages.2.30.block.token_channel_mixer.m.q.c.weight",
      "stages.2.30.block.token_channel_mixer.m.k.c.weight",
      "stages.2.30.block.token_channel_mixer.m.proj.c.weight",
      "stages.2.30.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.2.31.block.channel_mixer.m.0.c.weight",
      "stages.2.31.block.channel_mixer.m.2.c.weight",
      "stages.2.32.block.token_channel_mixer.m.0.c.weight",
      "stages.2.32.block.token_channel_mixer.m.1.c.weight",
      "stages.2.32.block.token_channel_mixer.m.3.c.weight",
      "stages.3.0.block.cpe.m.c.weight",
      "stages.3.1.block.token_channel_mixer.m.q.c.weight",
      "stages.3.1.block.token_channel_mixer.m.k.c.weight",
      "stages.3.1.block.token_channel_mixer.m.proj.c.weight",
      "stages.3.1.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.3.2.block.channel_mixer.m.0.c.weight",
      "stages.3.2.block.channel_mixer.m.2.c.weight",
      "stages.3.3.block.cpe.m.c.weight",
      "stages.3.4.block.token_channel_mixer.m.q.c.weight",
      "stages.3.4.block.token_channel_mixer.m.k.c.weight",
      "stages.3.4.block.token_channel_mixer.m.proj.c.weight",
      "stages.3.4.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.3.5.block.channel_mixer.m.0.c.weight",
      "stages.3.5.block.channel_mixer.m.2.c.weight",
      "classifier.classifier.l.weight"
    ],
    "lr_scale": 1.0
  },
  "no_decay": {
    "weight_decay": 0.0,
    "params": [
      "downsample_layers.0.0.bn.weight",
      "downsample_layers.0.0.bn.bias",
      "downsample_layers.0.2.conv_exp_bn1.bn.weight",
      "downsample_layers.0.2.conv_exp_bn1.bn.bias",
      "downsample_layers.0.2.conv_pwl_bn2.bn.weight",
      "downsample_layers.0.2.conv_pwl_bn2.bn.bias",
      "downsample_layers.1.0.bn.weight",
      "downsample_layers.1.0.bn.bias",
      "downsample_layers.2.0.bn.weight",
      "downsample_layers.2.0.bn.bias",
      "downsample_layers.3.0.bn.weight",
      "downsample_layers.3.0.bn.bias",
      "stages.0.0.block.token_channel_mixer.m.0.bn.weight",
      "stages.0.0.block.token_channel_mixer.m.0.bn.bias",
      "stages.0.0.block.token_channel_mixer.m.1.bn.weight",
      "stages.0.0.block.token_channel_mixer.m.1.bn.bias",
      "stages.0.0.block.token_channel_mixer.m.3.bn.weight",
      "stages.0.0.block.token_channel_mixer.m.3.bn.bias",
      "stages.0.1.block.token_channel_mixer.m.0.bn.weight",
      "stages.0.1.block.token_channel_mixer.m.0.bn.bias",
      "stages.0.1.block.token_channel_mixer.m.1.bn.weight",
      "stages.0.1.block.token_channel_mixer.m.1.bn.bias",
      "stages.0.1.block.token_channel_mixer.m.3.bn.weight",
      "stages.0.1.block.token_channel_mixer.m.3.bn.bias",
      "stages.1.0.block.token_channel_mixer.m.0.bn.weight",
      "stages.1.0.block.token_channel_mixer.m.0.bn.bias",
      "stages.1.0.block.token_channel_mixer.m.1.bn.weight",
      "stages.1.0.block.token_channel_mixer.m.1.bn.bias",
      "stages.1.0.block.token_channel_mixer.m.3.bn.weight",
      "stages.1.0.block.token_channel_mixer.m.3.bn.bias",
      "stages.1.1.block.token_channel_mixer.m.0.bn.weight",
      "stages.1.1.block.token_channel_mixer.m.0.bn.bias",
      "stages.1.1.block.token_channel_mixer.m.1.bn.weight",
      "stages.1.1.block.token_channel_mixer.m.1.bn.bias",
      "stages.1.1.block.token_channel_mixer.m.3.bn.weight",
      "stages.1.1.block.token_channel_mixer.m.3.bn.bias",
      "stages.2.0.block.token_channel_mixer.m.0.bn.weight",
      "stages.2.0.block.token_channel_mixer.m.0.bn.bias",
      "stages.2.0.block.token_channel_mixer.m.1.bn.weight",
      "stages.2.0.block.token_channel_mixer.m.1.bn.bias",
      "stages.2.0.block.token_channel_mixer.m.3.bn.weight",
      "stages.2.0.block.token_channel_mixer.m.3.bn.bias",
      "stages.2.1.block.token_channel_mixer.m.0.bn.weight",
      "stages.2.1.block.token_channel_mixer.m.0.bn.bias",
      "stages.2.1.block.token_channel_mixer.m.1.bn.weight",
      "stages.2.1.block.token_channel_mixer.m.1.bn.bias",
      "stages.2.1.block.token_channel_mixer.m.3.bn.weight",
      "stages.2.1.block.token_channel_mixer.m.3.bn.bias",
      "stages.2.2.block.token_channel_mixer.m.0.bn.weight",
      "stages.2.2.block.token_channel_mixer.m.0.bn.bias",
      "stages.2.2.block.token_channel_mixer.m.1.bn.weight",
      "stages.2.2.block.token_channel_mixer.m.1.bn.bias",
      "stages.2.2.block.token_channel_mixer.m.3.bn.weight",
      "stages.2.2.block.token_channel_mixer.m.3.bn.bias",
      "stages.2.3.block.token_channel_mixer.m.0.bn.weight",
      "stages.2.3.block.token_channel_mixer.m.0.bn.bias",
      "stages.2.3.block.token_channel_mixer.m.1.bn.weight",
      "stages.2.3.block.token_channel_mixer.m.1.bn.bias",
      "stages.2.3.block.token_channel_mixer.m.3.bn.weight",
      "stages.2.3.block.token_channel_mixer.m.3.bn.bias",
      "stages.2.4.block.token_channel_mixer.m.0.bn.weight",
      "stages.2.4.block.token_channel_mixer.m.0.bn.bias",
      "stages.2.4.block.token_channel_mixer.m.1.bn.weight",
      "stages.2.4.block.token_channel_mixer.m.1.bn.bias",
      "stages.2.4.block.token_channel_mixer.m.3.bn.weight",
      "stages.2.4.block.token_channel_mixer.m.3.bn.bias",
      "stages.2.5.block.token_channel_mixer.m.0.bn.weight",
      "stages.2.5.block.token_channel_mixer.m.0.bn.bias",
      "stages.2.5.block.token_channel_mixer.m.1.bn.weight",
      "stages.2.5.block.token_channel_mixer.m.1.bn.bias",
      "stages.2.5.block.token_channel_mixer.m.3.bn.weight",
      "stages.2.5.block.token_channel_mixer.m.3.bn.bias",
      "stages.2.6.block.token_channel_mixer.m.0.bn.weight",
      "stages.2.6.block.token_channel_mixer.m.0.bn.bias",
      "stages.2.6.block.token_channel_mixer.m.1.bn.weight",
      "stages.2.6.block.token_channel_mixer.m.1.bn.bias",
      "stages.2.6.block.token_channel_mixer.m.3.bn.weight",
      "stages.2.6.block.token_channel_mixer.m.3.bn.bias",
      "stages.2.7.block.token_channel_mixer.m.0.bn.weight",
      "stages.2.7.block.token_channel_mixer.m.0.bn.bias",
      "stages.2.7.block.token_channel_mixer.m.1.bn.weight",
      "stages.2.7.block.token_channel_mixer.m.1.bn.bias",
      "stages.2.7.block.token_channel_mixer.m.3.bn.weight",
      "stages.2.7.block.token_channel_mixer.m.3.bn.bias",
      "stages.2.8.block.cpe.m.bn.weight",
      "stages.2.8.block.cpe.m.bn.bias",
      "stages.2.9.block.token_channel_mixer.m.q.bn.weight",
      "stages.2.9.block.token_channel_mixer.m.q.bn.bias",
      "stages.2.9.block.token_channel_mixer.m.k.bn.weight",
      "stages.2.9.block.token_channel_mixer.m.k.bn.bias",
      "stages.2.9.block.token_channel_mixer.m.proj.bn.weight",
      "stages.2.9.block.token_channel_mixer.m.proj.bn.bias",
      "stages.2.9.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.2.9.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.2.10.block.channel_mixer.m.0.bn.weight",
      "stages.2.10.block.channel_mixer.m.0.bn.bias",
      "stages.2.10.block.channel_mixer.m.2.bn.weight",
      "stages.2.10.block.channel_mixer.m.2.bn.bias",
      "stages.2.11.block.cpe.m.bn.weight",
      "stages.2.11.block.cpe.m.bn.bias",
      "stages.2.12.block.token_channel_mixer.m.q.bn.weight",
      "stages.2.12.block.token_channel_mixer.m.q.bn.bias",
      "stages.2.12.block.token_channel_mixer.m.k.bn.weight",
      "stages.2.12.block.token_channel_mixer.m.k.bn.bias",
      "stages.2.12.block.token_channel_mixer.m.proj.bn.weight",
      "stages.2.12.block.token_channel_mixer.m.proj.bn.bias",
      "stages.2.12.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.2.12.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.2.13.block.channel_mixer.m.0.bn.weight",
      "stages.2.13.block.channel_mixer.m.0.bn.bias",
      "stages.2.13.block.channel_mixer.m.2.bn.weight",
      "stages.2.13.block.channel_mixer.m.2.bn.bias",
      "stages.2.14.block.cpe.m.bn.weight",
      "stages.2.14.block.cpe.m.bn.bias",
      "stages.2.15.block.token_channel_mixer.m.q.bn.weight",
      "stages.2.15.block.token_channel_mixer.m.q.bn.bias",
      "stages.2.15.block.token_channel_mixer.m.k.bn.weight",
      "stages.2.15.block.token_channel_mixer.m.k.bn.bias",
      "stages.2.15.block.token_channel_mixer.m.proj.bn.weight",
      "stages.2.15.block.token_channel_mixer.m.proj.bn.bias",
      "stages.2.15.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.2.15.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.2.16.block.channel_mixer.m.0.bn.weight",
      "stages.2.16.block.channel_mixer.m.0.bn.bias",
      "stages.2.16.block.channel_mixer.m.2.bn.weight",
      "stages.2.16.block.channel_mixer.m.2.bn.bias",
      "stages.2.17.block.cpe.m.bn.weight",
      "stages.2.17.block.cpe.m.bn.bias",
      "stages.2.18.block.token_channel_mixer.m.q.bn.weight",
      "stages.2.18.block.token_channel_mixer.m.q.bn.bias",
      "stages.2.18.block.token_channel_mixer.m.k.bn.weight",
      "stages.2.18.block.token_channel_mixer.m.k.bn.bias",
      "stages.2.18.block.token_channel_mixer.m.proj.bn.weight",
      "stages.2.18.block.token_channel_mixer.m.proj.bn.bias",
      "stages.2.18.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.2.18.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.2.19.block.channel_mixer.m.0.bn.weight",
      "stages.2.19.block.channel_mixer.m.0.bn.bias",
      "stages.2.19.block.channel_mixer.m.2.bn.weight",
      "stages.2.19.block.channel_mixer.m.2.bn.bias",
      "stages.2.20.block.cpe.m.bn.weight",
      "stages.2.20.block.cpe.m.bn.bias",
      "stages.2.21.block.token_channel_mixer.m.q.bn.weight",
      "stages.2.21.block.token_channel_mixer.m.q.bn.bias",
      "stages.2.21.block.token_channel_mixer.m.k.bn.weight",
      "stages.2.21.block.token_channel_mixer.m.k.bn.bias",
      "stages.2.21.block.token_channel_mixer.m.proj.bn.weight",
      "stages.2.21.block.token_channel_mixer.m.proj.bn.bias",
      "stages.2.21.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.2.21.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.2.22.block.channel_mixer.m.0.bn.weight",
      "stages.2.22.block.channel_mixer.m.0.bn.bias",
      "stages.2.22.block.channel_mixer.m.2.bn.weight",
      "stages.2.22.block.channel_mixer.m.2.bn.bias",
      "stages.2.23.block.cpe.m.bn.weight",
      "stages.2.23.block.cpe.m.bn.bias",
      "stages.2.24.block.token_channel_mixer.m.q.bn.weight",
      "stages.2.24.block.token_channel_mixer.m.q.bn.bias",
      "stages.2.24.block.token_channel_mixer.m.k.bn.weight",
      "stages.2.24.block.token_channel_mixer.m.k.bn.bias",
      "stages.2.24.block.token_channel_mixer.m.proj.bn.weight",
      "stages.2.24.block.token_channel_mixer.m.proj.bn.bias",
      "stages.2.24.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.2.24.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.2.25.block.channel_mixer.m.0.bn.weight",
      "stages.2.25.block.channel_mixer.m.0.bn.bias",
      "stages.2.25.block.channel_mixer.m.2.bn.weight",
      "stages.2.25.block.channel_mixer.m.2.bn.bias",
      "stages.2.26.block.cpe.m.bn.weight",
      "stages.2.26.block.cpe.m.bn.bias",
      "stages.2.27.block.token_channel_mixer.m.q.bn.weight",
      "stages.2.27.block.token_channel_mixer.m.q.bn.bias",
      "stages.2.27.block.token_channel_mixer.m.k.bn.weight",
      "stages.2.27.block.token_channel_mixer.m.k.bn.bias",
      "stages.2.27.block.token_channel_mixer.m.proj.bn.weight",
      "stages.2.27.block.token_channel_mixer.m.proj.bn.bias",
      "stages.2.27.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.2.27.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.2.28.block.channel_mixer.m.0.bn.weight",
      "stages.2.28.block.channel_mixer.m.0.bn.bias",
      "stages.2.28.block.channel_mixer.m.2.bn.weight",
      "stages.2.28.block.channel_mixer.m.2.bn.bias",
      "stages.2.29.block.cpe.m.bn.weight",
      "stages.2.29.block.cpe.m.bn.bias",
      "stages.2.30.block.token_channel_mixer.m.q.bn.weight",
      "stages.2.30.block.token_channel_mixer.m.q.bn.bias",
      "stages.2.30.block.token_channel_mixer.m.k.bn.weight",
      "stages.2.30.block.token_channel_mixer.m.k.bn.bias",
      "stages.2.30.block.token_channel_mixer.m.proj.bn.weight",
      "stages.2.30.block.token_channel_mixer.m.proj.bn.bias",
      "stages.2.30.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.2.30.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.2.31.block.channel_mixer.m.0.bn.weight",
      "stages.2.31.block.channel_mixer.m.0.bn.bias",
      "stages.2.31.block.channel_mixer.m.2.bn.weight",
      "stages.2.31.block.channel_mixer.m.2.bn.bias",
      "stages.2.32.block.token_channel_mixer.m.0.bn.weight",
      "stages.2.32.block.token_channel_mixer.m.0.bn.bias",
      "stages.2.32.block.token_channel_mixer.m.1.bn.weight",
      "stages.2.32.block.token_channel_mixer.m.1.bn.bias",
      "stages.2.32.block.token_channel_mixer.m.3.bn.weight",
      "stages.2.32.block.token_channel_mixer.m.3.bn.bias",
      "stages.3.0.block.cpe.m.bn.weight",
      "stages.3.0.block.cpe.m.bn.bias",
      "stages.3.1.block.token_channel_mixer.m.q.bn.weight",
      "stages.3.1.block.token_channel_mixer.m.q.bn.bias",
      "stages.3.1.block.token_channel_mixer.m.k.bn.weight",
      "stages.3.1.block.token_channel_mixer.m.k.bn.bias",
      "stages.3.1.block.token_channel_mixer.m.proj.bn.weight",
      "stages.3.1.block.token_channel_mixer.m.proj.bn.bias",
      "stages.3.1.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.3.1.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.3.2.block.channel_mixer.m.0.bn.weight",
      "stages.3.2.block.channel_mixer.m.0.bn.bias",
      "stages.3.2.block.channel_mixer.m.2.bn.weight",
      "stages.3.2.block.channel_mixer.m.2.bn.bias",
      "stages.3.3.block.cpe.m.bn.weight",
      "stages.3.3.block.cpe.m.bn.bias",
      "stages.3.4.block.token_channel_mixer.m.q.bn.weight",
      "stages.3.4.block.token_channel_mixer.m.q.bn.bias",
      "stages.3.4.block.token_channel_mixer.m.k.bn.weight",
      "stages.3.4.block.token_channel_mixer.m.k.bn.bias",
      "stages.3.4.block.token_channel_mixer.m.proj.bn.weight",
      "stages.3.4.block.token_channel_mixer.m.proj.bn.bias",
      "stages.3.4.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.3.4.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.3.5.block.channel_mixer.m.0.bn.weight",
      "stages.3.5.block.channel_mixer.m.0.bn.bias",
      "stages.3.5.block.channel_mixer.m.2.bn.weight",
      "stages.3.5.block.channel_mixer.m.2.bn.bias",
      "classifier.classifier.bn.weight",
      "classifier.classifier.bn.bias",
      "classifier.classifier.l.bias"
    ],
    "lr_scale": 1.0
  }
}
Use Cosine LR scheduler
Set warmup steps = 3120
Set warmup steps = 0
Max WD = 0.0500000, Min WD = 0.0500000
criterion = SoftTargetCrossEntropy()
Auto resume checkpoint: 
Start training for 300 epochs
Epoch: [0]  [  0/156]  eta: 2:39:57  lr: 0.000000  min_lr: 0.000000  loss: 6.9882 (6.9882)  weight_decay: 0.0500 (0.0500)  time: 61.5198  data: 18.0429  max mem: 55573
Epoch: [0]  [ 10/156]  eta: 0:17:51  lr: 0.000026  min_lr: 0.000026  loss: 6.9841 (6.9752)  weight_decay: 0.0500 (0.0500)  time: 7.3382  data: 1.6407  max mem: 55573
Epoch: [0]  [ 20/156]  eta: 0:09:20  lr: 0.000051  min_lr: 0.000051  loss: 6.9558 (6.9610)  weight_decay: 0.0500 (0.0500)  time: 1.2545  data: 0.0004  max mem: 55573
Epoch: [0]  [ 30/156]  eta: 0:06:15  lr: 0.000077  min_lr: 0.000077  loss: 6.9283 (6.9458)  weight_decay: 0.0500 (0.0500)  time: 0.5880  data: 0.0004  max mem: 55573
Epoch: [0]  [ 40/156]  eta: 0:04:38  lr: 0.000103  min_lr: 0.000103  loss: 6.9006 (6.9314)  weight_decay: 0.0500 (0.0500)  time: 0.5878  data: 0.0004  max mem: 55573
Epoch: [0]  [ 50/156]  eta: 0:03:36  lr: 0.000128  min_lr: 0.000128  loss: 6.8754 (6.9182)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0005  max mem: 55573
Epoch: [0]  [ 60/156]  eta: 0:02:59  lr: 0.000154  min_lr: 0.000154  loss: 6.8639 (6.9062)  weight_decay: 0.0500 (0.0500)  time: 0.7854  data: 0.0004  max mem: 55573
Epoch: [0]  [ 70/156]  eta: 0:02:26  lr: 0.000180  min_lr: 0.000180  loss: 6.8312 (6.8930)  weight_decay: 0.0500 (0.0500)  time: 0.8393  data: 0.0004  max mem: 55573
Epoch: [0]  [ 80/156]  eta: 0:02:05  lr: 0.000205  min_lr: 0.000205  loss: 6.8060 (6.8813)  weight_decay: 0.0500 (0.0500)  time: 1.0035  data: 0.0004  max mem: 55573
Epoch: [0]  [ 90/156]  eta: 0:01:45  lr: 0.000231  min_lr: 0.000231  loss: 6.7895 (6.8692)  weight_decay: 0.0500 (0.0500)  time: 1.2319  data: 0.0005  max mem: 55573
Epoch: [0]  [100/156]  eta: 0:01:25  lr: 0.000256  min_lr: 0.000256  loss: 6.7680 (6.8571)  weight_decay: 0.0500 (0.0500)  time: 0.9804  data: 0.0005  max mem: 55573
Epoch: [0]  [110/156]  eta: 0:01:08  lr: 0.000282  min_lr: 0.000282  loss: 6.7350 (6.8447)  weight_decay: 0.0500 (0.0500)  time: 1.0022  data: 0.0004  max mem: 55573
Epoch: [0]  [120/156]  eta: 0:00:51  lr: 0.000308  min_lr: 0.000308  loss: 6.7350 (6.8340)  weight_decay: 0.0500 (0.0500)  time: 0.9828  data: 0.0004  max mem: 55573
Epoch: [0]  [130/156]  eta: 0:00:37  lr: 0.000333  min_lr: 0.000333  loss: 6.7077 (6.8211)  weight_decay: 0.0500 (0.0500)  time: 1.0519  data: 0.0010  max mem: 55573
Epoch: [0]  [140/156]  eta: 0:00:22  lr: 0.000359  min_lr: 0.000359  loss: 6.6225 (6.8079)  weight_decay: 0.0500 (0.0500)  time: 1.2889  data: 0.0009  max mem: 55573
Epoch: [0]  [150/156]  eta: 0:00:08  lr: 0.000385  min_lr: 0.000385  loss: 6.6308 (6.7964)  weight_decay: 0.0500 (0.0500)  time: 0.9255  data: 0.0002  max mem: 55573
Epoch: [0]  [155/156]  eta: 0:00:01  lr: 0.000398  min_lr: 0.000398  loss: 6.6308 (6.7923)  weight_decay: 0.0500 (0.0500)  time: 0.9246  data: 0.0002  max mem: 55573
Epoch: [0] Total time: 0:03:29 (1.3419 s / it)
Averaged stats: lr: 0.000398  min_lr: 0.000398  loss: 6.6308 (6.7902)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:01:00  loss: 6.0855 (6.0855)  acc1: 1.5625 (1.5625)  acc5: 6.3802 (6.3802)  time: 12.0442  data: 7.0845  max mem: 55573
Test:  [4/5]  eta: 0:00:02  loss: 6.0897 (5.9683)  acc1: 2.4740 (2.5920)  acc5: 7.8125 (7.7120)  time: 2.8886  data: 1.4170  max mem: 55573
Test: Total time: 0:00:14 (2.9156 s / it)
* Acc@1 2.226 Acc@5 7.746 loss 6.005
Accuracy of the model on the 50000 test images: 2.2%
Max accuracy: 2.23%
Test:  [0/5]  eta: 0:00:36  loss: 6.9148 (6.9148)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 7.3080  data: 6.5321  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9148 (6.9127)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.6068  data: 1.3065  max mem: 55573
Test: Total time: 0:00:08 (1.6238 s / it)
* Acc@1 0.098 Acc@5 0.494 loss 6.911
Accuracy of the model EMA on 50000 test images: 0.1%
Max EMA accuracy: 0.10%
Epoch: [1]  [  0/156]  eta: 1:06:57  lr: 0.000400  min_lr: 0.000400  loss: 6.6372 (6.6372)  weight_decay: 0.0500 (0.0500)  time: 25.7538  data: 10.8100  max mem: 55573
Epoch: [1]  [ 10/156]  eta: 0:08:29  lr: 0.000426  min_lr: 0.000426  loss: 6.6414 (6.6380)  weight_decay: 0.0500 (0.0500)  time: 3.4909  data: 0.9830  max mem: 55573
Epoch: [1]  [ 20/156]  eta: 0:05:30  lr: 0.000451  min_lr: 0.000451  loss: 6.6266 (6.6014)  weight_decay: 0.0500 (0.0500)  time: 1.2617  data: 0.0004  max mem: 55573
Epoch: [1]  [ 30/156]  eta: 0:04:03  lr: 0.000477  min_lr: 0.000477  loss: 6.5608 (6.5878)  weight_decay: 0.0500 (0.0500)  time: 1.0777  data: 0.0004  max mem: 55573
Epoch: [1]  [ 40/156]  eta: 0:03:06  lr: 0.000503  min_lr: 0.000503  loss: 6.5216 (6.5705)  weight_decay: 0.0500 (0.0500)  time: 0.7420  data: 0.0004  max mem: 55573
Epoch: [1]  [ 50/156]  eta: 0:02:29  lr: 0.000528  min_lr: 0.000528  loss: 6.5068 (6.5513)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0005  max mem: 55573
Epoch: [1]  [ 60/156]  eta: 0:02:02  lr: 0.000554  min_lr: 0.000554  loss: 6.5115 (6.5436)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [1]  [ 70/156]  eta: 0:01:41  lr: 0.000580  min_lr: 0.000580  loss: 6.5306 (6.5378)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0006  max mem: 55573
Epoch: [1]  [ 80/156]  eta: 0:01:23  lr: 0.000605  min_lr: 0.000605  loss: 6.4794 (6.5272)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [1]  [ 90/156]  eta: 0:01:09  lr: 0.000631  min_lr: 0.000631  loss: 6.5008 (6.5230)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [1]  [100/156]  eta: 0:00:56  lr: 0.000657  min_lr: 0.000657  loss: 6.3610 (6.5055)  weight_decay: 0.0500 (0.0500)  time: 0.6214  data: 0.0005  max mem: 55573
Epoch: [1]  [110/156]  eta: 0:00:45  lr: 0.000682  min_lr: 0.000682  loss: 6.3056 (6.4924)  weight_decay: 0.0500 (0.0500)  time: 0.7288  data: 0.0004  max mem: 55573
Epoch: [1]  [120/156]  eta: 0:00:34  lr: 0.000708  min_lr: 0.000708  loss: 6.4525 (6.4931)  weight_decay: 0.0500 (0.0500)  time: 0.7764  data: 0.0004  max mem: 55573
Epoch: [1]  [130/156]  eta: 0:00:24  lr: 0.000734  min_lr: 0.000734  loss: 6.5321 (6.4860)  weight_decay: 0.0500 (0.0500)  time: 0.7510  data: 0.0010  max mem: 55573
Epoch: [1]  [140/156]  eta: 0:00:14  lr: 0.000759  min_lr: 0.000759  loss: 6.3089 (6.4717)  weight_decay: 0.0500 (0.0500)  time: 0.7211  data: 0.0008  max mem: 55573
Epoch: [1]  [150/156]  eta: 0:00:05  lr: 0.000785  min_lr: 0.000785  loss: 6.3089 (6.4631)  weight_decay: 0.0500 (0.0500)  time: 0.6461  data: 0.0001  max mem: 55573
Epoch: [1]  [155/156]  eta: 0:00:00  lr: 0.000798  min_lr: 0.000798  loss: 6.3103 (6.4619)  weight_decay: 0.0500 (0.0500)  time: 0.6253  data: 0.0001  max mem: 55573
Epoch: [1] Total time: 0:02:21 (0.9093 s / it)
Averaged stats: lr: 0.000798  min_lr: 0.000798  loss: 6.3103 (6.4843)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 5.1609 (5.1609)  acc1: 6.7708 (6.7708)  acc5: 20.4427 (20.4427)  time: 7.1668  data: 6.9294  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 5.2887 (5.0925)  acc1: 6.7708 (7.0720)  acc5: 20.0521 (19.6160)  time: 1.5799  data: 1.3860  max mem: 55573
Test: Total time: 0:00:08 (1.6146 s / it)
* Acc@1 7.006 Acc@5 19.644 loss 5.148
Accuracy of the model on the 50000 test images: 7.0%
Max accuracy: 7.01%
Test:  [0/5]  eta: 0:00:34  loss: 6.9132 (6.9132)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.8970  data: 6.6611  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9132 (6.9115)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5247  data: 1.3323  max mem: 55573
Test: Total time: 0:00:07 (1.5397 s / it)
* Acc@1 0.096 Acc@5 0.498 loss 6.910
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [2]  [  0/156]  eta: 0:34:50  lr: 0.000800  min_lr: 0.000800  loss: 6.5567 (6.5567)  weight_decay: 0.0500 (0.0500)  time: 13.4002  data: 9.3431  max mem: 55573
Epoch: [2]  [ 10/156]  eta: 0:04:19  lr: 0.000826  min_lr: 0.000826  loss: 6.4973 (6.4743)  weight_decay: 0.0500 (0.0500)  time: 1.7756  data: 0.8498  max mem: 55573
Epoch: [2]  [ 20/156]  eta: 0:02:44  lr: 0.000852  min_lr: 0.000852  loss: 6.4505 (6.3945)  weight_decay: 0.0500 (0.0500)  time: 0.6034  data: 0.0004  max mem: 55573
Epoch: [2]  [ 30/156]  eta: 0:02:07  lr: 0.000877  min_lr: 0.000877  loss: 6.3530 (6.3838)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0004  max mem: 55573
Epoch: [2]  [ 40/156]  eta: 0:01:45  lr: 0.000903  min_lr: 0.000903  loss: 6.3529 (6.3603)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0004  max mem: 55573
Epoch: [2]  [ 50/156]  eta: 0:01:29  lr: 0.000929  min_lr: 0.000929  loss: 6.3168 (6.3350)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [2]  [ 60/156]  eta: 0:01:17  lr: 0.000954  min_lr: 0.000954  loss: 6.2585 (6.3105)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [2]  [ 70/156]  eta: 0:01:06  lr: 0.000980  min_lr: 0.000980  loss: 6.2585 (6.2891)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [2]  [ 80/156]  eta: 0:00:57  lr: 0.001005  min_lr: 0.001005  loss: 6.1635 (6.2726)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0005  max mem: 55573
Epoch: [2]  [ 90/156]  eta: 0:00:48  lr: 0.001031  min_lr: 0.001031  loss: 6.3098 (6.2770)  weight_decay: 0.0500 (0.0500)  time: 0.5937  data: 0.0004  max mem: 55573
Epoch: [2]  [100/156]  eta: 0:00:40  lr: 0.001057  min_lr: 0.001057  loss: 6.2947 (6.2762)  weight_decay: 0.0500 (0.0500)  time: 0.5942  data: 0.0004  max mem: 55573
Epoch: [2]  [110/156]  eta: 0:00:32  lr: 0.001082  min_lr: 0.001082  loss: 6.1876 (6.2597)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0005  max mem: 55573
Epoch: [2]  [120/156]  eta: 0:00:25  lr: 0.001108  min_lr: 0.001108  loss: 6.2815 (6.2639)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [2]  [130/156]  eta: 0:00:17  lr: 0.001134  min_lr: 0.001134  loss: 6.2815 (6.2589)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0011  max mem: 55573
Epoch: [2]  [140/156]  eta: 0:00:10  lr: 0.001159  min_lr: 0.001159  loss: 6.1512 (6.2414)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0009  max mem: 55573
Epoch: [2]  [150/156]  eta: 0:00:04  lr: 0.001185  min_lr: 0.001185  loss: 5.9349 (6.2238)  weight_decay: 0.0500 (0.0500)  time: 0.5857  data: 0.0001  max mem: 55573
Epoch: [2]  [155/156]  eta: 0:00:00  lr: 0.001198  min_lr: 0.001198  loss: 5.9132 (6.2125)  weight_decay: 0.0500 (0.0500)  time: 0.5862  data: 0.0001  max mem: 55573
Epoch: [2] Total time: 0:01:46 (0.6816 s / it)
Averaged stats: lr: 0.001198  min_lr: 0.001198  loss: 5.9132 (6.2210)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 4.5472 (4.5472)  acc1: 12.7604 (12.7604)  acc5: 33.7240 (33.7240)  time: 6.7376  data: 6.5007  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 4.6459 (4.5515)  acc1: 12.7604 (13.1520)  acc5: 32.1615 (31.5840)  time: 1.4928  data: 1.3002  max mem: 55573
Test: Total time: 0:00:07 (1.5156 s / it)
* Acc@1 13.466 Acc@5 32.376 loss 4.588
Accuracy of the model on the 50000 test images: 13.5%
Max accuracy: 13.47%
Test:  [0/5]  eta: 0:00:34  loss: 6.9125 (6.9125)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 6.9821  data: 6.7459  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9125 (6.9112)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5416  data: 1.3493  max mem: 55573
Test: Total time: 0:00:07 (1.5586 s / it)
* Acc@1 0.098 Acc@5 0.494 loss 6.909
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [3]  [  0/156]  eta: 0:33:09  lr: 0.001200  min_lr: 0.001200  loss: 5.4995 (5.4995)  weight_decay: 0.0500 (0.0500)  time: 12.7523  data: 9.8772  max mem: 55573
Epoch: [3]  [ 10/156]  eta: 0:04:09  lr: 0.001226  min_lr: 0.001226  loss: 6.0262 (6.0100)  weight_decay: 0.0500 (0.0500)  time: 1.7092  data: 0.8983  max mem: 55573
Epoch: [3]  [ 20/156]  eta: 0:02:40  lr: 0.001252  min_lr: 0.001252  loss: 6.0837 (6.0797)  weight_decay: 0.0500 (0.0500)  time: 0.5978  data: 0.0004  max mem: 55573
Epoch: [3]  [ 30/156]  eta: 0:02:04  lr: 0.001277  min_lr: 0.001277  loss: 5.9958 (6.0507)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [3]  [ 40/156]  eta: 0:01:43  lr: 0.001303  min_lr: 0.001303  loss: 6.1164 (6.0677)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [3]  [ 50/156]  eta: 0:01:28  lr: 0.001329  min_lr: 0.001329  loss: 6.1642 (6.0808)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [3]  [ 60/156]  eta: 0:01:16  lr: 0.001354  min_lr: 0.001354  loss: 6.0965 (6.0342)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [3]  [ 70/156]  eta: 0:01:05  lr: 0.001380  min_lr: 0.001380  loss: 5.9905 (6.0262)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [3]  [ 80/156]  eta: 0:00:56  lr: 0.001406  min_lr: 0.001406  loss: 6.1164 (6.0223)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [3]  [ 90/156]  eta: 0:00:47  lr: 0.001431  min_lr: 0.001431  loss: 6.1260 (6.0286)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [3]  [100/156]  eta: 0:00:39  lr: 0.001457  min_lr: 0.001457  loss: 6.1993 (6.0315)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [3]  [110/156]  eta: 0:00:32  lr: 0.001483  min_lr: 0.001483  loss: 6.1564 (6.0311)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [3]  [120/156]  eta: 0:00:24  lr: 0.001508  min_lr: 0.001508  loss: 6.1283 (6.0259)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [3]  [130/156]  eta: 0:00:17  lr: 0.001534  min_lr: 0.001534  loss: 6.0463 (6.0195)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0010  max mem: 55573
Epoch: [3]  [140/156]  eta: 0:00:10  lr: 0.001559  min_lr: 0.001559  loss: 5.8850 (6.0068)  weight_decay: 0.0500 (0.0500)  time: 0.5867  data: 0.0008  max mem: 55573
Epoch: [3]  [150/156]  eta: 0:00:04  lr: 0.001585  min_lr: 0.001585  loss: 5.9052 (5.9991)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [3]  [155/156]  eta: 0:00:00  lr: 0.001598  min_lr: 0.001598  loss: 5.9807 (6.0035)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [3] Total time: 0:01:45 (0.6753 s / it)
Averaged stats: lr: 0.001598  min_lr: 0.001598  loss: 5.9807 (5.9695)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 3.9729 (3.9729)  acc1: 20.0521 (20.0521)  acc5: 45.9635 (45.9635)  time: 7.1147  data: 6.8773  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 4.0859 (3.9698)  acc1: 20.0521 (19.2320)  acc5: 43.7500 (40.8960)  time: 1.5686  data: 1.3756  max mem: 55573
Test: Total time: 0:00:08 (1.6058 s / it)
* Acc@1 19.768 Acc@5 41.254 loss 3.958
Accuracy of the model on the 50000 test images: 19.8%
Max accuracy: 19.77%
Test:  [0/5]  eta: 0:00:36  loss: 6.9120 (6.9120)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.2487  data: 7.0127  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9120 (6.9114)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.5950  data: 1.4026  max mem: 55573
Test: Total time: 0:00:08 (1.6137 s / it)
* Acc@1 0.098 Acc@5 0.492 loss 6.909
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [4]  [  0/156]  eta: 0:31:45  lr: 0.001601  min_lr: 0.001601  loss: 6.0239 (6.0239)  weight_decay: 0.0500 (0.0500)  time: 12.2145  data: 11.1444  max mem: 55573
Epoch: [4]  [ 10/156]  eta: 0:04:04  lr: 0.001626  min_lr: 0.001626  loss: 5.7407 (5.8413)  weight_decay: 0.0500 (0.0500)  time: 1.6738  data: 1.0135  max mem: 55573
Epoch: [4]  [ 20/156]  eta: 0:02:37  lr: 0.001652  min_lr: 0.001652  loss: 5.6432 (5.7308)  weight_decay: 0.0500 (0.0500)  time: 0.6059  data: 0.0005  max mem: 55573
Epoch: [4]  [ 30/156]  eta: 0:02:02  lr: 0.001677  min_lr: 0.001677  loss: 5.6773 (5.7710)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [4]  [ 40/156]  eta: 0:01:42  lr: 0.001703  min_lr: 0.001703  loss: 5.6765 (5.7286)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [4]  [ 50/156]  eta: 0:01:27  lr: 0.001729  min_lr: 0.001729  loss: 5.9544 (5.7858)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [4]  [ 60/156]  eta: 0:01:15  lr: 0.001754  min_lr: 0.001754  loss: 5.9651 (5.7807)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [4]  [ 70/156]  eta: 0:01:05  lr: 0.001780  min_lr: 0.001780  loss: 5.6288 (5.7597)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [4]  [ 80/156]  eta: 0:00:56  lr: 0.001806  min_lr: 0.001806  loss: 5.4777 (5.7458)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [4]  [ 90/156]  eta: 0:00:47  lr: 0.001831  min_lr: 0.001831  loss: 5.3874 (5.7143)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [4]  [100/156]  eta: 0:00:39  lr: 0.001857  min_lr: 0.001857  loss: 5.5030 (5.7062)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [4]  [110/156]  eta: 0:00:32  lr: 0.001883  min_lr: 0.001883  loss: 5.7600 (5.7179)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [4]  [120/156]  eta: 0:00:24  lr: 0.001908  min_lr: 0.001908  loss: 5.8973 (5.7232)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [4]  [130/156]  eta: 0:00:17  lr: 0.001934  min_lr: 0.001934  loss: 5.8772 (5.7087)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0011  max mem: 55573
Epoch: [4]  [140/156]  eta: 0:00:10  lr: 0.001960  min_lr: 0.001960  loss: 5.7799 (5.7196)  weight_decay: 0.0500 (0.0500)  time: 0.5855  data: 0.0010  max mem: 55573
Epoch: [4]  [150/156]  eta: 0:00:04  lr: 0.001985  min_lr: 0.001985  loss: 5.8654 (5.7217)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [4]  [155/156]  eta: 0:00:00  lr: 0.001998  min_lr: 0.001998  loss: 5.7799 (5.7202)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [4] Total time: 0:01:44 (0.6726 s / it)
Averaged stats: lr: 0.001998  min_lr: 0.001998  loss: 5.7799 (5.7241)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 3.3159 (3.3159)  acc1: 32.1615 (32.1615)  acc5: 57.1615 (57.1615)  time: 7.1629  data: 6.9250  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 3.5164 (3.4945)  acc1: 30.2083 (28.1920)  acc5: 52.6042 (50.6240)  time: 1.5792  data: 1.3851  max mem: 55573
Test: Total time: 0:00:08 (1.6140 s / it)
* Acc@1 27.444 Acc@5 51.622 loss 3.505
Accuracy of the model on the 50000 test images: 27.4%
Max accuracy: 27.44%
Test:  [0/5]  eta: 0:00:34  loss: 6.9118 (6.9118)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9262  data: 6.6902  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9118 (6.9117)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.5305  data: 1.3381  max mem: 55573
Test: Total time: 0:00:07 (1.5466 s / it)
* Acc@1 0.098 Acc@5 0.488 loss 6.910
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [5]  [  0/156]  eta: 0:35:56  lr: 0.002001  min_lr: 0.002001  loss: 6.0397 (6.0397)  weight_decay: 0.0500 (0.0500)  time: 13.8261  data: 6.3486  max mem: 55573
Epoch: [5]  [ 10/156]  eta: 0:04:23  lr: 0.002026  min_lr: 0.002026  loss: 5.6057 (5.5751)  weight_decay: 0.0500 (0.0500)  time: 1.8032  data: 0.5777  max mem: 55573
Epoch: [5]  [ 20/156]  eta: 0:02:46  lr: 0.002052  min_lr: 0.002052  loss: 5.6515 (5.6107)  weight_decay: 0.0500 (0.0500)  time: 0.5968  data: 0.0005  max mem: 55573
Epoch: [5]  [ 30/156]  eta: 0:02:08  lr: 0.002078  min_lr: 0.002078  loss: 5.2684 (5.4593)  weight_decay: 0.0500 (0.0500)  time: 0.5931  data: 0.0004  max mem: 55573
Epoch: [5]  [ 40/156]  eta: 0:01:46  lr: 0.002103  min_lr: 0.002103  loss: 5.2684 (5.5243)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [5]  [ 50/156]  eta: 0:01:30  lr: 0.002129  min_lr: 0.002129  loss: 5.7081 (5.4807)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [5]  [ 60/156]  eta: 0:01:17  lr: 0.002155  min_lr: 0.002155  loss: 5.1724 (5.4796)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [5]  [ 70/156]  eta: 0:01:07  lr: 0.002180  min_lr: 0.002180  loss: 5.6109 (5.4798)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [5]  [ 80/156]  eta: 0:00:57  lr: 0.002206  min_lr: 0.002206  loss: 5.6766 (5.4824)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [5]  [ 90/156]  eta: 0:00:48  lr: 0.002231  min_lr: 0.002231  loss: 5.5143 (5.4879)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [5]  [100/156]  eta: 0:00:40  lr: 0.002257  min_lr: 0.002257  loss: 5.4080 (5.4700)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [5]  [110/156]  eta: 0:00:32  lr: 0.002283  min_lr: 0.002283  loss: 5.3709 (5.4618)  weight_decay: 0.0500 (0.0500)  time: 0.5887  data: 0.0005  max mem: 55573
Epoch: [5]  [120/156]  eta: 0:00:25  lr: 0.002308  min_lr: 0.002308  loss: 5.6112 (5.4626)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [5]  [130/156]  eta: 0:00:18  lr: 0.002334  min_lr: 0.002334  loss: 5.6644 (5.4629)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0010  max mem: 55573
Epoch: [5]  [140/156]  eta: 0:00:10  lr: 0.002360  min_lr: 0.002360  loss: 5.7219 (5.4711)  weight_decay: 0.0500 (0.0500)  time: 0.5861  data: 0.0009  max mem: 55573
Epoch: [5]  [150/156]  eta: 0:00:04  lr: 0.002385  min_lr: 0.002385  loss: 5.6716 (5.4863)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [5]  [155/156]  eta: 0:00:00  lr: 0.002398  min_lr: 0.002398  loss: 5.5963 (5.4720)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [5] Total time: 0:01:46 (0.6818 s / it)
Averaged stats: lr: 0.002398  min_lr: 0.002398  loss: 5.5963 (5.5087)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 2.8956 (2.8956)  acc1: 37.2396 (37.2396)  acc5: 64.3229 (64.3229)  time: 6.8871  data: 6.6498  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 3.0522 (2.9907)  acc1: 35.4167 (32.6400)  acc5: 63.5417 (58.3040)  time: 1.5228  data: 1.3300  max mem: 55573
Test: Total time: 0:00:07 (1.5381 s / it)
* Acc@1 33.658 Acc@5 59.640 loss 3.020
Accuracy of the model on the 50000 test images: 33.7%
Max accuracy: 33.66%
Test:  [0/5]  eta: 0:00:34  loss: 6.9117 (6.9117)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.8973  data: 6.6612  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9117 (6.9123)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.5248  data: 1.3324  max mem: 55573
Test: Total time: 0:00:07 (1.5427 s / it)
* Acc@1 0.098 Acc@5 0.492 loss 6.910
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [6]  [  0/156]  eta: 0:30:46  lr: 0.002401  min_lr: 0.002401  loss: 5.8381 (5.8381)  weight_decay: 0.0500 (0.0500)  time: 11.8366  data: 11.2461  max mem: 55573
Epoch: [6]  [ 10/156]  eta: 0:04:18  lr: 0.002426  min_lr: 0.002426  loss: 5.4912 (5.3132)  weight_decay: 0.0500 (0.0500)  time: 1.7704  data: 1.0227  max mem: 55573
Epoch: [6]  [ 20/156]  eta: 0:02:44  lr: 0.002452  min_lr: 0.002452  loss: 5.2879 (5.2546)  weight_decay: 0.0500 (0.0500)  time: 0.6771  data: 0.0004  max mem: 55573
Epoch: [6]  [ 30/156]  eta: 0:02:07  lr: 0.002478  min_lr: 0.002478  loss: 5.3311 (5.3310)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [6]  [ 40/156]  eta: 0:01:45  lr: 0.002503  min_lr: 0.002503  loss: 5.5733 (5.3727)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [6]  [ 50/156]  eta: 0:01:29  lr: 0.002529  min_lr: 0.002529  loss: 5.5733 (5.3655)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [6]  [ 60/156]  eta: 0:01:17  lr: 0.002555  min_lr: 0.002555  loss: 5.5075 (5.3485)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [6]  [ 70/156]  eta: 0:01:06  lr: 0.002580  min_lr: 0.002580  loss: 5.3876 (5.3456)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [6]  [ 80/156]  eta: 0:00:57  lr: 0.002606  min_lr: 0.002606  loss: 5.5036 (5.3662)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [6]  [ 90/156]  eta: 0:00:48  lr: 0.002632  min_lr: 0.002632  loss: 5.4555 (5.3300)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0005  max mem: 55573
Epoch: [6]  [100/156]  eta: 0:00:40  lr: 0.002657  min_lr: 0.002657  loss: 5.2874 (5.3323)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [6]  [110/156]  eta: 0:00:32  lr: 0.002683  min_lr: 0.002683  loss: 5.5990 (5.3370)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [6]  [120/156]  eta: 0:00:25  lr: 0.002709  min_lr: 0.002709  loss: 5.5856 (5.3435)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [6]  [130/156]  eta: 0:00:17  lr: 0.002734  min_lr: 0.002734  loss: 5.5202 (5.3409)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0010  max mem: 55573
Epoch: [6]  [140/156]  eta: 0:00:10  lr: 0.002760  min_lr: 0.002760  loss: 5.1448 (5.3225)  weight_decay: 0.0500 (0.0500)  time: 0.5857  data: 0.0008  max mem: 55573
Epoch: [6]  [150/156]  eta: 0:00:04  lr: 0.002786  min_lr: 0.002786  loss: 5.0486 (5.3137)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0002  max mem: 55573
Epoch: [6]  [155/156]  eta: 0:00:00  lr: 0.002798  min_lr: 0.002798  loss: 5.1384 (5.3067)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0002  max mem: 55573
Epoch: [6] Total time: 0:01:45 (0.6795 s / it)
Averaged stats: lr: 0.002798  min_lr: 0.002798  loss: 5.1384 (5.3091)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 2.6750 (2.6750)  acc1: 43.6198 (43.6198)  acc5: 70.4427 (70.4427)  time: 6.8400  data: 6.6028  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.7894 (2.7218)  acc1: 41.0156 (38.3040)  acc5: 68.3594 (63.9680)  time: 1.5146  data: 1.3206  max mem: 55573
Test: Total time: 0:00:07 (1.5429 s / it)
* Acc@1 38.372 Acc@5 64.556 loss 2.750
Accuracy of the model on the 50000 test images: 38.4%
Max accuracy: 38.37%
Test:  [0/5]  eta: 0:00:34  loss: 6.9116 (6.9116)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9757  data: 6.7396  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9116 (6.9131)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.5405  data: 1.3480  max mem: 55573
Test: Total time: 0:00:07 (1.5556 s / it)
* Acc@1 0.098 Acc@5 0.494 loss 6.911
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [7]  [  0/156]  eta: 0:37:13  lr: 0.002801  min_lr: 0.002801  loss: 5.5152 (5.5152)  weight_decay: 0.0500 (0.0500)  time: 14.3186  data: 10.4473  max mem: 55573
Epoch: [7]  [ 10/156]  eta: 0:04:29  lr: 0.002827  min_lr: 0.002827  loss: 5.3660 (5.1670)  weight_decay: 0.0500 (0.0500)  time: 1.8480  data: 0.9501  max mem: 55573
Epoch: [7]  [ 20/156]  eta: 0:02:49  lr: 0.002852  min_lr: 0.002852  loss: 5.2283 (5.1960)  weight_decay: 0.0500 (0.0500)  time: 0.5956  data: 0.0004  max mem: 55573
Epoch: [7]  [ 30/156]  eta: 0:02:10  lr: 0.002878  min_lr: 0.002878  loss: 5.4239 (5.2517)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [7]  [ 40/156]  eta: 0:01:47  lr: 0.002903  min_lr: 0.002903  loss: 5.3792 (5.2436)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [7]  [ 50/156]  eta: 0:01:31  lr: 0.002929  min_lr: 0.002929  loss: 5.3339 (5.2421)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [7]  [ 60/156]  eta: 0:01:18  lr: 0.002955  min_lr: 0.002955  loss: 5.2355 (5.2310)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [7]  [ 70/156]  eta: 0:01:07  lr: 0.002980  min_lr: 0.002980  loss: 5.0080 (5.1935)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [7]  [ 80/156]  eta: 0:00:57  lr: 0.003006  min_lr: 0.003006  loss: 5.1562 (5.1975)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [7]  [ 90/156]  eta: 0:00:49  lr: 0.003032  min_lr: 0.003032  loss: 5.5073 (5.2387)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [7]  [100/156]  eta: 0:00:40  lr: 0.003057  min_lr: 0.003057  loss: 5.5048 (5.2380)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [7]  [110/156]  eta: 0:00:32  lr: 0.003083  min_lr: 0.003083  loss: 5.2972 (5.2413)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [7]  [120/156]  eta: 0:00:25  lr: 0.003109  min_lr: 0.003109  loss: 5.2620 (5.2350)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [7]  [130/156]  eta: 0:00:18  lr: 0.003134  min_lr: 0.003134  loss: 5.0427 (5.2177)  weight_decay: 0.0500 (0.0500)  time: 0.5933  data: 0.0011  max mem: 55573
Epoch: [7]  [140/156]  eta: 0:00:11  lr: 0.003160  min_lr: 0.003160  loss: 5.1017 (5.2059)  weight_decay: 0.0500 (0.0500)  time: 0.5885  data: 0.0009  max mem: 55573
Epoch: [7]  [150/156]  eta: 0:00:04  lr: 0.003186  min_lr: 0.003186  loss: 5.2094 (5.2031)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [7]  [155/156]  eta: 0:00:00  lr: 0.003198  min_lr: 0.003198  loss: 5.2094 (5.1929)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [7] Total time: 0:01:46 (0.6850 s / it)
Averaged stats: lr: 0.003198  min_lr: 0.003198  loss: 5.2094 (5.1540)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 2.5308 (2.5308)  acc1: 46.6146 (46.6146)  acc5: 72.1354 (72.1354)  time: 7.0156  data: 6.7783  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.5308 (2.5762)  acc1: 46.6146 (43.1360)  acc5: 72.1354 (67.5520)  time: 1.5495  data: 1.3557  max mem: 55573
Test: Total time: 0:00:07 (1.5873 s / it)
* Acc@1 43.032 Acc@5 69.178 loss 2.571
Accuracy of the model on the 50000 test images: 43.0%
Max accuracy: 43.03%
Test:  [0/5]  eta: 0:00:33  loss: 6.9113 (6.9113)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.7438  data: 6.5079  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9113 (6.9141)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.4941  data: 1.3017  max mem: 55573
Test: Total time: 0:00:07 (1.5116 s / it)
* Acc@1 0.096 Acc@5 0.492 loss 6.911
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [8]  [  0/156]  eta: 0:36:31  lr: 0.003201  min_lr: 0.003201  loss: 3.9776 (3.9776)  weight_decay: 0.0500 (0.0500)  time: 14.0467  data: 10.3489  max mem: 55573
Epoch: [8]  [ 10/156]  eta: 0:04:28  lr: 0.003227  min_lr: 0.003227  loss: 5.2855 (4.9991)  weight_decay: 0.0500 (0.0500)  time: 1.8366  data: 0.9411  max mem: 55573
Epoch: [8]  [ 20/156]  eta: 0:02:49  lr: 0.003252  min_lr: 0.003252  loss: 5.2855 (5.1360)  weight_decay: 0.0500 (0.0500)  time: 0.6030  data: 0.0004  max mem: 55573
Epoch: [8]  [ 30/156]  eta: 0:02:10  lr: 0.003278  min_lr: 0.003278  loss: 5.3199 (5.1471)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [8]  [ 40/156]  eta: 0:01:47  lr: 0.003304  min_lr: 0.003304  loss: 5.3709 (5.1700)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [8]  [ 50/156]  eta: 0:01:31  lr: 0.003329  min_lr: 0.003329  loss: 5.2418 (5.1474)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [8]  [ 60/156]  eta: 0:01:18  lr: 0.003355  min_lr: 0.003355  loss: 5.0524 (5.1170)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [8]  [ 70/156]  eta: 0:01:07  lr: 0.003381  min_lr: 0.003381  loss: 4.7696 (5.0684)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [8]  [ 80/156]  eta: 0:00:57  lr: 0.003406  min_lr: 0.003406  loss: 4.7772 (5.0540)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [8]  [ 90/156]  eta: 0:00:48  lr: 0.003432  min_lr: 0.003432  loss: 5.1843 (5.0541)  weight_decay: 0.0500 (0.0500)  time: 0.5887  data: 0.0004  max mem: 55573
Epoch: [8]  [100/156]  eta: 0:00:40  lr: 0.003458  min_lr: 0.003458  loss: 5.1790 (5.0553)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [8]  [110/156]  eta: 0:00:32  lr: 0.003483  min_lr: 0.003483  loss: 5.1538 (5.0691)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [8]  [120/156]  eta: 0:00:25  lr: 0.003509  min_lr: 0.003509  loss: 5.2927 (5.0823)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [8]  [130/156]  eta: 0:00:18  lr: 0.003534  min_lr: 0.003534  loss: 5.2927 (5.0901)  weight_decay: 0.0500 (0.0500)  time: 0.5936  data: 0.0011  max mem: 55573
Epoch: [8]  [140/156]  eta: 0:00:10  lr: 0.003560  min_lr: 0.003560  loss: 5.3431 (5.0917)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0009  max mem: 55573
Epoch: [8]  [150/156]  eta: 0:00:04  lr: 0.003586  min_lr: 0.003586  loss: 5.1436 (5.0813)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [8]  [155/156]  eta: 0:00:00  lr: 0.003599  min_lr: 0.003599  loss: 5.2154 (5.0908)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [8] Total time: 0:01:46 (0.6841 s / it)
Averaged stats: lr: 0.003599  min_lr: 0.003599  loss: 5.2154 (5.0060)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 2.3922 (2.3922)  acc1: 50.0000 (50.0000)  acc5: 76.4323 (76.4323)  time: 7.0623  data: 6.8249  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.4245 (2.4973)  acc1: 48.9583 (45.3120)  acc5: 76.3021 (70.9120)  time: 1.5588  data: 1.3651  max mem: 55573
Test: Total time: 0:00:08 (1.6021 s / it)
* Acc@1 45.262 Acc@5 71.348 loss 2.529
Accuracy of the model on the 50000 test images: 45.3%
Max accuracy: 45.26%
Test:  [0/5]  eta: 0:00:33  loss: 6.9110 (6.9110)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.7333  data: 6.4974  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9110 (6.9151)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.4920  data: 1.2996  max mem: 55573
Test: Total time: 0:00:07 (1.5077 s / it)
* Acc@1 0.096 Acc@5 0.496 loss 6.912
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [9]  [  0/156]  eta: 0:39:09  lr: 0.003601  min_lr: 0.003601  loss: 5.3790 (5.3790)  weight_decay: 0.0500 (0.0500)  time: 15.0603  data: 8.8702  max mem: 55573
Epoch: [9]  [ 10/156]  eta: 0:04:41  lr: 0.003627  min_lr: 0.003627  loss: 5.2249 (5.1180)  weight_decay: 0.0500 (0.0500)  time: 1.9259  data: 0.8068  max mem: 55573
Epoch: [9]  [ 20/156]  eta: 0:02:55  lr: 0.003652  min_lr: 0.003652  loss: 5.1351 (5.0197)  weight_decay: 0.0500 (0.0500)  time: 0.6031  data: 0.0004  max mem: 55573
Epoch: [9]  [ 30/156]  eta: 0:02:14  lr: 0.003678  min_lr: 0.003678  loss: 5.1196 (4.9613)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [9]  [ 40/156]  eta: 0:01:50  lr: 0.003704  min_lr: 0.003704  loss: 4.9508 (4.9587)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [9]  [ 50/156]  eta: 0:01:33  lr: 0.003729  min_lr: 0.003729  loss: 4.8207 (4.9487)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [9]  [ 60/156]  eta: 0:01:19  lr: 0.003755  min_lr: 0.003755  loss: 5.1701 (4.9918)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [9]  [ 70/156]  eta: 0:01:08  lr: 0.003781  min_lr: 0.003781  loss: 5.2281 (4.9582)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [9]  [ 80/156]  eta: 0:00:58  lr: 0.003806  min_lr: 0.003806  loss: 4.4322 (4.8905)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [9]  [ 90/156]  eta: 0:00:49  lr: 0.003832  min_lr: 0.003832  loss: 4.3604 (4.8560)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [9]  [100/156]  eta: 0:00:41  lr: 0.003858  min_lr: 0.003858  loss: 4.8846 (4.8675)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [9]  [110/156]  eta: 0:00:33  lr: 0.003883  min_lr: 0.003883  loss: 4.9952 (4.8802)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [9]  [120/156]  eta: 0:00:25  lr: 0.003909  min_lr: 0.003909  loss: 5.0464 (4.8972)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [9]  [130/156]  eta: 0:00:18  lr: 0.003935  min_lr: 0.003935  loss: 4.9664 (4.8771)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0010  max mem: 55573
Epoch: [9]  [140/156]  eta: 0:00:11  lr: 0.003960  min_lr: 0.003960  loss: 4.7380 (4.8683)  weight_decay: 0.0500 (0.0500)  time: 0.5871  data: 0.0009  max mem: 55573
Epoch: [9]  [150/156]  eta: 0:00:04  lr: 0.003986  min_lr: 0.003986  loss: 4.8673 (4.8644)  weight_decay: 0.0500 (0.0500)  time: 0.5829  data: 0.0001  max mem: 55573
Epoch: [9]  [155/156]  eta: 0:00:00  lr: 0.003999  min_lr: 0.003999  loss: 4.7244 (4.8625)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [9] Total time: 0:01:47 (0.6909 s / it)
Averaged stats: lr: 0.003999  min_lr: 0.003999  loss: 4.7244 (4.8817)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 2.1465 (2.1465)  acc1: 54.0365 (54.0365)  acc5: 79.6875 (79.6875)  time: 6.7304  data: 6.4934  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.1465 (2.2317)  acc1: 53.3854 (49.5680)  acc5: 79.6875 (75.0400)  time: 1.4926  data: 1.2988  max mem: 55573
Test: Total time: 0:00:07 (1.5313 s / it)
* Acc@1 50.574 Acc@5 75.910 loss 2.212
Accuracy of the model on the 50000 test images: 50.6%
Max accuracy: 50.57%
Test:  [0/5]  eta: 0:00:34  loss: 6.9105 (6.9105)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.8871  data: 6.6510  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9105 (6.9165)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5228  data: 1.3303  max mem: 55573
Test: Total time: 0:00:07 (1.5417 s / it)
* Acc@1 0.096 Acc@5 0.496 loss 6.913
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [10]  [  0/156]  eta: 0:40:51  lr: 0.004001  min_lr: 0.004001  loss: 3.7319 (3.7319)  weight_decay: 0.0500 (0.0500)  time: 15.7133  data: 7.3530  max mem: 55573
Epoch: [10]  [ 10/156]  eta: 0:04:46  lr: 0.004027  min_lr: 0.004027  loss: 4.5041 (4.4067)  weight_decay: 0.0500 (0.0500)  time: 1.9618  data: 0.6688  max mem: 55573
Epoch: [10]  [ 20/156]  eta: 0:02:58  lr: 0.004053  min_lr: 0.004053  loss: 4.6282 (4.5596)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0004  max mem: 55573
Epoch: [10]  [ 30/156]  eta: 0:02:15  lr: 0.004078  min_lr: 0.004078  loss: 4.9352 (4.7115)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [10]  [ 40/156]  eta: 0:01:51  lr: 0.004104  min_lr: 0.004104  loss: 5.0999 (4.8077)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [10]  [ 50/156]  eta: 0:01:33  lr: 0.004130  min_lr: 0.004130  loss: 5.0864 (4.8378)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [10]  [ 60/156]  eta: 0:01:20  lr: 0.004155  min_lr: 0.004155  loss: 4.8141 (4.8192)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [10]  [ 70/156]  eta: 0:01:09  lr: 0.004181  min_lr: 0.004181  loss: 4.9645 (4.8279)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [10]  [ 80/156]  eta: 0:00:59  lr: 0.004206  min_lr: 0.004206  loss: 4.9801 (4.8249)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [10]  [ 90/156]  eta: 0:00:49  lr: 0.004232  min_lr: 0.004232  loss: 4.5474 (4.7633)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [10]  [100/156]  eta: 0:00:41  lr: 0.004258  min_lr: 0.004258  loss: 4.1438 (4.7214)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [10]  [110/156]  eta: 0:00:33  lr: 0.004283  min_lr: 0.004283  loss: 4.5830 (4.7268)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [10]  [120/156]  eta: 0:00:25  lr: 0.004309  min_lr: 0.004309  loss: 4.9860 (4.7291)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0005  max mem: 55573
Epoch: [10]  [130/156]  eta: 0:00:18  lr: 0.004335  min_lr: 0.004335  loss: 4.9665 (4.7275)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0011  max mem: 55573
Epoch: [10]  [140/156]  eta: 0:00:11  lr: 0.004360  min_lr: 0.004360  loss: 5.0809 (4.7555)  weight_decay: 0.0500 (0.0500)  time: 0.5859  data: 0.0008  max mem: 55573
Epoch: [10]  [150/156]  eta: 0:00:04  lr: 0.004386  min_lr: 0.004386  loss: 5.0596 (4.7486)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [10]  [155/156]  eta: 0:00:00  lr: 0.004399  min_lr: 0.004399  loss: 4.8895 (4.7478)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [10] Total time: 0:01:48 (0.6925 s / it)
Averaged stats: lr: 0.004399  min_lr: 0.004399  loss: 4.8895 (4.7655)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.9734 (1.9734)  acc1: 59.6354 (59.6354)  acc5: 82.4219 (82.4219)  time: 6.8791  data: 6.6420  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.0561 (2.0826)  acc1: 56.3802 (53.5040)  acc5: 82.4219 (78.2720)  time: 1.5223  data: 1.3285  max mem: 55573
Test: Total time: 0:00:07 (1.5614 s / it)
* Acc@1 53.688 Acc@5 78.540 loss 2.102
Accuracy of the model on the 50000 test images: 53.7%
Max accuracy: 53.69%
Test:  [0/5]  eta: 0:00:34  loss: 6.9102 (6.9102)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9843  data: 6.7478  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9102 (6.9179)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5424  data: 1.3497  max mem: 55573
Test: Total time: 0:00:07 (1.5612 s / it)
* Acc@1 0.098 Acc@5 0.494 loss 6.914
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [11]  [  0/156]  eta: 0:40:41  lr: 0.004401  min_lr: 0.004401  loss: 4.9920 (4.9920)  weight_decay: 0.0500 (0.0500)  time: 15.6527  data: 9.1926  max mem: 55573
Epoch: [11]  [ 10/156]  eta: 0:04:45  lr: 0.004427  min_lr: 0.004427  loss: 4.8404 (4.5700)  weight_decay: 0.0500 (0.0500)  time: 1.9581  data: 0.8360  max mem: 55573
Epoch: [11]  [ 20/156]  eta: 0:02:57  lr: 0.004453  min_lr: 0.004453  loss: 4.8404 (4.6481)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [11]  [ 30/156]  eta: 0:02:16  lr: 0.004478  min_lr: 0.004478  loss: 4.9524 (4.6359)  weight_decay: 0.0500 (0.0500)  time: 0.5977  data: 0.0005  max mem: 55573
Epoch: [11]  [ 40/156]  eta: 0:01:51  lr: 0.004504  min_lr: 0.004504  loss: 4.4032 (4.5749)  weight_decay: 0.0500 (0.0500)  time: 0.5981  data: 0.0005  max mem: 55573
Epoch: [11]  [ 50/156]  eta: 0:01:34  lr: 0.004530  min_lr: 0.004530  loss: 4.7041 (4.6350)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [11]  [ 60/156]  eta: 0:01:20  lr: 0.004555  min_lr: 0.004555  loss: 4.9571 (4.6724)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [11]  [ 70/156]  eta: 0:01:09  lr: 0.004581  min_lr: 0.004581  loss: 4.7652 (4.6406)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [11]  [ 80/156]  eta: 0:00:59  lr: 0.004607  min_lr: 0.004607  loss: 4.8883 (4.6929)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [11]  [ 90/156]  eta: 0:00:49  lr: 0.004632  min_lr: 0.004632  loss: 4.9231 (4.6835)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [11]  [100/156]  eta: 0:00:41  lr: 0.004658  min_lr: 0.004658  loss: 4.6043 (4.6779)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [11]  [110/156]  eta: 0:00:33  lr: 0.004684  min_lr: 0.004684  loss: 4.7918 (4.6857)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [11]  [120/156]  eta: 0:00:25  lr: 0.004709  min_lr: 0.004709  loss: 4.8434 (4.6950)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [11]  [130/156]  eta: 0:00:18  lr: 0.004735  min_lr: 0.004735  loss: 4.7659 (4.6716)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0011  max mem: 55573
Epoch: [11]  [140/156]  eta: 0:00:11  lr: 0.004761  min_lr: 0.004761  loss: 4.5506 (4.6578)  weight_decay: 0.0500 (0.0500)  time: 0.5861  data: 0.0010  max mem: 55573
Epoch: [11]  [150/156]  eta: 0:00:04  lr: 0.004786  min_lr: 0.004786  loss: 4.5711 (4.6456)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [11]  [155/156]  eta: 0:00:00  lr: 0.004799  min_lr: 0.004799  loss: 4.5506 (4.6395)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [11] Total time: 0:01:48 (0.6934 s / it)
Averaged stats: lr: 0.004799  min_lr: 0.004799  loss: 4.5506 (4.6550)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.8946 (1.8946)  acc1: 59.7656 (59.7656)  acc5: 85.9375 (85.9375)  time: 7.1672  data: 6.9298  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.9879 (2.0284)  acc1: 57.5521 (53.4720)  acc5: 82.6823 (79.3280)  time: 1.5799  data: 1.3860  max mem: 55573
Test: Total time: 0:00:08 (1.6224 s / it)
* Acc@1 54.814 Acc@5 79.254 loss 2.030
Accuracy of the model on the 50000 test images: 54.8%
Max accuracy: 54.81%
Test:  [0/5]  eta: 0:00:34  loss: 6.9098 (6.9098)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9525  data: 6.7166  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9098 (6.9193)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5358  data: 1.3434  max mem: 55573
Test: Total time: 0:00:07 (1.5570 s / it)
* Acc@1 0.100 Acc@5 0.496 loss 6.915
Accuracy of the model EMA on 50000 test images: 0.1%
Max EMA accuracy: 0.10%
Epoch: [12]  [  0/156]  eta: 0:32:24  lr: 0.004802  min_lr: 0.004802  loss: 5.4818 (5.4818)  weight_decay: 0.0500 (0.0500)  time: 12.4618  data: 7.7381  max mem: 55573
Epoch: [12]  [ 10/156]  eta: 0:04:04  lr: 0.004827  min_lr: 0.004827  loss: 5.0961 (4.8765)  weight_decay: 0.0500 (0.0500)  time: 1.6717  data: 0.7038  max mem: 55573
Epoch: [12]  [ 20/156]  eta: 0:02:37  lr: 0.004853  min_lr: 0.004853  loss: 4.8127 (4.7189)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [12]  [ 30/156]  eta: 0:02:02  lr: 0.004878  min_lr: 0.004878  loss: 4.6095 (4.6540)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [12]  [ 40/156]  eta: 0:01:42  lr: 0.004904  min_lr: 0.004904  loss: 4.6095 (4.6423)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0004  max mem: 55573
Epoch: [12]  [ 50/156]  eta: 0:01:27  lr: 0.004930  min_lr: 0.004930  loss: 4.3094 (4.5428)  weight_decay: 0.0500 (0.0500)  time: 0.5932  data: 0.0004  max mem: 55573
Epoch: [12]  [ 60/156]  eta: 0:01:15  lr: 0.004955  min_lr: 0.004955  loss: 4.2687 (4.5175)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [12]  [ 70/156]  eta: 0:01:05  lr: 0.004981  min_lr: 0.004981  loss: 4.5342 (4.5177)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0004  max mem: 55573
Epoch: [12]  [ 80/156]  eta: 0:00:56  lr: 0.005007  min_lr: 0.005007  loss: 4.5342 (4.4892)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [12]  [ 90/156]  eta: 0:00:47  lr: 0.005032  min_lr: 0.005032  loss: 4.5729 (4.4981)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [12]  [100/156]  eta: 0:00:39  lr: 0.005058  min_lr: 0.005058  loss: 4.6141 (4.4839)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [12]  [110/156]  eta: 0:00:32  lr: 0.005084  min_lr: 0.005084  loss: 4.4440 (4.4840)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [12]  [120/156]  eta: 0:00:24  lr: 0.005109  min_lr: 0.005109  loss: 4.5360 (4.4840)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [12]  [130/156]  eta: 0:00:17  lr: 0.005135  min_lr: 0.005135  loss: 4.6980 (4.4880)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0010  max mem: 55573
Epoch: [12]  [140/156]  eta: 0:00:10  lr: 0.005161  min_lr: 0.005161  loss: 4.7625 (4.5110)  weight_decay: 0.0500 (0.0500)  time: 0.5874  data: 0.0009  max mem: 55573
Epoch: [12]  [150/156]  eta: 0:00:04  lr: 0.005186  min_lr: 0.005186  loss: 4.8243 (4.5148)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [12]  [155/156]  eta: 0:00:00  lr: 0.005199  min_lr: 0.005199  loss: 4.5914 (4.5176)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [12] Total time: 0:01:44 (0.6728 s / it)
Averaged stats: lr: 0.005199  min_lr: 0.005199  loss: 4.5914 (4.5784)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.8941 (1.8941)  acc1: 61.5885 (61.5885)  acc5: 86.1979 (86.1979)  time: 6.9301  data: 6.6929  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.0881 (2.1031)  acc1: 57.4219 (55.7760)  acc5: 82.5521 (79.7440)  time: 1.5326  data: 1.3387  max mem: 55573
Test: Total time: 0:00:07 (1.5749 s / it)
* Acc@1 55.166 Acc@5 79.774 loss 2.074
Accuracy of the model on the 50000 test images: 55.2%
Max accuracy: 55.17%
Test:  [0/5]  eta: 0:00:32  loss: 6.9092 (6.9092)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.5912  data: 6.3552  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9092 (6.9207)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.4636  data: 1.2712  max mem: 55573
Test: Total time: 0:00:07 (1.4794 s / it)
* Acc@1 0.100 Acc@5 0.498 loss 6.915
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [13]  [  0/156]  eta: 0:33:41  lr: 0.005202  min_lr: 0.005202  loss: 4.8295 (4.8295)  weight_decay: 0.0500 (0.0500)  time: 12.9583  data: 11.1965  max mem: 55573
Epoch: [13]  [ 10/156]  eta: 0:04:10  lr: 0.005227  min_lr: 0.005227  loss: 4.9281 (4.7366)  weight_decay: 0.0500 (0.0500)  time: 1.7173  data: 1.0182  max mem: 55573
Epoch: [13]  [ 20/156]  eta: 0:02:40  lr: 0.005253  min_lr: 0.005253  loss: 4.6482 (4.5791)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [13]  [ 30/156]  eta: 0:02:04  lr: 0.005279  min_lr: 0.005279  loss: 4.4575 (4.5306)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [13]  [ 40/156]  eta: 0:01:43  lr: 0.005304  min_lr: 0.005304  loss: 4.3990 (4.5171)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [13]  [ 50/156]  eta: 0:01:28  lr: 0.005330  min_lr: 0.005330  loss: 4.5699 (4.5514)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [13]  [ 60/156]  eta: 0:01:16  lr: 0.005356  min_lr: 0.005356  loss: 4.7034 (4.5576)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [13]  [ 70/156]  eta: 0:01:05  lr: 0.005381  min_lr: 0.005381  loss: 4.7335 (4.5455)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0005  max mem: 55573
Epoch: [13]  [ 80/156]  eta: 0:00:56  lr: 0.005407  min_lr: 0.005407  loss: 4.5970 (4.5395)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0005  max mem: 55573
Epoch: [13]  [ 90/156]  eta: 0:00:48  lr: 0.005433  min_lr: 0.005433  loss: 4.5970 (4.5359)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [13]  [100/156]  eta: 0:00:39  lr: 0.005458  min_lr: 0.005458  loss: 4.4146 (4.5277)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [13]  [110/156]  eta: 0:00:32  lr: 0.005484  min_lr: 0.005484  loss: 4.7969 (4.5557)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [13]  [120/156]  eta: 0:00:24  lr: 0.005509  min_lr: 0.005509  loss: 4.8554 (4.5487)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [13]  [130/156]  eta: 0:00:17  lr: 0.005535  min_lr: 0.005535  loss: 4.7270 (4.5568)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0010  max mem: 55573
Epoch: [13]  [140/156]  eta: 0:00:10  lr: 0.005561  min_lr: 0.005561  loss: 4.8001 (4.5562)  weight_decay: 0.0500 (0.0500)  time: 0.5881  data: 0.0008  max mem: 55573
Epoch: [13]  [150/156]  eta: 0:00:04  lr: 0.005586  min_lr: 0.005586  loss: 4.7131 (4.5506)  weight_decay: 0.0500 (0.0500)  time: 0.5830  data: 0.0001  max mem: 55573
Epoch: [13]  [155/156]  eta: 0:00:00  lr: 0.005599  min_lr: 0.005599  loss: 4.6612 (4.5415)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [13] Total time: 0:01:45 (0.6760 s / it)
Averaged stats: lr: 0.005599  min_lr: 0.005599  loss: 4.6612 (4.5049)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.9438 (1.9438)  acc1: 58.9844 (58.9844)  acc5: 84.7656 (84.7656)  time: 7.2876  data: 7.0502  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.0056 (2.0698)  acc1: 55.5990 (53.4080)  acc5: 81.6406 (78.6240)  time: 1.6041  data: 1.4101  max mem: 55573
Test: Total time: 0:00:08 (1.6429 s / it)
* Acc@1 54.470 Acc@5 79.108 loss 2.058
Accuracy of the model on the 50000 test images: 54.5%
Max accuracy: 55.17%
Test:  [0/5]  eta: 0:00:34  loss: 6.9084 (6.9084)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9898  data: 6.7538  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9084 (6.9223)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5433  data: 1.3508  max mem: 55573
Test: Total time: 0:00:07 (1.5593 s / it)
* Acc@1 0.100 Acc@5 0.502 loss 6.916
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [14]  [  0/156]  eta: 0:33:51  lr: 0.005602  min_lr: 0.005602  loss: 4.9298 (4.9298)  weight_decay: 0.0500 (0.0500)  time: 13.0222  data: 10.1353  max mem: 55573
Epoch: [14]  [ 10/156]  eta: 0:04:17  lr: 0.005627  min_lr: 0.005627  loss: 4.9298 (4.7823)  weight_decay: 0.0500 (0.0500)  time: 1.7662  data: 0.9218  max mem: 55573
Epoch: [14]  [ 20/156]  eta: 0:02:44  lr: 0.005653  min_lr: 0.005653  loss: 4.5458 (4.4764)  weight_decay: 0.0500 (0.0500)  time: 0.6157  data: 0.0004  max mem: 55573
Epoch: [14]  [ 30/156]  eta: 0:02:06  lr: 0.005679  min_lr: 0.005679  loss: 4.2501 (4.4756)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [14]  [ 40/156]  eta: 0:01:45  lr: 0.005704  min_lr: 0.005704  loss: 4.2094 (4.3705)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [14]  [ 50/156]  eta: 0:01:29  lr: 0.005730  min_lr: 0.005730  loss: 4.2094 (4.3555)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [14]  [ 60/156]  eta: 0:01:17  lr: 0.005756  min_lr: 0.005756  loss: 4.2787 (4.3528)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [14]  [ 70/156]  eta: 0:01:06  lr: 0.005781  min_lr: 0.005781  loss: 4.2403 (4.3534)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [14]  [ 80/156]  eta: 0:00:57  lr: 0.005807  min_lr: 0.005807  loss: 4.5974 (4.3884)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [14]  [ 90/156]  eta: 0:00:48  lr: 0.005833  min_lr: 0.005833  loss: 4.6873 (4.4109)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [14]  [100/156]  eta: 0:00:40  lr: 0.005858  min_lr: 0.005858  loss: 4.4425 (4.3954)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [14]  [110/156]  eta: 0:00:32  lr: 0.005884  min_lr: 0.005884  loss: 4.3789 (4.3957)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [14]  [120/156]  eta: 0:00:25  lr: 0.005910  min_lr: 0.005910  loss: 4.4347 (4.3918)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [14]  [130/156]  eta: 0:00:17  lr: 0.005935  min_lr: 0.005935  loss: 4.5331 (4.3972)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0011  max mem: 55573
Epoch: [14]  [140/156]  eta: 0:00:10  lr: 0.005961  min_lr: 0.005961  loss: 4.6499 (4.4161)  weight_decay: 0.0500 (0.0500)  time: 0.5872  data: 0.0009  max mem: 55573
Epoch: [14]  [150/156]  eta: 0:00:04  lr: 0.005987  min_lr: 0.005987  loss: 4.6891 (4.4279)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0001  max mem: 55573
Epoch: [14]  [155/156]  eta: 0:00:00  lr: 0.005999  min_lr: 0.005999  loss: 4.6891 (4.4232)  weight_decay: 0.0500 (0.0500)  time: 0.5830  data: 0.0001  max mem: 55573
Epoch: [14] Total time: 0:01:45 (0.6788 s / it)
Averaged stats: lr: 0.005999  min_lr: 0.005999  loss: 4.6891 (4.4297)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.7202 (1.7202)  acc1: 63.2812 (63.2812)  acc5: 86.0677 (86.0677)  time: 7.1056  data: 6.8685  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.8811 (1.8745)  acc1: 61.5885 (57.5680)  acc5: 84.2448 (81.4400)  time: 1.5675  data: 1.3738  max mem: 55573
Test: Total time: 0:00:08 (1.6045 s / it)
* Acc@1 56.956 Acc@5 81.802 loss 1.895
Accuracy of the model on the 50000 test images: 57.0%
Max accuracy: 56.96%
Test:  [0/5]  eta: 0:00:34  loss: 6.9075 (6.9075)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 6.9895  data: 6.7530  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9075 (6.9239)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5436  data: 1.3507  max mem: 55573
Test: Total time: 0:00:07 (1.5649 s / it)
* Acc@1 0.100 Acc@5 0.500 loss 6.917
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [15]  [  0/156]  eta: 0:36:01  lr: 0.006002  min_lr: 0.006002  loss: 4.4531 (4.4531)  weight_decay: 0.0500 (0.0500)  time: 13.8545  data: 8.8830  max mem: 55573
Epoch: [15]  [ 10/156]  eta: 0:04:26  lr: 0.006028  min_lr: 0.006028  loss: 4.3697 (4.4018)  weight_decay: 0.0500 (0.0500)  time: 1.8235  data: 0.8078  max mem: 55573
Epoch: [15]  [ 20/156]  eta: 0:02:48  lr: 0.006053  min_lr: 0.006053  loss: 4.2625 (4.3161)  weight_decay: 0.0500 (0.0500)  time: 0.6074  data: 0.0004  max mem: 55573
Epoch: [15]  [ 30/156]  eta: 0:02:09  lr: 0.006079  min_lr: 0.006079  loss: 4.2625 (4.3178)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [15]  [ 40/156]  eta: 0:01:46  lr: 0.006105  min_lr: 0.006105  loss: 4.4810 (4.3690)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [15]  [ 50/156]  eta: 0:01:30  lr: 0.006130  min_lr: 0.006130  loss: 4.4962 (4.3360)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [15]  [ 60/156]  eta: 0:01:18  lr: 0.006156  min_lr: 0.006156  loss: 4.1339 (4.3345)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0004  max mem: 55573
Epoch: [15]  [ 70/156]  eta: 0:01:07  lr: 0.006181  min_lr: 0.006181  loss: 4.6899 (4.3772)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0005  max mem: 55573
Epoch: [15]  [ 80/156]  eta: 0:00:57  lr: 0.006207  min_lr: 0.006207  loss: 4.4615 (4.3560)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [15]  [ 90/156]  eta: 0:00:48  lr: 0.006233  min_lr: 0.006233  loss: 4.1649 (4.3670)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0005  max mem: 55573
Epoch: [15]  [100/156]  eta: 0:00:40  lr: 0.006258  min_lr: 0.006258  loss: 4.5674 (4.3777)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [15]  [110/156]  eta: 0:00:32  lr: 0.006284  min_lr: 0.006284  loss: 4.5674 (4.3975)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [15]  [120/156]  eta: 0:00:25  lr: 0.006310  min_lr: 0.006310  loss: 4.7581 (4.4063)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [15]  [130/156]  eta: 0:00:18  lr: 0.006335  min_lr: 0.006335  loss: 4.7581 (4.3999)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0010  max mem: 55573
Epoch: [15]  [140/156]  eta: 0:00:10  lr: 0.006361  min_lr: 0.006361  loss: 4.1564 (4.3877)  weight_decay: 0.0500 (0.0500)  time: 0.5859  data: 0.0009  max mem: 55573
Epoch: [15]  [150/156]  eta: 0:00:04  lr: 0.006387  min_lr: 0.006387  loss: 4.3894 (4.4038)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [15]  [155/156]  eta: 0:00:00  lr: 0.006399  min_lr: 0.006399  loss: 4.3894 (4.3927)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [15] Total time: 0:01:46 (0.6825 s / it)
Averaged stats: lr: 0.006399  min_lr: 0.006399  loss: 4.3894 (4.4087)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.6790 (1.6790)  acc1: 63.2812 (63.2812)  acc5: 86.3281 (86.3281)  time: 7.3148  data: 7.0773  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.8268 (1.9031)  acc1: 59.5052 (56.4160)  acc5: 84.2448 (81.3440)  time: 1.6095  data: 1.4155  max mem: 55573
Test: Total time: 0:00:08 (1.6484 s / it)
* Acc@1 57.588 Acc@5 81.714 loss 1.864
Accuracy of the model on the 50000 test images: 57.6%
Max accuracy: 57.59%
Test:  [0/5]  eta: 0:00:34  loss: 6.9065 (6.9065)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 6.9628  data: 6.7262  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9065 (6.9255)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5382  data: 1.3454  max mem: 55573
Test: Total time: 0:00:07 (1.5564 s / it)
* Acc@1 0.102 Acc@5 0.500 loss 6.918
Accuracy of the model EMA on 50000 test images: 0.1%
Max EMA accuracy: 0.10%
Epoch: [16]  [  0/156]  eta: 0:36:31  lr: 0.006402  min_lr: 0.006402  loss: 3.7989 (3.7989)  weight_decay: 0.0500 (0.0500)  time: 14.0510  data: 10.4540  max mem: 55573
Epoch: [16]  [ 10/156]  eta: 0:04:26  lr: 0.006428  min_lr: 0.006428  loss: 3.9863 (4.1661)  weight_decay: 0.0500 (0.0500)  time: 1.8251  data: 0.9508  max mem: 55573
Epoch: [16]  [ 20/156]  eta: 0:02:48  lr: 0.006453  min_lr: 0.006453  loss: 4.3949 (4.3415)  weight_decay: 0.0500 (0.0500)  time: 0.5971  data: 0.0004  max mem: 55573
Epoch: [16]  [ 30/156]  eta: 0:02:09  lr: 0.006479  min_lr: 0.006479  loss: 4.5623 (4.4157)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [16]  [ 40/156]  eta: 0:01:46  lr: 0.006505  min_lr: 0.006505  loss: 4.6544 (4.4226)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [16]  [ 50/156]  eta: 0:01:30  lr: 0.006530  min_lr: 0.006530  loss: 4.3384 (4.4027)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [16]  [ 60/156]  eta: 0:01:18  lr: 0.006556  min_lr: 0.006556  loss: 4.5961 (4.4418)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [16]  [ 70/156]  eta: 0:01:07  lr: 0.006582  min_lr: 0.006582  loss: 4.6327 (4.4690)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0006  max mem: 55573
Epoch: [16]  [ 80/156]  eta: 0:00:57  lr: 0.006607  min_lr: 0.006607  loss: 4.4882 (4.4607)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0005  max mem: 55573
Epoch: [16]  [ 90/156]  eta: 0:00:48  lr: 0.006633  min_lr: 0.006633  loss: 4.4786 (4.4594)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [16]  [100/156]  eta: 0:00:40  lr: 0.006659  min_lr: 0.006659  loss: 4.3934 (4.4387)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [16]  [110/156]  eta: 0:00:32  lr: 0.006684  min_lr: 0.006684  loss: 4.3934 (4.4399)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [16]  [120/156]  eta: 0:00:25  lr: 0.006710  min_lr: 0.006710  loss: 4.5651 (4.4230)  weight_decay: 0.0500 (0.0500)  time: 0.5934  data: 0.0004  max mem: 55573
Epoch: [16]  [130/156]  eta: 0:00:18  lr: 0.006735  min_lr: 0.006735  loss: 4.3591 (4.4200)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0010  max mem: 55573
Epoch: [16]  [140/156]  eta: 0:00:10  lr: 0.006761  min_lr: 0.006761  loss: 4.3591 (4.4197)  weight_decay: 0.0500 (0.0500)  time: 0.5869  data: 0.0008  max mem: 55573
Epoch: [16]  [150/156]  eta: 0:00:04  lr: 0.006787  min_lr: 0.006787  loss: 4.6138 (4.4370)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [16]  [155/156]  eta: 0:00:00  lr: 0.006800  min_lr: 0.006800  loss: 4.6472 (4.4190)  weight_decay: 0.0500 (0.0500)  time: 0.5828  data: 0.0001  max mem: 55573
Epoch: [16] Total time: 0:01:46 (0.6835 s / it)
Averaged stats: lr: 0.006800  min_lr: 0.006800  loss: 4.6472 (4.3548)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.7706 (1.7706)  acc1: 65.4948 (65.4948)  acc5: 86.3281 (86.3281)  time: 6.7398  data: 6.5027  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.9889 (1.9403)  acc1: 57.5521 (56.6720)  acc5: 82.9427 (81.0240)  time: 1.4932  data: 1.3006  max mem: 55573
Test: Total time: 0:00:07 (1.5098 s / it)
* Acc@1 56.358 Acc@5 80.818 loss 1.954
Accuracy of the model on the 50000 test images: 56.4%
Max accuracy: 57.59%
Test:  [0/5]  eta: 0:00:35  loss: 6.9056 (6.9056)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 7.1035  data: 6.8670  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9063 (6.9270)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5661  data: 1.3735  max mem: 55573
Test: Total time: 0:00:07 (1.5888 s / it)
* Acc@1 0.102 Acc@5 0.502 loss 6.919
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [17]  [  0/156]  eta: 0:37:40  lr: 0.006802  min_lr: 0.006802  loss: 3.3034 (3.3034)  weight_decay: 0.0500 (0.0500)  time: 14.4913  data: 11.5777  max mem: 55573
Epoch: [17]  [ 10/156]  eta: 0:04:30  lr: 0.006828  min_lr: 0.006828  loss: 4.6115 (4.3792)  weight_decay: 0.0500 (0.0500)  time: 1.8546  data: 1.0528  max mem: 55573
Epoch: [17]  [ 20/156]  eta: 0:02:50  lr: 0.006853  min_lr: 0.006853  loss: 4.5655 (4.3761)  weight_decay: 0.0500 (0.0500)  time: 0.5939  data: 0.0003  max mem: 55573
Epoch: [17]  [ 30/156]  eta: 0:02:11  lr: 0.006879  min_lr: 0.006879  loss: 4.6254 (4.4206)  weight_decay: 0.0500 (0.0500)  time: 0.5939  data: 0.0003  max mem: 55573
Epoch: [17]  [ 40/156]  eta: 0:01:48  lr: 0.006905  min_lr: 0.006905  loss: 4.6500 (4.4193)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0004  max mem: 55573
Epoch: [17]  [ 50/156]  eta: 0:01:31  lr: 0.006930  min_lr: 0.006930  loss: 4.4969 (4.4496)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0004  max mem: 55573
Epoch: [17]  [ 60/156]  eta: 0:01:18  lr: 0.006956  min_lr: 0.006956  loss: 4.4966 (4.4727)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [17]  [ 70/156]  eta: 0:01:07  lr: 0.006982  min_lr: 0.006982  loss: 4.5452 (4.4453)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [17]  [ 80/156]  eta: 0:00:58  lr: 0.007007  min_lr: 0.007007  loss: 4.6041 (4.4731)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [17]  [ 90/156]  eta: 0:00:49  lr: 0.007033  min_lr: 0.007033  loss: 4.2574 (4.4108)  weight_decay: 0.0500 (0.0500)  time: 0.5938  data: 0.0005  max mem: 55573
Epoch: [17]  [100/156]  eta: 0:00:40  lr: 0.007059  min_lr: 0.007059  loss: 4.0771 (4.3906)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0004  max mem: 55573
Epoch: [17]  [110/156]  eta: 0:00:32  lr: 0.007084  min_lr: 0.007084  loss: 4.4569 (4.3821)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0004  max mem: 55573
Epoch: [17]  [120/156]  eta: 0:00:25  lr: 0.007110  min_lr: 0.007110  loss: 4.6497 (4.3957)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [17]  [130/156]  eta: 0:00:18  lr: 0.007136  min_lr: 0.007136  loss: 4.6227 (4.3841)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0010  max mem: 55573
Epoch: [17]  [140/156]  eta: 0:00:11  lr: 0.007161  min_lr: 0.007161  loss: 3.9257 (4.3530)  weight_decay: 0.0500 (0.0500)  time: 0.5856  data: 0.0009  max mem: 55573
Epoch: [17]  [150/156]  eta: 0:00:04  lr: 0.007187  min_lr: 0.007187  loss: 4.0566 (4.3473)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [17]  [155/156]  eta: 0:00:00  lr: 0.007200  min_lr: 0.007200  loss: 4.4000 (4.3475)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [17] Total time: 0:01:47 (0.6863 s / it)
Averaged stats: lr: 0.007200  min_lr: 0.007200  loss: 4.4000 (4.3042)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.8214 (1.8214)  acc1: 62.8906 (62.8906)  acc5: 86.4583 (86.4583)  time: 7.1516  data: 6.9132  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.9293 (2.0369)  acc1: 59.3750 (57.2800)  acc5: 84.5052 (80.9920)  time: 1.5768  data: 1.3827  max mem: 55573
Test: Total time: 0:00:08 (1.6139 s / it)
* Acc@1 57.308 Acc@5 81.436 loss 2.021
Accuracy of the model on the 50000 test images: 57.3%
Max accuracy: 57.59%
Test:  [0/5]  eta: 0:00:35  loss: 6.9046 (6.9046)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 7.0647  data: 6.8282  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9073 (6.9285)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5584  data: 1.3657  max mem: 55573
Test: Total time: 0:00:07 (1.5772 s / it)
* Acc@1 0.102 Acc@5 0.502 loss 6.920
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [18]  [  0/156]  eta: 0:35:08  lr: 0.007202  min_lr: 0.007202  loss: 4.3485 (4.3485)  weight_decay: 0.0500 (0.0500)  time: 13.5171  data: 10.2658  max mem: 55573
Epoch: [18]  [ 10/156]  eta: 0:04:18  lr: 0.007228  min_lr: 0.007228  loss: 4.6377 (4.3115)  weight_decay: 0.0500 (0.0500)  time: 1.7689  data: 0.9336  max mem: 55573
Epoch: [18]  [ 20/156]  eta: 0:02:44  lr: 0.007254  min_lr: 0.007254  loss: 4.5255 (4.2825)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [18]  [ 30/156]  eta: 0:02:07  lr: 0.007279  min_lr: 0.007279  loss: 4.2014 (4.2508)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [18]  [ 40/156]  eta: 0:01:45  lr: 0.007305  min_lr: 0.007305  loss: 4.1171 (4.2020)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [18]  [ 50/156]  eta: 0:01:29  lr: 0.007331  min_lr: 0.007331  loss: 3.7395 (4.1226)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0004  max mem: 55573
Epoch: [18]  [ 60/156]  eta: 0:01:17  lr: 0.007356  min_lr: 0.007356  loss: 3.8157 (4.1117)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0004  max mem: 55573
Epoch: [18]  [ 70/156]  eta: 0:01:06  lr: 0.007382  min_lr: 0.007382  loss: 4.0458 (4.1072)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [18]  [ 80/156]  eta: 0:00:57  lr: 0.007408  min_lr: 0.007408  loss: 4.0472 (4.0937)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [18]  [ 90/156]  eta: 0:00:48  lr: 0.007433  min_lr: 0.007433  loss: 4.4677 (4.1452)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [18]  [100/156]  eta: 0:00:40  lr: 0.007459  min_lr: 0.007459  loss: 4.4824 (4.1368)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [18]  [110/156]  eta: 0:00:32  lr: 0.007484  min_lr: 0.007484  loss: 4.3811 (4.1580)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [18]  [120/156]  eta: 0:00:25  lr: 0.007510  min_lr: 0.007510  loss: 4.5194 (4.1741)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [18]  [130/156]  eta: 0:00:17  lr: 0.007536  min_lr: 0.007536  loss: 4.5667 (4.2034)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0010  max mem: 55573
Epoch: [18]  [140/156]  eta: 0:00:10  lr: 0.007561  min_lr: 0.007561  loss: 4.3559 (4.1907)  weight_decay: 0.0500 (0.0500)  time: 0.5858  data: 0.0009  max mem: 55573
Epoch: [18]  [150/156]  eta: 0:00:04  lr: 0.007587  min_lr: 0.007587  loss: 4.0900 (4.1884)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [18]  [155/156]  eta: 0:00:00  lr: 0.007600  min_lr: 0.007600  loss: 4.0900 (4.1893)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [18] Total time: 0:01:46 (0.6797 s / it)
Averaged stats: lr: 0.007600  min_lr: 0.007600  loss: 4.0900 (4.2735)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.5740 (1.5740)  acc1: 63.0208 (63.0208)  acc5: 88.2812 (88.2812)  time: 7.0190  data: 6.7816  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.7820 (1.8589)  acc1: 61.7188 (58.3680)  acc5: 85.8073 (82.0480)  time: 1.5504  data: 1.3564  max mem: 55573
Test: Total time: 0:00:07 (1.5894 s / it)
* Acc@1 58.350 Acc@5 82.556 loss 1.874
Accuracy of the model on the 50000 test images: 58.4%
Max accuracy: 58.35%
Test:  [0/5]  eta: 0:00:32  loss: 6.9037 (6.9037)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 6.5158  data: 6.2791  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9084 (6.9300)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.4761  data: 1.2834  max mem: 55573
Test: Total time: 0:00:07 (1.4916 s / it)
* Acc@1 0.104 Acc@5 0.502 loss 6.921
Accuracy of the model EMA on 50000 test images: 0.1%
Max EMA accuracy: 0.10%
Epoch: [19]  [  0/156]  eta: 0:32:58  lr: 0.007602  min_lr: 0.007602  loss: 3.7124 (3.7124)  weight_decay: 0.0500 (0.0500)  time: 12.6842  data: 11.5792  max mem: 55573
Epoch: [19]  [ 10/156]  eta: 0:04:12  lr: 0.007628  min_lr: 0.007628  loss: 4.5908 (4.3943)  weight_decay: 0.0500 (0.0500)  time: 1.7271  data: 1.0531  max mem: 55573
Epoch: [19]  [ 20/156]  eta: 0:02:41  lr: 0.007654  min_lr: 0.007654  loss: 4.3464 (4.1883)  weight_decay: 0.0500 (0.0500)  time: 0.6109  data: 0.0005  max mem: 55573
Epoch: [19]  [ 30/156]  eta: 0:02:05  lr: 0.007679  min_lr: 0.007679  loss: 4.3464 (4.2723)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0005  max mem: 55573
Epoch: [19]  [ 40/156]  eta: 0:01:43  lr: 0.007705  min_lr: 0.007705  loss: 4.4619 (4.2709)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [19]  [ 50/156]  eta: 0:01:28  lr: 0.007731  min_lr: 0.007731  loss: 4.2020 (4.2618)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0004  max mem: 55573
Epoch: [19]  [ 60/156]  eta: 0:01:16  lr: 0.007756  min_lr: 0.007756  loss: 4.2020 (4.2401)  weight_decay: 0.0500 (0.0500)  time: 0.5941  data: 0.0005  max mem: 55573
Epoch: [19]  [ 70/156]  eta: 0:01:06  lr: 0.007782  min_lr: 0.007782  loss: 4.3172 (4.2408)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [19]  [ 80/156]  eta: 0:00:56  lr: 0.007808  min_lr: 0.007808  loss: 4.3253 (4.2396)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [19]  [ 90/156]  eta: 0:00:48  lr: 0.007833  min_lr: 0.007833  loss: 4.4804 (4.2495)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [19]  [100/156]  eta: 0:00:40  lr: 0.007859  min_lr: 0.007859  loss: 4.2002 (4.2243)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [19]  [110/156]  eta: 0:00:32  lr: 0.007885  min_lr: 0.007885  loss: 4.2002 (4.2231)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [19]  [120/156]  eta: 0:00:25  lr: 0.007910  min_lr: 0.007910  loss: 4.2540 (4.2029)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [19]  [130/156]  eta: 0:00:17  lr: 0.007936  min_lr: 0.007936  loss: 4.3281 (4.2134)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0012  max mem: 55573
Epoch: [19]  [140/156]  eta: 0:00:10  lr: 0.007962  min_lr: 0.007962  loss: 4.3180 (4.1982)  weight_decay: 0.0500 (0.0500)  time: 0.5860  data: 0.0011  max mem: 55573
Epoch: [19]  [150/156]  eta: 0:00:04  lr: 0.007987  min_lr: 0.007987  loss: 4.1359 (4.2023)  weight_decay: 0.0500 (0.0500)  time: 0.5829  data: 0.0001  max mem: 55573
Epoch: [19]  [155/156]  eta: 0:00:00  lr: 0.008000  min_lr: 0.008000  loss: 4.1359 (4.2114)  weight_decay: 0.0500 (0.0500)  time: 0.5828  data: 0.0001  max mem: 55573
Epoch: [19] Total time: 0:01:45 (0.6769 s / it)
Averaged stats: lr: 0.008000  min_lr: 0.008000  loss: 4.1359 (4.2360)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.8779 (1.8779)  acc1: 59.8958 (59.8958)  acc5: 85.0260 (85.0260)  time: 7.1843  data: 6.9468  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.9301 (1.9863)  acc1: 59.5052 (55.9360)  acc5: 85.0260 (80.6080)  time: 1.5834  data: 1.3894  max mem: 55573
Test: Total time: 0:00:08 (1.6262 s / it)
* Acc@1 56.212 Acc@5 80.678 loss 1.996
Accuracy of the model on the 50000 test images: 56.2%
Max accuracy: 58.35%
Test:  [0/5]  eta: 0:00:36  loss: 6.9032 (6.9032)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 7.3056  data: 7.0692  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9095 (6.9314)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.6067  data: 1.4139  max mem: 55573
Test: Total time: 0:00:08 (1.6292 s / it)
* Acc@1 0.104 Acc@5 0.506 loss 6.922
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [20]  [  0/156]  eta: 0:35:56  lr: 0.008000  min_lr: 0.008000  loss: 4.2855 (4.2855)  weight_decay: 0.0500 (0.0500)  time: 13.8216  data: 8.9870  max mem: 55573
Epoch: [20]  [ 10/156]  eta: 0:04:24  lr: 0.008000  min_lr: 0.008000  loss: 4.2176 (4.2147)  weight_decay: 0.0500 (0.0500)  time: 1.8127  data: 0.8174  max mem: 55573
Epoch: [20]  [ 20/156]  eta: 0:02:47  lr: 0.008000  min_lr: 0.008000  loss: 4.0041 (4.0595)  weight_decay: 0.0500 (0.0500)  time: 0.6004  data: 0.0004  max mem: 55573
Epoch: [20]  [ 30/156]  eta: 0:02:08  lr: 0.008000  min_lr: 0.008000  loss: 3.9495 (4.1045)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [20]  [ 40/156]  eta: 0:01:46  lr: 0.008000  min_lr: 0.008000  loss: 4.3076 (4.1069)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [20]  [ 50/156]  eta: 0:01:30  lr: 0.008000  min_lr: 0.008000  loss: 4.3551 (4.1390)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [20]  [ 60/156]  eta: 0:01:17  lr: 0.008000  min_lr: 0.008000  loss: 4.3631 (4.1593)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [20]  [ 70/156]  eta: 0:01:07  lr: 0.008000  min_lr: 0.008000  loss: 4.1741 (4.1359)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [20]  [ 80/156]  eta: 0:00:57  lr: 0.008000  min_lr: 0.008000  loss: 4.2377 (4.1637)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [20]  [ 90/156]  eta: 0:00:48  lr: 0.008000  min_lr: 0.008000  loss: 4.5812 (4.1745)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [20]  [100/156]  eta: 0:00:40  lr: 0.008000  min_lr: 0.008000  loss: 4.5812 (4.1909)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [20]  [110/156]  eta: 0:00:32  lr: 0.008000  min_lr: 0.008000  loss: 4.3684 (4.1981)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [20]  [120/156]  eta: 0:00:25  lr: 0.008000  min_lr: 0.008000  loss: 4.4961 (4.2331)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [20]  [130/156]  eta: 0:00:18  lr: 0.008000  min_lr: 0.008000  loss: 4.5009 (4.2348)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0010  max mem: 55573
Epoch: [20]  [140/156]  eta: 0:00:10  lr: 0.008000  min_lr: 0.008000  loss: 4.1065 (4.2371)  weight_decay: 0.0500 (0.0500)  time: 0.5864  data: 0.0008  max mem: 55573
Epoch: [20]  [150/156]  eta: 0:00:04  lr: 0.008000  min_lr: 0.008000  loss: 4.1065 (4.2297)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [20]  [155/156]  eta: 0:00:00  lr: 0.008000  min_lr: 0.008000  loss: 4.2666 (4.2296)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [20] Total time: 0:01:46 (0.6818 s / it)
Averaged stats: lr: 0.008000  min_lr: 0.008000  loss: 4.2666 (4.2070)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.7322 (1.7322)  acc1: 66.6667 (66.6667)  acc5: 87.6302 (87.6302)  time: 7.1410  data: 6.9040  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.8599 (1.9082)  acc1: 62.6302 (58.4960)  acc5: 85.6771 (82.0480)  time: 1.5747  data: 1.3809  max mem: 55573
Test: Total time: 0:00:08 (1.6212 s / it)
* Acc@1 58.806 Acc@5 82.494 loss 1.897
Accuracy of the model on the 50000 test images: 58.8%
Max accuracy: 58.81%
Test:  [0/5]  eta: 0:00:33  loss: 6.9027 (6.9027)  acc1: 0.3906 (0.3906)  acc5: 0.3906 (0.3906)  time: 6.6775  data: 6.4410  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9108 (6.9329)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.4812  data: 1.2883  max mem: 55573
Test: Total time: 0:00:07 (1.4988 s / it)
* Acc@1 0.102 Acc@5 0.504 loss 6.922
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [21]  [  0/156]  eta: 0:35:33  lr: 0.008000  min_lr: 0.008000  loss: 4.3258 (4.3258)  weight_decay: 0.0500 (0.0500)  time: 13.6772  data: 8.7018  max mem: 55573
Epoch: [21]  [ 10/156]  eta: 0:04:21  lr: 0.008000  min_lr: 0.008000  loss: 3.9713 (3.9941)  weight_decay: 0.0500 (0.0500)  time: 1.7931  data: 0.7915  max mem: 55573
Epoch: [21]  [ 20/156]  eta: 0:02:45  lr: 0.008000  min_lr: 0.008000  loss: 4.2187 (4.2220)  weight_decay: 0.0500 (0.0500)  time: 0.5971  data: 0.0004  max mem: 55573
Epoch: [21]  [ 30/156]  eta: 0:02:08  lr: 0.008000  min_lr: 0.008000  loss: 4.4579 (4.2390)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [21]  [ 40/156]  eta: 0:01:46  lr: 0.008000  min_lr: 0.008000  loss: 4.4579 (4.2353)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0004  max mem: 55573
Epoch: [21]  [ 50/156]  eta: 0:01:30  lr: 0.008000  min_lr: 0.008000  loss: 4.0083 (4.1594)  weight_decay: 0.0500 (0.0500)  time: 0.5984  data: 0.0004  max mem: 55573
Epoch: [21]  [ 60/156]  eta: 0:01:17  lr: 0.008000  min_lr: 0.008000  loss: 3.7893 (4.1196)  weight_decay: 0.0500 (0.0500)  time: 0.5953  data: 0.0004  max mem: 55573
Epoch: [21]  [ 70/156]  eta: 0:01:06  lr: 0.007999  min_lr: 0.007999  loss: 4.1699 (4.1304)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0005  max mem: 55573
Epoch: [21]  [ 80/156]  eta: 0:00:57  lr: 0.007999  min_lr: 0.007999  loss: 4.3223 (4.1286)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0005  max mem: 55573
Epoch: [21]  [ 90/156]  eta: 0:00:48  lr: 0.007999  min_lr: 0.007999  loss: 4.2963 (4.1334)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [21]  [100/156]  eta: 0:00:40  lr: 0.007999  min_lr: 0.007999  loss: 4.2127 (4.1405)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0005  max mem: 55573
Epoch: [21]  [110/156]  eta: 0:00:32  lr: 0.007999  min_lr: 0.007999  loss: 4.3516 (4.1584)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [21]  [120/156]  eta: 0:00:25  lr: 0.007999  min_lr: 0.007999  loss: 4.3058 (4.1540)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [21]  [130/156]  eta: 0:00:18  lr: 0.007999  min_lr: 0.007999  loss: 4.1344 (4.1614)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0011  max mem: 55573
Epoch: [21]  [140/156]  eta: 0:00:10  lr: 0.007999  min_lr: 0.007999  loss: 4.5062 (4.1620)  weight_decay: 0.0500 (0.0500)  time: 0.5859  data: 0.0009  max mem: 55573
Epoch: [21]  [150/156]  eta: 0:00:04  lr: 0.007999  min_lr: 0.007999  loss: 4.3090 (4.1549)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [21]  [155/156]  eta: 0:00:00  lr: 0.007999  min_lr: 0.007999  loss: 4.2480 (4.1457)  weight_decay: 0.0500 (0.0500)  time: 0.5829  data: 0.0001  max mem: 55573
Epoch: [21] Total time: 0:01:46 (0.6815 s / it)
Averaged stats: lr: 0.007999  min_lr: 0.007999  loss: 4.2480 (4.1580)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.8084 (1.8084)  acc1: 65.4948 (65.4948)  acc5: 87.3698 (87.3698)  time: 7.0090  data: 6.7717  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.8084 (1.8273)  acc1: 62.8906 (59.7120)  acc5: 87.3698 (83.5520)  time: 1.5482  data: 1.3544  max mem: 55573
Test: Total time: 0:00:07 (1.5913 s / it)
* Acc@1 60.202 Acc@5 83.816 loss 1.806
Accuracy of the model on the 50000 test images: 60.2%
Max accuracy: 60.20%
Test:  [0/5]  eta: 0:00:34  loss: 6.9022 (6.9022)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 6.9197  data: 6.6837  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9122 (6.9344)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5292  data: 1.3368  max mem: 55573
Test: Total time: 0:00:07 (1.5455 s / it)
* Acc@1 0.098 Acc@5 0.508 loss 6.923
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [22]  [  0/156]  eta: 0:34:15  lr: 0.007999  min_lr: 0.007999  loss: 4.6287 (4.6287)  weight_decay: 0.0500 (0.0500)  time: 13.1733  data: 5.9656  max mem: 55573
Epoch: [22]  [ 10/156]  eta: 0:04:15  lr: 0.007999  min_lr: 0.007999  loss: 3.8238 (4.0234)  weight_decay: 0.0500 (0.0500)  time: 1.7469  data: 0.5428  max mem: 55573
Epoch: [22]  [ 20/156]  eta: 0:02:42  lr: 0.007999  min_lr: 0.007999  loss: 3.9197 (4.0546)  weight_decay: 0.0500 (0.0500)  time: 0.5994  data: 0.0005  max mem: 55573
Epoch: [22]  [ 30/156]  eta: 0:02:06  lr: 0.007999  min_lr: 0.007999  loss: 4.4368 (4.1254)  weight_decay: 0.0500 (0.0500)  time: 0.5936  data: 0.0004  max mem: 55573
Epoch: [22]  [ 40/156]  eta: 0:01:44  lr: 0.007999  min_lr: 0.007999  loss: 3.9143 (4.0404)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [22]  [ 50/156]  eta: 0:01:29  lr: 0.007999  min_lr: 0.007999  loss: 4.0047 (4.0525)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0004  max mem: 55573
Epoch: [22]  [ 60/156]  eta: 0:01:16  lr: 0.007999  min_lr: 0.007999  loss: 4.0534 (4.0778)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [22]  [ 70/156]  eta: 0:01:06  lr: 0.007998  min_lr: 0.007998  loss: 4.1056 (4.0724)  weight_decay: 0.0500 (0.0500)  time: 0.5939  data: 0.0005  max mem: 55573
Epoch: [22]  [ 80/156]  eta: 0:00:56  lr: 0.007998  min_lr: 0.007998  loss: 4.2066 (4.0876)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0005  max mem: 55573
Epoch: [22]  [ 90/156]  eta: 0:00:48  lr: 0.007998  min_lr: 0.007998  loss: 4.0582 (4.0715)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [22]  [100/156]  eta: 0:00:40  lr: 0.007998  min_lr: 0.007998  loss: 4.0774 (4.0821)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0004  max mem: 55573
Epoch: [22]  [110/156]  eta: 0:00:32  lr: 0.007998  min_lr: 0.007998  loss: 4.0774 (4.0627)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [22]  [120/156]  eta: 0:00:25  lr: 0.007998  min_lr: 0.007998  loss: 3.9630 (4.0656)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [22]  [130/156]  eta: 0:00:17  lr: 0.007998  min_lr: 0.007998  loss: 4.4412 (4.0925)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0010  max mem: 55573
Epoch: [22]  [140/156]  eta: 0:00:10  lr: 0.007998  min_lr: 0.007998  loss: 4.3553 (4.0950)  weight_decay: 0.0500 (0.0500)  time: 0.5856  data: 0.0009  max mem: 55573
Epoch: [22]  [150/156]  eta: 0:00:04  lr: 0.007998  min_lr: 0.007998  loss: 4.2830 (4.1076)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [22]  [155/156]  eta: 0:00:00  lr: 0.007998  min_lr: 0.007998  loss: 4.2904 (4.1214)  weight_decay: 0.0500 (0.0500)  time: 0.5812  data: 0.0001  max mem: 55573
Epoch: [22] Total time: 0:01:45 (0.6779 s / it)
Averaged stats: lr: 0.007998  min_lr: 0.007998  loss: 4.2904 (4.1221)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.7081 (1.7081)  acc1: 64.9740 (64.9740)  acc5: 88.6719 (88.6719)  time: 6.8550  data: 6.6180  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.9066 (1.8631)  acc1: 61.7188 (58.7200)  acc5: 86.0677 (83.0080)  time: 1.5166  data: 1.3237  max mem: 55573
Test: Total time: 0:00:07 (1.5560 s / it)
* Acc@1 58.628 Acc@5 82.940 loss 1.879
Accuracy of the model on the 50000 test images: 58.6%
Max accuracy: 60.20%
Test:  [0/5]  eta: 0:00:36  loss: 6.9017 (6.9017)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 7.2160  data: 6.9801  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9138 (6.9358)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5886  data: 1.3961  max mem: 55573
Test: Total time: 0:00:08 (1.6074 s / it)
* Acc@1 0.098 Acc@5 0.508 loss 6.924
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [23]  [  0/156]  eta: 0:35:41  lr: 0.007998  min_lr: 0.007998  loss: 4.3032 (4.3032)  weight_decay: 0.0500 (0.0500)  time: 13.7304  data: 10.6293  max mem: 55573
Epoch: [23]  [ 10/156]  eta: 0:04:21  lr: 0.007998  min_lr: 0.007998  loss: 4.0780 (3.9850)  weight_decay: 0.0500 (0.0500)  time: 1.7882  data: 0.9667  max mem: 55573
Epoch: [23]  [ 20/156]  eta: 0:02:45  lr: 0.007998  min_lr: 0.007998  loss: 4.0780 (3.9727)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0004  max mem: 55573
Epoch: [23]  [ 30/156]  eta: 0:02:07  lr: 0.007997  min_lr: 0.007997  loss: 4.2574 (4.1122)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [23]  [ 40/156]  eta: 0:01:45  lr: 0.007997  min_lr: 0.007997  loss: 4.3475 (4.1755)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0004  max mem: 55573
Epoch: [23]  [ 50/156]  eta: 0:01:29  lr: 0.007997  min_lr: 0.007997  loss: 4.3193 (4.1739)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [23]  [ 60/156]  eta: 0:01:17  lr: 0.007997  min_lr: 0.007997  loss: 4.2670 (4.1839)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [23]  [ 70/156]  eta: 0:01:06  lr: 0.007997  min_lr: 0.007997  loss: 4.2737 (4.1622)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [23]  [ 80/156]  eta: 0:00:57  lr: 0.007997  min_lr: 0.007997  loss: 4.3859 (4.1767)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [23]  [ 90/156]  eta: 0:00:48  lr: 0.007997  min_lr: 0.007997  loss: 4.2502 (4.1653)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [23]  [100/156]  eta: 0:00:40  lr: 0.007997  min_lr: 0.007997  loss: 4.2137 (4.1710)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [23]  [110/156]  eta: 0:00:32  lr: 0.007997  min_lr: 0.007997  loss: 4.2376 (4.1516)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [23]  [120/156]  eta: 0:00:25  lr: 0.007996  min_lr: 0.007996  loss: 4.1558 (4.1536)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [23]  [130/156]  eta: 0:00:17  lr: 0.007996  min_lr: 0.007996  loss: 4.2273 (4.1592)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0007  max mem: 55573
Epoch: [23]  [140/156]  eta: 0:00:10  lr: 0.007996  min_lr: 0.007996  loss: 4.2729 (4.1576)  weight_decay: 0.0500 (0.0500)  time: 0.5866  data: 0.0006  max mem: 55573
Epoch: [23]  [150/156]  eta: 0:00:04  lr: 0.007996  min_lr: 0.007996  loss: 4.0856 (4.1563)  weight_decay: 0.0500 (0.0500)  time: 0.5812  data: 0.0001  max mem: 55573
Epoch: [23]  [155/156]  eta: 0:00:00  lr: 0.007996  min_lr: 0.007996  loss: 4.2562 (4.1634)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [23] Total time: 0:01:46 (0.6804 s / it)
Averaged stats: lr: 0.007996  min_lr: 0.007996  loss: 4.2562 (4.0931)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.5829 (1.5829)  acc1: 68.8802 (68.8802)  acc5: 90.4948 (90.4948)  time: 7.2406  data: 7.0021  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.8808 (1.9530)  acc1: 63.1510 (60.6400)  acc5: 87.1094 (83.6160)  time: 1.5946  data: 1.4005  max mem: 55573
Test: Total time: 0:00:08 (1.6534 s / it)
* Acc@1 61.114 Acc@5 83.824 loss 1.938
Accuracy of the model on the 50000 test images: 61.1%
Max accuracy: 61.11%
Test:  [0/5]  eta: 0:00:35  loss: 6.9015 (6.9015)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 7.0309  data: 6.7949  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9150 (6.9374)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5515  data: 1.3591  max mem: 55573
Test: Total time: 0:00:07 (1.5711 s / it)
* Acc@1 0.100 Acc@5 0.512 loss 6.925
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [24]  [  0/156]  eta: 0:35:28  lr: 0.007996  min_lr: 0.007996  loss: 4.1489 (4.1489)  weight_decay: 0.0500 (0.0500)  time: 13.6419  data: 8.8803  max mem: 55573
Epoch: [24]  [ 10/156]  eta: 0:04:19  lr: 0.007996  min_lr: 0.007996  loss: 4.0709 (3.9564)  weight_decay: 0.0500 (0.0500)  time: 1.7800  data: 0.8077  max mem: 55573
Epoch: [24]  [ 20/156]  eta: 0:02:45  lr: 0.007996  min_lr: 0.007996  loss: 4.2811 (4.0681)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0004  max mem: 55573
Epoch: [24]  [ 30/156]  eta: 0:02:07  lr: 0.007996  min_lr: 0.007996  loss: 4.3503 (4.0745)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [24]  [ 40/156]  eta: 0:01:45  lr: 0.007995  min_lr: 0.007995  loss: 4.2438 (4.0781)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [24]  [ 50/156]  eta: 0:01:29  lr: 0.007995  min_lr: 0.007995  loss: 4.1994 (4.0832)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [24]  [ 60/156]  eta: 0:01:17  lr: 0.007995  min_lr: 0.007995  loss: 4.1260 (4.0832)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [24]  [ 70/156]  eta: 0:01:06  lr: 0.007995  min_lr: 0.007995  loss: 3.9801 (4.0424)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [24]  [ 80/156]  eta: 0:00:57  lr: 0.007995  min_lr: 0.007995  loss: 4.2234 (4.0695)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [24]  [ 90/156]  eta: 0:00:48  lr: 0.007995  min_lr: 0.007995  loss: 4.3290 (4.0741)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [24]  [100/156]  eta: 0:00:40  lr: 0.007995  min_lr: 0.007995  loss: 4.2313 (4.0565)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [24]  [110/156]  eta: 0:00:32  lr: 0.007994  min_lr: 0.007994  loss: 4.2490 (4.0657)  weight_decay: 0.0500 (0.0500)  time: 0.5887  data: 0.0004  max mem: 55573
Epoch: [24]  [120/156]  eta: 0:00:25  lr: 0.007994  min_lr: 0.007994  loss: 4.1944 (4.0493)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [24]  [130/156]  eta: 0:00:17  lr: 0.007994  min_lr: 0.007994  loss: 4.1944 (4.0534)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0010  max mem: 55573
Epoch: [24]  [140/156]  eta: 0:00:10  lr: 0.007994  min_lr: 0.007994  loss: 4.2200 (4.0427)  weight_decay: 0.0500 (0.0500)  time: 0.5851  data: 0.0009  max mem: 55573
Epoch: [24]  [150/156]  eta: 0:00:04  lr: 0.007994  min_lr: 0.007994  loss: 4.0806 (4.0399)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [24]  [155/156]  eta: 0:00:00  lr: 0.007994  min_lr: 0.007994  loss: 4.1172 (4.0270)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [24] Total time: 0:01:45 (0.6794 s / it)
Averaged stats: lr: 0.007994  min_lr: 0.007994  loss: 4.1172 (4.0579)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.5592 (1.5592)  acc1: 66.4062 (66.4062)  acc5: 90.1042 (90.1042)  time: 7.1981  data: 6.9606  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.7756 (1.7966)  acc1: 63.0208 (60.3840)  acc5: 87.7604 (83.4560)  time: 1.5863  data: 1.3922  max mem: 55573
Test: Total time: 0:00:08 (1.6357 s / it)
* Acc@1 61.156 Acc@5 84.338 loss 1.779
Accuracy of the model on the 50000 test images: 61.2%
Max accuracy: 61.16%
Test:  [0/5]  eta: 0:00:34  loss: 6.9012 (6.9012)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 6.8748  data: 6.6388  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9165 (6.9390)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5203  data: 1.3279  max mem: 55573
Test: Total time: 0:00:07 (1.5370 s / it)
* Acc@1 0.100 Acc@5 0.514 loss 6.925
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [25]  [  0/156]  eta: 0:32:22  lr: 0.007994  min_lr: 0.007994  loss: 4.4994 (4.4994)  weight_decay: 0.0500 (0.0500)  time: 12.4514  data: 11.7960  max mem: 55573
Epoch: [25]  [ 10/156]  eta: 0:04:23  lr: 0.007994  min_lr: 0.007994  loss: 4.0514 (4.1087)  weight_decay: 0.0500 (0.0500)  time: 1.8014  data: 1.0727  max mem: 55573
Epoch: [25]  [ 20/156]  eta: 0:02:46  lr: 0.007993  min_lr: 0.007993  loss: 4.0514 (4.1224)  weight_decay: 0.0500 (0.0500)  time: 0.6634  data: 0.0004  max mem: 55573
Epoch: [25]  [ 30/156]  eta: 0:02:08  lr: 0.007993  min_lr: 0.007993  loss: 3.9340 (4.0304)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [25]  [ 40/156]  eta: 0:01:46  lr: 0.007993  min_lr: 0.007993  loss: 3.7872 (4.0008)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [25]  [ 50/156]  eta: 0:01:30  lr: 0.007993  min_lr: 0.007993  loss: 4.0677 (3.9764)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [25]  [ 60/156]  eta: 0:01:17  lr: 0.007993  min_lr: 0.007993  loss: 4.0677 (4.0053)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0004  max mem: 55573
Epoch: [25]  [ 70/156]  eta: 0:01:06  lr: 0.007993  min_lr: 0.007993  loss: 4.1872 (4.0557)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0004  max mem: 55573
Epoch: [25]  [ 80/156]  eta: 0:00:57  lr: 0.007992  min_lr: 0.007992  loss: 4.1424 (4.0607)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0004  max mem: 55573
Epoch: [25]  [ 90/156]  eta: 0:00:48  lr: 0.007992  min_lr: 0.007992  loss: 4.0127 (4.0650)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [25]  [100/156]  eta: 0:00:40  lr: 0.007992  min_lr: 0.007992  loss: 4.0127 (4.0601)  weight_decay: 0.0500 (0.0500)  time: 0.5987  data: 0.0005  max mem: 55573
Epoch: [25]  [110/156]  eta: 0:00:32  lr: 0.007992  min_lr: 0.007992  loss: 3.9260 (4.0493)  weight_decay: 0.0500 (0.0500)  time: 0.5975  data: 0.0005  max mem: 55573
Epoch: [25]  [120/156]  eta: 0:00:25  lr: 0.007992  min_lr: 0.007992  loss: 4.1455 (4.0563)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0005  max mem: 55573
Epoch: [25]  [130/156]  eta: 0:00:18  lr: 0.007991  min_lr: 0.007991  loss: 4.1455 (4.0550)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0010  max mem: 55573
Epoch: [25]  [140/156]  eta: 0:00:10  lr: 0.007991  min_lr: 0.007991  loss: 4.1892 (4.0667)  weight_decay: 0.0500 (0.0500)  time: 0.5859  data: 0.0009  max mem: 55573
Epoch: [25]  [150/156]  eta: 0:00:04  lr: 0.007991  min_lr: 0.007991  loss: 4.3140 (4.0868)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [25]  [155/156]  eta: 0:00:00  lr: 0.007991  min_lr: 0.007991  loss: 4.3140 (4.0954)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [25] Total time: 0:01:46 (0.6815 s / it)
Averaged stats: lr: 0.007991  min_lr: 0.007991  loss: 4.3140 (4.0429)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.6137 (1.6137)  acc1: 68.6198 (68.6198)  acc5: 91.2760 (91.2760)  time: 6.8377  data: 6.6001  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.8071 (1.8191)  acc1: 64.7135 (61.2800)  acc5: 89.0625 (85.3760)  time: 1.5140  data: 1.3201  max mem: 55573
Test: Total time: 0:00:07 (1.5580 s / it)
* Acc@1 62.102 Acc@5 84.940 loss 1.812
Accuracy of the model on the 50000 test images: 62.1%
Max accuracy: 62.10%
Test:  [0/5]  eta: 0:00:35  loss: 6.9011 (6.9011)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 7.0668  data: 6.8305  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9179 (6.9403)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5589  data: 1.3663  max mem: 55573
Test: Total time: 0:00:07 (1.5748 s / it)
* Acc@1 0.102 Acc@5 0.514 loss 6.926
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [26]  [  0/156]  eta: 0:37:31  lr: 0.007991  min_lr: 0.007991  loss: 4.1827 (4.1827)  weight_decay: 0.0500 (0.0500)  time: 14.4336  data: 10.4241  max mem: 55573
Epoch: [26]  [ 10/156]  eta: 0:04:30  lr: 0.007991  min_lr: 0.007991  loss: 3.6017 (3.7311)  weight_decay: 0.0500 (0.0500)  time: 1.8540  data: 0.9480  max mem: 55573
Epoch: [26]  [ 20/156]  eta: 0:02:50  lr: 0.007991  min_lr: 0.007991  loss: 3.6017 (3.7997)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0003  max mem: 55573
Epoch: [26]  [ 30/156]  eta: 0:02:10  lr: 0.007990  min_lr: 0.007990  loss: 3.5642 (3.7310)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [26]  [ 40/156]  eta: 0:01:47  lr: 0.007990  min_lr: 0.007990  loss: 3.9808 (3.8342)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [26]  [ 50/156]  eta: 0:01:31  lr: 0.007990  min_lr: 0.007990  loss: 4.0943 (3.8017)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [26]  [ 60/156]  eta: 0:01:18  lr: 0.007990  min_lr: 0.007990  loss: 3.9701 (3.8550)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [26]  [ 70/156]  eta: 0:01:07  lr: 0.007990  min_lr: 0.007990  loss: 4.0191 (3.8829)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [26]  [ 80/156]  eta: 0:00:57  lr: 0.007989  min_lr: 0.007989  loss: 3.9947 (3.8781)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [26]  [ 90/156]  eta: 0:00:49  lr: 0.007989  min_lr: 0.007989  loss: 3.9606 (3.8818)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [26]  [100/156]  eta: 0:00:40  lr: 0.007989  min_lr: 0.007989  loss: 4.0368 (3.9146)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [26]  [110/156]  eta: 0:00:32  lr: 0.007989  min_lr: 0.007989  loss: 4.1901 (3.9314)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [26]  [120/156]  eta: 0:00:25  lr: 0.007988  min_lr: 0.007988  loss: 4.2891 (3.9575)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0005  max mem: 55573
Epoch: [26]  [130/156]  eta: 0:00:18  lr: 0.007988  min_lr: 0.007988  loss: 4.1969 (3.9571)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0010  max mem: 55573
Epoch: [26]  [140/156]  eta: 0:00:11  lr: 0.007988  min_lr: 0.007988  loss: 4.0673 (3.9585)  weight_decay: 0.0500 (0.0500)  time: 0.5862  data: 0.0009  max mem: 55573
Epoch: [26]  [150/156]  eta: 0:00:04  lr: 0.007988  min_lr: 0.007988  loss: 4.1353 (3.9609)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [26]  [155/156]  eta: 0:00:00  lr: 0.007988  min_lr: 0.007988  loss: 4.1353 (3.9667)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [26] Total time: 0:01:46 (0.6842 s / it)
Averaged stats: lr: 0.007988  min_lr: 0.007988  loss: 4.1353 (4.0190)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.6240 (1.6240)  acc1: 69.0104 (69.0104)  acc5: 90.6250 (90.6250)  time: 6.9646  data: 6.7276  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.8326 (1.8489)  acc1: 65.3646 (62.7200)  acc5: 88.0208 (85.3760)  time: 1.5392  data: 1.3456  max mem: 55573
Test: Total time: 0:00:07 (1.5795 s / it)
* Acc@1 62.860 Acc@5 85.316 loss 1.847
Accuracy of the model on the 50000 test images: 62.9%
Max accuracy: 62.86%
Test:  [0/5]  eta: 0:00:33  loss: 6.9005 (6.9005)  acc1: 0.0000 (0.0000)  acc5: 0.7812 (0.7812)  time: 6.7469  data: 6.5109  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9198 (6.9416)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.4951  data: 1.3023  max mem: 55573
Test: Total time: 0:00:07 (1.5134 s / it)
* Acc@1 0.102 Acc@5 0.512 loss 6.927
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [27]  [  0/156]  eta: 0:34:26  lr: 0.007988  min_lr: 0.007988  loss: 3.3325 (3.3325)  weight_decay: 0.0500 (0.0500)  time: 13.2439  data: 9.8337  max mem: 55573
Epoch: [27]  [ 10/156]  eta: 0:04:18  lr: 0.007987  min_lr: 0.007987  loss: 4.1658 (4.0630)  weight_decay: 0.0500 (0.0500)  time: 1.7681  data: 0.8945  max mem: 55573
Epoch: [27]  [ 20/156]  eta: 0:02:44  lr: 0.007987  min_lr: 0.007987  loss: 4.1248 (3.8720)  weight_decay: 0.0500 (0.0500)  time: 0.6054  data: 0.0005  max mem: 55573
Epoch: [27]  [ 30/156]  eta: 0:02:07  lr: 0.007987  min_lr: 0.007987  loss: 3.8229 (3.9291)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [27]  [ 40/156]  eta: 0:01:45  lr: 0.007987  min_lr: 0.007987  loss: 4.1415 (3.9703)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [27]  [ 50/156]  eta: 0:01:29  lr: 0.007987  min_lr: 0.007987  loss: 4.1752 (3.9732)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [27]  [ 60/156]  eta: 0:01:17  lr: 0.007986  min_lr: 0.007986  loss: 3.8146 (3.9468)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [27]  [ 70/156]  eta: 0:01:06  lr: 0.007986  min_lr: 0.007986  loss: 3.9548 (3.9472)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [27]  [ 80/156]  eta: 0:00:57  lr: 0.007986  min_lr: 0.007986  loss: 3.9548 (3.9479)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [27]  [ 90/156]  eta: 0:00:48  lr: 0.007986  min_lr: 0.007986  loss: 3.9363 (3.9263)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0004  max mem: 55573
Epoch: [27]  [100/156]  eta: 0:00:40  lr: 0.007985  min_lr: 0.007985  loss: 4.1559 (3.9604)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [27]  [110/156]  eta: 0:00:32  lr: 0.007985  min_lr: 0.007985  loss: 4.2598 (3.9713)  weight_decay: 0.0500 (0.0500)  time: 0.5936  data: 0.0004  max mem: 55573
Epoch: [27]  [120/156]  eta: 0:00:25  lr: 0.007985  min_lr: 0.007985  loss: 4.1621 (3.9666)  weight_decay: 0.0500 (0.0500)  time: 0.5930  data: 0.0005  max mem: 55573
Epoch: [27]  [130/156]  eta: 0:00:17  lr: 0.007985  min_lr: 0.007985  loss: 4.0009 (3.9503)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0010  max mem: 55573
Epoch: [27]  [140/156]  eta: 0:00:10  lr: 0.007984  min_lr: 0.007984  loss: 4.0009 (3.9501)  weight_decay: 0.0500 (0.0500)  time: 0.5858  data: 0.0008  max mem: 55573
Epoch: [27]  [150/156]  eta: 0:00:04  lr: 0.007984  min_lr: 0.007984  loss: 4.1844 (3.9539)  weight_decay: 0.0500 (0.0500)  time: 0.5828  data: 0.0001  max mem: 55573
Epoch: [27]  [155/156]  eta: 0:00:00  lr: 0.007984  min_lr: 0.007984  loss: 4.2664 (3.9554)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [27] Total time: 0:01:45 (0.6792 s / it)
Averaged stats: lr: 0.007984  min_lr: 0.007984  loss: 4.2664 (4.0092)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.6575 (1.6575)  acc1: 67.5781 (67.5781)  acc5: 91.1458 (91.1458)  time: 6.8897  data: 6.6525  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.6827 (1.7312)  acc1: 66.0156 (62.4960)  acc5: 89.3229 (86.3040)  time: 1.5243  data: 1.3306  max mem: 55573
Test: Total time: 0:00:07 (1.5569 s / it)
* Acc@1 63.330 Acc@5 85.988 loss 1.726
Accuracy of the model on the 50000 test images: 63.3%
Max accuracy: 63.33%
Test:  [0/5]  eta: 0:00:35  loss: 6.9005 (6.9005)  acc1: 0.0000 (0.0000)  acc5: 0.7812 (0.7812)  time: 7.0139  data: 6.7779  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9213 (6.9428)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5481  data: 1.3557  max mem: 55573
Test: Total time: 0:00:07 (1.5626 s / it)
* Acc@1 0.102 Acc@5 0.514 loss 6.927
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [28]  [  0/156]  eta: 0:31:47  lr: 0.007984  min_lr: 0.007984  loss: 4.0935 (4.0935)  weight_decay: 0.0500 (0.0500)  time: 12.2290  data: 9.6337  max mem: 55573
Epoch: [28]  [ 10/156]  eta: 0:04:16  lr: 0.007984  min_lr: 0.007984  loss: 4.0888 (3.9062)  weight_decay: 0.0500 (0.0500)  time: 1.7573  data: 0.8762  max mem: 55573
Epoch: [28]  [ 20/156]  eta: 0:02:43  lr: 0.007983  min_lr: 0.007983  loss: 4.0571 (3.9362)  weight_decay: 0.0500 (0.0500)  time: 0.6503  data: 0.0004  max mem: 55573
Epoch: [28]  [ 30/156]  eta: 0:02:06  lr: 0.007983  min_lr: 0.007983  loss: 4.1829 (4.0220)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0003  max mem: 55573
Epoch: [28]  [ 40/156]  eta: 0:01:44  lr: 0.007983  min_lr: 0.007983  loss: 4.1829 (4.0459)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [28]  [ 50/156]  eta: 0:01:29  lr: 0.007983  min_lr: 0.007983  loss: 4.0589 (3.9911)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [28]  [ 60/156]  eta: 0:01:16  lr: 0.007982  min_lr: 0.007982  loss: 3.9352 (3.9654)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [28]  [ 70/156]  eta: 0:01:06  lr: 0.007982  min_lr: 0.007982  loss: 4.1539 (3.9834)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [28]  [ 80/156]  eta: 0:00:56  lr: 0.007982  min_lr: 0.007982  loss: 4.2916 (3.9999)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [28]  [ 90/156]  eta: 0:00:48  lr: 0.007981  min_lr: 0.007981  loss: 4.2336 (3.9624)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [28]  [100/156]  eta: 0:00:40  lr: 0.007981  min_lr: 0.007981  loss: 4.1373 (3.9947)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [28]  [110/156]  eta: 0:00:32  lr: 0.007981  min_lr: 0.007981  loss: 4.3270 (4.0037)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [28]  [120/156]  eta: 0:00:25  lr: 0.007981  min_lr: 0.007981  loss: 4.1607 (3.9957)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [28]  [130/156]  eta: 0:00:17  lr: 0.007980  min_lr: 0.007980  loss: 4.0847 (3.9925)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0010  max mem: 55573
Epoch: [28]  [140/156]  eta: 0:00:10  lr: 0.007980  min_lr: 0.007980  loss: 4.1724 (3.9989)  weight_decay: 0.0500 (0.0500)  time: 0.5855  data: 0.0008  max mem: 55573
Epoch: [28]  [150/156]  eta: 0:00:04  lr: 0.007980  min_lr: 0.007980  loss: 4.2949 (3.9925)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0001  max mem: 55573
Epoch: [28]  [155/156]  eta: 0:00:00  lr: 0.007980  min_lr: 0.007980  loss: 4.2949 (3.9937)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [28] Total time: 0:01:45 (0.6777 s / it)
Averaged stats: lr: 0.007980  min_lr: 0.007980  loss: 4.2949 (4.0063)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.5828 (1.5828)  acc1: 66.1458 (66.1458)  acc5: 89.9740 (89.9740)  time: 7.1002  data: 6.8624  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.7777 (1.7607)  acc1: 63.0208 (60.4800)  acc5: 87.7604 (84.2240)  time: 1.5665  data: 1.3726  max mem: 55573
Test: Total time: 0:00:08 (1.6147 s / it)
* Acc@1 61.036 Acc@5 84.250 loss 1.779
Accuracy of the model on the 50000 test images: 61.0%
Max accuracy: 63.33%
Test:  [0/5]  eta: 0:00:34  loss: 6.9007 (6.9007)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 6.9104  data: 6.6744  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9228 (6.9443)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.5274  data: 1.3350  max mem: 55573
Test: Total time: 0:00:07 (1.5441 s / it)
* Acc@1 0.102 Acc@5 0.514 loss 6.928
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [29]  [  0/156]  eta: 0:32:14  lr: 0.007980  min_lr: 0.007980  loss: 2.8638 (2.8638)  weight_decay: 0.0500 (0.0500)  time: 12.3998  data: 8.4643  max mem: 55573
Epoch: [29]  [ 10/156]  eta: 0:04:16  lr: 0.007979  min_lr: 0.007979  loss: 4.2574 (4.0782)  weight_decay: 0.0500 (0.0500)  time: 1.7585  data: 0.7699  max mem: 55573
Epoch: [29]  [ 20/156]  eta: 0:02:43  lr: 0.007979  min_lr: 0.007979  loss: 4.1295 (3.9222)  weight_decay: 0.0500 (0.0500)  time: 0.6430  data: 0.0004  max mem: 55573
Epoch: [29]  [ 30/156]  eta: 0:02:06  lr: 0.007979  min_lr: 0.007979  loss: 4.0121 (4.0040)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [29]  [ 40/156]  eta: 0:01:44  lr: 0.007978  min_lr: 0.007978  loss: 4.2753 (4.0268)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [29]  [ 50/156]  eta: 0:01:29  lr: 0.007978  min_lr: 0.007978  loss: 4.1676 (3.9948)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [29]  [ 60/156]  eta: 0:01:16  lr: 0.007978  min_lr: 0.007978  loss: 4.1818 (4.0319)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [29]  [ 70/156]  eta: 0:01:06  lr: 0.007978  min_lr: 0.007978  loss: 4.1818 (4.0289)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [29]  [ 80/156]  eta: 0:00:56  lr: 0.007977  min_lr: 0.007977  loss: 4.0584 (4.0314)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [29]  [ 90/156]  eta: 0:00:48  lr: 0.007977  min_lr: 0.007977  loss: 4.0538 (4.0211)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [29]  [100/156]  eta: 0:00:40  lr: 0.007977  min_lr: 0.007977  loss: 3.8530 (4.0079)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [29]  [110/156]  eta: 0:00:32  lr: 0.007976  min_lr: 0.007976  loss: 4.1211 (4.0151)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [29]  [120/156]  eta: 0:00:25  lr: 0.007976  min_lr: 0.007976  loss: 4.1335 (4.0208)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [29]  [130/156]  eta: 0:00:17  lr: 0.007976  min_lr: 0.007976  loss: 4.1085 (4.0257)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0011  max mem: 55573
Epoch: [29]  [140/156]  eta: 0:00:10  lr: 0.007975  min_lr: 0.007975  loss: 4.1844 (4.0324)  weight_decay: 0.0500 (0.0500)  time: 0.5854  data: 0.0009  max mem: 55573
Epoch: [29]  [150/156]  eta: 0:00:04  lr: 0.007975  min_lr: 0.007975  loss: 4.1844 (4.0425)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [29]  [155/156]  eta: 0:00:00  lr: 0.007975  min_lr: 0.007975  loss: 4.0908 (4.0388)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [29] Total time: 0:01:45 (0.6782 s / it)
Averaged stats: lr: 0.007975  min_lr: 0.007975  loss: 4.0908 (3.9740)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.6218 (1.6218)  acc1: 69.9219 (69.9219)  acc5: 91.1458 (91.1458)  time: 6.8469  data: 6.6100  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.7838 (1.8727)  acc1: 67.8385 (64.3840)  acc5: 88.9323 (86.1120)  time: 1.5158  data: 1.3221  max mem: 55573
Test: Total time: 0:00:07 (1.5632 s / it)
* Acc@1 63.440 Acc@5 85.954 loss 1.905
Accuracy of the model on the 50000 test images: 63.4%
Max accuracy: 63.44%
Test:  [0/5]  eta: 0:00:35  loss: 6.9012 (6.9012)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 7.1200  data: 6.8840  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9248 (6.9460)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.5692  data: 1.3769  max mem: 55573
Test: Total time: 0:00:07 (1.5860 s / it)
* Acc@1 0.102 Acc@5 0.510 loss 6.929
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [30]  [  0/156]  eta: 0:35:36  lr: 0.007975  min_lr: 0.007975  loss: 4.0529 (4.0529)  weight_decay: 0.0500 (0.0500)  time: 13.6966  data: 11.5007  max mem: 55573
Epoch: [30]  [ 10/156]  eta: 0:04:20  lr: 0.007975  min_lr: 0.007975  loss: 4.1267 (4.0341)  weight_decay: 0.0500 (0.0500)  time: 1.7836  data: 1.0459  max mem: 55573
Epoch: [30]  [ 20/156]  eta: 0:02:45  lr: 0.007974  min_lr: 0.007974  loss: 4.1267 (4.1404)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [30]  [ 30/156]  eta: 0:02:07  lr: 0.007974  min_lr: 0.007974  loss: 4.0204 (4.0563)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [30]  [ 40/156]  eta: 0:01:45  lr: 0.007974  min_lr: 0.007974  loss: 4.0764 (4.0546)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [30]  [ 50/156]  eta: 0:01:29  lr: 0.007973  min_lr: 0.007973  loss: 4.0764 (3.9942)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [30]  [ 60/156]  eta: 0:01:17  lr: 0.007973  min_lr: 0.007973  loss: 4.0317 (4.0177)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [30]  [ 70/156]  eta: 0:01:06  lr: 0.007973  min_lr: 0.007973  loss: 4.1959 (4.0045)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [30]  [ 80/156]  eta: 0:00:57  lr: 0.007972  min_lr: 0.007972  loss: 4.1156 (3.9785)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [30]  [ 90/156]  eta: 0:00:48  lr: 0.007972  min_lr: 0.007972  loss: 3.9550 (3.9730)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [30]  [100/156]  eta: 0:00:40  lr: 0.007972  min_lr: 0.007972  loss: 3.9598 (3.9710)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [30]  [110/156]  eta: 0:00:32  lr: 0.007971  min_lr: 0.007971  loss: 4.0934 (3.9747)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0004  max mem: 55573
Epoch: [30]  [120/156]  eta: 0:00:25  lr: 0.007971  min_lr: 0.007971  loss: 4.1383 (3.9703)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [30]  [130/156]  eta: 0:00:17  lr: 0.007970  min_lr: 0.007970  loss: 4.0658 (3.9722)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0015  max mem: 55573
Epoch: [30]  [140/156]  eta: 0:00:10  lr: 0.007970  min_lr: 0.007970  loss: 4.1089 (3.9720)  weight_decay: 0.0500 (0.0500)  time: 0.5860  data: 0.0013  max mem: 55573
Epoch: [30]  [150/156]  eta: 0:00:04  lr: 0.007970  min_lr: 0.007970  loss: 3.9966 (3.9596)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [30]  [155/156]  eta: 0:00:00  lr: 0.007970  min_lr: 0.007970  loss: 3.9966 (3.9674)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [30] Total time: 0:01:46 (0.6797 s / it)
Averaged stats: lr: 0.007970  min_lr: 0.007970  loss: 3.9966 (3.9523)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.5745 (1.5745)  acc1: 67.9688 (67.9688)  acc5: 89.8438 (89.8438)  time: 7.1304  data: 6.8933  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.6442 (1.7811)  acc1: 66.4062 (62.0160)  acc5: 89.8438 (84.8640)  time: 1.5723  data: 1.3787  max mem: 55573
Test: Total time: 0:00:08 (1.6146 s / it)
* Acc@1 62.216 Acc@5 85.046 loss 1.791
Accuracy of the model on the 50000 test images: 62.2%
Max accuracy: 63.44%
Test:  [0/5]  eta: 0:00:34  loss: 6.9019 (6.9019)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 6.9832  data: 6.7471  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9263 (6.9476)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.5420  data: 1.3495  max mem: 55573
Test: Total time: 0:00:07 (1.5613 s / it)
* Acc@1 0.102 Acc@5 0.510 loss 6.929
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [31]  [  0/156]  eta: 0:33:38  lr: 0.007970  min_lr: 0.007970  loss: 4.7284 (4.7284)  weight_decay: 0.0500 (0.0500)  time: 12.9390  data: 11.1059  max mem: 55573
Epoch: [31]  [ 10/156]  eta: 0:04:14  lr: 0.007969  min_lr: 0.007969  loss: 4.1389 (3.9806)  weight_decay: 0.0500 (0.0500)  time: 1.7437  data: 1.0100  max mem: 55573
Epoch: [31]  [ 20/156]  eta: 0:02:42  lr: 0.007969  min_lr: 0.007969  loss: 4.0278 (3.8391)  weight_decay: 0.0500 (0.0500)  time: 0.6070  data: 0.0004  max mem: 55573
Epoch: [31]  [ 30/156]  eta: 0:02:05  lr: 0.007969  min_lr: 0.007969  loss: 3.4589 (3.7237)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [31]  [ 40/156]  eta: 0:01:44  lr: 0.007968  min_lr: 0.007968  loss: 3.6847 (3.7563)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [31]  [ 50/156]  eta: 0:01:28  lr: 0.007968  min_lr: 0.007968  loss: 3.9102 (3.7863)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [31]  [ 60/156]  eta: 0:01:16  lr: 0.007967  min_lr: 0.007967  loss: 3.8521 (3.7736)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [31]  [ 70/156]  eta: 0:01:06  lr: 0.007967  min_lr: 0.007967  loss: 3.8994 (3.8178)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [31]  [ 80/156]  eta: 0:00:56  lr: 0.007967  min_lr: 0.007967  loss: 4.1076 (3.8242)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [31]  [ 90/156]  eta: 0:00:48  lr: 0.007966  min_lr: 0.007966  loss: 3.9521 (3.8311)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [31]  [100/156]  eta: 0:00:40  lr: 0.007966  min_lr: 0.007966  loss: 4.1410 (3.8707)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [31]  [110/156]  eta: 0:00:32  lr: 0.007966  min_lr: 0.007966  loss: 4.2060 (3.8832)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [31]  [120/156]  eta: 0:00:25  lr: 0.007965  min_lr: 0.007965  loss: 4.1603 (3.8879)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [31]  [130/156]  eta: 0:00:17  lr: 0.007965  min_lr: 0.007965  loss: 3.8358 (3.8804)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0010  max mem: 55573
Epoch: [31]  [140/156]  eta: 0:00:10  lr: 0.007964  min_lr: 0.007964  loss: 4.0148 (3.8964)  weight_decay: 0.0500 (0.0500)  time: 0.5876  data: 0.0008  max mem: 55573
Epoch: [31]  [150/156]  eta: 0:00:04  lr: 0.007964  min_lr: 0.007964  loss: 4.2039 (3.9153)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [31]  [155/156]  eta: 0:00:00  lr: 0.007964  min_lr: 0.007964  loss: 4.0914 (3.9100)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [31] Total time: 0:01:45 (0.6774 s / it)
Averaged stats: lr: 0.007964  min_lr: 0.007964  loss: 4.0914 (3.9297)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.3286 (1.3286)  acc1: 71.7448 (71.7448)  acc5: 91.0156 (91.0156)  time: 6.9226  data: 6.6855  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4694 (1.5508)  acc1: 67.7083 (63.7440)  acc5: 91.0156 (86.4320)  time: 1.5299  data: 1.3372  max mem: 55573
Test: Total time: 0:00:07 (1.5454 s / it)
* Acc@1 64.440 Acc@5 86.498 loss 1.569
Accuracy of the model on the 50000 test images: 64.4%
Max accuracy: 64.44%
Test:  [0/5]  eta: 0:00:33  loss: 6.9027 (6.9027)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 6.6443  data: 6.4077  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9280 (6.9492)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.4760  data: 1.2832  max mem: 55573
Test: Total time: 0:00:07 (1.4937 s / it)
* Acc@1 0.102 Acc@5 0.514 loss 6.930
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [32]  [  0/156]  eta: 0:32:50  lr: 0.007964  min_lr: 0.007964  loss: 3.8302 (3.8302)  weight_decay: 0.0500 (0.0500)  time: 12.6286  data: 11.3443  max mem: 55573
Epoch: [32]  [ 10/156]  eta: 0:04:11  lr: 0.007963  min_lr: 0.007963  loss: 4.1203 (4.0125)  weight_decay: 0.0500 (0.0500)  time: 1.7224  data: 1.0318  max mem: 55573
Epoch: [32]  [ 20/156]  eta: 0:02:40  lr: 0.007963  min_lr: 0.007963  loss: 4.1203 (4.0098)  weight_decay: 0.0500 (0.0500)  time: 0.6111  data: 0.0005  max mem: 55573
Epoch: [32]  [ 30/156]  eta: 0:02:04  lr: 0.007963  min_lr: 0.007963  loss: 3.9810 (3.9538)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [32]  [ 40/156]  eta: 0:01:43  lr: 0.007962  min_lr: 0.007962  loss: 3.8368 (3.9084)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [32]  [ 50/156]  eta: 0:01:28  lr: 0.007962  min_lr: 0.007962  loss: 3.9830 (3.9042)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0004  max mem: 55573
Epoch: [32]  [ 60/156]  eta: 0:01:16  lr: 0.007961  min_lr: 0.007961  loss: 3.9989 (3.8935)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [32]  [ 70/156]  eta: 0:01:05  lr: 0.007961  min_lr: 0.007961  loss: 4.1450 (3.9369)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [32]  [ 80/156]  eta: 0:00:56  lr: 0.007961  min_lr: 0.007961  loss: 4.1725 (3.9116)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [32]  [ 90/156]  eta: 0:00:47  lr: 0.007960  min_lr: 0.007960  loss: 4.0564 (3.9229)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [32]  [100/156]  eta: 0:00:39  lr: 0.007960  min_lr: 0.007960  loss: 3.9707 (3.9001)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [32]  [110/156]  eta: 0:00:32  lr: 0.007959  min_lr: 0.007959  loss: 3.7249 (3.9028)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [32]  [120/156]  eta: 0:00:24  lr: 0.007959  min_lr: 0.007959  loss: 3.8354 (3.9063)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [32]  [130/156]  eta: 0:00:17  lr: 0.007959  min_lr: 0.007959  loss: 3.8354 (3.8996)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0010  max mem: 55573
Epoch: [32]  [140/156]  eta: 0:00:10  lr: 0.007958  min_lr: 0.007958  loss: 4.0086 (3.9164)  weight_decay: 0.0500 (0.0500)  time: 0.5879  data: 0.0009  max mem: 55573
Epoch: [32]  [150/156]  eta: 0:00:04  lr: 0.007958  min_lr: 0.007958  loss: 4.2393 (3.9316)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [32]  [155/156]  eta: 0:00:00  lr: 0.007958  min_lr: 0.007958  loss: 4.1997 (3.9189)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [32] Total time: 0:01:45 (0.6754 s / it)
Averaged stats: lr: 0.007958  min_lr: 0.007958  loss: 4.1997 (3.9383)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.5006 (1.5006)  acc1: 70.9635 (70.9635)  acc5: 90.1042 (90.1042)  time: 7.2652  data: 7.0278  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5962 (1.6794)  acc1: 64.9740 (63.2960)  acc5: 89.4531 (86.0800)  time: 1.5995  data: 1.4056  max mem: 55573
Test: Total time: 0:00:08 (1.6575 s / it)
* Acc@1 63.570 Acc@5 86.108 loss 1.686
Accuracy of the model on the 50000 test images: 63.6%
Max accuracy: 64.44%
Test:  [0/5]  eta: 0:00:36  loss: 6.9029 (6.9029)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 7.2051  data: 6.9693  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9301 (6.9510)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.5863  data: 1.3940  max mem: 55573
Test: Total time: 0:00:08 (1.6052 s / it)
* Acc@1 0.100 Acc@5 0.514 loss 6.931
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [33]  [  0/156]  eta: 0:32:20  lr: 0.007958  min_lr: 0.007958  loss: 4.5243 (4.5243)  weight_decay: 0.0500 (0.0500)  time: 12.4375  data: 8.7119  max mem: 55573
Epoch: [33]  [ 10/156]  eta: 0:04:13  lr: 0.007957  min_lr: 0.007957  loss: 4.1227 (4.0923)  weight_decay: 0.0500 (0.0500)  time: 1.7332  data: 0.7924  max mem: 55573
Epoch: [33]  [ 20/156]  eta: 0:02:41  lr: 0.007957  min_lr: 0.007957  loss: 4.0931 (4.0945)  weight_decay: 0.0500 (0.0500)  time: 0.6270  data: 0.0004  max mem: 55573
Epoch: [33]  [ 30/156]  eta: 0:02:05  lr: 0.007956  min_lr: 0.007956  loss: 4.0766 (4.0107)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [33]  [ 40/156]  eta: 0:01:44  lr: 0.007956  min_lr: 0.007956  loss: 3.9179 (3.9766)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [33]  [ 50/156]  eta: 0:01:28  lr: 0.007955  min_lr: 0.007955  loss: 3.8618 (3.9539)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [33]  [ 60/156]  eta: 0:01:16  lr: 0.007955  min_lr: 0.007955  loss: 4.1914 (3.9846)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [33]  [ 70/156]  eta: 0:01:05  lr: 0.007955  min_lr: 0.007955  loss: 4.1528 (4.0042)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0005  max mem: 55573
Epoch: [33]  [ 80/156]  eta: 0:00:56  lr: 0.007954  min_lr: 0.007954  loss: 3.9981 (3.9654)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [33]  [ 90/156]  eta: 0:00:48  lr: 0.007954  min_lr: 0.007954  loss: 4.1180 (3.9983)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [33]  [100/156]  eta: 0:00:40  lr: 0.007953  min_lr: 0.007953  loss: 4.2165 (3.9850)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [33]  [110/156]  eta: 0:00:32  lr: 0.007953  min_lr: 0.007953  loss: 4.1041 (3.9938)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [33]  [120/156]  eta: 0:00:24  lr: 0.007952  min_lr: 0.007952  loss: 4.1489 (3.9872)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [33]  [130/156]  eta: 0:00:17  lr: 0.007952  min_lr: 0.007952  loss: 3.9516 (3.9865)  weight_decay: 0.0500 (0.0500)  time: 0.5881  data: 0.0010  max mem: 55573
Epoch: [33]  [140/156]  eta: 0:00:10  lr: 0.007951  min_lr: 0.007951  loss: 4.1491 (3.9963)  weight_decay: 0.0500 (0.0500)  time: 0.5849  data: 0.0008  max mem: 55573
Epoch: [33]  [150/156]  eta: 0:00:04  lr: 0.007951  min_lr: 0.007951  loss: 4.1722 (3.9998)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [33]  [155/156]  eta: 0:00:00  lr: 0.007951  min_lr: 0.007951  loss: 4.1764 (4.0051)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [33] Total time: 0:01:45 (0.6757 s / it)
Averaged stats: lr: 0.007951  min_lr: 0.007951  loss: 4.1764 (3.9170)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.4096 (1.4096)  acc1: 69.6615 (69.6615)  acc5: 90.4948 (90.4948)  time: 6.8967  data: 6.6595  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.6456 (1.6381)  acc1: 65.8854 (62.9760)  acc5: 88.8021 (86.3360)  time: 1.5258  data: 1.3320  max mem: 55573
Test: Total time: 0:00:07 (1.5694 s / it)
* Acc@1 64.274 Acc@5 86.562 loss 1.614
Accuracy of the model on the 50000 test images: 64.3%
Max accuracy: 64.44%
Test:  [0/5]  eta: 0:00:36  loss: 6.9028 (6.9028)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 7.3576  data: 7.1211  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9323 (6.9530)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.6174  data: 1.4243  max mem: 55573
Test: Total time: 0:00:08 (1.6356 s / it)
* Acc@1 0.100 Acc@5 0.508 loss 6.933
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [34]  [  0/156]  eta: 0:30:39  lr: 0.007951  min_lr: 0.007951  loss: 3.7036 (3.7036)  weight_decay: 0.0500 (0.0500)  time: 11.7933  data: 9.4561  max mem: 55573
Epoch: [34]  [ 10/156]  eta: 0:04:18  lr: 0.007950  min_lr: 0.007950  loss: 3.7036 (3.6902)  weight_decay: 0.0500 (0.0500)  time: 1.7734  data: 0.8601  max mem: 55573
Epoch: [34]  [ 20/156]  eta: 0:02:44  lr: 0.007950  min_lr: 0.007950  loss: 3.8188 (3.7520)  weight_decay: 0.0500 (0.0500)  time: 0.6807  data: 0.0005  max mem: 55573
Epoch: [34]  [ 30/156]  eta: 0:02:07  lr: 0.007949  min_lr: 0.007949  loss: 3.8188 (3.7398)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [34]  [ 40/156]  eta: 0:01:45  lr: 0.007949  min_lr: 0.007949  loss: 4.0815 (3.8101)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [34]  [ 50/156]  eta: 0:01:29  lr: 0.007948  min_lr: 0.007948  loss: 4.1681 (3.8586)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0004  max mem: 55573
Epoch: [34]  [ 60/156]  eta: 0:01:17  lr: 0.007948  min_lr: 0.007948  loss: 3.9746 (3.8518)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0005  max mem: 55573
Epoch: [34]  [ 70/156]  eta: 0:01:06  lr: 0.007948  min_lr: 0.007948  loss: 3.7215 (3.8419)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0005  max mem: 55573
Epoch: [34]  [ 80/156]  eta: 0:00:57  lr: 0.007947  min_lr: 0.007947  loss: 3.9190 (3.8360)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [34]  [ 90/156]  eta: 0:00:48  lr: 0.007947  min_lr: 0.007947  loss: 3.9190 (3.8383)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [34]  [100/156]  eta: 0:00:40  lr: 0.007946  min_lr: 0.007946  loss: 4.0646 (3.8632)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [34]  [110/156]  eta: 0:00:32  lr: 0.007946  min_lr: 0.007946  loss: 4.1066 (3.8711)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [34]  [120/156]  eta: 0:00:25  lr: 0.007945  min_lr: 0.007945  loss: 3.9027 (3.8476)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [34]  [130/156]  eta: 0:00:17  lr: 0.007945  min_lr: 0.007945  loss: 3.9121 (3.8491)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0010  max mem: 55573
Epoch: [34]  [140/156]  eta: 0:00:10  lr: 0.007944  min_lr: 0.007944  loss: 4.1295 (3.8505)  weight_decay: 0.0500 (0.0500)  time: 0.5863  data: 0.0009  max mem: 55573
Epoch: [34]  [150/156]  eta: 0:00:04  lr: 0.007944  min_lr: 0.007944  loss: 3.9719 (3.8579)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [34]  [155/156]  eta: 0:00:00  lr: 0.007944  min_lr: 0.007944  loss: 4.1492 (3.8588)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [34] Total time: 0:01:45 (0.6787 s / it)
Averaged stats: lr: 0.007944  min_lr: 0.007944  loss: 4.1492 (3.9114)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.5540 (1.5540)  acc1: 70.5729 (70.5729)  acc5: 90.7552 (90.7552)  time: 6.7660  data: 6.5281  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.6123 (1.6995)  acc1: 66.7969 (63.9040)  acc5: 89.7135 (86.0800)  time: 1.4998  data: 1.3057  max mem: 55573
Test: Total time: 0:00:07 (1.5436 s / it)
* Acc@1 64.036 Acc@5 86.398 loss 1.683
Accuracy of the model on the 50000 test images: 64.0%
Max accuracy: 64.44%
Test:  [0/5]  eta: 0:00:35  loss: 6.9027 (6.9027)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 7.0745  data: 6.8386  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9350 (6.9551)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.5601  data: 1.3678  max mem: 55573
Test: Total time: 0:00:07 (1.5793 s / it)
* Acc@1 0.100 Acc@5 0.506 loss 6.934
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [35]  [  0/156]  eta: 0:33:20  lr: 0.007943  min_lr: 0.007943  loss: 4.3675 (4.3675)  weight_decay: 0.0500 (0.0500)  time: 12.8215  data: 10.3399  max mem: 55573
Epoch: [35]  [ 10/156]  eta: 0:04:26  lr: 0.007943  min_lr: 0.007943  loss: 4.1266 (4.0783)  weight_decay: 0.0500 (0.0500)  time: 1.8232  data: 0.9404  max mem: 55573
Epoch: [35]  [ 20/156]  eta: 0:02:48  lr: 0.007943  min_lr: 0.007943  loss: 4.0234 (3.9470)  weight_decay: 0.0500 (0.0500)  time: 0.6564  data: 0.0003  max mem: 55573
Epoch: [35]  [ 30/156]  eta: 0:02:09  lr: 0.007942  min_lr: 0.007942  loss: 3.8776 (3.8948)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0003  max mem: 55573
Epoch: [35]  [ 40/156]  eta: 0:01:46  lr: 0.007942  min_lr: 0.007942  loss: 3.8776 (3.8647)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [35]  [ 50/156]  eta: 0:01:30  lr: 0.007941  min_lr: 0.007941  loss: 4.0763 (3.9005)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [35]  [ 60/156]  eta: 0:01:17  lr: 0.007941  min_lr: 0.007941  loss: 4.0591 (3.8974)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [35]  [ 70/156]  eta: 0:01:07  lr: 0.007940  min_lr: 0.007940  loss: 3.8981 (3.8903)  weight_decay: 0.0500 (0.0500)  time: 0.5881  data: 0.0005  max mem: 55573
Epoch: [35]  [ 80/156]  eta: 0:00:57  lr: 0.007940  min_lr: 0.007940  loss: 3.8008 (3.8741)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0005  max mem: 55573
Epoch: [35]  [ 90/156]  eta: 0:00:48  lr: 0.007939  min_lr: 0.007939  loss: 3.7913 (3.8355)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [35]  [100/156]  eta: 0:00:40  lr: 0.007939  min_lr: 0.007939  loss: 3.8418 (3.8276)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [35]  [110/156]  eta: 0:00:32  lr: 0.007938  min_lr: 0.007938  loss: 3.9976 (3.8530)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [35]  [120/156]  eta: 0:00:25  lr: 0.007938  min_lr: 0.007938  loss: 4.0409 (3.8464)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [35]  [130/156]  eta: 0:00:18  lr: 0.007937  min_lr: 0.007937  loss: 4.0739 (3.8653)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0011  max mem: 55573
Epoch: [35]  [140/156]  eta: 0:00:10  lr: 0.007937  min_lr: 0.007937  loss: 4.0575 (3.8620)  weight_decay: 0.0500 (0.0500)  time: 0.5861  data: 0.0009  max mem: 55573
Epoch: [35]  [150/156]  eta: 0:00:04  lr: 0.007936  min_lr: 0.007936  loss: 4.0010 (3.8606)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [35]  [155/156]  eta: 0:00:00  lr: 0.007936  min_lr: 0.007936  loss: 4.0852 (3.8594)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [35] Total time: 0:01:46 (0.6822 s / it)
Averaged stats: lr: 0.007936  min_lr: 0.007936  loss: 4.0852 (3.8772)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.4541 (1.4541)  acc1: 68.7500 (68.7500)  acc5: 91.2760 (91.2760)  time: 7.1459  data: 6.9085  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.6289 (1.7222)  acc1: 66.4062 (63.5200)  acc5: 88.6792 (86.5600)  time: 1.5755  data: 1.3818  max mem: 55573
Test: Total time: 0:00:08 (1.6173 s / it)
* Acc@1 63.878 Acc@5 86.618 loss 1.719
Accuracy of the model on the 50000 test images: 63.9%
Max accuracy: 64.44%
Test:  [0/5]  eta: 0:00:36  loss: 6.9032 (6.9032)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 7.3762  data: 7.1402  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9375 (6.9580)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.6205  data: 1.4282  max mem: 55573
Test: Total time: 0:00:08 (1.6398 s / it)
* Acc@1 0.098 Acc@5 0.512 loss 6.936
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [36]  [  0/156]  eta: 0:39:26  lr: 0.007936  min_lr: 0.007936  loss: 3.1379 (3.1379)  weight_decay: 0.0500 (0.0500)  time: 15.1713  data: 9.4588  max mem: 55573
Epoch: [36]  [ 10/156]  eta: 0:04:41  lr: 0.007935  min_lr: 0.007935  loss: 4.1412 (4.0353)  weight_decay: 0.0500 (0.0500)  time: 1.9300  data: 0.8603  max mem: 55573
Epoch: [36]  [ 20/156]  eta: 0:02:55  lr: 0.007935  min_lr: 0.007935  loss: 4.1412 (4.0850)  weight_decay: 0.0500 (0.0500)  time: 0.5987  data: 0.0004  max mem: 55573
Epoch: [36]  [ 30/156]  eta: 0:02:14  lr: 0.007934  min_lr: 0.007934  loss: 4.1199 (3.9811)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [36]  [ 40/156]  eta: 0:01:50  lr: 0.007934  min_lr: 0.007934  loss: 3.8677 (3.9185)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [36]  [ 50/156]  eta: 0:01:33  lr: 0.007933  min_lr: 0.007933  loss: 3.8486 (3.9306)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [36]  [ 60/156]  eta: 0:01:19  lr: 0.007933  min_lr: 0.007933  loss: 3.8431 (3.9116)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [36]  [ 70/156]  eta: 0:01:08  lr: 0.007932  min_lr: 0.007932  loss: 4.0550 (3.9217)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [36]  [ 80/156]  eta: 0:00:58  lr: 0.007932  min_lr: 0.007932  loss: 4.0596 (3.9311)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [36]  [ 90/156]  eta: 0:00:49  lr: 0.007931  min_lr: 0.007931  loss: 4.1431 (3.9288)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [36]  [100/156]  eta: 0:00:41  lr: 0.007930  min_lr: 0.007930  loss: 3.9484 (3.9044)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [36]  [110/156]  eta: 0:00:33  lr: 0.007930  min_lr: 0.007930  loss: 3.7784 (3.8993)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [36]  [120/156]  eta: 0:00:25  lr: 0.007929  min_lr: 0.007929  loss: 3.7784 (3.8796)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [36]  [130/156]  eta: 0:00:18  lr: 0.007929  min_lr: 0.007929  loss: 3.9321 (3.8808)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0010  max mem: 55573
Epoch: [36]  [140/156]  eta: 0:00:11  lr: 0.007928  min_lr: 0.007928  loss: 4.0445 (3.8880)  weight_decay: 0.0500 (0.0500)  time: 0.5860  data: 0.0009  max mem: 55573
Epoch: [36]  [150/156]  eta: 0:00:04  lr: 0.007928  min_lr: 0.007928  loss: 4.0653 (3.8918)  weight_decay: 0.0500 (0.0500)  time: 0.5808  data: 0.0001  max mem: 55573
Epoch: [36]  [155/156]  eta: 0:00:00  lr: 0.007928  min_lr: 0.007928  loss: 3.9706 (3.8935)  weight_decay: 0.0500 (0.0500)  time: 0.5811  data: 0.0001  max mem: 55573
Epoch: [36] Total time: 0:01:47 (0.6909 s / it)
Averaged stats: lr: 0.007928  min_lr: 0.007928  loss: 3.9706 (3.8850)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.6926 (1.6926)  acc1: 69.6615 (69.6615)  acc5: 90.8854 (90.8854)  time: 7.2076  data: 6.9701  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.9794 (2.0492)  acc1: 67.0573 (63.1040)  acc5: 88.9323 (85.4080)  time: 1.5877  data: 1.3941  max mem: 55573
Test: Total time: 0:00:08 (1.6266 s / it)
* Acc@1 63.326 Acc@5 85.734 loss 2.063
Accuracy of the model on the 50000 test images: 63.3%
Max accuracy: 64.44%
Test:  [0/5]  eta: 0:00:32  loss: 6.9048 (6.9048)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 6.5594  data: 6.3234  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9398 (6.9601)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.4612  data: 1.2688  max mem: 55573
Test: Total time: 0:00:07 (1.4804 s / it)
* Acc@1 0.100 Acc@5 0.516 loss 6.937
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [37]  [  0/156]  eta: 0:36:02  lr: 0.007927  min_lr: 0.007927  loss: 2.9370 (2.9370)  weight_decay: 0.0500 (0.0500)  time: 13.8621  data: 9.5656  max mem: 55573
Epoch: [37]  [ 10/156]  eta: 0:04:25  lr: 0.007927  min_lr: 0.007927  loss: 3.6975 (3.7159)  weight_decay: 0.0500 (0.0500)  time: 1.8167  data: 0.8700  max mem: 55573
Epoch: [37]  [ 20/156]  eta: 0:02:47  lr: 0.007926  min_lr: 0.007926  loss: 3.8209 (3.7205)  weight_decay: 0.0500 (0.0500)  time: 0.6013  data: 0.0004  max mem: 55573
Epoch: [37]  [ 30/156]  eta: 0:02:09  lr: 0.007926  min_lr: 0.007926  loss: 3.7957 (3.7932)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [37]  [ 40/156]  eta: 0:01:46  lr: 0.007925  min_lr: 0.007925  loss: 3.7957 (3.8185)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [37]  [ 50/156]  eta: 0:01:30  lr: 0.007925  min_lr: 0.007925  loss: 4.1329 (3.8423)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [37]  [ 60/156]  eta: 0:01:17  lr: 0.007924  min_lr: 0.007924  loss: 4.1590 (3.8739)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [37]  [ 70/156]  eta: 0:01:07  lr: 0.007924  min_lr: 0.007924  loss: 3.9903 (3.8530)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [37]  [ 80/156]  eta: 0:00:57  lr: 0.007923  min_lr: 0.007923  loss: 3.8599 (3.8454)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0006  max mem: 55573
Epoch: [37]  [ 90/156]  eta: 0:00:48  lr: 0.007922  min_lr: 0.007922  loss: 4.0552 (3.8810)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [37]  [100/156]  eta: 0:00:40  lr: 0.007922  min_lr: 0.007922  loss: 4.2634 (3.9202)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0004  max mem: 55573
Epoch: [37]  [110/156]  eta: 0:00:32  lr: 0.007921  min_lr: 0.007921  loss: 4.1435 (3.9271)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [37]  [120/156]  eta: 0:00:25  lr: 0.007921  min_lr: 0.007921  loss: 4.1397 (3.9294)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [37]  [130/156]  eta: 0:00:18  lr: 0.007920  min_lr: 0.007920  loss: 3.8708 (3.8953)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0013  max mem: 55573
Epoch: [37]  [140/156]  eta: 0:00:10  lr: 0.007920  min_lr: 0.007920  loss: 3.8708 (3.9098)  weight_decay: 0.0500 (0.0500)  time: 0.5867  data: 0.0011  max mem: 55573
Epoch: [37]  [150/156]  eta: 0:00:04  lr: 0.007919  min_lr: 0.007919  loss: 4.2078 (3.9152)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [37]  [155/156]  eta: 0:00:00  lr: 0.007919  min_lr: 0.007919  loss: 4.1963 (3.9083)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [37] Total time: 0:01:46 (0.6820 s / it)
Averaged stats: lr: 0.007919  min_lr: 0.007919  loss: 4.1963 (3.8795)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.5819 (1.5819)  acc1: 71.4844 (71.4844)  acc5: 89.0625 (89.0625)  time: 6.9539  data: 6.7165  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5819 (1.5995)  acc1: 68.8802 (65.7920)  acc5: 89.0625 (87.3600)  time: 1.5374  data: 1.3434  max mem: 55573
Test: Total time: 0:00:07 (1.5841 s / it)
* Acc@1 65.640 Acc@5 87.394 loss 1.605
Accuracy of the model on the 50000 test images: 65.6%
Max accuracy: 65.64%
Test:  [0/5]  eta: 0:00:35  loss: 6.9058 (6.9058)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 7.0229  data: 6.7868  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9419 (6.9621)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.5498  data: 1.3575  max mem: 55573
Test: Total time: 0:00:07 (1.5639 s / it)
* Acc@1 0.098 Acc@5 0.512 loss 6.938
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [38]  [  0/156]  eta: 0:33:13  lr: 0.007919  min_lr: 0.007919  loss: 4.1979 (4.1979)  weight_decay: 0.0500 (0.0500)  time: 12.7773  data: 10.9820  max mem: 55573
Epoch: [38]  [ 10/156]  eta: 0:04:10  lr: 0.007918  min_lr: 0.007918  loss: 4.1669 (4.0397)  weight_decay: 0.0500 (0.0500)  time: 1.7128  data: 0.9989  max mem: 55573
Epoch: [38]  [ 20/156]  eta: 0:02:40  lr: 0.007918  min_lr: 0.007918  loss: 4.0312 (3.9746)  weight_decay: 0.0500 (0.0500)  time: 0.5988  data: 0.0005  max mem: 55573
Epoch: [38]  [ 30/156]  eta: 0:02:04  lr: 0.007917  min_lr: 0.007917  loss: 3.9099 (3.8376)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0004  max mem: 55573
Epoch: [38]  [ 40/156]  eta: 0:01:43  lr: 0.007916  min_lr: 0.007916  loss: 4.0551 (3.8841)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0004  max mem: 55573
Epoch: [38]  [ 50/156]  eta: 0:01:28  lr: 0.007916  min_lr: 0.007916  loss: 4.1520 (3.8890)  weight_decay: 0.0500 (0.0500)  time: 0.5964  data: 0.0004  max mem: 55573
Epoch: [38]  [ 60/156]  eta: 0:01:16  lr: 0.007915  min_lr: 0.007915  loss: 4.0162 (3.9116)  weight_decay: 0.0500 (0.0500)  time: 0.5959  data: 0.0005  max mem: 55573
Epoch: [38]  [ 70/156]  eta: 0:01:05  lr: 0.007915  min_lr: 0.007915  loss: 3.9426 (3.9023)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [38]  [ 80/156]  eta: 0:00:56  lr: 0.007914  min_lr: 0.007914  loss: 4.0145 (3.9088)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [38]  [ 90/156]  eta: 0:00:48  lr: 0.007913  min_lr: 0.007913  loss: 4.0887 (3.9140)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [38]  [100/156]  eta: 0:00:40  lr: 0.007913  min_lr: 0.007913  loss: 3.9455 (3.8951)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [38]  [110/156]  eta: 0:00:32  lr: 0.007912  min_lr: 0.007912  loss: 3.7328 (3.8746)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [38]  [120/156]  eta: 0:00:24  lr: 0.007912  min_lr: 0.007912  loss: 3.7328 (3.8632)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [38]  [130/156]  eta: 0:00:17  lr: 0.007911  min_lr: 0.007911  loss: 3.7605 (3.8621)  weight_decay: 0.0500 (0.0500)  time: 0.5879  data: 0.0010  max mem: 55573
Epoch: [38]  [140/156]  eta: 0:00:10  lr: 0.007910  min_lr: 0.007910  loss: 4.1293 (3.8866)  weight_decay: 0.0500 (0.0500)  time: 0.5847  data: 0.0008  max mem: 55573
Epoch: [38]  [150/156]  eta: 0:00:04  lr: 0.007910  min_lr: 0.007910  loss: 4.1898 (3.8950)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [38]  [155/156]  eta: 0:00:00  lr: 0.007910  min_lr: 0.007910  loss: 4.1293 (3.8927)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [38] Total time: 0:01:45 (0.6757 s / it)
Averaged stats: lr: 0.007910  min_lr: 0.007910  loss: 4.1293 (3.8546)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.5498 (1.5498)  acc1: 69.0104 (69.0104)  acc5: 91.4062 (91.4062)  time: 6.8416  data: 6.6043  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.6334 (1.6642)  acc1: 66.9271 (64.0640)  acc5: 90.6250 (87.4240)  time: 1.5148  data: 1.3209  max mem: 55573
Test: Total time: 0:00:07 (1.5559 s / it)
* Acc@1 64.618 Acc@5 87.266 loss 1.667
Accuracy of the model on the 50000 test images: 64.6%
Max accuracy: 65.64%
Test:  [0/5]  eta: 0:00:36  loss: 6.9070 (6.9070)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 7.3092  data: 7.0723  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9439 (6.9640)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.6078  data: 1.4146  max mem: 55573
Test: Total time: 0:00:08 (1.6283 s / it)
* Acc@1 0.096 Acc@5 0.508 loss 6.939
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [39]  [  0/156]  eta: 0:37:34  lr: 0.007909  min_lr: 0.007909  loss: 3.3552 (3.3552)  weight_decay: 0.0500 (0.0500)  time: 14.4490  data: 9.5486  max mem: 55573
Epoch: [39]  [ 10/156]  eta: 0:04:30  lr: 0.007909  min_lr: 0.007909  loss: 4.0679 (3.8349)  weight_decay: 0.0500 (0.0500)  time: 1.8534  data: 0.8685  max mem: 55573
Epoch: [39]  [ 20/156]  eta: 0:02:50  lr: 0.007908  min_lr: 0.007908  loss: 3.8955 (3.8275)  weight_decay: 0.0500 (0.0500)  time: 0.5934  data: 0.0005  max mem: 55573
Epoch: [39]  [ 30/156]  eta: 0:02:11  lr: 0.007908  min_lr: 0.007908  loss: 3.5375 (3.7250)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [39]  [ 40/156]  eta: 0:01:47  lr: 0.007907  min_lr: 0.007907  loss: 3.5756 (3.7484)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [39]  [ 50/156]  eta: 0:01:31  lr: 0.007906  min_lr: 0.007906  loss: 3.9392 (3.7852)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [39]  [ 60/156]  eta: 0:01:18  lr: 0.007906  min_lr: 0.007906  loss: 3.6848 (3.7381)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [39]  [ 70/156]  eta: 0:01:07  lr: 0.007905  min_lr: 0.007905  loss: 3.6622 (3.7603)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [39]  [ 80/156]  eta: 0:00:57  lr: 0.007905  min_lr: 0.007905  loss: 3.9881 (3.7718)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [39]  [ 90/156]  eta: 0:00:49  lr: 0.007904  min_lr: 0.007904  loss: 4.0309 (3.8003)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [39]  [100/156]  eta: 0:00:40  lr: 0.007903  min_lr: 0.007903  loss: 3.9594 (3.7999)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [39]  [110/156]  eta: 0:00:32  lr: 0.007903  min_lr: 0.007903  loss: 4.0513 (3.8117)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [39]  [120/156]  eta: 0:00:25  lr: 0.007902  min_lr: 0.007902  loss: 4.0791 (3.8292)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [39]  [130/156]  eta: 0:00:18  lr: 0.007901  min_lr: 0.007901  loss: 3.6313 (3.7943)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0010  max mem: 55573
Epoch: [39]  [140/156]  eta: 0:00:11  lr: 0.007901  min_lr: 0.007901  loss: 3.7127 (3.8032)  weight_decay: 0.0500 (0.0500)  time: 0.5859  data: 0.0009  max mem: 55573
Epoch: [39]  [150/156]  eta: 0:00:04  lr: 0.007900  min_lr: 0.007900  loss: 3.8434 (3.7917)  weight_decay: 0.0500 (0.0500)  time: 0.5812  data: 0.0001  max mem: 55573
Epoch: [39]  [155/156]  eta: 0:00:00  lr: 0.007900  min_lr: 0.007900  loss: 3.9437 (3.8049)  weight_decay: 0.0500 (0.0500)  time: 0.5811  data: 0.0001  max mem: 55573
Epoch: [39] Total time: 0:01:46 (0.6853 s / it)
Averaged stats: lr: 0.007900  min_lr: 0.007900  loss: 3.9437 (3.8457)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.5038 (1.5038)  acc1: 72.2656 (72.2656)  acc5: 91.5365 (91.5365)  time: 6.8858  data: 6.6469  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5626 (1.6040)  acc1: 67.4479 (65.3120)  acc5: 90.4948 (87.3920)  time: 1.5238  data: 1.3295  max mem: 55573
Test: Total time: 0:00:07 (1.5738 s / it)
* Acc@1 66.260 Acc@5 87.454 loss 1.582
Accuracy of the model on the 50000 test images: 66.3%
Max accuracy: 66.26%
Test:  [0/5]  eta: 0:00:35  loss: 6.9077 (6.9077)  acc1: 0.0000 (0.0000)  acc5: 0.3906 (0.3906)  time: 7.1500  data: 6.9141  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9471 (6.9661)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.5752  data: 1.3829  max mem: 55573
Test: Total time: 0:00:07 (1.5918 s / it)
* Acc@1 0.096 Acc@5 0.504 loss 6.940
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [40]  [  0/156]  eta: 0:35:27  lr: 0.007900  min_lr: 0.007900  loss: 3.8804 (3.8804)  weight_decay: 0.0500 (0.0500)  time: 13.6359  data: 11.0192  max mem: 55573
Epoch: [40]  [ 10/156]  eta: 0:04:19  lr: 0.007899  min_lr: 0.007899  loss: 4.2174 (4.1050)  weight_decay: 0.0500 (0.0500)  time: 1.7796  data: 1.0022  max mem: 55573
Epoch: [40]  [ 20/156]  eta: 0:02:45  lr: 0.007898  min_lr: 0.007898  loss: 4.0518 (4.0012)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0005  max mem: 55573
Epoch: [40]  [ 30/156]  eta: 0:02:07  lr: 0.007898  min_lr: 0.007898  loss: 4.0356 (4.0520)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [40]  [ 40/156]  eta: 0:01:45  lr: 0.007897  min_lr: 0.007897  loss: 4.0488 (3.9868)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [40]  [ 50/156]  eta: 0:01:29  lr: 0.007896  min_lr: 0.007896  loss: 3.9284 (3.9848)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [40]  [ 60/156]  eta: 0:01:17  lr: 0.007896  min_lr: 0.007896  loss: 4.1112 (4.0125)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0004  max mem: 55573
Epoch: [40]  [ 70/156]  eta: 0:01:06  lr: 0.007895  min_lr: 0.007895  loss: 4.2110 (4.0452)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [40]  [ 80/156]  eta: 0:00:57  lr: 0.007895  min_lr: 0.007895  loss: 4.0877 (4.0027)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0004  max mem: 55573
Epoch: [40]  [ 90/156]  eta: 0:00:48  lr: 0.007894  min_lr: 0.007894  loss: 3.9907 (3.9923)  weight_decay: 0.0500 (0.0500)  time: 0.5884  data: 0.0005  max mem: 55573
Epoch: [40]  [100/156]  eta: 0:00:40  lr: 0.007893  min_lr: 0.007893  loss: 3.9640 (3.9788)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0005  max mem: 55573
Epoch: [40]  [110/156]  eta: 0:00:32  lr: 0.007893  min_lr: 0.007893  loss: 4.0080 (3.9641)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0005  max mem: 55573
Epoch: [40]  [120/156]  eta: 0:00:25  lr: 0.007892  min_lr: 0.007892  loss: 4.0566 (3.9525)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [40]  [130/156]  eta: 0:00:17  lr: 0.007891  min_lr: 0.007891  loss: 4.0878 (3.9516)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0010  max mem: 55573
Epoch: [40]  [140/156]  eta: 0:00:10  lr: 0.007891  min_lr: 0.007891  loss: 4.0928 (3.9599)  weight_decay: 0.0500 (0.0500)  time: 0.5859  data: 0.0009  max mem: 55573
Epoch: [40]  [150/156]  eta: 0:00:04  lr: 0.007890  min_lr: 0.007890  loss: 4.0776 (3.9483)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [40]  [155/156]  eta: 0:00:00  lr: 0.007890  min_lr: 0.007890  loss: 3.7692 (3.9492)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [40] Total time: 0:01:46 (0.6797 s / it)
Averaged stats: lr: 0.007890  min_lr: 0.007890  loss: 3.7692 (3.8451)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.6127 (1.6127)  acc1: 70.7031 (70.7031)  acc5: 91.0156 (91.0156)  time: 6.8432  data: 6.6053  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.7863 (1.8060)  acc1: 66.4062 (63.7760)  acc5: 88.4115 (85.9520)  time: 1.5151  data: 1.3212  max mem: 55573
Test: Total time: 0:00:07 (1.5579 s / it)
* Acc@1 64.718 Acc@5 86.814 loss 1.807
Accuracy of the model on the 50000 test images: 64.7%
Max accuracy: 66.26%
Test:  [0/5]  eta: 0:00:35  loss: 6.9092 (6.9092)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.1908  data: 6.9549  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9507 (6.9685)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.5835  data: 1.3911  max mem: 55573
Test: Total time: 0:00:08 (1.6063 s / it)
* Acc@1 0.096 Acc@5 0.508 loss 6.942
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [41]  [  0/156]  eta: 0:37:09  lr: 0.007889  min_lr: 0.007889  loss: 4.1249 (4.1249)  weight_decay: 0.0500 (0.0500)  time: 14.2924  data: 12.0192  max mem: 55573
Epoch: [41]  [ 10/156]  eta: 0:04:30  lr: 0.007889  min_lr: 0.007889  loss: 3.8268 (3.8527)  weight_decay: 0.0500 (0.0500)  time: 1.8549  data: 1.0930  max mem: 55573
Epoch: [41]  [ 20/156]  eta: 0:02:50  lr: 0.007888  min_lr: 0.007888  loss: 3.7609 (3.8465)  weight_decay: 0.0500 (0.0500)  time: 0.6016  data: 0.0004  max mem: 55573
Epoch: [41]  [ 30/156]  eta: 0:02:11  lr: 0.007887  min_lr: 0.007887  loss: 4.0573 (3.9043)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [41]  [ 40/156]  eta: 0:01:47  lr: 0.007887  min_lr: 0.007887  loss: 3.7810 (3.8127)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [41]  [ 50/156]  eta: 0:01:31  lr: 0.007886  min_lr: 0.007886  loss: 3.6182 (3.7700)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [41]  [ 60/156]  eta: 0:01:18  lr: 0.007885  min_lr: 0.007885  loss: 3.9982 (3.7722)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [41]  [ 70/156]  eta: 0:01:07  lr: 0.007885  min_lr: 0.007885  loss: 3.8091 (3.7366)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [41]  [ 80/156]  eta: 0:00:57  lr: 0.007884  min_lr: 0.007884  loss: 3.4140 (3.7205)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [41]  [ 90/156]  eta: 0:00:49  lr: 0.007883  min_lr: 0.007883  loss: 3.8200 (3.7380)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [41]  [100/156]  eta: 0:00:40  lr: 0.007883  min_lr: 0.007883  loss: 3.9871 (3.7432)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [41]  [110/156]  eta: 0:00:32  lr: 0.007882  min_lr: 0.007882  loss: 3.9133 (3.7457)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [41]  [120/156]  eta: 0:00:25  lr: 0.007881  min_lr: 0.007881  loss: 3.9396 (3.7603)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [41]  [130/156]  eta: 0:00:18  lr: 0.007881  min_lr: 0.007881  loss: 3.9396 (3.7575)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0010  max mem: 55573
Epoch: [41]  [140/156]  eta: 0:00:11  lr: 0.007880  min_lr: 0.007880  loss: 4.0687 (3.7703)  weight_decay: 0.0500 (0.0500)  time: 0.5862  data: 0.0009  max mem: 55573
Epoch: [41]  [150/156]  eta: 0:00:04  lr: 0.007879  min_lr: 0.007879  loss: 4.0816 (3.7832)  weight_decay: 0.0500 (0.0500)  time: 0.5831  data: 0.0001  max mem: 55573
Epoch: [41]  [155/156]  eta: 0:00:00  lr: 0.007879  min_lr: 0.007879  loss: 4.1181 (3.7949)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [41] Total time: 0:01:46 (0.6848 s / it)
Averaged stats: lr: 0.007879  min_lr: 0.007879  loss: 4.1181 (3.8207)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.6496 (1.6496)  acc1: 69.6615 (69.6615)  acc5: 90.4948 (90.4948)  time: 7.1259  data: 6.8889  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.7219 (1.7574)  acc1: 67.3177 (63.9360)  acc5: 90.1042 (87.0080)  time: 1.5715  data: 1.3779  max mem: 55573
Test: Total time: 0:00:08 (1.6225 s / it)
* Acc@1 65.092 Acc@5 86.582 loss 1.736
Accuracy of the model on the 50000 test images: 65.1%
Max accuracy: 66.26%
Test:  [0/5]  eta: 0:00:35  loss: 6.9092 (6.9092)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.1588  data: 6.9228  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9546 (6.9709)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.5770  data: 1.3847  max mem: 55573
Test: Total time: 0:00:07 (1.5947 s / it)
* Acc@1 0.094 Acc@5 0.506 loss 6.943
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [42]  [  0/156]  eta: 0:32:21  lr: 0.007879  min_lr: 0.007879  loss: 3.3599 (3.3599)  weight_decay: 0.0500 (0.0500)  time: 12.4469  data: 10.7423  max mem: 55573
Epoch: [42]  [ 10/156]  eta: 0:04:08  lr: 0.007878  min_lr: 0.007878  loss: 3.4847 (3.7164)  weight_decay: 0.0500 (0.0500)  time: 1.7049  data: 0.9770  max mem: 55573
Epoch: [42]  [ 20/156]  eta: 0:02:40  lr: 0.007877  min_lr: 0.007877  loss: 3.8078 (3.9080)  weight_decay: 0.0500 (0.0500)  time: 0.6144  data: 0.0006  max mem: 55573
Epoch: [42]  [ 30/156]  eta: 0:02:04  lr: 0.007877  min_lr: 0.007877  loss: 3.9619 (3.8031)  weight_decay: 0.0500 (0.0500)  time: 0.5946  data: 0.0005  max mem: 55573
Epoch: [42]  [ 40/156]  eta: 0:01:43  lr: 0.007876  min_lr: 0.007876  loss: 3.9197 (3.8539)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [42]  [ 50/156]  eta: 0:01:28  lr: 0.007875  min_lr: 0.007875  loss: 3.9707 (3.8455)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [42]  [ 60/156]  eta: 0:01:16  lr: 0.007875  min_lr: 0.007875  loss: 3.6706 (3.8037)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [42]  [ 70/156]  eta: 0:01:05  lr: 0.007874  min_lr: 0.007874  loss: 3.9559 (3.8376)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0004  max mem: 55573
Epoch: [42]  [ 80/156]  eta: 0:00:56  lr: 0.007873  min_lr: 0.007873  loss: 3.9559 (3.8100)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [42]  [ 90/156]  eta: 0:00:47  lr: 0.007872  min_lr: 0.007872  loss: 3.9370 (3.8197)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [42]  [100/156]  eta: 0:00:39  lr: 0.007872  min_lr: 0.007872  loss: 3.8672 (3.8194)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [42]  [110/156]  eta: 0:00:32  lr: 0.007871  min_lr: 0.007871  loss: 3.8865 (3.8132)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [42]  [120/156]  eta: 0:00:24  lr: 0.007870  min_lr: 0.007870  loss: 3.8865 (3.7990)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [42]  [130/156]  eta: 0:00:17  lr: 0.007869  min_lr: 0.007869  loss: 3.7478 (3.7853)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0013  max mem: 55573
Epoch: [42]  [140/156]  eta: 0:00:10  lr: 0.007869  min_lr: 0.007869  loss: 3.6228 (3.7706)  weight_decay: 0.0500 (0.0500)  time: 0.5858  data: 0.0011  max mem: 55573
Epoch: [42]  [150/156]  eta: 0:00:04  lr: 0.007868  min_lr: 0.007868  loss: 3.8377 (3.7904)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0001  max mem: 55573
Epoch: [42]  [155/156]  eta: 0:00:00  lr: 0.007868  min_lr: 0.007868  loss: 4.0923 (3.7989)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0001  max mem: 55573
Epoch: [42] Total time: 0:01:45 (0.6753 s / it)
Averaged stats: lr: 0.007868  min_lr: 0.007868  loss: 4.0923 (3.8069)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.4277 (1.4277)  acc1: 72.1354 (72.1354)  acc5: 92.0573 (92.0573)  time: 7.0471  data: 6.8097  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.6825 (1.6366)  acc1: 66.5365 (65.1840)  acc5: 90.1042 (88.1280)  time: 1.5560  data: 1.3620  max mem: 55573
Test: Total time: 0:00:08 (1.6043 s / it)
* Acc@1 65.844 Acc@5 87.834 loss 1.634
Accuracy of the model on the 50000 test images: 65.8%
Max accuracy: 66.26%
Test:  [0/5]  eta: 0:00:35  loss: 6.9090 (6.9090)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.0537  data: 6.8177  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9590 (6.9734)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.5561  data: 1.3636  max mem: 55573
Test: Total time: 0:00:07 (1.5722 s / it)
* Acc@1 0.094 Acc@5 0.504 loss 6.945
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [43]  [  0/156]  eta: 0:36:56  lr: 0.007868  min_lr: 0.007868  loss: 3.6134 (3.6134)  weight_decay: 0.0500 (0.0500)  time: 14.2053  data: 11.3915  max mem: 55573
Epoch: [43]  [ 10/156]  eta: 0:04:29  lr: 0.007867  min_lr: 0.007867  loss: 3.9085 (3.9599)  weight_decay: 0.0500 (0.0500)  time: 1.8437  data: 1.0359  max mem: 55573
Epoch: [43]  [ 20/156]  eta: 0:02:49  lr: 0.007866  min_lr: 0.007866  loss: 3.9973 (3.9403)  weight_decay: 0.0500 (0.0500)  time: 0.5993  data: 0.0004  max mem: 55573
Epoch: [43]  [ 30/156]  eta: 0:02:10  lr: 0.007865  min_lr: 0.007865  loss: 3.9030 (3.8935)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [43]  [ 40/156]  eta: 0:01:47  lr: 0.007865  min_lr: 0.007865  loss: 3.3381 (3.7191)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [43]  [ 50/156]  eta: 0:01:31  lr: 0.007864  min_lr: 0.007864  loss: 3.3260 (3.7419)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [43]  [ 60/156]  eta: 0:01:18  lr: 0.007863  min_lr: 0.007863  loss: 3.7889 (3.7235)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [43]  [ 70/156]  eta: 0:01:07  lr: 0.007862  min_lr: 0.007862  loss: 3.8261 (3.7666)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [43]  [ 80/156]  eta: 0:00:57  lr: 0.007862  min_lr: 0.007862  loss: 4.1444 (3.7996)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [43]  [ 90/156]  eta: 0:00:48  lr: 0.007861  min_lr: 0.007861  loss: 4.1319 (3.8055)  weight_decay: 0.0500 (0.0500)  time: 0.5887  data: 0.0004  max mem: 55573
Epoch: [43]  [100/156]  eta: 0:00:40  lr: 0.007860  min_lr: 0.007860  loss: 4.0954 (3.8022)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [43]  [110/156]  eta: 0:00:32  lr: 0.007859  min_lr: 0.007859  loss: 3.8918 (3.8038)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [43]  [120/156]  eta: 0:00:25  lr: 0.007859  min_lr: 0.007859  loss: 3.7965 (3.8094)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [43]  [130/156]  eta: 0:00:18  lr: 0.007858  min_lr: 0.007858  loss: 3.7810 (3.7959)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0010  max mem: 55573
Epoch: [43]  [140/156]  eta: 0:00:11  lr: 0.007857  min_lr: 0.007857  loss: 3.7810 (3.7975)  weight_decay: 0.0500 (0.0500)  time: 0.5852  data: 0.0009  max mem: 55573
Epoch: [43]  [150/156]  eta: 0:00:04  lr: 0.007856  min_lr: 0.007856  loss: 3.5130 (3.7754)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [43]  [155/156]  eta: 0:00:00  lr: 0.007856  min_lr: 0.007856  loss: 3.4231 (3.7791)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [43] Total time: 0:01:46 (0.6843 s / it)
Averaged stats: lr: 0.007856  min_lr: 0.007856  loss: 3.4231 (3.7971)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.5428 (1.5428)  acc1: 71.4844 (71.4844)  acc5: 91.2760 (91.2760)  time: 7.2828  data: 7.0454  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5598 (1.6263)  acc1: 69.6615 (65.4400)  acc5: 89.8438 (87.4240)  time: 1.6031  data: 1.4092  max mem: 55573
Test: Total time: 0:00:08 (1.6587 s / it)
* Acc@1 65.914 Acc@5 87.544 loss 1.654
Accuracy of the model on the 50000 test images: 65.9%
Max accuracy: 66.26%
Test:  [0/5]  eta: 0:00:35  loss: 6.9091 (6.9091)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.0492  data: 6.8133  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9645 (6.9763)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.5551  data: 1.3628  max mem: 55573
Test: Total time: 0:00:07 (1.5720 s / it)
* Acc@1 0.098 Acc@5 0.506 loss 6.947
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [44]  [  0/156]  eta: 0:34:14  lr: 0.007856  min_lr: 0.007856  loss: 3.4522 (3.4522)  weight_decay: 0.0500 (0.0500)  time: 13.1678  data: 10.4710  max mem: 55573
Epoch: [44]  [ 10/156]  eta: 0:04:15  lr: 0.007855  min_lr: 0.007855  loss: 3.8639 (3.8579)  weight_decay: 0.0500 (0.0500)  time: 1.7519  data: 0.9523  max mem: 55573
Epoch: [44]  [ 20/156]  eta: 0:02:43  lr: 0.007854  min_lr: 0.007854  loss: 3.7823 (3.7451)  weight_decay: 0.0500 (0.0500)  time: 0.6005  data: 0.0005  max mem: 55573
Epoch: [44]  [ 30/156]  eta: 0:02:06  lr: 0.007854  min_lr: 0.007854  loss: 3.8285 (3.7687)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [44]  [ 40/156]  eta: 0:01:44  lr: 0.007853  min_lr: 0.007853  loss: 3.8285 (3.7248)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [44]  [ 50/156]  eta: 0:01:29  lr: 0.007852  min_lr: 0.007852  loss: 3.7053 (3.7326)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0004  max mem: 55573
Epoch: [44]  [ 60/156]  eta: 0:01:17  lr: 0.007851  min_lr: 0.007851  loss: 3.6318 (3.7254)  weight_decay: 0.0500 (0.0500)  time: 0.5985  data: 0.0004  max mem: 55573
Epoch: [44]  [ 70/156]  eta: 0:01:06  lr: 0.007850  min_lr: 0.007850  loss: 3.7081 (3.7544)  weight_decay: 0.0500 (0.0500)  time: 0.5980  data: 0.0004  max mem: 55573
Epoch: [44]  [ 80/156]  eta: 0:00:56  lr: 0.007850  min_lr: 0.007850  loss: 3.7085 (3.7320)  weight_decay: 0.0500 (0.0500)  time: 0.5878  data: 0.0004  max mem: 55573
Epoch: [44]  [ 90/156]  eta: 0:00:48  lr: 0.007849  min_lr: 0.007849  loss: 3.7085 (3.7114)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [44]  [100/156]  eta: 0:00:40  lr: 0.007848  min_lr: 0.007848  loss: 3.8692 (3.7306)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [44]  [110/156]  eta: 0:00:32  lr: 0.007847  min_lr: 0.007847  loss: 3.8375 (3.7190)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [44]  [120/156]  eta: 0:00:25  lr: 0.007847  min_lr: 0.007847  loss: 3.7289 (3.7286)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0004  max mem: 55573
Epoch: [44]  [130/156]  eta: 0:00:17  lr: 0.007846  min_lr: 0.007846  loss: 3.8847 (3.7273)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0010  max mem: 55573
Epoch: [44]  [140/156]  eta: 0:00:10  lr: 0.007845  min_lr: 0.007845  loss: 3.4793 (3.7026)  weight_decay: 0.0500 (0.0500)  time: 0.5861  data: 0.0008  max mem: 55573
Epoch: [44]  [150/156]  eta: 0:00:04  lr: 0.007844  min_lr: 0.007844  loss: 3.5604 (3.7078)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [44]  [155/156]  eta: 0:00:00  lr: 0.007844  min_lr: 0.007844  loss: 3.7624 (3.7120)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [44] Total time: 0:01:45 (0.6782 s / it)
Averaged stats: lr: 0.007844  min_lr: 0.007844  loss: 3.7624 (3.7888)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.2774 (1.2774)  acc1: 73.3073 (73.3073)  acc5: 93.6198 (93.6198)  time: 7.1497  data: 6.9122  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5526 (1.5871)  acc1: 68.2292 (66.7200)  acc5: 89.8438 (87.8080)  time: 1.5765  data: 1.3825  max mem: 55573
Test: Total time: 0:00:08 (1.6173 s / it)
* Acc@1 66.512 Acc@5 87.896 loss 1.589
Accuracy of the model on the 50000 test images: 66.5%
Max accuracy: 66.51%
Test:  [0/5]  eta: 0:00:33  loss: 6.9095 (6.9095)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.7734  data: 6.5373  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9705 (6.9799)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.5000  data: 1.3076  max mem: 55573
Test: Total time: 0:00:07 (1.5201 s / it)
* Acc@1 0.100 Acc@5 0.508 loss 6.949
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [45]  [  0/156]  eta: 0:37:15  lr: 0.007844  min_lr: 0.007844  loss: 3.4263 (3.4263)  weight_decay: 0.0500 (0.0500)  time: 14.3271  data: 10.5767  max mem: 55573
Epoch: [45]  [ 10/156]  eta: 0:04:28  lr: 0.007843  min_lr: 0.007843  loss: 4.0582 (3.7897)  weight_decay: 0.0500 (0.0500)  time: 1.8414  data: 0.9619  max mem: 55573
Epoch: [45]  [ 20/156]  eta: 0:02:49  lr: 0.007842  min_lr: 0.007842  loss: 3.7817 (3.7275)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [45]  [ 30/156]  eta: 0:02:10  lr: 0.007841  min_lr: 0.007841  loss: 3.9123 (3.7163)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [45]  [ 40/156]  eta: 0:01:47  lr: 0.007840  min_lr: 0.007840  loss: 4.0048 (3.7358)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [45]  [ 50/156]  eta: 0:01:31  lr: 0.007840  min_lr: 0.007840  loss: 3.9427 (3.7600)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0005  max mem: 55573
Epoch: [45]  [ 60/156]  eta: 0:01:18  lr: 0.007839  min_lr: 0.007839  loss: 3.7522 (3.6987)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [45]  [ 70/156]  eta: 0:01:07  lr: 0.007838  min_lr: 0.007838  loss: 3.6746 (3.7187)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [45]  [ 80/156]  eta: 0:00:57  lr: 0.007837  min_lr: 0.007837  loss: 3.8787 (3.7462)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [45]  [ 90/156]  eta: 0:00:48  lr: 0.007836  min_lr: 0.007836  loss: 3.5823 (3.7095)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [45]  [100/156]  eta: 0:00:40  lr: 0.007836  min_lr: 0.007836  loss: 3.4573 (3.7256)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [45]  [110/156]  eta: 0:00:32  lr: 0.007835  min_lr: 0.007835  loss: 4.0694 (3.7364)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [45]  [120/156]  eta: 0:00:25  lr: 0.007834  min_lr: 0.007834  loss: 3.8887 (3.7336)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [45]  [130/156]  eta: 0:00:18  lr: 0.007833  min_lr: 0.007833  loss: 3.9425 (3.7609)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0012  max mem: 55573
Epoch: [45]  [140/156]  eta: 0:00:10  lr: 0.007832  min_lr: 0.007832  loss: 4.3430 (3.8006)  weight_decay: 0.0500 (0.0500)  time: 0.5850  data: 0.0010  max mem: 55573
Epoch: [45]  [150/156]  eta: 0:00:04  lr: 0.007832  min_lr: 0.007832  loss: 4.0825 (3.8142)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [45]  [155/156]  eta: 0:00:00  lr: 0.007831  min_lr: 0.007831  loss: 4.0524 (3.8110)  weight_decay: 0.0500 (0.0500)  time: 0.5812  data: 0.0001  max mem: 55573
Epoch: [45] Total time: 0:01:46 (0.6838 s / it)
Averaged stats: lr: 0.007831  min_lr: 0.007831  loss: 4.0524 (3.7865)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.4144 (1.4144)  acc1: 72.6562 (72.6562)  acc5: 91.9271 (91.9271)  time: 6.8671  data: 6.6297  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.6038 (1.6222)  acc1: 68.8802 (66.4320)  acc5: 91.9271 (88.8320)  time: 1.5188  data: 1.3260  max mem: 55573
Test: Total time: 0:00:07 (1.5443 s / it)
* Acc@1 67.310 Acc@5 88.576 loss 1.622
Accuracy of the model on the 50000 test images: 67.3%
Max accuracy: 67.31%
Test:  [0/5]  eta: 0:00:33  loss: 6.9115 (6.9115)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.6219  data: 6.3858  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9759 (6.9846)  acc1: 0.0000 (0.0960)  acc5: 0.1302 (0.5440)  time: 1.4697  data: 1.2773  max mem: 55573
Test: Total time: 0:00:07 (1.4874 s / it)
* Acc@1 0.100 Acc@5 0.510 loss 6.952
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [46]  [  0/156]  eta: 0:33:18  lr: 0.007831  min_lr: 0.007831  loss: 3.8008 (3.8008)  weight_decay: 0.0500 (0.0500)  time: 12.8108  data: 9.4432  max mem: 55573
Epoch: [46]  [ 10/156]  eta: 0:04:11  lr: 0.007830  min_lr: 0.007830  loss: 3.5282 (3.3861)  weight_decay: 0.0500 (0.0500)  time: 1.7250  data: 0.8590  max mem: 55573
Epoch: [46]  [ 20/156]  eta: 0:02:41  lr: 0.007829  min_lr: 0.007829  loss: 3.8772 (3.6498)  weight_decay: 0.0500 (0.0500)  time: 0.6026  data: 0.0006  max mem: 55573
Epoch: [46]  [ 30/156]  eta: 0:02:05  lr: 0.007829  min_lr: 0.007829  loss: 3.8772 (3.6380)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [46]  [ 40/156]  eta: 0:01:43  lr: 0.007828  min_lr: 0.007828  loss: 3.7414 (3.6873)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [46]  [ 50/156]  eta: 0:01:28  lr: 0.007827  min_lr: 0.007827  loss: 3.8049 (3.7028)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [46]  [ 60/156]  eta: 0:01:16  lr: 0.007826  min_lr: 0.007826  loss: 3.9236 (3.7272)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [46]  [ 70/156]  eta: 0:01:05  lr: 0.007825  min_lr: 0.007825  loss: 3.8498 (3.7362)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [46]  [ 80/156]  eta: 0:00:56  lr: 0.007824  min_lr: 0.007824  loss: 3.8740 (3.7547)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [46]  [ 90/156]  eta: 0:00:48  lr: 0.007823  min_lr: 0.007823  loss: 4.0488 (3.7945)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [46]  [100/156]  eta: 0:00:39  lr: 0.007823  min_lr: 0.007823  loss: 4.0488 (3.8006)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0004  max mem: 55573
Epoch: [46]  [110/156]  eta: 0:00:32  lr: 0.007822  min_lr: 0.007822  loss: 3.5977 (3.7634)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0004  max mem: 55573
Epoch: [46]  [120/156]  eta: 0:00:24  lr: 0.007821  min_lr: 0.007821  loss: 3.8292 (3.7690)  weight_decay: 0.0500 (0.0500)  time: 0.5885  data: 0.0004  max mem: 55573
Epoch: [46]  [130/156]  eta: 0:00:17  lr: 0.007820  min_lr: 0.007820  loss: 3.8490 (3.7756)  weight_decay: 0.0500 (0.0500)  time: 0.5884  data: 0.0010  max mem: 55573
Epoch: [46]  [140/156]  eta: 0:00:10  lr: 0.007819  min_lr: 0.007819  loss: 3.8978 (3.7867)  weight_decay: 0.0500 (0.0500)  time: 0.5853  data: 0.0008  max mem: 55573
Epoch: [46]  [150/156]  eta: 0:00:04  lr: 0.007818  min_lr: 0.007818  loss: 3.6464 (3.7801)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [46]  [155/156]  eta: 0:00:00  lr: 0.007818  min_lr: 0.007818  loss: 3.6114 (3.7781)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [46] Total time: 0:01:45 (0.6752 s / it)
Averaged stats: lr: 0.007818  min_lr: 0.007818  loss: 3.6114 (3.7811)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.4988 (1.4988)  acc1: 70.7031 (70.7031)  acc5: 92.3177 (92.3177)  time: 7.1342  data: 6.8967  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5418 (1.6788)  acc1: 69.1406 (65.1840)  acc5: 90.5660 (88.6400)  time: 1.5732  data: 1.3794  max mem: 55573
Test: Total time: 0:00:08 (1.6137 s / it)
* Acc@1 66.422 Acc@5 88.210 loss 1.644
Accuracy of the model on the 50000 test images: 66.4%
Max accuracy: 67.31%
Test:  [0/5]  eta: 0:00:34  loss: 6.9129 (6.9129)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9317  data: 6.6957  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9826 (6.9900)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.5316  data: 1.3392  max mem: 55573
Test: Total time: 0:00:07 (1.5556 s / it)
* Acc@1 0.100 Acc@5 0.514 loss 6.957
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [47]  [  0/156]  eta: 0:37:14  lr: 0.007818  min_lr: 0.007818  loss: 3.8641 (3.8641)  weight_decay: 0.0500 (0.0500)  time: 14.3239  data: 7.8135  max mem: 55573
Epoch: [47]  [ 10/156]  eta: 0:04:28  lr: 0.007817  min_lr: 0.007817  loss: 3.9462 (3.8788)  weight_decay: 0.0500 (0.0500)  time: 1.8390  data: 0.7107  max mem: 55573
Epoch: [47]  [ 20/156]  eta: 0:02:49  lr: 0.007816  min_lr: 0.007816  loss: 3.8319 (3.8624)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [47]  [ 30/156]  eta: 0:02:10  lr: 0.007815  min_lr: 0.007815  loss: 3.8741 (3.8476)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [47]  [ 40/156]  eta: 0:01:47  lr: 0.007814  min_lr: 0.007814  loss: 3.8054 (3.7832)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [47]  [ 50/156]  eta: 0:01:31  lr: 0.007814  min_lr: 0.007814  loss: 3.7891 (3.8088)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [47]  [ 60/156]  eta: 0:01:18  lr: 0.007813  min_lr: 0.007813  loss: 3.9941 (3.8268)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [47]  [ 70/156]  eta: 0:01:07  lr: 0.007812  min_lr: 0.007812  loss: 3.7265 (3.7547)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0004  max mem: 55573
Epoch: [47]  [ 80/156]  eta: 0:00:57  lr: 0.007811  min_lr: 0.007811  loss: 3.7265 (3.7706)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0005  max mem: 55573
Epoch: [47]  [ 90/156]  eta: 0:00:48  lr: 0.007810  min_lr: 0.007810  loss: 3.8841 (3.7701)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [47]  [100/156]  eta: 0:00:40  lr: 0.007809  min_lr: 0.007809  loss: 3.9108 (3.7765)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [47]  [110/156]  eta: 0:00:32  lr: 0.007808  min_lr: 0.007808  loss: 3.6263 (3.7491)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [47]  [120/156]  eta: 0:00:25  lr: 0.007807  min_lr: 0.007807  loss: 3.5549 (3.7386)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0004  max mem: 55573
Epoch: [47]  [130/156]  eta: 0:00:18  lr: 0.007807  min_lr: 0.007807  loss: 3.9562 (3.7520)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0011  max mem: 55573
Epoch: [47]  [140/156]  eta: 0:00:10  lr: 0.007806  min_lr: 0.007806  loss: 4.0011 (3.7645)  weight_decay: 0.0500 (0.0500)  time: 0.5839  data: 0.0009  max mem: 55573
Epoch: [47]  [150/156]  eta: 0:00:04  lr: 0.007805  min_lr: 0.007805  loss: 3.8608 (3.7725)  weight_decay: 0.0500 (0.0500)  time: 0.5810  data: 0.0001  max mem: 55573
Epoch: [47]  [155/156]  eta: 0:00:00  lr: 0.007804  min_lr: 0.007804  loss: 3.7052 (3.7593)  weight_decay: 0.0500 (0.0500)  time: 0.5810  data: 0.0001  max mem: 55573
Epoch: [47] Total time: 0:01:46 (0.6835 s / it)
Averaged stats: lr: 0.007804  min_lr: 0.007804  loss: 3.7052 (3.7670)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.4720 (1.4720)  acc1: 71.0938 (71.0938)  acc5: 92.0573 (92.0573)  time: 7.3096  data: 7.0722  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5652 (1.6448)  acc1: 68.4896 (65.0880)  acc5: 90.2344 (86.8800)  time: 1.6083  data: 1.4145  max mem: 55573
Test: Total time: 0:00:08 (1.6513 s / it)
* Acc@1 65.680 Acc@5 87.350 loss 1.645
Accuracy of the model on the 50000 test images: 65.7%
Max accuracy: 67.31%
Test:  [0/5]  eta: 0:00:33  loss: 6.9139 (6.9139)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.6981  data: 6.4622  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9910 (6.9948)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.5120)  time: 1.4849  data: 1.2926  max mem: 55573
Test: Total time: 0:00:07 (1.5021 s / it)
* Acc@1 0.100 Acc@5 0.510 loss 6.960
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [48]  [  0/156]  eta: 0:37:51  lr: 0.007804  min_lr: 0.007804  loss: 4.0850 (4.0850)  weight_decay: 0.0500 (0.0500)  time: 14.5591  data: 10.2769  max mem: 55573
Epoch: [48]  [ 10/156]  eta: 0:04:31  lr: 0.007803  min_lr: 0.007803  loss: 4.0375 (3.7508)  weight_decay: 0.0500 (0.0500)  time: 1.8593  data: 0.9346  max mem: 55573
Epoch: [48]  [ 20/156]  eta: 0:02:50  lr: 0.007802  min_lr: 0.007802  loss: 3.7175 (3.6487)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [48]  [ 30/156]  eta: 0:02:11  lr: 0.007802  min_lr: 0.007802  loss: 3.6595 (3.6747)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [48]  [ 40/156]  eta: 0:01:47  lr: 0.007801  min_lr: 0.007801  loss: 3.8061 (3.6805)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [48]  [ 50/156]  eta: 0:01:31  lr: 0.007800  min_lr: 0.007800  loss: 3.6735 (3.6651)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [48]  [ 60/156]  eta: 0:01:18  lr: 0.007799  min_lr: 0.007799  loss: 3.8399 (3.7137)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [48]  [ 70/156]  eta: 0:01:07  lr: 0.007798  min_lr: 0.007798  loss: 3.8399 (3.6751)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [48]  [ 80/156]  eta: 0:00:57  lr: 0.007797  min_lr: 0.007797  loss: 3.5175 (3.6447)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [48]  [ 90/156]  eta: 0:00:49  lr: 0.007796  min_lr: 0.007796  loss: 3.6889 (3.6637)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [48]  [100/156]  eta: 0:00:40  lr: 0.007795  min_lr: 0.007795  loss: 3.9456 (3.6833)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [48]  [110/156]  eta: 0:00:32  lr: 0.007794  min_lr: 0.007794  loss: 3.8398 (3.6797)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [48]  [120/156]  eta: 0:00:25  lr: 0.007793  min_lr: 0.007793  loss: 3.8298 (3.6938)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [48]  [130/156]  eta: 0:00:18  lr: 0.007793  min_lr: 0.007793  loss: 4.0864 (3.6933)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0012  max mem: 55573
Epoch: [48]  [140/156]  eta: 0:00:11  lr: 0.007792  min_lr: 0.007792  loss: 4.0130 (3.7031)  weight_decay: 0.0500 (0.0500)  time: 0.5873  data: 0.0010  max mem: 55573
Epoch: [48]  [150/156]  eta: 0:00:04  lr: 0.007791  min_lr: 0.007791  loss: 4.0055 (3.7038)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [48]  [155/156]  eta: 0:00:00  lr: 0.007790  min_lr: 0.007790  loss: 3.8419 (3.7031)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [48] Total time: 0:01:46 (0.6853 s / it)
Averaged stats: lr: 0.007790  min_lr: 0.007790  loss: 3.8419 (3.7460)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.2155 (1.2155)  acc1: 73.0469 (73.0469)  acc5: 93.8802 (93.8802)  time: 7.0381  data: 6.8009  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3920 (1.4498)  acc1: 68.0990 (66.3680)  acc5: 89.7135 (88.0320)  time: 1.5540  data: 1.3603  max mem: 55573
Test: Total time: 0:00:07 (1.5941 s / it)
* Acc@1 67.460 Acc@5 88.456 loss 1.427
Accuracy of the model on the 50000 test images: 67.5%
Max accuracy: 67.46%
Test:  [0/5]  eta: 0:00:34  loss: 6.9149 (6.9149)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9216  data: 6.6856  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9999 (7.0006)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.5120)  time: 1.5296  data: 1.3372  max mem: 55573
Test: Total time: 0:00:07 (1.5460 s / it)
* Acc@1 0.100 Acc@5 0.524 loss 6.965
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [49]  [  0/156]  eta: 0:34:54  lr: 0.007790  min_lr: 0.007790  loss: 3.5800 (3.5800)  weight_decay: 0.0500 (0.0500)  time: 13.4266  data: 10.6837  max mem: 55573
Epoch: [49]  [ 10/156]  eta: 0:04:18  lr: 0.007789  min_lr: 0.007789  loss: 3.8326 (3.7204)  weight_decay: 0.0500 (0.0500)  time: 1.7710  data: 0.9716  max mem: 55573
Epoch: [49]  [ 20/156]  eta: 0:02:44  lr: 0.007788  min_lr: 0.007788  loss: 3.9382 (3.8628)  weight_decay: 0.0500 (0.0500)  time: 0.5976  data: 0.0004  max mem: 55573
Epoch: [49]  [ 30/156]  eta: 0:02:07  lr: 0.007787  min_lr: 0.007787  loss: 4.0245 (3.8156)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [49]  [ 40/156]  eta: 0:01:45  lr: 0.007786  min_lr: 0.007786  loss: 3.6625 (3.7688)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [49]  [ 50/156]  eta: 0:01:29  lr: 0.007786  min_lr: 0.007786  loss: 3.7027 (3.7883)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [49]  [ 60/156]  eta: 0:01:17  lr: 0.007785  min_lr: 0.007785  loss: 4.0535 (3.8229)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [49]  [ 70/156]  eta: 0:01:06  lr: 0.007784  min_lr: 0.007784  loss: 3.8744 (3.7919)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [49]  [ 80/156]  eta: 0:00:57  lr: 0.007783  min_lr: 0.007783  loss: 3.3248 (3.7187)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0005  max mem: 55573
Epoch: [49]  [ 90/156]  eta: 0:00:48  lr: 0.007782  min_lr: 0.007782  loss: 3.5709 (3.7284)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [49]  [100/156]  eta: 0:00:40  lr: 0.007781  min_lr: 0.007781  loss: 3.6890 (3.7161)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [49]  [110/156]  eta: 0:00:32  lr: 0.007780  min_lr: 0.007780  loss: 3.6257 (3.7279)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [49]  [120/156]  eta: 0:00:25  lr: 0.007779  min_lr: 0.007779  loss: 3.8114 (3.7207)  weight_decay: 0.0500 (0.0500)  time: 0.5882  data: 0.0004  max mem: 55573
Epoch: [49]  [130/156]  eta: 0:00:17  lr: 0.007778  min_lr: 0.007778  loss: 3.8041 (3.7178)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0010  max mem: 55573
Epoch: [49]  [140/156]  eta: 0:00:10  lr: 0.007777  min_lr: 0.007777  loss: 3.6001 (3.7242)  weight_decay: 0.0500 (0.0500)  time: 0.5863  data: 0.0008  max mem: 55573
Epoch: [49]  [150/156]  eta: 0:00:04  lr: 0.007776  min_lr: 0.007776  loss: 3.8585 (3.7329)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [49]  [155/156]  eta: 0:00:00  lr: 0.007776  min_lr: 0.007776  loss: 3.6475 (3.7203)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [49] Total time: 0:01:45 (0.6785 s / it)
Averaged stats: lr: 0.007776  min_lr: 0.007776  loss: 3.6475 (3.7720)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.3519 (1.3519)  acc1: 73.9583 (73.9583)  acc5: 92.4479 (92.4479)  time: 6.9075  data: 6.6701  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5439 (1.6046)  acc1: 70.3125 (67.2000)  acc5: 90.8854 (88.2880)  time: 1.5280  data: 1.3341  max mem: 55573
Test: Total time: 0:00:07 (1.5750 s / it)
* Acc@1 67.566 Acc@5 88.680 loss 1.598
Accuracy of the model on the 50000 test images: 67.6%
Max accuracy: 67.57%
Test:  [0/5]  eta: 0:00:34  loss: 6.9164 (6.9164)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9047  data: 6.6687  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.0096 (7.0062)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.5120)  time: 1.5263  data: 1.3339  max mem: 55573
Test: Total time: 0:00:07 (1.5424 s / it)
* Acc@1 0.100 Acc@5 0.508 loss 6.969
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [50]  [  0/156]  eta: 0:37:30  lr: 0.007776  min_lr: 0.007776  loss: 4.6127 (4.6127)  weight_decay: 0.0500 (0.0500)  time: 14.4256  data: 9.2316  max mem: 55573
Epoch: [50]  [ 10/156]  eta: 0:04:30  lr: 0.007775  min_lr: 0.007775  loss: 3.8223 (3.8859)  weight_decay: 0.0500 (0.0500)  time: 1.8512  data: 0.8397  max mem: 55573
Epoch: [50]  [ 20/156]  eta: 0:02:50  lr: 0.007774  min_lr: 0.007774  loss: 3.8213 (3.8609)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [50]  [ 30/156]  eta: 0:02:10  lr: 0.007773  min_lr: 0.007773  loss: 3.7164 (3.7635)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [50]  [ 40/156]  eta: 0:01:47  lr: 0.007772  min_lr: 0.007772  loss: 3.7053 (3.7645)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [50]  [ 50/156]  eta: 0:01:31  lr: 0.007771  min_lr: 0.007771  loss: 3.9955 (3.7873)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [50]  [ 60/156]  eta: 0:01:18  lr: 0.007770  min_lr: 0.007770  loss: 3.8697 (3.7503)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [50]  [ 70/156]  eta: 0:01:07  lr: 0.007769  min_lr: 0.007769  loss: 3.8661 (3.7884)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [50]  [ 80/156]  eta: 0:00:57  lr: 0.007768  min_lr: 0.007768  loss: 3.8500 (3.7805)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [50]  [ 90/156]  eta: 0:00:49  lr: 0.007767  min_lr: 0.007767  loss: 3.7884 (3.7755)  weight_decay: 0.0500 (0.0500)  time: 0.5950  data: 0.0004  max mem: 55573
Epoch: [50]  [100/156]  eta: 0:00:40  lr: 0.007766  min_lr: 0.007766  loss: 3.9329 (3.7853)  weight_decay: 0.0500 (0.0500)  time: 0.5946  data: 0.0004  max mem: 55573
Epoch: [50]  [110/156]  eta: 0:00:32  lr: 0.007765  min_lr: 0.007765  loss: 3.9307 (3.8009)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [50]  [120/156]  eta: 0:00:25  lr: 0.007764  min_lr: 0.007764  loss: 3.6411 (3.7786)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [50]  [130/156]  eta: 0:00:18  lr: 0.007763  min_lr: 0.007763  loss: 3.6210 (3.7751)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0027  max mem: 55573
Epoch: [50]  [140/156]  eta: 0:00:11  lr: 0.007762  min_lr: 0.007762  loss: 3.9260 (3.7854)  weight_decay: 0.0500 (0.0500)  time: 0.5882  data: 0.0026  max mem: 55573
Epoch: [50]  [150/156]  eta: 0:00:04  lr: 0.007761  min_lr: 0.007761  loss: 3.9552 (3.7896)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [50]  [155/156]  eta: 0:00:00  lr: 0.007761  min_lr: 0.007761  loss: 3.8189 (3.7708)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [50] Total time: 0:01:46 (0.6853 s / it)
Averaged stats: lr: 0.007761  min_lr: 0.007761  loss: 3.8189 (3.7638)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.3404 (1.3404)  acc1: 73.4375 (73.4375)  acc5: 92.0573 (92.0573)  time: 6.8163  data: 6.5776  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4329 (1.4717)  acc1: 69.4010 (67.1040)  acc5: 90.8854 (88.2560)  time: 1.5100  data: 1.3156  max mem: 55573
Test: Total time: 0:00:07 (1.5539 s / it)
* Acc@1 67.128 Acc@5 88.424 loss 1.461
Accuracy of the model on the 50000 test images: 67.1%
Max accuracy: 67.57%
Test:  [0/5]  eta: 0:00:36  loss: 6.9191 (6.9191)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.2899  data: 7.0534  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.0187 (7.0134)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.5120)  time: 1.6036  data: 1.4108  max mem: 55573
Test: Total time: 0:00:08 (1.6198 s / it)
* Acc@1 0.100 Acc@5 0.508 loss 6.975
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [51]  [  0/156]  eta: 0:37:49  lr: 0.007761  min_lr: 0.007761  loss: 4.0468 (4.0468)  weight_decay: 0.0500 (0.0500)  time: 14.5503  data: 8.0450  max mem: 55573
Epoch: [51]  [ 10/156]  eta: 0:04:46  lr: 0.007760  min_lr: 0.007760  loss: 3.8744 (3.5760)  weight_decay: 0.0500 (0.0500)  time: 1.9655  data: 0.7318  max mem: 55573
Epoch: [51]  [ 20/156]  eta: 0:02:58  lr: 0.007759  min_lr: 0.007759  loss: 3.7807 (3.6358)  weight_decay: 0.0500 (0.0500)  time: 0.6487  data: 0.0005  max mem: 55573
Epoch: [51]  [ 30/156]  eta: 0:02:15  lr: 0.007758  min_lr: 0.007758  loss: 3.6712 (3.6425)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [51]  [ 40/156]  eta: 0:01:51  lr: 0.007757  min_lr: 0.007757  loss: 3.6712 (3.6403)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [51]  [ 50/156]  eta: 0:01:34  lr: 0.007756  min_lr: 0.007756  loss: 3.8292 (3.6697)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0004  max mem: 55573
Epoch: [51]  [ 60/156]  eta: 0:01:20  lr: 0.007755  min_lr: 0.007755  loss: 3.7276 (3.6595)  weight_decay: 0.0500 (0.0500)  time: 0.5938  data: 0.0004  max mem: 55573
Epoch: [51]  [ 70/156]  eta: 0:01:09  lr: 0.007754  min_lr: 0.007754  loss: 3.4917 (3.6383)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [51]  [ 80/156]  eta: 0:00:59  lr: 0.007753  min_lr: 0.007753  loss: 3.7645 (3.6627)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [51]  [ 90/156]  eta: 0:00:49  lr: 0.007752  min_lr: 0.007752  loss: 3.8903 (3.6579)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [51]  [100/156]  eta: 0:00:41  lr: 0.007751  min_lr: 0.007751  loss: 3.4514 (3.6232)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [51]  [110/156]  eta: 0:00:33  lr: 0.007750  min_lr: 0.007750  loss: 3.7006 (3.6373)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0004  max mem: 55573
Epoch: [51]  [120/156]  eta: 0:00:25  lr: 0.007749  min_lr: 0.007749  loss: 4.0665 (3.6670)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0004  max mem: 55573
Epoch: [51]  [130/156]  eta: 0:00:18  lr: 0.007748  min_lr: 0.007748  loss: 4.0310 (3.6687)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0010  max mem: 55573
Epoch: [51]  [140/156]  eta: 0:00:11  lr: 0.007747  min_lr: 0.007747  loss: 3.9264 (3.6974)  weight_decay: 0.0500 (0.0500)  time: 0.5862  data: 0.0008  max mem: 55573
Epoch: [51]  [150/156]  eta: 0:00:04  lr: 0.007746  min_lr: 0.007746  loss: 3.8878 (3.6822)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [51]  [155/156]  eta: 0:00:00  lr: 0.007745  min_lr: 0.007745  loss: 3.6592 (3.6831)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [51] Total time: 0:01:48 (0.6930 s / it)
Averaged stats: lr: 0.007745  min_lr: 0.007745  loss: 3.6592 (3.7518)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.2718 (1.2718)  acc1: 74.0885 (74.0885)  acc5: 93.2292 (93.2292)  time: 7.1008  data: 6.8634  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3920 (1.4053)  acc1: 70.5729 (67.8080)  acc5: 90.8854 (88.9600)  time: 1.5666  data: 1.3728  max mem: 55573
Test: Total time: 0:00:08 (1.6055 s / it)
* Acc@1 68.890 Acc@5 89.144 loss 1.375
Accuracy of the model on the 50000 test images: 68.9%
Max accuracy: 68.89%
Test:  [0/5]  eta: 0:00:34  loss: 6.9215 (6.9215)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9034  data: 6.6671  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.0290 (7.0218)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.5120)  time: 1.5260  data: 1.3335  max mem: 55573
Test: Total time: 0:00:07 (1.5435 s / it)
* Acc@1 0.100 Acc@5 0.502 loss 6.981
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [52]  [  0/156]  eta: 0:37:14  lr: 0.007745  min_lr: 0.007745  loss: 4.0576 (4.0576)  weight_decay: 0.0500 (0.0500)  time: 14.3255  data: 9.7790  max mem: 55573
Epoch: [52]  [ 10/156]  eta: 0:04:28  lr: 0.007744  min_lr: 0.007744  loss: 3.8367 (3.7163)  weight_decay: 0.0500 (0.0500)  time: 1.8393  data: 0.8894  max mem: 55573
Epoch: [52]  [ 20/156]  eta: 0:02:49  lr: 0.007743  min_lr: 0.007743  loss: 3.6809 (3.7139)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [52]  [ 30/156]  eta: 0:02:10  lr: 0.007742  min_lr: 0.007742  loss: 3.6742 (3.7192)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [52]  [ 40/156]  eta: 0:01:47  lr: 0.007741  min_lr: 0.007741  loss: 3.6742 (3.7018)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [52]  [ 50/156]  eta: 0:01:31  lr: 0.007740  min_lr: 0.007740  loss: 3.6067 (3.6874)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [52]  [ 60/156]  eta: 0:01:18  lr: 0.007739  min_lr: 0.007739  loss: 3.8046 (3.7105)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [52]  [ 70/156]  eta: 0:01:07  lr: 0.007738  min_lr: 0.007738  loss: 3.8683 (3.7180)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [52]  [ 80/156]  eta: 0:00:57  lr: 0.007737  min_lr: 0.007737  loss: 3.8963 (3.7201)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [52]  [ 90/156]  eta: 0:00:48  lr: 0.007736  min_lr: 0.007736  loss: 3.7743 (3.7195)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [52]  [100/156]  eta: 0:00:40  lr: 0.007735  min_lr: 0.007735  loss: 3.7703 (3.7171)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [52]  [110/156]  eta: 0:00:32  lr: 0.007734  min_lr: 0.007734  loss: 3.8111 (3.7145)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [52]  [120/156]  eta: 0:00:25  lr: 0.007733  min_lr: 0.007733  loss: 3.3198 (3.6909)  weight_decay: 0.0500 (0.0500)  time: 0.5970  data: 0.0004  max mem: 55573
Epoch: [52]  [130/156]  eta: 0:00:18  lr: 0.007732  min_lr: 0.007732  loss: 3.4583 (3.6901)  weight_decay: 0.0500 (0.0500)  time: 0.5975  data: 0.0010  max mem: 55573
Epoch: [52]  [140/156]  eta: 0:00:11  lr: 0.007731  min_lr: 0.007731  loss: 3.8577 (3.6984)  weight_decay: 0.0500 (0.0500)  time: 0.5869  data: 0.0008  max mem: 55573
Epoch: [52]  [150/156]  eta: 0:00:04  lr: 0.007730  min_lr: 0.007730  loss: 3.9935 (3.7149)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [52]  [155/156]  eta: 0:00:00  lr: 0.007729  min_lr: 0.007729  loss: 4.0777 (3.7263)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [52] Total time: 0:01:46 (0.6850 s / it)
Averaged stats: lr: 0.007729  min_lr: 0.007729  loss: 4.0777 (3.7284)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.3455 (1.3455)  acc1: 75.3906 (75.3906)  acc5: 92.9688 (92.9688)  time: 6.8814  data: 6.6441  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4775 (1.5077)  acc1: 69.9219 (68.8320)  acc5: 92.0573 (89.4720)  time: 1.5228  data: 1.3289  max mem: 55573
Test: Total time: 0:00:07 (1.5665 s / it)
* Acc@1 68.324 Acc@5 89.152 loss 1.519
Accuracy of the model on the 50000 test images: 68.3%
Max accuracy: 68.89%
Test:  [0/5]  eta: 0:00:34  loss: 6.9246 (6.9246)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9216  data: 6.6856  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.0399 (7.0301)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5296  data: 1.3372  max mem: 55573
Test: Total time: 0:00:07 (1.5495 s / it)
* Acc@1 0.100 Acc@5 0.516 loss 6.988
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [53]  [  0/156]  eta: 0:34:11  lr: 0.007729  min_lr: 0.007729  loss: 3.9868 (3.9868)  weight_decay: 0.0500 (0.0500)  time: 13.1500  data: 10.7853  max mem: 55573
Epoch: [53]  [ 10/156]  eta: 0:04:26  lr: 0.007728  min_lr: 0.007728  loss: 3.8338 (3.7809)  weight_decay: 0.0500 (0.0500)  time: 1.8220  data: 1.0617  max mem: 55573
Epoch: [53]  [ 20/156]  eta: 0:02:47  lr: 0.007727  min_lr: 0.007727  loss: 3.8257 (3.7988)  weight_decay: 0.0500 (0.0500)  time: 0.6393  data: 0.0449  max mem: 55573
Epoch: [53]  [ 30/156]  eta: 0:02:09  lr: 0.007726  min_lr: 0.007726  loss: 3.7013 (3.7460)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0006  max mem: 55573
Epoch: [53]  [ 40/156]  eta: 0:01:46  lr: 0.007725  min_lr: 0.007725  loss: 3.6286 (3.7341)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0007  max mem: 55573
Epoch: [53]  [ 50/156]  eta: 0:01:30  lr: 0.007724  min_lr: 0.007724  loss: 3.6286 (3.7113)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0007  max mem: 55573
Epoch: [53]  [ 60/156]  eta: 0:01:17  lr: 0.007723  min_lr: 0.007723  loss: 3.8822 (3.7320)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0006  max mem: 55573
Epoch: [53]  [ 70/156]  eta: 0:01:07  lr: 0.007722  min_lr: 0.007722  loss: 3.8778 (3.7378)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0006  max mem: 55573
Epoch: [53]  [ 80/156]  eta: 0:00:57  lr: 0.007721  min_lr: 0.007721  loss: 3.7225 (3.7009)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0006  max mem: 55573
Epoch: [53]  [ 90/156]  eta: 0:00:48  lr: 0.007720  min_lr: 0.007720  loss: 3.7271 (3.7167)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0006  max mem: 55573
Epoch: [53]  [100/156]  eta: 0:00:40  lr: 0.007718  min_lr: 0.007718  loss: 3.9819 (3.7137)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0006  max mem: 55573
Epoch: [53]  [110/156]  eta: 0:00:32  lr: 0.007717  min_lr: 0.007717  loss: 4.0150 (3.7328)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0006  max mem: 55573
Epoch: [53]  [120/156]  eta: 0:00:25  lr: 0.007716  min_lr: 0.007716  loss: 3.9610 (3.7470)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0006  max mem: 55573
Epoch: [53]  [130/156]  eta: 0:00:18  lr: 0.007715  min_lr: 0.007715  loss: 3.8053 (3.7474)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0024  max mem: 55573
Epoch: [53]  [140/156]  eta: 0:00:10  lr: 0.007714  min_lr: 0.007714  loss: 3.7151 (3.7559)  weight_decay: 0.0500 (0.0500)  time: 0.5873  data: 0.0022  max mem: 55573
Epoch: [53]  [150/156]  eta: 0:00:04  lr: 0.007713  min_lr: 0.007713  loss: 3.7151 (3.7355)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [53]  [155/156]  eta: 0:00:00  lr: 0.007713  min_lr: 0.007713  loss: 3.7265 (3.7375)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [53] Total time: 0:01:46 (0.6831 s / it)
Averaged stats: lr: 0.007713  min_lr: 0.007713  loss: 3.7265 (3.7332)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.3811 (1.3811)  acc1: 73.9583 (73.9583)  acc5: 92.7083 (92.7083)  time: 7.0324  data: 6.7951  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5518 (1.6278)  acc1: 69.9219 (66.6240)  acc5: 90.7552 (87.7440)  time: 1.5529  data: 1.3591  max mem: 55573
Test: Total time: 0:00:07 (1.5969 s / it)
* Acc@1 67.006 Acc@5 87.944 loss 1.625
Accuracy of the model on the 50000 test images: 67.0%
Max accuracy: 68.89%
Test:  [0/5]  eta: 0:00:36  loss: 6.9278 (6.9278)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.2127  data: 6.9767  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.0514 (7.0388)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.5878  data: 1.3954  max mem: 55573
Test: Total time: 0:00:08 (1.6097 s / it)
* Acc@1 0.100 Acc@5 0.520 loss 6.995
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [54]  [  0/156]  eta: 0:33:45  lr: 0.007712  min_lr: 0.007712  loss: 3.6926 (3.6926)  weight_decay: 0.0500 (0.0500)  time: 12.9869  data: 12.3963  max mem: 55573
Epoch: [54]  [ 10/156]  eta: 0:04:21  lr: 0.007711  min_lr: 0.007711  loss: 3.6926 (3.5525)  weight_decay: 0.0500 (0.0500)  time: 1.7919  data: 1.1273  max mem: 55573
Epoch: [54]  [ 20/156]  eta: 0:02:45  lr: 0.007710  min_lr: 0.007710  loss: 3.9273 (3.7717)  weight_decay: 0.0500 (0.0500)  time: 0.6314  data: 0.0004  max mem: 55573
Epoch: [54]  [ 30/156]  eta: 0:02:08  lr: 0.007709  min_lr: 0.007709  loss: 3.9388 (3.8103)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [54]  [ 40/156]  eta: 0:01:45  lr: 0.007708  min_lr: 0.007708  loss: 3.9225 (3.8083)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [54]  [ 50/156]  eta: 0:01:30  lr: 0.007707  min_lr: 0.007707  loss: 3.6822 (3.7456)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [54]  [ 60/156]  eta: 0:01:17  lr: 0.007706  min_lr: 0.007706  loss: 3.4983 (3.7105)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [54]  [ 70/156]  eta: 0:01:06  lr: 0.007705  min_lr: 0.007705  loss: 3.7253 (3.7469)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [54]  [ 80/156]  eta: 0:00:57  lr: 0.007704  min_lr: 0.007704  loss: 3.9519 (3.7616)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [54]  [ 90/156]  eta: 0:00:48  lr: 0.007703  min_lr: 0.007703  loss: 3.8644 (3.7654)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [54]  [100/156]  eta: 0:00:40  lr: 0.007702  min_lr: 0.007702  loss: 3.8384 (3.7621)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0004  max mem: 55573
Epoch: [54]  [110/156]  eta: 0:00:32  lr: 0.007701  min_lr: 0.007701  loss: 3.8299 (3.7613)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [54]  [120/156]  eta: 0:00:25  lr: 0.007700  min_lr: 0.007700  loss: 3.8878 (3.7743)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0004  max mem: 55573
Epoch: [54]  [130/156]  eta: 0:00:17  lr: 0.007698  min_lr: 0.007698  loss: 4.0217 (3.7870)  weight_decay: 0.0500 (0.0500)  time: 0.5932  data: 0.0010  max mem: 55573
Epoch: [54]  [140/156]  eta: 0:00:10  lr: 0.007697  min_lr: 0.007697  loss: 4.0314 (3.7874)  weight_decay: 0.0500 (0.0500)  time: 0.5865  data: 0.0009  max mem: 55573
Epoch: [54]  [150/156]  eta: 0:00:04  lr: 0.007696  min_lr: 0.007696  loss: 3.8666 (3.7846)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [54]  [155/156]  eta: 0:00:00  lr: 0.007696  min_lr: 0.007696  loss: 3.8498 (3.7736)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [54] Total time: 0:01:46 (0.6814 s / it)
Averaged stats: lr: 0.007696  min_lr: 0.007696  loss: 3.8498 (3.7331)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.5150 (1.5150)  acc1: 72.6562 (72.6562)  acc5: 91.7969 (91.7969)  time: 7.1821  data: 6.9446  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5150 (1.6194)  acc1: 70.7031 (67.2960)  acc5: 91.7969 (88.6080)  time: 1.5830  data: 1.3890  max mem: 55573
Test: Total time: 0:00:08 (1.6242 s / it)
* Acc@1 67.616 Acc@5 88.606 loss 1.596
Accuracy of the model on the 50000 test images: 67.6%
Max accuracy: 68.89%
Test:  [0/5]  eta: 0:00:36  loss: 6.9322 (6.9322)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.3162  data: 7.0802  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.0640 (7.0486)  acc1: 0.0000 (0.0960)  acc5: 0.3906 (0.4800)  time: 1.6086  data: 1.4161  max mem: 55573
Test: Total time: 0:00:08 (1.6320 s / it)
* Acc@1 0.100 Acc@5 0.510 loss 7.003
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [55]  [  0/156]  eta: 0:36:26  lr: 0.007696  min_lr: 0.007696  loss: 4.0795 (4.0795)  weight_decay: 0.0500 (0.0500)  time: 14.0135  data: 9.5808  max mem: 55573
Epoch: [55]  [ 10/156]  eta: 0:04:25  lr: 0.007694  min_lr: 0.007694  loss: 4.0007 (3.7109)  weight_decay: 0.0500 (0.0500)  time: 1.8218  data: 0.8714  max mem: 55573
Epoch: [55]  [ 20/156]  eta: 0:02:48  lr: 0.007693  min_lr: 0.007693  loss: 3.9110 (3.8183)  weight_decay: 0.0500 (0.0500)  time: 0.5978  data: 0.0005  max mem: 55573
Epoch: [55]  [ 30/156]  eta: 0:02:09  lr: 0.007692  min_lr: 0.007692  loss: 3.9110 (3.8331)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [55]  [ 40/156]  eta: 0:01:46  lr: 0.007691  min_lr: 0.007691  loss: 3.8815 (3.8385)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [55]  [ 50/156]  eta: 0:01:30  lr: 0.007690  min_lr: 0.007690  loss: 3.7799 (3.8273)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [55]  [ 60/156]  eta: 0:01:18  lr: 0.007689  min_lr: 0.007689  loss: 3.7263 (3.7930)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [55]  [ 70/156]  eta: 0:01:07  lr: 0.007688  min_lr: 0.007688  loss: 3.7263 (3.7801)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [55]  [ 80/156]  eta: 0:00:57  lr: 0.007687  min_lr: 0.007687  loss: 3.7158 (3.7445)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [55]  [ 90/156]  eta: 0:00:48  lr: 0.007686  min_lr: 0.007686  loss: 3.7158 (3.7407)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [55]  [100/156]  eta: 0:00:40  lr: 0.007684  min_lr: 0.007684  loss: 3.8344 (3.7300)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [55]  [110/156]  eta: 0:00:32  lr: 0.007683  min_lr: 0.007683  loss: 3.5782 (3.7128)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [55]  [120/156]  eta: 0:00:25  lr: 0.007682  min_lr: 0.007682  loss: 3.5782 (3.7103)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [55]  [130/156]  eta: 0:00:18  lr: 0.007681  min_lr: 0.007681  loss: 3.8252 (3.7188)  weight_decay: 0.0500 (0.0500)  time: 0.5934  data: 0.0011  max mem: 55573
Epoch: [55]  [140/156]  eta: 0:00:10  lr: 0.007680  min_lr: 0.007680  loss: 3.7439 (3.7176)  weight_decay: 0.0500 (0.0500)  time: 0.5879  data: 0.0009  max mem: 55573
Epoch: [55]  [150/156]  eta: 0:00:04  lr: 0.007679  min_lr: 0.007679  loss: 3.6881 (3.7149)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [55]  [155/156]  eta: 0:00:00  lr: 0.007678  min_lr: 0.007678  loss: 3.8751 (3.7200)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [55] Total time: 0:01:46 (0.6831 s / it)
Averaged stats: lr: 0.007678  min_lr: 0.007678  loss: 3.8751 (3.7153)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.3520 (1.3520)  acc1: 73.0469 (73.0469)  acc5: 92.9688 (92.9688)  time: 6.9387  data: 6.7013  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5345 (1.4971)  acc1: 68.3594 (67.0720)  acc5: 90.5660 (88.3520)  time: 1.5340  data: 1.3403  max mem: 55573
Test: Total time: 0:00:07 (1.5812 s / it)
* Acc@1 67.012 Acc@5 87.896 loss 1.509
Accuracy of the model on the 50000 test images: 67.0%
Max accuracy: 68.89%
Test:  [0/5]  eta: 0:00:34  loss: 6.9363 (6.9363)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.8797  data: 6.6435  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.0784 (7.0601)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.5213  data: 1.3289  max mem: 55573
Test: Total time: 0:00:07 (1.5417 s / it)
* Acc@1 0.100 Acc@5 0.506 loss 7.012
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [56]  [  0/156]  eta: 0:37:17  lr: 0.007678  min_lr: 0.007678  loss: 4.4975 (4.4975)  weight_decay: 0.0500 (0.0500)  time: 14.3407  data: 9.5337  max mem: 55573
Epoch: [56]  [ 10/156]  eta: 0:04:28  lr: 0.007677  min_lr: 0.007677  loss: 3.9072 (3.9071)  weight_decay: 0.0500 (0.0500)  time: 1.8404  data: 0.8671  max mem: 55573
Epoch: [56]  [ 20/156]  eta: 0:02:49  lr: 0.007676  min_lr: 0.007676  loss: 3.9072 (3.9027)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [56]  [ 30/156]  eta: 0:02:10  lr: 0.007675  min_lr: 0.007675  loss: 3.9330 (3.8739)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [56]  [ 40/156]  eta: 0:01:47  lr: 0.007674  min_lr: 0.007674  loss: 3.8306 (3.8539)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [56]  [ 50/156]  eta: 0:01:31  lr: 0.007672  min_lr: 0.007672  loss: 3.8584 (3.8388)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [56]  [ 60/156]  eta: 0:01:18  lr: 0.007671  min_lr: 0.007671  loss: 3.9664 (3.8353)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [56]  [ 70/156]  eta: 0:01:07  lr: 0.007670  min_lr: 0.007670  loss: 3.7292 (3.7910)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [56]  [ 80/156]  eta: 0:00:57  lr: 0.007669  min_lr: 0.007669  loss: 3.5945 (3.7775)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [56]  [ 90/156]  eta: 0:00:48  lr: 0.007668  min_lr: 0.007668  loss: 3.7132 (3.7760)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [56]  [100/156]  eta: 0:00:40  lr: 0.007667  min_lr: 0.007667  loss: 3.8055 (3.7761)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [56]  [110/156]  eta: 0:00:32  lr: 0.007666  min_lr: 0.007666  loss: 3.8833 (3.7875)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [56]  [120/156]  eta: 0:00:25  lr: 0.007664  min_lr: 0.007664  loss: 3.8639 (3.7742)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [56]  [130/156]  eta: 0:00:18  lr: 0.007663  min_lr: 0.007663  loss: 3.6378 (3.7646)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0010  max mem: 55573
Epoch: [56]  [140/156]  eta: 0:00:10  lr: 0.007662  min_lr: 0.007662  loss: 3.8659 (3.7622)  weight_decay: 0.0500 (0.0500)  time: 0.5869  data: 0.0009  max mem: 55573
Epoch: [56]  [150/156]  eta: 0:00:04  lr: 0.007661  min_lr: 0.007661  loss: 3.8659 (3.7583)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [56]  [155/156]  eta: 0:00:00  lr: 0.007660  min_lr: 0.007660  loss: 3.8649 (3.7504)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [56] Total time: 0:01:46 (0.6839 s / it)
Averaged stats: lr: 0.007660  min_lr: 0.007660  loss: 3.8649 (3.7154)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:32  loss: 1.2266 (1.2266)  acc1: 74.8698 (74.8698)  acc5: 93.7500 (93.7500)  time: 6.5501  data: 6.3128  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3855 (1.4537)  acc1: 71.0938 (67.8400)  acc5: 90.4948 (89.1840)  time: 1.4567  data: 1.2627  max mem: 55573
Test: Total time: 0:00:07 (1.5031 s / it)
* Acc@1 68.682 Acc@5 89.216 loss 1.458
Accuracy of the model on the 50000 test images: 68.7%
Max accuracy: 68.89%
Test:  [0/5]  eta: 0:00:36  loss: 6.9410 (6.9410)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.2304  data: 6.9943  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.0957 (7.0730)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.5914  data: 1.3989  max mem: 55573
Test: Total time: 0:00:08 (1.6102 s / it)
* Acc@1 0.100 Acc@5 0.514 loss 7.023
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [57]  [  0/156]  eta: 0:35:53  lr: 0.007660  min_lr: 0.007660  loss: 3.0118 (3.0118)  weight_decay: 0.0500 (0.0500)  time: 13.8035  data: 9.7881  max mem: 55573
Epoch: [57]  [ 10/156]  eta: 0:04:21  lr: 0.007659  min_lr: 0.007659  loss: 3.7405 (3.5113)  weight_decay: 0.0500 (0.0500)  time: 1.7923  data: 0.8902  max mem: 55573
Epoch: [57]  [ 20/156]  eta: 0:02:45  lr: 0.007658  min_lr: 0.007658  loss: 3.7873 (3.6860)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [57]  [ 30/156]  eta: 0:02:08  lr: 0.007657  min_lr: 0.007657  loss: 3.7823 (3.6923)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [57]  [ 40/156]  eta: 0:01:45  lr: 0.007656  min_lr: 0.007656  loss: 3.7575 (3.7388)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [57]  [ 50/156]  eta: 0:01:30  lr: 0.007654  min_lr: 0.007654  loss: 3.8775 (3.7320)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [57]  [ 60/156]  eta: 0:01:17  lr: 0.007653  min_lr: 0.007653  loss: 3.5811 (3.7006)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [57]  [ 70/156]  eta: 0:01:06  lr: 0.007652  min_lr: 0.007652  loss: 3.9129 (3.7410)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [57]  [ 80/156]  eta: 0:00:57  lr: 0.007651  min_lr: 0.007651  loss: 4.0346 (3.7744)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [57]  [ 90/156]  eta: 0:00:48  lr: 0.007650  min_lr: 0.007650  loss: 3.9268 (3.7374)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [57]  [100/156]  eta: 0:00:40  lr: 0.007649  min_lr: 0.007649  loss: 3.6933 (3.7469)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [57]  [110/156]  eta: 0:00:32  lr: 0.007647  min_lr: 0.007647  loss: 3.6657 (3.7028)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [57]  [120/156]  eta: 0:00:25  lr: 0.007646  min_lr: 0.007646  loss: 3.4639 (3.6761)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [57]  [130/156]  eta: 0:00:17  lr: 0.007645  min_lr: 0.007645  loss: 3.6886 (3.6972)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0010  max mem: 55573
Epoch: [57]  [140/156]  eta: 0:00:10  lr: 0.007644  min_lr: 0.007644  loss: 3.9501 (3.6998)  weight_decay: 0.0500 (0.0500)  time: 0.5852  data: 0.0009  max mem: 55573
Epoch: [57]  [150/156]  eta: 0:00:04  lr: 0.007643  min_lr: 0.007643  loss: 3.6747 (3.6988)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [57]  [155/156]  eta: 0:00:00  lr: 0.007642  min_lr: 0.007642  loss: 3.5376 (3.6857)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [57] Total time: 0:01:46 (0.6805 s / it)
Averaged stats: lr: 0.007642  min_lr: 0.007642  loss: 3.5376 (3.6825)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.1860 (1.1860)  acc1: 73.5677 (73.5677)  acc5: 93.7500 (93.7500)  time: 7.1287  data: 6.8857  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3865 (1.4332)  acc1: 69.4010 (67.7760)  acc5: 90.7552 (89.7600)  time: 1.5720  data: 1.3772  max mem: 55573
Test: Total time: 0:00:08 (1.6107 s / it)
* Acc@1 68.888 Acc@5 89.346 loss 1.419
Accuracy of the model on the 50000 test images: 68.9%
Max accuracy: 68.89%
Test:  [0/5]  eta: 0:00:33  loss: 6.9470 (6.9470)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.7973  data: 6.5614  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.1124 (7.0864)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.5049  data: 1.3124  max mem: 55573
Test: Total time: 0:00:07 (1.5279 s / it)
* Acc@1 0.100 Acc@5 0.506 loss 7.034
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [58]  [  0/156]  eta: 0:33:14  lr: 0.007642  min_lr: 0.007642  loss: 3.0269 (3.0269)  weight_decay: 0.0500 (0.0500)  time: 12.7858  data: 10.1727  max mem: 55573
Epoch: [58]  [ 10/156]  eta: 0:04:18  lr: 0.007641  min_lr: 0.007641  loss: 3.7025 (3.6045)  weight_decay: 0.0500 (0.0500)  time: 1.7707  data: 0.9253  max mem: 55573
Epoch: [58]  [ 20/156]  eta: 0:02:44  lr: 0.007640  min_lr: 0.007640  loss: 3.7035 (3.5973)  weight_decay: 0.0500 (0.0500)  time: 0.6312  data: 0.0005  max mem: 55573
Epoch: [58]  [ 30/156]  eta: 0:02:07  lr: 0.007638  min_lr: 0.007638  loss: 3.7759 (3.6597)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [58]  [ 40/156]  eta: 0:01:45  lr: 0.007637  min_lr: 0.007637  loss: 3.8012 (3.6791)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [58]  [ 50/156]  eta: 0:01:29  lr: 0.007636  min_lr: 0.007636  loss: 3.6738 (3.6878)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [58]  [ 60/156]  eta: 0:01:17  lr: 0.007635  min_lr: 0.007635  loss: 3.9464 (3.7349)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [58]  [ 70/156]  eta: 0:01:06  lr: 0.007634  min_lr: 0.007634  loss: 3.7090 (3.6894)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [58]  [ 80/156]  eta: 0:00:57  lr: 0.007632  min_lr: 0.007632  loss: 3.4980 (3.6846)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [58]  [ 90/156]  eta: 0:00:48  lr: 0.007631  min_lr: 0.007631  loss: 3.7213 (3.6783)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [58]  [100/156]  eta: 0:00:40  lr: 0.007630  min_lr: 0.007630  loss: 3.7950 (3.6751)  weight_decay: 0.0500 (0.0500)  time: 0.5885  data: 0.0004  max mem: 55573
Epoch: [58]  [110/156]  eta: 0:00:32  lr: 0.007629  min_lr: 0.007629  loss: 3.4505 (3.6581)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [58]  [120/156]  eta: 0:00:25  lr: 0.007628  min_lr: 0.007628  loss: 3.6453 (3.6701)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [58]  [130/156]  eta: 0:00:17  lr: 0.007626  min_lr: 0.007626  loss: 3.6938 (3.6522)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0010  max mem: 55573
Epoch: [58]  [140/156]  eta: 0:00:10  lr: 0.007625  min_lr: 0.007625  loss: 3.5870 (3.6500)  weight_decay: 0.0500 (0.0500)  time: 0.5871  data: 0.0009  max mem: 55573
Epoch: [58]  [150/156]  eta: 0:00:04  lr: 0.007624  min_lr: 0.007624  loss: 3.5870 (3.6564)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [58]  [155/156]  eta: 0:00:00  lr: 0.007623  min_lr: 0.007623  loss: 3.5870 (3.6532)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [58] Total time: 0:01:45 (0.6795 s / it)
Averaged stats: lr: 0.007623  min_lr: 0.007623  loss: 3.5870 (3.7024)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.5003 (1.5003)  acc1: 72.3958 (72.3958)  acc5: 92.8385 (92.8385)  time: 6.6232  data: 6.3860  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.6075 (1.6421)  acc1: 69.2708 (66.3680)  acc5: 90.3646 (88.7680)  time: 1.4714  data: 1.2773  max mem: 55573
Test: Total time: 0:00:07 (1.5100 s / it)
* Acc@1 67.854 Acc@5 88.758 loss 1.633
Accuracy of the model on the 50000 test images: 67.9%
Max accuracy: 68.89%
Test:  [0/5]  eta: 0:00:34  loss: 6.9543 (6.9543)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9774  data: 6.7414  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.1303 (7.1009)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.5120)  time: 1.5409  data: 1.3484  max mem: 55573
Test: Total time: 0:00:07 (1.5638 s / it)
* Acc@1 0.100 Acc@5 0.500 loss 7.046
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [59]  [  0/156]  eta: 0:34:04  lr: 0.007623  min_lr: 0.007623  loss: 4.1156 (4.1156)  weight_decay: 0.0500 (0.0500)  time: 13.1050  data: 11.7190  max mem: 55573
Epoch: [59]  [ 10/156]  eta: 0:04:18  lr: 0.007622  min_lr: 0.007622  loss: 3.9806 (3.7451)  weight_decay: 0.0500 (0.0500)  time: 1.7674  data: 1.0657  max mem: 55573
Epoch: [59]  [ 20/156]  eta: 0:02:44  lr: 0.007621  min_lr: 0.007621  loss: 3.9806 (3.8477)  weight_decay: 0.0500 (0.0500)  time: 0.6121  data: 0.0004  max mem: 55573
Epoch: [59]  [ 30/156]  eta: 0:02:06  lr: 0.007620  min_lr: 0.007620  loss: 3.6613 (3.7408)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [59]  [ 40/156]  eta: 0:01:45  lr: 0.007618  min_lr: 0.007618  loss: 3.5123 (3.6933)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [59]  [ 50/156]  eta: 0:01:29  lr: 0.007617  min_lr: 0.007617  loss: 3.6875 (3.7062)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [59]  [ 60/156]  eta: 0:01:17  lr: 0.007616  min_lr: 0.007616  loss: 3.8577 (3.7077)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [59]  [ 70/156]  eta: 0:01:06  lr: 0.007615  min_lr: 0.007615  loss: 3.8129 (3.7082)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [59]  [ 80/156]  eta: 0:00:56  lr: 0.007613  min_lr: 0.007613  loss: 3.8129 (3.7114)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0004  max mem: 55573
Epoch: [59]  [ 90/156]  eta: 0:00:48  lr: 0.007612  min_lr: 0.007612  loss: 3.8544 (3.7151)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [59]  [100/156]  eta: 0:00:40  lr: 0.007611  min_lr: 0.007611  loss: 3.8544 (3.6982)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [59]  [110/156]  eta: 0:00:32  lr: 0.007610  min_lr: 0.007610  loss: 3.7537 (3.6926)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [59]  [120/156]  eta: 0:00:25  lr: 0.007608  min_lr: 0.007608  loss: 3.8078 (3.6955)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [59]  [130/156]  eta: 0:00:17  lr: 0.007607  min_lr: 0.007607  loss: 3.6207 (3.6708)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0024  max mem: 55573
Epoch: [59]  [140/156]  eta: 0:00:10  lr: 0.007606  min_lr: 0.007606  loss: 3.5501 (3.6693)  weight_decay: 0.0500 (0.0500)  time: 0.5881  data: 0.0023  max mem: 55573
Epoch: [59]  [150/156]  eta: 0:00:04  lr: 0.007605  min_lr: 0.007605  loss: 3.6687 (3.6663)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [59]  [155/156]  eta: 0:00:00  lr: 0.007604  min_lr: 0.007604  loss: 3.6687 (3.6734)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [59] Total time: 0:01:45 (0.6789 s / it)
Averaged stats: lr: 0.007604  min_lr: 0.007604  loss: 3.6687 (3.6728)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.3541 (1.3541)  acc1: 75.0000 (75.0000)  acc5: 93.0990 (93.0990)  time: 6.7037  data: 6.4631  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5027 (1.5563)  acc1: 69.4010 (66.8480)  acc5: 91.2760 (88.7040)  time: 1.5172  data: 1.3238  max mem: 55573
Test: Total time: 0:00:07 (1.5683 s / it)
* Acc@1 67.868 Acc@5 88.620 loss 1.543
Accuracy of the model on the 50000 test images: 67.9%
Max accuracy: 68.89%
Test:  [0/5]  eta: 0:00:34  loss: 6.9614 (6.9614)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9314  data: 6.6949  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.1494 (7.1161)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.5317  data: 1.3391  max mem: 55573
Test: Total time: 0:00:07 (1.5496 s / it)
* Acc@1 0.100 Acc@5 0.502 loss 7.058
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [60]  [  0/156]  eta: 0:36:47  lr: 0.007604  min_lr: 0.007604  loss: 4.3507 (4.3507)  weight_decay: 0.0500 (0.0500)  time: 14.1534  data: 11.1545  max mem: 55573
Epoch: [60]  [ 10/156]  eta: 0:04:26  lr: 0.007603  min_lr: 0.007603  loss: 3.9603 (3.8837)  weight_decay: 0.0500 (0.0500)  time: 1.8237  data: 1.0144  max mem: 55573
Epoch: [60]  [ 20/156]  eta: 0:02:48  lr: 0.007601  min_lr: 0.007601  loss: 3.8238 (3.7444)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [60]  [ 30/156]  eta: 0:02:09  lr: 0.007600  min_lr: 0.007600  loss: 3.7112 (3.7437)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [60]  [ 40/156]  eta: 0:01:46  lr: 0.007599  min_lr: 0.007599  loss: 3.7964 (3.7323)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [60]  [ 50/156]  eta: 0:01:30  lr: 0.007598  min_lr: 0.007598  loss: 3.9687 (3.7668)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [60]  [ 60/156]  eta: 0:01:18  lr: 0.007596  min_lr: 0.007596  loss: 4.0583 (3.7931)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [60]  [ 70/156]  eta: 0:01:07  lr: 0.007595  min_lr: 0.007595  loss: 4.0225 (3.8111)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [60]  [ 80/156]  eta: 0:00:57  lr: 0.007594  min_lr: 0.007594  loss: 3.8357 (3.7956)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [60]  [ 90/156]  eta: 0:00:48  lr: 0.007593  min_lr: 0.007593  loss: 3.7966 (3.7933)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [60]  [100/156]  eta: 0:00:40  lr: 0.007591  min_lr: 0.007591  loss: 3.6093 (3.7655)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [60]  [110/156]  eta: 0:00:32  lr: 0.007590  min_lr: 0.007590  loss: 3.7508 (3.7750)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [60]  [120/156]  eta: 0:00:25  lr: 0.007589  min_lr: 0.007589  loss: 3.8868 (3.7841)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [60]  [130/156]  eta: 0:00:18  lr: 0.007588  min_lr: 0.007588  loss: 3.8366 (3.7838)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0011  max mem: 55573
Epoch: [60]  [140/156]  eta: 0:00:10  lr: 0.007586  min_lr: 0.007586  loss: 3.7266 (3.7797)  weight_decay: 0.0500 (0.0500)  time: 0.5864  data: 0.0009  max mem: 55573
Epoch: [60]  [150/156]  eta: 0:00:04  lr: 0.007585  min_lr: 0.007585  loss: 3.2387 (3.7362)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [60]  [155/156]  eta: 0:00:00  lr: 0.007584  min_lr: 0.007584  loss: 3.1213 (3.7204)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [60] Total time: 0:01:46 (0.6829 s / it)
Averaged stats: lr: 0.007584  min_lr: 0.007584  loss: 3.1213 (3.6824)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.3568 (1.3568)  acc1: 72.2656 (72.2656)  acc5: 92.4479 (92.4479)  time: 7.1658  data: 6.9284  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4452 (1.5126)  acc1: 69.2708 (66.4960)  acc5: 91.5365 (88.7040)  time: 1.5795  data: 1.3858  max mem: 55573
Test: Total time: 0:00:08 (1.6224 s / it)
* Acc@1 67.788 Acc@5 88.844 loss 1.497
Accuracy of the model on the 50000 test images: 67.8%
Max accuracy: 68.89%
Test:  [0/5]  eta: 0:00:34  loss: 6.9702 (6.9702)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9776  data: 6.7411  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.1697 (7.1320)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.5411  data: 1.3483  max mem: 55573
Test: Total time: 0:00:07 (1.5605 s / it)
* Acc@1 0.100 Acc@5 0.500 loss 7.071
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [61]  [  0/156]  eta: 0:30:31  lr: 0.007584  min_lr: 0.007584  loss: 4.1524 (4.1524)  weight_decay: 0.0500 (0.0500)  time: 11.7399  data: 9.7822  max mem: 55573
Epoch: [61]  [ 10/156]  eta: 0:04:10  lr: 0.007583  min_lr: 0.007583  loss: 3.8777 (3.7792)  weight_decay: 0.0500 (0.0500)  time: 1.7157  data: 0.8897  max mem: 55573
Epoch: [61]  [ 20/156]  eta: 0:02:40  lr: 0.007582  min_lr: 0.007582  loss: 3.8694 (3.8482)  weight_decay: 0.0500 (0.0500)  time: 0.6522  data: 0.0005  max mem: 55573
Epoch: [61]  [ 30/156]  eta: 0:02:04  lr: 0.007580  min_lr: 0.007580  loss: 3.6890 (3.7863)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [61]  [ 40/156]  eta: 0:01:43  lr: 0.007579  min_lr: 0.007579  loss: 3.6324 (3.7514)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [61]  [ 50/156]  eta: 0:01:28  lr: 0.007578  min_lr: 0.007578  loss: 3.6929 (3.7537)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [61]  [ 60/156]  eta: 0:01:16  lr: 0.007577  min_lr: 0.007577  loss: 3.6152 (3.7457)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [61]  [ 70/156]  eta: 0:01:05  lr: 0.007575  min_lr: 0.007575  loss: 3.7065 (3.7334)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0004  max mem: 55573
Epoch: [61]  [ 80/156]  eta: 0:00:56  lr: 0.007574  min_lr: 0.007574  loss: 3.6994 (3.6965)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [61]  [ 90/156]  eta: 0:00:47  lr: 0.007573  min_lr: 0.007573  loss: 3.5574 (3.6962)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [61]  [100/156]  eta: 0:00:39  lr: 0.007571  min_lr: 0.007571  loss: 3.5735 (3.6710)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [61]  [110/156]  eta: 0:00:32  lr: 0.007570  min_lr: 0.007570  loss: 3.7532 (3.6979)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0004  max mem: 55573
Epoch: [61]  [120/156]  eta: 0:00:24  lr: 0.007569  min_lr: 0.007569  loss: 3.7510 (3.6804)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [61]  [130/156]  eta: 0:00:17  lr: 0.007567  min_lr: 0.007567  loss: 3.5191 (3.6611)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0010  max mem: 55573
Epoch: [61]  [140/156]  eta: 0:00:10  lr: 0.007566  min_lr: 0.007566  loss: 3.7105 (3.6813)  weight_decay: 0.0500 (0.0500)  time: 0.5856  data: 0.0008  max mem: 55573
Epoch: [61]  [150/156]  eta: 0:00:04  lr: 0.007565  min_lr: 0.007565  loss: 3.9118 (3.6898)  weight_decay: 0.0500 (0.0500)  time: 0.5828  data: 0.0001  max mem: 55573
Epoch: [61]  [155/156]  eta: 0:00:00  lr: 0.007564  min_lr: 0.007564  loss: 3.8992 (3.7023)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0001  max mem: 55573
Epoch: [61] Total time: 0:01:45 (0.6747 s / it)
Averaged stats: lr: 0.007564  min_lr: 0.007564  loss: 3.8992 (3.6765)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.4131 (1.4131)  acc1: 75.0000 (75.0000)  acc5: 93.0990 (93.0990)  time: 6.9415  data: 6.7041  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5090 (1.5322)  acc1: 70.7031 (68.8640)  acc5: 91.0156 (89.0880)  time: 1.5345  data: 1.3409  max mem: 55573
Test: Total time: 0:00:07 (1.5783 s / it)
* Acc@1 69.228 Acc@5 89.492 loss 1.548
Accuracy of the model on the 50000 test images: 69.2%
Max accuracy: 69.23%
Test:  [0/5]  eta: 0:00:33  loss: 6.9825 (6.9825)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.6625  data: 6.4259  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.1899 (7.1504)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.4781  data: 1.2853  max mem: 55573
Test: Total time: 0:00:07 (1.4940 s / it)
* Acc@1 0.100 Acc@5 0.502 loss 7.087
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [62]  [  0/156]  eta: 0:32:30  lr: 0.007564  min_lr: 0.007564  loss: 3.3726 (3.3726)  weight_decay: 0.0500 (0.0500)  time: 12.5006  data: 8.4870  max mem: 55573
Epoch: [62]  [ 10/156]  eta: 0:04:05  lr: 0.007563  min_lr: 0.007563  loss: 3.8785 (3.8624)  weight_decay: 0.0500 (0.0500)  time: 1.6815  data: 0.7719  max mem: 55573
Epoch: [62]  [ 20/156]  eta: 0:02:38  lr: 0.007561  min_lr: 0.007561  loss: 3.8785 (3.8054)  weight_decay: 0.0500 (0.0500)  time: 0.5980  data: 0.0004  max mem: 55573
Epoch: [62]  [ 30/156]  eta: 0:02:03  lr: 0.007560  min_lr: 0.007560  loss: 3.7243 (3.7936)  weight_decay: 0.0500 (0.0500)  time: 0.5941  data: 0.0004  max mem: 55573
Epoch: [62]  [ 40/156]  eta: 0:01:42  lr: 0.007559  min_lr: 0.007559  loss: 3.8300 (3.8102)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [62]  [ 50/156]  eta: 0:01:27  lr: 0.007558  min_lr: 0.007558  loss: 3.8520 (3.7419)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0005  max mem: 55573
Epoch: [62]  [ 60/156]  eta: 0:01:15  lr: 0.007556  min_lr: 0.007556  loss: 3.8488 (3.7564)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [62]  [ 70/156]  eta: 0:01:05  lr: 0.007555  min_lr: 0.007555  loss: 3.8602 (3.7422)  weight_decay: 0.0500 (0.0500)  time: 0.5883  data: 0.0004  max mem: 55573
Epoch: [62]  [ 80/156]  eta: 0:00:56  lr: 0.007554  min_lr: 0.007554  loss: 3.6109 (3.7188)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0004  max mem: 55573
Epoch: [62]  [ 90/156]  eta: 0:00:47  lr: 0.007552  min_lr: 0.007552  loss: 3.7314 (3.7192)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [62]  [100/156]  eta: 0:00:39  lr: 0.007551  min_lr: 0.007551  loss: 3.6121 (3.6913)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [62]  [110/156]  eta: 0:00:32  lr: 0.007550  min_lr: 0.007550  loss: 3.6121 (3.6853)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [62]  [120/156]  eta: 0:00:24  lr: 0.007548  min_lr: 0.007548  loss: 3.7891 (3.6880)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [62]  [130/156]  eta: 0:00:17  lr: 0.007547  min_lr: 0.007547  loss: 3.6984 (3.6842)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0010  max mem: 55573
Epoch: [62]  [140/156]  eta: 0:00:10  lr: 0.007546  min_lr: 0.007546  loss: 3.6942 (3.6773)  weight_decay: 0.0500 (0.0500)  time: 0.5868  data: 0.0008  max mem: 55573
Epoch: [62]  [150/156]  eta: 0:00:04  lr: 0.007544  min_lr: 0.007544  loss: 3.8642 (3.6778)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [62]  [155/156]  eta: 0:00:00  lr: 0.007544  min_lr: 0.007544  loss: 3.8642 (3.6850)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [62] Total time: 0:01:45 (0.6735 s / it)
Averaged stats: lr: 0.007544  min_lr: 0.007544  loss: 3.8642 (3.6696)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.4726 (1.4726)  acc1: 72.9167 (72.9167)  acc5: 92.4479 (92.4479)  time: 7.2107  data: 6.9732  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5584 (1.6441)  acc1: 70.3125 (67.2640)  acc5: 92.4479 (89.1520)  time: 1.5886  data: 1.3947  max mem: 55573
Test: Total time: 0:00:08 (1.6336 s / it)
* Acc@1 68.172 Acc@5 88.904 loss 1.644
Accuracy of the model on the 50000 test images: 68.2%
Max accuracy: 69.23%
Test:  [0/5]  eta: 0:00:34  loss: 6.9964 (6.9964)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9454  data: 6.7094  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.2121 (7.1699)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.5343  data: 1.3420  max mem: 55573
Test: Total time: 0:00:07 (1.5563 s / it)
* Acc@1 0.100 Acc@5 0.510 loss 7.103
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [63]  [  0/156]  eta: 0:35:17  lr: 0.007543  min_lr: 0.007543  loss: 2.6704 (2.6704)  weight_decay: 0.0500 (0.0500)  time: 13.5762  data: 10.6600  max mem: 55573
Epoch: [63]  [ 10/156]  eta: 0:04:26  lr: 0.007542  min_lr: 0.007542  loss: 3.4398 (3.5437)  weight_decay: 0.0500 (0.0500)  time: 1.8262  data: 0.9695  max mem: 55573
Epoch: [63]  [ 20/156]  eta: 0:02:48  lr: 0.007541  min_lr: 0.007541  loss: 3.5610 (3.5539)  weight_decay: 0.0500 (0.0500)  time: 0.6210  data: 0.0005  max mem: 55573
Epoch: [63]  [ 30/156]  eta: 0:02:09  lr: 0.007539  min_lr: 0.007539  loss: 3.5610 (3.5082)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [63]  [ 40/156]  eta: 0:01:47  lr: 0.007538  min_lr: 0.007538  loss: 3.6408 (3.5547)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [63]  [ 50/156]  eta: 0:01:30  lr: 0.007537  min_lr: 0.007537  loss: 3.6408 (3.5628)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [63]  [ 60/156]  eta: 0:01:18  lr: 0.007535  min_lr: 0.007535  loss: 3.8152 (3.5758)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [63]  [ 70/156]  eta: 0:01:07  lr: 0.007534  min_lr: 0.007534  loss: 3.8073 (3.5782)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [63]  [ 80/156]  eta: 0:00:57  lr: 0.007533  min_lr: 0.007533  loss: 3.6082 (3.5827)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [63]  [ 90/156]  eta: 0:00:48  lr: 0.007531  min_lr: 0.007531  loss: 3.8815 (3.5979)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0005  max mem: 55573
Epoch: [63]  [100/156]  eta: 0:00:40  lr: 0.007530  min_lr: 0.007530  loss: 3.8225 (3.5946)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [63]  [110/156]  eta: 0:00:32  lr: 0.007529  min_lr: 0.007529  loss: 3.8133 (3.6033)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [63]  [120/156]  eta: 0:00:25  lr: 0.007527  min_lr: 0.007527  loss: 3.9610 (3.6280)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [63]  [130/156]  eta: 0:00:18  lr: 0.007526  min_lr: 0.007526  loss: 3.7798 (3.6096)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0010  max mem: 55573
Epoch: [63]  [140/156]  eta: 0:00:10  lr: 0.007525  min_lr: 0.007525  loss: 3.5145 (3.6145)  weight_decay: 0.0500 (0.0500)  time: 0.5866  data: 0.0009  max mem: 55573
Epoch: [63]  [150/156]  eta: 0:00:04  lr: 0.007523  min_lr: 0.007523  loss: 3.6314 (3.6135)  weight_decay: 0.0500 (0.0500)  time: 0.5830  data: 0.0001  max mem: 55573
Epoch: [63]  [155/156]  eta: 0:00:00  lr: 0.007523  min_lr: 0.007523  loss: 3.8886 (3.6223)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0001  max mem: 55573
Epoch: [63] Total time: 0:01:46 (0.6830 s / it)
Averaged stats: lr: 0.007523  min_lr: 0.007523  loss: 3.8886 (3.6536)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.3702 (1.3702)  acc1: 73.1771 (73.1771)  acc5: 92.5781 (92.5781)  time: 7.2350  data: 6.9976  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5157 (1.5342)  acc1: 69.0104 (67.3280)  acc5: 91.5365 (88.9600)  time: 1.5935  data: 1.3996  max mem: 55573
Test: Total time: 0:00:08 (1.6381 s / it)
* Acc@1 68.352 Acc@5 89.252 loss 1.526
Accuracy of the model on the 50000 test images: 68.4%
Max accuracy: 69.23%
Test:  [0/5]  eta: 0:00:36  loss: 7.0108 (7.0108)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.3432  data: 7.1073  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.2357 (7.1907)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.6141  data: 1.4216  max mem: 55573
Test: Total time: 0:00:08 (1.6363 s / it)
* Acc@1 0.100 Acc@5 0.512 loss 7.121
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [64]  [  0/156]  eta: 0:37:05  lr: 0.007522  min_lr: 0.007522  loss: 3.2708 (3.2708)  weight_decay: 0.0500 (0.0500)  time: 14.2668  data: 12.6012  max mem: 55573
Epoch: [64]  [ 10/156]  eta: 0:04:27  lr: 0.007521  min_lr: 0.007521  loss: 3.6016 (3.5085)  weight_decay: 0.0500 (0.0500)  time: 1.8344  data: 1.1460  max mem: 55573
Epoch: [64]  [ 20/156]  eta: 0:02:49  lr: 0.007520  min_lr: 0.007520  loss: 3.6194 (3.5699)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [64]  [ 30/156]  eta: 0:02:10  lr: 0.007518  min_lr: 0.007518  loss: 3.7709 (3.6633)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [64]  [ 40/156]  eta: 0:01:47  lr: 0.007517  min_lr: 0.007517  loss: 3.8483 (3.6895)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [64]  [ 50/156]  eta: 0:01:31  lr: 0.007516  min_lr: 0.007516  loss: 4.0223 (3.7458)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [64]  [ 60/156]  eta: 0:01:18  lr: 0.007514  min_lr: 0.007514  loss: 4.0301 (3.7446)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [64]  [ 70/156]  eta: 0:01:07  lr: 0.007513  min_lr: 0.007513  loss: 3.7700 (3.7338)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0005  max mem: 55573
Epoch: [64]  [ 80/156]  eta: 0:00:57  lr: 0.007511  min_lr: 0.007511  loss: 3.8461 (3.7574)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0004  max mem: 55573
Epoch: [64]  [ 90/156]  eta: 0:00:48  lr: 0.007510  min_lr: 0.007510  loss: 3.8792 (3.7560)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [64]  [100/156]  eta: 0:00:40  lr: 0.007509  min_lr: 0.007509  loss: 3.7884 (3.7636)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [64]  [110/156]  eta: 0:00:32  lr: 0.007507  min_lr: 0.007507  loss: 3.8517 (3.7545)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [64]  [120/156]  eta: 0:00:25  lr: 0.007506  min_lr: 0.007506  loss: 3.8094 (3.7342)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [64]  [130/156]  eta: 0:00:18  lr: 0.007505  min_lr: 0.007505  loss: 3.7600 (3.7419)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0010  max mem: 55573
Epoch: [64]  [140/156]  eta: 0:00:11  lr: 0.007503  min_lr: 0.007503  loss: 3.7813 (3.7423)  weight_decay: 0.0500 (0.0500)  time: 0.5869  data: 0.0008  max mem: 55573
Epoch: [64]  [150/156]  eta: 0:00:04  lr: 0.007502  min_lr: 0.007502  loss: 3.6702 (3.7330)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [64]  [155/156]  eta: 0:00:00  lr: 0.007501  min_lr: 0.007501  loss: 3.6702 (3.7290)  weight_decay: 0.0500 (0.0500)  time: 0.5812  data: 0.0001  max mem: 55573
Epoch: [64] Total time: 0:01:46 (0.6842 s / it)
Averaged stats: lr: 0.007501  min_lr: 0.007501  loss: 3.6702 (3.6503)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:37  loss: 1.3487 (1.3487)  acc1: 73.4375 (73.4375)  acc5: 92.7083 (92.7083)  time: 7.5023  data: 7.2649  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4762 (1.4940)  acc1: 70.7031 (68.1280)  acc5: 91.1458 (89.0880)  time: 1.6471  data: 1.4531  max mem: 55573
Test: Total time: 0:00:08 (1.6990 s / it)
* Acc@1 68.598 Acc@5 89.282 loss 1.515
Accuracy of the model on the 50000 test images: 68.6%
Max accuracy: 69.23%
Test:  [0/5]  eta: 0:00:35  loss: 7.0252 (7.0252)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.0609  data: 6.8247  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.2637 (7.2150)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.5575  data: 1.3651  max mem: 55573
Test: Total time: 0:00:07 (1.5793 s / it)
* Acc@1 0.100 Acc@5 0.512 loss 7.142
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [65]  [  0/156]  eta: 0:34:08  lr: 0.007501  min_lr: 0.007501  loss: 4.0552 (4.0552)  weight_decay: 0.0500 (0.0500)  time: 13.1284  data: 8.6262  max mem: 55573
Epoch: [65]  [ 10/156]  eta: 0:04:17  lr: 0.007500  min_lr: 0.007500  loss: 3.4529 (3.5803)  weight_decay: 0.0500 (0.0500)  time: 1.7665  data: 0.7853  max mem: 55573
Epoch: [65]  [ 20/156]  eta: 0:02:44  lr: 0.007498  min_lr: 0.007498  loss: 3.4430 (3.5408)  weight_decay: 0.0500 (0.0500)  time: 0.6111  data: 0.0011  max mem: 55573
Epoch: [65]  [ 30/156]  eta: 0:02:07  lr: 0.007497  min_lr: 0.007497  loss: 3.7263 (3.6366)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0009  max mem: 55573
Epoch: [65]  [ 40/156]  eta: 0:01:45  lr: 0.007495  min_lr: 0.007495  loss: 3.9839 (3.7017)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0009  max mem: 55573
Epoch: [65]  [ 50/156]  eta: 0:01:29  lr: 0.007494  min_lr: 0.007494  loss: 3.8426 (3.6549)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0009  max mem: 55573
Epoch: [65]  [ 60/156]  eta: 0:01:17  lr: 0.007493  min_lr: 0.007493  loss: 3.7749 (3.6841)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0009  max mem: 55573
Epoch: [65]  [ 70/156]  eta: 0:01:06  lr: 0.007491  min_lr: 0.007491  loss: 3.9036 (3.6874)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0009  max mem: 55573
Epoch: [65]  [ 80/156]  eta: 0:00:57  lr: 0.007490  min_lr: 0.007490  loss: 3.6815 (3.6574)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0009  max mem: 55573
Epoch: [65]  [ 90/156]  eta: 0:00:48  lr: 0.007488  min_lr: 0.007488  loss: 3.4978 (3.6405)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0009  max mem: 55573
Epoch: [65]  [100/156]  eta: 0:00:40  lr: 0.007487  min_lr: 0.007487  loss: 3.4978 (3.6271)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0009  max mem: 55573
Epoch: [65]  [110/156]  eta: 0:00:32  lr: 0.007486  min_lr: 0.007486  loss: 3.9659 (3.6748)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0009  max mem: 55573
Epoch: [65]  [120/156]  eta: 0:00:25  lr: 0.007484  min_lr: 0.007484  loss: 4.0678 (3.6956)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0009  max mem: 55573
Epoch: [65]  [130/156]  eta: 0:00:17  lr: 0.007483  min_lr: 0.007483  loss: 3.9444 (3.7007)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0020  max mem: 55573
Epoch: [65]  [140/156]  eta: 0:00:10  lr: 0.007481  min_lr: 0.007481  loss: 3.8485 (3.6940)  weight_decay: 0.0500 (0.0500)  time: 0.5853  data: 0.0016  max mem: 55573
Epoch: [65]  [150/156]  eta: 0:00:04  lr: 0.007480  min_lr: 0.007480  loss: 3.8485 (3.6983)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [65]  [155/156]  eta: 0:00:00  lr: 0.007479  min_lr: 0.007479  loss: 3.8897 (3.6958)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [65] Total time: 0:01:45 (0.6791 s / it)
Averaged stats: lr: 0.007479  min_lr: 0.007479  loss: 3.8897 (3.6613)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.3896 (1.3896)  acc1: 72.9167 (72.9167)  acc5: 92.0573 (92.0573)  time: 7.0447  data: 6.8068  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4506 (1.4701)  acc1: 70.1823 (67.7440)  acc5: 92.0573 (89.4080)  time: 1.5557  data: 1.3616  max mem: 55573
Test: Total time: 0:00:07 (1.5974 s / it)
* Acc@1 69.020 Acc@5 89.550 loss 1.461
Accuracy of the model on the 50000 test images: 69.0%
Max accuracy: 69.23%
Test:  [0/5]  eta: 0:00:35  loss: 7.0433 (7.0433)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.1731  data: 6.9370  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.2927 (7.2418)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.5799  data: 1.3875  max mem: 55573
Test: Total time: 0:00:07 (1.5987 s / it)
* Acc@1 0.100 Acc@5 0.514 loss 7.165
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [66]  [  0/156]  eta: 0:37:50  lr: 0.007479  min_lr: 0.007479  loss: 3.4993 (3.4993)  weight_decay: 0.0500 (0.0500)  time: 14.5570  data: 10.3143  max mem: 55573
Epoch: [66]  [ 10/156]  eta: 0:04:34  lr: 0.007478  min_lr: 0.007478  loss: 4.0905 (3.9692)  weight_decay: 0.0500 (0.0500)  time: 1.8817  data: 0.9381  max mem: 55573
Epoch: [66]  [ 20/156]  eta: 0:02:52  lr: 0.007476  min_lr: 0.007476  loss: 3.7041 (3.6499)  weight_decay: 0.0500 (0.0500)  time: 0.6023  data: 0.0004  max mem: 55573
Epoch: [66]  [ 30/156]  eta: 0:02:12  lr: 0.007475  min_lr: 0.007475  loss: 3.4754 (3.6424)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [66]  [ 40/156]  eta: 0:01:48  lr: 0.007473  min_lr: 0.007473  loss: 3.6056 (3.6548)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [66]  [ 50/156]  eta: 0:01:32  lr: 0.007472  min_lr: 0.007472  loss: 3.7541 (3.6970)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [66]  [ 60/156]  eta: 0:01:19  lr: 0.007470  min_lr: 0.007470  loss: 3.7883 (3.7106)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [66]  [ 70/156]  eta: 0:01:08  lr: 0.007469  min_lr: 0.007469  loss: 3.7688 (3.7141)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [66]  [ 80/156]  eta: 0:00:58  lr: 0.007468  min_lr: 0.007468  loss: 3.6345 (3.6799)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [66]  [ 90/156]  eta: 0:00:49  lr: 0.007466  min_lr: 0.007466  loss: 3.2086 (3.6527)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [66]  [100/156]  eta: 0:00:40  lr: 0.007465  min_lr: 0.007465  loss: 3.7385 (3.6620)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [66]  [110/156]  eta: 0:00:33  lr: 0.007463  min_lr: 0.007463  loss: 3.6709 (3.6339)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [66]  [120/156]  eta: 0:00:25  lr: 0.007462  min_lr: 0.007462  loss: 3.3792 (3.6451)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [66]  [130/156]  eta: 0:00:18  lr: 0.007460  min_lr: 0.007460  loss: 3.7873 (3.6419)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0010  max mem: 55573
Epoch: [66]  [140/156]  eta: 0:00:11  lr: 0.007459  min_lr: 0.007459  loss: 3.5415 (3.6321)  weight_decay: 0.0500 (0.0500)  time: 0.5865  data: 0.0009  max mem: 55573
Epoch: [66]  [150/156]  eta: 0:00:04  lr: 0.007458  min_lr: 0.007458  loss: 3.5415 (3.6274)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [66]  [155/156]  eta: 0:00:00  lr: 0.007457  min_lr: 0.007457  loss: 3.6145 (3.6327)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [66] Total time: 0:01:47 (0.6876 s / it)
Averaged stats: lr: 0.007457  min_lr: 0.007457  loss: 3.6145 (3.6554)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.2986 (1.2986)  acc1: 73.1771 (73.1771)  acc5: 91.9271 (91.9271)  time: 6.8549  data: 6.6175  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3871 (1.4808)  acc1: 72.7865 (67.9040)  acc5: 91.6667 (88.3840)  time: 1.5175  data: 1.3236  max mem: 55573
Test: Total time: 0:00:07 (1.5615 s / it)
* Acc@1 68.326 Acc@5 88.936 loss 1.462
Accuracy of the model on the 50000 test images: 68.3%
Max accuracy: 69.23%
Test:  [0/5]  eta: 0:00:36  loss: 7.0625 (7.0625)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.2123  data: 6.9762  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.3189 (7.2669)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.5760)  time: 1.5878  data: 1.3953  max mem: 55573
Test: Total time: 0:00:08 (1.6059 s / it)
* Acc@1 0.100 Acc@5 0.516 loss 7.187
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [67]  [  0/156]  eta: 0:39:05  lr: 0.007457  min_lr: 0.007457  loss: 3.4193 (3.4193)  weight_decay: 0.0500 (0.0500)  time: 15.0324  data: 10.1103  max mem: 55573
Epoch: [67]  [ 10/156]  eta: 0:04:37  lr: 0.007455  min_lr: 0.007455  loss: 3.4514 (3.4561)  weight_decay: 0.0500 (0.0500)  time: 1.9028  data: 0.9194  max mem: 55573
Epoch: [67]  [ 20/156]  eta: 0:02:53  lr: 0.007454  min_lr: 0.007454  loss: 3.8548 (3.6051)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0003  max mem: 55573
Epoch: [67]  [ 30/156]  eta: 0:02:13  lr: 0.007452  min_lr: 0.007452  loss: 3.8660 (3.6337)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0003  max mem: 55573
Epoch: [67]  [ 40/156]  eta: 0:01:49  lr: 0.007451  min_lr: 0.007451  loss: 3.8660 (3.6701)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [67]  [ 50/156]  eta: 0:01:32  lr: 0.007449  min_lr: 0.007449  loss: 3.8242 (3.6475)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [67]  [ 60/156]  eta: 0:01:19  lr: 0.007448  min_lr: 0.007448  loss: 3.8290 (3.6792)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [67]  [ 70/156]  eta: 0:01:08  lr: 0.007446  min_lr: 0.007446  loss: 3.7197 (3.6522)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [67]  [ 80/156]  eta: 0:00:58  lr: 0.007445  min_lr: 0.007445  loss: 3.4640 (3.6679)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [67]  [ 90/156]  eta: 0:00:49  lr: 0.007444  min_lr: 0.007444  loss: 3.6125 (3.6507)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [67]  [100/156]  eta: 0:00:41  lr: 0.007442  min_lr: 0.007442  loss: 3.8278 (3.6561)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [67]  [110/156]  eta: 0:00:33  lr: 0.007441  min_lr: 0.007441  loss: 3.3134 (3.6231)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [67]  [120/156]  eta: 0:00:25  lr: 0.007439  min_lr: 0.007439  loss: 3.2840 (3.6078)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [67]  [130/156]  eta: 0:00:18  lr: 0.007438  min_lr: 0.007438  loss: 3.7564 (3.6281)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0010  max mem: 55573
Epoch: [67]  [140/156]  eta: 0:00:11  lr: 0.007436  min_lr: 0.007436  loss: 3.6696 (3.6153)  weight_decay: 0.0500 (0.0500)  time: 0.5864  data: 0.0008  max mem: 55573
Epoch: [67]  [150/156]  eta: 0:00:04  lr: 0.007435  min_lr: 0.007435  loss: 3.5409 (3.6044)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [67]  [155/156]  eta: 0:00:00  lr: 0.007434  min_lr: 0.007434  loss: 3.5409 (3.6029)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [67] Total time: 0:01:47 (0.6884 s / it)
Averaged stats: lr: 0.007434  min_lr: 0.007434  loss: 3.5409 (3.6533)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.2583 (1.2583)  acc1: 72.9167 (72.9167)  acc5: 92.5781 (92.5781)  time: 7.1167  data: 6.8793  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3720 (1.3659)  acc1: 70.0521 (67.8720)  acc5: 91.2760 (89.7280)  time: 1.5701  data: 1.3760  max mem: 55573
Test: Total time: 0:00:08 (1.6109 s / it)
* Acc@1 69.668 Acc@5 89.988 loss 1.356
Accuracy of the model on the 50000 test images: 69.7%
Max accuracy: 69.67%
Test:  [0/5]  eta: 0:00:34  loss: 7.0829 (7.0829)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9900  data: 6.7522  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.3412 (7.2931)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.5440)  time: 1.5434  data: 1.3506  max mem: 55573
Test: Total time: 0:00:07 (1.5600 s / it)
* Acc@1 0.100 Acc@5 0.518 loss 7.211
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [68]  [  0/156]  eta: 0:34:07  lr: 0.007434  min_lr: 0.007434  loss: 3.9372 (3.9372)  weight_decay: 0.0500 (0.0500)  time: 13.1248  data: 11.8134  max mem: 55573
Epoch: [68]  [ 10/156]  eta: 0:04:17  lr: 0.007432  min_lr: 0.007432  loss: 3.5956 (3.5638)  weight_decay: 0.0500 (0.0500)  time: 1.7654  data: 1.0743  max mem: 55573
Epoch: [68]  [ 20/156]  eta: 0:02:44  lr: 0.007431  min_lr: 0.007431  loss: 3.5734 (3.5503)  weight_decay: 0.0500 (0.0500)  time: 0.6101  data: 0.0004  max mem: 55573
Epoch: [68]  [ 30/156]  eta: 0:02:06  lr: 0.007429  min_lr: 0.007429  loss: 3.8187 (3.7066)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [68]  [ 40/156]  eta: 0:01:45  lr: 0.007428  min_lr: 0.007428  loss: 3.8716 (3.6616)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [68]  [ 50/156]  eta: 0:01:29  lr: 0.007426  min_lr: 0.007426  loss: 3.5019 (3.6326)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [68]  [ 60/156]  eta: 0:01:17  lr: 0.007425  min_lr: 0.007425  loss: 3.5904 (3.6390)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0004  max mem: 55573
Epoch: [68]  [ 70/156]  eta: 0:01:06  lr: 0.007423  min_lr: 0.007423  loss: 3.7320 (3.6400)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [68]  [ 80/156]  eta: 0:00:57  lr: 0.007422  min_lr: 0.007422  loss: 3.8265 (3.6770)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [68]  [ 90/156]  eta: 0:00:48  lr: 0.007421  min_lr: 0.007421  loss: 3.7936 (3.6753)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [68]  [100/156]  eta: 0:00:40  lr: 0.007419  min_lr: 0.007419  loss: 3.6263 (3.6810)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [68]  [110/156]  eta: 0:00:32  lr: 0.007418  min_lr: 0.007418  loss: 3.8203 (3.6671)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [68]  [120/156]  eta: 0:00:25  lr: 0.007416  min_lr: 0.007416  loss: 3.8180 (3.6582)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [68]  [130/156]  eta: 0:00:17  lr: 0.007415  min_lr: 0.007415  loss: 3.8515 (3.6748)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0010  max mem: 55573
Epoch: [68]  [140/156]  eta: 0:00:10  lr: 0.007413  min_lr: 0.007413  loss: 3.9212 (3.6764)  weight_decay: 0.0500 (0.0500)  time: 0.5874  data: 0.0009  max mem: 55573
Epoch: [68]  [150/156]  eta: 0:00:04  lr: 0.007412  min_lr: 0.007412  loss: 3.6560 (3.6777)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [68]  [155/156]  eta: 0:00:00  lr: 0.007411  min_lr: 0.007411  loss: 3.5081 (3.6681)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [68] Total time: 0:01:46 (0.6796 s / it)
Averaged stats: lr: 0.007411  min_lr: 0.007411  loss: 3.5081 (3.6588)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.1781 (1.1781)  acc1: 74.4792 (74.4792)  acc5: 93.7500 (93.7500)  time: 6.8750  data: 6.6375  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3242 (1.3553)  acc1: 73.0469 (69.7600)  acc5: 92.4479 (90.1120)  time: 1.5216  data: 1.3276  max mem: 55573
Test: Total time: 0:00:07 (1.5688 s / it)
* Acc@1 70.164 Acc@5 90.088 loss 1.345
Accuracy of the model on the 50000 test images: 70.2%
Max accuracy: 70.16%
Test:  [0/5]  eta: 0:00:36  loss: 7.1014 (7.1014)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.2202  data: 6.9836  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.3647 (7.3195)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.5120)  time: 1.5896  data: 1.3968  max mem: 55573
Test: Total time: 0:00:08 (1.6095 s / it)
* Acc@1 0.100 Acc@5 0.518 loss 7.235
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [69]  [  0/156]  eta: 0:30:56  lr: 0.007411  min_lr: 0.007411  loss: 3.3690 (3.3690)  weight_decay: 0.0500 (0.0500)  time: 11.9022  data: 10.5920  max mem: 55573
Epoch: [69]  [ 10/156]  eta: 0:04:18  lr: 0.007409  min_lr: 0.007409  loss: 3.7891 (3.7014)  weight_decay: 0.0500 (0.0500)  time: 1.7718  data: 0.9633  max mem: 55573
Epoch: [69]  [ 20/156]  eta: 0:02:44  lr: 0.007408  min_lr: 0.007408  loss: 3.7549 (3.5930)  weight_decay: 0.0500 (0.0500)  time: 0.6760  data: 0.0004  max mem: 55573
Epoch: [69]  [ 30/156]  eta: 0:02:07  lr: 0.007406  min_lr: 0.007406  loss: 3.7549 (3.6275)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [69]  [ 40/156]  eta: 0:01:45  lr: 0.007405  min_lr: 0.007405  loss: 3.7486 (3.5899)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [69]  [ 50/156]  eta: 0:01:29  lr: 0.007403  min_lr: 0.007403  loss: 3.6994 (3.6072)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [69]  [ 60/156]  eta: 0:01:17  lr: 0.007402  min_lr: 0.007402  loss: 3.6570 (3.5762)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [69]  [ 70/156]  eta: 0:01:06  lr: 0.007400  min_lr: 0.007400  loss: 3.6656 (3.5755)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [69]  [ 80/156]  eta: 0:00:57  lr: 0.007399  min_lr: 0.007399  loss: 3.8029 (3.6001)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [69]  [ 90/156]  eta: 0:00:48  lr: 0.007397  min_lr: 0.007397  loss: 3.9576 (3.6323)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [69]  [100/156]  eta: 0:00:40  lr: 0.007396  min_lr: 0.007396  loss: 3.8362 (3.6478)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [69]  [110/156]  eta: 0:00:32  lr: 0.007394  min_lr: 0.007394  loss: 3.7992 (3.6568)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [69]  [120/156]  eta: 0:00:25  lr: 0.007392  min_lr: 0.007392  loss: 3.8058 (3.6567)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0005  max mem: 55573
Epoch: [69]  [130/156]  eta: 0:00:17  lr: 0.007391  min_lr: 0.007391  loss: 3.7921 (3.6518)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0010  max mem: 55573
Epoch: [69]  [140/156]  eta: 0:00:10  lr: 0.007389  min_lr: 0.007389  loss: 3.7412 (3.6411)  weight_decay: 0.0500 (0.0500)  time: 0.5865  data: 0.0009  max mem: 55573
Epoch: [69]  [150/156]  eta: 0:00:04  lr: 0.007388  min_lr: 0.007388  loss: 3.7412 (3.6421)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [69]  [155/156]  eta: 0:00:00  lr: 0.007387  min_lr: 0.007387  loss: 3.7412 (3.6272)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [69] Total time: 0:01:45 (0.6792 s / it)
Averaged stats: lr: 0.007387  min_lr: 0.007387  loss: 3.7412 (3.6601)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.2034 (1.2034)  acc1: 74.7396 (74.7396)  acc5: 93.2292 (93.2292)  time: 7.2372  data: 6.9998  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3396 (1.4097)  acc1: 72.1354 (68.8960)  acc5: 91.2760 (89.8240)  time: 1.5939  data: 1.4000  max mem: 55573
Test: Total time: 0:00:08 (1.6325 s / it)
* Acc@1 69.762 Acc@5 89.714 loss 1.402
Accuracy of the model on the 50000 test images: 69.8%
Max accuracy: 70.16%
Test:  [0/5]  eta: 0:00:34  loss: 7.1240 (7.1240)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9874  data: 6.7512  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.3882 (7.3462)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.5428  data: 1.3503  max mem: 55573
Test: Total time: 0:00:07 (1.5670 s / it)
* Acc@1 0.100 Acc@5 0.510 loss 7.260
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [70]  [  0/156]  eta: 0:35:43  lr: 0.007387  min_lr: 0.007387  loss: 3.1363 (3.1363)  weight_decay: 0.0500 (0.0500)  time: 13.7389  data: 8.8863  max mem: 55573
Epoch: [70]  [ 10/156]  eta: 0:04:20  lr: 0.007385  min_lr: 0.007385  loss: 3.3972 (3.4653)  weight_decay: 0.0500 (0.0500)  time: 1.7868  data: 0.8082  max mem: 55573
Epoch: [70]  [ 20/156]  eta: 0:02:45  lr: 0.007384  min_lr: 0.007384  loss: 3.4557 (3.5029)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [70]  [ 30/156]  eta: 0:02:07  lr: 0.007382  min_lr: 0.007382  loss: 3.7200 (3.6215)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0003  max mem: 55573
Epoch: [70]  [ 40/156]  eta: 0:01:45  lr: 0.007381  min_lr: 0.007381  loss: 3.7495 (3.6051)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [70]  [ 50/156]  eta: 0:01:29  lr: 0.007379  min_lr: 0.007379  loss: 3.7495 (3.6309)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [70]  [ 60/156]  eta: 0:01:17  lr: 0.007378  min_lr: 0.007378  loss: 3.8747 (3.6595)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [70]  [ 70/156]  eta: 0:01:06  lr: 0.007376  min_lr: 0.007376  loss: 3.6878 (3.6230)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0005  max mem: 55573
Epoch: [70]  [ 80/156]  eta: 0:00:57  lr: 0.007375  min_lr: 0.007375  loss: 3.6878 (3.6365)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [70]  [ 90/156]  eta: 0:00:48  lr: 0.007373  min_lr: 0.007373  loss: 3.7876 (3.6618)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [70]  [100/156]  eta: 0:00:40  lr: 0.007372  min_lr: 0.007372  loss: 3.8509 (3.6805)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [70]  [110/156]  eta: 0:00:32  lr: 0.007370  min_lr: 0.007370  loss: 3.7404 (3.6832)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [70]  [120/156]  eta: 0:00:25  lr: 0.007368  min_lr: 0.007368  loss: 3.7003 (3.6728)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [70]  [130/156]  eta: 0:00:17  lr: 0.007367  min_lr: 0.007367  loss: 3.5681 (3.6745)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0010  max mem: 55573
Epoch: [70]  [140/156]  eta: 0:00:10  lr: 0.007365  min_lr: 0.007365  loss: 3.7107 (3.6642)  weight_decay: 0.0500 (0.0500)  time: 0.5848  data: 0.0008  max mem: 55573
Epoch: [70]  [150/156]  eta: 0:00:04  lr: 0.007364  min_lr: 0.007364  loss: 3.8821 (3.6754)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [70]  [155/156]  eta: 0:00:00  lr: 0.007363  min_lr: 0.007363  loss: 3.8346 (3.6632)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [70] Total time: 0:01:46 (0.6807 s / it)
Averaged stats: lr: 0.007363  min_lr: 0.007363  loss: 3.8346 (3.6258)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.2324 (1.2324)  acc1: 74.7396 (74.7396)  acc5: 92.1875 (92.1875)  time: 7.0085  data: 6.7714  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4449 (1.4308)  acc1: 70.5729 (68.2560)  acc5: 90.7552 (89.1840)  time: 1.5482  data: 1.3544  max mem: 55573
Test: Total time: 0:00:07 (1.5890 s / it)
* Acc@1 69.432 Acc@5 89.758 loss 1.428
Accuracy of the model on the 50000 test images: 69.4%
Max accuracy: 70.16%
Test:  [0/5]  eta: 0:00:37  loss: 7.1504 (7.1504)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.4552  data: 7.2192  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.4115 (7.3761)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.6364  data: 1.4439  max mem: 55573
Test: Total time: 0:00:08 (1.6559 s / it)
* Acc@1 0.100 Acc@5 0.508 loss 7.288
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [71]  [  0/156]  eta: 0:34:39  lr: 0.007363  min_lr: 0.007363  loss: 3.9375 (3.9375)  weight_decay: 0.0500 (0.0500)  time: 13.3291  data: 11.2247  max mem: 55573
Epoch: [71]  [ 10/156]  eta: 0:04:22  lr: 0.007361  min_lr: 0.007361  loss: 3.7503 (3.5207)  weight_decay: 0.0500 (0.0500)  time: 1.8002  data: 1.0208  max mem: 55573
Epoch: [71]  [ 20/156]  eta: 0:02:46  lr: 0.007360  min_lr: 0.007360  loss: 3.7820 (3.6316)  weight_decay: 0.0500 (0.0500)  time: 0.6192  data: 0.0003  max mem: 55573
Epoch: [71]  [ 30/156]  eta: 0:02:08  lr: 0.007358  min_lr: 0.007358  loss: 3.7172 (3.5813)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0003  max mem: 55573
Epoch: [71]  [ 40/156]  eta: 0:01:46  lr: 0.007357  min_lr: 0.007357  loss: 3.6378 (3.6118)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [71]  [ 50/156]  eta: 0:01:30  lr: 0.007355  min_lr: 0.007355  loss: 3.7129 (3.6034)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [71]  [ 60/156]  eta: 0:01:17  lr: 0.007354  min_lr: 0.007354  loss: 3.7168 (3.6139)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0005  max mem: 55573
Epoch: [71]  [ 70/156]  eta: 0:01:06  lr: 0.007352  min_lr: 0.007352  loss: 3.8591 (3.6512)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [71]  [ 80/156]  eta: 0:00:57  lr: 0.007350  min_lr: 0.007350  loss: 3.6917 (3.6457)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [71]  [ 90/156]  eta: 0:00:48  lr: 0.007349  min_lr: 0.007349  loss: 3.5241 (3.6484)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [71]  [100/156]  eta: 0:00:40  lr: 0.007347  min_lr: 0.007347  loss: 3.6010 (3.6259)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [71]  [110/156]  eta: 0:00:32  lr: 0.007346  min_lr: 0.007346  loss: 3.6010 (3.6200)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [71]  [120/156]  eta: 0:00:25  lr: 0.007344  min_lr: 0.007344  loss: 3.6326 (3.6164)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [71]  [130/156]  eta: 0:00:17  lr: 0.007342  min_lr: 0.007342  loss: 3.7745 (3.6081)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0010  max mem: 55573
Epoch: [71]  [140/156]  eta: 0:00:10  lr: 0.007341  min_lr: 0.007341  loss: 3.7977 (3.6190)  weight_decay: 0.0500 (0.0500)  time: 0.5863  data: 0.0009  max mem: 55573
Epoch: [71]  [150/156]  eta: 0:00:04  lr: 0.007339  min_lr: 0.007339  loss: 3.6708 (3.6113)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [71]  [155/156]  eta: 0:00:00  lr: 0.007339  min_lr: 0.007339  loss: 3.8617 (3.6232)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [71] Total time: 0:01:46 (0.6812 s / it)
Averaged stats: lr: 0.007339  min_lr: 0.007339  loss: 3.8617 (3.6101)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.5641 (1.5641)  acc1: 72.9167 (72.9167)  acc5: 91.4062 (91.4062)  time: 7.0367  data: 6.7962  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5641 (1.6381)  acc1: 71.0938 (68.0640)  acc5: 91.4062 (88.9920)  time: 1.5527  data: 1.3593  max mem: 55573
Test: Total time: 0:00:08 (1.6014 s / it)
* Acc@1 69.288 Acc@5 89.550 loss 1.610
Accuracy of the model on the 50000 test images: 69.3%
Max accuracy: 70.16%
Test:  [0/5]  eta: 0:00:36  loss: 7.1715 (7.1715)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.2719  data: 7.0360  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.4349 (7.4035)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.4800)  time: 1.5997  data: 1.4073  max mem: 55573
Test: Total time: 0:00:08 (1.6176 s / it)
* Acc@1 0.100 Acc@5 0.514 loss 7.314
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [72]  [  0/156]  eta: 0:37:53  lr: 0.007338  min_lr: 0.007338  loss: 3.9429 (3.9429)  weight_decay: 0.0500 (0.0500)  time: 14.5738  data: 10.0827  max mem: 55573
Epoch: [72]  [ 10/156]  eta: 0:04:32  lr: 0.007337  min_lr: 0.007337  loss: 3.6694 (3.4890)  weight_decay: 0.0500 (0.0500)  time: 1.8632  data: 0.9170  max mem: 55573
Epoch: [72]  [ 20/156]  eta: 0:02:50  lr: 0.007335  min_lr: 0.007335  loss: 3.4105 (3.4939)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [72]  [ 30/156]  eta: 0:02:11  lr: 0.007334  min_lr: 0.007334  loss: 3.5831 (3.4913)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [72]  [ 40/156]  eta: 0:01:48  lr: 0.007332  min_lr: 0.007332  loss: 3.6773 (3.5289)  weight_decay: 0.0500 (0.0500)  time: 0.5939  data: 0.0005  max mem: 55573
Epoch: [72]  [ 50/156]  eta: 0:01:31  lr: 0.007330  min_lr: 0.007330  loss: 3.6025 (3.4939)  weight_decay: 0.0500 (0.0500)  time: 0.5931  data: 0.0004  max mem: 55573
Epoch: [72]  [ 60/156]  eta: 0:01:18  lr: 0.007329  min_lr: 0.007329  loss: 3.6729 (3.5421)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [72]  [ 70/156]  eta: 0:01:07  lr: 0.007327  min_lr: 0.007327  loss: 3.8317 (3.5678)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [72]  [ 80/156]  eta: 0:00:58  lr: 0.007326  min_lr: 0.007326  loss: 3.8079 (3.5795)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [72]  [ 90/156]  eta: 0:00:49  lr: 0.007324  min_lr: 0.007324  loss: 3.7313 (3.5868)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [72]  [100/156]  eta: 0:00:40  lr: 0.007322  min_lr: 0.007322  loss: 3.6661 (3.5895)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [72]  [110/156]  eta: 0:00:32  lr: 0.007321  min_lr: 0.007321  loss: 3.8003 (3.6130)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [72]  [120/156]  eta: 0:00:25  lr: 0.007319  min_lr: 0.007319  loss: 3.8706 (3.6334)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [72]  [130/156]  eta: 0:00:18  lr: 0.007318  min_lr: 0.007318  loss: 3.8425 (3.6292)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0021  max mem: 55573
Epoch: [72]  [140/156]  eta: 0:00:11  lr: 0.007316  min_lr: 0.007316  loss: 3.8425 (3.6357)  weight_decay: 0.0500 (0.0500)  time: 0.5858  data: 0.0020  max mem: 55573
Epoch: [72]  [150/156]  eta: 0:00:04  lr: 0.007314  min_lr: 0.007314  loss: 3.8448 (3.6357)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [72]  [155/156]  eta: 0:00:00  lr: 0.007314  min_lr: 0.007314  loss: 3.6331 (3.6335)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [72] Total time: 0:01:47 (0.6861 s / it)
Averaged stats: lr: 0.007314  min_lr: 0.007314  loss: 3.6331 (3.6244)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.2095 (1.2095)  acc1: 76.0417 (76.0417)  acc5: 93.6198 (93.6198)  time: 7.1556  data: 6.9182  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2989 (1.3752)  acc1: 72.2656 (70.4640)  acc5: 93.0990 (90.6240)  time: 1.5776  data: 1.3837  max mem: 55573
Test: Total time: 0:00:08 (1.6237 s / it)
* Acc@1 70.368 Acc@5 90.410 loss 1.376
Accuracy of the model on the 50000 test images: 70.4%
Max accuracy: 70.37%
Test:  [0/5]  eta: 0:00:34  loss: 7.1984 (7.1984)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.8916  data: 6.6555  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.4578 (7.4325)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.5120)  time: 1.5236  data: 1.3312  max mem: 55573
Test: Total time: 0:00:07 (1.5441 s / it)
* Acc@1 0.114 Acc@5 0.522 loss 7.342
Accuracy of the model EMA on 50000 test images: 0.1%
Max EMA accuracy: 0.11%
Epoch: [73]  [  0/156]  eta: 0:32:14  lr: 0.007313  min_lr: 0.007313  loss: 3.7661 (3.7661)  weight_decay: 0.0500 (0.0500)  time: 12.4000  data: 11.1616  max mem: 55573
Epoch: [73]  [ 10/156]  eta: 0:04:06  lr: 0.007312  min_lr: 0.007312  loss: 3.7661 (3.6184)  weight_decay: 0.0500 (0.0500)  time: 1.6896  data: 1.0151  max mem: 55573
Epoch: [73]  [ 20/156]  eta: 0:02:38  lr: 0.007310  min_lr: 0.007310  loss: 3.6157 (3.5254)  weight_decay: 0.0500 (0.0500)  time: 0.6053  data: 0.0004  max mem: 55573
Epoch: [73]  [ 30/156]  eta: 0:02:03  lr: 0.007309  min_lr: 0.007309  loss: 3.8138 (3.6614)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [73]  [ 40/156]  eta: 0:01:42  lr: 0.007307  min_lr: 0.007307  loss: 3.8138 (3.6243)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [73]  [ 50/156]  eta: 0:01:27  lr: 0.007305  min_lr: 0.007305  loss: 3.7385 (3.6529)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [73]  [ 60/156]  eta: 0:01:15  lr: 0.007304  min_lr: 0.007304  loss: 3.7737 (3.6417)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [73]  [ 70/156]  eta: 0:01:05  lr: 0.007302  min_lr: 0.007302  loss: 3.8307 (3.6504)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [73]  [ 80/156]  eta: 0:00:56  lr: 0.007300  min_lr: 0.007300  loss: 3.8164 (3.6637)  weight_decay: 0.0500 (0.0500)  time: 0.5885  data: 0.0004  max mem: 55573
Epoch: [73]  [ 90/156]  eta: 0:00:47  lr: 0.007299  min_lr: 0.007299  loss: 3.7101 (3.6507)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [73]  [100/156]  eta: 0:00:39  lr: 0.007297  min_lr: 0.007297  loss: 3.9606 (3.6804)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [73]  [110/156]  eta: 0:00:32  lr: 0.007296  min_lr: 0.007296  loss: 3.9916 (3.7047)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0004  max mem: 55573
Epoch: [73]  [120/156]  eta: 0:00:24  lr: 0.007294  min_lr: 0.007294  loss: 3.9677 (3.7217)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0004  max mem: 55573
Epoch: [73]  [130/156]  eta: 0:00:17  lr: 0.007292  min_lr: 0.007292  loss: 3.8582 (3.7159)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0010  max mem: 55573
Epoch: [73]  [140/156]  eta: 0:00:10  lr: 0.007291  min_lr: 0.007291  loss: 3.5597 (3.6933)  weight_decay: 0.0500 (0.0500)  time: 0.5855  data: 0.0009  max mem: 55573
Epoch: [73]  [150/156]  eta: 0:00:04  lr: 0.007289  min_lr: 0.007289  loss: 3.5597 (3.6868)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [73]  [155/156]  eta: 0:00:00  lr: 0.007288  min_lr: 0.007288  loss: 3.5597 (3.6838)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [73] Total time: 0:01:44 (0.6730 s / it)
Averaged stats: lr: 0.007288  min_lr: 0.007288  loss: 3.5597 (3.6356)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.2743 (1.2743)  acc1: 75.2604 (75.2604)  acc5: 94.0104 (94.0104)  time: 6.8011  data: 6.5636  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3073 (1.3654)  acc1: 73.1771 (70.6240)  acc5: 93.8802 (90.8480)  time: 1.5067  data: 1.3128  max mem: 55573
Test: Total time: 0:00:07 (1.5455 s / it)
* Acc@1 70.196 Acc@5 90.460 loss 1.376
Accuracy of the model on the 50000 test images: 70.2%
Max accuracy: 70.37%
Test:  [0/5]  eta: 0:00:35  loss: 7.2225 (7.2225)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.1891  data: 6.9531  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.4778 (7.4600)  acc1: 0.0000 (0.1600)  acc5: 0.0000 (0.5440)  time: 1.5831  data: 1.3907  max mem: 55573
Test: Total time: 0:00:07 (1.5993 s / it)
* Acc@1 0.110 Acc@5 0.524 loss 7.369
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [74]  [  0/156]  eta: 0:38:19  lr: 0.007288  min_lr: 0.007288  loss: 4.3539 (4.3539)  weight_decay: 0.0500 (0.0500)  time: 14.7412  data: 9.8608  max mem: 55573
Epoch: [74]  [ 10/156]  eta: 0:04:37  lr: 0.007286  min_lr: 0.007286  loss: 3.9037 (3.8512)  weight_decay: 0.0500 (0.0500)  time: 1.9011  data: 0.8969  max mem: 55573
Epoch: [74]  [ 20/156]  eta: 0:02:53  lr: 0.007285  min_lr: 0.007285  loss: 3.6509 (3.6553)  weight_decay: 0.0500 (0.0500)  time: 0.6041  data: 0.0004  max mem: 55573
Epoch: [74]  [ 30/156]  eta: 0:02:13  lr: 0.007283  min_lr: 0.007283  loss: 3.4550 (3.6216)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [74]  [ 40/156]  eta: 0:01:49  lr: 0.007282  min_lr: 0.007282  loss: 3.4550 (3.5391)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [74]  [ 50/156]  eta: 0:01:32  lr: 0.007280  min_lr: 0.007280  loss: 3.5199 (3.5373)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [74]  [ 60/156]  eta: 0:01:19  lr: 0.007278  min_lr: 0.007278  loss: 3.6701 (3.5641)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [74]  [ 70/156]  eta: 0:01:08  lr: 0.007277  min_lr: 0.007277  loss: 3.7158 (3.5687)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0004  max mem: 55573
Epoch: [74]  [ 80/156]  eta: 0:00:58  lr: 0.007275  min_lr: 0.007275  loss: 3.4557 (3.5432)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [74]  [ 90/156]  eta: 0:00:49  lr: 0.007273  min_lr: 0.007273  loss: 3.4374 (3.5478)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0005  max mem: 55573
Epoch: [74]  [100/156]  eta: 0:00:41  lr: 0.007272  min_lr: 0.007272  loss: 3.6476 (3.5615)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [74]  [110/156]  eta: 0:00:33  lr: 0.007270  min_lr: 0.007270  loss: 3.8412 (3.5546)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [74]  [120/156]  eta: 0:00:25  lr: 0.007268  min_lr: 0.007268  loss: 3.8706 (3.5879)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0004  max mem: 55573
Epoch: [74]  [130/156]  eta: 0:00:18  lr: 0.007267  min_lr: 0.007267  loss: 3.6029 (3.5750)  weight_decay: 0.0500 (0.0500)  time: 0.5932  data: 0.0010  max mem: 55573
Epoch: [74]  [140/156]  eta: 0:00:11  lr: 0.007265  min_lr: 0.007265  loss: 3.4204 (3.5747)  weight_decay: 0.0500 (0.0500)  time: 0.5876  data: 0.0009  max mem: 55573
Epoch: [74]  [150/156]  eta: 0:00:04  lr: 0.007263  min_lr: 0.007263  loss: 3.7218 (3.5857)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [74]  [155/156]  eta: 0:00:00  lr: 0.007263  min_lr: 0.007263  loss: 3.6792 (3.5907)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [74] Total time: 0:01:47 (0.6890 s / it)
Averaged stats: lr: 0.007263  min_lr: 0.007263  loss: 3.6792 (3.6126)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.2545 (1.2545)  acc1: 73.4375 (73.4375)  acc5: 92.3177 (92.3177)  time: 7.1047  data: 6.8673  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3370 (1.4115)  acc1: 73.1771 (68.2560)  acc5: 92.3177 (89.9840)  time: 1.5676  data: 1.3736  max mem: 55573
Test: Total time: 0:00:08 (1.6092 s / it)
* Acc@1 69.432 Acc@5 89.966 loss 1.399
Accuracy of the model on the 50000 test images: 69.4%
Max accuracy: 70.37%
Test:  [0/5]  eta: 0:00:33  loss: 7.2387 (7.2387)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.6109  data: 6.3748  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.4961 (7.4810)  acc1: 0.0000 (0.1280)  acc5: 0.0000 (0.5760)  time: 1.4675  data: 1.2751  max mem: 55573
Test: Total time: 0:00:07 (1.4862 s / it)
* Acc@1 0.114 Acc@5 0.534 loss 7.391
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [75]  [  0/156]  eta: 0:36:45  lr: 0.007262  min_lr: 0.007262  loss: 3.9622 (3.9622)  weight_decay: 0.0500 (0.0500)  time: 14.1351  data: 10.3581  max mem: 55573
Epoch: [75]  [ 10/156]  eta: 0:04:30  lr: 0.007261  min_lr: 0.007261  loss: 3.8577 (3.6639)  weight_decay: 0.0500 (0.0500)  time: 1.8517  data: 0.9419  max mem: 55573
Epoch: [75]  [ 20/156]  eta: 0:02:50  lr: 0.007259  min_lr: 0.007259  loss: 3.4439 (3.4239)  weight_decay: 0.0500 (0.0500)  time: 0.6071  data: 0.0003  max mem: 55573
Epoch: [75]  [ 30/156]  eta: 0:02:10  lr: 0.007257  min_lr: 0.007257  loss: 3.2928 (3.4288)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [75]  [ 40/156]  eta: 0:01:47  lr: 0.007256  min_lr: 0.007256  loss: 3.5321 (3.4534)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [75]  [ 50/156]  eta: 0:01:31  lr: 0.007254  min_lr: 0.007254  loss: 3.6503 (3.4501)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [75]  [ 60/156]  eta: 0:01:18  lr: 0.007252  min_lr: 0.007252  loss: 3.6503 (3.4709)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [75]  [ 70/156]  eta: 0:01:07  lr: 0.007251  min_lr: 0.007251  loss: 3.6665 (3.4936)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [75]  [ 80/156]  eta: 0:00:57  lr: 0.007249  min_lr: 0.007249  loss: 3.7852 (3.5256)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [75]  [ 90/156]  eta: 0:00:49  lr: 0.007247  min_lr: 0.007247  loss: 3.6147 (3.5284)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0004  max mem: 55573
Epoch: [75]  [100/156]  eta: 0:00:40  lr: 0.007246  min_lr: 0.007246  loss: 3.5612 (3.5237)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0005  max mem: 55573
Epoch: [75]  [110/156]  eta: 0:00:32  lr: 0.007244  min_lr: 0.007244  loss: 3.5958 (3.5556)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0005  max mem: 55573
Epoch: [75]  [120/156]  eta: 0:00:25  lr: 0.007242  min_lr: 0.007242  loss: 3.9157 (3.5614)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [75]  [130/156]  eta: 0:00:18  lr: 0.007241  min_lr: 0.007241  loss: 3.6506 (3.5600)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0011  max mem: 55573
Epoch: [75]  [140/156]  eta: 0:00:11  lr: 0.007239  min_lr: 0.007239  loss: 3.5971 (3.5629)  weight_decay: 0.0500 (0.0500)  time: 0.5853  data: 0.0009  max mem: 55573
Epoch: [75]  [150/156]  eta: 0:00:04  lr: 0.007237  min_lr: 0.007237  loss: 3.8299 (3.5812)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [75]  [155/156]  eta: 0:00:00  lr: 0.007236  min_lr: 0.007236  loss: 3.8299 (3.5788)  weight_decay: 0.0500 (0.0500)  time: 0.5811  data: 0.0001  max mem: 55573
Epoch: [75] Total time: 0:01:46 (0.6844 s / it)
Averaged stats: lr: 0.007236  min_lr: 0.007236  loss: 3.8299 (3.6207)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.2819 (1.2819)  acc1: 75.2604 (75.2604)  acc5: 93.0990 (93.0990)  time: 6.7442  data: 6.5067  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4129 (1.3855)  acc1: 71.7448 (70.0480)  acc5: 91.6667 (90.1440)  time: 1.4943  data: 1.3015  max mem: 55573
Test: Total time: 0:00:07 (1.5123 s / it)
* Acc@1 70.660 Acc@5 90.268 loss 1.390
Accuracy of the model on the 50000 test images: 70.7%
Max accuracy: 70.66%
Test:  [0/5]  eta: 0:00:33  loss: 7.2599 (7.2599)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.7269  data: 6.4908  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5158 (7.5049)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.5760)  time: 1.4909  data: 1.2983  max mem: 55573
Test: Total time: 0:00:07 (1.5085 s / it)
* Acc@1 0.114 Acc@5 0.546 loss 7.415
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [76]  [  0/156]  eta: 0:37:33  lr: 0.007236  min_lr: 0.007236  loss: 2.8566 (2.8566)  weight_decay: 0.0500 (0.0500)  time: 14.4473  data: 8.4237  max mem: 55573
Epoch: [76]  [ 10/156]  eta: 0:04:32  lr: 0.007234  min_lr: 0.007234  loss: 3.4879 (3.3928)  weight_decay: 0.0500 (0.0500)  time: 1.8685  data: 0.7662  max mem: 55573
Epoch: [76]  [ 20/156]  eta: 0:02:51  lr: 0.007233  min_lr: 0.007233  loss: 3.6725 (3.5007)  weight_decay: 0.0500 (0.0500)  time: 0.6001  data: 0.0004  max mem: 55573
Epoch: [76]  [ 30/156]  eta: 0:02:11  lr: 0.007231  min_lr: 0.007231  loss: 3.7497 (3.4912)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [76]  [ 40/156]  eta: 0:01:48  lr: 0.007229  min_lr: 0.007229  loss: 3.8281 (3.5341)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [76]  [ 50/156]  eta: 0:01:31  lr: 0.007228  min_lr: 0.007228  loss: 3.7870 (3.5726)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0004  max mem: 55573
Epoch: [76]  [ 60/156]  eta: 0:01:18  lr: 0.007226  min_lr: 0.007226  loss: 3.7561 (3.5729)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [76]  [ 70/156]  eta: 0:01:07  lr: 0.007224  min_lr: 0.007224  loss: 3.7091 (3.5945)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [76]  [ 80/156]  eta: 0:00:58  lr: 0.007223  min_lr: 0.007223  loss: 3.4858 (3.5650)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [76]  [ 90/156]  eta: 0:00:49  lr: 0.007221  min_lr: 0.007221  loss: 3.7231 (3.5790)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [76]  [100/156]  eta: 0:00:40  lr: 0.007219  min_lr: 0.007219  loss: 3.7884 (3.5766)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0004  max mem: 55573
Epoch: [76]  [110/156]  eta: 0:00:33  lr: 0.007217  min_lr: 0.007217  loss: 3.8052 (3.5894)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [76]  [120/156]  eta: 0:00:25  lr: 0.007216  min_lr: 0.007216  loss: 3.8052 (3.5956)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [76]  [130/156]  eta: 0:00:18  lr: 0.007214  min_lr: 0.007214  loss: 3.5218 (3.5821)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0010  max mem: 55573
Epoch: [76]  [140/156]  eta: 0:00:11  lr: 0.007212  min_lr: 0.007212  loss: 3.5218 (3.5830)  weight_decay: 0.0500 (0.0500)  time: 0.5879  data: 0.0008  max mem: 55573
Epoch: [76]  [150/156]  eta: 0:00:04  lr: 0.007211  min_lr: 0.007211  loss: 3.8880 (3.5962)  weight_decay: 0.0500 (0.0500)  time: 0.5851  data: 0.0001  max mem: 55573
Epoch: [76]  [155/156]  eta: 0:00:00  lr: 0.007210  min_lr: 0.007210  loss: 3.8493 (3.5976)  weight_decay: 0.0500 (0.0500)  time: 0.5848  data: 0.0001  max mem: 55573
Epoch: [76] Total time: 0:01:47 (0.6864 s / it)
Averaged stats: lr: 0.007210  min_lr: 0.007210  loss: 3.8493 (3.6160)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.2860 (1.2860)  acc1: 73.6979 (73.6979)  acc5: 93.7500 (93.7500)  time: 7.0630  data: 6.8256  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3774 (1.4006)  acc1: 72.9167 (69.7920)  acc5: 90.5660 (90.7520)  time: 1.5590  data: 1.3652  max mem: 55573
Test: Total time: 0:00:08 (1.6009 s / it)
* Acc@1 70.708 Acc@5 90.448 loss 1.380
Accuracy of the model on the 50000 test images: 70.7%
Max accuracy: 70.71%
Test:  [0/5]  eta: 0:00:34  loss: 7.2767 (7.2767)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.8262  data: 6.5897  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5326 (7.5257)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.5760)  time: 1.5107  data: 1.3180  max mem: 55573
Test: Total time: 0:00:07 (1.5281 s / it)
* Acc@1 0.106 Acc@5 0.556 loss 7.438
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [77]  [  0/156]  eta: 0:37:05  lr: 0.007210  min_lr: 0.007210  loss: 3.3552 (3.3552)  weight_decay: 0.0500 (0.0500)  time: 14.2688  data: 13.6136  max mem: 55573
Epoch: [77]  [ 10/156]  eta: 0:04:29  lr: 0.007208  min_lr: 0.007208  loss: 3.4367 (3.4936)  weight_decay: 0.0500 (0.0500)  time: 1.8425  data: 1.2380  max mem: 55573
Epoch: [77]  [ 20/156]  eta: 0:02:49  lr: 0.007206  min_lr: 0.007206  loss: 3.6484 (3.5472)  weight_decay: 0.0500 (0.0500)  time: 0.5945  data: 0.0004  max mem: 55573
Epoch: [77]  [ 30/156]  eta: 0:02:10  lr: 0.007204  min_lr: 0.007204  loss: 3.7936 (3.6539)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [77]  [ 40/156]  eta: 0:01:47  lr: 0.007203  min_lr: 0.007203  loss: 3.8917 (3.6245)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [77]  [ 50/156]  eta: 0:01:31  lr: 0.007201  min_lr: 0.007201  loss: 3.6951 (3.6533)  weight_decay: 0.0500 (0.0500)  time: 0.5933  data: 0.0004  max mem: 55573
Epoch: [77]  [ 60/156]  eta: 0:01:18  lr: 0.007199  min_lr: 0.007199  loss: 3.7767 (3.6663)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [77]  [ 70/156]  eta: 0:01:07  lr: 0.007198  min_lr: 0.007198  loss: 3.7298 (3.6443)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [77]  [ 80/156]  eta: 0:00:57  lr: 0.007196  min_lr: 0.007196  loss: 3.4840 (3.6240)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [77]  [ 90/156]  eta: 0:00:48  lr: 0.007194  min_lr: 0.007194  loss: 3.3895 (3.5856)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [77]  [100/156]  eta: 0:00:40  lr: 0.007192  min_lr: 0.007192  loss: 3.4160 (3.5816)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [77]  [110/156]  eta: 0:00:32  lr: 0.007191  min_lr: 0.007191  loss: 3.5165 (3.5800)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [77]  [120/156]  eta: 0:00:25  lr: 0.007189  min_lr: 0.007189  loss: 3.5165 (3.5741)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [77]  [130/156]  eta: 0:00:18  lr: 0.007187  min_lr: 0.007187  loss: 3.4085 (3.5577)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0012  max mem: 55573
Epoch: [77]  [140/156]  eta: 0:00:11  lr: 0.007185  min_lr: 0.007185  loss: 3.3815 (3.5594)  weight_decay: 0.0500 (0.0500)  time: 0.5866  data: 0.0010  max mem: 55573
Epoch: [77]  [150/156]  eta: 0:00:04  lr: 0.007184  min_lr: 0.007184  loss: 3.6415 (3.5638)  weight_decay: 0.0500 (0.0500)  time: 0.5830  data: 0.0001  max mem: 55573
Epoch: [77]  [155/156]  eta: 0:00:00  lr: 0.007183  min_lr: 0.007183  loss: 3.7290 (3.5637)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0001  max mem: 55573
Epoch: [77] Total time: 0:01:46 (0.6844 s / it)
Averaged stats: lr: 0.007183  min_lr: 0.007183  loss: 3.7290 (3.5776)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.2571 (1.2571)  acc1: 75.6510 (75.6510)  acc5: 93.6198 (93.6198)  time: 7.1803  data: 6.9424  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3528 (1.3939)  acc1: 71.2240 (69.3440)  acc5: 91.7969 (89.9200)  time: 1.5825  data: 1.3886  max mem: 55573
Test: Total time: 0:00:08 (1.6222 s / it)
* Acc@1 70.202 Acc@5 90.226 loss 1.372
Accuracy of the model on the 50000 test images: 70.2%
Max accuracy: 70.71%
Test:  [0/5]  eta: 0:00:34  loss: 7.2941 (7.2941)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.8884  data: 6.6524  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5474 (7.5449)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.5760)  time: 1.5231  data: 1.3306  max mem: 55573
Test: Total time: 0:00:07 (1.5476 s / it)
* Acc@1 0.106 Acc@5 0.546 loss 7.459
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [78]  [  0/156]  eta: 0:35:24  lr: 0.007183  min_lr: 0.007183  loss: 3.9012 (3.9012)  weight_decay: 0.0500 (0.0500)  time: 13.6159  data: 10.2409  max mem: 55573
Epoch: [78]  [ 10/156]  eta: 0:04:20  lr: 0.007181  min_lr: 0.007181  loss: 3.2908 (3.4612)  weight_decay: 0.0500 (0.0500)  time: 1.7851  data: 0.9314  max mem: 55573
Epoch: [78]  [ 20/156]  eta: 0:02:45  lr: 0.007179  min_lr: 0.007179  loss: 3.2866 (3.3928)  weight_decay: 0.0500 (0.0500)  time: 0.5947  data: 0.0004  max mem: 55573
Epoch: [78]  [ 30/156]  eta: 0:02:07  lr: 0.007177  min_lr: 0.007177  loss: 3.2866 (3.4010)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [78]  [ 40/156]  eta: 0:01:45  lr: 0.007176  min_lr: 0.007176  loss: 3.3999 (3.4047)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [78]  [ 50/156]  eta: 0:01:29  lr: 0.007174  min_lr: 0.007174  loss: 3.6096 (3.4392)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0005  max mem: 55573
Epoch: [78]  [ 60/156]  eta: 0:01:17  lr: 0.007172  min_lr: 0.007172  loss: 3.4942 (3.4568)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [78]  [ 70/156]  eta: 0:01:06  lr: 0.007170  min_lr: 0.007170  loss: 3.6304 (3.4810)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0004  max mem: 55573
Epoch: [78]  [ 80/156]  eta: 0:00:57  lr: 0.007169  min_lr: 0.007169  loss: 3.7949 (3.4958)  weight_decay: 0.0500 (0.0500)  time: 0.5883  data: 0.0004  max mem: 55573
Epoch: [78]  [ 90/156]  eta: 0:00:48  lr: 0.007167  min_lr: 0.007167  loss: 3.6249 (3.5184)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0004  max mem: 55573
Epoch: [78]  [100/156]  eta: 0:00:40  lr: 0.007165  min_lr: 0.007165  loss: 3.8634 (3.5410)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0004  max mem: 55573
Epoch: [78]  [110/156]  eta: 0:00:32  lr: 0.007163  min_lr: 0.007163  loss: 3.5090 (3.5220)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [78]  [120/156]  eta: 0:00:25  lr: 0.007162  min_lr: 0.007162  loss: 3.5090 (3.5466)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [78]  [130/156]  eta: 0:00:17  lr: 0.007160  min_lr: 0.007160  loss: 3.7808 (3.5575)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0010  max mem: 55573
Epoch: [78]  [140/156]  eta: 0:00:10  lr: 0.007158  min_lr: 0.007158  loss: 3.6996 (3.5641)  weight_decay: 0.0500 (0.0500)  time: 0.5859  data: 0.0008  max mem: 55573
Epoch: [78]  [150/156]  eta: 0:00:04  lr: 0.007156  min_lr: 0.007156  loss: 3.6847 (3.5683)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [78]  [155/156]  eta: 0:00:00  lr: 0.007155  min_lr: 0.007155  loss: 3.6847 (3.5779)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0001  max mem: 55573
Epoch: [78] Total time: 0:01:46 (0.6799 s / it)
Averaged stats: lr: 0.007155  min_lr: 0.007155  loss: 3.6847 (3.6017)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:37  loss: 1.2768 (1.2768)  acc1: 73.9583 (73.9583)  acc5: 92.4479 (92.4479)  time: 7.4706  data: 7.2325  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3649 (1.3977)  acc1: 71.6146 (69.4080)  acc5: 90.8854 (88.8960)  time: 1.6408  data: 1.4466  max mem: 55573
Test: Total time: 0:00:08 (1.6847 s / it)
* Acc@1 69.934 Acc@5 89.812 loss 1.390
Accuracy of the model on the 50000 test images: 69.9%
Max accuracy: 70.71%
Test:  [0/5]  eta: 0:00:34  loss: 7.3109 (7.3109)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.8926  data: 6.6561  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5596 (7.5619)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.5760)  time: 1.5636  data: 1.3709  max mem: 55573
Test: Total time: 0:00:07 (1.5805 s / it)
* Acc@1 0.106 Acc@5 0.540 loss 7.479
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [79]  [  0/156]  eta: 0:33:49  lr: 0.007155  min_lr: 0.007155  loss: 3.8152 (3.8152)  weight_decay: 0.0500 (0.0500)  time: 13.0109  data: 8.5316  max mem: 55573
Epoch: [79]  [ 10/156]  eta: 0:04:14  lr: 0.007153  min_lr: 0.007153  loss: 3.7850 (3.3369)  weight_decay: 0.0500 (0.0500)  time: 1.7462  data: 0.7761  max mem: 55573
Epoch: [79]  [ 20/156]  eta: 0:02:42  lr: 0.007152  min_lr: 0.007152  loss: 3.8502 (3.5834)  weight_decay: 0.0500 (0.0500)  time: 0.6049  data: 0.0005  max mem: 55573
Epoch: [79]  [ 30/156]  eta: 0:02:06  lr: 0.007150  min_lr: 0.007150  loss: 3.6924 (3.5269)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [79]  [ 40/156]  eta: 0:01:44  lr: 0.007148  min_lr: 0.007148  loss: 3.3924 (3.4867)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [79]  [ 50/156]  eta: 0:01:28  lr: 0.007146  min_lr: 0.007146  loss: 3.2189 (3.4399)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [79]  [ 60/156]  eta: 0:01:16  lr: 0.007145  min_lr: 0.007145  loss: 3.5730 (3.4738)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0005  max mem: 55573
Epoch: [79]  [ 70/156]  eta: 0:01:06  lr: 0.007143  min_lr: 0.007143  loss: 3.7176 (3.4730)  weight_decay: 0.0500 (0.0500)  time: 0.5930  data: 0.0004  max mem: 55573
Epoch: [79]  [ 80/156]  eta: 0:00:56  lr: 0.007141  min_lr: 0.007141  loss: 3.5448 (3.4746)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0005  max mem: 55573
Epoch: [79]  [ 90/156]  eta: 0:00:48  lr: 0.007139  min_lr: 0.007139  loss: 3.6530 (3.4950)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [79]  [100/156]  eta: 0:00:40  lr: 0.007137  min_lr: 0.007137  loss: 3.6530 (3.4901)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [79]  [110/156]  eta: 0:00:32  lr: 0.007136  min_lr: 0.007136  loss: 3.2663 (3.4809)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [79]  [120/156]  eta: 0:00:25  lr: 0.007134  min_lr: 0.007134  loss: 3.4419 (3.4769)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [79]  [130/156]  eta: 0:00:17  lr: 0.007132  min_lr: 0.007132  loss: 3.4998 (3.4692)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0010  max mem: 55573
Epoch: [79]  [140/156]  eta: 0:00:10  lr: 0.007130  min_lr: 0.007130  loss: 3.8141 (3.4926)  weight_decay: 0.0500 (0.0500)  time: 0.5862  data: 0.0009  max mem: 55573
Epoch: [79]  [150/156]  eta: 0:00:04  lr: 0.007129  min_lr: 0.007129  loss: 3.7091 (3.4979)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [79]  [155/156]  eta: 0:00:00  lr: 0.007128  min_lr: 0.007128  loss: 3.6486 (3.5034)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [79] Total time: 0:01:45 (0.6775 s / it)
Averaged stats: lr: 0.007128  min_lr: 0.007128  loss: 3.6486 (3.5952)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.3429 (1.3429)  acc1: 75.2604 (75.2604)  acc5: 92.9688 (92.9688)  time: 7.0685  data: 6.8312  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4741 (1.5447)  acc1: 72.7865 (69.0240)  acc5: 92.3177 (89.4400)  time: 1.5600  data: 1.3663  max mem: 55573
Test: Total time: 0:00:08 (1.6088 s / it)
* Acc@1 69.834 Acc@5 89.776 loss 1.563
Accuracy of the model on the 50000 test images: 69.8%
Max accuracy: 70.71%
Test:  [0/5]  eta: 0:00:35  loss: 7.3227 (7.3227)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.1454  data: 6.9087  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5677 (7.5742)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.5760)  time: 1.5746  data: 1.3818  max mem: 55573
Test: Total time: 0:00:07 (1.5977 s / it)
* Acc@1 0.110 Acc@5 0.548 loss 7.494
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [80]  [  0/156]  eta: 0:35:06  lr: 0.007127  min_lr: 0.007127  loss: 3.3273 (3.3273)  weight_decay: 0.0500 (0.0500)  time: 13.5045  data: 12.0208  max mem: 55573
Epoch: [80]  [ 10/156]  eta: 0:04:19  lr: 0.007126  min_lr: 0.007126  loss: 3.5671 (3.4123)  weight_decay: 0.0500 (0.0500)  time: 1.7786  data: 1.0932  max mem: 55573
Epoch: [80]  [ 20/156]  eta: 0:02:44  lr: 0.007124  min_lr: 0.007124  loss: 3.5671 (3.5006)  weight_decay: 0.0500 (0.0500)  time: 0.5976  data: 0.0004  max mem: 55573
Epoch: [80]  [ 30/156]  eta: 0:02:07  lr: 0.007122  min_lr: 0.007122  loss: 3.7568 (3.5738)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [80]  [ 40/156]  eta: 0:01:45  lr: 0.007120  min_lr: 0.007120  loss: 3.7568 (3.6239)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [80]  [ 50/156]  eta: 0:01:29  lr: 0.007118  min_lr: 0.007118  loss: 3.7337 (3.6306)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [80]  [ 60/156]  eta: 0:01:17  lr: 0.007117  min_lr: 0.007117  loss: 3.6221 (3.6014)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [80]  [ 70/156]  eta: 0:01:06  lr: 0.007115  min_lr: 0.007115  loss: 3.5315 (3.6011)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [80]  [ 80/156]  eta: 0:00:57  lr: 0.007113  min_lr: 0.007113  loss: 3.6112 (3.5872)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [80]  [ 90/156]  eta: 0:00:48  lr: 0.007111  min_lr: 0.007111  loss: 3.5790 (3.5726)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [80]  [100/156]  eta: 0:00:40  lr: 0.007109  min_lr: 0.007109  loss: 3.6918 (3.5952)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [80]  [110/156]  eta: 0:00:32  lr: 0.007108  min_lr: 0.007108  loss: 3.9052 (3.6124)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [80]  [120/156]  eta: 0:00:25  lr: 0.007106  min_lr: 0.007106  loss: 3.6389 (3.6039)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [80]  [130/156]  eta: 0:00:17  lr: 0.007104  min_lr: 0.007104  loss: 3.4302 (3.5941)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0010  max mem: 55573
Epoch: [80]  [140/156]  eta: 0:00:10  lr: 0.007102  min_lr: 0.007102  loss: 3.8058 (3.5969)  weight_decay: 0.0500 (0.0500)  time: 0.5860  data: 0.0009  max mem: 55573
Epoch: [80]  [150/156]  eta: 0:00:04  lr: 0.007100  min_lr: 0.007100  loss: 3.8058 (3.5993)  weight_decay: 0.0500 (0.0500)  time: 0.5812  data: 0.0001  max mem: 55573
Epoch: [80]  [155/156]  eta: 0:00:00  lr: 0.007099  min_lr: 0.007099  loss: 3.7216 (3.5952)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [80] Total time: 0:01:45 (0.6792 s / it)
Averaged stats: lr: 0.007099  min_lr: 0.007099  loss: 3.7216 (3.5985)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.2343 (1.2343)  acc1: 73.5677 (73.5677)  acc5: 92.5781 (92.5781)  time: 7.1086  data: 6.8711  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3173 (1.4009)  acc1: 73.5677 (69.3120)  acc5: 92.5781 (89.7920)  time: 1.5681  data: 1.3743  max mem: 55573
Test: Total time: 0:00:08 (1.6098 s / it)
* Acc@1 70.106 Acc@5 89.996 loss 1.387
Accuracy of the model on the 50000 test images: 70.1%
Max accuracy: 70.71%
Test:  [0/5]  eta: 0:00:36  loss: 7.3325 (7.3325)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.3303  data: 7.0939  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5746 (7.5852)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.5760)  time: 1.6115  data: 1.4189  max mem: 55573
Test: Total time: 0:00:08 (1.6438 s / it)
* Acc@1 0.112 Acc@5 0.554 loss 7.508
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [81]  [  0/156]  eta: 0:39:25  lr: 0.007099  min_lr: 0.007099  loss: 2.8104 (2.8104)  weight_decay: 0.0500 (0.0500)  time: 15.1660  data: 7.8838  max mem: 55573
Epoch: [81]  [ 10/156]  eta: 0:04:40  lr: 0.007097  min_lr: 0.007097  loss: 3.8179 (3.5600)  weight_decay: 0.0500 (0.0500)  time: 1.9183  data: 0.7172  max mem: 55573
Epoch: [81]  [ 20/156]  eta: 0:02:54  lr: 0.007096  min_lr: 0.007096  loss: 3.8179 (3.6088)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [81]  [ 30/156]  eta: 0:02:13  lr: 0.007094  min_lr: 0.007094  loss: 3.6875 (3.6065)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [81]  [ 40/156]  eta: 0:01:49  lr: 0.007092  min_lr: 0.007092  loss: 3.7121 (3.6479)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [81]  [ 50/156]  eta: 0:01:32  lr: 0.007090  min_lr: 0.007090  loss: 3.7782 (3.6497)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [81]  [ 60/156]  eta: 0:01:19  lr: 0.007088  min_lr: 0.007088  loss: 3.7880 (3.6488)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [81]  [ 70/156]  eta: 0:01:08  lr: 0.007086  min_lr: 0.007086  loss: 3.8081 (3.6526)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [81]  [ 80/156]  eta: 0:00:58  lr: 0.007085  min_lr: 0.007085  loss: 3.7272 (3.6563)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [81]  [ 90/156]  eta: 0:00:49  lr: 0.007083  min_lr: 0.007083  loss: 3.8729 (3.6679)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0005  max mem: 55573
Epoch: [81]  [100/156]  eta: 0:00:41  lr: 0.007081  min_lr: 0.007081  loss: 3.8519 (3.6537)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [81]  [110/156]  eta: 0:00:33  lr: 0.007079  min_lr: 0.007079  loss: 3.4186 (3.6345)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [81]  [120/156]  eta: 0:00:25  lr: 0.007077  min_lr: 0.007077  loss: 3.6972 (3.6351)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [81]  [130/156]  eta: 0:00:18  lr: 0.007075  min_lr: 0.007075  loss: 3.7662 (3.6429)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0026  max mem: 55573
Epoch: [81]  [140/156]  eta: 0:00:11  lr: 0.007074  min_lr: 0.007074  loss: 3.8053 (3.6346)  weight_decay: 0.0500 (0.0500)  time: 0.5869  data: 0.0025  max mem: 55573
Epoch: [81]  [150/156]  eta: 0:00:04  lr: 0.007072  min_lr: 0.007072  loss: 3.3396 (3.6231)  weight_decay: 0.0500 (0.0500)  time: 0.5831  data: 0.0001  max mem: 55573
Epoch: [81]  [155/156]  eta: 0:00:00  lr: 0.007071  min_lr: 0.007071  loss: 3.3396 (3.6184)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [81] Total time: 0:01:47 (0.6899 s / it)
Averaged stats: lr: 0.007071  min_lr: 0.007071  loss: 3.3396 (3.5925)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.2966 (1.2966)  acc1: 76.3021 (76.3021)  acc5: 93.3594 (93.3594)  time: 6.9791  data: 6.7422  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3943 (1.4441)  acc1: 71.3542 (69.4720)  acc5: 92.1875 (90.2080)  time: 1.5422  data: 1.3485  max mem: 55573
Test: Total time: 0:00:07 (1.5838 s / it)
* Acc@1 70.110 Acc@5 90.102 loss 1.435
Accuracy of the model on the 50000 test images: 70.1%
Max accuracy: 70.71%
Test:  [0/5]  eta: 0:00:35  loss: 7.3442 (7.3442)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.1132  data: 6.8773  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5828 (7.5981)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.5760)  time: 1.5679  data: 1.3755  max mem: 55573
Test: Total time: 0:00:07 (1.5864 s / it)
* Acc@1 0.106 Acc@5 0.558 loss 7.524
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [82]  [  0/156]  eta: 0:39:56  lr: 0.007071  min_lr: 0.007071  loss: 2.9143 (2.9143)  weight_decay: 0.0500 (0.0500)  time: 15.3601  data: 12.1954  max mem: 55573
Epoch: [82]  [ 10/156]  eta: 0:04:41  lr: 0.007069  min_lr: 0.007069  loss: 3.0157 (3.3353)  weight_decay: 0.0500 (0.0500)  time: 1.9307  data: 1.1090  max mem: 55573
Epoch: [82]  [ 20/156]  eta: 0:02:55  lr: 0.007067  min_lr: 0.007067  loss: 3.6908 (3.5114)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0004  max mem: 55573
Epoch: [82]  [ 30/156]  eta: 0:02:14  lr: 0.007065  min_lr: 0.007065  loss: 3.8300 (3.5313)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [82]  [ 40/156]  eta: 0:01:50  lr: 0.007063  min_lr: 0.007063  loss: 3.7589 (3.5957)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [82]  [ 50/156]  eta: 0:01:33  lr: 0.007061  min_lr: 0.007061  loss: 3.7031 (3.5564)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [82]  [ 60/156]  eta: 0:01:19  lr: 0.007060  min_lr: 0.007060  loss: 3.6339 (3.5997)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [82]  [ 70/156]  eta: 0:01:08  lr: 0.007058  min_lr: 0.007058  loss: 3.4668 (3.5536)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [82]  [ 80/156]  eta: 0:00:58  lr: 0.007056  min_lr: 0.007056  loss: 3.5331 (3.5845)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [82]  [ 90/156]  eta: 0:00:49  lr: 0.007054  min_lr: 0.007054  loss: 3.7231 (3.5876)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [82]  [100/156]  eta: 0:00:41  lr: 0.007052  min_lr: 0.007052  loss: 3.8312 (3.5913)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0005  max mem: 55573
Epoch: [82]  [110/156]  eta: 0:00:33  lr: 0.007050  min_lr: 0.007050  loss: 3.7936 (3.6040)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [82]  [120/156]  eta: 0:00:25  lr: 0.007048  min_lr: 0.007048  loss: 3.7936 (3.6148)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [82]  [130/156]  eta: 0:00:18  lr: 0.007047  min_lr: 0.007047  loss: 3.8286 (3.6306)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0010  max mem: 55573
Epoch: [82]  [140/156]  eta: 0:00:11  lr: 0.007045  min_lr: 0.007045  loss: 3.5965 (3.6115)  weight_decay: 0.0500 (0.0500)  time: 0.5867  data: 0.0009  max mem: 55573
Epoch: [82]  [150/156]  eta: 0:00:04  lr: 0.007043  min_lr: 0.007043  loss: 3.4767 (3.6161)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [82]  [155/156]  eta: 0:00:00  lr: 0.007042  min_lr: 0.007042  loss: 3.5698 (3.6154)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [82] Total time: 0:01:47 (0.6904 s / it)
Averaged stats: lr: 0.007042  min_lr: 0.007042  loss: 3.5698 (3.5701)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.3095 (1.3095)  acc1: 75.6510 (75.6510)  acc5: 93.3594 (93.3594)  time: 6.8435  data: 6.6062  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3980 (1.4860)  acc1: 72.9167 (69.6640)  acc5: 90.5660 (89.3440)  time: 1.5140  data: 1.3213  max mem: 55573
Test: Total time: 0:00:07 (1.5565 s / it)
* Acc@1 70.558 Acc@5 90.048 loss 1.475
Accuracy of the model on the 50000 test images: 70.6%
Max accuracy: 70.71%
Test:  [0/5]  eta: 0:00:37  loss: 7.3551 (7.3551)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.4588  data: 7.2223  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5864 (7.6086)  acc1: 0.0000 (0.1280)  acc5: 0.0000 (0.5120)  time: 1.6374  data: 1.4446  max mem: 55573
Test: Total time: 0:00:08 (1.6543 s / it)
* Acc@1 0.108 Acc@5 0.560 loss 7.538
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [83]  [  0/156]  eta: 0:35:04  lr: 0.007042  min_lr: 0.007042  loss: 2.4782 (2.4782)  weight_decay: 0.0500 (0.0500)  time: 13.4921  data: 10.3025  max mem: 55573
Epoch: [83]  [ 10/156]  eta: 0:04:21  lr: 0.007040  min_lr: 0.007040  loss: 3.6784 (3.5960)  weight_decay: 0.0500 (0.0500)  time: 1.7884  data: 0.9371  max mem: 55573
Epoch: [83]  [ 20/156]  eta: 0:02:45  lr: 0.007038  min_lr: 0.007038  loss: 3.7208 (3.6701)  weight_decay: 0.0500 (0.0500)  time: 0.6041  data: 0.0005  max mem: 55573
Epoch: [83]  [ 30/156]  eta: 0:02:07  lr: 0.007036  min_lr: 0.007036  loss: 3.7342 (3.6105)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [83]  [ 40/156]  eta: 0:01:45  lr: 0.007034  min_lr: 0.007034  loss: 3.3953 (3.5200)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [83]  [ 50/156]  eta: 0:01:29  lr: 0.007032  min_lr: 0.007032  loss: 3.3452 (3.5051)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [83]  [ 60/156]  eta: 0:01:17  lr: 0.007031  min_lr: 0.007031  loss: 3.6523 (3.5273)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [83]  [ 70/156]  eta: 0:01:06  lr: 0.007029  min_lr: 0.007029  loss: 3.8700 (3.5759)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [83]  [ 80/156]  eta: 0:00:57  lr: 0.007027  min_lr: 0.007027  loss: 3.8541 (3.6059)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [83]  [ 90/156]  eta: 0:00:48  lr: 0.007025  min_lr: 0.007025  loss: 3.7959 (3.6003)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [83]  [100/156]  eta: 0:00:40  lr: 0.007023  min_lr: 0.007023  loss: 3.8262 (3.6193)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [83]  [110/156]  eta: 0:00:32  lr: 0.007021  min_lr: 0.007021  loss: 3.8033 (3.6150)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [83]  [120/156]  eta: 0:00:25  lr: 0.007019  min_lr: 0.007019  loss: 3.5518 (3.5993)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [83]  [130/156]  eta: 0:00:17  lr: 0.007017  min_lr: 0.007017  loss: 3.3693 (3.5752)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0010  max mem: 55573
Epoch: [83]  [140/156]  eta: 0:00:10  lr: 0.007015  min_lr: 0.007015  loss: 3.7645 (3.5842)  weight_decay: 0.0500 (0.0500)  time: 0.5852  data: 0.0009  max mem: 55573
Epoch: [83]  [150/156]  eta: 0:00:04  lr: 0.007014  min_lr: 0.007014  loss: 3.7934 (3.5840)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [83]  [155/156]  eta: 0:00:00  lr: 0.007013  min_lr: 0.007013  loss: 3.8201 (3.6029)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [83] Total time: 0:01:46 (0.6797 s / it)
Averaged stats: lr: 0.007013  min_lr: 0.007013  loss: 3.8201 (3.5816)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.3522 (1.3522)  acc1: 75.5208 (75.5208)  acc5: 93.3594 (93.3594)  time: 7.0476  data: 6.8102  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5040 (1.5640)  acc1: 69.7917 (68.7360)  acc5: 91.1458 (89.3120)  time: 1.5559  data: 1.3621  max mem: 55573
Test: Total time: 0:00:08 (1.6007 s / it)
* Acc@1 70.186 Acc@5 89.770 loss 1.548
Accuracy of the model on the 50000 test images: 70.2%
Max accuracy: 70.71%
Test:  [0/5]  eta: 0:00:36  loss: 7.3696 (7.3696)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.2607  data: 7.0247  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5918 (7.6214)  acc1: 0.0000 (0.1280)  acc5: 0.0000 (0.5120)  time: 1.5975  data: 1.4050  max mem: 55573
Test: Total time: 0:00:08 (1.6160 s / it)
* Acc@1 0.112 Acc@5 0.554 loss 7.555
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [84]  [  0/156]  eta: 0:35:36  lr: 0.007012  min_lr: 0.007012  loss: 4.2180 (4.2180)  weight_decay: 0.0500 (0.0500)  time: 13.6953  data: 9.7590  max mem: 55573
Epoch: [84]  [ 10/156]  eta: 0:04:21  lr: 0.007011  min_lr: 0.007011  loss: 3.4610 (3.4506)  weight_decay: 0.0500 (0.0500)  time: 1.7890  data: 0.8876  max mem: 55573
Epoch: [84]  [ 20/156]  eta: 0:02:45  lr: 0.007009  min_lr: 0.007009  loss: 3.4610 (3.5247)  weight_decay: 0.0500 (0.0500)  time: 0.5944  data: 0.0004  max mem: 55573
Epoch: [84]  [ 30/156]  eta: 0:02:07  lr: 0.007007  min_lr: 0.007007  loss: 3.6279 (3.5073)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [84]  [ 40/156]  eta: 0:01:45  lr: 0.007005  min_lr: 0.007005  loss: 3.6994 (3.5386)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [84]  [ 50/156]  eta: 0:01:29  lr: 0.007003  min_lr: 0.007003  loss: 3.8320 (3.6145)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [84]  [ 60/156]  eta: 0:01:17  lr: 0.007001  min_lr: 0.007001  loss: 3.6832 (3.5677)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0005  max mem: 55573
Epoch: [84]  [ 70/156]  eta: 0:01:06  lr: 0.006999  min_lr: 0.006999  loss: 3.4123 (3.5602)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [84]  [ 80/156]  eta: 0:00:57  lr: 0.006997  min_lr: 0.006997  loss: 3.7485 (3.5840)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [84]  [ 90/156]  eta: 0:00:48  lr: 0.006995  min_lr: 0.006995  loss: 3.5918 (3.5630)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [84]  [100/156]  eta: 0:00:40  lr: 0.006993  min_lr: 0.006993  loss: 3.4900 (3.5590)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [84]  [110/156]  eta: 0:00:32  lr: 0.006991  min_lr: 0.006991  loss: 3.7950 (3.6001)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [84]  [120/156]  eta: 0:00:25  lr: 0.006990  min_lr: 0.006990  loss: 3.7787 (3.6057)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0004  max mem: 55573
Epoch: [84]  [130/156]  eta: 0:00:17  lr: 0.006988  min_lr: 0.006988  loss: 3.6092 (3.6053)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0010  max mem: 55573
Epoch: [84]  [140/156]  eta: 0:00:10  lr: 0.006986  min_lr: 0.006986  loss: 3.3904 (3.5785)  weight_decay: 0.0500 (0.0500)  time: 0.5873  data: 0.0009  max mem: 55573
Epoch: [84]  [150/156]  eta: 0:00:04  lr: 0.006984  min_lr: 0.006984  loss: 3.4917 (3.5805)  weight_decay: 0.0500 (0.0500)  time: 0.5828  data: 0.0001  max mem: 55573
Epoch: [84]  [155/156]  eta: 0:00:00  lr: 0.006983  min_lr: 0.006983  loss: 3.4917 (3.5750)  weight_decay: 0.0500 (0.0500)  time: 0.5831  data: 0.0001  max mem: 55573
Epoch: [84] Total time: 0:01:46 (0.6808 s / it)
Averaged stats: lr: 0.006983  min_lr: 0.006983  loss: 3.4917 (3.5881)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.2621 (1.2621)  acc1: 72.7865 (72.7865)  acc5: 92.8385 (92.8385)  time: 7.0437  data: 6.8033  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2990 (1.3808)  acc1: 70.5729 (68.2560)  acc5: 90.5660 (89.4720)  time: 1.5551  data: 1.3608  max mem: 55573
Test: Total time: 0:00:07 (1.5889 s / it)
* Acc@1 69.326 Acc@5 89.684 loss 1.347
Accuracy of the model on the 50000 test images: 69.3%
Max accuracy: 70.71%
Test:  [0/5]  eta: 0:00:37  loss: 7.3839 (7.3839)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.4826  data: 7.2465  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5968 (7.6310)  acc1: 0.0000 (0.0960)  acc5: 0.0000 (0.5120)  time: 1.6419  data: 1.4494  max mem: 55573
Test: Total time: 0:00:08 (1.6638 s / it)
* Acc@1 0.120 Acc@5 0.552 loss 7.567
Accuracy of the model EMA on 50000 test images: 0.1%
Max EMA accuracy: 0.12%
Epoch: [85]  [  0/156]  eta: 0:35:22  lr: 0.006983  min_lr: 0.006983  loss: 3.4805 (3.4805)  weight_decay: 0.0500 (0.0500)  time: 13.6059  data: 13.0191  max mem: 55573
Epoch: [85]  [ 10/156]  eta: 0:04:19  lr: 0.006981  min_lr: 0.006981  loss: 3.3862 (3.4345)  weight_decay: 0.0500 (0.0500)  time: 1.7799  data: 1.1841  max mem: 55573
Epoch: [85]  [ 20/156]  eta: 0:02:45  lr: 0.006979  min_lr: 0.006979  loss: 3.3989 (3.4581)  weight_decay: 0.0500 (0.0500)  time: 0.5944  data: 0.0006  max mem: 55573
Epoch: [85]  [ 30/156]  eta: 0:02:07  lr: 0.006977  min_lr: 0.006977  loss: 3.6327 (3.5587)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0007  max mem: 55573
Epoch: [85]  [ 40/156]  eta: 0:01:45  lr: 0.006975  min_lr: 0.006975  loss: 3.6717 (3.5392)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0007  max mem: 55573
Epoch: [85]  [ 50/156]  eta: 0:01:29  lr: 0.006973  min_lr: 0.006973  loss: 3.4184 (3.4487)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0007  max mem: 55573
Epoch: [85]  [ 60/156]  eta: 0:01:17  lr: 0.006971  min_lr: 0.006971  loss: 3.1355 (3.4553)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0006  max mem: 55573
Epoch: [85]  [ 70/156]  eta: 0:01:06  lr: 0.006969  min_lr: 0.006969  loss: 3.6218 (3.4762)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0006  max mem: 55573
Epoch: [85]  [ 80/156]  eta: 0:00:57  lr: 0.006967  min_lr: 0.006967  loss: 3.6218 (3.4882)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0006  max mem: 55573
Epoch: [85]  [ 90/156]  eta: 0:00:48  lr: 0.006965  min_lr: 0.006965  loss: 3.6279 (3.5031)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [85]  [100/156]  eta: 0:00:40  lr: 0.006963  min_lr: 0.006963  loss: 3.6279 (3.5078)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0006  max mem: 55573
Epoch: [85]  [110/156]  eta: 0:00:32  lr: 0.006962  min_lr: 0.006962  loss: 3.8645 (3.5342)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0007  max mem: 55573
Epoch: [85]  [120/156]  eta: 0:00:25  lr: 0.006960  min_lr: 0.006960  loss: 3.8981 (3.5638)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0006  max mem: 55573
Epoch: [85]  [130/156]  eta: 0:00:17  lr: 0.006958  min_lr: 0.006958  loss: 3.6927 (3.5602)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0012  max mem: 55573
Epoch: [85]  [140/156]  eta: 0:00:10  lr: 0.006956  min_lr: 0.006956  loss: 3.2311 (3.5311)  weight_decay: 0.0500 (0.0500)  time: 0.5855  data: 0.0010  max mem: 55573
Epoch: [85]  [150/156]  eta: 0:00:04  lr: 0.006954  min_lr: 0.006954  loss: 2.9387 (3.5027)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0002  max mem: 55573
Epoch: [85]  [155/156]  eta: 0:00:00  lr: 0.006953  min_lr: 0.006953  loss: 3.2304 (3.5141)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0002  max mem: 55573
Epoch: [85] Total time: 0:01:45 (0.6793 s / it)
Averaged stats: lr: 0.006953  min_lr: 0.006953  loss: 3.2304 (3.5832)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.3300 (1.3300)  acc1: 74.3490 (74.3490)  acc5: 91.9271 (91.9271)  time: 6.9800  data: 6.7419  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3694 (1.4101)  acc1: 72.1354 (70.2720)  acc5: 91.9271 (89.9840)  time: 1.5427  data: 1.3485  max mem: 55573
Test: Total time: 0:00:07 (1.5863 s / it)
* Acc@1 70.622 Acc@5 90.216 loss 1.402
Accuracy of the model on the 50000 test images: 70.6%
Max accuracy: 70.71%
Test:  [0/5]  eta: 0:00:37  loss: 7.3914 (7.3914)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.4768  data: 7.2408  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5949 (7.6333)  acc1: 0.0000 (0.1280)  acc5: 0.0000 (0.5120)  time: 1.6407  data: 1.4483  max mem: 55573
Test: Total time: 0:00:08 (1.6608 s / it)
* Acc@1 0.120 Acc@5 0.556 loss 7.572
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [86]  [  0/156]  eta: 0:34:21  lr: 0.006953  min_lr: 0.006953  loss: 3.7692 (3.7692)  weight_decay: 0.0500 (0.0500)  time: 13.2171  data: 9.4016  max mem: 55573
Epoch: [86]  [ 10/156]  eta: 0:04:17  lr: 0.006951  min_lr: 0.006951  loss: 3.5025 (3.4605)  weight_decay: 0.0500 (0.0500)  time: 1.7634  data: 0.8554  max mem: 55573
Epoch: [86]  [ 20/156]  eta: 0:02:43  lr: 0.006949  min_lr: 0.006949  loss: 3.6880 (3.6719)  weight_decay: 0.0500 (0.0500)  time: 0.6038  data: 0.0007  max mem: 55573
Epoch: [86]  [ 30/156]  eta: 0:02:06  lr: 0.006947  min_lr: 0.006947  loss: 3.8077 (3.7063)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0006  max mem: 55573
Epoch: [86]  [ 40/156]  eta: 0:01:44  lr: 0.006945  min_lr: 0.006945  loss: 3.5712 (3.6630)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0007  max mem: 55573
Epoch: [86]  [ 50/156]  eta: 0:01:29  lr: 0.006943  min_lr: 0.006943  loss: 3.8027 (3.6763)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0007  max mem: 55573
Epoch: [86]  [ 60/156]  eta: 0:01:17  lr: 0.006941  min_lr: 0.006941  loss: 3.7202 (3.6410)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0006  max mem: 55573
Epoch: [86]  [ 70/156]  eta: 0:01:06  lr: 0.006939  min_lr: 0.006939  loss: 3.3164 (3.6057)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0008  max mem: 55573
Epoch: [86]  [ 80/156]  eta: 0:00:57  lr: 0.006937  min_lr: 0.006937  loss: 3.6195 (3.6217)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0008  max mem: 55573
Epoch: [86]  [ 90/156]  eta: 0:00:48  lr: 0.006935  min_lr: 0.006935  loss: 3.7546 (3.6137)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0007  max mem: 55573
Epoch: [86]  [100/156]  eta: 0:00:40  lr: 0.006933  min_lr: 0.006933  loss: 3.6067 (3.6081)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0006  max mem: 55573
Epoch: [86]  [110/156]  eta: 0:00:32  lr: 0.006931  min_lr: 0.006931  loss: 3.2387 (3.5563)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0007  max mem: 55573
Epoch: [86]  [120/156]  eta: 0:00:25  lr: 0.006929  min_lr: 0.006929  loss: 3.4942 (3.5679)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0007  max mem: 55573
Epoch: [86]  [130/156]  eta: 0:00:17  lr: 0.006927  min_lr: 0.006927  loss: 3.5239 (3.5545)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0016  max mem: 55573
Epoch: [86]  [140/156]  eta: 0:00:10  lr: 0.006925  min_lr: 0.006925  loss: 3.5774 (3.5588)  weight_decay: 0.0500 (0.0500)  time: 0.5860  data: 0.0013  max mem: 55573
Epoch: [86]  [150/156]  eta: 0:00:04  lr: 0.006923  min_lr: 0.006923  loss: 3.5788 (3.5498)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0002  max mem: 55573
Epoch: [86]  [155/156]  eta: 0:00:00  lr: 0.006922  min_lr: 0.006922  loss: 3.4837 (3.5408)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0002  max mem: 55573
Epoch: [86] Total time: 0:01:45 (0.6787 s / it)
Averaged stats: lr: 0.006922  min_lr: 0.006922  loss: 3.4837 (3.5565)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.2330 (1.2330)  acc1: 77.0833 (77.0833)  acc5: 93.2292 (93.2292)  time: 7.1223  data: 6.8850  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3083 (1.3568)  acc1: 75.5208 (70.7840)  acc5: 92.9688 (90.0800)  time: 1.5708  data: 1.3771  max mem: 55573
Test: Total time: 0:00:08 (1.6152 s / it)
* Acc@1 70.666 Acc@5 90.204 loss 1.359
Accuracy of the model on the 50000 test images: 70.7%
Max accuracy: 70.71%
Test:  [0/5]  eta: 0:00:33  loss: 7.4025 (7.4025)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.7088  data: 6.4727  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5963 (7.6394)  acc1: 0.0000 (0.1280)  acc5: 0.0000 (0.5120)  time: 1.5031  data: 1.3105  max mem: 55573
Test: Total time: 0:00:07 (1.5193 s / it)
* Acc@1 0.132 Acc@5 0.554 loss 7.581
Accuracy of the model EMA on 50000 test images: 0.1%
Max EMA accuracy: 0.13%
Epoch: [87]  [  0/156]  eta: 0:32:37  lr: 0.006922  min_lr: 0.006922  loss: 4.1604 (4.1604)  weight_decay: 0.0500 (0.0500)  time: 12.5472  data: 7.6935  max mem: 55573
Epoch: [87]  [ 10/156]  eta: 0:04:05  lr: 0.006920  min_lr: 0.006920  loss: 3.7208 (3.6360)  weight_decay: 0.0500 (0.0500)  time: 1.6797  data: 0.7000  max mem: 55573
Epoch: [87]  [ 20/156]  eta: 0:02:37  lr: 0.006918  min_lr: 0.006918  loss: 3.5311 (3.4910)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0007  max mem: 55573
Epoch: [87]  [ 30/156]  eta: 0:02:03  lr: 0.006916  min_lr: 0.006916  loss: 3.3975 (3.4988)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0006  max mem: 55573
Epoch: [87]  [ 40/156]  eta: 0:01:42  lr: 0.006914  min_lr: 0.006914  loss: 3.5704 (3.4728)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0007  max mem: 55573
Epoch: [87]  [ 50/156]  eta: 0:01:27  lr: 0.006912  min_lr: 0.006912  loss: 3.6792 (3.5052)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0007  max mem: 55573
Epoch: [87]  [ 60/156]  eta: 0:01:15  lr: 0.006910  min_lr: 0.006910  loss: 3.8001 (3.5273)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0006  max mem: 55573
Epoch: [87]  [ 70/156]  eta: 0:01:05  lr: 0.006908  min_lr: 0.006908  loss: 3.3892 (3.4774)  weight_decay: 0.0500 (0.0500)  time: 0.5958  data: 0.0006  max mem: 55573
Epoch: [87]  [ 80/156]  eta: 0:00:56  lr: 0.006906  min_lr: 0.006906  loss: 3.0165 (3.4477)  weight_decay: 0.0500 (0.0500)  time: 0.5972  data: 0.0006  max mem: 55573
Epoch: [87]  [ 90/156]  eta: 0:00:47  lr: 0.006904  min_lr: 0.006904  loss: 3.3302 (3.4450)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0006  max mem: 55573
Epoch: [87]  [100/156]  eta: 0:00:39  lr: 0.006902  min_lr: 0.006902  loss: 3.7197 (3.4669)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0006  max mem: 55573
Epoch: [87]  [110/156]  eta: 0:00:32  lr: 0.006900  min_lr: 0.006900  loss: 3.8585 (3.4923)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0007  max mem: 55573
Epoch: [87]  [120/156]  eta: 0:00:24  lr: 0.006898  min_lr: 0.006898  loss: 3.7808 (3.4834)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0006  max mem: 55573
Epoch: [87]  [130/156]  eta: 0:00:17  lr: 0.006896  min_lr: 0.006896  loss: 3.4484 (3.4757)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0015  max mem: 55573
Epoch: [87]  [140/156]  eta: 0:00:10  lr: 0.006894  min_lr: 0.006894  loss: 3.6752 (3.4951)  weight_decay: 0.0500 (0.0500)  time: 0.5862  data: 0.0013  max mem: 55573
Epoch: [87]  [150/156]  eta: 0:00:04  lr: 0.006893  min_lr: 0.006893  loss: 3.7252 (3.5065)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0002  max mem: 55573
Epoch: [87]  [155/156]  eta: 0:00:00  lr: 0.006892  min_lr: 0.006892  loss: 3.7204 (3.5011)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0002  max mem: 55573
Epoch: [87] Total time: 0:01:45 (0.6738 s / it)
Averaged stats: lr: 0.006892  min_lr: 0.006892  loss: 3.7204 (3.5477)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.1353 (1.1353)  acc1: 76.9531 (76.9531)  acc5: 94.1406 (94.1406)  time: 7.3052  data: 7.0675  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2179 (1.2225)  acc1: 72.5260 (70.7520)  acc5: 92.4479 (91.1360)  time: 1.6073  data: 1.4136  max mem: 55573
Test: Total time: 0:00:08 (1.6501 s / it)
* Acc@1 71.598 Acc@5 90.960 loss 1.238
Accuracy of the model on the 50000 test images: 71.6%
Max accuracy: 71.60%
Test:  [0/5]  eta: 0:00:35  loss: 7.4056 (7.4056)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.0927  data: 6.8564  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5936 (7.6399)  acc1: 0.0000 (0.1600)  acc5: 0.0000 (0.5120)  time: 1.5639  data: 1.3714  max mem: 55573
Test: Total time: 0:00:07 (1.5808 s / it)
* Acc@1 0.150 Acc@5 0.552 loss 7.584
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.15%
Epoch: [88]  [  0/156]  eta: 0:37:10  lr: 0.006891  min_lr: 0.006891  loss: 3.7348 (3.7348)  weight_decay: 0.0500 (0.0500)  time: 14.2981  data: 10.2054  max mem: 55573
Epoch: [88]  [ 10/156]  eta: 0:04:27  lr: 0.006889  min_lr: 0.006889  loss: 3.5443 (3.5204)  weight_decay: 0.0500 (0.0500)  time: 1.8336  data: 0.9284  max mem: 55573
Epoch: [88]  [ 20/156]  eta: 0:02:48  lr: 0.006887  min_lr: 0.006887  loss: 3.6154 (3.5992)  weight_decay: 0.0500 (0.0500)  time: 0.5887  data: 0.0007  max mem: 55573
Epoch: [88]  [ 30/156]  eta: 0:02:09  lr: 0.006885  min_lr: 0.006885  loss: 3.6416 (3.6113)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0007  max mem: 55573
Epoch: [88]  [ 40/156]  eta: 0:01:47  lr: 0.006883  min_lr: 0.006883  loss: 3.5867 (3.6242)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [88]  [ 50/156]  eta: 0:01:30  lr: 0.006881  min_lr: 0.006881  loss: 3.8090 (3.6712)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0006  max mem: 55573
Epoch: [88]  [ 60/156]  eta: 0:01:18  lr: 0.006879  min_lr: 0.006879  loss: 3.8090 (3.6477)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0006  max mem: 55573
Epoch: [88]  [ 70/156]  eta: 0:01:07  lr: 0.006877  min_lr: 0.006877  loss: 3.5176 (3.6416)  weight_decay: 0.0500 (0.0500)  time: 0.5940  data: 0.0006  max mem: 55573
Epoch: [88]  [ 80/156]  eta: 0:00:57  lr: 0.006875  min_lr: 0.006875  loss: 3.3873 (3.5998)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0007  max mem: 55573
Epoch: [88]  [ 90/156]  eta: 0:00:48  lr: 0.006873  min_lr: 0.006873  loss: 3.4234 (3.5824)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0007  max mem: 55573
Epoch: [88]  [100/156]  eta: 0:00:40  lr: 0.006871  min_lr: 0.006871  loss: 3.5660 (3.5685)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0006  max mem: 55573
Epoch: [88]  [110/156]  eta: 0:00:32  lr: 0.006869  min_lr: 0.006869  loss: 3.6002 (3.5659)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0007  max mem: 55573
Epoch: [88]  [120/156]  eta: 0:00:25  lr: 0.006867  min_lr: 0.006867  loss: 3.7216 (3.5660)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0006  max mem: 55573
Epoch: [88]  [130/156]  eta: 0:00:18  lr: 0.006865  min_lr: 0.006865  loss: 3.7254 (3.5844)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0015  max mem: 55573
Epoch: [88]  [140/156]  eta: 0:00:11  lr: 0.006863  min_lr: 0.006863  loss: 3.6985 (3.5807)  weight_decay: 0.0500 (0.0500)  time: 0.5878  data: 0.0014  max mem: 55573
Epoch: [88]  [150/156]  eta: 0:00:04  lr: 0.006861  min_lr: 0.006861  loss: 3.6370 (3.5765)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0002  max mem: 55573
Epoch: [88]  [155/156]  eta: 0:00:00  lr: 0.006860  min_lr: 0.006860  loss: 3.6542 (3.5715)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0002  max mem: 55573
Epoch: [88] Total time: 0:01:46 (0.6843 s / it)
Averaged stats: lr: 0.006860  min_lr: 0.006860  loss: 3.6542 (3.5642)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.2736 (1.2736)  acc1: 75.2604 (75.2604)  acc5: 93.2292 (93.2292)  time: 7.0917  data: 6.8537  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3377 (1.3797)  acc1: 73.1771 (70.1120)  acc5: 92.0573 (90.0800)  time: 1.5649  data: 1.3708  max mem: 55573
Test: Total time: 0:00:08 (1.6164 s / it)
* Acc@1 70.428 Acc@5 90.272 loss 1.395
Accuracy of the model on the 50000 test images: 70.4%
Max accuracy: 71.60%
Test:  [0/5]  eta: 0:00:34  loss: 7.4049 (7.4049)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9622  data: 6.7261  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5837 (7.6340)  acc1: 0.0000 (0.1600)  acc5: 0.0000 (0.5120)  time: 1.5377  data: 1.3453  max mem: 55573
Test: Total time: 0:00:07 (1.5585 s / it)
* Acc@1 0.158 Acc@5 0.542 loss 7.581
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.16%
Epoch: [89]  [  0/156]  eta: 0:30:33  lr: 0.006860  min_lr: 0.006860  loss: 2.6832 (2.6832)  weight_decay: 0.0500 (0.0500)  time: 11.7552  data: 11.0447  max mem: 55573
Epoch: [89]  [ 10/156]  eta: 0:04:09  lr: 0.006858  min_lr: 0.006858  loss: 3.8554 (3.6095)  weight_decay: 0.0500 (0.0500)  time: 1.7072  data: 1.0605  max mem: 55573
Epoch: [89]  [ 20/156]  eta: 0:02:39  lr: 0.006856  min_lr: 0.006856  loss: 3.6512 (3.5461)  weight_decay: 0.0500 (0.0500)  time: 0.6473  data: 0.0313  max mem: 55573
Epoch: [89]  [ 30/156]  eta: 0:02:05  lr: 0.006854  min_lr: 0.006854  loss: 3.6331 (3.5763)  weight_decay: 0.0500 (0.0500)  time: 0.5991  data: 0.0006  max mem: 55573
Epoch: [89]  [ 40/156]  eta: 0:01:43  lr: 0.006852  min_lr: 0.006852  loss: 3.6193 (3.5006)  weight_decay: 0.0500 (0.0500)  time: 0.5976  data: 0.0006  max mem: 55573
Epoch: [89]  [ 50/156]  eta: 0:01:28  lr: 0.006850  min_lr: 0.006850  loss: 3.4308 (3.5058)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0006  max mem: 55573
Epoch: [89]  [ 60/156]  eta: 0:01:16  lr: 0.006848  min_lr: 0.006848  loss: 3.6337 (3.5288)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0007  max mem: 55573
Epoch: [89]  [ 70/156]  eta: 0:01:05  lr: 0.006846  min_lr: 0.006846  loss: 3.6867 (3.5101)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0007  max mem: 55573
Epoch: [89]  [ 80/156]  eta: 0:00:56  lr: 0.006844  min_lr: 0.006844  loss: 3.3618 (3.4745)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0006  max mem: 55573
Epoch: [89]  [ 90/156]  eta: 0:00:47  lr: 0.006842  min_lr: 0.006842  loss: 3.4297 (3.4980)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0006  max mem: 55573
Epoch: [89]  [100/156]  eta: 0:00:39  lr: 0.006840  min_lr: 0.006840  loss: 3.7841 (3.5032)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0007  max mem: 55573
Epoch: [89]  [110/156]  eta: 0:00:32  lr: 0.006838  min_lr: 0.006838  loss: 3.8193 (3.5329)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0008  max mem: 55573
Epoch: [89]  [120/156]  eta: 0:00:24  lr: 0.006836  min_lr: 0.006836  loss: 3.8853 (3.5573)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0007  max mem: 55573
Epoch: [89]  [130/156]  eta: 0:00:17  lr: 0.006834  min_lr: 0.006834  loss: 3.7809 (3.5551)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0019  max mem: 55573
Epoch: [89]  [140/156]  eta: 0:00:10  lr: 0.006832  min_lr: 0.006832  loss: 3.6426 (3.5673)  weight_decay: 0.0500 (0.0500)  time: 0.5873  data: 0.0017  max mem: 55573
Epoch: [89]  [150/156]  eta: 0:00:04  lr: 0.006830  min_lr: 0.006830  loss: 3.5979 (3.5665)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0002  max mem: 55573
Epoch: [89]  [155/156]  eta: 0:00:00  lr: 0.006829  min_lr: 0.006829  loss: 3.5979 (3.5702)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0002  max mem: 55573
Epoch: [89] Total time: 0:01:45 (0.6759 s / it)
Averaged stats: lr: 0.006829  min_lr: 0.006829  loss: 3.5979 (3.5462)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.3525 (1.3525)  acc1: 74.7396 (74.7396)  acc5: 92.9688 (92.9688)  time: 6.7035  data: 6.4655  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4314 (1.4609)  acc1: 72.9167 (69.6000)  acc5: 92.9688 (90.0800)  time: 1.4873  data: 1.2932  max mem: 55573
Test: Total time: 0:00:07 (1.5282 s / it)
* Acc@1 70.438 Acc@5 90.220 loss 1.436
Accuracy of the model on the 50000 test images: 70.4%
Max accuracy: 71.60%
Test:  [0/5]  eta: 0:00:37  loss: 7.3976 (7.3976)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.4066  data: 7.1707  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5713 (7.6237)  acc1: 0.0000 (0.1600)  acc5: 0.0000 (0.5120)  time: 1.6266  data: 1.4342  max mem: 55573
Test: Total time: 0:00:08 (1.6449 s / it)
* Acc@1 0.174 Acc@5 0.540 loss 7.574
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.17%
Epoch: [90]  [  0/156]  eta: 0:35:54  lr: 0.006829  min_lr: 0.006829  loss: 2.6781 (2.6781)  weight_decay: 0.0500 (0.0500)  time: 13.8110  data: 12.6045  max mem: 55573
Epoch: [90]  [ 10/156]  eta: 0:04:25  lr: 0.006827  min_lr: 0.006827  loss: 3.5112 (3.5169)  weight_decay: 0.0500 (0.0500)  time: 1.8194  data: 1.1465  max mem: 55573
Epoch: [90]  [ 20/156]  eta: 0:02:47  lr: 0.006825  min_lr: 0.006825  loss: 3.4810 (3.4534)  weight_decay: 0.0500 (0.0500)  time: 0.6059  data: 0.0006  max mem: 55573
Epoch: [90]  [ 30/156]  eta: 0:02:09  lr: 0.006822  min_lr: 0.006822  loss: 3.5843 (3.5752)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0006  max mem: 55573
Epoch: [90]  [ 40/156]  eta: 0:01:46  lr: 0.006820  min_lr: 0.006820  loss: 3.9058 (3.6186)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0007  max mem: 55573
Epoch: [90]  [ 50/156]  eta: 0:01:30  lr: 0.006818  min_lr: 0.006818  loss: 3.7393 (3.6642)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0006  max mem: 55573
Epoch: [90]  [ 60/156]  eta: 0:01:17  lr: 0.006816  min_lr: 0.006816  loss: 3.6988 (3.6426)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0007  max mem: 55573
Epoch: [90]  [ 70/156]  eta: 0:01:07  lr: 0.006814  min_lr: 0.006814  loss: 3.4116 (3.6093)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0009  max mem: 55573
Epoch: [90]  [ 80/156]  eta: 0:00:57  lr: 0.006812  min_lr: 0.006812  loss: 3.3385 (3.5901)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0008  max mem: 55573
Epoch: [90]  [ 90/156]  eta: 0:00:48  lr: 0.006810  min_lr: 0.006810  loss: 3.6255 (3.5972)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0008  max mem: 55573
Epoch: [90]  [100/156]  eta: 0:00:40  lr: 0.006808  min_lr: 0.006808  loss: 3.7996 (3.5953)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0007  max mem: 55573
Epoch: [90]  [110/156]  eta: 0:00:32  lr: 0.006806  min_lr: 0.006806  loss: 3.8426 (3.6135)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0006  max mem: 55573
Epoch: [90]  [120/156]  eta: 0:00:25  lr: 0.006804  min_lr: 0.006804  loss: 3.7834 (3.6066)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0007  max mem: 55573
Epoch: [90]  [130/156]  eta: 0:00:18  lr: 0.006802  min_lr: 0.006802  loss: 3.5808 (3.5946)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0021  max mem: 55573
Epoch: [90]  [140/156]  eta: 0:00:10  lr: 0.006800  min_lr: 0.006800  loss: 3.3426 (3.5832)  weight_decay: 0.0500 (0.0500)  time: 0.5869  data: 0.0019  max mem: 55573
Epoch: [90]  [150/156]  eta: 0:00:04  lr: 0.006798  min_lr: 0.006798  loss: 3.7988 (3.6004)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0002  max mem: 55573
Epoch: [90]  [155/156]  eta: 0:00:00  lr: 0.006797  min_lr: 0.006797  loss: 3.8259 (3.6061)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0002  max mem: 55573
Epoch: [90] Total time: 0:01:46 (0.6826 s / it)
Averaged stats: lr: 0.006797  min_lr: 0.006797  loss: 3.8259 (3.5612)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.3589 (1.3589)  acc1: 74.8698 (74.8698)  acc5: 93.0990 (93.0990)  time: 7.1502  data: 6.9125  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4793 (1.4890)  acc1: 72.0052 (69.8880)  acc5: 92.4479 (90.2720)  time: 1.5764  data: 1.3826  max mem: 55573
Test: Total time: 0:00:08 (1.6121 s / it)
* Acc@1 71.134 Acc@5 90.718 loss 1.494
Accuracy of the model on the 50000 test images: 71.1%
Max accuracy: 71.60%
Test:  [0/5]  eta: 0:00:35  loss: 7.3910 (7.3910)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.0097  data: 6.7736  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5621 (7.6157)  acc1: 0.0000 (0.2240)  acc5: 0.0000 (0.5120)  time: 1.5473  data: 1.3548  max mem: 55573
Test: Total time: 0:00:07 (1.5665 s / it)
* Acc@1 0.178 Acc@5 0.536 loss 7.569
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.18%
Epoch: [91]  [  0/156]  eta: 0:26:27  lr: 0.006797  min_lr: 0.006797  loss: 3.7694 (3.7694)  weight_decay: 0.0500 (0.0500)  time: 10.1738  data: 8.4450  max mem: 55573
Epoch: [91]  [ 10/156]  eta: 0:03:58  lr: 0.006795  min_lr: 0.006795  loss: 3.5228 (3.4325)  weight_decay: 0.0500 (0.0500)  time: 1.6340  data: 0.8495  max mem: 55573
Epoch: [91]  [ 20/156]  eta: 0:02:34  lr: 0.006793  min_lr: 0.006793  loss: 3.3679 (3.3587)  weight_decay: 0.0500 (0.0500)  time: 0.6845  data: 0.0452  max mem: 55573
Epoch: [91]  [ 30/156]  eta: 0:02:01  lr: 0.006790  min_lr: 0.006790  loss: 3.4854 (3.3546)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [91]  [ 40/156]  eta: 0:01:40  lr: 0.006788  min_lr: 0.006788  loss: 3.6781 (3.4061)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [91]  [ 50/156]  eta: 0:01:26  lr: 0.006786  min_lr: 0.006786  loss: 3.7155 (3.4558)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0007  max mem: 55573
Epoch: [91]  [ 60/156]  eta: 0:01:14  lr: 0.006784  min_lr: 0.006784  loss: 3.6401 (3.4683)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0006  max mem: 55573
Epoch: [91]  [ 70/156]  eta: 0:01:04  lr: 0.006782  min_lr: 0.006782  loss: 3.5889 (3.4672)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [91]  [ 80/156]  eta: 0:00:55  lr: 0.006780  min_lr: 0.006780  loss: 3.5567 (3.4878)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0006  max mem: 55573
Epoch: [91]  [ 90/156]  eta: 0:00:47  lr: 0.006778  min_lr: 0.006778  loss: 3.5511 (3.4883)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0006  max mem: 55573
Epoch: [91]  [100/156]  eta: 0:00:39  lr: 0.006776  min_lr: 0.006776  loss: 3.5153 (3.4985)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0005  max mem: 55573
Epoch: [91]  [110/156]  eta: 0:00:31  lr: 0.006774  min_lr: 0.006774  loss: 3.4458 (3.5021)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [91]  [120/156]  eta: 0:00:24  lr: 0.006772  min_lr: 0.006772  loss: 3.4617 (3.4931)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [91]  [130/156]  eta: 0:00:17  lr: 0.006770  min_lr: 0.006770  loss: 3.6449 (3.5075)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0012  max mem: 55573
Epoch: [91]  [140/156]  eta: 0:00:10  lr: 0.006768  min_lr: 0.006768  loss: 3.5981 (3.5068)  weight_decay: 0.0500 (0.0500)  time: 0.5878  data: 0.0010  max mem: 55573
Epoch: [91]  [150/156]  eta: 0:00:03  lr: 0.006766  min_lr: 0.006766  loss: 3.5736 (3.5158)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [91]  [155/156]  eta: 0:00:00  lr: 0.006765  min_lr: 0.006765  loss: 3.5842 (3.5230)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [91] Total time: 0:01:44 (0.6691 s / it)
Averaged stats: lr: 0.006765  min_lr: 0.006765  loss: 3.5842 (3.5523)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.2748 (1.2748)  acc1: 75.6510 (75.6510)  acc5: 94.6615 (94.6615)  time: 6.8877  data: 6.6507  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3994 (1.3732)  acc1: 72.9167 (70.4320)  acc5: 92.1875 (90.7520)  time: 1.5241  data: 1.3302  max mem: 55573
Test: Total time: 0:00:07 (1.5615 s / it)
* Acc@1 70.704 Acc@5 90.284 loss 1.375
Accuracy of the model on the 50000 test images: 70.7%
Max accuracy: 71.60%
Test:  [0/5]  eta: 0:00:34  loss: 7.3851 (7.3851)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9765  data: 6.7404  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5510 (7.6048)  acc1: 0.0000 (0.2560)  acc5: 0.0000 (0.5120)  time: 1.5589  data: 1.3665  max mem: 55573
Test: Total time: 0:00:07 (1.5772 s / it)
* Acc@1 0.190 Acc@5 0.534 loss 7.560
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.19%
Epoch: [92]  [  0/156]  eta: 0:31:28  lr: 0.006764  min_lr: 0.006764  loss: 3.4494 (3.4494)  weight_decay: 0.0500 (0.0500)  time: 12.1064  data: 8.0062  max mem: 55573
Epoch: [92]  [ 10/156]  eta: 0:04:14  lr: 0.006762  min_lr: 0.006762  loss: 3.7564 (3.4418)  weight_decay: 0.0500 (0.0500)  time: 1.7447  data: 0.7285  max mem: 55573
Epoch: [92]  [ 20/156]  eta: 0:02:42  lr: 0.006760  min_lr: 0.006760  loss: 3.7564 (3.5407)  weight_decay: 0.0500 (0.0500)  time: 0.6496  data: 0.0007  max mem: 55573
Epoch: [92]  [ 30/156]  eta: 0:02:06  lr: 0.006758  min_lr: 0.006758  loss: 3.8301 (3.5772)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0006  max mem: 55573
Epoch: [92]  [ 40/156]  eta: 0:01:44  lr: 0.006756  min_lr: 0.006756  loss: 3.7953 (3.5741)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0006  max mem: 55573
Epoch: [92]  [ 50/156]  eta: 0:01:29  lr: 0.006754  min_lr: 0.006754  loss: 3.6429 (3.6185)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0006  max mem: 55573
Epoch: [92]  [ 60/156]  eta: 0:01:16  lr: 0.006752  min_lr: 0.006752  loss: 3.7564 (3.6310)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0006  max mem: 55573
Epoch: [92]  [ 70/156]  eta: 0:01:06  lr: 0.006750  min_lr: 0.006750  loss: 3.7588 (3.6155)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [92]  [ 80/156]  eta: 0:00:56  lr: 0.006748  min_lr: 0.006748  loss: 3.7322 (3.6318)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [92]  [ 90/156]  eta: 0:00:48  lr: 0.006746  min_lr: 0.006746  loss: 3.7350 (3.6166)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [92]  [100/156]  eta: 0:00:40  lr: 0.006744  min_lr: 0.006744  loss: 3.7553 (3.6328)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0006  max mem: 55573
Epoch: [92]  [110/156]  eta: 0:00:32  lr: 0.006741  min_lr: 0.006741  loss: 3.8622 (3.6342)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0006  max mem: 55573
Epoch: [92]  [120/156]  eta: 0:00:25  lr: 0.006739  min_lr: 0.006739  loss: 3.8806 (3.6630)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0006  max mem: 55573
Epoch: [92]  [130/156]  eta: 0:00:17  lr: 0.006737  min_lr: 0.006737  loss: 3.7881 (3.6492)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0011  max mem: 55573
Epoch: [92]  [140/156]  eta: 0:00:10  lr: 0.006735  min_lr: 0.006735  loss: 3.6620 (3.6594)  weight_decay: 0.0500 (0.0500)  time: 0.5859  data: 0.0009  max mem: 55573
Epoch: [92]  [150/156]  eta: 0:00:04  lr: 0.006733  min_lr: 0.006733  loss: 3.6620 (3.6506)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0002  max mem: 55573
Epoch: [92]  [155/156]  eta: 0:00:00  lr: 0.006732  min_lr: 0.006732  loss: 3.5981 (3.6465)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0002  max mem: 55573
Epoch: [92] Total time: 0:01:45 (0.6774 s / it)
Averaged stats: lr: 0.006732  min_lr: 0.006732  loss: 3.5981 (3.5494)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.3255 (1.3255)  acc1: 74.8698 (74.8698)  acc5: 92.7083 (92.7083)  time: 7.0276  data: 6.7903  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3632 (1.4150)  acc1: 70.7031 (69.6000)  acc5: 92.0573 (89.9200)  time: 1.5521  data: 1.3581  max mem: 55573
Test: Total time: 0:00:07 (1.5890 s / it)
* Acc@1 70.390 Acc@5 90.460 loss 1.417
Accuracy of the model on the 50000 test images: 70.4%
Max accuracy: 71.60%
Test:  [0/5]  eta: 0:00:34  loss: 7.3721 (7.3721)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9443  data: 6.7081  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5366 (7.5913)  acc1: 0.0000 (0.2880)  acc5: 0.0000 (0.5440)  time: 1.5341  data: 1.3417  max mem: 55573
Test: Total time: 0:00:07 (1.5653 s / it)
* Acc@1 0.200 Acc@5 0.542 loss 7.549
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.20%
Epoch: [93]  [  0/156]  eta: 0:31:52  lr: 0.006732  min_lr: 0.006732  loss: 3.3785 (3.3785)  weight_decay: 0.0500 (0.0500)  time: 12.2613  data: 11.6415  max mem: 55573
Epoch: [93]  [ 10/156]  eta: 0:04:12  lr: 0.006730  min_lr: 0.006730  loss: 3.7472 (3.6943)  weight_decay: 0.0500 (0.0500)  time: 1.7322  data: 1.0794  max mem: 55573
Epoch: [93]  [ 20/156]  eta: 0:02:41  lr: 0.006728  min_lr: 0.006728  loss: 3.7472 (3.6036)  weight_decay: 0.0500 (0.0500)  time: 0.6345  data: 0.0119  max mem: 55573
Epoch: [93]  [ 30/156]  eta: 0:02:05  lr: 0.006725  min_lr: 0.006725  loss: 3.7886 (3.6322)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [93]  [ 40/156]  eta: 0:01:44  lr: 0.006723  min_lr: 0.006723  loss: 3.5720 (3.5584)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0006  max mem: 55573
Epoch: [93]  [ 50/156]  eta: 0:01:28  lr: 0.006721  min_lr: 0.006721  loss: 3.5720 (3.5961)  weight_decay: 0.0500 (0.0500)  time: 0.5933  data: 0.0007  max mem: 55573
Epoch: [93]  [ 60/156]  eta: 0:01:16  lr: 0.006719  min_lr: 0.006719  loss: 3.6537 (3.5690)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0006  max mem: 55573
Epoch: [93]  [ 70/156]  eta: 0:01:06  lr: 0.006717  min_lr: 0.006717  loss: 3.6537 (3.5666)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0006  max mem: 55573
Epoch: [93]  [ 80/156]  eta: 0:00:56  lr: 0.006715  min_lr: 0.006715  loss: 3.6671 (3.5694)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0006  max mem: 55573
Epoch: [93]  [ 90/156]  eta: 0:00:48  lr: 0.006713  min_lr: 0.006713  loss: 3.7406 (3.5760)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [93]  [100/156]  eta: 0:00:40  lr: 0.006711  min_lr: 0.006711  loss: 3.7946 (3.5964)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0006  max mem: 55573
Epoch: [93]  [110/156]  eta: 0:00:32  lr: 0.006709  min_lr: 0.006709  loss: 3.8819 (3.6207)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0006  max mem: 55573
Epoch: [93]  [120/156]  eta: 0:00:25  lr: 0.006706  min_lr: 0.006706  loss: 3.7618 (3.6134)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0006  max mem: 55573
Epoch: [93]  [130/156]  eta: 0:00:17  lr: 0.006704  min_lr: 0.006704  loss: 3.5684 (3.5941)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0012  max mem: 55573
Epoch: [93]  [140/156]  eta: 0:00:10  lr: 0.006702  min_lr: 0.006702  loss: 3.5684 (3.5797)  weight_decay: 0.0500 (0.0500)  time: 0.5852  data: 0.0010  max mem: 55573
Epoch: [93]  [150/156]  eta: 0:00:04  lr: 0.006700  min_lr: 0.006700  loss: 3.6782 (3.5904)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0002  max mem: 55573
Epoch: [93]  [155/156]  eta: 0:00:00  lr: 0.006699  min_lr: 0.006699  loss: 3.6632 (3.5788)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0002  max mem: 55573
Epoch: [93] Total time: 0:01:45 (0.6766 s / it)
Averaged stats: lr: 0.006699  min_lr: 0.006699  loss: 3.6632 (3.5508)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.1788 (1.1788)  acc1: 77.2135 (77.2135)  acc5: 94.6615 (94.6615)  time: 6.9852  data: 6.7479  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3453 (1.3834)  acc1: 72.7865 (70.8800)  acc5: 91.1458 (90.0480)  time: 1.5431  data: 1.3498  max mem: 55573
Test: Total time: 0:00:07 (1.5843 s / it)
* Acc@1 71.428 Acc@5 90.442 loss 1.345
Accuracy of the model on the 50000 test images: 71.4%
Max accuracy: 71.60%
Test:  [0/5]  eta: 0:00:37  loss: 7.3660 (7.3660)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.5785  data: 7.3420  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5250 (7.5813)  acc1: 0.0000 (0.2880)  acc5: 0.0000 (0.5440)  time: 1.6612  data: 1.4685  max mem: 55573
Test: Total time: 0:00:08 (1.6795 s / it)
* Acc@1 0.206 Acc@5 0.542 loss 7.541
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.21%
Epoch: [94]  [  0/156]  eta: 0:30:15  lr: 0.006699  min_lr: 0.006699  loss: 3.7628 (3.7628)  weight_decay: 0.0500 (0.0500)  time: 11.6410  data: 10.7380  max mem: 55573
Epoch: [94]  [ 10/156]  eta: 0:04:08  lr: 0.006697  min_lr: 0.006697  loss: 3.6545 (3.4709)  weight_decay: 0.0500 (0.0500)  time: 1.6992  data: 0.9768  max mem: 55573
Epoch: [94]  [ 20/156]  eta: 0:02:39  lr: 0.006695  min_lr: 0.006695  loss: 3.5358 (3.3901)  weight_decay: 0.0500 (0.0500)  time: 0.6477  data: 0.0006  max mem: 55573
Epoch: [94]  [ 30/156]  eta: 0:02:03  lr: 0.006692  min_lr: 0.006692  loss: 3.1403 (3.2794)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0007  max mem: 55573
Epoch: [94]  [ 40/156]  eta: 0:01:43  lr: 0.006690  min_lr: 0.006690  loss: 3.3206 (3.3132)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0007  max mem: 55573
Epoch: [94]  [ 50/156]  eta: 0:01:27  lr: 0.006688  min_lr: 0.006688  loss: 3.5670 (3.3481)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0007  max mem: 55573
Epoch: [94]  [ 60/156]  eta: 0:01:15  lr: 0.006686  min_lr: 0.006686  loss: 3.5670 (3.3506)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0006  max mem: 55573
Epoch: [94]  [ 70/156]  eta: 0:01:05  lr: 0.006684  min_lr: 0.006684  loss: 3.5819 (3.3711)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0006  max mem: 55573
Epoch: [94]  [ 80/156]  eta: 0:00:56  lr: 0.006682  min_lr: 0.006682  loss: 3.6485 (3.3876)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0007  max mem: 55573
Epoch: [94]  [ 90/156]  eta: 0:00:47  lr: 0.006680  min_lr: 0.006680  loss: 3.6168 (3.4083)  weight_decay: 0.0500 (0.0500)  time: 0.5936  data: 0.0008  max mem: 55573
Epoch: [94]  [100/156]  eta: 0:00:39  lr: 0.006678  min_lr: 0.006678  loss: 3.6828 (3.4104)  weight_decay: 0.0500 (0.0500)  time: 0.5940  data: 0.0008  max mem: 55573
Epoch: [94]  [110/156]  eta: 0:00:32  lr: 0.006675  min_lr: 0.006675  loss: 3.3354 (3.4019)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0008  max mem: 55573
Epoch: [94]  [120/156]  eta: 0:00:24  lr: 0.006673  min_lr: 0.006673  loss: 3.4742 (3.4084)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0007  max mem: 55573
Epoch: [94]  [130/156]  eta: 0:00:17  lr: 0.006671  min_lr: 0.006671  loss: 3.6243 (3.4362)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0022  max mem: 55573
Epoch: [94]  [140/156]  eta: 0:00:10  lr: 0.006669  min_lr: 0.006669  loss: 3.7145 (3.4506)  weight_decay: 0.0500 (0.0500)  time: 0.5868  data: 0.0020  max mem: 55573
Epoch: [94]  [150/156]  eta: 0:00:04  lr: 0.006667  min_lr: 0.006667  loss: 3.6419 (3.4567)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0002  max mem: 55573
Epoch: [94]  [155/156]  eta: 0:00:00  lr: 0.006666  min_lr: 0.006666  loss: 3.5381 (3.4563)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0002  max mem: 55573
Epoch: [94] Total time: 0:01:45 (0.6749 s / it)
Averaged stats: lr: 0.006666  min_lr: 0.006666  loss: 3.5381 (3.5447)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.1710 (1.1710)  acc1: 75.0000 (75.0000)  acc5: 94.1406 (94.1406)  time: 7.3282  data: 7.0908  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2986 (1.2826)  acc1: 72.2656 (70.2080)  acc5: 92.9688 (91.0720)  time: 1.6119  data: 1.4182  max mem: 55573
Test: Total time: 0:00:08 (1.6572 s / it)
* Acc@1 71.878 Acc@5 91.128 loss 1.237
Accuracy of the model on the 50000 test images: 71.9%
Max accuracy: 71.88%
Test:  [0/5]  eta: 0:00:35  loss: 7.3611 (7.3611)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.0480  data: 6.8113  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5166 (7.5743)  acc1: 0.0000 (0.2880)  acc5: 0.0000 (0.6080)  time: 1.5552  data: 1.3624  max mem: 55573
Test: Total time: 0:00:07 (1.5747 s / it)
* Acc@1 0.212 Acc@5 0.558 loss 7.535
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.21%
Epoch: [95]  [  0/156]  eta: 0:35:39  lr: 0.006666  min_lr: 0.006666  loss: 2.8651 (2.8651)  weight_decay: 0.0500 (0.0500)  time: 13.7125  data: 13.1282  max mem: 55573
Epoch: [95]  [ 10/156]  eta: 0:04:20  lr: 0.006663  min_lr: 0.006663  loss: 3.4265 (3.3282)  weight_decay: 0.0500 (0.0500)  time: 1.7813  data: 1.1938  max mem: 55573
Epoch: [95]  [ 20/156]  eta: 0:02:45  lr: 0.006661  min_lr: 0.006661  loss: 3.5377 (3.4726)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [95]  [ 30/156]  eta: 0:02:07  lr: 0.006659  min_lr: 0.006659  loss: 3.8204 (3.5590)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [95]  [ 40/156]  eta: 0:01:45  lr: 0.006657  min_lr: 0.006657  loss: 3.5810 (3.5320)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0007  max mem: 55573
Epoch: [95]  [ 50/156]  eta: 0:01:29  lr: 0.006655  min_lr: 0.006655  loss: 3.4604 (3.5467)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0008  max mem: 55573
Epoch: [95]  [ 60/156]  eta: 0:01:17  lr: 0.006653  min_lr: 0.006653  loss: 3.7159 (3.5469)  weight_decay: 0.0500 (0.0500)  time: 0.5932  data: 0.0007  max mem: 55573
Epoch: [95]  [ 70/156]  eta: 0:01:06  lr: 0.006651  min_lr: 0.006651  loss: 3.7194 (3.5446)  weight_decay: 0.0500 (0.0500)  time: 0.5938  data: 0.0006  max mem: 55573
Epoch: [95]  [ 80/156]  eta: 0:00:57  lr: 0.006648  min_lr: 0.006648  loss: 3.7712 (3.5438)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0008  max mem: 55573
Epoch: [95]  [ 90/156]  eta: 0:00:48  lr: 0.006646  min_lr: 0.006646  loss: 3.7712 (3.5482)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0007  max mem: 55573
Epoch: [95]  [100/156]  eta: 0:00:40  lr: 0.006644  min_lr: 0.006644  loss: 3.8091 (3.5563)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0007  max mem: 55573
Epoch: [95]  [110/156]  eta: 0:00:32  lr: 0.006642  min_lr: 0.006642  loss: 3.6754 (3.5345)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0008  max mem: 55573
Epoch: [95]  [120/156]  eta: 0:00:25  lr: 0.006640  min_lr: 0.006640  loss: 3.7165 (3.5491)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0006  max mem: 55573
Epoch: [95]  [130/156]  eta: 0:00:17  lr: 0.006638  min_lr: 0.006638  loss: 3.7470 (3.5476)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0013  max mem: 55573
Epoch: [95]  [140/156]  eta: 0:00:10  lr: 0.006635  min_lr: 0.006635  loss: 3.5032 (3.5490)  weight_decay: 0.0500 (0.0500)  time: 0.5856  data: 0.0011  max mem: 55573
Epoch: [95]  [150/156]  eta: 0:00:04  lr: 0.006633  min_lr: 0.006633  loss: 3.4795 (3.5421)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0002  max mem: 55573
Epoch: [95]  [155/156]  eta: 0:00:00  lr: 0.006632  min_lr: 0.006632  loss: 3.4169 (3.5390)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0002  max mem: 55573
Epoch: [95] Total time: 0:01:46 (0.6801 s / it)
Averaged stats: lr: 0.006632  min_lr: 0.006632  loss: 3.4169 (3.5240)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.2612 (1.2612)  acc1: 75.1302 (75.1302)  acc5: 93.4896 (93.4896)  time: 7.3837  data: 7.1463  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3669 (1.3644)  acc1: 71.6146 (69.7280)  acc5: 91.5365 (90.2080)  time: 1.6232  data: 1.4293  max mem: 55573
Test: Total time: 0:00:08 (1.6664 s / it)
* Acc@1 71.016 Acc@5 90.426 loss 1.352
Accuracy of the model on the 50000 test images: 71.0%
Max accuracy: 71.88%
Test:  [0/5]  eta: 0:00:36  loss: 7.3564 (7.3564)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.2543  data: 7.0177  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.5076 (7.5667)  acc1: 0.0000 (0.2880)  acc5: 0.0000 (0.6080)  time: 1.5964  data: 1.4037  max mem: 55573
Test: Total time: 0:00:08 (1.6113 s / it)
* Acc@1 0.216 Acc@5 0.568 loss 7.530
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.22%
Epoch: [96]  [  0/156]  eta: 0:38:32  lr: 0.006632  min_lr: 0.006632  loss: 3.3143 (3.3143)  weight_decay: 0.0500 (0.0500)  time: 14.8245  data: 14.2398  max mem: 55573
Epoch: [96]  [ 10/156]  eta: 0:04:35  lr: 0.006630  min_lr: 0.006630  loss: 3.6872 (3.5248)  weight_decay: 0.0500 (0.0500)  time: 1.8900  data: 1.2950  max mem: 55573
Epoch: [96]  [ 20/156]  eta: 0:02:52  lr: 0.006628  min_lr: 0.006628  loss: 3.7639 (3.5456)  weight_decay: 0.0500 (0.0500)  time: 0.5940  data: 0.0005  max mem: 55573
Epoch: [96]  [ 30/156]  eta: 0:02:12  lr: 0.006625  min_lr: 0.006625  loss: 3.7944 (3.6075)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [96]  [ 40/156]  eta: 0:01:48  lr: 0.006623  min_lr: 0.006623  loss: 3.7231 (3.5828)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0006  max mem: 55573
Epoch: [96]  [ 50/156]  eta: 0:01:32  lr: 0.006621  min_lr: 0.006621  loss: 3.4952 (3.5425)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0006  max mem: 55573
Epoch: [96]  [ 60/156]  eta: 0:01:19  lr: 0.006619  min_lr: 0.006619  loss: 3.5816 (3.5482)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0006  max mem: 55573
Epoch: [96]  [ 70/156]  eta: 0:01:08  lr: 0.006617  min_lr: 0.006617  loss: 3.7043 (3.5333)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [96]  [ 80/156]  eta: 0:00:58  lr: 0.006615  min_lr: 0.006615  loss: 3.4610 (3.5233)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0006  max mem: 55573
Epoch: [96]  [ 90/156]  eta: 0:00:49  lr: 0.006612  min_lr: 0.006612  loss: 3.4278 (3.4950)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0006  max mem: 55573
Epoch: [96]  [100/156]  eta: 0:00:41  lr: 0.006610  min_lr: 0.006610  loss: 3.6417 (3.5075)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [96]  [110/156]  eta: 0:00:33  lr: 0.006608  min_lr: 0.006608  loss: 3.6760 (3.5007)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0005  max mem: 55573
Epoch: [96]  [120/156]  eta: 0:00:25  lr: 0.006606  min_lr: 0.006606  loss: 3.7193 (3.5236)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [96]  [130/156]  eta: 0:00:18  lr: 0.006604  min_lr: 0.006604  loss: 3.7551 (3.5036)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0011  max mem: 55573
Epoch: [96]  [140/156]  eta: 0:00:11  lr: 0.006601  min_lr: 0.006601  loss: 3.7366 (3.5269)  weight_decay: 0.0500 (0.0500)  time: 0.5859  data: 0.0009  max mem: 55573
Epoch: [96]  [150/156]  eta: 0:00:04  lr: 0.006599  min_lr: 0.006599  loss: 3.8236 (3.5247)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [96]  [155/156]  eta: 0:00:00  lr: 0.006598  min_lr: 0.006598  loss: 3.4593 (3.5086)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [96] Total time: 0:01:47 (0.6879 s / it)
Averaged stats: lr: 0.006598  min_lr: 0.006598  loss: 3.4593 (3.5446)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.1760 (1.1760)  acc1: 73.0469 (73.0469)  acc5: 93.8802 (93.8802)  time: 7.3042  data: 7.0668  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2398 (1.2794)  acc1: 72.5260 (70.1440)  acc5: 93.0990 (90.9440)  time: 1.6074  data: 1.4134  max mem: 55573
Test: Total time: 0:00:08 (1.6686 s / it)
* Acc@1 71.050 Acc@5 90.774 loss 1.293
Accuracy of the model on the 50000 test images: 71.1%
Max accuracy: 71.88%
Test:  [0/5]  eta: 0:00:32  loss: 7.3554 (7.3554)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.4246  data: 6.1886  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.4980 (7.5598)  acc1: 0.0000 (0.2880)  acc5: 0.0000 (0.6720)  time: 1.4807  data: 1.2884  max mem: 55573
Test: Total time: 0:00:07 (1.5005 s / it)
* Acc@1 0.218 Acc@5 0.600 loss 7.524
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.22%
Epoch: [97]  [  0/156]  eta: 0:35:02  lr: 0.006598  min_lr: 0.006598  loss: 3.2738 (3.2738)  weight_decay: 0.0500 (0.0500)  time: 13.4783  data: 12.8864  max mem: 55573
Epoch: [97]  [ 10/156]  eta: 0:04:19  lr: 0.006596  min_lr: 0.006596  loss: 3.7798 (3.5207)  weight_decay: 0.0500 (0.0500)  time: 1.7801  data: 1.1719  max mem: 55573
Epoch: [97]  [ 20/156]  eta: 0:02:45  lr: 0.006594  min_lr: 0.006594  loss: 3.7635 (3.6852)  weight_decay: 0.0500 (0.0500)  time: 0.6000  data: 0.0005  max mem: 55573
Epoch: [97]  [ 30/156]  eta: 0:02:07  lr: 0.006591  min_lr: 0.006591  loss: 3.7165 (3.6186)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [97]  [ 40/156]  eta: 0:01:45  lr: 0.006589  min_lr: 0.006589  loss: 3.6586 (3.6095)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [97]  [ 50/156]  eta: 0:01:29  lr: 0.006587  min_lr: 0.006587  loss: 3.6602 (3.6270)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [97]  [ 60/156]  eta: 0:01:17  lr: 0.006585  min_lr: 0.006585  loss: 3.6215 (3.5861)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [97]  [ 70/156]  eta: 0:01:06  lr: 0.006583  min_lr: 0.006583  loss: 3.6208 (3.5692)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0006  max mem: 55573
Epoch: [97]  [ 80/156]  eta: 0:00:57  lr: 0.006580  min_lr: 0.006580  loss: 3.6907 (3.5775)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0006  max mem: 55573
Epoch: [97]  [ 90/156]  eta: 0:00:48  lr: 0.006578  min_lr: 0.006578  loss: 3.6717 (3.5749)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0006  max mem: 55573
Epoch: [97]  [100/156]  eta: 0:00:40  lr: 0.006576  min_lr: 0.006576  loss: 3.5779 (3.5739)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0006  max mem: 55573
Epoch: [97]  [110/156]  eta: 0:00:32  lr: 0.006574  min_lr: 0.006574  loss: 3.4441 (3.5588)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [97]  [120/156]  eta: 0:00:25  lr: 0.006572  min_lr: 0.006572  loss: 3.7326 (3.5689)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0006  max mem: 55573
Epoch: [97]  [130/156]  eta: 0:00:17  lr: 0.006569  min_lr: 0.006569  loss: 3.7985 (3.5686)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0011  max mem: 55573
Epoch: [97]  [140/156]  eta: 0:00:10  lr: 0.006567  min_lr: 0.006567  loss: 3.6297 (3.5702)  weight_decay: 0.0500 (0.0500)  time: 0.5858  data: 0.0009  max mem: 55573
Epoch: [97]  [150/156]  eta: 0:00:04  lr: 0.006565  min_lr: 0.006565  loss: 3.7583 (3.5872)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0002  max mem: 55573
Epoch: [97]  [155/156]  eta: 0:00:00  lr: 0.006564  min_lr: 0.006564  loss: 3.7296 (3.5786)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [97] Total time: 0:01:46 (0.6798 s / it)
Averaged stats: lr: 0.006564  min_lr: 0.006564  loss: 3.7296 (3.5348)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.1493 (1.1493)  acc1: 75.5208 (75.5208)  acc5: 93.6198 (93.6198)  time: 6.9973  data: 6.7595  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2146 (1.2402)  acc1: 73.5677 (71.8720)  acc5: 93.6198 (91.2960)  time: 1.5460  data: 1.3520  max mem: 55573
Test: Total time: 0:00:07 (1.5872 s / it)
* Acc@1 72.068 Acc@5 91.224 loss 1.257
Accuracy of the model on the 50000 test images: 72.1%
Max accuracy: 72.07%
Test:  [0/5]  eta: 0:00:34  loss: 7.3578 (7.3578)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.9066  data: 6.6700  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.4876 (7.5534)  acc1: 0.0000 (0.3200)  acc5: 0.0000 (0.7040)  time: 1.5269  data: 1.3341  max mem: 55573
Test: Total time: 0:00:07 (1.5451 s / it)
* Acc@1 0.234 Acc@5 0.626 loss 7.518
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.23%
Epoch: [98]  [  0/156]  eta: 0:31:29  lr: 0.006564  min_lr: 0.006564  loss: 4.0497 (4.0497)  weight_decay: 0.0500 (0.0500)  time: 12.1142  data: 8.7939  max mem: 55573
Epoch: [98]  [ 10/156]  eta: 0:04:18  lr: 0.006561  min_lr: 0.006561  loss: 3.8331 (3.7579)  weight_decay: 0.0500 (0.0500)  time: 1.7685  data: 0.8004  max mem: 55573
Epoch: [98]  [ 20/156]  eta: 0:02:44  lr: 0.006559  min_lr: 0.006559  loss: 3.5809 (3.5107)  weight_decay: 0.0500 (0.0500)  time: 0.6625  data: 0.0008  max mem: 55573
Epoch: [98]  [ 30/156]  eta: 0:02:07  lr: 0.006557  min_lr: 0.006557  loss: 3.5809 (3.5188)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0006  max mem: 55573
Epoch: [98]  [ 40/156]  eta: 0:01:45  lr: 0.006555  min_lr: 0.006555  loss: 3.6361 (3.4596)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0006  max mem: 55573
Epoch: [98]  [ 50/156]  eta: 0:01:29  lr: 0.006553  min_lr: 0.006553  loss: 3.4044 (3.4518)  weight_decay: 0.0500 (0.0500)  time: 0.5883  data: 0.0006  max mem: 55573
Epoch: [98]  [ 60/156]  eta: 0:01:17  lr: 0.006550  min_lr: 0.006550  loss: 3.4044 (3.4455)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0006  max mem: 55573
Epoch: [98]  [ 70/156]  eta: 0:01:06  lr: 0.006548  min_lr: 0.006548  loss: 3.7241 (3.5032)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0006  max mem: 55573
Epoch: [98]  [ 80/156]  eta: 0:00:57  lr: 0.006546  min_lr: 0.006546  loss: 3.7747 (3.5076)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0006  max mem: 55573
Epoch: [98]  [ 90/156]  eta: 0:00:48  lr: 0.006544  min_lr: 0.006544  loss: 3.6203 (3.5263)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0006  max mem: 55573
Epoch: [98]  [100/156]  eta: 0:00:40  lr: 0.006542  min_lr: 0.006542  loss: 3.7048 (3.5262)  weight_decay: 0.0500 (0.0500)  time: 0.5931  data: 0.0005  max mem: 55573
Epoch: [98]  [110/156]  eta: 0:00:32  lr: 0.006539  min_lr: 0.006539  loss: 3.4292 (3.4971)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0005  max mem: 55573
Epoch: [98]  [120/156]  eta: 0:00:25  lr: 0.006537  min_lr: 0.006537  loss: 3.1856 (3.4958)  weight_decay: 0.0500 (0.0500)  time: 0.5945  data: 0.0005  max mem: 55573
Epoch: [98]  [130/156]  eta: 0:00:17  lr: 0.006535  min_lr: 0.006535  loss: 3.5460 (3.5098)  weight_decay: 0.0500 (0.0500)  time: 0.5956  data: 0.0015  max mem: 55573
Epoch: [98]  [140/156]  eta: 0:00:10  lr: 0.006533  min_lr: 0.006533  loss: 3.5460 (3.4954)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0013  max mem: 55573
Epoch: [98]  [150/156]  eta: 0:00:04  lr: 0.006530  min_lr: 0.006530  loss: 3.3750 (3.4956)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0002  max mem: 55573
Epoch: [98]  [155/156]  eta: 0:00:00  lr: 0.006529  min_lr: 0.006529  loss: 3.5070 (3.4996)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0002  max mem: 55573
Epoch: [98] Total time: 0:01:46 (0.6806 s / it)
Averaged stats: lr: 0.006529  min_lr: 0.006529  loss: 3.5070 (3.5177)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.3889 (1.3889)  acc1: 75.9115 (75.9115)  acc5: 93.3594 (93.3594)  time: 7.1850  data: 6.9476  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4239 (1.4651)  acc1: 73.3073 (70.6880)  acc5: 93.3594 (91.1040)  time: 1.5834  data: 1.3896  max mem: 55573
Test: Total time: 0:00:08 (1.6184 s / it)
* Acc@1 71.180 Acc@5 90.434 loss 1.479
Accuracy of the model on the 50000 test images: 71.2%
Max accuracy: 72.07%
Test:  [0/5]  eta: 0:00:36  loss: 7.3554 (7.3554)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.3481  data: 7.1115  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.4745 (7.5431)  acc1: 0.0000 (0.3200)  acc5: 0.0000 (0.7360)  time: 1.6151  data: 1.4224  max mem: 55573
Test: Total time: 0:00:08 (1.6377 s / it)
* Acc@1 0.238 Acc@5 0.682 loss 7.510
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.24%
Epoch: [99]  [  0/156]  eta: 0:34:50  lr: 0.006529  min_lr: 0.006529  loss: 3.4162 (3.4162)  weight_decay: 0.0500 (0.0500)  time: 13.3979  data: 12.8119  max mem: 55573
Epoch: [99]  [ 10/156]  eta: 0:04:18  lr: 0.006527  min_lr: 0.006527  loss: 3.7542 (3.4532)  weight_decay: 0.0500 (0.0500)  time: 1.7678  data: 1.1656  max mem: 55573
Epoch: [99]  [ 20/156]  eta: 0:02:44  lr: 0.006525  min_lr: 0.006525  loss: 3.7542 (3.4165)  weight_decay: 0.0500 (0.0500)  time: 0.5968  data: 0.0008  max mem: 55573
Epoch: [99]  [ 30/156]  eta: 0:02:07  lr: 0.006522  min_lr: 0.006522  loss: 3.0335 (3.3452)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [99]  [ 40/156]  eta: 0:01:45  lr: 0.006520  min_lr: 0.006520  loss: 3.3180 (3.4107)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [99]  [ 50/156]  eta: 0:01:29  lr: 0.006518  min_lr: 0.006518  loss: 3.8494 (3.4950)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [99]  [ 60/156]  eta: 0:01:17  lr: 0.006516  min_lr: 0.006516  loss: 3.7652 (3.5033)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [99]  [ 70/156]  eta: 0:01:06  lr: 0.006513  min_lr: 0.006513  loss: 3.6751 (3.5127)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0006  max mem: 55573
Epoch: [99]  [ 80/156]  eta: 0:00:57  lr: 0.006511  min_lr: 0.006511  loss: 3.7312 (3.5328)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [99]  [ 90/156]  eta: 0:00:48  lr: 0.006509  min_lr: 0.006509  loss: 3.6664 (3.5312)  weight_decay: 0.0500 (0.0500)  time: 0.5945  data: 0.0005  max mem: 55573
Epoch: [99]  [100/156]  eta: 0:00:40  lr: 0.006507  min_lr: 0.006507  loss: 3.6669 (3.5337)  weight_decay: 0.0500 (0.0500)  time: 0.5940  data: 0.0006  max mem: 55573
Epoch: [99]  [110/156]  eta: 0:00:32  lr: 0.006504  min_lr: 0.006504  loss: 3.6758 (3.5433)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0006  max mem: 55573
Epoch: [99]  [120/156]  eta: 0:00:25  lr: 0.006502  min_lr: 0.006502  loss: 3.6374 (3.5402)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [99]  [130/156]  eta: 0:00:17  lr: 0.006500  min_lr: 0.006500  loss: 3.7376 (3.5581)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0011  max mem: 55573
Epoch: [99]  [140/156]  eta: 0:00:10  lr: 0.006498  min_lr: 0.006498  loss: 3.8850 (3.5479)  weight_decay: 0.0500 (0.0500)  time: 0.5877  data: 0.0010  max mem: 55573
Epoch: [99]  [150/156]  eta: 0:00:04  lr: 0.006495  min_lr: 0.006495  loss: 3.1770 (3.5289)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0002  max mem: 55573
Epoch: [99]  [155/156]  eta: 0:00:00  lr: 0.006494  min_lr: 0.006494  loss: 3.4122 (3.5362)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0002  max mem: 55573
Epoch: [99] Total time: 0:01:46 (0.6799 s / it)
Averaged stats: lr: 0.006494  min_lr: 0.006494  loss: 3.4122 (3.5224)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.1976 (1.1976)  acc1: 75.0000 (75.0000)  acc5: 92.8385 (92.8385)  time: 7.1792  data: 6.9421  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2474 (1.2969)  acc1: 71.6146 (70.0800)  acc5: 92.8385 (90.1440)  time: 1.5823  data: 1.3885  max mem: 55573
Test: Total time: 0:00:08 (1.6210 s / it)
* Acc@1 70.694 Acc@5 90.342 loss 1.303
Accuracy of the model on the 50000 test images: 70.7%
Max accuracy: 72.07%
Test:  [0/5]  eta: 0:00:36  loss: 7.3476 (7.3476)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.2327  data: 6.9962  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.4597 (7.5317)  acc1: 0.0000 (0.3200)  acc5: 0.1302 (0.8640)  time: 1.5921  data: 1.3993  max mem: 55573
Test: Total time: 0:00:08 (1.6126 s / it)
* Acc@1 0.240 Acc@5 0.714 loss 7.502
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.24%
Epoch: [100]  [  0/156]  eta: 0:32:27  lr: 0.006494  min_lr: 0.006494  loss: 3.3762 (3.3762)  weight_decay: 0.0500 (0.0500)  time: 12.4850  data: 7.8882  max mem: 55573
Epoch: [100]  [ 10/156]  eta: 0:04:06  lr: 0.006492  min_lr: 0.006492  loss: 3.6403 (3.4975)  weight_decay: 0.0500 (0.0500)  time: 1.6898  data: 0.7178  max mem: 55573
Epoch: [100]  [ 20/156]  eta: 0:02:38  lr: 0.006490  min_lr: 0.006490  loss: 3.6925 (3.5727)  weight_decay: 0.0500 (0.0500)  time: 0.6001  data: 0.0007  max mem: 55573
Epoch: [100]  [ 30/156]  eta: 0:02:03  lr: 0.006487  min_lr: 0.006487  loss: 3.6977 (3.6218)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0006  max mem: 55573
Epoch: [100]  [ 40/156]  eta: 0:01:42  lr: 0.006485  min_lr: 0.006485  loss: 3.4689 (3.5829)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [100]  [ 50/156]  eta: 0:01:27  lr: 0.006483  min_lr: 0.006483  loss: 3.2825 (3.4997)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [100]  [ 60/156]  eta: 0:01:15  lr: 0.006481  min_lr: 0.006481  loss: 3.4548 (3.5374)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0006  max mem: 55573
Epoch: [100]  [ 70/156]  eta: 0:01:05  lr: 0.006478  min_lr: 0.006478  loss: 3.7065 (3.5362)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0006  max mem: 55573
Epoch: [100]  [ 80/156]  eta: 0:00:56  lr: 0.006476  min_lr: 0.006476  loss: 3.6290 (3.5359)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0005  max mem: 55573
Epoch: [100]  [ 90/156]  eta: 0:00:47  lr: 0.006474  min_lr: 0.006474  loss: 3.3552 (3.5093)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [100]  [100/156]  eta: 0:00:39  lr: 0.006472  min_lr: 0.006472  loss: 3.6226 (3.5151)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [100]  [110/156]  eta: 0:00:32  lr: 0.006469  min_lr: 0.006469  loss: 3.6848 (3.5271)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [100]  [120/156]  eta: 0:00:24  lr: 0.006467  min_lr: 0.006467  loss: 3.7371 (3.5391)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [100]  [130/156]  eta: 0:00:17  lr: 0.006465  min_lr: 0.006465  loss: 3.5852 (3.5242)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0012  max mem: 55573
Epoch: [100]  [140/156]  eta: 0:00:10  lr: 0.006463  min_lr: 0.006463  loss: 3.5852 (3.5242)  weight_decay: 0.0500 (0.0500)  time: 0.5852  data: 0.0010  max mem: 55573
Epoch: [100]  [150/156]  eta: 0:00:04  lr: 0.006460  min_lr: 0.006460  loss: 3.6374 (3.5336)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [100]  [155/156]  eta: 0:00:00  lr: 0.006459  min_lr: 0.006459  loss: 3.5079 (3.5334)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [100] Total time: 0:01:45 (0.6736 s / it)
Averaged stats: lr: 0.006459  min_lr: 0.006459  loss: 3.5079 (3.5223)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.2705 (1.2705)  acc1: 73.4375 (73.4375)  acc5: 92.3177 (92.3177)  time: 6.9885  data: 6.7516  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3398 (1.3622)  acc1: 73.1771 (69.5360)  acc5: 91.7969 (90.1120)  time: 1.5441  data: 1.3504  max mem: 55573
Test: Total time: 0:00:07 (1.5900 s / it)
* Acc@1 70.328 Acc@5 90.278 loss 1.363
Accuracy of the model on the 50000 test images: 70.3%
Max accuracy: 72.07%
Test:  [0/5]  eta: 0:00:35  loss: 7.3376 (7.3376)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.0920  data: 6.8560  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.4415 (7.5169)  acc1: 0.0000 (0.3200)  acc5: 0.2604 (0.9280)  time: 1.5638  data: 1.3713  max mem: 55573
Test: Total time: 0:00:07 (1.5864 s / it)
* Acc@1 0.256 Acc@5 0.766 loss 7.491
Accuracy of the model EMA on 50000 test images: 0.3%
Max EMA accuracy: 0.26%
Epoch: [101]  [  0/156]  eta: 0:29:35  lr: 0.006459  min_lr: 0.006459  loss: 3.3351 (3.3351)  weight_decay: 0.0500 (0.0500)  time: 11.3821  data: 8.1261  max mem: 55573
Epoch: [101]  [ 10/156]  eta: 0:03:55  lr: 0.006457  min_lr: 0.006457  loss: 3.5261 (3.3052)  weight_decay: 0.0500 (0.0500)  time: 1.6125  data: 0.7727  max mem: 55573
Epoch: [101]  [ 20/156]  eta: 0:02:33  lr: 0.006454  min_lr: 0.006454  loss: 3.7065 (3.4988)  weight_decay: 0.0500 (0.0500)  time: 0.6128  data: 0.0190  max mem: 55573
Epoch: [101]  [ 30/156]  eta: 0:02:00  lr: 0.006452  min_lr: 0.006452  loss: 3.7711 (3.4542)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0005  max mem: 55573
Epoch: [101]  [ 40/156]  eta: 0:01:40  lr: 0.006450  min_lr: 0.006450  loss: 3.7711 (3.4830)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0005  max mem: 55573
Epoch: [101]  [ 50/156]  eta: 0:01:26  lr: 0.006448  min_lr: 0.006448  loss: 3.6455 (3.4787)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [101]  [ 60/156]  eta: 0:01:14  lr: 0.006445  min_lr: 0.006445  loss: 3.5447 (3.4934)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0006  max mem: 55573
Epoch: [101]  [ 70/156]  eta: 0:01:04  lr: 0.006443  min_lr: 0.006443  loss: 3.5005 (3.5085)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0006  max mem: 55573
Epoch: [101]  [ 80/156]  eta: 0:00:55  lr: 0.006441  min_lr: 0.006441  loss: 3.5610 (3.5342)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0006  max mem: 55573
Epoch: [101]  [ 90/156]  eta: 0:00:47  lr: 0.006438  min_lr: 0.006438  loss: 3.7379 (3.5507)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0006  max mem: 55573
Epoch: [101]  [100/156]  eta: 0:00:39  lr: 0.006436  min_lr: 0.006436  loss: 3.7007 (3.5354)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0007  max mem: 55573
Epoch: [101]  [110/156]  eta: 0:00:31  lr: 0.006434  min_lr: 0.006434  loss: 3.6743 (3.5473)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0006  max mem: 55573
Epoch: [101]  [120/156]  eta: 0:00:24  lr: 0.006432  min_lr: 0.006432  loss: 3.7842 (3.5626)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0008  max mem: 55573
Epoch: [101]  [130/156]  eta: 0:00:17  lr: 0.006429  min_lr: 0.006429  loss: 3.7842 (3.5670)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0014  max mem: 55573
Epoch: [101]  [140/156]  eta: 0:00:10  lr: 0.006427  min_lr: 0.006427  loss: 3.7472 (3.5703)  weight_decay: 0.0500 (0.0500)  time: 0.5851  data: 0.0010  max mem: 55573
Epoch: [101]  [150/156]  eta: 0:00:03  lr: 0.006425  min_lr: 0.006425  loss: 3.7472 (3.5846)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0002  max mem: 55573
Epoch: [101]  [155/156]  eta: 0:00:00  lr: 0.006424  min_lr: 0.006424  loss: 3.5546 (3.5642)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0002  max mem: 55573
Epoch: [101] Total time: 0:01:44 (0.6687 s / it)
Averaged stats: lr: 0.006424  min_lr: 0.006424  loss: 3.5546 (3.5207)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.2763 (1.2763)  acc1: 74.6094 (74.6094)  acc5: 92.8385 (92.8385)  time: 6.9669  data: 6.7294  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3521 (1.3851)  acc1: 70.1823 (69.2800)  acc5: 90.5660 (90.6240)  time: 1.5400  data: 1.3460  max mem: 55573
Test: Total time: 0:00:07 (1.5750 s / it)
* Acc@1 70.650 Acc@5 90.552 loss 1.351
Accuracy of the model on the 50000 test images: 70.7%
Max accuracy: 72.07%
Test:  [0/5]  eta: 0:00:37  loss: 7.3227 (7.3227)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 7.4686  data: 7.2327  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.4190 (7.4965)  acc1: 0.0000 (0.3520)  acc5: 0.1302 (0.9280)  time: 1.6390  data: 1.4467  max mem: 55573
Test: Total time: 0:00:08 (1.6618 s / it)
* Acc@1 0.258 Acc@5 0.820 loss 7.476
Accuracy of the model EMA on 50000 test images: 0.3%
Max EMA accuracy: 0.26%
Epoch: [102]  [  0/156]  eta: 0:34:24  lr: 0.006423  min_lr: 0.006423  loss: 3.8816 (3.8816)  weight_decay: 0.0500 (0.0500)  time: 13.2336  data: 7.9062  max mem: 55573
Epoch: [102]  [ 10/156]  eta: 0:04:20  lr: 0.006421  min_lr: 0.006421  loss: 3.5478 (3.4532)  weight_decay: 0.0500 (0.0500)  time: 1.7866  data: 0.7253  max mem: 55573
Epoch: [102]  [ 20/156]  eta: 0:02:45  lr: 0.006419  min_lr: 0.006419  loss: 3.0804 (3.2489)  weight_decay: 0.0500 (0.0500)  time: 0.6155  data: 0.0039  max mem: 55573
Epoch: [102]  [ 30/156]  eta: 0:02:07  lr: 0.006416  min_lr: 0.006416  loss: 3.3809 (3.3497)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0007  max mem: 55573
Epoch: [102]  [ 40/156]  eta: 0:01:45  lr: 0.006414  min_lr: 0.006414  loss: 3.5204 (3.3437)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0008  max mem: 55573
Epoch: [102]  [ 50/156]  eta: 0:01:29  lr: 0.006412  min_lr: 0.006412  loss: 3.4934 (3.3699)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0008  max mem: 55573
Epoch: [102]  [ 60/156]  eta: 0:01:17  lr: 0.006410  min_lr: 0.006410  loss: 3.7495 (3.3856)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0007  max mem: 55573
Epoch: [102]  [ 70/156]  eta: 0:01:06  lr: 0.006407  min_lr: 0.006407  loss: 3.6624 (3.3811)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [102]  [ 80/156]  eta: 0:00:57  lr: 0.006405  min_lr: 0.006405  loss: 3.2780 (3.3749)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0006  max mem: 55573
Epoch: [102]  [ 90/156]  eta: 0:00:48  lr: 0.006403  min_lr: 0.006403  loss: 3.4601 (3.4085)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [102]  [100/156]  eta: 0:00:40  lr: 0.006400  min_lr: 0.006400  loss: 3.6891 (3.4046)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [102]  [110/156]  eta: 0:00:32  lr: 0.006398  min_lr: 0.006398  loss: 3.5596 (3.4198)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0006  max mem: 55573
Epoch: [102]  [120/156]  eta: 0:00:25  lr: 0.006396  min_lr: 0.006396  loss: 3.5862 (3.4181)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0007  max mem: 55573
Epoch: [102]  [130/156]  eta: 0:00:17  lr: 0.006393  min_lr: 0.006393  loss: 3.5894 (3.4277)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0013  max mem: 55573
Epoch: [102]  [140/156]  eta: 0:00:10  lr: 0.006391  min_lr: 0.006391  loss: 3.6785 (3.4328)  weight_decay: 0.0500 (0.0500)  time: 0.5853  data: 0.0010  max mem: 55573
Epoch: [102]  [150/156]  eta: 0:00:04  lr: 0.006389  min_lr: 0.006389  loss: 3.5058 (3.4307)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [102]  [155/156]  eta: 0:00:00  lr: 0.006388  min_lr: 0.006388  loss: 3.5058 (3.4363)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [102] Total time: 0:01:46 (0.6802 s / it)
Averaged stats: lr: 0.006388  min_lr: 0.006388  loss: 3.5058 (3.5319)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.3337 (1.3337)  acc1: 75.7812 (75.7812)  acc5: 93.7500 (93.7500)  time: 6.8825  data: 6.6455  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3914 (1.4123)  acc1: 71.3542 (70.4960)  acc5: 92.4528 (91.2640)  time: 1.5231  data: 1.3292  max mem: 55573
Test: Total time: 0:00:07 (1.5704 s / it)
* Acc@1 71.728 Acc@5 90.988 loss 1.397
Accuracy of the model on the 50000 test images: 71.7%
Max accuracy: 72.07%
Test:  [0/5]  eta: 0:00:32  loss: 7.3012 (7.3012)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.4052  data: 6.1690  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.3857 (7.4675)  acc1: 0.0000 (0.3520)  acc5: 0.2604 (0.9920)  time: 1.4943  data: 1.3018  max mem: 55573
Test: Total time: 0:00:07 (1.5175 s / it)
* Acc@1 0.290 Acc@5 0.882 loss 7.454
Accuracy of the model EMA on 50000 test images: 0.3%
Max EMA accuracy: 0.29%
Epoch: [103]  [  0/156]  eta: 0:35:48  lr: 0.006388  min_lr: 0.006388  loss: 2.5335 (2.5335)  weight_decay: 0.0500 (0.0500)  time: 13.7732  data: 13.1568  max mem: 55573
Epoch: [103]  [ 10/156]  eta: 0:04:22  lr: 0.006385  min_lr: 0.006385  loss: 3.8120 (3.6249)  weight_decay: 0.0500 (0.0500)  time: 1.7957  data: 1.1964  max mem: 55573
Epoch: [103]  [ 20/156]  eta: 0:02:46  lr: 0.006383  min_lr: 0.006383  loss: 3.7687 (3.5986)  weight_decay: 0.0500 (0.0500)  time: 0.5932  data: 0.0004  max mem: 55573
Epoch: [103]  [ 30/156]  eta: 0:02:08  lr: 0.006381  min_lr: 0.006381  loss: 3.6850 (3.5119)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [103]  [ 40/156]  eta: 0:01:46  lr: 0.006378  min_lr: 0.006378  loss: 3.4322 (3.4836)  weight_decay: 0.0500 (0.0500)  time: 0.5937  data: 0.0005  max mem: 55573
Epoch: [103]  [ 50/156]  eta: 0:01:30  lr: 0.006376  min_lr: 0.006376  loss: 3.4491 (3.5038)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0005  max mem: 55573
Epoch: [103]  [ 60/156]  eta: 0:01:17  lr: 0.006374  min_lr: 0.006374  loss: 3.4885 (3.5040)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [103]  [ 70/156]  eta: 0:01:06  lr: 0.006371  min_lr: 0.006371  loss: 3.7830 (3.5356)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0006  max mem: 55573
Epoch: [103]  [ 80/156]  eta: 0:00:57  lr: 0.006369  min_lr: 0.006369  loss: 3.8508 (3.5531)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0006  max mem: 55573
Epoch: [103]  [ 90/156]  eta: 0:00:48  lr: 0.006367  min_lr: 0.006367  loss: 3.6506 (3.5572)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [103]  [100/156]  eta: 0:00:40  lr: 0.006364  min_lr: 0.006364  loss: 3.6506 (3.5480)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0006  max mem: 55573
Epoch: [103]  [110/156]  eta: 0:00:32  lr: 0.006362  min_lr: 0.006362  loss: 3.5826 (3.5238)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0006  max mem: 55573
Epoch: [103]  [120/156]  eta: 0:00:25  lr: 0.006360  min_lr: 0.006360  loss: 3.3605 (3.5130)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0006  max mem: 55573
Epoch: [103]  [130/156]  eta: 0:00:17  lr: 0.006357  min_lr: 0.006357  loss: 3.4195 (3.5149)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0015  max mem: 55573
Epoch: [103]  [140/156]  eta: 0:00:10  lr: 0.006355  min_lr: 0.006355  loss: 3.7763 (3.5406)  weight_decay: 0.0500 (0.0500)  time: 0.5866  data: 0.0012  max mem: 55573
Epoch: [103]  [150/156]  eta: 0:00:04  lr: 0.006353  min_lr: 0.006353  loss: 3.7040 (3.5241)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [103]  [155/156]  eta: 0:00:00  lr: 0.006352  min_lr: 0.006352  loss: 3.7007 (3.5292)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [103] Total time: 0:01:46 (0.6813 s / it)
Averaged stats: lr: 0.006352  min_lr: 0.006352  loss: 3.7007 (3.5160)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.2553 (1.2553)  acc1: 77.3438 (77.3438)  acc5: 94.2708 (94.2708)  time: 6.8774  data: 6.6405  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3405 (1.3897)  acc1: 74.6094 (71.9360)  acc5: 93.6198 (91.3600)  time: 1.5219  data: 1.3282  max mem: 55573
Test: Total time: 0:00:07 (1.5574 s / it)
* Acc@1 72.216 Acc@5 91.356 loss 1.375
Accuracy of the model on the 50000 test images: 72.2%
Max accuracy: 72.22%
Test:  [0/5]  eta: 0:00:33  loss: 7.2698 (7.2698)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.7807  data: 6.5444  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.3500 (7.4275)  acc1: 0.0000 (0.3520)  acc5: 0.2604 (1.1200)  time: 1.5015  data: 1.3090  max mem: 55573
Test: Total time: 0:00:07 (1.5189 s / it)
* Acc@1 0.310 Acc@5 0.956 loss 7.425
Accuracy of the model EMA on 50000 test images: 0.3%
Max EMA accuracy: 0.31%
Epoch: [104]  [  0/156]  eta: 0:33:31  lr: 0.006351  min_lr: 0.006351  loss: 3.5156 (3.5156)  weight_decay: 0.0500 (0.0500)  time: 12.8968  data: 8.5858  max mem: 55573
Epoch: [104]  [ 10/156]  eta: 0:04:11  lr: 0.006349  min_lr: 0.006349  loss: 3.5926 (3.4568)  weight_decay: 0.0500 (0.0500)  time: 1.7229  data: 0.7812  max mem: 55573
Epoch: [104]  [ 20/156]  eta: 0:02:40  lr: 0.006347  min_lr: 0.006347  loss: 3.5462 (3.4544)  weight_decay: 0.0500 (0.0500)  time: 0.5976  data: 0.0007  max mem: 55573
Epoch: [104]  [ 30/156]  eta: 0:02:05  lr: 0.006344  min_lr: 0.006344  loss: 3.3159 (3.4129)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0006  max mem: 55573
Epoch: [104]  [ 40/156]  eta: 0:01:43  lr: 0.006342  min_lr: 0.006342  loss: 3.4221 (3.4733)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0006  max mem: 55573
Epoch: [104]  [ 50/156]  eta: 0:01:28  lr: 0.006340  min_lr: 0.006340  loss: 3.6057 (3.4326)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0007  max mem: 55573
Epoch: [104]  [ 60/156]  eta: 0:01:16  lr: 0.006337  min_lr: 0.006337  loss: 3.6057 (3.4571)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [104]  [ 70/156]  eta: 0:01:05  lr: 0.006335  min_lr: 0.006335  loss: 3.4079 (3.4557)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0006  max mem: 55573
Epoch: [104]  [ 80/156]  eta: 0:00:56  lr: 0.006333  min_lr: 0.006333  loss: 3.3932 (3.4422)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0006  max mem: 55573
Epoch: [104]  [ 90/156]  eta: 0:00:48  lr: 0.006330  min_lr: 0.006330  loss: 3.6053 (3.4657)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [104]  [100/156]  eta: 0:00:39  lr: 0.006328  min_lr: 0.006328  loss: 3.6538 (3.4743)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [104]  [110/156]  eta: 0:00:32  lr: 0.006326  min_lr: 0.006326  loss: 3.5994 (3.4770)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [104]  [120/156]  eta: 0:00:24  lr: 0.006323  min_lr: 0.006323  loss: 3.6913 (3.4681)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [104]  [130/156]  eta: 0:00:17  lr: 0.006321  min_lr: 0.006321  loss: 3.2643 (3.4512)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0014  max mem: 55573
Epoch: [104]  [140/156]  eta: 0:00:10  lr: 0.006319  min_lr: 0.006319  loss: 3.5667 (3.4554)  weight_decay: 0.0500 (0.0500)  time: 0.5862  data: 0.0013  max mem: 55573
Epoch: [104]  [150/156]  eta: 0:00:04  lr: 0.006316  min_lr: 0.006316  loss: 3.6679 (3.4570)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [104]  [155/156]  eta: 0:00:00  lr: 0.006315  min_lr: 0.006315  loss: 3.6631 (3.4646)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [104] Total time: 0:01:45 (0.6757 s / it)
Averaged stats: lr: 0.006315  min_lr: 0.006315  loss: 3.6631 (3.5108)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.1745 (1.1745)  acc1: 76.9531 (76.9531)  acc5: 94.7917 (94.7917)  time: 6.8139  data: 6.5770  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3722 (1.3575)  acc1: 72.6562 (71.2000)  acc5: 93.0990 (91.1680)  time: 1.5102  data: 1.3158  max mem: 55573
Test: Total time: 0:00:07 (1.5505 s / it)
* Acc@1 72.166 Acc@5 91.242 loss 1.354
Accuracy of the model on the 50000 test images: 72.2%
Max accuracy: 72.22%
Test:  [0/5]  eta: 0:00:33  loss: 7.2293 (7.2293)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 6.7004  data: 6.4643  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.3143 (7.3783)  acc1: 0.0000 (0.3840)  acc5: 0.2604 (1.2160)  time: 1.4854  data: 1.2930  max mem: 55573
Test: Total time: 0:00:07 (1.5076 s / it)
* Acc@1 0.330 Acc@5 1.038 loss 7.390
Accuracy of the model EMA on 50000 test images: 0.3%
Max EMA accuracy: 0.33%
Epoch: [105]  [  0/156]  eta: 0:31:05  lr: 0.006315  min_lr: 0.006315  loss: 2.4087 (2.4087)  weight_decay: 0.0500 (0.0500)  time: 11.9611  data: 10.3034  max mem: 55573
Epoch: [105]  [ 10/156]  eta: 0:04:10  lr: 0.006313  min_lr: 0.006313  loss: 3.5131 (3.3917)  weight_decay: 0.0500 (0.0500)  time: 1.7162  data: 0.9372  max mem: 55573
Epoch: [105]  [ 20/156]  eta: 0:02:40  lr: 0.006310  min_lr: 0.006310  loss: 3.5131 (3.3838)  weight_decay: 0.0500 (0.0500)  time: 0.6409  data: 0.0006  max mem: 55573
Epoch: [105]  [ 30/156]  eta: 0:02:04  lr: 0.006308  min_lr: 0.006308  loss: 3.5275 (3.4647)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [105]  [ 40/156]  eta: 0:01:43  lr: 0.006306  min_lr: 0.006306  loss: 3.4917 (3.4377)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [105]  [ 50/156]  eta: 0:01:28  lr: 0.006303  min_lr: 0.006303  loss: 3.4061 (3.4620)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [105]  [ 60/156]  eta: 0:01:16  lr: 0.006301  min_lr: 0.006301  loss: 3.6771 (3.5071)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [105]  [ 70/156]  eta: 0:01:05  lr: 0.006298  min_lr: 0.006298  loss: 3.6771 (3.5068)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0006  max mem: 55573
Epoch: [105]  [ 80/156]  eta: 0:00:56  lr: 0.006296  min_lr: 0.006296  loss: 3.5218 (3.4846)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [105]  [ 90/156]  eta: 0:00:47  lr: 0.006294  min_lr: 0.006294  loss: 3.5135 (3.4837)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [105]  [100/156]  eta: 0:00:39  lr: 0.006291  min_lr: 0.006291  loss: 3.5239 (3.4830)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0006  max mem: 55573
Epoch: [105]  [110/156]  eta: 0:00:32  lr: 0.006289  min_lr: 0.006289  loss: 3.5875 (3.4790)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [105]  [120/156]  eta: 0:00:24  lr: 0.006287  min_lr: 0.006287  loss: 3.7452 (3.5027)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [105]  [130/156]  eta: 0:00:17  lr: 0.006284  min_lr: 0.006284  loss: 3.7584 (3.5139)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0015  max mem: 55573
Epoch: [105]  [140/156]  eta: 0:00:10  lr: 0.006282  min_lr: 0.006282  loss: 3.7499 (3.5156)  weight_decay: 0.0500 (0.0500)  time: 0.5856  data: 0.0013  max mem: 55573
Epoch: [105]  [150/156]  eta: 0:00:04  lr: 0.006280  min_lr: 0.006280  loss: 3.6516 (3.5100)  weight_decay: 0.0500 (0.0500)  time: 0.5812  data: 0.0001  max mem: 55573
Epoch: [105]  [155/156]  eta: 0:00:00  lr: 0.006278  min_lr: 0.006278  loss: 3.1556 (3.4988)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [105] Total time: 0:01:45 (0.6755 s / it)
Averaged stats: lr: 0.006278  min_lr: 0.006278  loss: 3.1556 (3.4976)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:37  loss: 1.0753 (1.0753)  acc1: 76.9531 (76.9531)  acc5: 94.0104 (94.0104)  time: 7.4464  data: 7.2090  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2284 (1.2641)  acc1: 74.6094 (71.8400)  acc5: 92.8385 (90.7200)  time: 1.6357  data: 1.4419  max mem: 55573
Test: Total time: 0:00:08 (1.6749 s / it)
* Acc@1 72.370 Acc@5 91.456 loss 1.254
Accuracy of the model on the 50000 test images: 72.4%
Max accuracy: 72.37%
Test:  [0/5]  eta: 0:00:34  loss: 7.1751 (7.1751)  acc1: 0.0000 (0.0000)  acc5: 0.1302 (0.1302)  time: 6.9113  data: 6.6752  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.2703 (7.3199)  acc1: 0.0000 (0.4160)  acc5: 1.8868 (1.5040)  time: 1.5320  data: 1.3396  max mem: 55573
Test: Total time: 0:00:07 (1.5533 s / it)
* Acc@1 0.374 Acc@5 1.156 loss 7.347
Accuracy of the model EMA on 50000 test images: 0.4%
Max EMA accuracy: 0.37%
Epoch: [106]  [  0/156]  eta: 0:30:55  lr: 0.006278  min_lr: 0.006278  loss: 3.6701 (3.6701)  weight_decay: 0.0500 (0.0500)  time: 11.8971  data: 7.6838  max mem: 55573
Epoch: [106]  [ 10/156]  eta: 0:03:56  lr: 0.006276  min_lr: 0.006276  loss: 3.6207 (3.4223)  weight_decay: 0.0500 (0.0500)  time: 1.6227  data: 0.6991  max mem: 55573
Epoch: [106]  [ 20/156]  eta: 0:02:33  lr: 0.006273  min_lr: 0.006273  loss: 3.3305 (3.2914)  weight_decay: 0.0500 (0.0500)  time: 0.5940  data: 0.0005  max mem: 55573
Epoch: [106]  [ 30/156]  eta: 0:02:00  lr: 0.006271  min_lr: 0.006271  loss: 3.3941 (3.3725)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [106]  [ 40/156]  eta: 0:01:40  lr: 0.006269  min_lr: 0.006269  loss: 3.6420 (3.4124)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [106]  [ 50/156]  eta: 0:01:26  lr: 0.006266  min_lr: 0.006266  loss: 3.8589 (3.4873)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [106]  [ 60/156]  eta: 0:01:14  lr: 0.006264  min_lr: 0.006264  loss: 3.7894 (3.4574)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0006  max mem: 55573
Epoch: [106]  [ 70/156]  eta: 0:01:04  lr: 0.006262  min_lr: 0.006262  loss: 3.3807 (3.4650)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0006  max mem: 55573
Epoch: [106]  [ 80/156]  eta: 0:00:55  lr: 0.006259  min_lr: 0.006259  loss: 3.5251 (3.4659)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0006  max mem: 55573
Epoch: [106]  [ 90/156]  eta: 0:00:47  lr: 0.006257  min_lr: 0.006257  loss: 3.6513 (3.4702)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0008  max mem: 55573
Epoch: [106]  [100/156]  eta: 0:00:39  lr: 0.006254  min_lr: 0.006254  loss: 3.6179 (3.4735)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0007  max mem: 55573
Epoch: [106]  [110/156]  eta: 0:00:31  lr: 0.006252  min_lr: 0.006252  loss: 3.7364 (3.4829)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0006  max mem: 55573
Epoch: [106]  [120/156]  eta: 0:00:24  lr: 0.006250  min_lr: 0.006250  loss: 3.7838 (3.4741)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0006  max mem: 55573
Epoch: [106]  [130/156]  eta: 0:00:17  lr: 0.006247  min_lr: 0.006247  loss: 3.6624 (3.4723)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0010  max mem: 55573
Epoch: [106]  [140/156]  eta: 0:00:10  lr: 0.006245  min_lr: 0.006245  loss: 3.4453 (3.4673)  weight_decay: 0.0500 (0.0500)  time: 0.5870  data: 0.0009  max mem: 55573
Epoch: [106]  [150/156]  eta: 0:00:03  lr: 0.006243  min_lr: 0.006243  loss: 3.3972 (3.4726)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0002  max mem: 55573
Epoch: [106]  [155/156]  eta: 0:00:00  lr: 0.006241  min_lr: 0.006241  loss: 3.3972 (3.4746)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0002  max mem: 55573
Epoch: [106] Total time: 0:01:44 (0.6694 s / it)
Averaged stats: lr: 0.006241  min_lr: 0.006241  loss: 3.3972 (3.5046)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.1263 (1.1263)  acc1: 77.3438 (77.3438)  acc5: 94.9219 (94.9219)  time: 6.8838  data: 6.6454  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2711 (1.2394)  acc1: 72.7865 (72.0320)  acc5: 92.8385 (90.8160)  time: 1.5234  data: 1.3292  max mem: 55573
Test: Total time: 0:00:07 (1.5619 s / it)
* Acc@1 72.238 Acc@5 91.160 loss 1.253
Accuracy of the model on the 50000 test images: 72.2%
Max accuracy: 72.37%
Test:  [0/5]  eta: 0:00:36  loss: 7.1005 (7.1005)  acc1: 0.1302 (0.1302)  acc5: 0.3906 (0.3906)  time: 7.3814  data: 7.1455  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.2109 (7.2461)  acc1: 0.1302 (0.5120)  acc5: 1.9531 (1.7280)  time: 1.6216  data: 1.4292  max mem: 55573
Test: Total time: 0:00:08 (1.6406 s / it)
* Acc@1 0.430 Acc@5 1.352 loss 7.293
Accuracy of the model EMA on 50000 test images: 0.4%
Max EMA accuracy: 0.43%
Epoch: [107]  [  0/156]  eta: 0:35:21  lr: 0.006241  min_lr: 0.006241  loss: 4.2462 (4.2462)  weight_decay: 0.0500 (0.0500)  time: 13.5977  data: 12.9786  max mem: 55573
Epoch: [107]  [ 10/156]  eta: 0:04:18  lr: 0.006239  min_lr: 0.006239  loss: 3.8584 (3.7923)  weight_decay: 0.0500 (0.0500)  time: 1.7736  data: 1.1804  max mem: 55573
Epoch: [107]  [ 20/156]  eta: 0:02:44  lr: 0.006236  min_lr: 0.006236  loss: 3.7035 (3.7247)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [107]  [ 30/156]  eta: 0:02:07  lr: 0.006234  min_lr: 0.006234  loss: 3.5193 (3.5622)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [107]  [ 40/156]  eta: 0:01:45  lr: 0.006232  min_lr: 0.006232  loss: 3.3467 (3.5187)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [107]  [ 50/156]  eta: 0:01:29  lr: 0.006229  min_lr: 0.006229  loss: 3.5820 (3.5197)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [107]  [ 60/156]  eta: 0:01:17  lr: 0.006227  min_lr: 0.006227  loss: 3.6006 (3.5484)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0007  max mem: 55573
Epoch: [107]  [ 70/156]  eta: 0:01:06  lr: 0.006224  min_lr: 0.006224  loss: 3.7053 (3.5647)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0007  max mem: 55573
Epoch: [107]  [ 80/156]  eta: 0:00:57  lr: 0.006222  min_lr: 0.006222  loss: 3.7793 (3.6043)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [107]  [ 90/156]  eta: 0:00:48  lr: 0.006220  min_lr: 0.006220  loss: 3.7074 (3.5915)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [107]  [100/156]  eta: 0:00:40  lr: 0.006217  min_lr: 0.006217  loss: 3.1775 (3.5450)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0005  max mem: 55573
Epoch: [107]  [110/156]  eta: 0:00:32  lr: 0.006215  min_lr: 0.006215  loss: 3.1775 (3.5355)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [107]  [120/156]  eta: 0:00:25  lr: 0.006212  min_lr: 0.006212  loss: 3.6572 (3.5183)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [107]  [130/156]  eta: 0:00:17  lr: 0.006210  min_lr: 0.006210  loss: 3.4951 (3.5164)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0013  max mem: 55573
Epoch: [107]  [140/156]  eta: 0:00:10  lr: 0.006208  min_lr: 0.006208  loss: 3.7711 (3.5284)  weight_decay: 0.0500 (0.0500)  time: 0.5873  data: 0.0011  max mem: 55573
Epoch: [107]  [150/156]  eta: 0:00:04  lr: 0.006205  min_lr: 0.006205  loss: 3.7711 (3.5362)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [107]  [155/156]  eta: 0:00:00  lr: 0.006204  min_lr: 0.006204  loss: 3.6791 (3.5406)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [107] Total time: 0:01:45 (0.6794 s / it)
Averaged stats: lr: 0.006204  min_lr: 0.006204  loss: 3.6791 (3.4990)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.1480 (1.1480)  acc1: 74.7396 (74.7396)  acc5: 93.8802 (93.8802)  time: 7.2441  data: 7.0068  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3458 (1.3813)  acc1: 71.3542 (70.4320)  acc5: 91.7969 (90.8480)  time: 1.5952  data: 1.4015  max mem: 55573
Test: Total time: 0:00:08 (1.6377 s / it)
* Acc@1 71.804 Acc@5 91.106 loss 1.358
Accuracy of the model on the 50000 test images: 71.8%
Max accuracy: 72.37%
Test:  [0/5]  eta: 0:00:35  loss: 7.0062 (7.0062)  acc1: 0.1302 (0.1302)  acc5: 0.6510 (0.6510)  time: 7.0318  data: 6.7957  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.1391 (7.1587)  acc1: 0.5208 (0.6400)  acc5: 2.3438 (2.0160)  time: 1.5517  data: 1.3592  max mem: 55573
Test: Total time: 0:00:07 (1.5686 s / it)
* Acc@1 0.520 Acc@5 1.580 loss 7.228
Accuracy of the model EMA on 50000 test images: 0.5%
Max EMA accuracy: 0.52%
Epoch: [108]  [  0/156]  eta: 0:35:26  lr: 0.006204  min_lr: 0.006204  loss: 4.0922 (4.0922)  weight_decay: 0.0500 (0.0500)  time: 13.6342  data: 10.0179  max mem: 55573
Epoch: [108]  [ 10/156]  eta: 0:04:19  lr: 0.006201  min_lr: 0.006201  loss: 3.4284 (3.3607)  weight_decay: 0.0500 (0.0500)  time: 1.7765  data: 0.9113  max mem: 55573
Epoch: [108]  [ 20/156]  eta: 0:02:44  lr: 0.006199  min_lr: 0.006199  loss: 3.6271 (3.5210)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [108]  [ 30/156]  eta: 0:02:07  lr: 0.006197  min_lr: 0.006197  loss: 3.6490 (3.5155)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [108]  [ 40/156]  eta: 0:01:45  lr: 0.006194  min_lr: 0.006194  loss: 3.5654 (3.4950)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [108]  [ 50/156]  eta: 0:01:29  lr: 0.006192  min_lr: 0.006192  loss: 3.5654 (3.4919)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [108]  [ 60/156]  eta: 0:01:17  lr: 0.006189  min_lr: 0.006189  loss: 3.3624 (3.4400)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [108]  [ 70/156]  eta: 0:01:06  lr: 0.006187  min_lr: 0.006187  loss: 3.6888 (3.4774)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0005  max mem: 55573
Epoch: [108]  [ 80/156]  eta: 0:00:57  lr: 0.006185  min_lr: 0.006185  loss: 3.6888 (3.4607)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [108]  [ 90/156]  eta: 0:00:48  lr: 0.006182  min_lr: 0.006182  loss: 3.3513 (3.4393)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0005  max mem: 55573
Epoch: [108]  [100/156]  eta: 0:00:40  lr: 0.006180  min_lr: 0.006180  loss: 3.4852 (3.4656)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0006  max mem: 55573
Epoch: [108]  [110/156]  eta: 0:00:32  lr: 0.006177  min_lr: 0.006177  loss: 3.8160 (3.4780)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0007  max mem: 55573
Epoch: [108]  [120/156]  eta: 0:00:25  lr: 0.006175  min_lr: 0.006175  loss: 3.5404 (3.4650)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0006  max mem: 55573
Epoch: [108]  [130/156]  eta: 0:00:17  lr: 0.006173  min_lr: 0.006173  loss: 3.5972 (3.4770)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0011  max mem: 55573
Epoch: [108]  [140/156]  eta: 0:00:10  lr: 0.006170  min_lr: 0.006170  loss: 3.5960 (3.4646)  weight_decay: 0.0500 (0.0500)  time: 0.5865  data: 0.0009  max mem: 55573
Epoch: [108]  [150/156]  eta: 0:00:04  lr: 0.006168  min_lr: 0.006168  loss: 3.4850 (3.4737)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [108]  [155/156]  eta: 0:00:00  lr: 0.006166  min_lr: 0.006166  loss: 3.6677 (3.4794)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [108] Total time: 0:01:46 (0.6799 s / it)
Averaged stats: lr: 0.006166  min_lr: 0.006166  loss: 3.6677 (3.5066)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.1228 (1.1228)  acc1: 74.3490 (74.3490)  acc5: 94.2708 (94.2708)  time: 6.8862  data: 6.6483  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1772 (1.2026)  acc1: 72.3958 (70.4320)  acc5: 93.0990 (90.8800)  time: 1.5238  data: 1.3297  max mem: 55573
Test: Total time: 0:00:07 (1.5576 s / it)
* Acc@1 71.740 Acc@5 91.084 loss 1.193
Accuracy of the model on the 50000 test images: 71.7%
Max accuracy: 72.37%
Test:  [0/5]  eta: 0:00:34  loss: 6.8944 (6.8944)  acc1: 0.5208 (0.5208)  acc5: 1.3021 (1.3021)  time: 6.9439  data: 6.7077  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 7.0528 (7.0558)  acc1: 0.7812 (0.8640)  acc5: 2.9948 (2.7840)  time: 1.5341  data: 1.3417  max mem: 55573
Test: Total time: 0:00:07 (1.5579 s / it)
* Acc@1 0.654 Acc@5 1.942 loss 7.150
Accuracy of the model EMA on 50000 test images: 0.7%
Max EMA accuracy: 0.65%
Epoch: [109]  [  0/156]  eta: 0:33:07  lr: 0.006166  min_lr: 0.006166  loss: 3.6316 (3.6316)  weight_decay: 0.0500 (0.0500)  time: 12.7421  data: 11.8819  max mem: 55573
Epoch: [109]  [ 10/156]  eta: 0:04:07  lr: 0.006164  min_lr: 0.006164  loss: 3.5888 (3.3142)  weight_decay: 0.0500 (0.0500)  time: 1.6983  data: 1.0805  max mem: 55573
Epoch: [109]  [ 20/156]  eta: 0:02:39  lr: 0.006161  min_lr: 0.006161  loss: 3.2037 (3.3024)  weight_decay: 0.0500 (0.0500)  time: 0.5936  data: 0.0004  max mem: 55573
Epoch: [109]  [ 30/156]  eta: 0:02:04  lr: 0.006159  min_lr: 0.006159  loss: 3.2037 (3.2925)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [109]  [ 40/156]  eta: 0:01:43  lr: 0.006157  min_lr: 0.006157  loss: 3.4766 (3.3518)  weight_decay: 0.0500 (0.0500)  time: 0.5935  data: 0.0004  max mem: 55573
Epoch: [109]  [ 50/156]  eta: 0:01:28  lr: 0.006154  min_lr: 0.006154  loss: 3.5784 (3.3439)  weight_decay: 0.0500 (0.0500)  time: 0.5936  data: 0.0004  max mem: 55573
Epoch: [109]  [ 60/156]  eta: 0:01:16  lr: 0.006152  min_lr: 0.006152  loss: 3.3691 (3.3677)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [109]  [ 70/156]  eta: 0:01:05  lr: 0.006149  min_lr: 0.006149  loss: 3.6415 (3.3905)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [109]  [ 80/156]  eta: 0:00:56  lr: 0.006147  min_lr: 0.006147  loss: 3.6714 (3.4153)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [109]  [ 90/156]  eta: 0:00:47  lr: 0.006144  min_lr: 0.006144  loss: 3.6412 (3.4229)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0005  max mem: 55573
Epoch: [109]  [100/156]  eta: 0:00:39  lr: 0.006142  min_lr: 0.006142  loss: 3.6624 (3.4422)  weight_decay: 0.0500 (0.0500)  time: 0.5931  data: 0.0004  max mem: 55573
Epoch: [109]  [110/156]  eta: 0:00:32  lr: 0.006140  min_lr: 0.006140  loss: 3.6624 (3.4495)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [109]  [120/156]  eta: 0:00:24  lr: 0.006137  min_lr: 0.006137  loss: 3.6478 (3.4688)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [109]  [130/156]  eta: 0:00:17  lr: 0.006135  min_lr: 0.006135  loss: 3.7436 (3.4734)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0012  max mem: 55573
Epoch: [109]  [140/156]  eta: 0:00:10  lr: 0.006132  min_lr: 0.006132  loss: 3.5500 (3.4621)  weight_decay: 0.0500 (0.0500)  time: 0.5867  data: 0.0010  max mem: 55573
Epoch: [109]  [150/156]  eta: 0:00:04  lr: 0.006130  min_lr: 0.006130  loss: 3.5620 (3.4744)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [109]  [155/156]  eta: 0:00:00  lr: 0.006129  min_lr: 0.006129  loss: 3.8341 (3.4858)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [109] Total time: 0:01:45 (0.6746 s / it)
Averaged stats: lr: 0.006129  min_lr: 0.006129  loss: 3.8341 (3.4953)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.1619 (1.1619)  acc1: 77.0833 (77.0833)  acc5: 93.0990 (93.0990)  time: 7.3279  data: 7.0905  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2805 (1.2872)  acc1: 74.7396 (71.4880)  acc5: 93.0990 (90.7520)  time: 1.6121  data: 1.4182  max mem: 55573
Test: Total time: 0:00:08 (1.6533 s / it)
* Acc@1 72.108 Acc@5 91.126 loss 1.281
Accuracy of the model on the 50000 test images: 72.1%
Max accuracy: 72.37%
Test:  [0/5]  eta: 0:00:37  loss: 6.7696 (6.7696)  acc1: 0.9115 (0.9115)  acc5: 1.8229 (1.8229)  time: 7.4638  data: 7.2279  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.9542 (6.9430)  acc1: 0.9115 (1.0240)  acc5: 4.1667 (3.6160)  time: 1.6381  data: 1.4457  max mem: 55573
Test: Total time: 0:00:08 (1.6549 s / it)
* Acc@1 0.846 Acc@5 2.420 loss 7.063
Accuracy of the model EMA on 50000 test images: 0.8%
Max EMA accuracy: 0.85%
Epoch: [110]  [  0/156]  eta: 0:29:03  lr: 0.006128  min_lr: 0.006128  loss: 4.1141 (4.1141)  weight_decay: 0.0500 (0.0500)  time: 11.1752  data: 10.2931  max mem: 55573
Epoch: [110]  [ 10/156]  eta: 0:04:14  lr: 0.006126  min_lr: 0.006126  loss: 3.5671 (3.4967)  weight_decay: 0.0500 (0.0500)  time: 1.7449  data: 0.9823  max mem: 55573
Epoch: [110]  [ 20/156]  eta: 0:02:42  lr: 0.006123  min_lr: 0.006123  loss: 3.4732 (3.4265)  weight_decay: 0.0500 (0.0500)  time: 0.6966  data: 0.0258  max mem: 55573
Epoch: [110]  [ 30/156]  eta: 0:02:06  lr: 0.006121  min_lr: 0.006121  loss: 3.1733 (3.3213)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [110]  [ 40/156]  eta: 0:01:44  lr: 0.006119  min_lr: 0.006119  loss: 3.4837 (3.3875)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [110]  [ 50/156]  eta: 0:01:29  lr: 0.006116  min_lr: 0.006116  loss: 3.5875 (3.4181)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [110]  [ 60/156]  eta: 0:01:16  lr: 0.006114  min_lr: 0.006114  loss: 3.5934 (3.4182)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0006  max mem: 55573
Epoch: [110]  [ 70/156]  eta: 0:01:06  lr: 0.006111  min_lr: 0.006111  loss: 3.3759 (3.4021)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0006  max mem: 55573
Epoch: [110]  [ 80/156]  eta: 0:00:56  lr: 0.006109  min_lr: 0.006109  loss: 3.3759 (3.4135)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0006  max mem: 55573
Epoch: [110]  [ 90/156]  eta: 0:00:48  lr: 0.006106  min_lr: 0.006106  loss: 3.6962 (3.4406)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0006  max mem: 55573
Epoch: [110]  [100/156]  eta: 0:00:40  lr: 0.006104  min_lr: 0.006104  loss: 3.6686 (3.4581)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0006  max mem: 55573
Epoch: [110]  [110/156]  eta: 0:00:32  lr: 0.006102  min_lr: 0.006102  loss: 3.6090 (3.4512)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [110]  [120/156]  eta: 0:00:25  lr: 0.006099  min_lr: 0.006099  loss: 3.4941 (3.4425)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [110]  [130/156]  eta: 0:00:17  lr: 0.006097  min_lr: 0.006097  loss: 3.7215 (3.4629)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0012  max mem: 55573
Epoch: [110]  [140/156]  eta: 0:00:10  lr: 0.006094  min_lr: 0.006094  loss: 3.6660 (3.4651)  weight_decay: 0.0500 (0.0500)  time: 0.5860  data: 0.0010  max mem: 55573
Epoch: [110]  [150/156]  eta: 0:00:04  lr: 0.006092  min_lr: 0.006092  loss: 3.6103 (3.4701)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [110]  [155/156]  eta: 0:00:00  lr: 0.006090  min_lr: 0.006090  loss: 3.6050 (3.4708)  weight_decay: 0.0500 (0.0500)  time: 0.5828  data: 0.0001  max mem: 55573
Epoch: [110] Total time: 0:01:45 (0.6779 s / it)
Averaged stats: lr: 0.006090  min_lr: 0.006090  loss: 3.6050 (3.4875)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.0735 (1.0735)  acc1: 78.6458 (78.6458)  acc5: 95.5729 (95.5729)  time: 6.8698  data: 6.6310  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2824 (1.2781)  acc1: 72.6562 (71.2640)  acc5: 92.9688 (91.3600)  time: 1.5209  data: 1.3264  max mem: 55573
Test: Total time: 0:00:07 (1.5683 s / it)
* Acc@1 72.306 Acc@5 91.048 loss 1.278
Accuracy of the model on the 50000 test images: 72.3%
Max accuracy: 72.37%
Test:  [0/5]  eta: 0:00:36  loss: 6.6196 (6.6196)  acc1: 1.5625 (1.5625)  acc5: 4.0365 (4.0365)  time: 7.3748  data: 7.1388  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.8090 (6.8096)  acc1: 1.5625 (1.4400)  acc5: 5.2083 (4.8960)  time: 1.6203  data: 1.4279  max mem: 55573
Test: Total time: 0:00:08 (1.6368 s / it)
* Acc@1 1.060 Acc@5 3.076 loss 6.955
Accuracy of the model EMA on 50000 test images: 1.1%
Max EMA accuracy: 1.06%
Epoch: [111]  [  0/156]  eta: 0:34:35  lr: 0.006090  min_lr: 0.006090  loss: 3.9365 (3.9365)  weight_decay: 0.0500 (0.0500)  time: 13.3036  data: 12.5295  max mem: 55573
Epoch: [111]  [ 10/156]  eta: 0:04:19  lr: 0.006088  min_lr: 0.006088  loss: 3.6873 (3.6133)  weight_decay: 0.0500 (0.0500)  time: 1.7804  data: 1.1395  max mem: 55573
Epoch: [111]  [ 20/156]  eta: 0:02:45  lr: 0.006085  min_lr: 0.006085  loss: 3.6103 (3.5407)  weight_decay: 0.0500 (0.0500)  time: 0.6097  data: 0.0005  max mem: 55573
Epoch: [111]  [ 30/156]  eta: 0:02:07  lr: 0.006083  min_lr: 0.006083  loss: 3.5147 (3.4549)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [111]  [ 40/156]  eta: 0:01:45  lr: 0.006080  min_lr: 0.006080  loss: 3.5167 (3.4845)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0006  max mem: 55573
Epoch: [111]  [ 50/156]  eta: 0:01:29  lr: 0.006078  min_lr: 0.006078  loss: 3.5095 (3.4311)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0007  max mem: 55573
Epoch: [111]  [ 60/156]  eta: 0:01:17  lr: 0.006075  min_lr: 0.006075  loss: 3.5209 (3.4467)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0006  max mem: 55573
Epoch: [111]  [ 70/156]  eta: 0:01:06  lr: 0.006073  min_lr: 0.006073  loss: 3.5209 (3.4152)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [111]  [ 80/156]  eta: 0:00:57  lr: 0.006071  min_lr: 0.006071  loss: 3.2755 (3.4176)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [111]  [ 90/156]  eta: 0:00:48  lr: 0.006068  min_lr: 0.006068  loss: 3.5635 (3.4445)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0006  max mem: 55573
Epoch: [111]  [100/156]  eta: 0:00:40  lr: 0.006066  min_lr: 0.006066  loss: 3.6631 (3.4598)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [111]  [110/156]  eta: 0:00:32  lr: 0.006063  min_lr: 0.006063  loss: 3.6171 (3.4735)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [111]  [120/156]  eta: 0:00:25  lr: 0.006061  min_lr: 0.006061  loss: 3.5703 (3.4717)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [111]  [130/156]  eta: 0:00:17  lr: 0.006058  min_lr: 0.006058  loss: 3.4727 (3.4641)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0012  max mem: 55573
Epoch: [111]  [140/156]  eta: 0:00:10  lr: 0.006056  min_lr: 0.006056  loss: 3.3490 (3.4405)  weight_decay: 0.0500 (0.0500)  time: 0.5858  data: 0.0010  max mem: 55573
Epoch: [111]  [150/156]  eta: 0:00:04  lr: 0.006053  min_lr: 0.006053  loss: 3.4018 (3.4507)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0002  max mem: 55573
Epoch: [111]  [155/156]  eta: 0:00:00  lr: 0.006052  min_lr: 0.006052  loss: 3.5128 (3.4470)  weight_decay: 0.0500 (0.0500)  time: 0.5829  data: 0.0002  max mem: 55573
Epoch: [111] Total time: 0:01:46 (0.6803 s / it)
Averaged stats: lr: 0.006052  min_lr: 0.006052  loss: 3.5128 (3.4566)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.1416 (1.1416)  acc1: 77.0833 (77.0833)  acc5: 94.6615 (94.6615)  time: 7.3886  data: 7.1511  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2656 (1.2053)  acc1: 70.3125 (71.1680)  acc5: 92.9688 (90.8160)  time: 1.6242  data: 1.4303  max mem: 55573
Test: Total time: 0:00:08 (1.6634 s / it)
* Acc@1 72.588 Acc@5 91.396 loss 1.204
Accuracy of the model on the 50000 test images: 72.6%
Max accuracy: 72.59%
Test:  [0/5]  eta: 0:00:35  loss: 6.4590 (6.4590)  acc1: 2.4740 (2.4740)  acc5: 5.4688 (5.4688)  time: 7.0831  data: 6.8471  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.6535 (6.6670)  acc1: 2.4740 (2.1760)  acc5: 5.9896 (5.8880)  time: 1.5619  data: 1.3695  max mem: 55573
Test: Total time: 0:00:07 (1.5797 s / it)
* Acc@1 1.412 Acc@5 3.992 loss 6.835
Accuracy of the model EMA on 50000 test images: 1.4%
Max EMA accuracy: 1.41%
Epoch: [112]  [  0/156]  eta: 0:34:18  lr: 0.006052  min_lr: 0.006052  loss: 3.3098 (3.3098)  weight_decay: 0.0500 (0.0500)  time: 13.1929  data: 9.2715  max mem: 55573
Epoch: [112]  [ 10/156]  eta: 0:04:18  lr: 0.006049  min_lr: 0.006049  loss: 3.6961 (3.4964)  weight_decay: 0.0500 (0.0500)  time: 1.7734  data: 0.8443  max mem: 55573
Epoch: [112]  [ 20/156]  eta: 0:02:44  lr: 0.006047  min_lr: 0.006047  loss: 3.7265 (3.5204)  weight_decay: 0.0500 (0.0500)  time: 0.6114  data: 0.0011  max mem: 55573
Epoch: [112]  [ 30/156]  eta: 0:02:07  lr: 0.006044  min_lr: 0.006044  loss: 3.5109 (3.4044)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0007  max mem: 55573
Epoch: [112]  [ 40/156]  eta: 0:01:45  lr: 0.006042  min_lr: 0.006042  loss: 3.4426 (3.4437)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0007  max mem: 55573
Epoch: [112]  [ 50/156]  eta: 0:01:29  lr: 0.006039  min_lr: 0.006039  loss: 3.7088 (3.4977)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [112]  [ 60/156]  eta: 0:01:17  lr: 0.006037  min_lr: 0.006037  loss: 3.7347 (3.4544)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [112]  [ 70/156]  eta: 0:01:06  lr: 0.006035  min_lr: 0.006035  loss: 3.7347 (3.4846)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0006  max mem: 55573
Epoch: [112]  [ 80/156]  eta: 0:00:57  lr: 0.006032  min_lr: 0.006032  loss: 3.7532 (3.4582)  weight_decay: 0.0500 (0.0500)  time: 0.5948  data: 0.0006  max mem: 55573
Epoch: [112]  [ 90/156]  eta: 0:00:48  lr: 0.006030  min_lr: 0.006030  loss: 3.2924 (3.4403)  weight_decay: 0.0500 (0.0500)  time: 0.5949  data: 0.0006  max mem: 55573
Epoch: [112]  [100/156]  eta: 0:00:40  lr: 0.006027  min_lr: 0.006027  loss: 3.6791 (3.4679)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0008  max mem: 55573
Epoch: [112]  [110/156]  eta: 0:00:32  lr: 0.006025  min_lr: 0.006025  loss: 3.7430 (3.4695)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0008  max mem: 55573
Epoch: [112]  [120/156]  eta: 0:00:25  lr: 0.006022  min_lr: 0.006022  loss: 3.7236 (3.4811)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0008  max mem: 55573
Epoch: [112]  [130/156]  eta: 0:00:17  lr: 0.006020  min_lr: 0.006020  loss: 3.6255 (3.4850)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0015  max mem: 55573
Epoch: [112]  [140/156]  eta: 0:00:10  lr: 0.006017  min_lr: 0.006017  loss: 3.5440 (3.4912)  weight_decay: 0.0500 (0.0500)  time: 0.5856  data: 0.0012  max mem: 55573
Epoch: [112]  [150/156]  eta: 0:00:04  lr: 0.006015  min_lr: 0.006015  loss: 3.7245 (3.4886)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0002  max mem: 55573
Epoch: [112]  [155/156]  eta: 0:00:00  lr: 0.006013  min_lr: 0.006013  loss: 3.7245 (3.4874)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [112] Total time: 0:01:46 (0.6805 s / it)
Averaged stats: lr: 0.006013  min_lr: 0.006013  loss: 3.7245 (3.4581)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.1072 (1.1072)  acc1: 77.6042 (77.6042)  acc5: 93.8802 (93.8802)  time: 6.8426  data: 6.6055  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2111 (1.2341)  acc1: 74.0885 (72.2560)  acc5: 93.4896 (91.2960)  time: 1.5150  data: 1.3212  max mem: 55573
Test: Total time: 0:00:07 (1.5486 s / it)
* Acc@1 72.444 Acc@5 91.470 loss 1.213
Accuracy of the model on the 50000 test images: 72.4%
Max accuracy: 72.59%
Test:  [0/5]  eta: 0:00:36  loss: 6.2673 (6.2673)  acc1: 2.8646 (2.8646)  acc5: 7.5521 (7.5521)  time: 7.2804  data: 7.0439  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.4818 (6.4994)  acc1: 2.8646 (2.6880)  acc5: 7.5521 (7.6160)  time: 1.6015  data: 1.4089  max mem: 55573
Test: Total time: 0:00:08 (1.6199 s / it)
* Acc@1 1.952 Acc@5 5.272 loss 6.691
Accuracy of the model EMA on 50000 test images: 2.0%
Max EMA accuracy: 1.95%
Epoch: [113]  [  0/156]  eta: 0:31:26  lr: 0.006013  min_lr: 0.006013  loss: 2.2979 (2.2979)  weight_decay: 0.0500 (0.0500)  time: 12.0912  data: 11.4866  max mem: 55573
Epoch: [113]  [ 10/156]  eta: 0:04:01  lr: 0.006011  min_lr: 0.006011  loss: 3.6502 (3.5055)  weight_decay: 0.0500 (0.0500)  time: 1.6519  data: 1.0447  max mem: 55573
Epoch: [113]  [ 20/156]  eta: 0:02:36  lr: 0.006008  min_lr: 0.006008  loss: 3.5825 (3.4278)  weight_decay: 0.0500 (0.0500)  time: 0.6006  data: 0.0005  max mem: 55573
Epoch: [113]  [ 30/156]  eta: 0:02:01  lr: 0.006006  min_lr: 0.006006  loss: 3.5825 (3.4788)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [113]  [ 40/156]  eta: 0:01:41  lr: 0.006003  min_lr: 0.006003  loss: 3.6148 (3.4538)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [113]  [ 50/156]  eta: 0:01:26  lr: 0.006001  min_lr: 0.006001  loss: 3.6717 (3.4733)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0006  max mem: 55573
Epoch: [113]  [ 60/156]  eta: 0:01:15  lr: 0.005998  min_lr: 0.005998  loss: 3.3437 (3.4025)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [113]  [ 70/156]  eta: 0:01:04  lr: 0.005996  min_lr: 0.005996  loss: 2.9886 (3.3775)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0006  max mem: 55573
Epoch: [113]  [ 80/156]  eta: 0:00:55  lr: 0.005993  min_lr: 0.005993  loss: 3.5825 (3.4123)  weight_decay: 0.0500 (0.0500)  time: 0.5930  data: 0.0007  max mem: 55573
Epoch: [113]  [ 90/156]  eta: 0:00:47  lr: 0.005991  min_lr: 0.005991  loss: 3.7266 (3.4218)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0006  max mem: 55573
Epoch: [113]  [100/156]  eta: 0:00:39  lr: 0.005988  min_lr: 0.005988  loss: 3.3774 (3.4036)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [113]  [110/156]  eta: 0:00:32  lr: 0.005986  min_lr: 0.005986  loss: 3.4022 (3.4219)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [113]  [120/156]  eta: 0:00:24  lr: 0.005983  min_lr: 0.005983  loss: 3.6713 (3.4291)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [113]  [130/156]  eta: 0:00:17  lr: 0.005981  min_lr: 0.005981  loss: 3.7001 (3.4408)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0011  max mem: 55573
Epoch: [113]  [140/156]  eta: 0:00:10  lr: 0.005978  min_lr: 0.005978  loss: 3.7206 (3.4420)  weight_decay: 0.0500 (0.0500)  time: 0.5859  data: 0.0009  max mem: 55573
Epoch: [113]  [150/156]  eta: 0:00:04  lr: 0.005976  min_lr: 0.005976  loss: 3.6037 (3.4411)  weight_decay: 0.0500 (0.0500)  time: 0.5829  data: 0.0001  max mem: 55573
Epoch: [113]  [155/156]  eta: 0:00:00  lr: 0.005975  min_lr: 0.005975  loss: 3.5262 (3.4378)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [113] Total time: 0:01:44 (0.6710 s / it)
Averaged stats: lr: 0.005975  min_lr: 0.005975  loss: 3.5262 (3.4836)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.2023 (1.2023)  acc1: 75.1302 (75.1302)  acc5: 93.3594 (93.3594)  time: 7.2427  data: 7.0052  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2295 (1.2555)  acc1: 74.2188 (71.4240)  acc5: 93.3594 (91.2320)  time: 1.5949  data: 1.4011  max mem: 55573
Test: Total time: 0:00:08 (1.6364 s / it)
* Acc@1 72.876 Acc@5 91.482 loss 1.251
Accuracy of the model on the 50000 test images: 72.9%
Max accuracy: 72.88%
Test:  [0/5]  eta: 0:00:37  loss: 6.0626 (6.0626)  acc1: 4.5573 (4.5573)  acc5: 10.5469 (10.5469)  time: 7.4240  data: 7.1874  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.2947 (6.3173)  acc1: 4.5573 (3.7440)  acc5: 9.6354 (9.5680)  time: 1.6304  data: 1.4376  max mem: 55573
Test: Total time: 0:00:08 (1.6513 s / it)
* Acc@1 2.638 Acc@5 6.918 loss 6.529
Accuracy of the model EMA on 50000 test images: 2.6%
Max EMA accuracy: 2.64%
Epoch: [114]  [  0/156]  eta: 0:28:35  lr: 0.005974  min_lr: 0.005974  loss: 2.5949 (2.5949)  weight_decay: 0.0500 (0.0500)  time: 10.9942  data: 9.9900  max mem: 55573
Epoch: [114]  [ 10/156]  eta: 0:04:08  lr: 0.005972  min_lr: 0.005972  loss: 3.6422 (3.3772)  weight_decay: 0.0500 (0.0500)  time: 1.7050  data: 1.0028  max mem: 55573
Epoch: [114]  [ 20/156]  eta: 0:02:39  lr: 0.005969  min_lr: 0.005969  loss: 3.6408 (3.4143)  weight_decay: 0.0500 (0.0500)  time: 0.6828  data: 0.0524  max mem: 55573
Epoch: [114]  [ 30/156]  eta: 0:02:04  lr: 0.005967  min_lr: 0.005967  loss: 3.6408 (3.4465)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0006  max mem: 55573
Epoch: [114]  [ 40/156]  eta: 0:01:43  lr: 0.005964  min_lr: 0.005964  loss: 3.4828 (3.3951)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [114]  [ 50/156]  eta: 0:01:28  lr: 0.005962  min_lr: 0.005962  loss: 3.2947 (3.4008)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0006  max mem: 55573
Epoch: [114]  [ 60/156]  eta: 0:01:15  lr: 0.005959  min_lr: 0.005959  loss: 3.5746 (3.4268)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0006  max mem: 55573
Epoch: [114]  [ 70/156]  eta: 0:01:05  lr: 0.005957  min_lr: 0.005957  loss: 3.4879 (3.4088)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0006  max mem: 55573
Epoch: [114]  [ 80/156]  eta: 0:00:56  lr: 0.005954  min_lr: 0.005954  loss: 3.4533 (3.4154)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [114]  [ 90/156]  eta: 0:00:47  lr: 0.005952  min_lr: 0.005952  loss: 3.6091 (3.4331)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0005  max mem: 55573
Epoch: [114]  [100/156]  eta: 0:00:39  lr: 0.005949  min_lr: 0.005949  loss: 3.5165 (3.4166)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0008  max mem: 55573
Epoch: [114]  [110/156]  eta: 0:00:32  lr: 0.005947  min_lr: 0.005947  loss: 3.5072 (3.4266)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0008  max mem: 55573
Epoch: [114]  [120/156]  eta: 0:00:24  lr: 0.005944  min_lr: 0.005944  loss: 3.7442 (3.4343)  weight_decay: 0.0500 (0.0500)  time: 0.5978  data: 0.0005  max mem: 55573
Epoch: [114]  [130/156]  eta: 0:00:17  lr: 0.005942  min_lr: 0.005942  loss: 3.6882 (3.4489)  weight_decay: 0.0500 (0.0500)  time: 0.5990  data: 0.0007  max mem: 55573
Epoch: [114]  [140/156]  eta: 0:00:10  lr: 0.005939  min_lr: 0.005939  loss: 3.6882 (3.4560)  weight_decay: 0.0500 (0.0500)  time: 0.5866  data: 0.0005  max mem: 55573
Epoch: [114]  [150/156]  eta: 0:00:04  lr: 0.005937  min_lr: 0.005937  loss: 3.3799 (3.4439)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [114]  [155/156]  eta: 0:00:00  lr: 0.005935  min_lr: 0.005935  loss: 3.4993 (3.4473)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [114] Total time: 0:01:45 (0.6758 s / it)
Averaged stats: lr: 0.005935  min_lr: 0.005935  loss: 3.4993 (3.4889)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.1396 (1.1396)  acc1: 76.8229 (76.8229)  acc5: 93.6198 (93.6198)  time: 6.7472  data: 6.5101  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2559 (1.2358)  acc1: 71.7448 (71.0400)  acc5: 92.4479 (90.7840)  time: 1.4958  data: 1.3021  max mem: 55573
Test: Total time: 0:00:07 (1.5296 s / it)
* Acc@1 71.808 Acc@5 91.054 loss 1.228
Accuracy of the model on the 50000 test images: 71.8%
Max accuracy: 72.88%
Test:  [0/5]  eta: 0:00:33  loss: 5.8458 (5.8458)  acc1: 5.7292 (5.7292)  acc5: 13.4115 (13.4115)  time: 6.7487  data: 6.5126  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 6.1053 (6.1232)  acc1: 5.4688 (4.5760)  acc5: 10.9375 (11.6800)  time: 1.4951  data: 1.3026  max mem: 55573
Test: Total time: 0:00:07 (1.5368 s / it)
* Acc@1 3.462 Acc@5 8.872 loss 6.352
Accuracy of the model EMA on 50000 test images: 3.5%
Max EMA accuracy: 3.46%
Epoch: [115]  [  0/156]  eta: 0:34:33  lr: 0.005935  min_lr: 0.005935  loss: 3.4633 (3.4633)  weight_decay: 0.0500 (0.0500)  time: 13.2915  data: 9.7036  max mem: 55573
Epoch: [115]  [ 10/156]  eta: 0:04:20  lr: 0.005933  min_lr: 0.005933  loss: 3.3890 (3.3916)  weight_decay: 0.0500 (0.0500)  time: 1.7847  data: 0.8832  max mem: 55573
Epoch: [115]  [ 20/156]  eta: 0:02:45  lr: 0.005930  min_lr: 0.005930  loss: 3.3750 (3.3970)  weight_decay: 0.0500 (0.0500)  time: 0.6113  data: 0.0009  max mem: 55573
Epoch: [115]  [ 30/156]  eta: 0:02:07  lr: 0.005928  min_lr: 0.005928  loss: 3.6170 (3.4642)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0006  max mem: 55573
Epoch: [115]  [ 40/156]  eta: 0:01:45  lr: 0.005925  min_lr: 0.005925  loss: 3.6322 (3.4074)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [115]  [ 50/156]  eta: 0:01:29  lr: 0.005923  min_lr: 0.005923  loss: 3.4098 (3.3790)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [115]  [ 60/156]  eta: 0:01:17  lr: 0.005920  min_lr: 0.005920  loss: 3.4502 (3.4062)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0006  max mem: 55573
Epoch: [115]  [ 70/156]  eta: 0:01:06  lr: 0.005917  min_lr: 0.005917  loss: 3.6190 (3.4230)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0006  max mem: 55573
Epoch: [115]  [ 80/156]  eta: 0:00:57  lr: 0.005915  min_lr: 0.005915  loss: 3.7152 (3.4450)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0006  max mem: 55573
Epoch: [115]  [ 90/156]  eta: 0:00:48  lr: 0.005912  min_lr: 0.005912  loss: 3.5177 (3.4312)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0006  max mem: 55573
Epoch: [115]  [100/156]  eta: 0:00:40  lr: 0.005910  min_lr: 0.005910  loss: 3.4324 (3.4448)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [115]  [110/156]  eta: 0:00:32  lr: 0.005907  min_lr: 0.005907  loss: 3.5830 (3.4361)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [115]  [120/156]  eta: 0:00:25  lr: 0.005905  min_lr: 0.005905  loss: 3.5830 (3.4336)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0005  max mem: 55573
Epoch: [115]  [130/156]  eta: 0:00:17  lr: 0.005902  min_lr: 0.005902  loss: 3.6624 (3.4335)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0011  max mem: 55573
Epoch: [115]  [140/156]  eta: 0:00:10  lr: 0.005900  min_lr: 0.005900  loss: 3.7332 (3.4505)  weight_decay: 0.0500 (0.0500)  time: 0.5861  data: 0.0010  max mem: 55573
Epoch: [115]  [150/156]  eta: 0:00:04  lr: 0.005897  min_lr: 0.005897  loss: 3.7391 (3.4551)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [115]  [155/156]  eta: 0:00:00  lr: 0.005896  min_lr: 0.005896  loss: 3.8033 (3.4695)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [115] Total time: 0:01:46 (0.6799 s / it)
Averaged stats: lr: 0.005896  min_lr: 0.005896  loss: 3.8033 (3.4751)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.4149 (1.4149)  acc1: 76.4323 (76.4323)  acc5: 93.2292 (93.2292)  time: 7.1561  data: 6.9185  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5264 (1.5459)  acc1: 72.3958 (71.2640)  acc5: 92.9688 (90.7840)  time: 1.5779  data: 1.3838  max mem: 55573
Test: Total time: 0:00:08 (1.6333 s / it)
* Acc@1 71.914 Acc@5 90.842 loss 1.538
Accuracy of the model on the 50000 test images: 71.9%
Max accuracy: 72.88%
Test:  [0/5]  eta: 0:00:36  loss: 5.6213 (5.6213)  acc1: 7.2917 (7.2917)  acc5: 17.3177 (17.3177)  time: 7.2774  data: 7.0408  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 5.9243 (5.9152)  acc1: 6.2500 (5.7600)  acc5: 13.2812 (14.8800)  time: 1.6009  data: 1.4083  max mem: 55573
Test: Total time: 0:00:08 (1.6209 s / it)
* Acc@1 4.424 Acc@5 11.102 loss 6.159
Accuracy of the model EMA on 50000 test images: 4.4%
Max EMA accuracy: 4.42%
Epoch: [116]  [  0/156]  eta: 0:31:47  lr: 0.005896  min_lr: 0.005896  loss: 3.8366 (3.8366)  weight_decay: 0.0500 (0.0500)  time: 12.2257  data: 10.4274  max mem: 55573
Epoch: [116]  [ 10/156]  eta: 0:04:06  lr: 0.005893  min_lr: 0.005893  loss: 3.6851 (3.5631)  weight_decay: 0.0500 (0.0500)  time: 1.6897  data: 0.9485  max mem: 55573
Epoch: [116]  [ 20/156]  eta: 0:02:38  lr: 0.005891  min_lr: 0.005891  loss: 3.6422 (3.5101)  weight_decay: 0.0500 (0.0500)  time: 0.6133  data: 0.0006  max mem: 55573
Epoch: [116]  [ 30/156]  eta: 0:02:03  lr: 0.005888  min_lr: 0.005888  loss: 3.5315 (3.5040)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0006  max mem: 55573
Epoch: [116]  [ 40/156]  eta: 0:01:42  lr: 0.005886  min_lr: 0.005886  loss: 3.4726 (3.4787)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0006  max mem: 55573
Epoch: [116]  [ 50/156]  eta: 0:01:27  lr: 0.005883  min_lr: 0.005883  loss: 3.5270 (3.4816)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0007  max mem: 55573
Epoch: [116]  [ 60/156]  eta: 0:01:15  lr: 0.005881  min_lr: 0.005881  loss: 3.5478 (3.4847)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0007  max mem: 55573
Epoch: [116]  [ 70/156]  eta: 0:01:05  lr: 0.005878  min_lr: 0.005878  loss: 3.5791 (3.4717)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [116]  [ 80/156]  eta: 0:00:56  lr: 0.005875  min_lr: 0.005875  loss: 3.5916 (3.4937)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [116]  [ 90/156]  eta: 0:00:47  lr: 0.005873  min_lr: 0.005873  loss: 3.5916 (3.4880)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [116]  [100/156]  eta: 0:00:39  lr: 0.005870  min_lr: 0.005870  loss: 3.6523 (3.4982)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [116]  [110/156]  eta: 0:00:32  lr: 0.005868  min_lr: 0.005868  loss: 3.6712 (3.5011)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0006  max mem: 55573
Epoch: [116]  [120/156]  eta: 0:00:24  lr: 0.005865  min_lr: 0.005865  loss: 3.5666 (3.5183)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0006  max mem: 55573
Epoch: [116]  [130/156]  eta: 0:00:17  lr: 0.005863  min_lr: 0.005863  loss: 3.6771 (3.5209)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0012  max mem: 55573
Epoch: [116]  [140/156]  eta: 0:00:10  lr: 0.005860  min_lr: 0.005860  loss: 3.7148 (3.5240)  weight_decay: 0.0500 (0.0500)  time: 0.5869  data: 0.0010  max mem: 55573
Epoch: [116]  [150/156]  eta: 0:00:04  lr: 0.005858  min_lr: 0.005858  loss: 3.7556 (3.5275)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [116]  [155/156]  eta: 0:00:00  lr: 0.005856  min_lr: 0.005856  loss: 3.7556 (3.5266)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [116] Total time: 0:01:45 (0.6738 s / it)
Averaged stats: lr: 0.005856  min_lr: 0.005856  loss: 3.7556 (3.4670)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.3068 (1.3068)  acc1: 76.3021 (76.3021)  acc5: 93.7500 (93.7500)  time: 6.9671  data: 6.7296  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3068 (1.3952)  acc1: 74.8698 (71.5840)  acc5: 93.7500 (90.8160)  time: 1.5459  data: 1.3532  max mem: 55573
Test: Total time: 0:00:07 (1.5869 s / it)
* Acc@1 71.756 Acc@5 90.912 loss 1.394
Accuracy of the model on the 50000 test images: 71.8%
Max accuracy: 72.88%
Test:  [0/5]  eta: 0:00:35  loss: 5.3873 (5.3873)  acc1: 8.5938 (8.5938)  acc5: 20.5729 (20.5729)  time: 7.1362  data: 6.9001  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 5.7017 (5.6963)  acc1: 7.2917 (7.0400)  acc5: 18.8679 (17.9200)  time: 1.5726  data: 1.3801  max mem: 55573
Test: Total time: 0:00:07 (1.5918 s / it)
* Acc@1 5.548 Acc@5 13.826 loss 5.951
Accuracy of the model EMA on 50000 test images: 5.5%
Max EMA accuracy: 5.55%
Epoch: [117]  [  0/156]  eta: 0:31:12  lr: 0.005856  min_lr: 0.005856  loss: 4.0609 (4.0609)  weight_decay: 0.0500 (0.0500)  time: 12.0010  data: 9.3068  max mem: 55573
Epoch: [117]  [ 10/156]  eta: 0:04:07  lr: 0.005854  min_lr: 0.005854  loss: 3.9304 (3.8105)  weight_decay: 0.0500 (0.0500)  time: 1.6921  data: 0.8667  max mem: 55573
Epoch: [117]  [ 20/156]  eta: 0:02:38  lr: 0.005851  min_lr: 0.005851  loss: 3.7401 (3.7495)  weight_decay: 0.0500 (0.0500)  time: 0.6264  data: 0.0116  max mem: 55573
Epoch: [117]  [ 30/156]  eta: 0:02:03  lr: 0.005848  min_lr: 0.005848  loss: 3.6627 (3.6798)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [117]  [ 40/156]  eta: 0:01:42  lr: 0.005846  min_lr: 0.005846  loss: 3.5729 (3.6017)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0006  max mem: 55573
Epoch: [117]  [ 50/156]  eta: 0:01:27  lr: 0.005843  min_lr: 0.005843  loss: 3.5366 (3.5981)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0008  max mem: 55573
Epoch: [117]  [ 60/156]  eta: 0:01:15  lr: 0.005841  min_lr: 0.005841  loss: 3.4655 (3.5835)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0007  max mem: 55573
Epoch: [117]  [ 70/156]  eta: 0:01:05  lr: 0.005838  min_lr: 0.005838  loss: 3.5656 (3.5803)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0006  max mem: 55573
Epoch: [117]  [ 80/156]  eta: 0:00:56  lr: 0.005836  min_lr: 0.005836  loss: 3.6944 (3.5607)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0006  max mem: 55573
Epoch: [117]  [ 90/156]  eta: 0:00:47  lr: 0.005833  min_lr: 0.005833  loss: 3.4821 (3.5462)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0007  max mem: 55573
Epoch: [117]  [100/156]  eta: 0:00:39  lr: 0.005831  min_lr: 0.005831  loss: 3.3279 (3.5266)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0008  max mem: 55573
Epoch: [117]  [110/156]  eta: 0:00:32  lr: 0.005828  min_lr: 0.005828  loss: 3.3661 (3.5075)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0006  max mem: 55573
Epoch: [117]  [120/156]  eta: 0:00:24  lr: 0.005825  min_lr: 0.005825  loss: 3.4725 (3.5033)  weight_decay: 0.0500 (0.0500)  time: 0.5942  data: 0.0006  max mem: 55573
Epoch: [117]  [130/156]  eta: 0:00:17  lr: 0.005823  min_lr: 0.005823  loss: 3.6039 (3.5038)  weight_decay: 0.0500 (0.0500)  time: 0.5935  data: 0.0011  max mem: 55573
Epoch: [117]  [140/156]  eta: 0:00:10  lr: 0.005820  min_lr: 0.005820  loss: 3.7482 (3.5227)  weight_decay: 0.0500 (0.0500)  time: 0.5853  data: 0.0009  max mem: 55573
Epoch: [117]  [150/156]  eta: 0:00:04  lr: 0.005818  min_lr: 0.005818  loss: 3.7765 (3.5178)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0002  max mem: 55573
Epoch: [117]  [155/156]  eta: 0:00:00  lr: 0.005816  min_lr: 0.005816  loss: 3.5927 (3.5163)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0002  max mem: 55573
Epoch: [117] Total time: 0:01:45 (0.6741 s / it)
Averaged stats: lr: 0.005816  min_lr: 0.005816  loss: 3.5927 (3.4763)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.1227 (1.1227)  acc1: 77.3438 (77.3438)  acc5: 94.6615 (94.6615)  time: 7.1453  data: 6.9079  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1842 (1.2090)  acc1: 75.5208 (72.0640)  acc5: 93.8802 (91.9040)  time: 1.5756  data: 1.3817  max mem: 55573
Test: Total time: 0:00:08 (1.6225 s / it)
* Acc@1 72.886 Acc@5 91.718 loss 1.211
Accuracy of the model on the 50000 test images: 72.9%
Max accuracy: 72.89%
Test:  [0/5]  eta: 0:00:36  loss: 5.1627 (5.1627)  acc1: 9.7656 (9.7656)  acc5: 24.3490 (24.3490)  time: 7.2261  data: 6.9894  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 5.4713 (5.4811)  acc1: 8.4635 (8.5760)  acc5: 21.2240 (20.8000)  time: 1.5908  data: 1.3980  max mem: 55573
Test: Total time: 0:00:08 (1.6106 s / it)
* Acc@1 6.882 Acc@5 16.620 loss 5.742
Accuracy of the model EMA on 50000 test images: 6.9%
Max EMA accuracy: 6.88%
Epoch: [118]  [  0/156]  eta: 0:35:16  lr: 0.005816  min_lr: 0.005816  loss: 3.3487 (3.3487)  weight_decay: 0.0500 (0.0500)  time: 13.5666  data: 8.7588  max mem: 55573
Epoch: [118]  [ 10/156]  eta: 0:04:21  lr: 0.005814  min_lr: 0.005814  loss: 3.8120 (3.5675)  weight_decay: 0.0500 (0.0500)  time: 1.7879  data: 0.7969  max mem: 55573
Epoch: [118]  [ 20/156]  eta: 0:02:45  lr: 0.005811  min_lr: 0.005811  loss: 3.4638 (3.4330)  weight_decay: 0.0500 (0.0500)  time: 0.6011  data: 0.0006  max mem: 55573
Epoch: [118]  [ 30/156]  eta: 0:02:08  lr: 0.005809  min_lr: 0.005809  loss: 3.4683 (3.4348)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0006  max mem: 55573
Epoch: [118]  [ 40/156]  eta: 0:01:45  lr: 0.005806  min_lr: 0.005806  loss: 3.5423 (3.4200)  weight_decay: 0.0500 (0.0500)  time: 0.5932  data: 0.0006  max mem: 55573
Epoch: [118]  [ 50/156]  eta: 0:01:30  lr: 0.005803  min_lr: 0.005803  loss: 3.1876 (3.3929)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0006  max mem: 55573
Epoch: [118]  [ 60/156]  eta: 0:01:17  lr: 0.005801  min_lr: 0.005801  loss: 3.2180 (3.3744)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0007  max mem: 55573
Epoch: [118]  [ 70/156]  eta: 0:01:06  lr: 0.005798  min_lr: 0.005798  loss: 3.5711 (3.4348)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0006  max mem: 55573
Epoch: [118]  [ 80/156]  eta: 0:00:57  lr: 0.005796  min_lr: 0.005796  loss: 3.7200 (3.4329)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0006  max mem: 55573
Epoch: [118]  [ 90/156]  eta: 0:00:48  lr: 0.005793  min_lr: 0.005793  loss: 3.4381 (3.4498)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0007  max mem: 55573
Epoch: [118]  [100/156]  eta: 0:00:40  lr: 0.005791  min_lr: 0.005791  loss: 3.6345 (3.4559)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0008  max mem: 55573
Epoch: [118]  [110/156]  eta: 0:00:32  lr: 0.005788  min_lr: 0.005788  loss: 3.6403 (3.4773)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0009  max mem: 55573
Epoch: [118]  [120/156]  eta: 0:00:25  lr: 0.005785  min_lr: 0.005785  loss: 3.6222 (3.4797)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0008  max mem: 55573
Epoch: [118]  [130/156]  eta: 0:00:17  lr: 0.005783  min_lr: 0.005783  loss: 3.2361 (3.4619)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0012  max mem: 55573
Epoch: [118]  [140/156]  eta: 0:00:10  lr: 0.005780  min_lr: 0.005780  loss: 3.0885 (3.4470)  weight_decay: 0.0500 (0.0500)  time: 0.5851  data: 0.0009  max mem: 55573
Epoch: [118]  [150/156]  eta: 0:00:04  lr: 0.005778  min_lr: 0.005778  loss: 3.5492 (3.4633)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0002  max mem: 55573
Epoch: [118]  [155/156]  eta: 0:00:00  lr: 0.005776  min_lr: 0.005776  loss: 3.5492 (3.4555)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0002  max mem: 55573
Epoch: [118] Total time: 0:01:46 (0.6813 s / it)
Averaged stats: lr: 0.005776  min_lr: 0.005776  loss: 3.5492 (3.4702)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.1607 (1.1607)  acc1: 77.7344 (77.7344)  acc5: 94.4010 (94.4010)  time: 7.1258  data: 6.8884  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2664 (1.2999)  acc1: 74.3490 (71.5520)  acc5: 93.2292 (91.5520)  time: 1.5715  data: 1.3777  max mem: 55573
Test: Total time: 0:00:08 (1.6270 s / it)
* Acc@1 72.422 Acc@5 91.570 loss 1.280
Accuracy of the model on the 50000 test images: 72.4%
Max accuracy: 72.89%
Test:  [0/5]  eta: 0:00:35  loss: 4.9356 (4.9356)  acc1: 12.2396 (12.2396)  acc5: 27.7344 (27.7344)  time: 7.0864  data: 6.8504  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 5.2310 (5.2598)  acc1: 10.6771 (10.7200)  acc5: 22.7865 (23.9040)  time: 1.5626  data: 1.3702  max mem: 55573
Test: Total time: 0:00:07 (1.5796 s / it)
* Acc@1 8.414 Acc@5 19.788 loss 5.526
Accuracy of the model EMA on 50000 test images: 8.4%
Max EMA accuracy: 8.41%
Epoch: [119]  [  0/156]  eta: 0:26:24  lr: 0.005776  min_lr: 0.005776  loss: 2.5564 (2.5564)  weight_decay: 0.0500 (0.0500)  time: 10.1575  data: 8.4803  max mem: 55573
Epoch: [119]  [ 10/156]  eta: 0:04:17  lr: 0.005774  min_lr: 0.005774  loss: 3.3588 (3.3302)  weight_decay: 0.0500 (0.0500)  time: 1.7608  data: 0.8922  max mem: 55573
Epoch: [119]  [ 20/156]  eta: 0:02:43  lr: 0.005771  min_lr: 0.005771  loss: 3.3761 (3.3811)  weight_decay: 0.0500 (0.0500)  time: 0.7556  data: 0.0670  max mem: 55573
Epoch: [119]  [ 30/156]  eta: 0:02:06  lr: 0.005768  min_lr: 0.005768  loss: 3.3761 (3.3739)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0007  max mem: 55573
Epoch: [119]  [ 40/156]  eta: 0:01:44  lr: 0.005766  min_lr: 0.005766  loss: 3.5837 (3.3916)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0006  max mem: 55573
Epoch: [119]  [ 50/156]  eta: 0:01:29  lr: 0.005763  min_lr: 0.005763  loss: 3.6158 (3.4063)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [119]  [ 60/156]  eta: 0:01:16  lr: 0.005761  min_lr: 0.005761  loss: 3.4906 (3.3722)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [119]  [ 70/156]  eta: 0:01:06  lr: 0.005758  min_lr: 0.005758  loss: 3.1343 (3.3513)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0005  max mem: 55573
Epoch: [119]  [ 80/156]  eta: 0:00:56  lr: 0.005755  min_lr: 0.005755  loss: 3.5055 (3.3919)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [119]  [ 90/156]  eta: 0:00:48  lr: 0.005753  min_lr: 0.005753  loss: 3.6279 (3.3948)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0006  max mem: 55573
Epoch: [119]  [100/156]  eta: 0:00:40  lr: 0.005750  min_lr: 0.005750  loss: 3.4787 (3.3999)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0006  max mem: 55573
Epoch: [119]  [110/156]  eta: 0:00:32  lr: 0.005748  min_lr: 0.005748  loss: 3.4381 (3.3957)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0007  max mem: 55573
Epoch: [119]  [120/156]  eta: 0:00:25  lr: 0.005745  min_lr: 0.005745  loss: 3.5099 (3.4019)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0007  max mem: 55573
Epoch: [119]  [130/156]  eta: 0:00:17  lr: 0.005743  min_lr: 0.005743  loss: 3.7229 (3.4277)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0013  max mem: 55573
Epoch: [119]  [140/156]  eta: 0:00:10  lr: 0.005740  min_lr: 0.005740  loss: 3.6348 (3.4220)  weight_decay: 0.0500 (0.0500)  time: 0.5868  data: 0.0010  max mem: 55573
Epoch: [119]  [150/156]  eta: 0:00:04  lr: 0.005737  min_lr: 0.005737  loss: 3.2959 (3.4121)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0002  max mem: 55573
Epoch: [119]  [155/156]  eta: 0:00:00  lr: 0.005736  min_lr: 0.005736  loss: 3.2343 (3.4090)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [119] Total time: 0:01:45 (0.6793 s / it)
Averaged stats: lr: 0.005736  min_lr: 0.005736  loss: 3.2343 (3.4517)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.2107 (1.2107)  acc1: 75.3906 (75.3906)  acc5: 93.7500 (93.7500)  time: 6.7812  data: 6.5442  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3694 (1.3917)  acc1: 73.0469 (70.7840)  acc5: 92.8385 (90.4000)  time: 1.5028  data: 1.3089  max mem: 55573
Test: Total time: 0:00:07 (1.5453 s / it)
* Acc@1 71.784 Acc@5 90.854 loss 1.368
Accuracy of the model on the 50000 test images: 71.8%
Max accuracy: 72.89%
Test:  [0/5]  eta: 0:00:33  loss: 4.7147 (4.7147)  acc1: 14.4531 (14.4531)  acc5: 30.3385 (30.3385)  time: 6.7306  data: 6.4946  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 4.9976 (5.0382)  acc1: 12.7604 (12.7680)  acc5: 26.8229 (27.2320)  time: 1.5699  data: 1.3775  max mem: 55573
Test: Total time: 0:00:07 (1.5933 s / it)
* Acc@1 10.164 Acc@5 23.016 loss 5.306
Accuracy of the model EMA on 50000 test images: 10.2%
Max EMA accuracy: 10.16%
Epoch: [120]  [  0/156]  eta: 0:35:29  lr: 0.005736  min_lr: 0.005736  loss: 3.4372 (3.4372)  weight_decay: 0.0500 (0.0500)  time: 13.6490  data: 9.9843  max mem: 55573
Epoch: [120]  [ 10/156]  eta: 0:04:22  lr: 0.005733  min_lr: 0.005733  loss: 3.4372 (3.4656)  weight_decay: 0.0500 (0.0500)  time: 1.8003  data: 0.9082  max mem: 55573
Epoch: [120]  [ 20/156]  eta: 0:02:46  lr: 0.005731  min_lr: 0.005731  loss: 3.3260 (3.3381)  weight_decay: 0.0500 (0.0500)  time: 0.6019  data: 0.0005  max mem: 55573
Epoch: [120]  [ 30/156]  eta: 0:02:08  lr: 0.005728  min_lr: 0.005728  loss: 3.4157 (3.3578)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [120]  [ 40/156]  eta: 0:01:46  lr: 0.005725  min_lr: 0.005725  loss: 3.3341 (3.3406)  weight_decay: 0.0500 (0.0500)  time: 0.5930  data: 0.0006  max mem: 55573
Epoch: [120]  [ 50/156]  eta: 0:01:30  lr: 0.005723  min_lr: 0.005723  loss: 3.4564 (3.3787)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [120]  [ 60/156]  eta: 0:01:17  lr: 0.005720  min_lr: 0.005720  loss: 3.6444 (3.4055)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [120]  [ 70/156]  eta: 0:01:06  lr: 0.005718  min_lr: 0.005718  loss: 3.4733 (3.4212)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0006  max mem: 55573
Epoch: [120]  [ 80/156]  eta: 0:00:57  lr: 0.005715  min_lr: 0.005715  loss: 3.5909 (3.4504)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [120]  [ 90/156]  eta: 0:00:48  lr: 0.005712  min_lr: 0.005712  loss: 3.6158 (3.4582)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0007  max mem: 55573
Epoch: [120]  [100/156]  eta: 0:00:40  lr: 0.005710  min_lr: 0.005710  loss: 3.3263 (3.4450)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0007  max mem: 55573
Epoch: [120]  [110/156]  eta: 0:00:32  lr: 0.005707  min_lr: 0.005707  loss: 3.3492 (3.4378)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0006  max mem: 55573
Epoch: [120]  [120/156]  eta: 0:00:25  lr: 0.005705  min_lr: 0.005705  loss: 3.5891 (3.4317)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0007  max mem: 55573
Epoch: [120]  [130/156]  eta: 0:00:18  lr: 0.005702  min_lr: 0.005702  loss: 3.4125 (3.4102)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0015  max mem: 55573
Epoch: [120]  [140/156]  eta: 0:00:10  lr: 0.005699  min_lr: 0.005699  loss: 3.4125 (3.4129)  weight_decay: 0.0500 (0.0500)  time: 0.5865  data: 0.0012  max mem: 55573
Epoch: [120]  [150/156]  eta: 0:00:04  lr: 0.005697  min_lr: 0.005697  loss: 3.6841 (3.4255)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0002  max mem: 55573
Epoch: [120]  [155/156]  eta: 0:00:00  lr: 0.005696  min_lr: 0.005696  loss: 3.6841 (3.4240)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [120] Total time: 0:01:46 (0.6822 s / it)
Averaged stats: lr: 0.005696  min_lr: 0.005696  loss: 3.6841 (3.4670)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.1259 (1.1259)  acc1: 77.8646 (77.8646)  acc5: 93.2292 (93.2292)  time: 6.9987  data: 6.7613  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1753 (1.2338)  acc1: 75.9115 (73.1520)  acc5: 93.2292 (92.0960)  time: 1.5468  data: 1.3526  max mem: 55573
Test: Total time: 0:00:07 (1.5952 s / it)
* Acc@1 73.672 Acc@5 92.256 loss 1.209
Accuracy of the model on the 50000 test images: 73.7%
Max accuracy: 73.67%
Test:  [0/5]  eta: 0:00:34  loss: 4.4953 (4.4953)  acc1: 16.0156 (16.0156)  acc5: 33.7240 (33.7240)  time: 6.9333  data: 6.6972  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 4.7522 (4.8120)  acc1: 14.9740 (14.6880)  acc5: 32.0755 (31.4560)  time: 1.5320  data: 1.3396  max mem: 55573
Test: Total time: 0:00:07 (1.5494 s / it)
* Acc@1 12.038 Acc@5 26.514 loss 5.078
Accuracy of the model EMA on 50000 test images: 12.0%
Max EMA accuracy: 12.04%
Epoch: [121]  [  0/156]  eta: 0:33:30  lr: 0.005695  min_lr: 0.005695  loss: 3.8528 (3.8528)  weight_decay: 0.0500 (0.0500)  time: 12.8895  data: 12.2308  max mem: 55573
Epoch: [121]  [ 10/156]  eta: 0:04:09  lr: 0.005693  min_lr: 0.005693  loss: 3.6669 (3.5650)  weight_decay: 0.0500 (0.0500)  time: 1.7113  data: 1.1124  max mem: 55573
Epoch: [121]  [ 20/156]  eta: 0:02:40  lr: 0.005690  min_lr: 0.005690  loss: 3.6669 (3.6794)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0006  max mem: 55573
Epoch: [121]  [ 30/156]  eta: 0:02:04  lr: 0.005687  min_lr: 0.005687  loss: 3.5926 (3.5750)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0006  max mem: 55573
Epoch: [121]  [ 40/156]  eta: 0:01:43  lr: 0.005685  min_lr: 0.005685  loss: 3.3887 (3.5149)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0006  max mem: 55573
Epoch: [121]  [ 50/156]  eta: 0:01:28  lr: 0.005682  min_lr: 0.005682  loss: 3.5314 (3.5265)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0006  max mem: 55573
Epoch: [121]  [ 60/156]  eta: 0:01:16  lr: 0.005680  min_lr: 0.005680  loss: 3.6012 (3.4905)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0006  max mem: 55573
Epoch: [121]  [ 70/156]  eta: 0:01:05  lr: 0.005677  min_lr: 0.005677  loss: 3.3106 (3.4578)  weight_decay: 0.0500 (0.0500)  time: 0.5987  data: 0.0006  max mem: 55573
Epoch: [121]  [ 80/156]  eta: 0:00:56  lr: 0.005674  min_lr: 0.005674  loss: 3.2830 (3.4258)  weight_decay: 0.0500 (0.0500)  time: 0.5971  data: 0.0006  max mem: 55573
Epoch: [121]  [ 90/156]  eta: 0:00:48  lr: 0.005672  min_lr: 0.005672  loss: 3.3520 (3.4263)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0006  max mem: 55573
Epoch: [121]  [100/156]  eta: 0:00:39  lr: 0.005669  min_lr: 0.005669  loss: 3.5988 (3.4539)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0006  max mem: 55573
Epoch: [121]  [110/156]  eta: 0:00:32  lr: 0.005667  min_lr: 0.005667  loss: 3.6708 (3.4450)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0007  max mem: 55573
Epoch: [121]  [120/156]  eta: 0:00:24  lr: 0.005664  min_lr: 0.005664  loss: 3.6024 (3.4408)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0006  max mem: 55573
Epoch: [121]  [130/156]  eta: 0:00:17  lr: 0.005661  min_lr: 0.005661  loss: 3.6530 (3.4399)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0012  max mem: 55573
Epoch: [121]  [140/156]  eta: 0:00:10  lr: 0.005659  min_lr: 0.005659  loss: 3.6342 (3.4439)  weight_decay: 0.0500 (0.0500)  time: 0.5852  data: 0.0010  max mem: 55573
Epoch: [121]  [150/156]  eta: 0:00:04  lr: 0.005656  min_lr: 0.005656  loss: 3.5548 (3.4411)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [121]  [155/156]  eta: 0:00:00  lr: 0.005655  min_lr: 0.005655  loss: 3.5613 (3.4345)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [121] Total time: 0:01:45 (0.6757 s / it)
Averaged stats: lr: 0.005655  min_lr: 0.005655  loss: 3.5613 (3.4491)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.0238 (1.0238)  acc1: 77.3438 (77.3438)  acc5: 95.7031 (95.7031)  time: 6.8798  data: 6.6427  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1447 (1.2249)  acc1: 76.4323 (72.2240)  acc5: 94.7917 (92.3200)  time: 1.5224  data: 1.3286  max mem: 55573
Test: Total time: 0:00:07 (1.5589 s / it)
* Acc@1 73.090 Acc@5 91.888 loss 1.224
Accuracy of the model on the 50000 test images: 73.1%
Max accuracy: 73.67%
Test:  [0/5]  eta: 0:00:35  loss: 4.2887 (4.2887)  acc1: 17.7083 (17.7083)  acc5: 37.3698 (37.3698)  time: 7.0664  data: 6.8305  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 4.5158 (4.5971)  acc1: 16.9811 (16.3840)  acc5: 33.9623 (34.8800)  time: 1.5586  data: 1.3662  max mem: 55573
Test: Total time: 0:00:07 (1.5761 s / it)
* Acc@1 13.952 Acc@5 30.212 loss 4.859
Accuracy of the model EMA on 50000 test images: 14.0%
Max EMA accuracy: 13.95%
Epoch: [122]  [  0/156]  eta: 0:37:45  lr: 0.005655  min_lr: 0.005655  loss: 3.8062 (3.8062)  weight_decay: 0.0500 (0.0500)  time: 14.5200  data: 13.9028  max mem: 55573
Epoch: [122]  [ 10/156]  eta: 0:04:30  lr: 0.005652  min_lr: 0.005652  loss: 3.1310 (3.1823)  weight_decay: 0.0500 (0.0500)  time: 1.8547  data: 1.2643  max mem: 55573
Epoch: [122]  [ 20/156]  eta: 0:02:50  lr: 0.005649  min_lr: 0.005649  loss: 3.2188 (3.2834)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [122]  [ 30/156]  eta: 0:02:10  lr: 0.005647  min_lr: 0.005647  loss: 3.4532 (3.3648)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0006  max mem: 55573
Epoch: [122]  [ 40/156]  eta: 0:01:47  lr: 0.005644  min_lr: 0.005644  loss: 3.6090 (3.4258)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [122]  [ 50/156]  eta: 0:01:31  lr: 0.005641  min_lr: 0.005641  loss: 3.7525 (3.4571)  weight_decay: 0.0500 (0.0500)  time: 0.5933  data: 0.0005  max mem: 55573
Epoch: [122]  [ 60/156]  eta: 0:01:18  lr: 0.005639  min_lr: 0.005639  loss: 3.5317 (3.4251)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0006  max mem: 55573
Epoch: [122]  [ 70/156]  eta: 0:01:07  lr: 0.005636  min_lr: 0.005636  loss: 3.5350 (3.4702)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0008  max mem: 55573
Epoch: [122]  [ 80/156]  eta: 0:00:57  lr: 0.005634  min_lr: 0.005634  loss: 3.7132 (3.4937)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0007  max mem: 55573
Epoch: [122]  [ 90/156]  eta: 0:00:49  lr: 0.005631  min_lr: 0.005631  loss: 3.5867 (3.4727)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0006  max mem: 55573
Epoch: [122]  [100/156]  eta: 0:00:40  lr: 0.005628  min_lr: 0.005628  loss: 3.2673 (3.4371)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0006  max mem: 55573
Epoch: [122]  [110/156]  eta: 0:00:32  lr: 0.005626  min_lr: 0.005626  loss: 3.2670 (3.4251)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0005  max mem: 55573
Epoch: [122]  [120/156]  eta: 0:00:25  lr: 0.005623  min_lr: 0.005623  loss: 3.5429 (3.4286)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [122]  [130/156]  eta: 0:00:18  lr: 0.005620  min_lr: 0.005620  loss: 3.5529 (3.4232)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0012  max mem: 55573
Epoch: [122]  [140/156]  eta: 0:00:11  lr: 0.005618  min_lr: 0.005618  loss: 3.7482 (3.4405)  weight_decay: 0.0500 (0.0500)  time: 0.5868  data: 0.0010  max mem: 55573
Epoch: [122]  [150/156]  eta: 0:00:04  lr: 0.005615  min_lr: 0.005615  loss: 3.7547 (3.4477)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0002  max mem: 55573
Epoch: [122]  [155/156]  eta: 0:00:00  lr: 0.005614  min_lr: 0.005614  loss: 3.7547 (3.4631)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0002  max mem: 55573
Epoch: [122] Total time: 0:01:46 (0.6856 s / it)
Averaged stats: lr: 0.005614  min_lr: 0.005614  loss: 3.7547 (3.4566)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.2550 (1.2550)  acc1: 76.0417 (76.0417)  acc5: 94.0104 (94.0104)  time: 7.2103  data: 6.9728  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3540 (1.3826)  acc1: 73.0469 (71.5200)  acc5: 93.6198 (91.5840)  time: 1.5887  data: 1.3947  max mem: 55573
Test: Total time: 0:00:08 (1.6273 s / it)
* Acc@1 72.856 Acc@5 91.630 loss 1.363
Accuracy of the model on the 50000 test images: 72.9%
Max accuracy: 73.67%
Test:  [0/5]  eta: 0:00:34  loss: 4.0998 (4.0998)  acc1: 20.4427 (20.4427)  acc5: 40.4948 (40.4948)  time: 6.9342  data: 6.6976  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 4.3014 (4.4001)  acc1: 20.3125 (18.9440)  acc5: 36.4583 (37.8240)  time: 1.5421  data: 1.3493  max mem: 55573
Test: Total time: 0:00:07 (1.5626 s / it)
* Acc@1 16.016 Acc@5 33.574 loss 4.654
Accuracy of the model EMA on 50000 test images: 16.0%
Max EMA accuracy: 16.02%
Epoch: [123]  [  0/156]  eta: 0:32:02  lr: 0.005614  min_lr: 0.005614  loss: 3.6126 (3.6126)  weight_decay: 0.0500 (0.0500)  time: 12.3226  data: 10.7436  max mem: 55573
Epoch: [123]  [ 10/156]  eta: 0:04:05  lr: 0.005611  min_lr: 0.005611  loss: 3.3895 (3.3322)  weight_decay: 0.0500 (0.0500)  time: 1.6790  data: 0.9772  max mem: 55573
Epoch: [123]  [ 20/156]  eta: 0:02:37  lr: 0.005608  min_lr: 0.005608  loss: 3.4017 (3.3708)  weight_decay: 0.0500 (0.0500)  time: 0.6030  data: 0.0005  max mem: 55573
Epoch: [123]  [ 30/156]  eta: 0:02:03  lr: 0.005606  min_lr: 0.005606  loss: 3.6412 (3.4904)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [123]  [ 40/156]  eta: 0:01:42  lr: 0.005603  min_lr: 0.005603  loss: 3.6409 (3.5224)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [123]  [ 50/156]  eta: 0:01:27  lr: 0.005600  min_lr: 0.005600  loss: 3.6339 (3.5208)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [123]  [ 60/156]  eta: 0:01:15  lr: 0.005598  min_lr: 0.005598  loss: 3.5975 (3.5110)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [123]  [ 70/156]  eta: 0:01:05  lr: 0.005595  min_lr: 0.005595  loss: 3.1776 (3.4364)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [123]  [ 80/156]  eta: 0:00:56  lr: 0.005592  min_lr: 0.005592  loss: 2.8652 (3.4351)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [123]  [ 90/156]  eta: 0:00:47  lr: 0.005590  min_lr: 0.005590  loss: 3.6345 (3.4394)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0006  max mem: 55573
Epoch: [123]  [100/156]  eta: 0:00:39  lr: 0.005587  min_lr: 0.005587  loss: 3.3702 (3.4274)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0006  max mem: 55573
Epoch: [123]  [110/156]  eta: 0:00:32  lr: 0.005585  min_lr: 0.005585  loss: 3.3538 (3.4268)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0007  max mem: 55573
Epoch: [123]  [120/156]  eta: 0:00:24  lr: 0.005582  min_lr: 0.005582  loss: 3.4953 (3.4310)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0008  max mem: 55573
Epoch: [123]  [130/156]  eta: 0:00:17  lr: 0.005579  min_lr: 0.005579  loss: 3.4953 (3.4272)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0019  max mem: 55573
Epoch: [123]  [140/156]  eta: 0:00:10  lr: 0.005577  min_lr: 0.005577  loss: 3.3656 (3.4229)  weight_decay: 0.0500 (0.0500)  time: 0.5870  data: 0.0016  max mem: 55573
Epoch: [123]  [150/156]  eta: 0:00:04  lr: 0.005574  min_lr: 0.005574  loss: 3.4826 (3.4203)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0002  max mem: 55573
Epoch: [123]  [155/156]  eta: 0:00:00  lr: 0.005573  min_lr: 0.005573  loss: 3.5358 (3.4294)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0002  max mem: 55573
Epoch: [123] Total time: 0:01:44 (0.6727 s / it)
Averaged stats: lr: 0.005573  min_lr: 0.005573  loss: 3.5358 (3.4621)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.1748 (1.1748)  acc1: 78.1250 (78.1250)  acc5: 94.1406 (94.1406)  time: 7.1475  data: 6.9094  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2558 (1.3181)  acc1: 76.9531 (73.8240)  acc5: 94.1406 (91.5520)  time: 1.5760  data: 1.3820  max mem: 55573
Test: Total time: 0:00:08 (1.6108 s / it)
* Acc@1 73.598 Acc@5 92.100 loss 1.307
Accuracy of the model on the 50000 test images: 73.6%
Max accuracy: 73.67%
Test:  [0/5]  eta: 0:00:36  loss: 3.9155 (3.9155)  acc1: 22.1354 (22.1354)  acc5: 43.4896 (43.4896)  time: 7.2053  data: 6.9692  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 4.0827 (4.2033)  acc1: 22.1354 (21.0880)  acc5: 40.1042 (41.3120)  time: 1.5865  data: 1.3940  max mem: 55573
Test: Total time: 0:00:08 (1.6042 s / it)
* Acc@1 18.070 Acc@5 37.000 loss 4.451
Accuracy of the model EMA on 50000 test images: 18.1%
Max EMA accuracy: 18.07%
Epoch: [124]  [  0/156]  eta: 0:31:56  lr: 0.005572  min_lr: 0.005572  loss: 3.8701 (3.8701)  weight_decay: 0.0500 (0.0500)  time: 12.2833  data: 9.7771  max mem: 55573
Epoch: [124]  [ 10/156]  eta: 0:04:24  lr: 0.005570  min_lr: 0.005570  loss: 3.6566 (3.4835)  weight_decay: 0.0500 (0.0500)  time: 1.8083  data: 0.9236  max mem: 55573
Epoch: [124]  [ 20/156]  eta: 0:02:47  lr: 0.005567  min_lr: 0.005567  loss: 3.5969 (3.4967)  weight_decay: 0.0500 (0.0500)  time: 0.6761  data: 0.0193  max mem: 55573
Epoch: [124]  [ 30/156]  eta: 0:02:08  lr: 0.005564  min_lr: 0.005564  loss: 3.3457 (3.3931)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [124]  [ 40/156]  eta: 0:01:46  lr: 0.005562  min_lr: 0.005562  loss: 3.3027 (3.3637)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [124]  [ 50/156]  eta: 0:01:30  lr: 0.005559  min_lr: 0.005559  loss: 3.2385 (3.3078)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [124]  [ 60/156]  eta: 0:01:17  lr: 0.005557  min_lr: 0.005557  loss: 3.3802 (3.3517)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [124]  [ 70/156]  eta: 0:01:07  lr: 0.005554  min_lr: 0.005554  loss: 3.5918 (3.3854)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0005  max mem: 55573
Epoch: [124]  [ 80/156]  eta: 0:00:57  lr: 0.005551  min_lr: 0.005551  loss: 3.7706 (3.4404)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [124]  [ 90/156]  eta: 0:00:48  lr: 0.005549  min_lr: 0.005549  loss: 3.7328 (3.4331)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [124]  [100/156]  eta: 0:00:40  lr: 0.005546  min_lr: 0.005546  loss: 3.5275 (3.4521)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0005  max mem: 55573
Epoch: [124]  [110/156]  eta: 0:00:32  lr: 0.005543  min_lr: 0.005543  loss: 3.6994 (3.4734)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [124]  [120/156]  eta: 0:00:25  lr: 0.005541  min_lr: 0.005541  loss: 3.6994 (3.4745)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [124]  [130/156]  eta: 0:00:18  lr: 0.005538  min_lr: 0.005538  loss: 3.6629 (3.4756)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0011  max mem: 55573
Epoch: [124]  [140/156]  eta: 0:00:10  lr: 0.005535  min_lr: 0.005535  loss: 3.7410 (3.4906)  weight_decay: 0.0500 (0.0500)  time: 0.5867  data: 0.0010  max mem: 55573
Epoch: [124]  [150/156]  eta: 0:00:04  lr: 0.005533  min_lr: 0.005533  loss: 3.7201 (3.4901)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0002  max mem: 55573
Epoch: [124]  [155/156]  eta: 0:00:00  lr: 0.005531  min_lr: 0.005531  loss: 3.6804 (3.4876)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0002  max mem: 55573
Epoch: [124] Total time: 0:01:46 (0.6818 s / it)
Averaged stats: lr: 0.005531  min_lr: 0.005531  loss: 3.6804 (3.4558)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.2924 (1.2924)  acc1: 78.7760 (78.7760)  acc5: 94.0104 (94.0104)  time: 7.1271  data: 6.8896  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3603 (1.4054)  acc1: 75.2604 (72.7360)  acc5: 93.2292 (91.2960)  time: 1.5720  data: 1.3780  max mem: 55573
Test: Total time: 0:00:08 (1.6176 s / it)
* Acc@1 72.460 Acc@5 91.636 loss 1.381
Accuracy of the model on the 50000 test images: 72.5%
Max accuracy: 73.67%
Test:  [0/5]  eta: 0:00:36  loss: 3.7405 (3.7405)  acc1: 24.4792 (24.4792)  acc5: 46.4844 (46.4844)  time: 7.3077  data: 7.0715  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 3.8843 (4.0147)  acc1: 22.6415 (23.0400)  acc5: 43.3962 (44.3200)  time: 1.6069  data: 1.4144  max mem: 55573
Test: Total time: 0:00:08 (1.6270 s / it)
* Acc@1 20.078 Acc@5 40.322 loss 4.255
Accuracy of the model EMA on 50000 test images: 20.1%
Max EMA accuracy: 20.08%
Epoch: [125]  [  0/156]  eta: 0:32:59  lr: 0.005531  min_lr: 0.005531  loss: 3.1081 (3.1081)  weight_decay: 0.0500 (0.0500)  time: 12.6869  data: 10.0959  max mem: 55573
Epoch: [125]  [ 10/156]  eta: 0:04:07  lr: 0.005528  min_lr: 0.005528  loss: 3.1451 (3.2051)  weight_decay: 0.0500 (0.0500)  time: 1.6954  data: 0.9182  max mem: 55573
Epoch: [125]  [ 20/156]  eta: 0:02:39  lr: 0.005526  min_lr: 0.005526  loss: 3.5250 (3.3328)  weight_decay: 0.0500 (0.0500)  time: 0.5946  data: 0.0005  max mem: 55573
Epoch: [125]  [ 30/156]  eta: 0:02:04  lr: 0.005523  min_lr: 0.005523  loss: 3.5216 (3.3493)  weight_decay: 0.0500 (0.0500)  time: 0.5947  data: 0.0005  max mem: 55573
Epoch: [125]  [ 40/156]  eta: 0:01:43  lr: 0.005520  min_lr: 0.005520  loss: 3.6035 (3.4071)  weight_decay: 0.0500 (0.0500)  time: 0.5949  data: 0.0005  max mem: 55573
Epoch: [125]  [ 50/156]  eta: 0:01:28  lr: 0.005518  min_lr: 0.005518  loss: 3.6680 (3.4102)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0005  max mem: 55573
Epoch: [125]  [ 60/156]  eta: 0:01:16  lr: 0.005515  min_lr: 0.005515  loss: 3.5545 (3.4116)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [125]  [ 70/156]  eta: 0:01:05  lr: 0.005512  min_lr: 0.005512  loss: 3.6114 (3.4655)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [125]  [ 80/156]  eta: 0:00:56  lr: 0.005510  min_lr: 0.005510  loss: 3.6973 (3.4769)  weight_decay: 0.0500 (0.0500)  time: 0.5931  data: 0.0005  max mem: 55573
Epoch: [125]  [ 90/156]  eta: 0:00:47  lr: 0.005507  min_lr: 0.005507  loss: 3.5256 (3.4585)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0005  max mem: 55573
Epoch: [125]  [100/156]  eta: 0:00:39  lr: 0.005504  min_lr: 0.005504  loss: 3.3531 (3.4478)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [125]  [110/156]  eta: 0:00:32  lr: 0.005502  min_lr: 0.005502  loss: 3.5445 (3.4610)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [125]  [120/156]  eta: 0:00:24  lr: 0.005499  min_lr: 0.005499  loss: 3.5988 (3.4638)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [125]  [130/156]  eta: 0:00:17  lr: 0.005496  min_lr: 0.005496  loss: 3.6568 (3.4730)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0011  max mem: 55573
Epoch: [125]  [140/156]  eta: 0:00:10  lr: 0.005494  min_lr: 0.005494  loss: 3.6795 (3.4676)  weight_decay: 0.0500 (0.0500)  time: 0.5867  data: 0.0009  max mem: 55573
Epoch: [125]  [150/156]  eta: 0:00:04  lr: 0.005491  min_lr: 0.005491  loss: 3.3142 (3.4489)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [125]  [155/156]  eta: 0:00:00  lr: 0.005490  min_lr: 0.005490  loss: 3.3038 (3.4396)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [125] Total time: 0:01:45 (0.6751 s / it)
Averaged stats: lr: 0.005490  min_lr: 0.005490  loss: 3.3038 (3.4466)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.1947 (1.1947)  acc1: 76.6927 (76.6927)  acc5: 93.4896 (93.4896)  time: 7.0065  data: 6.7695  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2923 (1.2914)  acc1: 74.3490 (72.2240)  acc5: 93.0990 (91.3600)  time: 1.5478  data: 1.3540  max mem: 55573
Test: Total time: 0:00:08 (1.6114 s / it)
* Acc@1 73.162 Acc@5 91.714 loss 1.305
Accuracy of the model on the 50000 test images: 73.2%
Max accuracy: 73.67%
Test:  [0/5]  eta: 0:00:34  loss: 3.5753 (3.5753)  acc1: 25.5208 (25.5208)  acc5: 49.7396 (49.7396)  time: 6.8670  data: 6.6308  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 3.6983 (3.8349)  acc1: 24.5283 (24.7360)  acc5: 47.1698 (46.8800)  time: 1.5274  data: 1.3349  max mem: 55573
Test: Total time: 0:00:07 (1.5483 s / it)
* Acc@1 22.026 Acc@5 43.390 loss 4.069
Accuracy of the model EMA on 50000 test images: 22.0%
Max EMA accuracy: 22.03%
Epoch: [126]  [  0/156]  eta: 0:33:05  lr: 0.005489  min_lr: 0.005489  loss: 2.8353 (2.8353)  weight_decay: 0.0500 (0.0500)  time: 12.7303  data: 7.6435  max mem: 55573
Epoch: [126]  [ 10/156]  eta: 0:04:07  lr: 0.005487  min_lr: 0.005487  loss: 3.5135 (3.3157)  weight_decay: 0.0500 (0.0500)  time: 1.6964  data: 0.6952  max mem: 55573
Epoch: [126]  [ 20/156]  eta: 0:02:39  lr: 0.005484  min_lr: 0.005484  loss: 3.5135 (3.3340)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [126]  [ 30/156]  eta: 0:02:03  lr: 0.005481  min_lr: 0.005481  loss: 3.4694 (3.3526)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [126]  [ 40/156]  eta: 0:01:42  lr: 0.005479  min_lr: 0.005479  loss: 3.4891 (3.3815)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [126]  [ 50/156]  eta: 0:01:27  lr: 0.005476  min_lr: 0.005476  loss: 3.5947 (3.4282)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [126]  [ 60/156]  eta: 0:01:15  lr: 0.005473  min_lr: 0.005473  loss: 3.4067 (3.3784)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [126]  [ 70/156]  eta: 0:01:05  lr: 0.005471  min_lr: 0.005471  loss: 3.4067 (3.4176)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [126]  [ 80/156]  eta: 0:00:56  lr: 0.005468  min_lr: 0.005468  loss: 3.6085 (3.4226)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [126]  [ 90/156]  eta: 0:00:47  lr: 0.005465  min_lr: 0.005465  loss: 3.5424 (3.4277)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [126]  [100/156]  eta: 0:00:39  lr: 0.005463  min_lr: 0.005463  loss: 3.5424 (3.4197)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [126]  [110/156]  eta: 0:00:32  lr: 0.005460  min_lr: 0.005460  loss: 3.7095 (3.4468)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [126]  [120/156]  eta: 0:00:24  lr: 0.005457  min_lr: 0.005457  loss: 3.6630 (3.4286)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [126]  [130/156]  eta: 0:00:17  lr: 0.005455  min_lr: 0.005455  loss: 3.3177 (3.4256)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0011  max mem: 55573
Epoch: [126]  [140/156]  eta: 0:00:10  lr: 0.005452  min_lr: 0.005452  loss: 3.3177 (3.4084)  weight_decay: 0.0500 (0.0500)  time: 0.5859  data: 0.0009  max mem: 55573
Epoch: [126]  [150/156]  eta: 0:00:04  lr: 0.005449  min_lr: 0.005449  loss: 3.4682 (3.4079)  weight_decay: 0.0500 (0.0500)  time: 0.5832  data: 0.0001  max mem: 55573
Epoch: [126]  [155/156]  eta: 0:00:00  lr: 0.005448  min_lr: 0.005448  loss: 3.4420 (3.4049)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [126] Total time: 0:01:45 (0.6741 s / it)
Averaged stats: lr: 0.005448  min_lr: 0.005448  loss: 3.4420 (3.4339)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.1381 (1.1381)  acc1: 77.2135 (77.2135)  acc5: 93.6198 (93.6198)  time: 6.9261  data: 6.6892  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2437 (1.2269)  acc1: 73.4375 (72.4480)  acc5: 93.2292 (91.1680)  time: 1.5306  data: 1.3379  max mem: 55573
Test: Total time: 0:00:07 (1.5474 s / it)
* Acc@1 73.590 Acc@5 91.870 loss 1.226
Accuracy of the model on the 50000 test images: 73.6%
Max accuracy: 73.67%
Test:  [0/5]  eta: 0:00:35  loss: 3.4222 (3.4222)  acc1: 27.2135 (27.2135)  acc5: 52.4740 (52.4740)  time: 7.1265  data: 6.8905  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 3.5273 (3.6660)  acc1: 26.4151 (26.4320)  acc5: 49.0566 (49.7280)  time: 1.5706  data: 1.3782  max mem: 55573
Test: Total time: 0:00:07 (1.5899 s / it)
* Acc@1 24.020 Acc@5 46.490 loss 3.892
Accuracy of the model EMA on 50000 test images: 24.0%
Max EMA accuracy: 24.02%
Epoch: [127]  [  0/156]  eta: 0:34:28  lr: 0.005448  min_lr: 0.005448  loss: 3.2270 (3.2270)  weight_decay: 0.0500 (0.0500)  time: 13.2623  data: 11.4714  max mem: 55573
Epoch: [127]  [ 10/156]  eta: 0:04:20  lr: 0.005445  min_lr: 0.005445  loss: 3.5844 (3.3883)  weight_decay: 0.0500 (0.0500)  time: 1.7870  data: 1.0432  max mem: 55573
Epoch: [127]  [ 20/156]  eta: 0:02:45  lr: 0.005442  min_lr: 0.005442  loss: 3.4925 (3.3111)  weight_decay: 0.0500 (0.0500)  time: 0.6147  data: 0.0004  max mem: 55573
Epoch: [127]  [ 30/156]  eta: 0:02:07  lr: 0.005440  min_lr: 0.005440  loss: 3.4748 (3.3490)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [127]  [ 40/156]  eta: 0:01:45  lr: 0.005437  min_lr: 0.005437  loss: 3.3540 (3.3028)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0005  max mem: 55573
Epoch: [127]  [ 50/156]  eta: 0:01:30  lr: 0.005434  min_lr: 0.005434  loss: 3.5032 (3.3658)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [127]  [ 60/156]  eta: 0:01:17  lr: 0.005432  min_lr: 0.005432  loss: 3.5543 (3.3686)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [127]  [ 70/156]  eta: 0:01:06  lr: 0.005429  min_lr: 0.005429  loss: 3.7218 (3.3958)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0004  max mem: 55573
Epoch: [127]  [ 80/156]  eta: 0:00:57  lr: 0.005426  min_lr: 0.005426  loss: 3.7228 (3.4049)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [127]  [ 90/156]  eta: 0:00:48  lr: 0.005424  min_lr: 0.005424  loss: 3.6538 (3.4228)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [127]  [100/156]  eta: 0:00:40  lr: 0.005421  min_lr: 0.005421  loss: 3.3990 (3.4071)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [127]  [110/156]  eta: 0:00:32  lr: 0.005418  min_lr: 0.005418  loss: 3.4356 (3.4261)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [127]  [120/156]  eta: 0:00:25  lr: 0.005416  min_lr: 0.005416  loss: 3.6743 (3.4337)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0005  max mem: 55573
Epoch: [127]  [130/156]  eta: 0:00:17  lr: 0.005413  min_lr: 0.005413  loss: 3.4705 (3.4395)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0011  max mem: 55573
Epoch: [127]  [140/156]  eta: 0:00:10  lr: 0.005410  min_lr: 0.005410  loss: 3.4213 (3.4352)  weight_decay: 0.0500 (0.0500)  time: 0.5855  data: 0.0009  max mem: 55573
Epoch: [127]  [150/156]  eta: 0:00:04  lr: 0.005407  min_lr: 0.005407  loss: 3.5466 (3.4231)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [127]  [155/156]  eta: 0:00:00  lr: 0.005406  min_lr: 0.005406  loss: 3.6005 (3.4343)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [127] Total time: 0:01:46 (0.6803 s / it)
Averaged stats: lr: 0.005406  min_lr: 0.005406  loss: 3.6005 (3.4259)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.1132 (1.1132)  acc1: 77.9948 (77.9948)  acc5: 93.6198 (93.6198)  time: 6.8553  data: 6.6177  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1516 (1.2088)  acc1: 74.2188 (72.5120)  acc5: 93.6198 (91.2960)  time: 1.5175  data: 1.3236  max mem: 55573
Test: Total time: 0:00:07 (1.5627 s / it)
* Acc@1 73.518 Acc@5 92.044 loss 1.177
Accuracy of the model on the 50000 test images: 73.5%
Max accuracy: 73.67%
Test:  [0/5]  eta: 0:00:35  loss: 3.2751 (3.2751)  acc1: 29.2969 (29.2969)  acc5: 56.2500 (56.2500)  time: 7.1857  data: 6.9497  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 3.3645 (3.5025)  acc1: 29.2969 (28.8960)  acc5: 54.7170 (52.5120)  time: 1.5831  data: 1.3902  max mem: 55573
Test: Total time: 0:00:08 (1.6031 s / it)
* Acc@1 26.210 Acc@5 49.390 loss 3.721
Accuracy of the model EMA on 50000 test images: 26.2%
Max EMA accuracy: 26.21%
Epoch: [128]  [  0/156]  eta: 0:37:18  lr: 0.005406  min_lr: 0.005406  loss: 3.4356 (3.4356)  weight_decay: 0.0500 (0.0500)  time: 14.3469  data: 10.2503  max mem: 55573
Epoch: [128]  [ 10/156]  eta: 0:04:33  lr: 0.005403  min_lr: 0.005403  loss: 3.3105 (3.3497)  weight_decay: 0.0500 (0.0500)  time: 1.8765  data: 0.9707  max mem: 55573
Epoch: [128]  [ 20/156]  eta: 0:02:51  lr: 0.005400  min_lr: 0.005400  loss: 3.3105 (3.3046)  weight_decay: 0.0500 (0.0500)  time: 0.6097  data: 0.0216  max mem: 55573
Epoch: [128]  [ 30/156]  eta: 0:02:11  lr: 0.005398  min_lr: 0.005398  loss: 3.3565 (3.3161)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [128]  [ 40/156]  eta: 0:01:48  lr: 0.005395  min_lr: 0.005395  loss: 3.3565 (3.3363)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [128]  [ 50/156]  eta: 0:01:31  lr: 0.005392  min_lr: 0.005392  loss: 3.5754 (3.3835)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [128]  [ 60/156]  eta: 0:01:18  lr: 0.005390  min_lr: 0.005390  loss: 3.6426 (3.3804)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [128]  [ 70/156]  eta: 0:01:07  lr: 0.005387  min_lr: 0.005387  loss: 3.6650 (3.4248)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [128]  [ 80/156]  eta: 0:00:58  lr: 0.005384  min_lr: 0.005384  loss: 3.7378 (3.4474)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [128]  [ 90/156]  eta: 0:00:49  lr: 0.005382  min_lr: 0.005382  loss: 3.5677 (3.4384)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0005  max mem: 55573
Epoch: [128]  [100/156]  eta: 0:00:40  lr: 0.005379  min_lr: 0.005379  loss: 3.5681 (3.4525)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [128]  [110/156]  eta: 0:00:33  lr: 0.005376  min_lr: 0.005376  loss: 3.7013 (3.4751)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [128]  [120/156]  eta: 0:00:25  lr: 0.005373  min_lr: 0.005373  loss: 3.4854 (3.4634)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [128]  [130/156]  eta: 0:00:18  lr: 0.005371  min_lr: 0.005371  loss: 3.5374 (3.4850)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0011  max mem: 55573
Epoch: [128]  [140/156]  eta: 0:00:11  lr: 0.005368  min_lr: 0.005368  loss: 3.6962 (3.4791)  weight_decay: 0.0500 (0.0500)  time: 0.5851  data: 0.0009  max mem: 55573
Epoch: [128]  [150/156]  eta: 0:00:04  lr: 0.005365  min_lr: 0.005365  loss: 3.2180 (3.4541)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [128]  [155/156]  eta: 0:00:00  lr: 0.005364  min_lr: 0.005364  loss: 3.1090 (3.4458)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [128] Total time: 0:01:47 (0.6870 s / it)
Averaged stats: lr: 0.005364  min_lr: 0.005364  loss: 3.1090 (3.4380)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.1135 (1.1135)  acc1: 76.9531 (76.9531)  acc5: 94.9219 (94.9219)  time: 7.2825  data: 7.0450  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1691 (1.2008)  acc1: 76.4323 (73.4400)  acc5: 93.6198 (92.0960)  time: 1.6030  data: 1.4091  max mem: 55573
Test: Total time: 0:00:08 (1.6522 s / it)
* Acc@1 73.894 Acc@5 92.274 loss 1.205
Accuracy of the model on the 50000 test images: 73.9%
Max accuracy: 73.89%
Test:  [0/5]  eta: 0:00:35  loss: 3.1412 (3.1412)  acc1: 30.9896 (30.9896)  acc5: 59.1146 (59.1146)  time: 7.0384  data: 6.8023  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 3.2147 (3.3515)  acc1: 30.9896 (30.3360)  acc5: 58.4906 (54.8160)  time: 1.5530  data: 1.3606  max mem: 55573
Test: Total time: 0:00:07 (1.5712 s / it)
* Acc@1 28.082 Acc@5 51.964 loss 3.563
Accuracy of the model EMA on 50000 test images: 28.1%
Max EMA accuracy: 28.08%
Epoch: [129]  [  0/156]  eta: 0:35:55  lr: 0.005364  min_lr: 0.005364  loss: 4.2157 (4.2157)  weight_decay: 0.0500 (0.0500)  time: 13.8188  data: 9.0143  max mem: 55573
Epoch: [129]  [ 10/156]  eta: 0:04:23  lr: 0.005361  min_lr: 0.005361  loss: 3.5320 (3.5059)  weight_decay: 0.0500 (0.0500)  time: 1.8048  data: 0.8199  max mem: 55573
Epoch: [129]  [ 20/156]  eta: 0:02:46  lr: 0.005358  min_lr: 0.005358  loss: 3.3651 (3.4173)  weight_decay: 0.0500 (0.0500)  time: 0.5969  data: 0.0005  max mem: 55573
Epoch: [129]  [ 30/156]  eta: 0:02:08  lr: 0.005356  min_lr: 0.005356  loss: 3.4301 (3.4317)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [129]  [ 40/156]  eta: 0:01:46  lr: 0.005353  min_lr: 0.005353  loss: 3.4561 (3.4144)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [129]  [ 50/156]  eta: 0:01:30  lr: 0.005350  min_lr: 0.005350  loss: 3.5540 (3.4405)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [129]  [ 60/156]  eta: 0:01:17  lr: 0.005347  min_lr: 0.005347  loss: 3.2729 (3.3576)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [129]  [ 70/156]  eta: 0:01:06  lr: 0.005345  min_lr: 0.005345  loss: 3.1980 (3.3704)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [129]  [ 80/156]  eta: 0:00:57  lr: 0.005342  min_lr: 0.005342  loss: 3.4207 (3.3718)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [129]  [ 90/156]  eta: 0:00:48  lr: 0.005339  min_lr: 0.005339  loss: 3.4207 (3.3703)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [129]  [100/156]  eta: 0:00:40  lr: 0.005337  min_lr: 0.005337  loss: 3.4339 (3.3745)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0005  max mem: 55573
Epoch: [129]  [110/156]  eta: 0:00:32  lr: 0.005334  min_lr: 0.005334  loss: 3.5953 (3.3798)  weight_decay: 0.0500 (0.0500)  time: 0.5953  data: 0.0005  max mem: 55573
Epoch: [129]  [120/156]  eta: 0:00:25  lr: 0.005331  min_lr: 0.005331  loss: 3.6774 (3.3993)  weight_decay: 0.0500 (0.0500)  time: 0.5930  data: 0.0005  max mem: 55573
Epoch: [129]  [130/156]  eta: 0:00:18  lr: 0.005329  min_lr: 0.005329  loss: 3.5978 (3.3782)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0010  max mem: 55573
Epoch: [129]  [140/156]  eta: 0:00:10  lr: 0.005326  min_lr: 0.005326  loss: 3.1545 (3.3811)  weight_decay: 0.0500 (0.0500)  time: 0.5854  data: 0.0009  max mem: 55573
Epoch: [129]  [150/156]  eta: 0:00:04  lr: 0.005323  min_lr: 0.005323  loss: 3.5188 (3.3822)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [129]  [155/156]  eta: 0:00:00  lr: 0.005322  min_lr: 0.005322  loss: 3.6009 (3.3866)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [129] Total time: 0:01:46 (0.6823 s / it)
Averaged stats: lr: 0.005322  min_lr: 0.005322  loss: 3.6009 (3.4169)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.1174 (1.1174)  acc1: 78.2552 (78.2552)  acc5: 94.0104 (94.0104)  time: 6.7744  data: 6.5365  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1966 (1.2679)  acc1: 75.6510 (72.5440)  acc5: 94.0104 (91.7440)  time: 1.5002  data: 1.3074  max mem: 55573
Test: Total time: 0:00:07 (1.5178 s / it)
* Acc@1 72.810 Acc@5 91.508 loss 1.270
Accuracy of the model on the 50000 test images: 72.8%
Max accuracy: 73.89%
Test:  [0/5]  eta: 0:00:35  loss: 3.0198 (3.0198)  acc1: 32.2917 (32.2917)  acc5: 61.4583 (61.4583)  time: 7.1238  data: 6.8879  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 3.0808 (3.2135)  acc1: 32.2917 (32.1280)  acc5: 61.0677 (57.2480)  time: 1.5701  data: 1.3777  max mem: 55573
Test: Total time: 0:00:08 (1.6006 s / it)
* Acc@1 29.920 Acc@5 54.408 loss 3.418
Accuracy of the model EMA on 50000 test images: 29.9%
Max EMA accuracy: 29.92%
Epoch: [130]  [  0/156]  eta: 0:31:46  lr: 0.005321  min_lr: 0.005321  loss: 3.9553 (3.9553)  weight_decay: 0.0500 (0.0500)  time: 12.2217  data: 8.2492  max mem: 55573
Epoch: [130]  [ 10/156]  eta: 0:04:12  lr: 0.005319  min_lr: 0.005319  loss: 3.7141 (3.5385)  weight_decay: 0.0500 (0.0500)  time: 1.7284  data: 0.7504  max mem: 55573
Epoch: [130]  [ 20/156]  eta: 0:02:41  lr: 0.005316  min_lr: 0.005316  loss: 3.6431 (3.5003)  weight_decay: 0.0500 (0.0500)  time: 0.6347  data: 0.0005  max mem: 55573
Epoch: [130]  [ 30/156]  eta: 0:02:05  lr: 0.005313  min_lr: 0.005313  loss: 3.5533 (3.4494)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [130]  [ 40/156]  eta: 0:01:43  lr: 0.005311  min_lr: 0.005311  loss: 3.4815 (3.4354)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [130]  [ 50/156]  eta: 0:01:28  lr: 0.005308  min_lr: 0.005308  loss: 3.4376 (3.4289)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [130]  [ 60/156]  eta: 0:01:16  lr: 0.005305  min_lr: 0.005305  loss: 3.4466 (3.4671)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [130]  [ 70/156]  eta: 0:01:05  lr: 0.005302  min_lr: 0.005302  loss: 3.2459 (3.4083)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [130]  [ 80/156]  eta: 0:00:56  lr: 0.005300  min_lr: 0.005300  loss: 3.0474 (3.3746)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [130]  [ 90/156]  eta: 0:00:48  lr: 0.005297  min_lr: 0.005297  loss: 3.2653 (3.4045)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [130]  [100/156]  eta: 0:00:40  lr: 0.005294  min_lr: 0.005294  loss: 3.5904 (3.4106)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [130]  [110/156]  eta: 0:00:32  lr: 0.005292  min_lr: 0.005292  loss: 3.5904 (3.4246)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [130]  [120/156]  eta: 0:00:24  lr: 0.005289  min_lr: 0.005289  loss: 3.4462 (3.4211)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0005  max mem: 55573
Epoch: [130]  [130/156]  eta: 0:00:17  lr: 0.005286  min_lr: 0.005286  loss: 3.4014 (3.4082)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0007  max mem: 55573
Epoch: [130]  [140/156]  eta: 0:00:10  lr: 0.005283  min_lr: 0.005283  loss: 3.5326 (3.4193)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0005  max mem: 55573
Epoch: [130]  [150/156]  eta: 0:00:04  lr: 0.005281  min_lr: 0.005281  loss: 3.6963 (3.4313)  weight_decay: 0.0500 (0.0500)  time: 0.5828  data: 0.0002  max mem: 55573
Epoch: [130]  [155/156]  eta: 0:00:00  lr: 0.005279  min_lr: 0.005279  loss: 3.6567 (3.4314)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0002  max mem: 55573
Epoch: [130] Total time: 0:01:45 (0.6766 s / it)
Averaged stats: lr: 0.005279  min_lr: 0.005279  loss: 3.6567 (3.4331)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.1280 (1.1280)  acc1: 77.0833 (77.0833)  acc5: 94.5312 (94.5312)  time: 7.3047  data: 7.0671  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2802 (1.2445)  acc1: 73.3073 (72.0000)  acc5: 92.8385 (91.5200)  time: 1.6083  data: 1.4137  max mem: 55573
Test: Total time: 0:00:08 (1.6610 s / it)
* Acc@1 73.380 Acc@5 91.938 loss 1.210
Accuracy of the model on the 50000 test images: 73.4%
Max accuracy: 73.89%
Test:  [0/5]  eta: 0:00:35  loss: 2.9075 (2.9075)  acc1: 34.5052 (34.5052)  acc5: 63.4115 (63.4115)  time: 7.1632  data: 6.9272  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.9586 (3.0854)  acc1: 34.5052 (34.0160)  acc5: 62.3698 (59.3280)  time: 1.5780  data: 1.3855  max mem: 55573
Test: Total time: 0:00:07 (1.5935 s / it)
* Acc@1 31.782 Acc@5 56.704 loss 3.283
Accuracy of the model EMA on 50000 test images: 31.8%
Max EMA accuracy: 31.78%
Epoch: [131]  [  0/156]  eta: 0:38:01  lr: 0.005279  min_lr: 0.005279  loss: 3.3231 (3.3231)  weight_decay: 0.0500 (0.0500)  time: 14.6245  data: 10.1662  max mem: 55573
Epoch: [131]  [ 10/156]  eta: 0:04:32  lr: 0.005276  min_lr: 0.005276  loss: 3.3231 (3.2244)  weight_decay: 0.0500 (0.0500)  time: 1.8630  data: 0.9247  max mem: 55573
Epoch: [131]  [ 20/156]  eta: 0:02:51  lr: 0.005274  min_lr: 0.005274  loss: 3.3810 (3.3555)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [131]  [ 30/156]  eta: 0:02:11  lr: 0.005271  min_lr: 0.005271  loss: 3.4479 (3.3762)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [131]  [ 40/156]  eta: 0:01:48  lr: 0.005268  min_lr: 0.005268  loss: 3.5097 (3.4092)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [131]  [ 50/156]  eta: 0:01:31  lr: 0.005265  min_lr: 0.005265  loss: 3.5586 (3.4338)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [131]  [ 60/156]  eta: 0:01:18  lr: 0.005263  min_lr: 0.005263  loss: 3.4774 (3.4153)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [131]  [ 70/156]  eta: 0:01:07  lr: 0.005260  min_lr: 0.005260  loss: 3.4774 (3.4171)  weight_decay: 0.0500 (0.0500)  time: 0.5930  data: 0.0005  max mem: 55573
Epoch: [131]  [ 80/156]  eta: 0:00:58  lr: 0.005257  min_lr: 0.005257  loss: 3.6349 (3.4280)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [131]  [ 90/156]  eta: 0:00:49  lr: 0.005254  min_lr: 0.005254  loss: 3.6349 (3.4292)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [131]  [100/156]  eta: 0:00:40  lr: 0.005252  min_lr: 0.005252  loss: 3.6360 (3.4319)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [131]  [110/156]  eta: 0:00:32  lr: 0.005249  min_lr: 0.005249  loss: 3.5109 (3.4437)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [131]  [120/156]  eta: 0:00:25  lr: 0.005246  min_lr: 0.005246  loss: 3.3486 (3.4224)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [131]  [130/156]  eta: 0:00:18  lr: 0.005244  min_lr: 0.005244  loss: 3.2951 (3.4345)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0010  max mem: 55573
Epoch: [131]  [140/156]  eta: 0:00:11  lr: 0.005241  min_lr: 0.005241  loss: 3.6640 (3.4442)  weight_decay: 0.0500 (0.0500)  time: 0.5841  data: 0.0009  max mem: 55573
Epoch: [131]  [150/156]  eta: 0:00:04  lr: 0.005238  min_lr: 0.005238  loss: 3.6600 (3.4602)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [131]  [155/156]  eta: 0:00:00  lr: 0.005237  min_lr: 0.005237  loss: 3.6014 (3.4559)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [131] Total time: 0:01:47 (0.6861 s / it)
Averaged stats: lr: 0.005237  min_lr: 0.005237  loss: 3.6014 (3.4356)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.2474 (1.2474)  acc1: 77.7344 (77.7344)  acc5: 94.1406 (94.1406)  time: 7.2972  data: 7.0597  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3605 (1.3556)  acc1: 73.8281 (72.3840)  acc5: 94.1406 (91.9360)  time: 1.6066  data: 1.4122  max mem: 55573
Test: Total time: 0:00:08 (1.6490 s / it)
* Acc@1 73.714 Acc@5 92.022 loss 1.346
Accuracy of the model on the 50000 test images: 73.7%
Max accuracy: 73.89%
Test:  [0/5]  eta: 0:00:33  loss: 2.7979 (2.7979)  acc1: 36.4583 (36.4583)  acc5: 65.7552 (65.7552)  time: 6.7201  data: 6.4839  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.8423 (2.9628)  acc1: 36.4583 (35.7760)  acc5: 63.8021 (61.3760)  time: 1.5048  data: 1.3125  max mem: 55573
Test: Total time: 0:00:07 (1.5272 s / it)
* Acc@1 33.622 Acc@5 58.814 loss 3.153
Accuracy of the model EMA on 50000 test images: 33.6%
Max EMA accuracy: 33.62%
Epoch: [132]  [  0/156]  eta: 0:33:58  lr: 0.005236  min_lr: 0.005236  loss: 3.6914 (3.6914)  weight_decay: 0.0500 (0.0500)  time: 13.0658  data: 9.5003  max mem: 55573
Epoch: [132]  [ 10/156]  eta: 0:04:24  lr: 0.005234  min_lr: 0.005234  loss: 3.6914 (3.5896)  weight_decay: 0.0500 (0.0500)  time: 1.8120  data: 0.8641  max mem: 55573
Epoch: [132]  [ 20/156]  eta: 0:02:47  lr: 0.005231  min_lr: 0.005231  loss: 3.6276 (3.5921)  weight_decay: 0.0500 (0.0500)  time: 0.6391  data: 0.0005  max mem: 55573
Epoch: [132]  [ 30/156]  eta: 0:02:09  lr: 0.005228  min_lr: 0.005228  loss: 3.6276 (3.5674)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [132]  [ 40/156]  eta: 0:01:46  lr: 0.005225  min_lr: 0.005225  loss: 3.4762 (3.4883)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [132]  [ 50/156]  eta: 0:01:30  lr: 0.005223  min_lr: 0.005223  loss: 3.4139 (3.4509)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [132]  [ 60/156]  eta: 0:01:17  lr: 0.005220  min_lr: 0.005220  loss: 3.5697 (3.4740)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [132]  [ 70/156]  eta: 0:01:07  lr: 0.005217  min_lr: 0.005217  loss: 3.6572 (3.4774)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0004  max mem: 55573
Epoch: [132]  [ 80/156]  eta: 0:00:57  lr: 0.005215  min_lr: 0.005215  loss: 3.3808 (3.4578)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [132]  [ 90/156]  eta: 0:00:48  lr: 0.005212  min_lr: 0.005212  loss: 3.5228 (3.4684)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [132]  [100/156]  eta: 0:00:40  lr: 0.005209  min_lr: 0.005209  loss: 3.5214 (3.4591)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [132]  [110/156]  eta: 0:00:32  lr: 0.005206  min_lr: 0.005206  loss: 3.5174 (3.4665)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [132]  [120/156]  eta: 0:00:25  lr: 0.005204  min_lr: 0.005204  loss: 3.5142 (3.4609)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [132]  [130/156]  eta: 0:00:18  lr: 0.005201  min_lr: 0.005201  loss: 3.5037 (3.4535)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0010  max mem: 55573
Epoch: [132]  [140/156]  eta: 0:00:10  lr: 0.005198  min_lr: 0.005198  loss: 3.5066 (3.4530)  weight_decay: 0.0500 (0.0500)  time: 0.5865  data: 0.0008  max mem: 55573
Epoch: [132]  [150/156]  eta: 0:00:04  lr: 0.005195  min_lr: 0.005195  loss: 3.4020 (3.4430)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [132]  [155/156]  eta: 0:00:00  lr: 0.005194  min_lr: 0.005194  loss: 3.2651 (3.4483)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [132] Total time: 0:01:46 (0.6822 s / it)
Averaged stats: lr: 0.005194  min_lr: 0.005194  loss: 3.2651 (3.4226)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.1740 (1.1740)  acc1: 78.6458 (78.6458)  acc5: 94.5312 (94.5312)  time: 7.0993  data: 6.8619  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3060 (1.3327)  acc1: 75.1302 (72.5440)  acc5: 93.7500 (91.8400)  time: 1.5664  data: 1.3725  max mem: 55573
Test: Total time: 0:00:08 (1.6066 s / it)
* Acc@1 73.526 Acc@5 92.060 loss 1.304
Accuracy of the model on the 50000 test images: 73.5%
Max accuracy: 73.89%
Test:  [0/5]  eta: 0:00:35  loss: 2.6944 (2.6944)  acc1: 38.5417 (38.5417)  acc5: 67.3177 (67.3177)  time: 7.0955  data: 6.8596  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.7321 (2.8451)  acc1: 38.5417 (37.5360)  acc5: 65.7552 (63.1680)  time: 1.5644  data: 1.3720  max mem: 55573
Test: Total time: 0:00:07 (1.5877 s / it)
* Acc@1 35.352 Acc@5 60.788 loss 3.028
Accuracy of the model EMA on 50000 test images: 35.4%
Max EMA accuracy: 35.35%
Epoch: [133]  [  0/156]  eta: 0:31:17  lr: 0.005194  min_lr: 0.005194  loss: 3.6948 (3.6948)  weight_decay: 0.0500 (0.0500)  time: 12.0321  data: 9.4314  max mem: 55573
Epoch: [133]  [ 10/156]  eta: 0:04:06  lr: 0.005191  min_lr: 0.005191  loss: 3.5615 (3.3795)  weight_decay: 0.0500 (0.0500)  time: 1.6868  data: 0.8578  max mem: 55573
Epoch: [133]  [ 20/156]  eta: 0:02:38  lr: 0.005188  min_lr: 0.005188  loss: 3.5918 (3.4613)  weight_decay: 0.0500 (0.0500)  time: 0.6223  data: 0.0004  max mem: 55573
Epoch: [133]  [ 30/156]  eta: 0:02:03  lr: 0.005185  min_lr: 0.005185  loss: 3.6189 (3.4302)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [133]  [ 40/156]  eta: 0:01:42  lr: 0.005183  min_lr: 0.005183  loss: 3.6440 (3.5012)  weight_decay: 0.0500 (0.0500)  time: 0.5953  data: 0.0004  max mem: 55573
Epoch: [133]  [ 50/156]  eta: 0:01:27  lr: 0.005180  min_lr: 0.005180  loss: 3.6440 (3.4325)  weight_decay: 0.0500 (0.0500)  time: 0.5963  data: 0.0004  max mem: 55573
Epoch: [133]  [ 60/156]  eta: 0:01:15  lr: 0.005177  min_lr: 0.005177  loss: 3.2443 (3.4066)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [133]  [ 70/156]  eta: 0:01:05  lr: 0.005174  min_lr: 0.005174  loss: 3.6011 (3.4259)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0004  max mem: 55573
Epoch: [133]  [ 80/156]  eta: 0:00:56  lr: 0.005172  min_lr: 0.005172  loss: 3.6011 (3.4198)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [133]  [ 90/156]  eta: 0:00:47  lr: 0.005169  min_lr: 0.005169  loss: 3.6175 (3.4422)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [133]  [100/156]  eta: 0:00:39  lr: 0.005166  min_lr: 0.005166  loss: 3.5788 (3.4428)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [133]  [110/156]  eta: 0:00:32  lr: 0.005163  min_lr: 0.005163  loss: 3.5272 (3.4257)  weight_decay: 0.0500 (0.0500)  time: 0.5887  data: 0.0004  max mem: 55573
Epoch: [133]  [120/156]  eta: 0:00:24  lr: 0.005161  min_lr: 0.005161  loss: 3.2500 (3.4007)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [133]  [130/156]  eta: 0:00:17  lr: 0.005158  min_lr: 0.005158  loss: 3.5360 (3.4206)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0010  max mem: 55573
Epoch: [133]  [140/156]  eta: 0:00:10  lr: 0.005155  min_lr: 0.005155  loss: 3.4542 (3.4041)  weight_decay: 0.0500 (0.0500)  time: 0.5881  data: 0.0008  max mem: 55573
Epoch: [133]  [150/156]  eta: 0:00:04  lr: 0.005152  min_lr: 0.005152  loss: 3.4176 (3.4080)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [133]  [155/156]  eta: 0:00:00  lr: 0.005151  min_lr: 0.005151  loss: 3.4589 (3.4108)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [133] Total time: 0:01:45 (0.6743 s / it)
Averaged stats: lr: 0.005151  min_lr: 0.005151  loss: 3.4589 (3.4212)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.0729 (1.0729)  acc1: 77.3438 (77.3438)  acc5: 94.7917 (94.7917)  time: 6.8596  data: 6.6222  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2047 (1.2235)  acc1: 75.6510 (72.8000)  acc5: 94.6615 (92.0000)  time: 1.5184  data: 1.3245  max mem: 55573
Test: Total time: 0:00:07 (1.5633 s / it)
* Acc@1 74.146 Acc@5 92.282 loss 1.209
Accuracy of the model on the 50000 test images: 74.1%
Max accuracy: 74.15%
Test:  [0/5]  eta: 0:00:33  loss: 2.5982 (2.5982)  acc1: 39.9740 (39.9740)  acc5: 68.4896 (68.4896)  time: 6.7673  data: 6.5310  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.6300 (2.7349)  acc1: 39.9740 (38.9760)  acc5: 68.0990 (65.0880)  time: 1.4988  data: 1.3063  max mem: 55573
Test: Total time: 0:00:07 (1.5137 s / it)
* Acc@1 36.910 Acc@5 62.774 loss 2.911
Accuracy of the model EMA on 50000 test images: 36.9%
Max EMA accuracy: 36.91%
Epoch: [134]  [  0/156]  eta: 0:36:41  lr: 0.005151  min_lr: 0.005151  loss: 2.9701 (2.9701)  weight_decay: 0.0500 (0.0500)  time: 14.1116  data: 11.3810  max mem: 55573
Epoch: [134]  [ 10/156]  eta: 0:04:25  lr: 0.005148  min_lr: 0.005148  loss: 3.1675 (3.2966)  weight_decay: 0.0500 (0.0500)  time: 1.8181  data: 1.0351  max mem: 55573
Epoch: [134]  [ 20/156]  eta: 0:02:47  lr: 0.005145  min_lr: 0.005145  loss: 3.5826 (3.4008)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0005  max mem: 55573
Epoch: [134]  [ 30/156]  eta: 0:02:09  lr: 0.005142  min_lr: 0.005142  loss: 3.6236 (3.3472)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [134]  [ 40/156]  eta: 0:01:46  lr: 0.005140  min_lr: 0.005140  loss: 3.0853 (3.2870)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [134]  [ 50/156]  eta: 0:01:30  lr: 0.005137  min_lr: 0.005137  loss: 3.4757 (3.3443)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0006  max mem: 55573
Epoch: [134]  [ 60/156]  eta: 0:01:17  lr: 0.005134  min_lr: 0.005134  loss: 3.5799 (3.3433)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0005  max mem: 55573
Epoch: [134]  [ 70/156]  eta: 0:01:07  lr: 0.005131  min_lr: 0.005131  loss: 3.5457 (3.3421)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0005  max mem: 55573
Epoch: [134]  [ 80/156]  eta: 0:00:57  lr: 0.005129  min_lr: 0.005129  loss: 3.5343 (3.3397)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [134]  [ 90/156]  eta: 0:00:48  lr: 0.005126  min_lr: 0.005126  loss: 3.4586 (3.3507)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0006  max mem: 55573
Epoch: [134]  [100/156]  eta: 0:00:40  lr: 0.005123  min_lr: 0.005123  loss: 3.5675 (3.3705)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0006  max mem: 55573
Epoch: [134]  [110/156]  eta: 0:00:32  lr: 0.005120  min_lr: 0.005120  loss: 3.5165 (3.3624)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [134]  [120/156]  eta: 0:00:25  lr: 0.005118  min_lr: 0.005118  loss: 3.3668 (3.3447)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [134]  [130/156]  eta: 0:00:18  lr: 0.005115  min_lr: 0.005115  loss: 3.3668 (3.3438)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0010  max mem: 55573
Epoch: [134]  [140/156]  eta: 0:00:10  lr: 0.005112  min_lr: 0.005112  loss: 3.4799 (3.3564)  weight_decay: 0.0500 (0.0500)  time: 0.5847  data: 0.0009  max mem: 55573
Epoch: [134]  [150/156]  eta: 0:00:04  lr: 0.005109  min_lr: 0.005109  loss: 3.4568 (3.3564)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [134]  [155/156]  eta: 0:00:00  lr: 0.005108  min_lr: 0.005108  loss: 3.4839 (3.3572)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [134] Total time: 0:01:46 (0.6821 s / it)
Averaged stats: lr: 0.005108  min_lr: 0.005108  loss: 3.4839 (3.3995)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.1719 (1.1719)  acc1: 76.6927 (76.6927)  acc5: 93.4896 (93.4896)  time: 7.1589  data: 6.9214  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2603 (1.2462)  acc1: 74.2188 (72.3520)  acc5: 93.3594 (91.1680)  time: 1.5782  data: 1.3844  max mem: 55573
Test: Total time: 0:00:08 (1.6179 s / it)
* Acc@1 73.274 Acc@5 91.724 loss 1.233
Accuracy of the model on the 50000 test images: 73.3%
Max accuracy: 74.15%
Test:  [0/5]  eta: 0:00:33  loss: 2.5074 (2.5074)  acc1: 41.9271 (41.9271)  acc5: 69.9219 (69.9219)  time: 6.7953  data: 6.5593  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.5348 (2.6312)  acc1: 41.9271 (40.3840)  acc5: 69.9219 (66.7520)  time: 1.5044  data: 1.3120  max mem: 55573
Test: Total time: 0:00:07 (1.5233 s / it)
* Acc@1 38.446 Acc@5 64.536 loss 2.802
Accuracy of the model EMA on 50000 test images: 38.4%
Max EMA accuracy: 38.45%
Epoch: [135]  [  0/156]  eta: 0:31:23  lr: 0.005108  min_lr: 0.005108  loss: 3.6188 (3.6188)  weight_decay: 0.0500 (0.0500)  time: 12.0758  data: 8.8046  max mem: 55573
Epoch: [135]  [ 10/156]  eta: 0:04:19  lr: 0.005105  min_lr: 0.005105  loss: 3.3493 (3.3597)  weight_decay: 0.0500 (0.0500)  time: 1.7797  data: 0.9405  max mem: 55573
Epoch: [135]  [ 20/156]  eta: 0:02:44  lr: 0.005102  min_lr: 0.005102  loss: 3.3333 (3.2669)  weight_decay: 0.0500 (0.0500)  time: 0.6701  data: 0.0773  max mem: 55573
Epoch: [135]  [ 30/156]  eta: 0:02:07  lr: 0.005099  min_lr: 0.005099  loss: 3.4652 (3.3252)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [135]  [ 40/156]  eta: 0:01:45  lr: 0.005097  min_lr: 0.005097  loss: 3.5351 (3.3226)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0004  max mem: 55573
Epoch: [135]  [ 50/156]  eta: 0:01:29  lr: 0.005094  min_lr: 0.005094  loss: 3.3529 (3.3004)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [135]  [ 60/156]  eta: 0:01:17  lr: 0.005091  min_lr: 0.005091  loss: 3.2870 (3.3039)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [135]  [ 70/156]  eta: 0:01:06  lr: 0.005088  min_lr: 0.005088  loss: 3.3326 (3.3063)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [135]  [ 80/156]  eta: 0:00:57  lr: 0.005086  min_lr: 0.005086  loss: 3.3282 (3.2988)  weight_decay: 0.0500 (0.0500)  time: 0.5936  data: 0.0004  max mem: 55573
Epoch: [135]  [ 90/156]  eta: 0:00:48  lr: 0.005083  min_lr: 0.005083  loss: 3.6224 (3.3410)  weight_decay: 0.0500 (0.0500)  time: 0.5935  data: 0.0004  max mem: 55573
Epoch: [135]  [100/156]  eta: 0:00:40  lr: 0.005080  min_lr: 0.005080  loss: 3.6448 (3.3516)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [135]  [110/156]  eta: 0:00:32  lr: 0.005077  min_lr: 0.005077  loss: 3.4838 (3.3427)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [135]  [120/156]  eta: 0:00:25  lr: 0.005074  min_lr: 0.005074  loss: 3.4052 (3.3379)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [135]  [130/156]  eta: 0:00:17  lr: 0.005072  min_lr: 0.005072  loss: 3.5516 (3.3574)  weight_decay: 0.0500 (0.0500)  time: 0.5945  data: 0.0011  max mem: 55573
Epoch: [135]  [140/156]  eta: 0:00:10  lr: 0.005069  min_lr: 0.005069  loss: 3.7234 (3.3674)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0010  max mem: 55573
Epoch: [135]  [150/156]  eta: 0:00:04  lr: 0.005066  min_lr: 0.005066  loss: 3.7739 (3.3859)  weight_decay: 0.0500 (0.0500)  time: 0.5828  data: 0.0001  max mem: 55573
Epoch: [135]  [155/156]  eta: 0:00:00  lr: 0.005065  min_lr: 0.005065  loss: 3.6944 (3.3870)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [135] Total time: 0:01:46 (0.6812 s / it)
Averaged stats: lr: 0.005065  min_lr: 0.005065  loss: 3.6944 (3.4095)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.2982 (1.2982)  acc1: 76.1719 (76.1719)  acc5: 94.4010 (94.4010)  time: 6.9635  data: 6.7265  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3240 (1.3698)  acc1: 75.0000 (72.6400)  acc5: 94.4010 (91.7120)  time: 1.5391  data: 1.3454  max mem: 55573
Test: Total time: 0:00:07 (1.5790 s / it)
* Acc@1 73.194 Acc@5 91.836 loss 1.374
Accuracy of the model on the 50000 test images: 73.2%
Max accuracy: 74.15%
Test:  [0/5]  eta: 0:00:35  loss: 2.4267 (2.4267)  acc1: 43.3594 (43.3594)  acc5: 71.2240 (71.2240)  time: 7.1928  data: 6.9568  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.4502 (2.5400)  acc1: 43.3594 (41.6960)  acc5: 71.2240 (68.2880)  time: 1.5839  data: 1.3915  max mem: 55573
Test: Total time: 0:00:08 (1.6082 s / it)
* Acc@1 39.938 Acc@5 66.202 loss 2.704
Accuracy of the model EMA on 50000 test images: 39.9%
Max EMA accuracy: 39.94%
Epoch: [136]  [  0/156]  eta: 0:31:39  lr: 0.005065  min_lr: 0.005065  loss: 3.4973 (3.4973)  weight_decay: 0.0500 (0.0500)  time: 12.1765  data: 11.5748  max mem: 55573
Epoch: [136]  [ 10/156]  eta: 0:04:04  lr: 0.005062  min_lr: 0.005062  loss: 3.5683 (3.4011)  weight_decay: 0.0500 (0.0500)  time: 1.6737  data: 1.0527  max mem: 55573
Epoch: [136]  [ 20/156]  eta: 0:02:37  lr: 0.005059  min_lr: 0.005059  loss: 3.5849 (3.3942)  weight_decay: 0.0500 (0.0500)  time: 0.6071  data: 0.0005  max mem: 55573
Epoch: [136]  [ 30/156]  eta: 0:02:02  lr: 0.005056  min_lr: 0.005056  loss: 3.5849 (3.4379)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [136]  [ 40/156]  eta: 0:01:42  lr: 0.005053  min_lr: 0.005053  loss: 3.4783 (3.4211)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [136]  [ 50/156]  eta: 0:01:27  lr: 0.005051  min_lr: 0.005051  loss: 3.4796 (3.4128)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [136]  [ 60/156]  eta: 0:01:15  lr: 0.005048  min_lr: 0.005048  loss: 3.4796 (3.4210)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [136]  [ 70/156]  eta: 0:01:05  lr: 0.005045  min_lr: 0.005045  loss: 3.6315 (3.4411)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0005  max mem: 55573
Epoch: [136]  [ 80/156]  eta: 0:00:56  lr: 0.005042  min_lr: 0.005042  loss: 3.4469 (3.4101)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0005  max mem: 55573
Epoch: [136]  [ 90/156]  eta: 0:00:47  lr: 0.005040  min_lr: 0.005040  loss: 3.2628 (3.3814)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [136]  [100/156]  eta: 0:00:39  lr: 0.005037  min_lr: 0.005037  loss: 2.9583 (3.3330)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [136]  [110/156]  eta: 0:00:32  lr: 0.005034  min_lr: 0.005034  loss: 3.1271 (3.3510)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [136]  [120/156]  eta: 0:00:24  lr: 0.005031  min_lr: 0.005031  loss: 3.6190 (3.3618)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [136]  [130/156]  eta: 0:00:17  lr: 0.005028  min_lr: 0.005028  loss: 3.5370 (3.3727)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0010  max mem: 55573
Epoch: [136]  [140/156]  eta: 0:00:10  lr: 0.005026  min_lr: 0.005026  loss: 3.5842 (3.3856)  weight_decay: 0.0500 (0.0500)  time: 0.5855  data: 0.0009  max mem: 55573
Epoch: [136]  [150/156]  eta: 0:00:04  lr: 0.005023  min_lr: 0.005023  loss: 3.6588 (3.3993)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [136]  [155/156]  eta: 0:00:00  lr: 0.005021  min_lr: 0.005021  loss: 3.6696 (3.4008)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [136] Total time: 0:01:44 (0.6726 s / it)
Averaged stats: lr: 0.005021  min_lr: 0.005021  loss: 3.6696 (3.4095)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.1246 (1.1246)  acc1: 77.4740 (77.4740)  acc5: 94.4010 (94.4010)  time: 7.1868  data: 6.9494  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2261 (1.2617)  acc1: 75.2604 (72.5120)  acc5: 93.6198 (91.4560)  time: 1.5836  data: 1.3899  max mem: 55573
Test: Total time: 0:00:08 (1.6267 s / it)
* Acc@1 73.442 Acc@5 91.914 loss 1.228
Accuracy of the model on the 50000 test images: 73.4%
Max accuracy: 74.15%
Test:  [0/5]  eta: 0:00:35  loss: 2.3502 (2.3502)  acc1: 45.0521 (45.0521)  acc5: 72.2656 (72.2656)  time: 7.1513  data: 6.9152  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.3716 (2.4545)  acc1: 45.0521 (43.1680)  acc5: 72.2656 (69.5360)  time: 1.5756  data: 1.3832  max mem: 55573
Test: Total time: 0:00:07 (1.5953 s / it)
* Acc@1 41.284 Acc@5 67.584 loss 2.613
Accuracy of the model EMA on 50000 test images: 41.3%
Max EMA accuracy: 41.28%
Epoch: [137]  [  0/156]  eta: 0:36:55  lr: 0.005021  min_lr: 0.005021  loss: 4.0033 (4.0033)  weight_decay: 0.0500 (0.0500)  time: 14.1987  data: 11.7106  max mem: 55573
Epoch: [137]  [ 10/156]  eta: 0:04:26  lr: 0.005018  min_lr: 0.005018  loss: 3.0694 (3.2148)  weight_decay: 0.0500 (0.0500)  time: 1.8279  data: 1.0650  max mem: 55573
Epoch: [137]  [ 20/156]  eta: 0:02:48  lr: 0.005016  min_lr: 0.005016  loss: 3.2831 (3.3319)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [137]  [ 30/156]  eta: 0:02:09  lr: 0.005013  min_lr: 0.005013  loss: 3.5954 (3.3968)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [137]  [ 40/156]  eta: 0:01:47  lr: 0.005010  min_lr: 0.005010  loss: 3.1988 (3.3138)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [137]  [ 50/156]  eta: 0:01:30  lr: 0.005007  min_lr: 0.005007  loss: 3.2558 (3.3355)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0005  max mem: 55573
Epoch: [137]  [ 60/156]  eta: 0:01:18  lr: 0.005004  min_lr: 0.005004  loss: 3.4780 (3.3194)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [137]  [ 70/156]  eta: 0:01:07  lr: 0.005002  min_lr: 0.005002  loss: 3.2661 (3.3050)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [137]  [ 80/156]  eta: 0:00:57  lr: 0.004999  min_lr: 0.004999  loss: 3.3919 (3.3380)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [137]  [ 90/156]  eta: 0:00:48  lr: 0.004996  min_lr: 0.004996  loss: 3.3919 (3.3056)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [137]  [100/156]  eta: 0:00:40  lr: 0.004993  min_lr: 0.004993  loss: 3.5112 (3.3189)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0005  max mem: 55573
Epoch: [137]  [110/156]  eta: 0:00:32  lr: 0.004991  min_lr: 0.004991  loss: 3.6124 (3.3298)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [137]  [120/156]  eta: 0:00:25  lr: 0.004988  min_lr: 0.004988  loss: 3.4258 (3.3287)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [137]  [130/156]  eta: 0:00:18  lr: 0.004985  min_lr: 0.004985  loss: 3.2593 (3.3148)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0015  max mem: 55573
Epoch: [137]  [140/156]  eta: 0:00:10  lr: 0.004982  min_lr: 0.004982  loss: 3.2593 (3.3137)  weight_decay: 0.0500 (0.0500)  time: 0.5876  data: 0.0013  max mem: 55573
Epoch: [137]  [150/156]  eta: 0:00:04  lr: 0.004979  min_lr: 0.004979  loss: 3.5470 (3.3207)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [137]  [155/156]  eta: 0:00:00  lr: 0.004978  min_lr: 0.004978  loss: 3.5470 (3.3268)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [137] Total time: 0:01:46 (0.6837 s / it)
Averaged stats: lr: 0.004978  min_lr: 0.004978  loss: 3.5470 (3.4002)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.1940 (1.1940)  acc1: 77.2135 (77.2135)  acc5: 93.8802 (93.8802)  time: 6.8638  data: 6.6266  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3060 (1.2974)  acc1: 76.8229 (73.2480)  acc5: 92.9688 (91.6800)  time: 1.5181  data: 1.3254  max mem: 55573
Test: Total time: 0:00:07 (1.5354 s / it)
* Acc@1 73.470 Acc@5 91.918 loss 1.294
Accuracy of the model on the 50000 test images: 73.5%
Max accuracy: 74.15%
Test:  [0/5]  eta: 0:00:35  loss: 2.2785 (2.2785)  acc1: 46.4844 (46.4844)  acc5: 73.1771 (73.1771)  time: 7.1299  data: 6.8939  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.2986 (2.3749)  acc1: 46.4844 (44.2880)  acc5: 73.1771 (70.3040)  time: 1.5713  data: 1.3789  max mem: 55573
Test: Total time: 0:00:07 (1.5896 s / it)
* Acc@1 42.702 Acc@5 68.922 loss 2.527
Accuracy of the model EMA on 50000 test images: 42.7%
Max EMA accuracy: 42.70%
Epoch: [138]  [  0/156]  eta: 0:31:26  lr: 0.004978  min_lr: 0.004978  loss: 2.6817 (2.6817)  weight_decay: 0.0500 (0.0500)  time: 12.0950  data: 10.3102  max mem: 55573
Epoch: [138]  [ 10/156]  eta: 0:04:22  lr: 0.004975  min_lr: 0.004975  loss: 2.7884 (3.0561)  weight_decay: 0.0500 (0.0500)  time: 1.7962  data: 0.9376  max mem: 55573
Epoch: [138]  [ 20/156]  eta: 0:02:46  lr: 0.004972  min_lr: 0.004972  loss: 2.8315 (3.1297)  weight_decay: 0.0500 (0.0500)  time: 0.6780  data: 0.0004  max mem: 55573
Epoch: [138]  [ 30/156]  eta: 0:02:08  lr: 0.004969  min_lr: 0.004969  loss: 3.1125 (3.1227)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [138]  [ 40/156]  eta: 0:01:46  lr: 0.004967  min_lr: 0.004967  loss: 3.4907 (3.2365)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [138]  [ 50/156]  eta: 0:01:30  lr: 0.004964  min_lr: 0.004964  loss: 3.5929 (3.2792)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [138]  [ 60/156]  eta: 0:01:17  lr: 0.004961  min_lr: 0.004961  loss: 3.5955 (3.3017)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [138]  [ 70/156]  eta: 0:01:06  lr: 0.004958  min_lr: 0.004958  loss: 3.3375 (3.2543)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [138]  [ 80/156]  eta: 0:00:57  lr: 0.004955  min_lr: 0.004955  loss: 2.8134 (3.2500)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [138]  [ 90/156]  eta: 0:00:48  lr: 0.004953  min_lr: 0.004953  loss: 3.0597 (3.2409)  weight_decay: 0.0500 (0.0500)  time: 0.5947  data: 0.0004  max mem: 55573
Epoch: [138]  [100/156]  eta: 0:00:40  lr: 0.004950  min_lr: 0.004950  loss: 3.4624 (3.2683)  weight_decay: 0.0500 (0.0500)  time: 0.5964  data: 0.0004  max mem: 55573
Epoch: [138]  [110/156]  eta: 0:00:32  lr: 0.004947  min_lr: 0.004947  loss: 3.6209 (3.2904)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0004  max mem: 55573
Epoch: [138]  [120/156]  eta: 0:00:25  lr: 0.004944  min_lr: 0.004944  loss: 3.6772 (3.3095)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [138]  [130/156]  eta: 0:00:18  lr: 0.004941  min_lr: 0.004941  loss: 3.3837 (3.3058)  weight_decay: 0.0500 (0.0500)  time: 0.5883  data: 0.0010  max mem: 55573
Epoch: [138]  [140/156]  eta: 0:00:10  lr: 0.004939  min_lr: 0.004939  loss: 3.2634 (3.3085)  weight_decay: 0.0500 (0.0500)  time: 0.5849  data: 0.0009  max mem: 55573
Epoch: [138]  [150/156]  eta: 0:00:04  lr: 0.004936  min_lr: 0.004936  loss: 3.3603 (3.3214)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [138]  [155/156]  eta: 0:00:00  lr: 0.004934  min_lr: 0.004934  loss: 3.3680 (3.3220)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [138] Total time: 0:01:46 (0.6816 s / it)
Averaged stats: lr: 0.004934  min_lr: 0.004934  loss: 3.3680 (3.4030)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.2375 (1.2375)  acc1: 78.3854 (78.3854)  acc5: 94.7917 (94.7917)  time: 7.2189  data: 6.9816  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3480 (1.3673)  acc1: 75.6510 (73.3120)  acc5: 93.6198 (91.3600)  time: 1.5902  data: 1.3964  max mem: 55573
Test: Total time: 0:00:08 (1.6352 s / it)
* Acc@1 74.010 Acc@5 92.184 loss 1.335
Accuracy of the model on the 50000 test images: 74.0%
Max accuracy: 74.15%
Test:  [0/5]  eta: 0:00:35  loss: 2.2097 (2.2097)  acc1: 47.9167 (47.9167)  acc5: 74.8698 (74.8698)  time: 7.1341  data: 6.8980  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.2296 (2.3001)  acc1: 47.9167 (45.5040)  acc5: 74.8698 (71.4880)  time: 1.5722  data: 1.3797  max mem: 55573
Test: Total time: 0:00:07 (1.5915 s / it)
* Acc@1 44.010 Acc@5 70.194 loss 2.447
Accuracy of the model EMA on 50000 test images: 44.0%
Max EMA accuracy: 44.01%
Epoch: [139]  [  0/156]  eta: 0:34:49  lr: 0.004934  min_lr: 0.004934  loss: 2.7249 (2.7249)  weight_decay: 0.0500 (0.0500)  time: 13.3946  data: 10.2349  max mem: 55573
Epoch: [139]  [ 10/156]  eta: 0:04:17  lr: 0.004931  min_lr: 0.004931  loss: 3.6492 (3.5529)  weight_decay: 0.0500 (0.0500)  time: 1.7644  data: 0.9309  max mem: 55573
Epoch: [139]  [ 20/156]  eta: 0:02:43  lr: 0.004929  min_lr: 0.004929  loss: 3.5365 (3.4367)  weight_decay: 0.0500 (0.0500)  time: 0.5958  data: 0.0005  max mem: 55573
Epoch: [139]  [ 30/156]  eta: 0:02:06  lr: 0.004926  min_lr: 0.004926  loss: 3.5250 (3.4740)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [139]  [ 40/156]  eta: 0:01:45  lr: 0.004923  min_lr: 0.004923  loss: 3.5825 (3.4782)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [139]  [ 50/156]  eta: 0:01:29  lr: 0.004920  min_lr: 0.004920  loss: 3.5760 (3.4630)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [139]  [ 60/156]  eta: 0:01:17  lr: 0.004917  min_lr: 0.004917  loss: 3.4170 (3.4257)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [139]  [ 70/156]  eta: 0:01:06  lr: 0.004915  min_lr: 0.004915  loss: 3.4170 (3.4146)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0005  max mem: 55573
Epoch: [139]  [ 80/156]  eta: 0:00:57  lr: 0.004912  min_lr: 0.004912  loss: 3.6928 (3.4664)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [139]  [ 90/156]  eta: 0:00:48  lr: 0.004909  min_lr: 0.004909  loss: 3.7743 (3.4506)  weight_decay: 0.0500 (0.0500)  time: 0.5941  data: 0.0004  max mem: 55573
Epoch: [139]  [100/156]  eta: 0:00:40  lr: 0.004906  min_lr: 0.004906  loss: 3.2561 (3.4343)  weight_decay: 0.0500 (0.0500)  time: 0.5942  data: 0.0004  max mem: 55573
Epoch: [139]  [110/156]  eta: 0:00:32  lr: 0.004903  min_lr: 0.004903  loss: 3.2561 (3.4160)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [139]  [120/156]  eta: 0:00:25  lr: 0.004901  min_lr: 0.004901  loss: 3.5903 (3.4206)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [139]  [130/156]  eta: 0:00:17  lr: 0.004898  min_lr: 0.004898  loss: 3.5903 (3.4329)  weight_decay: 0.0500 (0.0500)  time: 0.5885  data: 0.0011  max mem: 55573
Epoch: [139]  [140/156]  eta: 0:00:10  lr: 0.004895  min_lr: 0.004895  loss: 3.6313 (3.4397)  weight_decay: 0.0500 (0.0500)  time: 0.5846  data: 0.0009  max mem: 55573
Epoch: [139]  [150/156]  eta: 0:00:04  lr: 0.004892  min_lr: 0.004892  loss: 3.6470 (3.4494)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [139]  [155/156]  eta: 0:00:00  lr: 0.004891  min_lr: 0.004891  loss: 3.4035 (3.4377)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [139] Total time: 0:01:45 (0.6791 s / it)
Averaged stats: lr: 0.004891  min_lr: 0.004891  loss: 3.4035 (3.4025)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.0642 (1.0642)  acc1: 78.5156 (78.5156)  acc5: 94.6615 (94.6615)  time: 6.6488  data: 6.4117  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1766 (1.1905)  acc1: 75.1302 (73.1840)  acc5: 94.3396 (92.6400)  time: 1.5523  data: 1.3596  max mem: 55573
Test: Total time: 0:00:07 (1.5842 s / it)
* Acc@1 74.212 Acc@5 92.294 loss 1.179
Accuracy of the model on the 50000 test images: 74.2%
Max accuracy: 74.21%
Test:  [0/5]  eta: 0:00:32  loss: 2.1467 (2.1467)  acc1: 49.7396 (49.7396)  acc5: 75.5208 (75.5208)  time: 6.5638  data: 6.3279  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.1666 (2.2322)  acc1: 49.7396 (46.7520)  acc5: 75.5208 (72.5120)  time: 1.4581  data: 1.2657  max mem: 55573
Test: Total time: 0:00:07 (1.4796 s / it)
* Acc@1 45.230 Acc@5 71.368 loss 2.374
Accuracy of the model EMA on 50000 test images: 45.2%
Max EMA accuracy: 45.23%
Epoch: [140]  [  0/156]  eta: 0:37:25  lr: 0.004890  min_lr: 0.004890  loss: 3.8380 (3.8380)  weight_decay: 0.0500 (0.0500)  time: 14.3950  data: 11.9485  max mem: 55573
Epoch: [140]  [ 10/156]  eta: 0:04:30  lr: 0.004888  min_lr: 0.004888  loss: 3.6261 (3.3332)  weight_decay: 0.0500 (0.0500)  time: 1.8552  data: 1.0866  max mem: 55573
Epoch: [140]  [ 20/156]  eta: 0:02:50  lr: 0.004885  min_lr: 0.004885  loss: 3.0788 (3.2810)  weight_decay: 0.0500 (0.0500)  time: 0.5961  data: 0.0004  max mem: 55573
Epoch: [140]  [ 30/156]  eta: 0:02:10  lr: 0.004882  min_lr: 0.004882  loss: 3.4116 (3.3873)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [140]  [ 40/156]  eta: 0:01:47  lr: 0.004879  min_lr: 0.004879  loss: 3.5902 (3.4061)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [140]  [ 50/156]  eta: 0:01:31  lr: 0.004876  min_lr: 0.004876  loss: 3.5576 (3.3859)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [140]  [ 60/156]  eta: 0:01:18  lr: 0.004874  min_lr: 0.004874  loss: 3.5576 (3.4227)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [140]  [ 70/156]  eta: 0:01:07  lr: 0.004871  min_lr: 0.004871  loss: 3.4371 (3.3785)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [140]  [ 80/156]  eta: 0:00:57  lr: 0.004868  min_lr: 0.004868  loss: 3.4137 (3.4069)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [140]  [ 90/156]  eta: 0:00:49  lr: 0.004865  min_lr: 0.004865  loss: 3.6583 (3.4184)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [140]  [100/156]  eta: 0:00:40  lr: 0.004862  min_lr: 0.004862  loss: 3.4837 (3.4150)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [140]  [110/156]  eta: 0:00:32  lr: 0.004860  min_lr: 0.004860  loss: 3.4551 (3.4041)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [140]  [120/156]  eta: 0:00:25  lr: 0.004857  min_lr: 0.004857  loss: 3.4174 (3.4003)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [140]  [130/156]  eta: 0:00:18  lr: 0.004854  min_lr: 0.004854  loss: 3.2138 (3.3881)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0010  max mem: 55573
Epoch: [140]  [140/156]  eta: 0:00:11  lr: 0.004851  min_lr: 0.004851  loss: 3.3292 (3.4022)  weight_decay: 0.0500 (0.0500)  time: 0.5860  data: 0.0008  max mem: 55573
Epoch: [140]  [150/156]  eta: 0:00:04  lr: 0.004848  min_lr: 0.004848  loss: 3.6011 (3.4090)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [140]  [155/156]  eta: 0:00:00  lr: 0.004847  min_lr: 0.004847  loss: 3.5496 (3.4116)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [140] Total time: 0:01:46 (0.6858 s / it)
Averaged stats: lr: 0.004847  min_lr: 0.004847  loss: 3.5496 (3.4027)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.2517 (1.2517)  acc1: 78.1250 (78.1250)  acc5: 93.8802 (93.8802)  time: 7.3237  data: 7.0863  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3050 (1.3590)  acc1: 75.0000 (72.2880)  acc5: 93.0990 (91.4240)  time: 1.6109  data: 1.4173  max mem: 55573
Test: Total time: 0:00:08 (1.6586 s / it)
* Acc@1 73.372 Acc@5 91.932 loss 1.344
Accuracy of the model on the 50000 test images: 73.4%
Max accuracy: 74.21%
Test:  [0/5]  eta: 0:00:36  loss: 2.0845 (2.0845)  acc1: 50.3906 (50.3906)  acc5: 76.5625 (76.5625)  time: 7.2698  data: 7.0337  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.1068 (2.1673)  acc1: 50.3906 (47.7120)  acc5: 76.5625 (73.6640)  time: 1.6017  data: 1.4093  max mem: 55573
Test: Total time: 0:00:08 (1.6193 s / it)
* Acc@1 46.350 Acc@5 72.528 loss 2.303
Accuracy of the model EMA on 50000 test images: 46.4%
Max EMA accuracy: 46.35%
Epoch: [141]  [  0/156]  eta: 0:35:35  lr: 0.004847  min_lr: 0.004847  loss: 3.6784 (3.6784)  weight_decay: 0.0500 (0.0500)  time: 13.6883  data: 9.3604  max mem: 55573
Epoch: [141]  [ 10/156]  eta: 0:04:21  lr: 0.004844  min_lr: 0.004844  loss: 3.6784 (3.6256)  weight_decay: 0.0500 (0.0500)  time: 1.7887  data: 0.8514  max mem: 55573
Epoch: [141]  [ 20/156]  eta: 0:02:45  lr: 0.004841  min_lr: 0.004841  loss: 3.6348 (3.5139)  weight_decay: 0.0500 (0.0500)  time: 0.5946  data: 0.0004  max mem: 55573
Epoch: [141]  [ 30/156]  eta: 0:02:07  lr: 0.004838  min_lr: 0.004838  loss: 3.2135 (3.3878)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [141]  [ 40/156]  eta: 0:01:45  lr: 0.004835  min_lr: 0.004835  loss: 3.1596 (3.3128)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [141]  [ 50/156]  eta: 0:01:29  lr: 0.004833  min_lr: 0.004833  loss: 3.4294 (3.3385)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [141]  [ 60/156]  eta: 0:01:17  lr: 0.004830  min_lr: 0.004830  loss: 3.5694 (3.3555)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [141]  [ 70/156]  eta: 0:01:06  lr: 0.004827  min_lr: 0.004827  loss: 3.5583 (3.3806)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [141]  [ 80/156]  eta: 0:00:57  lr: 0.004824  min_lr: 0.004824  loss: 3.5584 (3.3715)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [141]  [ 90/156]  eta: 0:00:48  lr: 0.004821  min_lr: 0.004821  loss: 3.5584 (3.3693)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0005  max mem: 55573
Epoch: [141]  [100/156]  eta: 0:00:40  lr: 0.004819  min_lr: 0.004819  loss: 3.2033 (3.3629)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [141]  [110/156]  eta: 0:00:32  lr: 0.004816  min_lr: 0.004816  loss: 3.3351 (3.3753)  weight_decay: 0.0500 (0.0500)  time: 0.5954  data: 0.0005  max mem: 55573
Epoch: [141]  [120/156]  eta: 0:00:25  lr: 0.004813  min_lr: 0.004813  loss: 3.3351 (3.3488)  weight_decay: 0.0500 (0.0500)  time: 0.5959  data: 0.0005  max mem: 55573
Epoch: [141]  [130/156]  eta: 0:00:17  lr: 0.004810  min_lr: 0.004810  loss: 3.1434 (3.3490)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0013  max mem: 55573
Epoch: [141]  [140/156]  eta: 0:00:10  lr: 0.004807  min_lr: 0.004807  loss: 3.3979 (3.3522)  weight_decay: 0.0500 (0.0500)  time: 0.5844  data: 0.0012  max mem: 55573
Epoch: [141]  [150/156]  eta: 0:00:04  lr: 0.004804  min_lr: 0.004804  loss: 3.3842 (3.3442)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [141]  [155/156]  eta: 0:00:00  lr: 0.004803  min_lr: 0.004803  loss: 3.4783 (3.3567)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [141] Total time: 0:01:46 (0.6813 s / it)
Averaged stats: lr: 0.004803  min_lr: 0.004803  loss: 3.4783 (3.3904)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.3531 (1.3531)  acc1: 75.7812 (75.7812)  acc5: 93.3594 (93.3594)  time: 7.0670  data: 6.8296  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4292 (1.4805)  acc1: 74.7396 (71.9040)  acc5: 93.3594 (91.1040)  time: 1.5598  data: 1.3660  max mem: 55573
Test: Total time: 0:00:08 (1.6097 s / it)
* Acc@1 73.500 Acc@5 91.736 loss 1.472
Accuracy of the model on the 50000 test images: 73.5%
Max accuracy: 74.21%
Test:  [0/5]  eta: 0:00:33  loss: 2.0274 (2.0274)  acc1: 51.8229 (51.8229)  acc5: 77.8646 (77.8646)  time: 6.6867  data: 6.4505  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 2.0526 (2.1088)  acc1: 51.8229 (48.7040)  acc5: 77.7344 (74.7520)  time: 1.5394  data: 1.3470  max mem: 55573
Test: Total time: 0:00:07 (1.5555 s / it)
* Acc@1 47.380 Acc@5 73.544 loss 2.240
Accuracy of the model EMA on 50000 test images: 47.4%
Max EMA accuracy: 47.38%
Epoch: [142]  [  0/156]  eta: 0:33:41  lr: 0.004803  min_lr: 0.004803  loss: 2.6037 (2.6037)  weight_decay: 0.0500 (0.0500)  time: 12.9561  data: 8.9688  max mem: 55573
Epoch: [142]  [ 10/156]  eta: 0:04:11  lr: 0.004800  min_lr: 0.004800  loss: 3.1093 (3.1875)  weight_decay: 0.0500 (0.0500)  time: 1.7235  data: 0.8158  max mem: 55573
Epoch: [142]  [ 20/156]  eta: 0:02:40  lr: 0.004797  min_lr: 0.004797  loss: 3.5464 (3.3204)  weight_decay: 0.0500 (0.0500)  time: 0.5951  data: 0.0005  max mem: 55573
Epoch: [142]  [ 30/156]  eta: 0:02:05  lr: 0.004794  min_lr: 0.004794  loss: 3.7239 (3.3860)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [142]  [ 40/156]  eta: 0:01:43  lr: 0.004791  min_lr: 0.004791  loss: 3.6908 (3.3572)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [142]  [ 50/156]  eta: 0:01:28  lr: 0.004789  min_lr: 0.004789  loss: 3.5567 (3.3608)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0005  max mem: 55573
Epoch: [142]  [ 60/156]  eta: 0:01:16  lr: 0.004786  min_lr: 0.004786  loss: 3.5567 (3.3503)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [142]  [ 70/156]  eta: 0:01:05  lr: 0.004783  min_lr: 0.004783  loss: 3.5339 (3.3685)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [142]  [ 80/156]  eta: 0:00:56  lr: 0.004780  min_lr: 0.004780  loss: 3.5168 (3.3640)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [142]  [ 90/156]  eta: 0:00:48  lr: 0.004777  min_lr: 0.004777  loss: 3.4120 (3.3612)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [142]  [100/156]  eta: 0:00:39  lr: 0.004775  min_lr: 0.004775  loss: 3.5480 (3.3827)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [142]  [110/156]  eta: 0:00:32  lr: 0.004772  min_lr: 0.004772  loss: 3.6857 (3.4110)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [142]  [120/156]  eta: 0:00:24  lr: 0.004769  min_lr: 0.004769  loss: 3.7082 (3.4343)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [142]  [130/156]  eta: 0:00:17  lr: 0.004766  min_lr: 0.004766  loss: 3.6407 (3.4509)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0011  max mem: 55573
Epoch: [142]  [140/156]  eta: 0:00:10  lr: 0.004763  min_lr: 0.004763  loss: 3.4905 (3.4302)  weight_decay: 0.0500 (0.0500)  time: 0.5860  data: 0.0009  max mem: 55573
Epoch: [142]  [150/156]  eta: 0:00:04  lr: 0.004760  min_lr: 0.004760  loss: 3.1779 (3.4160)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [142]  [155/156]  eta: 0:00:00  lr: 0.004759  min_lr: 0.004759  loss: 3.3745 (3.4144)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [142] Total time: 0:01:45 (0.6762 s / it)
Averaged stats: lr: 0.004759  min_lr: 0.004759  loss: 3.3745 (3.3976)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.1411 (1.1411)  acc1: 75.7812 (75.7812)  acc5: 94.2708 (94.2708)  time: 7.0892  data: 6.8519  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2229 (1.2798)  acc1: 75.7812 (72.8960)  acc5: 92.4528 (92.3200)  time: 1.5642  data: 1.3704  max mem: 55573
Test: Total time: 0:00:08 (1.6046 s / it)
* Acc@1 74.026 Acc@5 92.164 loss 1.266
Accuracy of the model on the 50000 test images: 74.0%
Max accuracy: 74.21%
Test:  [0/5]  eta: 0:00:34  loss: 1.9720 (1.9720)  acc1: 52.4740 (52.4740)  acc5: 78.9062 (78.9062)  time: 6.8591  data: 6.6230  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.9994 (2.0517)  acc1: 52.4740 (49.6000)  acc5: 78.7760 (75.7760)  time: 1.5171  data: 1.3247  max mem: 55573
Test: Total time: 0:00:07 (1.5478 s / it)
* Acc@1 48.452 Acc@5 74.536 loss 2.177
Accuracy of the model EMA on 50000 test images: 48.5%
Max EMA accuracy: 48.45%
Epoch: [143]  [  0/156]  eta: 0:36:34  lr: 0.004759  min_lr: 0.004759  loss: 3.2024 (3.2024)  weight_decay: 0.0500 (0.0500)  time: 14.0676  data: 8.1886  max mem: 55573
Epoch: [143]  [ 10/156]  eta: 0:04:24  lr: 0.004756  min_lr: 0.004756  loss: 3.6043 (3.5381)  weight_decay: 0.0500 (0.0500)  time: 1.8133  data: 0.7448  max mem: 55573
Epoch: [143]  [ 20/156]  eta: 0:02:47  lr: 0.004753  min_lr: 0.004753  loss: 3.7328 (3.5291)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0004  max mem: 55573
Epoch: [143]  [ 30/156]  eta: 0:02:09  lr: 0.004750  min_lr: 0.004750  loss: 3.7354 (3.5783)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [143]  [ 40/156]  eta: 0:01:46  lr: 0.004747  min_lr: 0.004747  loss: 3.6861 (3.5492)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [143]  [ 50/156]  eta: 0:01:30  lr: 0.004745  min_lr: 0.004745  loss: 3.4360 (3.5183)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [143]  [ 60/156]  eta: 0:01:17  lr: 0.004742  min_lr: 0.004742  loss: 3.4671 (3.5010)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [143]  [ 70/156]  eta: 0:01:07  lr: 0.004739  min_lr: 0.004739  loss: 3.5806 (3.4913)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [143]  [ 80/156]  eta: 0:00:57  lr: 0.004736  min_lr: 0.004736  loss: 3.5612 (3.4980)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0005  max mem: 55573
Epoch: [143]  [ 90/156]  eta: 0:00:48  lr: 0.004733  min_lr: 0.004733  loss: 3.5374 (3.4973)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0005  max mem: 55573
Epoch: [143]  [100/156]  eta: 0:00:40  lr: 0.004730  min_lr: 0.004730  loss: 3.4958 (3.4768)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [143]  [110/156]  eta: 0:00:32  lr: 0.004728  min_lr: 0.004728  loss: 3.6421 (3.4892)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0004  max mem: 55573
Epoch: [143]  [120/156]  eta: 0:00:25  lr: 0.004725  min_lr: 0.004725  loss: 3.5379 (3.4694)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0004  max mem: 55573
Epoch: [143]  [130/156]  eta: 0:00:18  lr: 0.004722  min_lr: 0.004722  loss: 3.5379 (3.4769)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0010  max mem: 55573
Epoch: [143]  [140/156]  eta: 0:00:10  lr: 0.004719  min_lr: 0.004719  loss: 3.6453 (3.4826)  weight_decay: 0.0500 (0.0500)  time: 0.5860  data: 0.0008  max mem: 55573
Epoch: [143]  [150/156]  eta: 0:00:04  lr: 0.004716  min_lr: 0.004716  loss: 3.6405 (3.4795)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [143]  [155/156]  eta: 0:00:00  lr: 0.004715  min_lr: 0.004715  loss: 3.6244 (3.4907)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [143] Total time: 0:01:46 (0.6824 s / it)
Averaged stats: lr: 0.004715  min_lr: 0.004715  loss: 3.6244 (3.3947)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.3399 (1.3399)  acc1: 78.5156 (78.5156)  acc5: 94.1406 (94.1406)  time: 7.1510  data: 6.9135  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3595 (1.4063)  acc1: 75.5208 (73.8240)  acc5: 94.1406 (91.9040)  time: 1.5765  data: 1.3828  max mem: 55573
Test: Total time: 0:00:08 (1.6278 s / it)
* Acc@1 74.204 Acc@5 92.306 loss 1.394
Accuracy of the model on the 50000 test images: 74.2%
Max accuracy: 74.21%
Test:  [0/5]  eta: 0:00:35  loss: 1.9209 (1.9209)  acc1: 52.9948 (52.9948)  acc5: 79.6875 (79.6875)  time: 7.1736  data: 6.9376  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.9506 (1.9996)  acc1: 52.9948 (50.3360)  acc5: 79.4271 (76.6720)  time: 1.5800  data: 1.3876  max mem: 55573
Test: Total time: 0:00:07 (1.5965 s / it)
* Acc@1 49.468 Acc@5 75.476 loss 2.120
Accuracy of the model EMA on 50000 test images: 49.5%
Max EMA accuracy: 49.47%
Epoch: [144]  [  0/156]  eta: 0:37:24  lr: 0.004715  min_lr: 0.004715  loss: 2.7820 (2.7820)  weight_decay: 0.0500 (0.0500)  time: 14.3891  data: 13.4251  max mem: 55573
Epoch: [144]  [ 10/156]  eta: 0:04:28  lr: 0.004712  min_lr: 0.004712  loss: 3.4056 (3.2741)  weight_decay: 0.0500 (0.0500)  time: 1.8421  data: 1.2210  max mem: 55573
Epoch: [144]  [ 20/156]  eta: 0:02:49  lr: 0.004709  min_lr: 0.004709  loss: 3.6079 (3.3675)  weight_decay: 0.0500 (0.0500)  time: 0.5880  data: 0.0005  max mem: 55573
Epoch: [144]  [ 30/156]  eta: 0:02:10  lr: 0.004706  min_lr: 0.004706  loss: 3.5363 (3.3383)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [144]  [ 40/156]  eta: 0:01:47  lr: 0.004703  min_lr: 0.004703  loss: 3.3991 (3.3654)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [144]  [ 50/156]  eta: 0:01:31  lr: 0.004700  min_lr: 0.004700  loss: 3.6326 (3.4370)  weight_decay: 0.0500 (0.0500)  time: 0.5887  data: 0.0004  max mem: 55573
Epoch: [144]  [ 60/156]  eta: 0:01:18  lr: 0.004698  min_lr: 0.004698  loss: 3.5794 (3.4339)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [144]  [ 70/156]  eta: 0:01:07  lr: 0.004695  min_lr: 0.004695  loss: 3.3822 (3.4190)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [144]  [ 80/156]  eta: 0:00:57  lr: 0.004692  min_lr: 0.004692  loss: 3.2864 (3.3918)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [144]  [ 90/156]  eta: 0:00:48  lr: 0.004689  min_lr: 0.004689  loss: 3.4274 (3.3920)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [144]  [100/156]  eta: 0:00:40  lr: 0.004686  min_lr: 0.004686  loss: 3.4248 (3.3778)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [144]  [110/156]  eta: 0:00:32  lr: 0.004683  min_lr: 0.004683  loss: 3.4248 (3.3932)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [144]  [120/156]  eta: 0:00:25  lr: 0.004681  min_lr: 0.004681  loss: 3.6008 (3.3881)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [144]  [130/156]  eta: 0:00:18  lr: 0.004678  min_lr: 0.004678  loss: 2.9199 (3.3575)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0011  max mem: 55573
Epoch: [144]  [140/156]  eta: 0:00:10  lr: 0.004675  min_lr: 0.004675  loss: 3.2919 (3.3727)  weight_decay: 0.0500 (0.0500)  time: 0.5864  data: 0.0010  max mem: 55573
Epoch: [144]  [150/156]  eta: 0:00:04  lr: 0.004672  min_lr: 0.004672  loss: 3.5902 (3.3762)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0002  max mem: 55573
Epoch: [144]  [155/156]  eta: 0:00:00  lr: 0.004671  min_lr: 0.004671  loss: 3.4961 (3.3722)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [144] Total time: 0:01:46 (0.6840 s / it)
Averaged stats: lr: 0.004671  min_lr: 0.004671  loss: 3.4961 (3.3768)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.0768 (1.0768)  acc1: 76.9531 (76.9531)  acc5: 95.3125 (95.3125)  time: 7.2163  data: 6.9790  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2772 (1.2713)  acc1: 75.3906 (72.5760)  acc5: 93.2292 (91.8720)  time: 1.5898  data: 1.3959  max mem: 55573
Test: Total time: 0:00:08 (1.6343 s / it)
* Acc@1 73.816 Acc@5 92.278 loss 1.243
Accuracy of the model on the 50000 test images: 73.8%
Max accuracy: 74.21%
Test:  [0/5]  eta: 0:00:33  loss: 1.8751 (1.8751)  acc1: 54.8177 (54.8177)  acc5: 79.9479 (79.9479)  time: 6.7376  data: 6.5016  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.9063 (1.9516)  acc1: 54.8177 (51.5840)  acc5: 79.9479 (77.4080)  time: 1.4929  data: 1.3005  max mem: 55573
Test: Total time: 0:00:07 (1.5129 s / it)
* Acc@1 50.434 Acc@5 76.284 loss 2.067
Accuracy of the model EMA on 50000 test images: 50.4%
Max EMA accuracy: 50.43%
Epoch: [145]  [  0/156]  eta: 0:39:42  lr: 0.004670  min_lr: 0.004670  loss: 3.6732 (3.6732)  weight_decay: 0.0500 (0.0500)  time: 15.2752  data: 7.2197  max mem: 55573
Epoch: [145]  [ 10/156]  eta: 0:04:41  lr: 0.004668  min_lr: 0.004668  loss: 3.6389 (3.5724)  weight_decay: 0.0500 (0.0500)  time: 1.9248  data: 0.6567  max mem: 55573
Epoch: [145]  [ 20/156]  eta: 0:02:55  lr: 0.004665  min_lr: 0.004665  loss: 3.5872 (3.5874)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [145]  [ 30/156]  eta: 0:02:14  lr: 0.004662  min_lr: 0.004662  loss: 3.4787 (3.4898)  weight_decay: 0.0500 (0.0500)  time: 0.5940  data: 0.0004  max mem: 55573
Epoch: [145]  [ 40/156]  eta: 0:01:50  lr: 0.004659  min_lr: 0.004659  loss: 3.6392 (3.5049)  weight_decay: 0.0500 (0.0500)  time: 0.6005  data: 0.0004  max mem: 55573
Epoch: [145]  [ 50/156]  eta: 0:01:33  lr: 0.004656  min_lr: 0.004656  loss: 3.6212 (3.4503)  weight_decay: 0.0500 (0.0500)  time: 0.5982  data: 0.0005  max mem: 55573
Epoch: [145]  [ 60/156]  eta: 0:01:20  lr: 0.004653  min_lr: 0.004653  loss: 3.2523 (3.4177)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [145]  [ 70/156]  eta: 0:01:08  lr: 0.004651  min_lr: 0.004651  loss: 2.9220 (3.3302)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [145]  [ 80/156]  eta: 0:00:58  lr: 0.004648  min_lr: 0.004648  loss: 3.0568 (3.3362)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0005  max mem: 55573
Epoch: [145]  [ 90/156]  eta: 0:00:49  lr: 0.004645  min_lr: 0.004645  loss: 3.3617 (3.3288)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [145]  [100/156]  eta: 0:00:41  lr: 0.004642  min_lr: 0.004642  loss: 3.5230 (3.3384)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [145]  [110/156]  eta: 0:00:33  lr: 0.004639  min_lr: 0.004639  loss: 3.5319 (3.3385)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [145]  [120/156]  eta: 0:00:25  lr: 0.004636  min_lr: 0.004636  loss: 3.4425 (3.3376)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [145]  [130/156]  eta: 0:00:18  lr: 0.004634  min_lr: 0.004634  loss: 3.5622 (3.3471)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0010  max mem: 55573
Epoch: [145]  [140/156]  eta: 0:00:11  lr: 0.004631  min_lr: 0.004631  loss: 3.5622 (3.3555)  weight_decay: 0.0500 (0.0500)  time: 0.5857  data: 0.0009  max mem: 55573
Epoch: [145]  [150/156]  eta: 0:00:04  lr: 0.004628  min_lr: 0.004628  loss: 3.4447 (3.3547)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [145]  [155/156]  eta: 0:00:00  lr: 0.004626  min_lr: 0.004626  loss: 3.3853 (3.3459)  weight_decay: 0.0500 (0.0500)  time: 0.5812  data: 0.0001  max mem: 55573
Epoch: [145] Total time: 0:01:47 (0.6920 s / it)
Averaged stats: lr: 0.004626  min_lr: 0.004626  loss: 3.3853 (3.3668)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.1238 (1.1238)  acc1: 78.5156 (78.5156)  acc5: 94.9219 (94.9219)  time: 6.8535  data: 6.6163  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2730 (1.2905)  acc1: 75.9115 (73.2480)  acc5: 94.2708 (91.8720)  time: 1.5172  data: 1.3233  max mem: 55573
Test: Total time: 0:00:07 (1.5635 s / it)
* Acc@1 73.570 Acc@5 92.000 loss 1.295
Accuracy of the model on the 50000 test images: 73.6%
Max accuracy: 74.21%
Test:  [0/5]  eta: 0:00:35  loss: 1.8319 (1.8319)  acc1: 56.9010 (56.9010)  acc5: 80.5990 (80.5990)  time: 7.1767  data: 6.9408  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.8650 (1.9071)  acc1: 55.9896 (52.6080)  acc5: 80.5990 (78.0480)  time: 1.5806  data: 1.3882  max mem: 55573
Test: Total time: 0:00:07 (1.5993 s / it)
* Acc@1 51.378 Acc@5 77.024 loss 2.017
Accuracy of the model EMA on 50000 test images: 51.4%
Max EMA accuracy: 51.38%
Epoch: [146]  [  0/156]  eta: 0:34:18  lr: 0.004626  min_lr: 0.004626  loss: 4.0272 (4.0272)  weight_decay: 0.0500 (0.0500)  time: 13.1982  data: 9.6638  max mem: 55573
Epoch: [146]  [ 10/156]  eta: 0:04:15  lr: 0.004623  min_lr: 0.004623  loss: 3.5909 (3.6565)  weight_decay: 0.0500 (0.0500)  time: 1.7509  data: 0.8789  max mem: 55573
Epoch: [146]  [ 20/156]  eta: 0:02:42  lr: 0.004620  min_lr: 0.004620  loss: 3.5554 (3.4657)  weight_decay: 0.0500 (0.0500)  time: 0.5979  data: 0.0004  max mem: 55573
Epoch: [146]  [ 30/156]  eta: 0:02:06  lr: 0.004618  min_lr: 0.004618  loss: 3.6599 (3.5400)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [146]  [ 40/156]  eta: 0:01:44  lr: 0.004615  min_lr: 0.004615  loss: 3.6777 (3.5165)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [146]  [ 50/156]  eta: 0:01:29  lr: 0.004612  min_lr: 0.004612  loss: 3.3735 (3.4298)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [146]  [ 60/156]  eta: 0:01:16  lr: 0.004609  min_lr: 0.004609  loss: 3.3103 (3.4274)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [146]  [ 70/156]  eta: 0:01:06  lr: 0.004606  min_lr: 0.004606  loss: 3.4147 (3.4190)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [146]  [ 80/156]  eta: 0:00:56  lr: 0.004603  min_lr: 0.004603  loss: 3.5595 (3.4196)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [146]  [ 90/156]  eta: 0:00:48  lr: 0.004601  min_lr: 0.004601  loss: 3.5595 (3.3920)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [146]  [100/156]  eta: 0:00:40  lr: 0.004598  min_lr: 0.004598  loss: 3.0780 (3.3640)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [146]  [110/156]  eta: 0:00:32  lr: 0.004595  min_lr: 0.004595  loss: 3.2583 (3.3529)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [146]  [120/156]  eta: 0:00:25  lr: 0.004592  min_lr: 0.004592  loss: 3.4323 (3.3600)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [146]  [130/156]  eta: 0:00:17  lr: 0.004589  min_lr: 0.004589  loss: 3.5991 (3.3584)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0011  max mem: 55573
Epoch: [146]  [140/156]  eta: 0:00:10  lr: 0.004586  min_lr: 0.004586  loss: 3.5586 (3.3744)  weight_decay: 0.0500 (0.0500)  time: 0.5870  data: 0.0010  max mem: 55573
Epoch: [146]  [150/156]  eta: 0:00:04  lr: 0.004584  min_lr: 0.004584  loss: 3.3622 (3.3648)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [146]  [155/156]  eta: 0:00:00  lr: 0.004582  min_lr: 0.004582  loss: 3.4010 (3.3668)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [146] Total time: 0:01:45 (0.6784 s / it)
Averaged stats: lr: 0.004582  min_lr: 0.004582  loss: 3.4010 (3.3672)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.0120 (1.0120)  acc1: 79.4271 (79.4271)  acc5: 96.0938 (96.0938)  time: 7.0080  data: 6.7700  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1756 (1.1480)  acc1: 75.9115 (74.2080)  acc5: 93.8802 (92.5440)  time: 1.5480  data: 1.3541  max mem: 55573
Test: Total time: 0:00:07 (1.5816 s / it)
* Acc@1 74.516 Acc@5 92.438 loss 1.142
Accuracy of the model on the 50000 test images: 74.5%
Max accuracy: 74.52%
Test:  [0/5]  eta: 0:00:34  loss: 1.7924 (1.7924)  acc1: 57.5521 (57.5521)  acc5: 81.1198 (81.1198)  time: 6.8190  data: 6.5829  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.8271 (1.8669)  acc1: 56.1198 (53.2160)  acc5: 81.1198 (78.5280)  time: 1.5092  data: 1.3167  max mem: 55573
Test: Total time: 0:00:07 (1.5261 s / it)
* Acc@1 52.200 Acc@5 77.656 loss 1.972
Accuracy of the model EMA on 50000 test images: 52.2%
Max EMA accuracy: 52.20%
Epoch: [147]  [  0/156]  eta: 0:32:06  lr: 0.004582  min_lr: 0.004582  loss: 3.6801 (3.6801)  weight_decay: 0.0500 (0.0500)  time: 12.3501  data: 9.5425  max mem: 55573
Epoch: [147]  [ 10/156]  eta: 0:04:24  lr: 0.004579  min_lr: 0.004579  loss: 3.5410 (3.3388)  weight_decay: 0.0500 (0.0500)  time: 1.8092  data: 0.9112  max mem: 55573
Epoch: [147]  [ 20/156]  eta: 0:02:47  lr: 0.004576  min_lr: 0.004576  loss: 3.4657 (3.3535)  weight_decay: 0.0500 (0.0500)  time: 0.6718  data: 0.0243  max mem: 55573
Epoch: [147]  [ 30/156]  eta: 0:02:08  lr: 0.004573  min_lr: 0.004573  loss: 3.5843 (3.3995)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [147]  [ 40/156]  eta: 0:01:46  lr: 0.004570  min_lr: 0.004570  loss: 3.3994 (3.3294)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [147]  [ 50/156]  eta: 0:01:30  lr: 0.004568  min_lr: 0.004568  loss: 3.3994 (3.3649)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [147]  [ 60/156]  eta: 0:01:17  lr: 0.004565  min_lr: 0.004565  loss: 3.4987 (3.3357)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [147]  [ 70/156]  eta: 0:01:06  lr: 0.004562  min_lr: 0.004562  loss: 3.4899 (3.3399)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [147]  [ 80/156]  eta: 0:00:57  lr: 0.004559  min_lr: 0.004559  loss: 3.6371 (3.3818)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [147]  [ 90/156]  eta: 0:00:48  lr: 0.004556  min_lr: 0.004556  loss: 3.4984 (3.3740)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [147]  [100/156]  eta: 0:00:40  lr: 0.004553  min_lr: 0.004553  loss: 3.3483 (3.3604)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [147]  [110/156]  eta: 0:00:32  lr: 0.004550  min_lr: 0.004550  loss: 3.4396 (3.3542)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [147]  [120/156]  eta: 0:00:25  lr: 0.004548  min_lr: 0.004548  loss: 3.3584 (3.3390)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [147]  [130/156]  eta: 0:00:18  lr: 0.004545  min_lr: 0.004545  loss: 2.9680 (3.3189)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0010  max mem: 55573
Epoch: [147]  [140/156]  eta: 0:00:10  lr: 0.004542  min_lr: 0.004542  loss: 3.1954 (3.3198)  weight_decay: 0.0500 (0.0500)  time: 0.5884  data: 0.0009  max mem: 55573
Epoch: [147]  [150/156]  eta: 0:00:04  lr: 0.004539  min_lr: 0.004539  loss: 3.6082 (3.3305)  weight_decay: 0.0500 (0.0500)  time: 0.5832  data: 0.0001  max mem: 55573
Epoch: [147]  [155/156]  eta: 0:00:00  lr: 0.004538  min_lr: 0.004538  loss: 3.6082 (3.3372)  weight_decay: 0.0500 (0.0500)  time: 0.5828  data: 0.0001  max mem: 55573
Epoch: [147] Total time: 0:01:46 (0.6824 s / it)
Averaged stats: lr: 0.004538  min_lr: 0.004538  loss: 3.6082 (3.3890)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.1327 (1.1327)  acc1: 78.2552 (78.2552)  acc5: 94.9219 (94.9219)  time: 6.8996  data: 6.6625  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2649 (1.2981)  acc1: 75.6510 (73.0560)  acc5: 94.6615 (91.9040)  time: 1.5264  data: 1.3326  max mem: 55573
Test: Total time: 0:00:07 (1.5729 s / it)
* Acc@1 73.762 Acc@5 92.214 loss 1.291
Accuracy of the model on the 50000 test images: 73.8%
Max accuracy: 74.52%
Test:  [0/5]  eta: 0:00:35  loss: 1.7561 (1.7561)  acc1: 58.3333 (58.3333)  acc5: 81.9010 (81.9010)  time: 7.0317  data: 6.7955  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.7922 (1.8291)  acc1: 56.3802 (53.7280)  acc5: 81.9010 (79.1040)  time: 1.5517  data: 1.3592  max mem: 55573
Test: Total time: 0:00:07 (1.5710 s / it)
* Acc@1 52.976 Acc@5 78.234 loss 1.929
Accuracy of the model EMA on 50000 test images: 53.0%
Max EMA accuracy: 52.98%
Epoch: [148]  [  0/156]  eta: 0:33:18  lr: 0.004537  min_lr: 0.004537  loss: 3.1926 (3.1926)  weight_decay: 0.0500 (0.0500)  time: 12.8106  data: 12.1979  max mem: 55573
Epoch: [148]  [ 10/156]  eta: 0:04:09  lr: 0.004535  min_lr: 0.004535  loss: 3.4614 (3.3038)  weight_decay: 0.0500 (0.0500)  time: 1.7112  data: 1.1092  max mem: 55573
Epoch: [148]  [ 20/156]  eta: 0:02:40  lr: 0.004532  min_lr: 0.004532  loss: 3.6304 (3.4203)  weight_decay: 0.0500 (0.0500)  time: 0.5955  data: 0.0004  max mem: 55573
Epoch: [148]  [ 30/156]  eta: 0:02:04  lr: 0.004529  min_lr: 0.004529  loss: 3.5890 (3.3076)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [148]  [ 40/156]  eta: 0:01:43  lr: 0.004526  min_lr: 0.004526  loss: 3.2188 (3.3225)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [148]  [ 50/156]  eta: 0:01:28  lr: 0.004523  min_lr: 0.004523  loss: 3.2282 (3.3000)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [148]  [ 60/156]  eta: 0:01:16  lr: 0.004520  min_lr: 0.004520  loss: 3.4373 (3.3191)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [148]  [ 70/156]  eta: 0:01:05  lr: 0.004517  min_lr: 0.004517  loss: 3.5487 (3.3380)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [148]  [ 80/156]  eta: 0:00:56  lr: 0.004515  min_lr: 0.004515  loss: 3.3703 (3.3196)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [148]  [ 90/156]  eta: 0:00:47  lr: 0.004512  min_lr: 0.004512  loss: 3.4136 (3.3271)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [148]  [100/156]  eta: 0:00:39  lr: 0.004509  min_lr: 0.004509  loss: 3.4240 (3.3133)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [148]  [110/156]  eta: 0:00:32  lr: 0.004506  min_lr: 0.004506  loss: 3.5164 (3.3320)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [148]  [120/156]  eta: 0:00:24  lr: 0.004503  min_lr: 0.004503  loss: 3.6274 (3.3253)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0005  max mem: 55573
Epoch: [148]  [130/156]  eta: 0:00:17  lr: 0.004500  min_lr: 0.004500  loss: 3.3523 (3.3286)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0010  max mem: 55573
Epoch: [148]  [140/156]  eta: 0:00:10  lr: 0.004497  min_lr: 0.004497  loss: 3.4990 (3.3417)  weight_decay: 0.0500 (0.0500)  time: 0.5857  data: 0.0009  max mem: 55573
Epoch: [148]  [150/156]  eta: 0:00:04  lr: 0.004495  min_lr: 0.004495  loss: 3.4186 (3.3520)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [148]  [155/156]  eta: 0:00:00  lr: 0.004493  min_lr: 0.004493  loss: 3.4081 (3.3483)  weight_decay: 0.0500 (0.0500)  time: 0.5835  data: 0.0001  max mem: 55573
Epoch: [148] Total time: 0:01:45 (0.6752 s / it)
Averaged stats: lr: 0.004493  min_lr: 0.004493  loss: 3.4081 (3.3810)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.0904 (1.0904)  acc1: 77.8646 (77.8646)  acc5: 94.9219 (94.9219)  time: 7.0277  data: 6.7903  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1515 (1.1864)  acc1: 75.5208 (73.6960)  acc5: 94.2708 (92.5760)  time: 1.5521  data: 1.3581  max mem: 55573
Test: Total time: 0:00:07 (1.5950 s / it)
* Acc@1 74.474 Acc@5 92.516 loss 1.195
Accuracy of the model on the 50000 test images: 74.5%
Max accuracy: 74.52%
Test:  [0/5]  eta: 0:00:35  loss: 1.7221 (1.7221)  acc1: 59.1146 (59.1146)  acc5: 82.2917 (82.2917)  time: 7.1153  data: 6.8789  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.7585 (1.7933)  acc1: 56.6406 (54.4000)  acc5: 82.2917 (79.5520)  time: 1.5683  data: 1.3759  max mem: 55573
Test: Total time: 0:00:07 (1.5974 s / it)
* Acc@1 53.690 Acc@5 78.828 loss 1.889
Accuracy of the model EMA on 50000 test images: 53.7%
Max EMA accuracy: 53.69%
Epoch: [149]  [  0/156]  eta: 0:37:53  lr: 0.004493  min_lr: 0.004493  loss: 2.9407 (2.9407)  weight_decay: 0.0500 (0.0500)  time: 14.5767  data: 13.0068  max mem: 55573
Epoch: [149]  [ 10/156]  eta: 0:04:38  lr: 0.004490  min_lr: 0.004490  loss: 3.3730 (3.3282)  weight_decay: 0.0500 (0.0500)  time: 1.9082  data: 1.1827  max mem: 55573
Epoch: [149]  [ 20/156]  eta: 0:02:54  lr: 0.004487  min_lr: 0.004487  loss: 3.4671 (3.4403)  weight_decay: 0.0500 (0.0500)  time: 0.6167  data: 0.0003  max mem: 55573
Epoch: [149]  [ 30/156]  eta: 0:02:13  lr: 0.004484  min_lr: 0.004484  loss: 3.4966 (3.4373)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [149]  [ 40/156]  eta: 0:01:49  lr: 0.004481  min_lr: 0.004481  loss: 3.4087 (3.3824)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [149]  [ 50/156]  eta: 0:01:32  lr: 0.004479  min_lr: 0.004479  loss: 3.4128 (3.3505)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [149]  [ 60/156]  eta: 0:01:19  lr: 0.004476  min_lr: 0.004476  loss: 3.4888 (3.3408)  weight_decay: 0.0500 (0.0500)  time: 0.5952  data: 0.0005  max mem: 55573
Epoch: [149]  [ 70/156]  eta: 0:01:08  lr: 0.004473  min_lr: 0.004473  loss: 3.6141 (3.3863)  weight_decay: 0.0500 (0.0500)  time: 0.5931  data: 0.0005  max mem: 55573
Epoch: [149]  [ 80/156]  eta: 0:00:58  lr: 0.004470  min_lr: 0.004470  loss: 3.6540 (3.3814)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [149]  [ 90/156]  eta: 0:00:49  lr: 0.004467  min_lr: 0.004467  loss: 3.5051 (3.3658)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [149]  [100/156]  eta: 0:00:41  lr: 0.004464  min_lr: 0.004464  loss: 3.5051 (3.3593)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [149]  [110/156]  eta: 0:00:33  lr: 0.004461  min_lr: 0.004461  loss: 3.5527 (3.3712)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0004  max mem: 55573
Epoch: [149]  [120/156]  eta: 0:00:25  lr: 0.004459  min_lr: 0.004459  loss: 3.6935 (3.3794)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [149]  [130/156]  eta: 0:00:18  lr: 0.004456  min_lr: 0.004456  loss: 3.4936 (3.3597)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0010  max mem: 55573
Epoch: [149]  [140/156]  eta: 0:00:11  lr: 0.004453  min_lr: 0.004453  loss: 2.9530 (3.3475)  weight_decay: 0.0500 (0.0500)  time: 0.5866  data: 0.0008  max mem: 55573
Epoch: [149]  [150/156]  eta: 0:00:04  lr: 0.004450  min_lr: 0.004450  loss: 3.4855 (3.3534)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [149]  [155/156]  eta: 0:00:00  lr: 0.004449  min_lr: 0.004449  loss: 2.9530 (3.3418)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [149] Total time: 0:01:47 (0.6903 s / it)
Averaged stats: lr: 0.004449  min_lr: 0.004449  loss: 2.9530 (3.3566)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.0111 (1.0111)  acc1: 77.8646 (77.8646)  acc5: 94.7917 (94.7917)  time: 7.1382  data: 6.8973  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0976 (1.1140)  acc1: 76.3021 (74.4000)  acc5: 94.5312 (92.6400)  time: 1.5740  data: 1.3796  max mem: 55573
Test: Total time: 0:00:08 (1.6190 s / it)
* Acc@1 74.328 Acc@5 92.372 loss 1.113
Accuracy of the model on the 50000 test images: 74.3%
Max accuracy: 74.52%
Test:  [0/5]  eta: 0:00:33  loss: 1.6902 (1.6902)  acc1: 60.0260 (60.0260)  acc5: 82.5521 (82.5521)  time: 6.7381  data: 6.5019  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.7270 (1.7597)  acc1: 57.1615 (55.2000)  acc5: 82.5521 (79.9360)  time: 1.4929  data: 1.3005  max mem: 55573
Test: Total time: 0:00:07 (1.5145 s / it)
* Acc@1 54.468 Acc@5 79.414 loss 1.851
Accuracy of the model EMA on 50000 test images: 54.5%
Max EMA accuracy: 54.47%
Epoch: [150]  [  0/156]  eta: 0:35:06  lr: 0.004448  min_lr: 0.004448  loss: 3.7174 (3.7174)  weight_decay: 0.0500 (0.0500)  time: 13.5019  data: 11.0447  max mem: 55573
Epoch: [150]  [ 10/156]  eta: 0:04:21  lr: 0.004445  min_lr: 0.004445  loss: 3.4160 (3.2174)  weight_decay: 0.0500 (0.0500)  time: 1.7880  data: 1.0044  max mem: 55573
Epoch: [150]  [ 20/156]  eta: 0:02:45  lr: 0.004443  min_lr: 0.004443  loss: 3.1288 (3.1835)  weight_decay: 0.0500 (0.0500)  time: 0.6042  data: 0.0004  max mem: 55573
Epoch: [150]  [ 30/156]  eta: 0:02:08  lr: 0.004440  min_lr: 0.004440  loss: 3.3392 (3.1777)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [150]  [ 40/156]  eta: 0:01:45  lr: 0.004437  min_lr: 0.004437  loss: 3.3392 (3.2025)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [150]  [ 50/156]  eta: 0:01:30  lr: 0.004434  min_lr: 0.004434  loss: 3.3655 (3.2467)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [150]  [ 60/156]  eta: 0:01:17  lr: 0.004431  min_lr: 0.004431  loss: 3.5706 (3.3070)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [150]  [ 70/156]  eta: 0:01:06  lr: 0.004428  min_lr: 0.004428  loss: 3.6207 (3.3202)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [150]  [ 80/156]  eta: 0:00:57  lr: 0.004425  min_lr: 0.004425  loss: 3.4409 (3.3379)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [150]  [ 90/156]  eta: 0:00:48  lr: 0.004423  min_lr: 0.004423  loss: 3.5107 (3.3327)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [150]  [100/156]  eta: 0:00:40  lr: 0.004420  min_lr: 0.004420  loss: 3.2644 (3.3058)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [150]  [110/156]  eta: 0:00:32  lr: 0.004417  min_lr: 0.004417  loss: 3.0118 (3.3077)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [150]  [120/156]  eta: 0:00:25  lr: 0.004414  min_lr: 0.004414  loss: 3.4591 (3.3159)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [150]  [130/156]  eta: 0:00:17  lr: 0.004411  min_lr: 0.004411  loss: 3.4889 (3.3241)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0011  max mem: 55573
Epoch: [150]  [140/156]  eta: 0:00:10  lr: 0.004408  min_lr: 0.004408  loss: 3.5390 (3.3227)  weight_decay: 0.0500 (0.0500)  time: 0.5866  data: 0.0009  max mem: 55573
Epoch: [150]  [150/156]  eta: 0:00:04  lr: 0.004405  min_lr: 0.004405  loss: 3.3683 (3.3289)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [150]  [155/156]  eta: 0:00:00  lr: 0.004404  min_lr: 0.004404  loss: 3.2929 (3.3307)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [150] Total time: 0:01:46 (0.6806 s / it)
Averaged stats: lr: 0.004404  min_lr: 0.004404  loss: 3.2929 (3.3586)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.0492 (1.0492)  acc1: 78.6458 (78.6458)  acc5: 94.4010 (94.4010)  time: 7.1734  data: 6.9360  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1290 (1.1516)  acc1: 77.2135 (74.3360)  acc5: 94.4010 (92.9920)  time: 1.5809  data: 1.3873  max mem: 55573
Test: Total time: 0:00:08 (1.6254 s / it)
* Acc@1 74.832 Acc@5 92.678 loss 1.162
Accuracy of the model on the 50000 test images: 74.8%
Max accuracy: 74.83%
Test:  [0/5]  eta: 0:00:34  loss: 1.6611 (1.6611)  acc1: 60.1562 (60.1562)  acc5: 82.8125 (82.8125)  time: 6.8066  data: 6.5706  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.6978 (1.7287)  acc1: 57.5521 (55.7760)  acc5: 82.8125 (80.2240)  time: 1.5067  data: 1.3142  max mem: 55573
Test: Total time: 0:00:07 (1.5233 s / it)
* Acc@1 55.122 Acc@5 79.996 loss 1.816
Accuracy of the model EMA on 50000 test images: 55.1%
Max EMA accuracy: 55.12%
Epoch: [151]  [  0/156]  eta: 0:34:19  lr: 0.004404  min_lr: 0.004404  loss: 3.2982 (3.2982)  weight_decay: 0.0500 (0.0500)  time: 13.2038  data: 8.6946  max mem: 55573
Epoch: [151]  [ 10/156]  eta: 0:04:15  lr: 0.004401  min_lr: 0.004401  loss: 3.3602 (3.3960)  weight_decay: 0.0500 (0.0500)  time: 1.7506  data: 0.8022  max mem: 55573
Epoch: [151]  [ 20/156]  eta: 0:02:42  lr: 0.004398  min_lr: 0.004398  loss: 3.4837 (3.4542)  weight_decay: 0.0500 (0.0500)  time: 0.5979  data: 0.0067  max mem: 55573
Epoch: [151]  [ 30/156]  eta: 0:02:06  lr: 0.004395  min_lr: 0.004395  loss: 3.4577 (3.4320)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [151]  [ 40/156]  eta: 0:01:44  lr: 0.004392  min_lr: 0.004392  loss: 3.4801 (3.4542)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [151]  [ 50/156]  eta: 0:01:29  lr: 0.004389  min_lr: 0.004389  loss: 3.5331 (3.4614)  weight_decay: 0.0500 (0.0500)  time: 0.5937  data: 0.0004  max mem: 55573
Epoch: [151]  [ 60/156]  eta: 0:01:16  lr: 0.004387  min_lr: 0.004387  loss: 3.6562 (3.4766)  weight_decay: 0.0500 (0.0500)  time: 0.5932  data: 0.0005  max mem: 55573
Epoch: [151]  [ 70/156]  eta: 0:01:06  lr: 0.004384  min_lr: 0.004384  loss: 3.5596 (3.4791)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [151]  [ 80/156]  eta: 0:00:56  lr: 0.004381  min_lr: 0.004381  loss: 3.5359 (3.4599)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [151]  [ 90/156]  eta: 0:00:48  lr: 0.004378  min_lr: 0.004378  loss: 3.3860 (3.4465)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [151]  [100/156]  eta: 0:00:40  lr: 0.004375  min_lr: 0.004375  loss: 3.2480 (3.4112)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [151]  [110/156]  eta: 0:00:32  lr: 0.004372  min_lr: 0.004372  loss: 3.4709 (3.4236)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [151]  [120/156]  eta: 0:00:25  lr: 0.004369  min_lr: 0.004369  loss: 3.6554 (3.4297)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [151]  [130/156]  eta: 0:00:17  lr: 0.004366  min_lr: 0.004366  loss: 3.5353 (3.4222)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0011  max mem: 55573
Epoch: [151]  [140/156]  eta: 0:00:10  lr: 0.004364  min_lr: 0.004364  loss: 3.4002 (3.4180)  weight_decay: 0.0500 (0.0500)  time: 0.5852  data: 0.0009  max mem: 55573
Epoch: [151]  [150/156]  eta: 0:00:04  lr: 0.004361  min_lr: 0.004361  loss: 3.4010 (3.4050)  weight_decay: 0.0500 (0.0500)  time: 0.5828  data: 0.0002  max mem: 55573
Epoch: [151]  [155/156]  eta: 0:00:00  lr: 0.004359  min_lr: 0.004359  loss: 3.4591 (3.4089)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0002  max mem: 55573
Epoch: [151] Total time: 0:01:45 (0.6783 s / it)
Averaged stats: lr: 0.004359  min_lr: 0.004359  loss: 3.4591 (3.3589)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.0927 (1.0927)  acc1: 79.4271 (79.4271)  acc5: 94.5312 (94.5312)  time: 7.1466  data: 6.9092  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2713 (1.2334)  acc1: 75.9115 (74.9760)  acc5: 92.9688 (92.0960)  time: 1.5758  data: 1.3819  max mem: 55573
Test: Total time: 0:00:08 (1.6286 s / it)
* Acc@1 74.900 Acc@5 92.516 loss 1.229
Accuracy of the model on the 50000 test images: 74.9%
Max accuracy: 74.90%
Test:  [0/5]  eta: 0:00:34  loss: 1.6322 (1.6322)  acc1: 60.8073 (60.8073)  acc5: 83.2031 (83.2031)  time: 6.8854  data: 6.6492  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.6696 (1.6983)  acc1: 58.5938 (56.2880)  acc5: 83.2031 (80.7360)  time: 1.5224  data: 1.3300  max mem: 55573
Test: Total time: 0:00:07 (1.5406 s / it)
* Acc@1 55.756 Acc@5 80.516 loss 1.782
Accuracy of the model EMA on 50000 test images: 55.8%
Max EMA accuracy: 55.76%
Epoch: [152]  [  0/156]  eta: 0:34:03  lr: 0.004359  min_lr: 0.004359  loss: 3.4798 (3.4798)  weight_decay: 0.0500 (0.0500)  time: 13.0990  data: 9.1865  max mem: 55573
Epoch: [152]  [ 10/156]  eta: 0:04:13  lr: 0.004356  min_lr: 0.004356  loss: 3.5854 (3.4543)  weight_decay: 0.0500 (0.0500)  time: 1.7339  data: 0.8355  max mem: 55573
Epoch: [152]  [ 20/156]  eta: 0:02:41  lr: 0.004353  min_lr: 0.004353  loss: 3.3970 (3.3238)  weight_decay: 0.0500 (0.0500)  time: 0.5944  data: 0.0004  max mem: 55573
Epoch: [152]  [ 30/156]  eta: 0:02:05  lr: 0.004350  min_lr: 0.004350  loss: 3.2547 (3.3204)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [152]  [ 40/156]  eta: 0:01:44  lr: 0.004348  min_lr: 0.004348  loss: 3.2942 (3.3175)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [152]  [ 50/156]  eta: 0:01:28  lr: 0.004345  min_lr: 0.004345  loss: 3.3868 (3.3149)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0004  max mem: 55573
Epoch: [152]  [ 60/156]  eta: 0:01:16  lr: 0.004342  min_lr: 0.004342  loss: 3.4666 (3.3280)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [152]  [ 70/156]  eta: 0:01:06  lr: 0.004339  min_lr: 0.004339  loss: 3.1916 (3.3127)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [152]  [ 80/156]  eta: 0:00:56  lr: 0.004336  min_lr: 0.004336  loss: 3.0701 (3.2938)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [152]  [ 90/156]  eta: 0:00:48  lr: 0.004333  min_lr: 0.004333  loss: 3.3032 (3.2862)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [152]  [100/156]  eta: 0:00:40  lr: 0.004330  min_lr: 0.004330  loss: 3.3323 (3.2928)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [152]  [110/156]  eta: 0:00:32  lr: 0.004327  min_lr: 0.004327  loss: 3.3289 (3.2893)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [152]  [120/156]  eta: 0:00:25  lr: 0.004325  min_lr: 0.004325  loss: 3.4401 (3.3060)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [152]  [130/156]  eta: 0:00:17  lr: 0.004322  min_lr: 0.004322  loss: 3.5862 (3.3049)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0016  max mem: 55573
Epoch: [152]  [140/156]  eta: 0:00:10  lr: 0.004319  min_lr: 0.004319  loss: 3.2267 (3.2906)  weight_decay: 0.0500 (0.0500)  time: 0.5852  data: 0.0015  max mem: 55573
Epoch: [152]  [150/156]  eta: 0:00:04  lr: 0.004316  min_lr: 0.004316  loss: 3.3440 (3.3092)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [152]  [155/156]  eta: 0:00:00  lr: 0.004315  min_lr: 0.004315  loss: 3.6815 (3.3244)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [152] Total time: 0:01:45 (0.6766 s / it)
Averaged stats: lr: 0.004315  min_lr: 0.004315  loss: 3.6815 (3.3645)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.1433 (1.1433)  acc1: 77.3438 (77.3438)  acc5: 94.5312 (94.5312)  time: 7.1758  data: 6.9383  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2152 (1.2443)  acc1: 74.4792 (72.7680)  acc5: 93.7500 (91.7120)  time: 1.5817  data: 1.3877  max mem: 55573
Test: Total time: 0:00:08 (1.6207 s / it)
* Acc@1 73.866 Acc@5 91.932 loss 1.233
Accuracy of the model on the 50000 test images: 73.9%
Max accuracy: 74.90%
Test:  [0/5]  eta: 0:00:33  loss: 1.6058 (1.6058)  acc1: 61.4583 (61.4583)  acc5: 83.3333 (83.3333)  time: 6.6767  data: 6.4405  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.6436 (1.6699)  acc1: 58.4635 (56.6400)  acc5: 83.3333 (81.1840)  time: 1.4919  data: 1.2995  max mem: 55573
Test: Total time: 0:00:07 (1.5162 s / it)
* Acc@1 56.368 Acc@5 81.014 loss 1.750
Accuracy of the model EMA on 50000 test images: 56.4%
Max EMA accuracy: 56.37%
Epoch: [153]  [  0/156]  eta: 0:37:26  lr: 0.004314  min_lr: 0.004314  loss: 3.3260 (3.3260)  weight_decay: 0.0500 (0.0500)  time: 14.4000  data: 9.0028  max mem: 55573
Epoch: [153]  [ 10/156]  eta: 0:04:30  lr: 0.004311  min_lr: 0.004311  loss: 3.3628 (3.1649)  weight_decay: 0.0500 (0.0500)  time: 1.8548  data: 0.8188  max mem: 55573
Epoch: [153]  [ 20/156]  eta: 0:02:50  lr: 0.004309  min_lr: 0.004309  loss: 3.4595 (3.3114)  weight_decay: 0.0500 (0.0500)  time: 0.5965  data: 0.0004  max mem: 55573
Epoch: [153]  [ 30/156]  eta: 0:02:11  lr: 0.004306  min_lr: 0.004306  loss: 3.6723 (3.3447)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [153]  [ 40/156]  eta: 0:01:47  lr: 0.004303  min_lr: 0.004303  loss: 3.4846 (3.3440)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [153]  [ 50/156]  eta: 0:01:31  lr: 0.004300  min_lr: 0.004300  loss: 3.4846 (3.3489)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [153]  [ 60/156]  eta: 0:01:18  lr: 0.004297  min_lr: 0.004297  loss: 3.4551 (3.3444)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0005  max mem: 55573
Epoch: [153]  [ 70/156]  eta: 0:01:07  lr: 0.004294  min_lr: 0.004294  loss: 3.3874 (3.3692)  weight_decay: 0.0500 (0.0500)  time: 0.5936  data: 0.0004  max mem: 55573
Epoch: [153]  [ 80/156]  eta: 0:00:58  lr: 0.004291  min_lr: 0.004291  loss: 3.3206 (3.3555)  weight_decay: 0.0500 (0.0500)  time: 0.5991  data: 0.0005  max mem: 55573
Epoch: [153]  [ 90/156]  eta: 0:00:49  lr: 0.004288  min_lr: 0.004288  loss: 3.4429 (3.3648)  weight_decay: 0.0500 (0.0500)  time: 0.5992  data: 0.0005  max mem: 55573
Epoch: [153]  [100/156]  eta: 0:00:40  lr: 0.004286  min_lr: 0.004286  loss: 3.5221 (3.3861)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0005  max mem: 55573
Epoch: [153]  [110/156]  eta: 0:00:33  lr: 0.004283  min_lr: 0.004283  loss: 3.6213 (3.3751)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [153]  [120/156]  eta: 0:00:25  lr: 0.004280  min_lr: 0.004280  loss: 3.4019 (3.3759)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0004  max mem: 55573
Epoch: [153]  [130/156]  eta: 0:00:18  lr: 0.004277  min_lr: 0.004277  loss: 3.3937 (3.3692)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0010  max mem: 55573
Epoch: [153]  [140/156]  eta: 0:00:11  lr: 0.004274  min_lr: 0.004274  loss: 3.6439 (3.3899)  weight_decay: 0.0500 (0.0500)  time: 0.5858  data: 0.0009  max mem: 55573
Epoch: [153]  [150/156]  eta: 0:00:04  lr: 0.004271  min_lr: 0.004271  loss: 3.6321 (3.3888)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [153]  [155/156]  eta: 0:00:00  lr: 0.004270  min_lr: 0.004270  loss: 3.5585 (3.3873)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [153] Total time: 0:01:47 (0.6871 s / it)
Averaged stats: lr: 0.004270  min_lr: 0.004270  loss: 3.5585 (3.3445)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.1000 (1.1000)  acc1: 78.5156 (78.5156)  acc5: 95.0521 (95.0521)  time: 7.0241  data: 6.7867  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2504 (1.2992)  acc1: 75.5208 (73.7280)  acc5: 94.5312 (92.3200)  time: 1.5512  data: 1.3574  max mem: 55573
Test: Total time: 0:00:08 (1.6082 s / it)
* Acc@1 74.652 Acc@5 92.548 loss 1.272
Accuracy of the model on the 50000 test images: 74.7%
Max accuracy: 74.90%
Test:  [0/5]  eta: 0:00:36  loss: 1.5799 (1.5799)  acc1: 61.7188 (61.7188)  acc5: 83.8542 (83.8542)  time: 7.3271  data: 7.0911  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.6186 (1.6426)  acc1: 59.5052 (57.1840)  acc5: 83.8542 (81.4720)  time: 1.6108  data: 1.4183  max mem: 55573
Test: Total time: 0:00:08 (1.6307 s / it)
* Acc@1 56.886 Acc@5 81.446 loss 1.720
Accuracy of the model EMA on 50000 test images: 56.9%
Max EMA accuracy: 56.89%
Epoch: [154]  [  0/156]  eta: 0:35:28  lr: 0.004270  min_lr: 0.004270  loss: 3.8963 (3.8963)  weight_decay: 0.0500 (0.0500)  time: 13.6435  data: 13.0491  max mem: 55573
Epoch: [154]  [ 10/156]  eta: 0:04:21  lr: 0.004267  min_lr: 0.004267  loss: 2.9742 (3.0502)  weight_decay: 0.0500 (0.0500)  time: 1.7928  data: 1.1866  max mem: 55573
Epoch: [154]  [ 20/156]  eta: 0:02:45  lr: 0.004264  min_lr: 0.004264  loss: 3.0872 (3.2134)  weight_decay: 0.0500 (0.0500)  time: 0.5991  data: 0.0004  max mem: 55573
Epoch: [154]  [ 30/156]  eta: 0:02:08  lr: 0.004261  min_lr: 0.004261  loss: 3.4872 (3.2959)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [154]  [ 40/156]  eta: 0:01:45  lr: 0.004258  min_lr: 0.004258  loss: 3.3440 (3.2690)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [154]  [ 50/156]  eta: 0:01:30  lr: 0.004255  min_lr: 0.004255  loss: 3.3440 (3.2737)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [154]  [ 60/156]  eta: 0:01:17  lr: 0.004252  min_lr: 0.004252  loss: 3.3488 (3.2719)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [154]  [ 70/156]  eta: 0:01:06  lr: 0.004249  min_lr: 0.004249  loss: 3.5503 (3.3279)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [154]  [ 80/156]  eta: 0:00:57  lr: 0.004247  min_lr: 0.004247  loss: 3.5503 (3.3307)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [154]  [ 90/156]  eta: 0:00:48  lr: 0.004244  min_lr: 0.004244  loss: 3.5203 (3.3347)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [154]  [100/156]  eta: 0:00:40  lr: 0.004241  min_lr: 0.004241  loss: 3.5203 (3.3164)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [154]  [110/156]  eta: 0:00:32  lr: 0.004238  min_lr: 0.004238  loss: 3.4005 (3.3251)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [154]  [120/156]  eta: 0:00:25  lr: 0.004235  min_lr: 0.004235  loss: 3.4674 (3.3437)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [154]  [130/156]  eta: 0:00:17  lr: 0.004232  min_lr: 0.004232  loss: 3.5937 (3.3462)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0022  max mem: 55573
Epoch: [154]  [140/156]  eta: 0:00:10  lr: 0.004229  min_lr: 0.004229  loss: 3.6308 (3.3706)  weight_decay: 0.0500 (0.0500)  time: 0.5863  data: 0.0020  max mem: 55573
Epoch: [154]  [150/156]  eta: 0:00:04  lr: 0.004226  min_lr: 0.004226  loss: 3.5870 (3.3539)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [154]  [155/156]  eta: 0:00:00  lr: 0.004225  min_lr: 0.004225  loss: 3.4475 (3.3602)  weight_decay: 0.0500 (0.0500)  time: 0.5829  data: 0.0001  max mem: 55573
Epoch: [154] Total time: 0:01:46 (0.6815 s / it)
Averaged stats: lr: 0.004225  min_lr: 0.004225  loss: 3.4475 (3.3562)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.1239 (1.1239)  acc1: 76.8229 (76.8229)  acc5: 94.7917 (94.7917)  time: 6.7301  data: 6.4928  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2286 (1.2626)  acc1: 76.8229 (73.6640)  acc5: 94.2708 (92.3520)  time: 1.4915  data: 1.2987  max mem: 55573
Test: Total time: 0:00:07 (1.5379 s / it)
* Acc@1 74.962 Acc@5 92.694 loss 1.252
Accuracy of the model on the 50000 test images: 75.0%
Max accuracy: 74.96%
Test:  [0/5]  eta: 0:00:35  loss: 1.5547 (1.5547)  acc1: 62.1094 (62.1094)  acc5: 84.6354 (84.6354)  time: 7.1584  data: 6.9219  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5941 (1.6161)  acc1: 60.1562 (57.8240)  acc5: 84.6354 (81.9840)  time: 1.5772  data: 1.3845  max mem: 55573
Test: Total time: 0:00:07 (1.5945 s / it)
* Acc@1 57.466 Acc@5 81.930 loss 1.690
Accuracy of the model EMA on 50000 test images: 57.5%
Max EMA accuracy: 57.47%
Epoch: [155]  [  0/156]  eta: 0:36:00  lr: 0.004225  min_lr: 0.004225  loss: 3.6642 (3.6642)  weight_decay: 0.0500 (0.0500)  time: 13.8508  data: 8.8443  max mem: 55573
Epoch: [155]  [ 10/156]  eta: 0:04:22  lr: 0.004222  min_lr: 0.004222  loss: 3.4187 (3.3385)  weight_decay: 0.0500 (0.0500)  time: 1.7947  data: 0.8045  max mem: 55573
Epoch: [155]  [ 20/156]  eta: 0:02:46  lr: 0.004219  min_lr: 0.004219  loss: 3.5122 (3.4436)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [155]  [ 30/156]  eta: 0:02:08  lr: 0.004216  min_lr: 0.004216  loss: 3.5619 (3.3991)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [155]  [ 40/156]  eta: 0:01:45  lr: 0.004213  min_lr: 0.004213  loss: 3.3498 (3.3521)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [155]  [ 50/156]  eta: 0:01:30  lr: 0.004210  min_lr: 0.004210  loss: 3.3498 (3.3400)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [155]  [ 60/156]  eta: 0:01:17  lr: 0.004208  min_lr: 0.004208  loss: 3.4078 (3.3108)  weight_decay: 0.0500 (0.0500)  time: 0.5931  data: 0.0004  max mem: 55573
Epoch: [155]  [ 70/156]  eta: 0:01:06  lr: 0.004205  min_lr: 0.004205  loss: 3.4078 (3.3190)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [155]  [ 80/156]  eta: 0:00:57  lr: 0.004202  min_lr: 0.004202  loss: 3.4204 (3.3211)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [155]  [ 90/156]  eta: 0:00:48  lr: 0.004199  min_lr: 0.004199  loss: 3.4738 (3.3498)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [155]  [100/156]  eta: 0:00:40  lr: 0.004196  min_lr: 0.004196  loss: 3.5588 (3.3614)  weight_decay: 0.0500 (0.0500)  time: 0.5977  data: 0.0005  max mem: 55573
Epoch: [155]  [110/156]  eta: 0:00:32  lr: 0.004193  min_lr: 0.004193  loss: 3.4541 (3.3457)  weight_decay: 0.0500 (0.0500)  time: 0.5975  data: 0.0005  max mem: 55573
Epoch: [155]  [120/156]  eta: 0:00:25  lr: 0.004190  min_lr: 0.004190  loss: 3.2817 (3.3487)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0004  max mem: 55573
Epoch: [155]  [130/156]  eta: 0:00:18  lr: 0.004187  min_lr: 0.004187  loss: 3.5254 (3.3678)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0010  max mem: 55573
Epoch: [155]  [140/156]  eta: 0:00:10  lr: 0.004185  min_lr: 0.004185  loss: 3.5754 (3.3710)  weight_decay: 0.0500 (0.0500)  time: 0.5850  data: 0.0008  max mem: 55573
Epoch: [155]  [150/156]  eta: 0:00:04  lr: 0.004182  min_lr: 0.004182  loss: 3.4050 (3.3694)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [155]  [155/156]  eta: 0:00:00  lr: 0.004180  min_lr: 0.004180  loss: 3.3944 (3.3653)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [155] Total time: 0:01:46 (0.6822 s / it)
Averaged stats: lr: 0.004180  min_lr: 0.004180  loss: 3.3944 (3.3585)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.0264 (1.0264)  acc1: 77.6042 (77.6042)  acc5: 95.5729 (95.5729)  time: 6.9553  data: 6.7179  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1634 (1.2008)  acc1: 76.1719 (73.2800)  acc5: 95.0521 (92.7360)  time: 1.5373  data: 1.3436  max mem: 55573
Test: Total time: 0:00:07 (1.5853 s / it)
* Acc@1 74.960 Acc@5 92.698 loss 1.197
Accuracy of the model on the 50000 test images: 75.0%
Max accuracy: 74.96%
Test:  [0/5]  eta: 0:00:35  loss: 1.5318 (1.5318)  acc1: 62.6302 (62.6302)  acc5: 85.0260 (85.0260)  time: 7.0691  data: 6.8332  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5718 (1.5924)  acc1: 60.5469 (58.2720)  acc5: 85.0260 (82.2080)  time: 1.5592  data: 1.3667  max mem: 55573
Test: Total time: 0:00:07 (1.5788 s / it)
* Acc@1 58.004 Acc@5 82.322 loss 1.663
Accuracy of the model EMA on 50000 test images: 58.0%
Max EMA accuracy: 58.00%
Epoch: [156]  [  0/156]  eta: 0:35:25  lr: 0.004180  min_lr: 0.004180  loss: 2.1633 (2.1633)  weight_decay: 0.0500 (0.0500)  time: 13.6218  data: 10.4366  max mem: 55573
Epoch: [156]  [ 10/156]  eta: 0:04:25  lr: 0.004177  min_lr: 0.004177  loss: 3.5722 (3.3424)  weight_decay: 0.0500 (0.0500)  time: 1.8191  data: 0.9953  max mem: 55573
Epoch: [156]  [ 20/156]  eta: 0:02:47  lr: 0.004174  min_lr: 0.004174  loss: 3.4604 (3.2767)  weight_decay: 0.0500 (0.0500)  time: 0.6143  data: 0.0258  max mem: 55573
Epoch: [156]  [ 30/156]  eta: 0:02:09  lr: 0.004171  min_lr: 0.004171  loss: 3.2561 (3.2865)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [156]  [ 40/156]  eta: 0:01:46  lr: 0.004168  min_lr: 0.004168  loss: 3.2562 (3.3036)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [156]  [ 50/156]  eta: 0:01:30  lr: 0.004166  min_lr: 0.004166  loss: 2.9923 (3.2317)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [156]  [ 60/156]  eta: 0:01:17  lr: 0.004163  min_lr: 0.004163  loss: 3.2453 (3.2802)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [156]  [ 70/156]  eta: 0:01:07  lr: 0.004160  min_lr: 0.004160  loss: 3.4633 (3.2898)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [156]  [ 80/156]  eta: 0:00:57  lr: 0.004157  min_lr: 0.004157  loss: 3.4633 (3.3045)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [156]  [ 90/156]  eta: 0:00:48  lr: 0.004154  min_lr: 0.004154  loss: 3.3266 (3.2894)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0005  max mem: 55573
Epoch: [156]  [100/156]  eta: 0:00:40  lr: 0.004151  min_lr: 0.004151  loss: 3.3020 (3.2814)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [156]  [110/156]  eta: 0:00:32  lr: 0.004148  min_lr: 0.004148  loss: 3.2526 (3.2700)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [156]  [120/156]  eta: 0:00:25  lr: 0.004145  min_lr: 0.004145  loss: 2.9713 (3.2556)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0005  max mem: 55573
Epoch: [156]  [130/156]  eta: 0:00:18  lr: 0.004143  min_lr: 0.004143  loss: 3.1934 (3.2635)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0010  max mem: 55573
Epoch: [156]  [140/156]  eta: 0:00:10  lr: 0.004140  min_lr: 0.004140  loss: 3.4586 (3.2869)  weight_decay: 0.0500 (0.0500)  time: 0.5852  data: 0.0009  max mem: 55573
Epoch: [156]  [150/156]  eta: 0:00:04  lr: 0.004137  min_lr: 0.004137  loss: 3.4586 (3.2869)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [156]  [155/156]  eta: 0:00:00  lr: 0.004135  min_lr: 0.004135  loss: 3.2869 (3.2908)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [156] Total time: 0:01:46 (0.6826 s / it)
Averaged stats: lr: 0.004135  min_lr: 0.004135  loss: 3.2869 (3.3355)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.1741 (1.1741)  acc1: 76.0417 (76.0417)  acc5: 94.9219 (94.9219)  time: 6.6372  data: 6.4002  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2458 (1.2645)  acc1: 76.0417 (73.5680)  acc5: 94.4010 (92.7680)  time: 1.4727  data: 1.2801  max mem: 55573
Test: Total time: 0:00:07 (1.4893 s / it)
* Acc@1 74.880 Acc@5 92.726 loss 1.253
Accuracy of the model on the 50000 test images: 74.9%
Max accuracy: 74.96%
Test:  [0/5]  eta: 0:00:35  loss: 1.5105 (1.5105)  acc1: 62.7604 (62.7604)  acc5: 85.2865 (85.2865)  time: 7.1970  data: 6.9610  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5514 (1.5701)  acc1: 61.1979 (58.6880)  acc5: 85.2865 (82.4320)  time: 1.5847  data: 1.3923  max mem: 55573
Test: Total time: 0:00:08 (1.6035 s / it)
* Acc@1 58.492 Acc@5 82.676 loss 1.638
Accuracy of the model EMA on 50000 test images: 58.5%
Max EMA accuracy: 58.49%
Epoch: [157]  [  0/156]  eta: 0:37:13  lr: 0.004135  min_lr: 0.004135  loss: 3.3797 (3.3797)  weight_decay: 0.0500 (0.0500)  time: 14.3144  data: 8.7082  max mem: 55573
Epoch: [157]  [ 10/156]  eta: 0:04:29  lr: 0.004132  min_lr: 0.004132  loss: 3.4636 (3.2446)  weight_decay: 0.0500 (0.0500)  time: 1.8432  data: 0.7920  max mem: 55573
Epoch: [157]  [ 20/156]  eta: 0:02:49  lr: 0.004129  min_lr: 0.004129  loss: 3.4636 (3.2810)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0004  max mem: 55573
Epoch: [157]  [ 30/156]  eta: 0:02:10  lr: 0.004126  min_lr: 0.004126  loss: 3.3513 (3.3032)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [157]  [ 40/156]  eta: 0:01:47  lr: 0.004124  min_lr: 0.004124  loss: 3.4893 (3.3299)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [157]  [ 50/156]  eta: 0:01:31  lr: 0.004121  min_lr: 0.004121  loss: 3.6100 (3.3785)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [157]  [ 60/156]  eta: 0:01:18  lr: 0.004118  min_lr: 0.004118  loss: 3.6584 (3.3988)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [157]  [ 70/156]  eta: 0:01:07  lr: 0.004115  min_lr: 0.004115  loss: 3.4704 (3.3637)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [157]  [ 80/156]  eta: 0:00:57  lr: 0.004112  min_lr: 0.004112  loss: 3.5251 (3.3744)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [157]  [ 90/156]  eta: 0:00:48  lr: 0.004109  min_lr: 0.004109  loss: 3.5337 (3.3716)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [157]  [100/156]  eta: 0:00:40  lr: 0.004106  min_lr: 0.004106  loss: 3.5346 (3.3679)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [157]  [110/156]  eta: 0:00:32  lr: 0.004103  min_lr: 0.004103  loss: 3.3755 (3.3622)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [157]  [120/156]  eta: 0:00:25  lr: 0.004101  min_lr: 0.004101  loss: 3.2884 (3.3446)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [157]  [130/156]  eta: 0:00:18  lr: 0.004098  min_lr: 0.004098  loss: 3.3794 (3.3527)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0010  max mem: 55573
Epoch: [157]  [140/156]  eta: 0:00:11  lr: 0.004095  min_lr: 0.004095  loss: 3.5380 (3.3593)  weight_decay: 0.0500 (0.0500)  time: 0.5863  data: 0.0008  max mem: 55573
Epoch: [157]  [150/156]  eta: 0:00:04  lr: 0.004092  min_lr: 0.004092  loss: 3.5041 (3.3596)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [157]  [155/156]  eta: 0:00:00  lr: 0.004091  min_lr: 0.004091  loss: 3.4701 (3.3668)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [157] Total time: 0:01:46 (0.6844 s / it)
Averaged stats: lr: 0.004091  min_lr: 0.004091  loss: 3.4701 (3.3321)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.0789 (1.0789)  acc1: 78.7760 (78.7760)  acc5: 95.4427 (95.4427)  time: 7.0855  data: 6.8481  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1188 (1.2187)  acc1: 76.4323 (73.6320)  acc5: 95.3125 (93.1840)  time: 1.5636  data: 1.3697  max mem: 55573
Test: Total time: 0:00:08 (1.6126 s / it)
* Acc@1 75.110 Acc@5 92.860 loss 1.190
Accuracy of the model on the 50000 test images: 75.1%
Max accuracy: 75.11%
Test:  [0/5]  eta: 0:00:35  loss: 1.4898 (1.4898)  acc1: 63.2812 (63.2812)  acc5: 85.4167 (85.4167)  time: 7.1926  data: 6.9561  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5314 (1.5484)  acc1: 61.8490 (59.2000)  acc5: 85.4167 (82.7200)  time: 1.5840  data: 1.3913  max mem: 55573
Test: Total time: 0:00:08 (1.6009 s / it)
* Acc@1 58.926 Acc@5 83.030 loss 1.614
Accuracy of the model EMA on 50000 test images: 58.9%
Max EMA accuracy: 58.93%
Epoch: [158]  [  0/156]  eta: 0:33:07  lr: 0.004090  min_lr: 0.004090  loss: 2.8991 (2.8991)  weight_decay: 0.0500 (0.0500)  time: 12.7378  data: 10.5927  max mem: 55573
Epoch: [158]  [ 10/156]  eta: 0:04:37  lr: 0.004087  min_lr: 0.004087  loss: 3.6656 (3.4584)  weight_decay: 0.0500 (0.0500)  time: 1.8992  data: 1.1721  max mem: 55573
Epoch: [158]  [ 20/156]  eta: 0:02:53  lr: 0.004084  min_lr: 0.004084  loss: 3.6656 (3.4398)  weight_decay: 0.0500 (0.0500)  time: 0.7024  data: 0.1153  max mem: 55573
Epoch: [158]  [ 30/156]  eta: 0:02:12  lr: 0.004082  min_lr: 0.004082  loss: 3.3965 (3.3972)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [158]  [ 40/156]  eta: 0:01:49  lr: 0.004079  min_lr: 0.004079  loss: 3.3625 (3.3865)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [158]  [ 50/156]  eta: 0:01:32  lr: 0.004076  min_lr: 0.004076  loss: 3.4963 (3.3973)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [158]  [ 60/156]  eta: 0:01:19  lr: 0.004073  min_lr: 0.004073  loss: 3.4963 (3.3833)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [158]  [ 70/156]  eta: 0:01:08  lr: 0.004070  min_lr: 0.004070  loss: 3.5568 (3.3879)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [158]  [ 80/156]  eta: 0:00:58  lr: 0.004067  min_lr: 0.004067  loss: 3.5541 (3.3896)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [158]  [ 90/156]  eta: 0:00:49  lr: 0.004064  min_lr: 0.004064  loss: 3.6715 (3.4304)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [158]  [100/156]  eta: 0:00:41  lr: 0.004061  min_lr: 0.004061  loss: 3.7210 (3.4534)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [158]  [110/156]  eta: 0:00:33  lr: 0.004059  min_lr: 0.004059  loss: 3.5836 (3.4494)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0004  max mem: 55573
Epoch: [158]  [120/156]  eta: 0:00:25  lr: 0.004056  min_lr: 0.004056  loss: 3.4081 (3.4441)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [158]  [130/156]  eta: 0:00:18  lr: 0.004053  min_lr: 0.004053  loss: 3.4506 (3.4496)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0011  max mem: 55573
Epoch: [158]  [140/156]  eta: 0:00:11  lr: 0.004050  min_lr: 0.004050  loss: 2.9979 (3.3993)  weight_decay: 0.0500 (0.0500)  time: 0.5855  data: 0.0009  max mem: 55573
Epoch: [158]  [150/156]  eta: 0:00:04  lr: 0.004047  min_lr: 0.004047  loss: 2.9791 (3.3897)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [158]  [155/156]  eta: 0:00:00  lr: 0.004046  min_lr: 0.004046  loss: 3.2156 (3.3851)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [158] Total time: 0:01:47 (0.6881 s / it)
Averaged stats: lr: 0.004046  min_lr: 0.004046  loss: 3.2156 (3.3329)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:32  loss: 1.0838 (1.0838)  acc1: 78.1250 (78.1250)  acc5: 95.0521 (95.0521)  time: 6.5672  data: 6.3302  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1148 (1.1533)  acc1: 76.5625 (74.5600)  acc5: 95.0521 (93.0240)  time: 1.4588  data: 1.2661  max mem: 55573
Test: Total time: 0:00:07 (1.4765 s / it)
* Acc@1 75.120 Acc@5 92.730 loss 1.166
Accuracy of the model on the 50000 test images: 75.1%
Max accuracy: 75.12%
Test:  [0/5]  eta: 0:00:33  loss: 1.4697 (1.4697)  acc1: 63.9323 (63.9323)  acc5: 85.8073 (85.8073)  time: 6.7186  data: 6.4827  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.5119 (1.5275)  acc1: 62.1094 (59.5520)  acc5: 85.8073 (83.0080)  time: 1.4890  data: 1.2966  max mem: 55573
Test: Total time: 0:00:07 (1.5044 s / it)
* Acc@1 59.408 Acc@5 83.322 loss 1.591
Accuracy of the model EMA on 50000 test images: 59.4%
Max EMA accuracy: 59.41%
Epoch: [159]  [  0/156]  eta: 0:37:04  lr: 0.004045  min_lr: 0.004045  loss: 3.1934 (3.1934)  weight_decay: 0.0500 (0.0500)  time: 14.2586  data: 11.8814  max mem: 55573
Epoch: [159]  [ 10/156]  eta: 0:04:28  lr: 0.004042  min_lr: 0.004042  loss: 3.4524 (3.3234)  weight_decay: 0.0500 (0.0500)  time: 1.8356  data: 1.0805  max mem: 55573
Epoch: [159]  [ 20/156]  eta: 0:02:49  lr: 0.004040  min_lr: 0.004040  loss: 3.4661 (3.3620)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0004  max mem: 55573
Epoch: [159]  [ 30/156]  eta: 0:02:10  lr: 0.004037  min_lr: 0.004037  loss: 3.2157 (3.2385)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [159]  [ 40/156]  eta: 0:01:47  lr: 0.004034  min_lr: 0.004034  loss: 3.1221 (3.2570)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [159]  [ 50/156]  eta: 0:01:31  lr: 0.004031  min_lr: 0.004031  loss: 3.4374 (3.2896)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [159]  [ 60/156]  eta: 0:01:18  lr: 0.004028  min_lr: 0.004028  loss: 3.5180 (3.3408)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [159]  [ 70/156]  eta: 0:01:07  lr: 0.004025  min_lr: 0.004025  loss: 3.5776 (3.3348)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [159]  [ 80/156]  eta: 0:00:57  lr: 0.004022  min_lr: 0.004022  loss: 3.0741 (3.3329)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [159]  [ 90/156]  eta: 0:00:48  lr: 0.004019  min_lr: 0.004019  loss: 3.3023 (3.3274)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [159]  [100/156]  eta: 0:00:40  lr: 0.004017  min_lr: 0.004017  loss: 3.3023 (3.3124)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [159]  [110/156]  eta: 0:00:32  lr: 0.004014  min_lr: 0.004014  loss: 3.3472 (3.3089)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0005  max mem: 55573
Epoch: [159]  [120/156]  eta: 0:00:25  lr: 0.004011  min_lr: 0.004011  loss: 3.3665 (3.3193)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0005  max mem: 55573
Epoch: [159]  [130/156]  eta: 0:00:18  lr: 0.004008  min_lr: 0.004008  loss: 3.4906 (3.3094)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0010  max mem: 55573
Epoch: [159]  [140/156]  eta: 0:00:11  lr: 0.004005  min_lr: 0.004005  loss: 3.5810 (3.3284)  weight_decay: 0.0500 (0.0500)  time: 0.5863  data: 0.0008  max mem: 55573
Epoch: [159]  [150/156]  eta: 0:00:04  lr: 0.004002  min_lr: 0.004002  loss: 3.5802 (3.3341)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [159]  [155/156]  eta: 0:00:00  lr: 0.004001  min_lr: 0.004001  loss: 3.5792 (3.3458)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0001  max mem: 55573
Epoch: [159] Total time: 0:01:46 (0.6843 s / it)
Averaged stats: lr: 0.004001  min_lr: 0.004001  loss: 3.5792 (3.3317)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.1676 (1.1676)  acc1: 78.5156 (78.5156)  acc5: 94.9219 (94.9219)  time: 7.0468  data: 6.8094  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2375 (1.2825)  acc1: 76.1719 (73.9520)  acc5: 94.7917 (92.3520)  time: 1.5557  data: 1.3620  max mem: 55573
Test: Total time: 0:00:07 (1.6000 s / it)
* Acc@1 74.546 Acc@5 92.480 loss 1.288
Accuracy of the model on the 50000 test images: 74.5%
Max accuracy: 75.12%
Test:  [0/5]  eta: 0:00:33  loss: 1.4508 (1.4508)  acc1: 64.3229 (64.3229)  acc5: 86.0677 (86.0677)  time: 6.6910  data: 6.4548  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4935 (1.5079)  acc1: 62.3698 (60.0000)  acc5: 86.0677 (83.2000)  time: 1.4933  data: 1.3008  max mem: 55573
Test: Total time: 0:00:07 (1.5122 s / it)
* Acc@1 59.794 Acc@5 83.582 loss 1.569
Accuracy of the model EMA on 50000 test images: 59.8%
Max EMA accuracy: 59.79%
Epoch: [160]  [  0/156]  eta: 0:34:45  lr: 0.004001  min_lr: 0.004001  loss: 3.5150 (3.5150)  weight_decay: 0.0500 (0.0500)  time: 13.3707  data: 9.1671  max mem: 55573
Epoch: [160]  [ 10/156]  eta: 0:04:17  lr: 0.003998  min_lr: 0.003998  loss: 3.5150 (3.3202)  weight_decay: 0.0500 (0.0500)  time: 1.7609  data: 0.8337  max mem: 55573
Epoch: [160]  [ 20/156]  eta: 0:02:44  lr: 0.003995  min_lr: 0.003995  loss: 3.5244 (3.4634)  weight_decay: 0.0500 (0.0500)  time: 0.5978  data: 0.0004  max mem: 55573
Epoch: [160]  [ 30/156]  eta: 0:02:06  lr: 0.003992  min_lr: 0.003992  loss: 3.3883 (3.3349)  weight_decay: 0.0500 (0.0500)  time: 0.5934  data: 0.0004  max mem: 55573
Epoch: [160]  [ 40/156]  eta: 0:01:45  lr: 0.003989  min_lr: 0.003989  loss: 3.4317 (3.3718)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [160]  [ 50/156]  eta: 0:01:29  lr: 0.003986  min_lr: 0.003986  loss: 3.4903 (3.3712)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [160]  [ 60/156]  eta: 0:01:17  lr: 0.003983  min_lr: 0.003983  loss: 3.4972 (3.3430)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [160]  [ 70/156]  eta: 0:01:06  lr: 0.003980  min_lr: 0.003980  loss: 3.5297 (3.3495)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [160]  [ 80/156]  eta: 0:00:56  lr: 0.003977  min_lr: 0.003977  loss: 3.5297 (3.3577)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0005  max mem: 55573
Epoch: [160]  [ 90/156]  eta: 0:00:48  lr: 0.003975  min_lr: 0.003975  loss: 3.4328 (3.3500)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0005  max mem: 55573
Epoch: [160]  [100/156]  eta: 0:00:40  lr: 0.003972  min_lr: 0.003972  loss: 3.1997 (3.3174)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [160]  [110/156]  eta: 0:00:32  lr: 0.003969  min_lr: 0.003969  loss: 3.0763 (3.3102)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [160]  [120/156]  eta: 0:00:25  lr: 0.003966  min_lr: 0.003966  loss: 3.3773 (3.3154)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [160]  [130/156]  eta: 0:00:17  lr: 0.003963  min_lr: 0.003963  loss: 3.5036 (3.3257)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0010  max mem: 55573
Epoch: [160]  [140/156]  eta: 0:00:10  lr: 0.003960  min_lr: 0.003960  loss: 3.5036 (3.3313)  weight_decay: 0.0500 (0.0500)  time: 0.5860  data: 0.0009  max mem: 55573
Epoch: [160]  [150/156]  eta: 0:00:04  lr: 0.003957  min_lr: 0.003957  loss: 3.4223 (3.3193)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [160]  [155/156]  eta: 0:00:00  lr: 0.003956  min_lr: 0.003956  loss: 3.0646 (3.3090)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [160] Total time: 0:01:45 (0.6792 s / it)
Averaged stats: lr: 0.003956  min_lr: 0.003956  loss: 3.0646 (3.3241)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9448 (0.9448)  acc1: 78.2552 (78.2552)  acc5: 95.5729 (95.5729)  time: 6.9733  data: 6.7360  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0401 (1.0908)  acc1: 78.2552 (74.6560)  acc5: 94.2708 (92.7360)  time: 1.5410  data: 1.3473  max mem: 55573
Test: Total time: 0:00:07 (1.5785 s / it)
* Acc@1 75.876 Acc@5 93.330 loss 1.080
Accuracy of the model on the 50000 test images: 75.9%
Max accuracy: 75.88%
Test:  [0/5]  eta: 0:00:34  loss: 1.4334 (1.4334)  acc1: 64.9740 (64.9740)  acc5: 86.0677 (86.0677)  time: 6.9208  data: 6.6848  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4766 (1.4900)  acc1: 62.7604 (60.5760)  acc5: 86.0677 (83.3280)  time: 1.5300  data: 1.3371  max mem: 55573
Test: Total time: 0:00:07 (1.5497 s / it)
* Acc@1 60.200 Acc@5 83.834 loss 1.549
Accuracy of the model EMA on 50000 test images: 60.2%
Max EMA accuracy: 60.20%
Epoch: [161]  [  0/156]  eta: 0:32:07  lr: 0.003956  min_lr: 0.003956  loss: 2.6711 (2.6711)  weight_decay: 0.0500 (0.0500)  time: 12.3526  data: 8.5741  max mem: 55573
Epoch: [161]  [ 10/156]  eta: 0:04:06  lr: 0.003953  min_lr: 0.003953  loss: 3.1873 (3.0974)  weight_decay: 0.0500 (0.0500)  time: 1.6909  data: 0.7799  max mem: 55573
Epoch: [161]  [ 20/156]  eta: 0:02:38  lr: 0.003950  min_lr: 0.003950  loss: 3.2669 (3.1645)  weight_decay: 0.0500 (0.0500)  time: 0.6079  data: 0.0004  max mem: 55573
Epoch: [161]  [ 30/156]  eta: 0:02:03  lr: 0.003947  min_lr: 0.003947  loss: 3.2669 (3.1963)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [161]  [ 40/156]  eta: 0:01:42  lr: 0.003944  min_lr: 0.003944  loss: 3.2352 (3.2159)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [161]  [ 50/156]  eta: 0:01:27  lr: 0.003941  min_lr: 0.003941  loss: 3.4446 (3.2667)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [161]  [ 60/156]  eta: 0:01:15  lr: 0.003938  min_lr: 0.003938  loss: 3.5001 (3.2763)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [161]  [ 70/156]  eta: 0:01:05  lr: 0.003935  min_lr: 0.003935  loss: 3.5107 (3.2474)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [161]  [ 80/156]  eta: 0:00:56  lr: 0.003933  min_lr: 0.003933  loss: 3.4039 (3.2641)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [161]  [ 90/156]  eta: 0:00:47  lr: 0.003930  min_lr: 0.003930  loss: 3.4054 (3.2963)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [161]  [100/156]  eta: 0:00:39  lr: 0.003927  min_lr: 0.003927  loss: 3.5276 (3.3260)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [161]  [110/156]  eta: 0:00:32  lr: 0.003924  min_lr: 0.003924  loss: 3.5391 (3.3458)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [161]  [120/156]  eta: 0:00:24  lr: 0.003921  min_lr: 0.003921  loss: 3.5100 (3.3564)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [161]  [130/156]  eta: 0:00:17  lr: 0.003918  min_lr: 0.003918  loss: 3.4268 (3.3451)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0011  max mem: 55573
Epoch: [161]  [140/156]  eta: 0:00:10  lr: 0.003915  min_lr: 0.003915  loss: 3.4268 (3.3530)  weight_decay: 0.0500 (0.0500)  time: 0.5872  data: 0.0010  max mem: 55573
Epoch: [161]  [150/156]  eta: 0:00:04  lr: 0.003912  min_lr: 0.003912  loss: 3.3726 (3.3384)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [161]  [155/156]  eta: 0:00:00  lr: 0.003911  min_lr: 0.003911  loss: 3.3384 (3.3350)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [161] Total time: 0:01:45 (0.6740 s / it)
Averaged stats: lr: 0.003911  min_lr: 0.003911  loss: 3.3384 (3.3634)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9383 (0.9383)  acc1: 79.5573 (79.5573)  acc5: 95.3125 (95.3125)  time: 6.9889  data: 6.7513  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0335 (1.0735)  acc1: 76.8229 (74.6560)  acc5: 95.1823 (92.9600)  time: 1.5442  data: 1.3504  max mem: 55573
Test: Total time: 0:00:07 (1.5887 s / it)
* Acc@1 75.428 Acc@5 92.782 loss 1.067
Accuracy of the model on the 50000 test images: 75.4%
Max accuracy: 75.88%
Test:  [0/5]  eta: 0:00:35  loss: 1.4166 (1.4166)  acc1: 65.3646 (65.3646)  acc5: 86.3281 (86.3281)  time: 7.1590  data: 6.9229  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4602 (1.4726)  acc1: 63.1510 (61.0560)  acc5: 86.3281 (83.7120)  time: 1.5771  data: 1.3847  max mem: 55573
Test: Total time: 0:00:07 (1.5942 s / it)
* Acc@1 60.620 Acc@5 84.126 loss 1.530
Accuracy of the model EMA on 50000 test images: 60.6%
Max EMA accuracy: 60.62%
Epoch: [162]  [  0/156]  eta: 0:34:56  lr: 0.003911  min_lr: 0.003911  loss: 3.8140 (3.8140)  weight_decay: 0.0500 (0.0500)  time: 13.4371  data: 12.8500  max mem: 55573
Epoch: [162]  [ 10/156]  eta: 0:04:16  lr: 0.003908  min_lr: 0.003908  loss: 3.5238 (3.4496)  weight_decay: 0.0500 (0.0500)  time: 1.7582  data: 1.1685  max mem: 55573
Epoch: [162]  [ 20/156]  eta: 0:02:43  lr: 0.003905  min_lr: 0.003905  loss: 3.4695 (3.3913)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [162]  [ 30/156]  eta: 0:02:06  lr: 0.003902  min_lr: 0.003902  loss: 3.4732 (3.3938)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [162]  [ 40/156]  eta: 0:01:44  lr: 0.003899  min_lr: 0.003899  loss: 3.3810 (3.3690)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [162]  [ 50/156]  eta: 0:01:29  lr: 0.003896  min_lr: 0.003896  loss: 3.2104 (3.3414)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [162]  [ 60/156]  eta: 0:01:17  lr: 0.003894  min_lr: 0.003894  loss: 3.3313 (3.3400)  weight_decay: 0.0500 (0.0500)  time: 0.5937  data: 0.0005  max mem: 55573
Epoch: [162]  [ 70/156]  eta: 0:01:06  lr: 0.003891  min_lr: 0.003891  loss: 3.4458 (3.3395)  weight_decay: 0.0500 (0.0500)  time: 0.5940  data: 0.0004  max mem: 55573
Epoch: [162]  [ 80/156]  eta: 0:00:56  lr: 0.003888  min_lr: 0.003888  loss: 3.5648 (3.3566)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [162]  [ 90/156]  eta: 0:00:48  lr: 0.003885  min_lr: 0.003885  loss: 3.5648 (3.3646)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [162]  [100/156]  eta: 0:00:40  lr: 0.003882  min_lr: 0.003882  loss: 3.5432 (3.3484)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [162]  [110/156]  eta: 0:00:32  lr: 0.003879  min_lr: 0.003879  loss: 3.2857 (3.3305)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [162]  [120/156]  eta: 0:00:25  lr: 0.003876  min_lr: 0.003876  loss: 3.2036 (3.3259)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [162]  [130/156]  eta: 0:00:17  lr: 0.003873  min_lr: 0.003873  loss: 3.4265 (3.3316)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0024  max mem: 55573
Epoch: [162]  [140/156]  eta: 0:00:10  lr: 0.003871  min_lr: 0.003871  loss: 3.3203 (3.3197)  weight_decay: 0.0500 (0.0500)  time: 0.5873  data: 0.0023  max mem: 55573
Epoch: [162]  [150/156]  eta: 0:00:04  lr: 0.003868  min_lr: 0.003868  loss: 3.1625 (3.3047)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0001  max mem: 55573
Epoch: [162]  [155/156]  eta: 0:00:00  lr: 0.003866  min_lr: 0.003866  loss: 3.1625 (3.3090)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [162] Total time: 0:01:45 (0.6787 s / it)
Averaged stats: lr: 0.003866  min_lr: 0.003866  loss: 3.1625 (3.3072)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.0478 (1.0478)  acc1: 78.6458 (78.6458)  acc5: 94.7917 (94.7917)  time: 6.9062  data: 6.6692  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1178 (1.1167)  acc1: 76.6927 (74.3680)  acc5: 94.1406 (92.3520)  time: 1.5276  data: 1.3339  max mem: 55573
Test: Total time: 0:00:07 (1.5666 s / it)
* Acc@1 75.206 Acc@5 92.818 loss 1.115
Accuracy of the model on the 50000 test images: 75.2%
Max accuracy: 75.88%
Test:  [0/5]  eta: 0:00:34  loss: 1.4016 (1.4016)  acc1: 66.0156 (66.0156)  acc5: 86.4583 (86.4583)  time: 6.8483  data: 6.6123  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4454 (1.4567)  acc1: 64.1927 (61.7600)  acc5: 86.4583 (83.9680)  time: 1.5150  data: 1.3226  max mem: 55573
Test: Total time: 0:00:07 (1.5329 s / it)
* Acc@1 61.050 Acc@5 84.402 loss 1.512
Accuracy of the model EMA on 50000 test images: 61.1%
Max EMA accuracy: 61.05%
Epoch: [163]  [  0/156]  eta: 0:32:00  lr: 0.003866  min_lr: 0.003866  loss: 2.1192 (2.1192)  weight_decay: 0.0500 (0.0500)  time: 12.3118  data: 8.8474  max mem: 55573
Epoch: [163]  [ 10/156]  eta: 0:04:12  lr: 0.003863  min_lr: 0.003863  loss: 3.7415 (3.6000)  weight_decay: 0.0500 (0.0500)  time: 1.7308  data: 0.8047  max mem: 55573
Epoch: [163]  [ 20/156]  eta: 0:02:41  lr: 0.003860  min_lr: 0.003860  loss: 3.6856 (3.5885)  weight_decay: 0.0500 (0.0500)  time: 0.6329  data: 0.0004  max mem: 55573
Epoch: [163]  [ 30/156]  eta: 0:02:05  lr: 0.003857  min_lr: 0.003857  loss: 3.5686 (3.4284)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [163]  [ 40/156]  eta: 0:01:44  lr: 0.003854  min_lr: 0.003854  loss: 2.9775 (3.3186)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [163]  [ 50/156]  eta: 0:01:28  lr: 0.003852  min_lr: 0.003852  loss: 3.3290 (3.3289)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [163]  [ 60/156]  eta: 0:01:16  lr: 0.003849  min_lr: 0.003849  loss: 3.5623 (3.3323)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [163]  [ 70/156]  eta: 0:01:05  lr: 0.003846  min_lr: 0.003846  loss: 3.4911 (3.2969)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [163]  [ 80/156]  eta: 0:00:56  lr: 0.003843  min_lr: 0.003843  loss: 3.0311 (3.2777)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0004  max mem: 55573
Epoch: [163]  [ 90/156]  eta: 0:00:48  lr: 0.003840  min_lr: 0.003840  loss: 3.2617 (3.2507)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [163]  [100/156]  eta: 0:00:40  lr: 0.003837  min_lr: 0.003837  loss: 3.1476 (3.2228)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [163]  [110/156]  eta: 0:00:32  lr: 0.003834  min_lr: 0.003834  loss: 2.8841 (3.2021)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [163]  [120/156]  eta: 0:00:24  lr: 0.003831  min_lr: 0.003831  loss: 3.1616 (3.2177)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [163]  [130/156]  eta: 0:00:17  lr: 0.003829  min_lr: 0.003829  loss: 3.4301 (3.2129)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0011  max mem: 55573
Epoch: [163]  [140/156]  eta: 0:00:10  lr: 0.003826  min_lr: 0.003826  loss: 3.4973 (3.2310)  weight_decay: 0.0500 (0.0500)  time: 0.5858  data: 0.0010  max mem: 55573
Epoch: [163]  [150/156]  eta: 0:00:04  lr: 0.003823  min_lr: 0.003823  loss: 3.4973 (3.2357)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [163]  [155/156]  eta: 0:00:00  lr: 0.003821  min_lr: 0.003821  loss: 3.4314 (3.2414)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [163] Total time: 0:01:45 (0.6760 s / it)
Averaged stats: lr: 0.003821  min_lr: 0.003821  loss: 3.4314 (3.3012)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.1205 (1.1205)  acc1: 78.3854 (78.3854)  acc5: 94.9219 (94.9219)  time: 6.8136  data: 6.5763  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1838 (1.1810)  acc1: 77.6042 (74.5920)  acc5: 94.1406 (93.2160)  time: 1.5089  data: 1.3153  max mem: 55573
Test: Total time: 0:00:07 (1.5511 s / it)
* Acc@1 75.906 Acc@5 93.150 loss 1.158
Accuracy of the model on the 50000 test images: 75.9%
Max accuracy: 75.91%
Test:  [0/5]  eta: 0:00:34  loss: 1.3859 (1.3859)  acc1: 66.2760 (66.2760)  acc5: 86.5885 (86.5885)  time: 6.8137  data: 6.5777  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4304 (1.4407)  acc1: 64.4531 (61.9840)  acc5: 86.5885 (84.1280)  time: 1.5082  data: 1.3157  max mem: 55573
Test: Total time: 0:00:07 (1.5250 s / it)
* Acc@1 61.422 Acc@5 84.628 loss 1.494
Accuracy of the model EMA on 50000 test images: 61.4%
Max EMA accuracy: 61.42%
Epoch: [164]  [  0/156]  eta: 0:34:26  lr: 0.003821  min_lr: 0.003821  loss: 2.8559 (2.8559)  weight_decay: 0.0500 (0.0500)  time: 13.2499  data: 11.8291  max mem: 55573
Epoch: [164]  [ 10/156]  eta: 0:04:19  lr: 0.003818  min_lr: 0.003818  loss: 3.5464 (3.4248)  weight_decay: 0.0500 (0.0500)  time: 1.7771  data: 1.0757  max mem: 55573
Epoch: [164]  [ 20/156]  eta: 0:02:44  lr: 0.003815  min_lr: 0.003815  loss: 3.3421 (3.2783)  weight_decay: 0.0500 (0.0500)  time: 0.6107  data: 0.0005  max mem: 55573
Epoch: [164]  [ 30/156]  eta: 0:02:07  lr: 0.003812  min_lr: 0.003812  loss: 3.4328 (3.3957)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [164]  [ 40/156]  eta: 0:01:45  lr: 0.003810  min_lr: 0.003810  loss: 3.5066 (3.3717)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [164]  [ 50/156]  eta: 0:01:29  lr: 0.003807  min_lr: 0.003807  loss: 3.4769 (3.3625)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [164]  [ 60/156]  eta: 0:01:17  lr: 0.003804  min_lr: 0.003804  loss: 3.4287 (3.3277)  weight_decay: 0.0500 (0.0500)  time: 0.5958  data: 0.0004  max mem: 55573
Epoch: [164]  [ 70/156]  eta: 0:01:06  lr: 0.003801  min_lr: 0.003801  loss: 3.4287 (3.3302)  weight_decay: 0.0500 (0.0500)  time: 0.5953  data: 0.0004  max mem: 55573
Epoch: [164]  [ 80/156]  eta: 0:00:57  lr: 0.003798  min_lr: 0.003798  loss: 3.4912 (3.3233)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [164]  [ 90/156]  eta: 0:00:48  lr: 0.003795  min_lr: 0.003795  loss: 3.4125 (3.3076)  weight_decay: 0.0500 (0.0500)  time: 0.5965  data: 0.0004  max mem: 55573
Epoch: [164]  [100/156]  eta: 0:00:40  lr: 0.003792  min_lr: 0.003792  loss: 3.0624 (3.2868)  weight_decay: 0.0500 (0.0500)  time: 0.5966  data: 0.0004  max mem: 55573
Epoch: [164]  [110/156]  eta: 0:00:32  lr: 0.003789  min_lr: 0.003789  loss: 3.0624 (3.2717)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [164]  [120/156]  eta: 0:00:25  lr: 0.003787  min_lr: 0.003787  loss: 3.2079 (3.2782)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [164]  [130/156]  eta: 0:00:18  lr: 0.003784  min_lr: 0.003784  loss: 3.4750 (3.2884)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0010  max mem: 55573
Epoch: [164]  [140/156]  eta: 0:00:10  lr: 0.003781  min_lr: 0.003781  loss: 3.4750 (3.2996)  weight_decay: 0.0500 (0.0500)  time: 0.5873  data: 0.0009  max mem: 55573
Epoch: [164]  [150/156]  eta: 0:00:04  lr: 0.003778  min_lr: 0.003778  loss: 3.7124 (3.3080)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [164]  [155/156]  eta: 0:00:00  lr: 0.003777  min_lr: 0.003777  loss: 3.4169 (3.3085)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [164] Total time: 0:01:46 (0.6818 s / it)
Averaged stats: lr: 0.003777  min_lr: 0.003777  loss: 3.4169 (3.3138)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.1630 (1.1630)  acc1: 77.2135 (77.2135)  acc5: 93.6198 (93.6198)  time: 6.7671  data: 6.5297  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1630 (1.2110)  acc1: 77.2135 (74.4000)  acc5: 93.6198 (92.5440)  time: 1.4999  data: 1.3060  max mem: 55573
Test: Total time: 0:00:07 (1.5311 s / it)
* Acc@1 75.404 Acc@5 92.938 loss 1.181
Accuracy of the model on the 50000 test images: 75.4%
Max accuracy: 75.91%
Test:  [0/5]  eta: 0:00:33  loss: 1.3707 (1.3707)  acc1: 66.4062 (66.4062)  acc5: 86.7188 (86.7188)  time: 6.7900  data: 6.5540  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4154 (1.4247)  acc1: 64.7135 (62.0800)  acc5: 86.7188 (84.3840)  time: 1.5033  data: 1.3109  max mem: 55573
Test: Total time: 0:00:07 (1.5197 s / it)
* Acc@1 61.770 Acc@5 84.874 loss 1.477
Accuracy of the model EMA on 50000 test images: 61.8%
Max EMA accuracy: 61.77%
Epoch: [165]  [  0/156]  eta: 0:35:57  lr: 0.003776  min_lr: 0.003776  loss: 3.4557 (3.4557)  weight_decay: 0.0500 (0.0500)  time: 13.8305  data: 10.6897  max mem: 55573
Epoch: [165]  [ 10/156]  eta: 0:04:22  lr: 0.003773  min_lr: 0.003773  loss: 3.2027 (3.1556)  weight_decay: 0.0500 (0.0500)  time: 1.8003  data: 0.9721  max mem: 55573
Epoch: [165]  [ 20/156]  eta: 0:02:46  lr: 0.003771  min_lr: 0.003771  loss: 3.5290 (3.2980)  weight_decay: 0.0500 (0.0500)  time: 0.5944  data: 0.0004  max mem: 55573
Epoch: [165]  [ 30/156]  eta: 0:02:08  lr: 0.003768  min_lr: 0.003768  loss: 3.4120 (3.1918)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [165]  [ 40/156]  eta: 0:01:46  lr: 0.003765  min_lr: 0.003765  loss: 3.0809 (3.1801)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [165]  [ 50/156]  eta: 0:01:30  lr: 0.003762  min_lr: 0.003762  loss: 3.1490 (3.2004)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [165]  [ 60/156]  eta: 0:01:17  lr: 0.003759  min_lr: 0.003759  loss: 3.2569 (3.2258)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [165]  [ 70/156]  eta: 0:01:06  lr: 0.003756  min_lr: 0.003756  loss: 3.4256 (3.2574)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [165]  [ 80/156]  eta: 0:00:57  lr: 0.003753  min_lr: 0.003753  loss: 3.3895 (3.2572)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [165]  [ 90/156]  eta: 0:00:48  lr: 0.003750  min_lr: 0.003750  loss: 3.2368 (3.2575)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0004  max mem: 55573
Epoch: [165]  [100/156]  eta: 0:00:40  lr: 0.003748  min_lr: 0.003748  loss: 3.2211 (3.2623)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [165]  [110/156]  eta: 0:00:32  lr: 0.003745  min_lr: 0.003745  loss: 3.4596 (3.2880)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [165]  [120/156]  eta: 0:00:25  lr: 0.003742  min_lr: 0.003742  loss: 3.5647 (3.2960)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [165]  [130/156]  eta: 0:00:17  lr: 0.003739  min_lr: 0.003739  loss: 3.5647 (3.3111)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0010  max mem: 55573
Epoch: [165]  [140/156]  eta: 0:00:10  lr: 0.003736  min_lr: 0.003736  loss: 3.4354 (3.2917)  weight_decay: 0.0500 (0.0500)  time: 0.5874  data: 0.0009  max mem: 55573
Epoch: [165]  [150/156]  eta: 0:00:04  lr: 0.003733  min_lr: 0.003733  loss: 3.0079 (3.2695)  weight_decay: 0.0500 (0.0500)  time: 0.5830  data: 0.0001  max mem: 55573
Epoch: [165]  [155/156]  eta: 0:00:00  lr: 0.003732  min_lr: 0.003732  loss: 3.0079 (3.2754)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [165] Total time: 0:01:46 (0.6813 s / it)
Averaged stats: lr: 0.003732  min_lr: 0.003732  loss: 3.0079 (3.2913)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.0191 (1.0191)  acc1: 78.3854 (78.3854)  acc5: 94.9219 (94.9219)  time: 6.7478  data: 6.5105  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0437 (1.1355)  acc1: 78.3854 (75.2960)  acc5: 94.9219 (93.1200)  time: 1.4960  data: 1.3022  max mem: 55573
Test: Total time: 0:00:07 (1.5325 s / it)
* Acc@1 75.798 Acc@5 93.234 loss 1.134
Accuracy of the model on the 50000 test images: 75.8%
Max accuracy: 75.91%
Test:  [0/5]  eta: 0:00:34  loss: 1.3555 (1.3555)  acc1: 66.9271 (66.9271)  acc5: 86.7188 (86.7188)  time: 6.8724  data: 6.6364  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.4010 (1.4096)  acc1: 65.1042 (62.4000)  acc5: 86.7188 (84.4800)  time: 1.5199  data: 1.3274  max mem: 55573
Test: Total time: 0:00:07 (1.5385 s / it)
* Acc@1 62.090 Acc@5 85.106 loss 1.460
Accuracy of the model EMA on 50000 test images: 62.1%
Max EMA accuracy: 62.09%
Epoch: [166]  [  0/156]  eta: 0:34:16  lr: 0.003731  min_lr: 0.003731  loss: 3.6388 (3.6388)  weight_decay: 0.0500 (0.0500)  time: 13.1825  data: 8.2911  max mem: 55573
Epoch: [166]  [ 10/156]  eta: 0:04:15  lr: 0.003729  min_lr: 0.003729  loss: 3.0961 (3.0583)  weight_decay: 0.0500 (0.0500)  time: 1.7481  data: 0.7541  max mem: 55573
Epoch: [166]  [ 20/156]  eta: 0:02:42  lr: 0.003726  min_lr: 0.003726  loss: 3.2052 (3.1767)  weight_decay: 0.0500 (0.0500)  time: 0.5963  data: 0.0004  max mem: 55573
Epoch: [166]  [ 30/156]  eta: 0:02:06  lr: 0.003723  min_lr: 0.003723  loss: 3.3370 (3.2145)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [166]  [ 40/156]  eta: 0:01:44  lr: 0.003720  min_lr: 0.003720  loss: 3.3864 (3.2094)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [166]  [ 50/156]  eta: 0:01:29  lr: 0.003717  min_lr: 0.003717  loss: 3.4410 (3.2306)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [166]  [ 60/156]  eta: 0:01:16  lr: 0.003714  min_lr: 0.003714  loss: 3.0939 (3.2019)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [166]  [ 70/156]  eta: 0:01:06  lr: 0.003711  min_lr: 0.003711  loss: 3.1417 (3.2283)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [166]  [ 80/156]  eta: 0:00:56  lr: 0.003709  min_lr: 0.003709  loss: 3.4487 (3.2183)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [166]  [ 90/156]  eta: 0:00:48  lr: 0.003706  min_lr: 0.003706  loss: 3.2349 (3.2040)  weight_decay: 0.0500 (0.0500)  time: 0.5933  data: 0.0005  max mem: 55573
Epoch: [166]  [100/156]  eta: 0:00:40  lr: 0.003703  min_lr: 0.003703  loss: 3.1943 (3.1994)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [166]  [110/156]  eta: 0:00:32  lr: 0.003700  min_lr: 0.003700  loss: 3.3146 (3.2059)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [166]  [120/156]  eta: 0:00:25  lr: 0.003697  min_lr: 0.003697  loss: 3.3656 (3.2055)  weight_decay: 0.0500 (0.0500)  time: 0.5883  data: 0.0005  max mem: 55573
Epoch: [166]  [130/156]  eta: 0:00:17  lr: 0.003694  min_lr: 0.003694  loss: 3.3656 (3.2203)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0015  max mem: 55573
Epoch: [166]  [140/156]  eta: 0:00:10  lr: 0.003691  min_lr: 0.003691  loss: 3.4753 (3.2266)  weight_decay: 0.0500 (0.0500)  time: 0.5865  data: 0.0013  max mem: 55573
Epoch: [166]  [150/156]  eta: 0:00:04  lr: 0.003688  min_lr: 0.003688  loss: 3.3232 (3.2219)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [166]  [155/156]  eta: 0:00:00  lr: 0.003687  min_lr: 0.003687  loss: 3.2483 (3.2186)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [166] Total time: 0:01:45 (0.6776 s / it)
Averaged stats: lr: 0.003687  min_lr: 0.003687  loss: 3.2483 (3.3036)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.0476 (1.0476)  acc1: 79.2969 (79.2969)  acc5: 95.0521 (95.0521)  time: 7.1269  data: 6.8895  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1646 (1.1433)  acc1: 75.1302 (74.7200)  acc5: 94.2708 (92.6720)  time: 1.5718  data: 1.3780  max mem: 55573
Test: Total time: 0:00:08 (1.6170 s / it)
* Acc@1 75.436 Acc@5 93.096 loss 1.142
Accuracy of the model on the 50000 test images: 75.4%
Max accuracy: 75.91%
Test:  [0/5]  eta: 0:00:33  loss: 1.3417 (1.3417)  acc1: 67.4479 (67.4479)  acc5: 86.8490 (86.8490)  time: 6.7971  data: 6.5604  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3878 (1.3958)  acc1: 65.3646 (62.6880)  acc5: 86.8490 (84.6720)  time: 1.5049  data: 1.3124  max mem: 55573
Test: Total time: 0:00:07 (1.5259 s / it)
* Acc@1 62.384 Acc@5 85.340 loss 1.445
Accuracy of the model EMA on 50000 test images: 62.4%
Max EMA accuracy: 62.38%
Epoch: [167]  [  0/156]  eta: 0:34:12  lr: 0.003687  min_lr: 0.003687  loss: 3.5207 (3.5207)  weight_decay: 0.0500 (0.0500)  time: 13.1544  data: 8.6535  max mem: 55573
Epoch: [167]  [ 10/156]  eta: 0:04:14  lr: 0.003684  min_lr: 0.003684  loss: 3.5207 (3.2422)  weight_decay: 0.0500 (0.0500)  time: 1.7410  data: 0.7872  max mem: 55573
Epoch: [167]  [ 20/156]  eta: 0:02:42  lr: 0.003681  min_lr: 0.003681  loss: 3.3930 (3.1998)  weight_decay: 0.0500 (0.0500)  time: 0.5947  data: 0.0005  max mem: 55573
Epoch: [167]  [ 30/156]  eta: 0:02:05  lr: 0.003678  min_lr: 0.003678  loss: 3.3930 (3.2360)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [167]  [ 40/156]  eta: 0:01:44  lr: 0.003675  min_lr: 0.003675  loss: 3.3745 (3.2420)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [167]  [ 50/156]  eta: 0:01:28  lr: 0.003672  min_lr: 0.003672  loss: 3.3080 (3.2379)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [167]  [ 60/156]  eta: 0:01:16  lr: 0.003669  min_lr: 0.003669  loss: 3.4001 (3.2687)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [167]  [ 70/156]  eta: 0:01:06  lr: 0.003667  min_lr: 0.003667  loss: 3.3229 (3.2327)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [167]  [ 80/156]  eta: 0:00:56  lr: 0.003664  min_lr: 0.003664  loss: 3.0303 (3.2424)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [167]  [ 90/156]  eta: 0:00:48  lr: 0.003661  min_lr: 0.003661  loss: 3.1970 (3.2441)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [167]  [100/156]  eta: 0:00:40  lr: 0.003658  min_lr: 0.003658  loss: 3.2448 (3.2492)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [167]  [110/156]  eta: 0:00:32  lr: 0.003655  min_lr: 0.003655  loss: 3.2448 (3.2275)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [167]  [120/156]  eta: 0:00:25  lr: 0.003652  min_lr: 0.003652  loss: 3.5295 (3.2504)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [167]  [130/156]  eta: 0:00:17  lr: 0.003649  min_lr: 0.003649  loss: 3.5810 (3.2580)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0010  max mem: 55573
Epoch: [167]  [140/156]  eta: 0:00:10  lr: 0.003647  min_lr: 0.003647  loss: 3.4335 (3.2632)  weight_decay: 0.0500 (0.0500)  time: 0.5850  data: 0.0009  max mem: 55573
Epoch: [167]  [150/156]  eta: 0:00:04  lr: 0.003644  min_lr: 0.003644  loss: 3.4335 (3.2621)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [167]  [155/156]  eta: 0:00:00  lr: 0.003642  min_lr: 0.003642  loss: 3.2748 (3.2584)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [167] Total time: 0:01:45 (0.6775 s / it)
Averaged stats: lr: 0.003642  min_lr: 0.003642  loss: 3.2748 (3.3016)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.0242 (1.0242)  acc1: 79.4271 (79.4271)  acc5: 95.9635 (95.9635)  time: 7.0074  data: 6.7699  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1097 (1.1288)  acc1: 75.3906 (74.0800)  acc5: 95.0521 (92.8960)  time: 1.5473  data: 1.3541  max mem: 55573
Test: Total time: 0:00:07 (1.5888 s / it)
* Acc@1 75.062 Acc@5 92.674 loss 1.126
Accuracy of the model on the 50000 test images: 75.1%
Max accuracy: 75.91%
Test:  [0/5]  eta: 0:00:35  loss: 1.3279 (1.3279)  acc1: 67.9688 (67.9688)  acc5: 86.9792 (86.9792)  time: 7.1457  data: 6.9096  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3747 (1.3822)  acc1: 65.4948 (62.9760)  acc5: 86.9792 (84.8320)  time: 1.5745  data: 1.3820  max mem: 55573
Test: Total time: 0:00:07 (1.5966 s / it)
* Acc@1 62.688 Acc@5 85.504 loss 1.430
Accuracy of the model EMA on 50000 test images: 62.7%
Max EMA accuracy: 62.69%
Epoch: [168]  [  0/156]  eta: 0:30:08  lr: 0.003642  min_lr: 0.003642  loss: 2.9821 (2.9821)  weight_decay: 0.0500 (0.0500)  time: 11.5900  data: 9.0171  max mem: 55573
Epoch: [168]  [ 10/156]  eta: 0:04:20  lr: 0.003639  min_lr: 0.003639  loss: 3.4381 (3.3178)  weight_decay: 0.0500 (0.0500)  time: 1.7866  data: 0.9894  max mem: 55573
Epoch: [168]  [ 20/156]  eta: 0:02:45  lr: 0.003636  min_lr: 0.003636  loss: 3.3657 (3.2722)  weight_decay: 0.0500 (0.0500)  time: 0.6981  data: 0.0935  max mem: 55573
Epoch: [168]  [ 30/156]  eta: 0:02:07  lr: 0.003633  min_lr: 0.003633  loss: 3.3657 (3.2609)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [168]  [ 40/156]  eta: 0:01:45  lr: 0.003631  min_lr: 0.003631  loss: 3.4609 (3.2884)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0004  max mem: 55573
Epoch: [168]  [ 50/156]  eta: 0:01:29  lr: 0.003628  min_lr: 0.003628  loss: 3.5119 (3.3326)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [168]  [ 60/156]  eta: 0:01:17  lr: 0.003625  min_lr: 0.003625  loss: 3.5478 (3.3464)  weight_decay: 0.0500 (0.0500)  time: 0.5945  data: 0.0005  max mem: 55573
Epoch: [168]  [ 70/156]  eta: 0:01:06  lr: 0.003622  min_lr: 0.003622  loss: 3.4985 (3.3554)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0005  max mem: 55573
Epoch: [168]  [ 80/156]  eta: 0:00:57  lr: 0.003619  min_lr: 0.003619  loss: 3.5526 (3.3823)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [168]  [ 90/156]  eta: 0:00:48  lr: 0.003616  min_lr: 0.003616  loss: 3.4161 (3.3565)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [168]  [100/156]  eta: 0:00:40  lr: 0.003613  min_lr: 0.003613  loss: 3.3823 (3.3582)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [168]  [110/156]  eta: 0:00:32  lr: 0.003610  min_lr: 0.003610  loss: 3.2659 (3.3419)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [168]  [120/156]  eta: 0:00:25  lr: 0.003608  min_lr: 0.003608  loss: 3.3288 (3.3488)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [168]  [130/156]  eta: 0:00:17  lr: 0.003605  min_lr: 0.003605  loss: 3.4269 (3.3478)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0011  max mem: 55573
Epoch: [168]  [140/156]  eta: 0:00:10  lr: 0.003602  min_lr: 0.003602  loss: 3.3153 (3.3542)  weight_decay: 0.0500 (0.0500)  time: 0.5855  data: 0.0009  max mem: 55573
Epoch: [168]  [150/156]  eta: 0:00:04  lr: 0.003599  min_lr: 0.003599  loss: 3.3153 (3.3404)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [168]  [155/156]  eta: 0:00:00  lr: 0.003598  min_lr: 0.003598  loss: 3.2990 (3.3431)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [168] Total time: 0:01:46 (0.6804 s / it)
Averaged stats: lr: 0.003598  min_lr: 0.003598  loss: 3.2990 (3.2846)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 1.1285 (1.1285)  acc1: 77.2135 (77.2135)  acc5: 96.0938 (96.0938)  time: 7.3436  data: 7.1061  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2146 (1.2592)  acc1: 76.5625 (73.9840)  acc5: 94.7917 (93.2160)  time: 1.6151  data: 1.4213  max mem: 55573
Test: Total time: 0:00:08 (1.6605 s / it)
* Acc@1 75.566 Acc@5 93.022 loss 1.262
Accuracy of the model on the 50000 test images: 75.6%
Max accuracy: 75.91%
Test:  [0/5]  eta: 0:00:36  loss: 1.3150 (1.3150)  acc1: 68.0990 (68.0990)  acc5: 87.1094 (87.1094)  time: 7.3314  data: 7.0953  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3623 (1.3692)  acc1: 65.7552 (63.2000)  acc5: 87.1094 (84.9280)  time: 1.6116  data: 1.4192  max mem: 55573
Test: Total time: 0:00:08 (1.6273 s / it)
* Acc@1 62.948 Acc@5 85.722 loss 1.416
Accuracy of the model EMA on 50000 test images: 62.9%
Max EMA accuracy: 62.95%
Epoch: [169]  [  0/156]  eta: 0:35:52  lr: 0.003597  min_lr: 0.003597  loss: 3.8414 (3.8414)  weight_decay: 0.0500 (0.0500)  time: 13.8007  data: 10.9082  max mem: 55573
Epoch: [169]  [ 10/156]  eta: 0:04:23  lr: 0.003594  min_lr: 0.003594  loss: 3.5515 (3.4287)  weight_decay: 0.0500 (0.0500)  time: 1.8033  data: 0.9920  max mem: 55573
Epoch: [169]  [ 20/156]  eta: 0:02:46  lr: 0.003592  min_lr: 0.003592  loss: 3.4649 (3.3885)  weight_decay: 0.0500 (0.0500)  time: 0.5990  data: 0.0004  max mem: 55573
Epoch: [169]  [ 30/156]  eta: 0:02:08  lr: 0.003589  min_lr: 0.003589  loss: 3.2838 (3.3803)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0004  max mem: 55573
Epoch: [169]  [ 40/156]  eta: 0:01:46  lr: 0.003586  min_lr: 0.003586  loss: 3.2838 (3.3229)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [169]  [ 50/156]  eta: 0:01:30  lr: 0.003583  min_lr: 0.003583  loss: 3.2997 (3.3169)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [169]  [ 60/156]  eta: 0:01:17  lr: 0.003580  min_lr: 0.003580  loss: 3.4389 (3.3469)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [169]  [ 70/156]  eta: 0:01:07  lr: 0.003577  min_lr: 0.003577  loss: 3.4389 (3.3230)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0005  max mem: 55573
Epoch: [169]  [ 80/156]  eta: 0:00:57  lr: 0.003574  min_lr: 0.003574  loss: 3.3858 (3.3339)  weight_decay: 0.0500 (0.0500)  time: 0.5933  data: 0.0005  max mem: 55573
Epoch: [169]  [ 90/156]  eta: 0:00:48  lr: 0.003572  min_lr: 0.003572  loss: 3.4142 (3.3304)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [169]  [100/156]  eta: 0:00:40  lr: 0.003569  min_lr: 0.003569  loss: 3.4166 (3.3332)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [169]  [110/156]  eta: 0:00:32  lr: 0.003566  min_lr: 0.003566  loss: 3.3614 (3.3240)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [169]  [120/156]  eta: 0:00:25  lr: 0.003563  min_lr: 0.003563  loss: 3.2351 (3.3112)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [169]  [130/156]  eta: 0:00:18  lr: 0.003560  min_lr: 0.003560  loss: 3.1327 (3.2993)  weight_decay: 0.0500 (0.0500)  time: 0.5885  data: 0.0010  max mem: 55573
Epoch: [169]  [140/156]  eta: 0:00:10  lr: 0.003557  min_lr: 0.003557  loss: 3.1327 (3.2958)  weight_decay: 0.0500 (0.0500)  time: 0.5846  data: 0.0009  max mem: 55573
Epoch: [169]  [150/156]  eta: 0:00:04  lr: 0.003554  min_lr: 0.003554  loss: 3.5203 (3.3017)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [169]  [155/156]  eta: 0:00:00  lr: 0.003553  min_lr: 0.003553  loss: 3.5203 (3.2990)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [169] Total time: 0:01:46 (0.6824 s / it)
Averaged stats: lr: 0.003553  min_lr: 0.003553  loss: 3.5203 (3.2743)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.1328 (1.1328)  acc1: 78.5156 (78.5156)  acc5: 94.9219 (94.9219)  time: 7.0755  data: 6.8379  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2125 (1.2411)  acc1: 77.3438 (74.5600)  acc5: 94.9219 (93.1520)  time: 1.5617  data: 1.3677  max mem: 55573
Test: Total time: 0:00:08 (1.6032 s / it)
* Acc@1 75.498 Acc@5 93.046 loss 1.235
Accuracy of the model on the 50000 test images: 75.5%
Max accuracy: 75.91%
Test:  [0/5]  eta: 0:00:34  loss: 1.3034 (1.3034)  acc1: 68.2292 (68.2292)  acc5: 87.3698 (87.3698)  time: 6.9257  data: 6.6896  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3510 (1.3573)  acc1: 65.6250 (63.2640)  acc5: 87.3698 (85.1520)  time: 1.5305  data: 1.3380  max mem: 55573
Test: Total time: 0:00:07 (1.5529 s / it)
* Acc@1 63.172 Acc@5 85.920 loss 1.404
Accuracy of the model EMA on 50000 test images: 63.2%
Max EMA accuracy: 63.17%
Epoch: [170]  [  0/156]  eta: 0:36:23  lr: 0.003553  min_lr: 0.003553  loss: 3.1506 (3.1506)  weight_decay: 0.0500 (0.0500)  time: 13.9980  data: 6.9488  max mem: 55573
Epoch: [170]  [ 10/156]  eta: 0:04:24  lr: 0.003550  min_lr: 0.003550  loss: 3.2551 (3.3371)  weight_decay: 0.0500 (0.0500)  time: 1.8086  data: 0.6320  max mem: 55573
Epoch: [170]  [ 20/156]  eta: 0:02:47  lr: 0.003547  min_lr: 0.003547  loss: 3.4668 (3.4467)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [170]  [ 30/156]  eta: 0:02:08  lr: 0.003544  min_lr: 0.003544  loss: 3.5501 (3.4317)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0007  max mem: 55573
Epoch: [170]  [ 40/156]  eta: 0:01:46  lr: 0.003541  min_lr: 0.003541  loss: 3.4264 (3.3936)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0007  max mem: 55573
Epoch: [170]  [ 50/156]  eta: 0:01:30  lr: 0.003538  min_lr: 0.003538  loss: 3.5620 (3.3952)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [170]  [ 60/156]  eta: 0:01:17  lr: 0.003536  min_lr: 0.003536  loss: 3.3671 (3.3731)  weight_decay: 0.0500 (0.0500)  time: 0.5937  data: 0.0005  max mem: 55573
Epoch: [170]  [ 70/156]  eta: 0:01:07  lr: 0.003533  min_lr: 0.003533  loss: 3.3671 (3.3664)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0004  max mem: 55573
Epoch: [170]  [ 80/156]  eta: 0:00:57  lr: 0.003530  min_lr: 0.003530  loss: 3.4728 (3.3488)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [170]  [ 90/156]  eta: 0:00:48  lr: 0.003527  min_lr: 0.003527  loss: 3.4111 (3.3274)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [170]  [100/156]  eta: 0:00:40  lr: 0.003524  min_lr: 0.003524  loss: 3.2208 (3.3175)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [170]  [110/156]  eta: 0:00:32  lr: 0.003521  min_lr: 0.003521  loss: 3.2208 (3.3007)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [170]  [120/156]  eta: 0:00:25  lr: 0.003518  min_lr: 0.003518  loss: 3.0416 (3.2752)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [170]  [130/156]  eta: 0:00:18  lr: 0.003516  min_lr: 0.003516  loss: 3.1865 (3.2797)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0010  max mem: 55573
Epoch: [170]  [140/156]  eta: 0:00:10  lr: 0.003513  min_lr: 0.003513  loss: 3.5606 (3.3036)  weight_decay: 0.0500 (0.0500)  time: 0.5851  data: 0.0008  max mem: 55573
Epoch: [170]  [150/156]  eta: 0:00:04  lr: 0.003510  min_lr: 0.003510  loss: 3.6638 (3.3165)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [170]  [155/156]  eta: 0:00:00  lr: 0.003508  min_lr: 0.003508  loss: 3.5606 (3.3171)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [170] Total time: 0:01:46 (0.6823 s / it)
Averaged stats: lr: 0.003508  min_lr: 0.003508  loss: 3.5606 (3.2755)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.0318 (1.0318)  acc1: 80.7292 (80.7292)  acc5: 95.1823 (95.1823)  time: 7.1560  data: 6.9187  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1700 (1.2054)  acc1: 75.6510 (74.8800)  acc5: 94.1406 (92.8000)  time: 1.5777  data: 1.3838  max mem: 55573
Test: Total time: 0:00:08 (1.6320 s / it)
* Acc@1 75.520 Acc@5 92.826 loss 1.199
Accuracy of the model on the 50000 test images: 75.5%
Max accuracy: 75.91%
Test:  [0/5]  eta: 0:00:37  loss: 1.2919 (1.2919)  acc1: 68.3594 (68.3594)  acc5: 87.6302 (87.6302)  time: 7.4454  data: 7.2095  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3401 (1.3457)  acc1: 66.1458 (63.5840)  acc5: 87.6302 (85.3120)  time: 1.6344  data: 1.4420  max mem: 55573
Test: Total time: 0:00:08 (1.6539 s / it)
* Acc@1 63.496 Acc@5 86.048 loss 1.391
Accuracy of the model EMA on 50000 test images: 63.5%
Max EMA accuracy: 63.50%
Epoch: [171]  [  0/156]  eta: 0:33:38  lr: 0.003508  min_lr: 0.003508  loss: 2.9451 (2.9451)  weight_decay: 0.0500 (0.0500)  time: 12.9418  data: 9.2864  max mem: 55573
Epoch: [171]  [ 10/156]  eta: 0:04:16  lr: 0.003505  min_lr: 0.003505  loss: 3.3314 (3.2869)  weight_decay: 0.0500 (0.0500)  time: 1.7598  data: 0.8447  max mem: 55573
Epoch: [171]  [ 20/156]  eta: 0:02:43  lr: 0.003502  min_lr: 0.003502  loss: 3.3698 (3.3240)  weight_decay: 0.0500 (0.0500)  time: 0.6171  data: 0.0005  max mem: 55573
Epoch: [171]  [ 30/156]  eta: 0:02:06  lr: 0.003500  min_lr: 0.003500  loss: 3.5589 (3.3892)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0005  max mem: 55573
Epoch: [171]  [ 40/156]  eta: 0:01:45  lr: 0.003497  min_lr: 0.003497  loss: 3.3238 (3.3464)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0005  max mem: 55573
Epoch: [171]  [ 50/156]  eta: 0:01:29  lr: 0.003494  min_lr: 0.003494  loss: 3.3238 (3.3673)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [171]  [ 60/156]  eta: 0:01:17  lr: 0.003491  min_lr: 0.003491  loss: 3.4305 (3.3387)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [171]  [ 70/156]  eta: 0:01:06  lr: 0.003488  min_lr: 0.003488  loss: 3.3949 (3.3390)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [171]  [ 80/156]  eta: 0:00:56  lr: 0.003485  min_lr: 0.003485  loss: 3.2025 (3.3116)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [171]  [ 90/156]  eta: 0:00:48  lr: 0.003482  min_lr: 0.003482  loss: 3.2025 (3.2985)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [171]  [100/156]  eta: 0:00:40  lr: 0.003480  min_lr: 0.003480  loss: 3.2891 (3.2947)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [171]  [110/156]  eta: 0:00:32  lr: 0.003477  min_lr: 0.003477  loss: 3.4487 (3.3126)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [171]  [120/156]  eta: 0:00:25  lr: 0.003474  min_lr: 0.003474  loss: 3.5013 (3.3207)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [171]  [130/156]  eta: 0:00:17  lr: 0.003471  min_lr: 0.003471  loss: 3.5013 (3.3330)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0011  max mem: 55573
Epoch: [171]  [140/156]  eta: 0:00:10  lr: 0.003468  min_lr: 0.003468  loss: 3.3254 (3.3154)  weight_decay: 0.0500 (0.0500)  time: 0.5865  data: 0.0009  max mem: 55573
Epoch: [171]  [150/156]  eta: 0:00:04  lr: 0.003465  min_lr: 0.003465  loss: 3.4172 (3.3339)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [171]  [155/156]  eta: 0:00:00  lr: 0.003464  min_lr: 0.003464  loss: 3.4172 (3.3265)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [171] Total time: 0:01:45 (0.6787 s / it)
Averaged stats: lr: 0.003464  min_lr: 0.003464  loss: 3.4172 (3.2968)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.1004 (1.1004)  acc1: 77.3438 (77.3438)  acc5: 94.2708 (94.2708)  time: 6.9444  data: 6.7071  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1641 (1.1487)  acc1: 75.3906 (74.2080)  acc5: 94.2708 (92.7360)  time: 1.5354  data: 1.3415  max mem: 55573
Test: Total time: 0:00:07 (1.5781 s / it)
* Acc@1 75.808 Acc@5 93.198 loss 1.127
Accuracy of the model on the 50000 test images: 75.8%
Max accuracy: 75.91%
Test:  [0/5]  eta: 0:00:35  loss: 1.2805 (1.2805)  acc1: 68.7500 (68.7500)  acc5: 87.8906 (87.8906)  time: 7.1946  data: 6.9585  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3287 (1.3342)  acc1: 66.2760 (63.7760)  acc5: 87.8906 (85.6320)  time: 1.5843  data: 1.3918  max mem: 55573
Test: Total time: 0:00:08 (1.6063 s / it)
* Acc@1 63.724 Acc@5 86.234 loss 1.378
Accuracy of the model EMA on 50000 test images: 63.7%
Max EMA accuracy: 63.72%
Epoch: [172]  [  0/156]  eta: 0:34:36  lr: 0.003464  min_lr: 0.003464  loss: 1.9981 (1.9981)  weight_decay: 0.0500 (0.0500)  time: 13.3121  data: 10.6206  max mem: 55573
Epoch: [172]  [ 10/156]  eta: 0:04:16  lr: 0.003461  min_lr: 0.003461  loss: 3.0789 (3.0152)  weight_decay: 0.0500 (0.0500)  time: 1.7582  data: 0.9659  max mem: 55573
Epoch: [172]  [ 20/156]  eta: 0:02:43  lr: 0.003458  min_lr: 0.003458  loss: 3.3347 (3.2198)  weight_decay: 0.0500 (0.0500)  time: 0.5963  data: 0.0004  max mem: 55573
Epoch: [172]  [ 30/156]  eta: 0:02:06  lr: 0.003455  min_lr: 0.003455  loss: 3.4907 (3.3058)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [172]  [ 40/156]  eta: 0:01:44  lr: 0.003452  min_lr: 0.003452  loss: 3.5233 (3.3276)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0004  max mem: 55573
Epoch: [172]  [ 50/156]  eta: 0:01:29  lr: 0.003449  min_lr: 0.003449  loss: 3.4476 (3.3417)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [172]  [ 60/156]  eta: 0:01:16  lr: 0.003447  min_lr: 0.003447  loss: 3.3819 (3.2983)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [172]  [ 70/156]  eta: 0:01:06  lr: 0.003444  min_lr: 0.003444  loss: 3.3731 (3.3269)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [172]  [ 80/156]  eta: 0:00:56  lr: 0.003441  min_lr: 0.003441  loss: 3.4036 (3.3169)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [172]  [ 90/156]  eta: 0:00:48  lr: 0.003438  min_lr: 0.003438  loss: 3.4426 (3.3204)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [172]  [100/156]  eta: 0:00:40  lr: 0.003435  min_lr: 0.003435  loss: 3.4426 (3.3029)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [172]  [110/156]  eta: 0:00:32  lr: 0.003432  min_lr: 0.003432  loss: 3.2991 (3.3034)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [172]  [120/156]  eta: 0:00:25  lr: 0.003429  min_lr: 0.003429  loss: 3.4305 (3.3150)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [172]  [130/156]  eta: 0:00:17  lr: 0.003427  min_lr: 0.003427  loss: 3.5930 (3.3192)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0011  max mem: 55573
Epoch: [172]  [140/156]  eta: 0:00:10  lr: 0.003424  min_lr: 0.003424  loss: 3.4827 (3.3188)  weight_decay: 0.0500 (0.0500)  time: 0.5854  data: 0.0009  max mem: 55573
Epoch: [172]  [150/156]  eta: 0:00:04  lr: 0.003421  min_lr: 0.003421  loss: 3.2962 (3.3084)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [172]  [155/156]  eta: 0:00:00  lr: 0.003419  min_lr: 0.003419  loss: 3.2962 (3.3137)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [172] Total time: 0:01:45 (0.6783 s / it)
Averaged stats: lr: 0.003419  min_lr: 0.003419  loss: 3.2962 (3.2792)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.0943 (1.0943)  acc1: 79.1667 (79.1667)  acc5: 95.8333 (95.8333)  time: 6.6853  data: 6.4481  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2018 (1.2069)  acc1: 77.4740 (75.5200)  acc5: 95.4427 (93.5040)  time: 1.4836  data: 1.2897  max mem: 55573
Test: Total time: 0:00:07 (1.5302 s / it)
* Acc@1 75.894 Acc@5 93.172 loss 1.208
Accuracy of the model on the 50000 test images: 75.9%
Max accuracy: 75.91%
Test:  [0/5]  eta: 0:00:35  loss: 1.2695 (1.2695)  acc1: 69.2708 (69.2708)  acc5: 88.4115 (88.4115)  time: 7.0569  data: 6.8209  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3176 (1.3228)  acc1: 66.5365 (64.0640)  acc5: 88.4115 (85.7920)  time: 1.5567  data: 1.3643  max mem: 55573
Test: Total time: 0:00:07 (1.5717 s / it)
* Acc@1 64.024 Acc@5 86.426 loss 1.366
Accuracy of the model EMA on 50000 test images: 64.0%
Max EMA accuracy: 64.02%
Epoch: [173]  [  0/156]  eta: 0:33:39  lr: 0.003419  min_lr: 0.003419  loss: 2.9573 (2.9573)  weight_decay: 0.0500 (0.0500)  time: 12.9480  data: 11.8584  max mem: 55573
Epoch: [173]  [ 10/156]  eta: 0:04:10  lr: 0.003416  min_lr: 0.003416  loss: 3.3653 (3.3091)  weight_decay: 0.0500 (0.0500)  time: 1.7191  data: 1.0784  max mem: 55573
Epoch: [173]  [ 20/156]  eta: 0:02:40  lr: 0.003414  min_lr: 0.003414  loss: 3.1539 (3.2005)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0004  max mem: 55573
Epoch: [173]  [ 30/156]  eta: 0:02:04  lr: 0.003411  min_lr: 0.003411  loss: 3.1092 (3.1935)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [173]  [ 40/156]  eta: 0:01:43  lr: 0.003408  min_lr: 0.003408  loss: 3.3489 (3.2424)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [173]  [ 50/156]  eta: 0:01:28  lr: 0.003405  min_lr: 0.003405  loss: 3.4181 (3.2194)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [173]  [ 60/156]  eta: 0:01:16  lr: 0.003402  min_lr: 0.003402  loss: 3.3974 (3.2653)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [173]  [ 70/156]  eta: 0:01:05  lr: 0.003399  min_lr: 0.003399  loss: 3.3974 (3.2505)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [173]  [ 80/156]  eta: 0:00:56  lr: 0.003396  min_lr: 0.003396  loss: 3.2467 (3.2520)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [173]  [ 90/156]  eta: 0:00:47  lr: 0.003394  min_lr: 0.003394  loss: 3.4566 (3.2544)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [173]  [100/156]  eta: 0:00:39  lr: 0.003391  min_lr: 0.003391  loss: 3.4908 (3.2806)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [173]  [110/156]  eta: 0:00:32  lr: 0.003388  min_lr: 0.003388  loss: 3.5601 (3.2988)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [173]  [120/156]  eta: 0:00:24  lr: 0.003385  min_lr: 0.003385  loss: 3.4639 (3.2993)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [173]  [130/156]  eta: 0:00:17  lr: 0.003382  min_lr: 0.003382  loss: 3.4639 (3.3122)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0010  max mem: 55573
Epoch: [173]  [140/156]  eta: 0:00:10  lr: 0.003379  min_lr: 0.003379  loss: 3.5360 (3.3220)  weight_decay: 0.0500 (0.0500)  time: 0.5863  data: 0.0008  max mem: 55573
Epoch: [173]  [150/156]  eta: 0:00:04  lr: 0.003377  min_lr: 0.003377  loss: 3.5317 (3.3169)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [173]  [155/156]  eta: 0:00:00  lr: 0.003375  min_lr: 0.003375  loss: 3.4867 (3.3121)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [173] Total time: 0:01:45 (0.6754 s / it)
Averaged stats: lr: 0.003375  min_lr: 0.003375  loss: 3.4867 (3.2824)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.0118 (1.0118)  acc1: 80.0781 (80.0781)  acc5: 94.7917 (94.7917)  time: 7.1260  data: 6.8885  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1646 (1.1686)  acc1: 76.5625 (75.3280)  acc5: 94.3396 (92.5760)  time: 1.5717  data: 1.3778  max mem: 55573
Test: Total time: 0:00:08 (1.6105 s / it)
* Acc@1 76.220 Acc@5 93.182 loss 1.138
Accuracy of the model on the 50000 test images: 76.2%
Max accuracy: 76.22%
Test:  [0/5]  eta: 0:00:35  loss: 1.2592 (1.2592)  acc1: 69.1406 (69.1406)  acc5: 88.6719 (88.6719)  time: 7.1866  data: 6.9505  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.3073 (1.3125)  acc1: 66.6667 (64.2880)  acc5: 88.6719 (85.9840)  time: 1.5826  data: 1.3902  max mem: 55573
Test: Total time: 0:00:07 (1.6000 s / it)
* Acc@1 64.290 Acc@5 86.546 loss 1.355
Accuracy of the model EMA on 50000 test images: 64.3%
Max EMA accuracy: 64.29%
Epoch: [174]  [  0/156]  eta: 0:36:34  lr: 0.003375  min_lr: 0.003375  loss: 3.1354 (3.1354)  weight_decay: 0.0500 (0.0500)  time: 14.0689  data: 13.4762  max mem: 55573
Epoch: [174]  [ 10/156]  eta: 0:04:26  lr: 0.003372  min_lr: 0.003372  loss: 3.4660 (3.5060)  weight_decay: 0.0500 (0.0500)  time: 1.8262  data: 1.2254  max mem: 55573
Epoch: [174]  [ 20/156]  eta: 0:02:48  lr: 0.003369  min_lr: 0.003369  loss: 3.5441 (3.4649)  weight_decay: 0.0500 (0.0500)  time: 0.5962  data: 0.0003  max mem: 55573
Epoch: [174]  [ 30/156]  eta: 0:02:09  lr: 0.003366  min_lr: 0.003366  loss: 3.3503 (3.3885)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [174]  [ 40/156]  eta: 0:01:47  lr: 0.003363  min_lr: 0.003363  loss: 3.1823 (3.3401)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [174]  [ 50/156]  eta: 0:01:30  lr: 0.003361  min_lr: 0.003361  loss: 3.4232 (3.3539)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [174]  [ 60/156]  eta: 0:01:18  lr: 0.003358  min_lr: 0.003358  loss: 3.4349 (3.3592)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [174]  [ 70/156]  eta: 0:01:07  lr: 0.003355  min_lr: 0.003355  loss: 3.1996 (3.3145)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [174]  [ 80/156]  eta: 0:00:57  lr: 0.003352  min_lr: 0.003352  loss: 3.0489 (3.2838)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [174]  [ 90/156]  eta: 0:00:48  lr: 0.003349  min_lr: 0.003349  loss: 3.0246 (3.2564)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [174]  [100/156]  eta: 0:00:40  lr: 0.003346  min_lr: 0.003346  loss: 3.0199 (3.2347)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0004  max mem: 55573
Epoch: [174]  [110/156]  eta: 0:00:32  lr: 0.003344  min_lr: 0.003344  loss: 3.3272 (3.2579)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [174]  [120/156]  eta: 0:00:25  lr: 0.003341  min_lr: 0.003341  loss: 3.5025 (3.2795)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [174]  [130/156]  eta: 0:00:18  lr: 0.003338  min_lr: 0.003338  loss: 3.4988 (3.2813)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0012  max mem: 55573
Epoch: [174]  [140/156]  eta: 0:00:10  lr: 0.003335  min_lr: 0.003335  loss: 3.2764 (3.2741)  weight_decay: 0.0500 (0.0500)  time: 0.5859  data: 0.0011  max mem: 55573
Epoch: [174]  [150/156]  eta: 0:00:04  lr: 0.003332  min_lr: 0.003332  loss: 3.4330 (3.2713)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0002  max mem: 55573
Epoch: [174]  [155/156]  eta: 0:00:00  lr: 0.003331  min_lr: 0.003331  loss: 3.4330 (3.2715)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [174] Total time: 0:01:46 (0.6829 s / it)
Averaged stats: lr: 0.003331  min_lr: 0.003331  loss: 3.4330 (3.2832)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9766 (0.9766)  acc1: 78.6458 (78.6458)  acc5: 95.0521 (95.0521)  time: 6.8768  data: 6.6381  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0571 (1.1047)  acc1: 77.4740 (74.5280)  acc5: 94.6615 (92.7040)  time: 1.5221  data: 1.3277  max mem: 55573
Test: Total time: 0:00:07 (1.5687 s / it)
* Acc@1 75.816 Acc@5 93.344 loss 1.060
Accuracy of the model on the 50000 test images: 75.8%
Max accuracy: 76.22%
Test:  [0/5]  eta: 0:00:36  loss: 1.2497 (1.2497)  acc1: 69.1406 (69.1406)  acc5: 88.9323 (88.9323)  time: 7.2079  data: 6.9719  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2977 (1.3027)  acc1: 66.9271 (64.4160)  acc5: 88.9323 (86.1120)  time: 1.5870  data: 1.3945  max mem: 55573
Test: Total time: 0:00:08 (1.6059 s / it)
* Acc@1 64.544 Acc@5 86.674 loss 1.345
Accuracy of the model EMA on 50000 test images: 64.5%
Max EMA accuracy: 64.54%
Epoch: [175]  [  0/156]  eta: 0:34:36  lr: 0.003331  min_lr: 0.003331  loss: 3.1027 (3.1027)  weight_decay: 0.0500 (0.0500)  time: 13.3134  data: 7.5627  max mem: 55573
Epoch: [175]  [ 10/156]  eta: 0:04:15  lr: 0.003328  min_lr: 0.003328  loss: 3.1883 (3.1366)  weight_decay: 0.0500 (0.0500)  time: 1.7513  data: 0.6880  max mem: 55573
Epoch: [175]  [ 20/156]  eta: 0:02:42  lr: 0.003325  min_lr: 0.003325  loss: 3.3915 (3.2378)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0005  max mem: 55573
Epoch: [175]  [ 30/156]  eta: 0:02:06  lr: 0.003322  min_lr: 0.003322  loss: 3.5003 (3.2963)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [175]  [ 40/156]  eta: 0:01:44  lr: 0.003319  min_lr: 0.003319  loss: 3.3575 (3.2458)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [175]  [ 50/156]  eta: 0:01:29  lr: 0.003316  min_lr: 0.003316  loss: 3.1779 (3.2560)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [175]  [ 60/156]  eta: 0:01:16  lr: 0.003314  min_lr: 0.003314  loss: 3.5103 (3.2500)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [175]  [ 70/156]  eta: 0:01:06  lr: 0.003311  min_lr: 0.003311  loss: 3.1300 (3.2308)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [175]  [ 80/156]  eta: 0:00:56  lr: 0.003308  min_lr: 0.003308  loss: 3.0766 (3.2538)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [175]  [ 90/156]  eta: 0:00:48  lr: 0.003305  min_lr: 0.003305  loss: 3.5213 (3.2723)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [175]  [100/156]  eta: 0:00:40  lr: 0.003302  min_lr: 0.003302  loss: 3.4727 (3.2774)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [175]  [110/156]  eta: 0:00:32  lr: 0.003299  min_lr: 0.003299  loss: 3.4463 (3.2674)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [175]  [120/156]  eta: 0:00:25  lr: 0.003297  min_lr: 0.003297  loss: 3.4256 (3.2642)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [175]  [130/156]  eta: 0:00:17  lr: 0.003294  min_lr: 0.003294  loss: 3.3418 (3.2566)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0010  max mem: 55573
Epoch: [175]  [140/156]  eta: 0:00:10  lr: 0.003291  min_lr: 0.003291  loss: 3.2226 (3.2663)  weight_decay: 0.0500 (0.0500)  time: 0.5866  data: 0.0009  max mem: 55573
Epoch: [175]  [150/156]  eta: 0:00:04  lr: 0.003288  min_lr: 0.003288  loss: 3.3414 (3.2593)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [175]  [155/156]  eta: 0:00:00  lr: 0.003287  min_lr: 0.003287  loss: 3.4696 (3.2616)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [175] Total time: 0:01:45 (0.6774 s / it)
Averaged stats: lr: 0.003287  min_lr: 0.003287  loss: 3.4696 (3.2734)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.0584 (1.0584)  acc1: 80.4688 (80.4688)  acc5: 94.7917 (94.7917)  time: 6.7403  data: 6.5015  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1665 (1.1906)  acc1: 76.6927 (75.3280)  acc5: 92.4528 (93.3760)  time: 1.4946  data: 1.3004  max mem: 55573
Test: Total time: 0:00:07 (1.5343 s / it)
* Acc@1 76.074 Acc@5 93.350 loss 1.179
Accuracy of the model on the 50000 test images: 76.1%
Max accuracy: 76.22%
Test:  [0/5]  eta: 0:00:33  loss: 1.2408 (1.2408)  acc1: 69.7917 (69.7917)  acc5: 88.9323 (88.9323)  time: 6.7549  data: 6.5188  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2889 (1.2935)  acc1: 67.0573 (64.7360)  acc5: 88.9323 (86.3040)  time: 1.5034  data: 1.3110  max mem: 55573
Test: Total time: 0:00:07 (1.5244 s / it)
* Acc@1 64.796 Acc@5 86.796 loss 1.335
Accuracy of the model EMA on 50000 test images: 64.8%
Max EMA accuracy: 64.80%
Epoch: [176]  [  0/156]  eta: 0:30:50  lr: 0.003286  min_lr: 0.003286  loss: 2.2741 (2.2741)  weight_decay: 0.0500 (0.0500)  time: 11.8596  data: 11.2657  max mem: 55573
Epoch: [176]  [ 10/156]  eta: 0:04:17  lr: 0.003284  min_lr: 0.003284  loss: 2.8888 (2.9603)  weight_decay: 0.0500 (0.0500)  time: 1.7665  data: 1.0245  max mem: 55573
Epoch: [176]  [ 20/156]  eta: 0:02:44  lr: 0.003281  min_lr: 0.003281  loss: 3.2091 (3.0977)  weight_decay: 0.0500 (0.0500)  time: 0.6735  data: 0.0004  max mem: 55573
Epoch: [176]  [ 30/156]  eta: 0:02:06  lr: 0.003278  min_lr: 0.003278  loss: 3.3040 (3.1744)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [176]  [ 40/156]  eta: 0:01:45  lr: 0.003275  min_lr: 0.003275  loss: 3.5474 (3.2061)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [176]  [ 50/156]  eta: 0:01:29  lr: 0.003272  min_lr: 0.003272  loss: 3.3302 (3.2094)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [176]  [ 60/156]  eta: 0:01:17  lr: 0.003269  min_lr: 0.003269  loss: 3.2380 (3.1904)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [176]  [ 70/156]  eta: 0:01:06  lr: 0.003267  min_lr: 0.003267  loss: 3.4476 (3.2058)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [176]  [ 80/156]  eta: 0:00:57  lr: 0.003264  min_lr: 0.003264  loss: 3.4541 (3.2250)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [176]  [ 90/156]  eta: 0:00:48  lr: 0.003261  min_lr: 0.003261  loss: 3.5296 (3.2374)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [176]  [100/156]  eta: 0:00:40  lr: 0.003258  min_lr: 0.003258  loss: 3.5583 (3.2590)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [176]  [110/156]  eta: 0:00:32  lr: 0.003255  min_lr: 0.003255  loss: 3.4813 (3.2486)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [176]  [120/156]  eta: 0:00:25  lr: 0.003252  min_lr: 0.003252  loss: 3.1747 (3.2366)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [176]  [130/156]  eta: 0:00:17  lr: 0.003250  min_lr: 0.003250  loss: 3.2247 (3.2421)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0010  max mem: 55573
Epoch: [176]  [140/156]  eta: 0:00:10  lr: 0.003247  min_lr: 0.003247  loss: 3.2247 (3.2356)  weight_decay: 0.0500 (0.0500)  time: 0.5851  data: 0.0008  max mem: 55573
Epoch: [176]  [150/156]  eta: 0:00:04  lr: 0.003244  min_lr: 0.003244  loss: 3.2452 (3.2419)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0001  max mem: 55573
Epoch: [176]  [155/156]  eta: 0:00:00  lr: 0.003243  min_lr: 0.003243  loss: 3.4042 (3.2463)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [176] Total time: 0:01:45 (0.6788 s / it)
Averaged stats: lr: 0.003243  min_lr: 0.003243  loss: 3.4042 (3.2728)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.0005 (1.0005)  acc1: 79.8177 (79.8177)  acc5: 95.4427 (95.4427)  time: 6.9840  data: 6.7461  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0924 (1.1486)  acc1: 77.0833 (74.8160)  acc5: 94.7917 (93.0880)  time: 1.5434  data: 1.3493  max mem: 55573
Test: Total time: 0:00:07 (1.5847 s / it)
* Acc@1 76.128 Acc@5 93.500 loss 1.134
Accuracy of the model on the 50000 test images: 76.1%
Max accuracy: 76.22%
Test:  [0/5]  eta: 0:00:34  loss: 1.2325 (1.2325)  acc1: 70.0521 (70.0521)  acc5: 88.9323 (88.9323)  time: 6.9733  data: 6.7371  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2799 (1.2845)  acc1: 67.3177 (64.9600)  acc5: 88.9323 (86.4000)  time: 1.5399  data: 1.3475  max mem: 55573
Test: Total time: 0:00:07 (1.5565 s / it)
* Acc@1 65.020 Acc@5 86.944 loss 1.325
Accuracy of the model EMA on 50000 test images: 65.0%
Max EMA accuracy: 65.02%
Epoch: [177]  [  0/156]  eta: 0:36:45  lr: 0.003242  min_lr: 0.003242  loss: 3.3135 (3.3135)  weight_decay: 0.0500 (0.0500)  time: 14.1375  data: 13.5464  max mem: 55573
Epoch: [177]  [ 10/156]  eta: 0:04:27  lr: 0.003239  min_lr: 0.003239  loss: 2.9435 (3.0139)  weight_decay: 0.0500 (0.0500)  time: 1.8296  data: 1.2318  max mem: 55573
Epoch: [177]  [ 20/156]  eta: 0:02:48  lr: 0.003237  min_lr: 0.003237  loss: 3.1837 (3.1866)  weight_decay: 0.0500 (0.0500)  time: 0.5946  data: 0.0004  max mem: 55573
Epoch: [177]  [ 30/156]  eta: 0:02:09  lr: 0.003234  min_lr: 0.003234  loss: 3.2648 (3.1899)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [177]  [ 40/156]  eta: 0:01:47  lr: 0.003231  min_lr: 0.003231  loss: 3.2025 (3.1833)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [177]  [ 50/156]  eta: 0:01:30  lr: 0.003228  min_lr: 0.003228  loss: 3.3416 (3.2419)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [177]  [ 60/156]  eta: 0:01:18  lr: 0.003225  min_lr: 0.003225  loss: 3.5048 (3.2502)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [177]  [ 70/156]  eta: 0:01:07  lr: 0.003222  min_lr: 0.003222  loss: 3.3541 (3.2173)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [177]  [ 80/156]  eta: 0:00:57  lr: 0.003220  min_lr: 0.003220  loss: 3.3767 (3.2290)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [177]  [ 90/156]  eta: 0:00:48  lr: 0.003217  min_lr: 0.003217  loss: 3.4290 (3.2409)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [177]  [100/156]  eta: 0:00:40  lr: 0.003214  min_lr: 0.003214  loss: 3.3581 (3.2217)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [177]  [110/156]  eta: 0:00:32  lr: 0.003211  min_lr: 0.003211  loss: 3.2712 (3.2328)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0006  max mem: 55573
Epoch: [177]  [120/156]  eta: 0:00:25  lr: 0.003208  min_lr: 0.003208  loss: 3.4100 (3.2336)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [177]  [130/156]  eta: 0:00:18  lr: 0.003206  min_lr: 0.003206  loss: 3.1775 (3.2193)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0010  max mem: 55573
Epoch: [177]  [140/156]  eta: 0:00:10  lr: 0.003203  min_lr: 0.003203  loss: 3.1551 (3.2200)  weight_decay: 0.0500 (0.0500)  time: 0.5855  data: 0.0008  max mem: 55573
Epoch: [177]  [150/156]  eta: 0:00:04  lr: 0.003200  min_lr: 0.003200  loss: 3.2598 (3.2208)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [177]  [155/156]  eta: 0:00:00  lr: 0.003199  min_lr: 0.003199  loss: 3.0845 (3.2153)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [177] Total time: 0:01:46 (0.6834 s / it)
Averaged stats: lr: 0.003199  min_lr: 0.003199  loss: 3.0845 (3.2573)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9392 (0.9392)  acc1: 80.4688 (80.4688)  acc5: 96.4844 (96.4844)  time: 6.9673  data: 6.7299  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0856 (1.0878)  acc1: 76.6927 (76.2560)  acc5: 94.3396 (93.8560)  time: 1.5398  data: 1.3461  max mem: 55573
Test: Total time: 0:00:07 (1.5866 s / it)
* Acc@1 76.670 Acc@5 93.562 loss 1.056
Accuracy of the model on the 50000 test images: 76.7%
Max accuracy: 76.67%
Test:  [0/5]  eta: 0:00:35  loss: 1.2245 (1.2245)  acc1: 70.1823 (70.1823)  acc5: 88.8021 (88.8021)  time: 7.1398  data: 6.9038  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2715 (1.2759)  acc1: 67.4479 (65.1840)  acc5: 88.8021 (86.3680)  time: 1.5732  data: 1.3809  max mem: 55573
Test: Total time: 0:00:07 (1.5909 s / it)
* Acc@1 65.230 Acc@5 87.042 loss 1.316
Accuracy of the model EMA on 50000 test images: 65.2%
Max EMA accuracy: 65.23%
Epoch: [178]  [  0/156]  eta: 0:30:43  lr: 0.003198  min_lr: 0.003198  loss: 3.8978 (3.8978)  weight_decay: 0.0500 (0.0500)  time: 11.8202  data: 11.1327  max mem: 55573
Epoch: [178]  [ 10/156]  eta: 0:04:17  lr: 0.003195  min_lr: 0.003195  loss: 3.4391 (3.3445)  weight_decay: 0.0500 (0.0500)  time: 1.7658  data: 1.1576  max mem: 55573
Epoch: [178]  [ 20/156]  eta: 0:02:44  lr: 0.003193  min_lr: 0.003193  loss: 3.3267 (3.2882)  weight_decay: 0.0500 (0.0500)  time: 0.6758  data: 0.0802  max mem: 55573
Epoch: [178]  [ 30/156]  eta: 0:02:07  lr: 0.003190  min_lr: 0.003190  loss: 3.2584 (3.2715)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [178]  [ 40/156]  eta: 0:01:45  lr: 0.003187  min_lr: 0.003187  loss: 3.3984 (3.3107)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [178]  [ 50/156]  eta: 0:01:29  lr: 0.003184  min_lr: 0.003184  loss: 3.4751 (3.3009)  weight_decay: 0.0500 (0.0500)  time: 0.5934  data: 0.0005  max mem: 55573
Epoch: [178]  [ 60/156]  eta: 0:01:17  lr: 0.003181  min_lr: 0.003181  loss: 3.4751 (3.3146)  weight_decay: 0.0500 (0.0500)  time: 0.5944  data: 0.0005  max mem: 55573
Epoch: [178]  [ 70/156]  eta: 0:01:06  lr: 0.003179  min_lr: 0.003179  loss: 3.4792 (3.3157)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [178]  [ 80/156]  eta: 0:00:57  lr: 0.003176  min_lr: 0.003176  loss: 3.4922 (3.3336)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [178]  [ 90/156]  eta: 0:00:48  lr: 0.003173  min_lr: 0.003173  loss: 3.5301 (3.3507)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [178]  [100/156]  eta: 0:00:40  lr: 0.003170  min_lr: 0.003170  loss: 3.5301 (3.3532)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [178]  [110/156]  eta: 0:00:32  lr: 0.003167  min_lr: 0.003167  loss: 3.4994 (3.3586)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [178]  [120/156]  eta: 0:00:25  lr: 0.003164  min_lr: 0.003164  loss: 3.3173 (3.3428)  weight_decay: 0.0500 (0.0500)  time: 0.5934  data: 0.0005  max mem: 55573
Epoch: [178]  [130/156]  eta: 0:00:17  lr: 0.003162  min_lr: 0.003162  loss: 3.2322 (3.3241)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0011  max mem: 55573
Epoch: [178]  [140/156]  eta: 0:00:10  lr: 0.003159  min_lr: 0.003159  loss: 3.1083 (3.3030)  weight_decay: 0.0500 (0.0500)  time: 0.5850  data: 0.0009  max mem: 55573
Epoch: [178]  [150/156]  eta: 0:00:04  lr: 0.003156  min_lr: 0.003156  loss: 3.1298 (3.3003)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [178]  [155/156]  eta: 0:00:00  lr: 0.003155  min_lr: 0.003155  loss: 3.2504 (3.3022)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [178] Total time: 0:01:45 (0.6790 s / it)
Averaged stats: lr: 0.003155  min_lr: 0.003155  loss: 3.2504 (3.2503)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.0119 (1.0119)  acc1: 79.4271 (79.4271)  acc5: 96.2240 (96.2240)  time: 6.9978  data: 6.7609  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0559 (1.0775)  acc1: 78.5156 (76.0960)  acc5: 94.3396 (94.0160)  time: 1.5459  data: 1.3522  max mem: 55573
Test: Total time: 0:00:07 (1.5887 s / it)
* Acc@1 76.692 Acc@5 93.662 loss 1.067
Accuracy of the model on the 50000 test images: 76.7%
Max accuracy: 76.69%
Test:  [0/5]  eta: 0:00:34  loss: 1.2167 (1.2167)  acc1: 70.3125 (70.3125)  acc5: 88.8021 (88.8021)  time: 6.8675  data: 6.6313  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2630 (1.2676)  acc1: 67.7083 (65.5040)  acc5: 88.8021 (86.4640)  time: 1.5189  data: 1.3264  max mem: 55573
Test: Total time: 0:00:07 (1.5341 s / it)
* Acc@1 65.470 Acc@5 87.162 loss 1.307
Accuracy of the model EMA on 50000 test images: 65.5%
Max EMA accuracy: 65.47%
Epoch: [179]  [  0/156]  eta: 0:35:14  lr: 0.003154  min_lr: 0.003154  loss: 3.7500 (3.7500)  weight_decay: 0.0500 (0.0500)  time: 13.5523  data: 9.3909  max mem: 55573
Epoch: [179]  [ 10/156]  eta: 0:04:19  lr: 0.003152  min_lr: 0.003152  loss: 3.2805 (3.2471)  weight_decay: 0.0500 (0.0500)  time: 1.7768  data: 0.8541  max mem: 55573
Epoch: [179]  [ 20/156]  eta: 0:02:44  lr: 0.003149  min_lr: 0.003149  loss: 3.1178 (3.1645)  weight_decay: 0.0500 (0.0500)  time: 0.5947  data: 0.0004  max mem: 55573
Epoch: [179]  [ 30/156]  eta: 0:02:07  lr: 0.003146  min_lr: 0.003146  loss: 3.2448 (3.1877)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [179]  [ 40/156]  eta: 0:01:45  lr: 0.003143  min_lr: 0.003143  loss: 3.3944 (3.2358)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [179]  [ 50/156]  eta: 0:01:29  lr: 0.003140  min_lr: 0.003140  loss: 3.4239 (3.2637)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [179]  [ 60/156]  eta: 0:01:17  lr: 0.003137  min_lr: 0.003137  loss: 3.3902 (3.2418)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [179]  [ 70/156]  eta: 0:01:06  lr: 0.003135  min_lr: 0.003135  loss: 3.1629 (3.2435)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [179]  [ 80/156]  eta: 0:00:57  lr: 0.003132  min_lr: 0.003132  loss: 3.2138 (3.2530)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [179]  [ 90/156]  eta: 0:00:48  lr: 0.003129  min_lr: 0.003129  loss: 3.3655 (3.2405)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [179]  [100/156]  eta: 0:00:40  lr: 0.003126  min_lr: 0.003126  loss: 3.3510 (3.2380)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [179]  [110/156]  eta: 0:00:32  lr: 0.003123  min_lr: 0.003123  loss: 3.3510 (3.2357)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [179]  [120/156]  eta: 0:00:25  lr: 0.003121  min_lr: 0.003121  loss: 3.3592 (3.2391)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [179]  [130/156]  eta: 0:00:17  lr: 0.003118  min_lr: 0.003118  loss: 3.3588 (3.2358)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0010  max mem: 55573
Epoch: [179]  [140/156]  eta: 0:00:10  lr: 0.003115  min_lr: 0.003115  loss: 3.2952 (3.2326)  weight_decay: 0.0500 (0.0500)  time: 0.5848  data: 0.0009  max mem: 55573
Epoch: [179]  [150/156]  eta: 0:00:04  lr: 0.003112  min_lr: 0.003112  loss: 3.3565 (3.2305)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [179]  [155/156]  eta: 0:00:00  lr: 0.003111  min_lr: 0.003111  loss: 3.3778 (3.2419)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [179] Total time: 0:01:46 (0.6800 s / it)
Averaged stats: lr: 0.003111  min_lr: 0.003111  loss: 3.3778 (3.2505)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.1875 (1.1875)  acc1: 79.1667 (79.1667)  acc5: 94.5312 (94.5312)  time: 6.6622  data: 6.4249  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2378 (1.2913)  acc1: 76.9531 (74.4960)  acc5: 94.5312 (92.9920)  time: 1.5158  data: 1.3230  max mem: 55573
Test: Total time: 0:00:07 (1.5331 s / it)
* Acc@1 75.910 Acc@5 93.216 loss 1.293
Accuracy of the model on the 50000 test images: 75.9%
Max accuracy: 76.69%
Test:  [0/5]  eta: 0:00:34  loss: 1.2094 (1.2094)  acc1: 70.5729 (70.5729)  acc5: 89.0625 (89.0625)  time: 6.9014  data: 6.6654  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2551 (1.2597)  acc1: 67.8385 (65.6640)  acc5: 89.0625 (86.6880)  time: 1.5256  data: 1.3332  max mem: 55573
Test: Total time: 0:00:07 (1.5468 s / it)
* Acc@1 65.654 Acc@5 87.280 loss 1.298
Accuracy of the model EMA on 50000 test images: 65.7%
Max EMA accuracy: 65.65%
Epoch: [180]  [  0/156]  eta: 0:34:33  lr: 0.003111  min_lr: 0.003111  loss: 2.7331 (2.7331)  weight_decay: 0.0500 (0.0500)  time: 13.2908  data: 12.6910  max mem: 55573
Epoch: [180]  [ 10/156]  eta: 0:04:14  lr: 0.003108  min_lr: 0.003108  loss: 3.5256 (3.5054)  weight_decay: 0.0500 (0.0500)  time: 1.7459  data: 1.1541  max mem: 55573
Epoch: [180]  [ 20/156]  eta: 0:02:42  lr: 0.003105  min_lr: 0.003105  loss: 3.4225 (3.2326)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [180]  [ 30/156]  eta: 0:02:06  lr: 0.003102  min_lr: 0.003102  loss: 3.1784 (3.2419)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [180]  [ 40/156]  eta: 0:01:44  lr: 0.003099  min_lr: 0.003099  loss: 3.2303 (3.2277)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [180]  [ 50/156]  eta: 0:01:29  lr: 0.003097  min_lr: 0.003097  loss: 3.2303 (3.2251)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [180]  [ 60/156]  eta: 0:01:16  lr: 0.003094  min_lr: 0.003094  loss: 3.0405 (3.1715)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [180]  [ 70/156]  eta: 0:01:06  lr: 0.003091  min_lr: 0.003091  loss: 3.3871 (3.2284)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [180]  [ 80/156]  eta: 0:00:56  lr: 0.003088  min_lr: 0.003088  loss: 3.5254 (3.2172)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [180]  [ 90/156]  eta: 0:00:48  lr: 0.003085  min_lr: 0.003085  loss: 3.2835 (3.2224)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [180]  [100/156]  eta: 0:00:40  lr: 0.003083  min_lr: 0.003083  loss: 3.3057 (3.2360)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0005  max mem: 55573
Epoch: [180]  [110/156]  eta: 0:00:32  lr: 0.003080  min_lr: 0.003080  loss: 3.3057 (3.2285)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [180]  [120/156]  eta: 0:00:25  lr: 0.003077  min_lr: 0.003077  loss: 3.2532 (3.2218)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [180]  [130/156]  eta: 0:00:17  lr: 0.003074  min_lr: 0.003074  loss: 3.3593 (3.2270)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0011  max mem: 55573
Epoch: [180]  [140/156]  eta: 0:00:10  lr: 0.003071  min_lr: 0.003071  loss: 3.1723 (3.2194)  weight_decay: 0.0500 (0.0500)  time: 0.5859  data: 0.0009  max mem: 55573
Epoch: [180]  [150/156]  eta: 0:00:04  lr: 0.003069  min_lr: 0.003069  loss: 3.3338 (3.2348)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [180]  [155/156]  eta: 0:00:00  lr: 0.003067  min_lr: 0.003067  loss: 3.3246 (3.2286)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [180] Total time: 0:01:45 (0.6773 s / it)
Averaged stats: lr: 0.003067  min_lr: 0.003067  loss: 3.3246 (3.2505)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.1408 (1.1408)  acc1: 80.2083 (80.2083)  acc5: 95.5729 (95.5729)  time: 6.9384  data: 6.7011  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2563 (1.2629)  acc1: 78.5156 (75.9360)  acc5: 94.5312 (92.8320)  time: 1.5346  data: 1.3410  max mem: 55573
Test: Total time: 0:00:07 (1.5759 s / it)
* Acc@1 76.466 Acc@5 93.314 loss 1.258
Accuracy of the model on the 50000 test images: 76.5%
Max accuracy: 76.69%
Test:  [0/5]  eta: 0:00:36  loss: 1.2014 (1.2014)  acc1: 70.7031 (70.7031)  acc5: 89.0625 (89.0625)  time: 7.2497  data: 7.0137  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2470 (1.2517)  acc1: 67.8385 (65.6640)  acc5: 89.0625 (86.7520)  time: 1.5955  data: 1.4028  max mem: 55573
Test: Total time: 0:00:08 (1.6208 s / it)
* Acc@1 65.812 Acc@5 87.378 loss 1.290
Accuracy of the model EMA on 50000 test images: 65.8%
Max EMA accuracy: 65.81%
Epoch: [181]  [  0/156]  eta: 0:30:43  lr: 0.003067  min_lr: 0.003067  loss: 3.8795 (3.8795)  weight_decay: 0.0500 (0.0500)  time: 11.8165  data: 8.0116  max mem: 55573
Epoch: [181]  [ 10/156]  eta: 0:04:25  lr: 0.003064  min_lr: 0.003064  loss: 3.1953 (3.0868)  weight_decay: 0.0500 (0.0500)  time: 1.8181  data: 0.9163  max mem: 55573
Epoch: [181]  [ 20/156]  eta: 0:02:47  lr: 0.003061  min_lr: 0.003061  loss: 3.2930 (3.2422)  weight_decay: 0.0500 (0.0500)  time: 0.7034  data: 0.1036  max mem: 55573
Epoch: [181]  [ 30/156]  eta: 0:02:09  lr: 0.003058  min_lr: 0.003058  loss: 3.4727 (3.3112)  weight_decay: 0.0500 (0.0500)  time: 0.5985  data: 0.0005  max mem: 55573
Epoch: [181]  [ 40/156]  eta: 0:01:47  lr: 0.003056  min_lr: 0.003056  loss: 3.3798 (3.3053)  weight_decay: 0.0500 (0.0500)  time: 0.6012  data: 0.0005  max mem: 55573
Epoch: [181]  [ 50/156]  eta: 0:01:31  lr: 0.003053  min_lr: 0.003053  loss: 3.1702 (3.2594)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0005  max mem: 55573
Epoch: [181]  [ 60/156]  eta: 0:01:18  lr: 0.003050  min_lr: 0.003050  loss: 3.2154 (3.2493)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [181]  [ 70/156]  eta: 0:01:07  lr: 0.003047  min_lr: 0.003047  loss: 3.3312 (3.2508)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [181]  [ 80/156]  eta: 0:00:57  lr: 0.003044  min_lr: 0.003044  loss: 3.3312 (3.2487)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [181]  [ 90/156]  eta: 0:00:48  lr: 0.003042  min_lr: 0.003042  loss: 3.3662 (3.2483)  weight_decay: 0.0500 (0.0500)  time: 0.5933  data: 0.0004  max mem: 55573
Epoch: [181]  [100/156]  eta: 0:00:40  lr: 0.003039  min_lr: 0.003039  loss: 3.3909 (3.2391)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0004  max mem: 55573
Epoch: [181]  [110/156]  eta: 0:00:32  lr: 0.003036  min_lr: 0.003036  loss: 3.2581 (3.2302)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [181]  [120/156]  eta: 0:00:25  lr: 0.003033  min_lr: 0.003033  loss: 3.3226 (3.2322)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [181]  [130/156]  eta: 0:00:18  lr: 0.003031  min_lr: 0.003031  loss: 3.3870 (3.2419)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0013  max mem: 55573
Epoch: [181]  [140/156]  eta: 0:00:10  lr: 0.003028  min_lr: 0.003028  loss: 3.3601 (3.2421)  weight_decay: 0.0500 (0.0500)  time: 0.5859  data: 0.0012  max mem: 55573
Epoch: [181]  [150/156]  eta: 0:00:04  lr: 0.003025  min_lr: 0.003025  loss: 3.3109 (3.2349)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [181]  [155/156]  eta: 0:00:00  lr: 0.003024  min_lr: 0.003024  loss: 3.3109 (3.2290)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [181] Total time: 0:01:46 (0.6842 s / it)
Averaged stats: lr: 0.003024  min_lr: 0.003024  loss: 3.3109 (3.2564)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9303 (0.9303)  acc1: 81.2500 (81.2500)  acc5: 95.7031 (95.7031)  time: 6.9971  data: 6.7598  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0760 (1.0633)  acc1: 77.4740 (76.5120)  acc5: 95.1823 (94.0480)  time: 1.5460  data: 1.3520  max mem: 55573
Test: Total time: 0:00:07 (1.5872 s / it)
* Acc@1 76.872 Acc@5 93.734 loss 1.070
Accuracy of the model on the 50000 test images: 76.9%
Max accuracy: 76.87%
Test:  [0/5]  eta: 0:00:35  loss: 1.1945 (1.1945)  acc1: 71.0938 (71.0938)  acc5: 89.4531 (89.4531)  time: 7.0515  data: 6.8150  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2397 (1.2444)  acc1: 68.4896 (66.1120)  acc5: 89.4531 (86.9120)  time: 1.5556  data: 1.3631  max mem: 55573
Test: Total time: 0:00:07 (1.5731 s / it)
* Acc@1 66.014 Acc@5 87.524 loss 1.282
Accuracy of the model EMA on 50000 test images: 66.0%
Max EMA accuracy: 66.01%
Epoch: [182]  [  0/156]  eta: 0:32:18  lr: 0.003023  min_lr: 0.003023  loss: 3.0108 (3.0108)  weight_decay: 0.0500 (0.0500)  time: 12.4232  data: 8.1473  max mem: 55573
Epoch: [182]  [ 10/156]  eta: 0:04:11  lr: 0.003020  min_lr: 0.003020  loss: 3.4079 (3.2103)  weight_decay: 0.0500 (0.0500)  time: 1.7235  data: 0.7411  max mem: 55573
Epoch: [182]  [ 20/156]  eta: 0:02:41  lr: 0.003018  min_lr: 0.003018  loss: 3.3198 (3.2118)  weight_decay: 0.0500 (0.0500)  time: 0.6219  data: 0.0004  max mem: 55573
Epoch: [182]  [ 30/156]  eta: 0:02:05  lr: 0.003015  min_lr: 0.003015  loss: 3.2976 (3.2019)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [182]  [ 40/156]  eta: 0:01:43  lr: 0.003012  min_lr: 0.003012  loss: 3.2976 (3.2004)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [182]  [ 50/156]  eta: 0:01:28  lr: 0.003009  min_lr: 0.003009  loss: 3.2143 (3.1786)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [182]  [ 60/156]  eta: 0:01:16  lr: 0.003007  min_lr: 0.003007  loss: 3.2522 (3.1861)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [182]  [ 70/156]  eta: 0:01:05  lr: 0.003004  min_lr: 0.003004  loss: 3.3403 (3.2079)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [182]  [ 80/156]  eta: 0:00:56  lr: 0.003001  min_lr: 0.003001  loss: 3.4297 (3.2138)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [182]  [ 90/156]  eta: 0:00:48  lr: 0.002998  min_lr: 0.002998  loss: 3.2822 (3.2052)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [182]  [100/156]  eta: 0:00:40  lr: 0.002995  min_lr: 0.002995  loss: 3.2822 (3.2084)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [182]  [110/156]  eta: 0:00:32  lr: 0.002993  min_lr: 0.002993  loss: 3.4121 (3.2113)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [182]  [120/156]  eta: 0:00:24  lr: 0.002990  min_lr: 0.002990  loss: 3.4462 (3.2314)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [182]  [130/156]  eta: 0:00:17  lr: 0.002987  min_lr: 0.002987  loss: 3.2780 (3.2272)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0011  max mem: 55573
Epoch: [182]  [140/156]  eta: 0:00:10  lr: 0.002984  min_lr: 0.002984  loss: 3.2710 (3.2267)  weight_decay: 0.0500 (0.0500)  time: 0.5855  data: 0.0010  max mem: 55573
Epoch: [182]  [150/156]  eta: 0:00:04  lr: 0.002981  min_lr: 0.002981  loss: 3.3509 (3.2337)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0002  max mem: 55573
Epoch: [182]  [155/156]  eta: 0:00:00  lr: 0.002980  min_lr: 0.002980  loss: 3.4053 (3.2448)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [182] Total time: 0:01:45 (0.6761 s / it)
Averaged stats: lr: 0.002980  min_lr: 0.002980  loss: 3.4053 (3.2321)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:32  loss: 1.0684 (1.0684)  acc1: 79.0365 (79.0365)  acc5: 95.9635 (95.9635)  time: 6.5946  data: 6.3528  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1802 (1.2200)  acc1: 78.5156 (75.6480)  acc5: 94.3396 (93.6640)  time: 1.4642  data: 1.2706  max mem: 55573
Test: Total time: 0:00:07 (1.4808 s / it)
* Acc@1 76.650 Acc@5 93.520 loss 1.218
Accuracy of the model on the 50000 test images: 76.7%
Max accuracy: 76.87%
Test:  [0/5]  eta: 0:00:36  loss: 1.1881 (1.1881)  acc1: 70.9635 (70.9635)  acc5: 89.4531 (89.4531)  time: 7.3287  data: 7.0926  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2323 (1.2376)  acc1: 68.4896 (66.1440)  acc5: 89.4531 (87.0720)  time: 1.6114  data: 1.4186  max mem: 55573
Test: Total time: 0:00:08 (1.6285 s / it)
* Acc@1 66.164 Acc@5 87.628 loss 1.274
Accuracy of the model EMA on 50000 test images: 66.2%
Max EMA accuracy: 66.16%
Epoch: [183]  [  0/156]  eta: 0:35:33  lr: 0.002980  min_lr: 0.002980  loss: 2.2130 (2.2130)  weight_decay: 0.0500 (0.0500)  time: 13.6775  data: 13.0242  max mem: 55573
Epoch: [183]  [ 10/156]  eta: 0:04:19  lr: 0.002977  min_lr: 0.002977  loss: 3.3503 (3.2007)  weight_decay: 0.0500 (0.0500)  time: 1.7790  data: 1.1844  max mem: 55573
Epoch: [183]  [ 20/156]  eta: 0:02:45  lr: 0.002974  min_lr: 0.002974  loss: 3.4577 (3.2364)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [183]  [ 30/156]  eta: 0:02:07  lr: 0.002971  min_lr: 0.002971  loss: 3.4483 (3.2228)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [183]  [ 40/156]  eta: 0:01:45  lr: 0.002969  min_lr: 0.002969  loss: 3.4067 (3.2531)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [183]  [ 50/156]  eta: 0:01:29  lr: 0.002966  min_lr: 0.002966  loss: 3.3724 (3.2576)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [183]  [ 60/156]  eta: 0:01:17  lr: 0.002963  min_lr: 0.002963  loss: 3.2861 (3.2539)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [183]  [ 70/156]  eta: 0:01:06  lr: 0.002960  min_lr: 0.002960  loss: 3.3110 (3.2809)  weight_decay: 0.0500 (0.0500)  time: 0.5945  data: 0.0004  max mem: 55573
Epoch: [183]  [ 80/156]  eta: 0:00:57  lr: 0.002958  min_lr: 0.002958  loss: 3.4234 (3.2807)  weight_decay: 0.0500 (0.0500)  time: 0.5942  data: 0.0005  max mem: 55573
Epoch: [183]  [ 90/156]  eta: 0:00:48  lr: 0.002955  min_lr: 0.002955  loss: 3.5089 (3.2881)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [183]  [100/156]  eta: 0:00:40  lr: 0.002952  min_lr: 0.002952  loss: 3.4098 (3.2751)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [183]  [110/156]  eta: 0:00:32  lr: 0.002949  min_lr: 0.002949  loss: 3.4098 (3.2882)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [183]  [120/156]  eta: 0:00:25  lr: 0.002946  min_lr: 0.002946  loss: 3.3333 (3.2783)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [183]  [130/156]  eta: 0:00:17  lr: 0.002944  min_lr: 0.002944  loss: 3.1872 (3.2764)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0010  max mem: 55573
Epoch: [183]  [140/156]  eta: 0:00:10  lr: 0.002941  min_lr: 0.002941  loss: 3.2095 (3.2748)  weight_decay: 0.0500 (0.0500)  time: 0.5847  data: 0.0009  max mem: 55573
Epoch: [183]  [150/156]  eta: 0:00:04  lr: 0.002938  min_lr: 0.002938  loss: 3.3796 (3.2853)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [183]  [155/156]  eta: 0:00:00  lr: 0.002937  min_lr: 0.002937  loss: 3.5447 (3.2882)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [183] Total time: 0:01:46 (0.6802 s / it)
Averaged stats: lr: 0.002937  min_lr: 0.002937  loss: 3.5447 (3.2490)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.0098 (1.0098)  acc1: 79.5573 (79.5573)  acc5: 95.8333 (95.8333)  time: 6.9975  data: 6.7603  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1053 (1.1461)  acc1: 78.2552 (75.4240)  acc5: 94.3396 (93.4400)  time: 1.5459  data: 1.3521  max mem: 55573
Test: Total time: 0:00:07 (1.5891 s / it)
* Acc@1 76.308 Acc@5 93.374 loss 1.130
Accuracy of the model on the 50000 test images: 76.3%
Max accuracy: 76.87%
Test:  [0/5]  eta: 0:00:35  loss: 1.1814 (1.1814)  acc1: 70.9635 (70.9635)  acc5: 89.4531 (89.4531)  time: 7.1935  data: 6.9575  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2248 (1.2305)  acc1: 68.6198 (66.2080)  acc5: 89.4531 (87.1680)  time: 1.5839  data: 1.3916  max mem: 55573
Test: Total time: 0:00:08 (1.6404 s / it)
* Acc@1 66.368 Acc@5 87.740 loss 1.267
Accuracy of the model EMA on 50000 test images: 66.4%
Max EMA accuracy: 66.37%
Epoch: [184]  [  0/156]  eta: 0:33:16  lr: 0.002936  min_lr: 0.002936  loss: 3.8038 (3.8038)  weight_decay: 0.0500 (0.0500)  time: 12.7990  data: 11.5828  max mem: 55573
Epoch: [184]  [ 10/156]  eta: 0:04:11  lr: 0.002934  min_lr: 0.002934  loss: 3.4077 (3.1375)  weight_decay: 0.0500 (0.0500)  time: 1.7193  data: 1.0534  max mem: 55573
Epoch: [184]  [ 20/156]  eta: 0:02:40  lr: 0.002931  min_lr: 0.002931  loss: 3.4077 (3.2460)  weight_decay: 0.0500 (0.0500)  time: 0.6012  data: 0.0004  max mem: 55573
Epoch: [184]  [ 30/156]  eta: 0:02:04  lr: 0.002928  min_lr: 0.002928  loss: 3.6354 (3.3181)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [184]  [ 40/156]  eta: 0:01:43  lr: 0.002925  min_lr: 0.002925  loss: 3.4021 (3.2635)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [184]  [ 50/156]  eta: 0:01:28  lr: 0.002923  min_lr: 0.002923  loss: 3.4021 (3.2677)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [184]  [ 60/156]  eta: 0:01:16  lr: 0.002920  min_lr: 0.002920  loss: 3.2744 (3.2238)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [184]  [ 70/156]  eta: 0:01:05  lr: 0.002917  min_lr: 0.002917  loss: 3.2348 (3.1896)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [184]  [ 80/156]  eta: 0:00:56  lr: 0.002914  min_lr: 0.002914  loss: 3.2022 (3.1858)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [184]  [ 90/156]  eta: 0:00:47  lr: 0.002912  min_lr: 0.002912  loss: 3.3696 (3.2028)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [184]  [100/156]  eta: 0:00:39  lr: 0.002909  min_lr: 0.002909  loss: 3.4283 (3.2156)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [184]  [110/156]  eta: 0:00:32  lr: 0.002906  min_lr: 0.002906  loss: 3.3826 (3.2118)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [184]  [120/156]  eta: 0:00:24  lr: 0.002903  min_lr: 0.002903  loss: 3.3474 (3.2079)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [184]  [130/156]  eta: 0:00:17  lr: 0.002900  min_lr: 0.002900  loss: 3.3527 (3.2122)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0011  max mem: 55573
Epoch: [184]  [140/156]  eta: 0:00:10  lr: 0.002898  min_lr: 0.002898  loss: 3.3512 (3.2037)  weight_decay: 0.0500 (0.0500)  time: 0.5860  data: 0.0009  max mem: 55573
Epoch: [184]  [150/156]  eta: 0:00:04  lr: 0.002895  min_lr: 0.002895  loss: 3.3180 (3.2163)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [184]  [155/156]  eta: 0:00:00  lr: 0.002894  min_lr: 0.002894  loss: 3.2268 (3.2198)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [184] Total time: 0:01:45 (0.6753 s / it)
Averaged stats: lr: 0.002894  min_lr: 0.002894  loss: 3.2268 (3.2357)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9881 (0.9881)  acc1: 80.7292 (80.7292)  acc5: 95.7031 (95.7031)  time: 6.9944  data: 6.7570  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0513 (1.1286)  acc1: 78.9062 (76.8000)  acc5: 94.3396 (93.7280)  time: 1.5451  data: 1.3515  max mem: 55573
Test: Total time: 0:00:07 (1.5890 s / it)
* Acc@1 76.480 Acc@5 93.368 loss 1.119
Accuracy of the model on the 50000 test images: 76.5%
Max accuracy: 76.87%
Test:  [0/5]  eta: 0:00:33  loss: 1.1752 (1.1752)  acc1: 71.0938 (71.0938)  acc5: 89.4531 (89.4531)  time: 6.7697  data: 6.5337  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2178 (1.2239)  acc1: 68.6198 (66.3360)  acc5: 89.4531 (87.2640)  time: 1.5557  data: 1.3634  max mem: 55573
Test: Total time: 0:00:07 (1.5769 s / it)
* Acc@1 66.534 Acc@5 87.846 loss 1.259
Accuracy of the model EMA on 50000 test images: 66.5%
Max EMA accuracy: 66.53%
Epoch: [185]  [  0/156]  eta: 0:31:09  lr: 0.002893  min_lr: 0.002893  loss: 3.5611 (3.5611)  weight_decay: 0.0500 (0.0500)  time: 11.9842  data: 9.9705  max mem: 55573
Epoch: [185]  [ 10/156]  eta: 0:03:57  lr: 0.002891  min_lr: 0.002891  loss: 3.5167 (3.4833)  weight_decay: 0.0500 (0.0500)  time: 1.6289  data: 0.9069  max mem: 55573
Epoch: [185]  [ 20/156]  eta: 0:02:34  lr: 0.002888  min_lr: 0.002888  loss: 3.4463 (3.3564)  weight_decay: 0.0500 (0.0500)  time: 0.5944  data: 0.0005  max mem: 55573
Epoch: [185]  [ 30/156]  eta: 0:02:01  lr: 0.002885  min_lr: 0.002885  loss: 3.1794 (3.2729)  weight_decay: 0.0500 (0.0500)  time: 0.5941  data: 0.0005  max mem: 55573
Epoch: [185]  [ 40/156]  eta: 0:01:41  lr: 0.002882  min_lr: 0.002882  loss: 3.1548 (3.2093)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [185]  [ 50/156]  eta: 0:01:26  lr: 0.002879  min_lr: 0.002879  loss: 3.1190 (3.1750)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [185]  [ 60/156]  eta: 0:01:14  lr: 0.002877  min_lr: 0.002877  loss: 3.3157 (3.1945)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [185]  [ 70/156]  eta: 0:01:04  lr: 0.002874  min_lr: 0.002874  loss: 3.3157 (3.1680)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [185]  [ 80/156]  eta: 0:00:55  lr: 0.002871  min_lr: 0.002871  loss: 2.9700 (3.1594)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [185]  [ 90/156]  eta: 0:00:47  lr: 0.002868  min_lr: 0.002868  loss: 3.0236 (3.1349)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [185]  [100/156]  eta: 0:00:39  lr: 0.002866  min_lr: 0.002866  loss: 3.1305 (3.1537)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [185]  [110/156]  eta: 0:00:31  lr: 0.002863  min_lr: 0.002863  loss: 3.4949 (3.1680)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [185]  [120/156]  eta: 0:00:24  lr: 0.002860  min_lr: 0.002860  loss: 3.3425 (3.1795)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [185]  [130/156]  eta: 0:00:17  lr: 0.002857  min_lr: 0.002857  loss: 3.3425 (3.1844)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0008  max mem: 55573
Epoch: [185]  [140/156]  eta: 0:00:10  lr: 0.002855  min_lr: 0.002855  loss: 3.3929 (3.1810)  weight_decay: 0.0500 (0.0500)  time: 0.5871  data: 0.0007  max mem: 55573
Epoch: [185]  [150/156]  eta: 0:00:03  lr: 0.002852  min_lr: 0.002852  loss: 3.3929 (3.1804)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [185]  [155/156]  eta: 0:00:00  lr: 0.002851  min_lr: 0.002851  loss: 3.3365 (3.1797)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [185] Total time: 0:01:44 (0.6701 s / it)
Averaged stats: lr: 0.002851  min_lr: 0.002851  loss: 3.3365 (3.2328)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.9899 (0.9899)  acc1: 80.9896 (80.9896)  acc5: 95.9635 (95.9635)  time: 7.0366  data: 6.7992  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1795 (1.1571)  acc1: 77.3438 (76.0640)  acc5: 95.0521 (93.6640)  time: 1.5535  data: 1.3599  max mem: 55573
Test: Total time: 0:00:08 (1.6050 s / it)
* Acc@1 76.920 Acc@5 93.532 loss 1.140
Accuracy of the model on the 50000 test images: 76.9%
Max accuracy: 76.92%
Test:  [0/5]  eta: 0:00:34  loss: 1.1688 (1.1688)  acc1: 71.0938 (71.0938)  acc5: 89.4531 (89.4531)  time: 6.8455  data: 6.6096  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2106 (1.2172)  acc1: 68.8802 (66.4960)  acc5: 89.4531 (87.3280)  time: 1.5144  data: 1.3220  max mem: 55573
Test: Total time: 0:00:07 (1.5296 s / it)
* Acc@1 66.658 Acc@5 87.962 loss 1.252
Accuracy of the model EMA on 50000 test images: 66.7%
Max EMA accuracy: 66.66%
Epoch: [186]  [  0/156]  eta: 0:35:53  lr: 0.002850  min_lr: 0.002850  loss: 3.5596 (3.5596)  weight_decay: 0.0500 (0.0500)  time: 13.8052  data: 13.2242  max mem: 55573
Epoch: [186]  [ 10/156]  eta: 0:04:21  lr: 0.002847  min_lr: 0.002847  loss: 3.4514 (3.2920)  weight_decay: 0.0500 (0.0500)  time: 1.7893  data: 1.2025  max mem: 55573
Epoch: [186]  [ 20/156]  eta: 0:02:45  lr: 0.002845  min_lr: 0.002845  loss: 3.3248 (3.2776)  weight_decay: 0.0500 (0.0500)  time: 0.5882  data: 0.0004  max mem: 55573
Epoch: [186]  [ 30/156]  eta: 0:02:07  lr: 0.002842  min_lr: 0.002842  loss: 3.3028 (3.2486)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [186]  [ 40/156]  eta: 0:01:45  lr: 0.002839  min_lr: 0.002839  loss: 3.3076 (3.2460)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [186]  [ 50/156]  eta: 0:01:29  lr: 0.002836  min_lr: 0.002836  loss: 3.3202 (3.2676)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [186]  [ 60/156]  eta: 0:01:17  lr: 0.002834  min_lr: 0.002834  loss: 3.3136 (3.2363)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [186]  [ 70/156]  eta: 0:01:06  lr: 0.002831  min_lr: 0.002831  loss: 3.3415 (3.2634)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [186]  [ 80/156]  eta: 0:00:57  lr: 0.002828  min_lr: 0.002828  loss: 3.3858 (3.2579)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [186]  [ 90/156]  eta: 0:00:48  lr: 0.002825  min_lr: 0.002825  loss: 3.2667 (3.2434)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [186]  [100/156]  eta: 0:00:40  lr: 0.002823  min_lr: 0.002823  loss: 3.3298 (3.2401)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0005  max mem: 55573
Epoch: [186]  [110/156]  eta: 0:00:32  lr: 0.002820  min_lr: 0.002820  loss: 3.2344 (3.2295)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0005  max mem: 55573
Epoch: [186]  [120/156]  eta: 0:00:25  lr: 0.002817  min_lr: 0.002817  loss: 3.2458 (3.2254)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [186]  [130/156]  eta: 0:00:17  lr: 0.002814  min_lr: 0.002814  loss: 3.3755 (3.2402)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0010  max mem: 55573
Epoch: [186]  [140/156]  eta: 0:00:10  lr: 0.002812  min_lr: 0.002812  loss: 3.5053 (3.2421)  weight_decay: 0.0500 (0.0500)  time: 0.5868  data: 0.0009  max mem: 55573
Epoch: [186]  [150/156]  eta: 0:00:04  lr: 0.002809  min_lr: 0.002809  loss: 3.3024 (3.2417)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [186]  [155/156]  eta: 0:00:00  lr: 0.002808  min_lr: 0.002808  loss: 3.3024 (3.2468)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [186] Total time: 0:01:46 (0.6803 s / it)
Averaged stats: lr: 0.002808  min_lr: 0.002808  loss: 3.3024 (3.2308)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.9392 (0.9392)  acc1: 81.7708 (81.7708)  acc5: 96.3542 (96.3542)  time: 6.7609  data: 6.5236  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0668 (1.0585)  acc1: 77.3438 (76.6720)  acc5: 95.0521 (93.9840)  time: 1.4983  data: 1.3048  max mem: 55573
Test: Total time: 0:00:07 (1.5444 s / it)
* Acc@1 76.788 Acc@5 93.778 loss 1.054
Accuracy of the model on the 50000 test images: 76.8%
Max accuracy: 76.92%
Test:  [0/5]  eta: 0:00:36  loss: 1.1630 (1.1630)  acc1: 71.2240 (71.2240)  acc5: 89.7135 (89.7135)  time: 7.3076  data: 7.0715  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.2039 (1.2110)  acc1: 69.1406 (66.6880)  acc5: 89.7135 (87.4560)  time: 1.6069  data: 1.4144  max mem: 55573
Test: Total time: 0:00:08 (1.6247 s / it)
* Acc@1 66.828 Acc@5 88.064 loss 1.245
Accuracy of the model EMA on 50000 test images: 66.8%
Max EMA accuracy: 66.83%
Epoch: [187]  [  0/156]  eta: 0:36:28  lr: 0.002807  min_lr: 0.002807  loss: 2.0667 (2.0667)  weight_decay: 0.0500 (0.0500)  time: 14.0274  data: 9.1959  max mem: 55573
Epoch: [187]  [ 10/156]  eta: 0:04:24  lr: 0.002805  min_lr: 0.002805  loss: 3.0681 (3.0275)  weight_decay: 0.0500 (0.0500)  time: 1.8141  data: 0.8363  max mem: 55573
Epoch: [187]  [ 20/156]  eta: 0:02:47  lr: 0.002802  min_lr: 0.002802  loss: 3.0681 (3.0709)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0003  max mem: 55573
Epoch: [187]  [ 30/156]  eta: 0:02:09  lr: 0.002799  min_lr: 0.002799  loss: 3.3357 (3.0963)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [187]  [ 40/156]  eta: 0:01:46  lr: 0.002796  min_lr: 0.002796  loss: 3.3704 (3.1222)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [187]  [ 50/156]  eta: 0:01:30  lr: 0.002794  min_lr: 0.002794  loss: 3.3704 (3.1512)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [187]  [ 60/156]  eta: 0:01:17  lr: 0.002791  min_lr: 0.002791  loss: 3.3523 (3.1463)  weight_decay: 0.0500 (0.0500)  time: 0.5931  data: 0.0004  max mem: 55573
Epoch: [187]  [ 70/156]  eta: 0:01:07  lr: 0.002788  min_lr: 0.002788  loss: 3.3704 (3.1536)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [187]  [ 80/156]  eta: 0:00:57  lr: 0.002785  min_lr: 0.002785  loss: 3.4753 (3.1657)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [187]  [ 90/156]  eta: 0:00:48  lr: 0.002783  min_lr: 0.002783  loss: 3.3368 (3.1652)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [187]  [100/156]  eta: 0:00:40  lr: 0.002780  min_lr: 0.002780  loss: 3.2172 (3.1714)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [187]  [110/156]  eta: 0:00:32  lr: 0.002777  min_lr: 0.002777  loss: 3.3266 (3.2033)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [187]  [120/156]  eta: 0:00:25  lr: 0.002774  min_lr: 0.002774  loss: 3.5355 (3.2293)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [187]  [130/156]  eta: 0:00:18  lr: 0.002772  min_lr: 0.002772  loss: 3.4696 (3.2269)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0010  max mem: 55573
Epoch: [187]  [140/156]  eta: 0:00:10  lr: 0.002769  min_lr: 0.002769  loss: 3.2161 (3.2305)  weight_decay: 0.0500 (0.0500)  time: 0.5860  data: 0.0009  max mem: 55573
Epoch: [187]  [150/156]  eta: 0:00:04  lr: 0.002766  min_lr: 0.002766  loss: 3.2161 (3.2280)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [187]  [155/156]  eta: 0:00:00  lr: 0.002765  min_lr: 0.002765  loss: 3.2098 (3.2340)  weight_decay: 0.0500 (0.0500)  time: 0.5812  data: 0.0001  max mem: 55573
Epoch: [187] Total time: 0:01:46 (0.6824 s / it)
Averaged stats: lr: 0.002765  min_lr: 0.002765  loss: 3.2098 (3.2215)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.0909 (1.0909)  acc1: 78.3854 (78.3854)  acc5: 95.5729 (95.5729)  time: 6.9602  data: 6.7229  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1016 (1.1206)  acc1: 78.3854 (75.7120)  acc5: 94.9219 (93.4080)  time: 1.5385  data: 1.3446  max mem: 55573
Test: Total time: 0:00:07 (1.5804 s / it)
* Acc@1 77.026 Acc@5 93.730 loss 1.085
Accuracy of the model on the 50000 test images: 77.0%
Max accuracy: 77.03%
Test:  [0/5]  eta: 0:00:35  loss: 1.1574 (1.1574)  acc1: 71.4844 (71.4844)  acc5: 89.7135 (89.7135)  time: 7.0831  data: 6.8470  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1974 (1.2053)  acc1: 69.4010 (66.9440)  acc5: 89.7135 (87.4880)  time: 1.5620  data: 1.3695  max mem: 55573
Test: Total time: 0:00:07 (1.5772 s / it)
* Acc@1 66.986 Acc@5 88.160 loss 1.239
Accuracy of the model EMA on 50000 test images: 67.0%
Max EMA accuracy: 66.99%
Epoch: [188]  [  0/156]  eta: 0:34:18  lr: 0.002765  min_lr: 0.002765  loss: 3.5915 (3.5915)  weight_decay: 0.0500 (0.0500)  time: 13.1977  data: 8.0941  max mem: 55573
Epoch: [188]  [ 10/156]  eta: 0:04:14  lr: 0.002762  min_lr: 0.002762  loss: 3.5683 (3.4056)  weight_decay: 0.0500 (0.0500)  time: 1.7426  data: 0.7427  max mem: 55573
Epoch: [188]  [ 20/156]  eta: 0:02:42  lr: 0.002759  min_lr: 0.002759  loss: 3.4918 (3.3817)  weight_decay: 0.0500 (0.0500)  time: 0.5936  data: 0.0040  max mem: 55573
Epoch: [188]  [ 30/156]  eta: 0:02:05  lr: 0.002756  min_lr: 0.002756  loss: 3.4228 (3.2887)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [188]  [ 40/156]  eta: 0:01:44  lr: 0.002754  min_lr: 0.002754  loss: 3.1838 (3.2655)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [188]  [ 50/156]  eta: 0:01:28  lr: 0.002751  min_lr: 0.002751  loss: 3.2439 (3.2638)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [188]  [ 60/156]  eta: 0:01:16  lr: 0.002748  min_lr: 0.002748  loss: 3.2607 (3.2262)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [188]  [ 70/156]  eta: 0:01:06  lr: 0.002745  min_lr: 0.002745  loss: 3.3995 (3.2509)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [188]  [ 80/156]  eta: 0:00:56  lr: 0.002743  min_lr: 0.002743  loss: 3.3107 (3.2302)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [188]  [ 90/156]  eta: 0:00:48  lr: 0.002740  min_lr: 0.002740  loss: 3.1691 (3.2281)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [188]  [100/156]  eta: 0:00:40  lr: 0.002737  min_lr: 0.002737  loss: 3.4045 (3.2285)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [188]  [110/156]  eta: 0:00:32  lr: 0.002735  min_lr: 0.002735  loss: 3.4793 (3.2268)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [188]  [120/156]  eta: 0:00:25  lr: 0.002732  min_lr: 0.002732  loss: 3.0874 (3.2025)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [188]  [130/156]  eta: 0:00:17  lr: 0.002729  min_lr: 0.002729  loss: 3.0874 (3.2060)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0010  max mem: 55573
Epoch: [188]  [140/156]  eta: 0:00:10  lr: 0.002726  min_lr: 0.002726  loss: 3.4273 (3.2095)  weight_decay: 0.0500 (0.0500)  time: 0.5867  data: 0.0009  max mem: 55573
Epoch: [188]  [150/156]  eta: 0:00:04  lr: 0.002724  min_lr: 0.002724  loss: 3.4721 (3.2221)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [188]  [155/156]  eta: 0:00:00  lr: 0.002722  min_lr: 0.002722  loss: 3.4721 (3.2259)  weight_decay: 0.0500 (0.0500)  time: 0.5833  data: 0.0001  max mem: 55573
Epoch: [188] Total time: 0:01:45 (0.6770 s / it)
Averaged stats: lr: 0.002722  min_lr: 0.002722  loss: 3.4721 (3.2076)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9819 (0.9819)  acc1: 81.6406 (81.6406)  acc5: 95.8333 (95.8333)  time: 6.8679  data: 6.6304  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1305 (1.1283)  acc1: 78.5156 (77.2160)  acc5: 95.4427 (93.8880)  time: 1.5201  data: 1.3262  max mem: 55573
Test: Total time: 0:00:07 (1.5664 s / it)
* Acc@1 76.980 Acc@5 93.738 loss 1.119
Accuracy of the model on the 50000 test images: 77.0%
Max accuracy: 77.03%
Test:  [0/5]  eta: 0:00:36  loss: 1.1518 (1.1518)  acc1: 71.7448 (71.7448)  acc5: 89.7135 (89.7135)  time: 7.2111  data: 6.9749  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1911 (1.1996)  acc1: 69.2708 (67.0400)  acc5: 89.7135 (87.6160)  time: 1.5881  data: 1.3953  max mem: 55573
Test: Total time: 0:00:08 (1.6109 s / it)
* Acc@1 67.126 Acc@5 88.226 loss 1.232
Accuracy of the model EMA on 50000 test images: 67.1%
Max EMA accuracy: 67.13%
Epoch: [189]  [  0/156]  eta: 0:31:30  lr: 0.002722  min_lr: 0.002722  loss: 1.9360 (1.9360)  weight_decay: 0.0500 (0.0500)  time: 12.1157  data: 11.5220  max mem: 55573
Epoch: [189]  [ 10/156]  eta: 0:04:16  lr: 0.002719  min_lr: 0.002719  loss: 3.5613 (3.2376)  weight_decay: 0.0500 (0.0500)  time: 1.7538  data: 1.0479  max mem: 55573
Epoch: [189]  [ 20/156]  eta: 0:02:43  lr: 0.002717  min_lr: 0.002717  loss: 3.4457 (3.2704)  weight_decay: 0.0500 (0.0500)  time: 0.6544  data: 0.0005  max mem: 55573
Epoch: [189]  [ 30/156]  eta: 0:02:06  lr: 0.002714  min_lr: 0.002714  loss: 3.3625 (3.3048)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [189]  [ 40/156]  eta: 0:01:44  lr: 0.002711  min_lr: 0.002711  loss: 3.1551 (3.2103)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [189]  [ 50/156]  eta: 0:01:29  lr: 0.002708  min_lr: 0.002708  loss: 2.8943 (3.1838)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [189]  [ 60/156]  eta: 0:01:16  lr: 0.002706  min_lr: 0.002706  loss: 3.2175 (3.2106)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [189]  [ 70/156]  eta: 0:01:06  lr: 0.002703  min_lr: 0.002703  loss: 3.3108 (3.1875)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [189]  [ 80/156]  eta: 0:00:56  lr: 0.002700  min_lr: 0.002700  loss: 3.1881 (3.2044)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0005  max mem: 55573
Epoch: [189]  [ 90/156]  eta: 0:00:48  lr: 0.002697  min_lr: 0.002697  loss: 3.3450 (3.2173)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [189]  [100/156]  eta: 0:00:40  lr: 0.002695  min_lr: 0.002695  loss: 3.3450 (3.2016)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [189]  [110/156]  eta: 0:00:32  lr: 0.002692  min_lr: 0.002692  loss: 3.1717 (3.1954)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [189]  [120/156]  eta: 0:00:25  lr: 0.002689  min_lr: 0.002689  loss: 3.1299 (3.1845)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [189]  [130/156]  eta: 0:00:17  lr: 0.002687  min_lr: 0.002687  loss: 3.0536 (3.1764)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0011  max mem: 55573
Epoch: [189]  [140/156]  eta: 0:00:10  lr: 0.002684  min_lr: 0.002684  loss: 2.9250 (3.1613)  weight_decay: 0.0500 (0.0500)  time: 0.5876  data: 0.0009  max mem: 55573
Epoch: [189]  [150/156]  eta: 0:00:04  lr: 0.002681  min_lr: 0.002681  loss: 3.0987 (3.1745)  weight_decay: 0.0500 (0.0500)  time: 0.5828  data: 0.0001  max mem: 55573
Epoch: [189]  [155/156]  eta: 0:00:00  lr: 0.002680  min_lr: 0.002680  loss: 3.3348 (3.1828)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0002  max mem: 55573
Epoch: [189] Total time: 0:01:45 (0.6787 s / it)
Averaged stats: lr: 0.002680  min_lr: 0.002680  loss: 3.3348 (3.2260)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.0570 (1.0570)  acc1: 81.5104 (81.5104)  acc5: 96.4844 (96.4844)  time: 6.8337  data: 6.5957  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1187 (1.1703)  acc1: 79.2969 (76.2880)  acc5: 92.7083 (93.8880)  time: 1.5134  data: 1.3192  max mem: 55573
Test: Total time: 0:00:07 (1.5543 s / it)
* Acc@1 77.166 Acc@5 93.948 loss 1.130
Accuracy of the model on the 50000 test images: 77.2%
Max accuracy: 77.17%
Test:  [0/5]  eta: 0:00:34  loss: 1.1464 (1.1464)  acc1: 71.8750 (71.8750)  acc5: 89.8438 (89.8438)  time: 6.8328  data: 6.5968  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1851 (1.1941)  acc1: 69.7917 (67.2000)  acc5: 89.8438 (87.7120)  time: 1.5121  data: 1.3195  max mem: 55573
Test: Total time: 0:00:07 (1.5264 s / it)
* Acc@1 67.294 Acc@5 88.330 loss 1.226
Accuracy of the model EMA on 50000 test images: 67.3%
Max EMA accuracy: 67.29%
Epoch: [190]  [  0/156]  eta: 0:33:36  lr: 0.002680  min_lr: 0.002680  loss: 3.1399 (3.1399)  weight_decay: 0.0500 (0.0500)  time: 12.9246  data: 9.3370  max mem: 55573
Epoch: [190]  [ 10/156]  eta: 0:04:11  lr: 0.002677  min_lr: 0.002677  loss: 3.2426 (3.2261)  weight_decay: 0.0500 (0.0500)  time: 1.7202  data: 0.8493  max mem: 55573
Epoch: [190]  [ 20/156]  eta: 0:02:40  lr: 0.002674  min_lr: 0.002674  loss: 3.2092 (3.1566)  weight_decay: 0.0500 (0.0500)  time: 0.5954  data: 0.0005  max mem: 55573
Epoch: [190]  [ 30/156]  eta: 0:02:04  lr: 0.002671  min_lr: 0.002671  loss: 3.3151 (3.2010)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [190]  [ 40/156]  eta: 0:01:43  lr: 0.002669  min_lr: 0.002669  loss: 3.3151 (3.2313)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [190]  [ 50/156]  eta: 0:01:28  lr: 0.002666  min_lr: 0.002666  loss: 3.3165 (3.2510)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [190]  [ 60/156]  eta: 0:01:16  lr: 0.002663  min_lr: 0.002663  loss: 3.4989 (3.2658)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [190]  [ 70/156]  eta: 0:01:06  lr: 0.002661  min_lr: 0.002661  loss: 3.3541 (3.2537)  weight_decay: 0.0500 (0.0500)  time: 0.5971  data: 0.0004  max mem: 55573
Epoch: [190]  [ 80/156]  eta: 0:00:56  lr: 0.002658  min_lr: 0.002658  loss: 3.3283 (3.2591)  weight_decay: 0.0500 (0.0500)  time: 0.5963  data: 0.0004  max mem: 55573
Epoch: [190]  [ 90/156]  eta: 0:00:48  lr: 0.002655  min_lr: 0.002655  loss: 3.2152 (3.2452)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [190]  [100/156]  eta: 0:00:40  lr: 0.002652  min_lr: 0.002652  loss: 3.1488 (3.2335)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [190]  [110/156]  eta: 0:00:32  lr: 0.002650  min_lr: 0.002650  loss: 3.2181 (3.2294)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0005  max mem: 55573
Epoch: [190]  [120/156]  eta: 0:00:24  lr: 0.002647  min_lr: 0.002647  loss: 3.2196 (3.2238)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [190]  [130/156]  eta: 0:00:17  lr: 0.002644  min_lr: 0.002644  loss: 3.3169 (3.2262)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0010  max mem: 55573
Epoch: [190]  [140/156]  eta: 0:00:10  lr: 0.002642  min_lr: 0.002642  loss: 3.0724 (3.2010)  weight_decay: 0.0500 (0.0500)  time: 0.5858  data: 0.0008  max mem: 55573
Epoch: [190]  [150/156]  eta: 0:00:04  lr: 0.002639  min_lr: 0.002639  loss: 3.0698 (3.2103)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [190]  [155/156]  eta: 0:00:00  lr: 0.002638  min_lr: 0.002638  loss: 3.1872 (3.2135)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [190] Total time: 0:01:45 (0.6766 s / it)
Averaged stats: lr: 0.002638  min_lr: 0.002638  loss: 3.1872 (3.2202)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9742 (0.9742)  acc1: 79.0365 (79.0365)  acc5: 96.0938 (96.0938)  time: 6.9862  data: 6.7488  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0895 (1.0707)  acc1: 76.5625 (75.4240)  acc5: 94.3396 (93.8560)  time: 1.5438  data: 1.3498  max mem: 55573
Test: Total time: 0:00:07 (1.5881 s / it)
* Acc@1 76.704 Acc@5 93.926 loss 1.045
Accuracy of the model on the 50000 test images: 76.7%
Max accuracy: 77.17%
Test:  [0/5]  eta: 0:00:35  loss: 1.1415 (1.1415)  acc1: 71.7448 (71.7448)  acc5: 90.2344 (90.2344)  time: 7.0130  data: 6.7764  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1796 (1.1891)  acc1: 70.0521 (67.3280)  acc5: 90.2344 (87.8400)  time: 1.5482  data: 1.3554  max mem: 55573
Test: Total time: 0:00:07 (1.5684 s / it)
* Acc@1 67.472 Acc@5 88.434 loss 1.221
Accuracy of the model EMA on 50000 test images: 67.5%
Max EMA accuracy: 67.47%
Epoch: [191]  [  0/156]  eta: 0:39:07  lr: 0.002637  min_lr: 0.002637  loss: 3.6013 (3.6013)  weight_decay: 0.0500 (0.0500)  time: 15.0451  data: 12.1409  max mem: 55573
Epoch: [191]  [ 10/156]  eta: 0:04:37  lr: 0.002635  min_lr: 0.002635  loss: 3.3266 (3.3452)  weight_decay: 0.0500 (0.0500)  time: 1.9023  data: 1.1040  max mem: 55573
Epoch: [191]  [ 20/156]  eta: 0:02:53  lr: 0.002632  min_lr: 0.002632  loss: 3.3266 (3.2858)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [191]  [ 30/156]  eta: 0:02:13  lr: 0.002629  min_lr: 0.002629  loss: 3.4833 (3.2951)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [191]  [ 40/156]  eta: 0:01:49  lr: 0.002626  min_lr: 0.002626  loss: 3.4182 (3.2816)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [191]  [ 50/156]  eta: 0:01:32  lr: 0.002624  min_lr: 0.002624  loss: 3.3461 (3.2842)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0005  max mem: 55573
Epoch: [191]  [ 60/156]  eta: 0:01:19  lr: 0.002621  min_lr: 0.002621  loss: 3.3426 (3.2708)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0005  max mem: 55573
Epoch: [191]  [ 70/156]  eta: 0:01:08  lr: 0.002618  min_lr: 0.002618  loss: 3.3053 (3.2586)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [191]  [ 80/156]  eta: 0:00:58  lr: 0.002616  min_lr: 0.002616  loss: 3.2522 (3.2523)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0005  max mem: 55573
Epoch: [191]  [ 90/156]  eta: 0:00:49  lr: 0.002613  min_lr: 0.002613  loss: 3.0390 (3.2256)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [191]  [100/156]  eta: 0:00:41  lr: 0.002610  min_lr: 0.002610  loss: 2.9266 (3.2170)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [191]  [110/156]  eta: 0:00:33  lr: 0.002608  min_lr: 0.002608  loss: 3.3290 (3.2178)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [191]  [120/156]  eta: 0:00:25  lr: 0.002605  min_lr: 0.002605  loss: 3.3273 (3.2095)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [191]  [130/156]  eta: 0:00:18  lr: 0.002602  min_lr: 0.002602  loss: 3.1458 (3.2040)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0011  max mem: 55573
Epoch: [191]  [140/156]  eta: 0:00:11  lr: 0.002599  min_lr: 0.002599  loss: 3.1147 (3.1950)  weight_decay: 0.0500 (0.0500)  time: 0.5851  data: 0.0009  max mem: 55573
Epoch: [191]  [150/156]  eta: 0:00:04  lr: 0.002597  min_lr: 0.002597  loss: 3.2805 (3.2008)  weight_decay: 0.0500 (0.0500)  time: 0.5829  data: 0.0001  max mem: 55573
Epoch: [191]  [155/156]  eta: 0:00:00  lr: 0.002595  min_lr: 0.002595  loss: 3.4516 (3.2145)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0001  max mem: 55573
Epoch: [191] Total time: 0:01:47 (0.6896 s / it)
Averaged stats: lr: 0.002595  min_lr: 0.002595  loss: 3.4516 (3.2164)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 1.0109 (1.0109)  acc1: 79.1667 (79.1667)  acc5: 95.7031 (95.7031)  time: 6.7377  data: 6.5008  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0695 (1.0803)  acc1: 78.5156 (76.2880)  acc5: 95.5729 (93.9840)  time: 1.4940  data: 1.3002  max mem: 55573
Test: Total time: 0:00:07 (1.5410 s / it)
* Acc@1 77.460 Acc@5 94.098 loss 1.058
Accuracy of the model on the 50000 test images: 77.5%
Max accuracy: 77.46%
Test:  [0/5]  eta: 0:00:33  loss: 1.1365 (1.1365)  acc1: 71.7448 (71.7448)  acc5: 90.2344 (90.2344)  time: 6.6712  data: 6.4352  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1739 (1.1839)  acc1: 70.1823 (67.3600)  acc5: 90.2344 (87.8400)  time: 1.4798  data: 1.2872  max mem: 55573
Test: Total time: 0:00:07 (1.5160 s / it)
* Acc@1 67.598 Acc@5 88.512 loss 1.215
Accuracy of the model EMA on 50000 test images: 67.6%
Max EMA accuracy: 67.60%
Epoch: [192]  [  0/156]  eta: 0:33:56  lr: 0.002595  min_lr: 0.002595  loss: 3.8815 (3.8815)  weight_decay: 0.0500 (0.0500)  time: 13.0549  data: 12.3962  max mem: 55573
Epoch: [192]  [ 10/156]  eta: 0:04:12  lr: 0.002592  min_lr: 0.002592  loss: 3.5053 (3.3257)  weight_decay: 0.0500 (0.0500)  time: 1.7325  data: 1.1273  max mem: 55573
Epoch: [192]  [ 20/156]  eta: 0:02:41  lr: 0.002590  min_lr: 0.002590  loss: 3.4964 (3.3681)  weight_decay: 0.0500 (0.0500)  time: 0.5946  data: 0.0004  max mem: 55573
Epoch: [192]  [ 30/156]  eta: 0:02:05  lr: 0.002587  min_lr: 0.002587  loss: 3.4487 (3.2427)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [192]  [ 40/156]  eta: 0:01:44  lr: 0.002584  min_lr: 0.002584  loss: 3.3523 (3.2497)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [192]  [ 50/156]  eta: 0:01:28  lr: 0.002582  min_lr: 0.002582  loss: 3.3523 (3.2641)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0005  max mem: 55573
Epoch: [192]  [ 60/156]  eta: 0:01:16  lr: 0.002579  min_lr: 0.002579  loss: 3.2216 (3.2604)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [192]  [ 70/156]  eta: 0:01:06  lr: 0.002576  min_lr: 0.002576  loss: 3.3593 (3.2548)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [192]  [ 80/156]  eta: 0:00:56  lr: 0.002574  min_lr: 0.002574  loss: 3.2720 (3.2417)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [192]  [ 90/156]  eta: 0:00:48  lr: 0.002571  min_lr: 0.002571  loss: 3.0868 (3.2237)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [192]  [100/156]  eta: 0:00:40  lr: 0.002568  min_lr: 0.002568  loss: 3.3129 (3.2274)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0005  max mem: 55573
Epoch: [192]  [110/156]  eta: 0:00:32  lr: 0.002566  min_lr: 0.002566  loss: 3.3813 (3.2407)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [192]  [120/156]  eta: 0:00:24  lr: 0.002563  min_lr: 0.002563  loss: 3.3388 (3.2259)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [192]  [130/156]  eta: 0:00:17  lr: 0.002560  min_lr: 0.002560  loss: 3.0472 (3.2104)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0010  max mem: 55573
Epoch: [192]  [140/156]  eta: 0:00:10  lr: 0.002558  min_lr: 0.002558  loss: 3.3000 (3.2088)  weight_decay: 0.0500 (0.0500)  time: 0.5872  data: 0.0009  max mem: 55573
Epoch: [192]  [150/156]  eta: 0:00:04  lr: 0.002555  min_lr: 0.002555  loss: 3.4346 (3.2284)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [192]  [155/156]  eta: 0:00:00  lr: 0.002554  min_lr: 0.002554  loss: 3.3399 (3.2217)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [192] Total time: 0:01:45 (0.6764 s / it)
Averaged stats: lr: 0.002554  min_lr: 0.002554  loss: 3.3399 (3.2187)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9590 (0.9590)  acc1: 80.4688 (80.4688)  acc5: 95.5729 (95.5729)  time: 6.9760  data: 6.7386  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0488 (1.0613)  acc1: 78.5156 (76.4480)  acc5: 94.6615 (93.9200)  time: 1.5417  data: 1.3478  max mem: 55573
Test: Total time: 0:00:07 (1.5853 s / it)
* Acc@1 77.262 Acc@5 93.948 loss 1.052
Accuracy of the model on the 50000 test images: 77.3%
Max accuracy: 77.46%
Test:  [0/5]  eta: 0:00:35  loss: 1.1318 (1.1318)  acc1: 71.8750 (71.8750)  acc5: 90.4948 (90.4948)  time: 7.0321  data: 6.7960  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1685 (1.1790)  acc1: 70.1823 (67.4880)  acc5: 90.4948 (88.1920)  time: 1.5518  data: 1.3593  max mem: 55573
Test: Total time: 0:00:07 (1.5724 s / it)
* Acc@1 67.768 Acc@5 88.606 loss 1.210
Accuracy of the model EMA on 50000 test images: 67.8%
Max EMA accuracy: 67.77%
Epoch: [193]  [  0/156]  eta: 0:31:22  lr: 0.002553  min_lr: 0.002553  loss: 3.7320 (3.7320)  weight_decay: 0.0500 (0.0500)  time: 12.0658  data: 7.4114  max mem: 55573
Epoch: [193]  [ 10/156]  eta: 0:04:19  lr: 0.002551  min_lr: 0.002551  loss: 3.4360 (3.4249)  weight_decay: 0.0500 (0.0500)  time: 1.7740  data: 0.8031  max mem: 55573
Epoch: [193]  [ 20/156]  eta: 0:02:44  lr: 0.002548  min_lr: 0.002548  loss: 3.4360 (3.3710)  weight_decay: 0.0500 (0.0500)  time: 0.6682  data: 0.0713  max mem: 55573
Epoch: [193]  [ 30/156]  eta: 0:02:07  lr: 0.002545  min_lr: 0.002545  loss: 3.4057 (3.3568)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [193]  [ 40/156]  eta: 0:01:45  lr: 0.002543  min_lr: 0.002543  loss: 3.3367 (3.3031)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [193]  [ 50/156]  eta: 0:01:29  lr: 0.002540  min_lr: 0.002540  loss: 3.3447 (3.3000)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [193]  [ 60/156]  eta: 0:01:17  lr: 0.002537  min_lr: 0.002537  loss: 3.3896 (3.2692)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [193]  [ 70/156]  eta: 0:01:06  lr: 0.002534  min_lr: 0.002534  loss: 3.3669 (3.2821)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [193]  [ 80/156]  eta: 0:00:57  lr: 0.002532  min_lr: 0.002532  loss: 3.4105 (3.2809)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [193]  [ 90/156]  eta: 0:00:48  lr: 0.002529  min_lr: 0.002529  loss: 3.3024 (3.2780)  weight_decay: 0.0500 (0.0500)  time: 0.5933  data: 0.0005  max mem: 55573
Epoch: [193]  [100/156]  eta: 0:00:40  lr: 0.002526  min_lr: 0.002526  loss: 3.3058 (3.2895)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0004  max mem: 55573
Epoch: [193]  [110/156]  eta: 0:00:32  lr: 0.002524  min_lr: 0.002524  loss: 3.4545 (3.2766)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [193]  [120/156]  eta: 0:00:25  lr: 0.002521  min_lr: 0.002521  loss: 2.9233 (3.2300)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [193]  [130/156]  eta: 0:00:17  lr: 0.002518  min_lr: 0.002518  loss: 2.9402 (3.2338)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0010  max mem: 55573
Epoch: [193]  [140/156]  eta: 0:00:10  lr: 0.002516  min_lr: 0.002516  loss: 3.2175 (3.2314)  weight_decay: 0.0500 (0.0500)  time: 0.5855  data: 0.0009  max mem: 55573
Epoch: [193]  [150/156]  eta: 0:00:04  lr: 0.002513  min_lr: 0.002513  loss: 3.3746 (3.2458)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [193]  [155/156]  eta: 0:00:00  lr: 0.002512  min_lr: 0.002512  loss: 3.3966 (3.2483)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [193] Total time: 0:01:46 (0.6799 s / it)
Averaged stats: lr: 0.002512  min_lr: 0.002512  loss: 3.3966 (3.2096)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.0452 (1.0452)  acc1: 80.2083 (80.2083)  acc5: 96.0938 (96.0938)  time: 6.8145  data: 6.5771  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1297 (1.1634)  acc1: 78.5156 (75.9040)  acc5: 94.3396 (94.3680)  time: 1.5093  data: 1.3155  max mem: 55573
Test: Total time: 0:00:07 (1.5525 s / it)
* Acc@1 77.184 Acc@5 93.912 loss 1.141
Accuracy of the model on the 50000 test images: 77.2%
Max accuracy: 77.46%
Test:  [0/5]  eta: 0:00:35  loss: 1.1272 (1.1272)  acc1: 72.0052 (72.0052)  acc5: 90.4948 (90.4948)  time: 7.0445  data: 6.8084  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1632 (1.1740)  acc1: 70.0521 (67.6160)  acc5: 90.4948 (88.2560)  time: 1.5542  data: 1.3618  max mem: 55573
Test: Total time: 0:00:07 (1.5748 s / it)
* Acc@1 67.868 Acc@5 88.692 loss 1.204
Accuracy of the model EMA on 50000 test images: 67.9%
Max EMA accuracy: 67.87%
Epoch: [194]  [  0/156]  eta: 0:33:01  lr: 0.002512  min_lr: 0.002512  loss: 3.1631 (3.1631)  weight_decay: 0.0500 (0.0500)  time: 12.7024  data: 11.9698  max mem: 55573
Epoch: [194]  [ 10/156]  eta: 0:04:12  lr: 0.002509  min_lr: 0.002509  loss: 3.2741 (3.1570)  weight_decay: 0.0500 (0.0500)  time: 1.7299  data: 1.0885  max mem: 55573
Epoch: [194]  [ 20/156]  eta: 0:02:41  lr: 0.002506  min_lr: 0.002506  loss: 3.2381 (3.1009)  weight_decay: 0.0500 (0.0500)  time: 0.6130  data: 0.0004  max mem: 55573
Epoch: [194]  [ 30/156]  eta: 0:02:05  lr: 0.002504  min_lr: 0.002504  loss: 3.1094 (3.0951)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0004  max mem: 55573
Epoch: [194]  [ 40/156]  eta: 0:01:44  lr: 0.002501  min_lr: 0.002501  loss: 3.3718 (3.1512)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [194]  [ 50/156]  eta: 0:01:28  lr: 0.002498  min_lr: 0.002498  loss: 3.3873 (3.1628)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [194]  [ 60/156]  eta: 0:01:16  lr: 0.002496  min_lr: 0.002496  loss: 3.2858 (3.1807)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [194]  [ 70/156]  eta: 0:01:06  lr: 0.002493  min_lr: 0.002493  loss: 3.2858 (3.1829)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [194]  [ 80/156]  eta: 0:00:56  lr: 0.002490  min_lr: 0.002490  loss: 3.4258 (3.1924)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [194]  [ 90/156]  eta: 0:00:48  lr: 0.002488  min_lr: 0.002488  loss: 3.2744 (3.1802)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [194]  [100/156]  eta: 0:00:40  lr: 0.002485  min_lr: 0.002485  loss: 3.3547 (3.2005)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [194]  [110/156]  eta: 0:00:32  lr: 0.002482  min_lr: 0.002482  loss: 3.2581 (3.1896)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [194]  [120/156]  eta: 0:00:25  lr: 0.002480  min_lr: 0.002480  loss: 3.1892 (3.1862)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [194]  [130/156]  eta: 0:00:17  lr: 0.002477  min_lr: 0.002477  loss: 3.2395 (3.1917)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0011  max mem: 55573
Epoch: [194]  [140/156]  eta: 0:00:10  lr: 0.002474  min_lr: 0.002474  loss: 3.2573 (3.1784)  weight_decay: 0.0500 (0.0500)  time: 0.5861  data: 0.0009  max mem: 55573
Epoch: [194]  [150/156]  eta: 0:00:04  lr: 0.002472  min_lr: 0.002472  loss: 2.8678 (3.1621)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [194]  [155/156]  eta: 0:00:00  lr: 0.002470  min_lr: 0.002470  loss: 2.9706 (3.1749)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [194] Total time: 0:01:45 (0.6766 s / it)
Averaged stats: lr: 0.002470  min_lr: 0.002470  loss: 2.9706 (3.1925)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9751 (0.9751)  acc1: 81.3802 (81.3802)  acc5: 95.3125 (95.3125)  time: 6.8338  data: 6.5962  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1542 (1.1502)  acc1: 75.0000 (75.3920)  acc5: 94.3396 (93.7280)  time: 1.5132  data: 1.3193  max mem: 55573
Test: Total time: 0:00:07 (1.5601 s / it)
* Acc@1 77.064 Acc@5 93.736 loss 1.116
Accuracy of the model on the 50000 test images: 77.1%
Max accuracy: 77.46%
Test:  [0/5]  eta: 0:00:34  loss: 1.1228 (1.1228)  acc1: 71.8750 (71.8750)  acc5: 90.7552 (90.7552)  time: 6.8604  data: 6.6241  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1583 (1.1695)  acc1: 70.1823 (67.6480)  acc5: 90.7552 (88.4160)  time: 1.5175  data: 1.3249  max mem: 55573
Test: Total time: 0:00:07 (1.5345 s / it)
* Acc@1 68.000 Acc@5 88.784 loss 1.199
Accuracy of the model EMA on 50000 test images: 68.0%
Max EMA accuracy: 68.00%
Epoch: [195]  [  0/156]  eta: 0:31:43  lr: 0.002470  min_lr: 0.002470  loss: 3.4368 (3.4368)  weight_decay: 0.0500 (0.0500)  time: 12.2019  data: 10.9759  max mem: 55573
Epoch: [195]  [ 10/156]  eta: 0:04:13  lr: 0.002467  min_lr: 0.002467  loss: 3.2946 (3.2547)  weight_decay: 0.0500 (0.0500)  time: 1.7367  data: 1.0902  max mem: 55573
Epoch: [195]  [ 20/156]  eta: 0:02:41  lr: 0.002465  min_lr: 0.002465  loss: 3.2979 (3.2534)  weight_decay: 0.0500 (0.0500)  time: 0.6403  data: 0.0511  max mem: 55573
Epoch: [195]  [ 30/156]  eta: 0:02:05  lr: 0.002462  min_lr: 0.002462  loss: 3.2979 (3.1595)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0005  max mem: 55573
Epoch: [195]  [ 40/156]  eta: 0:01:44  lr: 0.002459  min_lr: 0.002459  loss: 3.1676 (3.1580)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [195]  [ 50/156]  eta: 0:01:28  lr: 0.002457  min_lr: 0.002457  loss: 3.3173 (3.1854)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [195]  [ 60/156]  eta: 0:01:16  lr: 0.002454  min_lr: 0.002454  loss: 3.4038 (3.2061)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [195]  [ 70/156]  eta: 0:01:06  lr: 0.002451  min_lr: 0.002451  loss: 3.4667 (3.2372)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [195]  [ 80/156]  eta: 0:00:56  lr: 0.002449  min_lr: 0.002449  loss: 3.3488 (3.2289)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [195]  [ 90/156]  eta: 0:00:48  lr: 0.002446  min_lr: 0.002446  loss: 3.0412 (3.2064)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [195]  [100/156]  eta: 0:00:40  lr: 0.002443  min_lr: 0.002443  loss: 2.8850 (3.1887)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0005  max mem: 55573
Epoch: [195]  [110/156]  eta: 0:00:32  lr: 0.002441  min_lr: 0.002441  loss: 2.8376 (3.1585)  weight_decay: 0.0500 (0.0500)  time: 0.5946  data: 0.0005  max mem: 55573
Epoch: [195]  [120/156]  eta: 0:00:25  lr: 0.002438  min_lr: 0.002438  loss: 2.8376 (3.1663)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0005  max mem: 55573
Epoch: [195]  [130/156]  eta: 0:00:17  lr: 0.002435  min_lr: 0.002435  loss: 3.3049 (3.1690)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0011  max mem: 55573
Epoch: [195]  [140/156]  eta: 0:00:10  lr: 0.002433  min_lr: 0.002433  loss: 3.0706 (3.1571)  weight_decay: 0.0500 (0.0500)  time: 0.5873  data: 0.0009  max mem: 55573
Epoch: [195]  [150/156]  eta: 0:00:04  lr: 0.002430  min_lr: 0.002430  loss: 3.0581 (3.1466)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [195]  [155/156]  eta: 0:00:00  lr: 0.002429  min_lr: 0.002429  loss: 3.0182 (3.1430)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [195] Total time: 0:01:45 (0.6778 s / it)
Averaged stats: lr: 0.002429  min_lr: 0.002429  loss: 3.0182 (3.1971)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.0491 (1.0491)  acc1: 80.3385 (80.3385)  acc5: 95.4427 (95.4427)  time: 6.8767  data: 6.6393  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0591 (1.0708)  acc1: 78.9062 (77.1840)  acc5: 95.4427 (94.2400)  time: 1.5218  data: 1.3279  max mem: 55573
Test: Total time: 0:00:07 (1.5663 s / it)
* Acc@1 77.902 Acc@5 94.064 loss 1.069
Accuracy of the model on the 50000 test images: 77.9%
Max accuracy: 77.90%
Test:  [0/5]  eta: 0:00:33  loss: 1.1186 (1.1186)  acc1: 71.8750 (71.8750)  acc5: 90.7552 (90.7552)  time: 6.7053  data: 6.4693  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1535 (1.1648)  acc1: 70.1823 (67.7120)  acc5: 90.7552 (88.5440)  time: 1.4864  data: 1.2940  max mem: 55573
Test: Total time: 0:00:07 (1.5025 s / it)
* Acc@1 68.124 Acc@5 88.874 loss 1.194
Accuracy of the model EMA on 50000 test images: 68.1%
Max EMA accuracy: 68.12%
Epoch: [196]  [  0/156]  eta: 0:34:18  lr: 0.002429  min_lr: 0.002429  loss: 3.4004 (3.4004)  weight_decay: 0.0500 (0.0500)  time: 13.1935  data: 7.8697  max mem: 55573
Epoch: [196]  [ 10/156]  eta: 0:04:14  lr: 0.002426  min_lr: 0.002426  loss: 3.2792 (3.1829)  weight_decay: 0.0500 (0.0500)  time: 1.7418  data: 0.7164  max mem: 55573
Epoch: [196]  [ 20/156]  eta: 0:02:42  lr: 0.002423  min_lr: 0.002423  loss: 3.2792 (3.2306)  weight_decay: 0.0500 (0.0500)  time: 0.5940  data: 0.0008  max mem: 55573
Epoch: [196]  [ 30/156]  eta: 0:02:05  lr: 0.002421  min_lr: 0.002421  loss: 3.3236 (3.2092)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [196]  [ 40/156]  eta: 0:01:44  lr: 0.002418  min_lr: 0.002418  loss: 3.3154 (3.2165)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [196]  [ 50/156]  eta: 0:01:28  lr: 0.002415  min_lr: 0.002415  loss: 3.3744 (3.2319)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [196]  [ 60/156]  eta: 0:01:16  lr: 0.002413  min_lr: 0.002413  loss: 3.2714 (3.2016)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [196]  [ 70/156]  eta: 0:01:06  lr: 0.002410  min_lr: 0.002410  loss: 3.1755 (3.1955)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [196]  [ 80/156]  eta: 0:00:56  lr: 0.002407  min_lr: 0.002407  loss: 3.4343 (3.2301)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [196]  [ 90/156]  eta: 0:00:48  lr: 0.002405  min_lr: 0.002405  loss: 3.4393 (3.2346)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [196]  [100/156]  eta: 0:00:40  lr: 0.002402  min_lr: 0.002402  loss: 3.3632 (3.2146)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [196]  [110/156]  eta: 0:00:32  lr: 0.002400  min_lr: 0.002400  loss: 2.7672 (3.1717)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [196]  [120/156]  eta: 0:00:25  lr: 0.002397  min_lr: 0.002397  loss: 2.8517 (3.1718)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [196]  [130/156]  eta: 0:00:17  lr: 0.002394  min_lr: 0.002394  loss: 3.3091 (3.1677)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0024  max mem: 55573
Epoch: [196]  [140/156]  eta: 0:00:10  lr: 0.002392  min_lr: 0.002392  loss: 3.3511 (3.1792)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0022  max mem: 55573
Epoch: [196]  [150/156]  eta: 0:00:04  lr: 0.002389  min_lr: 0.002389  loss: 3.3389 (3.1857)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [196]  [155/156]  eta: 0:00:00  lr: 0.002388  min_lr: 0.002388  loss: 3.2413 (3.1877)  weight_decay: 0.0500 (0.0500)  time: 0.5812  data: 0.0001  max mem: 55573
Epoch: [196] Total time: 0:01:45 (0.6772 s / it)
Averaged stats: lr: 0.002388  min_lr: 0.002388  loss: 3.2413 (3.1814)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.0246 (1.0246)  acc1: 79.9479 (79.9479)  acc5: 96.4844 (96.4844)  time: 7.0283  data: 6.7908  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1213 (1.0939)  acc1: 78.5156 (76.4160)  acc5: 95.3125 (94.2400)  time: 1.5522  data: 1.3583  max mem: 55573
Test: Total time: 0:00:07 (1.5969 s / it)
* Acc@1 77.768 Acc@5 94.232 loss 1.083
Accuracy of the model on the 50000 test images: 77.8%
Max accuracy: 77.90%
Test:  [0/5]  eta: 0:00:36  loss: 1.1147 (1.1147)  acc1: 72.0052 (72.0052)  acc5: 90.7552 (90.7552)  time: 7.3874  data: 7.1514  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1491 (1.1606)  acc1: 70.0521 (67.6800)  acc5: 90.7552 (88.5120)  time: 1.6228  data: 1.4304  max mem: 55573
Test: Total time: 0:00:08 (1.6435 s / it)
* Acc@1 68.252 Acc@5 88.930 loss 1.189
Accuracy of the model EMA on 50000 test images: 68.3%
Max EMA accuracy: 68.25%
Epoch: [197]  [  0/156]  eta: 0:33:04  lr: 0.002387  min_lr: 0.002387  loss: 2.9719 (2.9719)  weight_decay: 0.0500 (0.0500)  time: 12.7199  data: 11.2353  max mem: 55573
Epoch: [197]  [ 10/156]  eta: 0:04:09  lr: 0.002385  min_lr: 0.002385  loss: 3.2471 (3.1098)  weight_decay: 0.0500 (0.0500)  time: 1.7113  data: 1.0217  max mem: 55573
Epoch: [197]  [ 20/156]  eta: 0:02:40  lr: 0.002382  min_lr: 0.002382  loss: 3.2734 (3.1847)  weight_decay: 0.0500 (0.0500)  time: 0.5998  data: 0.0004  max mem: 55573
Epoch: [197]  [ 30/156]  eta: 0:02:04  lr: 0.002380  min_lr: 0.002380  loss: 3.4286 (3.2111)  weight_decay: 0.0500 (0.0500)  time: 0.5887  data: 0.0004  max mem: 55573
Epoch: [197]  [ 40/156]  eta: 0:01:43  lr: 0.002377  min_lr: 0.002377  loss: 3.4139 (3.2442)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [197]  [ 50/156]  eta: 0:01:28  lr: 0.002374  min_lr: 0.002374  loss: 3.4073 (3.2395)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [197]  [ 60/156]  eta: 0:01:16  lr: 0.002372  min_lr: 0.002372  loss: 3.2567 (3.2557)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [197]  [ 70/156]  eta: 0:01:05  lr: 0.002369  min_lr: 0.002369  loss: 3.3944 (3.2669)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [197]  [ 80/156]  eta: 0:00:56  lr: 0.002366  min_lr: 0.002366  loss: 3.4803 (3.2894)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [197]  [ 90/156]  eta: 0:00:47  lr: 0.002364  min_lr: 0.002364  loss: 3.4791 (3.3069)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [197]  [100/156]  eta: 0:00:39  lr: 0.002361  min_lr: 0.002361  loss: 3.4712 (3.3115)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [197]  [110/156]  eta: 0:00:32  lr: 0.002359  min_lr: 0.002359  loss: 3.4712 (3.3022)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [197]  [120/156]  eta: 0:00:24  lr: 0.002356  min_lr: 0.002356  loss: 3.3071 (3.3000)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [197]  [130/156]  eta: 0:00:17  lr: 0.002353  min_lr: 0.002353  loss: 3.3538 (3.3043)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0010  max mem: 55573
Epoch: [197]  [140/156]  eta: 0:00:10  lr: 0.002351  min_lr: 0.002351  loss: 3.4081 (3.3140)  weight_decay: 0.0500 (0.0500)  time: 0.5863  data: 0.0009  max mem: 55573
Epoch: [197]  [150/156]  eta: 0:00:04  lr: 0.002348  min_lr: 0.002348  loss: 3.3769 (3.3015)  weight_decay: 0.0500 (0.0500)  time: 0.5828  data: 0.0001  max mem: 55573
Epoch: [197]  [155/156]  eta: 0:00:00  lr: 0.002347  min_lr: 0.002347  loss: 3.3586 (3.3020)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [197] Total time: 0:01:45 (0.6749 s / it)
Averaged stats: lr: 0.002347  min_lr: 0.002347  loss: 3.3586 (3.1811)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9462 (0.9462)  acc1: 80.5990 (80.5990)  acc5: 96.2240 (96.2240)  time: 6.8391  data: 6.6019  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0211 (1.0297)  acc1: 77.4740 (76.2880)  acc5: 95.1823 (93.7600)  time: 1.5143  data: 1.3205  max mem: 55573
Test: Total time: 0:00:07 (1.5577 s / it)
* Acc@1 77.788 Acc@5 94.078 loss 0.991
Accuracy of the model on the 50000 test images: 77.8%
Max accuracy: 77.90%
Test:  [0/5]  eta: 0:00:35  loss: 1.1108 (1.1108)  acc1: 72.1354 (72.1354)  acc5: 90.7552 (90.7552)  time: 7.1176  data: 6.8816  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1446 (1.1563)  acc1: 70.0521 (67.7760)  acc5: 90.7552 (88.5760)  time: 1.5689  data: 1.3764  max mem: 55573
Test: Total time: 0:00:07 (1.5985 s / it)
* Acc@1 68.368 Acc@5 89.012 loss 1.185
Accuracy of the model EMA on 50000 test images: 68.4%
Max EMA accuracy: 68.37%
Epoch: [198]  [  0/156]  eta: 0:35:41  lr: 0.002346  min_lr: 0.002346  loss: 3.2688 (3.2688)  weight_decay: 0.0500 (0.0500)  time: 13.7250  data: 7.8232  max mem: 55573
Epoch: [198]  [ 10/156]  eta: 0:04:20  lr: 0.002344  min_lr: 0.002344  loss: 3.1401 (3.0362)  weight_decay: 0.0500 (0.0500)  time: 1.7840  data: 0.7116  max mem: 55573
Epoch: [198]  [ 20/156]  eta: 0:02:45  lr: 0.002341  min_lr: 0.002341  loss: 3.1947 (3.1191)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [198]  [ 30/156]  eta: 0:02:07  lr: 0.002339  min_lr: 0.002339  loss: 3.3506 (3.1763)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [198]  [ 40/156]  eta: 0:01:45  lr: 0.002336  min_lr: 0.002336  loss: 3.2985 (3.1718)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [198]  [ 50/156]  eta: 0:01:29  lr: 0.002333  min_lr: 0.002333  loss: 3.3139 (3.1798)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [198]  [ 60/156]  eta: 0:01:17  lr: 0.002331  min_lr: 0.002331  loss: 3.3942 (3.1651)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [198]  [ 70/156]  eta: 0:01:06  lr: 0.002328  min_lr: 0.002328  loss: 3.2112 (3.1412)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [198]  [ 80/156]  eta: 0:00:57  lr: 0.002326  min_lr: 0.002326  loss: 3.1987 (3.1262)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [198]  [ 90/156]  eta: 0:00:48  lr: 0.002323  min_lr: 0.002323  loss: 3.2460 (3.1424)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [198]  [100/156]  eta: 0:00:40  lr: 0.002320  min_lr: 0.002320  loss: 3.2617 (3.1391)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [198]  [110/156]  eta: 0:00:32  lr: 0.002318  min_lr: 0.002318  loss: 3.2389 (3.1393)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [198]  [120/156]  eta: 0:00:25  lr: 0.002315  min_lr: 0.002315  loss: 3.1864 (3.1355)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [198]  [130/156]  eta: 0:00:17  lr: 0.002312  min_lr: 0.002312  loss: 3.2383 (3.1416)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0009  max mem: 55573
Epoch: [198]  [140/156]  eta: 0:00:10  lr: 0.002310  min_lr: 0.002310  loss: 3.2370 (3.1436)  weight_decay: 0.0500 (0.0500)  time: 0.5849  data: 0.0008  max mem: 55573
Epoch: [198]  [150/156]  eta: 0:00:04  lr: 0.002307  min_lr: 0.002307  loss: 3.1694 (3.1485)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [198]  [155/156]  eta: 0:00:00  lr: 0.002306  min_lr: 0.002306  loss: 3.3500 (3.1642)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [198] Total time: 0:01:46 (0.6804 s / it)
Averaged stats: lr: 0.002306  min_lr: 0.002306  loss: 3.3500 (3.1597)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.9912 (0.9912)  acc1: 79.4271 (79.4271)  acc5: 95.7031 (95.7031)  time: 6.6440  data: 6.4056  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0772 (1.1039)  acc1: 78.3854 (76.4480)  acc5: 95.7031 (94.0160)  time: 1.4756  data: 1.2813  max mem: 55573
Test: Total time: 0:00:07 (1.5187 s / it)
* Acc@1 77.674 Acc@5 93.964 loss 1.090
Accuracy of the model on the 50000 test images: 77.7%
Max accuracy: 77.90%
Test:  [0/5]  eta: 0:00:36  loss: 1.1075 (1.1075)  acc1: 72.3958 (72.3958)  acc5: 90.7552 (90.7552)  time: 7.3137  data: 7.0777  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1406 (1.1525)  acc1: 69.9219 (68.1280)  acc5: 90.7552 (88.7040)  time: 1.6081  data: 1.4157  max mem: 55573
Test: Total time: 0:00:08 (1.6280 s / it)
* Acc@1 68.484 Acc@5 89.088 loss 1.180
Accuracy of the model EMA on 50000 test images: 68.5%
Max EMA accuracy: 68.48%
Epoch: [199]  [  0/156]  eta: 0:34:45  lr: 0.002306  min_lr: 0.002306  loss: 3.4276 (3.4276)  weight_decay: 0.0500 (0.0500)  time: 13.3662  data: 8.4396  max mem: 55573
Epoch: [199]  [ 10/156]  eta: 0:04:17  lr: 0.002303  min_lr: 0.002303  loss: 3.4276 (3.3329)  weight_decay: 0.0500 (0.0500)  time: 1.7630  data: 0.7676  max mem: 55573
Epoch: [199]  [ 20/156]  eta: 0:02:43  lr: 0.002301  min_lr: 0.002301  loss: 3.4383 (3.3307)  weight_decay: 0.0500 (0.0500)  time: 0.5959  data: 0.0004  max mem: 55573
Epoch: [199]  [ 30/156]  eta: 0:02:06  lr: 0.002298  min_lr: 0.002298  loss: 3.4018 (3.2858)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [199]  [ 40/156]  eta: 0:01:44  lr: 0.002295  min_lr: 0.002295  loss: 3.3156 (3.3120)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [199]  [ 50/156]  eta: 0:01:29  lr: 0.002293  min_lr: 0.002293  loss: 3.3797 (3.3227)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [199]  [ 60/156]  eta: 0:01:16  lr: 0.002290  min_lr: 0.002290  loss: 3.3797 (3.3268)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [199]  [ 70/156]  eta: 0:01:06  lr: 0.002288  min_lr: 0.002288  loss: 3.3458 (3.3219)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [199]  [ 80/156]  eta: 0:00:56  lr: 0.002285  min_lr: 0.002285  loss: 3.3432 (3.3103)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [199]  [ 90/156]  eta: 0:00:48  lr: 0.002282  min_lr: 0.002282  loss: 3.3707 (3.3204)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [199]  [100/156]  eta: 0:00:40  lr: 0.002280  min_lr: 0.002280  loss: 3.4297 (3.3212)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0005  max mem: 55573
Epoch: [199]  [110/156]  eta: 0:00:32  lr: 0.002277  min_lr: 0.002277  loss: 3.4297 (3.3163)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [199]  [120/156]  eta: 0:00:25  lr: 0.002275  min_lr: 0.002275  loss: 3.4113 (3.2954)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0004  max mem: 55573
Epoch: [199]  [130/156]  eta: 0:00:17  lr: 0.002272  min_lr: 0.002272  loss: 3.2418 (3.2922)  weight_decay: 0.0500 (0.0500)  time: 0.5936  data: 0.0010  max mem: 55573
Epoch: [199]  [140/156]  eta: 0:00:10  lr: 0.002269  min_lr: 0.002269  loss: 3.1603 (3.2741)  weight_decay: 0.0500 (0.0500)  time: 0.5887  data: 0.0009  max mem: 55573
Epoch: [199]  [150/156]  eta: 0:00:04  lr: 0.002267  min_lr: 0.002267  loss: 3.0150 (3.2542)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [199]  [155/156]  eta: 0:00:00  lr: 0.002265  min_lr: 0.002265  loss: 3.0823 (3.2528)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [199] Total time: 0:01:45 (0.6788 s / it)
Averaged stats: lr: 0.002265  min_lr: 0.002265  loss: 3.0823 (3.1837)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.9412 (0.9412)  acc1: 82.0312 (82.0312)  acc5: 96.8750 (96.8750)  time: 6.7189  data: 6.4819  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0488 (1.0516)  acc1: 77.9948 (76.6080)  acc5: 95.3125 (94.1440)  time: 1.4904  data: 1.2965  max mem: 55573
Test: Total time: 0:00:07 (1.5329 s / it)
* Acc@1 77.810 Acc@5 94.264 loss 1.044
Accuracy of the model on the 50000 test images: 77.8%
Max accuracy: 77.90%
Test:  [0/5]  eta: 0:00:34  loss: 1.1036 (1.1036)  acc1: 72.3958 (72.3958)  acc5: 90.7552 (90.7552)  time: 6.9047  data: 6.6687  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1362 (1.1484)  acc1: 69.9219 (68.1600)  acc5: 90.7552 (88.7680)  time: 1.5270  data: 1.3346  max mem: 55573
Test: Total time: 0:00:07 (1.5496 s / it)
* Acc@1 68.600 Acc@5 89.140 loss 1.176
Accuracy of the model EMA on 50000 test images: 68.6%
Max EMA accuracy: 68.60%
Epoch: [200]  [  0/156]  eta: 0:35:51  lr: 0.002265  min_lr: 0.002265  loss: 2.3705 (2.3705)  weight_decay: 0.0500 (0.0500)  time: 13.7893  data: 10.5216  max mem: 55573
Epoch: [200]  [ 10/156]  eta: 0:04:23  lr: 0.002263  min_lr: 0.002263  loss: 3.0911 (2.9640)  weight_decay: 0.0500 (0.0500)  time: 1.8024  data: 0.9568  max mem: 55573
Epoch: [200]  [ 20/156]  eta: 0:02:46  lr: 0.002260  min_lr: 0.002260  loss: 3.1619 (3.0533)  weight_decay: 0.0500 (0.0500)  time: 0.5994  data: 0.0004  max mem: 55573
Epoch: [200]  [ 30/156]  eta: 0:02:08  lr: 0.002257  min_lr: 0.002257  loss: 3.2936 (3.1186)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [200]  [ 40/156]  eta: 0:01:46  lr: 0.002255  min_lr: 0.002255  loss: 3.3707 (3.1640)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [200]  [ 50/156]  eta: 0:01:30  lr: 0.002252  min_lr: 0.002252  loss: 3.3865 (3.1818)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [200]  [ 60/156]  eta: 0:01:17  lr: 0.002250  min_lr: 0.002250  loss: 3.3963 (3.1715)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [200]  [ 70/156]  eta: 0:01:06  lr: 0.002247  min_lr: 0.002247  loss: 3.2504 (3.1693)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [200]  [ 80/156]  eta: 0:00:57  lr: 0.002244  min_lr: 0.002244  loss: 3.3572 (3.1834)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [200]  [ 90/156]  eta: 0:00:48  lr: 0.002242  min_lr: 0.002242  loss: 3.3492 (3.1818)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [200]  [100/156]  eta: 0:00:40  lr: 0.002239  min_lr: 0.002239  loss: 2.8973 (3.1462)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [200]  [110/156]  eta: 0:00:32  lr: 0.002237  min_lr: 0.002237  loss: 3.2032 (3.1548)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [200]  [120/156]  eta: 0:00:25  lr: 0.002234  min_lr: 0.002234  loss: 3.2032 (3.1324)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0005  max mem: 55573
Epoch: [200]  [130/156]  eta: 0:00:18  lr: 0.002232  min_lr: 0.002232  loss: 3.1666 (3.1448)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0010  max mem: 55573
Epoch: [200]  [140/156]  eta: 0:00:10  lr: 0.002229  min_lr: 0.002229  loss: 3.3805 (3.1680)  weight_decay: 0.0500 (0.0500)  time: 0.5858  data: 0.0009  max mem: 55573
Epoch: [200]  [150/156]  eta: 0:00:04  lr: 0.002226  min_lr: 0.002226  loss: 3.3074 (3.1576)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [200]  [155/156]  eta: 0:00:00  lr: 0.002225  min_lr: 0.002225  loss: 3.2587 (3.1537)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [200] Total time: 0:01:46 (0.6822 s / it)
Averaged stats: lr: 0.002225  min_lr: 0.002225  loss: 3.2587 (3.1742)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 0.9264 (0.9264)  acc1: 83.0729 (83.0729)  acc5: 96.4844 (96.4844)  time: 7.2012  data: 6.9638  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0494 (1.0349)  acc1: 78.3854 (76.8640)  acc5: 96.4844 (94.4000)  time: 1.5868  data: 1.3928  max mem: 55573
Test: Total time: 0:00:08 (1.6308 s / it)
* Acc@1 78.102 Acc@5 94.266 loss 1.036
Accuracy of the model on the 50000 test images: 78.1%
Max accuracy: 78.10%
Test:  [0/5]  eta: 0:00:33  loss: 1.0998 (1.0998)  acc1: 72.1354 (72.1354)  acc5: 90.8854 (90.8854)  time: 6.7816  data: 6.5454  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1323 (1.1447)  acc1: 69.9219 (68.1280)  acc5: 90.8854 (88.8640)  time: 1.5016  data: 1.3092  max mem: 55573
Test: Total time: 0:00:07 (1.5187 s / it)
* Acc@1 68.728 Acc@5 89.206 loss 1.172
Accuracy of the model EMA on 50000 test images: 68.7%
Max EMA accuracy: 68.73%
Epoch: [201]  [  0/156]  eta: 0:33:15  lr: 0.002225  min_lr: 0.002225  loss: 3.0937 (3.0937)  weight_decay: 0.0500 (0.0500)  time: 12.7918  data: 8.3029  max mem: 55573
Epoch: [201]  [ 10/156]  eta: 0:04:17  lr: 0.002222  min_lr: 0.002222  loss: 3.4455 (3.1910)  weight_decay: 0.0500 (0.0500)  time: 1.7648  data: 0.8070  max mem: 55573
Epoch: [201]  [ 20/156]  eta: 0:02:43  lr: 0.002220  min_lr: 0.002220  loss: 3.3113 (3.2003)  weight_decay: 0.0500 (0.0500)  time: 0.6263  data: 0.0289  max mem: 55573
Epoch: [201]  [ 30/156]  eta: 0:02:06  lr: 0.002217  min_lr: 0.002217  loss: 3.3113 (3.2429)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [201]  [ 40/156]  eta: 0:01:45  lr: 0.002215  min_lr: 0.002215  loss: 3.3717 (3.2091)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [201]  [ 50/156]  eta: 0:01:29  lr: 0.002212  min_lr: 0.002212  loss: 3.2744 (3.2015)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0005  max mem: 55573
Epoch: [201]  [ 60/156]  eta: 0:01:17  lr: 0.002209  min_lr: 0.002209  loss: 3.1519 (3.1913)  weight_decay: 0.0500 (0.0500)  time: 0.5938  data: 0.0005  max mem: 55573
Epoch: [201]  [ 70/156]  eta: 0:01:06  lr: 0.002207  min_lr: 0.002207  loss: 3.4164 (3.2306)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0004  max mem: 55573
Epoch: [201]  [ 80/156]  eta: 0:00:57  lr: 0.002204  min_lr: 0.002204  loss: 3.5206 (3.2336)  weight_decay: 0.0500 (0.0500)  time: 0.5939  data: 0.0004  max mem: 55573
Epoch: [201]  [ 90/156]  eta: 0:00:48  lr: 0.002202  min_lr: 0.002202  loss: 3.3027 (3.2144)  weight_decay: 0.0500 (0.0500)  time: 0.5940  data: 0.0004  max mem: 55573
Epoch: [201]  [100/156]  eta: 0:00:40  lr: 0.002199  min_lr: 0.002199  loss: 3.1643 (3.2063)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0005  max mem: 55573
Epoch: [201]  [110/156]  eta: 0:00:32  lr: 0.002197  min_lr: 0.002197  loss: 3.3621 (3.2173)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0005  max mem: 55573
Epoch: [201]  [120/156]  eta: 0:00:25  lr: 0.002194  min_lr: 0.002194  loss: 3.4017 (3.2256)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0004  max mem: 55573
Epoch: [201]  [130/156]  eta: 0:00:17  lr: 0.002191  min_lr: 0.002191  loss: 3.3310 (3.2204)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0010  max mem: 55573
Epoch: [201]  [140/156]  eta: 0:00:10  lr: 0.002189  min_lr: 0.002189  loss: 3.3310 (3.2139)  weight_decay: 0.0500 (0.0500)  time: 0.5854  data: 0.0009  max mem: 55573
Epoch: [201]  [150/156]  eta: 0:00:04  lr: 0.002186  min_lr: 0.002186  loss: 3.3787 (3.2119)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [201]  [155/156]  eta: 0:00:00  lr: 0.002185  min_lr: 0.002185  loss: 3.3466 (3.2048)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [201] Total time: 0:01:46 (0.6801 s / it)
Averaged stats: lr: 0.002185  min_lr: 0.002185  loss: 3.3466 (3.1602)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.0282 (1.0282)  acc1: 80.2083 (80.2083)  acc5: 95.5729 (95.5729)  time: 7.0999  data: 6.8625  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1335 (1.1146)  acc1: 77.6042 (76.8640)  acc5: 95.5729 (94.2080)  time: 1.5667  data: 1.3726  max mem: 55573
Test: Total time: 0:00:08 (1.6139 s / it)
* Acc@1 77.604 Acc@5 94.058 loss 1.109
Accuracy of the model on the 50000 test images: 77.6%
Max accuracy: 78.10%
Test:  [0/5]  eta: 0:00:35  loss: 1.0958 (1.0958)  acc1: 72.3958 (72.3958)  acc5: 90.8854 (90.8854)  time: 7.1427  data: 6.9064  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1281 (1.1406)  acc1: 70.0521 (68.3200)  acc5: 90.8854 (88.9280)  time: 1.5738  data: 1.3814  max mem: 55573
Test: Total time: 0:00:07 (1.5953 s / it)
* Acc@1 68.836 Acc@5 89.284 loss 1.168
Accuracy of the model EMA on 50000 test images: 68.8%
Max EMA accuracy: 68.84%
Epoch: [202]  [  0/156]  eta: 0:37:03  lr: 0.002185  min_lr: 0.002185  loss: 3.0723 (3.0723)  weight_decay: 0.0500 (0.0500)  time: 14.2536  data: 13.6746  max mem: 55573
Epoch: [202]  [ 10/156]  eta: 0:04:29  lr: 0.002182  min_lr: 0.002182  loss: 3.3788 (3.3455)  weight_decay: 0.0500 (0.0500)  time: 1.8453  data: 1.2435  max mem: 55573
Epoch: [202]  [ 20/156]  eta: 0:02:49  lr: 0.002180  min_lr: 0.002180  loss: 3.2982 (3.2663)  weight_decay: 0.0500 (0.0500)  time: 0.5985  data: 0.0004  max mem: 55573
Epoch: [202]  [ 30/156]  eta: 0:02:10  lr: 0.002177  min_lr: 0.002177  loss: 3.2272 (3.1804)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [202]  [ 40/156]  eta: 0:01:47  lr: 0.002175  min_lr: 0.002175  loss: 2.6154 (3.1181)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [202]  [ 50/156]  eta: 0:01:31  lr: 0.002172  min_lr: 0.002172  loss: 3.1227 (3.1118)  weight_decay: 0.0500 (0.0500)  time: 0.5945  data: 0.0005  max mem: 55573
Epoch: [202]  [ 60/156]  eta: 0:01:18  lr: 0.002169  min_lr: 0.002169  loss: 3.2846 (3.1393)  weight_decay: 0.0500 (0.0500)  time: 0.5963  data: 0.0005  max mem: 55573
Epoch: [202]  [ 70/156]  eta: 0:01:07  lr: 0.002167  min_lr: 0.002167  loss: 3.3999 (3.1712)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [202]  [ 80/156]  eta: 0:00:57  lr: 0.002164  min_lr: 0.002164  loss: 3.4238 (3.1899)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [202]  [ 90/156]  eta: 0:00:49  lr: 0.002162  min_lr: 0.002162  loss: 3.3560 (3.1787)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [202]  [100/156]  eta: 0:00:40  lr: 0.002159  min_lr: 0.002159  loss: 2.9044 (3.1548)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [202]  [110/156]  eta: 0:00:32  lr: 0.002157  min_lr: 0.002157  loss: 2.8731 (3.1531)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [202]  [120/156]  eta: 0:00:25  lr: 0.002154  min_lr: 0.002154  loss: 3.1289 (3.1470)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [202]  [130/156]  eta: 0:00:18  lr: 0.002152  min_lr: 0.002152  loss: 3.2330 (3.1584)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0012  max mem: 55573
Epoch: [202]  [140/156]  eta: 0:00:11  lr: 0.002149  min_lr: 0.002149  loss: 3.3453 (3.1697)  weight_decay: 0.0500 (0.0500)  time: 0.5849  data: 0.0011  max mem: 55573
Epoch: [202]  [150/156]  eta: 0:00:04  lr: 0.002146  min_lr: 0.002146  loss: 3.3241 (3.1612)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [202]  [155/156]  eta: 0:00:00  lr: 0.002145  min_lr: 0.002145  loss: 3.2790 (3.1641)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [202] Total time: 0:01:46 (0.6853 s / it)
Averaged stats: lr: 0.002145  min_lr: 0.002145  loss: 3.2790 (3.1660)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9807 (0.9807)  acc1: 79.6875 (79.6875)  acc5: 96.3542 (96.3542)  time: 6.9257  data: 6.6884  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0838 (1.0768)  acc1: 76.6927 (76.6080)  acc5: 95.7031 (94.5600)  time: 1.5316  data: 1.3378  max mem: 55573
Test: Total time: 0:00:07 (1.5747 s / it)
* Acc@1 78.048 Acc@5 94.368 loss 1.054
Accuracy of the model on the 50000 test images: 78.0%
Max accuracy: 78.10%
Test:  [0/5]  eta: 0:00:36  loss: 1.0923 (1.0923)  acc1: 72.5260 (72.5260)  acc5: 90.8854 (90.8854)  time: 7.3976  data: 7.1616  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1241 (1.1368)  acc1: 70.0521 (68.4160)  acc5: 90.8854 (88.9600)  time: 1.6248  data: 1.4324  max mem: 55573
Test: Total time: 0:00:08 (1.6442 s / it)
* Acc@1 68.938 Acc@5 89.372 loss 1.163
Accuracy of the model EMA on 50000 test images: 68.9%
Max EMA accuracy: 68.94%
Epoch: [203]  [  0/156]  eta: 0:30:52  lr: 0.002145  min_lr: 0.002145  loss: 3.5272 (3.5272)  weight_decay: 0.0500 (0.0500)  time: 11.8759  data: 8.9975  max mem: 55573
Epoch: [203]  [ 10/156]  eta: 0:04:08  lr: 0.002142  min_lr: 0.002142  loss: 3.3233 (3.2705)  weight_decay: 0.0500 (0.0500)  time: 1.7038  data: 0.8836  max mem: 55573
Epoch: [203]  [ 20/156]  eta: 0:02:39  lr: 0.002140  min_lr: 0.002140  loss: 3.3233 (3.2598)  weight_decay: 0.0500 (0.0500)  time: 0.6383  data: 0.0363  max mem: 55573
Epoch: [203]  [ 30/156]  eta: 0:02:04  lr: 0.002137  min_lr: 0.002137  loss: 3.4116 (3.2300)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [203]  [ 40/156]  eta: 0:01:43  lr: 0.002135  min_lr: 0.002135  loss: 3.4116 (3.2541)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [203]  [ 50/156]  eta: 0:01:28  lr: 0.002132  min_lr: 0.002132  loss: 3.3588 (3.2331)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [203]  [ 60/156]  eta: 0:01:16  lr: 0.002130  min_lr: 0.002130  loss: 3.3588 (3.2334)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0005  max mem: 55573
Epoch: [203]  [ 70/156]  eta: 0:01:05  lr: 0.002127  min_lr: 0.002127  loss: 3.1398 (3.1937)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [203]  [ 80/156]  eta: 0:00:56  lr: 0.002125  min_lr: 0.002125  loss: 3.0675 (3.1899)  weight_decay: 0.0500 (0.0500)  time: 0.6004  data: 0.0004  max mem: 55573
Epoch: [203]  [ 90/156]  eta: 0:00:48  lr: 0.002122  min_lr: 0.002122  loss: 3.2544 (3.1950)  weight_decay: 0.0500 (0.0500)  time: 0.5997  data: 0.0005  max mem: 55573
Epoch: [203]  [100/156]  eta: 0:00:39  lr: 0.002119  min_lr: 0.002119  loss: 3.4657 (3.2157)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [203]  [110/156]  eta: 0:00:32  lr: 0.002117  min_lr: 0.002117  loss: 3.4873 (3.2119)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [203]  [120/156]  eta: 0:00:24  lr: 0.002114  min_lr: 0.002114  loss: 3.3537 (3.2139)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [203]  [130/156]  eta: 0:00:17  lr: 0.002112  min_lr: 0.002112  loss: 3.3537 (3.2334)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0011  max mem: 55573
Epoch: [203]  [140/156]  eta: 0:00:10  lr: 0.002109  min_lr: 0.002109  loss: 3.2723 (3.2190)  weight_decay: 0.0500 (0.0500)  time: 0.5868  data: 0.0009  max mem: 55573
Epoch: [203]  [150/156]  eta: 0:00:04  lr: 0.002107  min_lr: 0.002107  loss: 3.2779 (3.2158)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [203]  [155/156]  eta: 0:00:00  lr: 0.002106  min_lr: 0.002106  loss: 3.2736 (3.2152)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [203] Total time: 0:01:45 (0.6758 s / it)
Averaged stats: lr: 0.002106  min_lr: 0.002106  loss: 3.2736 (3.1672)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:32  loss: 0.9962 (0.9962)  acc1: 81.7708 (81.7708)  acc5: 96.2240 (96.2240)  time: 6.4186  data: 6.1816  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0960 (1.0894)  acc1: 78.5156 (76.5760)  acc5: 94.3396 (94.5280)  time: 1.4794  data: 1.2867  max mem: 55573
Test: Total time: 0:00:07 (1.5125 s / it)
* Acc@1 77.496 Acc@5 94.088 loss 1.075
Accuracy of the model on the 50000 test images: 77.5%
Max accuracy: 78.10%
Test:  [0/5]  eta: 0:00:37  loss: 1.0890 (1.0890)  acc1: 72.6562 (72.6562)  acc5: 90.8854 (90.8854)  time: 7.4994  data: 7.2633  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1202 (1.1335)  acc1: 70.0521 (68.5440)  acc5: 90.8854 (88.9920)  time: 1.6453  data: 1.4528  max mem: 55573
Test: Total time: 0:00:08 (1.6656 s / it)
* Acc@1 69.032 Acc@5 89.416 loss 1.160
Accuracy of the model EMA on 50000 test images: 69.0%
Max EMA accuracy: 69.03%
Epoch: [204]  [  0/156]  eta: 0:34:36  lr: 0.002105  min_lr: 0.002105  loss: 3.7590 (3.7590)  weight_decay: 0.0500 (0.0500)  time: 13.3097  data: 12.7007  max mem: 55573
Epoch: [204]  [ 10/156]  eta: 0:04:16  lr: 0.002103  min_lr: 0.002103  loss: 3.4372 (3.2891)  weight_decay: 0.0500 (0.0500)  time: 1.7576  data: 1.1552  max mem: 55573
Epoch: [204]  [ 20/156]  eta: 0:02:43  lr: 0.002100  min_lr: 0.002100  loss: 3.1794 (3.1846)  weight_decay: 0.0500 (0.0500)  time: 0.5962  data: 0.0006  max mem: 55573
Epoch: [204]  [ 30/156]  eta: 0:02:06  lr: 0.002098  min_lr: 0.002098  loss: 3.0461 (3.1003)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [204]  [ 40/156]  eta: 0:01:44  lr: 0.002095  min_lr: 0.002095  loss: 3.2007 (3.1128)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [204]  [ 50/156]  eta: 0:01:29  lr: 0.002093  min_lr: 0.002093  loss: 3.2854 (3.1286)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [204]  [ 60/156]  eta: 0:01:16  lr: 0.002090  min_lr: 0.002090  loss: 3.1414 (3.1182)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [204]  [ 70/156]  eta: 0:01:06  lr: 0.002088  min_lr: 0.002088  loss: 3.0275 (3.1139)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [204]  [ 80/156]  eta: 0:00:57  lr: 0.002085  min_lr: 0.002085  loss: 3.0275 (3.1173)  weight_decay: 0.0500 (0.0500)  time: 0.5953  data: 0.0004  max mem: 55573
Epoch: [204]  [ 90/156]  eta: 0:00:48  lr: 0.002083  min_lr: 0.002083  loss: 3.2065 (3.1264)  weight_decay: 0.0500 (0.0500)  time: 0.5957  data: 0.0004  max mem: 55573
Epoch: [204]  [100/156]  eta: 0:00:40  lr: 0.002080  min_lr: 0.002080  loss: 3.3588 (3.1433)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [204]  [110/156]  eta: 0:00:32  lr: 0.002077  min_lr: 0.002077  loss: 3.3307 (3.1532)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [204]  [120/156]  eta: 0:00:25  lr: 0.002075  min_lr: 0.002075  loss: 3.3036 (3.1600)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [204]  [130/156]  eta: 0:00:17  lr: 0.002072  min_lr: 0.002072  loss: 3.4102 (3.1702)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0010  max mem: 55573
Epoch: [204]  [140/156]  eta: 0:00:10  lr: 0.002070  min_lr: 0.002070  loss: 3.2683 (3.1586)  weight_decay: 0.0500 (0.0500)  time: 0.5870  data: 0.0009  max mem: 55573
Epoch: [204]  [150/156]  eta: 0:00:04  lr: 0.002067  min_lr: 0.002067  loss: 3.2109 (3.1530)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [204]  [155/156]  eta: 0:00:00  lr: 0.002066  min_lr: 0.002066  loss: 3.2109 (3.1439)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [204] Total time: 0:01:45 (0.6787 s / it)
Averaged stats: lr: 0.002066  min_lr: 0.002066  loss: 3.2109 (3.1467)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9725 (0.9725)  acc1: 80.7292 (80.7292)  acc5: 96.2240 (96.2240)  time: 6.9708  data: 6.7333  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0318 (1.0763)  acc1: 79.1667 (76.8960)  acc5: 95.5729 (94.3040)  time: 1.5406  data: 1.3467  max mem: 55573
Test: Total time: 0:00:07 (1.5830 s / it)
* Acc@1 77.786 Acc@5 94.210 loss 1.035
Accuracy of the model on the 50000 test images: 77.8%
Max accuracy: 78.10%
Test:  [0/5]  eta: 0:00:34  loss: 1.0851 (1.0851)  acc1: 72.7865 (72.7865)  acc5: 91.0156 (91.0156)  time: 6.9418  data: 6.7057  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1158 (1.1296)  acc1: 70.0521 (68.6080)  acc5: 91.0156 (89.1840)  time: 1.5340  data: 1.3414  max mem: 55573
Test: Total time: 0:00:07 (1.5559 s / it)
* Acc@1 69.116 Acc@5 89.504 loss 1.155
Accuracy of the model EMA on 50000 test images: 69.1%
Max EMA accuracy: 69.12%
Epoch: [205]  [  0/156]  eta: 0:37:45  lr: 0.002066  min_lr: 0.002066  loss: 3.4646 (3.4646)  weight_decay: 0.0500 (0.0500)  time: 14.5196  data: 10.9171  max mem: 55573
Epoch: [205]  [ 10/156]  eta: 0:04:30  lr: 0.002063  min_lr: 0.002063  loss: 3.1164 (3.0108)  weight_decay: 0.0500 (0.0500)  time: 1.8536  data: 0.9928  max mem: 55573
Epoch: [205]  [ 20/156]  eta: 0:02:50  lr: 0.002061  min_lr: 0.002061  loss: 2.8680 (3.0119)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0004  max mem: 55573
Epoch: [205]  [ 30/156]  eta: 0:02:11  lr: 0.002058  min_lr: 0.002058  loss: 3.2336 (3.0582)  weight_decay: 0.0500 (0.0500)  time: 0.5953  data: 0.0004  max mem: 55573
Epoch: [205]  [ 40/156]  eta: 0:01:48  lr: 0.002056  min_lr: 0.002056  loss: 3.2762 (3.0697)  weight_decay: 0.0500 (0.0500)  time: 0.5952  data: 0.0005  max mem: 55573
Epoch: [205]  [ 50/156]  eta: 0:01:31  lr: 0.002053  min_lr: 0.002053  loss: 3.3293 (3.0735)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [205]  [ 60/156]  eta: 0:01:18  lr: 0.002051  min_lr: 0.002051  loss: 2.9210 (3.0615)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [205]  [ 70/156]  eta: 0:01:07  lr: 0.002048  min_lr: 0.002048  loss: 3.0254 (3.0577)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [205]  [ 80/156]  eta: 0:00:58  lr: 0.002046  min_lr: 0.002046  loss: 3.1423 (3.0812)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [205]  [ 90/156]  eta: 0:00:49  lr: 0.002043  min_lr: 0.002043  loss: 3.2512 (3.0993)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [205]  [100/156]  eta: 0:00:40  lr: 0.002041  min_lr: 0.002041  loss: 3.2651 (3.1159)  weight_decay: 0.0500 (0.0500)  time: 0.5931  data: 0.0005  max mem: 55573
Epoch: [205]  [110/156]  eta: 0:00:32  lr: 0.002038  min_lr: 0.002038  loss: 3.2519 (3.1061)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [205]  [120/156]  eta: 0:00:25  lr: 0.002036  min_lr: 0.002036  loss: 3.2011 (3.1157)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [205]  [130/156]  eta: 0:00:18  lr: 0.002033  min_lr: 0.002033  loss: 3.3224 (3.1317)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0011  max mem: 55573
Epoch: [205]  [140/156]  eta: 0:00:11  lr: 0.002031  min_lr: 0.002031  loss: 3.3333 (3.1461)  weight_decay: 0.0500 (0.0500)  time: 0.5866  data: 0.0009  max mem: 55573
Epoch: [205]  [150/156]  eta: 0:00:04  lr: 0.002028  min_lr: 0.002028  loss: 3.3472 (3.1558)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [205]  [155/156]  eta: 0:00:00  lr: 0.002027  min_lr: 0.002027  loss: 3.2980 (3.1512)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [205] Total time: 0:01:47 (0.6864 s / it)
Averaged stats: lr: 0.002027  min_lr: 0.002027  loss: 3.2980 (3.1566)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 0.9824 (0.9824)  acc1: 81.5104 (81.5104)  acc5: 96.3542 (96.3542)  time: 7.2348  data: 6.9975  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0644 (1.0918)  acc1: 79.0365 (77.0240)  acc5: 94.3396 (94.3040)  time: 1.5935  data: 1.3996  max mem: 55573
Test: Total time: 0:00:08 (1.6326 s / it)
* Acc@1 78.154 Acc@5 94.342 loss 1.049
Accuracy of the model on the 50000 test images: 78.2%
Max accuracy: 78.15%
Test:  [0/5]  eta: 0:00:34  loss: 1.0817 (1.0817)  acc1: 72.9167 (72.9167)  acc5: 91.1458 (91.1458)  time: 6.9899  data: 6.7538  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1121 (1.1262)  acc1: 70.0521 (68.6080)  acc5: 91.1458 (89.2160)  time: 1.5435  data: 1.3509  max mem: 55573
Test: Total time: 0:00:07 (1.5622 s / it)
* Acc@1 69.186 Acc@5 89.534 loss 1.152
Accuracy of the model EMA on 50000 test images: 69.2%
Max EMA accuracy: 69.19%
Epoch: [206]  [  0/156]  eta: 0:34:33  lr: 0.002027  min_lr: 0.002027  loss: 3.7574 (3.7574)  weight_decay: 0.0500 (0.0500)  time: 13.2889  data: 9.4575  max mem: 55573
Epoch: [206]  [ 10/156]  eta: 0:04:15  lr: 0.002024  min_lr: 0.002024  loss: 3.4241 (3.3171)  weight_decay: 0.0500 (0.0500)  time: 1.7493  data: 0.8602  max mem: 55573
Epoch: [206]  [ 20/156]  eta: 0:02:42  lr: 0.002022  min_lr: 0.002022  loss: 3.4327 (3.2802)  weight_decay: 0.0500 (0.0500)  time: 0.5939  data: 0.0004  max mem: 55573
Epoch: [206]  [ 30/156]  eta: 0:02:06  lr: 0.002019  min_lr: 0.002019  loss: 3.3849 (3.2853)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0004  max mem: 55573
Epoch: [206]  [ 40/156]  eta: 0:01:44  lr: 0.002017  min_lr: 0.002017  loss: 3.3220 (3.2793)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [206]  [ 50/156]  eta: 0:01:29  lr: 0.002014  min_lr: 0.002014  loss: 3.3745 (3.2727)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [206]  [ 60/156]  eta: 0:01:16  lr: 0.002012  min_lr: 0.002012  loss: 3.3022 (3.2625)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [206]  [ 70/156]  eta: 0:01:06  lr: 0.002009  min_lr: 0.002009  loss: 3.2462 (3.2454)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [206]  [ 80/156]  eta: 0:00:56  lr: 0.002007  min_lr: 0.002007  loss: 3.2408 (3.2474)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [206]  [ 90/156]  eta: 0:00:48  lr: 0.002004  min_lr: 0.002004  loss: 3.3793 (3.2468)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [206]  [100/156]  eta: 0:00:40  lr: 0.002002  min_lr: 0.002002  loss: 3.2542 (3.2177)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [206]  [110/156]  eta: 0:00:32  lr: 0.001999  min_lr: 0.001999  loss: 3.3020 (3.2220)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [206]  [120/156]  eta: 0:00:25  lr: 0.001997  min_lr: 0.001997  loss: 3.3054 (3.2139)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [206]  [130/156]  eta: 0:00:17  lr: 0.001994  min_lr: 0.001994  loss: 3.3054 (3.2141)  weight_decay: 0.0500 (0.0500)  time: 0.5887  data: 0.0010  max mem: 55573
Epoch: [206]  [140/156]  eta: 0:00:10  lr: 0.001992  min_lr: 0.001992  loss: 3.3563 (3.2151)  weight_decay: 0.0500 (0.0500)  time: 0.5856  data: 0.0008  max mem: 55573
Epoch: [206]  [150/156]  eta: 0:00:04  lr: 0.001989  min_lr: 0.001989  loss: 3.3269 (3.2046)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [206]  [155/156]  eta: 0:00:00  lr: 0.001988  min_lr: 0.001988  loss: 2.9841 (3.1999)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [206] Total time: 0:01:45 (0.6776 s / it)
Averaged stats: lr: 0.001988  min_lr: 0.001988  loss: 2.9841 (3.1443)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8816 (0.8816)  acc1: 81.6406 (81.6406)  acc5: 96.8750 (96.8750)  time: 6.9754  data: 6.7380  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9718 (1.0324)  acc1: 79.4271 (77.9520)  acc5: 94.3396 (94.7200)  time: 1.5415  data: 1.3477  max mem: 55573
Test: Total time: 0:00:07 (1.5884 s / it)
* Acc@1 78.614 Acc@5 94.450 loss 0.999
Accuracy of the model on the 50000 test images: 78.6%
Max accuracy: 78.61%
Test:  [0/5]  eta: 0:00:34  loss: 1.0791 (1.0791)  acc1: 73.0469 (73.0469)  acc5: 91.2760 (91.2760)  time: 6.8999  data: 6.6639  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1090 (1.1232)  acc1: 69.9219 (68.6080)  acc5: 91.2760 (89.3120)  time: 1.5254  data: 1.3329  max mem: 55573
Test: Total time: 0:00:07 (1.5451 s / it)
* Acc@1 69.286 Acc@5 89.598 loss 1.148
Accuracy of the model EMA on 50000 test images: 69.3%
Max EMA accuracy: 69.29%
Epoch: [207]  [  0/156]  eta: 0:28:51  lr: 0.001988  min_lr: 0.001988  loss: 3.4963 (3.4963)  weight_decay: 0.0500 (0.0500)  time: 11.1023  data: 10.3627  max mem: 55573
Epoch: [207]  [ 10/156]  eta: 0:04:06  lr: 0.001985  min_lr: 0.001985  loss: 3.4446 (3.2823)  weight_decay: 0.0500 (0.0500)  time: 1.6873  data: 0.9479  max mem: 55573
Epoch: [207]  [ 20/156]  eta: 0:02:38  lr: 0.001983  min_lr: 0.001983  loss: 3.3179 (3.2431)  weight_decay: 0.0500 (0.0500)  time: 0.6695  data: 0.0034  max mem: 55573
Epoch: [207]  [ 30/156]  eta: 0:02:03  lr: 0.001980  min_lr: 0.001980  loss: 3.4041 (3.3167)  weight_decay: 0.0500 (0.0500)  time: 0.5943  data: 0.0004  max mem: 55573
Epoch: [207]  [ 40/156]  eta: 0:01:42  lr: 0.001978  min_lr: 0.001978  loss: 3.5015 (3.2918)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [207]  [ 50/156]  eta: 0:01:27  lr: 0.001975  min_lr: 0.001975  loss: 3.3859 (3.2985)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [207]  [ 60/156]  eta: 0:01:15  lr: 0.001973  min_lr: 0.001973  loss: 3.3883 (3.2791)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [207]  [ 70/156]  eta: 0:01:05  lr: 0.001970  min_lr: 0.001970  loss: 3.1001 (3.2644)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [207]  [ 80/156]  eta: 0:00:56  lr: 0.001968  min_lr: 0.001968  loss: 3.0928 (3.2259)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0005  max mem: 55573
Epoch: [207]  [ 90/156]  eta: 0:00:47  lr: 0.001965  min_lr: 0.001965  loss: 3.1793 (3.2205)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [207]  [100/156]  eta: 0:00:39  lr: 0.001963  min_lr: 0.001963  loss: 3.2989 (3.2172)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [207]  [110/156]  eta: 0:00:32  lr: 0.001961  min_lr: 0.001961  loss: 3.2936 (3.2037)  weight_decay: 0.0500 (0.0500)  time: 0.5950  data: 0.0005  max mem: 55573
Epoch: [207]  [120/156]  eta: 0:00:24  lr: 0.001958  min_lr: 0.001958  loss: 3.3338 (3.2107)  weight_decay: 0.0500 (0.0500)  time: 0.5938  data: 0.0004  max mem: 55573
Epoch: [207]  [130/156]  eta: 0:00:17  lr: 0.001956  min_lr: 0.001956  loss: 3.3746 (3.2022)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0012  max mem: 55573
Epoch: [207]  [140/156]  eta: 0:00:10  lr: 0.001953  min_lr: 0.001953  loss: 3.0355 (3.1881)  weight_decay: 0.0500 (0.0500)  time: 0.5861  data: 0.0010  max mem: 55573
Epoch: [207]  [150/156]  eta: 0:00:04  lr: 0.001951  min_lr: 0.001951  loss: 3.3647 (3.2095)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [207]  [155/156]  eta: 0:00:00  lr: 0.001949  min_lr: 0.001949  loss: 3.4030 (3.1998)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [207] Total time: 0:01:45 (0.6742 s / it)
Averaged stats: lr: 0.001949  min_lr: 0.001949  loss: 3.4030 (3.1422)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.9650 (0.9650)  acc1: 81.2500 (81.2500)  acc5: 95.9635 (95.9635)  time: 6.6668  data: 6.4299  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0851 (1.0908)  acc1: 78.7760 (77.0560)  acc5: 95.9635 (94.4960)  time: 1.4796  data: 1.2861  max mem: 55573
Test: Total time: 0:00:07 (1.5250 s / it)
* Acc@1 78.116 Acc@5 94.318 loss 1.081
Accuracy of the model on the 50000 test images: 78.1%
Max accuracy: 78.61%
Test:  [0/5]  eta: 0:00:35  loss: 1.0763 (1.0763)  acc1: 73.0469 (73.0469)  acc5: 91.4062 (91.4062)  time: 7.0868  data: 6.8509  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1058 (1.1203)  acc1: 70.0521 (68.6080)  acc5: 91.4062 (89.3760)  time: 1.5627  data: 1.3703  max mem: 55573
Test: Total time: 0:00:07 (1.5815 s / it)
* Acc@1 69.332 Acc@5 89.640 loss 1.145
Accuracy of the model EMA on 50000 test images: 69.3%
Max EMA accuracy: 69.33%
Epoch: [208]  [  0/156]  eta: 0:33:29  lr: 0.001949  min_lr: 0.001949  loss: 2.4708 (2.4708)  weight_decay: 0.0500 (0.0500)  time: 12.8799  data: 11.1613  max mem: 55573
Epoch: [208]  [ 10/156]  eta: 0:04:09  lr: 0.001947  min_lr: 0.001947  loss: 2.9779 (3.0305)  weight_decay: 0.0500 (0.0500)  time: 1.7082  data: 1.0151  max mem: 55573
Epoch: [208]  [ 20/156]  eta: 0:02:39  lr: 0.001944  min_lr: 0.001944  loss: 3.0252 (3.0875)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [208]  [ 30/156]  eta: 0:02:04  lr: 0.001942  min_lr: 0.001942  loss: 3.4010 (3.1195)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [208]  [ 40/156]  eta: 0:01:43  lr: 0.001939  min_lr: 0.001939  loss: 3.3063 (3.0873)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [208]  [ 50/156]  eta: 0:01:28  lr: 0.001937  min_lr: 0.001937  loss: 2.8788 (3.0705)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [208]  [ 60/156]  eta: 0:01:15  lr: 0.001934  min_lr: 0.001934  loss: 3.0866 (3.0880)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [208]  [ 70/156]  eta: 0:01:05  lr: 0.001932  min_lr: 0.001932  loss: 3.0608 (3.0864)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [208]  [ 80/156]  eta: 0:00:56  lr: 0.001929  min_lr: 0.001929  loss: 3.0562 (3.0834)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [208]  [ 90/156]  eta: 0:00:47  lr: 0.001927  min_lr: 0.001927  loss: 3.1270 (3.0818)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [208]  [100/156]  eta: 0:00:39  lr: 0.001925  min_lr: 0.001925  loss: 3.1270 (3.0738)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [208]  [110/156]  eta: 0:00:32  lr: 0.001922  min_lr: 0.001922  loss: 3.0974 (3.0586)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [208]  [120/156]  eta: 0:00:24  lr: 0.001920  min_lr: 0.001920  loss: 3.2718 (3.0817)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [208]  [130/156]  eta: 0:00:17  lr: 0.001917  min_lr: 0.001917  loss: 3.3758 (3.0969)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0011  max mem: 55573
Epoch: [208]  [140/156]  eta: 0:00:10  lr: 0.001915  min_lr: 0.001915  loss: 3.3332 (3.1032)  weight_decay: 0.0500 (0.0500)  time: 0.5853  data: 0.0009  max mem: 55573
Epoch: [208]  [150/156]  eta: 0:00:04  lr: 0.001912  min_lr: 0.001912  loss: 3.1068 (3.0982)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [208]  [155/156]  eta: 0:00:00  lr: 0.001911  min_lr: 0.001911  loss: 2.9517 (3.0930)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [208] Total time: 0:01:45 (0.6744 s / it)
Averaged stats: lr: 0.001911  min_lr: 0.001911  loss: 2.9517 (3.1491)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9631 (0.9631)  acc1: 81.2500 (81.2500)  acc5: 95.5729 (95.5729)  time: 6.8543  data: 6.6172  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9761 (1.0095)  acc1: 80.0781 (77.6640)  acc5: 95.5729 (94.2080)  time: 1.5173  data: 1.3235  max mem: 55573
Test: Total time: 0:00:07 (1.5645 s / it)
* Acc@1 78.166 Acc@5 94.404 loss 0.998
Accuracy of the model on the 50000 test images: 78.2%
Max accuracy: 78.61%
Test:  [0/5]  eta: 0:00:36  loss: 1.0734 (1.0734)  acc1: 73.0469 (73.0469)  acc5: 91.5365 (91.5365)  time: 7.2293  data: 6.9934  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1026 (1.1173)  acc1: 70.1823 (68.6080)  acc5: 91.5365 (89.4080)  time: 1.5911  data: 1.3988  max mem: 55573
Test: Total time: 0:00:08 (1.6162 s / it)
* Acc@1 69.408 Acc@5 89.692 loss 1.142
Accuracy of the model EMA on 50000 test images: 69.4%
Max EMA accuracy: 69.41%
Epoch: [209]  [  0/156]  eta: 0:34:21  lr: 0.001911  min_lr: 0.001911  loss: 3.6690 (3.6690)  weight_decay: 0.0500 (0.0500)  time: 13.2140  data: 10.4374  max mem: 55573
Epoch: [209]  [ 10/156]  eta: 0:04:15  lr: 0.001908  min_lr: 0.001908  loss: 3.2353 (3.1254)  weight_decay: 0.0500 (0.0500)  time: 1.7511  data: 0.9493  max mem: 55573
Epoch: [209]  [ 20/156]  eta: 0:02:42  lr: 0.001906  min_lr: 0.001906  loss: 3.0032 (3.0763)  weight_decay: 0.0500 (0.0500)  time: 0.5970  data: 0.0004  max mem: 55573
Epoch: [209]  [ 30/156]  eta: 0:02:06  lr: 0.001903  min_lr: 0.001903  loss: 3.2847 (3.1201)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [209]  [ 40/156]  eta: 0:01:44  lr: 0.001901  min_lr: 0.001901  loss: 3.4727 (3.1834)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [209]  [ 50/156]  eta: 0:01:29  lr: 0.001899  min_lr: 0.001899  loss: 3.3518 (3.1506)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [209]  [ 60/156]  eta: 0:01:16  lr: 0.001896  min_lr: 0.001896  loss: 3.1544 (3.1475)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [209]  [ 70/156]  eta: 0:01:06  lr: 0.001894  min_lr: 0.001894  loss: 3.2288 (3.1499)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [209]  [ 80/156]  eta: 0:00:56  lr: 0.001891  min_lr: 0.001891  loss: 3.3405 (3.1569)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0005  max mem: 55573
Epoch: [209]  [ 90/156]  eta: 0:00:48  lr: 0.001889  min_lr: 0.001889  loss: 3.0906 (3.1289)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [209]  [100/156]  eta: 0:00:40  lr: 0.001886  min_lr: 0.001886  loss: 2.9302 (3.1016)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [209]  [110/156]  eta: 0:00:32  lr: 0.001884  min_lr: 0.001884  loss: 3.2180 (3.1239)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [209]  [120/156]  eta: 0:00:25  lr: 0.001881  min_lr: 0.001881  loss: 3.3017 (3.1314)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [209]  [130/156]  eta: 0:00:17  lr: 0.001879  min_lr: 0.001879  loss: 3.3347 (3.1563)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0024  max mem: 55573
Epoch: [209]  [140/156]  eta: 0:00:10  lr: 0.001877  min_lr: 0.001877  loss: 3.2794 (3.1434)  weight_decay: 0.0500 (0.0500)  time: 0.5871  data: 0.0022  max mem: 55573
Epoch: [209]  [150/156]  eta: 0:00:04  lr: 0.001874  min_lr: 0.001874  loss: 3.0019 (3.1330)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [209]  [155/156]  eta: 0:00:00  lr: 0.001873  min_lr: 0.001873  loss: 3.0067 (3.1443)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [209] Total time: 0:01:45 (0.6779 s / it)
Averaged stats: lr: 0.001873  min_lr: 0.001873  loss: 3.0067 (3.1496)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.9831 (0.9831)  acc1: 80.0781 (80.0781)  acc5: 96.0938 (96.0938)  time: 6.6074  data: 6.3692  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0735 (1.1035)  acc1: 78.2552 (76.9920)  acc5: 94.3396 (94.2720)  time: 1.4669  data: 1.2740  max mem: 55573
Test: Total time: 0:00:07 (1.5141 s / it)
* Acc@1 77.966 Acc@5 94.366 loss 1.072
Accuracy of the model on the 50000 test images: 78.0%
Max accuracy: 78.61%
Test:  [0/5]  eta: 0:00:32  loss: 1.0708 (1.0708)  acc1: 73.0469 (73.0469)  acc5: 91.5365 (91.5365)  time: 6.5380  data: 6.3019  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0996 (1.1146)  acc1: 70.5729 (68.7360)  acc5: 91.5365 (89.5360)  time: 1.5080  data: 1.3157  max mem: 55573
Test: Total time: 0:00:07 (1.5269 s / it)
* Acc@1 69.514 Acc@5 89.734 loss 1.138
Accuracy of the model EMA on 50000 test images: 69.5%
Max EMA accuracy: 69.51%
Epoch: [210]  [  0/156]  eta: 0:29:41  lr: 0.001873  min_lr: 0.001873  loss: 2.0959 (2.0959)  weight_decay: 0.0500 (0.0500)  time: 11.4205  data: 9.1617  max mem: 55573
Epoch: [210]  [ 10/156]  eta: 0:04:04  lr: 0.001870  min_lr: 0.001870  loss: 3.3309 (3.0669)  weight_decay: 0.0500 (0.0500)  time: 1.6769  data: 0.9129  max mem: 55573
Epoch: [210]  [ 20/156]  eta: 0:02:37  lr: 0.001868  min_lr: 0.001868  loss: 3.3289 (3.1228)  weight_decay: 0.0500 (0.0500)  time: 0.6471  data: 0.0442  max mem: 55573
Epoch: [210]  [ 30/156]  eta: 0:02:03  lr: 0.001865  min_lr: 0.001865  loss: 3.1541 (3.1505)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [210]  [ 40/156]  eta: 0:01:42  lr: 0.001863  min_lr: 0.001863  loss: 3.3196 (3.1643)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [210]  [ 50/156]  eta: 0:01:27  lr: 0.001860  min_lr: 0.001860  loss: 3.3196 (3.1454)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0004  max mem: 55573
Epoch: [210]  [ 60/156]  eta: 0:01:15  lr: 0.001858  min_lr: 0.001858  loss: 3.2585 (3.1659)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [210]  [ 70/156]  eta: 0:01:05  lr: 0.001856  min_lr: 0.001856  loss: 3.3563 (3.2044)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [210]  [ 80/156]  eta: 0:00:56  lr: 0.001853  min_lr: 0.001853  loss: 3.3563 (3.1772)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [210]  [ 90/156]  eta: 0:00:47  lr: 0.001851  min_lr: 0.001851  loss: 3.3470 (3.2091)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [210]  [100/156]  eta: 0:00:39  lr: 0.001848  min_lr: 0.001848  loss: 3.3504 (3.2131)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [210]  [110/156]  eta: 0:00:32  lr: 0.001846  min_lr: 0.001846  loss: 3.3256 (3.2012)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [210]  [120/156]  eta: 0:00:24  lr: 0.001843  min_lr: 0.001843  loss: 3.0427 (3.1804)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [210]  [130/156]  eta: 0:00:17  lr: 0.001841  min_lr: 0.001841  loss: 3.0427 (3.1736)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0011  max mem: 55573
Epoch: [210]  [140/156]  eta: 0:00:10  lr: 0.001839  min_lr: 0.001839  loss: 3.1942 (3.1638)  weight_decay: 0.0500 (0.0500)  time: 0.5865  data: 0.0009  max mem: 55573
Epoch: [210]  [150/156]  eta: 0:00:04  lr: 0.001836  min_lr: 0.001836  loss: 3.1942 (3.1618)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [210]  [155/156]  eta: 0:00:00  lr: 0.001835  min_lr: 0.001835  loss: 3.1942 (3.1635)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [210] Total time: 0:01:45 (0.6735 s / it)
Averaged stats: lr: 0.001835  min_lr: 0.001835  loss: 3.1942 (3.1191)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.0920 (1.0920)  acc1: 80.9896 (80.9896)  acc5: 96.7448 (96.7448)  time: 6.9672  data: 6.7297  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1419 (1.1153)  acc1: 79.2969 (77.6640)  acc5: 95.8333 (94.7200)  time: 1.5396  data: 1.3460  max mem: 55573
Test: Total time: 0:00:07 (1.5849 s / it)
* Acc@1 78.314 Acc@5 94.408 loss 1.106
Accuracy of the model on the 50000 test images: 78.3%
Max accuracy: 78.61%
Test:  [0/5]  eta: 0:00:33  loss: 1.0685 (1.0685)  acc1: 73.0469 (73.0469)  acc5: 91.6667 (91.6667)  time: 6.6596  data: 6.4235  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0971 (1.1121)  acc1: 70.5729 (68.8320)  acc5: 91.6667 (89.6320)  time: 1.4773  data: 1.2848  max mem: 55573
Test: Total time: 0:00:07 (1.4923 s / it)
* Acc@1 69.586 Acc@5 89.764 loss 1.136
Accuracy of the model EMA on 50000 test images: 69.6%
Max EMA accuracy: 69.59%
Epoch: [211]  [  0/156]  eta: 0:35:57  lr: 0.001835  min_lr: 0.001835  loss: 2.3913 (2.3913)  weight_decay: 0.0500 (0.0500)  time: 13.8286  data: 12.7412  max mem: 55573
Epoch: [211]  [ 10/156]  eta: 0:04:22  lr: 0.001832  min_lr: 0.001832  loss: 3.1645 (3.0139)  weight_decay: 0.0500 (0.0500)  time: 1.7997  data: 1.1586  max mem: 55573
Epoch: [211]  [ 20/156]  eta: 0:02:46  lr: 0.001830  min_lr: 0.001830  loss: 3.2544 (3.0907)  weight_decay: 0.0500 (0.0500)  time: 0.5952  data: 0.0004  max mem: 55573
Epoch: [211]  [ 30/156]  eta: 0:02:08  lr: 0.001828  min_lr: 0.001828  loss: 3.3390 (3.1218)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [211]  [ 40/156]  eta: 0:01:46  lr: 0.001825  min_lr: 0.001825  loss: 3.1527 (3.1032)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0005  max mem: 55573
Epoch: [211]  [ 50/156]  eta: 0:01:30  lr: 0.001823  min_lr: 0.001823  loss: 3.0350 (3.0817)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [211]  [ 60/156]  eta: 0:01:17  lr: 0.001820  min_lr: 0.001820  loss: 3.1466 (3.0876)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0005  max mem: 55573
Epoch: [211]  [ 70/156]  eta: 0:01:06  lr: 0.001818  min_lr: 0.001818  loss: 2.9999 (3.0465)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [211]  [ 80/156]  eta: 0:00:57  lr: 0.001815  min_lr: 0.001815  loss: 2.9878 (3.0590)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [211]  [ 90/156]  eta: 0:00:48  lr: 0.001813  min_lr: 0.001813  loss: 3.1868 (3.0660)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [211]  [100/156]  eta: 0:00:40  lr: 0.001811  min_lr: 0.001811  loss: 3.1868 (3.0688)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [211]  [110/156]  eta: 0:00:32  lr: 0.001808  min_lr: 0.001808  loss: 3.3567 (3.0996)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [211]  [120/156]  eta: 0:00:25  lr: 0.001806  min_lr: 0.001806  loss: 3.3516 (3.1045)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0004  max mem: 55573
Epoch: [211]  [130/156]  eta: 0:00:17  lr: 0.001803  min_lr: 0.001803  loss: 3.3195 (3.1233)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0011  max mem: 55573
Epoch: [211]  [140/156]  eta: 0:00:10  lr: 0.001801  min_lr: 0.001801  loss: 3.3195 (3.1383)  weight_decay: 0.0500 (0.0500)  time: 0.5874  data: 0.0010  max mem: 55573
Epoch: [211]  [150/156]  eta: 0:00:04  lr: 0.001799  min_lr: 0.001799  loss: 3.2773 (3.1448)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [211]  [155/156]  eta: 0:00:00  lr: 0.001797  min_lr: 0.001797  loss: 3.1939 (3.1315)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [211] Total time: 0:01:46 (0.6813 s / it)
Averaged stats: lr: 0.001797  min_lr: 0.001797  loss: 3.1939 (3.1266)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9255 (0.9255)  acc1: 81.7708 (81.7708)  acc5: 96.7448 (96.7448)  time: 6.9882  data: 6.7502  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9730 (0.9559)  acc1: 78.9062 (77.3120)  acc5: 95.5729 (94.6560)  time: 1.5441  data: 1.3501  max mem: 55573
Test: Total time: 0:00:07 (1.5872 s / it)
* Acc@1 78.652 Acc@5 94.518 loss 0.955
Accuracy of the model on the 50000 test images: 78.7%
Max accuracy: 78.65%
Test:  [0/5]  eta: 0:00:35  loss: 1.0660 (1.0660)  acc1: 73.1771 (73.1771)  acc5: 91.9271 (91.9271)  time: 7.0900  data: 6.8539  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0944 (1.1095)  acc1: 70.8333 (68.9920)  acc5: 91.9271 (89.7280)  time: 1.5634  data: 1.3709  max mem: 55573
Test: Total time: 0:00:07 (1.5853 s / it)
* Acc@1 69.664 Acc@5 89.812 loss 1.133
Accuracy of the model EMA on 50000 test images: 69.7%
Max EMA accuracy: 69.66%
Epoch: [212]  [  0/156]  eta: 0:36:33  lr: 0.001797  min_lr: 0.001797  loss: 2.7085 (2.7085)  weight_decay: 0.0500 (0.0500)  time: 14.0600  data: 12.6390  max mem: 55573
Epoch: [212]  [ 10/156]  eta: 0:04:26  lr: 0.001795  min_lr: 0.001795  loss: 3.2344 (3.2011)  weight_decay: 0.0500 (0.0500)  time: 1.8266  data: 1.1493  max mem: 55573
Epoch: [212]  [ 20/156]  eta: 0:02:48  lr: 0.001792  min_lr: 0.001792  loss: 3.3377 (3.2170)  weight_decay: 0.0500 (0.0500)  time: 0.5973  data: 0.0004  max mem: 55573
Epoch: [212]  [ 30/156]  eta: 0:02:09  lr: 0.001790  min_lr: 0.001790  loss: 3.2983 (3.1623)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [212]  [ 40/156]  eta: 0:01:46  lr: 0.001788  min_lr: 0.001788  loss: 3.3371 (3.2168)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [212]  [ 50/156]  eta: 0:01:30  lr: 0.001785  min_lr: 0.001785  loss: 3.3583 (3.2174)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [212]  [ 60/156]  eta: 0:01:18  lr: 0.001783  min_lr: 0.001783  loss: 3.2825 (3.2037)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [212]  [ 70/156]  eta: 0:01:07  lr: 0.001780  min_lr: 0.001780  loss: 2.9533 (3.1330)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [212]  [ 80/156]  eta: 0:00:57  lr: 0.001778  min_lr: 0.001778  loss: 2.8816 (3.1092)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [212]  [ 90/156]  eta: 0:00:48  lr: 0.001776  min_lr: 0.001776  loss: 3.1878 (3.1240)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [212]  [100/156]  eta: 0:00:40  lr: 0.001773  min_lr: 0.001773  loss: 3.1878 (3.1286)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [212]  [110/156]  eta: 0:00:32  lr: 0.001771  min_lr: 0.001771  loss: 3.2324 (3.1239)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [212]  [120/156]  eta: 0:00:25  lr: 0.001768  min_lr: 0.001768  loss: 3.3208 (3.1313)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [212]  [130/156]  eta: 0:00:18  lr: 0.001766  min_lr: 0.001766  loss: 3.2068 (3.1336)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0010  max mem: 55573
Epoch: [212]  [140/156]  eta: 0:00:10  lr: 0.001764  min_lr: 0.001764  loss: 3.1470 (3.1226)  weight_decay: 0.0500 (0.0500)  time: 0.5863  data: 0.0009  max mem: 55573
Epoch: [212]  [150/156]  eta: 0:00:04  lr: 0.001761  min_lr: 0.001761  loss: 3.0337 (3.1111)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [212]  [155/156]  eta: 0:00:00  lr: 0.001760  min_lr: 0.001760  loss: 2.9615 (3.1158)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [212] Total time: 0:01:46 (0.6831 s / it)
Averaged stats: lr: 0.001760  min_lr: 0.001760  loss: 2.9615 (3.1152)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.9168 (0.9168)  acc1: 81.6406 (81.6406)  acc5: 96.0938 (96.0938)  time: 7.1023  data: 6.8649  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0173 (1.0017)  acc1: 79.5573 (77.7920)  acc5: 96.0938 (94.8480)  time: 1.5670  data: 1.3730  max mem: 55573
Test: Total time: 0:00:08 (1.6102 s / it)
* Acc@1 78.838 Acc@5 94.634 loss 0.985
Accuracy of the model on the 50000 test images: 78.8%
Max accuracy: 78.84%
Test:  [0/5]  eta: 0:00:33  loss: 1.0634 (1.0634)  acc1: 73.1771 (73.1771)  acc5: 91.9271 (91.9271)  time: 6.7499  data: 6.5133  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0916 (1.1070)  acc1: 70.8333 (68.9920)  acc5: 91.9271 (89.6960)  time: 1.4960  data: 1.3029  max mem: 55573
Test: Total time: 0:00:07 (1.5145 s / it)
* Acc@1 69.732 Acc@5 89.836 loss 1.130
Accuracy of the model EMA on 50000 test images: 69.7%
Max EMA accuracy: 69.73%
Epoch: [213]  [  0/156]  eta: 0:33:36  lr: 0.001760  min_lr: 0.001760  loss: 3.0784 (3.0784)  weight_decay: 0.0500 (0.0500)  time: 12.9233  data: 9.3686  max mem: 55573
Epoch: [213]  [ 10/156]  eta: 0:04:14  lr: 0.001757  min_lr: 0.001757  loss: 3.1413 (3.1599)  weight_decay: 0.0500 (0.0500)  time: 1.7406  data: 0.8520  max mem: 55573
Epoch: [213]  [ 20/156]  eta: 0:02:42  lr: 0.001755  min_lr: 0.001755  loss: 3.3922 (3.2307)  weight_decay: 0.0500 (0.0500)  time: 0.6062  data: 0.0004  max mem: 55573
Epoch: [213]  [ 30/156]  eta: 0:02:05  lr: 0.001753  min_lr: 0.001753  loss: 3.3571 (3.1660)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [213]  [ 40/156]  eta: 0:01:44  lr: 0.001750  min_lr: 0.001750  loss: 3.1094 (3.1780)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [213]  [ 50/156]  eta: 0:01:28  lr: 0.001748  min_lr: 0.001748  loss: 3.1200 (3.1489)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [213]  [ 60/156]  eta: 0:01:16  lr: 0.001746  min_lr: 0.001746  loss: 3.2226 (3.1607)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0005  max mem: 55573
Epoch: [213]  [ 70/156]  eta: 0:01:06  lr: 0.001743  min_lr: 0.001743  loss: 3.3075 (3.1746)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0004  max mem: 55573
Epoch: [213]  [ 80/156]  eta: 0:00:56  lr: 0.001741  min_lr: 0.001741  loss: 3.2598 (3.1758)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [213]  [ 90/156]  eta: 0:00:48  lr: 0.001738  min_lr: 0.001738  loss: 3.2812 (3.1849)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [213]  [100/156]  eta: 0:00:40  lr: 0.001736  min_lr: 0.001736  loss: 3.3209 (3.1903)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [213]  [110/156]  eta: 0:00:32  lr: 0.001734  min_lr: 0.001734  loss: 3.2692 (3.1919)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [213]  [120/156]  eta: 0:00:25  lr: 0.001731  min_lr: 0.001731  loss: 3.1429 (3.1840)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [213]  [130/156]  eta: 0:00:17  lr: 0.001729  min_lr: 0.001729  loss: 3.0150 (3.1823)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0011  max mem: 55573
Epoch: [213]  [140/156]  eta: 0:00:10  lr: 0.001727  min_lr: 0.001727  loss: 3.0771 (3.1649)  weight_decay: 0.0500 (0.0500)  time: 0.5852  data: 0.0009  max mem: 55573
Epoch: [213]  [150/156]  eta: 0:00:04  lr: 0.001724  min_lr: 0.001724  loss: 3.0788 (3.1627)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [213]  [155/156]  eta: 0:00:00  lr: 0.001723  min_lr: 0.001723  loss: 3.1482 (3.1652)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [213] Total time: 0:01:45 (0.6764 s / it)
Averaged stats: lr: 0.001723  min_lr: 0.001723  loss: 3.1482 (3.1216)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 0.9814 (0.9814)  acc1: 80.8594 (80.8594)  acc5: 96.6146 (96.6146)  time: 7.2083  data: 6.9709  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0367 (1.0449)  acc1: 79.8177 (77.0560)  acc5: 95.9635 (94.6240)  time: 1.5881  data: 1.3943  max mem: 55573
Test: Total time: 0:00:08 (1.6284 s / it)
* Acc@1 78.454 Acc@5 94.612 loss 1.015
Accuracy of the model on the 50000 test images: 78.5%
Max accuracy: 78.84%
Test:  [0/5]  eta: 0:00:34  loss: 1.0611 (1.0611)  acc1: 73.3073 (73.3073)  acc5: 91.9271 (91.9271)  time: 6.9508  data: 6.7147  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0893 (1.1048)  acc1: 70.8333 (69.0560)  acc5: 91.9271 (89.6960)  time: 1.5354  data: 1.3430  max mem: 55573
Test: Total time: 0:00:07 (1.5543 s / it)
* Acc@1 69.828 Acc@5 89.870 loss 1.127
Accuracy of the model EMA on 50000 test images: 69.8%
Max EMA accuracy: 69.83%
Epoch: [214]  [  0/156]  eta: 0:38:02  lr: 0.001723  min_lr: 0.001723  loss: 3.5207 (3.5207)  weight_decay: 0.0500 (0.0500)  time: 14.6312  data: 9.5946  max mem: 55573
Epoch: [214]  [ 10/156]  eta: 0:04:33  lr: 0.001720  min_lr: 0.001720  loss: 3.2970 (3.1041)  weight_decay: 0.0500 (0.0500)  time: 1.8717  data: 0.8726  max mem: 55573
Epoch: [214]  [ 20/156]  eta: 0:02:51  lr: 0.001718  min_lr: 0.001718  loss: 3.1082 (3.0622)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0003  max mem: 55573
Epoch: [214]  [ 30/156]  eta: 0:02:11  lr: 0.001716  min_lr: 0.001716  loss: 2.8791 (2.9601)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [214]  [ 40/156]  eta: 0:01:48  lr: 0.001713  min_lr: 0.001713  loss: 3.0819 (2.9951)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [214]  [ 50/156]  eta: 0:01:31  lr: 0.001711  min_lr: 0.001711  loss: 3.1240 (2.9942)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [214]  [ 60/156]  eta: 0:01:18  lr: 0.001709  min_lr: 0.001709  loss: 3.0816 (3.0036)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [214]  [ 70/156]  eta: 0:01:07  lr: 0.001706  min_lr: 0.001706  loss: 3.1293 (3.0124)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [214]  [ 80/156]  eta: 0:00:58  lr: 0.001704  min_lr: 0.001704  loss: 3.1293 (3.0294)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [214]  [ 90/156]  eta: 0:00:49  lr: 0.001702  min_lr: 0.001702  loss: 3.2887 (3.0675)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [214]  [100/156]  eta: 0:00:40  lr: 0.001699  min_lr: 0.001699  loss: 3.3691 (3.0988)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [214]  [110/156]  eta: 0:00:32  lr: 0.001697  min_lr: 0.001697  loss: 3.3346 (3.0873)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [214]  [120/156]  eta: 0:00:25  lr: 0.001695  min_lr: 0.001695  loss: 3.3139 (3.1101)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [214]  [130/156]  eta: 0:00:18  lr: 0.001692  min_lr: 0.001692  loss: 3.3611 (3.1234)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0010  max mem: 55573
Epoch: [214]  [140/156]  eta: 0:00:11  lr: 0.001690  min_lr: 0.001690  loss: 3.3584 (3.1303)  weight_decay: 0.0500 (0.0500)  time: 0.5868  data: 0.0008  max mem: 55573
Epoch: [214]  [150/156]  eta: 0:00:04  lr: 0.001688  min_lr: 0.001688  loss: 3.3108 (3.1370)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [214]  [155/156]  eta: 0:00:00  lr: 0.001686  min_lr: 0.001686  loss: 3.3009 (3.1334)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [214] Total time: 0:01:47 (0.6862 s / it)
Averaged stats: lr: 0.001686  min_lr: 0.001686  loss: 3.3009 (3.0996)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.9133 (0.9133)  acc1: 81.1198 (81.1198)  acc5: 96.3542 (96.3542)  time: 7.0429  data: 6.8054  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9660 (0.9932)  acc1: 79.6875 (77.4720)  acc5: 96.2264 (94.5280)  time: 1.5550  data: 1.3612  max mem: 55573
Test: Total time: 0:00:08 (1.6169 s / it)
* Acc@1 78.810 Acc@5 94.690 loss 0.990
Accuracy of the model on the 50000 test images: 78.8%
Max accuracy: 78.84%
Test:  [0/5]  eta: 0:00:36  loss: 1.0587 (1.0587)  acc1: 73.4375 (73.4375)  acc5: 92.0573 (92.0573)  time: 7.3359  data: 7.0998  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0868 (1.1025)  acc1: 71.2240 (69.2160)  acc5: 92.0573 (89.7280)  time: 1.6126  data: 1.4201  max mem: 55573
Test: Total time: 0:00:08 (1.6332 s / it)
* Acc@1 69.906 Acc@5 89.906 loss 1.125
Accuracy of the model EMA on 50000 test images: 69.9%
Max EMA accuracy: 69.91%
Epoch: [215]  [  0/156]  eta: 0:30:23  lr: 0.001686  min_lr: 0.001686  loss: 3.4968 (3.4968)  weight_decay: 0.0500 (0.0500)  time: 11.6875  data: 9.7937  max mem: 55573
Epoch: [215]  [ 10/156]  eta: 0:04:16  lr: 0.001684  min_lr: 0.001684  loss: 3.4175 (3.1414)  weight_decay: 0.0500 (0.0500)  time: 1.7590  data: 0.8907  max mem: 55573
Epoch: [215]  [ 20/156]  eta: 0:02:43  lr: 0.001681  min_lr: 0.001681  loss: 3.3409 (3.2307)  weight_decay: 0.0500 (0.0500)  time: 0.6792  data: 0.0004  max mem: 55573
Epoch: [215]  [ 30/156]  eta: 0:02:06  lr: 0.001679  min_lr: 0.001679  loss: 3.3023 (3.1564)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [215]  [ 40/156]  eta: 0:01:44  lr: 0.001677  min_lr: 0.001677  loss: 3.1451 (3.1838)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [215]  [ 50/156]  eta: 0:01:29  lr: 0.001674  min_lr: 0.001674  loss: 3.2846 (3.1691)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [215]  [ 60/156]  eta: 0:01:16  lr: 0.001672  min_lr: 0.001672  loss: 3.2832 (3.1867)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [215]  [ 70/156]  eta: 0:01:06  lr: 0.001670  min_lr: 0.001670  loss: 3.3194 (3.1957)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [215]  [ 80/156]  eta: 0:00:56  lr: 0.001667  min_lr: 0.001667  loss: 3.2864 (3.1868)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [215]  [ 90/156]  eta: 0:00:48  lr: 0.001665  min_lr: 0.001665  loss: 3.4295 (3.2122)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [215]  [100/156]  eta: 0:00:40  lr: 0.001663  min_lr: 0.001663  loss: 3.3292 (3.1779)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [215]  [110/156]  eta: 0:00:32  lr: 0.001660  min_lr: 0.001660  loss: 2.7526 (3.1438)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [215]  [120/156]  eta: 0:00:25  lr: 0.001658  min_lr: 0.001658  loss: 3.1280 (3.1461)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [215]  [130/156]  eta: 0:00:17  lr: 0.001656  min_lr: 0.001656  loss: 3.0126 (3.1309)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0010  max mem: 55573
Epoch: [215]  [140/156]  eta: 0:00:10  lr: 0.001653  min_lr: 0.001653  loss: 3.1092 (3.1377)  weight_decay: 0.0500 (0.0500)  time: 0.5863  data: 0.0009  max mem: 55573
Epoch: [215]  [150/156]  eta: 0:00:04  lr: 0.001651  min_lr: 0.001651  loss: 3.1092 (3.1202)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [215]  [155/156]  eta: 0:00:00  lr: 0.001650  min_lr: 0.001650  loss: 3.0807 (3.1247)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [215] Total time: 0:01:45 (0.6784 s / it)
Averaged stats: lr: 0.001650  min_lr: 0.001650  loss: 3.0807 (3.1125)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9190 (0.9190)  acc1: 82.5521 (82.5521)  acc5: 95.9635 (95.9635)  time: 6.9906  data: 6.7533  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0183 (1.0213)  acc1: 79.4271 (77.8240)  acc5: 94.3396 (94.4640)  time: 1.5445  data: 1.3507  max mem: 55573
Test: Total time: 0:00:07 (1.5843 s / it)
* Acc@1 78.648 Acc@5 94.648 loss 1.006
Accuracy of the model on the 50000 test images: 78.6%
Max accuracy: 78.84%
Test:  [0/5]  eta: 0:00:35  loss: 1.0564 (1.0564)  acc1: 73.4375 (73.4375)  acc5: 92.1875 (92.1875)  time: 7.0016  data: 6.7656  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0846 (1.1003)  acc1: 71.2240 (69.2160)  acc5: 92.1875 (89.7600)  time: 1.5890  data: 1.3966  max mem: 55573
Test: Total time: 0:00:08 (1.6108 s / it)
* Acc@1 69.992 Acc@5 89.942 loss 1.122
Accuracy of the model EMA on 50000 test images: 70.0%
Max EMA accuracy: 69.99%
Epoch: [216]  [  0/156]  eta: 0:31:46  lr: 0.001650  min_lr: 0.001650  loss: 3.3019 (3.3019)  weight_decay: 0.0500 (0.0500)  time: 12.2226  data: 11.6346  max mem: 55573
Epoch: [216]  [ 10/156]  eta: 0:04:03  lr: 0.001647  min_lr: 0.001647  loss: 3.3019 (3.0478)  weight_decay: 0.0500 (0.0500)  time: 1.6683  data: 1.0581  max mem: 55573
Epoch: [216]  [ 20/156]  eta: 0:02:37  lr: 0.001645  min_lr: 0.001645  loss: 3.2251 (3.0778)  weight_decay: 0.0500 (0.0500)  time: 0.6049  data: 0.0004  max mem: 55573
Epoch: [216]  [ 30/156]  eta: 0:02:02  lr: 0.001643  min_lr: 0.001643  loss: 2.9045 (2.9681)  weight_decay: 0.0500 (0.0500)  time: 0.5935  data: 0.0004  max mem: 55573
Epoch: [216]  [ 40/156]  eta: 0:01:42  lr: 0.001640  min_lr: 0.001640  loss: 2.8963 (2.9995)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [216]  [ 50/156]  eta: 0:01:27  lr: 0.001638  min_lr: 0.001638  loss: 3.3549 (3.0728)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [216]  [ 60/156]  eta: 0:01:15  lr: 0.001636  min_lr: 0.001636  loss: 3.1543 (3.0433)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [216]  [ 70/156]  eta: 0:01:05  lr: 0.001633  min_lr: 0.001633  loss: 2.9654 (3.0307)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [216]  [ 80/156]  eta: 0:00:56  lr: 0.001631  min_lr: 0.001631  loss: 3.0545 (3.0375)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [216]  [ 90/156]  eta: 0:00:47  lr: 0.001629  min_lr: 0.001629  loss: 3.0796 (3.0214)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [216]  [100/156]  eta: 0:00:39  lr: 0.001626  min_lr: 0.001626  loss: 3.0796 (3.0181)  weight_decay: 0.0500 (0.0500)  time: 0.6030  data: 0.0004  max mem: 55573
Epoch: [216]  [110/156]  eta: 0:00:32  lr: 0.001624  min_lr: 0.001624  loss: 3.3750 (3.0569)  weight_decay: 0.0500 (0.0500)  time: 0.6025  data: 0.0004  max mem: 55573
Epoch: [216]  [120/156]  eta: 0:00:24  lr: 0.001622  min_lr: 0.001622  loss: 3.4004 (3.0658)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [216]  [130/156]  eta: 0:00:17  lr: 0.001620  min_lr: 0.001620  loss: 3.2722 (3.0725)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0010  max mem: 55573
Epoch: [216]  [140/156]  eta: 0:00:10  lr: 0.001617  min_lr: 0.001617  loss: 3.3209 (3.0755)  weight_decay: 0.0500 (0.0500)  time: 0.5866  data: 0.0009  max mem: 55573
Epoch: [216]  [150/156]  eta: 0:00:04  lr: 0.001615  min_lr: 0.001615  loss: 3.3481 (3.0819)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [216]  [155/156]  eta: 0:00:00  lr: 0.001614  min_lr: 0.001614  loss: 3.3474 (3.0902)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [216] Total time: 0:01:45 (0.6737 s / it)
Averaged stats: lr: 0.001614  min_lr: 0.001614  loss: 3.3474 (3.0937)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 1.0484 (1.0484)  acc1: 81.7708 (81.7708)  acc5: 96.7448 (96.7448)  time: 7.0718  data: 6.8343  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.1035 (1.1054)  acc1: 79.4271 (78.0800)  acc5: 96.2264 (95.1680)  time: 1.5608  data: 1.3670  max mem: 55573
Test: Total time: 0:00:08 (1.6048 s / it)
* Acc@1 79.098 Acc@5 94.884 loss 1.082
Accuracy of the model on the 50000 test images: 79.1%
Max accuracy: 79.10%
Test:  [0/5]  eta: 0:00:33  loss: 1.0542 (1.0542)  acc1: 73.4375 (73.4375)  acc5: 92.1875 (92.1875)  time: 6.6600  data: 6.4238  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0825 (1.0982)  acc1: 71.3542 (69.2800)  acc5: 92.1875 (89.8240)  time: 1.4774  data: 1.2849  max mem: 55573
Test: Total time: 0:00:07 (1.4976 s / it)
* Acc@1 70.066 Acc@5 90.000 loss 1.120
Accuracy of the model EMA on 50000 test images: 70.1%
Max EMA accuracy: 70.07%
Epoch: [217]  [  0/156]  eta: 0:31:28  lr: 0.001613  min_lr: 0.001613  loss: 2.9173 (2.9173)  weight_decay: 0.0500 (0.0500)  time: 12.1040  data: 9.1032  max mem: 55573
Epoch: [217]  [ 10/156]  eta: 0:04:09  lr: 0.001611  min_lr: 0.001611  loss: 3.3209 (3.2667)  weight_decay: 0.0500 (0.0500)  time: 1.7105  data: 0.9016  max mem: 55573
Epoch: [217]  [ 20/156]  eta: 0:02:40  lr: 0.001609  min_lr: 0.001609  loss: 3.3010 (3.2396)  weight_decay: 0.0500 (0.0500)  time: 0.6301  data: 0.0410  max mem: 55573
Epoch: [217]  [ 30/156]  eta: 0:02:04  lr: 0.001607  min_lr: 0.001607  loss: 3.2626 (3.2161)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [217]  [ 40/156]  eta: 0:01:43  lr: 0.001604  min_lr: 0.001604  loss: 3.1680 (3.1901)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [217]  [ 50/156]  eta: 0:01:28  lr: 0.001602  min_lr: 0.001602  loss: 3.2731 (3.1772)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0007  max mem: 55573
Epoch: [217]  [ 60/156]  eta: 0:01:16  lr: 0.001600  min_lr: 0.001600  loss: 3.1267 (3.1162)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0007  max mem: 55573
Epoch: [217]  [ 70/156]  eta: 0:01:05  lr: 0.001597  min_lr: 0.001597  loss: 2.8833 (3.0758)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [217]  [ 80/156]  eta: 0:00:56  lr: 0.001595  min_lr: 0.001595  loss: 3.0442 (3.0974)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [217]  [ 90/156]  eta: 0:00:47  lr: 0.001593  min_lr: 0.001593  loss: 3.2931 (3.0884)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [217]  [100/156]  eta: 0:00:39  lr: 0.001590  min_lr: 0.001590  loss: 2.7470 (3.0515)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [217]  [110/156]  eta: 0:00:32  lr: 0.001588  min_lr: 0.001588  loss: 2.7427 (3.0537)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [217]  [120/156]  eta: 0:00:24  lr: 0.001586  min_lr: 0.001586  loss: 3.0582 (3.0505)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [217]  [130/156]  eta: 0:00:17  lr: 0.001584  min_lr: 0.001584  loss: 3.1775 (3.0656)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0015  max mem: 55573
Epoch: [217]  [140/156]  eta: 0:00:10  lr: 0.001581  min_lr: 0.001581  loss: 3.2510 (3.0581)  weight_decay: 0.0500 (0.0500)  time: 0.5874  data: 0.0014  max mem: 55573
Epoch: [217]  [150/156]  eta: 0:00:04  lr: 0.001579  min_lr: 0.001579  loss: 3.2061 (3.0598)  weight_decay: 0.0500 (0.0500)  time: 0.5834  data: 0.0002  max mem: 55573
Epoch: [217]  [155/156]  eta: 0:00:00  lr: 0.001578  min_lr: 0.001578  loss: 3.2061 (3.0627)  weight_decay: 0.0500 (0.0500)  time: 0.5833  data: 0.0001  max mem: 55573
Epoch: [217] Total time: 0:01:45 (0.6750 s / it)
Averaged stats: lr: 0.001578  min_lr: 0.001578  loss: 3.2061 (3.1009)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.8942 (0.8942)  acc1: 81.7708 (81.7708)  acc5: 96.3542 (96.3542)  time: 6.7488  data: 6.5114  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9953 (1.0009)  acc1: 80.8594 (78.1120)  acc5: 95.8333 (94.4640)  time: 1.4953  data: 1.3024  max mem: 55573
Test: Total time: 0:00:07 (1.5144 s / it)
* Acc@1 78.890 Acc@5 94.752 loss 0.986
Accuracy of the model on the 50000 test images: 78.9%
Max accuracy: 79.10%
Test:  [0/5]  eta: 0:00:34  loss: 1.0521 (1.0521)  acc1: 73.4375 (73.4375)  acc5: 92.1875 (92.1875)  time: 6.8657  data: 6.6295  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0804 (1.0961)  acc1: 71.3542 (69.3440)  acc5: 92.1875 (89.8560)  time: 1.5183  data: 1.3260  max mem: 55573
Test: Total time: 0:00:07 (1.5358 s / it)
* Acc@1 70.124 Acc@5 90.054 loss 1.118
Accuracy of the model EMA on 50000 test images: 70.1%
Max EMA accuracy: 70.12%
Epoch: [218]  [  0/156]  eta: 0:36:41  lr: 0.001578  min_lr: 0.001578  loss: 2.4926 (2.4926)  weight_decay: 0.0500 (0.0500)  time: 14.1109  data: 9.5639  max mem: 55573
Epoch: [218]  [ 10/156]  eta: 0:04:25  lr: 0.001575  min_lr: 0.001575  loss: 3.2472 (3.0367)  weight_decay: 0.0500 (0.0500)  time: 1.8205  data: 0.8699  max mem: 55573
Epoch: [218]  [ 20/156]  eta: 0:02:47  lr: 0.001573  min_lr: 0.001573  loss: 3.1608 (3.0535)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [218]  [ 30/156]  eta: 0:02:09  lr: 0.001571  min_lr: 0.001571  loss: 3.1470 (3.0806)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [218]  [ 40/156]  eta: 0:01:46  lr: 0.001568  min_lr: 0.001568  loss: 3.2326 (3.0822)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0005  max mem: 55573
Epoch: [218]  [ 50/156]  eta: 0:01:30  lr: 0.001566  min_lr: 0.001566  loss: 3.1881 (3.0642)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [218]  [ 60/156]  eta: 0:01:17  lr: 0.001564  min_lr: 0.001564  loss: 3.1698 (3.0881)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0004  max mem: 55573
Epoch: [218]  [ 70/156]  eta: 0:01:07  lr: 0.001562  min_lr: 0.001562  loss: 3.2108 (3.0966)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [218]  [ 80/156]  eta: 0:00:57  lr: 0.001559  min_lr: 0.001559  loss: 2.9678 (3.0782)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [218]  [ 90/156]  eta: 0:00:48  lr: 0.001557  min_lr: 0.001557  loss: 3.1357 (3.0962)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0005  max mem: 55573
Epoch: [218]  [100/156]  eta: 0:00:40  lr: 0.001555  min_lr: 0.001555  loss: 3.0928 (3.0710)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0007  max mem: 55573
Epoch: [218]  [110/156]  eta: 0:00:32  lr: 0.001553  min_lr: 0.001553  loss: 3.0385 (3.0724)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0007  max mem: 55573
Epoch: [218]  [120/156]  eta: 0:00:25  lr: 0.001550  min_lr: 0.001550  loss: 3.1632 (3.0741)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [218]  [130/156]  eta: 0:00:18  lr: 0.001548  min_lr: 0.001548  loss: 3.2455 (3.0906)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0017  max mem: 55573
Epoch: [218]  [140/156]  eta: 0:00:10  lr: 0.001546  min_lr: 0.001546  loss: 3.1546 (3.0755)  weight_decay: 0.0500 (0.0500)  time: 0.5860  data: 0.0016  max mem: 55573
Epoch: [218]  [150/156]  eta: 0:00:04  lr: 0.001543  min_lr: 0.001543  loss: 3.0195 (3.0836)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [218]  [155/156]  eta: 0:00:00  lr: 0.001542  min_lr: 0.001542  loss: 3.0396 (3.0868)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [218] Total time: 0:01:46 (0.6828 s / it)
Averaged stats: lr: 0.001542  min_lr: 0.001542  loss: 3.0396 (3.0910)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.9451 (0.9451)  acc1: 83.0729 (83.0729)  acc5: 96.7448 (96.7448)  time: 6.7343  data: 6.4972  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0424 (1.0551)  acc1: 79.5573 (78.2080)  acc5: 96.2264 (94.7520)  time: 1.4927  data: 1.2996  max mem: 55573
Test: Total time: 0:00:07 (1.5245 s / it)
* Acc@1 79.088 Acc@5 94.710 loss 1.049
Accuracy of the model on the 50000 test images: 79.1%
Max accuracy: 79.10%
Test:  [0/5]  eta: 0:00:36  loss: 1.0501 (1.0501)  acc1: 73.5677 (73.5677)  acc5: 92.1875 (92.1875)  time: 7.3706  data: 7.1345  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0784 (1.0942)  acc1: 71.3542 (69.3760)  acc5: 92.1875 (89.8880)  time: 1.6195  data: 1.4270  max mem: 55573
Test: Total time: 0:00:08 (1.6351 s / it)
* Acc@1 70.186 Acc@5 90.096 loss 1.116
Accuracy of the model EMA on 50000 test images: 70.2%
Max EMA accuracy: 70.19%
Epoch: [219]  [  0/156]  eta: 0:31:28  lr: 0.001542  min_lr: 0.001542  loss: 3.4998 (3.4998)  weight_decay: 0.0500 (0.0500)  time: 12.1065  data: 8.2062  max mem: 55573
Epoch: [219]  [ 10/156]  eta: 0:04:16  lr: 0.001540  min_lr: 0.001540  loss: 2.9813 (2.9555)  weight_decay: 0.0500 (0.0500)  time: 1.7572  data: 0.8639  max mem: 55573
Epoch: [219]  [ 20/156]  eta: 0:02:43  lr: 0.001538  min_lr: 0.001538  loss: 3.1987 (3.0181)  weight_decay: 0.0500 (0.0500)  time: 0.6561  data: 0.0650  max mem: 55573
Epoch: [219]  [ 30/156]  eta: 0:02:06  lr: 0.001535  min_lr: 0.001535  loss: 3.1987 (3.0151)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [219]  [ 40/156]  eta: 0:01:44  lr: 0.001533  min_lr: 0.001533  loss: 3.0810 (2.9436)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [219]  [ 50/156]  eta: 0:01:29  lr: 0.001531  min_lr: 0.001531  loss: 2.7775 (2.9590)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [219]  [ 60/156]  eta: 0:01:16  lr: 0.001529  min_lr: 0.001529  loss: 2.7869 (2.9537)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [219]  [ 70/156]  eta: 0:01:06  lr: 0.001526  min_lr: 0.001526  loss: 2.9095 (2.9490)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [219]  [ 80/156]  eta: 0:00:56  lr: 0.001524  min_lr: 0.001524  loss: 3.3059 (2.9711)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [219]  [ 90/156]  eta: 0:00:48  lr: 0.001522  min_lr: 0.001522  loss: 3.2101 (2.9744)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [219]  [100/156]  eta: 0:00:40  lr: 0.001519  min_lr: 0.001519  loss: 3.1955 (2.9901)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [219]  [110/156]  eta: 0:00:32  lr: 0.001517  min_lr: 0.001517  loss: 3.2248 (3.0062)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [219]  [120/156]  eta: 0:00:25  lr: 0.001515  min_lr: 0.001515  loss: 3.2248 (3.0185)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [219]  [130/156]  eta: 0:00:17  lr: 0.001513  min_lr: 0.001513  loss: 3.1717 (3.0236)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0011  max mem: 55573
Epoch: [219]  [140/156]  eta: 0:00:10  lr: 0.001510  min_lr: 0.001510  loss: 3.1013 (3.0370)  weight_decay: 0.0500 (0.0500)  time: 0.5873  data: 0.0010  max mem: 55573
Epoch: [219]  [150/156]  eta: 0:00:04  lr: 0.001508  min_lr: 0.001508  loss: 3.1013 (3.0368)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [219]  [155/156]  eta: 0:00:00  lr: 0.001507  min_lr: 0.001507  loss: 3.3211 (3.0485)  weight_decay: 0.0500 (0.0500)  time: 0.5829  data: 0.0001  max mem: 55573
Epoch: [219] Total time: 0:01:45 (0.6783 s / it)
Averaged stats: lr: 0.001507  min_lr: 0.001507  loss: 3.3211 (3.0949)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.9831 (0.9831)  acc1: 81.9010 (81.9010)  acc5: 96.7448 (96.7448)  time: 7.1023  data: 6.8649  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0688 (1.0889)  acc1: 79.8177 (77.9840)  acc5: 95.7031 (94.6880)  time: 1.5666  data: 1.3731  max mem: 55573
Test: Total time: 0:00:08 (1.6180 s / it)
* Acc@1 78.978 Acc@5 94.882 loss 1.068
Accuracy of the model on the 50000 test images: 79.0%
Max accuracy: 79.10%
Test:  [0/5]  eta: 0:00:35  loss: 1.0484 (1.0484)  acc1: 73.5677 (73.5677)  acc5: 92.1875 (92.1875)  time: 7.1107  data: 6.8745  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0767 (1.0926)  acc1: 71.3542 (69.3760)  acc5: 92.1875 (89.9200)  time: 1.5675  data: 1.3750  max mem: 55573
Test: Total time: 0:00:07 (1.5904 s / it)
* Acc@1 70.276 Acc@5 90.134 loss 1.114
Accuracy of the model EMA on 50000 test images: 70.3%
Max EMA accuracy: 70.28%
Epoch: [220]  [  0/156]  eta: 0:35:38  lr: 0.001507  min_lr: 0.001507  loss: 3.3132 (3.3132)  weight_decay: 0.0500 (0.0500)  time: 13.7105  data: 13.0661  max mem: 55573
Epoch: [220]  [ 10/156]  eta: 0:04:20  lr: 0.001505  min_lr: 0.001505  loss: 3.3232 (3.2094)  weight_decay: 0.0500 (0.0500)  time: 1.7850  data: 1.1881  max mem: 55573
Epoch: [220]  [ 20/156]  eta: 0:02:45  lr: 0.001502  min_lr: 0.001502  loss: 3.2477 (3.0088)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [220]  [ 30/156]  eta: 0:02:07  lr: 0.001500  min_lr: 0.001500  loss: 2.8550 (3.0047)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [220]  [ 40/156]  eta: 0:01:45  lr: 0.001498  min_lr: 0.001498  loss: 3.2623 (3.0668)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [220]  [ 50/156]  eta: 0:01:30  lr: 0.001496  min_lr: 0.001496  loss: 3.3678 (3.1127)  weight_decay: 0.0500 (0.0500)  time: 0.5937  data: 0.0004  max mem: 55573
Epoch: [220]  [ 60/156]  eta: 0:01:17  lr: 0.001493  min_lr: 0.001493  loss: 3.2868 (3.1285)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0005  max mem: 55573
Epoch: [220]  [ 70/156]  eta: 0:01:06  lr: 0.001491  min_lr: 0.001491  loss: 3.1371 (3.1170)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [220]  [ 80/156]  eta: 0:00:57  lr: 0.001489  min_lr: 0.001489  loss: 3.0428 (3.1251)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [220]  [ 90/156]  eta: 0:00:48  lr: 0.001487  min_lr: 0.001487  loss: 3.2204 (3.1360)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [220]  [100/156]  eta: 0:00:40  lr: 0.001484  min_lr: 0.001484  loss: 3.1594 (3.1136)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [220]  [110/156]  eta: 0:00:32  lr: 0.001482  min_lr: 0.001482  loss: 2.9661 (3.1010)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [220]  [120/156]  eta: 0:00:25  lr: 0.001480  min_lr: 0.001480  loss: 3.0293 (3.1001)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [220]  [130/156]  eta: 0:00:17  lr: 0.001478  min_lr: 0.001478  loss: 3.2821 (3.1044)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0011  max mem: 55573
Epoch: [220]  [140/156]  eta: 0:00:10  lr: 0.001475  min_lr: 0.001475  loss: 3.0675 (3.0942)  weight_decay: 0.0500 (0.0500)  time: 0.5853  data: 0.0009  max mem: 55573
Epoch: [220]  [150/156]  eta: 0:00:04  lr: 0.001473  min_lr: 0.001473  loss: 2.9366 (3.0823)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [220]  [155/156]  eta: 0:00:00  lr: 0.001472  min_lr: 0.001472  loss: 3.0675 (3.0907)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [220] Total time: 0:01:46 (0.6808 s / it)
Averaged stats: lr: 0.001472  min_lr: 0.001472  loss: 3.0675 (3.0775)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.8588 (0.8588)  acc1: 81.9010 (81.9010)  acc5: 97.0052 (97.0052)  time: 6.7872  data: 6.5502  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9503 (0.9631)  acc1: 79.2969 (78.2080)  acc5: 95.9635 (94.9760)  time: 1.5040  data: 1.3101  max mem: 55573
Test: Total time: 0:00:07 (1.5509 s / it)
* Acc@1 79.362 Acc@5 94.986 loss 0.950
Accuracy of the model on the 50000 test images: 79.4%
Max accuracy: 79.36%
Test:  [0/5]  eta: 0:00:34  loss: 1.0465 (1.0465)  acc1: 73.5677 (73.5677)  acc5: 92.1875 (92.1875)  time: 6.8669  data: 6.6308  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0747 (1.0908)  acc1: 71.3542 (69.3440)  acc5: 92.1875 (89.9200)  time: 1.5187  data: 1.3263  max mem: 55573
Test: Total time: 0:00:07 (1.5362 s / it)
* Acc@1 70.312 Acc@5 90.154 loss 1.113
Accuracy of the model EMA on 50000 test images: 70.3%
Max EMA accuracy: 70.31%
Epoch: [221]  [  0/156]  eta: 0:32:15  lr: 0.001472  min_lr: 0.001472  loss: 3.0050 (3.0050)  weight_decay: 0.0500 (0.0500)  time: 12.4081  data: 10.9137  max mem: 55573
Epoch: [221]  [ 10/156]  eta: 0:04:05  lr: 0.001470  min_lr: 0.001470  loss: 3.3590 (3.2396)  weight_decay: 0.0500 (0.0500)  time: 1.6841  data: 0.9926  max mem: 55573
Epoch: [221]  [ 20/156]  eta: 0:02:38  lr: 0.001467  min_lr: 0.001467  loss: 3.2841 (3.1706)  weight_decay: 0.0500 (0.0500)  time: 0.6014  data: 0.0004  max mem: 55573
Epoch: [221]  [ 30/156]  eta: 0:02:03  lr: 0.001465  min_lr: 0.001465  loss: 3.2090 (3.1463)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [221]  [ 40/156]  eta: 0:01:42  lr: 0.001463  min_lr: 0.001463  loss: 3.1260 (3.1055)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [221]  [ 50/156]  eta: 0:01:27  lr: 0.001461  min_lr: 0.001461  loss: 3.0490 (3.0919)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [221]  [ 60/156]  eta: 0:01:15  lr: 0.001459  min_lr: 0.001459  loss: 3.0490 (3.0787)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [221]  [ 70/156]  eta: 0:01:05  lr: 0.001456  min_lr: 0.001456  loss: 2.9179 (3.0591)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [221]  [ 80/156]  eta: 0:00:56  lr: 0.001454  min_lr: 0.001454  loss: 3.2821 (3.0723)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0005  max mem: 55573
Epoch: [221]  [ 90/156]  eta: 0:00:47  lr: 0.001452  min_lr: 0.001452  loss: 3.2060 (3.0731)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [221]  [100/156]  eta: 0:00:39  lr: 0.001450  min_lr: 0.001450  loss: 3.1849 (3.0763)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [221]  [110/156]  eta: 0:00:32  lr: 0.001447  min_lr: 0.001447  loss: 3.1874 (3.0836)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [221]  [120/156]  eta: 0:00:24  lr: 0.001445  min_lr: 0.001445  loss: 3.1119 (3.0664)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [221]  [130/156]  eta: 0:00:17  lr: 0.001443  min_lr: 0.001443  loss: 3.0463 (3.0730)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0011  max mem: 55573
Epoch: [221]  [140/156]  eta: 0:00:10  lr: 0.001441  min_lr: 0.001441  loss: 3.0765 (3.0510)  weight_decay: 0.0500 (0.0500)  time: 0.5849  data: 0.0009  max mem: 55573
Epoch: [221]  [150/156]  eta: 0:00:04  lr: 0.001439  min_lr: 0.001439  loss: 3.0765 (3.0417)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [221]  [155/156]  eta: 0:00:00  lr: 0.001438  min_lr: 0.001438  loss: 3.2079 (3.0461)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [221] Total time: 0:01:44 (0.6728 s / it)
Averaged stats: lr: 0.001438  min_lr: 0.001438  loss: 3.2079 (3.0881)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8578 (0.8578)  acc1: 82.2917 (82.2917)  acc5: 96.3542 (96.3542)  time: 6.9172  data: 6.6799  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9290 (0.9376)  acc1: 81.2500 (78.4640)  acc5: 96.3542 (95.0400)  time: 1.5300  data: 1.3361  max mem: 55573
Test: Total time: 0:00:07 (1.5730 s / it)
* Acc@1 79.484 Acc@5 95.018 loss 0.919
Accuracy of the model on the 50000 test images: 79.5%
Max accuracy: 79.48%
Test:  [0/5]  eta: 0:00:34  loss: 1.0451 (1.0451)  acc1: 73.5677 (73.5677)  acc5: 92.0573 (92.0573)  time: 6.9333  data: 6.6972  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0733 (1.0894)  acc1: 71.3542 (69.3440)  acc5: 92.0573 (89.9200)  time: 1.5320  data: 1.3395  max mem: 55573
Test: Total time: 0:00:07 (1.5491 s / it)
* Acc@1 70.354 Acc@5 90.172 loss 1.111
Accuracy of the model EMA on 50000 test images: 70.4%
Max EMA accuracy: 70.35%
Epoch: [222]  [  0/156]  eta: 0:29:38  lr: 0.001437  min_lr: 0.001437  loss: 3.4519 (3.4519)  weight_decay: 0.0500 (0.0500)  time: 11.4030  data: 10.8178  max mem: 55573
Epoch: [222]  [ 10/156]  eta: 0:04:20  lr: 0.001435  min_lr: 0.001435  loss: 2.8442 (2.9033)  weight_decay: 0.0500 (0.0500)  time: 1.7815  data: 1.1869  max mem: 55573
Epoch: [222]  [ 20/156]  eta: 0:02:45  lr: 0.001433  min_lr: 0.001433  loss: 2.8442 (2.9453)  weight_decay: 0.0500 (0.0500)  time: 0.7044  data: 0.1121  max mem: 55573
Epoch: [222]  [ 30/156]  eta: 0:02:07  lr: 0.001431  min_lr: 0.001431  loss: 3.0235 (2.9628)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [222]  [ 40/156]  eta: 0:01:45  lr: 0.001428  min_lr: 0.001428  loss: 3.1882 (3.0432)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [222]  [ 50/156]  eta: 0:01:29  lr: 0.001426  min_lr: 0.001426  loss: 3.2110 (3.0052)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0004  max mem: 55573
Epoch: [222]  [ 60/156]  eta: 0:01:17  lr: 0.001424  min_lr: 0.001424  loss: 3.2533 (3.0781)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [222]  [ 70/156]  eta: 0:01:06  lr: 0.001422  min_lr: 0.001422  loss: 3.3662 (3.1116)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [222]  [ 80/156]  eta: 0:00:57  lr: 0.001420  min_lr: 0.001420  loss: 3.3522 (3.1463)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [222]  [ 90/156]  eta: 0:00:48  lr: 0.001417  min_lr: 0.001417  loss: 3.2783 (3.1372)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [222]  [100/156]  eta: 0:00:40  lr: 0.001415  min_lr: 0.001415  loss: 3.1423 (3.1371)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [222]  [110/156]  eta: 0:00:32  lr: 0.001413  min_lr: 0.001413  loss: 2.9690 (3.1148)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [222]  [120/156]  eta: 0:00:25  lr: 0.001411  min_lr: 0.001411  loss: 3.0090 (3.1178)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [222]  [130/156]  eta: 0:00:17  lr: 0.001409  min_lr: 0.001409  loss: 3.2359 (3.1222)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0012  max mem: 55573
Epoch: [222]  [140/156]  eta: 0:00:10  lr: 0.001407  min_lr: 0.001407  loss: 3.3590 (3.1319)  weight_decay: 0.0500 (0.0500)  time: 0.5864  data: 0.0011  max mem: 55573
Epoch: [222]  [150/156]  eta: 0:00:04  lr: 0.001404  min_lr: 0.001404  loss: 3.3879 (3.1378)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [222]  [155/156]  eta: 0:00:00  lr: 0.001403  min_lr: 0.001403  loss: 3.1612 (3.1244)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [222] Total time: 0:01:46 (0.6804 s / it)
Averaged stats: lr: 0.001403  min_lr: 0.001403  loss: 3.1612 (3.0765)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.9204 (0.9204)  acc1: 83.4635 (83.4635)  acc5: 96.6146 (96.6146)  time: 6.7965  data: 6.5584  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0008 (0.9924)  acc1: 80.8594 (78.8160)  acc5: 96.3542 (95.0400)  time: 1.5057  data: 1.3118  max mem: 55573
Test: Total time: 0:00:07 (1.5521 s / it)
* Acc@1 79.414 Acc@5 94.944 loss 0.991
Accuracy of the model on the 50000 test images: 79.4%
Max accuracy: 79.48%
Test:  [0/5]  eta: 0:00:34  loss: 1.0431 (1.0431)  acc1: 73.8281 (73.8281)  acc5: 92.1875 (92.1875)  time: 6.9747  data: 6.7386  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0714 (1.0877)  acc1: 71.3542 (69.5040)  acc5: 92.1875 (89.9200)  time: 1.5406  data: 1.3478  max mem: 55573
Test: Total time: 0:00:07 (1.5630 s / it)
* Acc@1 70.426 Acc@5 90.202 loss 1.109
Accuracy of the model EMA on 50000 test images: 70.4%
Max EMA accuracy: 70.43%
Epoch: [223]  [  0/156]  eta: 0:36:54  lr: 0.001403  min_lr: 0.001403  loss: 3.1219 (3.1219)  weight_decay: 0.0500 (0.0500)  time: 14.1944  data: 7.4531  max mem: 55573
Epoch: [223]  [ 10/156]  eta: 0:04:27  lr: 0.001401  min_lr: 0.001401  loss: 3.1041 (2.9581)  weight_decay: 0.0500 (0.0500)  time: 1.8297  data: 0.6782  max mem: 55573
Epoch: [223]  [ 20/156]  eta: 0:02:48  lr: 0.001399  min_lr: 0.001399  loss: 3.2144 (3.0248)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0006  max mem: 55573
Epoch: [223]  [ 30/156]  eta: 0:02:09  lr: 0.001396  min_lr: 0.001396  loss: 3.2817 (3.0663)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [223]  [ 40/156]  eta: 0:01:47  lr: 0.001394  min_lr: 0.001394  loss: 3.0579 (3.0489)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [223]  [ 50/156]  eta: 0:01:30  lr: 0.001392  min_lr: 0.001392  loss: 3.0579 (3.0716)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [223]  [ 60/156]  eta: 0:01:18  lr: 0.001390  min_lr: 0.001390  loss: 3.2482 (3.0608)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [223]  [ 70/156]  eta: 0:01:07  lr: 0.001388  min_lr: 0.001388  loss: 3.1407 (3.0718)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [223]  [ 80/156]  eta: 0:00:57  lr: 0.001386  min_lr: 0.001386  loss: 3.2461 (3.0884)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [223]  [ 90/156]  eta: 0:00:48  lr: 0.001383  min_lr: 0.001383  loss: 3.3058 (3.1108)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [223]  [100/156]  eta: 0:00:40  lr: 0.001381  min_lr: 0.001381  loss: 3.2627 (3.0787)  weight_decay: 0.0500 (0.0500)  time: 0.5882  data: 0.0004  max mem: 55573
Epoch: [223]  [110/156]  eta: 0:00:32  lr: 0.001379  min_lr: 0.001379  loss: 2.9406 (3.0698)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [223]  [120/156]  eta: 0:00:25  lr: 0.001377  min_lr: 0.001377  loss: 3.1233 (3.0809)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [223]  [130/156]  eta: 0:00:18  lr: 0.001375  min_lr: 0.001375  loss: 3.2467 (3.0865)  weight_decay: 0.0500 (0.0500)  time: 0.5879  data: 0.0011  max mem: 55573
Epoch: [223]  [140/156]  eta: 0:00:10  lr: 0.001373  min_lr: 0.001373  loss: 3.0770 (3.0908)  weight_decay: 0.0500 (0.0500)  time: 0.5849  data: 0.0009  max mem: 55573
Epoch: [223]  [150/156]  eta: 0:00:04  lr: 0.001370  min_lr: 0.001370  loss: 3.1268 (3.0883)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [223]  [155/156]  eta: 0:00:00  lr: 0.001369  min_lr: 0.001369  loss: 3.0422 (3.0906)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [223] Total time: 0:01:46 (0.6834 s / it)
Averaged stats: lr: 0.001369  min_lr: 0.001369  loss: 3.0422 (3.0706)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9181 (0.9181)  acc1: 82.5521 (82.5521)  acc5: 96.3542 (96.3542)  time: 6.9728  data: 6.7353  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9862 (0.9869)  acc1: 80.0781 (78.4320)  acc5: 96.3542 (95.1360)  time: 1.5411  data: 1.3471  max mem: 55573
Test: Total time: 0:00:07 (1.5791 s / it)
* Acc@1 79.408 Acc@5 94.888 loss 0.982
Accuracy of the model on the 50000 test images: 79.4%
Max accuracy: 79.48%
Test:  [0/5]  eta: 0:00:35  loss: 1.0416 (1.0416)  acc1: 73.6979 (73.6979)  acc5: 92.1875 (92.1875)  time: 7.0293  data: 6.7932  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0700 (1.0863)  acc1: 71.4844 (69.4720)  acc5: 92.1875 (90.0160)  time: 1.5512  data: 1.3588  max mem: 55573
Test: Total time: 0:00:07 (1.5708 s / it)
* Acc@1 70.462 Acc@5 90.234 loss 1.107
Accuracy of the model EMA on 50000 test images: 70.5%
Max EMA accuracy: 70.46%
Epoch: [224]  [  0/156]  eta: 0:37:03  lr: 0.001369  min_lr: 0.001369  loss: 3.2471 (3.2471)  weight_decay: 0.0500 (0.0500)  time: 14.2553  data: 13.6460  max mem: 55573
Epoch: [224]  [ 10/156]  eta: 0:04:27  lr: 0.001367  min_lr: 0.001367  loss: 3.3013 (3.2853)  weight_decay: 0.0500 (0.0500)  time: 1.8322  data: 1.2409  max mem: 55573
Epoch: [224]  [ 20/156]  eta: 0:02:48  lr: 0.001365  min_lr: 0.001365  loss: 3.2880 (3.2296)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0003  max mem: 55573
Epoch: [224]  [ 30/156]  eta: 0:02:09  lr: 0.001363  min_lr: 0.001363  loss: 3.2424 (3.1973)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [224]  [ 40/156]  eta: 0:01:47  lr: 0.001360  min_lr: 0.001360  loss: 2.9306 (3.0832)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [224]  [ 50/156]  eta: 0:01:30  lr: 0.001358  min_lr: 0.001358  loss: 2.7687 (3.0651)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [224]  [ 60/156]  eta: 0:01:18  lr: 0.001356  min_lr: 0.001356  loss: 3.1587 (3.0831)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [224]  [ 70/156]  eta: 0:01:07  lr: 0.001354  min_lr: 0.001354  loss: 3.1459 (3.0551)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [224]  [ 80/156]  eta: 0:00:57  lr: 0.001352  min_lr: 0.001352  loss: 2.7833 (3.0464)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [224]  [ 90/156]  eta: 0:00:48  lr: 0.001350  min_lr: 0.001350  loss: 3.0762 (3.0444)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [224]  [100/156]  eta: 0:00:40  lr: 0.001347  min_lr: 0.001347  loss: 3.1884 (3.0404)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0004  max mem: 55573
Epoch: [224]  [110/156]  eta: 0:00:32  lr: 0.001345  min_lr: 0.001345  loss: 3.1281 (3.0455)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0004  max mem: 55573
Epoch: [224]  [120/156]  eta: 0:00:25  lr: 0.001343  min_lr: 0.001343  loss: 3.0520 (3.0404)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [224]  [130/156]  eta: 0:00:18  lr: 0.001341  min_lr: 0.001341  loss: 3.1804 (3.0539)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0010  max mem: 55573
Epoch: [224]  [140/156]  eta: 0:00:10  lr: 0.001339  min_lr: 0.001339  loss: 3.0952 (3.0371)  weight_decay: 0.0500 (0.0500)  time: 0.5863  data: 0.0008  max mem: 55573
Epoch: [224]  [150/156]  eta: 0:00:04  lr: 0.001337  min_lr: 0.001337  loss: 3.0372 (3.0520)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [224]  [155/156]  eta: 0:00:00  lr: 0.001336  min_lr: 0.001336  loss: 3.2438 (3.0449)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [224] Total time: 0:01:46 (0.6839 s / it)
Averaged stats: lr: 0.001336  min_lr: 0.001336  loss: 3.2438 (3.0661)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.9301 (0.9301)  acc1: 82.6823 (82.6823)  acc5: 96.0938 (96.0938)  time: 7.1031  data: 6.8657  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9707 (0.9777)  acc1: 81.2500 (78.8480)  acc5: 96.0938 (94.9120)  time: 1.5672  data: 1.3732  max mem: 55573
Test: Total time: 0:00:08 (1.6143 s / it)
* Acc@1 79.518 Acc@5 94.962 loss 0.966
Accuracy of the model on the 50000 test images: 79.5%
Max accuracy: 79.52%
Test:  [0/5]  eta: 0:00:34  loss: 1.0401 (1.0401)  acc1: 73.8281 (73.8281)  acc5: 92.1875 (92.1875)  time: 6.8706  data: 6.6343  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0685 (1.0850)  acc1: 71.3542 (69.4720)  acc5: 92.1875 (90.0800)  time: 1.5195  data: 1.3270  max mem: 55573
Test: Total time: 0:00:07 (1.5392 s / it)
* Acc@1 70.498 Acc@5 90.268 loss 1.106
Accuracy of the model EMA on 50000 test images: 70.5%
Max EMA accuracy: 70.50%
Epoch: [225]  [  0/156]  eta: 0:35:30  lr: 0.001335  min_lr: 0.001335  loss: 3.3241 (3.3241)  weight_decay: 0.0500 (0.0500)  time: 13.6564  data: 7.5853  max mem: 55573
Epoch: [225]  [ 10/156]  eta: 0:04:20  lr: 0.001333  min_lr: 0.001333  loss: 3.1505 (2.9794)  weight_decay: 0.0500 (0.0500)  time: 1.7873  data: 0.6900  max mem: 55573
Epoch: [225]  [ 20/156]  eta: 0:02:45  lr: 0.001331  min_lr: 0.001331  loss: 3.1901 (3.1242)  weight_decay: 0.0500 (0.0500)  time: 0.5973  data: 0.0005  max mem: 55573
Epoch: [225]  [ 30/156]  eta: 0:02:08  lr: 0.001329  min_lr: 0.001329  loss: 3.3561 (3.1914)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0005  max mem: 55573
Epoch: [225]  [ 40/156]  eta: 0:01:45  lr: 0.001327  min_lr: 0.001327  loss: 3.3561 (3.2060)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [225]  [ 50/156]  eta: 0:01:30  lr: 0.001325  min_lr: 0.001325  loss: 3.2305 (3.1625)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [225]  [ 60/156]  eta: 0:01:17  lr: 0.001323  min_lr: 0.001323  loss: 3.2365 (3.1558)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [225]  [ 70/156]  eta: 0:01:06  lr: 0.001320  min_lr: 0.001320  loss: 3.3337 (3.1561)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [225]  [ 80/156]  eta: 0:00:57  lr: 0.001318  min_lr: 0.001318  loss: 3.3068 (3.1695)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [225]  [ 90/156]  eta: 0:00:48  lr: 0.001316  min_lr: 0.001316  loss: 3.3269 (3.1836)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [225]  [100/156]  eta: 0:00:40  lr: 0.001314  min_lr: 0.001314  loss: 3.3269 (3.1910)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [225]  [110/156]  eta: 0:00:32  lr: 0.001312  min_lr: 0.001312  loss: 3.2622 (3.1901)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [225]  [120/156]  eta: 0:00:25  lr: 0.001310  min_lr: 0.001310  loss: 3.2622 (3.1814)  weight_decay: 0.0500 (0.0500)  time: 0.5882  data: 0.0005  max mem: 55573
Epoch: [225]  [130/156]  eta: 0:00:17  lr: 0.001308  min_lr: 0.001308  loss: 3.1079 (3.1802)  weight_decay: 0.0500 (0.0500)  time: 0.5878  data: 0.0010  max mem: 55573
Epoch: [225]  [140/156]  eta: 0:00:10  lr: 0.001306  min_lr: 0.001306  loss: 3.0948 (3.1692)  weight_decay: 0.0500 (0.0500)  time: 0.5845  data: 0.0009  max mem: 55573
Epoch: [225]  [150/156]  eta: 0:00:04  lr: 0.001303  min_lr: 0.001303  loss: 3.3313 (3.1857)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [225]  [155/156]  eta: 0:00:00  lr: 0.001302  min_lr: 0.001302  loss: 3.3361 (3.1848)  weight_decay: 0.0500 (0.0500)  time: 0.5812  data: 0.0001  max mem: 55573
Epoch: [225] Total time: 0:01:46 (0.6804 s / it)
Averaged stats: lr: 0.001302  min_lr: 0.001302  loss: 3.3361 (3.0759)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9752 (0.9752)  acc1: 82.2917 (82.2917)  acc5: 96.7448 (96.7448)  time: 6.9131  data: 6.6761  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9865 (1.0339)  acc1: 81.5104 (78.9120)  acc5: 96.2264 (95.1680)  time: 1.5289  data: 1.3353  max mem: 55573
Test: Total time: 0:00:07 (1.5752 s / it)
* Acc@1 79.286 Acc@5 94.990 loss 1.014
Accuracy of the model on the 50000 test images: 79.3%
Max accuracy: 79.52%
Test:  [0/5]  eta: 0:00:35  loss: 1.0388 (1.0388)  acc1: 73.8281 (73.8281)  acc5: 92.0573 (92.0573)  time: 7.1785  data: 6.9424  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0670 (1.0837)  acc1: 71.6146 (69.4720)  acc5: 92.0573 (90.0480)  time: 1.5810  data: 1.3886  max mem: 55573
Test: Total time: 0:00:08 (1.6007 s / it)
* Acc@1 70.554 Acc@5 90.284 loss 1.105
Accuracy of the model EMA on 50000 test images: 70.6%
Max EMA accuracy: 70.55%
Epoch: [226]  [  0/156]  eta: 0:32:34  lr: 0.001302  min_lr: 0.001302  loss: 3.4337 (3.4337)  weight_decay: 0.0500 (0.0500)  time: 12.5261  data: 8.3005  max mem: 55573
Epoch: [226]  [ 10/156]  eta: 0:04:05  lr: 0.001300  min_lr: 0.001300  loss: 3.1413 (3.0090)  weight_decay: 0.0500 (0.0500)  time: 1.6790  data: 0.7550  max mem: 55573
Epoch: [226]  [ 20/156]  eta: 0:02:37  lr: 0.001298  min_lr: 0.001298  loss: 3.0162 (2.9641)  weight_decay: 0.0500 (0.0500)  time: 0.5931  data: 0.0004  max mem: 55573
Epoch: [226]  [ 30/156]  eta: 0:02:03  lr: 0.001296  min_lr: 0.001296  loss: 2.9583 (2.9477)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [226]  [ 40/156]  eta: 0:01:42  lr: 0.001294  min_lr: 0.001294  loss: 2.9671 (2.9474)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [226]  [ 50/156]  eta: 0:01:27  lr: 0.001292  min_lr: 0.001292  loss: 2.9671 (2.9600)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [226]  [ 60/156]  eta: 0:01:15  lr: 0.001289  min_lr: 0.001289  loss: 3.0280 (2.9996)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [226]  [ 70/156]  eta: 0:01:05  lr: 0.001287  min_lr: 0.001287  loss: 3.0280 (2.9961)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [226]  [ 80/156]  eta: 0:00:56  lr: 0.001285  min_lr: 0.001285  loss: 3.2521 (3.0028)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0005  max mem: 55573
Epoch: [226]  [ 90/156]  eta: 0:00:47  lr: 0.001283  min_lr: 0.001283  loss: 3.0709 (2.9998)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [226]  [100/156]  eta: 0:00:39  lr: 0.001281  min_lr: 0.001281  loss: 2.8813 (2.9864)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [226]  [110/156]  eta: 0:00:32  lr: 0.001279  min_lr: 0.001279  loss: 2.8813 (2.9860)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [226]  [120/156]  eta: 0:00:24  lr: 0.001277  min_lr: 0.001277  loss: 2.9877 (2.9839)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0005  max mem: 55573
Epoch: [226]  [130/156]  eta: 0:00:17  lr: 0.001275  min_lr: 0.001275  loss: 3.0524 (2.9850)  weight_decay: 0.0500 (0.0500)  time: 0.5934  data: 0.0010  max mem: 55573
Epoch: [226]  [140/156]  eta: 0:00:10  lr: 0.001273  min_lr: 0.001273  loss: 2.7070 (2.9606)  weight_decay: 0.0500 (0.0500)  time: 0.5864  data: 0.0008  max mem: 55573
Epoch: [226]  [150/156]  eta: 0:00:04  lr: 0.001270  min_lr: 0.001270  loss: 2.7070 (2.9644)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [226]  [155/156]  eta: 0:00:00  lr: 0.001269  min_lr: 0.001269  loss: 3.0981 (2.9662)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [226] Total time: 0:01:45 (0.6733 s / it)
Averaged stats: lr: 0.001269  min_lr: 0.001269  loss: 3.0981 (3.0601)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8513 (0.8513)  acc1: 80.9896 (80.9896)  acc5: 97.3958 (97.3958)  time: 6.9844  data: 6.7471  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9092 (0.9486)  acc1: 80.9896 (78.5280)  acc5: 96.2264 (95.3920)  time: 1.5431  data: 1.3495  max mem: 55573
Test: Total time: 0:00:07 (1.5880 s / it)
* Acc@1 79.620 Acc@5 95.104 loss 0.933
Accuracy of the model on the 50000 test images: 79.6%
Max accuracy: 79.62%
Test:  [0/5]  eta: 0:00:34  loss: 1.0373 (1.0373)  acc1: 73.8281 (73.8281)  acc5: 92.0573 (92.0573)  time: 6.8620  data: 6.6262  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0654 (1.0823)  acc1: 71.7448 (69.6000)  acc5: 92.0573 (90.1120)  time: 1.5177  data: 1.3253  max mem: 55573
Test: Total time: 0:00:07 (1.5330 s / it)
* Acc@1 70.616 Acc@5 90.306 loss 1.103
Accuracy of the model EMA on 50000 test images: 70.6%
Max EMA accuracy: 70.62%
Epoch: [227]  [  0/156]  eta: 0:32:46  lr: 0.001269  min_lr: 0.001269  loss: 2.8163 (2.8163)  weight_decay: 0.0500 (0.0500)  time: 12.6084  data: 8.7523  max mem: 55573
Epoch: [227]  [ 10/156]  eta: 0:04:12  lr: 0.001267  min_lr: 0.001267  loss: 3.0555 (3.0883)  weight_decay: 0.0500 (0.0500)  time: 1.7294  data: 0.7961  max mem: 55573
Epoch: [227]  [ 20/156]  eta: 0:02:41  lr: 0.001265  min_lr: 0.001265  loss: 3.1923 (3.1120)  weight_decay: 0.0500 (0.0500)  time: 0.6154  data: 0.0005  max mem: 55573
Epoch: [227]  [ 30/156]  eta: 0:02:05  lr: 0.001263  min_lr: 0.001263  loss: 3.1501 (3.0868)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [227]  [ 40/156]  eta: 0:01:43  lr: 0.001261  min_lr: 0.001261  loss: 3.0520 (3.0497)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [227]  [ 50/156]  eta: 0:01:28  lr: 0.001259  min_lr: 0.001259  loss: 2.9510 (3.0307)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [227]  [ 60/156]  eta: 0:01:16  lr: 0.001257  min_lr: 0.001257  loss: 3.2136 (3.0595)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [227]  [ 70/156]  eta: 0:01:05  lr: 0.001255  min_lr: 0.001255  loss: 3.3157 (3.0850)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [227]  [ 80/156]  eta: 0:00:56  lr: 0.001252  min_lr: 0.001252  loss: 3.2110 (3.0716)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [227]  [ 90/156]  eta: 0:00:48  lr: 0.001250  min_lr: 0.001250  loss: 3.2026 (3.0709)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [227]  [100/156]  eta: 0:00:40  lr: 0.001248  min_lr: 0.001248  loss: 3.2314 (3.0657)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [227]  [110/156]  eta: 0:00:32  lr: 0.001246  min_lr: 0.001246  loss: 3.2205 (3.0602)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [227]  [120/156]  eta: 0:00:25  lr: 0.001244  min_lr: 0.001244  loss: 3.0966 (3.0397)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0005  max mem: 55573
Epoch: [227]  [130/156]  eta: 0:00:17  lr: 0.001242  min_lr: 0.001242  loss: 2.7140 (3.0326)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0010  max mem: 55573
Epoch: [227]  [140/156]  eta: 0:00:10  lr: 0.001240  min_lr: 0.001240  loss: 2.9966 (3.0287)  weight_decay: 0.0500 (0.0500)  time: 0.5862  data: 0.0009  max mem: 55573
Epoch: [227]  [150/156]  eta: 0:00:04  lr: 0.001238  min_lr: 0.001238  loss: 3.0855 (3.0380)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [227]  [155/156]  eta: 0:00:00  lr: 0.001237  min_lr: 0.001237  loss: 3.2267 (3.0457)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [227] Total time: 0:01:45 (0.6768 s / it)
Averaged stats: lr: 0.001237  min_lr: 0.001237  loss: 3.2267 (3.0449)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.9990 (0.9990)  acc1: 81.1198 (81.1198)  acc5: 96.4844 (96.4844)  time: 7.0983  data: 6.8608  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0250 (1.0608)  acc1: 80.8594 (78.5920)  acc5: 94.3396 (94.8480)  time: 1.5660  data: 1.3722  max mem: 55573
Test: Total time: 0:00:08 (1.6094 s / it)
* Acc@1 79.512 Acc@5 95.066 loss 1.041
Accuracy of the model on the 50000 test images: 79.5%
Max accuracy: 79.62%
Test:  [0/5]  eta: 0:00:35  loss: 1.0358 (1.0358)  acc1: 74.0885 (74.0885)  acc5: 92.0573 (92.0573)  time: 7.0268  data: 6.7908  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0638 (1.0809)  acc1: 72.0052 (69.7600)  acc5: 92.0573 (90.0480)  time: 1.5511  data: 1.3584  max mem: 55573
Test: Total time: 0:00:07 (1.5748 s / it)
* Acc@1 70.666 Acc@5 90.320 loss 1.102
Accuracy of the model EMA on 50000 test images: 70.7%
Max EMA accuracy: 70.67%
Epoch: [228]  [  0/156]  eta: 0:31:34  lr: 0.001237  min_lr: 0.001237  loss: 2.5131 (2.5131)  weight_decay: 0.0500 (0.0500)  time: 12.1455  data: 11.5528  max mem: 55573
Epoch: [228]  [ 10/156]  eta: 0:04:07  lr: 0.001235  min_lr: 0.001235  loss: 3.1729 (3.0734)  weight_decay: 0.0500 (0.0500)  time: 1.6980  data: 1.0506  max mem: 55573
Epoch: [228]  [ 20/156]  eta: 0:02:39  lr: 0.001232  min_lr: 0.001232  loss: 3.3168 (3.0791)  weight_decay: 0.0500 (0.0500)  time: 0.6218  data: 0.0004  max mem: 55573
Epoch: [228]  [ 30/156]  eta: 0:02:03  lr: 0.001230  min_lr: 0.001230  loss: 3.3114 (3.1195)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [228]  [ 40/156]  eta: 0:01:42  lr: 0.001228  min_lr: 0.001228  loss: 3.3497 (3.1793)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [228]  [ 50/156]  eta: 0:01:27  lr: 0.001226  min_lr: 0.001226  loss: 3.2987 (3.1422)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [228]  [ 60/156]  eta: 0:01:15  lr: 0.001224  min_lr: 0.001224  loss: 3.2287 (3.1787)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [228]  [ 70/156]  eta: 0:01:05  lr: 0.001222  min_lr: 0.001222  loss: 3.2703 (3.1558)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [228]  [ 80/156]  eta: 0:00:56  lr: 0.001220  min_lr: 0.001220  loss: 2.9027 (3.1039)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [228]  [ 90/156]  eta: 0:00:47  lr: 0.001218  min_lr: 0.001218  loss: 3.1989 (3.1175)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0004  max mem: 55573
Epoch: [228]  [100/156]  eta: 0:00:39  lr: 0.001216  min_lr: 0.001216  loss: 3.1617 (3.0806)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [228]  [110/156]  eta: 0:00:32  lr: 0.001214  min_lr: 0.001214  loss: 2.7729 (3.0710)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [228]  [120/156]  eta: 0:00:24  lr: 0.001212  min_lr: 0.001212  loss: 3.2370 (3.0598)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [228]  [130/156]  eta: 0:00:17  lr: 0.001210  min_lr: 0.001210  loss: 3.2465 (3.0675)  weight_decay: 0.0500 (0.0500)  time: 0.5885  data: 0.0010  max mem: 55573
Epoch: [228]  [140/156]  eta: 0:00:10  lr: 0.001208  min_lr: 0.001208  loss: 3.2872 (3.0655)  weight_decay: 0.0500 (0.0500)  time: 0.5848  data: 0.0008  max mem: 55573
Epoch: [228]  [150/156]  eta: 0:00:04  lr: 0.001206  min_lr: 0.001206  loss: 2.8732 (3.0375)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [228]  [155/156]  eta: 0:00:00  lr: 0.001205  min_lr: 0.001205  loss: 3.0347 (3.0441)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [228] Total time: 0:01:45 (0.6742 s / it)
Averaged stats: lr: 0.001205  min_lr: 0.001205  loss: 3.0347 (3.0475)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8680 (0.8680)  acc1: 82.5521 (82.5521)  acc5: 96.6146 (96.6146)  time: 6.8899  data: 6.6524  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9535 (0.9494)  acc1: 80.4688 (79.2320)  acc5: 95.9635 (94.8800)  time: 1.5243  data: 1.3306  max mem: 55573
Test: Total time: 0:00:07 (1.5614 s / it)
* Acc@1 79.762 Acc@5 95.160 loss 0.932
Accuracy of the model on the 50000 test images: 79.8%
Max accuracy: 79.76%
Test:  [0/5]  eta: 0:00:33  loss: 1.0345 (1.0345)  acc1: 74.0885 (74.0885)  acc5: 91.9271 (91.9271)  time: 6.7594  data: 6.5232  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0625 (1.0797)  acc1: 72.0052 (69.7600)  acc5: 91.9271 (90.0160)  time: 1.4972  data: 1.3048  max mem: 55573
Test: Total time: 0:00:07 (1.5146 s / it)
* Acc@1 70.680 Acc@5 90.332 loss 1.101
Accuracy of the model EMA on 50000 test images: 70.7%
Max EMA accuracy: 70.68%
Epoch: [229]  [  0/156]  eta: 0:31:30  lr: 0.001204  min_lr: 0.001204  loss: 2.5137 (2.5137)  weight_decay: 0.0500 (0.0500)  time: 12.1189  data: 11.5339  max mem: 55573
Epoch: [229]  [ 10/156]  eta: 0:04:02  lr: 0.001202  min_lr: 0.001202  loss: 3.0896 (2.9449)  weight_decay: 0.0500 (0.0500)  time: 1.6592  data: 1.0489  max mem: 55573
Epoch: [229]  [ 20/156]  eta: 0:02:36  lr: 0.001200  min_lr: 0.001200  loss: 3.0896 (2.9743)  weight_decay: 0.0500 (0.0500)  time: 0.6023  data: 0.0004  max mem: 55573
Epoch: [229]  [ 30/156]  eta: 0:02:02  lr: 0.001198  min_lr: 0.001198  loss: 3.2674 (3.0803)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [229]  [ 40/156]  eta: 0:01:41  lr: 0.001196  min_lr: 0.001196  loss: 3.1836 (3.0064)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [229]  [ 50/156]  eta: 0:01:26  lr: 0.001194  min_lr: 0.001194  loss: 3.1144 (2.9876)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0005  max mem: 55573
Epoch: [229]  [ 60/156]  eta: 0:01:15  lr: 0.001192  min_lr: 0.001192  loss: 2.9827 (2.9846)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [229]  [ 70/156]  eta: 0:01:04  lr: 0.001190  min_lr: 0.001190  loss: 2.9855 (2.9972)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [229]  [ 80/156]  eta: 0:00:55  lr: 0.001188  min_lr: 0.001188  loss: 3.2722 (3.0290)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [229]  [ 90/156]  eta: 0:00:47  lr: 0.001186  min_lr: 0.001186  loss: 3.1640 (3.0102)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [229]  [100/156]  eta: 0:00:39  lr: 0.001184  min_lr: 0.001184  loss: 2.9256 (3.0158)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [229]  [110/156]  eta: 0:00:32  lr: 0.001182  min_lr: 0.001182  loss: 3.2222 (3.0250)  weight_decay: 0.0500 (0.0500)  time: 0.5947  data: 0.0004  max mem: 55573
Epoch: [229]  [120/156]  eta: 0:00:24  lr: 0.001180  min_lr: 0.001180  loss: 3.2183 (3.0377)  weight_decay: 0.0500 (0.0500)  time: 0.5942  data: 0.0005  max mem: 55573
Epoch: [229]  [130/156]  eta: 0:00:17  lr: 0.001178  min_lr: 0.001178  loss: 3.0658 (3.0213)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0011  max mem: 55573
Epoch: [229]  [140/156]  eta: 0:00:10  lr: 0.001176  min_lr: 0.001176  loss: 3.1340 (3.0235)  weight_decay: 0.0500 (0.0500)  time: 0.5863  data: 0.0009  max mem: 55573
Epoch: [229]  [150/156]  eta: 0:00:04  lr: 0.001174  min_lr: 0.001174  loss: 3.1340 (3.0159)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0002  max mem: 55573
Epoch: [229]  [155/156]  eta: 0:00:00  lr: 0.001173  min_lr: 0.001173  loss: 3.1264 (3.0129)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0002  max mem: 55573
Epoch: [229] Total time: 0:01:44 (0.6715 s / it)
Averaged stats: lr: 0.001173  min_lr: 0.001173  loss: 3.1264 (3.0354)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8798 (0.8798)  acc1: 82.2917 (82.2917)  acc5: 96.8750 (96.8750)  time: 6.9883  data: 6.7512  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9321 (0.9354)  acc1: 80.8594 (78.8160)  acc5: 96.7448 (95.3920)  time: 1.5441  data: 1.3503  max mem: 55573
Test: Total time: 0:00:07 (1.5870 s / it)
* Acc@1 79.666 Acc@5 95.112 loss 0.933
Accuracy of the model on the 50000 test images: 79.7%
Max accuracy: 79.76%
Test:  [0/5]  eta: 0:00:36  loss: 1.0333 (1.0333)  acc1: 74.0885 (74.0885)  acc5: 92.0573 (92.0573)  time: 7.2720  data: 7.0359  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0613 (1.0786)  acc1: 72.0052 (69.8880)  acc5: 92.0573 (90.0480)  time: 1.5997  data: 1.4073  max mem: 55573
Test: Total time: 0:00:08 (1.6207 s / it)
* Acc@1 70.700 Acc@5 90.368 loss 1.099
Accuracy of the model EMA on 50000 test images: 70.7%
Max EMA accuracy: 70.70%
Epoch: [230]  [  0/156]  eta: 0:34:28  lr: 0.001172  min_lr: 0.001172  loss: 2.4362 (2.4362)  weight_decay: 0.0500 (0.0500)  time: 13.2608  data: 7.6964  max mem: 55573
Epoch: [230]  [ 10/156]  eta: 0:04:14  lr: 0.001170  min_lr: 0.001170  loss: 2.4871 (2.7191)  weight_decay: 0.0500 (0.0500)  time: 1.7424  data: 0.7001  max mem: 55573
Epoch: [230]  [ 20/156]  eta: 0:02:42  lr: 0.001168  min_lr: 0.001168  loss: 3.0234 (2.8231)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [230]  [ 30/156]  eta: 0:02:05  lr: 0.001166  min_lr: 0.001166  loss: 3.1610 (2.9107)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [230]  [ 40/156]  eta: 0:01:44  lr: 0.001164  min_lr: 0.001164  loss: 3.2064 (2.9292)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [230]  [ 50/156]  eta: 0:01:28  lr: 0.001162  min_lr: 0.001162  loss: 2.9443 (2.9042)  weight_decay: 0.0500 (0.0500)  time: 0.5885  data: 0.0004  max mem: 55573
Epoch: [230]  [ 60/156]  eta: 0:01:16  lr: 0.001160  min_lr: 0.001160  loss: 3.0950 (2.9596)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [230]  [ 70/156]  eta: 0:01:06  lr: 0.001158  min_lr: 0.001158  loss: 3.1091 (2.9711)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [230]  [ 80/156]  eta: 0:00:56  lr: 0.001156  min_lr: 0.001156  loss: 3.2346 (3.0126)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0004  max mem: 55573
Epoch: [230]  [ 90/156]  eta: 0:00:48  lr: 0.001154  min_lr: 0.001154  loss: 3.2346 (2.9881)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [230]  [100/156]  eta: 0:00:40  lr: 0.001152  min_lr: 0.001152  loss: 2.4890 (2.9512)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [230]  [110/156]  eta: 0:00:32  lr: 0.001150  min_lr: 0.001150  loss: 3.0217 (2.9715)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [230]  [120/156]  eta: 0:00:25  lr: 0.001148  min_lr: 0.001148  loss: 3.1734 (2.9591)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [230]  [130/156]  eta: 0:00:17  lr: 0.001146  min_lr: 0.001146  loss: 2.9279 (2.9611)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0012  max mem: 55573
Epoch: [230]  [140/156]  eta: 0:00:10  lr: 0.001144  min_lr: 0.001144  loss: 3.0540 (2.9512)  weight_decay: 0.0500 (0.0500)  time: 0.5873  data: 0.0010  max mem: 55573
Epoch: [230]  [150/156]  eta: 0:00:04  lr: 0.001142  min_lr: 0.001142  loss: 3.0668 (2.9725)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [230]  [155/156]  eta: 0:00:00  lr: 0.001141  min_lr: 0.001141  loss: 3.1935 (2.9815)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [230] Total time: 0:01:45 (0.6770 s / it)
Averaged stats: lr: 0.001141  min_lr: 0.001141  loss: 3.1935 (3.0356)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 1.0018 (1.0018)  acc1: 81.5104 (81.5104)  acc5: 96.6146 (96.6146)  time: 6.9034  data: 6.6664  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0474 (1.0620)  acc1: 79.6875 (78.4320)  acc5: 95.8333 (95.0400)  time: 1.5270  data: 1.3334  max mem: 55573
Test: Total time: 0:00:07 (1.5746 s / it)
* Acc@1 79.730 Acc@5 95.038 loss 1.042
Accuracy of the model on the 50000 test images: 79.7%
Max accuracy: 79.76%
Test:  [0/5]  eta: 0:00:35  loss: 1.0322 (1.0322)  acc1: 74.0885 (74.0885)  acc5: 92.1875 (92.1875)  time: 7.0832  data: 6.8471  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0603 (1.0777)  acc1: 72.1354 (69.9200)  acc5: 92.1875 (90.0480)  time: 1.5620  data: 1.3695  max mem: 55573
Test: Total time: 0:00:07 (1.5783 s / it)
* Acc@1 70.746 Acc@5 90.372 loss 1.098
Accuracy of the model EMA on 50000 test images: 70.7%
Max EMA accuracy: 70.75%
Epoch: [231]  [  0/156]  eta: 0:35:41  lr: 0.001141  min_lr: 0.001141  loss: 2.1948 (2.1948)  weight_decay: 0.0500 (0.0500)  time: 13.7244  data: 11.6071  max mem: 55573
Epoch: [231]  [ 10/156]  eta: 0:04:21  lr: 0.001139  min_lr: 0.001139  loss: 3.1036 (3.0557)  weight_decay: 0.0500 (0.0500)  time: 1.7925  data: 1.0556  max mem: 55573
Epoch: [231]  [ 20/156]  eta: 0:02:45  lr: 0.001137  min_lr: 0.001137  loss: 3.0947 (2.9992)  weight_decay: 0.0500 (0.0500)  time: 0.5950  data: 0.0004  max mem: 55573
Epoch: [231]  [ 30/156]  eta: 0:02:08  lr: 0.001135  min_lr: 0.001135  loss: 3.0947 (3.0492)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [231]  [ 40/156]  eta: 0:01:45  lr: 0.001133  min_lr: 0.001133  loss: 3.0802 (3.0235)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [231]  [ 50/156]  eta: 0:01:30  lr: 0.001131  min_lr: 0.001131  loss: 3.0516 (2.9942)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [231]  [ 60/156]  eta: 0:01:17  lr: 0.001129  min_lr: 0.001129  loss: 3.0704 (3.0170)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [231]  [ 70/156]  eta: 0:01:06  lr: 0.001127  min_lr: 0.001127  loss: 3.0323 (2.9788)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [231]  [ 80/156]  eta: 0:00:57  lr: 0.001125  min_lr: 0.001125  loss: 2.8985 (2.9756)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [231]  [ 90/156]  eta: 0:00:48  lr: 0.001123  min_lr: 0.001123  loss: 3.0773 (2.9819)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [231]  [100/156]  eta: 0:00:40  lr: 0.001121  min_lr: 0.001121  loss: 3.0843 (2.9699)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [231]  [110/156]  eta: 0:00:32  lr: 0.001119  min_lr: 0.001119  loss: 3.0843 (2.9733)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [231]  [120/156]  eta: 0:00:25  lr: 0.001117  min_lr: 0.001117  loss: 3.1105 (2.9933)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [231]  [130/156]  eta: 0:00:17  lr: 0.001115  min_lr: 0.001115  loss: 3.2245 (2.9992)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0010  max mem: 55573
Epoch: [231]  [140/156]  eta: 0:00:10  lr: 0.001113  min_lr: 0.001113  loss: 3.0775 (3.0024)  weight_decay: 0.0500 (0.0500)  time: 0.5864  data: 0.0009  max mem: 55573
Epoch: [231]  [150/156]  eta: 0:00:04  lr: 0.001111  min_lr: 0.001111  loss: 2.9683 (2.9983)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0002  max mem: 55573
Epoch: [231]  [155/156]  eta: 0:00:00  lr: 0.001110  min_lr: 0.001110  loss: 2.9946 (3.0056)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [231] Total time: 0:01:46 (0.6805 s / it)
Averaged stats: lr: 0.001110  min_lr: 0.001110  loss: 2.9946 (3.0454)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.9001 (0.9001)  acc1: 82.1615 (82.1615)  acc5: 97.0052 (97.0052)  time: 6.7945  data: 6.5570  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9975 (1.0032)  acc1: 80.3385 (79.2000)  acc5: 96.0938 (95.3280)  time: 1.5055  data: 1.3115  max mem: 55573
Test: Total time: 0:00:07 (1.5488 s / it)
* Acc@1 79.838 Acc@5 95.056 loss 0.999
Accuracy of the model on the 50000 test images: 79.8%
Max accuracy: 79.84%
Test:  [0/5]  eta: 0:00:34  loss: 1.0308 (1.0308)  acc1: 74.2188 (74.2188)  acc5: 92.3177 (92.3177)  time: 6.8288  data: 6.5927  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0589 (1.0765)  acc1: 72.1354 (70.0480)  acc5: 92.3177 (90.1120)  time: 1.5111  data: 1.3187  max mem: 55573
Test: Total time: 0:00:07 (1.5289 s / it)
* Acc@1 70.798 Acc@5 90.398 loss 1.097
Accuracy of the model EMA on 50000 test images: 70.8%
Max EMA accuracy: 70.80%
Epoch: [232]  [  0/156]  eta: 0:32:47  lr: 0.001110  min_lr: 0.001110  loss: 3.1348 (3.1348)  weight_decay: 0.0500 (0.0500)  time: 12.6143  data: 11.1174  max mem: 55573
Epoch: [232]  [ 10/156]  eta: 0:04:16  lr: 0.001108  min_lr: 0.001108  loss: 2.9863 (2.8919)  weight_decay: 0.0500 (0.0500)  time: 1.7558  data: 1.0786  max mem: 55573
Epoch: [232]  [ 20/156]  eta: 0:02:43  lr: 0.001106  min_lr: 0.001106  loss: 2.9863 (2.9978)  weight_decay: 0.0500 (0.0500)  time: 0.6306  data: 0.0375  max mem: 55573
Epoch: [232]  [ 30/156]  eta: 0:02:06  lr: 0.001104  min_lr: 0.001104  loss: 3.1702 (3.0366)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [232]  [ 40/156]  eta: 0:01:44  lr: 0.001102  min_lr: 0.001102  loss: 3.0676 (2.9968)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [232]  [ 50/156]  eta: 0:01:29  lr: 0.001100  min_lr: 0.001100  loss: 3.0995 (3.0099)  weight_decay: 0.0500 (0.0500)  time: 0.5936  data: 0.0004  max mem: 55573
Epoch: [232]  [ 60/156]  eta: 0:01:17  lr: 0.001098  min_lr: 0.001098  loss: 3.1183 (3.0083)  weight_decay: 0.0500 (0.0500)  time: 0.5932  data: 0.0005  max mem: 55573
Epoch: [232]  [ 70/156]  eta: 0:01:06  lr: 0.001096  min_lr: 0.001096  loss: 3.1462 (3.0368)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [232]  [ 80/156]  eta: 0:00:56  lr: 0.001094  min_lr: 0.001094  loss: 3.2811 (3.0603)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [232]  [ 90/156]  eta: 0:00:48  lr: 0.001092  min_lr: 0.001092  loss: 3.1541 (3.0707)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [232]  [100/156]  eta: 0:00:40  lr: 0.001090  min_lr: 0.001090  loss: 3.1141 (3.0644)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [232]  [110/156]  eta: 0:00:32  lr: 0.001088  min_lr: 0.001088  loss: 3.2576 (3.0642)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0005  max mem: 55573
Epoch: [232]  [120/156]  eta: 0:00:25  lr: 0.001086  min_lr: 0.001086  loss: 2.9095 (3.0466)  weight_decay: 0.0500 (0.0500)  time: 0.5941  data: 0.0005  max mem: 55573
Epoch: [232]  [130/156]  eta: 0:00:17  lr: 0.001084  min_lr: 0.001084  loss: 2.7648 (3.0308)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0011  max mem: 55573
Epoch: [232]  [140/156]  eta: 0:00:10  lr: 0.001082  min_lr: 0.001082  loss: 2.8807 (3.0271)  weight_decay: 0.0500 (0.0500)  time: 0.5867  data: 0.0009  max mem: 55573
Epoch: [232]  [150/156]  eta: 0:00:04  lr: 0.001080  min_lr: 0.001080  loss: 2.8746 (3.0065)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [232]  [155/156]  eta: 0:00:00  lr: 0.001079  min_lr: 0.001079  loss: 2.5499 (3.0025)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [232] Total time: 0:01:45 (0.6791 s / it)
Averaged stats: lr: 0.001079  min_lr: 0.001079  loss: 2.5499 (3.0233)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:32  loss: 0.8828 (0.8828)  acc1: 81.6406 (81.6406)  acc5: 96.6146 (96.6146)  time: 6.5959  data: 6.3588  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9684 (0.9651)  acc1: 81.6406 (79.2640)  acc5: 96.0938 (95.2000)  time: 1.4656  data: 1.2719  max mem: 55573
Test: Total time: 0:00:07 (1.5205 s / it)
* Acc@1 80.096 Acc@5 95.214 loss 0.941
Accuracy of the model on the 50000 test images: 80.1%
Max accuracy: 80.10%
Test:  [0/5]  eta: 0:00:33  loss: 1.0298 (1.0298)  acc1: 74.4792 (74.4792)  acc5: 92.3177 (92.3177)  time: 6.7706  data: 6.5346  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0580 (1.0756)  acc1: 72.1354 (70.1120)  acc5: 92.3177 (90.1440)  time: 1.4995  data: 1.3070  max mem: 55573
Test: Total time: 0:00:07 (1.5164 s / it)
* Acc@1 70.844 Acc@5 90.408 loss 1.096
Accuracy of the model EMA on 50000 test images: 70.8%
Max EMA accuracy: 70.84%
Epoch: [233]  [  0/156]  eta: 0:35:07  lr: 0.001079  min_lr: 0.001079  loss: 3.3505 (3.3505)  weight_decay: 0.0500 (0.0500)  time: 13.5114  data: 8.3274  max mem: 55573
Epoch: [233]  [ 10/156]  eta: 0:04:17  lr: 0.001077  min_lr: 0.001077  loss: 3.1566 (3.0847)  weight_decay: 0.0500 (0.0500)  time: 1.7659  data: 0.7575  max mem: 55573
Epoch: [233]  [ 20/156]  eta: 0:02:44  lr: 0.001075  min_lr: 0.001075  loss: 3.1239 (3.0590)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [233]  [ 30/156]  eta: 0:02:07  lr: 0.001073  min_lr: 0.001073  loss: 3.1379 (3.0525)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [233]  [ 40/156]  eta: 0:01:45  lr: 0.001071  min_lr: 0.001071  loss: 3.0343 (3.0085)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [233]  [ 50/156]  eta: 0:01:29  lr: 0.001069  min_lr: 0.001069  loss: 3.1093 (3.0417)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [233]  [ 60/156]  eta: 0:01:17  lr: 0.001067  min_lr: 0.001067  loss: 3.2798 (3.0550)  weight_decay: 0.0500 (0.0500)  time: 0.5885  data: 0.0005  max mem: 55573
Epoch: [233]  [ 70/156]  eta: 0:01:06  lr: 0.001065  min_lr: 0.001065  loss: 3.2160 (3.0241)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [233]  [ 80/156]  eta: 0:00:56  lr: 0.001063  min_lr: 0.001063  loss: 3.0363 (3.0229)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [233]  [ 90/156]  eta: 0:00:48  lr: 0.001061  min_lr: 0.001061  loss: 3.2310 (3.0523)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0005  max mem: 55573
Epoch: [233]  [100/156]  eta: 0:00:40  lr: 0.001059  min_lr: 0.001059  loss: 3.2948 (3.0594)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [233]  [110/156]  eta: 0:00:32  lr: 0.001057  min_lr: 0.001057  loss: 3.0505 (3.0512)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [233]  [120/156]  eta: 0:00:25  lr: 0.001055  min_lr: 0.001055  loss: 2.9290 (3.0287)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [233]  [130/156]  eta: 0:00:17  lr: 0.001053  min_lr: 0.001053  loss: 2.8015 (3.0151)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0011  max mem: 55573
Epoch: [233]  [140/156]  eta: 0:00:10  lr: 0.001052  min_lr: 0.001052  loss: 3.1661 (3.0193)  weight_decay: 0.0500 (0.0500)  time: 0.5859  data: 0.0010  max mem: 55573
Epoch: [233]  [150/156]  eta: 0:00:04  lr: 0.001050  min_lr: 0.001050  loss: 3.2512 (3.0276)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [233]  [155/156]  eta: 0:00:00  lr: 0.001049  min_lr: 0.001049  loss: 3.1582 (3.0221)  weight_decay: 0.0500 (0.0500)  time: 0.5828  data: 0.0001  max mem: 55573
Epoch: [233] Total time: 0:01:45 (0.6787 s / it)
Averaged stats: lr: 0.001049  min_lr: 0.001049  loss: 3.1582 (3.0307)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.8899 (0.8899)  acc1: 82.4219 (82.4219)  acc5: 96.2240 (96.2240)  time: 7.1450  data: 6.9076  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9545 (0.9798)  acc1: 79.9479 (79.0080)  acc5: 96.2240 (95.2640)  time: 1.5755  data: 1.3816  max mem: 55573
Test: Total time: 0:00:08 (1.6167 s / it)
* Acc@1 79.954 Acc@5 95.242 loss 0.955
Accuracy of the model on the 50000 test images: 80.0%
Max accuracy: 80.10%
Test:  [0/5]  eta: 0:00:35  loss: 1.0286 (1.0286)  acc1: 74.4792 (74.4792)  acc5: 92.3177 (92.3177)  time: 7.1729  data: 6.9368  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0568 (1.0744)  acc1: 72.0052 (70.1440)  acc5: 92.3177 (90.1760)  time: 1.5800  data: 1.3875  max mem: 55573
Test: Total time: 0:00:08 (1.6004 s / it)
* Acc@1 70.886 Acc@5 90.436 loss 1.095
Accuracy of the model EMA on 50000 test images: 70.9%
Max EMA accuracy: 70.89%
Epoch: [234]  [  0/156]  eta: 0:32:31  lr: 0.001048  min_lr: 0.001048  loss: 3.6643 (3.6643)  weight_decay: 0.0500 (0.0500)  time: 12.5071  data: 11.8951  max mem: 55573
Epoch: [234]  [ 10/156]  eta: 0:04:14  lr: 0.001046  min_lr: 0.001046  loss: 3.2827 (3.0863)  weight_decay: 0.0500 (0.0500)  time: 1.7423  data: 1.1217  max mem: 55573
Epoch: [234]  [ 20/156]  eta: 0:02:42  lr: 0.001045  min_lr: 0.001045  loss: 3.2365 (3.1307)  weight_decay: 0.0500 (0.0500)  time: 0.6276  data: 0.0223  max mem: 55573
Epoch: [234]  [ 30/156]  eta: 0:02:05  lr: 0.001043  min_lr: 0.001043  loss: 3.1483 (3.0319)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [234]  [ 40/156]  eta: 0:01:44  lr: 0.001041  min_lr: 0.001041  loss: 3.0355 (3.0514)  weight_decay: 0.0500 (0.0500)  time: 0.5943  data: 0.0004  max mem: 55573
Epoch: [234]  [ 50/156]  eta: 0:01:29  lr: 0.001039  min_lr: 0.001039  loss: 3.0355 (3.0374)  weight_decay: 0.0500 (0.0500)  time: 0.5936  data: 0.0005  max mem: 55573
Epoch: [234]  [ 60/156]  eta: 0:01:16  lr: 0.001037  min_lr: 0.001037  loss: 2.9630 (3.0060)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [234]  [ 70/156]  eta: 0:01:06  lr: 0.001035  min_lr: 0.001035  loss: 3.1970 (3.0364)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [234]  [ 80/156]  eta: 0:00:56  lr: 0.001033  min_lr: 0.001033  loss: 3.2481 (3.0202)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [234]  [ 90/156]  eta: 0:00:48  lr: 0.001031  min_lr: 0.001031  loss: 3.2919 (3.0354)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [234]  [100/156]  eta: 0:00:40  lr: 0.001029  min_lr: 0.001029  loss: 3.2919 (3.0575)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [234]  [110/156]  eta: 0:00:32  lr: 0.001027  min_lr: 0.001027  loss: 3.2299 (3.0542)  weight_decay: 0.0500 (0.0500)  time: 0.5943  data: 0.0004  max mem: 55573
Epoch: [234]  [120/156]  eta: 0:00:25  lr: 0.001025  min_lr: 0.001025  loss: 3.0328 (3.0440)  weight_decay: 0.0500 (0.0500)  time: 0.5938  data: 0.0006  max mem: 55573
Epoch: [234]  [130/156]  eta: 0:00:17  lr: 0.001023  min_lr: 0.001023  loss: 2.9037 (3.0245)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0012  max mem: 55573
Epoch: [234]  [140/156]  eta: 0:00:10  lr: 0.001021  min_lr: 0.001021  loss: 2.8980 (3.0180)  weight_decay: 0.0500 (0.0500)  time: 0.5857  data: 0.0009  max mem: 55573
Epoch: [234]  [150/156]  eta: 0:00:04  lr: 0.001019  min_lr: 0.001019  loss: 3.2015 (3.0274)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [234]  [155/156]  eta: 0:00:00  lr: 0.001018  min_lr: 0.001018  loss: 3.2258 (3.0377)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [234] Total time: 0:01:45 (0.6780 s / it)
Averaged stats: lr: 0.001018  min_lr: 0.001018  loss: 3.2258 (3.0291)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.9435 (0.9435)  acc1: 81.7708 (81.7708)  acc5: 96.6146 (96.6146)  time: 6.6616  data: 6.4245  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0479 (1.0338)  acc1: 80.2083 (78.7840)  acc5: 96.2264 (95.1680)  time: 1.4789  data: 1.2850  max mem: 55573
Test: Total time: 0:00:07 (1.5195 s / it)
* Acc@1 79.936 Acc@5 95.278 loss 1.019
Accuracy of the model on the 50000 test images: 79.9%
Max accuracy: 80.10%
Test:  [0/5]  eta: 0:00:35  loss: 1.0276 (1.0276)  acc1: 74.6094 (74.6094)  acc5: 92.3177 (92.3177)  time: 7.0079  data: 6.7719  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0560 (1.0737)  acc1: 72.2656 (70.2720)  acc5: 92.3177 (90.1440)  time: 1.5469  data: 1.3545  max mem: 55573
Test: Total time: 0:00:07 (1.5731 s / it)
* Acc@1 70.946 Acc@5 90.462 loss 1.094
Accuracy of the model EMA on 50000 test images: 70.9%
Max EMA accuracy: 70.95%
Epoch: [235]  [  0/156]  eta: 0:34:39  lr: 0.001018  min_lr: 0.001018  loss: 3.4048 (3.4048)  weight_decay: 0.0500 (0.0500)  time: 13.3312  data: 8.8833  max mem: 55573
Epoch: [235]  [ 10/156]  eta: 0:04:16  lr: 0.001016  min_lr: 0.001016  loss: 3.1525 (3.1223)  weight_decay: 0.0500 (0.0500)  time: 1.7580  data: 0.8080  max mem: 55573
Epoch: [235]  [ 20/156]  eta: 0:02:43  lr: 0.001014  min_lr: 0.001014  loss: 3.1307 (3.0070)  weight_decay: 0.0500 (0.0500)  time: 0.5951  data: 0.0004  max mem: 55573
Epoch: [235]  [ 30/156]  eta: 0:02:06  lr: 0.001013  min_lr: 0.001013  loss: 3.0792 (2.9973)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [235]  [ 40/156]  eta: 0:01:44  lr: 0.001011  min_lr: 0.001011  loss: 2.8781 (2.9628)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [235]  [ 50/156]  eta: 0:01:29  lr: 0.001009  min_lr: 0.001009  loss: 3.0797 (2.9758)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [235]  [ 60/156]  eta: 0:01:16  lr: 0.001007  min_lr: 0.001007  loss: 3.0853 (2.9616)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [235]  [ 70/156]  eta: 0:01:06  lr: 0.001005  min_lr: 0.001005  loss: 2.7940 (2.9535)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [235]  [ 80/156]  eta: 0:00:56  lr: 0.001003  min_lr: 0.001003  loss: 3.0046 (2.9548)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [235]  [ 90/156]  eta: 0:00:48  lr: 0.001001  min_lr: 0.001001  loss: 3.1267 (2.9713)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [235]  [100/156]  eta: 0:00:40  lr: 0.000999  min_lr: 0.000999  loss: 3.3019 (3.0080)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [235]  [110/156]  eta: 0:00:32  lr: 0.000997  min_lr: 0.000997  loss: 3.3027 (3.0113)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [235]  [120/156]  eta: 0:00:25  lr: 0.000995  min_lr: 0.000995  loss: 3.0620 (3.0142)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [235]  [130/156]  eta: 0:00:17  lr: 0.000994  min_lr: 0.000994  loss: 3.1379 (3.0108)  weight_decay: 0.0500 (0.0500)  time: 0.5887  data: 0.0010  max mem: 55573
Epoch: [235]  [140/156]  eta: 0:00:10  lr: 0.000992  min_lr: 0.000992  loss: 2.9519 (2.9977)  weight_decay: 0.0500 (0.0500)  time: 0.5852  data: 0.0008  max mem: 55573
Epoch: [235]  [150/156]  eta: 0:00:04  lr: 0.000990  min_lr: 0.000990  loss: 2.9103 (2.9934)  weight_decay: 0.0500 (0.0500)  time: 0.5828  data: 0.0001  max mem: 55573
Epoch: [235]  [155/156]  eta: 0:00:00  lr: 0.000989  min_lr: 0.000989  loss: 2.9103 (2.9817)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [235] Total time: 0:01:45 (0.6778 s / it)
Averaged stats: lr: 0.000989  min_lr: 0.000989  loss: 2.9103 (3.0192)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 0.8577 (0.8577)  acc1: 82.5521 (82.5521)  acc5: 96.3542 (96.3542)  time: 7.3086  data: 7.0713  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9383 (0.9481)  acc1: 80.9896 (79.2960)  acc5: 96.3542 (95.2960)  time: 1.6079  data: 1.4143  max mem: 55573
Test: Total time: 0:00:08 (1.6606 s / it)
* Acc@1 80.202 Acc@5 95.328 loss 0.933
Accuracy of the model on the 50000 test images: 80.2%
Max accuracy: 80.20%
Test:  [0/5]  eta: 0:00:34  loss: 1.0267 (1.0267)  acc1: 74.6094 (74.6094)  acc5: 92.3177 (92.3177)  time: 6.8274  data: 6.5914  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0550 (1.0727)  acc1: 72.2656 (70.3040)  acc5: 92.3177 (90.1440)  time: 1.5108  data: 1.3184  max mem: 55573
Test: Total time: 0:00:07 (1.5290 s / it)
* Acc@1 70.984 Acc@5 90.470 loss 1.093
Accuracy of the model EMA on 50000 test images: 71.0%
Max EMA accuracy: 70.98%
Epoch: [236]  [  0/156]  eta: 0:32:40  lr: 0.000989  min_lr: 0.000989  loss: 3.3979 (3.3979)  weight_decay: 0.0500 (0.0500)  time: 12.5651  data: 11.0964  max mem: 55573
Epoch: [236]  [ 10/156]  eta: 0:04:09  lr: 0.000987  min_lr: 0.000987  loss: 3.2041 (3.0381)  weight_decay: 0.0500 (0.0500)  time: 1.7075  data: 1.0092  max mem: 55573
Epoch: [236]  [ 20/156]  eta: 0:02:40  lr: 0.000985  min_lr: 0.000985  loss: 3.0033 (2.9300)  weight_decay: 0.0500 (0.0500)  time: 0.6074  data: 0.0004  max mem: 55573
Epoch: [236]  [ 30/156]  eta: 0:02:04  lr: 0.000983  min_lr: 0.000983  loss: 3.1248 (2.9986)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [236]  [ 40/156]  eta: 0:01:43  lr: 0.000981  min_lr: 0.000981  loss: 3.1631 (2.9727)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [236]  [ 50/156]  eta: 0:01:28  lr: 0.000979  min_lr: 0.000979  loss: 3.0382 (3.0047)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [236]  [ 60/156]  eta: 0:01:16  lr: 0.000977  min_lr: 0.000977  loss: 3.1222 (3.0162)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [236]  [ 70/156]  eta: 0:01:05  lr: 0.000975  min_lr: 0.000975  loss: 3.1222 (3.0223)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0005  max mem: 55573
Epoch: [236]  [ 80/156]  eta: 0:00:56  lr: 0.000973  min_lr: 0.000973  loss: 3.1311 (3.0301)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0005  max mem: 55573
Epoch: [236]  [ 90/156]  eta: 0:00:47  lr: 0.000972  min_lr: 0.000972  loss: 3.1311 (3.0132)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [236]  [100/156]  eta: 0:00:39  lr: 0.000970  min_lr: 0.000970  loss: 3.2032 (3.0327)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [236]  [110/156]  eta: 0:00:32  lr: 0.000968  min_lr: 0.000968  loss: 3.1602 (3.0115)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [236]  [120/156]  eta: 0:00:24  lr: 0.000966  min_lr: 0.000966  loss: 3.1795 (3.0460)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [236]  [130/156]  eta: 0:00:17  lr: 0.000964  min_lr: 0.000964  loss: 3.2821 (3.0281)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0013  max mem: 55573
Epoch: [236]  [140/156]  eta: 0:00:10  lr: 0.000962  min_lr: 0.000962  loss: 2.8089 (3.0255)  weight_decay: 0.0500 (0.0500)  time: 0.5849  data: 0.0011  max mem: 55573
Epoch: [236]  [150/156]  eta: 0:00:04  lr: 0.000960  min_lr: 0.000960  loss: 3.1401 (3.0390)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [236]  [155/156]  eta: 0:00:00  lr: 0.000959  min_lr: 0.000959  loss: 3.2851 (3.0365)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [236] Total time: 0:01:45 (0.6746 s / it)
Averaged stats: lr: 0.000959  min_lr: 0.000959  loss: 3.2851 (3.0004)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 0.9586 (0.9586)  acc1: 82.2917 (82.2917)  acc5: 96.8750 (96.8750)  time: 7.3154  data: 7.0780  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0117 (1.0045)  acc1: 79.0365 (78.8480)  acc5: 96.7448 (95.3920)  time: 1.6096  data: 1.4157  max mem: 55573
Test: Total time: 0:00:08 (1.6522 s / it)
* Acc@1 79.952 Acc@5 95.234 loss 0.997
Accuracy of the model on the 50000 test images: 80.0%
Max accuracy: 80.20%
Test:  [0/5]  eta: 0:00:37  loss: 1.0259 (1.0259)  acc1: 74.3490 (74.3490)  acc5: 92.3177 (92.3177)  time: 7.4771  data: 7.2405  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0542 (1.0719)  acc1: 72.6562 (70.4000)  acc5: 92.3177 (90.1760)  time: 1.6407  data: 1.4482  max mem: 55573
Test: Total time: 0:00:08 (1.6598 s / it)
* Acc@1 71.026 Acc@5 90.482 loss 1.092
Accuracy of the model EMA on 50000 test images: 71.0%
Max EMA accuracy: 71.03%
Epoch: [237]  [  0/156]  eta: 0:36:57  lr: 0.000959  min_lr: 0.000959  loss: 2.1999 (2.1999)  weight_decay: 0.0500 (0.0500)  time: 14.2171  data: 13.6382  max mem: 55573
Epoch: [237]  [ 10/156]  eta: 0:04:26  lr: 0.000957  min_lr: 0.000957  loss: 3.1159 (2.9943)  weight_decay: 0.0500 (0.0500)  time: 1.8286  data: 1.2402  max mem: 55573
Epoch: [237]  [ 20/156]  eta: 0:02:48  lr: 0.000956  min_lr: 0.000956  loss: 3.2697 (3.1308)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [237]  [ 30/156]  eta: 0:02:09  lr: 0.000954  min_lr: 0.000954  loss: 3.2153 (3.1283)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [237]  [ 40/156]  eta: 0:01:46  lr: 0.000952  min_lr: 0.000952  loss: 3.1319 (3.0684)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [237]  [ 50/156]  eta: 0:01:30  lr: 0.000950  min_lr: 0.000950  loss: 2.8899 (3.0427)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [237]  [ 60/156]  eta: 0:01:18  lr: 0.000948  min_lr: 0.000948  loss: 3.0631 (3.0271)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0005  max mem: 55573
Epoch: [237]  [ 70/156]  eta: 0:01:07  lr: 0.000946  min_lr: 0.000946  loss: 2.9541 (3.0115)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [237]  [ 80/156]  eta: 0:00:57  lr: 0.000944  min_lr: 0.000944  loss: 2.8422 (2.9865)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [237]  [ 90/156]  eta: 0:00:48  lr: 0.000943  min_lr: 0.000943  loss: 3.1695 (3.0058)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [237]  [100/156]  eta: 0:00:40  lr: 0.000941  min_lr: 0.000941  loss: 3.1814 (2.9975)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [237]  [110/156]  eta: 0:00:32  lr: 0.000939  min_lr: 0.000939  loss: 3.0767 (3.0043)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [237]  [120/156]  eta: 0:00:25  lr: 0.000937  min_lr: 0.000937  loss: 2.9704 (2.9870)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [237]  [130/156]  eta: 0:00:18  lr: 0.000935  min_lr: 0.000935  loss: 2.8874 (2.9940)  weight_decay: 0.0500 (0.0500)  time: 0.5883  data: 0.0011  max mem: 55573
Epoch: [237]  [140/156]  eta: 0:00:10  lr: 0.000933  min_lr: 0.000933  loss: 3.0969 (2.9994)  weight_decay: 0.0500 (0.0500)  time: 0.5851  data: 0.0009  max mem: 55573
Epoch: [237]  [150/156]  eta: 0:00:04  lr: 0.000931  min_lr: 0.000931  loss: 3.0171 (2.9866)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [237]  [155/156]  eta: 0:00:00  lr: 0.000930  min_lr: 0.000930  loss: 3.0171 (2.9824)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [237] Total time: 0:01:46 (0.6830 s / it)
Averaged stats: lr: 0.000930  min_lr: 0.000930  loss: 3.0171 (3.0080)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8175 (0.8175)  acc1: 83.9844 (83.9844)  acc5: 96.4844 (96.4844)  time: 6.8026  data: 6.5654  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9224 (0.9133)  acc1: 80.3385 (79.5520)  acc5: 96.3542 (94.9760)  time: 1.5070  data: 1.3132  max mem: 55573
Test: Total time: 0:00:07 (1.5566 s / it)
* Acc@1 80.360 Acc@5 95.348 loss 0.883
Accuracy of the model on the 50000 test images: 80.4%
Max accuracy: 80.36%
Test:  [0/5]  eta: 0:00:34  loss: 1.0250 (1.0250)  acc1: 74.3490 (74.3490)  acc5: 92.5781 (92.5781)  time: 6.8909  data: 6.6548  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0532 (1.0709)  acc1: 72.5260 (70.4000)  acc5: 92.5781 (90.2400)  time: 1.5235  data: 1.3311  max mem: 55573
Test: Total time: 0:00:07 (1.5460 s / it)
* Acc@1 71.046 Acc@5 90.494 loss 1.092
Accuracy of the model EMA on 50000 test images: 71.0%
Max EMA accuracy: 71.05%
Epoch: [238]  [  0/156]  eta: 0:28:21  lr: 0.000930  min_lr: 0.000930  loss: 3.3425 (3.3425)  weight_decay: 0.0500 (0.0500)  time: 10.9044  data: 9.0314  max mem: 55573
Epoch: [238]  [ 10/156]  eta: 0:03:58  lr: 0.000928  min_lr: 0.000928  loss: 3.3451 (3.1910)  weight_decay: 0.0500 (0.0500)  time: 1.6321  data: 0.8587  max mem: 55573
Epoch: [238]  [ 20/156]  eta: 0:02:34  lr: 0.000927  min_lr: 0.000927  loss: 3.2249 (3.0664)  weight_decay: 0.0500 (0.0500)  time: 0.6501  data: 0.0209  max mem: 55573
Epoch: [238]  [ 30/156]  eta: 0:02:01  lr: 0.000925  min_lr: 0.000925  loss: 3.0973 (3.0900)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0004  max mem: 55573
Epoch: [238]  [ 40/156]  eta: 0:01:41  lr: 0.000923  min_lr: 0.000923  loss: 3.0731 (3.0330)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [238]  [ 50/156]  eta: 0:01:26  lr: 0.000921  min_lr: 0.000921  loss: 3.0731 (3.0753)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [238]  [ 60/156]  eta: 0:01:14  lr: 0.000919  min_lr: 0.000919  loss: 3.2486 (3.0574)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [238]  [ 70/156]  eta: 0:01:04  lr: 0.000917  min_lr: 0.000917  loss: 2.9300 (3.0168)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0004  max mem: 55573
Epoch: [238]  [ 80/156]  eta: 0:00:55  lr: 0.000916  min_lr: 0.000916  loss: 2.8110 (2.9884)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [238]  [ 90/156]  eta: 0:00:47  lr: 0.000914  min_lr: 0.000914  loss: 2.9812 (3.0042)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [238]  [100/156]  eta: 0:00:39  lr: 0.000912  min_lr: 0.000912  loss: 3.2001 (3.0058)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [238]  [110/156]  eta: 0:00:31  lr: 0.000910  min_lr: 0.000910  loss: 3.0272 (3.0090)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [238]  [120/156]  eta: 0:00:24  lr: 0.000908  min_lr: 0.000908  loss: 3.0263 (2.9980)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [238]  [130/156]  eta: 0:00:17  lr: 0.000906  min_lr: 0.000906  loss: 2.7779 (2.9961)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0010  max mem: 55573
Epoch: [238]  [140/156]  eta: 0:00:10  lr: 0.000905  min_lr: 0.000905  loss: 3.0270 (2.9993)  weight_decay: 0.0500 (0.0500)  time: 0.5858  data: 0.0008  max mem: 55573
Epoch: [238]  [150/156]  eta: 0:00:03  lr: 0.000903  min_lr: 0.000903  loss: 3.1117 (3.0027)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [238]  [155/156]  eta: 0:00:00  lr: 0.000902  min_lr: 0.000902  loss: 3.1117 (3.0084)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [238] Total time: 0:01:44 (0.6697 s / it)
Averaged stats: lr: 0.000902  min_lr: 0.000902  loss: 3.1117 (3.0070)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9160 (0.9160)  acc1: 83.0729 (83.0729)  acc5: 96.8750 (96.8750)  time: 6.8712  data: 6.6341  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0254 (1.0157)  acc1: 80.8594 (79.1680)  acc5: 96.2240 (95.2320)  time: 1.5205  data: 1.3269  max mem: 55573
Test: Total time: 0:00:07 (1.5738 s / it)
* Acc@1 80.090 Acc@5 95.288 loss 1.010
Accuracy of the model on the 50000 test images: 80.1%
Max accuracy: 80.36%
Test:  [0/5]  eta: 0:00:35  loss: 1.0242 (1.0242)  acc1: 74.4792 (74.4792)  acc5: 92.5781 (92.5781)  time: 7.0003  data: 6.7641  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0524 (1.0702)  acc1: 72.6562 (70.5280)  acc5: 92.5781 (90.2720)  time: 1.5454  data: 1.3529  max mem: 55573
Test: Total time: 0:00:07 (1.5633 s / it)
* Acc@1 71.062 Acc@5 90.510 loss 1.091
Accuracy of the model EMA on 50000 test images: 71.1%
Max EMA accuracy: 71.06%
Epoch: [239]  [  0/156]  eta: 0:34:09  lr: 0.000902  min_lr: 0.000902  loss: 3.3187 (3.3187)  weight_decay: 0.0500 (0.0500)  time: 13.1366  data: 9.9603  max mem: 55573
Epoch: [239]  [ 10/156]  eta: 0:04:12  lr: 0.000900  min_lr: 0.000900  loss: 3.2554 (2.9897)  weight_decay: 0.0500 (0.0500)  time: 1.7317  data: 0.9060  max mem: 55573
Epoch: [239]  [ 20/156]  eta: 0:02:41  lr: 0.000898  min_lr: 0.000898  loss: 2.9940 (2.9762)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [239]  [ 30/156]  eta: 0:02:05  lr: 0.000896  min_lr: 0.000896  loss: 3.0741 (3.0542)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [239]  [ 40/156]  eta: 0:01:44  lr: 0.000894  min_lr: 0.000894  loss: 3.2853 (3.0703)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [239]  [ 50/156]  eta: 0:01:28  lr: 0.000893  min_lr: 0.000893  loss: 3.1620 (3.0719)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [239]  [ 60/156]  eta: 0:01:16  lr: 0.000891  min_lr: 0.000891  loss: 3.0742 (3.0418)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [239]  [ 70/156]  eta: 0:01:06  lr: 0.000889  min_lr: 0.000889  loss: 3.0710 (3.0474)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [239]  [ 80/156]  eta: 0:00:56  lr: 0.000887  min_lr: 0.000887  loss: 3.2591 (3.0641)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [239]  [ 90/156]  eta: 0:00:48  lr: 0.000885  min_lr: 0.000885  loss: 2.9150 (3.0357)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [239]  [100/156]  eta: 0:00:40  lr: 0.000884  min_lr: 0.000884  loss: 2.8181 (3.0337)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [239]  [110/156]  eta: 0:00:32  lr: 0.000882  min_lr: 0.000882  loss: 3.2674 (3.0424)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0005  max mem: 55573
Epoch: [239]  [120/156]  eta: 0:00:25  lr: 0.000880  min_lr: 0.000880  loss: 2.9939 (3.0285)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [239]  [130/156]  eta: 0:00:17  lr: 0.000878  min_lr: 0.000878  loss: 2.9811 (3.0303)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0011  max mem: 55573
Epoch: [239]  [140/156]  eta: 0:00:10  lr: 0.000876  min_lr: 0.000876  loss: 3.1075 (3.0321)  weight_decay: 0.0500 (0.0500)  time: 0.5860  data: 0.0010  max mem: 55573
Epoch: [239]  [150/156]  eta: 0:00:04  lr: 0.000875  min_lr: 0.000875  loss: 3.0747 (3.0241)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [239]  [155/156]  eta: 0:00:00  lr: 0.000874  min_lr: 0.000874  loss: 3.0205 (3.0238)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [239] Total time: 0:01:45 (0.6766 s / it)
Averaged stats: lr: 0.000874  min_lr: 0.000874  loss: 3.0205 (3.0013)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8734 (0.8734)  acc1: 82.9427 (82.9427)  acc5: 96.8750 (96.8750)  time: 6.8741  data: 6.6367  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9880 (0.9537)  acc1: 81.5104 (80.0320)  acc5: 95.8333 (95.3920)  time: 1.5213  data: 1.3274  max mem: 55573
Test: Total time: 0:00:07 (1.5623 s / it)
* Acc@1 80.312 Acc@5 95.354 loss 0.954
Accuracy of the model on the 50000 test images: 80.3%
Max accuracy: 80.36%
Test:  [0/5]  eta: 0:00:34  loss: 1.0231 (1.0231)  acc1: 74.3490 (74.3490)  acc5: 92.5781 (92.5781)  time: 6.8800  data: 6.6439  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0515 (1.0692)  acc1: 72.6562 (70.4320)  acc5: 92.5781 (90.2720)  time: 1.5240  data: 1.3317  max mem: 55573
Test: Total time: 0:00:07 (1.5415 s / it)
* Acc@1 71.088 Acc@5 90.538 loss 1.090
Accuracy of the model EMA on 50000 test images: 71.1%
Max EMA accuracy: 71.09%
Epoch: [240]  [  0/156]  eta: 0:38:12  lr: 0.000874  min_lr: 0.000874  loss: 3.3980 (3.3980)  weight_decay: 0.0500 (0.0500)  time: 14.6971  data: 8.7712  max mem: 55573
Epoch: [240]  [ 10/156]  eta: 0:04:33  lr: 0.000872  min_lr: 0.000872  loss: 2.9356 (2.9409)  weight_decay: 0.0500 (0.0500)  time: 1.8700  data: 0.7977  max mem: 55573
Epoch: [240]  [ 20/156]  eta: 0:02:51  lr: 0.000870  min_lr: 0.000870  loss: 2.9447 (2.9405)  weight_decay: 0.0500 (0.0500)  time: 0.5881  data: 0.0005  max mem: 55573
Epoch: [240]  [ 30/156]  eta: 0:02:11  lr: 0.000868  min_lr: 0.000868  loss: 3.1113 (2.9153)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0005  max mem: 55573
Epoch: [240]  [ 40/156]  eta: 0:01:48  lr: 0.000866  min_lr: 0.000866  loss: 3.0330 (2.9265)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [240]  [ 50/156]  eta: 0:01:31  lr: 0.000865  min_lr: 0.000865  loss: 2.9506 (2.9259)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [240]  [ 60/156]  eta: 0:01:18  lr: 0.000863  min_lr: 0.000863  loss: 2.9410 (2.9047)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [240]  [ 70/156]  eta: 0:01:07  lr: 0.000861  min_lr: 0.000861  loss: 3.0100 (2.9196)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [240]  [ 80/156]  eta: 0:00:58  lr: 0.000859  min_lr: 0.000859  loss: 2.9925 (2.9168)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [240]  [ 90/156]  eta: 0:00:49  lr: 0.000857  min_lr: 0.000857  loss: 3.0491 (2.9331)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [240]  [100/156]  eta: 0:00:40  lr: 0.000856  min_lr: 0.000856  loss: 3.2209 (2.9435)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [240]  [110/156]  eta: 0:00:32  lr: 0.000854  min_lr: 0.000854  loss: 3.1894 (2.9450)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [240]  [120/156]  eta: 0:00:25  lr: 0.000852  min_lr: 0.000852  loss: 3.0388 (2.9504)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [240]  [130/156]  eta: 0:00:18  lr: 0.000850  min_lr: 0.000850  loss: 3.0539 (2.9530)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0026  max mem: 55573
Epoch: [240]  [140/156]  eta: 0:00:11  lr: 0.000849  min_lr: 0.000849  loss: 3.0896 (2.9631)  weight_decay: 0.0500 (0.0500)  time: 0.5862  data: 0.0024  max mem: 55573
Epoch: [240]  [150/156]  eta: 0:00:04  lr: 0.000847  min_lr: 0.000847  loss: 3.1142 (2.9701)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [240]  [155/156]  eta: 0:00:00  lr: 0.000846  min_lr: 0.000846  loss: 3.1142 (2.9636)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [240] Total time: 0:01:47 (0.6863 s / it)
Averaged stats: lr: 0.000846  min_lr: 0.000846  loss: 3.1142 (2.9940)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8776 (0.8776)  acc1: 82.9427 (82.9427)  acc5: 96.7448 (96.7448)  time: 6.8378  data: 6.6001  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9537 (0.9608)  acc1: 80.5990 (79.0400)  acc5: 96.7448 (95.6160)  time: 1.5144  data: 1.3201  max mem: 55573
Test: Total time: 0:00:07 (1.5628 s / it)
* Acc@1 80.326 Acc@5 95.380 loss 0.954
Accuracy of the model on the 50000 test images: 80.3%
Max accuracy: 80.36%
Test:  [0/5]  eta: 0:00:36  loss: 1.0224 (1.0224)  acc1: 74.3490 (74.3490)  acc5: 92.5781 (92.5781)  time: 7.3874  data: 7.1513  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0509 (1.0684)  acc1: 72.6562 (70.4320)  acc5: 92.5781 (90.2720)  time: 1.6228  data: 1.4303  max mem: 55573
Test: Total time: 0:00:08 (1.6483 s / it)
* Acc@1 71.120 Acc@5 90.564 loss 1.089
Accuracy of the model EMA on 50000 test images: 71.1%
Max EMA accuracy: 71.12%
Epoch: [241]  [  0/156]  eta: 0:36:18  lr: 0.000846  min_lr: 0.000846  loss: 3.1426 (3.1426)  weight_decay: 0.0500 (0.0500)  time: 13.9625  data: 13.3665  max mem: 55573
Epoch: [241]  [ 10/156]  eta: 0:04:23  lr: 0.000844  min_lr: 0.000844  loss: 3.1426 (2.9547)  weight_decay: 0.0500 (0.0500)  time: 1.8048  data: 1.2155  max mem: 55573
Epoch: [241]  [ 20/156]  eta: 0:02:46  lr: 0.000842  min_lr: 0.000842  loss: 3.0177 (2.9242)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [241]  [ 30/156]  eta: 0:02:08  lr: 0.000840  min_lr: 0.000840  loss: 3.0799 (2.9111)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [241]  [ 40/156]  eta: 0:01:46  lr: 0.000839  min_lr: 0.000839  loss: 3.1768 (2.9651)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [241]  [ 50/156]  eta: 0:01:30  lr: 0.000837  min_lr: 0.000837  loss: 3.2439 (3.0067)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [241]  [ 60/156]  eta: 0:01:17  lr: 0.000835  min_lr: 0.000835  loss: 3.2378 (3.0225)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [241]  [ 70/156]  eta: 0:01:06  lr: 0.000833  min_lr: 0.000833  loss: 3.1312 (3.0286)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [241]  [ 80/156]  eta: 0:00:57  lr: 0.000832  min_lr: 0.000832  loss: 2.9947 (3.0092)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [241]  [ 90/156]  eta: 0:00:48  lr: 0.000830  min_lr: 0.000830  loss: 2.8429 (3.0064)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [241]  [100/156]  eta: 0:00:40  lr: 0.000828  min_lr: 0.000828  loss: 3.0716 (3.0118)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [241]  [110/156]  eta: 0:00:32  lr: 0.000826  min_lr: 0.000826  loss: 3.2231 (3.0323)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [241]  [120/156]  eta: 0:00:25  lr: 0.000825  min_lr: 0.000825  loss: 3.2362 (3.0376)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [241]  [130/156]  eta: 0:00:17  lr: 0.000823  min_lr: 0.000823  loss: 3.0454 (3.0267)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0010  max mem: 55573
Epoch: [241]  [140/156]  eta: 0:00:10  lr: 0.000821  min_lr: 0.000821  loss: 3.1446 (3.0322)  weight_decay: 0.0500 (0.0500)  time: 0.5847  data: 0.0009  max mem: 55573
Epoch: [241]  [150/156]  eta: 0:00:04  lr: 0.000819  min_lr: 0.000819  loss: 2.9341 (3.0014)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [241]  [155/156]  eta: 0:00:00  lr: 0.000819  min_lr: 0.000819  loss: 2.5782 (2.9863)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [241] Total time: 0:01:46 (0.6815 s / it)
Averaged stats: lr: 0.000819  min_lr: 0.000819  loss: 2.5782 (2.9824)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8157 (0.8157)  acc1: 83.4635 (83.4635)  acc5: 96.8750 (96.8750)  time: 6.9084  data: 6.6708  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9151 (0.8980)  acc1: 79.5573 (79.3920)  acc5: 96.8750 (95.6160)  time: 1.5282  data: 1.3343  max mem: 55573
Test: Total time: 0:00:07 (1.5833 s / it)
* Acc@1 80.422 Acc@5 95.600 loss 0.890
Accuracy of the model on the 50000 test images: 80.4%
Max accuracy: 80.42%
Test:  [0/5]  eta: 0:00:33  loss: 1.0214 (1.0214)  acc1: 74.3490 (74.3490)  acc5: 92.4479 (92.4479)  time: 6.7254  data: 6.4893  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0500 (1.0676)  acc1: 72.9167 (70.5280)  acc5: 92.4479 (90.2400)  time: 1.4904  data: 1.2980  max mem: 55573
Test: Total time: 0:00:07 (1.5079 s / it)
* Acc@1 71.148 Acc@5 90.576 loss 1.088
Accuracy of the model EMA on 50000 test images: 71.1%
Max EMA accuracy: 71.15%
Epoch: [242]  [  0/156]  eta: 0:31:25  lr: 0.000818  min_lr: 0.000818  loss: 2.7945 (2.7945)  weight_decay: 0.0500 (0.0500)  time: 12.0856  data: 8.9755  max mem: 55573
Epoch: [242]  [ 10/156]  eta: 0:04:02  lr: 0.000817  min_lr: 0.000817  loss: 3.2077 (3.0510)  weight_decay: 0.0500 (0.0500)  time: 1.6632  data: 0.8297  max mem: 55573
Epoch: [242]  [ 20/156]  eta: 0:02:36  lr: 0.000815  min_lr: 0.000815  loss: 3.2077 (3.0223)  weight_decay: 0.0500 (0.0500)  time: 0.6051  data: 0.0078  max mem: 55573
Epoch: [242]  [ 30/156]  eta: 0:02:02  lr: 0.000813  min_lr: 0.000813  loss: 2.8132 (3.0070)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [242]  [ 40/156]  eta: 0:01:41  lr: 0.000811  min_lr: 0.000811  loss: 2.9504 (2.9950)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [242]  [ 50/156]  eta: 0:01:27  lr: 0.000810  min_lr: 0.000810  loss: 3.0321 (3.0186)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [242]  [ 60/156]  eta: 0:01:15  lr: 0.000808  min_lr: 0.000808  loss: 3.0670 (2.9828)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0005  max mem: 55573
Epoch: [242]  [ 70/156]  eta: 0:01:04  lr: 0.000806  min_lr: 0.000806  loss: 2.9167 (2.9859)  weight_decay: 0.0500 (0.0500)  time: 0.5882  data: 0.0005  max mem: 55573
Epoch: [242]  [ 80/156]  eta: 0:00:55  lr: 0.000805  min_lr: 0.000805  loss: 2.9167 (2.9679)  weight_decay: 0.0500 (0.0500)  time: 0.5887  data: 0.0005  max mem: 55573
Epoch: [242]  [ 90/156]  eta: 0:00:47  lr: 0.000803  min_lr: 0.000803  loss: 2.8917 (2.9524)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [242]  [100/156]  eta: 0:00:39  lr: 0.000801  min_lr: 0.000801  loss: 3.0962 (2.9727)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [242]  [110/156]  eta: 0:00:32  lr: 0.000799  min_lr: 0.000799  loss: 3.1545 (2.9711)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [242]  [120/156]  eta: 0:00:24  lr: 0.000798  min_lr: 0.000798  loss: 3.1514 (2.9780)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [242]  [130/156]  eta: 0:00:17  lr: 0.000796  min_lr: 0.000796  loss: 3.2966 (2.9971)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0010  max mem: 55573
Epoch: [242]  [140/156]  eta: 0:00:10  lr: 0.000794  min_lr: 0.000794  loss: 3.2589 (2.9956)  weight_decay: 0.0500 (0.0500)  time: 0.5844  data: 0.0008  max mem: 55573
Epoch: [242]  [150/156]  eta: 0:00:03  lr: 0.000792  min_lr: 0.000792  loss: 3.1391 (2.9873)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [242]  [155/156]  eta: 0:00:00  lr: 0.000792  min_lr: 0.000792  loss: 3.1605 (2.9902)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [242] Total time: 0:01:44 (0.6705 s / it)
Averaged stats: lr: 0.000792  min_lr: 0.000792  loss: 3.1605 (2.9962)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9076 (0.9076)  acc1: 82.1615 (82.1615)  acc5: 96.6146 (96.6146)  time: 6.8576  data: 6.6203  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9609 (0.9641)  acc1: 81.2500 (79.2320)  acc5: 96.6146 (95.5840)  time: 1.5179  data: 1.3241  max mem: 55573
Test: Total time: 0:00:07 (1.5607 s / it)
* Acc@1 80.460 Acc@5 95.506 loss 0.952
Accuracy of the model on the 50000 test images: 80.5%
Max accuracy: 80.46%
Test:  [0/5]  eta: 0:00:33  loss: 1.0207 (1.0207)  acc1: 74.4792 (74.4792)  acc5: 92.4479 (92.4479)  time: 6.7189  data: 6.4829  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0494 (1.0670)  acc1: 72.9167 (70.4960)  acc5: 92.4479 (90.2400)  time: 1.4891  data: 1.2967  max mem: 55573
Test: Total time: 0:00:07 (1.5063 s / it)
* Acc@1 71.186 Acc@5 90.580 loss 1.087
Accuracy of the model EMA on 50000 test images: 71.2%
Max EMA accuracy: 71.19%
Epoch: [243]  [  0/156]  eta: 0:32:13  lr: 0.000791  min_lr: 0.000791  loss: 2.8609 (2.8609)  weight_decay: 0.0500 (0.0500)  time: 12.3911  data: 11.6840  max mem: 55573
Epoch: [243]  [ 10/156]  eta: 0:04:04  lr: 0.000790  min_lr: 0.000790  loss: 3.1559 (2.9092)  weight_decay: 0.0500 (0.0500)  time: 1.6719  data: 1.0627  max mem: 55573
Epoch: [243]  [ 20/156]  eta: 0:02:37  lr: 0.000788  min_lr: 0.000788  loss: 3.1859 (2.9610)  weight_decay: 0.0500 (0.0500)  time: 0.5949  data: 0.0005  max mem: 55573
Epoch: [243]  [ 30/156]  eta: 0:02:02  lr: 0.000786  min_lr: 0.000786  loss: 3.0720 (2.9136)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [243]  [ 40/156]  eta: 0:01:42  lr: 0.000785  min_lr: 0.000785  loss: 2.7843 (2.8974)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [243]  [ 50/156]  eta: 0:01:27  lr: 0.000783  min_lr: 0.000783  loss: 2.9405 (2.8930)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [243]  [ 60/156]  eta: 0:01:15  lr: 0.000781  min_lr: 0.000781  loss: 3.0915 (2.9048)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [243]  [ 70/156]  eta: 0:01:05  lr: 0.000779  min_lr: 0.000779  loss: 3.0915 (2.9151)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [243]  [ 80/156]  eta: 0:00:56  lr: 0.000778  min_lr: 0.000778  loss: 2.8911 (2.8938)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [243]  [ 90/156]  eta: 0:00:47  lr: 0.000776  min_lr: 0.000776  loss: 3.0735 (2.9255)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [243]  [100/156]  eta: 0:00:39  lr: 0.000774  min_lr: 0.000774  loss: 3.0735 (2.9047)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0004  max mem: 55573
Epoch: [243]  [110/156]  eta: 0:00:32  lr: 0.000773  min_lr: 0.000773  loss: 2.9475 (2.9097)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0005  max mem: 55573
Epoch: [243]  [120/156]  eta: 0:00:24  lr: 0.000771  min_lr: 0.000771  loss: 2.9738 (2.9108)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [243]  [130/156]  eta: 0:00:17  lr: 0.000769  min_lr: 0.000769  loss: 2.9585 (2.9037)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0012  max mem: 55573
Epoch: [243]  [140/156]  eta: 0:00:10  lr: 0.000768  min_lr: 0.000768  loss: 3.1787 (2.9112)  weight_decay: 0.0500 (0.0500)  time: 0.5852  data: 0.0010  max mem: 55573
Epoch: [243]  [150/156]  eta: 0:00:04  lr: 0.000766  min_lr: 0.000766  loss: 2.9596 (2.9108)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [243]  [155/156]  eta: 0:00:00  lr: 0.000765  min_lr: 0.000765  loss: 2.9596 (2.9114)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [243] Total time: 0:01:44 (0.6715 s / it)
Averaged stats: lr: 0.000765  min_lr: 0.000765  loss: 2.9596 (2.9675)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.7999 (0.7999)  acc1: 84.6354 (84.6354)  acc5: 97.2656 (97.2656)  time: 6.9459  data: 6.7085  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9187 (0.8826)  acc1: 81.1198 (80.1600)  acc5: 95.9635 (95.6160)  time: 1.5356  data: 1.3418  max mem: 55573
Test: Total time: 0:00:07 (1.5738 s / it)
* Acc@1 80.684 Acc@5 95.530 loss 0.874
Accuracy of the model on the 50000 test images: 80.7%
Max accuracy: 80.68%
Test:  [0/5]  eta: 0:00:34  loss: 1.0196 (1.0196)  acc1: 74.4792 (74.4792)  acc5: 92.5781 (92.5781)  time: 6.8698  data: 6.6338  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0485 (1.0660)  acc1: 72.9167 (70.4320)  acc5: 92.5781 (90.2720)  time: 1.5194  data: 1.3269  max mem: 55573
Test: Total time: 0:00:07 (1.5341 s / it)
* Acc@1 71.206 Acc@5 90.590 loss 1.086
Accuracy of the model EMA on 50000 test images: 71.2%
Max EMA accuracy: 71.21%
Epoch: [244]  [  0/156]  eta: 0:37:17  lr: 0.000765  min_lr: 0.000765  loss: 3.3999 (3.3999)  weight_decay: 0.0500 (0.0500)  time: 14.3454  data: 10.2577  max mem: 55573
Epoch: [244]  [ 10/156]  eta: 0:04:29  lr: 0.000763  min_lr: 0.000763  loss: 3.2380 (2.9095)  weight_decay: 0.0500 (0.0500)  time: 1.8445  data: 0.9329  max mem: 55573
Epoch: [244]  [ 20/156]  eta: 0:02:49  lr: 0.000761  min_lr: 0.000761  loss: 2.8900 (2.8631)  weight_decay: 0.0500 (0.0500)  time: 0.5940  data: 0.0004  max mem: 55573
Epoch: [244]  [ 30/156]  eta: 0:02:10  lr: 0.000760  min_lr: 0.000760  loss: 2.9914 (2.9104)  weight_decay: 0.0500 (0.0500)  time: 0.5935  data: 0.0004  max mem: 55573
Epoch: [244]  [ 40/156]  eta: 0:01:47  lr: 0.000758  min_lr: 0.000758  loss: 3.1919 (2.9976)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0005  max mem: 55573
Epoch: [244]  [ 50/156]  eta: 0:01:31  lr: 0.000756  min_lr: 0.000756  loss: 3.1919 (3.0007)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [244]  [ 60/156]  eta: 0:01:18  lr: 0.000755  min_lr: 0.000755  loss: 3.2072 (3.0139)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [244]  [ 70/156]  eta: 0:01:07  lr: 0.000753  min_lr: 0.000753  loss: 3.2063 (3.0056)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [244]  [ 80/156]  eta: 0:00:57  lr: 0.000751  min_lr: 0.000751  loss: 3.1858 (3.0098)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [244]  [ 90/156]  eta: 0:00:49  lr: 0.000750  min_lr: 0.000750  loss: 3.2160 (3.0297)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [244]  [100/156]  eta: 0:00:40  lr: 0.000748  min_lr: 0.000748  loss: 3.2142 (3.0297)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0004  max mem: 55573
Epoch: [244]  [110/156]  eta: 0:00:32  lr: 0.000746  min_lr: 0.000746  loss: 3.0204 (3.0280)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [244]  [120/156]  eta: 0:00:25  lr: 0.000745  min_lr: 0.000745  loss: 2.9828 (3.0102)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [244]  [130/156]  eta: 0:00:18  lr: 0.000743  min_lr: 0.000743  loss: 2.9828 (3.0103)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0011  max mem: 55573
Epoch: [244]  [140/156]  eta: 0:00:11  lr: 0.000741  min_lr: 0.000741  loss: 3.0697 (3.0116)  weight_decay: 0.0500 (0.0500)  time: 0.5862  data: 0.0009  max mem: 55573
Epoch: [244]  [150/156]  eta: 0:00:04  lr: 0.000740  min_lr: 0.000740  loss: 3.2031 (3.0209)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [244]  [155/156]  eta: 0:00:00  lr: 0.000739  min_lr: 0.000739  loss: 3.1393 (3.0113)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [244] Total time: 0:01:46 (0.6852 s / it)
Averaged stats: lr: 0.000739  min_lr: 0.000739  loss: 3.1393 (2.9743)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.8708 (0.8708)  acc1: 83.2031 (83.2031)  acc5: 96.3542 (96.3542)  time: 7.0651  data: 6.8277  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9747 (0.9723)  acc1: 81.2500 (79.9680)  acc5: 96.3542 (95.6480)  time: 1.5593  data: 1.3656  max mem: 55573
Test: Total time: 0:00:08 (1.6038 s / it)
* Acc@1 80.662 Acc@5 95.568 loss 0.961
Accuracy of the model on the 50000 test images: 80.7%
Max accuracy: 80.68%
Test:  [0/5]  eta: 0:00:35  loss: 1.0190 (1.0190)  acc1: 74.4792 (74.4792)  acc5: 92.5781 (92.5781)  time: 7.0807  data: 6.8446  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0479 (1.0654)  acc1: 72.9167 (70.4960)  acc5: 92.5781 (90.2400)  time: 1.5614  data: 1.3690  max mem: 55573
Test: Total time: 0:00:07 (1.5852 s / it)
* Acc@1 71.222 Acc@5 90.616 loss 1.086
Accuracy of the model EMA on 50000 test images: 71.2%
Max EMA accuracy: 71.22%
Epoch: [245]  [  0/156]  eta: 0:35:58  lr: 0.000739  min_lr: 0.000739  loss: 3.3691 (3.3691)  weight_decay: 0.0500 (0.0500)  time: 13.8396  data: 11.0724  max mem: 55573
Epoch: [245]  [ 10/156]  eta: 0:04:23  lr: 0.000737  min_lr: 0.000737  loss: 3.0555 (2.9300)  weight_decay: 0.0500 (0.0500)  time: 1.8048  data: 1.0069  max mem: 55573
Epoch: [245]  [ 20/156]  eta: 0:02:46  lr: 0.000735  min_lr: 0.000735  loss: 3.1110 (3.0389)  weight_decay: 0.0500 (0.0500)  time: 0.5964  data: 0.0004  max mem: 55573
Epoch: [245]  [ 30/156]  eta: 0:02:08  lr: 0.000734  min_lr: 0.000734  loss: 3.1110 (2.9886)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [245]  [ 40/156]  eta: 0:01:46  lr: 0.000732  min_lr: 0.000732  loss: 3.0408 (3.0518)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [245]  [ 50/156]  eta: 0:01:30  lr: 0.000730  min_lr: 0.000730  loss: 3.2732 (3.0539)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [245]  [ 60/156]  eta: 0:01:17  lr: 0.000729  min_lr: 0.000729  loss: 3.2088 (3.0512)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [245]  [ 70/156]  eta: 0:01:06  lr: 0.000727  min_lr: 0.000727  loss: 3.0619 (3.0374)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [245]  [ 80/156]  eta: 0:00:57  lr: 0.000725  min_lr: 0.000725  loss: 2.9237 (3.0093)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [245]  [ 90/156]  eta: 0:00:48  lr: 0.000724  min_lr: 0.000724  loss: 2.8219 (2.9697)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [245]  [100/156]  eta: 0:00:40  lr: 0.000722  min_lr: 0.000722  loss: 2.8943 (2.9644)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [245]  [110/156]  eta: 0:00:32  lr: 0.000720  min_lr: 0.000720  loss: 2.9016 (2.9546)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [245]  [120/156]  eta: 0:00:25  lr: 0.000719  min_lr: 0.000719  loss: 2.9137 (2.9537)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [245]  [130/156]  eta: 0:00:18  lr: 0.000717  min_lr: 0.000717  loss: 3.0493 (2.9458)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0013  max mem: 55573
Epoch: [245]  [140/156]  eta: 0:00:10  lr: 0.000716  min_lr: 0.000716  loss: 2.9955 (2.9513)  weight_decay: 0.0500 (0.0500)  time: 0.5857  data: 0.0012  max mem: 55573
Epoch: [245]  [150/156]  eta: 0:00:04  lr: 0.000714  min_lr: 0.000714  loss: 3.0130 (2.9527)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [245]  [155/156]  eta: 0:00:00  lr: 0.000713  min_lr: 0.000713  loss: 3.0130 (2.9489)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [245] Total time: 0:01:46 (0.6814 s / it)
Averaged stats: lr: 0.000713  min_lr: 0.000713  loss: 3.0130 (2.9844)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.8974 (0.8974)  acc1: 83.3333 (83.3333)  acc5: 96.4844 (96.4844)  time: 6.6254  data: 6.3882  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9883 (0.9832)  acc1: 80.8594 (80.2880)  acc5: 96.3542 (95.3280)  time: 1.4703  data: 1.2777  max mem: 55573
Test: Total time: 0:00:07 (1.4867 s / it)
* Acc@1 80.540 Acc@5 95.402 loss 0.979
Accuracy of the model on the 50000 test images: 80.5%
Max accuracy: 80.68%
Test:  [0/5]  eta: 0:00:35  loss: 1.0181 (1.0181)  acc1: 74.3490 (74.3490)  acc5: 92.5781 (92.5781)  time: 7.1249  data: 6.8885  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0471 (1.0646)  acc1: 73.0469 (70.4960)  acc5: 92.5781 (90.2400)  time: 1.5702  data: 1.3778  max mem: 55573
Test: Total time: 0:00:07 (1.5937 s / it)
* Acc@1 71.258 Acc@5 90.636 loss 1.085
Accuracy of the model EMA on 50000 test images: 71.3%
Max EMA accuracy: 71.26%
Epoch: [246]  [  0/156]  eta: 0:33:02  lr: 0.000713  min_lr: 0.000713  loss: 2.6407 (2.6407)  weight_decay: 0.0500 (0.0500)  time: 12.7065  data: 9.8186  max mem: 55573
Epoch: [246]  [ 10/156]  eta: 0:04:08  lr: 0.000711  min_lr: 0.000711  loss: 3.1739 (2.9320)  weight_decay: 0.0500 (0.0500)  time: 1.7017  data: 0.8931  max mem: 55573
Epoch: [246]  [ 20/156]  eta: 0:02:39  lr: 0.000710  min_lr: 0.000710  loss: 3.1739 (2.9681)  weight_decay: 0.0500 (0.0500)  time: 0.5973  data: 0.0005  max mem: 55573
Epoch: [246]  [ 30/156]  eta: 0:02:04  lr: 0.000708  min_lr: 0.000708  loss: 3.1449 (2.9013)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0004  max mem: 55573
Epoch: [246]  [ 40/156]  eta: 0:01:43  lr: 0.000706  min_lr: 0.000706  loss: 2.8731 (2.8855)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [246]  [ 50/156]  eta: 0:01:28  lr: 0.000705  min_lr: 0.000705  loss: 3.0080 (2.9249)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [246]  [ 60/156]  eta: 0:01:15  lr: 0.000703  min_lr: 0.000703  loss: 3.2292 (2.9537)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [246]  [ 70/156]  eta: 0:01:05  lr: 0.000701  min_lr: 0.000701  loss: 3.0743 (2.9324)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [246]  [ 80/156]  eta: 0:00:56  lr: 0.000700  min_lr: 0.000700  loss: 2.9237 (2.9351)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [246]  [ 90/156]  eta: 0:00:47  lr: 0.000698  min_lr: 0.000698  loss: 2.9237 (2.9314)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [246]  [100/156]  eta: 0:00:39  lr: 0.000697  min_lr: 0.000697  loss: 3.0877 (2.9437)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [246]  [110/156]  eta: 0:00:32  lr: 0.000695  min_lr: 0.000695  loss: 2.9590 (2.9354)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [246]  [120/156]  eta: 0:00:24  lr: 0.000693  min_lr: 0.000693  loss: 2.8653 (2.9331)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [246]  [130/156]  eta: 0:00:17  lr: 0.000692  min_lr: 0.000692  loss: 3.1678 (2.9456)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0010  max mem: 55573
Epoch: [246]  [140/156]  eta: 0:00:10  lr: 0.000690  min_lr: 0.000690  loss: 3.2163 (2.9557)  weight_decay: 0.0500 (0.0500)  time: 0.5872  data: 0.0009  max mem: 55573
Epoch: [246]  [150/156]  eta: 0:00:04  lr: 0.000689  min_lr: 0.000689  loss: 3.1388 (2.9570)  weight_decay: 0.0500 (0.0500)  time: 0.5830  data: 0.0001  max mem: 55573
Epoch: [246]  [155/156]  eta: 0:00:00  lr: 0.000688  min_lr: 0.000688  loss: 3.1392 (2.9550)  weight_decay: 0.0500 (0.0500)  time: 0.5829  data: 0.0002  max mem: 55573
Epoch: [246] Total time: 0:01:45 (0.6748 s / it)
Averaged stats: lr: 0.000688  min_lr: 0.000688  loss: 3.1392 (2.9773)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:32  loss: 0.8925 (0.8925)  acc1: 83.3333 (83.3333)  acc5: 95.9635 (95.9635)  time: 6.5925  data: 6.3554  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9951 (0.9806)  acc1: 81.5104 (79.8080)  acc5: 95.9635 (95.4240)  time: 1.4638  data: 1.2712  max mem: 55573
Test: Total time: 0:00:07 (1.4982 s / it)
* Acc@1 80.752 Acc@5 95.538 loss 0.972
Accuracy of the model on the 50000 test images: 80.8%
Max accuracy: 80.75%
Test:  [0/5]  eta: 0:00:34  loss: 1.0175 (1.0175)  acc1: 74.4792 (74.4792)  acc5: 92.5781 (92.5781)  time: 6.8623  data: 6.6263  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0464 (1.0640)  acc1: 72.9167 (70.4960)  acc5: 92.5781 (90.2720)  time: 1.5177  data: 1.3254  max mem: 55573
Test: Total time: 0:00:07 (1.5398 s / it)
* Acc@1 71.284 Acc@5 90.654 loss 1.084
Accuracy of the model EMA on 50000 test images: 71.3%
Max EMA accuracy: 71.28%
Epoch: [247]  [  0/156]  eta: 0:33:39  lr: 0.000688  min_lr: 0.000688  loss: 2.8782 (2.8782)  weight_decay: 0.0500 (0.0500)  time: 12.9427  data: 12.3420  max mem: 55573
Epoch: [247]  [ 10/156]  eta: 0:04:10  lr: 0.000686  min_lr: 0.000686  loss: 3.2311 (2.9350)  weight_decay: 0.0500 (0.0500)  time: 1.7158  data: 1.1225  max mem: 55573
Epoch: [247]  [ 20/156]  eta: 0:02:40  lr: 0.000684  min_lr: 0.000684  loss: 2.7026 (2.8100)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [247]  [ 30/156]  eta: 0:02:04  lr: 0.000683  min_lr: 0.000683  loss: 2.8381 (2.8364)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [247]  [ 40/156]  eta: 0:01:43  lr: 0.000681  min_lr: 0.000681  loss: 2.9502 (2.8785)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [247]  [ 50/156]  eta: 0:01:28  lr: 0.000680  min_lr: 0.000680  loss: 3.0865 (2.9118)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [247]  [ 60/156]  eta: 0:01:16  lr: 0.000678  min_lr: 0.000678  loss: 3.0022 (2.8722)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [247]  [ 70/156]  eta: 0:01:05  lr: 0.000676  min_lr: 0.000676  loss: 2.9411 (2.8970)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [247]  [ 80/156]  eta: 0:00:56  lr: 0.000675  min_lr: 0.000675  loss: 3.0902 (2.8944)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [247]  [ 90/156]  eta: 0:00:47  lr: 0.000673  min_lr: 0.000673  loss: 2.9805 (2.9011)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [247]  [100/156]  eta: 0:00:39  lr: 0.000672  min_lr: 0.000672  loss: 2.9805 (2.8947)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [247]  [110/156]  eta: 0:00:32  lr: 0.000670  min_lr: 0.000670  loss: 2.9990 (2.8989)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [247]  [120/156]  eta: 0:00:24  lr: 0.000668  min_lr: 0.000668  loss: 2.9809 (2.8979)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0004  max mem: 55573
Epoch: [247]  [130/156]  eta: 0:00:17  lr: 0.000667  min_lr: 0.000667  loss: 2.9809 (2.9025)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0010  max mem: 55573
Epoch: [247]  [140/156]  eta: 0:00:10  lr: 0.000665  min_lr: 0.000665  loss: 3.1077 (2.9055)  weight_decay: 0.0500 (0.0500)  time: 0.5869  data: 0.0009  max mem: 55573
Epoch: [247]  [150/156]  eta: 0:00:04  lr: 0.000664  min_lr: 0.000664  loss: 3.0374 (2.9037)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0001  max mem: 55573
Epoch: [247]  [155/156]  eta: 0:00:00  lr: 0.000663  min_lr: 0.000663  loss: 3.0729 (2.9079)  weight_decay: 0.0500 (0.0500)  time: 0.5830  data: 0.0001  max mem: 55573
Epoch: [247] Total time: 0:01:45 (0.6750 s / it)
Averaged stats: lr: 0.000663  min_lr: 0.000663  loss: 3.0729 (2.9592)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8610 (0.8610)  acc1: 83.8542 (83.8542)  acc5: 96.7448 (96.7448)  time: 6.9495  data: 6.7121  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9372 (0.9487)  acc1: 81.3802 (79.8080)  acc5: 96.6146 (95.5520)  time: 1.5364  data: 1.3425  max mem: 55573
Test: Total time: 0:00:07 (1.5826 s / it)
* Acc@1 80.926 Acc@5 95.540 loss 0.929
Accuracy of the model on the 50000 test images: 80.9%
Max accuracy: 80.93%
Test:  [0/5]  eta: 0:00:32  loss: 1.0165 (1.0165)  acc1: 74.6094 (74.6094)  acc5: 92.9688 (92.9688)  time: 6.5870  data: 6.3509  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0455 (1.0631)  acc1: 72.9167 (70.5920)  acc5: 92.9688 (90.3680)  time: 1.4631  data: 1.2703  max mem: 55573
Test: Total time: 0:00:07 (1.4784 s / it)
* Acc@1 71.322 Acc@5 90.690 loss 1.084
Accuracy of the model EMA on 50000 test images: 71.3%
Max EMA accuracy: 71.32%
Epoch: [248]  [  0/156]  eta: 0:31:02  lr: 0.000663  min_lr: 0.000663  loss: 2.1806 (2.1806)  weight_decay: 0.0500 (0.0500)  time: 11.9395  data: 10.6523  max mem: 55573
Epoch: [248]  [ 10/156]  eta: 0:04:08  lr: 0.000661  min_lr: 0.000661  loss: 2.9802 (2.9472)  weight_decay: 0.0500 (0.0500)  time: 1.7016  data: 0.9688  max mem: 55573
Epoch: [248]  [ 20/156]  eta: 0:02:39  lr: 0.000659  min_lr: 0.000659  loss: 2.9802 (2.9646)  weight_decay: 0.0500 (0.0500)  time: 0.6342  data: 0.0004  max mem: 55573
Epoch: [248]  [ 30/156]  eta: 0:02:04  lr: 0.000658  min_lr: 0.000658  loss: 3.1493 (2.9802)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [248]  [ 40/156]  eta: 0:01:43  lr: 0.000656  min_lr: 0.000656  loss: 3.2330 (2.9500)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [248]  [ 50/156]  eta: 0:01:27  lr: 0.000655  min_lr: 0.000655  loss: 3.2184 (3.0073)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [248]  [ 60/156]  eta: 0:01:15  lr: 0.000653  min_lr: 0.000653  loss: 3.1678 (2.9907)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [248]  [ 70/156]  eta: 0:01:05  lr: 0.000652  min_lr: 0.000652  loss: 3.0659 (2.9861)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [248]  [ 80/156]  eta: 0:00:56  lr: 0.000650  min_lr: 0.000650  loss: 2.8058 (2.9423)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [248]  [ 90/156]  eta: 0:00:47  lr: 0.000648  min_lr: 0.000648  loss: 2.8035 (2.9423)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [248]  [100/156]  eta: 0:00:39  lr: 0.000647  min_lr: 0.000647  loss: 3.0561 (2.9374)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [248]  [110/156]  eta: 0:00:32  lr: 0.000645  min_lr: 0.000645  loss: 3.1550 (2.9463)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [248]  [120/156]  eta: 0:00:24  lr: 0.000644  min_lr: 0.000644  loss: 3.1226 (2.9425)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [248]  [130/156]  eta: 0:00:17  lr: 0.000642  min_lr: 0.000642  loss: 2.8107 (2.9193)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0010  max mem: 55573
Epoch: [248]  [140/156]  eta: 0:00:10  lr: 0.000641  min_lr: 0.000641  loss: 2.5655 (2.8975)  weight_decay: 0.0500 (0.0500)  time: 0.5863  data: 0.0008  max mem: 55573
Epoch: [248]  [150/156]  eta: 0:00:04  lr: 0.000639  min_lr: 0.000639  loss: 2.6729 (2.8937)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [248]  [155/156]  eta: 0:00:00  lr: 0.000638  min_lr: 0.000638  loss: 2.6904 (2.8941)  weight_decay: 0.0500 (0.0500)  time: 0.5831  data: 0.0001  max mem: 55573
Epoch: [248] Total time: 0:01:45 (0.6744 s / it)
Averaged stats: lr: 0.000638  min_lr: 0.000638  loss: 2.6904 (2.9557)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.8047 (0.8047)  acc1: 83.7240 (83.7240)  acc5: 96.3542 (96.3542)  time: 6.7582  data: 6.5212  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8819 (0.8884)  acc1: 81.2500 (79.7440)  acc5: 96.3542 (95.4240)  time: 1.4980  data: 1.3043  max mem: 55573
Test: Total time: 0:00:07 (1.5450 s / it)
* Acc@1 80.788 Acc@5 95.544 loss 0.869
Accuracy of the model on the 50000 test images: 80.8%
Max accuracy: 80.93%
Test:  [0/5]  eta: 0:00:36  loss: 1.0159 (1.0159)  acc1: 74.7396 (74.7396)  acc5: 92.8385 (92.8385)  time: 7.2882  data: 7.0522  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0448 (1.0624)  acc1: 72.9167 (70.5920)  acc5: 92.8385 (90.4000)  time: 1.6028  data: 1.4105  max mem: 55573
Test: Total time: 0:00:08 (1.6242 s / it)
* Acc@1 71.350 Acc@5 90.690 loss 1.083
Accuracy of the model EMA on 50000 test images: 71.4%
Max EMA accuracy: 71.35%
Epoch: [249]  [  0/156]  eta: 0:34:06  lr: 0.000638  min_lr: 0.000638  loss: 3.1490 (3.1490)  weight_decay: 0.0500 (0.0500)  time: 13.1209  data: 9.6996  max mem: 55573
Epoch: [249]  [ 10/156]  eta: 0:04:12  lr: 0.000637  min_lr: 0.000637  loss: 2.9597 (2.7495)  weight_decay: 0.0500 (0.0500)  time: 1.7328  data: 0.8822  max mem: 55573
Epoch: [249]  [ 20/156]  eta: 0:02:41  lr: 0.000635  min_lr: 0.000635  loss: 2.9597 (2.8714)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0005  max mem: 55573
Epoch: [249]  [ 30/156]  eta: 0:02:05  lr: 0.000633  min_lr: 0.000633  loss: 3.0604 (2.9169)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [249]  [ 40/156]  eta: 0:01:44  lr: 0.000632  min_lr: 0.000632  loss: 3.1562 (2.9564)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [249]  [ 50/156]  eta: 0:01:28  lr: 0.000630  min_lr: 0.000630  loss: 3.2065 (3.0046)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [249]  [ 60/156]  eta: 0:01:16  lr: 0.000629  min_lr: 0.000629  loss: 3.1668 (3.0245)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0005  max mem: 55573
Epoch: [249]  [ 70/156]  eta: 0:01:06  lr: 0.000627  min_lr: 0.000627  loss: 3.1482 (3.0206)  weight_decay: 0.0500 (0.0500)  time: 0.5935  data: 0.0005  max mem: 55573
Epoch: [249]  [ 80/156]  eta: 0:00:56  lr: 0.000626  min_lr: 0.000626  loss: 3.0365 (2.9783)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [249]  [ 90/156]  eta: 0:00:48  lr: 0.000624  min_lr: 0.000624  loss: 2.9328 (2.9896)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [249]  [100/156]  eta: 0:00:40  lr: 0.000623  min_lr: 0.000623  loss: 3.0039 (2.9723)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [249]  [110/156]  eta: 0:00:32  lr: 0.000621  min_lr: 0.000621  loss: 3.0492 (2.9870)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0005  max mem: 55573
Epoch: [249]  [120/156]  eta: 0:00:25  lr: 0.000620  min_lr: 0.000620  loss: 3.2353 (2.9849)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [249]  [130/156]  eta: 0:00:17  lr: 0.000618  min_lr: 0.000618  loss: 2.9726 (2.9711)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0010  max mem: 55573
Epoch: [249]  [140/156]  eta: 0:00:10  lr: 0.000616  min_lr: 0.000616  loss: 2.9929 (2.9847)  weight_decay: 0.0500 (0.0500)  time: 0.5857  data: 0.0009  max mem: 55573
Epoch: [249]  [150/156]  eta: 0:00:04  lr: 0.000615  min_lr: 0.000615  loss: 3.1881 (2.9904)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [249]  [155/156]  eta: 0:00:00  lr: 0.000614  min_lr: 0.000614  loss: 3.0254 (2.9835)  weight_decay: 0.0500 (0.0500)  time: 0.5812  data: 0.0001  max mem: 55573
Epoch: [249] Total time: 0:01:45 (0.6766 s / it)
Averaged stats: lr: 0.000614  min_lr: 0.000614  loss: 3.0254 (2.9614)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8010 (0.8010)  acc1: 82.4219 (82.4219)  acc5: 96.2240 (96.2240)  time: 6.8848  data: 6.6453  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8894 (0.8971)  acc1: 81.1198 (79.6480)  acc5: 96.2240 (95.2640)  time: 1.5238  data: 1.3292  max mem: 55573
Test: Total time: 0:00:07 (1.5682 s / it)
* Acc@1 80.894 Acc@5 95.680 loss 0.881
Accuracy of the model on the 50000 test images: 80.9%
Max accuracy: 80.93%
Test:  [0/5]  eta: 0:00:33  loss: 1.0154 (1.0154)  acc1: 74.7396 (74.7396)  acc5: 92.8385 (92.8385)  time: 6.7184  data: 6.4823  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0444 (1.0619)  acc1: 73.0469 (70.5920)  acc5: 92.8385 (90.4000)  time: 1.4890  data: 1.2966  max mem: 55573
Test: Total time: 0:00:07 (1.5152 s / it)
* Acc@1 71.372 Acc@5 90.708 loss 1.082
Accuracy of the model EMA on 50000 test images: 71.4%
Max EMA accuracy: 71.37%
Epoch: [250]  [  0/156]  eta: 0:33:51  lr: 0.000614  min_lr: 0.000614  loss: 3.3445 (3.3445)  weight_decay: 0.0500 (0.0500)  time: 13.0222  data: 8.0548  max mem: 55573
Epoch: [250]  [ 10/156]  eta: 0:04:12  lr: 0.000612  min_lr: 0.000612  loss: 2.8357 (2.8452)  weight_decay: 0.0500 (0.0500)  time: 1.7300  data: 0.7327  max mem: 55573
Epoch: [250]  [ 20/156]  eta: 0:02:41  lr: 0.000611  min_lr: 0.000611  loss: 2.6911 (2.7816)  weight_decay: 0.0500 (0.0500)  time: 0.5948  data: 0.0004  max mem: 55573
Epoch: [250]  [ 30/156]  eta: 0:02:05  lr: 0.000609  min_lr: 0.000609  loss: 2.9856 (2.8439)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [250]  [ 40/156]  eta: 0:01:43  lr: 0.000608  min_lr: 0.000608  loss: 3.1567 (2.8695)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0005  max mem: 55573
Epoch: [250]  [ 50/156]  eta: 0:01:28  lr: 0.000606  min_lr: 0.000606  loss: 3.2278 (2.9350)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [250]  [ 60/156]  eta: 0:01:16  lr: 0.000605  min_lr: 0.000605  loss: 3.1864 (2.9150)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [250]  [ 70/156]  eta: 0:01:05  lr: 0.000603  min_lr: 0.000603  loss: 3.1155 (2.9388)  weight_decay: 0.0500 (0.0500)  time: 0.5887  data: 0.0004  max mem: 55573
Epoch: [250]  [ 80/156]  eta: 0:00:56  lr: 0.000602  min_lr: 0.000602  loss: 3.1273 (2.9613)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [250]  [ 90/156]  eta: 0:00:47  lr: 0.000600  min_lr: 0.000600  loss: 3.1423 (2.9734)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [250]  [100/156]  eta: 0:00:39  lr: 0.000599  min_lr: 0.000599  loss: 3.1060 (2.9617)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [250]  [110/156]  eta: 0:00:32  lr: 0.000597  min_lr: 0.000597  loss: 3.1479 (2.9794)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [250]  [120/156]  eta: 0:00:24  lr: 0.000596  min_lr: 0.000596  loss: 3.1479 (2.9812)  weight_decay: 0.0500 (0.0500)  time: 0.5958  data: 0.0004  max mem: 55573
Epoch: [250]  [130/156]  eta: 0:00:17  lr: 0.000594  min_lr: 0.000594  loss: 3.1344 (2.9759)  weight_decay: 0.0500 (0.0500)  time: 0.5943  data: 0.0010  max mem: 55573
Epoch: [250]  [140/156]  eta: 0:00:10  lr: 0.000593  min_lr: 0.000593  loss: 3.0377 (2.9668)  weight_decay: 0.0500 (0.0500)  time: 0.5844  data: 0.0008  max mem: 55573
Epoch: [250]  [150/156]  eta: 0:00:04  lr: 0.000591  min_lr: 0.000591  loss: 2.7942 (2.9538)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [250]  [155/156]  eta: 0:00:00  lr: 0.000591  min_lr: 0.000591  loss: 3.0777 (2.9530)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [250] Total time: 0:01:45 (0.6759 s / it)
Averaged stats: lr: 0.000591  min_lr: 0.000591  loss: 3.0777 (2.9624)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:32  loss: 0.7916 (0.7916)  acc1: 84.7656 (84.7656)  acc5: 97.3958 (97.3958)  time: 6.5721  data: 6.3350  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8921 (0.8898)  acc1: 80.9896 (80.5760)  acc5: 96.6146 (95.6800)  time: 1.4610  data: 1.2671  max mem: 55573
Test: Total time: 0:00:07 (1.5100 s / it)
* Acc@1 81.088 Acc@5 95.628 loss 0.885
Accuracy of the model on the 50000 test images: 81.1%
Max accuracy: 81.09%
Test:  [0/5]  eta: 0:00:35  loss: 1.0143 (1.0143)  acc1: 74.7396 (74.7396)  acc5: 92.8385 (92.8385)  time: 7.0628  data: 6.8267  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0433 (1.0611)  acc1: 73.1771 (70.7200)  acc5: 92.8385 (90.4640)  time: 1.5579  data: 1.3655  max mem: 55573
Test: Total time: 0:00:07 (1.5745 s / it)
* Acc@1 71.402 Acc@5 90.720 loss 1.081
Accuracy of the model EMA on 50000 test images: 71.4%
Max EMA accuracy: 71.40%
Epoch: [251]  [  0/156]  eta: 0:36:03  lr: 0.000590  min_lr: 0.000590  loss: 2.2402 (2.2402)  weight_decay: 0.0500 (0.0500)  time: 13.8672  data: 8.5797  max mem: 55573
Epoch: [251]  [ 10/156]  eta: 0:04:24  lr: 0.000589  min_lr: 0.000589  loss: 2.8069 (2.7955)  weight_decay: 0.0500 (0.0500)  time: 1.8110  data: 0.7804  max mem: 55573
Epoch: [251]  [ 20/156]  eta: 0:02:47  lr: 0.000587  min_lr: 0.000587  loss: 3.1535 (2.9577)  weight_decay: 0.0500 (0.0500)  time: 0.5984  data: 0.0004  max mem: 55573
Epoch: [251]  [ 30/156]  eta: 0:02:09  lr: 0.000586  min_lr: 0.000586  loss: 3.1535 (2.9770)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [251]  [ 40/156]  eta: 0:01:46  lr: 0.000584  min_lr: 0.000584  loss: 3.0687 (2.9922)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [251]  [ 50/156]  eta: 0:01:30  lr: 0.000583  min_lr: 0.000583  loss: 2.9991 (2.9743)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [251]  [ 60/156]  eta: 0:01:17  lr: 0.000581  min_lr: 0.000581  loss: 3.2008 (2.9932)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [251]  [ 70/156]  eta: 0:01:07  lr: 0.000580  min_lr: 0.000580  loss: 3.2239 (2.9755)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [251]  [ 80/156]  eta: 0:00:57  lr: 0.000578  min_lr: 0.000578  loss: 3.2272 (2.9888)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [251]  [ 90/156]  eta: 0:00:48  lr: 0.000577  min_lr: 0.000577  loss: 3.0009 (2.9643)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [251]  [100/156]  eta: 0:00:40  lr: 0.000575  min_lr: 0.000575  loss: 2.9556 (2.9702)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [251]  [110/156]  eta: 0:00:32  lr: 0.000574  min_lr: 0.000574  loss: 3.1446 (2.9798)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [251]  [120/156]  eta: 0:00:25  lr: 0.000572  min_lr: 0.000572  loss: 2.9726 (2.9667)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [251]  [130/156]  eta: 0:00:18  lr: 0.000571  min_lr: 0.000571  loss: 2.8723 (2.9538)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0010  max mem: 55573
Epoch: [251]  [140/156]  eta: 0:00:10  lr: 0.000569  min_lr: 0.000569  loss: 2.7077 (2.9352)  weight_decay: 0.0500 (0.0500)  time: 0.5859  data: 0.0009  max mem: 55573
Epoch: [251]  [150/156]  eta: 0:00:04  lr: 0.000568  min_lr: 0.000568  loss: 2.8233 (2.9453)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [251]  [155/156]  eta: 0:00:00  lr: 0.000567  min_lr: 0.000567  loss: 2.9071 (2.9431)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [251] Total time: 0:01:46 (0.6820 s / it)
Averaged stats: lr: 0.000567  min_lr: 0.000567  loss: 2.9071 (2.9658)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.8097 (0.8097)  acc1: 84.3750 (84.3750)  acc5: 96.6146 (96.6146)  time: 7.0996  data: 6.8622  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9013 (0.8999)  acc1: 81.6406 (80.1920)  acc5: 96.6146 (95.6480)  time: 1.5664  data: 1.3725  max mem: 55573
Test: Total time: 0:00:08 (1.6097 s / it)
* Acc@1 80.938 Acc@5 95.652 loss 0.891
Accuracy of the model on the 50000 test images: 80.9%
Max accuracy: 81.09%
Test:  [0/5]  eta: 0:00:35  loss: 1.0137 (1.0137)  acc1: 74.7396 (74.7396)  acc5: 92.8385 (92.8385)  time: 7.0821  data: 6.8460  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0427 (1.0605)  acc1: 73.1771 (70.7200)  acc5: 92.8385 (90.4960)  time: 1.5681  data: 1.3757  max mem: 55573
Test: Total time: 0:00:07 (1.5912 s / it)
* Acc@1 71.446 Acc@5 90.728 loss 1.081
Accuracy of the model EMA on 50000 test images: 71.4%
Max EMA accuracy: 71.45%
Epoch: [252]  [  0/156]  eta: 0:33:14  lr: 0.000567  min_lr: 0.000567  loss: 2.9682 (2.9682)  weight_decay: 0.0500 (0.0500)  time: 12.7873  data: 8.0248  max mem: 55573
Epoch: [252]  [ 10/156]  eta: 0:04:08  lr: 0.000566  min_lr: 0.000566  loss: 3.0019 (3.0869)  weight_decay: 0.0500 (0.0500)  time: 1.6992  data: 0.7299  max mem: 55573
Epoch: [252]  [ 20/156]  eta: 0:02:39  lr: 0.000564  min_lr: 0.000564  loss: 2.9468 (2.8639)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [252]  [ 30/156]  eta: 0:02:03  lr: 0.000563  min_lr: 0.000563  loss: 2.6308 (2.8387)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [252]  [ 40/156]  eta: 0:01:43  lr: 0.000561  min_lr: 0.000561  loss: 2.8804 (2.8960)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [252]  [ 50/156]  eta: 0:01:27  lr: 0.000560  min_lr: 0.000560  loss: 3.0213 (2.9132)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [252]  [ 60/156]  eta: 0:01:15  lr: 0.000558  min_lr: 0.000558  loss: 2.9626 (2.9070)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [252]  [ 70/156]  eta: 0:01:05  lr: 0.000557  min_lr: 0.000557  loss: 3.1091 (2.9156)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [252]  [ 80/156]  eta: 0:00:56  lr: 0.000555  min_lr: 0.000555  loss: 3.1207 (2.9101)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [252]  [ 90/156]  eta: 0:00:47  lr: 0.000554  min_lr: 0.000554  loss: 3.1694 (2.9471)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [252]  [100/156]  eta: 0:00:39  lr: 0.000552  min_lr: 0.000552  loss: 3.1284 (2.9413)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [252]  [110/156]  eta: 0:00:32  lr: 0.000551  min_lr: 0.000551  loss: 3.1274 (2.9527)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [252]  [120/156]  eta: 0:00:24  lr: 0.000550  min_lr: 0.000550  loss: 3.1646 (2.9492)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [252]  [130/156]  eta: 0:00:17  lr: 0.000548  min_lr: 0.000548  loss: 2.9703 (2.9407)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0010  max mem: 55573
Epoch: [252]  [140/156]  eta: 0:00:10  lr: 0.000547  min_lr: 0.000547  loss: 3.0675 (2.9483)  weight_decay: 0.0500 (0.0500)  time: 0.5873  data: 0.0008  max mem: 55573
Epoch: [252]  [150/156]  eta: 0:00:04  lr: 0.000545  min_lr: 0.000545  loss: 3.1353 (2.9470)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [252]  [155/156]  eta: 0:00:00  lr: 0.000544  min_lr: 0.000544  loss: 2.9998 (2.9456)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [252] Total time: 0:01:45 (0.6743 s / it)
Averaged stats: lr: 0.000544  min_lr: 0.000544  loss: 2.9998 (2.9441)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8492 (0.8492)  acc1: 84.2448 (84.2448)  acc5: 96.7448 (96.7448)  time: 6.8070  data: 6.5699  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9133 (0.9162)  acc1: 82.2917 (80.0640)  acc5: 96.7448 (95.7440)  time: 1.5077  data: 1.3141  max mem: 55573
Test: Total time: 0:00:07 (1.5599 s / it)
* Acc@1 81.040 Acc@5 95.684 loss 0.907
Accuracy of the model on the 50000 test images: 81.0%
Max accuracy: 81.09%
Test:  [0/5]  eta: 0:00:35  loss: 1.0131 (1.0131)  acc1: 74.8698 (74.8698)  acc5: 92.8385 (92.8385)  time: 7.1325  data: 6.8964  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0419 (1.0599)  acc1: 73.1771 (70.7520)  acc5: 92.8385 (90.4960)  time: 1.5718  data: 1.3794  max mem: 55573
Test: Total time: 0:00:08 (1.6027 s / it)
* Acc@1 71.464 Acc@5 90.734 loss 1.080
Accuracy of the model EMA on 50000 test images: 71.5%
Max EMA accuracy: 71.46%
Epoch: [253]  [  0/156]  eta: 0:33:25  lr: 0.000544  min_lr: 0.000544  loss: 3.4221 (3.4221)  weight_decay: 0.0500 (0.0500)  time: 12.8526  data: 12.2643  max mem: 55573
Epoch: [253]  [ 10/156]  eta: 0:04:11  lr: 0.000543  min_lr: 0.000543  loss: 3.1473 (2.9906)  weight_decay: 0.0500 (0.0500)  time: 1.7233  data: 1.1155  max mem: 55573
Epoch: [253]  [ 20/156]  eta: 0:02:40  lr: 0.000541  min_lr: 0.000541  loss: 3.0944 (3.0063)  weight_decay: 0.0500 (0.0500)  time: 0.6001  data: 0.0007  max mem: 55573
Epoch: [253]  [ 30/156]  eta: 0:02:05  lr: 0.000540  min_lr: 0.000540  loss: 3.0308 (2.9520)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0007  max mem: 55573
Epoch: [253]  [ 40/156]  eta: 0:01:43  lr: 0.000539  min_lr: 0.000539  loss: 3.1300 (3.0006)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0007  max mem: 55573
Epoch: [253]  [ 50/156]  eta: 0:01:28  lr: 0.000537  min_lr: 0.000537  loss: 3.2214 (3.0107)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0006  max mem: 55573
Epoch: [253]  [ 60/156]  eta: 0:01:16  lr: 0.000536  min_lr: 0.000536  loss: 3.1802 (3.0262)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0006  max mem: 55573
Epoch: [253]  [ 70/156]  eta: 0:01:05  lr: 0.000534  min_lr: 0.000534  loss: 3.1178 (2.9938)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0006  max mem: 55573
Epoch: [253]  [ 80/156]  eta: 0:00:56  lr: 0.000533  min_lr: 0.000533  loss: 2.8278 (2.9713)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0006  max mem: 55573
Epoch: [253]  [ 90/156]  eta: 0:00:48  lr: 0.000531  min_lr: 0.000531  loss: 2.8278 (2.9503)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [253]  [100/156]  eta: 0:00:40  lr: 0.000530  min_lr: 0.000530  loss: 2.9784 (2.9393)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [253]  [110/156]  eta: 0:00:32  lr: 0.000529  min_lr: 0.000529  loss: 3.1659 (2.9627)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0005  max mem: 55573
Epoch: [253]  [120/156]  eta: 0:00:24  lr: 0.000527  min_lr: 0.000527  loss: 3.1787 (2.9729)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0005  max mem: 55573
Epoch: [253]  [130/156]  eta: 0:00:17  lr: 0.000526  min_lr: 0.000526  loss: 3.1232 (2.9736)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0022  max mem: 55573
Epoch: [253]  [140/156]  eta: 0:00:10  lr: 0.000524  min_lr: 0.000524  loss: 3.1073 (2.9703)  weight_decay: 0.0500 (0.0500)  time: 0.5866  data: 0.0020  max mem: 55573
Epoch: [253]  [150/156]  eta: 0:00:04  lr: 0.000523  min_lr: 0.000523  loss: 3.0690 (2.9697)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0002  max mem: 55573
Epoch: [253]  [155/156]  eta: 0:00:00  lr: 0.000522  min_lr: 0.000522  loss: 3.0690 (2.9754)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0002  max mem: 55573
Epoch: [253] Total time: 0:01:45 (0.6762 s / it)
Averaged stats: lr: 0.000522  min_lr: 0.000522  loss: 3.0690 (2.9484)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8757 (0.8757)  acc1: 83.9844 (83.9844)  acc5: 97.1354 (97.1354)  time: 6.9042  data: 6.6668  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9822 (0.9628)  acc1: 81.2500 (79.7440)  acc5: 96.4844 (95.5520)  time: 1.5274  data: 1.3334  max mem: 55573
Test: Total time: 0:00:07 (1.5685 s / it)
* Acc@1 80.950 Acc@5 95.638 loss 0.952
Accuracy of the model on the 50000 test images: 81.0%
Max accuracy: 81.09%
Test:  [0/5]  eta: 0:00:33  loss: 1.0123 (1.0123)  acc1: 74.8698 (74.8698)  acc5: 92.8385 (92.8385)  time: 6.7825  data: 6.5463  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0412 (1.0594)  acc1: 73.1771 (70.7200)  acc5: 92.8385 (90.4960)  time: 1.5024  data: 1.3096  max mem: 55573
Test: Total time: 0:00:07 (1.5227 s / it)
* Acc@1 71.494 Acc@5 90.752 loss 1.079
Accuracy of the model EMA on 50000 test images: 71.5%
Max EMA accuracy: 71.49%
Epoch: [254]  [  0/156]  eta: 0:34:59  lr: 0.000522  min_lr: 0.000522  loss: 2.8639 (2.8639)  weight_decay: 0.0500 (0.0500)  time: 13.4556  data: 9.2297  max mem: 55573
Epoch: [254]  [ 10/156]  eta: 0:04:17  lr: 0.000521  min_lr: 0.000521  loss: 2.8639 (2.9365)  weight_decay: 0.0500 (0.0500)  time: 1.7646  data: 0.8394  max mem: 55573
Epoch: [254]  [ 20/156]  eta: 0:02:43  lr: 0.000519  min_lr: 0.000519  loss: 3.0973 (2.9797)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0004  max mem: 55573
Epoch: [254]  [ 30/156]  eta: 0:02:06  lr: 0.000518  min_lr: 0.000518  loss: 3.1396 (2.9560)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [254]  [ 40/156]  eta: 0:01:45  lr: 0.000516  min_lr: 0.000516  loss: 2.9000 (2.9217)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [254]  [ 50/156]  eta: 0:01:29  lr: 0.000515  min_lr: 0.000515  loss: 3.0022 (2.9422)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [254]  [ 60/156]  eta: 0:01:17  lr: 0.000513  min_lr: 0.000513  loss: 3.1607 (2.9574)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [254]  [ 70/156]  eta: 0:01:06  lr: 0.000512  min_lr: 0.000512  loss: 3.1571 (2.9691)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [254]  [ 80/156]  eta: 0:00:56  lr: 0.000511  min_lr: 0.000511  loss: 3.0491 (2.9606)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [254]  [ 90/156]  eta: 0:00:48  lr: 0.000509  min_lr: 0.000509  loss: 3.0129 (2.9561)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [254]  [100/156]  eta: 0:00:40  lr: 0.000508  min_lr: 0.000508  loss: 3.1075 (2.9717)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [254]  [110/156]  eta: 0:00:32  lr: 0.000506  min_lr: 0.000506  loss: 3.1075 (2.9697)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [254]  [120/156]  eta: 0:00:25  lr: 0.000505  min_lr: 0.000505  loss: 3.1044 (2.9692)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0004  max mem: 55573
Epoch: [254]  [130/156]  eta: 0:00:17  lr: 0.000504  min_lr: 0.000504  loss: 3.1654 (2.9672)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0011  max mem: 55573
Epoch: [254]  [140/156]  eta: 0:00:10  lr: 0.000502  min_lr: 0.000502  loss: 3.1147 (2.9587)  weight_decay: 0.0500 (0.0500)  time: 0.5880  data: 0.0009  max mem: 55573
Epoch: [254]  [150/156]  eta: 0:00:04  lr: 0.000501  min_lr: 0.000501  loss: 2.9490 (2.9527)  weight_decay: 0.0500 (0.0500)  time: 0.5829  data: 0.0001  max mem: 55573
Epoch: [254]  [155/156]  eta: 0:00:00  lr: 0.000500  min_lr: 0.000500  loss: 2.8677 (2.9457)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [254] Total time: 0:01:45 (0.6793 s / it)
Averaged stats: lr: 0.000500  min_lr: 0.000500  loss: 2.8677 (2.9402)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8205 (0.8205)  acc1: 83.4635 (83.4635)  acc5: 96.8750 (96.8750)  time: 6.8401  data: 6.6029  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9191 (0.9095)  acc1: 80.0781 (79.3920)  acc5: 96.4844 (95.2960)  time: 1.5143  data: 1.3207  max mem: 55573
Test: Total time: 0:00:07 (1.5626 s / it)
* Acc@1 81.148 Acc@5 95.688 loss 0.889
Accuracy of the model on the 50000 test images: 81.1%
Max accuracy: 81.15%
Test:  [0/5]  eta: 0:00:35  loss: 1.0116 (1.0116)  acc1: 74.8698 (74.8698)  acc5: 92.8385 (92.8385)  time: 7.0763  data: 6.8402  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0406 (1.0587)  acc1: 73.1771 (70.7200)  acc5: 92.8385 (90.4960)  time: 1.5606  data: 1.3681  max mem: 55573
Test: Total time: 0:00:07 (1.5762 s / it)
* Acc@1 71.520 Acc@5 90.762 loss 1.079
Accuracy of the model EMA on 50000 test images: 71.5%
Max EMA accuracy: 71.52%
Epoch: [255]  [  0/156]  eta: 0:34:35  lr: 0.000500  min_lr: 0.000500  loss: 2.9535 (2.9535)  weight_decay: 0.0500 (0.0500)  time: 13.3055  data: 12.1994  max mem: 55573
Epoch: [255]  [ 10/156]  eta: 0:04:14  lr: 0.000499  min_lr: 0.000499  loss: 2.9525 (2.8866)  weight_decay: 0.0500 (0.0500)  time: 1.7463  data: 1.1101  max mem: 55573
Epoch: [255]  [ 20/156]  eta: 0:02:42  lr: 0.000497  min_lr: 0.000497  loss: 2.9538 (2.8507)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0008  max mem: 55573
Epoch: [255]  [ 30/156]  eta: 0:02:06  lr: 0.000496  min_lr: 0.000496  loss: 3.1051 (2.8840)  weight_decay: 0.0500 (0.0500)  time: 0.5931  data: 0.0005  max mem: 55573
Epoch: [255]  [ 40/156]  eta: 0:01:44  lr: 0.000494  min_lr: 0.000494  loss: 3.1073 (2.9318)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [255]  [ 50/156]  eta: 0:01:29  lr: 0.000493  min_lr: 0.000493  loss: 3.1127 (2.9452)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [255]  [ 60/156]  eta: 0:01:16  lr: 0.000492  min_lr: 0.000492  loss: 3.0280 (2.9338)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [255]  [ 70/156]  eta: 0:01:06  lr: 0.000490  min_lr: 0.000490  loss: 2.9144 (2.9130)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [255]  [ 80/156]  eta: 0:00:56  lr: 0.000489  min_lr: 0.000489  loss: 2.9144 (2.9254)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [255]  [ 90/156]  eta: 0:00:48  lr: 0.000488  min_lr: 0.000488  loss: 3.0562 (2.9446)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0004  max mem: 55573
Epoch: [255]  [100/156]  eta: 0:00:40  lr: 0.000486  min_lr: 0.000486  loss: 3.0562 (2.9399)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0005  max mem: 55573
Epoch: [255]  [110/156]  eta: 0:00:32  lr: 0.000485  min_lr: 0.000485  loss: 2.7112 (2.9092)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [255]  [120/156]  eta: 0:00:25  lr: 0.000483  min_lr: 0.000483  loss: 2.7353 (2.9191)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [255]  [130/156]  eta: 0:00:17  lr: 0.000482  min_lr: 0.000482  loss: 3.0040 (2.9082)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0011  max mem: 55573
Epoch: [255]  [140/156]  eta: 0:00:10  lr: 0.000481  min_lr: 0.000481  loss: 3.0637 (2.9178)  weight_decay: 0.0500 (0.0500)  time: 0.5876  data: 0.0010  max mem: 55573
Epoch: [255]  [150/156]  eta: 0:00:04  lr: 0.000479  min_lr: 0.000479  loss: 2.9804 (2.9109)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [255]  [155/156]  eta: 0:00:00  lr: 0.000479  min_lr: 0.000479  loss: 2.9681 (2.9084)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [255] Total time: 0:01:45 (0.6777 s / it)
Averaged stats: lr: 0.000479  min_lr: 0.000479  loss: 2.9681 (2.9234)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.7980 (0.7980)  acc1: 85.5469 (85.5469)  acc5: 97.0052 (97.0052)  time: 7.0333  data: 6.7960  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8964 (0.8791)  acc1: 81.5104 (80.4160)  acc5: 96.8750 (95.7760)  time: 1.5527  data: 1.3593  max mem: 55573
Test: Total time: 0:00:08 (1.6005 s / it)
* Acc@1 81.160 Acc@5 95.680 loss 0.870
Accuracy of the model on the 50000 test images: 81.2%
Max accuracy: 81.16%
Test:  [0/5]  eta: 0:00:34  loss: 1.0106 (1.0106)  acc1: 75.0000 (75.0000)  acc5: 92.9688 (92.9688)  time: 6.9980  data: 6.7620  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0398 (1.0579)  acc1: 73.0469 (70.7200)  acc5: 92.9688 (90.5280)  time: 1.5449  data: 1.3525  max mem: 55573
Test: Total time: 0:00:08 (1.6040 s / it)
* Acc@1 71.526 Acc@5 90.774 loss 1.078
Accuracy of the model EMA on 50000 test images: 71.5%
Max EMA accuracy: 71.53%
Epoch: [256]  [  0/156]  eta: 0:33:49  lr: 0.000479  min_lr: 0.000479  loss: 3.3918 (3.3918)  weight_decay: 0.0500 (0.0500)  time: 13.0121  data: 10.7335  max mem: 55573
Epoch: [256]  [ 10/156]  eta: 0:04:11  lr: 0.000477  min_lr: 0.000477  loss: 3.2328 (3.1635)  weight_decay: 0.0500 (0.0500)  time: 1.7217  data: 0.9762  max mem: 55573
Epoch: [256]  [ 20/156]  eta: 0:02:40  lr: 0.000476  min_lr: 0.000476  loss: 3.1182 (3.0441)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [256]  [ 30/156]  eta: 0:02:05  lr: 0.000474  min_lr: 0.000474  loss: 2.8632 (2.9648)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0005  max mem: 55573
Epoch: [256]  [ 40/156]  eta: 0:01:43  lr: 0.000473  min_lr: 0.000473  loss: 2.8632 (2.9331)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0005  max mem: 55573
Epoch: [256]  [ 50/156]  eta: 0:01:28  lr: 0.000472  min_lr: 0.000472  loss: 3.1343 (2.9705)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [256]  [ 60/156]  eta: 0:01:16  lr: 0.000470  min_lr: 0.000470  loss: 3.1530 (2.9864)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [256]  [ 70/156]  eta: 0:01:05  lr: 0.000469  min_lr: 0.000469  loss: 3.1530 (2.9752)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0005  max mem: 55573
Epoch: [256]  [ 80/156]  eta: 0:00:56  lr: 0.000468  min_lr: 0.000468  loss: 2.8329 (2.9581)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [256]  [ 90/156]  eta: 0:00:48  lr: 0.000466  min_lr: 0.000466  loss: 2.6293 (2.9269)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [256]  [100/156]  eta: 0:00:39  lr: 0.000465  min_lr: 0.000465  loss: 2.6572 (2.9197)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [256]  [110/156]  eta: 0:00:32  lr: 0.000464  min_lr: 0.000464  loss: 2.8957 (2.9084)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [256]  [120/156]  eta: 0:00:24  lr: 0.000462  min_lr: 0.000462  loss: 2.9314 (2.9062)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0006  max mem: 55573
Epoch: [256]  [130/156]  eta: 0:00:17  lr: 0.000461  min_lr: 0.000461  loss: 3.0481 (2.9170)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0011  max mem: 55573
Epoch: [256]  [140/156]  eta: 0:00:10  lr: 0.000460  min_lr: 0.000460  loss: 3.1691 (2.9231)  weight_decay: 0.0500 (0.0500)  time: 0.5870  data: 0.0009  max mem: 55573
Epoch: [256]  [150/156]  eta: 0:00:04  lr: 0.000458  min_lr: 0.000458  loss: 3.1533 (2.9224)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [256]  [155/156]  eta: 0:00:00  lr: 0.000458  min_lr: 0.000458  loss: 3.2344 (2.9331)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [256] Total time: 0:01:45 (0.6760 s / it)
Averaged stats: lr: 0.000458  min_lr: 0.000458  loss: 3.2344 (2.9279)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.8586 (0.8586)  acc1: 82.6823 (82.6823)  acc5: 96.8750 (96.8750)  time: 6.7346  data: 6.4975  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9384 (0.9228)  acc1: 80.5990 (79.3920)  acc5: 96.7448 (95.8400)  time: 1.4934  data: 1.2996  max mem: 55573
Test: Total time: 0:00:07 (1.5451 s / it)
* Acc@1 81.056 Acc@5 95.700 loss 0.910
Accuracy of the model on the 50000 test images: 81.1%
Max accuracy: 81.16%
Test:  [0/5]  eta: 0:00:35  loss: 1.0097 (1.0097)  acc1: 75.0000 (75.0000)  acc5: 92.9688 (92.9688)  time: 7.1635  data: 6.9274  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0389 (1.0571)  acc1: 72.9167 (70.7200)  acc5: 92.9688 (90.5600)  time: 1.5779  data: 1.3856  max mem: 55573
Test: Total time: 0:00:08 (1.6020 s / it)
* Acc@1 71.578 Acc@5 90.808 loss 1.077
Accuracy of the model EMA on 50000 test images: 71.6%
Max EMA accuracy: 71.58%
Epoch: [257]  [  0/156]  eta: 0:37:03  lr: 0.000458  min_lr: 0.000458  loss: 3.5251 (3.5251)  weight_decay: 0.0500 (0.0500)  time: 14.2519  data: 13.6495  max mem: 55573
Epoch: [257]  [ 10/156]  eta: 0:04:27  lr: 0.000456  min_lr: 0.000456  loss: 3.0016 (2.8137)  weight_decay: 0.0500 (0.0500)  time: 1.8300  data: 1.2411  max mem: 55573
Epoch: [257]  [ 20/156]  eta: 0:02:48  lr: 0.000455  min_lr: 0.000455  loss: 3.0507 (2.8634)  weight_decay: 0.0500 (0.0500)  time: 0.5884  data: 0.0004  max mem: 55573
Epoch: [257]  [ 30/156]  eta: 0:02:09  lr: 0.000454  min_lr: 0.000454  loss: 3.0507 (2.8281)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0004  max mem: 55573
Epoch: [257]  [ 40/156]  eta: 0:01:46  lr: 0.000452  min_lr: 0.000452  loss: 2.8542 (2.8531)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0004  max mem: 55573
Epoch: [257]  [ 50/156]  eta: 0:01:30  lr: 0.000451  min_lr: 0.000451  loss: 3.0111 (2.9003)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [257]  [ 60/156]  eta: 0:01:18  lr: 0.000450  min_lr: 0.000450  loss: 3.1115 (2.9130)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0004  max mem: 55573
Epoch: [257]  [ 70/156]  eta: 0:01:07  lr: 0.000448  min_lr: 0.000448  loss: 3.0531 (2.9105)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [257]  [ 80/156]  eta: 0:00:57  lr: 0.000447  min_lr: 0.000447  loss: 2.9950 (2.9250)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [257]  [ 90/156]  eta: 0:00:48  lr: 0.000446  min_lr: 0.000446  loss: 2.9439 (2.9291)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [257]  [100/156]  eta: 0:00:40  lr: 0.000444  min_lr: 0.000444  loss: 2.8294 (2.9148)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0004  max mem: 55573
Epoch: [257]  [110/156]  eta: 0:00:32  lr: 0.000443  min_lr: 0.000443  loss: 2.8769 (2.9223)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [257]  [120/156]  eta: 0:00:25  lr: 0.000442  min_lr: 0.000442  loss: 2.8769 (2.9109)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [257]  [130/156]  eta: 0:00:18  lr: 0.000440  min_lr: 0.000440  loss: 2.9277 (2.9182)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0010  max mem: 55573
Epoch: [257]  [140/156]  eta: 0:00:10  lr: 0.000439  min_lr: 0.000439  loss: 3.0314 (2.9226)  weight_decay: 0.0500 (0.0500)  time: 0.5840  data: 0.0009  max mem: 55573
Epoch: [257]  [150/156]  eta: 0:00:04  lr: 0.000438  min_lr: 0.000438  loss: 2.7961 (2.9055)  weight_decay: 0.0500 (0.0500)  time: 0.5811  data: 0.0001  max mem: 55573
Epoch: [257]  [155/156]  eta: 0:00:00  lr: 0.000437  min_lr: 0.000437  loss: 2.7961 (2.9028)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [257] Total time: 0:01:46 (0.6826 s / it)
Averaged stats: lr: 0.000437  min_lr: 0.000437  loss: 2.7961 (2.9313)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.7976 (0.7976)  acc1: 83.7240 (83.7240)  acc5: 96.7448 (96.7448)  time: 6.8944  data: 6.6569  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8990 (0.8710)  acc1: 81.1198 (79.6160)  acc5: 96.7448 (95.6160)  time: 1.5254  data: 1.3315  max mem: 55573
Test: Total time: 0:00:07 (1.5675 s / it)
* Acc@1 81.212 Acc@5 95.762 loss 0.861
Accuracy of the model on the 50000 test images: 81.2%
Max accuracy: 81.21%
Test:  [0/5]  eta: 0:00:34  loss: 1.0088 (1.0088)  acc1: 75.0000 (75.0000)  acc5: 92.9688 (92.9688)  time: 6.9383  data: 6.7022  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0380 (1.0563)  acc1: 73.0469 (70.7520)  acc5: 92.9688 (90.5600)  time: 1.5330  data: 1.3405  max mem: 55573
Test: Total time: 0:00:07 (1.5502 s / it)
* Acc@1 71.598 Acc@5 90.814 loss 1.076
Accuracy of the model EMA on 50000 test images: 71.6%
Max EMA accuracy: 71.60%
Epoch: [258]  [  0/156]  eta: 0:35:04  lr: 0.000437  min_lr: 0.000437  loss: 3.1609 (3.1609)  weight_decay: 0.0500 (0.0500)  time: 13.4923  data: 12.8979  max mem: 55573
Epoch: [258]  [ 10/156]  eta: 0:04:18  lr: 0.000436  min_lr: 0.000436  loss: 3.0857 (3.0078)  weight_decay: 0.0500 (0.0500)  time: 1.7680  data: 1.1729  max mem: 55573
Epoch: [258]  [ 20/156]  eta: 0:02:44  lr: 0.000434  min_lr: 0.000434  loss: 3.0857 (3.0714)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0004  max mem: 55573
Epoch: [258]  [ 30/156]  eta: 0:02:07  lr: 0.000433  min_lr: 0.000433  loss: 3.0856 (3.0823)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [258]  [ 40/156]  eta: 0:01:45  lr: 0.000432  min_lr: 0.000432  loss: 3.1378 (3.1148)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [258]  [ 50/156]  eta: 0:01:29  lr: 0.000430  min_lr: 0.000430  loss: 3.1415 (3.1042)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [258]  [ 60/156]  eta: 0:01:17  lr: 0.000429  min_lr: 0.000429  loss: 3.2020 (3.1005)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [258]  [ 70/156]  eta: 0:01:06  lr: 0.000428  min_lr: 0.000428  loss: 3.0974 (3.0729)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [258]  [ 80/156]  eta: 0:00:57  lr: 0.000427  min_lr: 0.000427  loss: 2.8465 (3.0470)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [258]  [ 90/156]  eta: 0:00:48  lr: 0.000425  min_lr: 0.000425  loss: 2.8811 (3.0166)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [258]  [100/156]  eta: 0:00:40  lr: 0.000424  min_lr: 0.000424  loss: 2.8724 (2.9978)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0005  max mem: 55573
Epoch: [258]  [110/156]  eta: 0:00:32  lr: 0.000423  min_lr: 0.000423  loss: 2.9120 (2.9882)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [258]  [120/156]  eta: 0:00:25  lr: 0.000421  min_lr: 0.000421  loss: 2.9853 (2.9854)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [258]  [130/156]  eta: 0:00:17  lr: 0.000420  min_lr: 0.000420  loss: 2.9863 (2.9824)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0010  max mem: 55573
Epoch: [258]  [140/156]  eta: 0:00:10  lr: 0.000419  min_lr: 0.000419  loss: 3.0808 (2.9917)  weight_decay: 0.0500 (0.0500)  time: 0.5860  data: 0.0009  max mem: 55573
Epoch: [258]  [150/156]  eta: 0:00:04  lr: 0.000418  min_lr: 0.000418  loss: 3.0985 (2.9838)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [258]  [155/156]  eta: 0:00:00  lr: 0.000417  min_lr: 0.000417  loss: 3.0808 (2.9702)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [258] Total time: 0:01:46 (0.6798 s / it)
Averaged stats: lr: 0.000417  min_lr: 0.000417  loss: 3.0808 (2.9329)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.7947 (0.7947)  acc1: 83.9844 (83.9844)  acc5: 97.2656 (97.2656)  time: 6.7196  data: 6.4825  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8978 (0.8649)  acc1: 82.1615 (80.5120)  acc5: 96.3542 (95.6480)  time: 1.4904  data: 1.2966  max mem: 55573
Test: Total time: 0:00:07 (1.5346 s / it)
* Acc@1 81.234 Acc@5 95.776 loss 0.865
Accuracy of the model on the 50000 test images: 81.2%
Max accuracy: 81.23%
Test:  [0/5]  eta: 0:00:36  loss: 1.0078 (1.0078)  acc1: 75.0000 (75.0000)  acc5: 93.0990 (93.0990)  time: 7.2707  data: 7.0346  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0371 (1.0555)  acc1: 73.0469 (70.7200)  acc5: 93.0990 (90.5600)  time: 1.5995  data: 1.4070  max mem: 55573
Test: Total time: 0:00:08 (1.6168 s / it)
* Acc@1 71.630 Acc@5 90.834 loss 1.075
Accuracy of the model EMA on 50000 test images: 71.6%
Max EMA accuracy: 71.63%
Epoch: [259]  [  0/156]  eta: 0:31:16  lr: 0.000417  min_lr: 0.000417  loss: 2.7398 (2.7398)  weight_decay: 0.0500 (0.0500)  time: 12.0281  data: 10.3321  max mem: 55573
Epoch: [259]  [ 10/156]  eta: 0:04:00  lr: 0.000415  min_lr: 0.000415  loss: 2.8272 (2.8356)  weight_decay: 0.0500 (0.0500)  time: 1.6496  data: 0.9397  max mem: 55573
Epoch: [259]  [ 20/156]  eta: 0:02:35  lr: 0.000414  min_lr: 0.000414  loss: 2.9253 (2.9164)  weight_decay: 0.0500 (0.0500)  time: 0.6009  data: 0.0005  max mem: 55573
Epoch: [259]  [ 30/156]  eta: 0:02:01  lr: 0.000413  min_lr: 0.000413  loss: 2.9181 (2.8343)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [259]  [ 40/156]  eta: 0:01:41  lr: 0.000412  min_lr: 0.000412  loss: 2.8941 (2.8687)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0004  max mem: 55573
Epoch: [259]  [ 50/156]  eta: 0:01:26  lr: 0.000410  min_lr: 0.000410  loss: 3.1082 (2.9165)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0005  max mem: 55573
Epoch: [259]  [ 60/156]  eta: 0:01:15  lr: 0.000409  min_lr: 0.000409  loss: 3.1785 (2.9675)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [259]  [ 70/156]  eta: 0:01:04  lr: 0.000408  min_lr: 0.000408  loss: 3.1441 (2.9631)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [259]  [ 80/156]  eta: 0:00:55  lr: 0.000407  min_lr: 0.000407  loss: 3.0342 (2.9823)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [259]  [ 90/156]  eta: 0:00:47  lr: 0.000405  min_lr: 0.000405  loss: 3.0444 (2.9808)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [259]  [100/156]  eta: 0:00:39  lr: 0.000404  min_lr: 0.000404  loss: 2.9600 (2.9724)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [259]  [110/156]  eta: 0:00:31  lr: 0.000403  min_lr: 0.000403  loss: 3.0743 (2.9819)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0005  max mem: 55573
Epoch: [259]  [120/156]  eta: 0:00:24  lr: 0.000402  min_lr: 0.000402  loss: 3.1718 (2.9947)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0005  max mem: 55573
Epoch: [259]  [130/156]  eta: 0:00:17  lr: 0.000400  min_lr: 0.000400  loss: 3.0573 (2.9772)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0022  max mem: 55573
Epoch: [259]  [140/156]  eta: 0:00:10  lr: 0.000399  min_lr: 0.000399  loss: 2.7366 (2.9590)  weight_decay: 0.0500 (0.0500)  time: 0.5862  data: 0.0021  max mem: 55573
Epoch: [259]  [150/156]  eta: 0:00:03  lr: 0.000398  min_lr: 0.000398  loss: 2.7334 (2.9472)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [259]  [155/156]  eta: 0:00:00  lr: 0.000397  min_lr: 0.000397  loss: 2.6201 (2.9324)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [259] Total time: 0:01:44 (0.6706 s / it)
Averaged stats: lr: 0.000397  min_lr: 0.000397  loss: 2.6201 (2.9241)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.7550 (0.7550)  acc1: 84.3750 (84.3750)  acc5: 97.2656 (97.2656)  time: 6.9502  data: 6.7128  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8701 (0.8482)  acc1: 81.9010 (80.5760)  acc5: 96.3542 (95.6800)  time: 1.5365  data: 1.3426  max mem: 55573
Test: Total time: 0:00:07 (1.5875 s / it)
* Acc@1 81.400 Acc@5 95.780 loss 0.844
Accuracy of the model on the 50000 test images: 81.4%
Max accuracy: 81.40%
Test:  [0/5]  eta: 0:00:33  loss: 1.0068 (1.0068)  acc1: 75.0000 (75.0000)  acc5: 93.0990 (93.0990)  time: 6.7884  data: 6.5523  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0362 (1.0546)  acc1: 73.1771 (70.7840)  acc5: 93.0990 (90.5600)  time: 1.5031  data: 1.3106  max mem: 55573
Test: Total time: 0:00:07 (1.5173 s / it)
* Acc@1 71.670 Acc@5 90.862 loss 1.074
Accuracy of the model EMA on 50000 test images: 71.7%
Max EMA accuracy: 71.67%
Epoch: [260]  [  0/156]  eta: 0:33:31  lr: 0.000397  min_lr: 0.000397  loss: 2.7067 (2.7067)  weight_decay: 0.0500 (0.0500)  time: 12.8913  data: 8.4963  max mem: 55573
Epoch: [260]  [ 10/156]  eta: 0:04:09  lr: 0.000396  min_lr: 0.000396  loss: 3.1695 (3.0916)  weight_decay: 0.0500 (0.0500)  time: 1.7093  data: 0.7728  max mem: 55573
Epoch: [260]  [ 20/156]  eta: 0:02:39  lr: 0.000395  min_lr: 0.000395  loss: 3.1695 (3.0567)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [260]  [ 30/156]  eta: 0:02:04  lr: 0.000393  min_lr: 0.000393  loss: 3.0272 (2.9721)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [260]  [ 40/156]  eta: 0:01:43  lr: 0.000392  min_lr: 0.000392  loss: 3.0272 (2.9480)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [260]  [ 50/156]  eta: 0:01:28  lr: 0.000391  min_lr: 0.000391  loss: 2.9547 (2.9194)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [260]  [ 60/156]  eta: 0:01:16  lr: 0.000390  min_lr: 0.000390  loss: 2.9547 (2.9192)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [260]  [ 70/156]  eta: 0:01:05  lr: 0.000388  min_lr: 0.000388  loss: 2.9820 (2.9082)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [260]  [ 80/156]  eta: 0:00:56  lr: 0.000387  min_lr: 0.000387  loss: 2.9820 (2.9017)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [260]  [ 90/156]  eta: 0:00:47  lr: 0.000386  min_lr: 0.000386  loss: 3.0424 (2.9050)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [260]  [100/156]  eta: 0:00:39  lr: 0.000385  min_lr: 0.000385  loss: 3.0424 (2.9086)  weight_decay: 0.0500 (0.0500)  time: 0.5942  data: 0.0005  max mem: 55573
Epoch: [260]  [110/156]  eta: 0:00:32  lr: 0.000383  min_lr: 0.000383  loss: 3.0122 (2.9271)  weight_decay: 0.0500 (0.0500)  time: 0.5952  data: 0.0004  max mem: 55573
Epoch: [260]  [120/156]  eta: 0:00:24  lr: 0.000382  min_lr: 0.000382  loss: 3.0122 (2.9194)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [260]  [130/156]  eta: 0:00:17  lr: 0.000381  min_lr: 0.000381  loss: 2.7703 (2.9040)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0010  max mem: 55573
Epoch: [260]  [140/156]  eta: 0:00:10  lr: 0.000380  min_lr: 0.000380  loss: 2.8755 (2.8947)  weight_decay: 0.0500 (0.0500)  time: 0.5874  data: 0.0009  max mem: 55573
Epoch: [260]  [150/156]  eta: 0:00:04  lr: 0.000379  min_lr: 0.000379  loss: 2.9898 (2.8945)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [260]  [155/156]  eta: 0:00:00  lr: 0.000378  min_lr: 0.000378  loss: 2.9950 (2.9021)  weight_decay: 0.0500 (0.0500)  time: 0.5811  data: 0.0001  max mem: 55573
Epoch: [260] Total time: 0:01:45 (0.6753 s / it)
Averaged stats: lr: 0.000378  min_lr: 0.000378  loss: 2.9950 (2.9215)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8271 (0.8271)  acc1: 83.3333 (83.3333)  acc5: 97.1354 (97.1354)  time: 6.9527  data: 6.7152  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9396 (0.9142)  acc1: 81.1198 (80.1280)  acc5: 96.6146 (95.6160)  time: 1.5370  data: 1.3431  max mem: 55573
Test: Total time: 0:00:07 (1.5838 s / it)
* Acc@1 81.320 Acc@5 95.762 loss 0.904
Accuracy of the model on the 50000 test images: 81.3%
Max accuracy: 81.40%
Test:  [0/5]  eta: 0:00:35  loss: 1.0056 (1.0056)  acc1: 75.1302 (75.1302)  acc5: 93.0990 (93.0990)  time: 7.0897  data: 6.8537  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0351 (1.0535)  acc1: 73.4375 (70.9120)  acc5: 93.0990 (90.5920)  time: 1.5632  data: 1.3708  max mem: 55573
Test: Total time: 0:00:07 (1.5806 s / it)
* Acc@1 71.690 Acc@5 90.872 loss 1.073
Accuracy of the model EMA on 50000 test images: 71.7%
Max EMA accuracy: 71.69%
Epoch: [261]  [  0/156]  eta: 0:33:23  lr: 0.000378  min_lr: 0.000378  loss: 3.2199 (3.2199)  weight_decay: 0.0500 (0.0500)  time: 12.8413  data: 11.0221  max mem: 55573
Epoch: [261]  [ 10/156]  eta: 0:04:11  lr: 0.000377  min_lr: 0.000377  loss: 3.1778 (3.1154)  weight_decay: 0.0500 (0.0500)  time: 1.7212  data: 1.0024  max mem: 55573
Epoch: [261]  [ 20/156]  eta: 0:02:40  lr: 0.000375  min_lr: 0.000375  loss: 2.9262 (2.9330)  weight_decay: 0.0500 (0.0500)  time: 0.5998  data: 0.0004  max mem: 55573
Epoch: [261]  [ 30/156]  eta: 0:02:04  lr: 0.000374  min_lr: 0.000374  loss: 2.9090 (2.9269)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [261]  [ 40/156]  eta: 0:01:43  lr: 0.000373  min_lr: 0.000373  loss: 3.0872 (2.9015)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0004  max mem: 55573
Epoch: [261]  [ 50/156]  eta: 0:01:28  lr: 0.000372  min_lr: 0.000372  loss: 2.9542 (2.9061)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [261]  [ 60/156]  eta: 0:01:16  lr: 0.000371  min_lr: 0.000371  loss: 2.7454 (2.8846)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [261]  [ 70/156]  eta: 0:01:05  lr: 0.000369  min_lr: 0.000369  loss: 2.7431 (2.8717)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [261]  [ 80/156]  eta: 0:00:56  lr: 0.000368  min_lr: 0.000368  loss: 2.8727 (2.8687)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [261]  [ 90/156]  eta: 0:00:47  lr: 0.000367  min_lr: 0.000367  loss: 2.8727 (2.8630)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [261]  [100/156]  eta: 0:00:39  lr: 0.000366  min_lr: 0.000366  loss: 2.9614 (2.8657)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0004  max mem: 55573
Epoch: [261]  [110/156]  eta: 0:00:32  lr: 0.000365  min_lr: 0.000365  loss: 2.8262 (2.8486)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [261]  [120/156]  eta: 0:00:24  lr: 0.000363  min_lr: 0.000363  loss: 2.6465 (2.8376)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [261]  [130/156]  eta: 0:00:17  lr: 0.000362  min_lr: 0.000362  loss: 2.7699 (2.8312)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0010  max mem: 55573
Epoch: [261]  [140/156]  eta: 0:00:10  lr: 0.000361  min_lr: 0.000361  loss: 2.8326 (2.8363)  weight_decay: 0.0500 (0.0500)  time: 0.5872  data: 0.0009  max mem: 55573
Epoch: [261]  [150/156]  eta: 0:00:04  lr: 0.000360  min_lr: 0.000360  loss: 3.0701 (2.8487)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0002  max mem: 55573
Epoch: [261]  [155/156]  eta: 0:00:00  lr: 0.000359  min_lr: 0.000359  loss: 3.2062 (2.8631)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0002  max mem: 55573
Epoch: [261] Total time: 0:01:45 (0.6762 s / it)
Averaged stats: lr: 0.000359  min_lr: 0.000359  loss: 3.2062 (2.9025)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.9068 (0.9068)  acc1: 83.8542 (83.8542)  acc5: 97.2656 (97.2656)  time: 6.8009  data: 6.5637  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9978 (0.9878)  acc1: 81.1198 (79.9680)  acc5: 96.4844 (95.8080)  time: 1.5066  data: 1.3128  max mem: 55573
Test: Total time: 0:00:07 (1.5480 s / it)
* Acc@1 81.202 Acc@5 95.678 loss 0.972
Accuracy of the model on the 50000 test images: 81.2%
Max accuracy: 81.40%
Test:  [0/5]  eta: 0:00:36  loss: 1.0044 (1.0044)  acc1: 75.2604 (75.2604)  acc5: 93.0990 (93.0990)  time: 7.3110  data: 7.0751  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0341 (1.0524)  acc1: 73.3073 (70.9760)  acc5: 93.0990 (90.5920)  time: 1.6075  data: 1.4151  max mem: 55573
Test: Total time: 0:00:08 (1.6259 s / it)
* Acc@1 71.738 Acc@5 90.894 loss 1.072
Accuracy of the model EMA on 50000 test images: 71.7%
Max EMA accuracy: 71.74%
Epoch: [262]  [  0/156]  eta: 0:32:12  lr: 0.000359  min_lr: 0.000359  loss: 2.1975 (2.1975)  weight_decay: 0.0500 (0.0500)  time: 12.3863  data: 11.7987  max mem: 55573
Epoch: [262]  [ 10/156]  eta: 0:04:05  lr: 0.000358  min_lr: 0.000358  loss: 2.9014 (2.8406)  weight_decay: 0.0500 (0.0500)  time: 1.6807  data: 1.0730  max mem: 55573
Epoch: [262]  [ 20/156]  eta: 0:02:38  lr: 0.000357  min_lr: 0.000357  loss: 2.9014 (2.7601)  weight_decay: 0.0500 (0.0500)  time: 0.6012  data: 0.0004  max mem: 55573
Epoch: [262]  [ 30/156]  eta: 0:02:03  lr: 0.000355  min_lr: 0.000355  loss: 2.9793 (2.8640)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [262]  [ 40/156]  eta: 0:01:42  lr: 0.000354  min_lr: 0.000354  loss: 3.1412 (2.8176)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [262]  [ 50/156]  eta: 0:01:27  lr: 0.000353  min_lr: 0.000353  loss: 2.8914 (2.8544)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [262]  [ 60/156]  eta: 0:01:15  lr: 0.000352  min_lr: 0.000352  loss: 2.9434 (2.8632)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [262]  [ 70/156]  eta: 0:01:05  lr: 0.000351  min_lr: 0.000351  loss: 2.7523 (2.8384)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [262]  [ 80/156]  eta: 0:00:56  lr: 0.000350  min_lr: 0.000350  loss: 2.8386 (2.8401)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [262]  [ 90/156]  eta: 0:00:47  lr: 0.000348  min_lr: 0.000348  loss: 2.9707 (2.8454)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [262]  [100/156]  eta: 0:00:39  lr: 0.000347  min_lr: 0.000347  loss: 2.9872 (2.8512)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [262]  [110/156]  eta: 0:00:32  lr: 0.000346  min_lr: 0.000346  loss: 2.9872 (2.8652)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [262]  [120/156]  eta: 0:00:24  lr: 0.000345  min_lr: 0.000345  loss: 3.0972 (2.8735)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [262]  [130/156]  eta: 0:00:17  lr: 0.000344  min_lr: 0.000344  loss: 3.1143 (2.8736)  weight_decay: 0.0500 (0.0500)  time: 0.5891  data: 0.0010  max mem: 55573
Epoch: [262]  [140/156]  eta: 0:00:10  lr: 0.000343  min_lr: 0.000343  loss: 3.0647 (2.8810)  weight_decay: 0.0500 (0.0500)  time: 0.5851  data: 0.0009  max mem: 55573
Epoch: [262]  [150/156]  eta: 0:00:04  lr: 0.000341  min_lr: 0.000341  loss: 3.0451 (2.8887)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [262]  [155/156]  eta: 0:00:00  lr: 0.000341  min_lr: 0.000341  loss: 3.0713 (2.8903)  weight_decay: 0.0500 (0.0500)  time: 0.5812  data: 0.0001  max mem: 55573
Epoch: [262] Total time: 0:01:44 (0.6725 s / it)
Averaged stats: lr: 0.000341  min_lr: 0.000341  loss: 3.0713 (2.8898)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 0.8446 (0.8446)  acc1: 83.9844 (83.9844)  acc5: 97.0052 (97.0052)  time: 7.2412  data: 7.0037  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9370 (0.9247)  acc1: 81.2500 (80.0640)  acc5: 96.6146 (95.8400)  time: 1.5946  data: 1.4008  max mem: 55573
Test: Total time: 0:00:08 (1.6372 s / it)
* Acc@1 81.370 Acc@5 95.744 loss 0.909
Accuracy of the model on the 50000 test images: 81.4%
Max accuracy: 81.40%
Test:  [0/5]  eta: 0:00:34  loss: 1.0033 (1.0033)  acc1: 75.2604 (75.2604)  acc5: 93.0990 (93.0990)  time: 6.8980  data: 6.6619  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0330 (1.0513)  acc1: 73.0469 (70.9120)  acc5: 93.0990 (90.6240)  time: 1.5249  data: 1.3325  max mem: 55573
Test: Total time: 0:00:07 (1.5474 s / it)
* Acc@1 71.746 Acc@5 90.910 loss 1.071
Accuracy of the model EMA on 50000 test images: 71.7%
Max EMA accuracy: 71.75%
Epoch: [263]  [  0/156]  eta: 0:33:46  lr: 0.000341  min_lr: 0.000341  loss: 3.1876 (3.1876)  weight_decay: 0.0500 (0.0500)  time: 12.9881  data: 9.0971  max mem: 55573
Epoch: [263]  [ 10/156]  eta: 0:04:14  lr: 0.000340  min_lr: 0.000340  loss: 3.0054 (2.9130)  weight_decay: 0.0500 (0.0500)  time: 1.7419  data: 0.8274  max mem: 55573
Epoch: [263]  [ 20/156]  eta: 0:02:42  lr: 0.000338  min_lr: 0.000338  loss: 3.0456 (3.0095)  weight_decay: 0.0500 (0.0500)  time: 0.6041  data: 0.0004  max mem: 55573
Epoch: [263]  [ 30/156]  eta: 0:02:05  lr: 0.000337  min_lr: 0.000337  loss: 3.0223 (2.9204)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [263]  [ 40/156]  eta: 0:01:44  lr: 0.000336  min_lr: 0.000336  loss: 3.0071 (2.9674)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [263]  [ 50/156]  eta: 0:01:28  lr: 0.000335  min_lr: 0.000335  loss: 3.1295 (2.9658)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0004  max mem: 55573
Epoch: [263]  [ 60/156]  eta: 0:01:16  lr: 0.000334  min_lr: 0.000334  loss: 2.9793 (2.9590)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [263]  [ 70/156]  eta: 0:01:06  lr: 0.000333  min_lr: 0.000333  loss: 3.1265 (2.9743)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [263]  [ 80/156]  eta: 0:00:56  lr: 0.000331  min_lr: 0.000331  loss: 3.0273 (2.9453)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [263]  [ 90/156]  eta: 0:00:48  lr: 0.000330  min_lr: 0.000330  loss: 3.1652 (2.9685)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [263]  [100/156]  eta: 0:00:40  lr: 0.000329  min_lr: 0.000329  loss: 3.2533 (2.9676)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [263]  [110/156]  eta: 0:00:32  lr: 0.000328  min_lr: 0.000328  loss: 3.1120 (2.9849)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [263]  [120/156]  eta: 0:00:25  lr: 0.000327  min_lr: 0.000327  loss: 3.1710 (2.9950)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [263]  [130/156]  eta: 0:00:17  lr: 0.000326  min_lr: 0.000326  loss: 3.0840 (2.9880)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0010  max mem: 55573
Epoch: [263]  [140/156]  eta: 0:00:10  lr: 0.000325  min_lr: 0.000325  loss: 2.8981 (2.9735)  weight_decay: 0.0500 (0.0500)  time: 0.5857  data: 0.0008  max mem: 55573
Epoch: [263]  [150/156]  eta: 0:00:04  lr: 0.000324  min_lr: 0.000324  loss: 2.8909 (2.9727)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [263]  [155/156]  eta: 0:00:00  lr: 0.000323  min_lr: 0.000323  loss: 3.0125 (2.9750)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [263] Total time: 0:01:45 (0.6771 s / it)
Averaged stats: lr: 0.000323  min_lr: 0.000323  loss: 3.0125 (2.8937)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.8238 (0.8238)  acc1: 85.1562 (85.1562)  acc5: 97.3958 (97.3958)  time: 7.1451  data: 6.9077  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9314 (0.9102)  acc1: 81.3802 (80.4800)  acc5: 96.2240 (95.8080)  time: 1.5754  data: 1.3816  max mem: 55573
Test: Total time: 0:00:08 (1.6197 s / it)
* Acc@1 81.438 Acc@5 95.792 loss 0.892
Accuracy of the model on the 50000 test images: 81.4%
Max accuracy: 81.44%
Test:  [0/5]  eta: 0:00:33  loss: 1.0021 (1.0021)  acc1: 75.2604 (75.2604)  acc5: 93.0990 (93.0990)  time: 6.7269  data: 6.4907  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0318 (1.0502)  acc1: 73.1771 (70.9760)  acc5: 93.0990 (90.6880)  time: 1.4906  data: 1.2982  max mem: 55573
Test: Total time: 0:00:07 (1.5088 s / it)
* Acc@1 71.770 Acc@5 90.936 loss 1.070
Accuracy of the model EMA on 50000 test images: 71.8%
Max EMA accuracy: 71.77%
Epoch: [264]  [  0/156]  eta: 0:34:55  lr: 0.000323  min_lr: 0.000323  loss: 2.1688 (2.1688)  weight_decay: 0.0500 (0.0500)  time: 13.4342  data: 12.2308  max mem: 55573
Epoch: [264]  [ 10/156]  eta: 0:04:18  lr: 0.000322  min_lr: 0.000322  loss: 3.0480 (2.8455)  weight_decay: 0.0500 (0.0500)  time: 1.7705  data: 1.1125  max mem: 55573
Epoch: [264]  [ 20/156]  eta: 0:02:44  lr: 0.000321  min_lr: 0.000321  loss: 3.0487 (2.8925)  weight_decay: 0.0500 (0.0500)  time: 0.5970  data: 0.0006  max mem: 55573
Epoch: [264]  [ 30/156]  eta: 0:02:07  lr: 0.000319  min_lr: 0.000319  loss: 3.0600 (2.9023)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [264]  [ 40/156]  eta: 0:01:45  lr: 0.000318  min_lr: 0.000318  loss: 2.7512 (2.8595)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [264]  [ 50/156]  eta: 0:01:29  lr: 0.000317  min_lr: 0.000317  loss: 2.7512 (2.8438)  weight_decay: 0.0500 (0.0500)  time: 0.5978  data: 0.0004  max mem: 55573
Epoch: [264]  [ 60/156]  eta: 0:01:17  lr: 0.000316  min_lr: 0.000316  loss: 2.9402 (2.8437)  weight_decay: 0.0500 (0.0500)  time: 0.5971  data: 0.0004  max mem: 55573
Epoch: [264]  [ 70/156]  eta: 0:01:06  lr: 0.000315  min_lr: 0.000315  loss: 2.9885 (2.8372)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0004  max mem: 55573
Epoch: [264]  [ 80/156]  eta: 0:00:57  lr: 0.000314  min_lr: 0.000314  loss: 2.8966 (2.8333)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [264]  [ 90/156]  eta: 0:00:48  lr: 0.000313  min_lr: 0.000313  loss: 2.9865 (2.8574)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [264]  [100/156]  eta: 0:00:40  lr: 0.000312  min_lr: 0.000312  loss: 3.1369 (2.8870)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [264]  [110/156]  eta: 0:00:32  lr: 0.000311  min_lr: 0.000311  loss: 3.1578 (2.8992)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [264]  [120/156]  eta: 0:00:25  lr: 0.000309  min_lr: 0.000309  loss: 3.0884 (2.9003)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [264]  [130/156]  eta: 0:00:17  lr: 0.000308  min_lr: 0.000308  loss: 3.0144 (2.9003)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0011  max mem: 55573
Epoch: [264]  [140/156]  eta: 0:00:10  lr: 0.000307  min_lr: 0.000307  loss: 2.9853 (2.8949)  weight_decay: 0.0500 (0.0500)  time: 0.5862  data: 0.0009  max mem: 55573
Epoch: [264]  [150/156]  eta: 0:00:04  lr: 0.000306  min_lr: 0.000306  loss: 2.9853 (2.8901)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [264]  [155/156]  eta: 0:00:00  lr: 0.000306  min_lr: 0.000306  loss: 2.9865 (2.8790)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [264] Total time: 0:01:46 (0.6801 s / it)
Averaged stats: lr: 0.000306  min_lr: 0.000306  loss: 2.9865 (2.9031)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.7493 (0.7493)  acc1: 85.4167 (85.4167)  acc5: 97.5260 (97.5260)  time: 6.6617  data: 6.4246  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8910 (0.8612)  acc1: 80.9896 (80.7040)  acc5: 97.2656 (95.9680)  time: 1.4787  data: 1.2850  max mem: 55573
Test: Total time: 0:00:07 (1.5255 s / it)
* Acc@1 81.594 Acc@5 95.906 loss 0.846
Accuracy of the model on the 50000 test images: 81.6%
Max accuracy: 81.59%
Test:  [0/5]  eta: 0:00:36  loss: 1.0008 (1.0008)  acc1: 75.2604 (75.2604)  acc5: 93.0990 (93.0990)  time: 7.2770  data: 7.0409  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0306 (1.0492)  acc1: 73.1771 (70.9440)  acc5: 93.0990 (90.6880)  time: 1.6008  data: 1.4083  max mem: 55573
Test: Total time: 0:00:08 (1.6206 s / it)
* Acc@1 71.792 Acc@5 90.954 loss 1.069
Accuracy of the model EMA on 50000 test images: 71.8%
Max EMA accuracy: 71.79%
Epoch: [265]  [  0/156]  eta: 0:31:33  lr: 0.000305  min_lr: 0.000305  loss: 2.0269 (2.0269)  weight_decay: 0.0500 (0.0500)  time: 12.1368  data: 10.6787  max mem: 55573
Epoch: [265]  [ 10/156]  eta: 0:04:08  lr: 0.000304  min_lr: 0.000304  loss: 2.9988 (2.8009)  weight_decay: 0.0500 (0.0500)  time: 1.7023  data: 1.0080  max mem: 55573
Epoch: [265]  [ 20/156]  eta: 0:02:39  lr: 0.000303  min_lr: 0.000303  loss: 2.9988 (2.8437)  weight_decay: 0.0500 (0.0500)  time: 0.6239  data: 0.0207  max mem: 55573
Epoch: [265]  [ 30/156]  eta: 0:02:04  lr: 0.000302  min_lr: 0.000302  loss: 3.1459 (2.8635)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [265]  [ 40/156]  eta: 0:01:43  lr: 0.000301  min_lr: 0.000301  loss: 3.1344 (2.9092)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [265]  [ 50/156]  eta: 0:01:27  lr: 0.000300  min_lr: 0.000300  loss: 2.9664 (2.9137)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [265]  [ 60/156]  eta: 0:01:15  lr: 0.000299  min_lr: 0.000299  loss: 2.9664 (2.9160)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [265]  [ 70/156]  eta: 0:01:05  lr: 0.000298  min_lr: 0.000298  loss: 3.0165 (2.9094)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [265]  [ 80/156]  eta: 0:00:56  lr: 0.000297  min_lr: 0.000297  loss: 2.8801 (2.8881)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [265]  [ 90/156]  eta: 0:00:47  lr: 0.000296  min_lr: 0.000296  loss: 2.8796 (2.8977)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [265]  [100/156]  eta: 0:00:39  lr: 0.000295  min_lr: 0.000295  loss: 2.8727 (2.8801)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0005  max mem: 55573
Epoch: [265]  [110/156]  eta: 0:00:32  lr: 0.000293  min_lr: 0.000293  loss: 2.9514 (2.8811)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [265]  [120/156]  eta: 0:00:24  lr: 0.000292  min_lr: 0.000292  loss: 3.0595 (2.8849)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [265]  [130/156]  eta: 0:00:17  lr: 0.000291  min_lr: 0.000291  loss: 3.0083 (2.8724)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0010  max mem: 55573
Epoch: [265]  [140/156]  eta: 0:00:10  lr: 0.000290  min_lr: 0.000290  loss: 3.0046 (2.8720)  weight_decay: 0.0500 (0.0500)  time: 0.5853  data: 0.0009  max mem: 55573
Epoch: [265]  [150/156]  eta: 0:00:04  lr: 0.000289  min_lr: 0.000289  loss: 3.0046 (2.8720)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [265]  [155/156]  eta: 0:00:00  lr: 0.000289  min_lr: 0.000289  loss: 3.0046 (2.8728)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [265] Total time: 0:01:45 (0.6742 s / it)
Averaged stats: lr: 0.000289  min_lr: 0.000289  loss: 3.0046 (2.9060)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.7970 (0.7970)  acc1: 84.2448 (84.2448)  acc5: 97.5260 (97.5260)  time: 7.0308  data: 6.7933  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9168 (0.9011)  acc1: 80.4688 (80.1600)  acc5: 97.1354 (96.1920)  time: 1.5526  data: 1.3587  max mem: 55573
Test: Total time: 0:00:08 (1.6007 s / it)
* Acc@1 81.428 Acc@5 95.764 loss 0.888
Accuracy of the model on the 50000 test images: 81.4%
Max accuracy: 81.59%
Test:  [0/5]  eta: 0:00:34  loss: 0.9991 (0.9991)  acc1: 75.2604 (75.2604)  acc5: 93.2292 (93.2292)  time: 6.8546  data: 6.6185  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0291 (1.0477)  acc1: 73.1771 (71.0080)  acc5: 93.2292 (90.7200)  time: 1.5163  data: 1.3238  max mem: 55573
Test: Total time: 0:00:07 (1.5341 s / it)
* Acc@1 71.820 Acc@5 90.992 loss 1.067
Accuracy of the model EMA on 50000 test images: 71.8%
Max EMA accuracy: 71.82%
Epoch: [266]  [  0/156]  eta: 0:34:10  lr: 0.000289  min_lr: 0.000289  loss: 1.9500 (1.9500)  weight_decay: 0.0500 (0.0500)  time: 13.1437  data: 7.9771  max mem: 55573
Epoch: [266]  [ 10/156]  eta: 0:04:14  lr: 0.000287  min_lr: 0.000287  loss: 2.9130 (2.6579)  weight_decay: 0.0500 (0.0500)  time: 1.7448  data: 0.7376  max mem: 55573
Epoch: [266]  [ 20/156]  eta: 0:02:42  lr: 0.000286  min_lr: 0.000286  loss: 3.0565 (2.8202)  weight_decay: 0.0500 (0.0500)  time: 0.5978  data: 0.0070  max mem: 55573
Epoch: [266]  [ 30/156]  eta: 0:02:05  lr: 0.000285  min_lr: 0.000285  loss: 3.0629 (2.7700)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [266]  [ 40/156]  eta: 0:01:44  lr: 0.000284  min_lr: 0.000284  loss: 2.9586 (2.7967)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [266]  [ 50/156]  eta: 0:01:28  lr: 0.000283  min_lr: 0.000283  loss: 2.9686 (2.7991)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [266]  [ 60/156]  eta: 0:01:16  lr: 0.000282  min_lr: 0.000282  loss: 2.9586 (2.8085)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [266]  [ 70/156]  eta: 0:01:06  lr: 0.000281  min_lr: 0.000281  loss: 3.1175 (2.8643)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [266]  [ 80/156]  eta: 0:00:56  lr: 0.000280  min_lr: 0.000280  loss: 3.0374 (2.8778)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [266]  [ 90/156]  eta: 0:00:48  lr: 0.000279  min_lr: 0.000279  loss: 3.0374 (2.8947)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [266]  [100/156]  eta: 0:00:40  lr: 0.000278  min_lr: 0.000278  loss: 3.0375 (2.9006)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [266]  [110/156]  eta: 0:00:32  lr: 0.000277  min_lr: 0.000277  loss: 2.8785 (2.8771)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [266]  [120/156]  eta: 0:00:25  lr: 0.000276  min_lr: 0.000276  loss: 2.8723 (2.8759)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [266]  [130/156]  eta: 0:00:17  lr: 0.000275  min_lr: 0.000275  loss: 2.9773 (2.8775)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0011  max mem: 55573
Epoch: [266]  [140/156]  eta: 0:00:10  lr: 0.000274  min_lr: 0.000274  loss: 2.8768 (2.8637)  weight_decay: 0.0500 (0.0500)  time: 0.5887  data: 0.0009  max mem: 55573
Epoch: [266]  [150/156]  eta: 0:00:04  lr: 0.000273  min_lr: 0.000273  loss: 2.7045 (2.8467)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [266]  [155/156]  eta: 0:00:00  lr: 0.000272  min_lr: 0.000272  loss: 2.7601 (2.8534)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [266] Total time: 0:01:45 (0.6774 s / it)
Averaged stats: lr: 0.000272  min_lr: 0.000272  loss: 2.7601 (2.8738)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.7754 (0.7754)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 7.0181  data: 6.7806  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8946 (0.8679)  acc1: 81.7708 (80.9600)  acc5: 96.2240 (95.8080)  time: 1.5500  data: 1.3562  max mem: 55573
Test: Total time: 0:00:07 (1.5962 s / it)
* Acc@1 81.570 Acc@5 95.860 loss 0.850
Accuracy of the model on the 50000 test images: 81.6%
Max accuracy: 81.59%
Test:  [0/5]  eta: 0:00:35  loss: 0.9978 (0.9978)  acc1: 75.5208 (75.5208)  acc5: 93.2292 (93.2292)  time: 7.1616  data: 6.9254  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0280 (1.0465)  acc1: 73.3073 (71.1360)  acc5: 93.2292 (90.7200)  time: 1.5777  data: 1.3852  max mem: 55573
Test: Total time: 0:00:07 (1.5943 s / it)
* Acc@1 71.876 Acc@5 91.004 loss 1.066
Accuracy of the model EMA on 50000 test images: 71.9%
Max EMA accuracy: 71.88%
Epoch: [267]  [  0/156]  eta: 0:32:56  lr: 0.000272  min_lr: 0.000272  loss: 3.1005 (3.1005)  weight_decay: 0.0500 (0.0500)  time: 12.6680  data: 12.0716  max mem: 55573
Epoch: [267]  [ 10/156]  eta: 0:04:09  lr: 0.000271  min_lr: 0.000271  loss: 3.0704 (2.8159)  weight_decay: 0.0500 (0.0500)  time: 1.7089  data: 1.0978  max mem: 55573
Epoch: [267]  [ 20/156]  eta: 0:02:40  lr: 0.000270  min_lr: 0.000270  loss: 3.0704 (2.8178)  weight_decay: 0.0500 (0.0500)  time: 0.6028  data: 0.0005  max mem: 55573
Epoch: [267]  [ 30/156]  eta: 0:02:04  lr: 0.000269  min_lr: 0.000269  loss: 3.0643 (2.8176)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0006  max mem: 55573
Epoch: [267]  [ 40/156]  eta: 0:01:43  lr: 0.000268  min_lr: 0.000268  loss: 2.8956 (2.8432)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [267]  [ 50/156]  eta: 0:01:28  lr: 0.000267  min_lr: 0.000267  loss: 2.7596 (2.8139)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [267]  [ 60/156]  eta: 0:01:16  lr: 0.000266  min_lr: 0.000266  loss: 2.8994 (2.8390)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [267]  [ 70/156]  eta: 0:01:05  lr: 0.000265  min_lr: 0.000265  loss: 2.8301 (2.8214)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [267]  [ 80/156]  eta: 0:00:56  lr: 0.000264  min_lr: 0.000264  loss: 2.8301 (2.8196)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0005  max mem: 55573
Epoch: [267]  [ 90/156]  eta: 0:00:47  lr: 0.000263  min_lr: 0.000263  loss: 2.9773 (2.8299)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0005  max mem: 55573
Epoch: [267]  [100/156]  eta: 0:00:39  lr: 0.000262  min_lr: 0.000262  loss: 2.9307 (2.8252)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [267]  [110/156]  eta: 0:00:32  lr: 0.000261  min_lr: 0.000261  loss: 3.0279 (2.8372)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [267]  [120/156]  eta: 0:00:24  lr: 0.000260  min_lr: 0.000260  loss: 2.9123 (2.8265)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [267]  [130/156]  eta: 0:00:17  lr: 0.000259  min_lr: 0.000259  loss: 2.9123 (2.8322)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0013  max mem: 55573
Epoch: [267]  [140/156]  eta: 0:00:10  lr: 0.000258  min_lr: 0.000258  loss: 3.1719 (2.8425)  weight_decay: 0.0500 (0.0500)  time: 0.5862  data: 0.0012  max mem: 55573
Epoch: [267]  [150/156]  eta: 0:00:04  lr: 0.000257  min_lr: 0.000257  loss: 2.9450 (2.8435)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0002  max mem: 55573
Epoch: [267]  [155/156]  eta: 0:00:00  lr: 0.000256  min_lr: 0.000256  loss: 2.8096 (2.8360)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [267] Total time: 0:01:45 (0.6751 s / it)
Averaged stats: lr: 0.000256  min_lr: 0.000256  loss: 2.8096 (2.8918)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.7924 (0.7924)  acc1: 84.8958 (84.8958)  acc5: 97.5260 (97.5260)  time: 7.0536  data: 6.8162  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8892 (0.8671)  acc1: 82.8125 (80.9600)  acc5: 96.2264 (96.0000)  time: 1.5570  data: 1.3633  max mem: 55573
Test: Total time: 0:00:07 (1.5996 s / it)
* Acc@1 81.624 Acc@5 95.834 loss 0.854
Accuracy of the model on the 50000 test images: 81.6%
Max accuracy: 81.62%
Test:  [0/5]  eta: 0:00:34  loss: 0.9963 (0.9963)  acc1: 75.5208 (75.5208)  acc5: 93.3594 (93.3594)  time: 6.9512  data: 6.7153  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0268 (1.0453)  acc1: 73.3073 (71.1360)  acc5: 93.3594 (90.7520)  time: 1.5356  data: 1.3432  max mem: 55573
Test: Total time: 0:00:07 (1.5553 s / it)
* Acc@1 71.900 Acc@5 91.020 loss 1.065
Accuracy of the model EMA on 50000 test images: 71.9%
Max EMA accuracy: 71.90%
Epoch: [268]  [  0/156]  eta: 0:35:04  lr: 0.000256  min_lr: 0.000256  loss: 3.2994 (3.2994)  weight_decay: 0.0500 (0.0500)  time: 13.4906  data: 12.0033  max mem: 55573
Epoch: [268]  [ 10/156]  eta: 0:04:24  lr: 0.000255  min_lr: 0.000255  loss: 2.9986 (2.8224)  weight_decay: 0.0500 (0.0500)  time: 1.8090  data: 1.1404  max mem: 55573
Epoch: [268]  [ 20/156]  eta: 0:02:47  lr: 0.000254  min_lr: 0.000254  loss: 2.9964 (2.7410)  weight_decay: 0.0500 (0.0500)  time: 0.6153  data: 0.0275  max mem: 55573
Epoch: [268]  [ 30/156]  eta: 0:02:08  lr: 0.000253  min_lr: 0.000253  loss: 2.5462 (2.7029)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0007  max mem: 55573
Epoch: [268]  [ 40/156]  eta: 0:01:46  lr: 0.000252  min_lr: 0.000252  loss: 2.8679 (2.7747)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0006  max mem: 55573
Epoch: [268]  [ 50/156]  eta: 0:01:30  lr: 0.000251  min_lr: 0.000251  loss: 3.0400 (2.7980)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0006  max mem: 55573
Epoch: [268]  [ 60/156]  eta: 0:01:17  lr: 0.000250  min_lr: 0.000250  loss: 3.0254 (2.8008)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0008  max mem: 55573
Epoch: [268]  [ 70/156]  eta: 0:01:07  lr: 0.000249  min_lr: 0.000249  loss: 3.0189 (2.8246)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0009  max mem: 55573
Epoch: [268]  [ 80/156]  eta: 0:00:57  lr: 0.000248  min_lr: 0.000248  loss: 3.0146 (2.8351)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0009  max mem: 55573
Epoch: [268]  [ 90/156]  eta: 0:00:48  lr: 0.000247  min_lr: 0.000247  loss: 3.0146 (2.8387)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0009  max mem: 55573
Epoch: [268]  [100/156]  eta: 0:00:40  lr: 0.000246  min_lr: 0.000246  loss: 3.1794 (2.8740)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0008  max mem: 55573
Epoch: [268]  [110/156]  eta: 0:00:32  lr: 0.000245  min_lr: 0.000245  loss: 3.1049 (2.8670)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0009  max mem: 55573
Epoch: [268]  [120/156]  eta: 0:00:25  lr: 0.000244  min_lr: 0.000244  loss: 2.8424 (2.8576)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0008  max mem: 55573
Epoch: [268]  [130/156]  eta: 0:00:18  lr: 0.000243  min_lr: 0.000243  loss: 2.8424 (2.8509)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0017  max mem: 55573
Epoch: [268]  [140/156]  eta: 0:00:10  lr: 0.000242  min_lr: 0.000242  loss: 2.9300 (2.8573)  weight_decay: 0.0500 (0.0500)  time: 0.5886  data: 0.0013  max mem: 55573
Epoch: [268]  [150/156]  eta: 0:00:04  lr: 0.000241  min_lr: 0.000241  loss: 2.9830 (2.8730)  weight_decay: 0.0500 (0.0500)  time: 0.5834  data: 0.0001  max mem: 55573
Epoch: [268]  [155/156]  eta: 0:00:00  lr: 0.000241  min_lr: 0.000241  loss: 2.9983 (2.8816)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0001  max mem: 55573
Epoch: [268] Total time: 0:01:46 (0.6822 s / it)
Averaged stats: lr: 0.000241  min_lr: 0.000241  loss: 2.9983 (2.8931)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.8810 (0.8810)  acc1: 84.7656 (84.7656)  acc5: 97.6562 (97.6562)  time: 6.6305  data: 6.3934  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9746 (0.9619)  acc1: 81.3802 (80.8000)  acc5: 96.8750 (96.0000)  time: 1.4726  data: 1.2788  max mem: 55573
Test: Total time: 0:00:07 (1.5256 s / it)
* Acc@1 81.458 Acc@5 95.938 loss 0.945
Accuracy of the model on the 50000 test images: 81.5%
Max accuracy: 81.62%
Test:  [0/5]  eta: 0:00:34  loss: 0.9948 (0.9948)  acc1: 75.5208 (75.5208)  acc5: 93.3594 (93.3594)  time: 6.9925  data: 6.7565  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0252 (1.0439)  acc1: 73.4375 (71.2000)  acc5: 93.3594 (90.7840)  time: 1.5438  data: 1.3514  max mem: 55573
Test: Total time: 0:00:07 (1.5635 s / it)
* Acc@1 71.944 Acc@5 91.040 loss 1.063
Accuracy of the model EMA on 50000 test images: 71.9%
Max EMA accuracy: 71.94%
Epoch: [269]  [  0/156]  eta: 0:35:53  lr: 0.000240  min_lr: 0.000240  loss: 3.1547 (3.1547)  weight_decay: 0.0500 (0.0500)  time: 13.8031  data: 8.6513  max mem: 55573
Epoch: [269]  [ 10/156]  eta: 0:04:21  lr: 0.000240  min_lr: 0.000240  loss: 2.9185 (2.6801)  weight_decay: 0.0500 (0.0500)  time: 1.7915  data: 0.7868  max mem: 55573
Epoch: [269]  [ 20/156]  eta: 0:02:45  lr: 0.000239  min_lr: 0.000239  loss: 3.0589 (2.8956)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [269]  [ 30/156]  eta: 0:02:08  lr: 0.000238  min_lr: 0.000238  loss: 3.1496 (2.9056)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [269]  [ 40/156]  eta: 0:01:45  lr: 0.000237  min_lr: 0.000237  loss: 3.0737 (2.9178)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [269]  [ 50/156]  eta: 0:01:30  lr: 0.000236  min_lr: 0.000236  loss: 2.8693 (2.8944)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [269]  [ 60/156]  eta: 0:01:17  lr: 0.000235  min_lr: 0.000235  loss: 2.8160 (2.8732)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [269]  [ 70/156]  eta: 0:01:06  lr: 0.000234  min_lr: 0.000234  loss: 2.9254 (2.8828)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [269]  [ 80/156]  eta: 0:00:57  lr: 0.000233  min_lr: 0.000233  loss: 3.0773 (2.8880)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [269]  [ 90/156]  eta: 0:00:48  lr: 0.000232  min_lr: 0.000232  loss: 3.0324 (2.8974)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [269]  [100/156]  eta: 0:00:40  lr: 0.000231  min_lr: 0.000231  loss: 3.0407 (2.8925)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0004  max mem: 55573
Epoch: [269]  [110/156]  eta: 0:00:32  lr: 0.000230  min_lr: 0.000230  loss: 3.1261 (2.8965)  weight_decay: 0.0500 (0.0500)  time: 0.5933  data: 0.0004  max mem: 55573
Epoch: [269]  [120/156]  eta: 0:00:25  lr: 0.000229  min_lr: 0.000229  loss: 2.9741 (2.8878)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0004  max mem: 55573
Epoch: [269]  [130/156]  eta: 0:00:17  lr: 0.000228  min_lr: 0.000228  loss: 2.9741 (2.8669)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0010  max mem: 55573
Epoch: [269]  [140/156]  eta: 0:00:10  lr: 0.000227  min_lr: 0.000227  loss: 3.0449 (2.8755)  weight_decay: 0.0500 (0.0500)  time: 0.5874  data: 0.0009  max mem: 55573
Epoch: [269]  [150/156]  eta: 0:00:04  lr: 0.000226  min_lr: 0.000226  loss: 3.0082 (2.8824)  weight_decay: 0.0500 (0.0500)  time: 0.5829  data: 0.0002  max mem: 55573
Epoch: [269]  [155/156]  eta: 0:00:00  lr: 0.000226  min_lr: 0.000226  loss: 2.8860 (2.8853)  weight_decay: 0.0500 (0.0500)  time: 0.5828  data: 0.0002  max mem: 55573
Epoch: [269] Total time: 0:01:46 (0.6812 s / it)
Averaged stats: lr: 0.000226  min_lr: 0.000226  loss: 2.8860 (2.8851)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8288 (0.8288)  acc1: 85.0260 (85.0260)  acc5: 97.3958 (97.3958)  time: 6.9102  data: 6.6706  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9264 (0.9141)  acc1: 81.5104 (80.8000)  acc5: 96.8750 (95.8080)  time: 1.5288  data: 1.3342  max mem: 55573
Test: Total time: 0:00:07 (1.5827 s / it)
* Acc@1 81.536 Acc@5 95.834 loss 0.900
Accuracy of the model on the 50000 test images: 81.5%
Max accuracy: 81.62%
Test:  [0/5]  eta: 0:00:36  loss: 0.9933 (0.9933)  acc1: 75.5208 (75.5208)  acc5: 93.3594 (93.3594)  time: 7.3407  data: 7.1047  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0239 (1.0425)  acc1: 73.5677 (71.2960)  acc5: 93.3594 (90.8480)  time: 1.6134  data: 1.4210  max mem: 55573
Test: Total time: 0:00:08 (1.6356 s / it)
* Acc@1 71.988 Acc@5 91.068 loss 1.062
Accuracy of the model EMA on 50000 test images: 72.0%
Max EMA accuracy: 71.99%
Epoch: [270]  [  0/156]  eta: 0:33:30  lr: 0.000225  min_lr: 0.000225  loss: 3.3509 (3.3509)  weight_decay: 0.0500 (0.0500)  time: 12.8899  data: 11.8927  max mem: 55573
Epoch: [270]  [ 10/156]  eta: 0:04:16  lr: 0.000224  min_lr: 0.000224  loss: 2.8087 (2.7720)  weight_decay: 0.0500 (0.0500)  time: 1.7546  data: 1.1157  max mem: 55573
Epoch: [270]  [ 20/156]  eta: 0:02:43  lr: 0.000224  min_lr: 0.000224  loss: 2.8701 (2.9050)  weight_decay: 0.0500 (0.0500)  time: 0.6163  data: 0.0192  max mem: 55573
Epoch: [270]  [ 30/156]  eta: 0:02:06  lr: 0.000223  min_lr: 0.000223  loss: 2.9750 (2.8910)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [270]  [ 40/156]  eta: 0:01:44  lr: 0.000222  min_lr: 0.000222  loss: 2.8659 (2.8468)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0004  max mem: 55573
Epoch: [270]  [ 50/156]  eta: 0:01:29  lr: 0.000221  min_lr: 0.000221  loss: 2.8659 (2.8432)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0004  max mem: 55573
Epoch: [270]  [ 60/156]  eta: 0:01:16  lr: 0.000220  min_lr: 0.000220  loss: 3.1727 (2.8808)  weight_decay: 0.0500 (0.0500)  time: 0.5938  data: 0.0004  max mem: 55573
Epoch: [270]  [ 70/156]  eta: 0:01:06  lr: 0.000219  min_lr: 0.000219  loss: 3.1105 (2.8865)  weight_decay: 0.0500 (0.0500)  time: 0.5931  data: 0.0004  max mem: 55573
Epoch: [270]  [ 80/156]  eta: 0:00:56  lr: 0.000218  min_lr: 0.000218  loss: 3.1105 (2.9081)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [270]  [ 90/156]  eta: 0:00:48  lr: 0.000217  min_lr: 0.000217  loss: 3.0267 (2.8943)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [270]  [100/156]  eta: 0:00:40  lr: 0.000216  min_lr: 0.000216  loss: 2.9689 (2.9033)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [270]  [110/156]  eta: 0:00:32  lr: 0.000215  min_lr: 0.000215  loss: 2.9698 (2.9019)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [270]  [120/156]  eta: 0:00:25  lr: 0.000214  min_lr: 0.000214  loss: 3.0649 (2.9093)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [270]  [130/156]  eta: 0:00:17  lr: 0.000213  min_lr: 0.000213  loss: 2.9174 (2.8961)  weight_decay: 0.0500 (0.0500)  time: 0.5889  data: 0.0010  max mem: 55573
Epoch: [270]  [140/156]  eta: 0:00:10  lr: 0.000212  min_lr: 0.000212  loss: 3.0142 (2.9157)  weight_decay: 0.0500 (0.0500)  time: 0.5854  data: 0.0009  max mem: 55573
Epoch: [270]  [150/156]  eta: 0:00:04  lr: 0.000211  min_lr: 0.000211  loss: 2.9378 (2.8910)  weight_decay: 0.0500 (0.0500)  time: 0.5832  data: 0.0001  max mem: 55573
Epoch: [270]  [155/156]  eta: 0:00:00  lr: 0.000211  min_lr: 0.000211  loss: 2.7498 (2.8863)  weight_decay: 0.0500 (0.0500)  time: 0.5830  data: 0.0001  max mem: 55573
Epoch: [270] Total time: 0:01:45 (0.6789 s / it)
Averaged stats: lr: 0.000211  min_lr: 0.000211  loss: 2.7498 (2.8759)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.7849 (0.7849)  acc1: 84.1146 (84.1146)  acc5: 97.1354 (97.1354)  time: 6.8922  data: 6.6548  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8878 (0.8763)  acc1: 81.1198 (80.3840)  acc5: 96.2264 (96.0320)  time: 1.5250  data: 1.3310  max mem: 55573
Test: Total time: 0:00:07 (1.5696 s / it)
* Acc@1 81.686 Acc@5 95.888 loss 0.860
Accuracy of the model on the 50000 test images: 81.7%
Max accuracy: 81.69%
Test:  [0/5]  eta: 0:00:35  loss: 0.9916 (0.9916)  acc1: 75.6510 (75.6510)  acc5: 93.3594 (93.3594)  time: 7.0966  data: 6.8606  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0223 (1.0410)  acc1: 73.6979 (71.3600)  acc5: 93.3594 (90.8800)  time: 1.5646  data: 1.3722  max mem: 55573
Test: Total time: 0:00:07 (1.5816 s / it)
* Acc@1 72.024 Acc@5 91.088 loss 1.060
Accuracy of the model EMA on 50000 test images: 72.0%
Max EMA accuracy: 72.02%
Epoch: [271]  [  0/156]  eta: 0:32:12  lr: 0.000211  min_lr: 0.000211  loss: 2.9731 (2.9731)  weight_decay: 0.0500 (0.0500)  time: 12.3903  data: 11.7915  max mem: 55573
Epoch: [271]  [ 10/156]  eta: 0:04:09  lr: 0.000210  min_lr: 0.000210  loss: 3.0375 (2.8147)  weight_decay: 0.0500 (0.0500)  time: 1.7093  data: 1.1157  max mem: 55573
Epoch: [271]  [ 20/156]  eta: 0:02:40  lr: 0.000209  min_lr: 0.000209  loss: 2.9679 (2.7672)  weight_decay: 0.0500 (0.0500)  time: 0.6163  data: 0.0243  max mem: 55573
Epoch: [271]  [ 30/156]  eta: 0:02:04  lr: 0.000208  min_lr: 0.000208  loss: 3.0723 (2.8565)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [271]  [ 40/156]  eta: 0:01:43  lr: 0.000207  min_lr: 0.000207  loss: 2.9674 (2.8729)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [271]  [ 50/156]  eta: 0:01:28  lr: 0.000206  min_lr: 0.000206  loss: 3.0335 (2.9276)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0005  max mem: 55573
Epoch: [271]  [ 60/156]  eta: 0:01:16  lr: 0.000205  min_lr: 0.000205  loss: 3.0697 (2.9339)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [271]  [ 70/156]  eta: 0:01:05  lr: 0.000204  min_lr: 0.000204  loss: 2.9706 (2.9379)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [271]  [ 80/156]  eta: 0:00:56  lr: 0.000204  min_lr: 0.000204  loss: 2.9828 (2.9189)  weight_decay: 0.0500 (0.0500)  time: 0.5885  data: 0.0005  max mem: 55573
Epoch: [271]  [ 90/156]  eta: 0:00:47  lr: 0.000203  min_lr: 0.000203  loss: 2.8744 (2.9080)  weight_decay: 0.0500 (0.0500)  time: 0.5887  data: 0.0004  max mem: 55573
Epoch: [271]  [100/156]  eta: 0:00:39  lr: 0.000202  min_lr: 0.000202  loss: 2.9927 (2.9161)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [271]  [110/156]  eta: 0:00:32  lr: 0.000201  min_lr: 0.000201  loss: 2.9831 (2.8890)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [271]  [120/156]  eta: 0:00:24  lr: 0.000200  min_lr: 0.000200  loss: 3.0192 (2.9066)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [271]  [130/156]  eta: 0:00:17  lr: 0.000199  min_lr: 0.000199  loss: 3.1113 (2.9054)  weight_decay: 0.0500 (0.0500)  time: 0.5888  data: 0.0012  max mem: 55573
Epoch: [271]  [140/156]  eta: 0:00:10  lr: 0.000198  min_lr: 0.000198  loss: 2.8702 (2.8951)  weight_decay: 0.0500 (0.0500)  time: 0.5850  data: 0.0010  max mem: 55573
Epoch: [271]  [150/156]  eta: 0:00:04  lr: 0.000197  min_lr: 0.000197  loss: 3.0250 (2.9004)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [271]  [155/156]  eta: 0:00:00  lr: 0.000197  min_lr: 0.000197  loss: 3.0069 (2.9007)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [271] Total time: 0:01:45 (0.6743 s / it)
Averaged stats: lr: 0.000197  min_lr: 0.000197  loss: 3.0069 (2.8870)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8084 (0.8084)  acc1: 84.1146 (84.1146)  acc5: 97.5260 (97.5260)  time: 6.8773  data: 6.6394  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9106 (0.8865)  acc1: 81.5104 (80.7680)  acc5: 97.2656 (96.1600)  time: 1.5221  data: 1.3280  max mem: 55573
Test: Total time: 0:00:07 (1.5657 s / it)
* Acc@1 81.602 Acc@5 95.970 loss 0.872
Accuracy of the model on the 50000 test images: 81.6%
Max accuracy: 81.69%
Test:  [0/5]  eta: 0:00:35  loss: 0.9901 (0.9901)  acc1: 75.6510 (75.6510)  acc5: 93.3594 (93.3594)  time: 7.1664  data: 6.9304  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0208 (1.0396)  acc1: 73.6979 (71.3600)  acc5: 93.3594 (90.8800)  time: 1.5786  data: 1.3862  max mem: 55573
Test: Total time: 0:00:07 (1.5958 s / it)
* Acc@1 72.050 Acc@5 91.102 loss 1.059
Accuracy of the model EMA on 50000 test images: 72.1%
Max EMA accuracy: 72.05%
Epoch: [272]  [  0/156]  eta: 0:35:23  lr: 0.000197  min_lr: 0.000197  loss: 3.0785 (3.0785)  weight_decay: 0.0500 (0.0500)  time: 13.6136  data: 9.8098  max mem: 55573
Epoch: [272]  [ 10/156]  eta: 0:04:19  lr: 0.000196  min_lr: 0.000196  loss: 2.8848 (2.7486)  weight_decay: 0.0500 (0.0500)  time: 1.7802  data: 0.8922  max mem: 55573
Epoch: [272]  [ 20/156]  eta: 0:02:44  lr: 0.000195  min_lr: 0.000195  loss: 2.8848 (2.8555)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0004  max mem: 55573
Epoch: [272]  [ 30/156]  eta: 0:02:07  lr: 0.000194  min_lr: 0.000194  loss: 3.0463 (2.8513)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [272]  [ 40/156]  eta: 0:01:45  lr: 0.000193  min_lr: 0.000193  loss: 3.0463 (2.8526)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [272]  [ 50/156]  eta: 0:01:29  lr: 0.000192  min_lr: 0.000192  loss: 3.0224 (2.8676)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0005  max mem: 55573
Epoch: [272]  [ 60/156]  eta: 0:01:17  lr: 0.000191  min_lr: 0.000191  loss: 3.0030 (2.8710)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [272]  [ 70/156]  eta: 0:01:06  lr: 0.000191  min_lr: 0.000191  loss: 3.1715 (2.9004)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [272]  [ 80/156]  eta: 0:00:57  lr: 0.000190  min_lr: 0.000190  loss: 3.1406 (2.8986)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0005  max mem: 55573
Epoch: [272]  [ 90/156]  eta: 0:00:48  lr: 0.000189  min_lr: 0.000189  loss: 2.9815 (2.9084)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [272]  [100/156]  eta: 0:00:40  lr: 0.000188  min_lr: 0.000188  loss: 2.8916 (2.8962)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [272]  [110/156]  eta: 0:00:32  lr: 0.000187  min_lr: 0.000187  loss: 2.9755 (2.9030)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [272]  [120/156]  eta: 0:00:25  lr: 0.000186  min_lr: 0.000186  loss: 3.0085 (2.8951)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [272]  [130/156]  eta: 0:00:17  lr: 0.000185  min_lr: 0.000185  loss: 2.4400 (2.8545)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0011  max mem: 55573
Epoch: [272]  [140/156]  eta: 0:00:10  lr: 0.000184  min_lr: 0.000184  loss: 2.8369 (2.8741)  weight_decay: 0.0500 (0.0500)  time: 0.5876  data: 0.0009  max mem: 55573
Epoch: [272]  [150/156]  eta: 0:00:04  lr: 0.000184  min_lr: 0.000184  loss: 3.0467 (2.8728)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [272]  [155/156]  eta: 0:00:00  lr: 0.000183  min_lr: 0.000183  loss: 3.0467 (2.8740)  weight_decay: 0.0500 (0.0500)  time: 0.5813  data: 0.0001  max mem: 55573
Epoch: [272] Total time: 0:01:46 (0.6798 s / it)
Averaged stats: lr: 0.000183  min_lr: 0.000183  loss: 3.0467 (2.8696)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.7914 (0.7914)  acc1: 84.7656 (84.7656)  acc5: 97.1354 (97.1354)  time: 7.0456  data: 6.8068  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8801 (0.8709)  acc1: 81.6406 (80.6080)  acc5: 96.8750 (96.0640)  time: 1.5558  data: 1.3615  max mem: 55573
Test: Total time: 0:00:08 (1.6015 s / it)
* Acc@1 81.722 Acc@5 95.936 loss 0.858
Accuracy of the model on the 50000 test images: 81.7%
Max accuracy: 81.72%
Test:  [0/5]  eta: 0:00:34  loss: 0.9883 (0.9883)  acc1: 75.6510 (75.6510)  acc5: 93.3594 (93.3594)  time: 6.9464  data: 6.7103  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0190 (1.0380)  acc1: 73.6979 (71.3600)  acc5: 93.3594 (90.9440)  time: 1.5346  data: 1.3422  max mem: 55573
Test: Total time: 0:00:07 (1.5537 s / it)
* Acc@1 72.084 Acc@5 91.140 loss 1.057
Accuracy of the model EMA on 50000 test images: 72.1%
Max EMA accuracy: 72.08%
Epoch: [273]  [  0/156]  eta: 0:35:44  lr: 0.000183  min_lr: 0.000183  loss: 3.0868 (3.0868)  weight_decay: 0.0500 (0.0500)  time: 13.7493  data: 8.6278  max mem: 55573
Epoch: [273]  [ 10/156]  eta: 0:04:21  lr: 0.000182  min_lr: 0.000182  loss: 2.9065 (2.7109)  weight_decay: 0.0500 (0.0500)  time: 1.7894  data: 0.7849  max mem: 55573
Epoch: [273]  [ 20/156]  eta: 0:02:45  lr: 0.000181  min_lr: 0.000181  loss: 2.9065 (2.7549)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0008  max mem: 55573
Epoch: [273]  [ 30/156]  eta: 0:02:08  lr: 0.000181  min_lr: 0.000181  loss: 2.9043 (2.7482)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0008  max mem: 55573
Epoch: [273]  [ 40/156]  eta: 0:01:45  lr: 0.000180  min_lr: 0.000180  loss: 2.9043 (2.7710)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0005  max mem: 55573
Epoch: [273]  [ 50/156]  eta: 0:01:30  lr: 0.000179  min_lr: 0.000179  loss: 3.0610 (2.8060)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [273]  [ 60/156]  eta: 0:01:17  lr: 0.000178  min_lr: 0.000178  loss: 3.0965 (2.8346)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [273]  [ 70/156]  eta: 0:01:06  lr: 0.000177  min_lr: 0.000177  loss: 3.0055 (2.8116)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [273]  [ 80/156]  eta: 0:00:57  lr: 0.000176  min_lr: 0.000176  loss: 3.0055 (2.8196)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0005  max mem: 55573
Epoch: [273]  [ 90/156]  eta: 0:00:48  lr: 0.000175  min_lr: 0.000175  loss: 3.1073 (2.8258)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [273]  [100/156]  eta: 0:00:40  lr: 0.000175  min_lr: 0.000175  loss: 3.0083 (2.8117)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [273]  [110/156]  eta: 0:00:32  lr: 0.000174  min_lr: 0.000174  loss: 2.7577 (2.8077)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0006  max mem: 55573
Epoch: [273]  [120/156]  eta: 0:00:25  lr: 0.000173  min_lr: 0.000173  loss: 2.7572 (2.8070)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [273]  [130/156]  eta: 0:00:17  lr: 0.000172  min_lr: 0.000172  loss: 2.7707 (2.8186)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0022  max mem: 55573
Epoch: [273]  [140/156]  eta: 0:00:10  lr: 0.000171  min_lr: 0.000171  loss: 2.9528 (2.8186)  weight_decay: 0.0500 (0.0500)  time: 0.5856  data: 0.0020  max mem: 55573
Epoch: [273]  [150/156]  eta: 0:00:04  lr: 0.000170  min_lr: 0.000170  loss: 3.0684 (2.8393)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [273]  [155/156]  eta: 0:00:00  lr: 0.000170  min_lr: 0.000170  loss: 3.1031 (2.8410)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [273] Total time: 0:01:46 (0.6808 s / it)
Averaged stats: lr: 0.000170  min_lr: 0.000170  loss: 3.1031 (2.8716)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.8527 (0.8527)  acc1: 84.2448 (84.2448)  acc5: 97.3958 (97.3958)  time: 6.7890  data: 6.5518  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9505 (0.9288)  acc1: 80.9896 (80.3840)  acc5: 96.7448 (95.9680)  time: 1.5046  data: 1.3105  max mem: 55573
Test: Total time: 0:00:07 (1.5489 s / it)
* Acc@1 81.670 Acc@5 95.908 loss 0.916
Accuracy of the model on the 50000 test images: 81.7%
Max accuracy: 81.72%
Test:  [0/5]  eta: 0:00:35  loss: 0.9865 (0.9865)  acc1: 75.6510 (75.6510)  acc5: 93.4896 (93.4896)  time: 7.1342  data: 6.8982  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0174 (1.0363)  acc1: 73.6979 (71.3920)  acc5: 93.4896 (90.9760)  time: 1.5722  data: 1.3797  max mem: 55573
Test: Total time: 0:00:07 (1.5879 s / it)
* Acc@1 72.158 Acc@5 91.170 loss 1.055
Accuracy of the model EMA on 50000 test images: 72.2%
Max EMA accuracy: 72.16%
Epoch: [274]  [  0/156]  eta: 0:36:00  lr: 0.000170  min_lr: 0.000170  loss: 3.2319 (3.2319)  weight_decay: 0.0500 (0.0500)  time: 13.8499  data: 11.5596  max mem: 55573
Epoch: [274]  [ 10/156]  eta: 0:04:22  lr: 0.000169  min_lr: 0.000169  loss: 2.8645 (2.7537)  weight_decay: 0.0500 (0.0500)  time: 1.8004  data: 1.0512  max mem: 55573
Epoch: [274]  [ 20/156]  eta: 0:02:46  lr: 0.000168  min_lr: 0.000168  loss: 2.8645 (2.8375)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0004  max mem: 55573
Epoch: [274]  [ 30/156]  eta: 0:02:08  lr: 0.000168  min_lr: 0.000168  loss: 3.0200 (2.8232)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [274]  [ 40/156]  eta: 0:01:46  lr: 0.000167  min_lr: 0.000167  loss: 2.8389 (2.7958)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [274]  [ 50/156]  eta: 0:01:30  lr: 0.000166  min_lr: 0.000166  loss: 2.8389 (2.8168)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [274]  [ 60/156]  eta: 0:01:17  lr: 0.000165  min_lr: 0.000165  loss: 3.0415 (2.8577)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [274]  [ 70/156]  eta: 0:01:06  lr: 0.000164  min_lr: 0.000164  loss: 2.9775 (2.8330)  weight_decay: 0.0500 (0.0500)  time: 0.5892  data: 0.0005  max mem: 55573
Epoch: [274]  [ 80/156]  eta: 0:00:57  lr: 0.000163  min_lr: 0.000163  loss: 2.9775 (2.8612)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [274]  [ 90/156]  eta: 0:00:48  lr: 0.000163  min_lr: 0.000163  loss: 2.9621 (2.8533)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [274]  [100/156]  eta: 0:00:40  lr: 0.000162  min_lr: 0.000162  loss: 2.9614 (2.8601)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [274]  [110/156]  eta: 0:00:32  lr: 0.000161  min_lr: 0.000161  loss: 2.9966 (2.8536)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [274]  [120/156]  eta: 0:00:25  lr: 0.000160  min_lr: 0.000160  loss: 3.0455 (2.8590)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [274]  [130/156]  eta: 0:00:17  lr: 0.000159  min_lr: 0.000159  loss: 3.1109 (2.8694)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0011  max mem: 55573
Epoch: [274]  [140/156]  eta: 0:00:10  lr: 0.000159  min_lr: 0.000159  loss: 3.0928 (2.8641)  weight_decay: 0.0500 (0.0500)  time: 0.5862  data: 0.0009  max mem: 55573
Epoch: [274]  [150/156]  eta: 0:00:04  lr: 0.000158  min_lr: 0.000158  loss: 2.5622 (2.8460)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0002  max mem: 55573
Epoch: [274]  [155/156]  eta: 0:00:00  lr: 0.000157  min_lr: 0.000157  loss: 2.7618 (2.8504)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0002  max mem: 55573
Epoch: [274] Total time: 0:01:46 (0.6809 s / it)
Averaged stats: lr: 0.000157  min_lr: 0.000157  loss: 2.7618 (2.8822)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.8005 (0.8005)  acc1: 84.3750 (84.3750)  acc5: 97.1354 (97.1354)  time: 7.0358  data: 6.7983  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8975 (0.8808)  acc1: 81.1198 (80.8640)  acc5: 97.1354 (96.0960)  time: 1.5536  data: 1.3597  max mem: 55573
Test: Total time: 0:00:07 (1.5936 s / it)
* Acc@1 81.718 Acc@5 95.984 loss 0.866
Accuracy of the model on the 50000 test images: 81.7%
Max accuracy: 81.72%
Test:  [0/5]  eta: 0:00:35  loss: 0.9846 (0.9846)  acc1: 75.9115 (75.9115)  acc5: 93.4896 (93.4896)  time: 7.0228  data: 6.7864  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0154 (1.0345)  acc1: 73.6979 (71.4880)  acc5: 93.4896 (91.0720)  time: 1.5499  data: 1.3574  max mem: 55573
Test: Total time: 0:00:07 (1.5710 s / it)
* Acc@1 72.204 Acc@5 91.196 loss 1.053
Accuracy of the model EMA on 50000 test images: 72.2%
Max EMA accuracy: 72.20%
Epoch: [275]  [  0/156]  eta: 0:30:31  lr: 0.000157  min_lr: 0.000157  loss: 2.6137 (2.6137)  weight_decay: 0.0500 (0.0500)  time: 11.7372  data: 8.6486  max mem: 55573
Epoch: [275]  [ 10/156]  eta: 0:04:00  lr: 0.000157  min_lr: 0.000157  loss: 2.8953 (2.8371)  weight_decay: 0.0500 (0.0500)  time: 1.6477  data: 0.7867  max mem: 55573
Epoch: [275]  [ 20/156]  eta: 0:02:35  lr: 0.000156  min_lr: 0.000156  loss: 2.8622 (2.8195)  weight_decay: 0.0500 (0.0500)  time: 0.6146  data: 0.0005  max mem: 55573
Epoch: [275]  [ 30/156]  eta: 0:02:01  lr: 0.000155  min_lr: 0.000155  loss: 2.8622 (2.8349)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [275]  [ 40/156]  eta: 0:01:41  lr: 0.000154  min_lr: 0.000154  loss: 2.6771 (2.7700)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0004  max mem: 55573
Epoch: [275]  [ 50/156]  eta: 0:01:26  lr: 0.000153  min_lr: 0.000153  loss: 2.5807 (2.7739)  weight_decay: 0.0500 (0.0500)  time: 0.5934  data: 0.0005  max mem: 55573
Epoch: [275]  [ 60/156]  eta: 0:01:15  lr: 0.000153  min_lr: 0.000153  loss: 2.7003 (2.7621)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0005  max mem: 55573
Epoch: [275]  [ 70/156]  eta: 0:01:04  lr: 0.000152  min_lr: 0.000152  loss: 2.9134 (2.7801)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [275]  [ 80/156]  eta: 0:00:55  lr: 0.000151  min_lr: 0.000151  loss: 2.9021 (2.7945)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [275]  [ 90/156]  eta: 0:00:47  lr: 0.000150  min_lr: 0.000150  loss: 3.0215 (2.8020)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0005  max mem: 55573
Epoch: [275]  [100/156]  eta: 0:00:39  lr: 0.000149  min_lr: 0.000149  loss: 3.0215 (2.7932)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0005  max mem: 55573
Epoch: [275]  [110/156]  eta: 0:00:32  lr: 0.000149  min_lr: 0.000149  loss: 2.8279 (2.8076)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0005  max mem: 55573
Epoch: [275]  [120/156]  eta: 0:00:24  lr: 0.000148  min_lr: 0.000148  loss: 2.8756 (2.8042)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [275]  [130/156]  eta: 0:00:17  lr: 0.000147  min_lr: 0.000147  loss: 2.9965 (2.8180)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0011  max mem: 55573
Epoch: [275]  [140/156]  eta: 0:00:10  lr: 0.000146  min_lr: 0.000146  loss: 3.0541 (2.8253)  weight_decay: 0.0500 (0.0500)  time: 0.5854  data: 0.0009  max mem: 55573
Epoch: [275]  [150/156]  eta: 0:00:04  lr: 0.000146  min_lr: 0.000146  loss: 2.9819 (2.8322)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [275]  [155/156]  eta: 0:00:00  lr: 0.000145  min_lr: 0.000145  loss: 2.9575 (2.8296)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [275] Total time: 0:01:44 (0.6711 s / it)
Averaged stats: lr: 0.000145  min_lr: 0.000145  loss: 2.9575 (2.8688)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8135 (0.8135)  acc1: 83.9844 (83.9844)  acc5: 97.3958 (97.3958)  time: 6.8242  data: 6.5870  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9098 (0.9029)  acc1: 81.6406 (80.5440)  acc5: 96.8750 (96.0000)  time: 1.5102  data: 1.3175  max mem: 55573
Test: Total time: 0:00:07 (1.5282 s / it)
* Acc@1 81.776 Acc@5 95.918 loss 0.888
Accuracy of the model on the 50000 test images: 81.8%
Max accuracy: 81.78%
Test:  [0/5]  eta: 0:00:32  loss: 0.9827 (0.9827)  acc1: 76.1719 (76.1719)  acc5: 93.4896 (93.4896)  time: 6.4678  data: 6.2318  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0136 (1.0328)  acc1: 73.6979 (71.6160)  acc5: 93.4896 (91.0720)  time: 1.4390  data: 1.2465  max mem: 55573
Test: Total time: 0:00:07 (1.4562 s / it)
* Acc@1 72.262 Acc@5 91.212 loss 1.051
Accuracy of the model EMA on 50000 test images: 72.3%
Max EMA accuracy: 72.26%
Epoch: [276]  [  0/156]  eta: 0:28:44  lr: 0.000145  min_lr: 0.000145  loss: 3.0251 (3.0251)  weight_decay: 0.0500 (0.0500)  time: 11.0558  data: 8.0143  max mem: 55573
Epoch: [276]  [ 10/156]  eta: 0:03:56  lr: 0.000144  min_lr: 0.000144  loss: 3.1052 (2.9076)  weight_decay: 0.0500 (0.0500)  time: 1.6186  data: 0.7290  max mem: 55573
Epoch: [276]  [ 20/156]  eta: 0:02:33  lr: 0.000144  min_lr: 0.000144  loss: 3.1052 (2.9666)  weight_decay: 0.0500 (0.0500)  time: 0.6332  data: 0.0005  max mem: 55573
Epoch: [276]  [ 30/156]  eta: 0:02:00  lr: 0.000143  min_lr: 0.000143  loss: 3.1001 (2.9804)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [276]  [ 40/156]  eta: 0:01:40  lr: 0.000142  min_lr: 0.000142  loss: 3.0867 (2.9893)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [276]  [ 50/156]  eta: 0:01:26  lr: 0.000141  min_lr: 0.000141  loss: 3.0753 (2.9518)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [276]  [ 60/156]  eta: 0:01:14  lr: 0.000141  min_lr: 0.000141  loss: 3.0753 (2.9585)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [276]  [ 70/156]  eta: 0:01:04  lr: 0.000140  min_lr: 0.000140  loss: 2.8266 (2.9242)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [276]  [ 80/156]  eta: 0:00:55  lr: 0.000139  min_lr: 0.000139  loss: 2.7481 (2.8809)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [276]  [ 90/156]  eta: 0:00:47  lr: 0.000138  min_lr: 0.000138  loss: 2.9235 (2.8942)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [276]  [100/156]  eta: 0:00:39  lr: 0.000138  min_lr: 0.000138  loss: 2.9235 (2.8796)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0005  max mem: 55573
Epoch: [276]  [110/156]  eta: 0:00:31  lr: 0.000137  min_lr: 0.000137  loss: 2.8841 (2.8694)  weight_decay: 0.0500 (0.0500)  time: 0.5934  data: 0.0005  max mem: 55573
Epoch: [276]  [120/156]  eta: 0:00:24  lr: 0.000136  min_lr: 0.000136  loss: 2.9660 (2.8864)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [276]  [130/156]  eta: 0:00:17  lr: 0.000135  min_lr: 0.000135  loss: 3.1339 (2.8915)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0010  max mem: 55573
Epoch: [276]  [140/156]  eta: 0:00:10  lr: 0.000135  min_lr: 0.000135  loss: 3.0665 (2.8919)  weight_decay: 0.0500 (0.0500)  time: 0.5878  data: 0.0009  max mem: 55573
Epoch: [276]  [150/156]  eta: 0:00:03  lr: 0.000134  min_lr: 0.000134  loss: 2.9620 (2.8959)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [276]  [155/156]  eta: 0:00:00  lr: 0.000134  min_lr: 0.000134  loss: 3.0800 (2.9020)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [276] Total time: 0:01:44 (0.6687 s / it)
Averaged stats: lr: 0.000134  min_lr: 0.000134  loss: 3.0800 (2.8731)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.8573 (0.8573)  acc1: 84.5052 (84.5052)  acc5: 97.0052 (97.0052)  time: 6.7333  data: 6.4964  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9505 (0.9351)  acc1: 81.6406 (80.5760)  acc5: 96.8750 (95.9680)  time: 1.4931  data: 1.2994  max mem: 55573
Test: Total time: 0:00:07 (1.5406 s / it)
* Acc@1 81.736 Acc@5 95.908 loss 0.921
Accuracy of the model on the 50000 test images: 81.7%
Max accuracy: 81.78%
Test:  [0/5]  eta: 0:00:34  loss: 0.9807 (0.9807)  acc1: 76.0417 (76.0417)  acc5: 93.6198 (93.6198)  time: 6.9378  data: 6.7017  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0116 (1.0308)  acc1: 73.6979 (71.5840)  acc5: 93.6198 (91.1680)  time: 1.5329  data: 1.3404  max mem: 55573
Test: Total time: 0:00:08 (1.6062 s / it)
* Acc@1 72.310 Acc@5 91.248 loss 1.049
Accuracy of the model EMA on 50000 test images: 72.3%
Max EMA accuracy: 72.31%
Epoch: [277]  [  0/156]  eta: 0:36:41  lr: 0.000133  min_lr: 0.000133  loss: 3.3783 (3.3783)  weight_decay: 0.0500 (0.0500)  time: 14.1091  data: 13.5285  max mem: 55573
Epoch: [277]  [ 10/156]  eta: 0:04:25  lr: 0.000133  min_lr: 0.000133  loss: 2.9981 (2.9181)  weight_decay: 0.0500 (0.0500)  time: 1.8176  data: 1.2302  max mem: 55573
Epoch: [277]  [ 20/156]  eta: 0:02:47  lr: 0.000132  min_lr: 0.000132  loss: 2.9981 (2.9133)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [277]  [ 30/156]  eta: 0:02:09  lr: 0.000131  min_lr: 0.000131  loss: 3.0500 (2.9532)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [277]  [ 40/156]  eta: 0:01:46  lr: 0.000131  min_lr: 0.000131  loss: 3.1262 (2.9792)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [277]  [ 50/156]  eta: 0:01:31  lr: 0.000130  min_lr: 0.000130  loss: 3.1377 (2.9543)  weight_decay: 0.0500 (0.0500)  time: 0.6034  data: 0.0004  max mem: 55573
Epoch: [277]  [ 60/156]  eta: 0:01:18  lr: 0.000129  min_lr: 0.000129  loss: 3.0087 (2.9553)  weight_decay: 0.0500 (0.0500)  time: 0.6031  data: 0.0005  max mem: 55573
Epoch: [277]  [ 70/156]  eta: 0:01:07  lr: 0.000128  min_lr: 0.000128  loss: 3.1087 (2.9526)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [277]  [ 80/156]  eta: 0:00:57  lr: 0.000128  min_lr: 0.000128  loss: 3.1242 (2.9746)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [277]  [ 90/156]  eta: 0:00:48  lr: 0.000127  min_lr: 0.000127  loss: 2.9561 (2.9403)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [277]  [100/156]  eta: 0:00:40  lr: 0.000126  min_lr: 0.000126  loss: 2.8543 (2.9275)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [277]  [110/156]  eta: 0:00:32  lr: 0.000125  min_lr: 0.000125  loss: 2.9188 (2.9131)  weight_decay: 0.0500 (0.0500)  time: 0.5943  data: 0.0005  max mem: 55573
Epoch: [277]  [120/156]  eta: 0:00:25  lr: 0.000125  min_lr: 0.000125  loss: 3.0062 (2.9164)  weight_decay: 0.0500 (0.0500)  time: 0.5934  data: 0.0004  max mem: 55573
Epoch: [277]  [130/156]  eta: 0:00:18  lr: 0.000124  min_lr: 0.000124  loss: 2.9783 (2.9227)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0011  max mem: 55573
Epoch: [277]  [140/156]  eta: 0:00:11  lr: 0.000123  min_lr: 0.000123  loss: 2.9664 (2.9200)  weight_decay: 0.0500 (0.0500)  time: 0.5868  data: 0.0009  max mem: 55573
Epoch: [277]  [150/156]  eta: 0:00:04  lr: 0.000123  min_lr: 0.000123  loss: 2.8675 (2.9050)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [277]  [155/156]  eta: 0:00:00  lr: 0.000122  min_lr: 0.000122  loss: 2.8675 (2.9089)  weight_decay: 0.0500 (0.0500)  time: 0.5812  data: 0.0001  max mem: 55573
Epoch: [277] Total time: 0:01:46 (0.6845 s / it)
Averaged stats: lr: 0.000122  min_lr: 0.000122  loss: 2.8675 (2.8693)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8202 (0.8202)  acc1: 84.2448 (84.2448)  acc5: 97.3958 (97.3958)  time: 6.9426  data: 6.7053  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9114 (0.8918)  acc1: 81.3802 (80.7360)  acc5: 97.1354 (96.1280)  time: 1.5349  data: 1.3411  max mem: 55573
Test: Total time: 0:00:07 (1.5796 s / it)
* Acc@1 81.774 Acc@5 95.936 loss 0.877
Accuracy of the model on the 50000 test images: 81.8%
Max accuracy: 81.78%
Test:  [0/5]  eta: 0:00:36  loss: 0.9786 (0.9786)  acc1: 76.0417 (76.0417)  acc5: 93.6198 (93.6198)  time: 7.3245  data: 7.0885  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0095 (1.0288)  acc1: 73.8281 (71.5840)  acc5: 93.6198 (91.2000)  time: 1.6102  data: 1.4178  max mem: 55573
Test: Total time: 0:00:08 (1.6292 s / it)
* Acc@1 72.360 Acc@5 91.268 loss 1.047
Accuracy of the model EMA on 50000 test images: 72.4%
Max EMA accuracy: 72.36%
Epoch: [278]  [  0/156]  eta: 0:32:32  lr: 0.000122  min_lr: 0.000122  loss: 3.3732 (3.3732)  weight_decay: 0.0500 (0.0500)  time: 12.5141  data: 10.4160  max mem: 55573
Epoch: [278]  [ 10/156]  eta: 0:04:10  lr: 0.000122  min_lr: 0.000122  loss: 2.9924 (2.8719)  weight_decay: 0.0500 (0.0500)  time: 1.7163  data: 0.9565  max mem: 55573
Epoch: [278]  [ 20/156]  eta: 0:02:40  lr: 0.000121  min_lr: 0.000121  loss: 2.9924 (2.9340)  weight_decay: 0.0500 (0.0500)  time: 0.6133  data: 0.0055  max mem: 55573
Epoch: [278]  [ 30/156]  eta: 0:02:04  lr: 0.000120  min_lr: 0.000120  loss: 3.0376 (2.9080)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [278]  [ 40/156]  eta: 0:01:43  lr: 0.000119  min_lr: 0.000119  loss: 3.0005 (2.9066)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [278]  [ 50/156]  eta: 0:01:28  lr: 0.000119  min_lr: 0.000119  loss: 3.0577 (2.9084)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0004  max mem: 55573
Epoch: [278]  [ 60/156]  eta: 0:01:16  lr: 0.000118  min_lr: 0.000118  loss: 3.0817 (2.9299)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [278]  [ 70/156]  eta: 0:01:05  lr: 0.000117  min_lr: 0.000117  loss: 3.0773 (2.9269)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [278]  [ 80/156]  eta: 0:00:56  lr: 0.000117  min_lr: 0.000117  loss: 2.9015 (2.9232)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0005  max mem: 55573
Epoch: [278]  [ 90/156]  eta: 0:00:47  lr: 0.000116  min_lr: 0.000116  loss: 2.9713 (2.9236)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [278]  [100/156]  eta: 0:00:39  lr: 0.000115  min_lr: 0.000115  loss: 2.9572 (2.9118)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [278]  [110/156]  eta: 0:00:32  lr: 0.000115  min_lr: 0.000115  loss: 2.8622 (2.9095)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [278]  [120/156]  eta: 0:00:24  lr: 0.000114  min_lr: 0.000114  loss: 2.8601 (2.8911)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0004  max mem: 55573
Epoch: [278]  [130/156]  eta: 0:00:17  lr: 0.000113  min_lr: 0.000113  loss: 2.6864 (2.8749)  weight_decay: 0.0500 (0.0500)  time: 0.5922  data: 0.0011  max mem: 55573
Epoch: [278]  [140/156]  eta: 0:00:10  lr: 0.000113  min_lr: 0.000113  loss: 2.8828 (2.8786)  weight_decay: 0.0500 (0.0500)  time: 0.5866  data: 0.0009  max mem: 55573
Epoch: [278]  [150/156]  eta: 0:00:04  lr: 0.000112  min_lr: 0.000112  loss: 2.8997 (2.8695)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [278]  [155/156]  eta: 0:00:00  lr: 0.000112  min_lr: 0.000112  loss: 2.9700 (2.8790)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [278] Total time: 0:01:45 (0.6761 s / it)
Averaged stats: lr: 0.000112  min_lr: 0.000112  loss: 2.9700 (2.8761)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8284 (0.8284)  acc1: 84.7656 (84.7656)  acc5: 97.2656 (97.2656)  time: 6.8877  data: 6.6502  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9250 (0.9071)  acc1: 80.8594 (80.8320)  acc5: 97.2656 (96.2560)  time: 1.5241  data: 1.3301  max mem: 55573
Test: Total time: 0:00:07 (1.5644 s / it)
* Acc@1 81.774 Acc@5 95.948 loss 0.896
Accuracy of the model on the 50000 test images: 81.8%
Max accuracy: 81.78%
Test:  [0/5]  eta: 0:00:36  loss: 0.9767 (0.9767)  acc1: 76.1719 (76.1719)  acc5: 93.6198 (93.6198)  time: 7.2233  data: 6.9874  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0075 (1.0268)  acc1: 73.9583 (71.6800)  acc5: 93.6198 (91.2000)  time: 1.5900  data: 1.3976  max mem: 55573
Test: Total time: 0:00:08 (1.6082 s / it)
* Acc@1 72.418 Acc@5 91.306 loss 1.045
Accuracy of the model EMA on 50000 test images: 72.4%
Max EMA accuracy: 72.42%
Epoch: [279]  [  0/156]  eta: 0:32:15  lr: 0.000112  min_lr: 0.000112  loss: 3.1425 (3.1425)  weight_decay: 0.0500 (0.0500)  time: 12.4087  data: 8.8257  max mem: 55573
Epoch: [279]  [ 10/156]  eta: 0:04:04  lr: 0.000111  min_lr: 0.000111  loss: 2.7046 (2.6851)  weight_decay: 0.0500 (0.0500)  time: 1.6730  data: 0.8028  max mem: 55573
Epoch: [279]  [ 20/156]  eta: 0:02:37  lr: 0.000110  min_lr: 0.000110  loss: 2.7398 (2.7332)  weight_decay: 0.0500 (0.0500)  time: 0.5951  data: 0.0004  max mem: 55573
Epoch: [279]  [ 30/156]  eta: 0:02:02  lr: 0.000110  min_lr: 0.000110  loss: 2.9136 (2.8033)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [279]  [ 40/156]  eta: 0:01:42  lr: 0.000109  min_lr: 0.000109  loss: 2.9491 (2.8089)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [279]  [ 50/156]  eta: 0:01:27  lr: 0.000108  min_lr: 0.000108  loss: 2.8078 (2.8036)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [279]  [ 60/156]  eta: 0:01:15  lr: 0.000108  min_lr: 0.000108  loss: 2.8078 (2.7795)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [279]  [ 70/156]  eta: 0:01:05  lr: 0.000107  min_lr: 0.000107  loss: 2.9958 (2.7999)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [279]  [ 80/156]  eta: 0:00:56  lr: 0.000106  min_lr: 0.000106  loss: 3.1198 (2.8321)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [279]  [ 90/156]  eta: 0:00:47  lr: 0.000106  min_lr: 0.000106  loss: 3.0668 (2.8269)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [279]  [100/156]  eta: 0:00:39  lr: 0.000105  min_lr: 0.000105  loss: 2.6445 (2.8105)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [279]  [110/156]  eta: 0:00:32  lr: 0.000104  min_lr: 0.000104  loss: 2.8839 (2.8216)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [279]  [120/156]  eta: 0:00:24  lr: 0.000104  min_lr: 0.000104  loss: 2.9987 (2.8163)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [279]  [130/156]  eta: 0:00:17  lr: 0.000103  min_lr: 0.000103  loss: 2.9983 (2.8309)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0011  max mem: 55573
Epoch: [279]  [140/156]  eta: 0:00:10  lr: 0.000102  min_lr: 0.000102  loss: 2.8902 (2.8361)  weight_decay: 0.0500 (0.0500)  time: 0.5880  data: 0.0010  max mem: 55573
Epoch: [279]  [150/156]  eta: 0:00:04  lr: 0.000102  min_lr: 0.000102  loss: 2.9304 (2.8404)  weight_decay: 0.0500 (0.0500)  time: 0.5837  data: 0.0001  max mem: 55573
Epoch: [279]  [155/156]  eta: 0:00:00  lr: 0.000101  min_lr: 0.000101  loss: 2.8936 (2.8429)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [279] Total time: 0:01:44 (0.6726 s / it)
Averaged stats: lr: 0.000101  min_lr: 0.000101  loss: 2.8936 (2.8742)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.8033 (0.8033)  acc1: 84.7656 (84.7656)  acc5: 97.1354 (97.1354)  time: 6.7411  data: 6.5042  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8964 (0.8874)  acc1: 82.0312 (80.9920)  acc5: 97.0052 (96.0640)  time: 1.4935  data: 1.3009  max mem: 55573
Test: Total time: 0:00:07 (1.5106 s / it)
* Acc@1 81.848 Acc@5 95.994 loss 0.874
Accuracy of the model on the 50000 test images: 81.8%
Max accuracy: 81.85%
Test:  [0/5]  eta: 0:00:34  loss: 0.9745 (0.9745)  acc1: 76.1719 (76.1719)  acc5: 93.6198 (93.6198)  time: 6.8288  data: 6.5927  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0054 (1.0248)  acc1: 74.0885 (71.7760)  acc5: 93.6198 (91.2000)  time: 1.5111  data: 1.3187  max mem: 55573
Test: Total time: 0:00:07 (1.5306 s / it)
* Acc@1 72.490 Acc@5 91.336 loss 1.042
Accuracy of the model EMA on 50000 test images: 72.5%
Max EMA accuracy: 72.49%
Epoch: [280]  [  0/156]  eta: 0:33:11  lr: 0.000101  min_lr: 0.000101  loss: 2.3962 (2.3962)  weight_decay: 0.0500 (0.0500)  time: 12.7652  data: 11.1993  max mem: 55573
Epoch: [280]  [ 10/156]  eta: 0:04:13  lr: 0.000101  min_lr: 0.000101  loss: 3.0947 (3.0152)  weight_decay: 0.0500 (0.0500)  time: 1.7384  data: 1.0559  max mem: 55573
Epoch: [280]  [ 20/156]  eta: 0:02:42  lr: 0.000100  min_lr: 0.000100  loss: 3.0747 (3.0042)  weight_decay: 0.0500 (0.0500)  time: 0.6144  data: 0.0210  max mem: 55573
Epoch: [280]  [ 30/156]  eta: 0:02:05  lr: 0.000099  min_lr: 0.000099  loss: 3.0030 (2.9311)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0005  max mem: 55573
Epoch: [280]  [ 40/156]  eta: 0:01:44  lr: 0.000099  min_lr: 0.000099  loss: 2.9751 (2.9596)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [280]  [ 50/156]  eta: 0:01:28  lr: 0.000098  min_lr: 0.000098  loss: 3.0363 (2.9141)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [280]  [ 60/156]  eta: 0:01:16  lr: 0.000097  min_lr: 0.000097  loss: 3.0363 (2.9013)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [280]  [ 70/156]  eta: 0:01:06  lr: 0.000097  min_lr: 0.000097  loss: 3.0472 (2.8983)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [280]  [ 80/156]  eta: 0:00:56  lr: 0.000096  min_lr: 0.000096  loss: 3.0177 (2.8970)  weight_decay: 0.0500 (0.0500)  time: 0.5935  data: 0.0004  max mem: 55573
Epoch: [280]  [ 90/156]  eta: 0:00:48  lr: 0.000096  min_lr: 0.000096  loss: 2.9567 (2.8680)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [280]  [100/156]  eta: 0:00:40  lr: 0.000095  min_lr: 0.000095  loss: 2.8161 (2.8537)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [280]  [110/156]  eta: 0:00:32  lr: 0.000094  min_lr: 0.000094  loss: 2.9916 (2.8603)  weight_decay: 0.0500 (0.0500)  time: 0.5947  data: 0.0004  max mem: 55573
Epoch: [280]  [120/156]  eta: 0:00:25  lr: 0.000094  min_lr: 0.000094  loss: 2.9916 (2.8345)  weight_decay: 0.0500 (0.0500)  time: 0.5934  data: 0.0004  max mem: 55573
Epoch: [280]  [130/156]  eta: 0:00:17  lr: 0.000093  min_lr: 0.000093  loss: 2.7402 (2.8309)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0010  max mem: 55573
Epoch: [280]  [140/156]  eta: 0:00:10  lr: 0.000093  min_lr: 0.000093  loss: 3.0753 (2.8456)  weight_decay: 0.0500 (0.0500)  time: 0.5862  data: 0.0009  max mem: 55573
Epoch: [280]  [150/156]  eta: 0:00:04  lr: 0.000092  min_lr: 0.000092  loss: 3.0745 (2.8548)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0002  max mem: 55573
Epoch: [280]  [155/156]  eta: 0:00:00  lr: 0.000092  min_lr: 0.000092  loss: 3.0479 (2.8493)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [280] Total time: 0:01:45 (0.6781 s / it)
Averaged stats: lr: 0.000092  min_lr: 0.000092  loss: 3.0479 (2.8715)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.7918 (0.7918)  acc1: 85.4167 (85.4167)  acc5: 97.3958 (97.3958)  time: 6.8633  data: 6.6259  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8881 (0.8745)  acc1: 81.2500 (80.8640)  acc5: 97.1354 (96.1280)  time: 1.5191  data: 1.3253  max mem: 55573
Test: Total time: 0:00:07 (1.5686 s / it)
* Acc@1 81.742 Acc@5 95.994 loss 0.860
Accuracy of the model on the 50000 test images: 81.7%
Max accuracy: 81.85%
Test:  [0/5]  eta: 0:00:36  loss: 0.9722 (0.9722)  acc1: 76.1719 (76.1719)  acc5: 93.6198 (93.6198)  time: 7.2602  data: 7.0242  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0031 (1.0224)  acc1: 74.2188 (71.8400)  acc5: 93.6198 (91.2000)  time: 1.5977  data: 1.4049  max mem: 55573
Test: Total time: 0:00:08 (1.6147 s / it)
* Acc@1 72.546 Acc@5 91.372 loss 1.040
Accuracy of the model EMA on 50000 test images: 72.5%
Max EMA accuracy: 72.55%
Epoch: [281]  [  0/156]  eta: 0:32:06  lr: 0.000092  min_lr: 0.000092  loss: 2.1736 (2.1736)  weight_decay: 0.0500 (0.0500)  time: 12.3490  data: 10.6832  max mem: 55573
Epoch: [281]  [ 10/156]  eta: 0:04:06  lr: 0.000091  min_lr: 0.000091  loss: 3.0667 (2.8989)  weight_decay: 0.0500 (0.0500)  time: 1.6903  data: 0.9716  max mem: 55573
Epoch: [281]  [ 20/156]  eta: 0:02:38  lr: 0.000090  min_lr: 0.000090  loss: 3.0390 (2.9200)  weight_decay: 0.0500 (0.0500)  time: 0.6096  data: 0.0004  max mem: 55573
Epoch: [281]  [ 30/156]  eta: 0:02:03  lr: 0.000090  min_lr: 0.000090  loss: 2.8059 (2.8474)  weight_decay: 0.0500 (0.0500)  time: 0.5956  data: 0.0004  max mem: 55573
Epoch: [281]  [ 40/156]  eta: 0:01:43  lr: 0.000089  min_lr: 0.000089  loss: 2.7517 (2.8580)  weight_decay: 0.0500 (0.0500)  time: 0.5934  data: 0.0005  max mem: 55573
Epoch: [281]  [ 50/156]  eta: 0:01:27  lr: 0.000089  min_lr: 0.000089  loss: 3.0054 (2.8560)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [281]  [ 60/156]  eta: 0:01:15  lr: 0.000088  min_lr: 0.000088  loss: 2.7002 (2.8261)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [281]  [ 70/156]  eta: 0:01:05  lr: 0.000087  min_lr: 0.000087  loss: 2.9783 (2.8512)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0006  max mem: 55573
Epoch: [281]  [ 80/156]  eta: 0:00:56  lr: 0.000087  min_lr: 0.000087  loss: 2.9783 (2.8571)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0006  max mem: 55573
Epoch: [281]  [ 90/156]  eta: 0:00:47  lr: 0.000086  min_lr: 0.000086  loss: 2.9241 (2.8747)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [281]  [100/156]  eta: 0:00:39  lr: 0.000086  min_lr: 0.000086  loss: 3.0839 (2.8883)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0004  max mem: 55573
Epoch: [281]  [110/156]  eta: 0:00:32  lr: 0.000085  min_lr: 0.000085  loss: 3.0449 (2.8735)  weight_decay: 0.0500 (0.0500)  time: 0.5936  data: 0.0004  max mem: 55573
Epoch: [281]  [120/156]  eta: 0:00:24  lr: 0.000084  min_lr: 0.000084  loss: 2.6024 (2.8660)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [281]  [130/156]  eta: 0:00:17  lr: 0.000084  min_lr: 0.000084  loss: 3.0602 (2.8673)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0012  max mem: 55573
Epoch: [281]  [140/156]  eta: 0:00:10  lr: 0.000083  min_lr: 0.000083  loss: 3.0709 (2.8785)  weight_decay: 0.0500 (0.0500)  time: 0.5860  data: 0.0011  max mem: 55573
Epoch: [281]  [150/156]  eta: 0:00:04  lr: 0.000083  min_lr: 0.000083  loss: 2.9858 (2.8694)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [281]  [155/156]  eta: 0:00:00  lr: 0.000082  min_lr: 0.000082  loss: 2.9858 (2.8795)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [281] Total time: 0:01:45 (0.6748 s / it)
Averaged stats: lr: 0.000082  min_lr: 0.000082  loss: 2.9858 (2.8654)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.8159 (0.8159)  acc1: 85.0260 (85.0260)  acc5: 97.1354 (97.1354)  time: 6.7516  data: 6.5145  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9061 (0.8973)  acc1: 81.6406 (80.8000)  acc5: 97.1354 (95.9360)  time: 1.4968  data: 1.3030  max mem: 55573
Test: Total time: 0:00:07 (1.5451 s / it)
* Acc@1 81.778 Acc@5 95.934 loss 0.879
Accuracy of the model on the 50000 test images: 81.8%
Max accuracy: 81.85%
Test:  [0/5]  eta: 0:00:36  loss: 0.9698 (0.9698)  acc1: 76.0417 (76.0417)  acc5: 93.6198 (93.6198)  time: 7.2125  data: 6.9765  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 1.0008 (1.0201)  acc1: 74.2188 (71.8080)  acc5: 93.6198 (91.2320)  time: 1.5878  data: 1.3954  max mem: 55573
Test: Total time: 0:00:08 (1.6046 s / it)
* Acc@1 72.590 Acc@5 91.394 loss 1.038
Accuracy of the model EMA on 50000 test images: 72.6%
Max EMA accuracy: 72.59%
Epoch: [282]  [  0/156]  eta: 0:36:14  lr: 0.000082  min_lr: 0.000082  loss: 3.0202 (3.0202)  weight_decay: 0.0500 (0.0500)  time: 13.9371  data: 11.6352  max mem: 55573
Epoch: [282]  [ 10/156]  eta: 0:04:23  lr: 0.000082  min_lr: 0.000082  loss: 2.9689 (2.9441)  weight_decay: 0.0500 (0.0500)  time: 1.8027  data: 1.0581  max mem: 55573
Epoch: [282]  [ 20/156]  eta: 0:02:46  lr: 0.000081  min_lr: 0.000081  loss: 2.8565 (2.8519)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [282]  [ 30/156]  eta: 0:02:08  lr: 0.000081  min_lr: 0.000081  loss: 2.7889 (2.7863)  weight_decay: 0.0500 (0.0500)  time: 0.5935  data: 0.0004  max mem: 55573
Epoch: [282]  [ 40/156]  eta: 0:01:46  lr: 0.000080  min_lr: 0.000080  loss: 2.8706 (2.8569)  weight_decay: 0.0500 (0.0500)  time: 0.5942  data: 0.0004  max mem: 55573
Epoch: [282]  [ 50/156]  eta: 0:01:30  lr: 0.000079  min_lr: 0.000079  loss: 2.9835 (2.8911)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [282]  [ 60/156]  eta: 0:01:17  lr: 0.000079  min_lr: 0.000079  loss: 3.0733 (2.8991)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [282]  [ 70/156]  eta: 0:01:07  lr: 0.000078  min_lr: 0.000078  loss: 2.9419 (2.8848)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [282]  [ 80/156]  eta: 0:00:57  lr: 0.000078  min_lr: 0.000078  loss: 2.8939 (2.8882)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [282]  [ 90/156]  eta: 0:00:48  lr: 0.000077  min_lr: 0.000077  loss: 3.0421 (2.8891)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [282]  [100/156]  eta: 0:00:40  lr: 0.000077  min_lr: 0.000077  loss: 3.0722 (2.8761)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0004  max mem: 55573
Epoch: [282]  [110/156]  eta: 0:00:32  lr: 0.000076  min_lr: 0.000076  loss: 2.9012 (2.8743)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [282]  [120/156]  eta: 0:00:25  lr: 0.000076  min_lr: 0.000076  loss: 3.0392 (2.8910)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [282]  [130/156]  eta: 0:00:18  lr: 0.000075  min_lr: 0.000075  loss: 3.0219 (2.8991)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0010  max mem: 55573
Epoch: [282]  [140/156]  eta: 0:00:10  lr: 0.000074  min_lr: 0.000074  loss: 3.0189 (2.8973)  weight_decay: 0.0500 (0.0500)  time: 0.5866  data: 0.0009  max mem: 55573
Epoch: [282]  [150/156]  eta: 0:00:04  lr: 0.000074  min_lr: 0.000074  loss: 3.0009 (2.8926)  weight_decay: 0.0500 (0.0500)  time: 0.5830  data: 0.0001  max mem: 55573
Epoch: [282]  [155/156]  eta: 0:00:00  lr: 0.000074  min_lr: 0.000074  loss: 2.9796 (2.8898)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0001  max mem: 55573
Epoch: [282] Total time: 0:01:46 (0.6821 s / it)
Averaged stats: lr: 0.000074  min_lr: 0.000074  loss: 2.9796 (2.8730)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.8014 (0.8014)  acc1: 84.6354 (84.6354)  acc5: 97.2656 (97.2656)  time: 7.1694  data: 6.9319  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8963 (0.8847)  acc1: 81.6406 (80.6400)  acc5: 97.2656 (96.0640)  time: 1.5804  data: 1.3865  max mem: 55573
Test: Total time: 0:00:08 (1.6311 s / it)
* Acc@1 81.884 Acc@5 95.984 loss 0.869
Accuracy of the model on the 50000 test images: 81.9%
Max accuracy: 81.88%
Test:  [0/5]  eta: 0:00:35  loss: 0.9674 (0.9674)  acc1: 76.0417 (76.0417)  acc5: 93.6198 (93.6198)  time: 7.0038  data: 6.7676  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9984 (1.0179)  acc1: 74.0885 (71.8080)  acc5: 93.6198 (91.2320)  time: 1.5462  data: 1.3536  max mem: 55573
Test: Total time: 0:00:07 (1.5661 s / it)
* Acc@1 72.644 Acc@5 91.416 loss 1.035
Accuracy of the model EMA on 50000 test images: 72.6%
Max EMA accuracy: 72.64%
Epoch: [283]  [  0/156]  eta: 0:32:09  lr: 0.000074  min_lr: 0.000074  loss: 3.0168 (3.0168)  weight_decay: 0.0500 (0.0500)  time: 12.3688  data: 11.7568  max mem: 55573
Epoch: [283]  [ 10/156]  eta: 0:04:15  lr: 0.000073  min_lr: 0.000073  loss: 3.1233 (3.0783)  weight_decay: 0.0500 (0.0500)  time: 1.7478  data: 1.1411  max mem: 55573
Epoch: [283]  [ 20/156]  eta: 0:02:42  lr: 0.000072  min_lr: 0.000072  loss: 3.1132 (3.0332)  weight_decay: 0.0500 (0.0500)  time: 0.6378  data: 0.0400  max mem: 55573
Epoch: [283]  [ 30/156]  eta: 0:02:06  lr: 0.000072  min_lr: 0.000072  loss: 3.0557 (2.9833)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [283]  [ 40/156]  eta: 0:01:44  lr: 0.000071  min_lr: 0.000071  loss: 3.1337 (3.0206)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0005  max mem: 55573
Epoch: [283]  [ 50/156]  eta: 0:01:29  lr: 0.000071  min_lr: 0.000071  loss: 3.2321 (3.0197)  weight_decay: 0.0500 (0.0500)  time: 0.5935  data: 0.0004  max mem: 55573
Epoch: [283]  [ 60/156]  eta: 0:01:16  lr: 0.000070  min_lr: 0.000070  loss: 2.9087 (2.9625)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0004  max mem: 55573
Epoch: [283]  [ 70/156]  eta: 0:01:06  lr: 0.000070  min_lr: 0.000070  loss: 2.8530 (2.9575)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [283]  [ 80/156]  eta: 0:00:56  lr: 0.000069  min_lr: 0.000069  loss: 3.0562 (2.9503)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0004  max mem: 55573
Epoch: [283]  [ 90/156]  eta: 0:00:48  lr: 0.000069  min_lr: 0.000069  loss: 2.9420 (2.9419)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [283]  [100/156]  eta: 0:00:40  lr: 0.000068  min_lr: 0.000068  loss: 3.0544 (2.9514)  weight_decay: 0.0500 (0.0500)  time: 0.5940  data: 0.0005  max mem: 55573
Epoch: [283]  [110/156]  eta: 0:00:32  lr: 0.000068  min_lr: 0.000068  loss: 3.0339 (2.9367)  weight_decay: 0.0500 (0.0500)  time: 0.5944  data: 0.0004  max mem: 55573
Epoch: [283]  [120/156]  eta: 0:00:25  lr: 0.000067  min_lr: 0.000067  loss: 2.8495 (2.9319)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [283]  [130/156]  eta: 0:00:17  lr: 0.000067  min_lr: 0.000067  loss: 2.9518 (2.9290)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0012  max mem: 55573
Epoch: [283]  [140/156]  eta: 0:00:10  lr: 0.000066  min_lr: 0.000066  loss: 2.7645 (2.9049)  weight_decay: 0.0500 (0.0500)  time: 0.5866  data: 0.0011  max mem: 55573
Epoch: [283]  [150/156]  eta: 0:00:04  lr: 0.000066  min_lr: 0.000066  loss: 2.9573 (2.9131)  weight_decay: 0.0500 (0.0500)  time: 0.5836  data: 0.0001  max mem: 55573
Epoch: [283]  [155/156]  eta: 0:00:00  lr: 0.000065  min_lr: 0.000065  loss: 3.0765 (2.9156)  weight_decay: 0.0500 (0.0500)  time: 0.5834  data: 0.0001  max mem: 55573
Epoch: [283] Total time: 0:01:45 (0.6783 s / it)
Averaged stats: lr: 0.000065  min_lr: 0.000065  loss: 3.0765 (2.8682)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8314 (0.8314)  acc1: 84.5052 (84.5052)  acc5: 97.3958 (97.3958)  time: 6.9747  data: 6.7373  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9257 (0.9127)  acc1: 81.6406 (80.7040)  acc5: 97.3958 (96.2240)  time: 1.5414  data: 1.3475  max mem: 55573
Test: Total time: 0:00:07 (1.5851 s / it)
* Acc@1 81.772 Acc@5 96.022 loss 0.896
Accuracy of the model on the 50000 test images: 81.8%
Max accuracy: 81.88%
Test:  [0/5]  eta: 0:00:36  loss: 0.9649 (0.9649)  acc1: 76.3021 (76.3021)  acc5: 93.6198 (93.6198)  time: 7.3321  data: 7.0960  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9959 (1.0154)  acc1: 74.3490 (72.0640)  acc5: 93.6198 (91.2320)  time: 1.6118  data: 1.4193  max mem: 55573
Test: Total time: 0:00:08 (1.6319 s / it)
* Acc@1 72.742 Acc@5 91.460 loss 1.032
Accuracy of the model EMA on 50000 test images: 72.7%
Max EMA accuracy: 72.74%
Epoch: [284]  [  0/156]  eta: 0:36:12  lr: 0.000065  min_lr: 0.000065  loss: 2.0253 (2.0253)  weight_decay: 0.0500 (0.0500)  time: 13.9276  data: 9.3562  max mem: 55573
Epoch: [284]  [ 10/156]  eta: 0:04:23  lr: 0.000065  min_lr: 0.000065  loss: 2.9073 (2.7983)  weight_decay: 0.0500 (0.0500)  time: 1.8067  data: 0.8509  max mem: 55573
Epoch: [284]  [ 20/156]  eta: 0:02:46  lr: 0.000064  min_lr: 0.000064  loss: 2.7498 (2.7712)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0004  max mem: 55573
Epoch: [284]  [ 30/156]  eta: 0:02:08  lr: 0.000064  min_lr: 0.000064  loss: 2.7498 (2.8487)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [284]  [ 40/156]  eta: 0:01:46  lr: 0.000063  min_lr: 0.000063  loss: 2.9884 (2.8287)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [284]  [ 50/156]  eta: 0:01:30  lr: 0.000063  min_lr: 0.000063  loss: 3.0362 (2.8830)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0004  max mem: 55573
Epoch: [284]  [ 60/156]  eta: 0:01:17  lr: 0.000062  min_lr: 0.000062  loss: 3.1333 (2.8965)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [284]  [ 70/156]  eta: 0:01:07  lr: 0.000062  min_lr: 0.000062  loss: 2.9779 (2.8773)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [284]  [ 80/156]  eta: 0:00:57  lr: 0.000061  min_lr: 0.000061  loss: 2.9779 (2.9010)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [284]  [ 90/156]  eta: 0:00:48  lr: 0.000061  min_lr: 0.000061  loss: 2.9566 (2.8996)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [284]  [100/156]  eta: 0:00:40  lr: 0.000060  min_lr: 0.000060  loss: 2.9100 (2.8927)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [284]  [110/156]  eta: 0:00:32  lr: 0.000060  min_lr: 0.000060  loss: 2.8789 (2.8678)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [284]  [120/156]  eta: 0:00:25  lr: 0.000059  min_lr: 0.000059  loss: 2.7212 (2.8564)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [284]  [130/156]  eta: 0:00:18  lr: 0.000059  min_lr: 0.000059  loss: 2.9992 (2.8617)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0024  max mem: 55573
Epoch: [284]  [140/156]  eta: 0:00:10  lr: 0.000058  min_lr: 0.000058  loss: 2.9992 (2.8591)  weight_decay: 0.0500 (0.0500)  time: 0.5863  data: 0.0022  max mem: 55573
Epoch: [284]  [150/156]  eta: 0:00:04  lr: 0.000058  min_lr: 0.000058  loss: 2.8909 (2.8476)  weight_decay: 0.0500 (0.0500)  time: 0.5832  data: 0.0001  max mem: 55573
Epoch: [284]  [155/156]  eta: 0:00:00  lr: 0.000058  min_lr: 0.000058  loss: 2.9291 (2.8522)  weight_decay: 0.0500 (0.0500)  time: 0.5827  data: 0.0001  max mem: 55573
Epoch: [284] Total time: 0:01:46 (0.6822 s / it)
Averaged stats: lr: 0.000058  min_lr: 0.000058  loss: 2.9291 (2.8765)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.7862 (0.7862)  acc1: 84.6354 (84.6354)  acc5: 97.2656 (97.2656)  time: 6.6780  data: 6.4408  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8816 (0.8692)  acc1: 81.9010 (80.8960)  acc5: 97.1354 (96.0960)  time: 1.4821  data: 1.2883  max mem: 55573
Test: Total time: 0:00:07 (1.5303 s / it)
* Acc@1 81.904 Acc@5 96.012 loss 0.854
Accuracy of the model on the 50000 test images: 81.9%
Max accuracy: 81.90%
Test:  [0/5]  eta: 0:00:34  loss: 0.9623 (0.9623)  acc1: 76.5625 (76.5625)  acc5: 93.7500 (93.7500)  time: 6.9725  data: 6.7365  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9935 (1.0130)  acc1: 74.2188 (72.1600)  acc5: 93.7500 (91.2640)  time: 1.5398  data: 1.3474  max mem: 55573
Test: Total time: 0:00:07 (1.5574 s / it)
* Acc@1 72.846 Acc@5 91.492 loss 1.030
Accuracy of the model EMA on 50000 test images: 72.8%
Max EMA accuracy: 72.85%
Epoch: [285]  [  0/156]  eta: 0:35:37  lr: 0.000058  min_lr: 0.000058  loss: 2.3760 (2.3760)  weight_decay: 0.0500 (0.0500)  time: 13.7018  data: 12.5015  max mem: 55573
Epoch: [285]  [ 10/156]  eta: 0:04:20  lr: 0.000057  min_lr: 0.000057  loss: 3.1132 (2.9954)  weight_decay: 0.0500 (0.0500)  time: 1.7849  data: 1.1369  max mem: 55573
Epoch: [285]  [ 20/156]  eta: 0:02:45  lr: 0.000057  min_lr: 0.000057  loss: 3.0209 (2.8511)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0004  max mem: 55573
Epoch: [285]  [ 30/156]  eta: 0:02:08  lr: 0.000056  min_lr: 0.000056  loss: 2.9831 (2.8225)  weight_decay: 0.0500 (0.0500)  time: 0.5954  data: 0.0004  max mem: 55573
Epoch: [285]  [ 40/156]  eta: 0:01:45  lr: 0.000056  min_lr: 0.000056  loss: 3.0889 (2.8674)  weight_decay: 0.0500 (0.0500)  time: 0.5940  data: 0.0004  max mem: 55573
Epoch: [285]  [ 50/156]  eta: 0:01:30  lr: 0.000055  min_lr: 0.000055  loss: 2.9176 (2.8097)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [285]  [ 60/156]  eta: 0:01:17  lr: 0.000055  min_lr: 0.000055  loss: 2.7905 (2.8267)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0005  max mem: 55573
Epoch: [285]  [ 70/156]  eta: 0:01:06  lr: 0.000054  min_lr: 0.000054  loss: 2.7905 (2.7906)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0004  max mem: 55573
Epoch: [285]  [ 80/156]  eta: 0:00:57  lr: 0.000054  min_lr: 0.000054  loss: 2.9389 (2.7998)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0004  max mem: 55573
Epoch: [285]  [ 90/156]  eta: 0:00:48  lr: 0.000053  min_lr: 0.000053  loss: 3.0323 (2.8112)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [285]  [100/156]  eta: 0:00:40  lr: 0.000053  min_lr: 0.000053  loss: 2.8982 (2.7953)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [285]  [110/156]  eta: 0:00:32  lr: 0.000052  min_lr: 0.000052  loss: 2.8344 (2.7935)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [285]  [120/156]  eta: 0:00:25  lr: 0.000052  min_lr: 0.000052  loss: 3.0923 (2.8005)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [285]  [130/156]  eta: 0:00:17  lr: 0.000051  min_lr: 0.000051  loss: 3.0472 (2.8067)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0010  max mem: 55573
Epoch: [285]  [140/156]  eta: 0:00:10  lr: 0.000051  min_lr: 0.000051  loss: 3.0472 (2.8179)  weight_decay: 0.0500 (0.0500)  time: 0.5864  data: 0.0009  max mem: 55573
Epoch: [285]  [150/156]  eta: 0:00:04  lr: 0.000051  min_lr: 0.000051  loss: 3.0727 (2.8310)  weight_decay: 0.0500 (0.0500)  time: 0.5824  data: 0.0001  max mem: 55573
Epoch: [285]  [155/156]  eta: 0:00:00  lr: 0.000050  min_lr: 0.000050  loss: 3.0727 (2.8288)  weight_decay: 0.0500 (0.0500)  time: 0.5829  data: 0.0001  max mem: 55573
Epoch: [285] Total time: 0:01:46 (0.6814 s / it)
Averaged stats: lr: 0.000050  min_lr: 0.000050  loss: 3.0727 (2.8518)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.8161 (0.8161)  acc1: 84.7656 (84.7656)  acc5: 97.1354 (97.1354)  time: 7.0845  data: 6.8471  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9113 (0.8990)  acc1: 82.0312 (80.8320)  acc5: 97.1354 (96.0320)  time: 1.5634  data: 1.3695  max mem: 55573
Test: Total time: 0:00:08 (1.6085 s / it)
* Acc@1 81.890 Acc@5 96.024 loss 0.884
Accuracy of the model on the 50000 test images: 81.9%
Max accuracy: 81.90%
Test:  [0/5]  eta: 0:00:35  loss: 0.9599 (0.9599)  acc1: 76.5625 (76.5625)  acc5: 93.7500 (93.7500)  time: 7.1427  data: 6.9066  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9911 (1.0105)  acc1: 74.2188 (72.2560)  acc5: 93.7500 (91.3600)  time: 1.5739  data: 1.3814  max mem: 55573
Test: Total time: 0:00:07 (1.5906 s / it)
* Acc@1 72.920 Acc@5 91.532 loss 1.027
Accuracy of the model EMA on 50000 test images: 72.9%
Max EMA accuracy: 72.92%
Epoch: [286]  [  0/156]  eta: 0:35:22  lr: 0.000050  min_lr: 0.000050  loss: 2.5102 (2.5102)  weight_decay: 0.0500 (0.0500)  time: 13.6086  data: 10.7925  max mem: 55573
Epoch: [286]  [ 10/156]  eta: 0:04:18  lr: 0.000050  min_lr: 0.000050  loss: 3.1600 (2.8889)  weight_decay: 0.0500 (0.0500)  time: 1.7730  data: 0.9816  max mem: 55573
Epoch: [286]  [ 20/156]  eta: 0:02:44  lr: 0.000049  min_lr: 0.000049  loss: 3.0893 (2.8497)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [286]  [ 30/156]  eta: 0:02:07  lr: 0.000049  min_lr: 0.000049  loss: 3.1179 (2.8952)  weight_decay: 0.0500 (0.0500)  time: 0.5893  data: 0.0005  max mem: 55573
Epoch: [286]  [ 40/156]  eta: 0:01:45  lr: 0.000048  min_lr: 0.000048  loss: 3.1147 (2.8701)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0004  max mem: 55573
Epoch: [286]  [ 50/156]  eta: 0:01:29  lr: 0.000048  min_lr: 0.000048  loss: 2.9239 (2.8835)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [286]  [ 60/156]  eta: 0:01:17  lr: 0.000048  min_lr: 0.000048  loss: 2.7915 (2.8409)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [286]  [ 70/156]  eta: 0:01:06  lr: 0.000047  min_lr: 0.000047  loss: 2.7915 (2.8474)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [286]  [ 80/156]  eta: 0:00:57  lr: 0.000047  min_lr: 0.000047  loss: 2.8335 (2.8279)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [286]  [ 90/156]  eta: 0:00:48  lr: 0.000046  min_lr: 0.000046  loss: 2.8335 (2.8383)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [286]  [100/156]  eta: 0:00:40  lr: 0.000046  min_lr: 0.000046  loss: 3.1145 (2.8624)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0006  max mem: 55573
Epoch: [286]  [110/156]  eta: 0:00:32  lr: 0.000045  min_lr: 0.000045  loss: 3.1145 (2.8739)  weight_decay: 0.0500 (0.0500)  time: 0.5890  data: 0.0005  max mem: 55573
Epoch: [286]  [120/156]  eta: 0:00:25  lr: 0.000045  min_lr: 0.000045  loss: 3.1064 (2.8774)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0004  max mem: 55573
Epoch: [286]  [130/156]  eta: 0:00:17  lr: 0.000045  min_lr: 0.000045  loss: 3.0182 (2.8813)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0010  max mem: 55573
Epoch: [286]  [140/156]  eta: 0:00:10  lr: 0.000044  min_lr: 0.000044  loss: 2.9897 (2.8798)  weight_decay: 0.0500 (0.0500)  time: 0.5865  data: 0.0009  max mem: 55573
Epoch: [286]  [150/156]  eta: 0:00:04  lr: 0.000044  min_lr: 0.000044  loss: 2.9253 (2.8724)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [286]  [155/156]  eta: 0:00:00  lr: 0.000044  min_lr: 0.000044  loss: 2.9253 (2.8699)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [286] Total time: 0:01:46 (0.6795 s / it)
Averaged stats: lr: 0.000044  min_lr: 0.000044  loss: 2.9253 (2.8477)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.7928 (0.7928)  acc1: 84.3750 (84.3750)  acc5: 97.1354 (97.1354)  time: 6.8678  data: 6.6306  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8845 (0.8724)  acc1: 82.0312 (80.9280)  acc5: 97.1354 (96.1600)  time: 1.5198  data: 1.3262  max mem: 55573
Test: Total time: 0:00:07 (1.5672 s / it)
* Acc@1 81.880 Acc@5 96.008 loss 0.857
Accuracy of the model on the 50000 test images: 81.9%
Max accuracy: 81.90%
Test:  [0/5]  eta: 0:00:35  loss: 0.9572 (0.9572)  acc1: 76.6927 (76.6927)  acc5: 93.7500 (93.7500)  time: 7.0800  data: 6.8415  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9883 (1.0078)  acc1: 74.3490 (72.3200)  acc5: 93.7500 (91.3600)  time: 1.5614  data: 1.3684  max mem: 55573
Test: Total time: 0:00:07 (1.5840 s / it)
* Acc@1 72.992 Acc@5 91.570 loss 1.024
Accuracy of the model EMA on 50000 test images: 73.0%
Max EMA accuracy: 72.99%
Epoch: [287]  [  0/156]  eta: 0:32:03  lr: 0.000043  min_lr: 0.000043  loss: 3.1491 (3.1491)  weight_decay: 0.0500 (0.0500)  time: 12.3294  data: 8.6382  max mem: 55573
Epoch: [287]  [ 10/156]  eta: 0:04:17  lr: 0.000043  min_lr: 0.000043  loss: 3.1491 (2.8690)  weight_decay: 0.0500 (0.0500)  time: 1.7654  data: 0.7859  max mem: 55573
Epoch: [287]  [ 20/156]  eta: 0:02:43  lr: 0.000043  min_lr: 0.000043  loss: 2.9434 (2.8666)  weight_decay: 0.0500 (0.0500)  time: 0.6497  data: 0.0005  max mem: 55573
Epoch: [287]  [ 30/156]  eta: 0:02:06  lr: 0.000042  min_lr: 0.000042  loss: 3.0048 (2.8840)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [287]  [ 40/156]  eta: 0:01:45  lr: 0.000042  min_lr: 0.000042  loss: 3.0175 (2.8688)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0005  max mem: 55573
Epoch: [287]  [ 50/156]  eta: 0:01:29  lr: 0.000041  min_lr: 0.000041  loss: 2.9220 (2.8303)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [287]  [ 60/156]  eta: 0:01:17  lr: 0.000041  min_lr: 0.000041  loss: 2.9421 (2.8355)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [287]  [ 70/156]  eta: 0:01:06  lr: 0.000041  min_lr: 0.000041  loss: 2.7052 (2.8216)  weight_decay: 0.0500 (0.0500)  time: 0.5926  data: 0.0005  max mem: 55573
Epoch: [287]  [ 80/156]  eta: 0:00:57  lr: 0.000040  min_lr: 0.000040  loss: 2.6833 (2.8006)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [287]  [ 90/156]  eta: 0:00:48  lr: 0.000040  min_lr: 0.000040  loss: 2.8434 (2.8047)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0005  max mem: 55573
Epoch: [287]  [100/156]  eta: 0:00:40  lr: 0.000039  min_lr: 0.000039  loss: 2.8239 (2.7876)  weight_decay: 0.0500 (0.0500)  time: 0.5898  data: 0.0005  max mem: 55573
Epoch: [287]  [110/156]  eta: 0:00:32  lr: 0.000039  min_lr: 0.000039  loss: 2.8239 (2.7935)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [287]  [120/156]  eta: 0:00:25  lr: 0.000039  min_lr: 0.000039  loss: 2.9736 (2.8074)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [287]  [130/156]  eta: 0:00:17  lr: 0.000038  min_lr: 0.000038  loss: 2.9736 (2.8018)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0011  max mem: 55573
Epoch: [287]  [140/156]  eta: 0:00:10  lr: 0.000038  min_lr: 0.000038  loss: 2.7893 (2.7998)  weight_decay: 0.0500 (0.0500)  time: 0.5881  data: 0.0009  max mem: 55573
Epoch: [287]  [150/156]  eta: 0:00:04  lr: 0.000037  min_lr: 0.000037  loss: 3.0011 (2.8013)  weight_decay: 0.0500 (0.0500)  time: 0.5829  data: 0.0001  max mem: 55573
Epoch: [287]  [155/156]  eta: 0:00:00  lr: 0.000037  min_lr: 0.000037  loss: 2.9931 (2.8003)  weight_decay: 0.0500 (0.0500)  time: 0.5823  data: 0.0001  max mem: 55573
Epoch: [287] Total time: 0:01:45 (0.6794 s / it)
Averaged stats: lr: 0.000037  min_lr: 0.000037  loss: 2.9931 (2.8369)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 0.7850 (0.7850)  acc1: 85.1562 (85.1562)  acc5: 97.1354 (97.1354)  time: 7.2335  data: 6.9961  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8828 (0.8641)  acc1: 81.6406 (80.9920)  acc5: 96.7448 (95.9680)  time: 1.5949  data: 1.3993  max mem: 55573
Test: Total time: 0:00:08 (1.6515 s / it)
* Acc@1 81.872 Acc@5 95.924 loss 0.849
Accuracy of the model on the 50000 test images: 81.9%
Max accuracy: 81.90%
Test:  [0/5]  eta: 0:00:36  loss: 0.9544 (0.9544)  acc1: 76.9531 (76.9531)  acc5: 93.7500 (93.7500)  time: 7.3306  data: 7.0946  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9856 (1.0051)  acc1: 74.6094 (72.4480)  acc5: 93.7500 (91.3920)  time: 1.6114  data: 1.4190  max mem: 55573
Test: Total time: 0:00:08 (1.6294 s / it)
* Acc@1 73.078 Acc@5 91.616 loss 1.021
Accuracy of the model EMA on 50000 test images: 73.1%
Max EMA accuracy: 73.08%
Epoch: [288]  [  0/156]  eta: 0:32:50  lr: 0.000037  min_lr: 0.000037  loss: 3.0875 (3.0875)  weight_decay: 0.0500 (0.0500)  time: 12.6303  data: 11.2286  max mem: 55573
Epoch: [288]  [ 10/156]  eta: 0:04:11  lr: 0.000037  min_lr: 0.000037  loss: 3.0723 (2.7767)  weight_decay: 0.0500 (0.0500)  time: 1.7203  data: 1.0214  max mem: 55573
Epoch: [288]  [ 20/156]  eta: 0:02:40  lr: 0.000036  min_lr: 0.000036  loss: 3.0515 (2.8677)  weight_decay: 0.0500 (0.0500)  time: 0.6103  data: 0.0005  max mem: 55573
Epoch: [288]  [ 30/156]  eta: 0:02:04  lr: 0.000036  min_lr: 0.000036  loss: 3.0456 (2.8959)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0004  max mem: 55573
Epoch: [288]  [ 40/156]  eta: 0:01:43  lr: 0.000036  min_lr: 0.000036  loss: 3.0456 (2.8856)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [288]  [ 50/156]  eta: 0:01:28  lr: 0.000035  min_lr: 0.000035  loss: 3.0522 (2.8742)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [288]  [ 60/156]  eta: 0:01:16  lr: 0.000035  min_lr: 0.000035  loss: 2.7860 (2.8451)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [288]  [ 70/156]  eta: 0:01:05  lr: 0.000035  min_lr: 0.000035  loss: 2.8750 (2.8649)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0005  max mem: 55573
Epoch: [288]  [ 80/156]  eta: 0:00:56  lr: 0.000034  min_lr: 0.000034  loss: 3.0016 (2.8752)  weight_decay: 0.0500 (0.0500)  time: 0.5935  data: 0.0005  max mem: 55573
Epoch: [288]  [ 90/156]  eta: 0:00:48  lr: 0.000034  min_lr: 0.000034  loss: 2.9417 (2.8646)  weight_decay: 0.0500 (0.0500)  time: 0.5940  data: 0.0005  max mem: 55573
Epoch: [288]  [100/156]  eta: 0:00:40  lr: 0.000033  min_lr: 0.000033  loss: 2.8958 (2.8622)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0005  max mem: 55573
Epoch: [288]  [110/156]  eta: 0:00:32  lr: 0.000033  min_lr: 0.000033  loss: 2.9906 (2.8730)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [288]  [120/156]  eta: 0:00:25  lr: 0.000033  min_lr: 0.000033  loss: 2.9906 (2.8747)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [288]  [130/156]  eta: 0:00:17  lr: 0.000032  min_lr: 0.000032  loss: 2.8610 (2.8681)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0010  max mem: 55573
Epoch: [288]  [140/156]  eta: 0:00:10  lr: 0.000032  min_lr: 0.000032  loss: 2.7924 (2.8670)  weight_decay: 0.0500 (0.0500)  time: 0.5865  data: 0.0009  max mem: 55573
Epoch: [288]  [150/156]  eta: 0:00:04  lr: 0.000032  min_lr: 0.000032  loss: 3.0264 (2.8810)  weight_decay: 0.0500 (0.0500)  time: 0.5821  data: 0.0001  max mem: 55573
Epoch: [288]  [155/156]  eta: 0:00:00  lr: 0.000031  min_lr: 0.000031  loss: 3.0339 (2.8875)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [288] Total time: 0:01:45 (0.6766 s / it)
Averaged stats: lr: 0.000031  min_lr: 0.000031  loss: 3.0339 (2.8531)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.8507 (0.8507)  acc1: 84.8958 (84.8958)  acc5: 97.2656 (97.2656)  time: 7.0260  data: 6.7886  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9471 (0.9332)  acc1: 81.2500 (80.6720)  acc5: 96.8750 (95.9680)  time: 1.5514  data: 1.3578  max mem: 55573
Test: Total time: 0:00:07 (1.5972 s / it)
* Acc@1 81.810 Acc@5 96.010 loss 0.917
Accuracy of the model on the 50000 test images: 81.8%
Max accuracy: 81.90%
Test:  [0/5]  eta: 0:00:36  loss: 0.9516 (0.9516)  acc1: 77.0833 (77.0833)  acc5: 93.7500 (93.7500)  time: 7.2111  data: 6.9750  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9830 (1.0024)  acc1: 74.6094 (72.5120)  acc5: 93.7500 (91.4240)  time: 1.5875  data: 1.3951  max mem: 55573
Test: Total time: 0:00:08 (1.6099 s / it)
* Acc@1 73.158 Acc@5 91.648 loss 1.018
Accuracy of the model EMA on 50000 test images: 73.2%
Max EMA accuracy: 73.16%
Epoch: [289]  [  0/156]  eta: 0:31:40  lr: 0.000031  min_lr: 0.000031  loss: 3.3554 (3.3554)  weight_decay: 0.0500 (0.0500)  time: 12.1840  data: 10.2032  max mem: 55573
Epoch: [289]  [ 10/156]  eta: 0:04:10  lr: 0.000031  min_lr: 0.000031  loss: 3.0822 (3.0132)  weight_decay: 0.0500 (0.0500)  time: 1.7170  data: 0.9280  max mem: 55573
Epoch: [289]  [ 20/156]  eta: 0:02:40  lr: 0.000031  min_lr: 0.000031  loss: 3.0480 (2.9721)  weight_decay: 0.0500 (0.0500)  time: 0.6304  data: 0.0005  max mem: 55573
Epoch: [289]  [ 30/156]  eta: 0:02:04  lr: 0.000030  min_lr: 0.000030  loss: 3.0946 (2.9647)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [289]  [ 40/156]  eta: 0:01:43  lr: 0.000030  min_lr: 0.000030  loss: 3.0263 (2.9218)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [289]  [ 50/156]  eta: 0:01:28  lr: 0.000030  min_lr: 0.000030  loss: 2.9952 (2.9236)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [289]  [ 60/156]  eta: 0:01:16  lr: 0.000029  min_lr: 0.000029  loss: 3.0477 (2.9167)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [289]  [ 70/156]  eta: 0:01:05  lr: 0.000029  min_lr: 0.000029  loss: 3.0477 (2.9165)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [289]  [ 80/156]  eta: 0:00:56  lr: 0.000029  min_lr: 0.000029  loss: 3.0076 (2.8888)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [289]  [ 90/156]  eta: 0:00:47  lr: 0.000028  min_lr: 0.000028  loss: 2.9021 (2.8940)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [289]  [100/156]  eta: 0:00:39  lr: 0.000028  min_lr: 0.000028  loss: 2.9760 (2.9047)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [289]  [110/156]  eta: 0:00:32  lr: 0.000028  min_lr: 0.000028  loss: 2.8991 (2.8903)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [289]  [120/156]  eta: 0:00:24  lr: 0.000027  min_lr: 0.000027  loss: 2.9122 (2.8848)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [289]  [130/156]  eta: 0:00:17  lr: 0.000027  min_lr: 0.000027  loss: 2.9122 (2.8825)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0012  max mem: 55573
Epoch: [289]  [140/156]  eta: 0:00:10  lr: 0.000027  min_lr: 0.000027  loss: 2.8368 (2.8846)  weight_decay: 0.0500 (0.0500)  time: 0.5862  data: 0.0010  max mem: 55573
Epoch: [289]  [150/156]  eta: 0:00:04  lr: 0.000026  min_lr: 0.000026  loss: 3.0024 (2.8832)  weight_decay: 0.0500 (0.0500)  time: 0.5822  data: 0.0001  max mem: 55573
Epoch: [289]  [155/156]  eta: 0:00:00  lr: 0.000026  min_lr: 0.000026  loss: 2.9468 (2.8817)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [289] Total time: 0:01:45 (0.6756 s / it)
Averaged stats: lr: 0.000026  min_lr: 0.000026  loss: 2.9468 (2.8537)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.7988 (0.7988)  acc1: 84.7656 (84.7656)  acc5: 97.2656 (97.2656)  time: 6.8551  data: 6.6180  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8995 (0.8808)  acc1: 81.3802 (80.8000)  acc5: 97.0052 (96.0320)  time: 1.5173  data: 1.3237  max mem: 55573
Test: Total time: 0:00:07 (1.5641 s / it)
* Acc@1 81.844 Acc@5 96.022 loss 0.867
Accuracy of the model on the 50000 test images: 81.8%
Max accuracy: 81.90%
Test:  [0/5]  eta: 0:00:34  loss: 0.9487 (0.9487)  acc1: 77.0833 (77.0833)  acc5: 93.7500 (93.7500)  time: 6.9733  data: 6.7371  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9803 (0.9995)  acc1: 74.6094 (72.5760)  acc5: 93.7500 (91.4560)  time: 1.5400  data: 1.3475  max mem: 55573
Test: Total time: 0:00:07 (1.5645 s / it)
* Acc@1 73.224 Acc@5 91.700 loss 1.015
Accuracy of the model EMA on 50000 test images: 73.2%
Max EMA accuracy: 73.22%
Epoch: [290]  [  0/156]  eta: 0:34:02  lr: 0.000026  min_lr: 0.000026  loss: 2.0972 (2.0972)  weight_decay: 0.0500 (0.0500)  time: 13.0960  data: 11.2859  max mem: 55573
Epoch: [290]  [ 10/156]  eta: 0:04:12  lr: 0.000026  min_lr: 0.000026  loss: 2.2559 (2.5711)  weight_decay: 0.0500 (0.0500)  time: 1.7300  data: 1.0264  max mem: 55573
Epoch: [290]  [ 20/156]  eta: 0:02:41  lr: 0.000026  min_lr: 0.000026  loss: 2.8997 (2.7508)  weight_decay: 0.0500 (0.0500)  time: 0.5931  data: 0.0005  max mem: 55573
Epoch: [290]  [ 30/156]  eta: 0:02:05  lr: 0.000025  min_lr: 0.000025  loss: 2.9835 (2.8070)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [290]  [ 40/156]  eta: 0:01:44  lr: 0.000025  min_lr: 0.000025  loss: 3.0443 (2.8403)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0005  max mem: 55573
Epoch: [290]  [ 50/156]  eta: 0:01:28  lr: 0.000025  min_lr: 0.000025  loss: 3.0395 (2.8589)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [290]  [ 60/156]  eta: 0:01:16  lr: 0.000024  min_lr: 0.000024  loss: 3.0842 (2.8842)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [290]  [ 70/156]  eta: 0:01:06  lr: 0.000024  min_lr: 0.000024  loss: 3.1245 (2.9165)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [290]  [ 80/156]  eta: 0:00:56  lr: 0.000024  min_lr: 0.000024  loss: 3.0605 (2.9162)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [290]  [ 90/156]  eta: 0:00:48  lr: 0.000023  min_lr: 0.000023  loss: 2.7673 (2.8780)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [290]  [100/156]  eta: 0:00:40  lr: 0.000023  min_lr: 0.000023  loss: 2.7673 (2.8760)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [290]  [110/156]  eta: 0:00:32  lr: 0.000023  min_lr: 0.000023  loss: 3.0256 (2.8713)  weight_decay: 0.0500 (0.0500)  time: 0.5938  data: 0.0004  max mem: 55573
Epoch: [290]  [120/156]  eta: 0:00:25  lr: 0.000022  min_lr: 0.000022  loss: 2.9084 (2.8601)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0004  max mem: 55573
Epoch: [290]  [130/156]  eta: 0:00:17  lr: 0.000022  min_lr: 0.000022  loss: 2.6264 (2.8512)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0010  max mem: 55573
Epoch: [290]  [140/156]  eta: 0:00:10  lr: 0.000022  min_lr: 0.000022  loss: 2.5372 (2.8217)  weight_decay: 0.0500 (0.0500)  time: 0.5859  data: 0.0009  max mem: 55573
Epoch: [290]  [150/156]  eta: 0:00:04  lr: 0.000022  min_lr: 0.000022  loss: 2.8465 (2.8413)  weight_decay: 0.0500 (0.0500)  time: 0.5828  data: 0.0001  max mem: 55573
Epoch: [290]  [155/156]  eta: 0:00:00  lr: 0.000021  min_lr: 0.000021  loss: 2.9433 (2.8471)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [290] Total time: 0:01:45 (0.6771 s / it)
Averaged stats: lr: 0.000021  min_lr: 0.000021  loss: 2.9433 (2.8529)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.8424 (0.8424)  acc1: 84.7656 (84.7656)  acc5: 97.0052 (97.0052)  time: 7.1317  data: 6.8937  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9341 (0.9199)  acc1: 81.5104 (80.8640)  acc5: 97.0052 (95.9680)  time: 1.5728  data: 1.3788  max mem: 55573
Test: Total time: 0:00:08 (1.6161 s / it)
* Acc@1 81.848 Acc@5 95.966 loss 0.908
Accuracy of the model on the 50000 test images: 81.8%
Max accuracy: 81.90%
Test:  [0/5]  eta: 0:00:35  loss: 0.9460 (0.9460)  acc1: 77.0833 (77.0833)  acc5: 93.7500 (93.7500)  time: 7.1119  data: 6.8758  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9776 (0.9967)  acc1: 74.8698 (72.6400)  acc5: 93.7500 (91.5200)  time: 1.5677  data: 1.3753  max mem: 55573
Test: Total time: 0:00:07 (1.5866 s / it)
* Acc@1 73.308 Acc@5 91.756 loss 1.012
Accuracy of the model EMA on 50000 test images: 73.3%
Max EMA accuracy: 73.31%
Epoch: [291]  [  0/156]  eta: 0:33:05  lr: 0.000021  min_lr: 0.000021  loss: 2.7024 (2.7024)  weight_decay: 0.0500 (0.0500)  time: 12.7284  data: 8.9519  max mem: 55573
Epoch: [291]  [ 10/156]  eta: 0:04:31  lr: 0.000021  min_lr: 0.000021  loss: 2.9492 (2.8120)  weight_decay: 0.0500 (0.0500)  time: 1.8567  data: 0.8142  max mem: 55573
Epoch: [291]  [ 20/156]  eta: 0:02:50  lr: 0.000021  min_lr: 0.000021  loss: 2.8525 (2.8054)  weight_decay: 0.0500 (0.0500)  time: 0.6800  data: 0.0004  max mem: 55573
Epoch: [291]  [ 30/156]  eta: 0:02:10  lr: 0.000021  min_lr: 0.000021  loss: 3.0332 (2.8431)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [291]  [ 40/156]  eta: 0:01:47  lr: 0.000020  min_lr: 0.000020  loss: 3.0845 (2.8431)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [291]  [ 50/156]  eta: 0:01:31  lr: 0.000020  min_lr: 0.000020  loss: 2.9757 (2.8458)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [291]  [ 60/156]  eta: 0:01:18  lr: 0.000020  min_lr: 0.000020  loss: 2.8873 (2.8316)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [291]  [ 70/156]  eta: 0:01:07  lr: 0.000019  min_lr: 0.000019  loss: 2.8873 (2.8301)  weight_decay: 0.0500 (0.0500)  time: 0.5929  data: 0.0005  max mem: 55573
Epoch: [291]  [ 80/156]  eta: 0:00:58  lr: 0.000019  min_lr: 0.000019  loss: 2.8086 (2.8173)  weight_decay: 0.0500 (0.0500)  time: 0.5925  data: 0.0005  max mem: 55573
Epoch: [291]  [ 90/156]  eta: 0:00:49  lr: 0.000019  min_lr: 0.000019  loss: 2.9851 (2.8454)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [291]  [100/156]  eta: 0:00:40  lr: 0.000019  min_lr: 0.000019  loss: 3.0320 (2.8265)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [291]  [110/156]  eta: 0:00:32  lr: 0.000018  min_lr: 0.000018  loss: 2.9695 (2.8355)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0005  max mem: 55573
Epoch: [291]  [120/156]  eta: 0:00:25  lr: 0.000018  min_lr: 0.000018  loss: 2.7717 (2.8188)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [291]  [130/156]  eta: 0:00:18  lr: 0.000018  min_lr: 0.000018  loss: 2.4876 (2.8157)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0014  max mem: 55573
Epoch: [291]  [140/156]  eta: 0:00:11  lr: 0.000018  min_lr: 0.000018  loss: 2.4876 (2.7896)  weight_decay: 0.0500 (0.0500)  time: 0.5868  data: 0.0012  max mem: 55573
Epoch: [291]  [150/156]  eta: 0:00:04  lr: 0.000017  min_lr: 0.000017  loss: 2.7828 (2.8030)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [291]  [155/156]  eta: 0:00:00  lr: 0.000017  min_lr: 0.000017  loss: 2.8447 (2.8027)  weight_decay: 0.0500 (0.0500)  time: 0.5830  data: 0.0001  max mem: 55573
Epoch: [291] Total time: 0:01:47 (0.6860 s / it)
Averaged stats: lr: 0.000017  min_lr: 0.000017  loss: 2.8447 (2.8537)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:36  loss: 0.7961 (0.7961)  acc1: 84.5052 (84.5052)  acc5: 97.2656 (97.2656)  time: 7.3960  data: 7.1586  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8927 (0.8782)  acc1: 81.9010 (80.9600)  acc5: 97.0052 (96.0640)  time: 1.6257  data: 1.4318  max mem: 55573
Test: Total time: 0:00:08 (1.6678 s / it)
* Acc@1 81.890 Acc@5 96.038 loss 0.864
Accuracy of the model on the 50000 test images: 81.9%
Max accuracy: 81.90%
Test:  [0/5]  eta: 0:00:34  loss: 0.9429 (0.9429)  acc1: 77.2135 (77.2135)  acc5: 93.7500 (93.7500)  time: 6.9735  data: 6.7371  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9748 (0.9937)  acc1: 74.7396 (72.6400)  acc5: 93.7500 (91.5520)  time: 1.5910  data: 1.3985  max mem: 55573
Test: Total time: 0:00:08 (1.6125 s / it)
* Acc@1 73.438 Acc@5 91.808 loss 1.009
Accuracy of the model EMA on 50000 test images: 73.4%
Max EMA accuracy: 73.44%
Epoch: [292]  [  0/156]  eta: 0:33:30  lr: 0.000017  min_lr: 0.000017  loss: 3.3552 (3.3552)  weight_decay: 0.0500 (0.0500)  time: 12.8871  data: 12.2592  max mem: 55573
Epoch: [292]  [ 10/156]  eta: 0:04:13  lr: 0.000017  min_lr: 0.000017  loss: 3.1688 (2.9962)  weight_decay: 0.0500 (0.0500)  time: 1.7383  data: 1.1149  max mem: 55573
Epoch: [292]  [ 20/156]  eta: 0:02:42  lr: 0.000017  min_lr: 0.000017  loss: 2.8273 (2.8098)  weight_decay: 0.0500 (0.0500)  time: 0.6076  data: 0.0004  max mem: 55573
Epoch: [292]  [ 30/156]  eta: 0:02:05  lr: 0.000016  min_lr: 0.000016  loss: 2.7934 (2.8302)  weight_decay: 0.0500 (0.0500)  time: 0.5917  data: 0.0004  max mem: 55573
Epoch: [292]  [ 40/156]  eta: 0:01:44  lr: 0.000016  min_lr: 0.000016  loss: 3.0566 (2.8688)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0004  max mem: 55573
Epoch: [292]  [ 50/156]  eta: 0:01:28  lr: 0.000016  min_lr: 0.000016  loss: 3.0120 (2.8432)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [292]  [ 60/156]  eta: 0:01:16  lr: 0.000016  min_lr: 0.000016  loss: 2.9895 (2.8602)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [292]  [ 70/156]  eta: 0:01:06  lr: 0.000015  min_lr: 0.000015  loss: 2.8553 (2.8443)  weight_decay: 0.0500 (0.0500)  time: 0.5895  data: 0.0005  max mem: 55573
Epoch: [292]  [ 80/156]  eta: 0:00:56  lr: 0.000015  min_lr: 0.000015  loss: 2.7597 (2.8301)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0005  max mem: 55573
Epoch: [292]  [ 90/156]  eta: 0:00:48  lr: 0.000015  min_lr: 0.000015  loss: 2.9450 (2.8240)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [292]  [100/156]  eta: 0:00:40  lr: 0.000015  min_lr: 0.000015  loss: 2.7164 (2.8015)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [292]  [110/156]  eta: 0:00:32  lr: 0.000014  min_lr: 0.000014  loss: 2.7164 (2.8103)  weight_decay: 0.0500 (0.0500)  time: 0.5927  data: 0.0005  max mem: 55573
Epoch: [292]  [120/156]  eta: 0:00:25  lr: 0.000014  min_lr: 0.000014  loss: 3.1511 (2.8291)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0004  max mem: 55573
Epoch: [292]  [130/156]  eta: 0:00:17  lr: 0.000014  min_lr: 0.000014  loss: 2.9493 (2.8365)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0010  max mem: 55573
Epoch: [292]  [140/156]  eta: 0:00:10  lr: 0.000014  min_lr: 0.000014  loss: 2.9234 (2.8423)  weight_decay: 0.0500 (0.0500)  time: 0.5863  data: 0.0009  max mem: 55573
Epoch: [292]  [150/156]  eta: 0:00:04  lr: 0.000013  min_lr: 0.000013  loss: 3.0988 (2.8602)  weight_decay: 0.0500 (0.0500)  time: 0.5831  data: 0.0002  max mem: 55573
Epoch: [292]  [155/156]  eta: 0:00:00  lr: 0.000013  min_lr: 0.000013  loss: 2.8884 (2.8471)  weight_decay: 0.0500 (0.0500)  time: 0.5825  data: 0.0001  max mem: 55573
Epoch: [292] Total time: 0:01:45 (0.6774 s / it)
Averaged stats: lr: 0.000013  min_lr: 0.000013  loss: 2.8884 (2.8553)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.7828 (0.7828)  acc1: 84.8958 (84.8958)  acc5: 97.2656 (97.2656)  time: 6.7270  data: 6.4900  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8845 (0.8684)  acc1: 81.5104 (80.8960)  acc5: 97.1354 (96.0320)  time: 1.4914  data: 1.2981  max mem: 55573
Test: Total time: 0:00:07 (1.5295 s / it)
* Acc@1 81.888 Acc@5 96.022 loss 0.854
Accuracy of the model on the 50000 test images: 81.9%
Max accuracy: 81.90%
Test:  [0/5]  eta: 0:00:36  loss: 0.9399 (0.9399)  acc1: 77.2135 (77.2135)  acc5: 93.8802 (93.8802)  time: 7.2509  data: 7.0149  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9719 (0.9907)  acc1: 75.0000 (72.7680)  acc5: 93.8802 (91.5840)  time: 1.5955  data: 1.4031  max mem: 55573
Test: Total time: 0:00:08 (1.6260 s / it)
* Acc@1 73.538 Acc@5 91.832 loss 1.006
Accuracy of the model EMA on 50000 test images: 73.5%
Max EMA accuracy: 73.54%
Epoch: [293]  [  0/156]  eta: 0:31:09  lr: 0.000013  min_lr: 0.000013  loss: 3.1664 (3.1664)  weight_decay: 0.0500 (0.0500)  time: 11.9865  data: 10.7557  max mem: 55573
Epoch: [293]  [ 10/156]  eta: 0:04:07  lr: 0.000013  min_lr: 0.000013  loss: 3.0679 (2.9283)  weight_decay: 0.0500 (0.0500)  time: 1.6925  data: 0.9782  max mem: 55573
Epoch: [293]  [ 20/156]  eta: 0:02:38  lr: 0.000013  min_lr: 0.000013  loss: 3.0037 (2.9609)  weight_decay: 0.0500 (0.0500)  time: 0.6269  data: 0.0005  max mem: 55573
Epoch: [293]  [ 30/156]  eta: 0:02:03  lr: 0.000013  min_lr: 0.000013  loss: 2.9378 (2.8642)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [293]  [ 40/156]  eta: 0:01:42  lr: 0.000012  min_lr: 0.000012  loss: 2.8914 (2.8820)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [293]  [ 50/156]  eta: 0:01:27  lr: 0.000012  min_lr: 0.000012  loss: 2.8585 (2.8478)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [293]  [ 60/156]  eta: 0:01:15  lr: 0.000012  min_lr: 0.000012  loss: 2.8590 (2.8530)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [293]  [ 70/156]  eta: 0:01:05  lr: 0.000012  min_lr: 0.000012  loss: 2.9792 (2.8673)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [293]  [ 80/156]  eta: 0:00:56  lr: 0.000012  min_lr: 0.000012  loss: 3.0380 (2.8762)  weight_decay: 0.0500 (0.0500)  time: 0.5923  data: 0.0004  max mem: 55573
Epoch: [293]  [ 90/156]  eta: 0:00:47  lr: 0.000011  min_lr: 0.000011  loss: 2.8843 (2.8363)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0004  max mem: 55573
Epoch: [293]  [100/156]  eta: 0:00:39  lr: 0.000011  min_lr: 0.000011  loss: 2.5096 (2.8184)  weight_decay: 0.0500 (0.0500)  time: 0.5920  data: 0.0004  max mem: 55573
Epoch: [293]  [110/156]  eta: 0:00:32  lr: 0.000011  min_lr: 0.000011  loss: 2.8958 (2.8242)  weight_decay: 0.0500 (0.0500)  time: 0.5939  data: 0.0005  max mem: 55573
Epoch: [293]  [120/156]  eta: 0:00:24  lr: 0.000011  min_lr: 0.000011  loss: 3.0192 (2.8396)  weight_decay: 0.0500 (0.0500)  time: 0.5957  data: 0.0004  max mem: 55573
Epoch: [293]  [130/156]  eta: 0:00:17  lr: 0.000011  min_lr: 0.000011  loss: 3.0357 (2.8336)  weight_decay: 0.0500 (0.0500)  time: 0.5938  data: 0.0011  max mem: 55573
Epoch: [293]  [140/156]  eta: 0:00:10  lr: 0.000010  min_lr: 0.000010  loss: 2.8001 (2.8309)  weight_decay: 0.0500 (0.0500)  time: 0.5861  data: 0.0010  max mem: 55573
Epoch: [293]  [150/156]  eta: 0:00:04  lr: 0.000010  min_lr: 0.000010  loss: 2.6256 (2.8197)  weight_decay: 0.0500 (0.0500)  time: 0.5820  data: 0.0001  max mem: 55573
Epoch: [293]  [155/156]  eta: 0:00:00  lr: 0.000010  min_lr: 0.000010  loss: 2.9359 (2.8148)  weight_decay: 0.0500 (0.0500)  time: 0.5818  data: 0.0001  max mem: 55573
Epoch: [293] Total time: 0:01:45 (0.6748 s / it)
Averaged stats: lr: 0.000010  min_lr: 0.000010  loss: 2.9359 (2.8447)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.7408 (0.7408)  acc1: 84.6354 (84.6354)  acc5: 97.2656 (97.2656)  time: 6.9570  data: 6.7197  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8427 (0.8257)  acc1: 82.4219 (81.1200)  acc5: 97.2656 (96.0960)  time: 1.5378  data: 1.3440  max mem: 55573
Test: Total time: 0:00:07 (1.5848 s / it)
* Acc@1 81.958 Acc@5 96.046 loss 0.809
Accuracy of the model on the 50000 test images: 82.0%
Max accuracy: 81.96%
Test:  [0/5]  eta: 0:00:33  loss: 0.9367 (0.9367)  acc1: 77.2135 (77.2135)  acc5: 93.8802 (93.8802)  time: 6.7995  data: 6.5634  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9690 (0.9875)  acc1: 75.3906 (72.8640)  acc5: 93.8802 (91.6160)  time: 1.5052  data: 1.3128  max mem: 55573
Test: Total time: 0:00:07 (1.5225 s / it)
* Acc@1 73.622 Acc@5 91.876 loss 1.003
Accuracy of the model EMA on 50000 test images: 73.6%
Max EMA accuracy: 73.62%
Epoch: [294]  [  0/156]  eta: 0:30:41  lr: 0.000010  min_lr: 0.000010  loss: 2.1934 (2.1934)  weight_decay: 0.0500 (0.0500)  time: 11.8042  data: 10.5507  max mem: 55573
Epoch: [294]  [ 10/156]  eta: 0:04:03  lr: 0.000010  min_lr: 0.000010  loss: 2.5740 (2.5848)  weight_decay: 0.0500 (0.0500)  time: 1.6663  data: 0.9638  max mem: 55573
Epoch: [294]  [ 20/156]  eta: 0:02:37  lr: 0.000010  min_lr: 0.000010  loss: 2.7381 (2.7451)  weight_decay: 0.0500 (0.0500)  time: 0.6226  data: 0.0028  max mem: 55573
Epoch: [294]  [ 30/156]  eta: 0:02:02  lr: 0.000009  min_lr: 0.000009  loss: 2.9648 (2.7495)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0005  max mem: 55573
Epoch: [294]  [ 40/156]  eta: 0:01:42  lr: 0.000009  min_lr: 0.000009  loss: 2.9194 (2.7514)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [294]  [ 50/156]  eta: 0:01:27  lr: 0.000009  min_lr: 0.000009  loss: 2.8902 (2.7494)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [294]  [ 60/156]  eta: 0:01:15  lr: 0.000009  min_lr: 0.000009  loss: 3.0295 (2.7813)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [294]  [ 70/156]  eta: 0:01:05  lr: 0.000009  min_lr: 0.000009  loss: 3.0529 (2.8079)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [294]  [ 80/156]  eta: 0:00:56  lr: 0.000009  min_lr: 0.000009  loss: 3.0601 (2.8344)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [294]  [ 90/156]  eta: 0:00:47  lr: 0.000008  min_lr: 0.000008  loss: 3.0654 (2.8370)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0005  max mem: 55573
Epoch: [294]  [100/156]  eta: 0:00:39  lr: 0.000008  min_lr: 0.000008  loss: 2.9787 (2.8429)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [294]  [110/156]  eta: 0:00:32  lr: 0.000008  min_lr: 0.000008  loss: 3.0866 (2.8529)  weight_decay: 0.0500 (0.0500)  time: 0.5918  data: 0.0005  max mem: 55573
Epoch: [294]  [120/156]  eta: 0:00:24  lr: 0.000008  min_lr: 0.000008  loss: 3.0026 (2.8445)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0005  max mem: 55573
Epoch: [294]  [130/156]  eta: 0:00:17  lr: 0.000008  min_lr: 0.000008  loss: 3.0026 (2.8460)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0011  max mem: 55573
Epoch: [294]  [140/156]  eta: 0:00:10  lr: 0.000008  min_lr: 0.000008  loss: 3.0376 (2.8545)  weight_decay: 0.0500 (0.0500)  time: 0.5878  data: 0.0009  max mem: 55573
Epoch: [294]  [150/156]  eta: 0:00:04  lr: 0.000007  min_lr: 0.000007  loss: 3.0376 (2.8525)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [294]  [155/156]  eta: 0:00:00  lr: 0.000007  min_lr: 0.000007  loss: 3.0042 (2.8550)  weight_decay: 0.0500 (0.0500)  time: 0.5819  data: 0.0001  max mem: 55573
Epoch: [294] Total time: 0:01:44 (0.6720 s / it)
Averaged stats: lr: 0.000007  min_lr: 0.000007  loss: 3.0042 (2.8448)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8175 (0.8175)  acc1: 84.7656 (84.7656)  acc5: 97.2656 (97.2656)  time: 6.9018  data: 6.6644  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9137 (0.8983)  acc1: 81.5104 (80.8000)  acc5: 97.0052 (96.0320)  time: 1.5268  data: 1.3329  max mem: 55573
Test: Total time: 0:00:07 (1.5661 s / it)
* Acc@1 81.872 Acc@5 96.010 loss 0.885
Accuracy of the model on the 50000 test images: 81.9%
Max accuracy: 81.96%
Test:  [0/5]  eta: 0:00:34  loss: 0.9337 (0.9337)  acc1: 77.3438 (77.3438)  acc5: 93.8802 (93.8802)  time: 6.8451  data: 6.6091  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9662 (0.9845)  acc1: 75.3906 (72.8960)  acc5: 93.8802 (91.6160)  time: 1.5144  data: 1.3219  max mem: 55573
Test: Total time: 0:00:07 (1.5349 s / it)
* Acc@1 73.720 Acc@5 91.926 loss 0.999
Accuracy of the model EMA on 50000 test images: 73.7%
Max EMA accuracy: 73.72%
Epoch: [295]  [  0/156]  eta: 0:30:57  lr: 0.000007  min_lr: 0.000007  loss: 2.8298 (2.8298)  weight_decay: 0.0500 (0.0500)  time: 11.9045  data: 7.7112  max mem: 55573
Epoch: [295]  [ 10/156]  eta: 0:04:00  lr: 0.000007  min_lr: 0.000007  loss: 2.9205 (2.8992)  weight_decay: 0.0500 (0.0500)  time: 1.6492  data: 0.7015  max mem: 55573
Epoch: [295]  [ 20/156]  eta: 0:02:35  lr: 0.000007  min_lr: 0.000007  loss: 2.9603 (2.8803)  weight_decay: 0.0500 (0.0500)  time: 0.6070  data: 0.0005  max mem: 55573
Epoch: [295]  [ 30/156]  eta: 0:02:01  lr: 0.000007  min_lr: 0.000007  loss: 2.9223 (2.8307)  weight_decay: 0.0500 (0.0500)  time: 0.5930  data: 0.0004  max mem: 55573
Epoch: [295]  [ 40/156]  eta: 0:01:41  lr: 0.000007  min_lr: 0.000007  loss: 2.9042 (2.8310)  weight_decay: 0.0500 (0.0500)  time: 0.5944  data: 0.0004  max mem: 55573
Epoch: [295]  [ 50/156]  eta: 0:01:27  lr: 0.000007  min_lr: 0.000007  loss: 2.9042 (2.8300)  weight_decay: 0.0500 (0.0500)  time: 0.5930  data: 0.0004  max mem: 55573
Epoch: [295]  [ 60/156]  eta: 0:01:15  lr: 0.000006  min_lr: 0.000006  loss: 2.9355 (2.8455)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0004  max mem: 55573
Epoch: [295]  [ 70/156]  eta: 0:01:05  lr: 0.000006  min_lr: 0.000006  loss: 2.9355 (2.8250)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0004  max mem: 55573
Epoch: [295]  [ 80/156]  eta: 0:00:55  lr: 0.000006  min_lr: 0.000006  loss: 3.0327 (2.8322)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [295]  [ 90/156]  eta: 0:00:47  lr: 0.000006  min_lr: 0.000006  loss: 3.0606 (2.8363)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [295]  [100/156]  eta: 0:00:39  lr: 0.000006  min_lr: 0.000006  loss: 3.0606 (2.8595)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [295]  [110/156]  eta: 0:00:32  lr: 0.000006  min_lr: 0.000006  loss: 3.0415 (2.8386)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [295]  [120/156]  eta: 0:00:24  lr: 0.000006  min_lr: 0.000006  loss: 2.8471 (2.8469)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0005  max mem: 55573
Epoch: [295]  [130/156]  eta: 0:00:17  lr: 0.000005  min_lr: 0.000005  loss: 3.0761 (2.8581)  weight_decay: 0.0500 (0.0500)  time: 0.5916  data: 0.0010  max mem: 55573
Epoch: [295]  [140/156]  eta: 0:00:10  lr: 0.000005  min_lr: 0.000005  loss: 3.0761 (2.8585)  weight_decay: 0.0500 (0.0500)  time: 0.5858  data: 0.0008  max mem: 55573
Epoch: [295]  [150/156]  eta: 0:00:04  lr: 0.000005  min_lr: 0.000005  loss: 2.9832 (2.8604)  weight_decay: 0.0500 (0.0500)  time: 0.5815  data: 0.0001  max mem: 55573
Epoch: [295]  [155/156]  eta: 0:00:00  lr: 0.000005  min_lr: 0.000005  loss: 3.0793 (2.8625)  weight_decay: 0.0500 (0.0500)  time: 0.5812  data: 0.0001  max mem: 55573
Epoch: [295] Total time: 0:01:44 (0.6713 s / it)
Averaged stats: lr: 0.000005  min_lr: 0.000005  loss: 3.0793 (2.8447)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.8398 (0.8398)  acc1: 84.5052 (84.5052)  acc5: 97.0052 (97.0052)  time: 6.6415  data: 6.4045  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9391 (0.9234)  acc1: 80.8594 (80.6720)  acc5: 96.8750 (95.9040)  time: 1.4737  data: 1.2810  max mem: 55573
Test: Total time: 0:00:07 (1.4941 s / it)
* Acc@1 81.816 Acc@5 95.898 loss 0.910
Accuracy of the model on the 50000 test images: 81.8%
Max accuracy: 81.96%
Test:  [0/5]  eta: 0:00:36  loss: 0.9307 (0.9307)  acc1: 77.3438 (77.3438)  acc5: 93.8802 (93.8802)  time: 7.3595  data: 7.1235  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9632 (0.9815)  acc1: 75.3906 (72.9280)  acc5: 93.8802 (91.6160)  time: 1.6172  data: 1.4248  max mem: 55573
Test: Total time: 0:00:08 (1.6436 s / it)
* Acc@1 73.824 Acc@5 91.966 loss 0.996
Accuracy of the model EMA on 50000 test images: 73.8%
Max EMA accuracy: 73.82%
Epoch: [296]  [  0/156]  eta: 0:32:43  lr: 0.000005  min_lr: 0.000005  loss: 2.9076 (2.9076)  weight_decay: 0.0500 (0.0500)  time: 12.5873  data: 10.3984  max mem: 55573
Epoch: [296]  [ 10/156]  eta: 0:04:07  lr: 0.000005  min_lr: 0.000005  loss: 2.9076 (2.8473)  weight_decay: 0.0500 (0.0500)  time: 1.6922  data: 0.9457  max mem: 55573
Epoch: [296]  [ 20/156]  eta: 0:02:38  lr: 0.000005  min_lr: 0.000005  loss: 3.0570 (2.9805)  weight_decay: 0.0500 (0.0500)  time: 0.5963  data: 0.0004  max mem: 55573
Epoch: [296]  [ 30/156]  eta: 0:02:03  lr: 0.000005  min_lr: 0.000005  loss: 2.9175 (2.8956)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [296]  [ 40/156]  eta: 0:01:42  lr: 0.000005  min_lr: 0.000005  loss: 2.7863 (2.8858)  weight_decay: 0.0500 (0.0500)  time: 0.5924  data: 0.0005  max mem: 55573
Epoch: [296]  [ 50/156]  eta: 0:01:27  lr: 0.000004  min_lr: 0.000004  loss: 2.9910 (2.8860)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [296]  [ 60/156]  eta: 0:01:15  lr: 0.000004  min_lr: 0.000004  loss: 2.9123 (2.8721)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [296]  [ 70/156]  eta: 0:01:05  lr: 0.000004  min_lr: 0.000004  loss: 2.8518 (2.8483)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0004  max mem: 55573
Epoch: [296]  [ 80/156]  eta: 0:00:56  lr: 0.000004  min_lr: 0.000004  loss: 2.7960 (2.8551)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0005  max mem: 55573
Epoch: [296]  [ 90/156]  eta: 0:00:47  lr: 0.000004  min_lr: 0.000004  loss: 2.8421 (2.8562)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [296]  [100/156]  eta: 0:00:39  lr: 0.000004  min_lr: 0.000004  loss: 2.9746 (2.8466)  weight_decay: 0.0500 (0.0500)  time: 0.5915  data: 0.0004  max mem: 55573
Epoch: [296]  [110/156]  eta: 0:00:32  lr: 0.000004  min_lr: 0.000004  loss: 3.0020 (2.8633)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0005  max mem: 55573
Epoch: [296]  [120/156]  eta: 0:00:24  lr: 0.000004  min_lr: 0.000004  loss: 3.0276 (2.8741)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0005  max mem: 55573
Epoch: [296]  [130/156]  eta: 0:00:17  lr: 0.000004  min_lr: 0.000004  loss: 2.7416 (2.8525)  weight_decay: 0.0500 (0.0500)  time: 0.5894  data: 0.0013  max mem: 55573
Epoch: [296]  [140/156]  eta: 0:00:10  lr: 0.000003  min_lr: 0.000003  loss: 2.9587 (2.8688)  weight_decay: 0.0500 (0.0500)  time: 0.5856  data: 0.0011  max mem: 55573
Epoch: [296]  [150/156]  eta: 0:00:04  lr: 0.000003  min_lr: 0.000003  loss: 3.0616 (2.8664)  weight_decay: 0.0500 (0.0500)  time: 0.5814  data: 0.0001  max mem: 55573
Epoch: [296]  [155/156]  eta: 0:00:00  lr: 0.000003  min_lr: 0.000003  loss: 3.0616 (2.8744)  weight_decay: 0.0500 (0.0500)  time: 0.5816  data: 0.0001  max mem: 55573
Epoch: [296] Total time: 0:01:45 (0.6738 s / it)
Averaged stats: lr: 0.000003  min_lr: 0.000003  loss: 3.0616 (2.8562)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:34  loss: 0.8350 (0.8350)  acc1: 84.7656 (84.7656)  acc5: 97.2656 (97.2656)  time: 6.9584  data: 6.7195  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9331 (0.9170)  acc1: 81.2500 (80.9280)  acc5: 96.8750 (95.9680)  time: 1.5383  data: 1.3440  max mem: 55573
Test: Total time: 0:00:07 (1.5860 s / it)
* Acc@1 81.832 Acc@5 95.970 loss 0.903
Accuracy of the model on the 50000 test images: 81.8%
Max accuracy: 81.96%
Test:  [0/5]  eta: 0:00:34  loss: 0.9274 (0.9274)  acc1: 77.4740 (77.4740)  acc5: 93.8802 (93.8802)  time: 6.9398  data: 6.7037  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9601 (0.9782)  acc1: 75.5208 (73.0240)  acc5: 93.8802 (91.6480)  time: 1.5333  data: 1.3408  max mem: 55573
Test: Total time: 0:00:07 (1.5578 s / it)
* Acc@1 73.920 Acc@5 92.020 loss 0.993
Accuracy of the model EMA on 50000 test images: 73.9%
Max EMA accuracy: 73.92%
Epoch: [297]  [  0/156]  eta: 0:30:27  lr: 0.000003  min_lr: 0.000003  loss: 3.3890 (3.3890)  weight_decay: 0.0500 (0.0500)  time: 11.7165  data: 8.3041  max mem: 55573
Epoch: [297]  [ 10/156]  eta: 0:04:05  lr: 0.000003  min_lr: 0.000003  loss: 3.0997 (3.0449)  weight_decay: 0.0500 (0.0500)  time: 1.6789  data: 0.7553  max mem: 55573
Epoch: [297]  [ 20/156]  eta: 0:02:37  lr: 0.000003  min_lr: 0.000003  loss: 2.9268 (2.8493)  weight_decay: 0.0500 (0.0500)  time: 0.6336  data: 0.0005  max mem: 55573
Epoch: [297]  [ 30/156]  eta: 0:02:03  lr: 0.000003  min_lr: 0.000003  loss: 2.9268 (2.8785)  weight_decay: 0.0500 (0.0500)  time: 0.5919  data: 0.0005  max mem: 55573
Epoch: [297]  [ 40/156]  eta: 0:01:42  lr: 0.000003  min_lr: 0.000003  loss: 3.1059 (2.9361)  weight_decay: 0.0500 (0.0500)  time: 0.5913  data: 0.0005  max mem: 55573
Epoch: [297]  [ 50/156]  eta: 0:01:27  lr: 0.000003  min_lr: 0.000003  loss: 2.9491 (2.8898)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [297]  [ 60/156]  eta: 0:01:15  lr: 0.000003  min_lr: 0.000003  loss: 2.8951 (2.8834)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [297]  [ 70/156]  eta: 0:01:05  lr: 0.000003  min_lr: 0.000003  loss: 2.9032 (2.8841)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [297]  [ 80/156]  eta: 0:00:56  lr: 0.000003  min_lr: 0.000003  loss: 3.0608 (2.8876)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [297]  [ 90/156]  eta: 0:00:47  lr: 0.000002  min_lr: 0.000002  loss: 2.9876 (2.8791)  weight_decay: 0.0500 (0.0500)  time: 0.5911  data: 0.0004  max mem: 55573
Epoch: [297]  [100/156]  eta: 0:00:39  lr: 0.000002  min_lr: 0.000002  loss: 2.9018 (2.8710)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0004  max mem: 55573
Epoch: [297]  [110/156]  eta: 0:00:32  lr: 0.000002  min_lr: 0.000002  loss: 2.7992 (2.8557)  weight_decay: 0.0500 (0.0500)  time: 0.5901  data: 0.0004  max mem: 55573
Epoch: [297]  [120/156]  eta: 0:00:24  lr: 0.000002  min_lr: 0.000002  loss: 2.9701 (2.8668)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0004  max mem: 55573
Epoch: [297]  [130/156]  eta: 0:00:17  lr: 0.000002  min_lr: 0.000002  loss: 2.9675 (2.8693)  weight_decay: 0.0500 (0.0500)  time: 0.5909  data: 0.0010  max mem: 55573
Epoch: [297]  [140/156]  eta: 0:00:10  lr: 0.000002  min_lr: 0.000002  loss: 2.9558 (2.8775)  weight_decay: 0.0500 (0.0500)  time: 0.5879  data: 0.0009  max mem: 55573
Epoch: [297]  [150/156]  eta: 0:00:04  lr: 0.000002  min_lr: 0.000002  loss: 2.9080 (2.8565)  weight_decay: 0.0500 (0.0500)  time: 0.5835  data: 0.0001  max mem: 55573
Epoch: [297]  [155/156]  eta: 0:00:00  lr: 0.000002  min_lr: 0.000002  loss: 2.8371 (2.8540)  weight_decay: 0.0500 (0.0500)  time: 0.5833  data: 0.0001  max mem: 55573
Epoch: [297] Total time: 0:01:45 (0.6735 s / it)
Averaged stats: lr: 0.000002  min_lr: 0.000002  loss: 2.8371 (2.8571)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:35  loss: 0.7598 (0.7598)  acc1: 84.6354 (84.6354)  acc5: 97.2656 (97.2656)  time: 7.0852  data: 6.8478  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.8607 (0.8456)  acc1: 82.1615 (81.0560)  acc5: 96.8750 (96.0320)  time: 1.5636  data: 1.3696  max mem: 55573
Test: Total time: 0:00:08 (1.6068 s / it)
* Acc@1 81.966 Acc@5 96.024 loss 0.830
Accuracy of the model on the 50000 test images: 82.0%
Max accuracy: 81.97%
Test:  [0/5]  eta: 0:00:33  loss: 0.9244 (0.9244)  acc1: 77.4740 (77.4740)  acc5: 93.8802 (93.8802)  time: 6.7108  data: 6.4748  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9574 (0.9752)  acc1: 75.5208 (72.9600)  acc5: 93.8802 (91.7120)  time: 1.4875  data: 1.2951  max mem: 55573
Test: Total time: 0:00:07 (1.5027 s / it)
* Acc@1 74.014 Acc@5 92.078 loss 0.989
Accuracy of the model EMA on 50000 test images: 74.0%
Max EMA accuracy: 74.01%
Epoch: [298]  [  0/156]  eta: 0:36:39  lr: 0.000002  min_lr: 0.000002  loss: 3.1110 (3.1110)  weight_decay: 0.0500 (0.0500)  time: 14.0979  data: 13.5134  max mem: 55573
Epoch: [298]  [ 10/156]  eta: 0:04:26  lr: 0.000002  min_lr: 0.000002  loss: 2.6630 (2.6563)  weight_decay: 0.0500 (0.0500)  time: 1.8227  data: 1.2288  max mem: 55573
Epoch: [298]  [ 20/156]  eta: 0:02:47  lr: 0.000002  min_lr: 0.000002  loss: 2.5014 (2.6700)  weight_decay: 0.0500 (0.0500)  time: 0.5921  data: 0.0003  max mem: 55573
Epoch: [298]  [ 30/156]  eta: 0:02:09  lr: 0.000002  min_lr: 0.000002  loss: 2.9936 (2.7282)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0004  max mem: 55573
Epoch: [298]  [ 40/156]  eta: 0:01:46  lr: 0.000002  min_lr: 0.000002  loss: 3.0350 (2.7538)  weight_decay: 0.0500 (0.0500)  time: 0.5914  data: 0.0005  max mem: 55573
Epoch: [298]  [ 50/156]  eta: 0:01:30  lr: 0.000002  min_lr: 0.000002  loss: 3.0350 (2.7672)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [298]  [ 60/156]  eta: 0:01:18  lr: 0.000002  min_lr: 0.000002  loss: 2.9931 (2.7770)  weight_decay: 0.0500 (0.0500)  time: 0.5908  data: 0.0005  max mem: 55573
Epoch: [298]  [ 70/156]  eta: 0:01:07  lr: 0.000002  min_lr: 0.000002  loss: 2.8794 (2.7603)  weight_decay: 0.0500 (0.0500)  time: 0.5907  data: 0.0005  max mem: 55573
Epoch: [298]  [ 80/156]  eta: 0:00:57  lr: 0.000002  min_lr: 0.000002  loss: 2.9347 (2.7858)  weight_decay: 0.0500 (0.0500)  time: 0.5905  data: 0.0005  max mem: 55573
Epoch: [298]  [ 90/156]  eta: 0:00:48  lr: 0.000002  min_lr: 0.000002  loss: 3.0463 (2.8106)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0005  max mem: 55573
Epoch: [298]  [100/156]  eta: 0:00:40  lr: 0.000001  min_lr: 0.000001  loss: 3.0166 (2.8102)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [298]  [110/156]  eta: 0:00:32  lr: 0.000001  min_lr: 0.000001  loss: 2.9596 (2.8126)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [298]  [120/156]  eta: 0:00:25  lr: 0.000001  min_lr: 0.000001  loss: 3.0473 (2.8286)  weight_decay: 0.0500 (0.0500)  time: 0.5906  data: 0.0004  max mem: 55573
Epoch: [298]  [130/156]  eta: 0:00:18  lr: 0.000001  min_lr: 0.000001  loss: 3.0072 (2.8131)  weight_decay: 0.0500 (0.0500)  time: 0.5900  data: 0.0011  max mem: 55573
Epoch: [298]  [140/156]  eta: 0:00:10  lr: 0.000001  min_lr: 0.000001  loss: 2.6989 (2.8125)  weight_decay: 0.0500 (0.0500)  time: 0.5861  data: 0.0009  max mem: 55573
Epoch: [298]  [150/156]  eta: 0:00:04  lr: 0.000001  min_lr: 0.000001  loss: 2.9530 (2.8272)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [298]  [155/156]  eta: 0:00:00  lr: 0.000001  min_lr: 0.000001  loss: 3.0692 (2.8344)  weight_decay: 0.0500 (0.0500)  time: 0.5817  data: 0.0001  max mem: 55573
Epoch: [298] Total time: 0:01:46 (0.6829 s / it)
Averaged stats: lr: 0.000001  min_lr: 0.000001  loss: 3.0692 (2.8486)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:33  loss: 0.8633 (0.8633)  acc1: 84.8958 (84.8958)  acc5: 97.2656 (97.2656)  time: 6.6935  data: 6.4565  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9571 (0.9422)  acc1: 81.3802 (80.8000)  acc5: 97.1354 (96.0640)  time: 1.4851  data: 1.2914  max mem: 55573
Test: Total time: 0:00:07 (1.5400 s / it)
* Acc@1 81.784 Acc@5 96.028 loss 0.929
Accuracy of the model on the 50000 test images: 81.8%
Max accuracy: 81.97%
Test:  [0/5]  eta: 0:00:35  loss: 0.9210 (0.9210)  acc1: 77.6042 (77.6042)  acc5: 94.0104 (94.0104)  time: 7.0315  data: 6.7951  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9542 (0.9717)  acc1: 75.5208 (72.9600)  acc5: 94.0104 (91.9040)  time: 1.5516  data: 1.3591  max mem: 55573
Test: Total time: 0:00:07 (1.5677 s / it)
* Acc@1 74.100 Acc@5 92.128 loss 0.986
Accuracy of the model EMA on 50000 test images: 74.1%
Max EMA accuracy: 74.10%
Epoch: [299]  [  0/156]  eta: 0:31:59  lr: 0.000001  min_lr: 0.000001  loss: 3.1710 (3.1710)  weight_decay: 0.0500 (0.0500)  time: 12.3018  data: 9.0384  max mem: 55573
Epoch: [299]  [ 10/156]  eta: 0:04:02  lr: 0.000001  min_lr: 0.000001  loss: 2.9251 (2.8009)  weight_decay: 0.0500 (0.0500)  time: 1.6597  data: 0.8221  max mem: 55573
Epoch: [299]  [ 20/156]  eta: 0:02:36  lr: 0.000001  min_lr: 0.000001  loss: 2.8581 (2.7667)  weight_decay: 0.0500 (0.0500)  time: 0.5928  data: 0.0005  max mem: 55573
Epoch: [299]  [ 30/156]  eta: 0:02:02  lr: 0.000001  min_lr: 0.000001  loss: 2.9887 (2.8109)  weight_decay: 0.0500 (0.0500)  time: 0.5912  data: 0.0005  max mem: 55573
Epoch: [299]  [ 40/156]  eta: 0:01:41  lr: 0.000001  min_lr: 0.000001  loss: 2.9588 (2.7911)  weight_decay: 0.0500 (0.0500)  time: 0.5904  data: 0.0005  max mem: 55573
Epoch: [299]  [ 50/156]  eta: 0:01:27  lr: 0.000001  min_lr: 0.000001  loss: 2.9431 (2.8041)  weight_decay: 0.0500 (0.0500)  time: 0.5896  data: 0.0005  max mem: 55573
Epoch: [299]  [ 60/156]  eta: 0:01:15  lr: 0.000001  min_lr: 0.000001  loss: 2.9431 (2.7788)  weight_decay: 0.0500 (0.0500)  time: 0.5910  data: 0.0005  max mem: 55573
Epoch: [299]  [ 70/156]  eta: 0:01:05  lr: 0.000001  min_lr: 0.000001  loss: 2.8625 (2.7956)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0005  max mem: 55573
Epoch: [299]  [ 80/156]  eta: 0:00:55  lr: 0.000001  min_lr: 0.000001  loss: 2.8719 (2.8019)  weight_decay: 0.0500 (0.0500)  time: 0.5899  data: 0.0004  max mem: 55573
Epoch: [299]  [ 90/156]  eta: 0:00:47  lr: 0.000001  min_lr: 0.000001  loss: 2.7964 (2.7788)  weight_decay: 0.0500 (0.0500)  time: 0.5902  data: 0.0004  max mem: 55573
Epoch: [299]  [100/156]  eta: 0:00:39  lr: 0.000001  min_lr: 0.000001  loss: 2.7309 (2.7743)  weight_decay: 0.0500 (0.0500)  time: 0.5903  data: 0.0004  max mem: 55573
Epoch: [299]  [110/156]  eta: 0:00:32  lr: 0.000001  min_lr: 0.000001  loss: 2.7309 (2.7686)  weight_decay: 0.0500 (0.0500)  time: 0.5939  data: 0.0004  max mem: 55573
Epoch: [299]  [120/156]  eta: 0:00:24  lr: 0.000001  min_lr: 0.000001  loss: 2.9476 (2.7835)  weight_decay: 0.0500 (0.0500)  time: 0.5930  data: 0.0004  max mem: 55573
Epoch: [299]  [130/156]  eta: 0:00:17  lr: 0.000001  min_lr: 0.000001  loss: 3.0920 (2.7981)  weight_decay: 0.0500 (0.0500)  time: 0.5897  data: 0.0010  max mem: 55573
Epoch: [299]  [140/156]  eta: 0:00:10  lr: 0.000001  min_lr: 0.000001  loss: 3.0089 (2.8152)  weight_decay: 0.0500 (0.0500)  time: 0.5869  data: 0.0009  max mem: 55573
Epoch: [299]  [150/156]  eta: 0:00:04  lr: 0.000001  min_lr: 0.000001  loss: 3.0798 (2.8276)  weight_decay: 0.0500 (0.0500)  time: 0.5830  data: 0.0001  max mem: 55573
Epoch: [299]  [155/156]  eta: 0:00:00  lr: 0.000001  min_lr: 0.000001  loss: 3.1008 (2.8296)  weight_decay: 0.0500 (0.0500)  time: 0.5826  data: 0.0001  max mem: 55573
Epoch: [299] Total time: 0:01:44 (0.6717 s / it)
Averaged stats: lr: 0.000001  min_lr: 0.000001  loss: 3.1008 (2.8573)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:32  loss: 0.8475 (0.8475)  acc1: 84.6354 (84.6354)  acc5: 97.0052 (97.0052)  time: 6.5963  data: 6.3595  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9444 (0.9285)  acc1: 81.1198 (80.7360)  acc5: 97.0052 (95.9680)  time: 1.4646  data: 1.2720  max mem: 55573
Test: Total time: 0:00:07 (1.4876 s / it)
* Acc@1 81.800 Acc@5 95.938 loss 0.915
Accuracy of the model on the 50000 test images: 81.8%
Max accuracy: 81.97%
Test:  [0/5]  eta: 0:00:35  loss: 0.9177 (0.9177)  acc1: 77.4740 (77.4740)  acc5: 94.0104 (94.0104)  time: 7.1367  data: 6.9008  max mem: 55573
Test:  [4/5]  eta: 0:00:01  loss: 0.9512 (0.9685)  acc1: 75.6510 (73.0880)  acc5: 94.0104 (91.9680)  time: 1.5727  data: 1.3802  max mem: 55573
Test: Total time: 0:00:07 (1.5884 s / it)
* Acc@1 74.194 Acc@5 92.164 loss 0.982
Accuracy of the model EMA on 50000 test images: 74.2%
Max EMA accuracy: 74.19%
Training time 10:21:37