Namespace(aa='rand-m9-mstd0.5-inc1', auto_resume=True, batch_size=128, cfg_path='iFormer_h.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=0.6, 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=1e-06, 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_h', 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=64)
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 0x7f61fe2a3520>
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 0x7f61fd5c7a60>
---------------------------
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 0x7f61fd5c7d00>
Warning: Enabling distributed evaluation with an eval dataset not divisible by process number. This will slightly alter validation results as extra duplicate entries are added to achieve equal num of samples per-process.
Mixup is activated!
Using EMA with decay = 0.99990000
Model = iFormer(
  (downsample_layers): ModuleList(
    (0): Sequential(
      (0): Conv2d_BN(
        (c): Conv2d(3, 48, kernel_size=(5, 5), stride=(2, 2), padding=(2, 2), bias=False)
        (bn): BatchNorm2d(48, 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(48, 192, kernel_size=(5, 5), stride=(2, 2), padding=(2, 2), bias=False)
          (bn): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        )
        (act): GELU(approximate='none')
        (conv_pwl_bn2): Conv2d_BN(
          (c): Conv2d(192, 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)
        )
      )
    )
    (1): Sequential(
      (0): Conv2d_BN(
        (c): Conv2d(96, 192, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
        (bn): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      )
    )
    (2): Sequential(
      (0): Conv2d_BN(
        (c): Conv2d(192, 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)
      )
    )
    (3): Sequential(
      (0): Conv2d_BN(
        (c): Conv2d(384, 768, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
        (bn): BatchNorm2d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      )
    )
  )
  (stages): ModuleList(
    (0): Sequential(
      (0): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (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): Identity()
          )
        )
      )
      (1): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (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): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (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()
          )
        )
      )
      (3): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (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()
          )
        )
      )
      (4): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (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): Sequential(
      (0): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(192, 192, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=192, bias=False)
                (bn): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(192, 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)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (1): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(192, 192, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=192, bias=False)
                (bn): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(192, 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)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (2): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(192, 192, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=192, bias=False)
                (bn): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(192, 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)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (3): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(192, 192, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=192, bias=False)
                (bn): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(192, 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)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (4): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(192, 192, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=192, bias=False)
                (bn): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(192, 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)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
    )
    (2): Sequential(
      (0): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 384, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=384, bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (1): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 384, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=384, bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (2): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 384, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=384, bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1536, 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): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 384, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=384, bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (4): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 384, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=384, bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (5): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 384, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=384, bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (6): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 384, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=384, bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (7): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 384, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=384, bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (8): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 384, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=384, bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (9): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 384, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=384, bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (10): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 384, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=384, bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (11): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 384, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=384, bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (12): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 384, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=384, bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (13): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 384, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=384, bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (14): 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)
            )
          )
        )
      )
      (15): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): 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)
              )
              (k): 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)
              )
              (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()
          )
        )
      )
      (16): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (17): 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)
            )
          )
        )
      )
      (18): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): 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)
              )
              (k): 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)
              )
              (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()
          )
        )
      )
      (19): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (20): 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)
            )
          )
        )
      )
      (21): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): 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)
              )
              (k): 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)
              )
              (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()
          )
        )
      )
      (22): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (23): 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)
            )
          )
        )
      )
      (24): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): 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)
              )
              (k): 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)
              )
              (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()
          )
        )
      )
      (25): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (26): 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)
            )
          )
        )
      )
      (27): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): 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)
              )
              (k): 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)
              )
              (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()
          )
        )
      )
      (28): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (29): 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)
            )
          )
        )
      )
      (30): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): 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)
              )
              (k): 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)
              )
              (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()
          )
        )
      )
      (31): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (32): 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)
            )
          )
        )
      )
      (33): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): 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)
              )
              (k): 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)
              )
              (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()
          )
        )
      )
      (34): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (35): 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)
            )
          )
        )
      )
      (36): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): 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)
              )
              (k): 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)
              )
              (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()
          )
        )
      )
      (37): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (38): 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)
            )
          )
        )
      )
      (39): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): 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)
              )
              (k): 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)
              )
              (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()
          )
        )
      )
      (40): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (41): 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)
            )
          )
        )
      )
      (42): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): 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)
              )
              (k): 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)
              )
              (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()
          )
        )
      )
      (43): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (44): 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)
            )
          )
        )
      )
      (45): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): 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)
              )
              (k): 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)
              )
              (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()
          )
        )
      )
      (46): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (47): 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)
            )
          )
        )
      )
      (48): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): 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)
              )
              (k): 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)
              )
              (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()
          )
        )
      )
      (49): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (50): 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)
            )
          )
        )
      )
      (51): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): 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)
              )
              (k): 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)
              )
              (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()
          )
        )
      )
      (52): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (53): 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)
            )
          )
        )
      )
      (54): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): 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)
              )
              (k): 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)
              )
              (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()
          )
        )
      )
      (55): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (56): 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)
            )
          )
        )
      )
      (57): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): 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)
              )
              (k): 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)
              )
              (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()
          )
        )
      )
      (58): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(1536, 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()
          )
        )
      )
      (59): BasicBlock(
        (block): ConvBlock(
          (token_channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(384, 384, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3), groups=384, bias=False)
                (bn): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): Conv2d_BN(
                (c): Conv2d(384, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (2): GELU(approximate='none')
              (3): Conv2d_BN(
                (c): Conv2d(1536, 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): Sequential(
      (0): BasicBlock(
        (block): RepCPE(
          (cpe): Residual(
            (m): Conv2d_BN(
              (c): Conv2d(768, 768, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=768, bias=False)
              (bn): BatchNorm2d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
        )
      )
      (1): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (k): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (gate_act): Sigmoid()
              (attn_drop): Dropout(p=0.0, inplace=False)
              (proj): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (v_gate): Conv2d_BN(
                (c): Conv2d(768, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (2): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(768, 3072, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(3072, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(3072, 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()
          )
        )
      )
      (3): BasicBlock(
        (block): RepCPE(
          (cpe): Residual(
            (m): Conv2d_BN(
              (c): Conv2d(768, 768, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=768, bias=False)
              (bn): BatchNorm2d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
        )
      )
      (4): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (k): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (gate_act): Sigmoid()
              (attn_drop): Dropout(p=0.0, inplace=False)
              (proj): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (v_gate): Conv2d_BN(
                (c): Conv2d(768, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (5): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(768, 3072, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(3072, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(3072, 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()
          )
        )
      )
      (6): BasicBlock(
        (block): RepCPE(
          (cpe): Residual(
            (m): Conv2d_BN(
              (c): Conv2d(768, 768, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=768, bias=False)
              (bn): BatchNorm2d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
        )
      )
      (7): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (k): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (gate_act): Sigmoid()
              (attn_drop): Dropout(p=0.0, inplace=False)
              (proj): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (v_gate): Conv2d_BN(
                (c): Conv2d(768, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (8): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(768, 3072, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(3072, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(3072, 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()
          )
        )
      )
      (9): BasicBlock(
        (block): RepCPE(
          (cpe): Residual(
            (m): Conv2d_BN(
              (c): Conv2d(768, 768, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=768, bias=False)
              (bn): BatchNorm2d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
        )
      )
      (10): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (k): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (gate_act): Sigmoid()
              (attn_drop): Dropout(p=0.0, inplace=False)
              (proj): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (v_gate): Conv2d_BN(
                (c): Conv2d(768, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (11): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(768, 3072, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(3072, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(3072, 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()
          )
        )
      )
      (12): BasicBlock(
        (block): RepCPE(
          (cpe): Residual(
            (m): Conv2d_BN(
              (c): Conv2d(768, 768, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=768, bias=False)
              (bn): BatchNorm2d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
        )
      )
      (13): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (k): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (gate_act): Sigmoid()
              (attn_drop): Dropout(p=0.0, inplace=False)
              (proj): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (v_gate): Conv2d_BN(
                (c): Conv2d(768, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (14): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(768, 3072, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(3072, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(3072, 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()
          )
        )
      )
      (15): BasicBlock(
        (block): RepCPE(
          (cpe): Residual(
            (m): Conv2d_BN(
              (c): Conv2d(768, 768, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=768, bias=False)
              (bn): BatchNorm2d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
        )
      )
      (16): BasicBlock(
        (block): SHMABlock(
          (token_channel_mixer): Residual_drop(
            (m): GAU2dv2(
              (q): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (k): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (gate_act): Sigmoid()
              (attn_drop): Dropout(p=0.0, inplace=False)
              (proj): Conv2d_BN(
                (c): Conv2d(768, 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)
              )
              (v_gate): Conv2d_BN(
                (c): Conv2d(768, 1536, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(1536, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
            )
            (drop_path): DropPath()
          )
        )
      )
      (17): BasicBlock(
        (block): FFN2d(
          (channel_mixer): Residual_drop(
            (m): Sequential(
              (0): Conv2d_BN(
                (c): Conv2d(768, 3072, kernel_size=(1, 1), stride=(1, 1), bias=False)
                (bn): BatchNorm2d(3072, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              )
              (1): GELU(approximate='none')
              (2): Conv2d_BN(
                (c): Conv2d(3072, 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()
          )
        )
      )
    )
  )
  (classifier): Classfier(
    (classifier): BN_Linear(
      (bn): BatchNorm1d(768, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      (l): Linear(in_features=768, out_features=1000, bias=True)
    )
  )
)
number of params: 99596632
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.gamma",
      "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.gamma",
      "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.0.2.block.token_channel_mixer.gamma",
      "stages.0.2.block.token_channel_mixer.m.0.c.weight",
      "stages.0.2.block.token_channel_mixer.m.1.c.weight",
      "stages.0.2.block.token_channel_mixer.m.3.c.weight",
      "stages.0.3.block.token_channel_mixer.gamma",
      "stages.0.3.block.token_channel_mixer.m.0.c.weight",
      "stages.0.3.block.token_channel_mixer.m.1.c.weight",
      "stages.0.3.block.token_channel_mixer.m.3.c.weight",
      "stages.0.4.block.token_channel_mixer.gamma",
      "stages.0.4.block.token_channel_mixer.m.0.c.weight",
      "stages.0.4.block.token_channel_mixer.m.1.c.weight",
      "stages.0.4.block.token_channel_mixer.m.3.c.weight",
      "stages.1.0.block.token_channel_mixer.gamma",
      "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.gamma",
      "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.1.2.block.token_channel_mixer.gamma",
      "stages.1.2.block.token_channel_mixer.m.0.c.weight",
      "stages.1.2.block.token_channel_mixer.m.1.c.weight",
      "stages.1.2.block.token_channel_mixer.m.3.c.weight",
      "stages.1.3.block.token_channel_mixer.gamma",
      "stages.1.3.block.token_channel_mixer.m.0.c.weight",
      "stages.1.3.block.token_channel_mixer.m.1.c.weight",
      "stages.1.3.block.token_channel_mixer.m.3.c.weight",
      "stages.1.4.block.token_channel_mixer.gamma",
      "stages.1.4.block.token_channel_mixer.m.0.c.weight",
      "stages.1.4.block.token_channel_mixer.m.1.c.weight",
      "stages.1.4.block.token_channel_mixer.m.3.c.weight",
      "stages.2.0.block.token_channel_mixer.gamma",
      "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.gamma",
      "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.gamma",
      "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.gamma",
      "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.gamma",
      "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.gamma",
      "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.gamma",
      "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.gamma",
      "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.token_channel_mixer.gamma",
      "stages.2.8.block.token_channel_mixer.m.0.c.weight",
      "stages.2.8.block.token_channel_mixer.m.1.c.weight",
      "stages.2.8.block.token_channel_mixer.m.3.c.weight",
      "stages.2.9.block.token_channel_mixer.gamma",
      "stages.2.9.block.token_channel_mixer.m.0.c.weight",
      "stages.2.9.block.token_channel_mixer.m.1.c.weight",
      "stages.2.9.block.token_channel_mixer.m.3.c.weight",
      "stages.2.10.block.token_channel_mixer.gamma",
      "stages.2.10.block.token_channel_mixer.m.0.c.weight",
      "stages.2.10.block.token_channel_mixer.m.1.c.weight",
      "stages.2.10.block.token_channel_mixer.m.3.c.weight",
      "stages.2.11.block.token_channel_mixer.gamma",
      "stages.2.11.block.token_channel_mixer.m.0.c.weight",
      "stages.2.11.block.token_channel_mixer.m.1.c.weight",
      "stages.2.11.block.token_channel_mixer.m.3.c.weight",
      "stages.2.12.block.token_channel_mixer.gamma",
      "stages.2.12.block.token_channel_mixer.m.0.c.weight",
      "stages.2.12.block.token_channel_mixer.m.1.c.weight",
      "stages.2.12.block.token_channel_mixer.m.3.c.weight",
      "stages.2.13.block.token_channel_mixer.gamma",
      "stages.2.13.block.token_channel_mixer.m.0.c.weight",
      "stages.2.13.block.token_channel_mixer.m.1.c.weight",
      "stages.2.13.block.token_channel_mixer.m.3.c.weight",
      "stages.2.14.block.cpe.m.c.weight",
      "stages.2.15.block.token_channel_mixer.gamma",
      "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.gamma",
      "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.gamma",
      "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.gamma",
      "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.gamma",
      "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.gamma",
      "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.gamma",
      "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.gamma",
      "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.gamma",
      "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.gamma",
      "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.gamma",
      "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.gamma",
      "stages.2.31.block.channel_mixer.m.0.c.weight",
      "stages.2.31.block.channel_mixer.m.2.c.weight",
      "stages.2.32.block.cpe.m.c.weight",
      "stages.2.33.block.token_channel_mixer.gamma",
      "stages.2.33.block.token_channel_mixer.m.q.c.weight",
      "stages.2.33.block.token_channel_mixer.m.k.c.weight",
      "stages.2.33.block.token_channel_mixer.m.proj.c.weight",
      "stages.2.33.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.2.34.block.channel_mixer.gamma",
      "stages.2.34.block.channel_mixer.m.0.c.weight",
      "stages.2.34.block.channel_mixer.m.2.c.weight",
      "stages.2.35.block.cpe.m.c.weight",
      "stages.2.36.block.token_channel_mixer.gamma",
      "stages.2.36.block.token_channel_mixer.m.q.c.weight",
      "stages.2.36.block.token_channel_mixer.m.k.c.weight",
      "stages.2.36.block.token_channel_mixer.m.proj.c.weight",
      "stages.2.36.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.2.37.block.channel_mixer.gamma",
      "stages.2.37.block.channel_mixer.m.0.c.weight",
      "stages.2.37.block.channel_mixer.m.2.c.weight",
      "stages.2.38.block.cpe.m.c.weight",
      "stages.2.39.block.token_channel_mixer.gamma",
      "stages.2.39.block.token_channel_mixer.m.q.c.weight",
      "stages.2.39.block.token_channel_mixer.m.k.c.weight",
      "stages.2.39.block.token_channel_mixer.m.proj.c.weight",
      "stages.2.39.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.2.40.block.channel_mixer.gamma",
      "stages.2.40.block.channel_mixer.m.0.c.weight",
      "stages.2.40.block.channel_mixer.m.2.c.weight",
      "stages.2.41.block.cpe.m.c.weight",
      "stages.2.42.block.token_channel_mixer.gamma",
      "stages.2.42.block.token_channel_mixer.m.q.c.weight",
      "stages.2.42.block.token_channel_mixer.m.k.c.weight",
      "stages.2.42.block.token_channel_mixer.m.proj.c.weight",
      "stages.2.42.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.2.43.block.channel_mixer.gamma",
      "stages.2.43.block.channel_mixer.m.0.c.weight",
      "stages.2.43.block.channel_mixer.m.2.c.weight",
      "stages.2.44.block.cpe.m.c.weight",
      "stages.2.45.block.token_channel_mixer.gamma",
      "stages.2.45.block.token_channel_mixer.m.q.c.weight",
      "stages.2.45.block.token_channel_mixer.m.k.c.weight",
      "stages.2.45.block.token_channel_mixer.m.proj.c.weight",
      "stages.2.45.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.2.46.block.channel_mixer.gamma",
      "stages.2.46.block.channel_mixer.m.0.c.weight",
      "stages.2.46.block.channel_mixer.m.2.c.weight",
      "stages.2.47.block.cpe.m.c.weight",
      "stages.2.48.block.token_channel_mixer.gamma",
      "stages.2.48.block.token_channel_mixer.m.q.c.weight",
      "stages.2.48.block.token_channel_mixer.m.k.c.weight",
      "stages.2.48.block.token_channel_mixer.m.proj.c.weight",
      "stages.2.48.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.2.49.block.channel_mixer.gamma",
      "stages.2.49.block.channel_mixer.m.0.c.weight",
      "stages.2.49.block.channel_mixer.m.2.c.weight",
      "stages.2.50.block.cpe.m.c.weight",
      "stages.2.51.block.token_channel_mixer.gamma",
      "stages.2.51.block.token_channel_mixer.m.q.c.weight",
      "stages.2.51.block.token_channel_mixer.m.k.c.weight",
      "stages.2.51.block.token_channel_mixer.m.proj.c.weight",
      "stages.2.51.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.2.52.block.channel_mixer.gamma",
      "stages.2.52.block.channel_mixer.m.0.c.weight",
      "stages.2.52.block.channel_mixer.m.2.c.weight",
      "stages.2.53.block.cpe.m.c.weight",
      "stages.2.54.block.token_channel_mixer.gamma",
      "stages.2.54.block.token_channel_mixer.m.q.c.weight",
      "stages.2.54.block.token_channel_mixer.m.k.c.weight",
      "stages.2.54.block.token_channel_mixer.m.proj.c.weight",
      "stages.2.54.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.2.55.block.channel_mixer.gamma",
      "stages.2.55.block.channel_mixer.m.0.c.weight",
      "stages.2.55.block.channel_mixer.m.2.c.weight",
      "stages.2.56.block.cpe.m.c.weight",
      "stages.2.57.block.token_channel_mixer.gamma",
      "stages.2.57.block.token_channel_mixer.m.q.c.weight",
      "stages.2.57.block.token_channel_mixer.m.k.c.weight",
      "stages.2.57.block.token_channel_mixer.m.proj.c.weight",
      "stages.2.57.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.2.58.block.channel_mixer.gamma",
      "stages.2.58.block.channel_mixer.m.0.c.weight",
      "stages.2.58.block.channel_mixer.m.2.c.weight",
      "stages.2.59.block.token_channel_mixer.gamma",
      "stages.2.59.block.token_channel_mixer.m.0.c.weight",
      "stages.2.59.block.token_channel_mixer.m.1.c.weight",
      "stages.2.59.block.token_channel_mixer.m.3.c.weight",
      "stages.3.0.block.cpe.m.c.weight",
      "stages.3.1.block.token_channel_mixer.gamma",
      "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.gamma",
      "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.gamma",
      "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.gamma",
      "stages.3.5.block.channel_mixer.m.0.c.weight",
      "stages.3.5.block.channel_mixer.m.2.c.weight",
      "stages.3.6.block.cpe.m.c.weight",
      "stages.3.7.block.token_channel_mixer.gamma",
      "stages.3.7.block.token_channel_mixer.m.q.c.weight",
      "stages.3.7.block.token_channel_mixer.m.k.c.weight",
      "stages.3.7.block.token_channel_mixer.m.proj.c.weight",
      "stages.3.7.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.3.8.block.channel_mixer.gamma",
      "stages.3.8.block.channel_mixer.m.0.c.weight",
      "stages.3.8.block.channel_mixer.m.2.c.weight",
      "stages.3.9.block.cpe.m.c.weight",
      "stages.3.10.block.token_channel_mixer.gamma",
      "stages.3.10.block.token_channel_mixer.m.q.c.weight",
      "stages.3.10.block.token_channel_mixer.m.k.c.weight",
      "stages.3.10.block.token_channel_mixer.m.proj.c.weight",
      "stages.3.10.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.3.11.block.channel_mixer.gamma",
      "stages.3.11.block.channel_mixer.m.0.c.weight",
      "stages.3.11.block.channel_mixer.m.2.c.weight",
      "stages.3.12.block.cpe.m.c.weight",
      "stages.3.13.block.token_channel_mixer.gamma",
      "stages.3.13.block.token_channel_mixer.m.q.c.weight",
      "stages.3.13.block.token_channel_mixer.m.k.c.weight",
      "stages.3.13.block.token_channel_mixer.m.proj.c.weight",
      "stages.3.13.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.3.14.block.channel_mixer.gamma",
      "stages.3.14.block.channel_mixer.m.0.c.weight",
      "stages.3.14.block.channel_mixer.m.2.c.weight",
      "stages.3.15.block.cpe.m.c.weight",
      "stages.3.16.block.token_channel_mixer.gamma",
      "stages.3.16.block.token_channel_mixer.m.q.c.weight",
      "stages.3.16.block.token_channel_mixer.m.k.c.weight",
      "stages.3.16.block.token_channel_mixer.m.proj.c.weight",
      "stages.3.16.block.token_channel_mixer.m.v_gate.c.weight",
      "stages.3.17.block.channel_mixer.gamma",
      "stages.3.17.block.channel_mixer.m.0.c.weight",
      "stages.3.17.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.0.2.block.token_channel_mixer.m.0.bn.weight",
      "stages.0.2.block.token_channel_mixer.m.0.bn.bias",
      "stages.0.2.block.token_channel_mixer.m.1.bn.weight",
      "stages.0.2.block.token_channel_mixer.m.1.bn.bias",
      "stages.0.2.block.token_channel_mixer.m.3.bn.weight",
      "stages.0.2.block.token_channel_mixer.m.3.bn.bias",
      "stages.0.3.block.token_channel_mixer.m.0.bn.weight",
      "stages.0.3.block.token_channel_mixer.m.0.bn.bias",
      "stages.0.3.block.token_channel_mixer.m.1.bn.weight",
      "stages.0.3.block.token_channel_mixer.m.1.bn.bias",
      "stages.0.3.block.token_channel_mixer.m.3.bn.weight",
      "stages.0.3.block.token_channel_mixer.m.3.bn.bias",
      "stages.0.4.block.token_channel_mixer.m.0.bn.weight",
      "stages.0.4.block.token_channel_mixer.m.0.bn.bias",
      "stages.0.4.block.token_channel_mixer.m.1.bn.weight",
      "stages.0.4.block.token_channel_mixer.m.1.bn.bias",
      "stages.0.4.block.token_channel_mixer.m.3.bn.weight",
      "stages.0.4.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.1.2.block.token_channel_mixer.m.0.bn.weight",
      "stages.1.2.block.token_channel_mixer.m.0.bn.bias",
      "stages.1.2.block.token_channel_mixer.m.1.bn.weight",
      "stages.1.2.block.token_channel_mixer.m.1.bn.bias",
      "stages.1.2.block.token_channel_mixer.m.3.bn.weight",
      "stages.1.2.block.token_channel_mixer.m.3.bn.bias",
      "stages.1.3.block.token_channel_mixer.m.0.bn.weight",
      "stages.1.3.block.token_channel_mixer.m.0.bn.bias",
      "stages.1.3.block.token_channel_mixer.m.1.bn.weight",
      "stages.1.3.block.token_channel_mixer.m.1.bn.bias",
      "stages.1.3.block.token_channel_mixer.m.3.bn.weight",
      "stages.1.3.block.token_channel_mixer.m.3.bn.bias",
      "stages.1.4.block.token_channel_mixer.m.0.bn.weight",
      "stages.1.4.block.token_channel_mixer.m.0.bn.bias",
      "stages.1.4.block.token_channel_mixer.m.1.bn.weight",
      "stages.1.4.block.token_channel_mixer.m.1.bn.bias",
      "stages.1.4.block.token_channel_mixer.m.3.bn.weight",
      "stages.1.4.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.token_channel_mixer.m.0.bn.weight",
      "stages.2.8.block.token_channel_mixer.m.0.bn.bias",
      "stages.2.8.block.token_channel_mixer.m.1.bn.weight",
      "stages.2.8.block.token_channel_mixer.m.1.bn.bias",
      "stages.2.8.block.token_channel_mixer.m.3.bn.weight",
      "stages.2.8.block.token_channel_mixer.m.3.bn.bias",
      "stages.2.9.block.token_channel_mixer.m.0.bn.weight",
      "stages.2.9.block.token_channel_mixer.m.0.bn.bias",
      "stages.2.9.block.token_channel_mixer.m.1.bn.weight",
      "stages.2.9.block.token_channel_mixer.m.1.bn.bias",
      "stages.2.9.block.token_channel_mixer.m.3.bn.weight",
      "stages.2.9.block.token_channel_mixer.m.3.bn.bias",
      "stages.2.10.block.token_channel_mixer.m.0.bn.weight",
      "stages.2.10.block.token_channel_mixer.m.0.bn.bias",
      "stages.2.10.block.token_channel_mixer.m.1.bn.weight",
      "stages.2.10.block.token_channel_mixer.m.1.bn.bias",
      "stages.2.10.block.token_channel_mixer.m.3.bn.weight",
      "stages.2.10.block.token_channel_mixer.m.3.bn.bias",
      "stages.2.11.block.token_channel_mixer.m.0.bn.weight",
      "stages.2.11.block.token_channel_mixer.m.0.bn.bias",
      "stages.2.11.block.token_channel_mixer.m.1.bn.weight",
      "stages.2.11.block.token_channel_mixer.m.1.bn.bias",
      "stages.2.11.block.token_channel_mixer.m.3.bn.weight",
      "stages.2.11.block.token_channel_mixer.m.3.bn.bias",
      "stages.2.12.block.token_channel_mixer.m.0.bn.weight",
      "stages.2.12.block.token_channel_mixer.m.0.bn.bias",
      "stages.2.12.block.token_channel_mixer.m.1.bn.weight",
      "stages.2.12.block.token_channel_mixer.m.1.bn.bias",
      "stages.2.12.block.token_channel_mixer.m.3.bn.weight",
      "stages.2.12.block.token_channel_mixer.m.3.bn.bias",
      "stages.2.13.block.token_channel_mixer.m.0.bn.weight",
      "stages.2.13.block.token_channel_mixer.m.0.bn.bias",
      "stages.2.13.block.token_channel_mixer.m.1.bn.weight",
      "stages.2.13.block.token_channel_mixer.m.1.bn.bias",
      "stages.2.13.block.token_channel_mixer.m.3.bn.weight",
      "stages.2.13.block.token_channel_mixer.m.3.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.cpe.m.bn.weight",
      "stages.2.32.block.cpe.m.bn.bias",
      "stages.2.33.block.token_channel_mixer.m.q.bn.weight",
      "stages.2.33.block.token_channel_mixer.m.q.bn.bias",
      "stages.2.33.block.token_channel_mixer.m.k.bn.weight",
      "stages.2.33.block.token_channel_mixer.m.k.bn.bias",
      "stages.2.33.block.token_channel_mixer.m.proj.bn.weight",
      "stages.2.33.block.token_channel_mixer.m.proj.bn.bias",
      "stages.2.33.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.2.33.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.2.34.block.channel_mixer.m.0.bn.weight",
      "stages.2.34.block.channel_mixer.m.0.bn.bias",
      "stages.2.34.block.channel_mixer.m.2.bn.weight",
      "stages.2.34.block.channel_mixer.m.2.bn.bias",
      "stages.2.35.block.cpe.m.bn.weight",
      "stages.2.35.block.cpe.m.bn.bias",
      "stages.2.36.block.token_channel_mixer.m.q.bn.weight",
      "stages.2.36.block.token_channel_mixer.m.q.bn.bias",
      "stages.2.36.block.token_channel_mixer.m.k.bn.weight",
      "stages.2.36.block.token_channel_mixer.m.k.bn.bias",
      "stages.2.36.block.token_channel_mixer.m.proj.bn.weight",
      "stages.2.36.block.token_channel_mixer.m.proj.bn.bias",
      "stages.2.36.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.2.36.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.2.37.block.channel_mixer.m.0.bn.weight",
      "stages.2.37.block.channel_mixer.m.0.bn.bias",
      "stages.2.37.block.channel_mixer.m.2.bn.weight",
      "stages.2.37.block.channel_mixer.m.2.bn.bias",
      "stages.2.38.block.cpe.m.bn.weight",
      "stages.2.38.block.cpe.m.bn.bias",
      "stages.2.39.block.token_channel_mixer.m.q.bn.weight",
      "stages.2.39.block.token_channel_mixer.m.q.bn.bias",
      "stages.2.39.block.token_channel_mixer.m.k.bn.weight",
      "stages.2.39.block.token_channel_mixer.m.k.bn.bias",
      "stages.2.39.block.token_channel_mixer.m.proj.bn.weight",
      "stages.2.39.block.token_channel_mixer.m.proj.bn.bias",
      "stages.2.39.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.2.39.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.2.40.block.channel_mixer.m.0.bn.weight",
      "stages.2.40.block.channel_mixer.m.0.bn.bias",
      "stages.2.40.block.channel_mixer.m.2.bn.weight",
      "stages.2.40.block.channel_mixer.m.2.bn.bias",
      "stages.2.41.block.cpe.m.bn.weight",
      "stages.2.41.block.cpe.m.bn.bias",
      "stages.2.42.block.token_channel_mixer.m.q.bn.weight",
      "stages.2.42.block.token_channel_mixer.m.q.bn.bias",
      "stages.2.42.block.token_channel_mixer.m.k.bn.weight",
      "stages.2.42.block.token_channel_mixer.m.k.bn.bias",
      "stages.2.42.block.token_channel_mixer.m.proj.bn.weight",
      "stages.2.42.block.token_channel_mixer.m.proj.bn.bias",
      "stages.2.42.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.2.42.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.2.43.block.channel_mixer.m.0.bn.weight",
      "stages.2.43.block.channel_mixer.m.0.bn.bias",
      "stages.2.43.block.channel_mixer.m.2.bn.weight",
      "stages.2.43.block.channel_mixer.m.2.bn.bias",
      "stages.2.44.block.cpe.m.bn.weight",
      "stages.2.44.block.cpe.m.bn.bias",
      "stages.2.45.block.token_channel_mixer.m.q.bn.weight",
      "stages.2.45.block.token_channel_mixer.m.q.bn.bias",
      "stages.2.45.block.token_channel_mixer.m.k.bn.weight",
      "stages.2.45.block.token_channel_mixer.m.k.bn.bias",
      "stages.2.45.block.token_channel_mixer.m.proj.bn.weight",
      "stages.2.45.block.token_channel_mixer.m.proj.bn.bias",
      "stages.2.45.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.2.45.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.2.46.block.channel_mixer.m.0.bn.weight",
      "stages.2.46.block.channel_mixer.m.0.bn.bias",
      "stages.2.46.block.channel_mixer.m.2.bn.weight",
      "stages.2.46.block.channel_mixer.m.2.bn.bias",
      "stages.2.47.block.cpe.m.bn.weight",
      "stages.2.47.block.cpe.m.bn.bias",
      "stages.2.48.block.token_channel_mixer.m.q.bn.weight",
      "stages.2.48.block.token_channel_mixer.m.q.bn.bias",
      "stages.2.48.block.token_channel_mixer.m.k.bn.weight",
      "stages.2.48.block.token_channel_mixer.m.k.bn.bias",
      "stages.2.48.block.token_channel_mixer.m.proj.bn.weight",
      "stages.2.48.block.token_channel_mixer.m.proj.bn.bias",
      "stages.2.48.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.2.48.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.2.49.block.channel_mixer.m.0.bn.weight",
      "stages.2.49.block.channel_mixer.m.0.bn.bias",
      "stages.2.49.block.channel_mixer.m.2.bn.weight",
      "stages.2.49.block.channel_mixer.m.2.bn.bias",
      "stages.2.50.block.cpe.m.bn.weight",
      "stages.2.50.block.cpe.m.bn.bias",
      "stages.2.51.block.token_channel_mixer.m.q.bn.weight",
      "stages.2.51.block.token_channel_mixer.m.q.bn.bias",
      "stages.2.51.block.token_channel_mixer.m.k.bn.weight",
      "stages.2.51.block.token_channel_mixer.m.k.bn.bias",
      "stages.2.51.block.token_channel_mixer.m.proj.bn.weight",
      "stages.2.51.block.token_channel_mixer.m.proj.bn.bias",
      "stages.2.51.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.2.51.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.2.52.block.channel_mixer.m.0.bn.weight",
      "stages.2.52.block.channel_mixer.m.0.bn.bias",
      "stages.2.52.block.channel_mixer.m.2.bn.weight",
      "stages.2.52.block.channel_mixer.m.2.bn.bias",
      "stages.2.53.block.cpe.m.bn.weight",
      "stages.2.53.block.cpe.m.bn.bias",
      "stages.2.54.block.token_channel_mixer.m.q.bn.weight",
      "stages.2.54.block.token_channel_mixer.m.q.bn.bias",
      "stages.2.54.block.token_channel_mixer.m.k.bn.weight",
      "stages.2.54.block.token_channel_mixer.m.k.bn.bias",
      "stages.2.54.block.token_channel_mixer.m.proj.bn.weight",
      "stages.2.54.block.token_channel_mixer.m.proj.bn.bias",
      "stages.2.54.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.2.54.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.2.55.block.channel_mixer.m.0.bn.weight",
      "stages.2.55.block.channel_mixer.m.0.bn.bias",
      "stages.2.55.block.channel_mixer.m.2.bn.weight",
      "stages.2.55.block.channel_mixer.m.2.bn.bias",
      "stages.2.56.block.cpe.m.bn.weight",
      "stages.2.56.block.cpe.m.bn.bias",
      "stages.2.57.block.token_channel_mixer.m.q.bn.weight",
      "stages.2.57.block.token_channel_mixer.m.q.bn.bias",
      "stages.2.57.block.token_channel_mixer.m.k.bn.weight",
      "stages.2.57.block.token_channel_mixer.m.k.bn.bias",
      "stages.2.57.block.token_channel_mixer.m.proj.bn.weight",
      "stages.2.57.block.token_channel_mixer.m.proj.bn.bias",
      "stages.2.57.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.2.57.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.2.58.block.channel_mixer.m.0.bn.weight",
      "stages.2.58.block.channel_mixer.m.0.bn.bias",
      "stages.2.58.block.channel_mixer.m.2.bn.weight",
      "stages.2.58.block.channel_mixer.m.2.bn.bias",
      "stages.2.59.block.token_channel_mixer.m.0.bn.weight",
      "stages.2.59.block.token_channel_mixer.m.0.bn.bias",
      "stages.2.59.block.token_channel_mixer.m.1.bn.weight",
      "stages.2.59.block.token_channel_mixer.m.1.bn.bias",
      "stages.2.59.block.token_channel_mixer.m.3.bn.weight",
      "stages.2.59.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",
      "stages.3.6.block.cpe.m.bn.weight",
      "stages.3.6.block.cpe.m.bn.bias",
      "stages.3.7.block.token_channel_mixer.m.q.bn.weight",
      "stages.3.7.block.token_channel_mixer.m.q.bn.bias",
      "stages.3.7.block.token_channel_mixer.m.k.bn.weight",
      "stages.3.7.block.token_channel_mixer.m.k.bn.bias",
      "stages.3.7.block.token_channel_mixer.m.proj.bn.weight",
      "stages.3.7.block.token_channel_mixer.m.proj.bn.bias",
      "stages.3.7.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.3.7.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.3.8.block.channel_mixer.m.0.bn.weight",
      "stages.3.8.block.channel_mixer.m.0.bn.bias",
      "stages.3.8.block.channel_mixer.m.2.bn.weight",
      "stages.3.8.block.channel_mixer.m.2.bn.bias",
      "stages.3.9.block.cpe.m.bn.weight",
      "stages.3.9.block.cpe.m.bn.bias",
      "stages.3.10.block.token_channel_mixer.m.q.bn.weight",
      "stages.3.10.block.token_channel_mixer.m.q.bn.bias",
      "stages.3.10.block.token_channel_mixer.m.k.bn.weight",
      "stages.3.10.block.token_channel_mixer.m.k.bn.bias",
      "stages.3.10.block.token_channel_mixer.m.proj.bn.weight",
      "stages.3.10.block.token_channel_mixer.m.proj.bn.bias",
      "stages.3.10.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.3.10.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.3.11.block.channel_mixer.m.0.bn.weight",
      "stages.3.11.block.channel_mixer.m.0.bn.bias",
      "stages.3.11.block.channel_mixer.m.2.bn.weight",
      "stages.3.11.block.channel_mixer.m.2.bn.bias",
      "stages.3.12.block.cpe.m.bn.weight",
      "stages.3.12.block.cpe.m.bn.bias",
      "stages.3.13.block.token_channel_mixer.m.q.bn.weight",
      "stages.3.13.block.token_channel_mixer.m.q.bn.bias",
      "stages.3.13.block.token_channel_mixer.m.k.bn.weight",
      "stages.3.13.block.token_channel_mixer.m.k.bn.bias",
      "stages.3.13.block.token_channel_mixer.m.proj.bn.weight",
      "stages.3.13.block.token_channel_mixer.m.proj.bn.bias",
      "stages.3.13.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.3.13.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.3.14.block.channel_mixer.m.0.bn.weight",
      "stages.3.14.block.channel_mixer.m.0.bn.bias",
      "stages.3.14.block.channel_mixer.m.2.bn.weight",
      "stages.3.14.block.channel_mixer.m.2.bn.bias",
      "stages.3.15.block.cpe.m.bn.weight",
      "stages.3.15.block.cpe.m.bn.bias",
      "stages.3.16.block.token_channel_mixer.m.q.bn.weight",
      "stages.3.16.block.token_channel_mixer.m.q.bn.bias",
      "stages.3.16.block.token_channel_mixer.m.k.bn.weight",
      "stages.3.16.block.token_channel_mixer.m.k.bn.bias",
      "stages.3.16.block.token_channel_mixer.m.proj.bn.weight",
      "stages.3.16.block.token_channel_mixer.m.proj.bn.bias",
      "stages.3.16.block.token_channel_mixer.m.v_gate.bn.weight",
      "stages.3.16.block.token_channel_mixer.m.v_gate.bn.bias",
      "stages.3.17.block.channel_mixer.m.0.bn.weight",
      "stages.3.17.block.channel_mixer.m.0.bn.bias",
      "stages.3.17.block.channel_mixer.m.2.bn.weight",
      "stages.3.17.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: 1:39:53  lr: 0.000000  min_lr: 0.000000  loss: 7.1039 (7.1039)  weight_decay: 0.0500 (0.0500)  time: 38.4193  data: 3.7894  max mem: 57114
Epoch: [0]  [ 10/156]  eta: 0:13:49  lr: 0.000026  min_lr: 0.000026  loss: 7.0620 (7.0630)  weight_decay: 0.0500 (0.0500)  time: 5.6824  data: 0.3449  max mem: 57114
Epoch: [0]  [ 20/156]  eta: 0:08:13  lr: 0.000051  min_lr: 0.000051  loss: 7.0482 (7.0528)  weight_decay: 0.0500 (0.0500)  time: 1.8881  data: 0.0004  max mem: 57114
Epoch: [0]  [ 30/156]  eta: 0:06:04  lr: 0.000077  min_lr: 0.000077  loss: 7.0064 (7.0232)  weight_decay: 0.0500 (0.0500)  time: 1.3567  data: 0.0004  max mem: 57114
Epoch: [0]  [ 40/156]  eta: 0:04:43  lr: 0.000103  min_lr: 0.000103  loss: 6.9370 (7.0009)  weight_decay: 0.0500 (0.0500)  time: 1.1988  data: 0.0005  max mem: 57114
Epoch: [0]  [ 50/156]  eta: 0:03:42  lr: 0.000128  min_lr: 0.000128  loss: 6.9130 (6.9794)  weight_decay: 0.0500 (0.0500)  time: 0.8605  data: 0.0004  max mem: 57114
Epoch: [0]  [ 60/156]  eta: 0:02:58  lr: 0.000154  min_lr: 0.000154  loss: 6.8703 (6.9557)  weight_decay: 0.0500 (0.0500)  time: 0.6704  data: 0.0004  max mem: 57114
Epoch: [0]  [ 70/156]  eta: 0:02:25  lr: 0.000180  min_lr: 0.000180  loss: 6.8283 (6.9370)  weight_decay: 0.0500 (0.0500)  time: 0.6716  data: 0.0004  max mem: 57114
Epoch: [0]  [ 80/156]  eta: 0:01:59  lr: 0.000205  min_lr: 0.000205  loss: 6.8090 (6.9192)  weight_decay: 0.0500 (0.0500)  time: 0.6742  data: 0.0004  max mem: 57114
Epoch: [0]  [ 90/156]  eta: 0:01:37  lr: 0.000231  min_lr: 0.000231  loss: 6.7934 (6.9051)  weight_decay: 0.0500 (0.0500)  time: 0.6845  data: 0.0004  max mem: 57114
Epoch: [0]  [100/156]  eta: 0:01:18  lr: 0.000256  min_lr: 0.000256  loss: 6.7964 (6.8928)  weight_decay: 0.0500 (0.0500)  time: 0.6841  data: 0.0004  max mem: 57114
Epoch: [0]  [110/156]  eta: 0:01:01  lr: 0.000282  min_lr: 0.000282  loss: 6.7762 (6.8792)  weight_decay: 0.0500 (0.0500)  time: 0.6761  data: 0.0004  max mem: 57114
Epoch: [0]  [120/156]  eta: 0:00:45  lr: 0.000308  min_lr: 0.000308  loss: 6.7329 (6.8658)  weight_decay: 0.0500 (0.0500)  time: 0.6750  data: 0.0004  max mem: 57114
Epoch: [0]  [130/156]  eta: 0:00:31  lr: 0.000333  min_lr: 0.000333  loss: 6.6943 (6.8526)  weight_decay: 0.0500 (0.0500)  time: 0.6743  data: 0.0009  max mem: 57114
Epoch: [0]  [140/156]  eta: 0:00:19  lr: 0.000359  min_lr: 0.000359  loss: 6.6845 (6.8383)  weight_decay: 0.0500 (0.0500)  time: 0.6695  data: 0.0007  max mem: 57114
Epoch: [0]  [150/156]  eta: 0:00:06  lr: 0.000385  min_lr: 0.000385  loss: 6.6398 (6.8251)  weight_decay: 0.0500 (0.0500)  time: 0.6648  data: 0.0001  max mem: 57114
Epoch: [0]  [155/156]  eta: 0:00:01  lr: 0.000398  min_lr: 0.000398  loss: 6.6470 (6.8193)  weight_decay: 0.0500 (0.0500)  time: 0.6691  data: 0.0001  max mem: 57114
Epoch: [0] Total time: 0:02:58 (1.1420 s / it)
Averaged stats: lr: 0.000398  min_lr: 0.000398  loss: 6.6470 (6.8177)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:27  loss: 5.8772 (5.8772)  acc1: 2.6042 (2.6042)  acc5: 12.5000 (12.5000)  time: 5.4747  data: 1.8190  max mem: 57114
Test:  [4/5]  eta: 0:00:01  loss: 5.9074 (5.7426)  acc1: 4.6875 (4.0921)  acc5: 12.5000 (12.7877)  time: 1.3068  data: 0.3640  max mem: 57114
Test: Total time: 0:00:06 (1.3182 s / it)
* Acc@1 3.587 Acc@5 10.746 loss 5.833
Accuracy of the model on the 50000 test images: 3.6%
Max accuracy: 3.59%
Test:  [0/5]  eta: 0:00:12  loss: 6.9078 (6.9078)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 2.4422  data: 1.8873  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9078 (6.9079)  acc1: 0.0000 (0.1279)  acc5: 0.5208 (0.5115)  time: 0.6391  data: 0.3776  max mem: 57114
Test: Total time: 0:00:03 (0.6505 s / it)
* Acc@1 0.104 Acc@5 0.523 loss 6.908
Accuracy of the model EMA on 50000 test images: 0.1%
Max EMA accuracy: 0.10%
Epoch: [1]  [  0/156]  eta: 0:26:53  lr: 0.000400  min_lr: 0.000400  loss: 6.5778 (6.5778)  weight_decay: 0.0500 (0.0500)  time: 10.3454  data: 3.3689  max mem: 57114
Epoch: [1]  [ 10/156]  eta: 0:05:11  lr: 0.000426  min_lr: 0.000426  loss: 6.6556 (6.6238)  weight_decay: 0.0500 (0.0500)  time: 2.1363  data: 0.3067  max mem: 57114
Epoch: [1]  [ 20/156]  eta: 0:03:57  lr: 0.000451  min_lr: 0.000451  loss: 6.6123 (6.5961)  weight_decay: 0.0500 (0.0500)  time: 1.3144  data: 0.0004  max mem: 57114
Epoch: [1]  [ 30/156]  eta: 0:03:22  lr: 0.000477  min_lr: 0.000477  loss: 6.5493 (6.5990)  weight_decay: 0.0500 (0.0500)  time: 1.3129  data: 0.0003  max mem: 57114
Epoch: [1]  [ 40/156]  eta: 0:02:47  lr: 0.000503  min_lr: 0.000503  loss: 6.5787 (6.5863)  weight_decay: 0.0500 (0.0500)  time: 1.1293  data: 0.0004  max mem: 57114
Epoch: [1]  [ 50/156]  eta: 0:02:17  lr: 0.000528  min_lr: 0.000528  loss: 6.4575 (6.5580)  weight_decay: 0.0500 (0.0500)  time: 0.8211  data: 0.0004  max mem: 57114
Epoch: [1]  [ 60/156]  eta: 0:01:55  lr: 0.000554  min_lr: 0.000554  loss: 6.4459 (6.5423)  weight_decay: 0.0500 (0.0500)  time: 0.6933  data: 0.0003  max mem: 57114
Epoch: [1]  [ 70/156]  eta: 0:01:37  lr: 0.000580  min_lr: 0.000580  loss: 6.4919 (6.5374)  weight_decay: 0.0500 (0.0500)  time: 0.6980  data: 0.0005  max mem: 57114
Epoch: [1]  [ 80/156]  eta: 0:01:21  lr: 0.000605  min_lr: 0.000605  loss: 6.4891 (6.5235)  weight_decay: 0.0500 (0.0500)  time: 0.6913  data: 0.0005  max mem: 57114
Epoch: [1]  [ 90/156]  eta: 0:01:07  lr: 0.000631  min_lr: 0.000631  loss: 6.4321 (6.5096)  weight_decay: 0.0500 (0.0500)  time: 0.6753  data: 0.0003  max mem: 57114
Epoch: [1]  [100/156]  eta: 0:00:55  lr: 0.000657  min_lr: 0.000657  loss: 6.3939 (6.4890)  weight_decay: 0.0500 (0.0500)  time: 0.6728  data: 0.0003  max mem: 57114
Epoch: [1]  [110/156]  eta: 0:00:44  lr: 0.000682  min_lr: 0.000682  loss: 6.2998 (6.4741)  weight_decay: 0.0500 (0.0500)  time: 0.6755  data: 0.0004  max mem: 57114
Epoch: [1]  [120/156]  eta: 0:00:33  lr: 0.000708  min_lr: 0.000708  loss: 6.4889 (6.4771)  weight_decay: 0.0500 (0.0500)  time: 0.6786  data: 0.0003  max mem: 57114
Epoch: [1]  [130/156]  eta: 0:00:23  lr: 0.000734  min_lr: 0.000734  loss: 6.4917 (6.4677)  weight_decay: 0.0500 (0.0500)  time: 0.6735  data: 0.0004  max mem: 57114
Epoch: [1]  [140/156]  eta: 0:00:14  lr: 0.000759  min_lr: 0.000759  loss: 6.2513 (6.4513)  weight_decay: 0.0500 (0.0500)  time: 0.6699  data: 0.0003  max mem: 57114
Epoch: [1]  [150/156]  eta: 0:00:05  lr: 0.000785  min_lr: 0.000785  loss: 6.2406 (6.4399)  weight_decay: 0.0500 (0.0500)  time: 0.6670  data: 0.0001  max mem: 57114
Epoch: [1]  [155/156]  eta: 0:00:00  lr: 0.000798  min_lr: 0.000798  loss: 6.2406 (6.4370)  weight_decay: 0.0500 (0.0500)  time: 0.6667  data: 0.0001  max mem: 57114
Epoch: [1] Total time: 0:02:17 (0.8824 s / it)
Averaged stats: lr: 0.000798  min_lr: 0.000798  loss: 6.2406 (6.4632)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 4.8223 (4.8223)  acc1: 9.8958 (9.8958)  acc5: 31.2500 (31.2500)  time: 2.0794  data: 1.8237  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 4.9562 (4.8510)  acc1: 10.9375 (11.2532)  acc5: 26.5625 (26.5985)  time: 0.5665  data: 0.3648  max mem: 57114
Test: Total time: 0:00:02 (0.5876 s / it)
* Acc@1 9.883 Acc@5 24.692 loss 4.918
Accuracy of the model on the 50000 test images: 9.9%
Max accuracy: 9.88%
Test:  [0/5]  eta: 0:00:10  loss: 6.9077 (6.9077)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.5208)  time: 2.0879  data: 1.8443  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9079 (6.9079)  acc1: 0.0000 (0.1279)  acc5: 0.5208 (0.6394)  time: 0.5682  data: 0.3690  max mem: 57114
Test: Total time: 0:00:02 (0.5821 s / it)
* Acc@1 0.094 Acc@5 0.478 loss 6.908
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [2]  [  0/156]  eta: 0:22:07  lr: 0.000800  min_lr: 0.000800  loss: 6.5611 (6.5611)  weight_decay: 0.0500 (0.0500)  time: 8.5087  data: 4.1362  max mem: 57114
Epoch: [2]  [ 10/156]  eta: 0:03:23  lr: 0.000826  min_lr: 0.000826  loss: 6.5089 (6.4570)  weight_decay: 0.0500 (0.0500)  time: 1.3905  data: 0.3763  max mem: 57114
Epoch: [2]  [ 20/156]  eta: 0:02:24  lr: 0.000852  min_lr: 0.000852  loss: 6.3899 (6.3559)  weight_decay: 0.0500 (0.0500)  time: 0.6885  data: 0.0003  max mem: 57114
Epoch: [2]  [ 30/156]  eta: 0:01:58  lr: 0.000877  min_lr: 0.000877  loss: 6.3528 (6.3473)  weight_decay: 0.0500 (0.0500)  time: 0.6876  data: 0.0003  max mem: 57114
Epoch: [2]  [ 40/156]  eta: 0:01:42  lr: 0.000903  min_lr: 0.000903  loss: 6.2996 (6.3262)  weight_decay: 0.0500 (0.0500)  time: 0.6899  data: 0.0003  max mem: 57114
Epoch: [2]  [ 50/156]  eta: 0:01:29  lr: 0.000929  min_lr: 0.000929  loss: 6.2132 (6.2965)  weight_decay: 0.0500 (0.0500)  time: 0.6973  data: 0.0004  max mem: 57114
Epoch: [2]  [ 60/156]  eta: 0:01:18  lr: 0.000954  min_lr: 0.000954  loss: 6.2350 (6.2734)  weight_decay: 0.0500 (0.0500)  time: 0.6812  data: 0.0004  max mem: 57114
Epoch: [2]  [ 70/156]  eta: 0:01:08  lr: 0.000980  min_lr: 0.000980  loss: 6.2350 (6.2559)  weight_decay: 0.0500 (0.0500)  time: 0.6729  data: 0.0004  max mem: 57114
Epoch: [2]  [ 80/156]  eta: 0:00:59  lr: 0.001005  min_lr: 0.001005  loss: 6.1314 (6.2363)  weight_decay: 0.0500 (0.0500)  time: 0.6757  data: 0.0004  max mem: 57114
Epoch: [2]  [ 90/156]  eta: 0:00:50  lr: 0.001031  min_lr: 0.001031  loss: 6.2395 (6.2381)  weight_decay: 0.0500 (0.0500)  time: 0.6736  data: 0.0004  max mem: 57114
Epoch: [2]  [100/156]  eta: 0:00:42  lr: 0.001057  min_lr: 0.001057  loss: 6.2119 (6.2300)  weight_decay: 0.0500 (0.0500)  time: 0.6773  data: 0.0004  max mem: 57114
Epoch: [2]  [110/156]  eta: 0:00:34  lr: 0.001082  min_lr: 0.001082  loss: 6.0858 (6.2110)  weight_decay: 0.0500 (0.0500)  time: 0.6830  data: 0.0004  max mem: 57114
Epoch: [2]  [120/156]  eta: 0:00:26  lr: 0.001108  min_lr: 0.001108  loss: 6.2229 (6.2165)  weight_decay: 0.0500 (0.0500)  time: 0.6780  data: 0.0004  max mem: 57114
Epoch: [2]  [130/156]  eta: 0:00:19  lr: 0.001134  min_lr: 0.001134  loss: 6.2307 (6.2152)  weight_decay: 0.0500 (0.0500)  time: 0.6729  data: 0.0009  max mem: 57114
Epoch: [2]  [140/156]  eta: 0:00:11  lr: 0.001159  min_lr: 0.001159  loss: 6.0501 (6.1945)  weight_decay: 0.0500 (0.0500)  time: 0.6713  data: 0.0007  max mem: 57114
Epoch: [2]  [150/156]  eta: 0:00:04  lr: 0.001185  min_lr: 0.001185  loss: 5.8137 (6.1751)  weight_decay: 0.0500 (0.0500)  time: 0.6685  data: 0.0001  max mem: 57114
Epoch: [2]  [155/156]  eta: 0:00:00  lr: 0.001198  min_lr: 0.001198  loss: 5.8137 (6.1651)  weight_decay: 0.0500 (0.0500)  time: 0.6667  data: 0.0001  max mem: 57114
Epoch: [2] Total time: 0:01:54 (0.7312 s / it)
Averaged stats: lr: 0.001198  min_lr: 0.001198  loss: 5.8137 (6.1638)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 4.4752 (4.4752)  acc1: 14.5833 (14.5833)  acc5: 37.5000 (37.5000)  time: 2.1620  data: 1.9065  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 4.5564 (4.4272)  acc1: 17.1875 (15.6010)  acc5: 35.4167 (33.8875)  time: 0.5832  data: 0.3814  max mem: 57114
Test: Total time: 0:00:03 (0.6065 s / it)
* Acc@1 13.621 Acc@5 30.882 loss 4.501
Accuracy of the model on the 50000 test images: 13.6%
Max accuracy: 13.62%
Test:  [0/5]  eta: 0:00:10  loss: 6.9080 (6.9080)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 2.0615  data: 1.8180  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9080 (6.9080)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.2558)  time: 0.5628  data: 0.3637  max mem: 57114
Test: Total time: 0:00:02 (0.5757 s / it)
* Acc@1 0.078 Acc@5 0.512 loss 6.908
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [3]  [  0/156]  eta: 0:21:25  lr: 0.001200  min_lr: 0.001200  loss: 5.1468 (5.1468)  weight_decay: 0.0500 (0.0500)  time: 8.2406  data: 4.8250  max mem: 57114
Epoch: [3]  [ 10/156]  eta: 0:03:23  lr: 0.001226  min_lr: 0.001226  loss: 5.9493 (5.8822)  weight_decay: 0.0500 (0.0500)  time: 1.3938  data: 0.4389  max mem: 57114
Epoch: [3]  [ 20/156]  eta: 0:02:22  lr: 0.001252  min_lr: 0.001252  loss: 6.0072 (5.9925)  weight_decay: 0.0500 (0.0500)  time: 0.6916  data: 0.0003  max mem: 57114
Epoch: [3]  [ 30/156]  eta: 0:01:58  lr: 0.001277  min_lr: 0.001277  loss: 6.0051 (5.9736)  weight_decay: 0.0500 (0.0500)  time: 0.6865  data: 0.0004  max mem: 57114
Epoch: [3]  [ 40/156]  eta: 0:01:41  lr: 0.001303  min_lr: 0.001303  loss: 6.0435 (5.9921)  weight_decay: 0.0500 (0.0500)  time: 0.6860  data: 0.0004  max mem: 57114
Epoch: [3]  [ 50/156]  eta: 0:01:28  lr: 0.001329  min_lr: 0.001329  loss: 6.1303 (6.0105)  weight_decay: 0.0500 (0.0500)  time: 0.6790  data: 0.0003  max mem: 57114
Epoch: [3]  [ 60/156]  eta: 0:01:17  lr: 0.001354  min_lr: 0.001354  loss: 6.0615 (5.9625)  weight_decay: 0.0500 (0.0500)  time: 0.6833  data: 0.0004  max mem: 57114
Epoch: [3]  [ 70/156]  eta: 0:01:08  lr: 0.001380  min_lr: 0.001380  loss: 5.9806 (5.9518)  weight_decay: 0.0500 (0.0500)  time: 0.6775  data: 0.0004  max mem: 57114
Epoch: [3]  [ 80/156]  eta: 0:00:59  lr: 0.001406  min_lr: 0.001406  loss: 6.0051 (5.9461)  weight_decay: 0.0500 (0.0500)  time: 0.6758  data: 0.0004  max mem: 57114
Epoch: [3]  [ 90/156]  eta: 0:00:50  lr: 0.001431  min_lr: 0.001431  loss: 6.0051 (5.9506)  weight_decay: 0.0500 (0.0500)  time: 0.6840  data: 0.0004  max mem: 57114
Epoch: [3]  [100/156]  eta: 0:00:42  lr: 0.001457  min_lr: 0.001457  loss: 6.1309 (5.9557)  weight_decay: 0.0500 (0.0500)  time: 0.6834  data: 0.0004  max mem: 57114
Epoch: [3]  [110/156]  eta: 0:00:34  lr: 0.001483  min_lr: 0.001483  loss: 5.9996 (5.9503)  weight_decay: 0.0500 (0.0500)  time: 0.6792  data: 0.0004  max mem: 57114
Epoch: [3]  [120/156]  eta: 0:00:26  lr: 0.001508  min_lr: 0.001508  loss: 5.9397 (5.9434)  weight_decay: 0.0500 (0.0500)  time: 0.6816  data: 0.0004  max mem: 57114
Epoch: [3]  [130/156]  eta: 0:00:19  lr: 0.001534  min_lr: 0.001534  loss: 5.9180 (5.9328)  weight_decay: 0.0500 (0.0500)  time: 0.6792  data: 0.0008  max mem: 57114
Epoch: [3]  [140/156]  eta: 0:00:11  lr: 0.001559  min_lr: 0.001559  loss: 5.7780 (5.9187)  weight_decay: 0.0500 (0.0500)  time: 0.6721  data: 0.0007  max mem: 57114
Epoch: [3]  [150/156]  eta: 0:00:04  lr: 0.001585  min_lr: 0.001585  loss: 5.8363 (5.9122)  weight_decay: 0.0500 (0.0500)  time: 0.6667  data: 0.0001  max mem: 57114
Epoch: [3]  [155/156]  eta: 0:00:00  lr: 0.001598  min_lr: 0.001598  loss: 5.8973 (5.9189)  weight_decay: 0.0500 (0.0500)  time: 0.6672  data: 0.0001  max mem: 57114
Epoch: [3] Total time: 0:01:54 (0.7311 s / it)
Averaged stats: lr: 0.001598  min_lr: 0.001598  loss: 5.8973 (5.8915)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 3.2672 (3.2672)  acc1: 28.6458 (28.6458)  acc5: 60.4167 (60.4167)  time: 2.2613  data: 1.9999  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 3.7789 (3.4579)  acc1: 27.0833 (27.4936)  acc5: 48.4375 (49.7442)  time: 0.6031  data: 0.4001  max mem: 57114
Test: Total time: 0:00:03 (0.6271 s / it)
* Acc@1 25.659 Acc@5 49.043 loss 3.565
Accuracy of the model on the 50000 test images: 25.7%
Max accuracy: 25.66%
Test:  [0/5]  eta: 0:00:10  loss: 6.9082 (6.9082)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 2.1406  data: 1.8973  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9082 (6.9081)  acc1: 0.0000 (0.1279)  acc5: 0.5208 (0.6394)  time: 0.5787  data: 0.3796  max mem: 57114
Test: Total time: 0:00:02 (0.5905 s / it)
* Acc@1 0.110 Acc@5 0.512 loss 6.908
Accuracy of the model EMA on 50000 test images: 0.1%
Max EMA accuracy: 0.11%
Epoch: [4]  [  0/156]  eta: 0:10:34  lr: 0.001601  min_lr: 0.001601  loss: 5.9406 (5.9406)  weight_decay: 0.0500 (0.0500)  time: 4.0668  data: 3.4213  max mem: 57114
Epoch: [4]  [ 10/156]  eta: 0:02:29  lr: 0.001626  min_lr: 0.001626  loss: 5.5890 (5.7499)  weight_decay: 0.0500 (0.0500)  time: 1.0266  data: 0.3113  max mem: 57114
Epoch: [4]  [ 20/156]  eta: 0:01:58  lr: 0.001652  min_lr: 0.001652  loss: 5.5055 (5.6140)  weight_decay: 0.0500 (0.0500)  time: 0.7112  data: 0.0003  max mem: 57114
Epoch: [4]  [ 30/156]  eta: 0:01:42  lr: 0.001677  min_lr: 0.001677  loss: 5.5502 (5.6366)  weight_decay: 0.0500 (0.0500)  time: 0.6976  data: 0.0003  max mem: 57114
Epoch: [4]  [ 40/156]  eta: 0:01:30  lr: 0.001703  min_lr: 0.001703  loss: 5.5502 (5.6043)  weight_decay: 0.0500 (0.0500)  time: 0.6862  data: 0.0003  max mem: 57114
Epoch: [4]  [ 50/156]  eta: 0:01:20  lr: 0.001729  min_lr: 0.001729  loss: 5.8068 (5.6708)  weight_decay: 0.0500 (0.0500)  time: 0.6759  data: 0.0003  max mem: 57114
Epoch: [4]  [ 60/156]  eta: 0:01:11  lr: 0.001754  min_lr: 0.001754  loss: 5.8324 (5.6714)  weight_decay: 0.0500 (0.0500)  time: 0.6740  data: 0.0004  max mem: 57114
Epoch: [4]  [ 70/156]  eta: 0:01:03  lr: 0.001780  min_lr: 0.001780  loss: 5.5681 (5.6467)  weight_decay: 0.0500 (0.0500)  time: 0.6770  data: 0.0004  max mem: 57114
Epoch: [4]  [ 80/156]  eta: 0:00:55  lr: 0.001806  min_lr: 0.001806  loss: 5.3869 (5.6387)  weight_decay: 0.0500 (0.0500)  time: 0.6819  data: 0.0004  max mem: 57114
Epoch: [4]  [ 90/156]  eta: 0:00:47  lr: 0.001831  min_lr: 0.001831  loss: 5.3869 (5.6104)  weight_decay: 0.0500 (0.0500)  time: 0.6868  data: 0.0004  max mem: 57114
Epoch: [4]  [100/156]  eta: 0:00:40  lr: 0.001857  min_lr: 0.001857  loss: 5.4439 (5.6037)  weight_decay: 0.0500 (0.0500)  time: 0.6836  data: 0.0004  max mem: 57114
Epoch: [4]  [110/156]  eta: 0:00:32  lr: 0.001883  min_lr: 0.001883  loss: 5.7090 (5.6176)  weight_decay: 0.0500 (0.0500)  time: 0.6778  data: 0.0004  max mem: 57114
Epoch: [4]  [120/156]  eta: 0:00:25  lr: 0.001908  min_lr: 0.001908  loss: 5.9107 (5.6281)  weight_decay: 0.0500 (0.0500)  time: 0.6812  data: 0.0003  max mem: 57114
Epoch: [4]  [130/156]  eta: 0:00:18  lr: 0.001934  min_lr: 0.001934  loss: 5.7810 (5.6136)  weight_decay: 0.0500 (0.0500)  time: 0.6779  data: 0.0008  max mem: 57114
Epoch: [4]  [140/156]  eta: 0:00:11  lr: 0.001960  min_lr: 0.001960  loss: 5.7441 (5.6202)  weight_decay: 0.0500 (0.0500)  time: 0.6714  data: 0.0007  max mem: 57114
Epoch: [4]  [150/156]  eta: 0:00:04  lr: 0.001985  min_lr: 0.001985  loss: 5.7324 (5.6209)  weight_decay: 0.0500 (0.0500)  time: 0.6713  data: 0.0001  max mem: 57114
Epoch: [4]  [155/156]  eta: 0:00:00  lr: 0.001998  min_lr: 0.001998  loss: 5.7291 (5.6157)  weight_decay: 0.0500 (0.0500)  time: 0.6699  data: 0.0001  max mem: 57114
Epoch: [4] Total time: 0:01:50 (0.7068 s / it)
Averaged stats: lr: 0.001998  min_lr: 0.001998  loss: 5.7291 (5.6442)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 2.8919 (2.8919)  acc1: 33.8542 (33.8542)  acc5: 63.0208 (63.0208)  time: 2.2529  data: 1.9971  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 3.3534 (3.0626)  acc1: 29.6875 (30.6905)  acc5: 57.2917 (57.0333)  time: 0.6013  data: 0.3995  max mem: 57114
Test: Total time: 0:00:03 (0.6224 s / it)
* Acc@1 32.021 Acc@5 56.811 loss 3.097
Accuracy of the model on the 50000 test images: 32.0%
Max accuracy: 32.02%
Test:  [0/5]  eta: 0:00:10  loss: 6.9083 (6.9083)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.5208)  time: 2.0619  data: 1.8185  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9083 (6.9083)  acc1: 0.0000 (0.1279)  acc5: 0.5208 (0.3836)  time: 0.5629  data: 0.3638  max mem: 57114
Test: Total time: 0:00:03 (0.6032 s / it)
* Acc@1 0.110 Acc@5 0.514 loss 6.908
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [5]  [  0/156]  eta: 0:20:02  lr: 0.002001  min_lr: 0.002001  loss: 5.9639 (5.9639)  weight_decay: 0.0500 (0.0500)  time: 7.7084  data: 4.3664  max mem: 57114
Epoch: [5]  [ 10/156]  eta: 0:03:12  lr: 0.002026  min_lr: 0.002026  loss: 5.6525 (5.5266)  weight_decay: 0.0500 (0.0500)  time: 1.3191  data: 0.3972  max mem: 57114
Epoch: [5]  [ 20/156]  eta: 0:02:19  lr: 0.002052  min_lr: 0.002052  loss: 5.6525 (5.5212)  weight_decay: 0.0500 (0.0500)  time: 0.6947  data: 0.0003  max mem: 57114
Epoch: [5]  [ 30/156]  eta: 0:01:55  lr: 0.002078  min_lr: 0.002078  loss: 5.1874 (5.3290)  weight_decay: 0.0500 (0.0500)  time: 0.6980  data: 0.0003  max mem: 57114
Epoch: [5]  [ 40/156]  eta: 0:01:39  lr: 0.002103  min_lr: 0.002103  loss: 5.1355 (5.4068)  weight_decay: 0.0500 (0.0500)  time: 0.6781  data: 0.0003  max mem: 57114
Epoch: [5]  [ 50/156]  eta: 0:01:27  lr: 0.002129  min_lr: 0.002129  loss: 5.6139 (5.3739)  weight_decay: 0.0500 (0.0500)  time: 0.6833  data: 0.0004  max mem: 57114
Epoch: [5]  [ 60/156]  eta: 0:01:17  lr: 0.002155  min_lr: 0.002155  loss: 5.2794 (5.3784)  weight_decay: 0.0500 (0.0500)  time: 0.6953  data: 0.0004  max mem: 57114
Epoch: [5]  [ 70/156]  eta: 0:01:07  lr: 0.002180  min_lr: 0.002180  loss: 5.5758 (5.3820)  weight_decay: 0.0500 (0.0500)  time: 0.6951  data: 0.0004  max mem: 57114
Epoch: [5]  [ 80/156]  eta: 0:00:58  lr: 0.002206  min_lr: 0.002206  loss: 5.6194 (5.3913)  weight_decay: 0.0500 (0.0500)  time: 0.6880  data: 0.0004  max mem: 57114
Epoch: [5]  [ 90/156]  eta: 0:00:50  lr: 0.002231  min_lr: 0.002231  loss: 5.5139 (5.3997)  weight_decay: 0.0500 (0.0500)  time: 0.6875  data: 0.0004  max mem: 57114
Epoch: [5]  [100/156]  eta: 0:00:42  lr: 0.002257  min_lr: 0.002257  loss: 5.3715 (5.3808)  weight_decay: 0.0500 (0.0500)  time: 0.6919  data: 0.0004  max mem: 57114
Epoch: [5]  [110/156]  eta: 0:00:34  lr: 0.002283  min_lr: 0.002283  loss: 5.3134 (5.3762)  weight_decay: 0.0500 (0.0500)  time: 0.6839  data: 0.0004  max mem: 57114
Epoch: [5]  [120/156]  eta: 0:00:26  lr: 0.002308  min_lr: 0.002308  loss: 5.4709 (5.3777)  weight_decay: 0.0500 (0.0500)  time: 0.6787  data: 0.0004  max mem: 57114
Epoch: [5]  [130/156]  eta: 0:00:19  lr: 0.002334  min_lr: 0.002334  loss: 5.4709 (5.3749)  weight_decay: 0.0500 (0.0500)  time: 0.6777  data: 0.0009  max mem: 57114
Epoch: [5]  [140/156]  eta: 0:00:11  lr: 0.002360  min_lr: 0.002360  loss: 5.6401 (5.3770)  weight_decay: 0.0500 (0.0500)  time: 0.6742  data: 0.0007  max mem: 57114
Epoch: [5]  [150/156]  eta: 0:00:04  lr: 0.002385  min_lr: 0.002385  loss: 5.6315 (5.3940)  weight_decay: 0.0500 (0.0500)  time: 0.6757  data: 0.0001  max mem: 57114
Epoch: [5]  [155/156]  eta: 0:00:00  lr: 0.002398  min_lr: 0.002398  loss: 5.4568 (5.3803)  weight_decay: 0.0500 (0.0500)  time: 0.6732  data: 0.0001  max mem: 57114
Epoch: [5] Total time: 0:01:54 (0.7322 s / it)
Averaged stats: lr: 0.002398  min_lr: 0.002398  loss: 5.4568 (5.4283)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:09  loss: 2.4418 (2.4418)  acc1: 44.7917 (44.7917)  acc5: 71.8750 (71.8750)  time: 1.9912  data: 1.7351  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 2.8786 (2.7478)  acc1: 39.0625 (38.3632)  acc5: 65.1042 (62.9156)  time: 0.5489  data: 0.3471  max mem: 57114
Test: Total time: 0:00:02 (0.5726 s / it)
* Acc@1 38.615 Acc@5 63.949 loss 2.790
Accuracy of the model on the 50000 test images: 38.6%
Max accuracy: 38.61%
Test:  [0/5]  eta: 0:00:09  loss: 6.9084 (6.9084)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 1.9533  data: 1.7097  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9084 (6.9083)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.5413  data: 0.3420  max mem: 57114
Test: Total time: 0:00:02 (0.5526 s / it)
* Acc@1 0.108 Acc@5 0.518 loss 6.909
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [6]  [  0/156]  eta: 0:18:58  lr: 0.002401  min_lr: 0.002401  loss: 5.8492 (5.8492)  weight_decay: 0.0500 (0.0500)  time: 7.2979  data: 4.1359  max mem: 57114
Epoch: [6]  [ 10/156]  eta: 0:03:12  lr: 0.002426  min_lr: 0.002426  loss: 5.1618 (5.0687)  weight_decay: 0.0500 (0.0500)  time: 1.3190  data: 0.3762  max mem: 57114
Epoch: [6]  [ 20/156]  eta: 0:02:18  lr: 0.002452  min_lr: 0.002452  loss: 5.1542 (5.0744)  weight_decay: 0.0500 (0.0500)  time: 0.7023  data: 0.0003  max mem: 57114
Epoch: [6]  [ 30/156]  eta: 0:01:54  lr: 0.002478  min_lr: 0.002478  loss: 5.3058 (5.1887)  weight_decay: 0.0500 (0.0500)  time: 0.6788  data: 0.0003  max mem: 57114
Epoch: [6]  [ 40/156]  eta: 0:01:39  lr: 0.002503  min_lr: 0.002503  loss: 5.5512 (5.2456)  weight_decay: 0.0500 (0.0500)  time: 0.6843  data: 0.0004  max mem: 57114
Epoch: [6]  [ 50/156]  eta: 0:01:26  lr: 0.002529  min_lr: 0.002529  loss: 5.5493 (5.2467)  weight_decay: 0.0500 (0.0500)  time: 0.6875  data: 0.0004  max mem: 57114
Epoch: [6]  [ 60/156]  eta: 0:01:17  lr: 0.002555  min_lr: 0.002555  loss: 5.4700 (5.2303)  weight_decay: 0.0500 (0.0500)  time: 0.7004  data: 0.0004  max mem: 57114
Epoch: [6]  [ 70/156]  eta: 0:01:07  lr: 0.002580  min_lr: 0.002580  loss: 5.2491 (5.2380)  weight_decay: 0.0500 (0.0500)  time: 0.7143  data: 0.0004  max mem: 57114
Epoch: [6]  [ 80/156]  eta: 0:00:58  lr: 0.002606  min_lr: 0.002606  loss: 5.5533 (5.2621)  weight_decay: 0.0500 (0.0500)  time: 0.6914  data: 0.0004  max mem: 57114
Epoch: [6]  [ 90/156]  eta: 0:00:50  lr: 0.002632  min_lr: 0.002632  loss: 5.4331 (5.2234)  weight_decay: 0.0500 (0.0500)  time: 0.6774  data: 0.0003  max mem: 57114
Epoch: [6]  [100/156]  eta: 0:00:42  lr: 0.002657  min_lr: 0.002657  loss: 5.1729 (5.2242)  weight_decay: 0.0500 (0.0500)  time: 0.6826  data: 0.0004  max mem: 57114
Epoch: [6]  [110/156]  eta: 0:00:34  lr: 0.002683  min_lr: 0.002683  loss: 5.4187 (5.2287)  weight_decay: 0.0500 (0.0500)  time: 0.6791  data: 0.0004  max mem: 57114
Epoch: [6]  [120/156]  eta: 0:00:26  lr: 0.002709  min_lr: 0.002709  loss: 5.5138 (5.2394)  weight_decay: 0.0500 (0.0500)  time: 0.6846  data: 0.0004  max mem: 57114
Epoch: [6]  [130/156]  eta: 0:00:19  lr: 0.002734  min_lr: 0.002734  loss: 5.4750 (5.2390)  weight_decay: 0.0500 (0.0500)  time: 0.6867  data: 0.0009  max mem: 57114
Epoch: [6]  [140/156]  eta: 0:00:11  lr: 0.002760  min_lr: 0.002760  loss: 5.1393 (5.2231)  weight_decay: 0.0500 (0.0500)  time: 0.6761  data: 0.0007  max mem: 57114
Epoch: [6]  [150/156]  eta: 0:00:04  lr: 0.002786  min_lr: 0.002786  loss: 5.0780 (5.2167)  weight_decay: 0.0500 (0.0500)  time: 0.6727  data: 0.0001  max mem: 57114
Epoch: [6]  [155/156]  eta: 0:00:00  lr: 0.002798  min_lr: 0.002798  loss: 5.1271 (5.2078)  weight_decay: 0.0500 (0.0500)  time: 0.6721  data: 0.0001  max mem: 57114
Epoch: [6] Total time: 0:01:54 (0.7311 s / it)
Averaged stats: lr: 0.002798  min_lr: 0.002798  loss: 5.1271 (5.2297)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 2.2497 (2.2497)  acc1: 49.4792 (49.4792)  acc5: 78.6458 (78.6458)  time: 2.0373  data: 1.7817  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 2.5918 (2.4409)  acc1: 45.3125 (44.3734)  acc5: 72.9167 (70.3325)  time: 0.5581  data: 0.3564  max mem: 57114
Test: Total time: 0:00:02 (0.5793 s / it)
* Acc@1 44.721 Acc@5 70.482 loss 2.504
Accuracy of the model on the 50000 test images: 44.7%
Max accuracy: 44.72%
Test:  [0/5]  eta: 0:00:10  loss: 6.9086 (6.9086)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 2.0737  data: 1.8304  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9086 (6.9088)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.3836)  time: 0.5653  data: 0.3662  max mem: 57114
Test: Total time: 0:00:02 (0.5752 s / it)
* Acc@1 0.104 Acc@5 0.527 loss 6.909
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [7]  [  0/156]  eta: 0:17:21  lr: 0.002801  min_lr: 0.002801  loss: 5.3940 (5.3940)  weight_decay: 0.0500 (0.0500)  time: 6.6789  data: 4.6488  max mem: 57114
Epoch: [7]  [ 10/156]  eta: 0:03:00  lr: 0.002827  min_lr: 0.002827  loss: 5.3940 (5.0810)  weight_decay: 0.0500 (0.0500)  time: 1.2373  data: 0.4228  max mem: 57114
Epoch: [7]  [ 20/156]  eta: 0:02:12  lr: 0.002852  min_lr: 0.002852  loss: 5.1281 (5.1146)  weight_decay: 0.0500 (0.0500)  time: 0.6906  data: 0.0003  max mem: 57114
Epoch: [7]  [ 30/156]  eta: 0:01:52  lr: 0.002878  min_lr: 0.002878  loss: 5.3059 (5.1869)  weight_decay: 0.0500 (0.0500)  time: 0.6988  data: 0.0003  max mem: 57114
Epoch: [7]  [ 40/156]  eta: 0:01:37  lr: 0.002903  min_lr: 0.002903  loss: 5.4876 (5.2008)  weight_decay: 0.0500 (0.0500)  time: 0.7010  data: 0.0003  max mem: 57114
Epoch: [7]  [ 50/156]  eta: 0:01:26  lr: 0.002929  min_lr: 0.002929  loss: 5.4197 (5.2047)  weight_decay: 0.0500 (0.0500)  time: 0.6916  data: 0.0004  max mem: 57114
Epoch: [7]  [ 60/156]  eta: 0:01:16  lr: 0.002955  min_lr: 0.002955  loss: 5.3034 (5.1837)  weight_decay: 0.0500 (0.0500)  time: 0.6947  data: 0.0004  max mem: 57114
Epoch: [7]  [ 70/156]  eta: 0:01:07  lr: 0.002980  min_lr: 0.002980  loss: 4.9472 (5.1409)  weight_decay: 0.0500 (0.0500)  time: 0.6963  data: 0.0004  max mem: 57114
Epoch: [7]  [ 80/156]  eta: 0:00:58  lr: 0.003006  min_lr: 0.003006  loss: 5.0532 (5.1394)  weight_decay: 0.0500 (0.0500)  time: 0.6892  data: 0.0004  max mem: 57114
Epoch: [7]  [ 90/156]  eta: 0:00:50  lr: 0.003032  min_lr: 0.003032  loss: 5.4746 (5.1880)  weight_decay: 0.0500 (0.0500)  time: 0.6829  data: 0.0004  max mem: 57114
Epoch: [7]  [100/156]  eta: 0:00:42  lr: 0.003057  min_lr: 0.003057  loss: 5.4798 (5.1866)  weight_decay: 0.0500 (0.0500)  time: 0.6801  data: 0.0004  max mem: 57114
Epoch: [7]  [110/156]  eta: 0:00:34  lr: 0.003083  min_lr: 0.003083  loss: 5.2562 (5.1851)  weight_decay: 0.0500 (0.0500)  time: 0.6872  data: 0.0004  max mem: 57114
Epoch: [7]  [120/156]  eta: 0:00:26  lr: 0.003109  min_lr: 0.003109  loss: 5.2149 (5.1792)  weight_decay: 0.0500 (0.0500)  time: 0.6844  data: 0.0003  max mem: 57114
Epoch: [7]  [130/156]  eta: 0:00:19  lr: 0.003134  min_lr: 0.003134  loss: 5.0034 (5.1518)  weight_decay: 0.0500 (0.0500)  time: 0.6732  data: 0.0008  max mem: 57114
Epoch: [7]  [140/156]  eta: 0:00:11  lr: 0.003160  min_lr: 0.003160  loss: 4.9023 (5.1306)  weight_decay: 0.0500 (0.0500)  time: 0.6687  data: 0.0007  max mem: 57114
Epoch: [7]  [150/156]  eta: 0:00:04  lr: 0.003186  min_lr: 0.003186  loss: 5.1397 (5.1284)  weight_decay: 0.0500 (0.0500)  time: 0.6663  data: 0.0001  max mem: 57114
Epoch: [7]  [155/156]  eta: 0:00:00  lr: 0.003198  min_lr: 0.003198  loss: 5.1397 (5.1186)  weight_decay: 0.0500 (0.0500)  time: 0.6676  data: 0.0001  max mem: 57114
Epoch: [7] Total time: 0:01:53 (0.7261 s / it)
Averaged stats: lr: 0.003198  min_lr: 0.003198  loss: 5.1397 (5.0694)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.9626 (1.9626)  acc1: 55.7292 (55.7292)  acc5: 80.7292 (80.7292)  time: 2.1840  data: 1.9283  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 2.4342 (2.2472)  acc1: 48.4375 (48.5934)  acc5: 76.0417 (74.0409)  time: 0.5875  data: 0.3857  max mem: 57114
Test: Total time: 0:00:03 (0.6103 s / it)
* Acc@1 48.298 Acc@5 73.699 loss 2.308
Accuracy of the model on the 50000 test images: 48.3%
Max accuracy: 48.30%
Test:  [0/5]  eta: 0:00:10  loss: 6.9091 (6.9091)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 2.0315  data: 1.7880  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9091 (6.9097)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.3836)  time: 0.5571  data: 0.3577  max mem: 57114
Test: Total time: 0:00:02 (0.5678 s / it)
* Acc@1 0.102 Acc@5 0.523 loss 6.909
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [8]  [  0/156]  eta: 0:16:14  lr: 0.003201  min_lr: 0.003201  loss: 3.8847 (3.8847)  weight_decay: 0.0500 (0.0500)  time: 6.2495  data: 4.3225  max mem: 57114
Epoch: [8]  [ 10/156]  eta: 0:02:57  lr: 0.003227  min_lr: 0.003227  loss: 5.1398 (4.9664)  weight_decay: 0.0500 (0.0500)  time: 1.2152  data: 0.3933  max mem: 57114
Epoch: [8]  [ 20/156]  eta: 0:02:10  lr: 0.003252  min_lr: 0.003252  loss: 5.1865 (5.0731)  weight_decay: 0.0500 (0.0500)  time: 0.6984  data: 0.0004  max mem: 57114
Epoch: [8]  [ 30/156]  eta: 0:01:51  lr: 0.003278  min_lr: 0.003278  loss: 5.3653 (5.0536)  weight_decay: 0.0500 (0.0500)  time: 0.7028  data: 0.0004  max mem: 57114
Epoch: [8]  [ 40/156]  eta: 0:01:37  lr: 0.003304  min_lr: 0.003304  loss: 5.1962 (5.0740)  weight_decay: 0.0500 (0.0500)  time: 0.7099  data: 0.0004  max mem: 57114
Epoch: [8]  [ 50/156]  eta: 0:01:25  lr: 0.003329  min_lr: 0.003329  loss: 5.0863 (5.0388)  weight_decay: 0.0500 (0.0500)  time: 0.6913  data: 0.0004  max mem: 57114
Epoch: [8]  [ 60/156]  eta: 0:01:15  lr: 0.003355  min_lr: 0.003355  loss: 4.8664 (5.0143)  weight_decay: 0.0500 (0.0500)  time: 0.6848  data: 0.0003  max mem: 57114
Epoch: [8]  [ 70/156]  eta: 0:01:06  lr: 0.003381  min_lr: 0.003381  loss: 4.6681 (4.9577)  weight_decay: 0.0500 (0.0500)  time: 0.6918  data: 0.0004  max mem: 57114
Epoch: [8]  [ 80/156]  eta: 0:00:58  lr: 0.003406  min_lr: 0.003406  loss: 4.6681 (4.9490)  weight_decay: 0.0500 (0.0500)  time: 0.6930  data: 0.0004  max mem: 57114
Epoch: [8]  [ 90/156]  eta: 0:00:49  lr: 0.003432  min_lr: 0.003432  loss: 5.1346 (4.9512)  weight_decay: 0.0500 (0.0500)  time: 0.6871  data: 0.0004  max mem: 57114
Epoch: [8]  [100/156]  eta: 0:00:42  lr: 0.003458  min_lr: 0.003458  loss: 5.1225 (4.9559)  weight_decay: 0.0500 (0.0500)  time: 0.6928  data: 0.0004  max mem: 57114
Epoch: [8]  [110/156]  eta: 0:00:34  lr: 0.003483  min_lr: 0.003483  loss: 5.1225 (4.9665)  weight_decay: 0.0500 (0.0500)  time: 0.6874  data: 0.0004  max mem: 57114
Epoch: [8]  [120/156]  eta: 0:00:26  lr: 0.003509  min_lr: 0.003509  loss: 5.2188 (4.9796)  weight_decay: 0.0500 (0.0500)  time: 0.6765  data: 0.0004  max mem: 57114
Epoch: [8]  [130/156]  eta: 0:00:19  lr: 0.003534  min_lr: 0.003534  loss: 5.2217 (4.9881)  weight_decay: 0.0500 (0.0500)  time: 0.6754  data: 0.0008  max mem: 57114
Epoch: [8]  [140/156]  eta: 0:00:11  lr: 0.003560  min_lr: 0.003560  loss: 5.1238 (4.9875)  weight_decay: 0.0500 (0.0500)  time: 0.6700  data: 0.0007  max mem: 57114
Epoch: [8]  [150/156]  eta: 0:00:04  lr: 0.003586  min_lr: 0.003586  loss: 4.9722 (4.9793)  weight_decay: 0.0500 (0.0500)  time: 0.6684  data: 0.0001  max mem: 57114
Epoch: [8]  [155/156]  eta: 0:00:00  lr: 0.003599  min_lr: 0.003599  loss: 5.1238 (4.9875)  weight_decay: 0.0500 (0.0500)  time: 0.6718  data: 0.0001  max mem: 57114
Epoch: [8] Total time: 0:01:53 (0.7253 s / it)
Averaged stats: lr: 0.003599  min_lr: 0.003599  loss: 5.1238 (4.8906)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 2.1142 (2.1142)  acc1: 52.6042 (52.6042)  acc5: 80.2083 (80.2083)  time: 2.1189  data: 1.8622  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 2.2420 (2.1686)  acc1: 52.6042 (48.2097)  acc5: 79.1667 (75.0639)  time: 0.5744  data: 0.3725  max mem: 57114
Test: Total time: 0:00:02 (0.5966 s / it)
* Acc@1 51.586 Acc@5 76.401 loss 2.229
Accuracy of the model on the 50000 test images: 51.6%
Max accuracy: 51.59%
Test:  [0/5]  eta: 0:00:09  loss: 6.9093 (6.9093)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 1.9969  data: 1.7534  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9093 (6.9103)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.3836)  time: 0.5500  data: 0.3508  max mem: 57114
Test: Total time: 0:00:02 (0.5621 s / it)
* Acc@1 0.098 Acc@5 0.518 loss 6.909
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [9]  [  0/156]  eta: 0:15:47  lr: 0.003601  min_lr: 0.003601  loss: 5.1432 (5.1432)  weight_decay: 0.0500 (0.0500)  time: 6.0715  data: 4.0109  max mem: 57114
Epoch: [9]  [ 10/156]  eta: 0:02:59  lr: 0.003627  min_lr: 0.003627  loss: 5.0773 (4.9432)  weight_decay: 0.0500 (0.0500)  time: 1.2269  data: 0.3649  max mem: 57114
Epoch: [9]  [ 20/156]  eta: 0:02:10  lr: 0.003652  min_lr: 0.003652  loss: 4.9699 (4.8684)  weight_decay: 0.0500 (0.0500)  time: 0.7062  data: 0.0004  max mem: 57114
Epoch: [9]  [ 30/156]  eta: 0:01:49  lr: 0.003678  min_lr: 0.003678  loss: 4.9699 (4.8163)  weight_decay: 0.0500 (0.0500)  time: 0.6767  data: 0.0004  max mem: 57114
Epoch: [9]  [ 40/156]  eta: 0:01:35  lr: 0.003704  min_lr: 0.003704  loss: 4.8599 (4.8146)  weight_decay: 0.0500 (0.0500)  time: 0.6837  data: 0.0004  max mem: 57114
Epoch: [9]  [ 50/156]  eta: 0:01:24  lr: 0.003729  min_lr: 0.003729  loss: 4.8600 (4.8159)  weight_decay: 0.0500 (0.0500)  time: 0.6830  data: 0.0004  max mem: 57114
Epoch: [9]  [ 60/156]  eta: 0:01:15  lr: 0.003755  min_lr: 0.003755  loss: 5.0165 (4.8562)  weight_decay: 0.0500 (0.0500)  time: 0.6951  data: 0.0004  max mem: 57114
Epoch: [9]  [ 70/156]  eta: 0:01:06  lr: 0.003781  min_lr: 0.003781  loss: 5.1001 (4.8128)  weight_decay: 0.0500 (0.0500)  time: 0.7079  data: 0.0004  max mem: 57114
Epoch: [9]  [ 80/156]  eta: 0:00:57  lr: 0.003806  min_lr: 0.003806  loss: 4.1551 (4.7379)  weight_decay: 0.0500 (0.0500)  time: 0.6963  data: 0.0004  max mem: 57114
Epoch: [9]  [ 90/156]  eta: 0:00:49  lr: 0.003832  min_lr: 0.003832  loss: 4.1551 (4.7057)  weight_decay: 0.0500 (0.0500)  time: 0.6820  data: 0.0003  max mem: 57114
Epoch: [9]  [100/156]  eta: 0:00:41  lr: 0.003858  min_lr: 0.003858  loss: 4.7114 (4.7220)  weight_decay: 0.0500 (0.0500)  time: 0.6798  data: 0.0003  max mem: 57114
Epoch: [9]  [110/156]  eta: 0:00:34  lr: 0.003883  min_lr: 0.003883  loss: 4.8707 (4.7305)  weight_decay: 0.0500 (0.0500)  time: 0.6818  data: 0.0004  max mem: 57114
Epoch: [9]  [120/156]  eta: 0:00:26  lr: 0.003909  min_lr: 0.003909  loss: 4.8271 (4.7453)  weight_decay: 0.0500 (0.0500)  time: 0.6842  data: 0.0003  max mem: 57114
Epoch: [9]  [130/156]  eta: 0:00:19  lr: 0.003935  min_lr: 0.003935  loss: 4.8126 (4.7299)  weight_decay: 0.0500 (0.0500)  time: 0.6868  data: 0.0008  max mem: 57114
Epoch: [9]  [140/156]  eta: 0:00:11  lr: 0.003960  min_lr: 0.003960  loss: 4.7693 (4.7257)  weight_decay: 0.0500 (0.0500)  time: 0.6773  data: 0.0007  max mem: 57114
Epoch: [9]  [150/156]  eta: 0:00:04  lr: 0.003986  min_lr: 0.003986  loss: 4.8290 (4.7242)  weight_decay: 0.0500 (0.0500)  time: 0.6642  data: 0.0001  max mem: 57114
Epoch: [9]  [155/156]  eta: 0:00:00  lr: 0.003999  min_lr: 0.003999  loss: 4.7403 (4.7190)  weight_decay: 0.0500 (0.0500)  time: 0.6647  data: 0.0001  max mem: 57114
Epoch: [9] Total time: 0:01:52 (0.7233 s / it)
Averaged stats: lr: 0.003999  min_lr: 0.003999  loss: 4.7403 (4.7397)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.7449 (1.7449)  acc1: 62.5000 (62.5000)  acc5: 85.9375 (85.9375)  time: 2.1259  data: 1.8692  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.9608 (1.9027)  acc1: 54.1667 (55.3708)  acc5: 82.2917 (79.2839)  time: 0.5760  data: 0.3739  max mem: 57114
Test: Total time: 0:00:03 (0.6000 s / it)
* Acc@1 55.525 Acc@5 80.125 loss 1.961
Accuracy of the model on the 50000 test images: 55.5%
Max accuracy: 55.52%
Test:  [0/5]  eta: 0:00:10  loss: 6.9096 (6.9096)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.5208)  time: 2.0321  data: 1.7886  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9096 (6.9110)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.3836)  time: 0.5571  data: 0.3578  max mem: 57114
Test: Total time: 0:00:02 (0.5691 s / it)
* Acc@1 0.098 Acc@5 0.510 loss 6.910
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [10]  [  0/156]  eta: 0:15:44  lr: 0.004001  min_lr: 0.004001  loss: 3.6188 (3.6188)  weight_decay: 0.0500 (0.0500)  time: 6.0555  data: 4.6256  max mem: 57114
Epoch: [10]  [ 10/156]  eta: 0:02:54  lr: 0.004027  min_lr: 0.004027  loss: 4.2387 (4.2581)  weight_decay: 0.0500 (0.0500)  time: 1.1932  data: 0.4208  max mem: 57114
Epoch: [10]  [ 20/156]  eta: 0:02:09  lr: 0.004053  min_lr: 0.004053  loss: 4.4351 (4.4081)  weight_decay: 0.0500 (0.0500)  time: 0.6960  data: 0.0003  max mem: 57114
Epoch: [10]  [ 30/156]  eta: 0:01:49  lr: 0.004078  min_lr: 0.004078  loss: 4.8167 (4.5418)  weight_decay: 0.0500 (0.0500)  time: 0.6951  data: 0.0003  max mem: 57114
Epoch: [10]  [ 40/156]  eta: 0:01:36  lr: 0.004104  min_lr: 0.004104  loss: 4.8801 (4.6249)  weight_decay: 0.0500 (0.0500)  time: 0.7060  data: 0.0003  max mem: 57114
Epoch: [10]  [ 50/156]  eta: 0:01:25  lr: 0.004130  min_lr: 0.004130  loss: 4.8801 (4.6611)  weight_decay: 0.0500 (0.0500)  time: 0.7007  data: 0.0004  max mem: 57114
Epoch: [10]  [ 60/156]  eta: 0:01:15  lr: 0.004155  min_lr: 0.004155  loss: 4.7794 (4.6616)  weight_decay: 0.0500 (0.0500)  time: 0.6895  data: 0.0004  max mem: 57114
Epoch: [10]  [ 70/156]  eta: 0:01:06  lr: 0.004181  min_lr: 0.004181  loss: 4.9151 (4.6728)  weight_decay: 0.0500 (0.0500)  time: 0.6800  data: 0.0004  max mem: 57114
Epoch: [10]  [ 80/156]  eta: 0:00:57  lr: 0.004206  min_lr: 0.004206  loss: 4.8788 (4.6724)  weight_decay: 0.0500 (0.0500)  time: 0.6777  data: 0.0004  max mem: 57114
Epoch: [10]  [ 90/156]  eta: 0:00:49  lr: 0.004232  min_lr: 0.004232  loss: 4.5888 (4.6042)  weight_decay: 0.0500 (0.0500)  time: 0.6822  data: 0.0004  max mem: 57114
Epoch: [10]  [100/156]  eta: 0:00:41  lr: 0.004258  min_lr: 0.004258  loss: 4.0322 (4.5689)  weight_decay: 0.0500 (0.0500)  time: 0.6853  data: 0.0004  max mem: 57114
Epoch: [10]  [110/156]  eta: 0:00:33  lr: 0.004283  min_lr: 0.004283  loss: 4.4005 (4.5794)  weight_decay: 0.0500 (0.0500)  time: 0.6855  data: 0.0004  max mem: 57114
Epoch: [10]  [120/156]  eta: 0:00:26  lr: 0.004309  min_lr: 0.004309  loss: 4.8883 (4.5790)  weight_decay: 0.0500 (0.0500)  time: 0.6861  data: 0.0004  max mem: 57114
Epoch: [10]  [130/156]  eta: 0:00:19  lr: 0.004335  min_lr: 0.004335  loss: 4.6563 (4.5763)  weight_decay: 0.0500 (0.0500)  time: 0.6887  data: 0.0008  max mem: 57114
Epoch: [10]  [140/156]  eta: 0:00:11  lr: 0.004360  min_lr: 0.004360  loss: 4.8933 (4.6079)  weight_decay: 0.0500 (0.0500)  time: 0.6773  data: 0.0007  max mem: 57114
Epoch: [10]  [150/156]  eta: 0:00:04  lr: 0.004386  min_lr: 0.004386  loss: 4.7839 (4.5990)  weight_decay: 0.0500 (0.0500)  time: 0.6656  data: 0.0001  max mem: 57114
Epoch: [10]  [155/156]  eta: 0:00:00  lr: 0.004399  min_lr: 0.004399  loss: 4.6794 (4.5962)  weight_decay: 0.0500 (0.0500)  time: 0.6655  data: 0.0001  max mem: 57114
Epoch: [10] Total time: 0:01:52 (0.7229 s / it)
Averaged stats: lr: 0.004399  min_lr: 0.004399  loss: 4.6794 (4.6198)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.6451 (1.6451)  acc1: 63.5417 (63.5417)  acc5: 89.0625 (89.0625)  time: 2.0794  data: 1.8240  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.8384 (1.7284)  acc1: 58.8542 (58.6957)  acc5: 82.8125 (82.2251)  time: 0.5665  data: 0.3649  max mem: 57114
Test: Total time: 0:00:02 (0.5897 s / it)
* Acc@1 58.588 Acc@5 82.223 loss 1.777
Accuracy of the model on the 50000 test images: 58.6%
Max accuracy: 58.59%
Test:  [0/5]  eta: 0:00:10  loss: 6.9098 (6.9098)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.5208)  time: 2.0778  data: 1.8343  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9098 (6.9116)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.3836)  time: 0.5661  data: 0.3669  max mem: 57114
Test: Total time: 0:00:02 (0.5786 s / it)
* Acc@1 0.098 Acc@5 0.518 loss 6.910
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [11]  [  0/156]  eta: 0:17:13  lr: 0.004401  min_lr: 0.004401  loss: 4.7723 (4.7723)  weight_decay: 0.0500 (0.0500)  time: 6.6228  data: 4.3958  max mem: 57114
Epoch: [11]  [ 10/156]  eta: 0:03:04  lr: 0.004427  min_lr: 0.004427  loss: 4.5982 (4.4395)  weight_decay: 0.0500 (0.0500)  time: 1.2617  data: 0.3999  max mem: 57114
Epoch: [11]  [ 20/156]  eta: 0:02:14  lr: 0.004453  min_lr: 0.004453  loss: 4.6713 (4.5439)  weight_decay: 0.0500 (0.0500)  time: 0.7077  data: 0.0003  max mem: 57114
Epoch: [11]  [ 30/156]  eta: 0:01:52  lr: 0.004478  min_lr: 0.004478  loss: 4.8827 (4.5346)  weight_decay: 0.0500 (0.0500)  time: 0.6940  data: 0.0003  max mem: 57114
Epoch: [11]  [ 40/156]  eta: 0:01:39  lr: 0.004504  min_lr: 0.004504  loss: 4.2066 (4.4452)  weight_decay: 0.0500 (0.0500)  time: 0.7138  data: 0.0004  max mem: 57114
Epoch: [11]  [ 50/156]  eta: 0:01:27  lr: 0.004530  min_lr: 0.004530  loss: 4.3688 (4.5010)  weight_decay: 0.0500 (0.0500)  time: 0.7059  data: 0.0004  max mem: 57114
Epoch: [11]  [ 60/156]  eta: 0:01:16  lr: 0.004555  min_lr: 0.004555  loss: 4.9140 (4.5416)  weight_decay: 0.0500 (0.0500)  time: 0.6801  data: 0.0005  max mem: 57114
Epoch: [11]  [ 70/156]  eta: 0:01:07  lr: 0.004581  min_lr: 0.004581  loss: 4.4657 (4.5074)  weight_decay: 0.0500 (0.0500)  time: 0.6831  data: 0.0005  max mem: 57114
Epoch: [11]  [ 80/156]  eta: 0:00:58  lr: 0.004607  min_lr: 0.004607  loss: 4.6939 (4.5458)  weight_decay: 0.0500 (0.0500)  time: 0.6847  data: 0.0004  max mem: 57114
Epoch: [11]  [ 90/156]  eta: 0:00:50  lr: 0.004632  min_lr: 0.004632  loss: 4.6545 (4.5347)  weight_decay: 0.0500 (0.0500)  time: 0.6837  data: 0.0004  max mem: 57114
Epoch: [11]  [100/156]  eta: 0:00:42  lr: 0.004658  min_lr: 0.004658  loss: 4.3636 (4.5236)  weight_decay: 0.0500 (0.0500)  time: 0.6949  data: 0.0004  max mem: 57114
Epoch: [11]  [110/156]  eta: 0:00:34  lr: 0.004684  min_lr: 0.004684  loss: 4.6084 (4.5358)  weight_decay: 0.0500 (0.0500)  time: 0.6987  data: 0.0004  max mem: 57114
Epoch: [11]  [120/156]  eta: 0:00:26  lr: 0.004709  min_lr: 0.004709  loss: 4.7154 (4.5508)  weight_decay: 0.0500 (0.0500)  time: 0.6882  data: 0.0004  max mem: 57114
Epoch: [11]  [130/156]  eta: 0:00:19  lr: 0.004735  min_lr: 0.004735  loss: 4.6399 (4.5273)  weight_decay: 0.0500 (0.0500)  time: 0.6786  data: 0.0009  max mem: 57114
Epoch: [11]  [140/156]  eta: 0:00:11  lr: 0.004761  min_lr: 0.004761  loss: 4.3947 (4.5158)  weight_decay: 0.0500 (0.0500)  time: 0.6695  data: 0.0007  max mem: 57114
Epoch: [11]  [150/156]  eta: 0:00:04  lr: 0.004786  min_lr: 0.004786  loss: 4.4052 (4.5043)  weight_decay: 0.0500 (0.0500)  time: 0.6671  data: 0.0001  max mem: 57114
Epoch: [11]  [155/156]  eta: 0:00:00  lr: 0.004799  min_lr: 0.004799  loss: 4.4484 (4.4977)  weight_decay: 0.0500 (0.0500)  time: 0.6693  data: 0.0001  max mem: 57114
Epoch: [11] Total time: 0:01:53 (0.7296 s / it)
Averaged stats: lr: 0.004799  min_lr: 0.004799  loss: 4.4484 (4.5192)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.6277 (1.6277)  acc1: 63.0208 (63.0208)  acc5: 91.1458 (91.1458)  time: 2.0886  data: 1.8326  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.7335 (1.7245)  acc1: 62.5000 (58.9514)  acc5: 84.8958 (83.1202)  time: 0.5684  data: 0.3666  max mem: 57114
Test: Total time: 0:00:02 (0.5922 s / it)
* Acc@1 59.039 Acc@5 82.757 loss 1.766
Accuracy of the model on the 50000 test images: 59.0%
Max accuracy: 59.04%
Test:  [0/5]  eta: 0:00:10  loss: 6.9102 (6.9102)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 2.0590  data: 1.8155  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9102 (6.9122)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.3836)  time: 0.5624  data: 0.3632  max mem: 57114
Test: Total time: 0:00:02 (0.5747 s / it)
* Acc@1 0.094 Acc@5 0.506 loss 6.911
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [12]  [  0/156]  eta: 0:16:07  lr: 0.004802  min_lr: 0.004802  loss: 5.3681 (5.3681)  weight_decay: 0.0500 (0.0500)  time: 6.2009  data: 4.7998  max mem: 57114
Epoch: [12]  [ 10/156]  eta: 0:03:02  lr: 0.004827  min_lr: 0.004827  loss: 4.7481 (4.6400)  weight_decay: 0.0500 (0.0500)  time: 1.2522  data: 0.4366  max mem: 57114
Epoch: [12]  [ 20/156]  eta: 0:02:15  lr: 0.004853  min_lr: 0.004853  loss: 4.5595 (4.5553)  weight_decay: 0.0500 (0.0500)  time: 0.7330  data: 0.0003  max mem: 57114
Epoch: [12]  [ 30/156]  eta: 0:01:53  lr: 0.004878  min_lr: 0.004878  loss: 4.5595 (4.5059)  weight_decay: 0.0500 (0.0500)  time: 0.7022  data: 0.0003  max mem: 57114
Epoch: [12]  [ 40/156]  eta: 0:01:37  lr: 0.004904  min_lr: 0.004904  loss: 4.5331 (4.4985)  weight_decay: 0.0500 (0.0500)  time: 0.6846  data: 0.0003  max mem: 57114
Epoch: [12]  [ 50/156]  eta: 0:01:26  lr: 0.004930  min_lr: 0.004930  loss: 4.1933 (4.3986)  weight_decay: 0.0500 (0.0500)  time: 0.6866  data: 0.0003  max mem: 57114
Epoch: [12]  [ 60/156]  eta: 0:01:16  lr: 0.004955  min_lr: 0.004955  loss: 4.0547 (4.3727)  weight_decay: 0.0500 (0.0500)  time: 0.6918  data: 0.0004  max mem: 57114
Epoch: [12]  [ 70/156]  eta: 0:01:06  lr: 0.004981  min_lr: 0.004981  loss: 4.4223 (4.3865)  weight_decay: 0.0500 (0.0500)  time: 0.6866  data: 0.0004  max mem: 57114
Epoch: [12]  [ 80/156]  eta: 0:00:58  lr: 0.005007  min_lr: 0.005007  loss: 4.3126 (4.3548)  weight_decay: 0.0500 (0.0500)  time: 0.6933  data: 0.0004  max mem: 57114
Epoch: [12]  [ 90/156]  eta: 0:00:50  lr: 0.005032  min_lr: 0.005032  loss: 4.2843 (4.3579)  weight_decay: 0.0500 (0.0500)  time: 0.6984  data: 0.0004  max mem: 57114
Epoch: [12]  [100/156]  eta: 0:00:42  lr: 0.005058  min_lr: 0.005058  loss: 4.3074 (4.3378)  weight_decay: 0.0500 (0.0500)  time: 0.6906  data: 0.0004  max mem: 57114
Epoch: [12]  [110/156]  eta: 0:00:34  lr: 0.005084  min_lr: 0.005084  loss: 4.2232 (4.3271)  weight_decay: 0.0500 (0.0500)  time: 0.6797  data: 0.0003  max mem: 57114
Epoch: [12]  [120/156]  eta: 0:00:26  lr: 0.005109  min_lr: 0.005109  loss: 4.2948 (4.3255)  weight_decay: 0.0500 (0.0500)  time: 0.6778  data: 0.0004  max mem: 57114
Epoch: [12]  [130/156]  eta: 0:00:19  lr: 0.005135  min_lr: 0.005135  loss: 4.5300 (4.3249)  weight_decay: 0.0500 (0.0500)  time: 0.6777  data: 0.0008  max mem: 57114
Epoch: [12]  [140/156]  eta: 0:00:11  lr: 0.005161  min_lr: 0.005161  loss: 4.6315 (4.3539)  weight_decay: 0.0500 (0.0500)  time: 0.6767  data: 0.0007  max mem: 57114
Epoch: [12]  [150/156]  eta: 0:00:04  lr: 0.005186  min_lr: 0.005186  loss: 4.5261 (4.3539)  weight_decay: 0.0500 (0.0500)  time: 0.6757  data: 0.0001  max mem: 57114
Epoch: [12]  [155/156]  eta: 0:00:00  lr: 0.005199  min_lr: 0.005199  loss: 4.4675 (4.3541)  weight_decay: 0.0500 (0.0500)  time: 0.6780  data: 0.0001  max mem: 57114
Epoch: [12] Total time: 0:01:53 (0.7285 s / it)
Averaged stats: lr: 0.005199  min_lr: 0.005199  loss: 4.4675 (4.4300)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.5187 (1.5187)  acc1: 63.5417 (63.5417)  acc5: 91.6667 (91.6667)  time: 2.0309  data: 1.7753  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.6583 (1.6781)  acc1: 63.0208 (59.7187)  acc5: 89.5833 (84.6547)  time: 0.5568  data: 0.3551  max mem: 57114
Test: Total time: 0:00:02 (0.5790 s / it)
* Acc@1 59.996 Acc@5 83.600 loss 1.734
Accuracy of the model on the 50000 test images: 60.0%
Max accuracy: 60.00%
Test:  [0/5]  eta: 0:00:10  loss: 6.9107 (6.9107)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 2.0618  data: 1.8183  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9107 (6.9128)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.3836)  time: 0.5630  data: 0.3637  max mem: 57114
Test: Total time: 0:00:02 (0.5751 s / it)
* Acc@1 0.092 Acc@5 0.510 loss 6.911
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [13]  [  0/156]  eta: 0:16:35  lr: 0.005202  min_lr: 0.005202  loss: 4.5402 (4.5402)  weight_decay: 0.0500 (0.0500)  time: 6.3793  data: 3.8439  max mem: 57114
Epoch: [13]  [ 10/156]  eta: 0:03:02  lr: 0.005227  min_lr: 0.005227  loss: 4.6508 (4.5840)  weight_decay: 0.0500 (0.0500)  time: 1.2469  data: 0.3498  max mem: 57114
Epoch: [13]  [ 20/156]  eta: 0:02:16  lr: 0.005253  min_lr: 0.005253  loss: 4.5138 (4.4470)  weight_decay: 0.0500 (0.0500)  time: 0.7321  data: 0.0003  max mem: 57114
Epoch: [13]  [ 30/156]  eta: 0:01:53  lr: 0.005279  min_lr: 0.005279  loss: 4.3785 (4.3992)  weight_decay: 0.0500 (0.0500)  time: 0.7061  data: 0.0003  max mem: 57114
Epoch: [13]  [ 40/156]  eta: 0:01:37  lr: 0.005304  min_lr: 0.005304  loss: 4.2946 (4.3747)  weight_decay: 0.0500 (0.0500)  time: 0.6801  data: 0.0004  max mem: 57114
Epoch: [13]  [ 50/156]  eta: 0:01:27  lr: 0.005330  min_lr: 0.005330  loss: 4.4639 (4.3911)  weight_decay: 0.0500 (0.0500)  time: 0.7028  data: 0.0004  max mem: 57114
Epoch: [13]  [ 60/156]  eta: 0:01:17  lr: 0.005356  min_lr: 0.005356  loss: 4.5273 (4.4103)  weight_decay: 0.0500 (0.0500)  time: 0.7310  data: 0.0004  max mem: 57114
Epoch: [13]  [ 70/156]  eta: 0:01:08  lr: 0.005381  min_lr: 0.005381  loss: 4.6722 (4.3915)  weight_decay: 0.0500 (0.0500)  time: 0.7160  data: 0.0004  max mem: 57114
Epoch: [13]  [ 80/156]  eta: 0:00:59  lr: 0.005407  min_lr: 0.005407  loss: 4.4926 (4.3853)  weight_decay: 0.0500 (0.0500)  time: 0.6993  data: 0.0003  max mem: 57114
Epoch: [13]  [ 90/156]  eta: 0:00:50  lr: 0.005433  min_lr: 0.005433  loss: 4.4926 (4.3788)  weight_decay: 0.0500 (0.0500)  time: 0.6968  data: 0.0004  max mem: 57114
Epoch: [13]  [100/156]  eta: 0:00:42  lr: 0.005458  min_lr: 0.005458  loss: 4.3244 (4.3757)  weight_decay: 0.0500 (0.0500)  time: 0.6842  data: 0.0004  max mem: 57114
Epoch: [13]  [110/156]  eta: 0:00:34  lr: 0.005484  min_lr: 0.005484  loss: 4.6445 (4.4035)  weight_decay: 0.0500 (0.0500)  time: 0.6773  data: 0.0004  max mem: 57114
Epoch: [13]  [120/156]  eta: 0:00:26  lr: 0.005509  min_lr: 0.005509  loss: 4.7577 (4.4009)  weight_decay: 0.0500 (0.0500)  time: 0.6891  data: 0.0004  max mem: 57114
Epoch: [13]  [130/156]  eta: 0:00:19  lr: 0.005535  min_lr: 0.005535  loss: 4.6034 (4.4121)  weight_decay: 0.0500 (0.0500)  time: 0.6940  data: 0.0008  max mem: 57114
Epoch: [13]  [140/156]  eta: 0:00:11  lr: 0.005561  min_lr: 0.005561  loss: 4.5908 (4.4131)  weight_decay: 0.0500 (0.0500)  time: 0.6881  data: 0.0007  max mem: 57114
Epoch: [13]  [150/156]  eta: 0:00:04  lr: 0.005586  min_lr: 0.005586  loss: 4.5576 (4.4045)  weight_decay: 0.0500 (0.0500)  time: 0.6846  data: 0.0001  max mem: 57114
Epoch: [13]  [155/156]  eta: 0:00:00  lr: 0.005599  min_lr: 0.005599  loss: 4.4230 (4.3926)  weight_decay: 0.0500 (0.0500)  time: 0.6893  data: 0.0001  max mem: 57114
Epoch: [13] Total time: 0:01:55 (0.7376 s / it)
Averaged stats: lr: 0.005599  min_lr: 0.005599  loss: 4.4230 (4.3569)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.5302 (1.5302)  acc1: 67.7083 (67.7083)  acc5: 91.6667 (91.6667)  time: 2.1176  data: 1.8621  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.7642 (1.6972)  acc1: 64.5833 (60.9974)  acc5: 88.0208 (85.6778)  time: 0.5742  data: 0.3725  max mem: 57114
Test: Total time: 0:00:02 (0.5939 s / it)
* Acc@1 61.141 Acc@5 83.905 loss 1.781
Accuracy of the model on the 50000 test images: 61.1%
Max accuracy: 61.14%
Test:  [0/5]  eta: 0:00:10  loss: 6.9113 (6.9113)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 2.0059  data: 1.7624  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9113 (6.9134)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.3836)  time: 0.5518  data: 0.3526  max mem: 57114
Test: Total time: 0:00:02 (0.5648 s / it)
* Acc@1 0.092 Acc@5 0.516 loss 6.912
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [14]  [  0/156]  eta: 0:13:57  lr: 0.005602  min_lr: 0.005602  loss: 4.7302 (4.7302)  weight_decay: 0.0500 (0.0500)  time: 5.3714  data: 4.1448  max mem: 57114
Epoch: [14]  [ 10/156]  eta: 0:02:45  lr: 0.005627  min_lr: 0.005627  loss: 4.7302 (4.5816)  weight_decay: 0.0500 (0.0500)  time: 1.1327  data: 0.3771  max mem: 57114
Epoch: [14]  [ 20/156]  eta: 0:02:04  lr: 0.005653  min_lr: 0.005653  loss: 4.3651 (4.2984)  weight_decay: 0.0500 (0.0500)  time: 0.6939  data: 0.0003  max mem: 57114
Epoch: [14]  [ 30/156]  eta: 0:01:47  lr: 0.005679  min_lr: 0.005679  loss: 4.0908 (4.3066)  weight_decay: 0.0500 (0.0500)  time: 0.7011  data: 0.0004  max mem: 57114
Epoch: [14]  [ 40/156]  eta: 0:01:35  lr: 0.005704  min_lr: 0.005704  loss: 4.0908 (4.2181)  weight_decay: 0.0500 (0.0500)  time: 0.7226  data: 0.0004  max mem: 57114
Epoch: [14]  [ 50/156]  eta: 0:01:24  lr: 0.005730  min_lr: 0.005730  loss: 4.1430 (4.2114)  weight_decay: 0.0500 (0.0500)  time: 0.7206  data: 0.0005  max mem: 57114
Epoch: [14]  [ 60/156]  eta: 0:01:16  lr: 0.005756  min_lr: 0.005756  loss: 4.1348 (4.2134)  weight_decay: 0.0500 (0.0500)  time: 0.7394  data: 0.0005  max mem: 57114
Epoch: [14]  [ 70/156]  eta: 0:01:07  lr: 0.005781  min_lr: 0.005781  loss: 4.1295 (4.1964)  weight_decay: 0.0500 (0.0500)  time: 0.7250  data: 0.0004  max mem: 57114
Epoch: [14]  [ 80/156]  eta: 0:00:58  lr: 0.005807  min_lr: 0.005807  loss: 4.3387 (4.2263)  weight_decay: 0.0500 (0.0500)  time: 0.6906  data: 0.0004  max mem: 57114
Epoch: [14]  [ 90/156]  eta: 0:00:50  lr: 0.005833  min_lr: 0.005833  loss: 4.3387 (4.2498)  weight_decay: 0.0500 (0.0500)  time: 0.6952  data: 0.0004  max mem: 57114
Epoch: [14]  [100/156]  eta: 0:00:42  lr: 0.005858  min_lr: 0.005858  loss: 4.2877 (4.2393)  weight_decay: 0.0500 (0.0500)  time: 0.6905  data: 0.0004  max mem: 57114
Epoch: [14]  [110/156]  eta: 0:00:34  lr: 0.005884  min_lr: 0.005884  loss: 4.2700 (4.2423)  weight_decay: 0.0500 (0.0500)  time: 0.6895  data: 0.0004  max mem: 57114
Epoch: [14]  [120/156]  eta: 0:00:26  lr: 0.005910  min_lr: 0.005910  loss: 4.2505 (4.2355)  weight_decay: 0.0500 (0.0500)  time: 0.7036  data: 0.0004  max mem: 57114
Epoch: [14]  [130/156]  eta: 0:00:19  lr: 0.005935  min_lr: 0.005935  loss: 4.2505 (4.2404)  weight_decay: 0.0500 (0.0500)  time: 0.6977  data: 0.0005  max mem: 57114
Epoch: [14]  [140/156]  eta: 0:00:11  lr: 0.005961  min_lr: 0.005961  loss: 4.4454 (4.2562)  weight_decay: 0.0500 (0.0500)  time: 0.6801  data: 0.0003  max mem: 57114
Epoch: [14]  [150/156]  eta: 0:00:04  lr: 0.005987  min_lr: 0.005987  loss: 4.5180 (4.2661)  weight_decay: 0.0500 (0.0500)  time: 0.6773  data: 0.0001  max mem: 57114
Epoch: [14]  [155/156]  eta: 0:00:00  lr: 0.005999  min_lr: 0.005999  loss: 4.5180 (4.2597)  weight_decay: 0.0500 (0.0500)  time: 0.6766  data: 0.0001  max mem: 57114
Epoch: [14] Total time: 0:01:54 (0.7328 s / it)
Averaged stats: lr: 0.005999  min_lr: 0.005999  loss: 4.5180 (4.2944)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.5211 (1.5211)  acc1: 70.3125 (70.3125)  acc5: 89.5833 (89.5833)  time: 2.1059  data: 1.8504  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.7095 (1.6033)  acc1: 65.6250 (64.5780)  acc5: 89.5833 (85.2941)  time: 0.5718  data: 0.3701  max mem: 57114
Test: Total time: 0:00:02 (0.5955 s / it)
* Acc@1 61.515 Acc@5 83.989 loss 1.768
Accuracy of the model on the 50000 test images: 61.5%
Max accuracy: 61.51%
Test:  [0/5]  eta: 0:00:09  loss: 6.9120 (6.9120)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 1.9540  data: 1.7104  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9120 (6.9139)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.2558)  time: 0.5415  data: 0.3422  max mem: 57114
Test: Total time: 0:00:02 (0.5543 s / it)
* Acc@1 0.094 Acc@5 0.508 loss 6.912
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [15]  [  0/156]  eta: 0:14:44  lr: 0.006002  min_lr: 0.006002  loss: 4.3763 (4.3763)  weight_decay: 0.0500 (0.0500)  time: 5.6721  data: 3.7482  max mem: 57114
Epoch: [15]  [ 10/156]  eta: 0:02:49  lr: 0.006028  min_lr: 0.006028  loss: 4.2275 (4.2601)  weight_decay: 0.0500 (0.0500)  time: 1.1619  data: 0.3410  max mem: 57114
Epoch: [15]  [ 20/156]  eta: 0:02:08  lr: 0.006053  min_lr: 0.006053  loss: 4.1850 (4.1846)  weight_decay: 0.0500 (0.0500)  time: 0.7070  data: 0.0003  max mem: 57114
Epoch: [15]  [ 30/156]  eta: 0:01:50  lr: 0.006079  min_lr: 0.006079  loss: 4.2131 (4.1899)  weight_decay: 0.0500 (0.0500)  time: 0.7201  data: 0.0003  max mem: 57114
Epoch: [15]  [ 40/156]  eta: 0:01:37  lr: 0.006105  min_lr: 0.006105  loss: 4.3292 (4.2263)  weight_decay: 0.0500 (0.0500)  time: 0.7266  data: 0.0004  max mem: 57114
Epoch: [15]  [ 50/156]  eta: 0:01:26  lr: 0.006130  min_lr: 0.006130  loss: 4.3614 (4.1919)  weight_decay: 0.0500 (0.0500)  time: 0.7205  data: 0.0004  max mem: 57114
Epoch: [15]  [ 60/156]  eta: 0:01:16  lr: 0.006156  min_lr: 0.006156  loss: 3.8396 (4.1721)  weight_decay: 0.0500 (0.0500)  time: 0.7126  data: 0.0003  max mem: 57114
Epoch: [15]  [ 70/156]  eta: 0:01:07  lr: 0.006181  min_lr: 0.006181  loss: 4.5889 (4.2220)  weight_decay: 0.0500 (0.0500)  time: 0.7095  data: 0.0003  max mem: 57114
Epoch: [15]  [ 80/156]  eta: 0:00:58  lr: 0.006207  min_lr: 0.006207  loss: 4.4255 (4.1946)  weight_decay: 0.0500 (0.0500)  time: 0.7078  data: 0.0003  max mem: 57114
Epoch: [15]  [ 90/156]  eta: 0:00:50  lr: 0.006233  min_lr: 0.006233  loss: 4.2510 (4.2136)  weight_decay: 0.0500 (0.0500)  time: 0.6929  data: 0.0004  max mem: 57114
Epoch: [15]  [100/156]  eta: 0:00:42  lr: 0.006258  min_lr: 0.006258  loss: 4.4419 (4.2302)  weight_decay: 0.0500 (0.0500)  time: 0.7018  data: 0.0003  max mem: 57114
Epoch: [15]  [110/156]  eta: 0:00:34  lr: 0.006284  min_lr: 0.006284  loss: 4.4036 (4.2478)  weight_decay: 0.0500 (0.0500)  time: 0.7093  data: 0.0004  max mem: 57114
Epoch: [15]  [120/156]  eta: 0:00:27  lr: 0.006310  min_lr: 0.006310  loss: 4.4928 (4.2541)  weight_decay: 0.0500 (0.0500)  time: 0.7046  data: 0.0004  max mem: 57114
Epoch: [15]  [130/156]  eta: 0:00:19  lr: 0.006335  min_lr: 0.006335  loss: 4.4928 (4.2451)  weight_decay: 0.0500 (0.0500)  time: 0.6992  data: 0.0005  max mem: 57114
Epoch: [15]  [140/156]  eta: 0:00:11  lr: 0.006361  min_lr: 0.006361  loss: 3.9288 (4.2330)  weight_decay: 0.0500 (0.0500)  time: 0.6882  data: 0.0003  max mem: 57114
Epoch: [15]  [150/156]  eta: 0:00:04  lr: 0.006387  min_lr: 0.006387  loss: 4.3350 (4.2466)  weight_decay: 0.0500 (0.0500)  time: 0.6780  data: 0.0001  max mem: 57114
Epoch: [15]  [155/156]  eta: 0:00:00  lr: 0.006399  min_lr: 0.006399  loss: 4.1822 (4.2338)  weight_decay: 0.0500 (0.0500)  time: 0.6764  data: 0.0001  max mem: 57114
Epoch: [15] Total time: 0:01:55 (0.7376 s / it)
Averaged stats: lr: 0.006399  min_lr: 0.006399  loss: 4.1822 (4.2456)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.4190 (1.4190)  acc1: 73.4375 (73.4375)  acc5: 92.1875 (92.1875)  time: 2.0678  data: 1.8123  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.7325 (1.5931)  acc1: 61.9792 (63.6829)  acc5: 86.9792 (85.8056)  time: 0.5644  data: 0.3625  max mem: 57114
Test: Total time: 0:00:02 (0.5872 s / it)
* Acc@1 62.182 Acc@5 84.877 loss 1.682
Accuracy of the model on the 50000 test images: 62.2%
Max accuracy: 62.18%
Test:  [0/5]  eta: 0:00:10  loss: 6.9124 (6.9124)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 2.1030  data: 1.8598  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9124 (6.9145)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.2558)  time: 0.5712  data: 0.3720  max mem: 57114
Test: Total time: 0:00:02 (0.5825 s / it)
* Acc@1 0.094 Acc@5 0.500 loss 6.913
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [16]  [  0/156]  eta: 0:14:10  lr: 0.006402  min_lr: 0.006402  loss: 3.5861 (3.5861)  weight_decay: 0.0500 (0.0500)  time: 5.4492  data: 4.0975  max mem: 57114
Epoch: [16]  [ 10/156]  eta: 0:02:50  lr: 0.006428  min_lr: 0.006428  loss: 3.7842 (4.0122)  weight_decay: 0.0500 (0.0500)  time: 1.1709  data: 0.3728  max mem: 57114
Epoch: [16]  [ 20/156]  eta: 0:02:10  lr: 0.006453  min_lr: 0.006453  loss: 4.2437 (4.2029)  weight_decay: 0.0500 (0.0500)  time: 0.7357  data: 0.0004  max mem: 57114
Epoch: [16]  [ 30/156]  eta: 0:01:51  lr: 0.006479  min_lr: 0.006479  loss: 4.4830 (4.2684)  weight_decay: 0.0500 (0.0500)  time: 0.7325  data: 0.0003  max mem: 57114
Epoch: [16]  [ 40/156]  eta: 0:01:38  lr: 0.006505  min_lr: 0.006505  loss: 4.4500 (4.2827)  weight_decay: 0.0500 (0.0500)  time: 0.7308  data: 0.0004  max mem: 57114
Epoch: [16]  [ 50/156]  eta: 0:01:26  lr: 0.006530  min_lr: 0.006530  loss: 4.2034 (4.2377)  weight_decay: 0.0500 (0.0500)  time: 0.7055  data: 0.0004  max mem: 57114
Epoch: [16]  [ 60/156]  eta: 0:01:17  lr: 0.006556  min_lr: 0.006556  loss: 4.4842 (4.2893)  weight_decay: 0.0500 (0.0500)  time: 0.7076  data: 0.0004  max mem: 57114
Epoch: [16]  [ 70/156]  eta: 0:01:07  lr: 0.006582  min_lr: 0.006582  loss: 4.6258 (4.3134)  weight_decay: 0.0500 (0.0500)  time: 0.7244  data: 0.0004  max mem: 57114
Epoch: [16]  [ 80/156]  eta: 0:00:59  lr: 0.006607  min_lr: 0.006607  loss: 4.4099 (4.3037)  weight_decay: 0.0500 (0.0500)  time: 0.7164  data: 0.0004  max mem: 57114
Epoch: [16]  [ 90/156]  eta: 0:00:50  lr: 0.006633  min_lr: 0.006633  loss: 4.2877 (4.2977)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0004  max mem: 57114
Epoch: [16]  [100/156]  eta: 0:00:42  lr: 0.006659  min_lr: 0.006659  loss: 4.0884 (4.2702)  weight_decay: 0.0500 (0.0500)  time: 0.7111  data: 0.0004  max mem: 57114
Epoch: [16]  [110/156]  eta: 0:00:35  lr: 0.006684  min_lr: 0.006684  loss: 4.1894 (4.2780)  weight_decay: 0.0500 (0.0500)  time: 0.7075  data: 0.0004  max mem: 57114
Epoch: [16]  [120/156]  eta: 0:00:27  lr: 0.006710  min_lr: 0.006710  loss: 4.3576 (4.2654)  weight_decay: 0.0500 (0.0500)  time: 0.6971  data: 0.0004  max mem: 57114
Epoch: [16]  [130/156]  eta: 0:00:19  lr: 0.006735  min_lr: 0.006735  loss: 4.1939 (4.2604)  weight_decay: 0.0500 (0.0500)  time: 0.6947  data: 0.0009  max mem: 57114
Epoch: [16]  [140/156]  eta: 0:00:11  lr: 0.006761  min_lr: 0.006761  loss: 4.1939 (4.2625)  weight_decay: 0.0500 (0.0500)  time: 0.6858  data: 0.0007  max mem: 57114
Epoch: [16]  [150/156]  eta: 0:00:04  lr: 0.006787  min_lr: 0.006787  loss: 4.3985 (4.2804)  weight_decay: 0.0500 (0.0500)  time: 0.6803  data: 0.0001  max mem: 57114
Epoch: [16]  [155/156]  eta: 0:00:00  lr: 0.006800  min_lr: 0.006800  loss: 4.4220 (4.2621)  weight_decay: 0.0500 (0.0500)  time: 0.6811  data: 0.0001  max mem: 57114
Epoch: [16] Total time: 0:01:55 (0.7415 s / it)
Averaged stats: lr: 0.006800  min_lr: 0.006800  loss: 4.4220 (4.1987)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.3990 (1.3990)  acc1: 71.3542 (71.3542)  acc5: 92.7083 (92.7083)  time: 2.0732  data: 1.8177  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.7268 (1.6207)  acc1: 64.0625 (61.8926)  acc5: 88.0208 (85.5499)  time: 0.5653  data: 0.3636  max mem: 57114
Test: Total time: 0:00:02 (0.5886 s / it)
* Acc@1 62.158 Acc@5 84.984 loss 1.724
Accuracy of the model on the 50000 test images: 62.2%
Max accuracy: 62.18%
Test:  [0/5]  eta: 0:00:11  loss: 6.9127 (6.9127)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 2.3583  data: 2.1150  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9127 (6.9151)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.2558)  time: 0.6223  data: 0.4231  max mem: 57114
Test: Total time: 0:00:03 (0.6324 s / it)
* Acc@1 0.094 Acc@5 0.496 loss 6.913
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [17]  [  0/156]  eta: 0:14:45  lr: 0.006802  min_lr: 0.006802  loss: 3.2083 (3.2083)  weight_decay: 0.0500 (0.0500)  time: 5.6747  data: 4.4156  max mem: 57114
Epoch: [17]  [ 10/156]  eta: 0:02:55  lr: 0.006828  min_lr: 0.006828  loss: 4.3897 (4.2701)  weight_decay: 0.0500 (0.0500)  time: 1.2043  data: 0.4017  max mem: 57114
Epoch: [17]  [ 20/156]  eta: 0:02:12  lr: 0.006853  min_lr: 0.006853  loss: 4.3897 (4.2103)  weight_decay: 0.0500 (0.0500)  time: 0.7371  data: 0.0003  max mem: 57114
Epoch: [17]  [ 30/156]  eta: 0:01:52  lr: 0.006879  min_lr: 0.006879  loss: 4.3925 (4.2199)  weight_decay: 0.0500 (0.0500)  time: 0.7168  data: 0.0003  max mem: 57114
Epoch: [17]  [ 40/156]  eta: 0:01:38  lr: 0.006905  min_lr: 0.006905  loss: 4.3702 (4.2318)  weight_decay: 0.0500 (0.0500)  time: 0.7194  data: 0.0003  max mem: 57114
Epoch: [17]  [ 50/156]  eta: 0:01:27  lr: 0.006930  min_lr: 0.006930  loss: 4.4279 (4.2763)  weight_decay: 0.0500 (0.0500)  time: 0.7142  data: 0.0004  max mem: 57114
Epoch: [17]  [ 60/156]  eta: 0:01:17  lr: 0.006956  min_lr: 0.006956  loss: 4.4699 (4.2928)  weight_decay: 0.0500 (0.0500)  time: 0.7137  data: 0.0004  max mem: 57114
Epoch: [17]  [ 70/156]  eta: 0:01:08  lr: 0.006982  min_lr: 0.006982  loss: 4.2058 (4.2642)  weight_decay: 0.0500 (0.0500)  time: 0.7231  data: 0.0004  max mem: 57114
Epoch: [17]  [ 80/156]  eta: 0:00:59  lr: 0.007007  min_lr: 0.007007  loss: 4.3176 (4.2954)  weight_decay: 0.0500 (0.0500)  time: 0.7216  data: 0.0004  max mem: 57114
Epoch: [17]  [ 90/156]  eta: 0:00:51  lr: 0.007033  min_lr: 0.007033  loss: 4.2535 (4.2346)  weight_decay: 0.0500 (0.0500)  time: 0.7104  data: 0.0004  max mem: 57114
Epoch: [17]  [100/156]  eta: 0:00:42  lr: 0.007059  min_lr: 0.007059  loss: 3.8849 (4.2152)  weight_decay: 0.0500 (0.0500)  time: 0.6968  data: 0.0004  max mem: 57114
Epoch: [17]  [110/156]  eta: 0:00:34  lr: 0.007084  min_lr: 0.007084  loss: 4.2134 (4.2026)  weight_decay: 0.0500 (0.0500)  time: 0.6952  data: 0.0003  max mem: 57114
Epoch: [17]  [120/156]  eta: 0:00:27  lr: 0.007110  min_lr: 0.007110  loss: 4.4383 (4.2190)  weight_decay: 0.0500 (0.0500)  time: 0.6988  data: 0.0003  max mem: 57114
Epoch: [17]  [130/156]  eta: 0:00:19  lr: 0.007136  min_lr: 0.007136  loss: 4.2249 (4.2090)  weight_decay: 0.0500 (0.0500)  time: 0.6976  data: 0.0009  max mem: 57114
Epoch: [17]  [140/156]  eta: 0:00:11  lr: 0.007161  min_lr: 0.007161  loss: 3.9044 (4.1813)  weight_decay: 0.0500 (0.0500)  time: 0.6920  data: 0.0007  max mem: 57114
Epoch: [17]  [150/156]  eta: 0:00:04  lr: 0.007187  min_lr: 0.007187  loss: 3.7496 (4.1717)  weight_decay: 0.0500 (0.0500)  time: 0.6831  data: 0.0002  max mem: 57114
Epoch: [17]  [155/156]  eta: 0:00:00  lr: 0.007200  min_lr: 0.007200  loss: 4.1252 (4.1722)  weight_decay: 0.0500 (0.0500)  time: 0.6836  data: 0.0002  max mem: 57114
Epoch: [17] Total time: 0:01:55 (0.7427 s / it)
Averaged stats: lr: 0.007200  min_lr: 0.007200  loss: 4.1252 (4.1599)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.6515 (1.6515)  acc1: 70.3125 (70.3125)  acc5: 92.1875 (92.1875)  time: 2.1043  data: 1.8484  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.8337 (1.8879)  acc1: 65.6250 (61.5090)  acc5: 88.0208 (84.7826)  time: 0.5715  data: 0.3698  max mem: 57114
Test: Total time: 0:00:02 (0.5942 s / it)
* Acc@1 60.578 Acc@5 83.506 loss 2.018
Accuracy of the model on the 50000 test images: 60.6%
Max accuracy: 62.18%
Test:  [0/5]  eta: 0:00:11  loss: 6.9131 (6.9131)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 2.2941  data: 2.0506  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9131 (6.9154)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.2558)  time: 0.6096  data: 0.4103  max mem: 57114
Test: Total time: 0:00:03 (0.6237 s / it)
* Acc@1 0.090 Acc@5 0.488 loss 6.914
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [18]  [  0/156]  eta: 0:14:08  lr: 0.007202  min_lr: 0.007202  loss: 4.2097 (4.2097)  weight_decay: 0.0500 (0.0500)  time: 5.4412  data: 4.3787  max mem: 57114
Epoch: [18]  [ 10/156]  eta: 0:02:48  lr: 0.007228  min_lr: 0.007228  loss: 4.2543 (4.2210)  weight_decay: 0.0500 (0.0500)  time: 1.1520  data: 0.3984  max mem: 57114
Epoch: [18]  [ 20/156]  eta: 0:02:07  lr: 0.007254  min_lr: 0.007254  loss: 4.2294 (4.1376)  weight_decay: 0.0500 (0.0500)  time: 0.7113  data: 0.0003  max mem: 57114
Epoch: [18]  [ 30/156]  eta: 0:01:49  lr: 0.007279  min_lr: 0.007279  loss: 4.0090 (4.0851)  weight_decay: 0.0500 (0.0500)  time: 0.7163  data: 0.0003  max mem: 57114
Epoch: [18]  [ 40/156]  eta: 0:01:37  lr: 0.007305  min_lr: 0.007305  loss: 3.9226 (4.0563)  weight_decay: 0.0500 (0.0500)  time: 0.7344  data: 0.0004  max mem: 57114
Epoch: [18]  [ 50/156]  eta: 0:01:26  lr: 0.007331  min_lr: 0.007331  loss: 3.7113 (3.9872)  weight_decay: 0.0500 (0.0500)  time: 0.7194  data: 0.0004  max mem: 57114
Epoch: [18]  [ 60/156]  eta: 0:01:16  lr: 0.007356  min_lr: 0.007356  loss: 3.6777 (3.9632)  weight_decay: 0.0500 (0.0500)  time: 0.7111  data: 0.0004  max mem: 57114
Epoch: [18]  [ 70/156]  eta: 0:01:07  lr: 0.007382  min_lr: 0.007382  loss: 3.7919 (3.9487)  weight_decay: 0.0500 (0.0500)  time: 0.7219  data: 0.0004  max mem: 57114
Epoch: [18]  [ 80/156]  eta: 0:00:58  lr: 0.007408  min_lr: 0.007408  loss: 3.9074 (3.9410)  weight_decay: 0.0500 (0.0500)  time: 0.7115  data: 0.0004  max mem: 57114
Epoch: [18]  [ 90/156]  eta: 0:00:50  lr: 0.007433  min_lr: 0.007433  loss: 4.3708 (3.9900)  weight_decay: 0.0500 (0.0500)  time: 0.7021  data: 0.0004  max mem: 57114
Epoch: [18]  [100/156]  eta: 0:00:42  lr: 0.007459  min_lr: 0.007459  loss: 4.2398 (3.9775)  weight_decay: 0.0500 (0.0500)  time: 0.7040  data: 0.0004  max mem: 57114
Epoch: [18]  [110/156]  eta: 0:00:34  lr: 0.007484  min_lr: 0.007484  loss: 4.0923 (3.9994)  weight_decay: 0.0500 (0.0500)  time: 0.7030  data: 0.0004  max mem: 57114
Epoch: [18]  [120/156]  eta: 0:00:27  lr: 0.007510  min_lr: 0.007510  loss: 4.3381 (4.0135)  weight_decay: 0.0500 (0.0500)  time: 0.6999  data: 0.0004  max mem: 57114
Epoch: [18]  [130/156]  eta: 0:00:19  lr: 0.007536  min_lr: 0.007536  loss: 4.4067 (4.0474)  weight_decay: 0.0500 (0.0500)  time: 0.6930  data: 0.0005  max mem: 57114
Epoch: [18]  [140/156]  eta: 0:00:11  lr: 0.007561  min_lr: 0.007561  loss: 4.2909 (4.0437)  weight_decay: 0.0500 (0.0500)  time: 0.6801  data: 0.0003  max mem: 57114
Epoch: [18]  [150/156]  eta: 0:00:04  lr: 0.007587  min_lr: 0.007587  loss: 4.0057 (4.0427)  weight_decay: 0.0500 (0.0500)  time: 0.6806  data: 0.0001  max mem: 57114
Epoch: [18]  [155/156]  eta: 0:00:00  lr: 0.007600  min_lr: 0.007600  loss: 3.9905 (4.0403)  weight_decay: 0.0500 (0.0500)  time: 0.6839  data: 0.0001  max mem: 57114
Epoch: [18] Total time: 0:01:55 (0.7380 s / it)
Averaged stats: lr: 0.007600  min_lr: 0.007600  loss: 3.9905 (4.1269)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.4056 (1.4056)  acc1: 68.7500 (68.7500)  acc5: 91.6667 (91.6667)  time: 2.1299  data: 1.8744  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.6288 (1.5463)  acc1: 66.1458 (62.6598)  acc5: 89.5833 (86.8286)  time: 0.5766  data: 0.3749  max mem: 57114
Test: Total time: 0:00:03 (0.6030 s / it)
* Acc@1 62.582 Acc@5 85.322 loss 1.673
Accuracy of the model on the 50000 test images: 62.6%
Max accuracy: 62.58%
Test:  [0/5]  eta: 0:00:10  loss: 6.9131 (6.9131)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 2.0376  data: 1.7942  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9131 (6.9156)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.1279)  time: 0.5581  data: 0.3589  max mem: 57114
Test: Total time: 0:00:02 (0.5729 s / it)
* Acc@1 0.092 Acc@5 0.482 loss 6.914
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [19]  [  0/156]  eta: 0:13:26  lr: 0.007602  min_lr: 0.007602  loss: 3.2453 (3.2453)  weight_decay: 0.0500 (0.0500)  time: 5.1717  data: 4.3763  max mem: 57114
Epoch: [19]  [ 10/156]  eta: 0:02:44  lr: 0.007628  min_lr: 0.007628  loss: 4.3939 (4.2462)  weight_decay: 0.0500 (0.0500)  time: 1.1252  data: 0.3981  max mem: 57114
Epoch: [19]  [ 20/156]  eta: 0:02:08  lr: 0.007654  min_lr: 0.007654  loss: 4.3011 (4.0756)  weight_decay: 0.0500 (0.0500)  time: 0.7299  data: 0.0003  max mem: 57114
Epoch: [19]  [ 30/156]  eta: 0:01:51  lr: 0.007679  min_lr: 0.007679  loss: 4.3011 (4.1555)  weight_decay: 0.0500 (0.0500)  time: 0.7481  data: 0.0003  max mem: 57114
Epoch: [19]  [ 40/156]  eta: 0:01:37  lr: 0.007705  min_lr: 0.007705  loss: 4.2574 (4.1396)  weight_decay: 0.0500 (0.0500)  time: 0.7299  data: 0.0004  max mem: 57114
Epoch: [19]  [ 50/156]  eta: 0:01:26  lr: 0.007731  min_lr: 0.007731  loss: 4.1608 (4.1317)  weight_decay: 0.0500 (0.0500)  time: 0.7040  data: 0.0004  max mem: 57114
Epoch: [19]  [ 60/156]  eta: 0:01:17  lr: 0.007756  min_lr: 0.007756  loss: 4.1131 (4.1097)  weight_decay: 0.0500 (0.0500)  time: 0.7332  data: 0.0004  max mem: 57114
Epoch: [19]  [ 70/156]  eta: 0:01:08  lr: 0.007782  min_lr: 0.007782  loss: 4.1131 (4.1099)  weight_decay: 0.0500 (0.0500)  time: 0.7390  data: 0.0003  max mem: 57114
Epoch: [19]  [ 80/156]  eta: 0:00:59  lr: 0.007808  min_lr: 0.007808  loss: 4.1617 (4.1117)  weight_decay: 0.0500 (0.0500)  time: 0.7088  data: 0.0004  max mem: 57114
Epoch: [19]  [ 90/156]  eta: 0:00:50  lr: 0.007833  min_lr: 0.007833  loss: 4.3236 (4.1172)  weight_decay: 0.0500 (0.0500)  time: 0.6962  data: 0.0004  max mem: 57114
Epoch: [19]  [100/156]  eta: 0:00:42  lr: 0.007859  min_lr: 0.007859  loss: 4.1616 (4.0887)  weight_decay: 0.0500 (0.0500)  time: 0.6986  data: 0.0004  max mem: 57114
Epoch: [19]  [110/156]  eta: 0:00:34  lr: 0.007885  min_lr: 0.007885  loss: 4.0414 (4.0874)  weight_decay: 0.0500 (0.0500)  time: 0.7010  data: 0.0004  max mem: 57114
Epoch: [19]  [120/156]  eta: 0:00:27  lr: 0.007910  min_lr: 0.007910  loss: 4.0414 (4.0655)  weight_decay: 0.0500 (0.0500)  time: 0.7009  data: 0.0004  max mem: 57114
Epoch: [19]  [130/156]  eta: 0:00:19  lr: 0.007936  min_lr: 0.007936  loss: 4.3433 (4.0810)  weight_decay: 0.0500 (0.0500)  time: 0.6992  data: 0.0005  max mem: 57114
Epoch: [19]  [140/156]  eta: 0:00:11  lr: 0.007962  min_lr: 0.007962  loss: 4.2178 (4.0658)  weight_decay: 0.0500 (0.0500)  time: 0.6839  data: 0.0003  max mem: 57114
Epoch: [19]  [150/156]  eta: 0:00:04  lr: 0.007987  min_lr: 0.007987  loss: 4.2096 (4.0730)  weight_decay: 0.0500 (0.0500)  time: 0.6795  data: 0.0001  max mem: 57114
Epoch: [19]  [155/156]  eta: 0:00:00  lr: 0.008000  min_lr: 0.008000  loss: 4.1025 (4.0801)  weight_decay: 0.0500 (0.0500)  time: 0.6845  data: 0.0001  max mem: 57114
Epoch: [19] Total time: 0:01:55 (0.7402 s / it)
Averaged stats: lr: 0.008000  min_lr: 0.008000  loss: 4.1025 (4.0941)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.5892 (1.5892)  acc1: 68.7500 (68.7500)  acc5: 94.7917 (94.7917)  time: 2.0027  data: 1.7473  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.7985 (1.7919)  acc1: 64.0625 (62.4041)  acc5: 86.4583 (86.4450)  time: 0.5512  data: 0.3495  max mem: 57114
Test: Total time: 0:00:02 (0.5723 s / it)
* Acc@1 61.281 Acc@5 84.861 loss 1.904
Accuracy of the model on the 50000 test images: 61.3%
Max accuracy: 62.58%
Test:  [0/5]  eta: 0:00:11  loss: 6.9136 (6.9136)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 2.3153  data: 2.0720  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9136 (6.9161)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.1279)  time: 0.6137  data: 0.4145  max mem: 57114
Test: Total time: 0:00:03 (0.6251 s / it)
* Acc@1 0.096 Acc@5 0.480 loss 6.915
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [20]  [  0/156]  eta: 0:14:13  lr: 0.008000  min_lr: 0.008000  loss: 4.0290 (4.0290)  weight_decay: 0.0500 (0.0500)  time: 5.4737  data: 3.7498  max mem: 57114
Epoch: [20]  [ 10/156]  eta: 0:02:53  lr: 0.008000  min_lr: 0.008000  loss: 4.0290 (4.0520)  weight_decay: 0.0500 (0.0500)  time: 1.1872  data: 0.3411  max mem: 57114
Epoch: [20]  [ 20/156]  eta: 0:02:10  lr: 0.008000  min_lr: 0.008000  loss: 3.9414 (3.8907)  weight_decay: 0.0500 (0.0500)  time: 0.7340  data: 0.0003  max mem: 57114
Epoch: [20]  [ 30/156]  eta: 0:01:51  lr: 0.008000  min_lr: 0.008000  loss: 3.9078 (3.9340)  weight_decay: 0.0500 (0.0500)  time: 0.7232  data: 0.0003  max mem: 57114
Epoch: [20]  [ 40/156]  eta: 0:01:37  lr: 0.008000  min_lr: 0.008000  loss: 4.0936 (3.9520)  weight_decay: 0.0500 (0.0500)  time: 0.7196  data: 0.0003  max mem: 57114
Epoch: [20]  [ 50/156]  eta: 0:01:26  lr: 0.008000  min_lr: 0.008000  loss: 4.1379 (3.9879)  weight_decay: 0.0500 (0.0500)  time: 0.7037  data: 0.0004  max mem: 57114
Epoch: [20]  [ 60/156]  eta: 0:01:16  lr: 0.008000  min_lr: 0.008000  loss: 4.2262 (3.9890)  weight_decay: 0.0500 (0.0500)  time: 0.7151  data: 0.0004  max mem: 57114
Epoch: [20]  [ 70/156]  eta: 0:01:07  lr: 0.008000  min_lr: 0.008000  loss: 3.9458 (3.9696)  weight_decay: 0.0500 (0.0500)  time: 0.7189  data: 0.0004  max mem: 57114
Epoch: [20]  [ 80/156]  eta: 0:00:59  lr: 0.008000  min_lr: 0.008000  loss: 4.0986 (4.0000)  weight_decay: 0.0500 (0.0500)  time: 0.7052  data: 0.0003  max mem: 57114
Epoch: [20]  [ 90/156]  eta: 0:00:50  lr: 0.008000  min_lr: 0.008000  loss: 4.2416 (4.0135)  weight_decay: 0.0500 (0.0500)  time: 0.7048  data: 0.0003  max mem: 57114
Epoch: [20]  [100/156]  eta: 0:00:42  lr: 0.008000  min_lr: 0.008000  loss: 4.3286 (4.0286)  weight_decay: 0.0500 (0.0500)  time: 0.7108  data: 0.0004  max mem: 57114
Epoch: [20]  [110/156]  eta: 0:00:34  lr: 0.008000  min_lr: 0.008000  loss: 4.1594 (4.0466)  weight_decay: 0.0500 (0.0500)  time: 0.6995  data: 0.0004  max mem: 57114
Epoch: [20]  [120/156]  eta: 0:00:27  lr: 0.008000  min_lr: 0.008000  loss: 4.3322 (4.0781)  weight_decay: 0.0500 (0.0500)  time: 0.6895  data: 0.0004  max mem: 57114
Epoch: [20]  [130/156]  eta: 0:00:19  lr: 0.008000  min_lr: 0.008000  loss: 4.2505 (4.0834)  weight_decay: 0.0500 (0.0500)  time: 0.6908  data: 0.0008  max mem: 57114
Epoch: [20]  [140/156]  eta: 0:00:11  lr: 0.008000  min_lr: 0.008000  loss: 3.9719 (4.0811)  weight_decay: 0.0500 (0.0500)  time: 0.6864  data: 0.0007  max mem: 57114
Epoch: [20]  [150/156]  eta: 0:00:04  lr: 0.008000  min_lr: 0.008000  loss: 3.9719 (4.0729)  weight_decay: 0.0500 (0.0500)  time: 0.6836  data: 0.0001  max mem: 57114
Epoch: [20]  [155/156]  eta: 0:00:00  lr: 0.008000  min_lr: 0.008000  loss: 3.9583 (4.0731)  weight_decay: 0.0500 (0.0500)  time: 0.6842  data: 0.0001  max mem: 57114
Epoch: [20] Total time: 0:01:55 (0.7390 s / it)
Averaged stats: lr: 0.008000  min_lr: 0.008000  loss: 3.9583 (4.0587)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.3891 (1.3891)  acc1: 74.4792 (74.4792)  acc5: 92.7083 (92.7083)  time: 2.1088  data: 1.8531  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.6327 (1.5960)  acc1: 66.1458 (63.9386)  acc5: 89.0625 (85.9335)  time: 0.5724  data: 0.3707  max mem: 57114
Test: Total time: 0:00:02 (0.5921 s / it)
* Acc@1 62.622 Acc@5 85.326 loss 1.728
Accuracy of the model on the 50000 test images: 62.6%
Max accuracy: 62.62%
Test:  [0/5]  eta: 0:00:09  loss: 6.9137 (6.9137)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 1.9627  data: 1.7193  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9137 (6.9167)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.1279)  time: 0.5432  data: 0.3439  max mem: 57114
Test: Total time: 0:00:02 (0.5538 s / it)
* Acc@1 0.094 Acc@5 0.488 loss 6.915
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [21]  [  0/156]  eta: 0:14:11  lr: 0.008000  min_lr: 0.008000  loss: 4.0519 (4.0519)  weight_decay: 0.0500 (0.0500)  time: 5.4586  data: 3.8907  max mem: 57114
Epoch: [21]  [ 10/156]  eta: 0:02:51  lr: 0.008000  min_lr: 0.008000  loss: 3.9884 (3.8412)  weight_decay: 0.0500 (0.0500)  time: 1.1773  data: 0.3540  max mem: 57114
Epoch: [21]  [ 20/156]  eta: 0:02:08  lr: 0.008000  min_lr: 0.008000  loss: 3.9884 (4.0779)  weight_decay: 0.0500 (0.0500)  time: 0.7180  data: 0.0003  max mem: 57114
Epoch: [21]  [ 30/156]  eta: 0:01:49  lr: 0.008000  min_lr: 0.008000  loss: 4.3641 (4.0831)  weight_decay: 0.0500 (0.0500)  time: 0.7052  data: 0.0003  max mem: 57114
Epoch: [21]  [ 40/156]  eta: 0:01:36  lr: 0.008000  min_lr: 0.008000  loss: 4.1796 (4.0676)  weight_decay: 0.0500 (0.0500)  time: 0.7228  data: 0.0004  max mem: 57114
Epoch: [21]  [ 50/156]  eta: 0:01:27  lr: 0.008000  min_lr: 0.008000  loss: 3.8136 (3.9918)  weight_decay: 0.0500 (0.0500)  time: 0.7466  data: 0.0004  max mem: 57114
Epoch: [21]  [ 60/156]  eta: 0:01:18  lr: 0.008000  min_lr: 0.008000  loss: 3.6541 (3.9356)  weight_decay: 0.0500 (0.0500)  time: 0.7653  data: 0.0004  max mem: 57114
Epoch: [21]  [ 70/156]  eta: 0:01:08  lr: 0.007999  min_lr: 0.007999  loss: 3.8990 (3.9479)  weight_decay: 0.0500 (0.0500)  time: 0.7435  data: 0.0004  max mem: 57114
Epoch: [21]  [ 80/156]  eta: 0:00:59  lr: 0.007999  min_lr: 0.007999  loss: 4.2274 (3.9567)  weight_decay: 0.0500 (0.0500)  time: 0.7172  data: 0.0004  max mem: 57114
Epoch: [21]  [ 90/156]  eta: 0:00:51  lr: 0.007999  min_lr: 0.007999  loss: 4.1621 (3.9629)  weight_decay: 0.0500 (0.0500)  time: 0.7007  data: 0.0004  max mem: 57114
Epoch: [21]  [100/156]  eta: 0:00:43  lr: 0.007999  min_lr: 0.007999  loss: 4.1397 (3.9620)  weight_decay: 0.0500 (0.0500)  time: 0.6945  data: 0.0004  max mem: 57114
Epoch: [21]  [110/156]  eta: 0:00:35  lr: 0.007999  min_lr: 0.007999  loss: 4.1632 (3.9831)  weight_decay: 0.0500 (0.0500)  time: 0.6950  data: 0.0004  max mem: 57114
Epoch: [21]  [120/156]  eta: 0:00:27  lr: 0.007999  min_lr: 0.007999  loss: 4.0781 (3.9852)  weight_decay: 0.0500 (0.0500)  time: 0.7001  data: 0.0004  max mem: 57114
Epoch: [21]  [130/156]  eta: 0:00:19  lr: 0.007999  min_lr: 0.007999  loss: 4.0781 (3.9930)  weight_decay: 0.0500 (0.0500)  time: 0.7012  data: 0.0009  max mem: 57114
Epoch: [21]  [140/156]  eta: 0:00:11  lr: 0.007999  min_lr: 0.007999  loss: 4.2535 (3.9940)  weight_decay: 0.0500 (0.0500)  time: 0.6893  data: 0.0007  max mem: 57114
Epoch: [21]  [150/156]  eta: 0:00:04  lr: 0.007999  min_lr: 0.007999  loss: 4.1484 (3.9903)  weight_decay: 0.0500 (0.0500)  time: 0.6799  data: 0.0001  max mem: 57114
Epoch: [21]  [155/156]  eta: 0:00:00  lr: 0.007999  min_lr: 0.007999  loss: 4.0884 (3.9782)  weight_decay: 0.0500 (0.0500)  time: 0.6810  data: 0.0001  max mem: 57114
Epoch: [21] Total time: 0:01:56 (0.7443 s / it)
Averaged stats: lr: 0.007999  min_lr: 0.007999  loss: 4.0884 (4.0108)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 1.2933 (1.2933)  acc1: 73.9583 (73.9583)  acc5: 94.7917 (94.7917)  time: 2.2525  data: 1.9962  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.5546 (1.5008)  acc1: 64.5833 (65.0895)  acc5: 91.6667 (87.7238)  time: 0.6012  data: 0.3993  max mem: 57114
Test: Total time: 0:00:03 (0.6242 s / it)
* Acc@1 63.785 Acc@5 85.979 loss 1.653
Accuracy of the model on the 50000 test images: 63.8%
Max accuracy: 63.78%
Test:  [0/5]  eta: 0:00:10  loss: 6.9138 (6.9138)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.0000)  time: 2.0521  data: 1.8086  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9138 (6.9179)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.2558)  time: 0.5610  data: 0.3618  max mem: 57114
Test: Total time: 0:00:02 (0.5731 s / it)
* Acc@1 0.102 Acc@5 0.504 loss 6.917
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [22]  [  0/156]  eta: 0:13:26  lr: 0.007999  min_lr: 0.007999  loss: 4.5682 (4.5682)  weight_decay: 0.0500 (0.0500)  time: 5.1676  data: 4.1159  max mem: 57114
Epoch: [22]  [ 10/156]  eta: 0:02:45  lr: 0.007999  min_lr: 0.007999  loss: 3.6516 (3.9382)  weight_decay: 0.0500 (0.0500)  time: 1.1334  data: 0.3747  max mem: 57114
Epoch: [22]  [ 20/156]  eta: 0:02:08  lr: 0.007999  min_lr: 0.007999  loss: 3.8041 (3.9448)  weight_decay: 0.0500 (0.0500)  time: 0.7369  data: 0.0005  max mem: 57114
Epoch: [22]  [ 30/156]  eta: 0:01:51  lr: 0.007999  min_lr: 0.007999  loss: 4.2082 (4.0119)  weight_decay: 0.0500 (0.0500)  time: 0.7486  data: 0.0003  max mem: 57114
Epoch: [22]  [ 40/156]  eta: 0:01:38  lr: 0.007999  min_lr: 0.007999  loss: 3.7499 (3.9217)  weight_decay: 0.0500 (0.0500)  time: 0.7380  data: 0.0004  max mem: 57114
Epoch: [22]  [ 50/156]  eta: 0:01:27  lr: 0.007999  min_lr: 0.007999  loss: 3.7499 (3.9247)  weight_decay: 0.0500 (0.0500)  time: 0.7402  data: 0.0004  max mem: 57114
Epoch: [22]  [ 60/156]  eta: 0:01:17  lr: 0.007999  min_lr: 0.007999  loss: 3.9444 (3.9500)  weight_decay: 0.0500 (0.0500)  time: 0.7359  data: 0.0004  max mem: 57114
Epoch: [22]  [ 70/156]  eta: 0:01:08  lr: 0.007998  min_lr: 0.007998  loss: 4.0086 (3.9408)  weight_decay: 0.0500 (0.0500)  time: 0.7016  data: 0.0004  max mem: 57114
Epoch: [22]  [ 80/156]  eta: 0:00:59  lr: 0.007998  min_lr: 0.007998  loss: 4.0617 (3.9537)  weight_decay: 0.0500 (0.0500)  time: 0.6938  data: 0.0004  max mem: 57114
Epoch: [22]  [ 90/156]  eta: 0:00:50  lr: 0.007998  min_lr: 0.007998  loss: 3.8378 (3.9335)  weight_decay: 0.0500 (0.0500)  time: 0.6988  data: 0.0004  max mem: 57114
Epoch: [22]  [100/156]  eta: 0:00:42  lr: 0.007998  min_lr: 0.007998  loss: 3.8562 (3.9492)  weight_decay: 0.0500 (0.0500)  time: 0.7030  data: 0.0004  max mem: 57114
Epoch: [22]  [110/156]  eta: 0:00:34  lr: 0.007998  min_lr: 0.007998  loss: 3.9819 (3.9369)  weight_decay: 0.0500 (0.0500)  time: 0.7008  data: 0.0004  max mem: 57114
Epoch: [22]  [120/156]  eta: 0:00:27  lr: 0.007998  min_lr: 0.007998  loss: 3.9819 (3.9404)  weight_decay: 0.0500 (0.0500)  time: 0.7019  data: 0.0004  max mem: 57114
Epoch: [22]  [130/156]  eta: 0:00:19  lr: 0.007998  min_lr: 0.007998  loss: 4.3525 (3.9628)  weight_decay: 0.0500 (0.0500)  time: 0.7024  data: 0.0008  max mem: 57114
Epoch: [22]  [140/156]  eta: 0:00:11  lr: 0.007998  min_lr: 0.007998  loss: 4.1706 (3.9624)  weight_decay: 0.0500 (0.0500)  time: 0.6868  data: 0.0007  max mem: 57114
Epoch: [22]  [150/156]  eta: 0:00:04  lr: 0.007998  min_lr: 0.007998  loss: 4.1671 (3.9729)  weight_decay: 0.0500 (0.0500)  time: 0.6816  data: 0.0001  max mem: 57114
Epoch: [22]  [155/156]  eta: 0:00:00  lr: 0.007998  min_lr: 0.007998  loss: 4.2100 (3.9858)  weight_decay: 0.0500 (0.0500)  time: 0.6841  data: 0.0001  max mem: 57114
Epoch: [22] Total time: 0:01:55 (0.7417 s / it)
Averaged stats: lr: 0.007998  min_lr: 0.007998  loss: 4.2100 (3.9777)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.6116 (1.6116)  acc1: 70.3125 (70.3125)  acc5: 91.1458 (91.1458)  time: 2.1167  data: 1.8611  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.7703 (1.6374)  acc1: 67.1875 (65.8568)  acc5: 87.5000 (85.8056)  time: 0.5741  data: 0.3723  max mem: 57114
Test: Total time: 0:00:02 (0.5953 s / it)
* Acc@1 64.468 Acc@5 86.283 loss 1.784
Accuracy of the model on the 50000 test images: 64.5%
Max accuracy: 64.47%
Test:  [0/5]  eta: 0:00:10  loss: 6.9145 (6.9145)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.5208)  time: 2.0886  data: 1.8451  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9145 (6.9190)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.2558)  time: 0.5683  data: 0.3691  max mem: 57114
Test: Total time: 0:00:02 (0.5798 s / it)
* Acc@1 0.098 Acc@5 0.496 loss 6.917
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [23]  [  0/156]  eta: 0:13:33  lr: 0.007998  min_lr: 0.007998  loss: 4.2850 (4.2850)  weight_decay: 0.0500 (0.0500)  time: 5.2164  data: 3.6513  max mem: 57114
Epoch: [23]  [ 10/156]  eta: 0:02:49  lr: 0.007998  min_lr: 0.007998  loss: 3.9023 (3.8744)  weight_decay: 0.0500 (0.0500)  time: 1.1579  data: 0.3322  max mem: 57114
Epoch: [23]  [ 20/156]  eta: 0:02:10  lr: 0.007998  min_lr: 0.007998  loss: 3.9023 (3.8452)  weight_decay: 0.0500 (0.0500)  time: 0.7471  data: 0.0003  max mem: 57114
Epoch: [23]  [ 30/156]  eta: 0:01:52  lr: 0.007997  min_lr: 0.007997  loss: 4.1270 (3.9887)  weight_decay: 0.0500 (0.0500)  time: 0.7469  data: 0.0003  max mem: 57114
Epoch: [23]  [ 40/156]  eta: 0:01:39  lr: 0.007997  min_lr: 0.007997  loss: 4.2081 (4.0422)  weight_decay: 0.0500 (0.0500)  time: 0.7461  data: 0.0004  max mem: 57114
Epoch: [23]  [ 50/156]  eta: 0:01:28  lr: 0.007997  min_lr: 0.007997  loss: 4.1507 (4.0211)  weight_decay: 0.0500 (0.0500)  time: 0.7384  data: 0.0004  max mem: 57114
Epoch: [23]  [ 60/156]  eta: 0:01:18  lr: 0.007997  min_lr: 0.007997  loss: 4.0525 (4.0320)  weight_decay: 0.0500 (0.0500)  time: 0.7299  data: 0.0004  max mem: 57114
Epoch: [23]  [ 70/156]  eta: 0:01:08  lr: 0.007997  min_lr: 0.007997  loss: 4.0994 (4.0063)  weight_decay: 0.0500 (0.0500)  time: 0.7194  data: 0.0004  max mem: 57114
Epoch: [23]  [ 80/156]  eta: 0:00:59  lr: 0.007997  min_lr: 0.007997  loss: 4.0994 (4.0192)  weight_decay: 0.0500 (0.0500)  time: 0.7107  data: 0.0003  max mem: 57114
Epoch: [23]  [ 90/156]  eta: 0:00:51  lr: 0.007997  min_lr: 0.007997  loss: 3.9851 (4.0034)  weight_decay: 0.0500 (0.0500)  time: 0.7069  data: 0.0003  max mem: 57114
Epoch: [23]  [100/156]  eta: 0:00:43  lr: 0.007997  min_lr: 0.007997  loss: 3.9851 (4.0119)  weight_decay: 0.0500 (0.0500)  time: 0.7083  data: 0.0004  max mem: 57114
Epoch: [23]  [110/156]  eta: 0:00:35  lr: 0.007997  min_lr: 0.007997  loss: 4.1765 (3.9958)  weight_decay: 0.0500 (0.0500)  time: 0.6992  data: 0.0004  max mem: 57114
Epoch: [23]  [120/156]  eta: 0:00:27  lr: 0.007996  min_lr: 0.007996  loss: 4.0171 (3.9963)  weight_decay: 0.0500 (0.0500)  time: 0.7005  data: 0.0004  max mem: 57114
Epoch: [23]  [130/156]  eta: 0:00:19  lr: 0.007996  min_lr: 0.007996  loss: 4.0171 (4.0047)  weight_decay: 0.0500 (0.0500)  time: 0.6979  data: 0.0008  max mem: 57114
Epoch: [23]  [140/156]  eta: 0:00:11  lr: 0.007996  min_lr: 0.007996  loss: 4.3166 (4.0061)  weight_decay: 0.0500 (0.0500)  time: 0.6838  data: 0.0007  max mem: 57114
Epoch: [23]  [150/156]  eta: 0:00:04  lr: 0.007996  min_lr: 0.007996  loss: 4.0111 (4.0030)  weight_decay: 0.0500 (0.0500)  time: 0.6776  data: 0.0001  max mem: 57114
Epoch: [23]  [155/156]  eta: 0:00:00  lr: 0.007996  min_lr: 0.007996  loss: 4.1693 (4.0112)  weight_decay: 0.0500 (0.0500)  time: 0.6815  data: 0.0001  max mem: 57114
Epoch: [23] Total time: 0:01:56 (0.7450 s / it)
Averaged stats: lr: 0.007996  min_lr: 0.007996  loss: 4.1693 (3.9493)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.6036 (1.6036)  acc1: 66.6667 (66.6667)  acc5: 91.6667 (91.6667)  time: 2.0510  data: 1.7953  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.6036 (1.7054)  acc1: 66.6667 (63.0435)  acc5: 91.6667 (87.5959)  time: 0.5609  data: 0.3591  max mem: 57114
Test: Total time: 0:00:02 (0.5844 s / it)
* Acc@1 63.297 Acc@5 85.922 loss 1.748
Accuracy of the model on the 50000 test images: 63.3%
Max accuracy: 64.47%
Test:  [0/5]  eta: 0:00:11  loss: 6.9158 (6.9158)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.5208)  time: 2.2776  data: 2.0341  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9158 (6.9201)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.2558)  time: 0.6061  data: 0.4069  max mem: 57114
Test: Total time: 0:00:03 (0.6189 s / it)
* Acc@1 0.096 Acc@5 0.496 loss 6.919
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [24]  [  0/156]  eta: 0:13:33  lr: 0.007996  min_lr: 0.007996  loss: 4.1404 (4.1404)  weight_decay: 0.0500 (0.0500)  time: 5.2132  data: 3.6261  max mem: 57114
Epoch: [24]  [ 10/156]  eta: 0:02:46  lr: 0.007996  min_lr: 0.007996  loss: 3.7474 (3.7405)  weight_decay: 0.0500 (0.0500)  time: 1.1410  data: 0.3300  max mem: 57114
Epoch: [24]  [ 20/156]  eta: 0:02:08  lr: 0.007996  min_lr: 0.007996  loss: 4.0824 (3.9129)  weight_decay: 0.0500 (0.0500)  time: 0.7298  data: 0.0003  max mem: 57114
Epoch: [24]  [ 30/156]  eta: 0:01:49  lr: 0.007996  min_lr: 0.007996  loss: 4.1396 (3.9327)  weight_decay: 0.0500 (0.0500)  time: 0.7221  data: 0.0004  max mem: 57114
Epoch: [24]  [ 40/156]  eta: 0:01:36  lr: 0.007995  min_lr: 0.007995  loss: 4.1908 (3.9482)  weight_decay: 0.0500 (0.0500)  time: 0.7158  data: 0.0004  max mem: 57114
Epoch: [24]  [ 50/156]  eta: 0:01:25  lr: 0.007995  min_lr: 0.007995  loss: 4.0363 (3.9439)  weight_decay: 0.0500 (0.0500)  time: 0.7122  data: 0.0004  max mem: 57114
Epoch: [24]  [ 60/156]  eta: 0:01:16  lr: 0.007995  min_lr: 0.007995  loss: 3.9619 (3.9425)  weight_decay: 0.0500 (0.0500)  time: 0.7115  data: 0.0004  max mem: 57114
Epoch: [24]  [ 70/156]  eta: 0:01:07  lr: 0.007995  min_lr: 0.007995  loss: 3.9817 (3.9041)  weight_decay: 0.0500 (0.0500)  time: 0.7243  data: 0.0004  max mem: 57114
Epoch: [24]  [ 80/156]  eta: 0:00:58  lr: 0.007995  min_lr: 0.007995  loss: 4.0483 (3.9322)  weight_decay: 0.0500 (0.0500)  time: 0.7241  data: 0.0004  max mem: 57114
Epoch: [24]  [ 90/156]  eta: 0:00:50  lr: 0.007995  min_lr: 0.007995  loss: 4.2413 (3.9454)  weight_decay: 0.0500 (0.0500)  time: 0.7050  data: 0.0004  max mem: 57114
Epoch: [24]  [100/156]  eta: 0:00:42  lr: 0.007995  min_lr: 0.007995  loss: 4.0651 (3.9332)  weight_decay: 0.0500 (0.0500)  time: 0.6979  data: 0.0004  max mem: 57114
Epoch: [24]  [110/156]  eta: 0:00:34  lr: 0.007994  min_lr: 0.007994  loss: 4.0115 (3.9372)  weight_decay: 0.0500 (0.0500)  time: 0.7017  data: 0.0004  max mem: 57114
Epoch: [24]  [120/156]  eta: 0:00:27  lr: 0.007994  min_lr: 0.007994  loss: 3.9867 (3.9167)  weight_decay: 0.0500 (0.0500)  time: 0.7041  data: 0.0003  max mem: 57114
Epoch: [24]  [130/156]  eta: 0:00:19  lr: 0.007994  min_lr: 0.007994  loss: 3.9867 (3.9148)  weight_decay: 0.0500 (0.0500)  time: 0.6964  data: 0.0008  max mem: 57114
Epoch: [24]  [140/156]  eta: 0:00:11  lr: 0.007994  min_lr: 0.007994  loss: 4.0414 (3.9053)  weight_decay: 0.0500 (0.0500)  time: 0.6873  data: 0.0007  max mem: 57114
Epoch: [24]  [150/156]  eta: 0:00:04  lr: 0.007994  min_lr: 0.007994  loss: 3.9165 (3.9004)  weight_decay: 0.0500 (0.0500)  time: 0.6841  data: 0.0001  max mem: 57114
Epoch: [24]  [155/156]  eta: 0:00:00  lr: 0.007994  min_lr: 0.007994  loss: 3.9165 (3.8867)  weight_decay: 0.0500 (0.0500)  time: 0.6829  data: 0.0001  max mem: 57114
Epoch: [24] Total time: 0:01:55 (0.7388 s / it)
Averaged stats: lr: 0.007994  min_lr: 0.007994  loss: 3.9165 (3.9126)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.4251 (1.4251)  acc1: 72.3958 (72.3958)  acc5: 93.7500 (93.7500)  time: 2.0772  data: 1.8211  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.5030 (1.5455)  acc1: 69.7917 (66.1125)  acc5: 92.1875 (88.8747)  time: 0.5662  data: 0.3643  max mem: 57114
Test: Total time: 0:00:02 (0.5897 s / it)
* Acc@1 65.147 Acc@5 86.981 loss 1.692
Accuracy of the model on the 50000 test images: 65.1%
Max accuracy: 65.15%
Test:  [0/5]  eta: 0:00:10  loss: 6.9173 (6.9173)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.5208)  time: 2.0411  data: 1.7976  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9173 (6.9211)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.2558)  time: 0.5588  data: 0.3596  max mem: 57114
Test: Total time: 0:00:02 (0.5699 s / it)
* Acc@1 0.088 Acc@5 0.492 loss 6.920
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [25]  [  0/156]  eta: 0:13:25  lr: 0.007994  min_lr: 0.007994  loss: 4.2126 (4.2126)  weight_decay: 0.0500 (0.0500)  time: 5.1660  data: 3.8111  max mem: 57114
Epoch: [25]  [ 10/156]  eta: 0:02:45  lr: 0.007994  min_lr: 0.007994  loss: 3.9795 (3.9035)  weight_decay: 0.0500 (0.0500)  time: 1.1312  data: 0.3468  max mem: 57114
Epoch: [25]  [ 20/156]  eta: 0:02:06  lr: 0.007993  min_lr: 0.007993  loss: 3.9772 (3.9579)  weight_decay: 0.0500 (0.0500)  time: 0.7198  data: 0.0005  max mem: 57114
Epoch: [25]  [ 30/156]  eta: 0:01:49  lr: 0.007993  min_lr: 0.007993  loss: 3.8957 (3.8764)  weight_decay: 0.0500 (0.0500)  time: 0.7293  data: 0.0005  max mem: 57114
Epoch: [25]  [ 40/156]  eta: 0:01:37  lr: 0.007993  min_lr: 0.007993  loss: 3.6146 (3.8661)  weight_decay: 0.0500 (0.0500)  time: 0.7514  data: 0.0004  max mem: 57114
Epoch: [25]  [ 50/156]  eta: 0:01:27  lr: 0.007993  min_lr: 0.007993  loss: 3.7940 (3.8319)  weight_decay: 0.0500 (0.0500)  time: 0.7512  data: 0.0004  max mem: 57114
Epoch: [25]  [ 60/156]  eta: 0:01:18  lr: 0.007993  min_lr: 0.007993  loss: 3.7940 (3.8420)  weight_decay: 0.0500 (0.0500)  time: 0.7491  data: 0.0004  max mem: 57114
Epoch: [25]  [ 70/156]  eta: 0:01:08  lr: 0.007993  min_lr: 0.007993  loss: 4.1369 (3.8978)  weight_decay: 0.0500 (0.0500)  time: 0.7349  data: 0.0004  max mem: 57114
Epoch: [25]  [ 80/156]  eta: 0:00:59  lr: 0.007992  min_lr: 0.007992  loss: 4.0223 (3.9081)  weight_decay: 0.0500 (0.0500)  time: 0.7153  data: 0.0004  max mem: 57114
Epoch: [25]  [ 90/156]  eta: 0:00:51  lr: 0.007992  min_lr: 0.007992  loss: 3.8032 (3.9023)  weight_decay: 0.0500 (0.0500)  time: 0.7024  data: 0.0004  max mem: 57114
Epoch: [25]  [100/156]  eta: 0:00:43  lr: 0.007992  min_lr: 0.007992  loss: 3.7742 (3.8904)  weight_decay: 0.0500 (0.0500)  time: 0.6959  data: 0.0004  max mem: 57114
Epoch: [25]  [110/156]  eta: 0:00:35  lr: 0.007992  min_lr: 0.007992  loss: 3.5502 (3.8794)  weight_decay: 0.0500 (0.0500)  time: 0.7068  data: 0.0003  max mem: 57114
Epoch: [25]  [120/156]  eta: 0:00:27  lr: 0.007992  min_lr: 0.007992  loss: 3.7894 (3.8827)  weight_decay: 0.0500 (0.0500)  time: 0.7056  data: 0.0004  max mem: 57114
Epoch: [25]  [130/156]  eta: 0:00:19  lr: 0.007991  min_lr: 0.007991  loss: 3.9673 (3.8848)  weight_decay: 0.0500 (0.0500)  time: 0.6951  data: 0.0008  max mem: 57114
Epoch: [25]  [140/156]  eta: 0:00:11  lr: 0.007991  min_lr: 0.007991  loss: 3.9829 (3.8987)  weight_decay: 0.0500 (0.0500)  time: 0.6884  data: 0.0007  max mem: 57114
Epoch: [25]  [150/156]  eta: 0:00:04  lr: 0.007991  min_lr: 0.007991  loss: 4.1517 (3.9199)  weight_decay: 0.0500 (0.0500)  time: 0.6830  data: 0.0001  max mem: 57114
Epoch: [25]  [155/156]  eta: 0:00:00  lr: 0.007991  min_lr: 0.007991  loss: 4.1301 (3.9266)  weight_decay: 0.0500 (0.0500)  time: 0.6817  data: 0.0001  max mem: 57114
Epoch: [25] Total time: 0:01:56 (0.7452 s / it)
Averaged stats: lr: 0.007991  min_lr: 0.007991  loss: 4.1301 (3.8837)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.4728 (1.4728)  acc1: 72.3958 (72.3958)  acc5: 91.6667 (91.6667)  time: 2.0499  data: 1.7922  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.5900 (1.5778)  acc1: 70.3125 (65.8568)  acc5: 89.0625 (86.4450)  time: 0.5607  data: 0.3585  max mem: 57114
Test: Total time: 0:00:02 (0.5835 s / it)
* Acc@1 64.282 Acc@5 86.451 loss 1.703
Accuracy of the model on the 50000 test images: 64.3%
Max accuracy: 65.15%
Test:  [0/5]  eta: 0:00:11  loss: 6.9198 (6.9198)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.5208)  time: 2.2215  data: 1.9780  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9198 (6.9221)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.2558)  time: 0.5949  data: 0.3957  max mem: 57114
Test: Total time: 0:00:03 (0.6068 s / it)
* Acc@1 0.090 Acc@5 0.484 loss 6.920
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [26]  [  0/156]  eta: 0:13:31  lr: 0.007991  min_lr: 0.007991  loss: 3.9487 (3.9487)  weight_decay: 0.0500 (0.0500)  time: 5.2004  data: 3.8813  max mem: 57114
Epoch: [26]  [ 10/156]  eta: 0:02:42  lr: 0.007991  min_lr: 0.007991  loss: 3.2667 (3.5472)  weight_decay: 0.0500 (0.0500)  time: 1.1142  data: 0.3531  max mem: 57114
Epoch: [26]  [ 20/156]  eta: 0:02:07  lr: 0.007991  min_lr: 0.007991  loss: 3.2667 (3.6112)  weight_decay: 0.0500 (0.0500)  time: 0.7235  data: 0.0003  max mem: 57114
Epoch: [26]  [ 30/156]  eta: 0:01:49  lr: 0.007990  min_lr: 0.007990  loss: 3.4242 (3.5425)  weight_decay: 0.0500 (0.0500)  time: 0.7291  data: 0.0003  max mem: 57114
Epoch: [26]  [ 40/156]  eta: 0:01:36  lr: 0.007990  min_lr: 0.007990  loss: 3.7398 (3.6327)  weight_decay: 0.0500 (0.0500)  time: 0.7240  data: 0.0004  max mem: 57114
Epoch: [26]  [ 50/156]  eta: 0:01:26  lr: 0.007990  min_lr: 0.007990  loss: 3.7710 (3.6103)  weight_decay: 0.0500 (0.0500)  time: 0.7295  data: 0.0004  max mem: 57114
Epoch: [26]  [ 60/156]  eta: 0:01:16  lr: 0.007990  min_lr: 0.007990  loss: 3.7517 (3.6553)  weight_decay: 0.0500 (0.0500)  time: 0.7262  data: 0.0004  max mem: 57114
Epoch: [26]  [ 70/156]  eta: 0:01:07  lr: 0.007990  min_lr: 0.007990  loss: 3.8275 (3.6950)  weight_decay: 0.0500 (0.0500)  time: 0.7065  data: 0.0004  max mem: 57114
Epoch: [26]  [ 80/156]  eta: 0:00:58  lr: 0.007989  min_lr: 0.007989  loss: 3.7945 (3.6913)  weight_decay: 0.0500 (0.0500)  time: 0.7033  data: 0.0004  max mem: 57114
Epoch: [26]  [ 90/156]  eta: 0:00:50  lr: 0.007989  min_lr: 0.007989  loss: 3.7708 (3.6993)  weight_decay: 0.0500 (0.0500)  time: 0.7101  data: 0.0003  max mem: 57114
Epoch: [26]  [100/156]  eta: 0:00:42  lr: 0.007989  min_lr: 0.007989  loss: 3.9531 (3.7295)  weight_decay: 0.0500 (0.0500)  time: 0.7049  data: 0.0003  max mem: 57114
Epoch: [26]  [110/156]  eta: 0:00:34  lr: 0.007989  min_lr: 0.007989  loss: 4.0270 (3.7453)  weight_decay: 0.0500 (0.0500)  time: 0.7173  data: 0.0004  max mem: 57114
Epoch: [26]  [120/156]  eta: 0:00:27  lr: 0.007988  min_lr: 0.007988  loss: 4.0129 (3.7761)  weight_decay: 0.0500 (0.0500)  time: 0.7099  data: 0.0004  max mem: 57114
Epoch: [26]  [130/156]  eta: 0:00:19  lr: 0.007988  min_lr: 0.007988  loss: 4.0129 (3.7778)  weight_decay: 0.0500 (0.0500)  time: 0.6950  data: 0.0009  max mem: 57114
Epoch: [26]  [140/156]  eta: 0:00:11  lr: 0.007988  min_lr: 0.007988  loss: 3.8276 (3.7750)  weight_decay: 0.0500 (0.0500)  time: 0.6919  data: 0.0008  max mem: 57114
Epoch: [26]  [150/156]  eta: 0:00:04  lr: 0.007988  min_lr: 0.007988  loss: 3.7458 (3.7771)  weight_decay: 0.0500 (0.0500)  time: 0.6820  data: 0.0001  max mem: 57114
Epoch: [26]  [155/156]  eta: 0:00:00  lr: 0.007988  min_lr: 0.007988  loss: 3.8459 (3.7806)  weight_decay: 0.0500 (0.0500)  time: 0.6829  data: 0.0001  max mem: 57114
Epoch: [26] Total time: 0:01:55 (0.7401 s / it)
Averaged stats: lr: 0.007988  min_lr: 0.007988  loss: 3.8459 (3.8570)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.3004 (1.3004)  acc1: 74.4792 (74.4792)  acc5: 94.7917 (94.7917)  time: 2.1010  data: 1.8455  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.5600 (1.4948)  acc1: 68.7500 (66.1125)  acc5: 90.6250 (87.7238)  time: 0.5709  data: 0.3692  max mem: 57114
Test: Total time: 0:00:02 (0.5937 s / it)
* Acc@1 64.922 Acc@5 86.943 loss 1.597
Accuracy of the model on the 50000 test images: 64.9%
Max accuracy: 65.15%
Test:  [0/5]  eta: 0:00:11  loss: 6.9223 (6.9223)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.5208)  time: 2.3030  data: 2.0596  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9223 (6.9232)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.2558)  time: 0.6112  data: 0.4120  max mem: 57114
Test: Total time: 0:00:03 (0.6256 s / it)
* Acc@1 0.092 Acc@5 0.482 loss 6.921
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [27]  [  0/156]  eta: 0:13:57  lr: 0.007988  min_lr: 0.007988  loss: 3.1739 (3.1739)  weight_decay: 0.0500 (0.0500)  time: 5.3714  data: 3.8088  max mem: 57114
Epoch: [27]  [ 10/156]  eta: 0:02:47  lr: 0.007987  min_lr: 0.007987  loss: 4.0995 (3.8914)  weight_decay: 0.0500 (0.0500)  time: 1.1460  data: 0.3465  max mem: 57114
Epoch: [27]  [ 20/156]  eta: 0:02:08  lr: 0.007987  min_lr: 0.007987  loss: 3.9258 (3.7148)  weight_decay: 0.0500 (0.0500)  time: 0.7199  data: 0.0003  max mem: 57114
Epoch: [27]  [ 30/156]  eta: 0:01:50  lr: 0.007987  min_lr: 0.007987  loss: 3.7404 (3.7730)  weight_decay: 0.0500 (0.0500)  time: 0.7248  data: 0.0003  max mem: 57114
Epoch: [27]  [ 40/156]  eta: 0:01:36  lr: 0.007987  min_lr: 0.007987  loss: 3.9083 (3.8083)  weight_decay: 0.0500 (0.0500)  time: 0.7251  data: 0.0003  max mem: 57114
Epoch: [27]  [ 50/156]  eta: 0:01:25  lr: 0.007987  min_lr: 0.007987  loss: 3.9083 (3.8145)  weight_decay: 0.0500 (0.0500)  time: 0.7135  data: 0.0003  max mem: 57114
Epoch: [27]  [ 60/156]  eta: 0:01:16  lr: 0.007986  min_lr: 0.007986  loss: 3.7070 (3.7857)  weight_decay: 0.0500 (0.0500)  time: 0.7072  data: 0.0003  max mem: 57114
Epoch: [27]  [ 70/156]  eta: 0:01:07  lr: 0.007986  min_lr: 0.007986  loss: 3.7851 (3.7948)  weight_decay: 0.0500 (0.0500)  time: 0.7049  data: 0.0004  max mem: 57114
Epoch: [27]  [ 80/156]  eta: 0:00:58  lr: 0.007986  min_lr: 0.007986  loss: 3.8384 (3.7953)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0004  max mem: 57114
Epoch: [27]  [ 90/156]  eta: 0:00:50  lr: 0.007986  min_lr: 0.007986  loss: 3.6555 (3.7831)  weight_decay: 0.0500 (0.0500)  time: 0.7110  data: 0.0004  max mem: 57114
Epoch: [27]  [100/156]  eta: 0:00:42  lr: 0.007985  min_lr: 0.007985  loss: 3.9999 (3.8158)  weight_decay: 0.0500 (0.0500)  time: 0.7094  data: 0.0005  max mem: 57114
Epoch: [27]  [110/156]  eta: 0:00:34  lr: 0.007985  min_lr: 0.007985  loss: 4.0418 (3.8278)  weight_decay: 0.0500 (0.0500)  time: 0.7087  data: 0.0005  max mem: 57114
Epoch: [27]  [120/156]  eta: 0:00:26  lr: 0.007985  min_lr: 0.007985  loss: 4.0402 (3.8251)  weight_decay: 0.0500 (0.0500)  time: 0.6962  data: 0.0004  max mem: 57114
Epoch: [27]  [130/156]  eta: 0:00:19  lr: 0.007985  min_lr: 0.007985  loss: 3.8465 (3.8032)  weight_decay: 0.0500 (0.0500)  time: 0.6870  data: 0.0008  max mem: 57114
Epoch: [27]  [140/156]  eta: 0:00:11  lr: 0.007984  min_lr: 0.007984  loss: 3.8465 (3.7975)  weight_decay: 0.0500 (0.0500)  time: 0.6832  data: 0.0007  max mem: 57114
Epoch: [27]  [150/156]  eta: 0:00:04  lr: 0.007984  min_lr: 0.007984  loss: 3.9235 (3.7977)  weight_decay: 0.0500 (0.0500)  time: 0.6785  data: 0.0001  max mem: 57114
Epoch: [27]  [155/156]  eta: 0:00:00  lr: 0.007984  min_lr: 0.007984  loss: 3.9803 (3.7971)  weight_decay: 0.0500 (0.0500)  time: 0.6796  data: 0.0001  max mem: 57114
Epoch: [27] Total time: 0:01:54 (0.7367 s / it)
Averaged stats: lr: 0.007984  min_lr: 0.007984  loss: 3.9803 (3.8432)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.2778 (1.2778)  acc1: 78.1250 (78.1250)  acc5: 93.7500 (93.7500)  time: 2.1081  data: 1.8525  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.4551 (1.4668)  acc1: 71.8750 (69.0537)  acc5: 90.6250 (89.2583)  time: 0.5722  data: 0.3706  max mem: 57114
Test: Total time: 0:00:02 (0.5942 s / it)
* Acc@1 66.764 Acc@5 87.882 loss 1.578
Accuracy of the model on the 50000 test images: 66.8%
Max accuracy: 66.76%
Test:  [0/5]  eta: 0:00:10  loss: 6.9241 (6.9241)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.5208)  time: 2.0302  data: 1.7865  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9241 (6.9248)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.1279)  time: 0.5567  data: 0.3574  max mem: 57114
Test: Total time: 0:00:02 (0.5669 s / it)
* Acc@1 0.092 Acc@5 0.488 loss 6.923
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [28]  [  0/156]  eta: 0:13:41  lr: 0.007984  min_lr: 0.007984  loss: 4.0943 (4.0943)  weight_decay: 0.0500 (0.0500)  time: 5.2656  data: 4.1147  max mem: 57114
Epoch: [28]  [ 10/156]  eta: 0:02:53  lr: 0.007984  min_lr: 0.007984  loss: 4.0293 (3.8238)  weight_decay: 0.0500 (0.0500)  time: 1.1882  data: 0.3744  max mem: 57114
Epoch: [28]  [ 20/156]  eta: 0:02:10  lr: 0.007983  min_lr: 0.007983  loss: 4.0149 (3.8182)  weight_decay: 0.0500 (0.0500)  time: 0.7455  data: 0.0004  max mem: 57114
Epoch: [28]  [ 30/156]  eta: 0:01:50  lr: 0.007983  min_lr: 0.007983  loss: 3.9766 (3.8891)  weight_decay: 0.0500 (0.0500)  time: 0.7096  data: 0.0004  max mem: 57114
Epoch: [28]  [ 40/156]  eta: 0:01:38  lr: 0.007983  min_lr: 0.007983  loss: 3.9766 (3.9157)  weight_decay: 0.0500 (0.0500)  time: 0.7274  data: 0.0004  max mem: 57114
Epoch: [28]  [ 50/156]  eta: 0:01:26  lr: 0.007983  min_lr: 0.007983  loss: 3.8478 (3.8458)  weight_decay: 0.0500 (0.0500)  time: 0.7283  data: 0.0003  max mem: 57114
Epoch: [28]  [ 60/156]  eta: 0:01:17  lr: 0.007982  min_lr: 0.007982  loss: 3.5695 (3.8396)  weight_decay: 0.0500 (0.0500)  time: 0.7204  data: 0.0004  max mem: 57114
Epoch: [28]  [ 70/156]  eta: 0:01:08  lr: 0.007982  min_lr: 0.007982  loss: 4.0957 (3.8609)  weight_decay: 0.0500 (0.0500)  time: 0.7350  data: 0.0004  max mem: 57114
Epoch: [28]  [ 80/156]  eta: 0:00:59  lr: 0.007982  min_lr: 0.007982  loss: 4.1326 (3.8706)  weight_decay: 0.0500 (0.0500)  time: 0.7316  data: 0.0004  max mem: 57114
Epoch: [28]  [ 90/156]  eta: 0:00:51  lr: 0.007981  min_lr: 0.007981  loss: 3.9249 (3.8332)  weight_decay: 0.0500 (0.0500)  time: 0.7132  data: 0.0004  max mem: 57114
Epoch: [28]  [100/156]  eta: 0:00:43  lr: 0.007981  min_lr: 0.007981  loss: 3.9615 (3.8547)  weight_decay: 0.0500 (0.0500)  time: 0.6967  data: 0.0004  max mem: 57114
Epoch: [28]  [110/156]  eta: 0:00:35  lr: 0.007981  min_lr: 0.007981  loss: 4.1304 (3.8606)  weight_decay: 0.0500 (0.0500)  time: 0.6938  data: 0.0004  max mem: 57114
Epoch: [28]  [120/156]  eta: 0:00:27  lr: 0.007981  min_lr: 0.007981  loss: 4.0432 (3.8563)  weight_decay: 0.0500 (0.0500)  time: 0.6957  data: 0.0004  max mem: 57114
Epoch: [28]  [130/156]  eta: 0:00:19  lr: 0.007980  min_lr: 0.007980  loss: 3.9167 (3.8439)  weight_decay: 0.0500 (0.0500)  time: 0.6912  data: 0.0005  max mem: 57114
Epoch: [28]  [140/156]  eta: 0:00:11  lr: 0.007980  min_lr: 0.007980  loss: 3.9521 (3.8470)  weight_decay: 0.0500 (0.0500)  time: 0.6853  data: 0.0003  max mem: 57114
Epoch: [28]  [150/156]  eta: 0:00:04  lr: 0.007980  min_lr: 0.007980  loss: 3.9595 (3.8371)  weight_decay: 0.0500 (0.0500)  time: 0.6850  data: 0.0001  max mem: 57114
Epoch: [28]  [155/156]  eta: 0:00:00  lr: 0.007980  min_lr: 0.007980  loss: 4.0059 (3.8384)  weight_decay: 0.0500 (0.0500)  time: 0.6874  data: 0.0001  max mem: 57114
Epoch: [28] Total time: 0:01:56 (0.7437 s / it)
Averaged stats: lr: 0.007980  min_lr: 0.007980  loss: 4.0059 (3.8187)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.3139 (1.3139)  acc1: 74.4792 (74.4792)  acc5: 90.1042 (90.1042)  time: 2.0650  data: 1.8094  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.6599 (1.5776)  acc1: 71.3542 (67.3913)  acc5: 89.5833 (86.8286)  time: 0.5637  data: 0.3620  max mem: 57114
Test: Total time: 0:00:02 (0.5886 s / it)
* Acc@1 65.617 Acc@5 86.883 loss 1.658
Accuracy of the model on the 50000 test images: 65.6%
Max accuracy: 66.76%
Test:  [0/5]  eta: 0:00:11  loss: 6.9262 (6.9262)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.2900  data: 2.0465  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9262 (6.9262)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.2558)  time: 0.6086  data: 0.4094  max mem: 57114
Test: Total time: 0:00:03 (0.6202 s / it)
* Acc@1 0.096 Acc@5 0.480 loss 6.924
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [29]  [  0/156]  eta: 0:12:59  lr: 0.007980  min_lr: 0.007980  loss: 2.7767 (2.7767)  weight_decay: 0.0500 (0.0500)  time: 4.9962  data: 3.5853  max mem: 57114
Epoch: [29]  [ 10/156]  eta: 0:02:40  lr: 0.007979  min_lr: 0.007979  loss: 4.0994 (3.9965)  weight_decay: 0.0500 (0.0500)  time: 1.0977  data: 0.3262  max mem: 57114
Epoch: [29]  [ 20/156]  eta: 0:02:04  lr: 0.007979  min_lr: 0.007979  loss: 4.0306 (3.7974)  weight_decay: 0.0500 (0.0500)  time: 0.7080  data: 0.0003  max mem: 57114
Epoch: [29]  [ 30/156]  eta: 0:01:46  lr: 0.007979  min_lr: 0.007979  loss: 3.7502 (3.8583)  weight_decay: 0.0500 (0.0500)  time: 0.7007  data: 0.0003  max mem: 57114
Epoch: [29]  [ 40/156]  eta: 0:01:34  lr: 0.007978  min_lr: 0.007978  loss: 3.9325 (3.8727)  weight_decay: 0.0500 (0.0500)  time: 0.7054  data: 0.0004  max mem: 57114
Epoch: [29]  [ 50/156]  eta: 0:01:24  lr: 0.007978  min_lr: 0.007978  loss: 3.9080 (3.8426)  weight_decay: 0.0500 (0.0500)  time: 0.7276  data: 0.0004  max mem: 57114
Epoch: [29]  [ 60/156]  eta: 0:01:15  lr: 0.007978  min_lr: 0.007978  loss: 3.9287 (3.8769)  weight_decay: 0.0500 (0.0500)  time: 0.7381  data: 0.0004  max mem: 57114
Epoch: [29]  [ 70/156]  eta: 0:01:06  lr: 0.007978  min_lr: 0.007978  loss: 3.9287 (3.8769)  weight_decay: 0.0500 (0.0500)  time: 0.7230  data: 0.0004  max mem: 57114
Epoch: [29]  [ 80/156]  eta: 0:00:58  lr: 0.007977  min_lr: 0.007977  loss: 3.9033 (3.8793)  weight_decay: 0.0500 (0.0500)  time: 0.7089  data: 0.0004  max mem: 57114
Epoch: [29]  [ 90/156]  eta: 0:00:50  lr: 0.007977  min_lr: 0.007977  loss: 3.8942 (3.8557)  weight_decay: 0.0500 (0.0500)  time: 0.6990  data: 0.0004  max mem: 57114
Epoch: [29]  [100/156]  eta: 0:00:42  lr: 0.007977  min_lr: 0.007977  loss: 3.8193 (3.8431)  weight_decay: 0.0500 (0.0500)  time: 0.6957  data: 0.0004  max mem: 57114
Epoch: [29]  [110/156]  eta: 0:00:34  lr: 0.007976  min_lr: 0.007976  loss: 3.9441 (3.8464)  weight_decay: 0.0500 (0.0500)  time: 0.6989  data: 0.0004  max mem: 57114
Epoch: [29]  [120/156]  eta: 0:00:26  lr: 0.007976  min_lr: 0.007976  loss: 3.8500 (3.8474)  weight_decay: 0.0500 (0.0500)  time: 0.6986  data: 0.0004  max mem: 57114
Epoch: [29]  [130/156]  eta: 0:00:19  lr: 0.007976  min_lr: 0.007976  loss: 3.8500 (3.8515)  weight_decay: 0.0500 (0.0500)  time: 0.6971  data: 0.0009  max mem: 57114
Epoch: [29]  [140/156]  eta: 0:00:11  lr: 0.007975  min_lr: 0.007975  loss: 4.0755 (3.8607)  weight_decay: 0.0500 (0.0500)  time: 0.6825  data: 0.0008  max mem: 57114
Epoch: [29]  [150/156]  eta: 0:00:04  lr: 0.007975  min_lr: 0.007975  loss: 4.0132 (3.8714)  weight_decay: 0.0500 (0.0500)  time: 0.6776  data: 0.0001  max mem: 57114
Epoch: [29]  [155/156]  eta: 0:00:00  lr: 0.007975  min_lr: 0.007975  loss: 3.9953 (3.8654)  weight_decay: 0.0500 (0.0500)  time: 0.6789  data: 0.0001  max mem: 57114
Epoch: [29] Total time: 0:01:54 (0.7323 s / it)
Averaged stats: lr: 0.007975  min_lr: 0.007975  loss: 3.9953 (3.7956)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.6090 (1.6090)  acc1: 75.5208 (75.5208)  acc5: 93.2292 (93.2292)  time: 2.0708  data: 1.8147  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.8204 (1.7830)  acc1: 63.5417 (63.0435)  acc5: 89.5833 (86.8286)  time: 0.5648  data: 0.3630  max mem: 57114
Test: Total time: 0:00:02 (0.5850 s / it)
* Acc@1 63.999 Acc@5 86.225 loss 1.884
Accuracy of the model on the 50000 test images: 64.0%
Max accuracy: 66.76%
Test:  [0/5]  eta: 0:00:11  loss: 6.9285 (6.9285)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.2696  data: 2.0261  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9285 (6.9275)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.2558)  time: 0.6045  data: 0.4053  max mem: 57114
Test: Total time: 0:00:03 (0.6157 s / it)
* Acc@1 0.098 Acc@5 0.472 loss 6.925
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [30]  [  0/156]  eta: 0:13:14  lr: 0.007975  min_lr: 0.007975  loss: 3.9167 (3.9167)  weight_decay: 0.0500 (0.0500)  time: 5.0952  data: 3.8931  max mem: 57114
Epoch: [30]  [ 10/156]  eta: 0:02:45  lr: 0.007975  min_lr: 0.007975  loss: 3.9715 (3.8715)  weight_decay: 0.0500 (0.0500)  time: 1.1330  data: 0.3542  max mem: 57114
Epoch: [30]  [ 20/156]  eta: 0:02:07  lr: 0.007974  min_lr: 0.007974  loss: 4.1074 (4.0107)  weight_decay: 0.0500 (0.0500)  time: 0.7270  data: 0.0003  max mem: 57114
Epoch: [30]  [ 30/156]  eta: 0:01:50  lr: 0.007974  min_lr: 0.007974  loss: 3.8612 (3.9097)  weight_decay: 0.0500 (0.0500)  time: 0.7337  data: 0.0004  max mem: 57114
Epoch: [30]  [ 40/156]  eta: 0:01:37  lr: 0.007974  min_lr: 0.007974  loss: 3.9601 (3.8988)  weight_decay: 0.0500 (0.0500)  time: 0.7405  data: 0.0004  max mem: 57114
Epoch: [30]  [ 50/156]  eta: 0:01:26  lr: 0.007973  min_lr: 0.007973  loss: 4.0111 (3.8440)  weight_decay: 0.0500 (0.0500)  time: 0.7196  data: 0.0004  max mem: 57114
Epoch: [30]  [ 60/156]  eta: 0:01:16  lr: 0.007973  min_lr: 0.007973  loss: 3.8022 (3.8480)  weight_decay: 0.0500 (0.0500)  time: 0.7109  data: 0.0004  max mem: 57114
Epoch: [30]  [ 70/156]  eta: 0:01:07  lr: 0.007973  min_lr: 0.007973  loss: 4.0092 (3.8435)  weight_decay: 0.0500 (0.0500)  time: 0.7112  data: 0.0004  max mem: 57114
Epoch: [30]  [ 80/156]  eta: 0:00:58  lr: 0.007972  min_lr: 0.007972  loss: 3.9162 (3.8069)  weight_decay: 0.0500 (0.0500)  time: 0.7008  data: 0.0004  max mem: 57114
Epoch: [30]  [ 90/156]  eta: 0:00:50  lr: 0.007972  min_lr: 0.007972  loss: 3.7935 (3.8057)  weight_decay: 0.0500 (0.0500)  time: 0.6932  data: 0.0004  max mem: 57114
Epoch: [30]  [100/156]  eta: 0:00:42  lr: 0.007972  min_lr: 0.007972  loss: 3.7712 (3.8077)  weight_decay: 0.0500 (0.0500)  time: 0.6995  data: 0.0004  max mem: 57114
Epoch: [30]  [110/156]  eta: 0:00:34  lr: 0.007971  min_lr: 0.007971  loss: 3.9713 (3.8170)  weight_decay: 0.0500 (0.0500)  time: 0.7044  data: 0.0004  max mem: 57114
Epoch: [30]  [120/156]  eta: 0:00:26  lr: 0.007971  min_lr: 0.007971  loss: 3.9713 (3.8164)  weight_decay: 0.0500 (0.0500)  time: 0.7019  data: 0.0004  max mem: 57114
Epoch: [30]  [130/156]  eta: 0:00:19  lr: 0.007970  min_lr: 0.007970  loss: 3.8336 (3.8117)  weight_decay: 0.0500 (0.0500)  time: 0.6924  data: 0.0009  max mem: 57114
Epoch: [30]  [140/156]  eta: 0:00:11  lr: 0.007970  min_lr: 0.007970  loss: 3.8460 (3.8135)  weight_decay: 0.0500 (0.0500)  time: 0.6827  data: 0.0007  max mem: 57114
Epoch: [30]  [150/156]  eta: 0:00:04  lr: 0.007970  min_lr: 0.007970  loss: 3.8431 (3.8036)  weight_decay: 0.0500 (0.0500)  time: 0.6809  data: 0.0001  max mem: 57114
Epoch: [30]  [155/156]  eta: 0:00:00  lr: 0.007970  min_lr: 0.007970  loss: 3.7716 (3.8097)  weight_decay: 0.0500 (0.0500)  time: 0.6819  data: 0.0001  max mem: 57114
Epoch: [30] Total time: 0:01:54 (0.7369 s / it)
Averaged stats: lr: 0.007970  min_lr: 0.007970  loss: 3.7716 (3.7764)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.3972 (1.3972)  acc1: 72.3958 (72.3958)  acc5: 92.7083 (92.7083)  time: 2.1536  data: 1.8981  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.5573 (1.5089)  acc1: 71.3542 (66.7519)  acc5: 91.1458 (88.6189)  time: 0.5813  data: 0.3797  max mem: 57114
Test: Total time: 0:00:03 (0.6056 s / it)
* Acc@1 65.072 Acc@5 87.302 loss 1.654
Accuracy of the model on the 50000 test images: 65.1%
Max accuracy: 66.76%
Test:  [0/5]  eta: 0:00:11  loss: 6.9323 (6.9323)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.3620  data: 2.1185  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9308 (6.9290)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.2558)  time: 0.6231  data: 0.4238  max mem: 57114
Test: Total time: 0:00:03 (0.6346 s / it)
* Acc@1 0.098 Acc@5 0.470 loss 6.926
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [31]  [  0/156]  eta: 0:12:54  lr: 0.007970  min_lr: 0.007970  loss: 4.5630 (4.5630)  weight_decay: 0.0500 (0.0500)  time: 4.9625  data: 3.9091  max mem: 57114
Epoch: [31]  [ 10/156]  eta: 0:02:43  lr: 0.007969  min_lr: 0.007969  loss: 3.9857 (3.7954)  weight_decay: 0.0500 (0.0500)  time: 1.1231  data: 0.3557  max mem: 57114
Epoch: [31]  [ 20/156]  eta: 0:02:05  lr: 0.007969  min_lr: 0.007969  loss: 3.7119 (3.6234)  weight_decay: 0.0500 (0.0500)  time: 0.7234  data: 0.0003  max mem: 57114
Epoch: [31]  [ 30/156]  eta: 0:01:48  lr: 0.007969  min_lr: 0.007969  loss: 3.3602 (3.5191)  weight_decay: 0.0500 (0.0500)  time: 0.7197  data: 0.0003  max mem: 57114
Epoch: [31]  [ 40/156]  eta: 0:01:36  lr: 0.007968  min_lr: 0.007968  loss: 3.5574 (3.5746)  weight_decay: 0.0500 (0.0500)  time: 0.7256  data: 0.0004  max mem: 57114
Epoch: [31]  [ 50/156]  eta: 0:01:25  lr: 0.007968  min_lr: 0.007968  loss: 3.9190 (3.6047)  weight_decay: 0.0500 (0.0500)  time: 0.7103  data: 0.0004  max mem: 57114
Epoch: [31]  [ 60/156]  eta: 0:01:16  lr: 0.007967  min_lr: 0.007967  loss: 3.6770 (3.5983)  weight_decay: 0.0500 (0.0500)  time: 0.7179  data: 0.0004  max mem: 57114
Epoch: [31]  [ 70/156]  eta: 0:01:06  lr: 0.007967  min_lr: 0.007967  loss: 3.7717 (3.6339)  weight_decay: 0.0500 (0.0500)  time: 0.7163  data: 0.0004  max mem: 57114
Epoch: [31]  [ 80/156]  eta: 0:00:58  lr: 0.007967  min_lr: 0.007967  loss: 3.8264 (3.6386)  weight_decay: 0.0500 (0.0500)  time: 0.7066  data: 0.0004  max mem: 57114
Epoch: [31]  [ 90/156]  eta: 0:00:50  lr: 0.007966  min_lr: 0.007966  loss: 3.7124 (3.6486)  weight_decay: 0.0500 (0.0500)  time: 0.7106  data: 0.0004  max mem: 57114
Epoch: [31]  [100/156]  eta: 0:00:42  lr: 0.007966  min_lr: 0.007966  loss: 3.9621 (3.6906)  weight_decay: 0.0500 (0.0500)  time: 0.7110  data: 0.0004  max mem: 57114
Epoch: [31]  [110/156]  eta: 0:00:34  lr: 0.007966  min_lr: 0.007966  loss: 4.0759 (3.7058)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0004  max mem: 57114
Epoch: [31]  [120/156]  eta: 0:00:26  lr: 0.007965  min_lr: 0.007965  loss: 4.0728 (3.7140)  weight_decay: 0.0500 (0.0500)  time: 0.7023  data: 0.0004  max mem: 57114
Epoch: [31]  [130/156]  eta: 0:00:19  lr: 0.007965  min_lr: 0.007965  loss: 3.9085 (3.7064)  weight_decay: 0.0500 (0.0500)  time: 0.6977  data: 0.0008  max mem: 57114
Epoch: [31]  [140/156]  eta: 0:00:11  lr: 0.007964  min_lr: 0.007964  loss: 3.9463 (3.7239)  weight_decay: 0.0500 (0.0500)  time: 0.6871  data: 0.0007  max mem: 57114
Epoch: [31]  [150/156]  eta: 0:00:04  lr: 0.007964  min_lr: 0.007964  loss: 4.0518 (3.7420)  weight_decay: 0.0500 (0.0500)  time: 0.6807  data: 0.0001  max mem: 57114
Epoch: [31]  [155/156]  eta: 0:00:00  lr: 0.007964  min_lr: 0.007964  loss: 3.9463 (3.7370)  weight_decay: 0.0500 (0.0500)  time: 0.6856  data: 0.0001  max mem: 57114
Epoch: [31] Total time: 0:01:55 (0.7374 s / it)
Averaged stats: lr: 0.007964  min_lr: 0.007964  loss: 3.9463 (3.7560)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.5930 (1.5930)  acc1: 71.3542 (71.3542)  acc5: 92.1875 (92.1875)  time: 2.1231  data: 1.8676  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.6890 (1.6459)  acc1: 69.7917 (66.4962)  acc5: 89.5833 (86.9565)  time: 0.5753  data: 0.3736  max mem: 57114
Test: Total time: 0:00:02 (0.5988 s / it)
* Acc@1 66.027 Acc@5 87.188 loss 1.738
Accuracy of the model on the 50000 test images: 66.0%
Max accuracy: 66.76%
Test:  [0/5]  eta: 0:00:11  loss: 6.9361 (6.9361)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.3802  data: 2.1367  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9328 (6.9308)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.2558)  time: 0.6267  data: 0.4274  max mem: 57114
Test: Total time: 0:00:03 (0.6398 s / it)
* Acc@1 0.094 Acc@5 0.470 loss 6.928
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [32]  [  0/156]  eta: 0:12:38  lr: 0.007964  min_lr: 0.007964  loss: 3.6025 (3.6025)  weight_decay: 0.0500 (0.0500)  time: 4.8632  data: 3.1122  max mem: 57114
Epoch: [32]  [ 10/156]  eta: 0:02:40  lr: 0.007963  min_lr: 0.007963  loss: 3.9496 (3.7952)  weight_decay: 0.0500 (0.0500)  time: 1.0977  data: 0.2833  max mem: 57114
Epoch: [32]  [ 20/156]  eta: 0:02:04  lr: 0.007963  min_lr: 0.007963  loss: 3.9496 (3.8267)  weight_decay: 0.0500 (0.0500)  time: 0.7157  data: 0.0004  max mem: 57114
Epoch: [32]  [ 30/156]  eta: 0:01:47  lr: 0.007963  min_lr: 0.007963  loss: 3.7261 (3.7474)  weight_decay: 0.0500 (0.0500)  time: 0.7133  data: 0.0003  max mem: 57114
Epoch: [32]  [ 40/156]  eta: 0:01:35  lr: 0.007962  min_lr: 0.007962  loss: 3.7250 (3.7110)  weight_decay: 0.0500 (0.0500)  time: 0.7220  data: 0.0004  max mem: 57114
Epoch: [32]  [ 50/156]  eta: 0:01:24  lr: 0.007962  min_lr: 0.007962  loss: 3.7250 (3.7123)  weight_decay: 0.0500 (0.0500)  time: 0.7116  data: 0.0004  max mem: 57114
Epoch: [32]  [ 60/156]  eta: 0:01:14  lr: 0.007961  min_lr: 0.007961  loss: 3.7062 (3.6961)  weight_decay: 0.0500 (0.0500)  time: 0.7004  data: 0.0004  max mem: 57114
Epoch: [32]  [ 70/156]  eta: 0:01:06  lr: 0.007961  min_lr: 0.007961  loss: 3.9869 (3.7317)  weight_decay: 0.0500 (0.0500)  time: 0.7146  data: 0.0004  max mem: 57114
Epoch: [32]  [ 80/156]  eta: 0:00:58  lr: 0.007961  min_lr: 0.007961  loss: 3.8539 (3.7093)  weight_decay: 0.0500 (0.0500)  time: 0.7145  data: 0.0004  max mem: 57114
Epoch: [32]  [ 90/156]  eta: 0:00:50  lr: 0.007960  min_lr: 0.007960  loss: 3.8242 (3.7193)  weight_decay: 0.0500 (0.0500)  time: 0.7078  data: 0.0004  max mem: 57114
Epoch: [32]  [100/156]  eta: 0:00:42  lr: 0.007960  min_lr: 0.007960  loss: 3.8242 (3.6967)  weight_decay: 0.0500 (0.0500)  time: 0.7026  data: 0.0004  max mem: 57114
Epoch: [32]  [110/156]  eta: 0:00:34  lr: 0.007959  min_lr: 0.007959  loss: 3.7176 (3.7019)  weight_decay: 0.0500 (0.0500)  time: 0.6989  data: 0.0004  max mem: 57114
Epoch: [32]  [120/156]  eta: 0:00:26  lr: 0.007959  min_lr: 0.007959  loss: 3.9420 (3.7101)  weight_decay: 0.0500 (0.0500)  time: 0.7010  data: 0.0004  max mem: 57114
Epoch: [32]  [130/156]  eta: 0:00:19  lr: 0.007959  min_lr: 0.007959  loss: 3.7780 (3.7061)  weight_decay: 0.0500 (0.0500)  time: 0.6980  data: 0.0008  max mem: 57114
Epoch: [32]  [140/156]  eta: 0:00:11  lr: 0.007958  min_lr: 0.007958  loss: 3.7780 (3.7245)  weight_decay: 0.0500 (0.0500)  time: 0.6849  data: 0.0007  max mem: 57114
Epoch: [32]  [150/156]  eta: 0:00:04  lr: 0.007958  min_lr: 0.007958  loss: 3.8998 (3.7362)  weight_decay: 0.0500 (0.0500)  time: 0.6778  data: 0.0001  max mem: 57114
Epoch: [32]  [155/156]  eta: 0:00:00  lr: 0.007958  min_lr: 0.007958  loss: 3.8506 (3.7230)  weight_decay: 0.0500 (0.0500)  time: 0.6794  data: 0.0001  max mem: 57114
Epoch: [32] Total time: 0:01:54 (0.7326 s / it)
Averaged stats: lr: 0.007958  min_lr: 0.007958  loss: 3.8506 (3.7440)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.2513 (1.2513)  acc1: 75.0000 (75.0000)  acc5: 94.7917 (94.7917)  time: 2.0910  data: 1.8351  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.5180 (1.4404)  acc1: 69.2708 (68.2864)  acc5: 90.1042 (88.4911)  time: 0.5688  data: 0.3671  max mem: 57114
Test: Total time: 0:00:02 (0.5936 s / it)
* Acc@1 68.071 Acc@5 88.681 loss 1.540
Accuracy of the model on the 50000 test images: 68.1%
Max accuracy: 68.07%
Test:  [0/5]  eta: 0:00:10  loss: 6.9416 (6.9416)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.0046  data: 1.7611  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9344 (6.9327)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.2558)  time: 0.5515  data: 0.3523  max mem: 57114
Test: Total time: 0:00:02 (0.5635 s / it)
* Acc@1 0.102 Acc@5 0.474 loss 6.930
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [33]  [  0/156]  eta: 0:13:56  lr: 0.007958  min_lr: 0.007958  loss: 4.4507 (4.4507)  weight_decay: 0.0500 (0.0500)  time: 5.3620  data: 3.8129  max mem: 57114
Epoch: [33]  [ 10/156]  eta: 0:02:52  lr: 0.007957  min_lr: 0.007957  loss: 3.9710 (3.9178)  weight_decay: 0.0500 (0.0500)  time: 1.1810  data: 0.3469  max mem: 57114
Epoch: [33]  [ 20/156]  eta: 0:02:13  lr: 0.007957  min_lr: 0.007957  loss: 3.7970 (3.8800)  weight_decay: 0.0500 (0.0500)  time: 0.7619  data: 0.0003  max mem: 57114
Epoch: [33]  [ 30/156]  eta: 0:01:53  lr: 0.007956  min_lr: 0.007956  loss: 3.7739 (3.8201)  weight_decay: 0.0500 (0.0500)  time: 0.7414  data: 0.0003  max mem: 57114
Epoch: [33]  [ 40/156]  eta: 0:01:38  lr: 0.007956  min_lr: 0.007956  loss: 3.6692 (3.7751)  weight_decay: 0.0500 (0.0500)  time: 0.7190  data: 0.0003  max mem: 57114
Epoch: [33]  [ 50/156]  eta: 0:01:28  lr: 0.007955  min_lr: 0.007955  loss: 3.6692 (3.7686)  weight_decay: 0.0500 (0.0500)  time: 0.7278  data: 0.0003  max mem: 57114
Epoch: [33]  [ 60/156]  eta: 0:01:18  lr: 0.007955  min_lr: 0.007955  loss: 3.9719 (3.8019)  weight_decay: 0.0500 (0.0500)  time: 0.7387  data: 0.0003  max mem: 57114
Epoch: [33]  [ 70/156]  eta: 0:01:08  lr: 0.007955  min_lr: 0.007955  loss: 4.0939 (3.8239)  weight_decay: 0.0500 (0.0500)  time: 0.7228  data: 0.0003  max mem: 57114
Epoch: [33]  [ 80/156]  eta: 0:01:00  lr: 0.007954  min_lr: 0.007954  loss: 3.9007 (3.8031)  weight_decay: 0.0500 (0.0500)  time: 0.7123  data: 0.0004  max mem: 57114
Epoch: [33]  [ 90/156]  eta: 0:00:51  lr: 0.007954  min_lr: 0.007954  loss: 3.9007 (3.8229)  weight_decay: 0.0500 (0.0500)  time: 0.7107  data: 0.0004  max mem: 57114
Epoch: [33]  [100/156]  eta: 0:00:43  lr: 0.007953  min_lr: 0.007953  loss: 3.8879 (3.8079)  weight_decay: 0.0500 (0.0500)  time: 0.7013  data: 0.0004  max mem: 57114
Epoch: [33]  [110/156]  eta: 0:00:35  lr: 0.007953  min_lr: 0.007953  loss: 3.7898 (3.8142)  weight_decay: 0.0500 (0.0500)  time: 0.7028  data: 0.0004  max mem: 57114
Epoch: [33]  [120/156]  eta: 0:00:27  lr: 0.007952  min_lr: 0.007952  loss: 3.7898 (3.8045)  weight_decay: 0.0500 (0.0500)  time: 0.7029  data: 0.0004  max mem: 57114
Epoch: [33]  [130/156]  eta: 0:00:19  lr: 0.007952  min_lr: 0.007952  loss: 3.7862 (3.8044)  weight_decay: 0.0500 (0.0500)  time: 0.6985  data: 0.0008  max mem: 57114
Epoch: [33]  [140/156]  eta: 0:00:12  lr: 0.007951  min_lr: 0.007951  loss: 3.8975 (3.8078)  weight_decay: 0.0500 (0.0500)  time: 0.6916  data: 0.0007  max mem: 57114
Epoch: [33]  [150/156]  eta: 0:00:04  lr: 0.007951  min_lr: 0.007951  loss: 3.8975 (3.8037)  weight_decay: 0.0500 (0.0500)  time: 0.6861  data: 0.0001  max mem: 57114
Epoch: [33]  [155/156]  eta: 0:00:00  lr: 0.007951  min_lr: 0.007951  loss: 3.9580 (3.8116)  weight_decay: 0.0500 (0.0500)  time: 0.6853  data: 0.0001  max mem: 57114
Epoch: [33] Total time: 0:01:56 (0.7470 s / it)
Averaged stats: lr: 0.007951  min_lr: 0.007951  loss: 3.9580 (3.7344)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.1119 (1.1119)  acc1: 81.2500 (81.2500)  acc5: 94.7917 (94.7917)  time: 2.0279  data: 1.7725  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.3044 (1.3333)  acc1: 72.3958 (70.7161)  acc5: 92.7083 (89.3862)  time: 0.5563  data: 0.3546  max mem: 57114
Test: Total time: 0:00:02 (0.5822 s / it)
* Acc@1 68.598 Acc@5 89.186 loss 1.411
Accuracy of the model on the 50000 test images: 68.6%
Max accuracy: 68.60%
Test:  [0/5]  eta: 0:00:10  loss: 6.9479 (6.9479)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.1013  data: 1.8578  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9359 (6.9349)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.3836)  time: 0.5709  data: 0.3717  max mem: 57114
Test: Total time: 0:00:02 (0.5837 s / it)
* Acc@1 0.098 Acc@5 0.474 loss 6.931
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [34]  [  0/156]  eta: 0:13:21  lr: 0.007951  min_lr: 0.007951  loss: 3.4468 (3.4468)  weight_decay: 0.0500 (0.0500)  time: 5.1360  data: 4.2764  max mem: 57114
Epoch: [34]  [ 10/156]  eta: 0:02:48  lr: 0.007950  min_lr: 0.007950  loss: 3.4468 (3.4604)  weight_decay: 0.0500 (0.0500)  time: 1.1537  data: 0.3891  max mem: 57114
Epoch: [34]  [ 20/156]  eta: 0:02:09  lr: 0.007950  min_lr: 0.007950  loss: 3.5855 (3.5394)  weight_decay: 0.0500 (0.0500)  time: 0.7452  data: 0.0003  max mem: 57114
Epoch: [34]  [ 30/156]  eta: 0:01:51  lr: 0.007949  min_lr: 0.007949  loss: 3.5290 (3.5308)  weight_decay: 0.0500 (0.0500)  time: 0.7345  data: 0.0003  max mem: 57114
Epoch: [34]  [ 40/156]  eta: 0:01:38  lr: 0.007949  min_lr: 0.007949  loss: 3.8317 (3.6150)  weight_decay: 0.0500 (0.0500)  time: 0.7316  data: 0.0003  max mem: 57114
Epoch: [34]  [ 50/156]  eta: 0:01:27  lr: 0.007948  min_lr: 0.007948  loss: 3.9799 (3.6551)  weight_decay: 0.0500 (0.0500)  time: 0.7261  data: 0.0004  max mem: 57114
Epoch: [34]  [ 60/156]  eta: 0:01:17  lr: 0.007948  min_lr: 0.007948  loss: 3.9158 (3.6535)  weight_decay: 0.0500 (0.0500)  time: 0.7294  data: 0.0004  max mem: 57114
Epoch: [34]  [ 70/156]  eta: 0:01:08  lr: 0.007948  min_lr: 0.007948  loss: 3.6094 (3.6431)  weight_decay: 0.0500 (0.0500)  time: 0.7269  data: 0.0004  max mem: 57114
Epoch: [34]  [ 80/156]  eta: 0:00:59  lr: 0.007947  min_lr: 0.007947  loss: 3.6094 (3.6344)  weight_decay: 0.0500 (0.0500)  time: 0.7308  data: 0.0004  max mem: 57114
Epoch: [34]  [ 90/156]  eta: 0:00:51  lr: 0.007947  min_lr: 0.007947  loss: 3.7552 (3.6392)  weight_decay: 0.0500 (0.0500)  time: 0.7257  data: 0.0004  max mem: 57114
Epoch: [34]  [100/156]  eta: 0:00:43  lr: 0.007946  min_lr: 0.007946  loss: 3.9331 (3.6647)  weight_decay: 0.0500 (0.0500)  time: 0.7103  data: 0.0004  max mem: 57114
Epoch: [34]  [110/156]  eta: 0:00:35  lr: 0.007946  min_lr: 0.007946  loss: 3.9067 (3.6760)  weight_decay: 0.0500 (0.0500)  time: 0.7106  data: 0.0004  max mem: 57114
Epoch: [34]  [120/156]  eta: 0:00:27  lr: 0.007945  min_lr: 0.007945  loss: 3.6396 (3.6538)  weight_decay: 0.0500 (0.0500)  time: 0.7128  data: 0.0004  max mem: 57114
Epoch: [34]  [130/156]  eta: 0:00:19  lr: 0.007945  min_lr: 0.007945  loss: 3.8057 (3.6621)  weight_decay: 0.0500 (0.0500)  time: 0.7014  data: 0.0009  max mem: 57114
Epoch: [34]  [140/156]  eta: 0:00:12  lr: 0.007944  min_lr: 0.007944  loss: 3.7750 (3.6666)  weight_decay: 0.0500 (0.0500)  time: 0.6874  data: 0.0007  max mem: 57114
Epoch: [34]  [150/156]  eta: 0:00:04  lr: 0.007944  min_lr: 0.007944  loss: 3.7750 (3.6724)  weight_decay: 0.0500 (0.0500)  time: 0.6878  data: 0.0001  max mem: 57114
Epoch: [34]  [155/156]  eta: 0:00:00  lr: 0.007944  min_lr: 0.007944  loss: 3.8798 (3.6700)  weight_decay: 0.0500 (0.0500)  time: 0.6900  data: 0.0001  max mem: 57114
Epoch: [34] Total time: 0:01:56 (0.7485 s / it)
Averaged stats: lr: 0.007944  min_lr: 0.007944  loss: 3.8798 (3.7164)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.1092 (1.1092)  acc1: 78.1250 (78.1250)  acc5: 95.3125 (95.3125)  time: 2.0324  data: 1.7768  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.3172 (1.3167)  acc1: 76.5625 (69.6931)  acc5: 92.1875 (89.7698)  time: 0.5572  data: 0.3554  max mem: 57114
Test: Total time: 0:00:02 (0.5768 s / it)
* Acc@1 68.824 Acc@5 89.520 loss 1.415
Accuracy of the model on the 50000 test images: 68.8%
Max accuracy: 68.82%
Test:  [0/5]  eta: 0:00:10  loss: 6.9542 (6.9542)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.0794  data: 1.8359  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9381 (6.9373)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.5115)  time: 0.5664  data: 0.3673  max mem: 57114
Test: Total time: 0:00:02 (0.5778 s / it)
* Acc@1 0.100 Acc@5 0.476 loss 6.934
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [35]  [  0/156]  eta: 0:14:03  lr: 0.007943  min_lr: 0.007943  loss: 4.0256 (4.0256)  weight_decay: 0.0500 (0.0500)  time: 5.4058  data: 3.9358  max mem: 57114
Epoch: [35]  [ 10/156]  eta: 0:02:49  lr: 0.007943  min_lr: 0.007943  loss: 3.9828 (3.9388)  weight_decay: 0.0500 (0.0500)  time: 1.1611  data: 0.3581  max mem: 57114
Epoch: [35]  [ 20/156]  eta: 0:02:09  lr: 0.007943  min_lr: 0.007943  loss: 3.9179 (3.8044)  weight_decay: 0.0500 (0.0500)  time: 0.7295  data: 0.0003  max mem: 57114
Epoch: [35]  [ 30/156]  eta: 0:01:52  lr: 0.007942  min_lr: 0.007942  loss: 3.6061 (3.7141)  weight_decay: 0.0500 (0.0500)  time: 0.7413  data: 0.0003  max mem: 57114
Epoch: [35]  [ 40/156]  eta: 0:01:38  lr: 0.007942  min_lr: 0.007942  loss: 3.5662 (3.6798)  weight_decay: 0.0500 (0.0500)  time: 0.7464  data: 0.0004  max mem: 57114
Epoch: [35]  [ 50/156]  eta: 0:01:27  lr: 0.007941  min_lr: 0.007941  loss: 3.8813 (3.7079)  weight_decay: 0.0500 (0.0500)  time: 0.7340  data: 0.0004  max mem: 57114
Epoch: [35]  [ 60/156]  eta: 0:01:18  lr: 0.007941  min_lr: 0.007941  loss: 3.8476 (3.6967)  weight_decay: 0.0500 (0.0500)  time: 0.7336  data: 0.0004  max mem: 57114
Epoch: [35]  [ 70/156]  eta: 0:01:08  lr: 0.007940  min_lr: 0.007940  loss: 3.8438 (3.6944)  weight_decay: 0.0500 (0.0500)  time: 0.7307  data: 0.0004  max mem: 57114
Epoch: [35]  [ 80/156]  eta: 0:00:59  lr: 0.007940  min_lr: 0.007940  loss: 3.6790 (3.6813)  weight_decay: 0.0500 (0.0500)  time: 0.7121  data: 0.0004  max mem: 57114
Epoch: [35]  [ 90/156]  eta: 0:00:51  lr: 0.007939  min_lr: 0.007939  loss: 3.6624 (3.6524)  weight_decay: 0.0500 (0.0500)  time: 0.6995  data: 0.0005  max mem: 57114
Epoch: [35]  [100/156]  eta: 0:00:43  lr: 0.007939  min_lr: 0.007939  loss: 3.7012 (3.6442)  weight_decay: 0.0500 (0.0500)  time: 0.7108  data: 0.0005  max mem: 57114
Epoch: [35]  [110/156]  eta: 0:00:35  lr: 0.007938  min_lr: 0.007938  loss: 3.7478 (3.6628)  weight_decay: 0.0500 (0.0500)  time: 0.7043  data: 0.0004  max mem: 57114
Epoch: [35]  [120/156]  eta: 0:00:27  lr: 0.007938  min_lr: 0.007938  loss: 3.9049 (3.6522)  weight_decay: 0.0500 (0.0500)  time: 0.6969  data: 0.0004  max mem: 57114
Epoch: [35]  [130/156]  eta: 0:00:19  lr: 0.007937  min_lr: 0.007937  loss: 3.9049 (3.6717)  weight_decay: 0.0500 (0.0500)  time: 0.7031  data: 0.0008  max mem: 57114
Epoch: [35]  [140/156]  eta: 0:00:11  lr: 0.007937  min_lr: 0.007937  loss: 3.7660 (3.6665)  weight_decay: 0.0500 (0.0500)  time: 0.6869  data: 0.0007  max mem: 57114
Epoch: [35]  [150/156]  eta: 0:00:04  lr: 0.007936  min_lr: 0.007936  loss: 3.7223 (3.6618)  weight_decay: 0.0500 (0.0500)  time: 0.6757  data: 0.0001  max mem: 57114
Epoch: [35]  [155/156]  eta: 0:00:00  lr: 0.007936  min_lr: 0.007936  loss: 3.8752 (3.6588)  weight_decay: 0.0500 (0.0500)  time: 0.6761  data: 0.0001  max mem: 57114
Epoch: [35] Total time: 0:01:56 (0.7449 s / it)
Averaged stats: lr: 0.007936  min_lr: 0.007936  loss: 3.8752 (3.6940)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.2144 (1.2144)  acc1: 77.6042 (77.6042)  acc5: 93.7500 (93.7500)  time: 2.0428  data: 1.7870  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.5003 (1.4708)  acc1: 73.4375 (67.9028)  acc5: 90.6250 (88.6189)  time: 0.5592  data: 0.3575  max mem: 57114
Test: Total time: 0:00:02 (0.5833 s / it)
* Acc@1 67.565 Acc@5 88.505 loss 1.538
Accuracy of the model on the 50000 test images: 67.6%
Max accuracy: 68.82%
Test:  [0/5]  eta: 0:00:11  loss: 6.9620 (6.9620)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.3284  data: 2.0849  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9404 (6.9403)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.5115)  time: 0.6163  data: 0.4171  max mem: 57114
Test: Total time: 0:00:03 (0.6298 s / it)
* Acc@1 0.096 Acc@5 0.488 loss 6.936
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [36]  [  0/156]  eta: 0:13:35  lr: 0.007936  min_lr: 0.007936  loss: 2.9476 (2.9476)  weight_decay: 0.0500 (0.0500)  time: 5.2257  data: 3.4122  max mem: 57114
Epoch: [36]  [ 10/156]  eta: 0:02:46  lr: 0.007935  min_lr: 0.007935  loss: 4.0111 (3.8144)  weight_decay: 0.0500 (0.0500)  time: 1.1397  data: 0.3105  max mem: 57114
Epoch: [36]  [ 20/156]  eta: 0:02:08  lr: 0.007935  min_lr: 0.007935  loss: 4.0503 (3.8929)  weight_decay: 0.0500 (0.0500)  time: 0.7341  data: 0.0003  max mem: 57114
Epoch: [36]  [ 30/156]  eta: 0:01:50  lr: 0.007934  min_lr: 0.007934  loss: 3.8711 (3.7776)  weight_decay: 0.0500 (0.0500)  time: 0.7384  data: 0.0003  max mem: 57114
Epoch: [36]  [ 40/156]  eta: 0:01:37  lr: 0.007934  min_lr: 0.007934  loss: 3.7295 (3.7160)  weight_decay: 0.0500 (0.0500)  time: 0.7277  data: 0.0003  max mem: 57114
Epoch: [36]  [ 50/156]  eta: 0:01:26  lr: 0.007933  min_lr: 0.007933  loss: 3.6506 (3.7392)  weight_decay: 0.0500 (0.0500)  time: 0.7075  data: 0.0004  max mem: 57114
Epoch: [36]  [ 60/156]  eta: 0:01:16  lr: 0.007933  min_lr: 0.007933  loss: 3.7391 (3.7254)  weight_decay: 0.0500 (0.0500)  time: 0.7108  data: 0.0004  max mem: 57114
Epoch: [36]  [ 70/156]  eta: 0:01:07  lr: 0.007932  min_lr: 0.007932  loss: 3.7391 (3.7271)  weight_decay: 0.0500 (0.0500)  time: 0.7099  data: 0.0004  max mem: 57114
Epoch: [36]  [ 80/156]  eta: 0:00:58  lr: 0.007932  min_lr: 0.007932  loss: 3.8442 (3.7338)  weight_decay: 0.0500 (0.0500)  time: 0.7005  data: 0.0004  max mem: 57114
Epoch: [36]  [ 90/156]  eta: 0:00:50  lr: 0.007931  min_lr: 0.007931  loss: 4.0621 (3.7321)  weight_decay: 0.0500 (0.0500)  time: 0.7125  data: 0.0004  max mem: 57114
Epoch: [36]  [100/156]  eta: 0:00:42  lr: 0.007930  min_lr: 0.007930  loss: 3.7939 (3.7091)  weight_decay: 0.0500 (0.0500)  time: 0.7195  data: 0.0003  max mem: 57114
Epoch: [36]  [110/156]  eta: 0:00:34  lr: 0.007930  min_lr: 0.007930  loss: 3.7172 (3.7034)  weight_decay: 0.0500 (0.0500)  time: 0.7048  data: 0.0003  max mem: 57114
Epoch: [36]  [120/156]  eta: 0:00:27  lr: 0.007929  min_lr: 0.007929  loss: 3.6282 (3.6786)  weight_decay: 0.0500 (0.0500)  time: 0.6957  data: 0.0003  max mem: 57114
Epoch: [36]  [130/156]  eta: 0:00:19  lr: 0.007929  min_lr: 0.007929  loss: 3.6600 (3.6855)  weight_decay: 0.0500 (0.0500)  time: 0.6980  data: 0.0008  max mem: 57114
Epoch: [36]  [140/156]  eta: 0:00:11  lr: 0.007928  min_lr: 0.007928  loss: 3.7901 (3.6927)  weight_decay: 0.0500 (0.0500)  time: 0.6906  data: 0.0007  max mem: 57114
Epoch: [36]  [150/156]  eta: 0:00:04  lr: 0.007928  min_lr: 0.007928  loss: 3.8182 (3.6970)  weight_decay: 0.0500 (0.0500)  time: 0.6825  data: 0.0001  max mem: 57114
Epoch: [36]  [155/156]  eta: 0:00:00  lr: 0.007928  min_lr: 0.007928  loss: 3.7901 (3.6960)  weight_decay: 0.0500 (0.0500)  time: 0.6768  data: 0.0001  max mem: 57114
Epoch: [36] Total time: 0:01:55 (0.7390 s / it)
Averaged stats: lr: 0.007928  min_lr: 0.007928  loss: 3.7901 (3.6858)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.1260 (1.1260)  acc1: 79.1667 (79.1667)  acc5: 94.7917 (94.7917)  time: 2.1516  data: 1.8953  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2824 (1.2360)  acc1: 72.9167 (72.2506)  acc5: 92.1875 (90.7928)  time: 0.5811  data: 0.3791  max mem: 57114
Test: Total time: 0:00:03 (0.6050 s / it)
* Acc@1 68.736 Acc@5 89.100 loss 1.415
Accuracy of the model on the 50000 test images: 68.7%
Max accuracy: 68.82%
Test:  [0/5]  eta: 0:00:11  loss: 6.9694 (6.9694)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.2865  data: 2.0431  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9421 (6.9423)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.5115)  time: 0.6079  data: 0.4087  max mem: 57114
Test: Total time: 0:00:03 (0.6179 s / it)
* Acc@1 0.090 Acc@5 0.494 loss 6.938
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [37]  [  0/156]  eta: 0:13:01  lr: 0.007927  min_lr: 0.007927  loss: 2.7302 (2.7302)  weight_decay: 0.0500 (0.0500)  time: 5.0096  data: 4.0693  max mem: 57114
Epoch: [37]  [ 10/156]  eta: 0:02:46  lr: 0.007927  min_lr: 0.007927  loss: 3.6475 (3.6055)  weight_decay: 0.0500 (0.0500)  time: 1.1376  data: 0.3702  max mem: 57114
Epoch: [37]  [ 20/156]  eta: 0:02:07  lr: 0.007926  min_lr: 0.007926  loss: 3.6475 (3.5942)  weight_decay: 0.0500 (0.0500)  time: 0.7367  data: 0.0004  max mem: 57114
Epoch: [37]  [ 30/156]  eta: 0:01:49  lr: 0.007926  min_lr: 0.007926  loss: 3.5704 (3.6485)  weight_decay: 0.0500 (0.0500)  time: 0.7183  data: 0.0003  max mem: 57114
Epoch: [37]  [ 40/156]  eta: 0:01:36  lr: 0.007925  min_lr: 0.007925  loss: 3.6311 (3.6666)  weight_decay: 0.0500 (0.0500)  time: 0.7198  data: 0.0003  max mem: 57114
Epoch: [37]  [ 50/156]  eta: 0:01:26  lr: 0.007925  min_lr: 0.007925  loss: 3.8632 (3.6784)  weight_decay: 0.0500 (0.0500)  time: 0.7273  data: 0.0004  max mem: 57114
Epoch: [37]  [ 60/156]  eta: 0:01:16  lr: 0.007924  min_lr: 0.007924  loss: 3.9054 (3.7118)  weight_decay: 0.0500 (0.0500)  time: 0.7248  data: 0.0004  max mem: 57114
Epoch: [37]  [ 70/156]  eta: 0:01:07  lr: 0.007924  min_lr: 0.007924  loss: 3.7960 (3.6902)  weight_decay: 0.0500 (0.0500)  time: 0.7140  data: 0.0005  max mem: 57114
Epoch: [37]  [ 80/156]  eta: 0:00:59  lr: 0.007923  min_lr: 0.007923  loss: 3.6517 (3.6811)  weight_decay: 0.0500 (0.0500)  time: 0.7128  data: 0.0004  max mem: 57114
Epoch: [37]  [ 90/156]  eta: 0:00:50  lr: 0.007922  min_lr: 0.007922  loss: 3.7964 (3.7048)  weight_decay: 0.0500 (0.0500)  time: 0.7017  data: 0.0004  max mem: 57114
Epoch: [37]  [100/156]  eta: 0:00:42  lr: 0.007922  min_lr: 0.007922  loss: 4.0008 (3.7446)  weight_decay: 0.0500 (0.0500)  time: 0.6990  data: 0.0004  max mem: 57114
Epoch: [37]  [110/156]  eta: 0:00:34  lr: 0.007921  min_lr: 0.007921  loss: 4.0830 (3.7582)  weight_decay: 0.0500 (0.0500)  time: 0.7069  data: 0.0004  max mem: 57114
Epoch: [37]  [120/156]  eta: 0:00:27  lr: 0.007921  min_lr: 0.007921  loss: 3.9228 (3.7531)  weight_decay: 0.0500 (0.0500)  time: 0.7142  data: 0.0004  max mem: 57114
Epoch: [37]  [130/156]  eta: 0:00:19  lr: 0.007920  min_lr: 0.007920  loss: 3.7931 (3.7270)  weight_decay: 0.0500 (0.0500)  time: 0.7111  data: 0.0009  max mem: 57114
Epoch: [37]  [140/156]  eta: 0:00:11  lr: 0.007920  min_lr: 0.007920  loss: 3.7863 (3.7402)  weight_decay: 0.0500 (0.0500)  time: 0.6887  data: 0.0007  max mem: 57114
Epoch: [37]  [150/156]  eta: 0:00:04  lr: 0.007919  min_lr: 0.007919  loss: 3.8762 (3.7439)  weight_decay: 0.0500 (0.0500)  time: 0.6836  data: 0.0001  max mem: 57114
Epoch: [37]  [155/156]  eta: 0:00:00  lr: 0.007919  min_lr: 0.007919  loss: 3.9530 (3.7340)  weight_decay: 0.0500 (0.0500)  time: 0.6857  data: 0.0001  max mem: 57114
Epoch: [37] Total time: 0:01:55 (0.7403 s / it)
Averaged stats: lr: 0.007919  min_lr: 0.007919  loss: 3.9530 (3.6728)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.2886 (1.2886)  acc1: 74.4792 (74.4792)  acc5: 96.3542 (96.3542)  time: 2.0736  data: 1.8179  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.4685 (1.4304)  acc1: 71.8750 (68.7980)  acc5: 92.1875 (91.3044)  time: 0.5655  data: 0.3637  max mem: 57114
Test: Total time: 0:00:02 (0.5889 s / it)
* Acc@1 69.545 Acc@5 89.808 loss 1.554
Accuracy of the model on the 50000 test images: 69.5%
Max accuracy: 69.55%
Test:  [0/5]  eta: 0:00:10  loss: 6.9769 (6.9769)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.0587  data: 1.8152  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9456 (6.9448)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.5115)  time: 0.5624  data: 0.3632  max mem: 57114
Test: Total time: 0:00:02 (0.5719 s / it)
* Acc@1 0.098 Acc@5 0.494 loss 6.941
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [38]  [  0/156]  eta: 0:14:14  lr: 0.007919  min_lr: 0.007919  loss: 4.0006 (4.0006)  weight_decay: 0.0500 (0.0500)  time: 5.4778  data: 4.4062  max mem: 57114
Epoch: [38]  [ 10/156]  eta: 0:02:53  lr: 0.007918  min_lr: 0.007918  loss: 3.8761 (3.8576)  weight_decay: 0.0500 (0.0500)  time: 1.1864  data: 0.4008  max mem: 57114
Epoch: [38]  [ 20/156]  eta: 0:02:13  lr: 0.007918  min_lr: 0.007918  loss: 3.8439 (3.7777)  weight_decay: 0.0500 (0.0500)  time: 0.7541  data: 0.0003  max mem: 57114
Epoch: [38]  [ 30/156]  eta: 0:01:53  lr: 0.007917  min_lr: 0.007917  loss: 3.6773 (3.6552)  weight_decay: 0.0500 (0.0500)  time: 0.7495  data: 0.0003  max mem: 57114
Epoch: [38]  [ 40/156]  eta: 0:01:40  lr: 0.007916  min_lr: 0.007916  loss: 3.7866 (3.7159)  weight_decay: 0.0500 (0.0500)  time: 0.7411  data: 0.0004  max mem: 57114
Epoch: [38]  [ 50/156]  eta: 0:01:28  lr: 0.007916  min_lr: 0.007916  loss: 3.8734 (3.6918)  weight_decay: 0.0500 (0.0500)  time: 0.7325  data: 0.0004  max mem: 57114
Epoch: [38]  [ 60/156]  eta: 0:01:19  lr: 0.007915  min_lr: 0.007915  loss: 3.8225 (3.7211)  weight_decay: 0.0500 (0.0500)  time: 0.7425  data: 0.0004  max mem: 57114
Epoch: [38]  [ 70/156]  eta: 0:01:09  lr: 0.007915  min_lr: 0.007915  loss: 3.7721 (3.7046)  weight_decay: 0.0500 (0.0500)  time: 0.7289  data: 0.0004  max mem: 57114
Epoch: [38]  [ 80/156]  eta: 0:01:00  lr: 0.007914  min_lr: 0.007914  loss: 3.7687 (3.7101)  weight_decay: 0.0500 (0.0500)  time: 0.7023  data: 0.0003  max mem: 57114
Epoch: [38]  [ 90/156]  eta: 0:00:51  lr: 0.007913  min_lr: 0.007913  loss: 3.7687 (3.7136)  weight_decay: 0.0500 (0.0500)  time: 0.7071  data: 0.0003  max mem: 57114
Epoch: [38]  [100/156]  eta: 0:00:43  lr: 0.007913  min_lr: 0.007913  loss: 3.7519 (3.6917)  weight_decay: 0.0500 (0.0500)  time: 0.7130  data: 0.0004  max mem: 57114
Epoch: [38]  [110/156]  eta: 0:00:35  lr: 0.007912  min_lr: 0.007912  loss: 3.5196 (3.6666)  weight_decay: 0.0500 (0.0500)  time: 0.7140  data: 0.0004  max mem: 57114
Epoch: [38]  [120/156]  eta: 0:00:27  lr: 0.007912  min_lr: 0.007912  loss: 3.6486 (3.6591)  weight_decay: 0.0500 (0.0500)  time: 0.7054  data: 0.0003  max mem: 57114
Epoch: [38]  [130/156]  eta: 0:00:19  lr: 0.007911  min_lr: 0.007911  loss: 3.6486 (3.6569)  weight_decay: 0.0500 (0.0500)  time: 0.6953  data: 0.0008  max mem: 57114
Epoch: [38]  [140/156]  eta: 0:00:12  lr: 0.007910  min_lr: 0.007910  loss: 3.9140 (3.6817)  weight_decay: 0.0500 (0.0500)  time: 0.6901  data: 0.0007  max mem: 57114
Epoch: [38]  [150/156]  eta: 0:00:04  lr: 0.007910  min_lr: 0.007910  loss: 3.9453 (3.6917)  weight_decay: 0.0500 (0.0500)  time: 0.6830  data: 0.0001  max mem: 57114
Epoch: [38]  [155/156]  eta: 0:00:00  lr: 0.007910  min_lr: 0.007910  loss: 3.9093 (3.6902)  weight_decay: 0.0500 (0.0500)  time: 0.6822  data: 0.0001  max mem: 57114
Epoch: [38] Total time: 0:01:56 (0.7500 s / it)
Averaged stats: lr: 0.007910  min_lr: 0.007910  loss: 3.9093 (3.6585)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.2714 (1.2714)  acc1: 78.1250 (78.1250)  acc5: 96.3542 (96.3542)  time: 2.0516  data: 1.7959  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.4674 (1.4146)  acc1: 73.9583 (70.4604)  acc5: 93.7500 (91.0486)  time: 0.5609  data: 0.3593  max mem: 57114
Test: Total time: 0:00:02 (0.5842 s / it)
* Acc@1 69.519 Acc@5 89.932 loss 1.532
Accuracy of the model on the 50000 test images: 69.5%
Max accuracy: 69.55%
Test:  [0/5]  eta: 0:00:11  loss: 6.9843 (6.9843)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.2873  data: 2.0438  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9496 (6.9479)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.5115)  time: 0.6081  data: 0.4089  max mem: 57114
Test: Total time: 0:00:03 (0.6204 s / it)
* Acc@1 0.096 Acc@5 0.502 loss 6.944
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [39]  [  0/156]  eta: 0:13:12  lr: 0.007909  min_lr: 0.007909  loss: 3.2476 (3.2476)  weight_decay: 0.0500 (0.0500)  time: 5.0824  data: 3.8984  max mem: 57114
Epoch: [39]  [ 10/156]  eta: 0:02:46  lr: 0.007909  min_lr: 0.007909  loss: 4.1147 (3.7284)  weight_decay: 0.0500 (0.0500)  time: 1.1414  data: 0.3547  max mem: 57114
Epoch: [39]  [ 20/156]  eta: 0:02:08  lr: 0.007908  min_lr: 0.007908  loss: 3.8930 (3.6963)  weight_decay: 0.0500 (0.0500)  time: 0.7398  data: 0.0003  max mem: 57114
Epoch: [39]  [ 30/156]  eta: 0:01:50  lr: 0.007908  min_lr: 0.007908  loss: 3.3102 (3.6115)  weight_decay: 0.0500 (0.0500)  time: 0.7323  data: 0.0003  max mem: 57114
Epoch: [39]  [ 40/156]  eta: 0:01:37  lr: 0.007907  min_lr: 0.007907  loss: 3.4952 (3.6177)  weight_decay: 0.0500 (0.0500)  time: 0.7303  data: 0.0003  max mem: 57114
Epoch: [39]  [ 50/156]  eta: 0:01:26  lr: 0.007906  min_lr: 0.007906  loss: 3.7782 (3.6425)  weight_decay: 0.0500 (0.0500)  time: 0.7165  data: 0.0004  max mem: 57114
Epoch: [39]  [ 60/156]  eta: 0:01:16  lr: 0.007906  min_lr: 0.007906  loss: 3.5006 (3.5844)  weight_decay: 0.0500 (0.0500)  time: 0.7142  data: 0.0004  max mem: 57114
Epoch: [39]  [ 70/156]  eta: 0:01:07  lr: 0.007905  min_lr: 0.007905  loss: 3.5576 (3.6099)  weight_decay: 0.0500 (0.0500)  time: 0.7126  data: 0.0004  max mem: 57114
Epoch: [39]  [ 80/156]  eta: 0:00:59  lr: 0.007905  min_lr: 0.007905  loss: 3.7617 (3.6226)  weight_decay: 0.0500 (0.0500)  time: 0.7057  data: 0.0004  max mem: 57114
Epoch: [39]  [ 90/156]  eta: 0:00:50  lr: 0.007904  min_lr: 0.007904  loss: 3.9628 (3.6560)  weight_decay: 0.0500 (0.0500)  time: 0.7130  data: 0.0003  max mem: 57114
Epoch: [39]  [100/156]  eta: 0:00:42  lr: 0.007903  min_lr: 0.007903  loss: 3.8155 (3.6574)  weight_decay: 0.0500 (0.0500)  time: 0.7138  data: 0.0003  max mem: 57114
Epoch: [39]  [110/156]  eta: 0:00:34  lr: 0.007903  min_lr: 0.007903  loss: 3.8155 (3.6710)  weight_decay: 0.0500 (0.0500)  time: 0.7058  data: 0.0003  max mem: 57114
Epoch: [39]  [120/156]  eta: 0:00:27  lr: 0.007902  min_lr: 0.007902  loss: 3.9575 (3.6831)  weight_decay: 0.0500 (0.0500)  time: 0.7018  data: 0.0003  max mem: 57114
Epoch: [39]  [130/156]  eta: 0:00:19  lr: 0.007901  min_lr: 0.007901  loss: 3.3313 (3.6401)  weight_decay: 0.0500 (0.0500)  time: 0.7029  data: 0.0007  max mem: 57114
Epoch: [39]  [140/156]  eta: 0:00:11  lr: 0.007901  min_lr: 0.007901  loss: 3.4074 (3.6484)  weight_decay: 0.0500 (0.0500)  time: 0.6911  data: 0.0006  max mem: 57114
Epoch: [39]  [150/156]  eta: 0:00:04  lr: 0.007900  min_lr: 0.007900  loss: 3.7368 (3.6378)  weight_decay: 0.0500 (0.0500)  time: 0.6831  data: 0.0001  max mem: 57114
Epoch: [39]  [155/156]  eta: 0:00:00  lr: 0.007900  min_lr: 0.007900  loss: 3.7819 (3.6481)  weight_decay: 0.0500 (0.0500)  time: 0.6832  data: 0.0001  max mem: 57114
Epoch: [39] Total time: 0:01:55 (0.7410 s / it)
Averaged stats: lr: 0.007900  min_lr: 0.007900  loss: 3.7819 (3.6425)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.2522 (1.2522)  acc1: 74.4792 (74.4792)  acc5: 95.3125 (95.3125)  time: 2.1595  data: 1.9036  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.4446 (1.4281)  acc1: 71.8750 (67.1356)  acc5: 92.1875 (89.6419)  time: 0.5826  data: 0.3808  max mem: 57114
Test: Total time: 0:00:03 (0.6080 s / it)
* Acc@1 68.926 Acc@5 89.158 loss 1.525
Accuracy of the model on the 50000 test images: 68.9%
Max accuracy: 69.55%
Test:  [0/5]  eta: 0:00:11  loss: 6.9944 (6.9944)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.2585  data: 2.0152  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9536 (6.9514)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.5115)  time: 0.6024  data: 0.4031  max mem: 57114
Test: Total time: 0:00:03 (0.6150 s / it)
* Acc@1 0.102 Acc@5 0.506 loss 6.947
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [40]  [  0/156]  eta: 0:13:20  lr: 0.007900  min_lr: 0.007900  loss: 3.8104 (3.8104)  weight_decay: 0.0500 (0.0500)  time: 5.1321  data: 4.0331  max mem: 57114
Epoch: [40]  [ 10/156]  eta: 0:02:47  lr: 0.007899  min_lr: 0.007899  loss: 3.8899 (3.9169)  weight_decay: 0.0500 (0.0500)  time: 1.1501  data: 0.3669  max mem: 57114
Epoch: [40]  [ 20/156]  eta: 0:02:07  lr: 0.007898  min_lr: 0.007898  loss: 3.8182 (3.7871)  weight_decay: 0.0500 (0.0500)  time: 0.7302  data: 0.0003  max mem: 57114
Epoch: [40]  [ 30/156]  eta: 0:01:49  lr: 0.007898  min_lr: 0.007898  loss: 3.8446 (3.8629)  weight_decay: 0.0500 (0.0500)  time: 0.7170  data: 0.0003  max mem: 57114
Epoch: [40]  [ 40/156]  eta: 0:01:37  lr: 0.007897  min_lr: 0.007897  loss: 3.9420 (3.7968)  weight_decay: 0.0500 (0.0500)  time: 0.7305  data: 0.0003  max mem: 57114
Epoch: [40]  [ 50/156]  eta: 0:01:26  lr: 0.007896  min_lr: 0.007896  loss: 3.7360 (3.8059)  weight_decay: 0.0500 (0.0500)  time: 0.7293  data: 0.0003  max mem: 57114
Epoch: [40]  [ 60/156]  eta: 0:01:16  lr: 0.007896  min_lr: 0.007896  loss: 3.8362 (3.8227)  weight_decay: 0.0500 (0.0500)  time: 0.7208  data: 0.0004  max mem: 57114
Epoch: [40]  [ 70/156]  eta: 0:01:07  lr: 0.007895  min_lr: 0.007895  loss: 4.0095 (3.8444)  weight_decay: 0.0500 (0.0500)  time: 0.7147  data: 0.0004  max mem: 57114
Epoch: [40]  [ 80/156]  eta: 0:00:59  lr: 0.007895  min_lr: 0.007895  loss: 3.8131 (3.7958)  weight_decay: 0.0500 (0.0500)  time: 0.7097  data: 0.0004  max mem: 57114
Epoch: [40]  [ 90/156]  eta: 0:00:50  lr: 0.007894  min_lr: 0.007894  loss: 3.7564 (3.7851)  weight_decay: 0.0500 (0.0500)  time: 0.7029  data: 0.0004  max mem: 57114
Epoch: [40]  [100/156]  eta: 0:00:42  lr: 0.007893  min_lr: 0.007893  loss: 3.7735 (3.7714)  weight_decay: 0.0500 (0.0500)  time: 0.7012  data: 0.0003  max mem: 57114
Epoch: [40]  [110/156]  eta: 0:00:34  lr: 0.007893  min_lr: 0.007893  loss: 3.7735 (3.7601)  weight_decay: 0.0500 (0.0500)  time: 0.7104  data: 0.0003  max mem: 57114
Epoch: [40]  [120/156]  eta: 0:00:27  lr: 0.007892  min_lr: 0.007892  loss: 3.8581 (3.7480)  weight_decay: 0.0500 (0.0500)  time: 0.7098  data: 0.0004  max mem: 57114
Epoch: [40]  [130/156]  eta: 0:00:19  lr: 0.007891  min_lr: 0.007891  loss: 3.8581 (3.7438)  weight_decay: 0.0500 (0.0500)  time: 0.7010  data: 0.0008  max mem: 57114
Epoch: [40]  [140/156]  eta: 0:00:11  lr: 0.007891  min_lr: 0.007891  loss: 3.8682 (3.7549)  weight_decay: 0.0500 (0.0500)  time: 0.6920  data: 0.0007  max mem: 57114
Epoch: [40]  [150/156]  eta: 0:00:04  lr: 0.007890  min_lr: 0.007890  loss: 3.9180 (3.7498)  weight_decay: 0.0500 (0.0500)  time: 0.6845  data: 0.0001  max mem: 57114
Epoch: [40]  [155/156]  eta: 0:00:00  lr: 0.007890  min_lr: 0.007890  loss: 3.7298 (3.7506)  weight_decay: 0.0500 (0.0500)  time: 0.6832  data: 0.0001  max mem: 57114
Epoch: [40] Total time: 0:01:55 (0.7408 s / it)
Averaged stats: lr: 0.007890  min_lr: 0.007890  loss: 3.7298 (3.6347)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.1549 (1.1549)  acc1: 77.0833 (77.0833)  acc5: 95.8333 (95.8333)  time: 2.0459  data: 1.7871  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.3533 (1.3181)  acc1: 72.3958 (72.1228)  acc5: 93.2292 (91.5601)  time: 0.5598  data: 0.3575  max mem: 57114
Test: Total time: 0:00:02 (0.5846 s / it)
* Acc@1 68.808 Acc@5 89.284 loss 1.480
Accuracy of the model on the 50000 test images: 68.8%
Max accuracy: 69.55%
Test:  [0/5]  eta: 0:00:11  loss: 7.0058 (7.0058)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.2465  data: 2.0029  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9577 (6.9557)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.5115)  time: 0.6000  data: 0.4007  max mem: 57114
Test: Total time: 0:00:03 (0.6146 s / it)
* Acc@1 0.100 Acc@5 0.500 loss 6.951
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [41]  [  0/156]  eta: 0:14:09  lr: 0.007889  min_lr: 0.007889  loss: 3.9790 (3.9790)  weight_decay: 0.0500 (0.0500)  time: 5.4451  data: 3.5614  max mem: 57114
Epoch: [41]  [ 10/156]  eta: 0:02:47  lr: 0.007889  min_lr: 0.007889  loss: 3.4852 (3.6099)  weight_decay: 0.0500 (0.0500)  time: 1.1448  data: 0.3241  max mem: 57114
Epoch: [41]  [ 20/156]  eta: 0:02:08  lr: 0.007888  min_lr: 0.007888  loss: 3.4852 (3.6177)  weight_decay: 0.0500 (0.0500)  time: 0.7186  data: 0.0004  max mem: 57114
Epoch: [41]  [ 30/156]  eta: 0:01:50  lr: 0.007887  min_lr: 0.007887  loss: 3.7686 (3.6746)  weight_decay: 0.0500 (0.0500)  time: 0.7353  data: 0.0003  max mem: 57114
Epoch: [41]  [ 40/156]  eta: 0:01:37  lr: 0.007887  min_lr: 0.007887  loss: 3.5939 (3.6114)  weight_decay: 0.0500 (0.0500)  time: 0.7339  data: 0.0004  max mem: 57114
Epoch: [41]  [ 50/156]  eta: 0:01:26  lr: 0.007886  min_lr: 0.007886  loss: 3.4629 (3.5622)  weight_decay: 0.0500 (0.0500)  time: 0.7254  data: 0.0004  max mem: 57114
Epoch: [41]  [ 60/156]  eta: 0:01:17  lr: 0.007885  min_lr: 0.007885  loss: 3.6673 (3.5676)  weight_decay: 0.0500 (0.0500)  time: 0.7232  data: 0.0004  max mem: 57114
Epoch: [41]  [ 70/156]  eta: 0:01:07  lr: 0.007885  min_lr: 0.007885  loss: 3.6673 (3.5388)  weight_decay: 0.0500 (0.0500)  time: 0.7165  data: 0.0004  max mem: 57114
Epoch: [41]  [ 80/156]  eta: 0:00:59  lr: 0.007884  min_lr: 0.007884  loss: 3.1708 (3.5236)  weight_decay: 0.0500 (0.0500)  time: 0.7166  data: 0.0003  max mem: 57114
Epoch: [41]  [ 90/156]  eta: 0:00:51  lr: 0.007883  min_lr: 0.007883  loss: 3.6706 (3.5451)  weight_decay: 0.0500 (0.0500)  time: 0.7114  data: 0.0004  max mem: 57114
Epoch: [41]  [100/156]  eta: 0:00:43  lr: 0.007883  min_lr: 0.007883  loss: 3.7885 (3.5523)  weight_decay: 0.0500 (0.0500)  time: 0.7140  data: 0.0004  max mem: 57114
Epoch: [41]  [110/156]  eta: 0:00:35  lr: 0.007882  min_lr: 0.007882  loss: 3.7885 (3.5547)  weight_decay: 0.0500 (0.0500)  time: 0.7162  data: 0.0004  max mem: 57114
Epoch: [41]  [120/156]  eta: 0:00:27  lr: 0.007881  min_lr: 0.007881  loss: 3.6860 (3.5617)  weight_decay: 0.0500 (0.0500)  time: 0.7138  data: 0.0004  max mem: 57114
Epoch: [41]  [130/156]  eta: 0:00:19  lr: 0.007881  min_lr: 0.007881  loss: 3.5894 (3.5593)  weight_decay: 0.0500 (0.0500)  time: 0.6981  data: 0.0009  max mem: 57114
Epoch: [41]  [140/156]  eta: 0:00:11  lr: 0.007880  min_lr: 0.007880  loss: 3.7426 (3.5680)  weight_decay: 0.0500 (0.0500)  time: 0.6817  data: 0.0007  max mem: 57114
Epoch: [41]  [150/156]  eta: 0:00:04  lr: 0.007879  min_lr: 0.007879  loss: 3.7426 (3.5786)  weight_decay: 0.0500 (0.0500)  time: 0.6797  data: 0.0001  max mem: 57114
Epoch: [41]  [155/156]  eta: 0:00:00  lr: 0.007879  min_lr: 0.007879  loss: 3.7551 (3.5891)  weight_decay: 0.0500 (0.0500)  time: 0.6797  data: 0.0001  max mem: 57114
Epoch: [41] Total time: 0:01:56 (0.7437 s / it)
Averaged stats: lr: 0.007879  min_lr: 0.007879  loss: 3.7551 (3.6234)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.3543 (1.3543)  acc1: 72.9167 (72.9167)  acc5: 93.7500 (93.7500)  time: 2.0973  data: 1.8416  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.4055 (1.4174)  acc1: 72.3958 (67.6471)  acc5: 93.7500 (89.5141)  time: 0.5701  data: 0.3684  max mem: 57114
Test: Total time: 0:00:02 (0.5928 s / it)
* Acc@1 69.068 Acc@5 89.336 loss 1.504
Accuracy of the model on the 50000 test images: 69.1%
Max accuracy: 69.55%
Test:  [0/5]  eta: 0:00:11  loss: 7.0176 (7.0176)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.3838  data: 2.1404  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9627 (6.9603)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.5115)  time: 0.6274  data: 0.4281  max mem: 57114
Test: Total time: 0:00:03 (0.6421 s / it)
* Acc@1 0.100 Acc@5 0.496 loss 6.955
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [42]  [  0/156]  eta: 0:13:56  lr: 0.007879  min_lr: 0.007879  loss: 3.2153 (3.2153)  weight_decay: 0.0500 (0.0500)  time: 5.3606  data: 3.8897  max mem: 57114
Epoch: [42]  [ 10/156]  eta: 0:02:51  lr: 0.007878  min_lr: 0.007878  loss: 3.3203 (3.4754)  weight_decay: 0.0500 (0.0500)  time: 1.1742  data: 0.3539  max mem: 57114
Epoch: [42]  [ 20/156]  eta: 0:02:10  lr: 0.007877  min_lr: 0.007877  loss: 3.6967 (3.7078)  weight_decay: 0.0500 (0.0500)  time: 0.7406  data: 0.0003  max mem: 57114
Epoch: [42]  [ 30/156]  eta: 0:01:51  lr: 0.007877  min_lr: 0.007877  loss: 3.7087 (3.6115)  weight_decay: 0.0500 (0.0500)  time: 0.7311  data: 0.0003  max mem: 57114
Epoch: [42]  [ 40/156]  eta: 0:01:38  lr: 0.007876  min_lr: 0.007876  loss: 3.7086 (3.6690)  weight_decay: 0.0500 (0.0500)  time: 0.7241  data: 0.0004  max mem: 57114
Epoch: [42]  [ 50/156]  eta: 0:01:27  lr: 0.007875  min_lr: 0.007875  loss: 3.7758 (3.6499)  weight_decay: 0.0500 (0.0500)  time: 0.7195  data: 0.0004  max mem: 57114
Epoch: [42]  [ 60/156]  eta: 0:01:17  lr: 0.007875  min_lr: 0.007875  loss: 3.3371 (3.5934)  weight_decay: 0.0500 (0.0500)  time: 0.7247  data: 0.0004  max mem: 57114
Epoch: [42]  [ 70/156]  eta: 0:01:08  lr: 0.007874  min_lr: 0.007874  loss: 3.5893 (3.6325)  weight_decay: 0.0500 (0.0500)  time: 0.7193  data: 0.0004  max mem: 57114
Epoch: [42]  [ 80/156]  eta: 0:00:59  lr: 0.007873  min_lr: 0.007873  loss: 3.7439 (3.6026)  weight_decay: 0.0500 (0.0500)  time: 0.7141  data: 0.0004  max mem: 57114
Epoch: [42]  [ 90/156]  eta: 0:00:51  lr: 0.007872  min_lr: 0.007872  loss: 3.7175 (3.6119)  weight_decay: 0.0500 (0.0500)  time: 0.7180  data: 0.0004  max mem: 57114
Epoch: [42]  [100/156]  eta: 0:00:43  lr: 0.007872  min_lr: 0.007872  loss: 3.7175 (3.6179)  weight_decay: 0.0500 (0.0500)  time: 0.7177  data: 0.0004  max mem: 57114
Epoch: [42]  [110/156]  eta: 0:00:35  lr: 0.007871  min_lr: 0.007871  loss: 3.6399 (3.6070)  weight_decay: 0.0500 (0.0500)  time: 0.7061  data: 0.0004  max mem: 57114
Epoch: [42]  [120/156]  eta: 0:00:27  lr: 0.007870  min_lr: 0.007870  loss: 3.5312 (3.5951)  weight_decay: 0.0500 (0.0500)  time: 0.6975  data: 0.0004  max mem: 57114
Epoch: [42]  [130/156]  eta: 0:00:19  lr: 0.007869  min_lr: 0.007869  loss: 3.5312 (3.5783)  weight_decay: 0.0500 (0.0500)  time: 0.6969  data: 0.0010  max mem: 57114
Epoch: [42]  [140/156]  eta: 0:00:11  lr: 0.007869  min_lr: 0.007869  loss: 3.5007 (3.5719)  weight_decay: 0.0500 (0.0500)  time: 0.6939  data: 0.0008  max mem: 57114
Epoch: [42]  [150/156]  eta: 0:00:04  lr: 0.007868  min_lr: 0.007868  loss: 3.7559 (3.5911)  weight_decay: 0.0500 (0.0500)  time: 0.6857  data: 0.0001  max mem: 57114
Epoch: [42]  [155/156]  eta: 0:00:00  lr: 0.007868  min_lr: 0.007868  loss: 3.7994 (3.5976)  weight_decay: 0.0500 (0.0500)  time: 0.6870  data: 0.0001  max mem: 57114
Epoch: [42] Total time: 0:01:56 (0.7452 s / it)
Averaged stats: lr: 0.007868  min_lr: 0.007868  loss: 3.7994 (3.6128)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.2029 (1.2029)  acc1: 75.5208 (75.5208)  acc5: 93.7500 (93.7500)  time: 2.0562  data: 1.8002  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.3718 (1.3995)  acc1: 72.9167 (70.5882)  acc5: 93.2292 (90.4092)  time: 0.5619  data: 0.3601  max mem: 57114
Test: Total time: 0:00:02 (0.5838 s / it)
* Acc@1 69.537 Acc@5 89.866 loss 1.489
Accuracy of the model on the 50000 test images: 69.5%
Max accuracy: 69.55%
Test:  [0/5]  eta: 0:00:11  loss: 7.0295 (7.0295)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.3127  data: 2.0692  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9684 (6.9654)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.5115)  time: 0.6132  data: 0.4139  max mem: 57114
Test: Total time: 0:00:03 (0.6245 s / it)
* Acc@1 0.098 Acc@5 0.498 loss 6.960
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [43]  [  0/156]  eta: 0:13:17  lr: 0.007868  min_lr: 0.007868  loss: 3.4967 (3.4967)  weight_decay: 0.0500 (0.0500)  time: 5.1103  data: 3.5846  max mem: 57114
Epoch: [43]  [ 10/156]  eta: 0:02:44  lr: 0.007867  min_lr: 0.007867  loss: 3.6675 (3.7243)  weight_decay: 0.0500 (0.0500)  time: 1.1249  data: 0.3262  max mem: 57114
Epoch: [43]  [ 20/156]  eta: 0:02:07  lr: 0.007866  min_lr: 0.007866  loss: 3.8050 (3.7579)  weight_decay: 0.0500 (0.0500)  time: 0.7262  data: 0.0003  max mem: 57114
Epoch: [43]  [ 30/156]  eta: 0:01:48  lr: 0.007865  min_lr: 0.007865  loss: 3.8254 (3.7123)  weight_decay: 0.0500 (0.0500)  time: 0.7218  data: 0.0003  max mem: 57114
Epoch: [43]  [ 40/156]  eta: 0:01:35  lr: 0.007865  min_lr: 0.007865  loss: 3.3205 (3.5467)  weight_decay: 0.0500 (0.0500)  time: 0.7143  data: 0.0004  max mem: 57114
Epoch: [43]  [ 50/156]  eta: 0:01:25  lr: 0.007864  min_lr: 0.007864  loss: 3.3079 (3.5655)  weight_decay: 0.0500 (0.0500)  time: 0.7234  data: 0.0004  max mem: 57114
Epoch: [43]  [ 60/156]  eta: 0:01:16  lr: 0.007863  min_lr: 0.007863  loss: 3.6824 (3.5640)  weight_decay: 0.0500 (0.0500)  time: 0.7309  data: 0.0004  max mem: 57114
Epoch: [43]  [ 70/156]  eta: 0:01:07  lr: 0.007862  min_lr: 0.007862  loss: 3.7014 (3.5957)  weight_decay: 0.0500 (0.0500)  time: 0.7168  data: 0.0004  max mem: 57114
Epoch: [43]  [ 80/156]  eta: 0:00:58  lr: 0.007862  min_lr: 0.007862  loss: 3.9096 (3.6213)  weight_decay: 0.0500 (0.0500)  time: 0.7052  data: 0.0004  max mem: 57114
Epoch: [43]  [ 90/156]  eta: 0:00:50  lr: 0.007861  min_lr: 0.007861  loss: 3.7983 (3.6157)  weight_decay: 0.0500 (0.0500)  time: 0.7109  data: 0.0004  max mem: 57114
Epoch: [43]  [100/156]  eta: 0:00:42  lr: 0.007860  min_lr: 0.007860  loss: 3.7839 (3.6155)  weight_decay: 0.0500 (0.0500)  time: 0.7031  data: 0.0004  max mem: 57114
Epoch: [43]  [110/156]  eta: 0:00:34  lr: 0.007859  min_lr: 0.007859  loss: 3.8582 (3.6148)  weight_decay: 0.0500 (0.0500)  time: 0.6971  data: 0.0004  max mem: 57114
Epoch: [43]  [120/156]  eta: 0:00:27  lr: 0.007859  min_lr: 0.007859  loss: 3.6100 (3.6181)  weight_decay: 0.0500 (0.0500)  time: 0.7046  data: 0.0003  max mem: 57114
Epoch: [43]  [130/156]  eta: 0:00:19  lr: 0.007858  min_lr: 0.007858  loss: 3.5787 (3.6063)  weight_decay: 0.0500 (0.0500)  time: 0.7003  data: 0.0008  max mem: 57114
Epoch: [43]  [140/156]  eta: 0:00:11  lr: 0.007857  min_lr: 0.007857  loss: 3.5397 (3.6076)  weight_decay: 0.0500 (0.0500)  time: 0.6884  data: 0.0007  max mem: 57114
Epoch: [43]  [150/156]  eta: 0:00:04  lr: 0.007856  min_lr: 0.007856  loss: 3.3224 (3.5886)  weight_decay: 0.0500 (0.0500)  time: 0.6831  data: 0.0001  max mem: 57114
Epoch: [43]  [155/156]  eta: 0:00:00  lr: 0.007856  min_lr: 0.007856  loss: 3.2921 (3.5903)  weight_decay: 0.0500 (0.0500)  time: 0.6821  data: 0.0001  max mem: 57114
Epoch: [43] Total time: 0:01:55 (0.7383 s / it)
Averaged stats: lr: 0.007856  min_lr: 0.007856  loss: 3.2921 (3.5992)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.3191 (1.3191)  acc1: 76.5625 (76.5625)  acc5: 93.7500 (93.7500)  time: 2.1010  data: 1.8454  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.3256 (1.3487)  acc1: 76.5625 (70.3325)  acc5: 93.7500 (89.8977)  time: 0.5709  data: 0.3691  max mem: 57114
Test: Total time: 0:00:02 (0.5923 s / it)
* Acc@1 70.031 Acc@5 89.816 loss 1.464
Accuracy of the model on the 50000 test images: 70.0%
Max accuracy: 70.03%
Test:  [0/5]  eta: 0:00:10  loss: 7.0438 (7.0438)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.0352  data: 1.7920  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9745 (6.9703)  acc1: 0.0000 (0.0000)  acc5: 0.0000 (0.5115)  time: 0.5574  data: 0.3585  max mem: 57114
Test: Total time: 0:00:02 (0.5699 s / it)
* Acc@1 0.098 Acc@5 0.502 loss 6.965
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [44]  [  0/156]  eta: 0:13:27  lr: 0.007856  min_lr: 0.007856  loss: 3.3795 (3.3795)  weight_decay: 0.0500 (0.0500)  time: 5.1790  data: 4.0850  max mem: 57114
Epoch: [44]  [ 10/156]  eta: 0:02:44  lr: 0.007855  min_lr: 0.007855  loss: 3.8556 (3.6566)  weight_decay: 0.0500 (0.0500)  time: 1.1297  data: 0.3717  max mem: 57114
Epoch: [44]  [ 20/156]  eta: 0:02:07  lr: 0.007854  min_lr: 0.007854  loss: 3.6876 (3.5175)  weight_decay: 0.0500 (0.0500)  time: 0.7259  data: 0.0003  max mem: 57114
Epoch: [44]  [ 30/156]  eta: 0:01:49  lr: 0.007854  min_lr: 0.007854  loss: 3.5665 (3.5469)  weight_decay: 0.0500 (0.0500)  time: 0.7225  data: 0.0004  max mem: 57114
Epoch: [44]  [ 40/156]  eta: 0:01:37  lr: 0.007853  min_lr: 0.007853  loss: 3.5665 (3.5056)  weight_decay: 0.0500 (0.0500)  time: 0.7332  data: 0.0004  max mem: 57114
Epoch: [44]  [ 50/156]  eta: 0:01:26  lr: 0.007852  min_lr: 0.007852  loss: 3.5503 (3.5254)  weight_decay: 0.0500 (0.0500)  time: 0.7399  data: 0.0004  max mem: 57114
Epoch: [44]  [ 60/156]  eta: 0:01:17  lr: 0.007851  min_lr: 0.007851  loss: 3.6035 (3.5265)  weight_decay: 0.0500 (0.0500)  time: 0.7442  data: 0.0003  max mem: 57114
Epoch: [44]  [ 70/156]  eta: 0:01:08  lr: 0.007850  min_lr: 0.007850  loss: 3.4794 (3.5465)  weight_decay: 0.0500 (0.0500)  time: 0.7420  data: 0.0003  max mem: 57114
Epoch: [44]  [ 80/156]  eta: 0:00:59  lr: 0.007850  min_lr: 0.007850  loss: 3.4794 (3.5285)  weight_decay: 0.0500 (0.0500)  time: 0.7210  data: 0.0004  max mem: 57114
Epoch: [44]  [ 90/156]  eta: 0:00:51  lr: 0.007849  min_lr: 0.007849  loss: 3.4771 (3.5185)  weight_decay: 0.0500 (0.0500)  time: 0.7064  data: 0.0004  max mem: 57114
Epoch: [44]  [100/156]  eta: 0:00:43  lr: 0.007848  min_lr: 0.007848  loss: 3.6269 (3.5370)  weight_decay: 0.0500 (0.0500)  time: 0.7103  data: 0.0004  max mem: 57114
Epoch: [44]  [110/156]  eta: 0:00:35  lr: 0.007847  min_lr: 0.007847  loss: 3.6269 (3.5234)  weight_decay: 0.0500 (0.0500)  time: 0.7156  data: 0.0004  max mem: 57114
Epoch: [44]  [120/156]  eta: 0:00:27  lr: 0.007847  min_lr: 0.007847  loss: 3.6316 (3.5315)  weight_decay: 0.0500 (0.0500)  time: 0.7058  data: 0.0004  max mem: 57114
Epoch: [44]  [130/156]  eta: 0:00:19  lr: 0.007846  min_lr: 0.007846  loss: 3.6254 (3.5283)  weight_decay: 0.0500 (0.0500)  time: 0.6946  data: 0.0005  max mem: 57114
Epoch: [44]  [140/156]  eta: 0:00:11  lr: 0.007845  min_lr: 0.007845  loss: 3.2878 (3.5009)  weight_decay: 0.0500 (0.0500)  time: 0.6851  data: 0.0003  max mem: 57114
Epoch: [44]  [150/156]  eta: 0:00:04  lr: 0.007844  min_lr: 0.007844  loss: 3.2281 (3.5061)  weight_decay: 0.0500 (0.0500)  time: 0.6836  data: 0.0001  max mem: 57114
Epoch: [44]  [155/156]  eta: 0:00:00  lr: 0.007844  min_lr: 0.007844  loss: 3.6013 (3.5118)  weight_decay: 0.0500 (0.0500)  time: 0.6856  data: 0.0001  max mem: 57114
Epoch: [44] Total time: 0:01:56 (0.7458 s / it)
Averaged stats: lr: 0.007844  min_lr: 0.007844  loss: 3.6013 (3.5893)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:09  loss: 1.0860 (1.0860)  acc1: 75.0000 (75.0000)  acc5: 94.2708 (94.2708)  time: 1.9977  data: 1.7421  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1862 (1.2253)  acc1: 73.4375 (70.9719)  acc5: 94.2708 (90.6650)  time: 0.5503  data: 0.3485  max mem: 57114
Test: Total time: 0:00:02 (0.5694 s / it)
* Acc@1 70.684 Acc@5 90.375 loss 1.321
Accuracy of the model on the 50000 test images: 70.7%
Max accuracy: 70.68%
Test:  [0/5]  eta: 0:00:09  loss: 7.0582 (7.0582)  acc1: 0.5208 (0.5208)  acc5: 1.0417 (1.0417)  time: 1.9834  data: 1.7399  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9759 (6.9763)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.5115)  time: 0.5472  data: 0.3481  max mem: 57114
Test: Total time: 0:00:02 (0.5569 s / it)
* Acc@1 0.100 Acc@5 0.512 loss 6.971
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [45]  [  0/156]  eta: 0:14:01  lr: 0.007844  min_lr: 0.007844  loss: 3.0866 (3.0866)  weight_decay: 0.0500 (0.0500)  time: 5.3972  data: 4.3448  max mem: 57114
Epoch: [45]  [ 10/156]  eta: 0:02:52  lr: 0.007843  min_lr: 0.007843  loss: 3.8410 (3.5653)  weight_decay: 0.0500 (0.0500)  time: 1.1798  data: 0.3953  max mem: 57114
Epoch: [45]  [ 20/156]  eta: 0:02:10  lr: 0.007842  min_lr: 0.007842  loss: 3.6438 (3.5080)  weight_decay: 0.0500 (0.0500)  time: 0.7394  data: 0.0004  max mem: 57114
Epoch: [45]  [ 30/156]  eta: 0:01:51  lr: 0.007841  min_lr: 0.007841  loss: 3.7275 (3.4953)  weight_decay: 0.0500 (0.0500)  time: 0.7284  data: 0.0004  max mem: 57114
Epoch: [45]  [ 40/156]  eta: 0:01:38  lr: 0.007840  min_lr: 0.007840  loss: 3.8051 (3.5321)  weight_decay: 0.0500 (0.0500)  time: 0.7332  data: 0.0004  max mem: 57114
Epoch: [45]  [ 50/156]  eta: 0:01:27  lr: 0.007840  min_lr: 0.007840  loss: 3.7874 (3.5696)  weight_decay: 0.0500 (0.0500)  time: 0.7274  data: 0.0003  max mem: 57114
Epoch: [45]  [ 60/156]  eta: 0:01:18  lr: 0.007839  min_lr: 0.007839  loss: 3.5106 (3.4939)  weight_decay: 0.0500 (0.0500)  time: 0.7395  data: 0.0004  max mem: 57114
Epoch: [45]  [ 70/156]  eta: 0:01:08  lr: 0.007838  min_lr: 0.007838  loss: 3.2942 (3.5102)  weight_decay: 0.0500 (0.0500)  time: 0.7267  data: 0.0004  max mem: 57114
Epoch: [45]  [ 80/156]  eta: 0:00:59  lr: 0.007837  min_lr: 0.007837  loss: 3.6893 (3.5276)  weight_decay: 0.0500 (0.0500)  time: 0.7053  data: 0.0004  max mem: 57114
Epoch: [45]  [ 90/156]  eta: 0:00:51  lr: 0.007836  min_lr: 0.007836  loss: 3.4270 (3.5015)  weight_decay: 0.0500 (0.0500)  time: 0.7200  data: 0.0004  max mem: 57114
Epoch: [45]  [100/156]  eta: 0:00:43  lr: 0.007836  min_lr: 0.007836  loss: 3.3269 (3.5155)  weight_decay: 0.0500 (0.0500)  time: 0.7190  data: 0.0004  max mem: 57114
Epoch: [45]  [110/156]  eta: 0:00:35  lr: 0.007835  min_lr: 0.007835  loss: 3.7344 (3.5191)  weight_decay: 0.0500 (0.0500)  time: 0.7101  data: 0.0004  max mem: 57114
Epoch: [45]  [120/156]  eta: 0:00:27  lr: 0.007834  min_lr: 0.007834  loss: 3.6706 (3.5156)  weight_decay: 0.0500 (0.0500)  time: 0.7044  data: 0.0004  max mem: 57114
Epoch: [45]  [130/156]  eta: 0:00:19  lr: 0.007833  min_lr: 0.007833  loss: 3.6800 (3.5411)  weight_decay: 0.0500 (0.0500)  time: 0.6946  data: 0.0009  max mem: 57114
Epoch: [45]  [140/156]  eta: 0:00:12  lr: 0.007832  min_lr: 0.007832  loss: 4.1154 (3.5764)  weight_decay: 0.0500 (0.0500)  time: 0.6831  data: 0.0007  max mem: 57114
Epoch: [45]  [150/156]  eta: 0:00:04  lr: 0.007832  min_lr: 0.007832  loss: 3.9132 (3.5897)  weight_decay: 0.0500 (0.0500)  time: 0.6768  data: 0.0001  max mem: 57114
Epoch: [45]  [155/156]  eta: 0:00:00  lr: 0.007831  min_lr: 0.007831  loss: 3.8282 (3.5867)  weight_decay: 0.0500 (0.0500)  time: 0.6793  data: 0.0001  max mem: 57114
Epoch: [45] Total time: 0:01:56 (0.7465 s / it)
Averaged stats: lr: 0.007831  min_lr: 0.007831  loss: 3.8282 (3.5742)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.2124 (1.2124)  acc1: 79.1667 (79.1667)  acc5: 95.3125 (95.3125)  time: 2.0264  data: 1.7708  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.3490 (1.3634)  acc1: 75.0000 (72.3785)  acc5: 95.3125 (91.9437)  time: 0.5559  data: 0.3542  max mem: 57114
Test: Total time: 0:00:02 (0.5767 s / it)
* Acc@1 70.598 Acc@5 90.715 loss 1.465
Accuracy of the model on the 50000 test images: 70.6%
Max accuracy: 70.68%
Test:  [0/5]  eta: 0:00:11  loss: 7.0714 (7.0714)  acc1: 0.5208 (0.5208)  acc5: 1.0417 (1.0417)  time: 2.3234  data: 2.0799  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9770 (6.9824)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.5115)  time: 0.6153  data: 0.4161  max mem: 57114
Test: Total time: 0:00:03 (0.6298 s / it)
* Acc@1 0.100 Acc@5 0.512 loss 6.977
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [46]  [  0/156]  eta: 0:13:53  lr: 0.007831  min_lr: 0.007831  loss: 3.4829 (3.4829)  weight_decay: 0.0500 (0.0500)  time: 5.3442  data: 3.6409  max mem: 57114
Epoch: [46]  [ 10/156]  eta: 0:02:53  lr: 0.007830  min_lr: 0.007830  loss: 3.3244 (3.1717)  weight_decay: 0.0500 (0.0500)  time: 1.1853  data: 0.3313  max mem: 57114
Epoch: [46]  [ 20/156]  eta: 0:02:10  lr: 0.007829  min_lr: 0.007829  loss: 3.5124 (3.4196)  weight_decay: 0.0500 (0.0500)  time: 0.7438  data: 0.0004  max mem: 57114
Epoch: [46]  [ 30/156]  eta: 0:01:51  lr: 0.007829  min_lr: 0.007829  loss: 3.5360 (3.4263)  weight_decay: 0.0500 (0.0500)  time: 0.7190  data: 0.0003  max mem: 57114
Epoch: [46]  [ 40/156]  eta: 0:01:37  lr: 0.007828  min_lr: 0.007828  loss: 3.5360 (3.4806)  weight_decay: 0.0500 (0.0500)  time: 0.7175  data: 0.0003  max mem: 57114
Epoch: [46]  [ 50/156]  eta: 0:01:26  lr: 0.007827  min_lr: 0.007827  loss: 3.6534 (3.5082)  weight_decay: 0.0500 (0.0500)  time: 0.7149  data: 0.0003  max mem: 57114
Epoch: [46]  [ 60/156]  eta: 0:01:16  lr: 0.007826  min_lr: 0.007826  loss: 3.5558 (3.5273)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0004  max mem: 57114
Epoch: [46]  [ 70/156]  eta: 0:01:07  lr: 0.007825  min_lr: 0.007825  loss: 3.4867 (3.5296)  weight_decay: 0.0500 (0.0500)  time: 0.7121  data: 0.0004  max mem: 57114
Epoch: [46]  [ 80/156]  eta: 0:00:59  lr: 0.007824  min_lr: 0.007824  loss: 3.6246 (3.5410)  weight_decay: 0.0500 (0.0500)  time: 0.7202  data: 0.0004  max mem: 57114
Epoch: [46]  [ 90/156]  eta: 0:00:50  lr: 0.007823  min_lr: 0.007823  loss: 3.8498 (3.5862)  weight_decay: 0.0500 (0.0500)  time: 0.7111  data: 0.0004  max mem: 57114
Epoch: [46]  [100/156]  eta: 0:00:42  lr: 0.007823  min_lr: 0.007823  loss: 3.8869 (3.5959)  weight_decay: 0.0500 (0.0500)  time: 0.7021  data: 0.0004  max mem: 57114
Epoch: [46]  [110/156]  eta: 0:00:34  lr: 0.007822  min_lr: 0.007822  loss: 3.2626 (3.5625)  weight_decay: 0.0500 (0.0500)  time: 0.7021  data: 0.0003  max mem: 57114
Epoch: [46]  [120/156]  eta: 0:00:27  lr: 0.007821  min_lr: 0.007821  loss: 3.5459 (3.5618)  weight_decay: 0.0500 (0.0500)  time: 0.7033  data: 0.0003  max mem: 57114
Epoch: [46]  [130/156]  eta: 0:00:19  lr: 0.007820  min_lr: 0.007820  loss: 3.6516 (3.5691)  weight_decay: 0.0500 (0.0500)  time: 0.6984  data: 0.0008  max mem: 57114
Epoch: [46]  [140/156]  eta: 0:00:11  lr: 0.007819  min_lr: 0.007819  loss: 3.7504 (3.5823)  weight_decay: 0.0500 (0.0500)  time: 0.6931  data: 0.0007  max mem: 57114
Epoch: [46]  [150/156]  eta: 0:00:04  lr: 0.007818  min_lr: 0.007818  loss: 3.5332 (3.5738)  weight_decay: 0.0500 (0.0500)  time: 0.6892  data: 0.0001  max mem: 57114
Epoch: [46]  [155/156]  eta: 0:00:00  lr: 0.007818  min_lr: 0.007818  loss: 3.4486 (3.5703)  weight_decay: 0.0500 (0.0500)  time: 0.6823  data: 0.0001  max mem: 57114
Epoch: [46] Total time: 0:01:55 (0.7416 s / it)
Averaged stats: lr: 0.007818  min_lr: 0.007818  loss: 3.4486 (3.5762)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.1581 (1.1581)  acc1: 79.1667 (79.1667)  acc5: 96.8750 (96.8750)  time: 2.0740  data: 1.8178  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2742 (1.2721)  acc1: 73.9583 (71.3555)  acc5: 93.7500 (91.1765)  time: 0.5655  data: 0.3636  max mem: 57114
Test: Total time: 0:00:02 (0.5914 s / it)
* Acc@1 71.589 Acc@5 91.546 loss 1.318
Accuracy of the model on the 50000 test images: 71.6%
Max accuracy: 71.59%
Test:  [0/5]  eta: 0:00:10  loss: 7.0848 (7.0848)  acc1: 0.5208 (0.5208)  acc5: 1.0417 (1.0417)  time: 2.1032  data: 1.8597  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9783 (6.9889)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.5115)  time: 0.5713  data: 0.3720  max mem: 57114
Test: Total time: 0:00:02 (0.5847 s / it)
* Acc@1 0.100 Acc@5 0.510 loss 6.983
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [47]  [  0/156]  eta: 0:13:34  lr: 0.007818  min_lr: 0.007818  loss: 3.7394 (3.7394)  weight_decay: 0.0500 (0.0500)  time: 5.2241  data: 3.7711  max mem: 57114
Epoch: [47]  [ 10/156]  eta: 0:02:47  lr: 0.007817  min_lr: 0.007817  loss: 3.7394 (3.6541)  weight_decay: 0.0500 (0.0500)  time: 1.1470  data: 0.3432  max mem: 57114
Epoch: [47]  [ 20/156]  eta: 0:02:07  lr: 0.007816  min_lr: 0.007816  loss: 3.7039 (3.7079)  weight_decay: 0.0500 (0.0500)  time: 0.7247  data: 0.0004  max mem: 57114
Epoch: [47]  [ 30/156]  eta: 0:01:49  lr: 0.007815  min_lr: 0.007815  loss: 3.7375 (3.6779)  weight_decay: 0.0500 (0.0500)  time: 0.7158  data: 0.0003  max mem: 57114
Epoch: [47]  [ 40/156]  eta: 0:01:37  lr: 0.007814  min_lr: 0.007814  loss: 3.5832 (3.5874)  weight_decay: 0.0500 (0.0500)  time: 0.7334  data: 0.0003  max mem: 57114
Epoch: [47]  [ 50/156]  eta: 0:01:27  lr: 0.007814  min_lr: 0.007814  loss: 3.5676 (3.5995)  weight_decay: 0.0500 (0.0500)  time: 0.7479  data: 0.0004  max mem: 57114
Epoch: [47]  [ 60/156]  eta: 0:01:17  lr: 0.007813  min_lr: 0.007813  loss: 3.8986 (3.6283)  weight_decay: 0.0500 (0.0500)  time: 0.7412  data: 0.0004  max mem: 57114
Epoch: [47]  [ 70/156]  eta: 0:01:08  lr: 0.007812  min_lr: 0.007812  loss: 3.5264 (3.5612)  weight_decay: 0.0500 (0.0500)  time: 0.7255  data: 0.0004  max mem: 57114
Epoch: [47]  [ 80/156]  eta: 0:00:59  lr: 0.007811  min_lr: 0.007811  loss: 3.5264 (3.5762)  weight_decay: 0.0500 (0.0500)  time: 0.7059  data: 0.0004  max mem: 57114
Epoch: [47]  [ 90/156]  eta: 0:00:51  lr: 0.007810  min_lr: 0.007810  loss: 3.7241 (3.5780)  weight_decay: 0.0500 (0.0500)  time: 0.6976  data: 0.0004  max mem: 57114
Epoch: [47]  [100/156]  eta: 0:00:42  lr: 0.007809  min_lr: 0.007809  loss: 3.7491 (3.5894)  weight_decay: 0.0500 (0.0500)  time: 0.7093  data: 0.0004  max mem: 57114
Epoch: [47]  [110/156]  eta: 0:00:35  lr: 0.007808  min_lr: 0.007808  loss: 3.6008 (3.5706)  weight_decay: 0.0500 (0.0500)  time: 0.7161  data: 0.0004  max mem: 57114
Epoch: [47]  [120/156]  eta: 0:00:27  lr: 0.007807  min_lr: 0.007807  loss: 3.3627 (3.5599)  weight_decay: 0.0500 (0.0500)  time: 0.7077  data: 0.0004  max mem: 57114
Epoch: [47]  [130/156]  eta: 0:00:19  lr: 0.007807  min_lr: 0.007807  loss: 3.6494 (3.5679)  weight_decay: 0.0500 (0.0500)  time: 0.7026  data: 0.0008  max mem: 57114
Epoch: [47]  [140/156]  eta: 0:00:11  lr: 0.007806  min_lr: 0.007806  loss: 3.7773 (3.5764)  weight_decay: 0.0500 (0.0500)  time: 0.6922  data: 0.0007  max mem: 57114
Epoch: [47]  [150/156]  eta: 0:00:04  lr: 0.007805  min_lr: 0.007805  loss: 3.5707 (3.5857)  weight_decay: 0.0500 (0.0500)  time: 0.6741  data: 0.0001  max mem: 57114
Epoch: [47]  [155/156]  eta: 0:00:00  lr: 0.007804  min_lr: 0.007804  loss: 3.5384 (3.5727)  weight_decay: 0.0500 (0.0500)  time: 0.6756  data: 0.0001  max mem: 57114
Epoch: [47] Total time: 0:01:55 (0.7434 s / it)
Averaged stats: lr: 0.007804  min_lr: 0.007804  loss: 3.5384 (3.5648)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8955 (0.8955)  acc1: 84.3750 (84.3750)  acc5: 95.8333 (95.8333)  time: 2.0146  data: 1.7577  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1862 (1.0820)  acc1: 76.0417 (73.9130)  acc5: 93.2292 (91.5601)  time: 0.5538  data: 0.3516  max mem: 57114
Test: Total time: 0:00:02 (0.5757 s / it)
* Acc@1 71.551 Acc@5 90.639 loss 1.232
Accuracy of the model on the 50000 test images: 71.6%
Max accuracy: 71.59%
Test:  [0/5]  eta: 0:00:11  loss: 7.0984 (7.0984)  acc1: 0.5208 (0.5208)  acc5: 1.0417 (1.0417)  time: 2.2707  data: 2.0272  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9818 (6.9962)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.6048  data: 0.4055  max mem: 57114
Test: Total time: 0:00:03 (0.6159 s / it)
* Acc@1 0.100 Acc@5 0.508 loss 6.989
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [48]  [  0/156]  eta: 0:12:19  lr: 0.007804  min_lr: 0.007804  loss: 3.8780 (3.8780)  weight_decay: 0.0500 (0.0500)  time: 4.7397  data: 3.4335  max mem: 57114
Epoch: [48]  [ 10/156]  eta: 0:02:42  lr: 0.007803  min_lr: 0.007803  loss: 3.7685 (3.5928)  weight_decay: 0.0500 (0.0500)  time: 1.1121  data: 0.3126  max mem: 57114
Epoch: [48]  [ 20/156]  eta: 0:02:06  lr: 0.007802  min_lr: 0.007802  loss: 3.5447 (3.5063)  weight_decay: 0.0500 (0.0500)  time: 0.7376  data: 0.0004  max mem: 57114
Epoch: [48]  [ 30/156]  eta: 0:01:48  lr: 0.007802  min_lr: 0.007802  loss: 3.5447 (3.5225)  weight_decay: 0.0500 (0.0500)  time: 0.7192  data: 0.0003  max mem: 57114
Epoch: [48]  [ 40/156]  eta: 0:01:36  lr: 0.007801  min_lr: 0.007801  loss: 3.7305 (3.5265)  weight_decay: 0.0500 (0.0500)  time: 0.7285  data: 0.0004  max mem: 57114
Epoch: [48]  [ 50/156]  eta: 0:01:25  lr: 0.007800  min_lr: 0.007800  loss: 3.4907 (3.5015)  weight_decay: 0.0500 (0.0500)  time: 0.7268  data: 0.0004  max mem: 57114
Epoch: [48]  [ 60/156]  eta: 0:01:16  lr: 0.007799  min_lr: 0.007799  loss: 3.6561 (3.5470)  weight_decay: 0.0500 (0.0500)  time: 0.7211  data: 0.0004  max mem: 57114
Epoch: [48]  [ 70/156]  eta: 0:01:07  lr: 0.007798  min_lr: 0.007798  loss: 3.7180 (3.5136)  weight_decay: 0.0500 (0.0500)  time: 0.7126  data: 0.0004  max mem: 57114
Epoch: [48]  [ 80/156]  eta: 0:00:58  lr: 0.007797  min_lr: 0.007797  loss: 3.3927 (3.4752)  weight_decay: 0.0500 (0.0500)  time: 0.6964  data: 0.0003  max mem: 57114
Epoch: [48]  [ 90/156]  eta: 0:00:50  lr: 0.007796  min_lr: 0.007796  loss: 3.5652 (3.4910)  weight_decay: 0.0500 (0.0500)  time: 0.7051  data: 0.0003  max mem: 57114
Epoch: [48]  [100/156]  eta: 0:00:42  lr: 0.007795  min_lr: 0.007795  loss: 3.6404 (3.5012)  weight_decay: 0.0500 (0.0500)  time: 0.7069  data: 0.0003  max mem: 57114
Epoch: [48]  [110/156]  eta: 0:00:34  lr: 0.007794  min_lr: 0.007794  loss: 3.5655 (3.4974)  weight_decay: 0.0500 (0.0500)  time: 0.7105  data: 0.0003  max mem: 57114
Epoch: [48]  [120/156]  eta: 0:00:27  lr: 0.007793  min_lr: 0.007793  loss: 3.6479 (3.5066)  weight_decay: 0.0500 (0.0500)  time: 0.7114  data: 0.0003  max mem: 57114
Epoch: [48]  [130/156]  eta: 0:00:19  lr: 0.007793  min_lr: 0.007793  loss: 3.7302 (3.5066)  weight_decay: 0.0500 (0.0500)  time: 0.7038  data: 0.0008  max mem: 57114
Epoch: [48]  [140/156]  eta: 0:00:11  lr: 0.007792  min_lr: 0.007792  loss: 3.6277 (3.5064)  weight_decay: 0.0500 (0.0500)  time: 0.6897  data: 0.0007  max mem: 57114
Epoch: [48]  [150/156]  eta: 0:00:04  lr: 0.007791  min_lr: 0.007791  loss: 3.6275 (3.5062)  weight_decay: 0.0500 (0.0500)  time: 0.6795  data: 0.0001  max mem: 57114
Epoch: [48]  [155/156]  eta: 0:00:00  lr: 0.007790  min_lr: 0.007790  loss: 3.5336 (3.5037)  weight_decay: 0.0500 (0.0500)  time: 0.6838  data: 0.0001  max mem: 57114
Epoch: [48] Total time: 0:01:55 (0.7382 s / it)
Averaged stats: lr: 0.007790  min_lr: 0.007790  loss: 3.5336 (3.5486)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0421 (1.0421)  acc1: 81.7708 (81.7708)  acc5: 96.3542 (96.3542)  time: 2.0362  data: 1.7806  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2486 (1.3213)  acc1: 76.5625 (72.6343)  acc5: 92.8571 (91.9437)  time: 0.5579  data: 0.3562  max mem: 57114
Test: Total time: 0:00:02 (0.5790 s / it)
* Acc@1 71.371 Acc@5 90.763 loss 1.394
Accuracy of the model on the 50000 test images: 71.4%
Max accuracy: 71.59%
Test:  [0/5]  eta: 0:00:11  loss: 7.1122 (7.1122)  acc1: 0.5208 (0.5208)  acc5: 1.0417 (1.0417)  time: 2.2981  data: 2.0546  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9881 (7.0046)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.6103  data: 0.4110  max mem: 57114
Test: Total time: 0:00:03 (0.6232 s / it)
* Acc@1 0.100 Acc@5 0.508 loss 6.997
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [49]  [  0/156]  eta: 0:12:45  lr: 0.007790  min_lr: 0.007790  loss: 3.5170 (3.5170)  weight_decay: 0.0500 (0.0500)  time: 4.9080  data: 4.0643  max mem: 57114
Epoch: [49]  [ 10/156]  eta: 0:02:41  lr: 0.007789  min_lr: 0.007789  loss: 3.5170 (3.5283)  weight_decay: 0.0500 (0.0500)  time: 1.1042  data: 0.3697  max mem: 57114
Epoch: [49]  [ 20/156]  eta: 0:02:06  lr: 0.007788  min_lr: 0.007788  loss: 3.7628 (3.6611)  weight_decay: 0.0500 (0.0500)  time: 0.7331  data: 0.0003  max mem: 57114
Epoch: [49]  [ 30/156]  eta: 0:01:48  lr: 0.007787  min_lr: 0.007787  loss: 3.8029 (3.6040)  weight_decay: 0.0500 (0.0500)  time: 0.7308  data: 0.0003  max mem: 57114
Epoch: [49]  [ 40/156]  eta: 0:01:36  lr: 0.007786  min_lr: 0.007786  loss: 3.4493 (3.5590)  weight_decay: 0.0500 (0.0500)  time: 0.7320  data: 0.0004  max mem: 57114
Epoch: [49]  [ 50/156]  eta: 0:01:25  lr: 0.007786  min_lr: 0.007786  loss: 3.5084 (3.5739)  weight_decay: 0.0500 (0.0500)  time: 0.7276  data: 0.0004  max mem: 57114
Epoch: [49]  [ 60/156]  eta: 0:01:16  lr: 0.007785  min_lr: 0.007785  loss: 3.7999 (3.6006)  weight_decay: 0.0500 (0.0500)  time: 0.7073  data: 0.0004  max mem: 57114
Epoch: [49]  [ 70/156]  eta: 0:01:07  lr: 0.007784  min_lr: 0.007784  loss: 3.6271 (3.5737)  weight_decay: 0.0500 (0.0500)  time: 0.7071  data: 0.0005  max mem: 57114
Epoch: [49]  [ 80/156]  eta: 0:00:58  lr: 0.007783  min_lr: 0.007783  loss: 3.1113 (3.4985)  weight_decay: 0.0500 (0.0500)  time: 0.7126  data: 0.0004  max mem: 57114
Epoch: [49]  [ 90/156]  eta: 0:00:50  lr: 0.007782  min_lr: 0.007782  loss: 3.3080 (3.5074)  weight_decay: 0.0500 (0.0500)  time: 0.7146  data: 0.0004  max mem: 57114
Epoch: [49]  [100/156]  eta: 0:00:42  lr: 0.007781  min_lr: 0.007781  loss: 3.4198 (3.4928)  weight_decay: 0.0500 (0.0500)  time: 0.7104  data: 0.0004  max mem: 57114
Epoch: [49]  [110/156]  eta: 0:00:34  lr: 0.007780  min_lr: 0.007780  loss: 3.3955 (3.5039)  weight_decay: 0.0500 (0.0500)  time: 0.7034  data: 0.0004  max mem: 57114
Epoch: [49]  [120/156]  eta: 0:00:27  lr: 0.007779  min_lr: 0.007779  loss: 3.3604 (3.4962)  weight_decay: 0.0500 (0.0500)  time: 0.7058  data: 0.0004  max mem: 57114
Epoch: [49]  [130/156]  eta: 0:00:19  lr: 0.007778  min_lr: 0.007778  loss: 3.3581 (3.4988)  weight_decay: 0.0500 (0.0500)  time: 0.7018  data: 0.0009  max mem: 57114
Epoch: [49]  [140/156]  eta: 0:00:11  lr: 0.007777  min_lr: 0.007777  loss: 3.4590 (3.5032)  weight_decay: 0.0500 (0.0500)  time: 0.6858  data: 0.0007  max mem: 57114
Epoch: [49]  [150/156]  eta: 0:00:04  lr: 0.007776  min_lr: 0.007776  loss: 3.5042 (3.5110)  weight_decay: 0.0500 (0.0500)  time: 0.6855  data: 0.0001  max mem: 57114
Epoch: [49]  [155/156]  eta: 0:00:00  lr: 0.007776  min_lr: 0.007776  loss: 3.5358 (3.5012)  weight_decay: 0.0500 (0.0500)  time: 0.6873  data: 0.0001  max mem: 57114
Epoch: [49] Total time: 0:01:55 (0.7396 s / it)
Averaged stats: lr: 0.007776  min_lr: 0.007776  loss: 3.5358 (3.5531)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 1.0978 (1.0978)  acc1: 80.7292 (80.7292)  acc5: 97.3958 (97.3958)  time: 2.2650  data: 2.0092  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2945 (1.3493)  acc1: 75.0000 (73.5294)  acc5: 92.8571 (91.5601)  time: 0.6037  data: 0.4019  max mem: 57114
Test: Total time: 0:00:03 (0.6271 s / it)
* Acc@1 71.741 Acc@5 91.160 loss 1.438
Accuracy of the model on the 50000 test images: 71.7%
Max accuracy: 71.74%
Test:  [0/5]  eta: 0:00:10  loss: 7.1272 (7.1272)  acc1: 0.5208 (0.5208)  acc5: 1.0417 (1.0417)  time: 2.0330  data: 1.7896  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9894 (7.0124)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.5572  data: 0.3580  max mem: 57114
Test: Total time: 0:00:02 (0.5679 s / it)
* Acc@1 0.100 Acc@5 0.506 loss 7.004
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [50]  [  0/156]  eta: 0:13:46  lr: 0.007776  min_lr: 0.007776  loss: 4.3633 (4.3633)  weight_decay: 0.0500 (0.0500)  time: 5.2990  data: 3.7504  max mem: 57114
Epoch: [50]  [ 10/156]  eta: 0:02:52  lr: 0.007775  min_lr: 0.007775  loss: 3.8204 (3.6912)  weight_decay: 0.0500 (0.0500)  time: 1.1842  data: 0.3412  max mem: 57114
Epoch: [50]  [ 20/156]  eta: 0:02:11  lr: 0.007774  min_lr: 0.007774  loss: 3.6086 (3.7001)  weight_decay: 0.0500 (0.0500)  time: 0.7510  data: 0.0003  max mem: 57114
Epoch: [50]  [ 30/156]  eta: 0:01:51  lr: 0.007773  min_lr: 0.007773  loss: 3.5601 (3.5936)  weight_decay: 0.0500 (0.0500)  time: 0.7257  data: 0.0004  max mem: 57114
Epoch: [50]  [ 40/156]  eta: 0:01:39  lr: 0.007772  min_lr: 0.007772  loss: 3.5352 (3.5641)  weight_decay: 0.0500 (0.0500)  time: 0.7413  data: 0.0004  max mem: 57114
Epoch: [50]  [ 50/156]  eta: 0:01:28  lr: 0.007771  min_lr: 0.007771  loss: 3.7451 (3.5782)  weight_decay: 0.0500 (0.0500)  time: 0.7450  data: 0.0004  max mem: 57114
Epoch: [50]  [ 60/156]  eta: 0:01:18  lr: 0.007770  min_lr: 0.007770  loss: 3.6609 (3.5456)  weight_decay: 0.0500 (0.0500)  time: 0.7350  data: 0.0004  max mem: 57114
Epoch: [50]  [ 70/156]  eta: 0:01:09  lr: 0.007769  min_lr: 0.007769  loss: 3.7746 (3.5961)  weight_decay: 0.0500 (0.0500)  time: 0.7278  data: 0.0004  max mem: 57114
Epoch: [50]  [ 80/156]  eta: 0:01:00  lr: 0.007768  min_lr: 0.007768  loss: 3.7746 (3.5892)  weight_decay: 0.0500 (0.0500)  time: 0.7178  data: 0.0004  max mem: 57114
Epoch: [50]  [ 90/156]  eta: 0:00:51  lr: 0.007767  min_lr: 0.007767  loss: 3.6053 (3.5882)  weight_decay: 0.0500 (0.0500)  time: 0.7132  data: 0.0004  max mem: 57114
Epoch: [50]  [100/156]  eta: 0:00:43  lr: 0.007766  min_lr: 0.007766  loss: 3.7891 (3.6015)  weight_decay: 0.0500 (0.0500)  time: 0.7096  data: 0.0004  max mem: 57114
Epoch: [50]  [110/156]  eta: 0:00:35  lr: 0.007765  min_lr: 0.007765  loss: 3.7578 (3.6183)  weight_decay: 0.0500 (0.0500)  time: 0.7055  data: 0.0004  max mem: 57114
Epoch: [50]  [120/156]  eta: 0:00:27  lr: 0.007764  min_lr: 0.007764  loss: 3.4600 (3.6009)  weight_decay: 0.0500 (0.0500)  time: 0.6979  data: 0.0004  max mem: 57114
Epoch: [50]  [130/156]  eta: 0:00:19  lr: 0.007763  min_lr: 0.007763  loss: 3.5661 (3.5957)  weight_decay: 0.0500 (0.0500)  time: 0.6939  data: 0.0008  max mem: 57114
Epoch: [50]  [140/156]  eta: 0:00:12  lr: 0.007762  min_lr: 0.007762  loss: 3.6797 (3.6018)  weight_decay: 0.0500 (0.0500)  time: 0.6859  data: 0.0007  max mem: 57114
Epoch: [50]  [150/156]  eta: 0:00:04  lr: 0.007761  min_lr: 0.007761  loss: 3.7217 (3.6016)  weight_decay: 0.0500 (0.0500)  time: 0.6814  data: 0.0001  max mem: 57114
Epoch: [50]  [155/156]  eta: 0:00:00  lr: 0.007761  min_lr: 0.007761  loss: 3.6315 (3.5819)  weight_decay: 0.0500 (0.0500)  time: 0.6816  data: 0.0001  max mem: 57114
Epoch: [50] Total time: 0:01:56 (0.7474 s / it)
Averaged stats: lr: 0.007761  min_lr: 0.007761  loss: 3.6315 (3.5365)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0245 (1.0245)  acc1: 79.1667 (79.1667)  acc5: 94.7917 (94.7917)  time: 2.1677  data: 1.9117  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1999 (1.2385)  acc1: 78.1250 (72.6343)  acc5: 94.7917 (91.3044)  time: 0.5842  data: 0.3824  max mem: 57114
Test: Total time: 0:00:03 (0.6098 s / it)
* Acc@1 71.735 Acc@5 91.194 loss 1.316
Accuracy of the model on the 50000 test images: 71.7%
Max accuracy: 71.74%
Test:  [0/5]  eta: 0:00:11  loss: 7.1426 (7.1426)  acc1: 0.5208 (0.5208)  acc5: 1.0417 (1.0417)  time: 2.3284  data: 2.0848  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9919 (7.0208)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.6163  data: 0.4170  max mem: 57114
Test: Total time: 0:00:03 (0.6263 s / it)
* Acc@1 0.100 Acc@5 0.504 loss 7.012
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [51]  [  0/156]  eta: 0:13:34  lr: 0.007761  min_lr: 0.007761  loss: 3.7779 (3.7779)  weight_decay: 0.0500 (0.0500)  time: 5.2190  data: 3.5217  max mem: 57114
Epoch: [51]  [ 10/156]  eta: 0:02:48  lr: 0.007760  min_lr: 0.007760  loss: 3.6214 (3.4166)  weight_decay: 0.0500 (0.0500)  time: 1.1517  data: 0.3205  max mem: 57114
Epoch: [51]  [ 20/156]  eta: 0:02:08  lr: 0.007759  min_lr: 0.007759  loss: 3.5170 (3.4408)  weight_decay: 0.0500 (0.0500)  time: 0.7292  data: 0.0003  max mem: 57114
Epoch: [51]  [ 30/156]  eta: 0:01:49  lr: 0.007758  min_lr: 0.007758  loss: 3.4948 (3.4471)  weight_decay: 0.0500 (0.0500)  time: 0.7197  data: 0.0003  max mem: 57114
Epoch: [51]  [ 40/156]  eta: 0:01:36  lr: 0.007757  min_lr: 0.007757  loss: 3.4948 (3.4561)  weight_decay: 0.0500 (0.0500)  time: 0.7225  data: 0.0003  max mem: 57114
Epoch: [51]  [ 50/156]  eta: 0:01:25  lr: 0.007756  min_lr: 0.007756  loss: 3.6575 (3.4868)  weight_decay: 0.0500 (0.0500)  time: 0.7154  data: 0.0004  max mem: 57114
Epoch: [51]  [ 60/156]  eta: 0:01:16  lr: 0.007755  min_lr: 0.007755  loss: 3.4215 (3.4651)  weight_decay: 0.0500 (0.0500)  time: 0.7157  data: 0.0004  max mem: 57114
Epoch: [51]  [ 70/156]  eta: 0:01:07  lr: 0.007754  min_lr: 0.007754  loss: 3.3497 (3.4331)  weight_decay: 0.0500 (0.0500)  time: 0.7236  data: 0.0004  max mem: 57114
Epoch: [51]  [ 80/156]  eta: 0:00:58  lr: 0.007753  min_lr: 0.007753  loss: 3.5868 (3.4571)  weight_decay: 0.0500 (0.0500)  time: 0.7142  data: 0.0003  max mem: 57114
Epoch: [51]  [ 90/156]  eta: 0:00:50  lr: 0.007752  min_lr: 0.007752  loss: 3.6269 (3.4564)  weight_decay: 0.0500 (0.0500)  time: 0.7019  data: 0.0003  max mem: 57114
Epoch: [51]  [100/156]  eta: 0:00:42  lr: 0.007751  min_lr: 0.007751  loss: 3.3410 (3.4223)  weight_decay: 0.0500 (0.0500)  time: 0.7019  data: 0.0004  max mem: 57114
Epoch: [51]  [110/156]  eta: 0:00:34  lr: 0.007750  min_lr: 0.007750  loss: 3.5691 (3.4359)  weight_decay: 0.0500 (0.0500)  time: 0.7116  data: 0.0004  max mem: 57114
Epoch: [51]  [120/156]  eta: 0:00:27  lr: 0.007749  min_lr: 0.007749  loss: 3.8314 (3.4646)  weight_decay: 0.0500 (0.0500)  time: 0.7136  data: 0.0004  max mem: 57114
Epoch: [51]  [130/156]  eta: 0:00:19  lr: 0.007748  min_lr: 0.007748  loss: 3.7877 (3.4668)  weight_decay: 0.0500 (0.0500)  time: 0.7037  data: 0.0008  max mem: 57114
Epoch: [51]  [140/156]  eta: 0:00:11  lr: 0.007747  min_lr: 0.007747  loss: 3.7364 (3.4904)  weight_decay: 0.0500 (0.0500)  time: 0.6920  data: 0.0007  max mem: 57114
Epoch: [51]  [150/156]  eta: 0:00:04  lr: 0.007746  min_lr: 0.007746  loss: 3.6250 (3.4768)  weight_decay: 0.0500 (0.0500)  time: 0.6811  data: 0.0001  max mem: 57114
Epoch: [51]  [155/156]  eta: 0:00:00  lr: 0.007745  min_lr: 0.007745  loss: 3.5607 (3.4768)  weight_decay: 0.0500 (0.0500)  time: 0.6830  data: 0.0001  max mem: 57114
Epoch: [51] Total time: 0:01:55 (0.7410 s / it)
Averaged stats: lr: 0.007745  min_lr: 0.007745  loss: 3.5607 (3.5392)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0980 (1.0980)  acc1: 76.5625 (76.5625)  acc5: 96.3542 (96.3542)  time: 2.0932  data: 1.8370  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2148 (1.2129)  acc1: 73.4375 (72.3785)  acc5: 96.3542 (92.8389)  time: 0.5693  data: 0.3675  max mem: 57114
Test: Total time: 0:00:02 (0.5944 s / it)
* Acc@1 72.235 Acc@5 91.097 loss 1.310
Accuracy of the model on the 50000 test images: 72.2%
Max accuracy: 72.23%
Test:  [0/5]  eta: 0:00:09  loss: 7.1607 (7.1607)  acc1: 0.5208 (0.5208)  acc5: 1.0417 (1.0417)  time: 1.9825  data: 1.7390  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9991 (7.0307)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.5471  data: 0.3479  max mem: 57114
Test: Total time: 0:00:02 (0.5587 s / it)
* Acc@1 0.100 Acc@5 0.508 loss 7.022
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [52]  [  0/156]  eta: 0:13:03  lr: 0.007745  min_lr: 0.007745  loss: 3.9533 (3.9533)  weight_decay: 0.0500 (0.0500)  time: 5.0196  data: 3.8772  max mem: 57114
Epoch: [52]  [ 10/156]  eta: 0:02:49  lr: 0.007744  min_lr: 0.007744  loss: 3.5453 (3.4957)  weight_decay: 0.0500 (0.0500)  time: 1.1640  data: 0.3527  max mem: 57114
Epoch: [52]  [ 20/156]  eta: 0:02:09  lr: 0.007743  min_lr: 0.007743  loss: 3.4076 (3.4864)  weight_decay: 0.0500 (0.0500)  time: 0.7496  data: 0.0003  max mem: 57114
Epoch: [52]  [ 30/156]  eta: 0:01:51  lr: 0.007742  min_lr: 0.007742  loss: 3.3618 (3.4660)  weight_decay: 0.0500 (0.0500)  time: 0.7345  data: 0.0003  max mem: 57114
Epoch: [52]  [ 40/156]  eta: 0:01:38  lr: 0.007741  min_lr: 0.007741  loss: 3.3675 (3.4391)  weight_decay: 0.0500 (0.0500)  time: 0.7375  data: 0.0004  max mem: 57114
Epoch: [52]  [ 50/156]  eta: 0:01:27  lr: 0.007740  min_lr: 0.007740  loss: 3.4022 (3.4283)  weight_decay: 0.0500 (0.0500)  time: 0.7376  data: 0.0004  max mem: 57114
Epoch: [52]  [ 60/156]  eta: 0:01:17  lr: 0.007739  min_lr: 0.007739  loss: 3.4897 (3.4512)  weight_decay: 0.0500 (0.0500)  time: 0.7359  data: 0.0004  max mem: 57114
Epoch: [52]  [ 70/156]  eta: 0:01:08  lr: 0.007738  min_lr: 0.007738  loss: 3.6668 (3.4705)  weight_decay: 0.0500 (0.0500)  time: 0.7169  data: 0.0003  max mem: 57114
Epoch: [52]  [ 80/156]  eta: 0:00:59  lr: 0.007737  min_lr: 0.007737  loss: 3.7344 (3.4810)  weight_decay: 0.0500 (0.0500)  time: 0.7107  data: 0.0003  max mem: 57114
Epoch: [52]  [ 90/156]  eta: 0:00:51  lr: 0.007736  min_lr: 0.007736  loss: 3.6370 (3.4849)  weight_decay: 0.0500 (0.0500)  time: 0.7069  data: 0.0004  max mem: 57114
Epoch: [52]  [100/156]  eta: 0:00:43  lr: 0.007735  min_lr: 0.007735  loss: 3.5115 (3.4943)  weight_decay: 0.0500 (0.0500)  time: 0.7088  data: 0.0004  max mem: 57114
Epoch: [52]  [110/156]  eta: 0:00:35  lr: 0.007734  min_lr: 0.007734  loss: 3.5115 (3.4883)  weight_decay: 0.0500 (0.0500)  time: 0.7089  data: 0.0004  max mem: 57114
Epoch: [52]  [120/156]  eta: 0:00:27  lr: 0.007733  min_lr: 0.007733  loss: 3.3252 (3.4691)  weight_decay: 0.0500 (0.0500)  time: 0.7055  data: 0.0004  max mem: 57114
Epoch: [52]  [130/156]  eta: 0:00:19  lr: 0.007732  min_lr: 0.007732  loss: 3.3252 (3.4668)  weight_decay: 0.0500 (0.0500)  time: 0.7063  data: 0.0009  max mem: 57114
Epoch: [52]  [140/156]  eta: 0:00:12  lr: 0.007731  min_lr: 0.007731  loss: 3.7761 (3.4726)  weight_decay: 0.0500 (0.0500)  time: 0.6974  data: 0.0007  max mem: 57114
Epoch: [52]  [150/156]  eta: 0:00:04  lr: 0.007730  min_lr: 0.007730  loss: 3.7741 (3.4851)  weight_decay: 0.0500 (0.0500)  time: 0.6829  data: 0.0001  max mem: 57114
Epoch: [52]  [155/156]  eta: 0:00:00  lr: 0.007729  min_lr: 0.007729  loss: 3.7800 (3.5007)  weight_decay: 0.0500 (0.0500)  time: 0.6830  data: 0.0001  max mem: 57114
Epoch: [52] Total time: 0:01:56 (0.7464 s / it)
Averaged stats: lr: 0.007729  min_lr: 0.007729  loss: 3.7800 (3.5188)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.1849 (1.1849)  acc1: 79.1667 (79.1667)  acc5: 95.3125 (95.3125)  time: 2.0518  data: 1.7962  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.3330 (1.3451)  acc1: 75.0000 (73.4015)  acc5: 93.2292 (91.0486)  time: 0.5611  data: 0.3593  max mem: 57114
Test: Total time: 0:00:02 (0.5828 s / it)
* Acc@1 72.051 Acc@5 91.073 loss 1.424
Accuracy of the model on the 50000 test images: 72.1%
Max accuracy: 72.23%
Test:  [0/5]  eta: 0:00:11  loss: 7.1776 (7.1776)  acc1: 0.5208 (0.5208)  acc5: 1.0417 (1.0417)  time: 2.2536  data: 2.0102  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.0003 (7.0398)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.6013  data: 0.4021  max mem: 57114
Test: Total time: 0:00:03 (0.6146 s / it)
* Acc@1 0.100 Acc@5 0.508 loss 7.031
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [53]  [  0/156]  eta: 0:13:03  lr: 0.007729  min_lr: 0.007729  loss: 3.8379 (3.8379)  weight_decay: 0.0500 (0.0500)  time: 5.0214  data: 3.0864  max mem: 57114
Epoch: [53]  [ 10/156]  eta: 0:02:45  lr: 0.007728  min_lr: 0.007728  loss: 3.6787 (3.6091)  weight_decay: 0.0500 (0.0500)  time: 1.1320  data: 0.2811  max mem: 57114
Epoch: [53]  [ 20/156]  eta: 0:02:06  lr: 0.007727  min_lr: 0.007727  loss: 3.6396 (3.5858)  weight_decay: 0.0500 (0.0500)  time: 0.7268  data: 0.0004  max mem: 57114
Epoch: [53]  [ 30/156]  eta: 0:01:49  lr: 0.007726  min_lr: 0.007726  loss: 3.5191 (3.5297)  weight_decay: 0.0500 (0.0500)  time: 0.7184  data: 0.0003  max mem: 57114
Epoch: [53]  [ 40/156]  eta: 0:01:36  lr: 0.007725  min_lr: 0.007725  loss: 3.3988 (3.4916)  weight_decay: 0.0500 (0.0500)  time: 0.7286  data: 0.0003  max mem: 57114
Epoch: [53]  [ 50/156]  eta: 0:01:25  lr: 0.007724  min_lr: 0.007724  loss: 3.3988 (3.4732)  weight_decay: 0.0500 (0.0500)  time: 0.7252  data: 0.0004  max mem: 57114
Epoch: [53]  [ 60/156]  eta: 0:01:16  lr: 0.007723  min_lr: 0.007723  loss: 3.6269 (3.5007)  weight_decay: 0.0500 (0.0500)  time: 0.7197  data: 0.0004  max mem: 57114
Epoch: [53]  [ 70/156]  eta: 0:01:07  lr: 0.007722  min_lr: 0.007722  loss: 3.7664 (3.5117)  weight_decay: 0.0500 (0.0500)  time: 0.7128  data: 0.0004  max mem: 57114
Epoch: [53]  [ 80/156]  eta: 0:00:58  lr: 0.007721  min_lr: 0.007721  loss: 3.7013 (3.4798)  weight_decay: 0.0500 (0.0500)  time: 0.7046  data: 0.0003  max mem: 57114
Epoch: [53]  [ 90/156]  eta: 0:00:50  lr: 0.007720  min_lr: 0.007720  loss: 3.4244 (3.4984)  weight_decay: 0.0500 (0.0500)  time: 0.7060  data: 0.0004  max mem: 57114
Epoch: [53]  [100/156]  eta: 0:00:42  lr: 0.007718  min_lr: 0.007718  loss: 3.7287 (3.4971)  weight_decay: 0.0500 (0.0500)  time: 0.7062  data: 0.0003  max mem: 57114
Epoch: [53]  [110/156]  eta: 0:00:34  lr: 0.007717  min_lr: 0.007717  loss: 3.5550 (3.5145)  weight_decay: 0.0500 (0.0500)  time: 0.7044  data: 0.0004  max mem: 57114
Epoch: [53]  [120/156]  eta: 0:00:27  lr: 0.007716  min_lr: 0.007716  loss: 3.5550 (3.5267)  weight_decay: 0.0500 (0.0500)  time: 0.7089  data: 0.0004  max mem: 57114
Epoch: [53]  [130/156]  eta: 0:00:19  lr: 0.007715  min_lr: 0.007715  loss: 3.6271 (3.5271)  weight_decay: 0.0500 (0.0500)  time: 0.7050  data: 0.0008  max mem: 57114
Epoch: [53]  [140/156]  eta: 0:00:11  lr: 0.007714  min_lr: 0.007714  loss: 3.6611 (3.5360)  weight_decay: 0.0500 (0.0500)  time: 0.6929  data: 0.0007  max mem: 57114
Epoch: [53]  [150/156]  eta: 0:00:04  lr: 0.007713  min_lr: 0.007713  loss: 3.5675 (3.5188)  weight_decay: 0.0500 (0.0500)  time: 0.6841  data: 0.0001  max mem: 57114
Epoch: [53]  [155/156]  eta: 0:00:00  lr: 0.007713  min_lr: 0.007713  loss: 3.5675 (3.5205)  weight_decay: 0.0500 (0.0500)  time: 0.6895  data: 0.0001  max mem: 57114
Epoch: [53] Total time: 0:01:55 (0.7396 s / it)
Averaged stats: lr: 0.007713  min_lr: 0.007713  loss: 3.5675 (3.5205)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0100 (1.0100)  acc1: 77.0833 (77.0833)  acc5: 94.2708 (94.2708)  time: 2.1529  data: 1.8970  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2150 (1.1545)  acc1: 73.4375 (72.7622)  acc5: 93.2292 (91.6880)  time: 0.5813  data: 0.3795  max mem: 57114
Test: Total time: 0:00:03 (0.6052 s / it)
* Acc@1 73.416 Acc@5 91.820 loss 1.223
Accuracy of the model on the 50000 test images: 73.4%
Max accuracy: 73.42%
Test:  [0/5]  eta: 0:00:09  loss: 7.1942 (7.1942)  acc1: 0.5208 (0.5208)  acc5: 1.0417 (1.0417)  time: 1.9569  data: 1.7135  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.0092 (7.0503)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.5420  data: 0.3428  max mem: 57114
Test: Total time: 0:00:02 (0.5515 s / it)
* Acc@1 0.100 Acc@5 0.510 loss 7.041
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [54]  [  0/156]  eta: 0:13:53  lr: 0.007712  min_lr: 0.007712  loss: 3.5323 (3.5323)  weight_decay: 0.0500 (0.0500)  time: 5.3444  data: 3.9844  max mem: 57114
Epoch: [54]  [ 10/156]  eta: 0:02:47  lr: 0.007711  min_lr: 0.007711  loss: 3.5323 (3.3232)  weight_decay: 0.0500 (0.0500)  time: 1.1477  data: 0.3625  max mem: 57114
Epoch: [54]  [ 20/156]  eta: 0:02:07  lr: 0.007710  min_lr: 0.007710  loss: 3.7126 (3.5257)  weight_decay: 0.0500 (0.0500)  time: 0.7190  data: 0.0004  max mem: 57114
Epoch: [54]  [ 30/156]  eta: 0:01:50  lr: 0.007709  min_lr: 0.007709  loss: 3.7359 (3.5583)  weight_decay: 0.0500 (0.0500)  time: 0.7284  data: 0.0003  max mem: 57114
Epoch: [54]  [ 40/156]  eta: 0:01:37  lr: 0.007708  min_lr: 0.007708  loss: 3.6262 (3.5398)  weight_decay: 0.0500 (0.0500)  time: 0.7377  data: 0.0003  max mem: 57114
Epoch: [54]  [ 50/156]  eta: 0:01:27  lr: 0.007707  min_lr: 0.007707  loss: 3.6262 (3.4954)  weight_decay: 0.0500 (0.0500)  time: 0.7401  data: 0.0004  max mem: 57114
Epoch: [54]  [ 60/156]  eta: 0:01:17  lr: 0.007706  min_lr: 0.007706  loss: 3.3456 (3.4752)  weight_decay: 0.0500 (0.0500)  time: 0.7388  data: 0.0004  max mem: 57114
Epoch: [54]  [ 70/156]  eta: 0:01:08  lr: 0.007705  min_lr: 0.007705  loss: 3.5471 (3.5198)  weight_decay: 0.0500 (0.0500)  time: 0.7207  data: 0.0004  max mem: 57114
Epoch: [54]  [ 80/156]  eta: 0:00:59  lr: 0.007704  min_lr: 0.007704  loss: 3.7100 (3.5361)  weight_decay: 0.0500 (0.0500)  time: 0.7170  data: 0.0004  max mem: 57114
Epoch: [54]  [ 90/156]  eta: 0:00:51  lr: 0.007703  min_lr: 0.007703  loss: 3.7015 (3.5434)  weight_decay: 0.0500 (0.0500)  time: 0.7140  data: 0.0004  max mem: 57114
Epoch: [54]  [100/156]  eta: 0:00:43  lr: 0.007702  min_lr: 0.007702  loss: 3.6382 (3.5383)  weight_decay: 0.0500 (0.0500)  time: 0.7078  data: 0.0006  max mem: 57114
Epoch: [54]  [110/156]  eta: 0:00:35  lr: 0.007701  min_lr: 0.007701  loss: 3.6382 (3.5443)  weight_decay: 0.0500 (0.0500)  time: 0.7062  data: 0.0006  max mem: 57114
Epoch: [54]  [120/156]  eta: 0:00:27  lr: 0.007700  min_lr: 0.007700  loss: 3.6984 (3.5555)  weight_decay: 0.0500 (0.0500)  time: 0.7037  data: 0.0003  max mem: 57114
Epoch: [54]  [130/156]  eta: 0:00:19  lr: 0.007698  min_lr: 0.007698  loss: 3.7329 (3.5670)  weight_decay: 0.0500 (0.0500)  time: 0.6983  data: 0.0009  max mem: 57114
Epoch: [54]  [140/156]  eta: 0:00:11  lr: 0.007697  min_lr: 0.007697  loss: 3.7329 (3.5723)  weight_decay: 0.0500 (0.0500)  time: 0.6932  data: 0.0007  max mem: 57114
Epoch: [54]  [150/156]  eta: 0:00:04  lr: 0.007696  min_lr: 0.007696  loss: 3.6960 (3.5696)  weight_decay: 0.0500 (0.0500)  time: 0.6842  data: 0.0001  max mem: 57114
Epoch: [54]  [155/156]  eta: 0:00:00  lr: 0.007696  min_lr: 0.007696  loss: 3.6609 (3.5606)  weight_decay: 0.0500 (0.0500)  time: 0.6765  data: 0.0001  max mem: 57114
Epoch: [54] Total time: 0:01:56 (0.7444 s / it)
Averaged stats: lr: 0.007696  min_lr: 0.007696  loss: 3.6609 (3.5185)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9536 (0.9536)  acc1: 79.6875 (79.6875)  acc5: 95.8333 (95.8333)  time: 2.1727  data: 1.9169  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1526 (1.1209)  acc1: 72.9167 (73.0179)  acc5: 93.7500 (91.6880)  time: 0.5852  data: 0.3835  max mem: 57114
Test: Total time: 0:00:03 (0.6082 s / it)
* Acc@1 72.159 Acc@5 91.242 loss 1.255
Accuracy of the model on the 50000 test images: 72.2%
Max accuracy: 73.42%
Test:  [0/5]  eta: 0:00:11  loss: 7.2119 (7.2119)  acc1: 0.5208 (0.5208)  acc5: 1.0417 (1.0417)  time: 2.3523  data: 2.1088  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.0198 (7.0619)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.6212  data: 0.4219  max mem: 57114
Test: Total time: 0:00:03 (0.6563 s / it)
* Acc@1 0.100 Acc@5 0.516 loss 7.053
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [55]  [  0/156]  eta: 0:16:46  lr: 0.007696  min_lr: 0.007696  loss: 3.8874 (3.8874)  weight_decay: 0.0500 (0.0500)  time: 6.4550  data: 5.1667  max mem: 57114
Epoch: [55]  [ 10/156]  eta: 0:03:08  lr: 0.007694  min_lr: 0.007694  loss: 3.8099 (3.4303)  weight_decay: 0.0500 (0.0500)  time: 1.2928  data: 0.4701  max mem: 57114
Epoch: [55]  [ 20/156]  eta: 0:02:18  lr: 0.007693  min_lr: 0.007693  loss: 3.6382 (3.5367)  weight_decay: 0.0500 (0.0500)  time: 0.7465  data: 0.0004  max mem: 57114
Epoch: [55]  [ 30/156]  eta: 0:01:56  lr: 0.007692  min_lr: 0.007692  loss: 3.6673 (3.5795)  weight_decay: 0.0500 (0.0500)  time: 0.7206  data: 0.0004  max mem: 57114
Epoch: [55]  [ 40/156]  eta: 0:01:41  lr: 0.007691  min_lr: 0.007691  loss: 3.6673 (3.6020)  weight_decay: 0.0500 (0.0500)  time: 0.7230  data: 0.0004  max mem: 57114
Epoch: [55]  [ 50/156]  eta: 0:01:29  lr: 0.007690  min_lr: 0.007690  loss: 3.6614 (3.5937)  weight_decay: 0.0500 (0.0500)  time: 0.7221  data: 0.0003  max mem: 57114
Epoch: [55]  [ 60/156]  eta: 0:01:19  lr: 0.007689  min_lr: 0.007689  loss: 3.4770 (3.5678)  weight_decay: 0.0500 (0.0500)  time: 0.7236  data: 0.0003  max mem: 57114
Epoch: [55]  [ 70/156]  eta: 0:01:09  lr: 0.007688  min_lr: 0.007688  loss: 3.6285 (3.5706)  weight_decay: 0.0500 (0.0500)  time: 0.7214  data: 0.0004  max mem: 57114
Epoch: [55]  [ 80/156]  eta: 0:01:00  lr: 0.007687  min_lr: 0.007687  loss: 3.6285 (3.5289)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0004  max mem: 57114
Epoch: [55]  [ 90/156]  eta: 0:00:51  lr: 0.007686  min_lr: 0.007686  loss: 3.3940 (3.5283)  weight_decay: 0.0500 (0.0500)  time: 0.7020  data: 0.0004  max mem: 57114
Epoch: [55]  [100/156]  eta: 0:00:43  lr: 0.007684  min_lr: 0.007684  loss: 3.5662 (3.5185)  weight_decay: 0.0500 (0.0500)  time: 0.7062  data: 0.0004  max mem: 57114
Epoch: [55]  [110/156]  eta: 0:00:35  lr: 0.007683  min_lr: 0.007683  loss: 3.4568 (3.5043)  weight_decay: 0.0500 (0.0500)  time: 0.7096  data: 0.0004  max mem: 57114
Epoch: [55]  [120/156]  eta: 0:00:27  lr: 0.007682  min_lr: 0.007682  loss: 3.4568 (3.4971)  weight_decay: 0.0500 (0.0500)  time: 0.7055  data: 0.0004  max mem: 57114
Epoch: [55]  [130/156]  eta: 0:00:19  lr: 0.007681  min_lr: 0.007681  loss: 3.6941 (3.5044)  weight_decay: 0.0500 (0.0500)  time: 0.6969  data: 0.0008  max mem: 57114
Epoch: [55]  [140/156]  eta: 0:00:12  lr: 0.007680  min_lr: 0.007680  loss: 3.5527 (3.5089)  weight_decay: 0.0500 (0.0500)  time: 0.6866  data: 0.0007  max mem: 57114
Epoch: [55]  [150/156]  eta: 0:00:04  lr: 0.007679  min_lr: 0.007679  loss: 3.5527 (3.5106)  weight_decay: 0.0500 (0.0500)  time: 0.6824  data: 0.0001  max mem: 57114
Epoch: [55]  [155/156]  eta: 0:00:00  lr: 0.007678  min_lr: 0.007678  loss: 3.7050 (3.5128)  weight_decay: 0.0500 (0.0500)  time: 0.6840  data: 0.0001  max mem: 57114
Epoch: [55] Total time: 0:01:57 (0.7505 s / it)
Averaged stats: lr: 0.007678  min_lr: 0.007678  loss: 3.7050 (3.5055)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.1624 (1.1624)  acc1: 81.7708 (81.7708)  acc5: 95.3125 (95.3125)  time: 2.0821  data: 1.8264  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.3949 (1.3190)  acc1: 73.9583 (72.5064)  acc5: 92.1875 (90.7928)  time: 0.5671  data: 0.3653  max mem: 57114
Test: Total time: 0:00:02 (0.5908 s / it)
* Acc@1 71.555 Acc@5 90.945 loss 1.415
Accuracy of the model on the 50000 test images: 71.6%
Max accuracy: 73.42%
Test:  [0/5]  eta: 0:00:11  loss: 7.2335 (7.2335)  acc1: 0.5208 (0.5208)  acc5: 1.0417 (1.0417)  time: 2.3985  data: 2.1550  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.0320 (7.0751)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.5115)  time: 0.6304  data: 0.4311  max mem: 57114
Test: Total time: 0:00:03 (0.6430 s / it)
* Acc@1 0.100 Acc@5 0.525 loss 7.066
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [56]  [  0/156]  eta: 0:13:45  lr: 0.007678  min_lr: 0.007678  loss: 4.2096 (4.2096)  weight_decay: 0.0500 (0.0500)  time: 5.2892  data: 4.1597  max mem: 57114
Epoch: [56]  [ 10/156]  eta: 0:02:48  lr: 0.007677  min_lr: 0.007677  loss: 3.8185 (3.7317)  weight_decay: 0.0500 (0.0500)  time: 1.1542  data: 0.3785  max mem: 57114
Epoch: [56]  [ 20/156]  eta: 0:02:09  lr: 0.007676  min_lr: 0.007676  loss: 3.7670 (3.6989)  weight_decay: 0.0500 (0.0500)  time: 0.7375  data: 0.0003  max mem: 57114
Epoch: [56]  [ 30/156]  eta: 0:01:49  lr: 0.007675  min_lr: 0.007675  loss: 3.6170 (3.6762)  weight_decay: 0.0500 (0.0500)  time: 0.7176  data: 0.0003  max mem: 57114
Epoch: [56]  [ 40/156]  eta: 0:01:37  lr: 0.007674  min_lr: 0.007674  loss: 3.5612 (3.6508)  weight_decay: 0.0500 (0.0500)  time: 0.7147  data: 0.0004  max mem: 57114
Epoch: [56]  [ 50/156]  eta: 0:01:26  lr: 0.007672  min_lr: 0.007672  loss: 3.6405 (3.6273)  weight_decay: 0.0500 (0.0500)  time: 0.7194  data: 0.0004  max mem: 57114
Epoch: [56]  [ 60/156]  eta: 0:01:16  lr: 0.007671  min_lr: 0.007671  loss: 3.6405 (3.6171)  weight_decay: 0.0500 (0.0500)  time: 0.7108  data: 0.0004  max mem: 57114
Epoch: [56]  [ 70/156]  eta: 0:01:07  lr: 0.007670  min_lr: 0.007670  loss: 3.4231 (3.5719)  weight_decay: 0.0500 (0.0500)  time: 0.7171  data: 0.0004  max mem: 57114
Epoch: [56]  [ 80/156]  eta: 0:00:58  lr: 0.007669  min_lr: 0.007669  loss: 3.4231 (3.5601)  weight_decay: 0.0500 (0.0500)  time: 0.7137  data: 0.0004  max mem: 57114
Epoch: [56]  [ 90/156]  eta: 0:00:50  lr: 0.007668  min_lr: 0.007668  loss: 3.4820 (3.5559)  weight_decay: 0.0500 (0.0500)  time: 0.7024  data: 0.0004  max mem: 57114
Epoch: [56]  [100/156]  eta: 0:00:42  lr: 0.007667  min_lr: 0.007667  loss: 3.5966 (3.5510)  weight_decay: 0.0500 (0.0500)  time: 0.7077  data: 0.0004  max mem: 57114
Epoch: [56]  [110/156]  eta: 0:00:34  lr: 0.007666  min_lr: 0.007666  loss: 3.6356 (3.5597)  weight_decay: 0.0500 (0.0500)  time: 0.7083  data: 0.0004  max mem: 57114
Epoch: [56]  [120/156]  eta: 0:00:27  lr: 0.007664  min_lr: 0.007664  loss: 3.5847 (3.5459)  weight_decay: 0.0500 (0.0500)  time: 0.7037  data: 0.0004  max mem: 57114
Epoch: [56]  [130/156]  eta: 0:00:19  lr: 0.007663  min_lr: 0.007663  loss: 3.4700 (3.5377)  weight_decay: 0.0500 (0.0500)  time: 0.7023  data: 0.0008  max mem: 57114
Epoch: [56]  [140/156]  eta: 0:00:11  lr: 0.007662  min_lr: 0.007662  loss: 3.8018 (3.5371)  weight_decay: 0.0500 (0.0500)  time: 0.6935  data: 0.0007  max mem: 57114
Epoch: [56]  [150/156]  eta: 0:00:04  lr: 0.007661  min_lr: 0.007661  loss: 3.7980 (3.5411)  weight_decay: 0.0500 (0.0500)  time: 0.6831  data: 0.0001  max mem: 57114
Epoch: [56]  [155/156]  eta: 0:00:00  lr: 0.007660  min_lr: 0.007660  loss: 3.5942 (3.5325)  weight_decay: 0.0500 (0.0500)  time: 0.6839  data: 0.0001  max mem: 57114
Epoch: [56] Total time: 0:01:55 (0.7403 s / it)
Averaged stats: lr: 0.007660  min_lr: 0.007660  loss: 3.5942 (3.4893)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0405 (1.0405)  acc1: 78.6458 (78.6458)  acc5: 95.3125 (95.3125)  time: 2.0991  data: 1.8434  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2091 (1.1908)  acc1: 75.5208 (73.6573)  acc5: 93.2292 (91.1765)  time: 0.5705  data: 0.3688  max mem: 57114
Test: Total time: 0:00:02 (0.5945 s / it)
* Acc@1 71.457 Acc@5 90.783 loss 1.337
Accuracy of the model on the 50000 test images: 71.5%
Max accuracy: 73.42%
Test:  [0/5]  eta: 0:00:11  loss: 7.2559 (7.2559)  acc1: 0.5208 (0.5208)  acc5: 1.0417 (1.0417)  time: 2.2787  data: 2.0352  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.0449 (7.0877)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.5115)  time: 0.6064  data: 0.4071  max mem: 57114
Test: Total time: 0:00:03 (0.6202 s / it)
* Acc@1 0.100 Acc@5 0.527 loss 7.078
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [57]  [  0/156]  eta: 0:13:04  lr: 0.007660  min_lr: 0.007660  loss: 2.7292 (2.7292)  weight_decay: 0.0500 (0.0500)  time: 5.0319  data: 3.4790  max mem: 57114
Epoch: [57]  [ 10/156]  eta: 0:02:44  lr: 0.007659  min_lr: 0.007659  loss: 3.4683 (3.2837)  weight_decay: 0.0500 (0.0500)  time: 1.1250  data: 0.3166  max mem: 57114
Epoch: [57]  [ 20/156]  eta: 0:02:07  lr: 0.007658  min_lr: 0.007658  loss: 3.6138 (3.4387)  weight_decay: 0.0500 (0.0500)  time: 0.7323  data: 0.0003  max mem: 57114
Epoch: [57]  [ 30/156]  eta: 0:01:48  lr: 0.007657  min_lr: 0.007657  loss: 3.5899 (3.4703)  weight_decay: 0.0500 (0.0500)  time: 0.7184  data: 0.0003  max mem: 57114
Epoch: [57]  [ 40/156]  eta: 0:01:36  lr: 0.007656  min_lr: 0.007656  loss: 3.5899 (3.5136)  weight_decay: 0.0500 (0.0500)  time: 0.7209  data: 0.0004  max mem: 57114
Epoch: [57]  [ 50/156]  eta: 0:01:25  lr: 0.007654  min_lr: 0.007654  loss: 3.6538 (3.5122)  weight_decay: 0.0500 (0.0500)  time: 0.7243  data: 0.0005  max mem: 57114
Epoch: [57]  [ 60/156]  eta: 0:01:16  lr: 0.007653  min_lr: 0.007653  loss: 3.1860 (3.4792)  weight_decay: 0.0500 (0.0500)  time: 0.7158  data: 0.0005  max mem: 57114
Epoch: [57]  [ 70/156]  eta: 0:01:07  lr: 0.007652  min_lr: 0.007652  loss: 3.5683 (3.5230)  weight_decay: 0.0500 (0.0500)  time: 0.7110  data: 0.0004  max mem: 57114
Epoch: [57]  [ 80/156]  eta: 0:00:58  lr: 0.007651  min_lr: 0.007651  loss: 3.8468 (3.5582)  weight_decay: 0.0500 (0.0500)  time: 0.7071  data: 0.0004  max mem: 57114
Epoch: [57]  [ 90/156]  eta: 0:00:50  lr: 0.007650  min_lr: 0.007650  loss: 3.6798 (3.5200)  weight_decay: 0.0500 (0.0500)  time: 0.7075  data: 0.0004  max mem: 57114
Epoch: [57]  [100/156]  eta: 0:00:42  lr: 0.007649  min_lr: 0.007649  loss: 3.4855 (3.5251)  weight_decay: 0.0500 (0.0500)  time: 0.7057  data: 0.0004  max mem: 57114
Epoch: [57]  [110/156]  eta: 0:00:34  lr: 0.007647  min_lr: 0.007647  loss: 3.4371 (3.4842)  weight_decay: 0.0500 (0.0500)  time: 0.7087  data: 0.0004  max mem: 57114
Epoch: [57]  [120/156]  eta: 0:00:27  lr: 0.007646  min_lr: 0.007646  loss: 3.2043 (3.4612)  weight_decay: 0.0500 (0.0500)  time: 0.7056  data: 0.0004  max mem: 57114
Epoch: [57]  [130/156]  eta: 0:00:19  lr: 0.007645  min_lr: 0.007645  loss: 3.7246 (3.4873)  weight_decay: 0.0500 (0.0500)  time: 0.6965  data: 0.0009  max mem: 57114
Epoch: [57]  [140/156]  eta: 0:00:11  lr: 0.007644  min_lr: 0.007644  loss: 3.7803 (3.4893)  weight_decay: 0.0500 (0.0500)  time: 0.6859  data: 0.0007  max mem: 57114
Epoch: [57]  [150/156]  eta: 0:00:04  lr: 0.007643  min_lr: 0.007643  loss: 3.4048 (3.4848)  weight_decay: 0.0500 (0.0500)  time: 0.6797  data: 0.0001  max mem: 57114
Epoch: [57]  [155/156]  eta: 0:00:00  lr: 0.007642  min_lr: 0.007642  loss: 3.3662 (3.4739)  weight_decay: 0.0500 (0.0500)  time: 0.6792  data: 0.0001  max mem: 57114
Epoch: [57] Total time: 0:01:55 (0.7375 s / it)
Averaged stats: lr: 0.007642  min_lr: 0.007642  loss: 3.3662 (3.4826)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8720 (0.8720)  acc1: 79.1667 (79.1667)  acc5: 96.8750 (96.8750)  time: 2.0528  data: 1.7975  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0847 (1.0073)  acc1: 76.0417 (75.1918)  acc5: 93.7500 (92.5831)  time: 0.5612  data: 0.3596  max mem: 57114
Test: Total time: 0:00:02 (0.5984 s / it)
* Acc@1 73.108 Acc@5 91.694 loss 1.121
Accuracy of the model on the 50000 test images: 73.1%
Max accuracy: 73.42%
Test:  [0/5]  eta: 0:00:11  loss: 7.2810 (7.2810)  acc1: 0.5208 (0.5208)  acc5: 1.0417 (1.0417)  time: 2.3510  data: 2.1075  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.0594 (7.1023)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.5115)  time: 0.6208  data: 0.4216  max mem: 57114
Test: Total time: 0:00:03 (0.6374 s / it)
* Acc@1 0.100 Acc@5 0.527 loss 7.092
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [58]  [  0/156]  eta: 0:12:44  lr: 0.007642  min_lr: 0.007642  loss: 2.8505 (2.8505)  weight_decay: 0.0500 (0.0500)  time: 4.8988  data: 3.6749  max mem: 57114
Epoch: [58]  [ 10/156]  eta: 0:02:45  lr: 0.007641  min_lr: 0.007641  loss: 3.4130 (3.3069)  weight_decay: 0.0500 (0.0500)  time: 1.1362  data: 0.3344  max mem: 57114
Epoch: [58]  [ 20/156]  eta: 0:02:07  lr: 0.007640  min_lr: 0.007640  loss: 3.4341 (3.3099)  weight_decay: 0.0500 (0.0500)  time: 0.7405  data: 0.0003  max mem: 57114
Epoch: [58]  [ 30/156]  eta: 0:01:49  lr: 0.007638  min_lr: 0.007638  loss: 3.4704 (3.3767)  weight_decay: 0.0500 (0.0500)  time: 0.7275  data: 0.0003  max mem: 57114
Epoch: [58]  [ 40/156]  eta: 0:01:36  lr: 0.007637  min_lr: 0.007637  loss: 3.5553 (3.4066)  weight_decay: 0.0500 (0.0500)  time: 0.7253  data: 0.0003  max mem: 57114
Epoch: [58]  [ 50/156]  eta: 0:01:25  lr: 0.007636  min_lr: 0.007636  loss: 3.4882 (3.4224)  weight_decay: 0.0500 (0.0500)  time: 0.7073  data: 0.0004  max mem: 57114
Epoch: [58]  [ 60/156]  eta: 0:01:16  lr: 0.007635  min_lr: 0.007635  loss: 3.6109 (3.4823)  weight_decay: 0.0500 (0.0500)  time: 0.7177  data: 0.0004  max mem: 57114
Epoch: [58]  [ 70/156]  eta: 0:01:07  lr: 0.007634  min_lr: 0.007634  loss: 3.5749 (3.4509)  weight_decay: 0.0500 (0.0500)  time: 0.7258  data: 0.0004  max mem: 57114
Epoch: [58]  [ 80/156]  eta: 0:00:58  lr: 0.007632  min_lr: 0.007632  loss: 3.3519 (3.4490)  weight_decay: 0.0500 (0.0500)  time: 0.7072  data: 0.0004  max mem: 57114
Epoch: [58]  [ 90/156]  eta: 0:00:50  lr: 0.007631  min_lr: 0.007631  loss: 3.3687 (3.4415)  weight_decay: 0.0500 (0.0500)  time: 0.7005  data: 0.0004  max mem: 57114
Epoch: [58]  [100/156]  eta: 0:00:42  lr: 0.007630  min_lr: 0.007630  loss: 3.5949 (3.4419)  weight_decay: 0.0500 (0.0500)  time: 0.7029  data: 0.0004  max mem: 57114
Epoch: [58]  [110/156]  eta: 0:00:34  lr: 0.007629  min_lr: 0.007629  loss: 3.3722 (3.4312)  weight_decay: 0.0500 (0.0500)  time: 0.7102  data: 0.0004  max mem: 57114
Epoch: [58]  [120/156]  eta: 0:00:27  lr: 0.007628  min_lr: 0.007628  loss: 3.3722 (3.4450)  weight_decay: 0.0500 (0.0500)  time: 0.7281  data: 0.0004  max mem: 57114
Epoch: [58]  [130/156]  eta: 0:00:19  lr: 0.007626  min_lr: 0.007626  loss: 3.4744 (3.4262)  weight_decay: 0.0500 (0.0500)  time: 0.7312  data: 0.0010  max mem: 57114
Epoch: [58]  [140/156]  eta: 0:00:11  lr: 0.007625  min_lr: 0.007625  loss: 3.3738 (3.4282)  weight_decay: 0.0500 (0.0500)  time: 0.7014  data: 0.0008  max mem: 57114
Epoch: [58]  [150/156]  eta: 0:00:04  lr: 0.007624  min_lr: 0.007624  loss: 3.4512 (3.4331)  weight_decay: 0.0500 (0.0500)  time: 0.6847  data: 0.0001  max mem: 57114
Epoch: [58]  [155/156]  eta: 0:00:00  lr: 0.007623  min_lr: 0.007623  loss: 3.4512 (3.4306)  weight_decay: 0.0500 (0.0500)  time: 0.6796  data: 0.0001  max mem: 57114
Epoch: [58] Total time: 0:01:55 (0.7430 s / it)
Averaged stats: lr: 0.007623  min_lr: 0.007623  loss: 3.4512 (3.4823)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.3089 (1.3089)  acc1: 78.6458 (78.6458)  acc5: 93.7500 (93.7500)  time: 2.0679  data: 1.8121  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.3878 (1.4330)  acc1: 73.9583 (70.7161)  acc5: 91.1458 (89.7698)  time: 0.5642  data: 0.3625  max mem: 57114
Test: Total time: 0:00:02 (0.5870 s / it)
* Acc@1 69.377 Acc@5 89.314 loss 1.484
Accuracy of the model on the 50000 test images: 69.4%
Max accuracy: 73.42%
Test:  [0/5]  eta: 0:00:11  loss: 7.3049 (7.3049)  acc1: 0.5208 (0.5208)  acc5: 1.0417 (1.0417)  time: 2.3106  data: 2.0671  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.0749 (7.1174)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.5115)  time: 0.6127  data: 0.4135  max mem: 57114
Test: Total time: 0:00:03 (0.6285 s / it)
* Acc@1 0.100 Acc@5 0.535 loss 7.106
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [59]  [  0/156]  eta: 0:12:53  lr: 0.007623  min_lr: 0.007623  loss: 3.7330 (3.7330)  weight_decay: 0.0500 (0.0500)  time: 4.9558  data: 3.6774  max mem: 57114
Epoch: [59]  [ 10/156]  eta: 0:02:44  lr: 0.007622  min_lr: 0.007622  loss: 3.8479 (3.5451)  weight_decay: 0.0500 (0.0500)  time: 1.1275  data: 0.3346  max mem: 57114
Epoch: [59]  [ 20/156]  eta: 0:02:07  lr: 0.007621  min_lr: 0.007621  loss: 3.8362 (3.6544)  weight_decay: 0.0500 (0.0500)  time: 0.7354  data: 0.0003  max mem: 57114
Epoch: [59]  [ 30/156]  eta: 0:01:49  lr: 0.007620  min_lr: 0.007620  loss: 3.4887 (3.5440)  weight_decay: 0.0500 (0.0500)  time: 0.7319  data: 0.0004  max mem: 57114
Epoch: [59]  [ 40/156]  eta: 0:01:37  lr: 0.007618  min_lr: 0.007618  loss: 3.3003 (3.4926)  weight_decay: 0.0500 (0.0500)  time: 0.7326  data: 0.0004  max mem: 57114
Epoch: [59]  [ 50/156]  eta: 0:01:25  lr: 0.007617  min_lr: 0.007617  loss: 3.4792 (3.5045)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0004  max mem: 57114
Epoch: [59]  [ 60/156]  eta: 0:01:16  lr: 0.007616  min_lr: 0.007616  loss: 3.7432 (3.4953)  weight_decay: 0.0500 (0.0500)  time: 0.7097  data: 0.0004  max mem: 57114
Epoch: [59]  [ 70/156]  eta: 0:01:07  lr: 0.007615  min_lr: 0.007615  loss: 3.5582 (3.4916)  weight_decay: 0.0500 (0.0500)  time: 0.7144  data: 0.0004  max mem: 57114
Epoch: [59]  [ 80/156]  eta: 0:00:59  lr: 0.007613  min_lr: 0.007613  loss: 3.5936 (3.4959)  weight_decay: 0.0500 (0.0500)  time: 0.7243  data: 0.0004  max mem: 57114
Epoch: [59]  [ 90/156]  eta: 0:00:50  lr: 0.007612  min_lr: 0.007612  loss: 3.7791 (3.5023)  weight_decay: 0.0500 (0.0500)  time: 0.7334  data: 0.0005  max mem: 57114
Epoch: [59]  [100/156]  eta: 0:00:43  lr: 0.007611  min_lr: 0.007611  loss: 3.6539 (3.4815)  weight_decay: 0.0500 (0.0500)  time: 0.7412  data: 0.0004  max mem: 57114
Epoch: [59]  [110/156]  eta: 0:00:35  lr: 0.007610  min_lr: 0.007610  loss: 3.3252 (3.4771)  weight_decay: 0.0500 (0.0500)  time: 0.7442  data: 0.0004  max mem: 57114
Epoch: [59]  [120/156]  eta: 0:00:27  lr: 0.007608  min_lr: 0.007608  loss: 3.4706 (3.4768)  weight_decay: 0.0500 (0.0500)  time: 0.7325  data: 0.0004  max mem: 57114
Epoch: [59]  [130/156]  eta: 0:00:19  lr: 0.007607  min_lr: 0.007607  loss: 3.3815 (3.4530)  weight_decay: 0.0500 (0.0500)  time: 0.7229  data: 0.0009  max mem: 57114
Epoch: [59]  [140/156]  eta: 0:00:12  lr: 0.007606  min_lr: 0.007606  loss: 3.1326 (3.4448)  weight_decay: 0.0500 (0.0500)  time: 0.7173  data: 0.0007  max mem: 57114
Epoch: [59]  [150/156]  eta: 0:00:04  lr: 0.007605  min_lr: 0.007605  loss: 3.5765 (3.4461)  weight_decay: 0.0500 (0.0500)  time: 0.7165  data: 0.0001  max mem: 57114
Epoch: [59]  [155/156]  eta: 0:00:00  lr: 0.007604  min_lr: 0.007604  loss: 3.5765 (3.4547)  weight_decay: 0.0500 (0.0500)  time: 0.7164  data: 0.0001  max mem: 57114
Epoch: [59] Total time: 0:01:57 (0.7558 s / it)
Averaged stats: lr: 0.007604  min_lr: 0.007604  loss: 3.5765 (3.4714)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:15  loss: 1.1505 (1.1505)  acc1: 80.2083 (80.2083)  acc5: 95.3125 (95.3125)  time: 3.0447  data: 2.7854  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2777 (1.2145)  acc1: 71.3542 (71.4834)  acc5: 94.7917 (91.6880)  time: 0.7615  data: 0.5572  max mem: 57114
Test: Total time: 0:00:04 (0.8108 s / it)
* Acc@1 71.825 Acc@5 91.202 loss 1.300
Accuracy of the model on the 50000 test images: 71.8%
Max accuracy: 73.42%
Test:  [0/5]  eta: 0:00:16  loss: 7.3262 (7.3262)  acc1: 0.5208 (0.5208)  acc5: 1.0417 (1.0417)  time: 3.3372  data: 3.0938  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.0908 (7.1314)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.5115)  time: 0.8181  data: 0.6189  max mem: 57114
Test: Total time: 0:00:04 (0.8406 s / it)
* Acc@1 0.100 Acc@5 0.529 loss 7.120
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [60]  [  0/156]  eta: 0:14:01  lr: 0.007604  min_lr: 0.007604  loss: 4.3764 (4.3764)  weight_decay: 0.0500 (0.0500)  time: 5.3932  data: 4.3267  max mem: 57114
Epoch: [60]  [ 10/156]  eta: 0:02:51  lr: 0.007603  min_lr: 0.007603  loss: 3.8104 (3.7247)  weight_decay: 0.0500 (0.0500)  time: 1.1749  data: 0.3936  max mem: 57114
Epoch: [60]  [ 20/156]  eta: 0:02:12  lr: 0.007601  min_lr: 0.007601  loss: 3.6127 (3.5432)  weight_decay: 0.0500 (0.0500)  time: 0.7535  data: 0.0004  max mem: 57114
Epoch: [60]  [ 30/156]  eta: 0:01:53  lr: 0.007600  min_lr: 0.007600  loss: 3.4359 (3.5370)  weight_decay: 0.0500 (0.0500)  time: 0.7527  data: 0.0007  max mem: 57114
Epoch: [60]  [ 40/156]  eta: 0:01:40  lr: 0.007599  min_lr: 0.007599  loss: 3.7351 (3.5424)  weight_decay: 0.0500 (0.0500)  time: 0.7517  data: 0.0007  max mem: 57114
Epoch: [60]  [ 50/156]  eta: 0:01:30  lr: 0.007598  min_lr: 0.007598  loss: 3.7471 (3.5655)  weight_decay: 0.0500 (0.0500)  time: 0.7757  data: 0.0004  max mem: 57114
Epoch: [60]  [ 60/156]  eta: 0:01:20  lr: 0.007596  min_lr: 0.007596  loss: 3.8245 (3.5919)  weight_decay: 0.0500 (0.0500)  time: 0.7891  data: 0.0004  max mem: 57114
Epoch: [60]  [ 70/156]  eta: 0:01:11  lr: 0.007595  min_lr: 0.007595  loss: 3.7669 (3.5942)  weight_decay: 0.0500 (0.0500)  time: 0.7683  data: 0.0006  max mem: 57114
Epoch: [60]  [ 80/156]  eta: 0:01:02  lr: 0.007594  min_lr: 0.007594  loss: 3.5981 (3.5846)  weight_decay: 0.0500 (0.0500)  time: 0.7660  data: 0.0006  max mem: 57114
Epoch: [60]  [ 90/156]  eta: 0:00:54  lr: 0.007593  min_lr: 0.007593  loss: 3.5995 (3.5774)  weight_decay: 0.0500 (0.0500)  time: 0.8024  data: 0.0004  max mem: 57114
Epoch: [60]  [100/156]  eta: 0:00:45  lr: 0.007591  min_lr: 0.007591  loss: 3.3381 (3.5473)  weight_decay: 0.0500 (0.0500)  time: 0.7891  data: 0.0004  max mem: 57114
Epoch: [60]  [110/156]  eta: 0:00:37  lr: 0.007590  min_lr: 0.007590  loss: 3.4493 (3.5526)  weight_decay: 0.0500 (0.0500)  time: 0.7426  data: 0.0004  max mem: 57114
Epoch: [60]  [120/156]  eta: 0:00:28  lr: 0.007589  min_lr: 0.007589  loss: 3.6022 (3.5595)  weight_decay: 0.0500 (0.0500)  time: 0.7208  data: 0.0004  max mem: 57114
Epoch: [60]  [130/156]  eta: 0:00:20  lr: 0.007588  min_lr: 0.007588  loss: 3.5330 (3.5645)  weight_decay: 0.0500 (0.0500)  time: 0.6983  data: 0.0009  max mem: 57114
Epoch: [60]  [140/156]  eta: 0:00:12  lr: 0.007586  min_lr: 0.007586  loss: 3.4426 (3.5615)  weight_decay: 0.0500 (0.0500)  time: 0.6855  data: 0.0007  max mem: 57114
Epoch: [60]  [150/156]  eta: 0:00:04  lr: 0.007585  min_lr: 0.007585  loss: 3.0824 (3.5164)  weight_decay: 0.0500 (0.0500)  time: 0.6812  data: 0.0001  max mem: 57114
Epoch: [60]  [155/156]  eta: 0:00:00  lr: 0.007584  min_lr: 0.007584  loss: 2.8628 (3.4963)  weight_decay: 0.0500 (0.0500)  time: 0.6873  data: 0.0001  max mem: 57114
Epoch: [60] Total time: 0:02:01 (0.7769 s / it)
Averaged stats: lr: 0.007584  min_lr: 0.007584  loss: 2.8628 (3.4692)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8561 (0.8561)  acc1: 79.1667 (79.1667)  acc5: 96.8750 (96.8750)  time: 2.0885  data: 1.8332  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0686 (1.0470)  acc1: 76.0417 (74.6803)  acc5: 94.2708 (91.8159)  time: 0.5684  data: 0.3667  max mem: 57114
Test: Total time: 0:00:02 (0.5907 s / it)
* Acc@1 73.422 Acc@5 91.886 loss 1.138
Accuracy of the model on the 50000 test images: 73.4%
Max accuracy: 73.42%
Test:  [0/5]  eta: 0:00:10  loss: 7.3463 (7.3463)  acc1: 0.5208 (0.5208)  acc5: 0.5208 (0.5208)  time: 2.0588  data: 1.8156  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.1069 (7.1467)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.5623  data: 0.3632  max mem: 57114
Test: Total time: 0:00:02 (0.5734 s / it)
* Acc@1 0.100 Acc@5 0.518 loss 7.135
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [61]  [  0/156]  eta: 0:13:07  lr: 0.007584  min_lr: 0.007584  loss: 4.0122 (4.0122)  weight_decay: 0.0500 (0.0500)  time: 5.0476  data: 3.6372  max mem: 57114
Epoch: [61]  [ 10/156]  eta: 0:02:49  lr: 0.007583  min_lr: 0.007583  loss: 3.6508 (3.5175)  weight_decay: 0.0500 (0.0500)  time: 1.1636  data: 0.3310  max mem: 57114
Epoch: [61]  [ 20/156]  eta: 0:02:10  lr: 0.007582  min_lr: 0.007582  loss: 3.6128 (3.6000)  weight_decay: 0.0500 (0.0500)  time: 0.7565  data: 0.0003  max mem: 57114
Epoch: [61]  [ 30/156]  eta: 0:01:51  lr: 0.007580  min_lr: 0.007580  loss: 3.5233 (3.5489)  weight_decay: 0.0500 (0.0500)  time: 0.7293  data: 0.0003  max mem: 57114
Epoch: [61]  [ 40/156]  eta: 0:01:37  lr: 0.007579  min_lr: 0.007579  loss: 3.4238 (3.5258)  weight_decay: 0.0500 (0.0500)  time: 0.7222  data: 0.0004  max mem: 57114
Epoch: [61]  [ 50/156]  eta: 0:01:27  lr: 0.007578  min_lr: 0.007578  loss: 3.5470 (3.5180)  weight_decay: 0.0500 (0.0500)  time: 0.7253  data: 0.0004  max mem: 57114
Epoch: [61]  [ 60/156]  eta: 0:01:17  lr: 0.007577  min_lr: 0.007577  loss: 3.5212 (3.5071)  weight_decay: 0.0500 (0.0500)  time: 0.7203  data: 0.0004  max mem: 57114
Epoch: [61]  [ 70/156]  eta: 0:01:08  lr: 0.007575  min_lr: 0.007575  loss: 3.5857 (3.5034)  weight_decay: 0.0500 (0.0500)  time: 0.7244  data: 0.0004  max mem: 57114
Epoch: [61]  [ 80/156]  eta: 0:00:59  lr: 0.007574  min_lr: 0.007574  loss: 3.4190 (3.4629)  weight_decay: 0.0500 (0.0500)  time: 0.7208  data: 0.0004  max mem: 57114
Epoch: [61]  [ 90/156]  eta: 0:00:51  lr: 0.007573  min_lr: 0.007573  loss: 3.3506 (3.4644)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0004  max mem: 57114
Epoch: [61]  [100/156]  eta: 0:00:43  lr: 0.007571  min_lr: 0.007571  loss: 3.3977 (3.4543)  weight_decay: 0.0500 (0.0500)  time: 0.7069  data: 0.0004  max mem: 57114
Epoch: [61]  [110/156]  eta: 0:00:35  lr: 0.007570  min_lr: 0.007570  loss: 3.6111 (3.4824)  weight_decay: 0.0500 (0.0500)  time: 0.6988  data: 0.0004  max mem: 57114
Epoch: [61]  [120/156]  eta: 0:00:27  lr: 0.007569  min_lr: 0.007569  loss: 3.6642 (3.4701)  weight_decay: 0.0500 (0.0500)  time: 0.7019  data: 0.0004  max mem: 57114
Epoch: [61]  [130/156]  eta: 0:00:19  lr: 0.007567  min_lr: 0.007567  loss: 3.2973 (3.4506)  weight_decay: 0.0500 (0.0500)  time: 0.6976  data: 0.0009  max mem: 57114
Epoch: [61]  [140/156]  eta: 0:00:11  lr: 0.007566  min_lr: 0.007566  loss: 3.4831 (3.4643)  weight_decay: 0.0500 (0.0500)  time: 0.6896  data: 0.0007  max mem: 57114
Epoch: [61]  [150/156]  eta: 0:00:04  lr: 0.007565  min_lr: 0.007565  loss: 3.6495 (3.4723)  weight_decay: 0.0500 (0.0500)  time: 0.6855  data: 0.0001  max mem: 57114
Epoch: [61]  [155/156]  eta: 0:00:00  lr: 0.007564  min_lr: 0.007564  loss: 3.6455 (3.4829)  weight_decay: 0.0500 (0.0500)  time: 0.6805  data: 0.0001  max mem: 57114
Epoch: [61] Total time: 0:01:55 (0.7431 s / it)
Averaged stats: lr: 0.007564  min_lr: 0.007564  loss: 3.6455 (3.4590)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0752 (1.0752)  acc1: 78.6458 (78.6458)  acc5: 96.8750 (96.8750)  time: 2.0551  data: 1.7993  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2075 (1.1987)  acc1: 76.5625 (73.9130)  acc5: 95.3125 (93.4783)  time: 0.5617  data: 0.3599  max mem: 57114
Test: Total time: 0:00:02 (0.5832 s / it)
* Acc@1 73.801 Acc@5 92.150 loss 1.281
Accuracy of the model on the 50000 test images: 73.8%
Max accuracy: 73.80%
Test:  [0/5]  eta: 0:00:11  loss: 7.3661 (7.3661)  acc1: 0.5208 (0.5208)  acc5: 0.5208 (0.5208)  time: 2.2012  data: 1.9577  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.1230 (7.1601)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.5909  data: 0.3916  max mem: 57114
Test: Total time: 0:00:03 (0.6047 s / it)
* Acc@1 0.100 Acc@5 0.525 loss 7.148
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [62]  [  0/156]  eta: 0:12:43  lr: 0.007564  min_lr: 0.007564  loss: 2.9844 (2.9844)  weight_decay: 0.0500 (0.0500)  time: 4.8956  data: 3.2870  max mem: 57114
Epoch: [62]  [ 10/156]  eta: 0:02:48  lr: 0.007563  min_lr: 0.007563  loss: 3.6327 (3.5558)  weight_decay: 0.0500 (0.0500)  time: 1.1544  data: 0.2992  max mem: 57114
Epoch: [62]  [ 20/156]  eta: 0:02:11  lr: 0.007561  min_lr: 0.007561  loss: 3.5925 (3.5325)  weight_decay: 0.0500 (0.0500)  time: 0.7678  data: 0.0004  max mem: 57114
Epoch: [62]  [ 30/156]  eta: 0:01:52  lr: 0.007560  min_lr: 0.007560  loss: 3.5738 (3.5556)  weight_decay: 0.0500 (0.0500)  time: 0.7517  data: 0.0004  max mem: 57114
Epoch: [62]  [ 40/156]  eta: 0:01:39  lr: 0.007559  min_lr: 0.007559  loss: 3.5944 (3.5825)  weight_decay: 0.0500 (0.0500)  time: 0.7411  data: 0.0003  max mem: 57114
Epoch: [62]  [ 50/156]  eta: 0:01:28  lr: 0.007558  min_lr: 0.007558  loss: 3.5674 (3.5248)  weight_decay: 0.0500 (0.0500)  time: 0.7419  data: 0.0003  max mem: 57114
Epoch: [62]  [ 60/156]  eta: 0:01:18  lr: 0.007556  min_lr: 0.007556  loss: 3.5975 (3.5515)  weight_decay: 0.0500 (0.0500)  time: 0.7370  data: 0.0004  max mem: 57114
Epoch: [62]  [ 70/156]  eta: 0:01:09  lr: 0.007555  min_lr: 0.007555  loss: 3.7241 (3.5397)  weight_decay: 0.0500 (0.0500)  time: 0.7202  data: 0.0004  max mem: 57114
Epoch: [62]  [ 80/156]  eta: 0:00:59  lr: 0.007554  min_lr: 0.007554  loss: 3.4491 (3.5215)  weight_decay: 0.0500 (0.0500)  time: 0.7065  data: 0.0004  max mem: 57114
Epoch: [62]  [ 90/156]  eta: 0:00:51  lr: 0.007552  min_lr: 0.007552  loss: 3.5814 (3.5194)  weight_decay: 0.0500 (0.0500)  time: 0.7030  data: 0.0004  max mem: 57114
Epoch: [62]  [100/156]  eta: 0:00:43  lr: 0.007551  min_lr: 0.007551  loss: 3.3620 (3.4887)  weight_decay: 0.0500 (0.0500)  time: 0.7039  data: 0.0004  max mem: 57114
Epoch: [62]  [110/156]  eta: 0:00:35  lr: 0.007550  min_lr: 0.007550  loss: 3.3053 (3.4779)  weight_decay: 0.0500 (0.0500)  time: 0.6998  data: 0.0003  max mem: 57114
Epoch: [62]  [120/156]  eta: 0:00:27  lr: 0.007548  min_lr: 0.007548  loss: 3.5120 (3.4829)  weight_decay: 0.0500 (0.0500)  time: 0.7015  data: 0.0004  max mem: 57114
Epoch: [62]  [130/156]  eta: 0:00:19  lr: 0.007547  min_lr: 0.007547  loss: 3.5767 (3.4784)  weight_decay: 0.0500 (0.0500)  time: 0.7002  data: 0.0008  max mem: 57114
Epoch: [62]  [140/156]  eta: 0:00:12  lr: 0.007546  min_lr: 0.007546  loss: 3.5426 (3.4684)  weight_decay: 0.0500 (0.0500)  time: 0.6905  data: 0.0007  max mem: 57114
Epoch: [62]  [150/156]  eta: 0:00:04  lr: 0.007544  min_lr: 0.007544  loss: 3.6128 (3.4696)  weight_decay: 0.0500 (0.0500)  time: 0.6834  data: 0.0001  max mem: 57114
Epoch: [62]  [155/156]  eta: 0:00:00  lr: 0.007544  min_lr: 0.007544  loss: 3.6076 (3.4761)  weight_decay: 0.0500 (0.0500)  time: 0.6819  data: 0.0001  max mem: 57114
Epoch: [62] Total time: 0:01:56 (0.7459 s / it)
Averaged stats: lr: 0.007544  min_lr: 0.007544  loss: 3.6076 (3.4614)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.1269 (1.1269)  acc1: 79.1667 (79.1667)  acc5: 97.3958 (97.3958)  time: 2.1350  data: 1.8783  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2209 (1.2244)  acc1: 78.1250 (74.9361)  acc5: 96.3542 (93.0946)  time: 0.5777  data: 0.3757  max mem: 57114
Test: Total time: 0:00:03 (0.6022 s / it)
* Acc@1 73.260 Acc@5 91.912 loss 1.382
Accuracy of the model on the 50000 test images: 73.3%
Max accuracy: 73.80%
Test:  [0/5]  eta: 0:00:11  loss: 7.3849 (7.3849)  acc1: 0.5208 (0.5208)  acc5: 0.5208 (0.5208)  time: 2.3378  data: 2.0943  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.1383 (7.1730)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.6182  data: 0.4190  max mem: 57114
Test: Total time: 0:00:03 (0.6319 s / it)
* Acc@1 0.100 Acc@5 0.533 loss 7.162
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [63]  [  0/156]  eta: 0:13:31  lr: 0.007543  min_lr: 0.007543  loss: 2.5639 (2.5639)  weight_decay: 0.0500 (0.0500)  time: 5.2018  data: 4.1853  max mem: 57114
Epoch: [63]  [ 10/156]  eta: 0:02:48  lr: 0.007542  min_lr: 0.007542  loss: 3.2097 (3.3851)  weight_decay: 0.0500 (0.0500)  time: 1.1508  data: 0.3807  max mem: 57114
Epoch: [63]  [ 20/156]  eta: 0:02:09  lr: 0.007541  min_lr: 0.007541  loss: 3.4171 (3.3697)  weight_decay: 0.0500 (0.0500)  time: 0.7378  data: 0.0003  max mem: 57114
Epoch: [63]  [ 30/156]  eta: 0:01:50  lr: 0.007539  min_lr: 0.007539  loss: 3.3896 (3.3123)  weight_decay: 0.0500 (0.0500)  time: 0.7295  data: 0.0003  max mem: 57114
Epoch: [63]  [ 40/156]  eta: 0:01:37  lr: 0.007538  min_lr: 0.007538  loss: 3.3896 (3.3321)  weight_decay: 0.0500 (0.0500)  time: 0.7288  data: 0.0004  max mem: 57114
Epoch: [63]  [ 50/156]  eta: 0:01:26  lr: 0.007537  min_lr: 0.007537  loss: 3.4281 (3.3427)  weight_decay: 0.0500 (0.0500)  time: 0.7254  data: 0.0004  max mem: 57114
Epoch: [63]  [ 60/156]  eta: 0:01:16  lr: 0.007535  min_lr: 0.007535  loss: 3.4281 (3.3406)  weight_decay: 0.0500 (0.0500)  time: 0.7184  data: 0.0004  max mem: 57114
Epoch: [63]  [ 70/156]  eta: 0:01:07  lr: 0.007534  min_lr: 0.007534  loss: 3.4519 (3.3423)  weight_decay: 0.0500 (0.0500)  time: 0.7074  data: 0.0004  max mem: 57114
Epoch: [63]  [ 80/156]  eta: 0:00:59  lr: 0.007533  min_lr: 0.007533  loss: 3.4113 (3.3529)  weight_decay: 0.0500 (0.0500)  time: 0.7106  data: 0.0004  max mem: 57114
Epoch: [63]  [ 90/156]  eta: 0:00:50  lr: 0.007531  min_lr: 0.007531  loss: 3.5383 (3.3651)  weight_decay: 0.0500 (0.0500)  time: 0.7130  data: 0.0004  max mem: 57114
Epoch: [63]  [100/156]  eta: 0:00:42  lr: 0.007530  min_lr: 0.007530  loss: 3.5383 (3.3597)  weight_decay: 0.0500 (0.0500)  time: 0.7086  data: 0.0003  max mem: 57114
Epoch: [63]  [110/156]  eta: 0:00:34  lr: 0.007529  min_lr: 0.007529  loss: 3.4910 (3.3698)  weight_decay: 0.0500 (0.0500)  time: 0.7123  data: 0.0004  max mem: 57114
Epoch: [63]  [120/156]  eta: 0:00:27  lr: 0.007527  min_lr: 0.007527  loss: 3.5920 (3.3941)  weight_decay: 0.0500 (0.0500)  time: 0.7133  data: 0.0004  max mem: 57114
Epoch: [63]  [130/156]  eta: 0:00:19  lr: 0.007526  min_lr: 0.007526  loss: 3.4403 (3.3786)  weight_decay: 0.0500 (0.0500)  time: 0.6980  data: 0.0010  max mem: 57114
Epoch: [63]  [140/156]  eta: 0:00:11  lr: 0.007525  min_lr: 0.007525  loss: 3.2826 (3.3875)  weight_decay: 0.0500 (0.0500)  time: 0.6812  data: 0.0008  max mem: 57114
Epoch: [63]  [150/156]  eta: 0:00:04  lr: 0.007523  min_lr: 0.007523  loss: 3.4751 (3.3854)  weight_decay: 0.0500 (0.0500)  time: 0.6839  data: 0.0001  max mem: 57114
Epoch: [63]  [155/156]  eta: 0:00:00  lr: 0.007523  min_lr: 0.007523  loss: 3.6069 (3.3943)  weight_decay: 0.0500 (0.0500)  time: 0.6832  data: 0.0001  max mem: 57114
Epoch: [63] Total time: 0:01:55 (0.7421 s / it)
Averaged stats: lr: 0.007523  min_lr: 0.007523  loss: 3.6069 (3.4436)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 1.0044 (1.0044)  acc1: 78.6458 (78.6458)  acc5: 96.3542 (96.3542)  time: 2.2185  data: 1.9588  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1195 (1.1347)  acc1: 76.5625 (73.7852)  acc5: 93.7500 (92.4552)  time: 0.5944  data: 0.3918  max mem: 57114
Test: Total time: 0:00:03 (0.6147 s / it)
* Acc@1 73.961 Acc@5 92.347 loss 1.227
Accuracy of the model on the 50000 test images: 74.0%
Max accuracy: 73.96%
Test:  [0/5]  eta: 0:00:09  loss: 7.4068 (7.4068)  acc1: 0.5208 (0.5208)  acc5: 0.5208 (0.5208)  time: 1.9703  data: 1.7269  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.1530 (7.1904)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.5447  data: 0.3455  max mem: 57114
Test: Total time: 0:00:02 (0.5595 s / it)
* Acc@1 0.100 Acc@5 0.533 loss 7.179
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [64]  [  0/156]  eta: 0:12:51  lr: 0.007522  min_lr: 0.007522  loss: 3.2357 (3.2357)  weight_decay: 0.0500 (0.0500)  time: 4.9437  data: 3.5323  max mem: 57114
Epoch: [64]  [ 10/156]  eta: 0:02:45  lr: 0.007521  min_lr: 0.007521  loss: 3.2718 (3.3631)  weight_decay: 0.0500 (0.0500)  time: 1.1340  data: 0.3214  max mem: 57114
Epoch: [64]  [ 20/156]  eta: 0:02:09  lr: 0.007520  min_lr: 0.007520  loss: 3.4827 (3.4199)  weight_decay: 0.0500 (0.0500)  time: 0.7533  data: 0.0003  max mem: 57114
Epoch: [64]  [ 30/156]  eta: 0:01:51  lr: 0.007518  min_lr: 0.007518  loss: 3.6737 (3.5071)  weight_decay: 0.0500 (0.0500)  time: 0.7501  data: 0.0003  max mem: 57114
Epoch: [64]  [ 40/156]  eta: 0:01:37  lr: 0.007517  min_lr: 0.007517  loss: 3.7694 (3.5266)  weight_decay: 0.0500 (0.0500)  time: 0.7250  data: 0.0003  max mem: 57114
Epoch: [64]  [ 50/156]  eta: 0:01:26  lr: 0.007516  min_lr: 0.007516  loss: 3.7877 (3.5639)  weight_decay: 0.0500 (0.0500)  time: 0.7087  data: 0.0004  max mem: 57114
Epoch: [64]  [ 60/156]  eta: 0:01:17  lr: 0.007514  min_lr: 0.007514  loss: 3.8431 (3.5676)  weight_decay: 0.0500 (0.0500)  time: 0.7225  data: 0.0004  max mem: 57114
Epoch: [64]  [ 70/156]  eta: 0:01:08  lr: 0.007513  min_lr: 0.007513  loss: 3.5258 (3.5419)  weight_decay: 0.0500 (0.0500)  time: 0.7348  data: 0.0004  max mem: 57114
Epoch: [64]  [ 80/156]  eta: 0:00:59  lr: 0.007511  min_lr: 0.007511  loss: 3.5626 (3.5602)  weight_decay: 0.0500 (0.0500)  time: 0.7269  data: 0.0004  max mem: 57114
Epoch: [64]  [ 90/156]  eta: 0:00:51  lr: 0.007510  min_lr: 0.007510  loss: 3.6489 (3.5656)  weight_decay: 0.0500 (0.0500)  time: 0.7108  data: 0.0004  max mem: 57114
Epoch: [64]  [100/156]  eta: 0:00:43  lr: 0.007509  min_lr: 0.007509  loss: 3.6489 (3.5697)  weight_decay: 0.0500 (0.0500)  time: 0.7039  data: 0.0004  max mem: 57114
Epoch: [64]  [110/156]  eta: 0:00:35  lr: 0.007507  min_lr: 0.007507  loss: 3.5678 (3.5568)  weight_decay: 0.0500 (0.0500)  time: 0.7052  data: 0.0004  max mem: 57114
Epoch: [64]  [120/156]  eta: 0:00:27  lr: 0.007506  min_lr: 0.007506  loss: 3.5512 (3.5332)  weight_decay: 0.0500 (0.0500)  time: 0.7028  data: 0.0004  max mem: 57114
Epoch: [64]  [130/156]  eta: 0:00:19  lr: 0.007505  min_lr: 0.007505  loss: 3.5746 (3.5381)  weight_decay: 0.0500 (0.0500)  time: 0.6983  data: 0.0009  max mem: 57114
Epoch: [64]  [140/156]  eta: 0:00:11  lr: 0.007503  min_lr: 0.007503  loss: 3.6485 (3.5350)  weight_decay: 0.0500 (0.0500)  time: 0.6922  data: 0.0007  max mem: 57114
Epoch: [64]  [150/156]  eta: 0:00:04  lr: 0.007502  min_lr: 0.007502  loss: 3.4407 (3.5268)  weight_decay: 0.0500 (0.0500)  time: 0.6848  data: 0.0001  max mem: 57114
Epoch: [64]  [155/156]  eta: 0:00:00  lr: 0.007501  min_lr: 0.007501  loss: 3.4646 (3.5253)  weight_decay: 0.0500 (0.0500)  time: 0.6892  data: 0.0001  max mem: 57114
Epoch: [64] Total time: 0:01:56 (0.7444 s / it)
Averaged stats: lr: 0.007501  min_lr: 0.007501  loss: 3.4646 (3.4449)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.1217 (1.1217)  acc1: 81.7708 (81.7708)  acc5: 96.3542 (96.3542)  time: 2.0772  data: 1.8217  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2825 (1.2477)  acc1: 75.5208 (73.5294)  acc5: 93.7500 (91.4322)  time: 0.5661  data: 0.3644  max mem: 57114
Test: Total time: 0:00:02 (0.5902 s / it)
* Acc@1 72.500 Acc@5 91.616 loss 1.351
Accuracy of the model on the 50000 test images: 72.5%
Max accuracy: 73.96%
Test:  [0/5]  eta: 0:00:11  loss: 7.4306 (7.4306)  acc1: 0.5208 (0.5208)  acc5: 0.5208 (0.5208)  time: 2.3603  data: 2.1167  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.1650 (7.2067)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.6227  data: 0.4234  max mem: 57114
Test: Total time: 0:00:03 (0.6381 s / it)
* Acc@1 0.100 Acc@5 0.531 loss 7.195
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [65]  [  0/156]  eta: 0:12:50  lr: 0.007501  min_lr: 0.007501  loss: 3.7206 (3.7206)  weight_decay: 0.0500 (0.0500)  time: 4.9401  data: 3.0192  max mem: 57114
Epoch: [65]  [ 10/156]  eta: 0:02:42  lr: 0.007500  min_lr: 0.007500  loss: 3.2739 (3.3084)  weight_decay: 0.0500 (0.0500)  time: 1.1140  data: 0.2748  max mem: 57114
Epoch: [65]  [ 20/156]  eta: 0:02:05  lr: 0.007498  min_lr: 0.007498  loss: 3.1618 (3.2829)  weight_decay: 0.0500 (0.0500)  time: 0.7233  data: 0.0004  max mem: 57114
Epoch: [65]  [ 30/156]  eta: 0:01:48  lr: 0.007497  min_lr: 0.007497  loss: 3.5126 (3.3928)  weight_decay: 0.0500 (0.0500)  time: 0.7203  data: 0.0004  max mem: 57114
Epoch: [65]  [ 40/156]  eta: 0:01:35  lr: 0.007495  min_lr: 0.007495  loss: 3.6479 (3.4477)  weight_decay: 0.0500 (0.0500)  time: 0.7154  data: 0.0003  max mem: 57114
Epoch: [65]  [ 50/156]  eta: 0:01:24  lr: 0.007494  min_lr: 0.007494  loss: 3.5806 (3.4075)  weight_decay: 0.0500 (0.0500)  time: 0.7100  data: 0.0004  max mem: 57114
Epoch: [65]  [ 60/156]  eta: 0:01:15  lr: 0.007493  min_lr: 0.007493  loss: 3.5806 (3.4481)  weight_decay: 0.0500 (0.0500)  time: 0.7173  data: 0.0004  max mem: 57114
Epoch: [65]  [ 70/156]  eta: 0:01:07  lr: 0.007491  min_lr: 0.007491  loss: 3.5620 (3.4558)  weight_decay: 0.0500 (0.0500)  time: 0.7289  data: 0.0004  max mem: 57114
Epoch: [65]  [ 80/156]  eta: 0:00:58  lr: 0.007490  min_lr: 0.007490  loss: 3.3744 (3.4199)  weight_decay: 0.0500 (0.0500)  time: 0.7275  data: 0.0004  max mem: 57114
Epoch: [65]  [ 90/156]  eta: 0:00:50  lr: 0.007488  min_lr: 0.007488  loss: 3.2310 (3.4115)  weight_decay: 0.0500 (0.0500)  time: 0.7110  data: 0.0004  max mem: 57114
Epoch: [65]  [100/156]  eta: 0:00:42  lr: 0.007487  min_lr: 0.007487  loss: 3.4211 (3.4027)  weight_decay: 0.0500 (0.0500)  time: 0.7003  data: 0.0004  max mem: 57114
Epoch: [65]  [110/156]  eta: 0:00:34  lr: 0.007486  min_lr: 0.007486  loss: 3.8165 (3.4504)  weight_decay: 0.0500 (0.0500)  time: 0.6977  data: 0.0004  max mem: 57114
Epoch: [65]  [120/156]  eta: 0:00:26  lr: 0.007484  min_lr: 0.007484  loss: 3.8554 (3.4741)  weight_decay: 0.0500 (0.0500)  time: 0.7034  data: 0.0004  max mem: 57114
Epoch: [65]  [130/156]  eta: 0:00:19  lr: 0.007483  min_lr: 0.007483  loss: 3.7752 (3.4802)  weight_decay: 0.0500 (0.0500)  time: 0.6997  data: 0.0005  max mem: 57114
Epoch: [65]  [140/156]  eta: 0:00:11  lr: 0.007481  min_lr: 0.007481  loss: 3.4284 (3.4708)  weight_decay: 0.0500 (0.0500)  time: 0.6858  data: 0.0003  max mem: 57114
Epoch: [65]  [150/156]  eta: 0:00:04  lr: 0.007480  min_lr: 0.007480  loss: 3.4686 (3.4747)  weight_decay: 0.0500 (0.0500)  time: 0.6825  data: 0.0001  max mem: 57114
Epoch: [65]  [155/156]  eta: 0:00:00  lr: 0.007479  min_lr: 0.007479  loss: 3.5462 (3.4722)  weight_decay: 0.0500 (0.0500)  time: 0.6832  data: 0.0001  max mem: 57114
Epoch: [65] Total time: 0:01:55 (0.7373 s / it)
Averaged stats: lr: 0.007479  min_lr: 0.007479  loss: 3.5462 (3.4379)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 1.0170 (1.0170)  acc1: 82.8125 (82.8125)  acc5: 96.8750 (96.8750)  time: 2.2462  data: 1.9908  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1371 (1.1430)  acc1: 78.1250 (75.8312)  acc5: 95.3125 (93.2225)  time: 0.5999  data: 0.3982  max mem: 57114
Test: Total time: 0:00:03 (0.6222 s / it)
* Acc@1 73.849 Acc@5 92.255 loss 1.227
Accuracy of the model on the 50000 test images: 73.8%
Max accuracy: 73.96%
Test:  [0/5]  eta: 0:00:11  loss: 7.4525 (7.4525)  acc1: 0.5208 (0.5208)  acc5: 0.5208 (0.5208)  time: 2.3436  data: 2.0996  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.1782 (7.2235)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.6196  data: 0.4200  max mem: 57114
Test: Total time: 0:00:03 (0.6347 s / it)
* Acc@1 0.100 Acc@5 0.529 loss 7.211
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [66]  [  0/156]  eta: 0:12:32  lr: 0.007479  min_lr: 0.007479  loss: 3.4151 (3.4151)  weight_decay: 0.0500 (0.0500)  time: 4.8223  data: 3.9308  max mem: 57114
Epoch: [66]  [ 10/156]  eta: 0:02:40  lr: 0.007478  min_lr: 0.007478  loss: 3.7948 (3.7117)  weight_decay: 0.0500 (0.0500)  time: 1.1021  data: 0.3577  max mem: 57114
Epoch: [66]  [ 20/156]  eta: 0:02:05  lr: 0.007476  min_lr: 0.007476  loss: 3.5231 (3.4225)  weight_decay: 0.0500 (0.0500)  time: 0.7265  data: 0.0004  max mem: 57114
Epoch: [66]  [ 30/156]  eta: 0:01:48  lr: 0.007475  min_lr: 0.007475  loss: 3.2047 (3.4074)  weight_decay: 0.0500 (0.0500)  time: 0.7296  data: 0.0003  max mem: 57114
Epoch: [66]  [ 40/156]  eta: 0:01:36  lr: 0.007473  min_lr: 0.007473  loss: 3.5045 (3.4406)  weight_decay: 0.0500 (0.0500)  time: 0.7329  data: 0.0004  max mem: 57114
Epoch: [66]  [ 50/156]  eta: 0:01:25  lr: 0.007472  min_lr: 0.007472  loss: 3.5550 (3.4860)  weight_decay: 0.0500 (0.0500)  time: 0.7168  data: 0.0004  max mem: 57114
Epoch: [66]  [ 60/156]  eta: 0:01:16  lr: 0.007470  min_lr: 0.007470  loss: 3.6775 (3.4922)  weight_decay: 0.0500 (0.0500)  time: 0.7184  data: 0.0004  max mem: 57114
Epoch: [66]  [ 70/156]  eta: 0:01:07  lr: 0.007469  min_lr: 0.007469  loss: 3.5866 (3.5010)  weight_decay: 0.0500 (0.0500)  time: 0.7209  data: 0.0004  max mem: 57114
Epoch: [66]  [ 80/156]  eta: 0:00:58  lr: 0.007468  min_lr: 0.007468  loss: 3.5174 (3.4624)  weight_decay: 0.0500 (0.0500)  time: 0.7056  data: 0.0005  max mem: 57114
Epoch: [66]  [ 90/156]  eta: 0:00:50  lr: 0.007466  min_lr: 0.007466  loss: 2.9512 (3.4377)  weight_decay: 0.0500 (0.0500)  time: 0.6992  data: 0.0004  max mem: 57114
Epoch: [66]  [100/156]  eta: 0:00:42  lr: 0.007465  min_lr: 0.007465  loss: 3.5550 (3.4481)  weight_decay: 0.0500 (0.0500)  time: 0.7117  data: 0.0003  max mem: 57114
Epoch: [66]  [110/156]  eta: 0:00:34  lr: 0.007463  min_lr: 0.007463  loss: 3.5550 (3.4207)  weight_decay: 0.0500 (0.0500)  time: 0.7139  data: 0.0003  max mem: 57114
Epoch: [66]  [120/156]  eta: 0:00:27  lr: 0.007462  min_lr: 0.007462  loss: 3.2264 (3.4356)  weight_decay: 0.0500 (0.0500)  time: 0.7037  data: 0.0004  max mem: 57114
Epoch: [66]  [130/156]  eta: 0:00:19  lr: 0.007460  min_lr: 0.007460  loss: 3.3241 (3.4312)  weight_decay: 0.0500 (0.0500)  time: 0.7025  data: 0.0009  max mem: 57114
Epoch: [66]  [140/156]  eta: 0:00:11  lr: 0.007459  min_lr: 0.007459  loss: 3.4067 (3.4273)  weight_decay: 0.0500 (0.0500)  time: 0.6867  data: 0.0007  max mem: 57114
Epoch: [66]  [150/156]  eta: 0:00:04  lr: 0.007458  min_lr: 0.007458  loss: 3.4456 (3.4241)  weight_decay: 0.0500 (0.0500)  time: 0.6774  data: 0.0001  max mem: 57114
Epoch: [66]  [155/156]  eta: 0:00:00  lr: 0.007457  min_lr: 0.007457  loss: 3.4456 (3.4275)  weight_decay: 0.0500 (0.0500)  time: 0.6787  data: 0.0001  max mem: 57114
Epoch: [66] Total time: 0:01:55 (0.7379 s / it)
Averaged stats: lr: 0.007457  min_lr: 0.007457  loss: 3.4456 (3.4368)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 1.1311 (1.1311)  acc1: 79.1667 (79.1667)  acc5: 95.3125 (95.3125)  time: 2.2302  data: 1.9736  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2768 (1.2977)  acc1: 76.0417 (72.8900)  acc5: 94.2708 (91.8159)  time: 0.5967  data: 0.3948  max mem: 57114
Test: Total time: 0:00:03 (0.6214 s / it)
* Acc@1 73.430 Acc@5 91.946 loss 1.375
Accuracy of the model on the 50000 test images: 73.4%
Max accuracy: 73.96%
Test:  [0/5]  eta: 0:00:11  loss: 7.4778 (7.4778)  acc1: 0.5208 (0.5208)  acc5: 0.5208 (0.5208)  time: 2.2781  data: 2.0346  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.1930 (7.2420)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.6063  data: 0.4070  max mem: 57114
Test: Total time: 0:00:03 (0.6201 s / it)
* Acc@1 0.104 Acc@5 0.525 loss 7.229
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [67]  [  0/156]  eta: 0:12:28  lr: 0.007457  min_lr: 0.007457  loss: 3.1947 (3.1947)  weight_decay: 0.0500 (0.0500)  time: 4.7958  data: 3.7243  max mem: 57114
Epoch: [67]  [ 10/156]  eta: 0:02:41  lr: 0.007455  min_lr: 0.007455  loss: 3.1947 (3.1808)  weight_decay: 0.0500 (0.0500)  time: 1.1040  data: 0.3389  max mem: 57114
Epoch: [67]  [ 20/156]  eta: 0:02:05  lr: 0.007454  min_lr: 0.007454  loss: 3.5212 (3.3867)  weight_decay: 0.0500 (0.0500)  time: 0.7268  data: 0.0003  max mem: 57114
Epoch: [67]  [ 30/156]  eta: 0:01:47  lr: 0.007452  min_lr: 0.007452  loss: 3.6829 (3.4177)  weight_decay: 0.0500 (0.0500)  time: 0.7135  data: 0.0003  max mem: 57114
Epoch: [67]  [ 40/156]  eta: 0:01:36  lr: 0.007451  min_lr: 0.007451  loss: 3.5996 (3.4615)  weight_decay: 0.0500 (0.0500)  time: 0.7299  data: 0.0005  max mem: 57114
Epoch: [67]  [ 50/156]  eta: 0:01:25  lr: 0.007449  min_lr: 0.007449  loss: 3.6586 (3.4483)  weight_decay: 0.0500 (0.0500)  time: 0.7413  data: 0.0005  max mem: 57114
Epoch: [67]  [ 60/156]  eta: 0:01:16  lr: 0.007448  min_lr: 0.007448  loss: 3.5703 (3.4739)  weight_decay: 0.0500 (0.0500)  time: 0.7190  data: 0.0004  max mem: 57114
Epoch: [67]  [ 70/156]  eta: 0:01:07  lr: 0.007446  min_lr: 0.007446  loss: 3.4212 (3.4457)  weight_decay: 0.0500 (0.0500)  time: 0.7065  data: 0.0003  max mem: 57114
Epoch: [67]  [ 80/156]  eta: 0:00:58  lr: 0.007445  min_lr: 0.007445  loss: 3.3742 (3.4520)  weight_decay: 0.0500 (0.0500)  time: 0.7112  data: 0.0004  max mem: 57114
Epoch: [67]  [ 90/156]  eta: 0:00:50  lr: 0.007444  min_lr: 0.007444  loss: 3.3742 (3.4418)  weight_decay: 0.0500 (0.0500)  time: 0.7056  data: 0.0004  max mem: 57114
Epoch: [67]  [100/156]  eta: 0:00:42  lr: 0.007442  min_lr: 0.007442  loss: 3.4478 (3.4452)  weight_decay: 0.0500 (0.0500)  time: 0.7045  data: 0.0004  max mem: 57114
Epoch: [67]  [110/156]  eta: 0:00:34  lr: 0.007441  min_lr: 0.007441  loss: 3.1942 (3.4215)  weight_decay: 0.0500 (0.0500)  time: 0.7133  data: 0.0004  max mem: 57114
Epoch: [67]  [120/156]  eta: 0:00:27  lr: 0.007439  min_lr: 0.007439  loss: 3.0578 (3.4013)  weight_decay: 0.0500 (0.0500)  time: 0.7085  data: 0.0004  max mem: 57114
Epoch: [67]  [130/156]  eta: 0:00:19  lr: 0.007438  min_lr: 0.007438  loss: 3.4314 (3.4125)  weight_decay: 0.0500 (0.0500)  time: 0.7022  data: 0.0008  max mem: 57114
Epoch: [67]  [140/156]  eta: 0:00:11  lr: 0.007436  min_lr: 0.007436  loss: 3.4523 (3.4023)  weight_decay: 0.0500 (0.0500)  time: 0.6898  data: 0.0007  max mem: 57114
Epoch: [67]  [150/156]  eta: 0:00:04  lr: 0.007435  min_lr: 0.007435  loss: 3.2824 (3.3920)  weight_decay: 0.0500 (0.0500)  time: 0.6801  data: 0.0001  max mem: 57114
Epoch: [67]  [155/156]  eta: 0:00:00  lr: 0.007434  min_lr: 0.007434  loss: 3.3300 (3.3932)  weight_decay: 0.0500 (0.0500)  time: 0.6820  data: 0.0001  max mem: 57114
Epoch: [67] Total time: 0:01:55 (0.7382 s / it)
Averaged stats: lr: 0.007434  min_lr: 0.007434  loss: 3.3300 (3.4355)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 1.0726 (1.0726)  acc1: 81.7708 (81.7708)  acc5: 94.7917 (94.7917)  time: 2.3634  data: 2.1067  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2586 (1.1945)  acc1: 76.0417 (73.4015)  acc5: 94.7917 (92.7110)  time: 0.6234  data: 0.4214  max mem: 57114
Test: Total time: 0:00:03 (0.6435 s / it)
* Acc@1 73.735 Acc@5 92.191 loss 1.279
Accuracy of the model on the 50000 test images: 73.7%
Max accuracy: 73.96%
Test:  [0/5]  eta: 0:00:11  loss: 7.5027 (7.5027)  acc1: 0.5208 (0.5208)  acc5: 0.5208 (0.5208)  time: 2.3188  data: 2.0754  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.2131 (7.2617)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.6144  data: 0.4152  max mem: 57114
Test: Total time: 0:00:03 (0.6279 s / it)
* Acc@1 0.114 Acc@5 0.521 loss 7.247
Accuracy of the model EMA on 50000 test images: 0.1%
Max EMA accuracy: 0.11%
Epoch: [68]  [  0/156]  eta: 0:08:13  lr: 0.007434  min_lr: 0.007434  loss: 3.8776 (3.8776)  weight_decay: 0.0500 (0.0500)  time: 3.1639  data: 2.3849  max mem: 57114
Epoch: [68]  [ 10/156]  eta: 0:02:17  lr: 0.007432  min_lr: 0.007432  loss: 3.4229 (3.3311)  weight_decay: 0.0500 (0.0500)  time: 0.9447  data: 0.2171  max mem: 57114
Epoch: [68]  [ 20/156]  eta: 0:01:53  lr: 0.007431  min_lr: 0.007431  loss: 3.3011 (3.3362)  weight_decay: 0.0500 (0.0500)  time: 0.7196  data: 0.0003  max mem: 57114
Epoch: [68]  [ 30/156]  eta: 0:01:40  lr: 0.007429  min_lr: 0.007429  loss: 3.6624 (3.4606)  weight_decay: 0.0500 (0.0500)  time: 0.7223  data: 0.0003  max mem: 57114
Epoch: [68]  [ 40/156]  eta: 0:01:30  lr: 0.007428  min_lr: 0.007428  loss: 3.6806 (3.4232)  weight_decay: 0.0500 (0.0500)  time: 0.7238  data: 0.0004  max mem: 57114
Epoch: [68]  [ 50/156]  eta: 0:01:21  lr: 0.007426  min_lr: 0.007426  loss: 3.2184 (3.3939)  weight_decay: 0.0500 (0.0500)  time: 0.7180  data: 0.0004  max mem: 57114
Epoch: [68]  [ 60/156]  eta: 0:01:12  lr: 0.007425  min_lr: 0.007425  loss: 3.2522 (3.4027)  weight_decay: 0.0500 (0.0500)  time: 0.7113  data: 0.0004  max mem: 57114
Epoch: [68]  [ 70/156]  eta: 0:01:04  lr: 0.007423  min_lr: 0.007423  loss: 3.5909 (3.4034)  weight_decay: 0.0500 (0.0500)  time: 0.7166  data: 0.0004  max mem: 57114
Epoch: [68]  [ 80/156]  eta: 0:00:56  lr: 0.007422  min_lr: 0.007422  loss: 3.6773 (3.4438)  weight_decay: 0.0500 (0.0500)  time: 0.7101  data: 0.0005  max mem: 57114
Epoch: [68]  [ 90/156]  eta: 0:00:49  lr: 0.007421  min_lr: 0.007421  loss: 3.6812 (3.4470)  weight_decay: 0.0500 (0.0500)  time: 0.7058  data: 0.0004  max mem: 57114
Epoch: [68]  [100/156]  eta: 0:00:41  lr: 0.007419  min_lr: 0.007419  loss: 3.4098 (3.4509)  weight_decay: 0.0500 (0.0500)  time: 0.7149  data: 0.0004  max mem: 57114
Epoch: [68]  [110/156]  eta: 0:00:33  lr: 0.007418  min_lr: 0.007418  loss: 3.4707 (3.4447)  weight_decay: 0.0500 (0.0500)  time: 0.7017  data: 0.0004  max mem: 57114
Epoch: [68]  [120/156]  eta: 0:00:26  lr: 0.007416  min_lr: 0.007416  loss: 3.4748 (3.4418)  weight_decay: 0.0500 (0.0500)  time: 0.6967  data: 0.0004  max mem: 57114
Epoch: [68]  [130/156]  eta: 0:00:18  lr: 0.007415  min_lr: 0.007415  loss: 3.7084 (3.4567)  weight_decay: 0.0500 (0.0500)  time: 0.6992  data: 0.0009  max mem: 57114
Epoch: [68]  [140/156]  eta: 0:00:11  lr: 0.007413  min_lr: 0.007413  loss: 3.6410 (3.4501)  weight_decay: 0.0500 (0.0500)  time: 0.6882  data: 0.0007  max mem: 57114
Epoch: [68]  [150/156]  eta: 0:00:04  lr: 0.007412  min_lr: 0.007412  loss: 3.4945 (3.4524)  weight_decay: 0.0500 (0.0500)  time: 0.6849  data: 0.0001  max mem: 57114
Epoch: [68]  [155/156]  eta: 0:00:00  lr: 0.007411  min_lr: 0.007411  loss: 3.4945 (3.4443)  weight_decay: 0.0500 (0.0500)  time: 0.6878  data: 0.0001  max mem: 57114
Epoch: [68] Total time: 0:01:53 (0.7253 s / it)
Averaged stats: lr: 0.007411  min_lr: 0.007411  loss: 3.4945 (3.4341)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 0.9635 (0.9635)  acc1: 79.1667 (79.1667)  acc5: 95.3125 (95.3125)  time: 2.2952  data: 2.0334  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1549 (1.0925)  acc1: 77.0833 (75.3197)  acc5: 94.7917 (92.4552)  time: 0.6097  data: 0.4067  max mem: 57114
Test: Total time: 0:00:03 (0.6554 s / it)
* Acc@1 73.258 Acc@5 91.544 loss 1.239
Accuracy of the model on the 50000 test images: 73.3%
Max accuracy: 73.96%
Test:  [0/5]  eta: 0:00:12  loss: 7.5248 (7.5248)  acc1: 0.5208 (0.5208)  acc5: 0.5208 (0.5208)  time: 2.5785  data: 2.3350  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.2360 (7.2831)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.6663  data: 0.4671  max mem: 57114
Test: Total time: 0:00:03 (0.6777 s / it)
* Acc@1 0.126 Acc@5 0.521 loss 7.267
Accuracy of the model EMA on 50000 test images: 0.1%
Max EMA accuracy: 0.13%
Epoch: [69]  [  0/156]  eta: 0:07:04  lr: 0.007411  min_lr: 0.007411  loss: 3.3803 (3.3803)  weight_decay: 0.0500 (0.0500)  time: 2.7182  data: 2.0675  max mem: 57114
Epoch: [69]  [ 10/156]  eta: 0:02:12  lr: 0.007409  min_lr: 0.007409  loss: 3.3803 (3.4505)  weight_decay: 0.0500 (0.0500)  time: 0.9082  data: 0.1882  max mem: 57114
Epoch: [69]  [ 20/156]  eta: 0:01:51  lr: 0.007408  min_lr: 0.007408  loss: 3.3606 (3.3675)  weight_decay: 0.0500 (0.0500)  time: 0.7245  data: 0.0003  max mem: 57114
Epoch: [69]  [ 30/156]  eta: 0:01:39  lr: 0.007406  min_lr: 0.007406  loss: 3.4896 (3.3995)  weight_decay: 0.0500 (0.0500)  time: 0.7243  data: 0.0003  max mem: 57114
Epoch: [69]  [ 40/156]  eta: 0:01:29  lr: 0.007405  min_lr: 0.007405  loss: 3.4455 (3.3609)  weight_decay: 0.0500 (0.0500)  time: 0.7157  data: 0.0004  max mem: 57114
Epoch: [69]  [ 50/156]  eta: 0:01:20  lr: 0.007403  min_lr: 0.007403  loss: 3.4155 (3.3698)  weight_decay: 0.0500 (0.0500)  time: 0.7156  data: 0.0004  max mem: 57114
Epoch: [69]  [ 60/156]  eta: 0:01:12  lr: 0.007402  min_lr: 0.007402  loss: 3.3953 (3.3344)  weight_decay: 0.0500 (0.0500)  time: 0.7288  data: 0.0004  max mem: 57114
Epoch: [69]  [ 70/156]  eta: 0:01:04  lr: 0.007400  min_lr: 0.007400  loss: 3.3953 (3.3396)  weight_decay: 0.0500 (0.0500)  time: 0.7261  data: 0.0004  max mem: 57114
Epoch: [69]  [ 80/156]  eta: 0:00:56  lr: 0.007399  min_lr: 0.007399  loss: 3.6050 (3.3702)  weight_decay: 0.0500 (0.0500)  time: 0.7272  data: 0.0004  max mem: 57114
Epoch: [69]  [ 90/156]  eta: 0:00:49  lr: 0.007397  min_lr: 0.007397  loss: 3.7078 (3.3885)  weight_decay: 0.0500 (0.0500)  time: 0.7169  data: 0.0004  max mem: 57114
Epoch: [69]  [100/156]  eta: 0:00:41  lr: 0.007396  min_lr: 0.007396  loss: 3.5926 (3.4006)  weight_decay: 0.0500 (0.0500)  time: 0.6983  data: 0.0004  max mem: 57114
Epoch: [69]  [110/156]  eta: 0:00:33  lr: 0.007394  min_lr: 0.007394  loss: 3.6156 (3.4156)  weight_decay: 0.0500 (0.0500)  time: 0.6969  data: 0.0003  max mem: 57114
Epoch: [69]  [120/156]  eta: 0:00:26  lr: 0.007392  min_lr: 0.007392  loss: 3.6532 (3.4217)  weight_decay: 0.0500 (0.0500)  time: 0.6963  data: 0.0004  max mem: 57114
Epoch: [69]  [130/156]  eta: 0:00:18  lr: 0.007391  min_lr: 0.007391  loss: 3.5699 (3.4175)  weight_decay: 0.0500 (0.0500)  time: 0.6988  data: 0.0009  max mem: 57114
Epoch: [69]  [140/156]  eta: 0:00:11  lr: 0.007389  min_lr: 0.007389  loss: 3.4763 (3.4113)  weight_decay: 0.0500 (0.0500)  time: 0.6917  data: 0.0007  max mem: 57114
Epoch: [69]  [150/156]  eta: 0:00:04  lr: 0.007388  min_lr: 0.007388  loss: 3.5009 (3.4107)  weight_decay: 0.0500 (0.0500)  time: 0.6866  data: 0.0001  max mem: 57114
Epoch: [69]  [155/156]  eta: 0:00:00  lr: 0.007387  min_lr: 0.007387  loss: 3.5009 (3.3952)  weight_decay: 0.0500 (0.0500)  time: 0.6878  data: 0.0001  max mem: 57114
Epoch: [69] Total time: 0:01:53 (0.7247 s / it)
Averaged stats: lr: 0.007387  min_lr: 0.007387  loss: 3.5009 (3.4346)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8554 (0.8554)  acc1: 79.6875 (79.6875)  acc5: 95.8333 (95.8333)  time: 2.1198  data: 1.8634  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9411 (1.0148)  acc1: 76.5625 (73.7852)  acc5: 95.3125 (92.7110)  time: 0.5746  data: 0.3728  max mem: 57114
Test: Total time: 0:00:02 (0.5985 s / it)
* Acc@1 74.423 Acc@5 92.475 loss 1.059
Accuracy of the model on the 50000 test images: 74.4%
Max accuracy: 74.42%
Test:  [0/5]  eta: 0:00:10  loss: 7.5487 (7.5487)  acc1: 0.5208 (0.5208)  acc5: 0.5208 (0.5208)  time: 2.1057  data: 1.8623  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.2587 (7.3053)  acc1: 0.0000 (0.2558)  acc5: 0.0000 (0.3836)  time: 0.5718  data: 0.3726  max mem: 57114
Test: Total time: 0:00:02 (0.5816 s / it)
* Acc@1 0.130 Acc@5 0.516 loss 7.288
Accuracy of the model EMA on 50000 test images: 0.1%
Max EMA accuracy: 0.13%
Epoch: [70]  [  0/156]  eta: 0:07:34  lr: 0.007387  min_lr: 0.007387  loss: 3.0696 (3.0696)  weight_decay: 0.0500 (0.0500)  time: 2.9105  data: 2.2611  max mem: 57114
Epoch: [70]  [ 10/156]  eta: 0:02:16  lr: 0.007385  min_lr: 0.007385  loss: 3.2893 (3.2758)  weight_decay: 0.0500 (0.0500)  time: 0.9343  data: 0.2059  max mem: 57114
Epoch: [70]  [ 20/156]  eta: 0:01:52  lr: 0.007384  min_lr: 0.007384  loss: 3.2893 (3.3014)  weight_decay: 0.0500 (0.0500)  time: 0.7260  data: 0.0004  max mem: 57114
Epoch: [70]  [ 30/156]  eta: 0:01:40  lr: 0.007382  min_lr: 0.007382  loss: 3.5321 (3.4280)  weight_decay: 0.0500 (0.0500)  time: 0.7281  data: 0.0004  max mem: 57114
Epoch: [70]  [ 40/156]  eta: 0:01:30  lr: 0.007381  min_lr: 0.007381  loss: 3.6007 (3.4228)  weight_decay: 0.0500 (0.0500)  time: 0.7354  data: 0.0004  max mem: 57114
Epoch: [70]  [ 50/156]  eta: 0:01:21  lr: 0.007379  min_lr: 0.007379  loss: 3.5148 (3.4388)  weight_decay: 0.0500 (0.0500)  time: 0.7269  data: 0.0003  max mem: 57114
Epoch: [70]  [ 60/156]  eta: 0:01:13  lr: 0.007378  min_lr: 0.007378  loss: 3.6861 (3.4696)  weight_decay: 0.0500 (0.0500)  time: 0.7437  data: 0.0004  max mem: 57114
Epoch: [70]  [ 70/156]  eta: 0:01:05  lr: 0.007376  min_lr: 0.007376  loss: 3.5586 (3.4301)  weight_decay: 0.0500 (0.0500)  time: 0.7343  data: 0.0004  max mem: 57114
Epoch: [70]  [ 80/156]  eta: 0:00:57  lr: 0.007375  min_lr: 0.007375  loss: 3.4280 (3.4354)  weight_decay: 0.0500 (0.0500)  time: 0.7163  data: 0.0004  max mem: 57114
Epoch: [70]  [ 90/156]  eta: 0:00:49  lr: 0.007373  min_lr: 0.007373  loss: 3.5417 (3.4551)  weight_decay: 0.0500 (0.0500)  time: 0.7136  data: 0.0004  max mem: 57114
Epoch: [70]  [100/156]  eta: 0:00:41  lr: 0.007372  min_lr: 0.007372  loss: 3.6126 (3.4725)  weight_decay: 0.0500 (0.0500)  time: 0.7066  data: 0.0004  max mem: 57114
Epoch: [70]  [110/156]  eta: 0:00:34  lr: 0.007370  min_lr: 0.007370  loss: 3.6173 (3.4810)  weight_decay: 0.0500 (0.0500)  time: 0.7084  data: 0.0004  max mem: 57114
Epoch: [70]  [120/156]  eta: 0:00:26  lr: 0.007368  min_lr: 0.007368  loss: 3.4245 (3.4690)  weight_decay: 0.0500 (0.0500)  time: 0.7027  data: 0.0003  max mem: 57114
Epoch: [70]  [130/156]  eta: 0:00:19  lr: 0.007367  min_lr: 0.007367  loss: 3.3337 (3.4682)  weight_decay: 0.0500 (0.0500)  time: 0.6994  data: 0.0008  max mem: 57114
Epoch: [70]  [140/156]  eta: 0:00:11  lr: 0.007365  min_lr: 0.007365  loss: 3.3634 (3.4615)  weight_decay: 0.0500 (0.0500)  time: 0.6898  data: 0.0007  max mem: 57114
Epoch: [70]  [150/156]  eta: 0:00:04  lr: 0.007364  min_lr: 0.007364  loss: 3.6305 (3.4732)  weight_decay: 0.0500 (0.0500)  time: 0.6829  data: 0.0001  max mem: 57114
Epoch: [70]  [155/156]  eta: 0:00:00  lr: 0.007363  min_lr: 0.007363  loss: 3.6302 (3.4648)  weight_decay: 0.0500 (0.0500)  time: 0.6843  data: 0.0001  max mem: 57114
Epoch: [70] Total time: 0:01:53 (0.7305 s / it)
Averaged stats: lr: 0.007363  min_lr: 0.007363  loss: 3.6302 (3.4150)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8982 (0.8982)  acc1: 81.7708 (81.7708)  acc5: 95.8333 (95.8333)  time: 2.0676  data: 1.8119  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1335 (1.0617)  acc1: 76.5625 (74.9361)  acc5: 94.7917 (92.9668)  time: 0.5641  data: 0.3624  max mem: 57114
Test: Total time: 0:00:02 (0.5885 s / it)
* Acc@1 74.405 Acc@5 92.485 loss 1.181
Accuracy of the model on the 50000 test images: 74.4%
Max accuracy: 74.42%
Test:  [0/5]  eta: 0:00:11  loss: 7.5719 (7.5719)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.5208)  time: 2.2880  data: 2.0439  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.2782 (7.3271)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.3836)  time: 0.6085  data: 0.4089  max mem: 57114
Test: Total time: 0:00:03 (0.6219 s / it)
* Acc@1 0.118 Acc@5 0.521 loss 7.309
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [71]  [  0/156]  eta: 0:12:41  lr: 0.007363  min_lr: 0.007363  loss: 3.7077 (3.7077)  weight_decay: 0.0500 (0.0500)  time: 4.8786  data: 3.7355  max mem: 57114
Epoch: [71]  [ 10/156]  eta: 0:02:42  lr: 0.007361  min_lr: 0.007361  loss: 3.5406 (3.3085)  weight_decay: 0.0500 (0.0500)  time: 1.1139  data: 0.3399  max mem: 57114
Epoch: [71]  [ 20/156]  eta: 0:02:06  lr: 0.007360  min_lr: 0.007360  loss: 3.5019 (3.3968)  weight_decay: 0.0500 (0.0500)  time: 0.7338  data: 0.0004  max mem: 57114
Epoch: [71]  [ 30/156]  eta: 0:01:47  lr: 0.007358  min_lr: 0.007358  loss: 3.3750 (3.3422)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0004  max mem: 57114
Epoch: [71]  [ 40/156]  eta: 0:01:35  lr: 0.007357  min_lr: 0.007357  loss: 3.3742 (3.3633)  weight_decay: 0.0500 (0.0500)  time: 0.7123  data: 0.0004  max mem: 57114
Epoch: [71]  [ 50/156]  eta: 0:01:25  lr: 0.007355  min_lr: 0.007355  loss: 3.4563 (3.3562)  weight_decay: 0.0500 (0.0500)  time: 0.7297  data: 0.0004  max mem: 57114
Epoch: [71]  [ 60/156]  eta: 0:01:15  lr: 0.007354  min_lr: 0.007354  loss: 3.4644 (3.3700)  weight_decay: 0.0500 (0.0500)  time: 0.7169  data: 0.0004  max mem: 57114
Epoch: [71]  [ 70/156]  eta: 0:01:07  lr: 0.007352  min_lr: 0.007352  loss: 3.6308 (3.4032)  weight_decay: 0.0500 (0.0500)  time: 0.7195  data: 0.0004  max mem: 57114
Epoch: [71]  [ 80/156]  eta: 0:00:58  lr: 0.007350  min_lr: 0.007350  loss: 3.5342 (3.3944)  weight_decay: 0.0500 (0.0500)  time: 0.7158  data: 0.0004  max mem: 57114
Epoch: [71]  [ 90/156]  eta: 0:00:50  lr: 0.007349  min_lr: 0.007349  loss: 3.3851 (3.3938)  weight_decay: 0.0500 (0.0500)  time: 0.6995  data: 0.0004  max mem: 57114
Epoch: [71]  [100/156]  eta: 0:00:42  lr: 0.007347  min_lr: 0.007347  loss: 3.3291 (3.3785)  weight_decay: 0.0500 (0.0500)  time: 0.6970  data: 0.0004  max mem: 57114
Epoch: [71]  [110/156]  eta: 0:00:34  lr: 0.007346  min_lr: 0.007346  loss: 3.3030 (3.3765)  weight_decay: 0.0500 (0.0500)  time: 0.7070  data: 0.0004  max mem: 57114
Epoch: [71]  [120/156]  eta: 0:00:26  lr: 0.007344  min_lr: 0.007344  loss: 3.5478 (3.3752)  weight_decay: 0.0500 (0.0500)  time: 0.7075  data: 0.0004  max mem: 57114
Epoch: [71]  [130/156]  eta: 0:00:19  lr: 0.007342  min_lr: 0.007342  loss: 3.5956 (3.3747)  weight_decay: 0.0500 (0.0500)  time: 0.6992  data: 0.0009  max mem: 57114
Epoch: [71]  [140/156]  eta: 0:00:11  lr: 0.007341  min_lr: 0.007341  loss: 3.6307 (3.3856)  weight_decay: 0.0500 (0.0500)  time: 0.6944  data: 0.0007  max mem: 57114
Epoch: [71]  [150/156]  eta: 0:00:04  lr: 0.007339  min_lr: 0.007339  loss: 3.3059 (3.3791)  weight_decay: 0.0500 (0.0500)  time: 0.6867  data: 0.0001  max mem: 57114
Epoch: [71]  [155/156]  eta: 0:00:00  lr: 0.007339  min_lr: 0.007339  loss: 3.6493 (3.3923)  weight_decay: 0.0500 (0.0500)  time: 0.6854  data: 0.0001  max mem: 57114
Epoch: [71] Total time: 0:01:55 (0.7376 s / it)
Averaged stats: lr: 0.007339  min_lr: 0.007339  loss: 3.6493 (3.4021)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 1.0732 (1.0732)  acc1: 80.7292 (80.7292)  acc5: 95.3125 (95.3125)  time: 2.2064  data: 1.9508  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.3182 (1.2362)  acc1: 72.3958 (72.2506)  acc5: 92.1875 (91.3044)  time: 0.5919  data: 0.3902  max mem: 57114
Test: Total time: 0:00:03 (0.6179 s / it)
* Acc@1 73.573 Acc@5 92.076 loss 1.292
Accuracy of the model on the 50000 test images: 73.6%
Max accuracy: 74.42%
Test:  [0/5]  eta: 0:00:11  loss: 7.5982 (7.5982)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.5208)  time: 2.3788  data: 2.1353  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.2977 (7.3516)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.5115)  time: 0.6264  data: 0.4271  max mem: 57114
Test: Total time: 0:00:03 (0.6365 s / it)
* Acc@1 0.108 Acc@5 0.523 loss 7.331
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [72]  [  0/156]  eta: 0:12:00  lr: 0.007338  min_lr: 0.007338  loss: 3.5625 (3.5625)  weight_decay: 0.0500 (0.0500)  time: 4.6203  data: 3.5817  max mem: 57114
Epoch: [72]  [ 10/156]  eta: 0:02:38  lr: 0.007337  min_lr: 0.007337  loss: 3.3503 (3.2530)  weight_decay: 0.0500 (0.0500)  time: 1.0870  data: 0.3259  max mem: 57114
Epoch: [72]  [ 20/156]  eta: 0:02:04  lr: 0.007335  min_lr: 0.007335  loss: 3.3098 (3.2698)  weight_decay: 0.0500 (0.0500)  time: 0.7316  data: 0.0003  max mem: 57114
Epoch: [72]  [ 30/156]  eta: 0:01:47  lr: 0.007334  min_lr: 0.007334  loss: 3.3779 (3.2812)  weight_decay: 0.0500 (0.0500)  time: 0.7253  data: 0.0003  max mem: 57114
Epoch: [72]  [ 40/156]  eta: 0:01:35  lr: 0.007332  min_lr: 0.007332  loss: 3.4339 (3.3173)  weight_decay: 0.0500 (0.0500)  time: 0.7247  data: 0.0003  max mem: 57114
Epoch: [72]  [ 50/156]  eta: 0:01:25  lr: 0.007330  min_lr: 0.007330  loss: 3.3904 (3.2606)  weight_decay: 0.0500 (0.0500)  time: 0.7282  data: 0.0004  max mem: 57114
Epoch: [72]  [ 60/156]  eta: 0:01:15  lr: 0.007329  min_lr: 0.007329  loss: 3.3904 (3.3106)  weight_decay: 0.0500 (0.0500)  time: 0.7259  data: 0.0004  max mem: 57114
Epoch: [72]  [ 70/156]  eta: 0:01:06  lr: 0.007327  min_lr: 0.007327  loss: 3.6223 (3.3461)  weight_decay: 0.0500 (0.0500)  time: 0.7121  data: 0.0004  max mem: 57114
Epoch: [72]  [ 80/156]  eta: 0:00:58  lr: 0.007326  min_lr: 0.007326  loss: 3.6267 (3.3682)  weight_decay: 0.0500 (0.0500)  time: 0.7157  data: 0.0004  max mem: 57114
Epoch: [72]  [ 90/156]  eta: 0:00:50  lr: 0.007324  min_lr: 0.007324  loss: 3.5976 (3.3766)  weight_decay: 0.0500 (0.0500)  time: 0.7166  data: 0.0004  max mem: 57114
Epoch: [72]  [100/156]  eta: 0:00:42  lr: 0.007322  min_lr: 0.007322  loss: 3.5473 (3.3821)  weight_decay: 0.0500 (0.0500)  time: 0.7069  data: 0.0004  max mem: 57114
Epoch: [72]  [110/156]  eta: 0:00:34  lr: 0.007321  min_lr: 0.007321  loss: 3.6312 (3.4062)  weight_decay: 0.0500 (0.0500)  time: 0.7079  data: 0.0003  max mem: 57114
Epoch: [72]  [120/156]  eta: 0:00:27  lr: 0.007319  min_lr: 0.007319  loss: 3.7051 (3.4159)  weight_decay: 0.0500 (0.0500)  time: 0.7035  data: 0.0004  max mem: 57114
Epoch: [72]  [130/156]  eta: 0:00:19  lr: 0.007318  min_lr: 0.007318  loss: 3.5175 (3.4122)  weight_decay: 0.0500 (0.0500)  time: 0.6930  data: 0.0009  max mem: 57114
Epoch: [72]  [140/156]  eta: 0:00:11  lr: 0.007316  min_lr: 0.007316  loss: 3.5364 (3.4256)  weight_decay: 0.0500 (0.0500)  time: 0.6839  data: 0.0008  max mem: 57114
Epoch: [72]  [150/156]  eta: 0:00:04  lr: 0.007314  min_lr: 0.007314  loss: 3.5289 (3.4255)  weight_decay: 0.0500 (0.0500)  time: 0.6890  data: 0.0001  max mem: 57114
Epoch: [72]  [155/156]  eta: 0:00:00  lr: 0.007314  min_lr: 0.007314  loss: 3.4713 (3.4218)  weight_decay: 0.0500 (0.0500)  time: 0.6897  data: 0.0001  max mem: 57114
Epoch: [72] Total time: 0:01:55 (0.7383 s / it)
Averaged stats: lr: 0.007314  min_lr: 0.007314  loss: 3.4713 (3.4059)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8885 (0.8885)  acc1: 82.2917 (82.2917)  acc5: 95.8333 (95.8333)  time: 2.0480  data: 1.7914  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0892 (1.0384)  acc1: 77.0833 (76.5985)  acc5: 94.7917 (93.2225)  time: 0.5604  data: 0.3584  max mem: 57114
Test: Total time: 0:00:02 (0.5816 s / it)
* Acc@1 74.944 Acc@5 92.817 loss 1.149
Accuracy of the model on the 50000 test images: 74.9%
Max accuracy: 74.94%
Test:  [0/5]  eta: 0:00:10  loss: 7.6266 (7.6266)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.5208)  time: 2.0085  data: 1.7650  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.3222 (7.3782)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.5115)  time: 0.5523  data: 0.3531  max mem: 57114
Test: Total time: 0:00:02 (0.5625 s / it)
* Acc@1 0.108 Acc@5 0.521 loss 7.355
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [73]  [  0/156]  eta: 0:12:50  lr: 0.007313  min_lr: 0.007313  loss: 3.6480 (3.6480)  weight_decay: 0.0500 (0.0500)  time: 4.9399  data: 3.9538  max mem: 57114
Epoch: [73]  [ 10/156]  eta: 0:02:46  lr: 0.007312  min_lr: 0.007312  loss: 3.5674 (3.3805)  weight_decay: 0.0500 (0.0500)  time: 1.1388  data: 0.3597  max mem: 57114
Epoch: [73]  [ 20/156]  eta: 0:02:08  lr: 0.007310  min_lr: 0.007310  loss: 3.4584 (3.3068)  weight_decay: 0.0500 (0.0500)  time: 0.7472  data: 0.0003  max mem: 57114
Epoch: [73]  [ 30/156]  eta: 0:01:50  lr: 0.007309  min_lr: 0.007309  loss: 3.6368 (3.4568)  weight_decay: 0.0500 (0.0500)  time: 0.7341  data: 0.0003  max mem: 57114
Epoch: [73]  [ 40/156]  eta: 0:01:37  lr: 0.007307  min_lr: 0.007307  loss: 3.6366 (3.4226)  weight_decay: 0.0500 (0.0500)  time: 0.7339  data: 0.0003  max mem: 57114
Epoch: [73]  [ 50/156]  eta: 0:01:26  lr: 0.007305  min_lr: 0.007305  loss: 3.4884 (3.4438)  weight_decay: 0.0500 (0.0500)  time: 0.7276  data: 0.0003  max mem: 57114
Epoch: [73]  [ 60/156]  eta: 0:01:17  lr: 0.007304  min_lr: 0.007304  loss: 3.5102 (3.4379)  weight_decay: 0.0500 (0.0500)  time: 0.7277  data: 0.0004  max mem: 57114
Epoch: [73]  [ 70/156]  eta: 0:01:08  lr: 0.007302  min_lr: 0.007302  loss: 3.5508 (3.4525)  weight_decay: 0.0500 (0.0500)  time: 0.7341  data: 0.0004  max mem: 57114
Epoch: [73]  [ 80/156]  eta: 0:00:59  lr: 0.007300  min_lr: 0.007300  loss: 3.5908 (3.4629)  weight_decay: 0.0500 (0.0500)  time: 0.7193  data: 0.0004  max mem: 57114
Epoch: [73]  [ 90/156]  eta: 0:00:51  lr: 0.007299  min_lr: 0.007299  loss: 3.5085 (3.4442)  weight_decay: 0.0500 (0.0500)  time: 0.7093  data: 0.0004  max mem: 57114
Epoch: [73]  [100/156]  eta: 0:00:43  lr: 0.007297  min_lr: 0.007297  loss: 3.6071 (3.4639)  weight_decay: 0.0500 (0.0500)  time: 0.7047  data: 0.0004  max mem: 57114
Epoch: [73]  [110/156]  eta: 0:00:35  lr: 0.007296  min_lr: 0.007296  loss: 3.7296 (3.4842)  weight_decay: 0.0500 (0.0500)  time: 0.6992  data: 0.0003  max mem: 57114
Epoch: [73]  [120/156]  eta: 0:00:27  lr: 0.007294  min_lr: 0.007294  loss: 3.7604 (3.4984)  weight_decay: 0.0500 (0.0500)  time: 0.7053  data: 0.0003  max mem: 57114
Epoch: [73]  [130/156]  eta: 0:00:19  lr: 0.007292  min_lr: 0.007292  loss: 3.5359 (3.4939)  weight_decay: 0.0500 (0.0500)  time: 0.7114  data: 0.0008  max mem: 57114
Epoch: [73]  [140/156]  eta: 0:00:11  lr: 0.007291  min_lr: 0.007291  loss: 3.4196 (3.4798)  weight_decay: 0.0500 (0.0500)  time: 0.7010  data: 0.0007  max mem: 57114
Epoch: [73]  [150/156]  eta: 0:00:04  lr: 0.007289  min_lr: 0.007289  loss: 3.4097 (3.4704)  weight_decay: 0.0500 (0.0500)  time: 0.6851  data: 0.0001  max mem: 57114
Epoch: [73]  [155/156]  eta: 0:00:00  lr: 0.007288  min_lr: 0.007288  loss: 3.4097 (3.4671)  weight_decay: 0.0500 (0.0500)  time: 0.6812  data: 0.0001  max mem: 57114
Epoch: [73] Total time: 0:01:56 (0.7454 s / it)
Averaged stats: lr: 0.007288  min_lr: 0.007288  loss: 3.4097 (3.4037)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0045 (1.0045)  acc1: 83.3333 (83.3333)  acc5: 96.3542 (96.3542)  time: 2.0478  data: 1.7922  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1430 (1.1016)  acc1: 78.1250 (76.9821)  acc5: 95.3125 (93.2225)  time: 0.5602  data: 0.3585  max mem: 57114
Test: Total time: 0:00:02 (0.5799 s / it)
* Acc@1 74.650 Acc@5 92.795 loss 1.205
Accuracy of the model on the 50000 test images: 74.7%
Max accuracy: 74.94%
Test:  [0/5]  eta: 0:00:11  loss: 7.6553 (7.6553)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.5208)  time: 2.3239  data: 2.0806  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.3499 (7.4063)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.5115)  time: 0.6154  data: 0.4162  max mem: 57114
Test: Total time: 0:00:03 (0.6282 s / it)
* Acc@1 0.108 Acc@5 0.529 loss 7.381
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [74]  [  0/156]  eta: 0:12:24  lr: 0.007288  min_lr: 0.007288  loss: 3.8801 (3.8801)  weight_decay: 0.0500 (0.0500)  time: 4.7700  data: 3.7131  max mem: 57114
Epoch: [74]  [ 10/156]  eta: 0:02:38  lr: 0.007286  min_lr: 0.007286  loss: 3.6656 (3.5517)  weight_decay: 0.0500 (0.0500)  time: 1.0866  data: 0.3378  max mem: 57114
Epoch: [74]  [ 20/156]  eta: 0:02:04  lr: 0.007285  min_lr: 0.007285  loss: 3.4759 (3.3748)  weight_decay: 0.0500 (0.0500)  time: 0.7231  data: 0.0003  max mem: 57114
Epoch: [74]  [ 30/156]  eta: 0:01:47  lr: 0.007283  min_lr: 0.007283  loss: 3.2539 (3.3652)  weight_decay: 0.0500 (0.0500)  time: 0.7250  data: 0.0003  max mem: 57114
Epoch: [74]  [ 40/156]  eta: 0:01:34  lr: 0.007282  min_lr: 0.007282  loss: 3.2014 (3.2893)  weight_decay: 0.0500 (0.0500)  time: 0.7162  data: 0.0003  max mem: 57114
Epoch: [74]  [ 50/156]  eta: 0:01:24  lr: 0.007280  min_lr: 0.007280  loss: 3.2729 (3.2861)  weight_decay: 0.0500 (0.0500)  time: 0.7186  data: 0.0003  max mem: 57114
Epoch: [74]  [ 60/156]  eta: 0:01:15  lr: 0.007278  min_lr: 0.007278  loss: 3.3757 (3.3151)  weight_decay: 0.0500 (0.0500)  time: 0.7313  data: 0.0004  max mem: 57114
Epoch: [74]  [ 70/156]  eta: 0:01:06  lr: 0.007277  min_lr: 0.007277  loss: 3.4843 (3.3282)  weight_decay: 0.0500 (0.0500)  time: 0.7193  data: 0.0004  max mem: 57114
Epoch: [74]  [ 80/156]  eta: 0:00:58  lr: 0.007275  min_lr: 0.007275  loss: 3.2850 (3.3170)  weight_decay: 0.0500 (0.0500)  time: 0.7128  data: 0.0003  max mem: 57114
Epoch: [74]  [ 90/156]  eta: 0:00:50  lr: 0.007273  min_lr: 0.007273  loss: 3.3134 (3.3234)  weight_decay: 0.0500 (0.0500)  time: 0.7202  data: 0.0003  max mem: 57114
Epoch: [74]  [100/156]  eta: 0:00:42  lr: 0.007272  min_lr: 0.007272  loss: 3.3263 (3.3297)  weight_decay: 0.0500 (0.0500)  time: 0.7093  data: 0.0004  max mem: 57114
Epoch: [74]  [110/156]  eta: 0:00:34  lr: 0.007270  min_lr: 0.007270  loss: 3.4845 (3.3192)  weight_decay: 0.0500 (0.0500)  time: 0.7038  data: 0.0004  max mem: 57114
Epoch: [74]  [120/156]  eta: 0:00:26  lr: 0.007268  min_lr: 0.007268  loss: 3.6721 (3.3529)  weight_decay: 0.0500 (0.0500)  time: 0.6988  data: 0.0004  max mem: 57114
Epoch: [74]  [130/156]  eta: 0:00:19  lr: 0.007267  min_lr: 0.007267  loss: 3.4697 (3.3452)  weight_decay: 0.0500 (0.0500)  time: 0.6902  data: 0.0009  max mem: 57114
Epoch: [74]  [140/156]  eta: 0:00:11  lr: 0.007265  min_lr: 0.007265  loss: 3.3989 (3.3475)  weight_decay: 0.0500 (0.0500)  time: 0.6875  data: 0.0007  max mem: 57114
Epoch: [74]  [150/156]  eta: 0:00:04  lr: 0.007263  min_lr: 0.007263  loss: 3.4621 (3.3573)  weight_decay: 0.0500 (0.0500)  time: 0.6858  data: 0.0001  max mem: 57114
Epoch: [74]  [155/156]  eta: 0:00:00  lr: 0.007263  min_lr: 0.007263  loss: 3.4621 (3.3633)  weight_decay: 0.0500 (0.0500)  time: 0.6847  data: 0.0001  max mem: 57114
Epoch: [74] Total time: 0:01:54 (0.7370 s / it)
Averaged stats: lr: 0.007263  min_lr: 0.007263  loss: 3.4621 (3.3994)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0235 (1.0235)  acc1: 81.2500 (81.2500)  acc5: 95.8333 (95.8333)  time: 2.0362  data: 1.7802  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2162 (1.1961)  acc1: 76.0417 (74.6803)  acc5: 95.8333 (93.2225)  time: 0.5580  data: 0.3561  max mem: 57114
Test: Total time: 0:00:02 (0.5795 s / it)
* Acc@1 74.516 Acc@5 92.435 loss 1.281
Accuracy of the model on the 50000 test images: 74.5%
Max accuracy: 74.94%
Test:  [0/5]  eta: 0:00:11  loss: 7.6854 (7.6854)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.5208)  time: 2.3188  data: 2.0751  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.3809 (7.4360)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.6394)  time: 0.6146  data: 0.4151  max mem: 57114
Test: Total time: 0:00:03 (0.6272 s / it)
* Acc@1 0.108 Acc@5 0.537 loss 7.408
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [75]  [  0/156]  eta: 0:12:08  lr: 0.007262  min_lr: 0.007262  loss: 3.6863 (3.6863)  weight_decay: 0.0500 (0.0500)  time: 4.6730  data: 3.4636  max mem: 57114
Epoch: [75]  [ 10/156]  eta: 0:02:41  lr: 0.007261  min_lr: 0.007261  loss: 3.5516 (3.4425)  weight_decay: 0.0500 (0.0500)  time: 1.1063  data: 0.3151  max mem: 57114
Epoch: [75]  [ 20/156]  eta: 0:02:04  lr: 0.007259  min_lr: 0.007259  loss: 3.3416 (3.2228)  weight_decay: 0.0500 (0.0500)  time: 0.7305  data: 0.0003  max mem: 57114
Epoch: [75]  [ 30/156]  eta: 0:01:46  lr: 0.007257  min_lr: 0.007257  loss: 3.2483 (3.2326)  weight_decay: 0.0500 (0.0500)  time: 0.7043  data: 0.0004  max mem: 57114
Epoch: [75]  [ 40/156]  eta: 0:01:35  lr: 0.007256  min_lr: 0.007256  loss: 3.3003 (3.2540)  weight_decay: 0.0500 (0.0500)  time: 0.7242  data: 0.0003  max mem: 57114
Epoch: [75]  [ 50/156]  eta: 0:01:26  lr: 0.007254  min_lr: 0.007254  loss: 3.3346 (3.2359)  weight_decay: 0.0500 (0.0500)  time: 0.7567  data: 0.0004  max mem: 57114
Epoch: [75]  [ 60/156]  eta: 0:01:16  lr: 0.007252  min_lr: 0.007252  loss: 3.3492 (3.2634)  weight_decay: 0.0500 (0.0500)  time: 0.7449  data: 0.0004  max mem: 57114
Epoch: [75]  [ 70/156]  eta: 0:01:07  lr: 0.007251  min_lr: 0.007251  loss: 3.4983 (3.2932)  weight_decay: 0.0500 (0.0500)  time: 0.7175  data: 0.0004  max mem: 57114
Epoch: [75]  [ 80/156]  eta: 0:00:58  lr: 0.007249  min_lr: 0.007249  loss: 3.4983 (3.3191)  weight_decay: 0.0500 (0.0500)  time: 0.7078  data: 0.0004  max mem: 57114
Epoch: [75]  [ 90/156]  eta: 0:00:50  lr: 0.007247  min_lr: 0.007247  loss: 3.3846 (3.3201)  weight_decay: 0.0500 (0.0500)  time: 0.7059  data: 0.0004  max mem: 57114
Epoch: [75]  [100/156]  eta: 0:00:42  lr: 0.007246  min_lr: 0.007246  loss: 3.2806 (3.3162)  weight_decay: 0.0500 (0.0500)  time: 0.7028  data: 0.0004  max mem: 57114
Epoch: [75]  [110/156]  eta: 0:00:34  lr: 0.007244  min_lr: 0.007244  loss: 3.4372 (3.3453)  weight_decay: 0.0500 (0.0500)  time: 0.7067  data: 0.0004  max mem: 57114
Epoch: [75]  [120/156]  eta: 0:00:27  lr: 0.007242  min_lr: 0.007242  loss: 3.6445 (3.3519)  weight_decay: 0.0500 (0.0500)  time: 0.7088  data: 0.0004  max mem: 57114
Epoch: [75]  [130/156]  eta: 0:00:19  lr: 0.007241  min_lr: 0.007241  loss: 3.5218 (3.3533)  weight_decay: 0.0500 (0.0500)  time: 0.7016  data: 0.0008  max mem: 57114
Epoch: [75]  [140/156]  eta: 0:00:11  lr: 0.007239  min_lr: 0.007239  loss: 3.3644 (3.3527)  weight_decay: 0.0500 (0.0500)  time: 0.6920  data: 0.0007  max mem: 57114
Epoch: [75]  [150/156]  eta: 0:00:04  lr: 0.007237  min_lr: 0.007237  loss: 3.5353 (3.3674)  weight_decay: 0.0500 (0.0500)  time: 0.7037  data: 0.0001  max mem: 57114
Epoch: [75]  [155/156]  eta: 0:00:00  lr: 0.007236  min_lr: 0.007236  loss: 3.4918 (3.3643)  weight_decay: 0.0500 (0.0500)  time: 0.7029  data: 0.0001  max mem: 57114
Epoch: [75] Total time: 0:01:55 (0.7423 s / it)
Averaged stats: lr: 0.007236  min_lr: 0.007236  loss: 3.4918 (3.3947)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.1273 (1.1273)  acc1: 80.7292 (80.7292)  acc5: 93.7500 (93.7500)  time: 2.0659  data: 1.8101  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2137 (1.2649)  acc1: 79.1667 (75.7033)  acc5: 93.7500 (92.7110)  time: 0.5639  data: 0.3621  max mem: 57114
Test: Total time: 0:00:02 (0.5865 s / it)
* Acc@1 74.500 Acc@5 92.431 loss 1.346
Accuracy of the model on the 50000 test images: 74.5%
Max accuracy: 74.94%
Test:  [0/5]  eta: 0:00:11  loss: 7.7223 (7.7223)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.5208)  time: 2.2899  data: 2.0465  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.4254 (7.4734)  acc1: 0.0000 (0.1279)  acc5: 0.0000 (0.6394)  time: 0.6087  data: 0.4094  max mem: 57114
Test: Total time: 0:00:03 (0.6202 s / it)
* Acc@1 0.108 Acc@5 0.591 loss 7.442
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [76]  [  0/156]  eta: 0:12:45  lr: 0.007236  min_lr: 0.007236  loss: 2.4127 (2.4127)  weight_decay: 0.0500 (0.0500)  time: 4.9078  data: 2.7365  max mem: 57114
Epoch: [76]  [ 10/156]  eta: 0:02:42  lr: 0.007234  min_lr: 0.007234  loss: 3.2767 (3.1805)  weight_decay: 0.0500 (0.0500)  time: 1.1154  data: 0.2491  max mem: 57114
Epoch: [76]  [ 20/156]  eta: 0:02:07  lr: 0.007233  min_lr: 0.007233  loss: 3.3795 (3.2597)  weight_decay: 0.0500 (0.0500)  time: 0.7393  data: 0.0004  max mem: 57114
Epoch: [76]  [ 30/156]  eta: 0:01:49  lr: 0.007231  min_lr: 0.007231  loss: 3.5217 (3.2492)  weight_decay: 0.0500 (0.0500)  time: 0.7343  data: 0.0004  max mem: 57114
Epoch: [76]  [ 40/156]  eta: 0:01:37  lr: 0.007229  min_lr: 0.007229  loss: 3.5803 (3.2838)  weight_decay: 0.0500 (0.0500)  time: 0.7380  data: 0.0004  max mem: 57114
Epoch: [76]  [ 50/156]  eta: 0:01:26  lr: 0.007228  min_lr: 0.007228  loss: 3.5872 (3.3316)  weight_decay: 0.0500 (0.0500)  time: 0.7360  data: 0.0004  max mem: 57114
Epoch: [76]  [ 60/156]  eta: 0:01:16  lr: 0.007226  min_lr: 0.007226  loss: 3.5278 (3.3447)  weight_decay: 0.0500 (0.0500)  time: 0.7196  data: 0.0004  max mem: 57114
Epoch: [76]  [ 70/156]  eta: 0:01:07  lr: 0.007224  min_lr: 0.007224  loss: 3.4442 (3.3722)  weight_decay: 0.0500 (0.0500)  time: 0.7194  data: 0.0003  max mem: 57114
Epoch: [76]  [ 80/156]  eta: 0:00:59  lr: 0.007223  min_lr: 0.007223  loss: 3.3559 (3.3448)  weight_decay: 0.0500 (0.0500)  time: 0.7106  data: 0.0004  max mem: 57114
Epoch: [76]  [ 90/156]  eta: 0:00:50  lr: 0.007221  min_lr: 0.007221  loss: 3.3498 (3.3556)  weight_decay: 0.0500 (0.0500)  time: 0.7025  data: 0.0004  max mem: 57114
Epoch: [76]  [100/156]  eta: 0:00:42  lr: 0.007219  min_lr: 0.007219  loss: 3.5026 (3.3470)  weight_decay: 0.0500 (0.0500)  time: 0.7091  data: 0.0003  max mem: 57114
Epoch: [76]  [110/156]  eta: 0:00:34  lr: 0.007217  min_lr: 0.007217  loss: 3.4909 (3.3687)  weight_decay: 0.0500 (0.0500)  time: 0.7117  data: 0.0004  max mem: 57114
Epoch: [76]  [120/156]  eta: 0:00:27  lr: 0.007216  min_lr: 0.007216  loss: 3.5152 (3.3806)  weight_decay: 0.0500 (0.0500)  time: 0.7054  data: 0.0004  max mem: 57114
Epoch: [76]  [130/156]  eta: 0:00:19  lr: 0.007214  min_lr: 0.007214  loss: 3.3381 (3.3651)  weight_decay: 0.0500 (0.0500)  time: 0.6992  data: 0.0008  max mem: 57114
Epoch: [76]  [140/156]  eta: 0:00:11  lr: 0.007212  min_lr: 0.007212  loss: 3.1609 (3.3624)  weight_decay: 0.0500 (0.0500)  time: 0.6888  data: 0.0007  max mem: 57114
Epoch: [76]  [150/156]  eta: 0:00:04  lr: 0.007211  min_lr: 0.007211  loss: 3.6188 (3.3742)  weight_decay: 0.0500 (0.0500)  time: 0.6803  data: 0.0001  max mem: 57114
Epoch: [76]  [155/156]  eta: 0:00:00  lr: 0.007210  min_lr: 0.007210  loss: 3.5637 (3.3735)  weight_decay: 0.0500 (0.0500)  time: 0.6823  data: 0.0001  max mem: 57114
Epoch: [76] Total time: 0:01:55 (0.7419 s / it)
Averaged stats: lr: 0.007210  min_lr: 0.007210  loss: 3.5637 (3.3908)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9252 (0.9252)  acc1: 80.7292 (80.7292)  acc5: 96.8750 (96.8750)  time: 2.0993  data: 1.8435  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1567 (1.0841)  acc1: 74.4792 (74.0409)  acc5: 96.8750 (93.2225)  time: 0.5706  data: 0.3688  max mem: 57114
Test: Total time: 0:00:02 (0.5933 s / it)
* Acc@1 73.881 Acc@5 92.223 loss 1.197
Accuracy of the model on the 50000 test images: 73.9%
Max accuracy: 74.94%
Test:  [0/5]  eta: 0:00:11  loss: 7.7666 (7.7666)  acc1: 0.0000 (0.0000)  acc5: 0.5208 (0.5208)  time: 2.3384  data: 2.0950  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.4726 (7.5157)  acc1: 0.0000 (0.1279)  acc5: 0.5208 (0.8951)  time: 0.6183  data: 0.4191  max mem: 57114
Test: Total time: 0:00:03 (0.6292 s / it)
* Acc@1 0.108 Acc@5 0.601 loss 7.482
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [77]  [  0/156]  eta: 0:12:49  lr: 0.007210  min_lr: 0.007210  loss: 3.0532 (3.0532)  weight_decay: 0.0500 (0.0500)  time: 4.9306  data: 3.7539  max mem: 57114
Epoch: [77]  [ 10/156]  eta: 0:02:44  lr: 0.007208  min_lr: 0.007208  loss: 3.3726 (3.3054)  weight_decay: 0.0500 (0.0500)  time: 1.1280  data: 0.3416  max mem: 57114
Epoch: [77]  [ 20/156]  eta: 0:02:07  lr: 0.007206  min_lr: 0.007206  loss: 3.3726 (3.3355)  weight_decay: 0.0500 (0.0500)  time: 0.7373  data: 0.0004  max mem: 57114
Epoch: [77]  [ 30/156]  eta: 0:01:49  lr: 0.007204  min_lr: 0.007204  loss: 3.5579 (3.4069)  weight_decay: 0.0500 (0.0500)  time: 0.7265  data: 0.0004  max mem: 57114
Epoch: [77]  [ 40/156]  eta: 0:01:37  lr: 0.007203  min_lr: 0.007203  loss: 3.5579 (3.3825)  weight_decay: 0.0500 (0.0500)  time: 0.7318  data: 0.0004  max mem: 57114
Epoch: [77]  [ 50/156]  eta: 0:01:26  lr: 0.007201  min_lr: 0.007201  loss: 3.5207 (3.4121)  weight_decay: 0.0500 (0.0500)  time: 0.7259  data: 0.0004  max mem: 57114
Epoch: [77]  [ 60/156]  eta: 0:01:16  lr: 0.007199  min_lr: 0.007199  loss: 3.6478 (3.4324)  weight_decay: 0.0500 (0.0500)  time: 0.7119  data: 0.0004  max mem: 57114
Epoch: [77]  [ 70/156]  eta: 0:01:07  lr: 0.007198  min_lr: 0.007198  loss: 3.4340 (3.4129)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0004  max mem: 57114
Epoch: [77]  [ 80/156]  eta: 0:00:59  lr: 0.007196  min_lr: 0.007196  loss: 3.2723 (3.3916)  weight_decay: 0.0500 (0.0500)  time: 0.7187  data: 0.0004  max mem: 57114
Epoch: [77]  [ 90/156]  eta: 0:00:50  lr: 0.007194  min_lr: 0.007194  loss: 3.2001 (3.3545)  weight_decay: 0.0500 (0.0500)  time: 0.7098  data: 0.0004  max mem: 57114
Epoch: [77]  [100/156]  eta: 0:00:42  lr: 0.007192  min_lr: 0.007192  loss: 3.2062 (3.3555)  weight_decay: 0.0500 (0.0500)  time: 0.7079  data: 0.0004  max mem: 57114
Epoch: [77]  [110/156]  eta: 0:00:34  lr: 0.007191  min_lr: 0.007191  loss: 3.3032 (3.3529)  weight_decay: 0.0500 (0.0500)  time: 0.7077  data: 0.0004  max mem: 57114
Epoch: [77]  [120/156]  eta: 0:00:27  lr: 0.007189  min_lr: 0.007189  loss: 3.4047 (3.3459)  weight_decay: 0.0500 (0.0500)  time: 0.7034  data: 0.0004  max mem: 57114
Epoch: [77]  [130/156]  eta: 0:00:19  lr: 0.007187  min_lr: 0.007187  loss: 3.2840 (3.3279)  weight_decay: 0.0500 (0.0500)  time: 0.6975  data: 0.0008  max mem: 57114
Epoch: [77]  [140/156]  eta: 0:00:11  lr: 0.007185  min_lr: 0.007185  loss: 3.2165 (3.3269)  weight_decay: 0.0500 (0.0500)  time: 0.6906  data: 0.0007  max mem: 57114
Epoch: [77]  [150/156]  eta: 0:00:04  lr: 0.007184  min_lr: 0.007184  loss: 3.4293 (3.3313)  weight_decay: 0.0500 (0.0500)  time: 0.6947  data: 0.0001  max mem: 57114
Epoch: [77]  [155/156]  eta: 0:00:00  lr: 0.007183  min_lr: 0.007183  loss: 3.4671 (3.3353)  weight_decay: 0.0500 (0.0500)  time: 0.6844  data: 0.0001  max mem: 57114
Epoch: [77] Total time: 0:01:55 (0.7408 s / it)
Averaged stats: lr: 0.007183  min_lr: 0.007183  loss: 3.4671 (3.3810)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0738 (1.0738)  acc1: 78.6458 (78.6458)  acc5: 94.7917 (94.7917)  time: 2.0595  data: 1.8026  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1493 (1.1800)  acc1: 78.1250 (73.7852)  acc5: 94.7917 (93.2225)  time: 0.5626  data: 0.3606  max mem: 57114
Test: Total time: 0:00:02 (0.5820 s / it)
* Acc@1 73.308 Acc@5 92.040 loss 1.265
Accuracy of the model on the 50000 test images: 73.3%
Max accuracy: 74.94%
Test:  [0/5]  eta: 0:00:11  loss: 7.8056 (7.8056)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.2946  data: 2.0510  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.5318 (7.5613)  acc1: 0.0000 (0.1279)  acc5: 1.0417 (1.1509)  time: 0.6096  data: 0.4103  max mem: 57114
Test: Total time: 0:00:03 (0.6234 s / it)
* Acc@1 0.106 Acc@5 0.685 loss 7.523
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [78]  [  0/156]  eta: 0:12:27  lr: 0.007183  min_lr: 0.007183  loss: 3.8536 (3.8536)  weight_decay: 0.0500 (0.0500)  time: 4.7910  data: 3.2710  max mem: 57114
Epoch: [78]  [ 10/156]  eta: 0:02:41  lr: 0.007181  min_lr: 0.007181  loss: 3.3207 (3.2888)  weight_decay: 0.0500 (0.0500)  time: 1.1055  data: 0.2977  max mem: 57114
Epoch: [78]  [ 20/156]  eta: 0:02:04  lr: 0.007179  min_lr: 0.007179  loss: 3.1133 (3.1960)  weight_decay: 0.0500 (0.0500)  time: 0.7251  data: 0.0003  max mem: 57114
Epoch: [78]  [ 30/156]  eta: 0:01:47  lr: 0.007177  min_lr: 0.007177  loss: 3.0731 (3.1808)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0003  max mem: 57114
Epoch: [78]  [ 40/156]  eta: 0:01:35  lr: 0.007176  min_lr: 0.007176  loss: 3.1950 (3.1856)  weight_decay: 0.0500 (0.0500)  time: 0.7241  data: 0.0004  max mem: 57114
Epoch: [78]  [ 50/156]  eta: 0:01:25  lr: 0.007174  min_lr: 0.007174  loss: 3.3601 (3.2145)  weight_decay: 0.0500 (0.0500)  time: 0.7394  data: 0.0004  max mem: 57114
Epoch: [78]  [ 60/156]  eta: 0:01:16  lr: 0.007172  min_lr: 0.007172  loss: 3.2530 (3.2339)  weight_decay: 0.0500 (0.0500)  time: 0.7285  data: 0.0004  max mem: 57114
Epoch: [78]  [ 70/156]  eta: 0:01:07  lr: 0.007170  min_lr: 0.007170  loss: 3.2944 (3.2625)  weight_decay: 0.0500 (0.0500)  time: 0.7123  data: 0.0004  max mem: 57114
Epoch: [78]  [ 80/156]  eta: 0:00:58  lr: 0.007169  min_lr: 0.007169  loss: 3.2944 (3.2754)  weight_decay: 0.0500 (0.0500)  time: 0.7154  data: 0.0003  max mem: 57114
Epoch: [78]  [ 90/156]  eta: 0:00:50  lr: 0.007167  min_lr: 0.007167  loss: 3.3977 (3.2919)  weight_decay: 0.0500 (0.0500)  time: 0.7125  data: 0.0003  max mem: 57114
Epoch: [78]  [100/156]  eta: 0:00:42  lr: 0.007165  min_lr: 0.007165  loss: 3.4754 (3.3115)  weight_decay: 0.0500 (0.0500)  time: 0.7035  data: 0.0004  max mem: 57114
Epoch: [78]  [110/156]  eta: 0:00:34  lr: 0.007163  min_lr: 0.007163  loss: 3.4334 (3.3001)  weight_decay: 0.0500 (0.0500)  time: 0.7075  data: 0.0004  max mem: 57114
Epoch: [78]  [120/156]  eta: 0:00:27  lr: 0.007162  min_lr: 0.007162  loss: 3.5088 (3.3266)  weight_decay: 0.0500 (0.0500)  time: 0.7105  data: 0.0004  max mem: 57114
Epoch: [78]  [130/156]  eta: 0:00:19  lr: 0.007160  min_lr: 0.007160  loss: 3.5965 (3.3429)  weight_decay: 0.0500 (0.0500)  time: 0.6987  data: 0.0007  max mem: 57114
Epoch: [78]  [140/156]  eta: 0:00:11  lr: 0.007158  min_lr: 0.007158  loss: 3.5551 (3.3464)  weight_decay: 0.0500 (0.0500)  time: 0.6882  data: 0.0006  max mem: 57114
Epoch: [78]  [150/156]  eta: 0:00:04  lr: 0.007156  min_lr: 0.007156  loss: 3.3650 (3.3485)  weight_decay: 0.0500 (0.0500)  time: 0.6835  data: 0.0001  max mem: 57114
Epoch: [78]  [155/156]  eta: 0:00:00  lr: 0.007155  min_lr: 0.007155  loss: 3.3841 (3.3547)  weight_decay: 0.0500 (0.0500)  time: 0.6789  data: 0.0001  max mem: 57114
Epoch: [78] Total time: 0:01:55 (0.7386 s / it)
Averaged stats: lr: 0.007155  min_lr: 0.007155  loss: 3.3841 (3.3742)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.1189 (1.1189)  acc1: 80.2083 (80.2083)  acc5: 95.3125 (95.3125)  time: 2.1335  data: 1.8773  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2619 (1.2465)  acc1: 77.0833 (73.9130)  acc5: 94.7917 (91.6880)  time: 0.5773  data: 0.3755  max mem: 57114
Test: Total time: 0:00:03 (0.6019 s / it)
* Acc@1 72.542 Acc@5 91.536 loss 1.333
Accuracy of the model on the 50000 test images: 72.5%
Max accuracy: 74.94%
Test:  [0/5]  eta: 0:00:11  loss: 7.8473 (7.8473)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.3236  data: 2.0797  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.5922 (7.6094)  acc1: 0.0000 (0.1279)  acc5: 1.0417 (1.0230)  time: 0.6155  data: 0.4160  max mem: 57114
Test: Total time: 0:00:03 (0.6300 s / it)
* Acc@1 0.106 Acc@5 0.727 loss 7.567
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [79]  [  0/156]  eta: 0:11:56  lr: 0.007155  min_lr: 0.007155  loss: 3.7084 (3.7084)  weight_decay: 0.0500 (0.0500)  time: 4.5956  data: 3.3689  max mem: 57114
Epoch: [79]  [ 10/156]  eta: 0:02:39  lr: 0.007153  min_lr: 0.007153  loss: 3.2150 (3.1123)  weight_decay: 0.0500 (0.0500)  time: 1.0918  data: 0.3065  max mem: 57114
Epoch: [79]  [ 20/156]  eta: 0:02:05  lr: 0.007152  min_lr: 0.007152  loss: 3.5602 (3.3778)  weight_decay: 0.0500 (0.0500)  time: 0.7426  data: 0.0003  max mem: 57114
Epoch: [79]  [ 30/156]  eta: 0:01:50  lr: 0.007150  min_lr: 0.007150  loss: 3.4743 (3.3134)  weight_decay: 0.0500 (0.0500)  time: 0.7618  data: 0.0003  max mem: 57114
Epoch: [79]  [ 40/156]  eta: 0:01:38  lr: 0.007148  min_lr: 0.007148  loss: 3.1619 (3.2799)  weight_decay: 0.0500 (0.0500)  time: 0.7705  data: 0.0005  max mem: 57114
Epoch: [79]  [ 50/156]  eta: 0:01:28  lr: 0.007146  min_lr: 0.007146  loss: 3.0429 (3.2369)  weight_decay: 0.0500 (0.0500)  time: 0.7594  data: 0.0007  max mem: 57114
Epoch: [79]  [ 60/156]  eta: 0:01:18  lr: 0.007145  min_lr: 0.007145  loss: 3.4668 (3.2661)  weight_decay: 0.0500 (0.0500)  time: 0.7442  data: 0.0006  max mem: 57114
Epoch: [79]  [ 70/156]  eta: 0:01:09  lr: 0.007143  min_lr: 0.007143  loss: 3.4685 (3.2710)  weight_decay: 0.0500 (0.0500)  time: 0.7433  data: 0.0005  max mem: 57114
Epoch: [79]  [ 80/156]  eta: 0:01:00  lr: 0.007141  min_lr: 0.007141  loss: 3.2430 (3.2701)  weight_decay: 0.0500 (0.0500)  time: 0.7507  data: 0.0004  max mem: 57114
Epoch: [79]  [ 90/156]  eta: 0:00:52  lr: 0.007139  min_lr: 0.007139  loss: 3.3776 (3.2866)  weight_decay: 0.0500 (0.0500)  time: 0.7369  data: 0.0004  max mem: 57114
Epoch: [79]  [100/156]  eta: 0:00:43  lr: 0.007137  min_lr: 0.007137  loss: 3.4196 (3.2867)  weight_decay: 0.0500 (0.0500)  time: 0.7194  data: 0.0004  max mem: 57114
Epoch: [79]  [110/156]  eta: 0:00:35  lr: 0.007136  min_lr: 0.007136  loss: 3.2954 (3.2793)  weight_decay: 0.0500 (0.0500)  time: 0.7091  data: 0.0004  max mem: 57114
Epoch: [79]  [120/156]  eta: 0:00:27  lr: 0.007134  min_lr: 0.007134  loss: 3.3295 (3.2784)  weight_decay: 0.0500 (0.0500)  time: 0.7033  data: 0.0004  max mem: 57114
Epoch: [79]  [130/156]  eta: 0:00:19  lr: 0.007132  min_lr: 0.007132  loss: 3.3374 (3.2767)  weight_decay: 0.0500 (0.0500)  time: 0.6959  data: 0.0009  max mem: 57114
Epoch: [79]  [140/156]  eta: 0:00:12  lr: 0.007130  min_lr: 0.007130  loss: 3.4701 (3.3029)  weight_decay: 0.0500 (0.0500)  time: 0.6848  data: 0.0007  max mem: 57114
Epoch: [79]  [150/156]  eta: 0:00:04  lr: 0.007129  min_lr: 0.007129  loss: 3.5398 (3.3073)  weight_decay: 0.0500 (0.0500)  time: 0.6824  data: 0.0001  max mem: 57114
Epoch: [79]  [155/156]  eta: 0:00:00  lr: 0.007128  min_lr: 0.007128  loss: 3.5398 (3.3102)  weight_decay: 0.0500 (0.0500)  time: 0.6852  data: 0.0001  max mem: 57114
Epoch: [79] Total time: 0:01:57 (0.7534 s / it)
Averaged stats: lr: 0.007128  min_lr: 0.007128  loss: 3.5398 (3.3807)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0097 (1.0097)  acc1: 83.8542 (83.8542)  acc5: 95.3125 (95.3125)  time: 2.0436  data: 1.7876  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1356 (1.1937)  acc1: 75.5208 (74.2967)  acc5: 95.3125 (92.1995)  time: 0.5594  data: 0.3576  max mem: 57114
Test: Total time: 0:00:02 (0.5818 s / it)
* Acc@1 73.783 Acc@5 92.128 loss 1.272
Accuracy of the model on the 50000 test images: 73.8%
Max accuracy: 74.94%
Test:  [0/5]  eta: 0:00:11  loss: 7.8978 (7.8978)  acc1: 0.0000 (0.0000)  acc5: 1.0417 (1.0417)  time: 2.2998  data: 2.0564  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.6606 (7.6674)  acc1: 0.0000 (0.1279)  acc5: 0.5208 (0.7673)  time: 0.6105  data: 0.4113  max mem: 57114
Test: Total time: 0:00:03 (0.6262 s / it)
* Acc@1 0.106 Acc@5 0.717 loss 7.619
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [80]  [  0/156]  eta: 0:12:36  lr: 0.007127  min_lr: 0.007127  loss: 3.0597 (3.0597)  weight_decay: 0.0500 (0.0500)  time: 4.8489  data: 3.6960  max mem: 57114
Epoch: [80]  [ 10/156]  eta: 0:02:45  lr: 0.007126  min_lr: 0.007126  loss: 3.4141 (3.2385)  weight_decay: 0.0500 (0.0500)  time: 1.1337  data: 0.3363  max mem: 57114
Epoch: [80]  [ 20/156]  eta: 0:02:07  lr: 0.007124  min_lr: 0.007124  loss: 3.3649 (3.2751)  weight_decay: 0.0500 (0.0500)  time: 0.7394  data: 0.0003  max mem: 57114
Epoch: [80]  [ 30/156]  eta: 0:01:49  lr: 0.007122  min_lr: 0.007122  loss: 3.4325 (3.3401)  weight_decay: 0.0500 (0.0500)  time: 0.7234  data: 0.0003  max mem: 57114
Epoch: [80]  [ 40/156]  eta: 0:01:36  lr: 0.007120  min_lr: 0.007120  loss: 3.5569 (3.3978)  weight_decay: 0.0500 (0.0500)  time: 0.7245  data: 0.0004  max mem: 57114
Epoch: [80]  [ 50/156]  eta: 0:01:25  lr: 0.007118  min_lr: 0.007118  loss: 3.4837 (3.4041)  weight_decay: 0.0500 (0.0500)  time: 0.7141  data: 0.0004  max mem: 57114
Epoch: [80]  [ 60/156]  eta: 0:01:16  lr: 0.007117  min_lr: 0.007117  loss: 3.2810 (3.3756)  weight_decay: 0.0500 (0.0500)  time: 0.7213  data: 0.0005  max mem: 57114
Epoch: [80]  [ 70/156]  eta: 0:01:07  lr: 0.007115  min_lr: 0.007115  loss: 3.2810 (3.3876)  weight_decay: 0.0500 (0.0500)  time: 0.7189  data: 0.0005  max mem: 57114
Epoch: [80]  [ 80/156]  eta: 0:00:59  lr: 0.007113  min_lr: 0.007113  loss: 3.5536 (3.3814)  weight_decay: 0.0500 (0.0500)  time: 0.7188  data: 0.0004  max mem: 57114
Epoch: [80]  [ 90/156]  eta: 0:00:50  lr: 0.007111  min_lr: 0.007111  loss: 3.2782 (3.3669)  weight_decay: 0.0500 (0.0500)  time: 0.7196  data: 0.0004  max mem: 57114
Epoch: [80]  [100/156]  eta: 0:00:42  lr: 0.007109  min_lr: 0.007109  loss: 3.3539 (3.3918)  weight_decay: 0.0500 (0.0500)  time: 0.7048  data: 0.0004  max mem: 57114
Epoch: [80]  [110/156]  eta: 0:00:34  lr: 0.007108  min_lr: 0.007108  loss: 3.7064 (3.4149)  weight_decay: 0.0500 (0.0500)  time: 0.7093  data: 0.0004  max mem: 57114
Epoch: [80]  [120/156]  eta: 0:00:27  lr: 0.007106  min_lr: 0.007106  loss: 3.5052 (3.4100)  weight_decay: 0.0500 (0.0500)  time: 0.6997  data: 0.0004  max mem: 57114
Epoch: [80]  [130/156]  eta: 0:00:19  lr: 0.007104  min_lr: 0.007104  loss: 3.3434 (3.4020)  weight_decay: 0.0500 (0.0500)  time: 0.6878  data: 0.0008  max mem: 57114
Epoch: [80]  [140/156]  eta: 0:00:11  lr: 0.007102  min_lr: 0.007102  loss: 3.4530 (3.3984)  weight_decay: 0.0500 (0.0500)  time: 0.6885  data: 0.0007  max mem: 57114
Epoch: [80]  [150/156]  eta: 0:00:04  lr: 0.007100  min_lr: 0.007100  loss: 3.4530 (3.4007)  weight_decay: 0.0500 (0.0500)  time: 0.6870  data: 0.0001  max mem: 57114
Epoch: [80]  [155/156]  eta: 0:00:00  lr: 0.007099  min_lr: 0.007099  loss: 3.3411 (3.3963)  weight_decay: 0.0500 (0.0500)  time: 0.6845  data: 0.0001  max mem: 57114
Epoch: [80] Total time: 0:01:55 (0.7399 s / it)
Averaged stats: lr: 0.007099  min_lr: 0.007099  loss: 3.3411 (3.3719)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9277 (0.9277)  acc1: 82.8125 (82.8125)  acc5: 95.8333 (95.8333)  time: 2.0358  data: 1.7798  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0143 (1.0781)  acc1: 81.7708 (76.7263)  acc5: 95.8333 (93.6061)  time: 0.5579  data: 0.3560  max mem: 57114
Test: Total time: 0:00:02 (0.5798 s / it)
* Acc@1 75.370 Acc@5 92.935 loss 1.170
Accuracy of the model on the 50000 test images: 75.4%
Max accuracy: 75.37%
Test:  [0/5]  eta: 0:00:10  loss: 7.9493 (7.9493)  acc1: 0.0000 (0.0000)  acc5: 1.5625 (1.5625)  time: 2.0411  data: 1.7978  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.7311 (7.7242)  acc1: 0.0000 (0.1279)  acc5: 0.5208 (0.8951)  time: 0.5588  data: 0.3596  max mem: 57114
Test: Total time: 0:00:02 (0.5718 s / it)
* Acc@1 0.110 Acc@5 0.691 loss 7.670
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [81]  [  0/156]  eta: 0:12:54  lr: 0.007099  min_lr: 0.007099  loss: 2.3271 (2.3271)  weight_decay: 0.0500 (0.0500)  time: 4.9677  data: 3.9402  max mem: 57114
Epoch: [81]  [ 10/156]  eta: 0:02:46  lr: 0.007097  min_lr: 0.007097  loss: 3.5035 (3.3105)  weight_decay: 0.0500 (0.0500)  time: 1.1378  data: 0.3585  max mem: 57114
Epoch: [81]  [ 20/156]  eta: 0:02:07  lr: 0.007096  min_lr: 0.007096  loss: 3.5647 (3.3735)  weight_decay: 0.0500 (0.0500)  time: 0.7336  data: 0.0003  max mem: 57114
Epoch: [81]  [ 30/156]  eta: 0:01:49  lr: 0.007094  min_lr: 0.007094  loss: 3.4847 (3.3802)  weight_decay: 0.0500 (0.0500)  time: 0.7266  data: 0.0003  max mem: 57114
Epoch: [81]  [ 40/156]  eta: 0:01:37  lr: 0.007092  min_lr: 0.007092  loss: 3.4847 (3.4226)  weight_decay: 0.0500 (0.0500)  time: 0.7382  data: 0.0003  max mem: 57114
Epoch: [81]  [ 50/156]  eta: 0:01:27  lr: 0.007090  min_lr: 0.007090  loss: 3.5470 (3.4219)  weight_decay: 0.0500 (0.0500)  time: 0.7429  data: 0.0004  max mem: 57114
Epoch: [81]  [ 60/156]  eta: 0:01:17  lr: 0.007088  min_lr: 0.007088  loss: 3.5090 (3.4190)  weight_decay: 0.0500 (0.0500)  time: 0.7374  data: 0.0004  max mem: 57114
Epoch: [81]  [ 70/156]  eta: 0:01:08  lr: 0.007086  min_lr: 0.007086  loss: 3.5728 (3.4162)  weight_decay: 0.0500 (0.0500)  time: 0.7395  data: 0.0004  max mem: 57114
Epoch: [81]  [ 80/156]  eta: 0:00:59  lr: 0.007085  min_lr: 0.007085  loss: 3.6206 (3.4213)  weight_decay: 0.0500 (0.0500)  time: 0.7362  data: 0.0004  max mem: 57114
Epoch: [81]  [ 90/156]  eta: 0:00:51  lr: 0.007083  min_lr: 0.007083  loss: 3.6427 (3.4271)  weight_decay: 0.0500 (0.0500)  time: 0.7022  data: 0.0003  max mem: 57114
Epoch: [81]  [100/156]  eta: 0:00:43  lr: 0.007081  min_lr: 0.007081  loss: 3.5234 (3.4124)  weight_decay: 0.0500 (0.0500)  time: 0.6930  data: 0.0004  max mem: 57114
Epoch: [81]  [110/156]  eta: 0:00:35  lr: 0.007079  min_lr: 0.007079  loss: 3.1161 (3.3987)  weight_decay: 0.0500 (0.0500)  time: 0.7012  data: 0.0004  max mem: 57114
Epoch: [81]  [120/156]  eta: 0:00:27  lr: 0.007077  min_lr: 0.007077  loss: 3.2918 (3.3983)  weight_decay: 0.0500 (0.0500)  time: 0.7122  data: 0.0004  max mem: 57114
Epoch: [81]  [130/156]  eta: 0:00:19  lr: 0.007075  min_lr: 0.007075  loss: 3.4690 (3.4021)  weight_decay: 0.0500 (0.0500)  time: 0.7083  data: 0.0009  max mem: 57114
Epoch: [81]  [140/156]  eta: 0:00:12  lr: 0.007074  min_lr: 0.007074  loss: 3.3892 (3.3956)  weight_decay: 0.0500 (0.0500)  time: 0.6900  data: 0.0007  max mem: 57114
Epoch: [81]  [150/156]  eta: 0:00:04  lr: 0.007072  min_lr: 0.007072  loss: 3.1742 (3.3850)  weight_decay: 0.0500 (0.0500)  time: 0.6829  data: 0.0001  max mem: 57114
Epoch: [81]  [155/156]  eta: 0:00:00  lr: 0.007071  min_lr: 0.007071  loss: 3.1742 (3.3825)  weight_decay: 0.0500 (0.0500)  time: 0.6820  data: 0.0001  max mem: 57114
Epoch: [81] Total time: 0:01:56 (0.7456 s / it)
Averaged stats: lr: 0.007071  min_lr: 0.007071  loss: 3.1742 (3.3672)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0042 (1.0042)  acc1: 82.2917 (82.2917)  acc5: 94.2708 (94.2708)  time: 2.0653  data: 1.8084  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0794 (1.1177)  acc1: 76.0417 (74.5524)  acc5: 94.2708 (92.1995)  time: 0.5637  data: 0.3618  max mem: 57114
Test: Total time: 0:00:02 (0.5847 s / it)
* Acc@1 74.345 Acc@5 92.493 loss 1.217
Accuracy of the model on the 50000 test images: 74.3%
Max accuracy: 75.37%
Test:  [0/5]  eta: 0:00:11  loss: 8.0114 (8.0114)  acc1: 0.0000 (0.0000)  acc5: 1.5625 (1.5625)  time: 2.3537  data: 2.1102  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.8060 (7.7903)  acc1: 0.0000 (0.1279)  acc5: 0.5208 (0.8951)  time: 0.6215  data: 0.4222  max mem: 57114
Test: Total time: 0:00:03 (0.6327 s / it)
* Acc@1 0.112 Acc@5 0.697 loss 7.730
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [82]  [  0/156]  eta: 0:13:07  lr: 0.007071  min_lr: 0.007071  loss: 2.6856 (2.6856)  weight_decay: 0.0500 (0.0500)  time: 5.0452  data: 4.0070  max mem: 57114
Epoch: [82]  [ 10/156]  eta: 0:02:43  lr: 0.007069  min_lr: 0.007069  loss: 2.7322 (3.1164)  weight_decay: 0.0500 (0.0500)  time: 1.1201  data: 0.3645  max mem: 57114
Epoch: [82]  [ 20/156]  eta: 0:02:06  lr: 0.007067  min_lr: 0.007067  loss: 3.4465 (3.3029)  weight_decay: 0.0500 (0.0500)  time: 0.7220  data: 0.0003  max mem: 57114
Epoch: [82]  [ 30/156]  eta: 0:01:48  lr: 0.007065  min_lr: 0.007065  loss: 3.5815 (3.3186)  weight_decay: 0.0500 (0.0500)  time: 0.7230  data: 0.0003  max mem: 57114
Epoch: [82]  [ 40/156]  eta: 0:01:36  lr: 0.007063  min_lr: 0.007063  loss: 3.5654 (3.3608)  weight_decay: 0.0500 (0.0500)  time: 0.7279  data: 0.0004  max mem: 57114
Epoch: [82]  [ 50/156]  eta: 0:01:25  lr: 0.007061  min_lr: 0.007061  loss: 3.3994 (3.3237)  weight_decay: 0.0500 (0.0500)  time: 0.7256  data: 0.0004  max mem: 57114
Epoch: [82]  [ 60/156]  eta: 0:01:16  lr: 0.007060  min_lr: 0.007060  loss: 3.3994 (3.3608)  weight_decay: 0.0500 (0.0500)  time: 0.7386  data: 0.0004  max mem: 57114
Epoch: [82]  [ 70/156]  eta: 0:01:07  lr: 0.007058  min_lr: 0.007058  loss: 3.2725 (3.3043)  weight_decay: 0.0500 (0.0500)  time: 0.7399  data: 0.0004  max mem: 57114
Epoch: [82]  [ 80/156]  eta: 0:00:59  lr: 0.007056  min_lr: 0.007056  loss: 3.4520 (3.3390)  weight_decay: 0.0500 (0.0500)  time: 0.7184  data: 0.0004  max mem: 57114
Epoch: [82]  [ 90/156]  eta: 0:00:50  lr: 0.007054  min_lr: 0.007054  loss: 3.5421 (3.3413)  weight_decay: 0.0500 (0.0500)  time: 0.7043  data: 0.0004  max mem: 57114
Epoch: [82]  [100/156]  eta: 0:00:42  lr: 0.007052  min_lr: 0.007052  loss: 3.4557 (3.3410)  weight_decay: 0.0500 (0.0500)  time: 0.6984  data: 0.0004  max mem: 57114
Epoch: [82]  [110/156]  eta: 0:00:34  lr: 0.007050  min_lr: 0.007050  loss: 3.5948 (3.3595)  weight_decay: 0.0500 (0.0500)  time: 0.7024  data: 0.0004  max mem: 57114
Epoch: [82]  [120/156]  eta: 0:00:27  lr: 0.007048  min_lr: 0.007048  loss: 3.6473 (3.3694)  weight_decay: 0.0500 (0.0500)  time: 0.7025  data: 0.0004  max mem: 57114
Epoch: [82]  [130/156]  eta: 0:00:19  lr: 0.007047  min_lr: 0.007047  loss: 3.5946 (3.3882)  weight_decay: 0.0500 (0.0500)  time: 0.6948  data: 0.0007  max mem: 57114
Epoch: [82]  [140/156]  eta: 0:00:11  lr: 0.007045  min_lr: 0.007045  loss: 3.4322 (3.3685)  weight_decay: 0.0500 (0.0500)  time: 0.6879  data: 0.0006  max mem: 57114
Epoch: [82]  [150/156]  eta: 0:00:04  lr: 0.007043  min_lr: 0.007043  loss: 3.3761 (3.3743)  weight_decay: 0.0500 (0.0500)  time: 0.6866  data: 0.0002  max mem: 57114
Epoch: [82]  [155/156]  eta: 0:00:00  lr: 0.007042  min_lr: 0.007042  loss: 3.1912 (3.3736)  weight_decay: 0.0500 (0.0500)  time: 0.6884  data: 0.0002  max mem: 57114
Epoch: [82] Total time: 0:01:55 (0.7412 s / it)
Averaged stats: lr: 0.007042  min_lr: 0.007042  loss: 3.1912 (3.3668)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0460 (1.0460)  acc1: 82.8125 (82.8125)  acc5: 95.8333 (95.8333)  time: 2.1316  data: 1.8758  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1457 (1.1111)  acc1: 76.0417 (76.5985)  acc5: 94.7917 (92.5831)  time: 0.5770  data: 0.3752  max mem: 57114
Test: Total time: 0:00:02 (0.5995 s / it)
* Acc@1 75.330 Acc@5 92.815 loss 1.209
Accuracy of the model on the 50000 test images: 75.3%
Max accuracy: 75.37%
Test:  [0/5]  eta: 0:00:11  loss: 8.0714 (8.0714)  acc1: 0.0000 (0.0000)  acc5: 1.5625 (1.5625)  time: 2.3253  data: 2.0819  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.8823 (7.8583)  acc1: 0.0000 (0.1279)  acc5: 1.0417 (1.0230)  time: 0.6157  data: 0.4165  max mem: 57114
Test: Total time: 0:00:03 (0.6385 s / it)
* Acc@1 0.112 Acc@5 0.713 loss 7.793
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [83]  [  0/156]  eta: 0:13:06  lr: 0.007042  min_lr: 0.007042  loss: 2.1852 (2.1852)  weight_decay: 0.0500 (0.0500)  time: 5.0427  data: 3.2440  max mem: 57114
Epoch: [83]  [ 10/156]  eta: 0:02:43  lr: 0.007040  min_lr: 0.007040  loss: 3.4595 (3.3855)  weight_decay: 0.0500 (0.0500)  time: 1.1195  data: 0.2952  max mem: 57114
Epoch: [83]  [ 20/156]  eta: 0:02:06  lr: 0.007038  min_lr: 0.007038  loss: 3.5536 (3.4473)  weight_decay: 0.0500 (0.0500)  time: 0.7265  data: 0.0003  max mem: 57114
Epoch: [83]  [ 30/156]  eta: 0:01:49  lr: 0.007036  min_lr: 0.007036  loss: 3.4404 (3.3892)  weight_decay: 0.0500 (0.0500)  time: 0.7278  data: 0.0003  max mem: 57114
Epoch: [83]  [ 40/156]  eta: 0:01:36  lr: 0.007034  min_lr: 0.007034  loss: 3.2832 (3.3064)  weight_decay: 0.0500 (0.0500)  time: 0.7264  data: 0.0003  max mem: 57114
Epoch: [83]  [ 50/156]  eta: 0:01:25  lr: 0.007032  min_lr: 0.007032  loss: 3.2443 (3.2997)  weight_decay: 0.0500 (0.0500)  time: 0.7161  data: 0.0004  max mem: 57114
Epoch: [83]  [ 60/156]  eta: 0:01:16  lr: 0.007031  min_lr: 0.007031  loss: 3.3920 (3.3275)  weight_decay: 0.0500 (0.0500)  time: 0.7150  data: 0.0004  max mem: 57114
Epoch: [83]  [ 70/156]  eta: 0:01:07  lr: 0.007029  min_lr: 0.007029  loss: 3.7123 (3.3743)  weight_decay: 0.0500 (0.0500)  time: 0.7229  data: 0.0004  max mem: 57114
Epoch: [83]  [ 80/156]  eta: 0:00:58  lr: 0.007027  min_lr: 0.007027  loss: 3.6018 (3.3990)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0004  max mem: 57114
Epoch: [83]  [ 90/156]  eta: 0:00:50  lr: 0.007025  min_lr: 0.007025  loss: 3.5460 (3.3911)  weight_decay: 0.0500 (0.0500)  time: 0.7034  data: 0.0004  max mem: 57114
Epoch: [83]  [100/156]  eta: 0:00:42  lr: 0.007023  min_lr: 0.007023  loss: 3.6066 (3.4049)  weight_decay: 0.0500 (0.0500)  time: 0.7010  data: 0.0004  max mem: 57114
Epoch: [83]  [110/156]  eta: 0:00:34  lr: 0.007021  min_lr: 0.007021  loss: 3.5928 (3.4025)  weight_decay: 0.0500 (0.0500)  time: 0.7085  data: 0.0004  max mem: 57114
Epoch: [83]  [120/156]  eta: 0:00:27  lr: 0.007019  min_lr: 0.007019  loss: 3.2334 (3.3863)  weight_decay: 0.0500 (0.0500)  time: 0.7104  data: 0.0004  max mem: 57114
Epoch: [83]  [130/156]  eta: 0:00:19  lr: 0.007017  min_lr: 0.007017  loss: 3.0567 (3.3600)  weight_decay: 0.0500 (0.0500)  time: 0.6916  data: 0.0008  max mem: 57114
Epoch: [83]  [140/156]  eta: 0:00:11  lr: 0.007015  min_lr: 0.007015  loss: 3.3487 (3.3592)  weight_decay: 0.0500 (0.0500)  time: 0.6854  data: 0.0007  max mem: 57114
Epoch: [83]  [150/156]  eta: 0:00:04  lr: 0.007014  min_lr: 0.007014  loss: 3.5513 (3.3637)  weight_decay: 0.0500 (0.0500)  time: 0.6883  data: 0.0001  max mem: 57114
Epoch: [83]  [155/156]  eta: 0:00:00  lr: 0.007013  min_lr: 0.007013  loss: 3.5772 (3.3810)  weight_decay: 0.0500 (0.0500)  time: 0.6890  data: 0.0001  max mem: 57114
Epoch: [83] Total time: 0:01:55 (0.7394 s / it)
Averaged stats: lr: 0.007013  min_lr: 0.007013  loss: 3.5772 (3.3659)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.1175 (1.1175)  acc1: 79.1667 (79.1667)  acc5: 97.9167 (97.9167)  time: 2.0922  data: 1.8363  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1990 (1.2427)  acc1: 76.5625 (73.7852)  acc5: 95.3125 (93.0946)  time: 0.5691  data: 0.3673  max mem: 57114
Test: Total time: 0:00:02 (0.5885 s / it)
* Acc@1 74.411 Acc@5 92.525 loss 1.327
Accuracy of the model on the 50000 test images: 74.4%
Max accuracy: 75.37%
Test:  [0/5]  eta: 0:00:11  loss: 8.1309 (8.1309)  acc1: 0.5208 (0.5208)  acc5: 1.5625 (1.5625)  time: 2.2657  data: 2.0222  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.9607 (7.9318)  acc1: 0.0000 (0.2558)  acc5: 1.0417 (1.0230)  time: 0.6038  data: 0.4045  max mem: 57114
Test: Total time: 0:00:03 (0.6178 s / it)
* Acc@1 0.118 Acc@5 0.733 loss 7.860
Accuracy of the model EMA on 50000 test images: 0.1%
Epoch: [84]  [  0/156]  eta: 0:12:27  lr: 0.007012  min_lr: 0.007012  loss: 3.7746 (3.7746)  weight_decay: 0.0500 (0.0500)  time: 4.7929  data: 3.5359  max mem: 57114
Epoch: [84]  [ 10/156]  eta: 0:02:40  lr: 0.007011  min_lr: 0.007011  loss: 3.1114 (3.2055)  weight_decay: 0.0500 (0.0500)  time: 1.0999  data: 0.3217  max mem: 57114
Epoch: [84]  [ 20/156]  eta: 0:02:05  lr: 0.007009  min_lr: 0.007009  loss: 3.1369 (3.2918)  weight_decay: 0.0500 (0.0500)  time: 0.7263  data: 0.0004  max mem: 57114
Epoch: [84]  [ 30/156]  eta: 0:01:48  lr: 0.007007  min_lr: 0.007007  loss: 3.3633 (3.2694)  weight_decay: 0.0500 (0.0500)  time: 0.7306  data: 0.0004  max mem: 57114
Epoch: [84]  [ 40/156]  eta: 0:01:35  lr: 0.007005  min_lr: 0.007005  loss: 3.3633 (3.2816)  weight_decay: 0.0500 (0.0500)  time: 0.7214  data: 0.0004  max mem: 57114
Epoch: [84]  [ 50/156]  eta: 0:01:24  lr: 0.007003  min_lr: 0.007003  loss: 3.5053 (3.3529)  weight_decay: 0.0500 (0.0500)  time: 0.7091  data: 0.0004  max mem: 57114
Epoch: [84]  [ 60/156]  eta: 0:01:16  lr: 0.007001  min_lr: 0.007001  loss: 3.4007 (3.3087)  weight_decay: 0.0500 (0.0500)  time: 0.7319  data: 0.0004  max mem: 57114
Epoch: [84]  [ 70/156]  eta: 0:01:07  lr: 0.006999  min_lr: 0.006999  loss: 3.1912 (3.3026)  weight_decay: 0.0500 (0.0500)  time: 0.7217  data: 0.0003  max mem: 57114
Epoch: [84]  [ 80/156]  eta: 0:00:58  lr: 0.006997  min_lr: 0.006997  loss: 3.4608 (3.3374)  weight_decay: 0.0500 (0.0500)  time: 0.7082  data: 0.0003  max mem: 57114
Epoch: [84]  [ 90/156]  eta: 0:00:50  lr: 0.006995  min_lr: 0.006995  loss: 3.2927 (3.3088)  weight_decay: 0.0500 (0.0500)  time: 0.7204  data: 0.0003  max mem: 57114
Epoch: [84]  [100/156]  eta: 0:00:42  lr: 0.006993  min_lr: 0.006993  loss: 3.2927 (3.3135)  weight_decay: 0.0500 (0.0500)  time: 0.7168  data: 0.0004  max mem: 57114
Epoch: [84]  [110/156]  eta: 0:00:34  lr: 0.006991  min_lr: 0.006991  loss: 3.6858 (3.3520)  weight_decay: 0.0500 (0.0500)  time: 0.6987  data: 0.0003  max mem: 57114
Epoch: [84]  [120/156]  eta: 0:00:27  lr: 0.006990  min_lr: 0.006990  loss: 3.5720 (3.3587)  weight_decay: 0.0500 (0.0500)  time: 0.7007  data: 0.0004  max mem: 57114
Epoch: [84]  [130/156]  eta: 0:00:19  lr: 0.006988  min_lr: 0.006988  loss: 3.4017 (3.3587)  weight_decay: 0.0500 (0.0500)  time: 0.7057  data: 0.0009  max mem: 57114
Epoch: [84]  [140/156]  eta: 0:00:11  lr: 0.006986  min_lr: 0.006986  loss: 3.3183 (3.3364)  weight_decay: 0.0500 (0.0500)  time: 0.6901  data: 0.0008  max mem: 57114
Epoch: [84]  [150/156]  eta: 0:00:04  lr: 0.006984  min_lr: 0.006984  loss: 3.3605 (3.3375)  weight_decay: 0.0500 (0.0500)  time: 0.6903  data: 0.0001  max mem: 57114
Epoch: [84]  [155/156]  eta: 0:00:00  lr: 0.006983  min_lr: 0.006983  loss: 3.3159 (3.3327)  weight_decay: 0.0500 (0.0500)  time: 0.6914  data: 0.0001  max mem: 57114
Epoch: [84] Total time: 0:01:55 (0.7398 s / it)
Averaged stats: lr: 0.006983  min_lr: 0.006983  loss: 3.3159 (3.3523)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 0.9403 (0.9403)  acc1: 80.2083 (80.2083)  acc5: 96.3542 (96.3542)  time: 2.2108  data: 1.9555  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1187 (1.0873)  acc1: 77.6042 (75.1918)  acc5: 95.3125 (92.4552)  time: 0.5928  data: 0.3912  max mem: 57114
Test: Total time: 0:00:03 (0.6168 s / it)
* Acc@1 75.278 Acc@5 92.871 loss 1.167
Accuracy of the model on the 50000 test images: 75.3%
Max accuracy: 75.37%
Test:  [0/5]  eta: 0:00:11  loss: 8.2027 (8.2027)  acc1: 0.5208 (0.5208)  acc5: 1.5625 (1.5625)  time: 2.3114  data: 2.0679  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.0499 (8.0133)  acc1: 0.0000 (0.2558)  acc5: 1.0417 (1.1509)  time: 0.6129  data: 0.4137  max mem: 57114
Test: Total time: 0:00:03 (0.6227 s / it)
* Acc@1 0.134 Acc@5 0.769 loss 7.935
Accuracy of the model EMA on 50000 test images: 0.1%
Max EMA accuracy: 0.13%
Epoch: [85]  [  0/156]  eta: 0:07:02  lr: 0.006983  min_lr: 0.006983  loss: 3.3636 (3.3636)  weight_decay: 0.0500 (0.0500)  time: 2.7111  data: 2.0514  max mem: 57114
Epoch: [85]  [ 10/156]  eta: 0:02:13  lr: 0.006981  min_lr: 0.006981  loss: 3.3636 (3.2338)  weight_decay: 0.0500 (0.0500)  time: 0.9155  data: 0.1868  max mem: 57114
Epoch: [85]  [ 20/156]  eta: 0:01:50  lr: 0.006979  min_lr: 0.006979  loss: 3.1278 (3.2392)  weight_decay: 0.0500 (0.0500)  time: 0.7179  data: 0.0004  max mem: 57114
Epoch: [85]  [ 30/156]  eta: 0:01:39  lr: 0.006977  min_lr: 0.006977  loss: 3.3016 (3.3167)  weight_decay: 0.0500 (0.0500)  time: 0.7182  data: 0.0004  max mem: 57114
Epoch: [85]  [ 40/156]  eta: 0:01:30  lr: 0.006975  min_lr: 0.006975  loss: 3.3016 (3.2994)  weight_decay: 0.0500 (0.0500)  time: 0.7393  data: 0.0004  max mem: 57114
Epoch: [85]  [ 50/156]  eta: 0:01:21  lr: 0.006973  min_lr: 0.006973  loss: 3.1313 (3.2013)  weight_decay: 0.0500 (0.0500)  time: 0.7380  data: 0.0004  max mem: 57114
Epoch: [85]  [ 60/156]  eta: 0:01:12  lr: 0.006971  min_lr: 0.006971  loss: 2.9835 (3.2125)  weight_decay: 0.0500 (0.0500)  time: 0.7252  data: 0.0004  max mem: 57114
Epoch: [85]  [ 70/156]  eta: 0:01:04  lr: 0.006969  min_lr: 0.006969  loss: 3.3656 (3.2391)  weight_decay: 0.0500 (0.0500)  time: 0.7158  data: 0.0004  max mem: 57114
Epoch: [85]  [ 80/156]  eta: 0:00:56  lr: 0.006967  min_lr: 0.006967  loss: 3.4139 (3.2481)  weight_decay: 0.0500 (0.0500)  time: 0.7107  data: 0.0004  max mem: 57114
Epoch: [85]  [ 90/156]  eta: 0:00:48  lr: 0.006965  min_lr: 0.006965  loss: 3.4337 (3.2671)  weight_decay: 0.0500 (0.0500)  time: 0.7010  data: 0.0004  max mem: 57114
Epoch: [85]  [100/156]  eta: 0:00:41  lr: 0.006963  min_lr: 0.006963  loss: 3.4951 (3.2749)  weight_decay: 0.0500 (0.0500)  time: 0.6961  data: 0.0003  max mem: 57114
Epoch: [85]  [110/156]  eta: 0:00:33  lr: 0.006962  min_lr: 0.006962  loss: 3.6256 (3.3047)  weight_decay: 0.0500 (0.0500)  time: 0.7062  data: 0.0004  max mem: 57114
Epoch: [85]  [120/156]  eta: 0:00:26  lr: 0.006960  min_lr: 0.006960  loss: 3.7180 (3.3326)  weight_decay: 0.0500 (0.0500)  time: 0.7062  data: 0.0004  max mem: 57114
Epoch: [85]  [130/156]  eta: 0:00:18  lr: 0.006958  min_lr: 0.006958  loss: 3.3959 (3.3325)  weight_decay: 0.0500 (0.0500)  time: 0.6971  data: 0.0008  max mem: 57114
Epoch: [85]  [140/156]  eta: 0:00:11  lr: 0.006956  min_lr: 0.006956  loss: 3.2210 (3.3024)  weight_decay: 0.0500 (0.0500)  time: 0.6887  data: 0.0007  max mem: 57114
Epoch: [85]  [150/156]  eta: 0:00:04  lr: 0.006954  min_lr: 0.006954  loss: 2.5802 (3.2696)  weight_decay: 0.0500 (0.0500)  time: 0.6811  data: 0.0001  max mem: 57114
Epoch: [85]  [155/156]  eta: 0:00:00  lr: 0.006953  min_lr: 0.006953  loss: 2.9878 (3.2815)  weight_decay: 0.0500 (0.0500)  time: 0.6839  data: 0.0001  max mem: 57114
Epoch: [85] Total time: 0:01:52 (0.7243 s / it)
Averaged stats: lr: 0.006953  min_lr: 0.006953  loss: 2.9878 (3.3520)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8885 (0.8885)  acc1: 82.2917 (82.2917)  acc5: 97.3958 (97.3958)  time: 2.0351  data: 1.7794  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0229 (1.0037)  acc1: 75.5208 (75.7033)  acc5: 95.8333 (94.2455)  time: 0.5577  data: 0.3559  max mem: 57114
Test: Total time: 0:00:02 (0.5797 s / it)
* Acc@1 75.284 Acc@5 93.197 loss 1.107
Accuracy of the model on the 50000 test images: 75.3%
Max accuracy: 75.37%
Test:  [0/5]  eta: 0:00:11  loss: 8.2771 (8.2771)  acc1: 0.5208 (0.5208)  acc5: 1.5625 (1.5625)  time: 2.3281  data: 2.0847  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.1433 (8.0974)  acc1: 0.0000 (0.2558)  acc5: 1.0417 (1.1509)  time: 0.6162  data: 0.4170  max mem: 57114
Test: Total time: 0:00:03 (0.6294 s / it)
* Acc@1 0.148 Acc@5 0.799 loss 8.012
Accuracy of the model EMA on 50000 test images: 0.1%
Max EMA accuracy: 0.15%
Epoch: [86]  [  0/156]  eta: 0:06:09  lr: 0.006953  min_lr: 0.006953  loss: 3.6072 (3.6072)  weight_decay: 0.0500 (0.0500)  time: 2.3692  data: 1.7097  max mem: 57114
Epoch: [86]  [ 10/156]  eta: 0:02:05  lr: 0.006951  min_lr: 0.006951  loss: 3.3059 (3.2104)  weight_decay: 0.0500 (0.0500)  time: 0.8602  data: 0.1557  max mem: 57114
Epoch: [86]  [ 20/156]  eta: 0:01:48  lr: 0.006949  min_lr: 0.006949  loss: 3.4895 (3.4170)  weight_decay: 0.0500 (0.0500)  time: 0.7156  data: 0.0004  max mem: 57114
Epoch: [86]  [ 30/156]  eta: 0:01:37  lr: 0.006947  min_lr: 0.006947  loss: 3.4895 (3.4356)  weight_decay: 0.0500 (0.0500)  time: 0.7267  data: 0.0004  max mem: 57114
Epoch: [86]  [ 40/156]  eta: 0:01:27  lr: 0.006945  min_lr: 0.006945  loss: 3.3060 (3.4031)  weight_decay: 0.0500 (0.0500)  time: 0.7181  data: 0.0004  max mem: 57114
Epoch: [86]  [ 50/156]  eta: 0:01:19  lr: 0.006943  min_lr: 0.006943  loss: 3.3393 (3.4147)  weight_decay: 0.0500 (0.0500)  time: 0.7125  data: 0.0004  max mem: 57114
Epoch: [86]  [ 60/156]  eta: 0:01:11  lr: 0.006941  min_lr: 0.006941  loss: 3.3393 (3.3794)  weight_decay: 0.0500 (0.0500)  time: 0.7268  data: 0.0004  max mem: 57114
Epoch: [86]  [ 70/156]  eta: 0:01:03  lr: 0.006939  min_lr: 0.006939  loss: 3.0807 (3.3456)  weight_decay: 0.0500 (0.0500)  time: 0.7269  data: 0.0004  max mem: 57114
Epoch: [86]  [ 80/156]  eta: 0:00:56  lr: 0.006937  min_lr: 0.006937  loss: 3.3370 (3.3712)  weight_decay: 0.0500 (0.0500)  time: 0.7199  data: 0.0004  max mem: 57114
Epoch: [86]  [ 90/156]  eta: 0:00:48  lr: 0.006935  min_lr: 0.006935  loss: 3.5401 (3.3714)  weight_decay: 0.0500 (0.0500)  time: 0.7053  data: 0.0004  max mem: 57114
Epoch: [86]  [100/156]  eta: 0:00:41  lr: 0.006933  min_lr: 0.006933  loss: 3.5401 (3.3702)  weight_decay: 0.0500 (0.0500)  time: 0.7032  data: 0.0004  max mem: 57114
Epoch: [86]  [110/156]  eta: 0:00:33  lr: 0.006931  min_lr: 0.006931  loss: 3.0258 (3.3196)  weight_decay: 0.0500 (0.0500)  time: 0.7124  data: 0.0004  max mem: 57114
Epoch: [86]  [120/156]  eta: 0:00:26  lr: 0.006929  min_lr: 0.006929  loss: 3.1517 (3.3304)  weight_decay: 0.0500 (0.0500)  time: 0.7098  data: 0.0004  max mem: 57114
Epoch: [86]  [130/156]  eta: 0:00:18  lr: 0.006927  min_lr: 0.006927  loss: 3.2784 (3.3173)  weight_decay: 0.0500 (0.0500)  time: 0.7021  data: 0.0008  max mem: 57114
Epoch: [86]  [140/156]  eta: 0:00:11  lr: 0.006925  min_lr: 0.006925  loss: 3.2867 (3.3230)  weight_decay: 0.0500 (0.0500)  time: 0.6881  data: 0.0007  max mem: 57114
Epoch: [86]  [150/156]  eta: 0:00:04  lr: 0.006923  min_lr: 0.006923  loss: 3.2867 (3.3187)  weight_decay: 0.0500 (0.0500)  time: 0.6869  data: 0.0001  max mem: 57114
Epoch: [86]  [155/156]  eta: 0:00:00  lr: 0.006922  min_lr: 0.006922  loss: 3.2994 (3.3139)  weight_decay: 0.0500 (0.0500)  time: 0.6835  data: 0.0001  max mem: 57114
Epoch: [86] Total time: 0:01:52 (0.7220 s / it)
Averaged stats: lr: 0.006922  min_lr: 0.006922  loss: 3.2994 (3.3546)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9147 (0.9147)  acc1: 80.7292 (80.7292)  acc5: 95.8333 (95.8333)  time: 2.1270  data: 1.8710  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9759 (1.0272)  acc1: 79.1667 (75.4476)  acc5: 95.8333 (92.9668)  time: 0.5760  data: 0.3743  max mem: 57114
Test: Total time: 0:00:02 (0.5986 s / it)
* Acc@1 75.885 Acc@5 93.394 loss 1.092
Accuracy of the model on the 50000 test images: 75.9%
Max accuracy: 75.89%
Test:  [0/5]  eta: 0:00:10  loss: 8.3458 (8.3458)  acc1: 0.5208 (0.5208)  acc5: 1.5625 (1.5625)  time: 2.0475  data: 1.8039  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.2306 (8.1753)  acc1: 0.0000 (0.3836)  acc5: 1.0417 (1.1509)  time: 0.5601  data: 0.3609  max mem: 57114
Test: Total time: 0:00:02 (0.5700 s / it)
* Acc@1 0.156 Acc@5 0.843 loss 8.085
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.16%
Epoch: [87]  [  0/156]  eta: 0:05:52  lr: 0.006922  min_lr: 0.006922  loss: 3.6836 (3.6836)  weight_decay: 0.0500 (0.0500)  time: 2.2593  data: 1.6061  max mem: 57114
Epoch: [87]  [ 10/156]  eta: 0:02:08  lr: 0.006920  min_lr: 0.006920  loss: 3.6836 (3.4491)  weight_decay: 0.0500 (0.0500)  time: 0.8828  data: 0.1463  max mem: 57114
Epoch: [87]  [ 20/156]  eta: 0:01:52  lr: 0.006918  min_lr: 0.006918  loss: 3.2180 (3.2804)  weight_decay: 0.0500 (0.0500)  time: 0.7554  data: 0.0003  max mem: 57114
Epoch: [87]  [ 30/156]  eta: 0:01:40  lr: 0.006916  min_lr: 0.006916  loss: 3.0075 (3.2879)  weight_decay: 0.0500 (0.0500)  time: 0.7542  data: 0.0003  max mem: 57114
Epoch: [87]  [ 40/156]  eta: 0:01:31  lr: 0.006914  min_lr: 0.006914  loss: 3.1474 (3.2660)  weight_decay: 0.0500 (0.0500)  time: 0.7468  data: 0.0004  max mem: 57114
Epoch: [87]  [ 50/156]  eta: 0:01:22  lr: 0.006912  min_lr: 0.006912  loss: 3.4774 (3.2818)  weight_decay: 0.0500 (0.0500)  time: 0.7482  data: 0.0004  max mem: 57114
Epoch: [87]  [ 60/156]  eta: 0:01:13  lr: 0.006910  min_lr: 0.006910  loss: 3.6267 (3.3238)  weight_decay: 0.0500 (0.0500)  time: 0.7298  data: 0.0004  max mem: 57114
Epoch: [87]  [ 70/156]  eta: 0:01:05  lr: 0.006908  min_lr: 0.006908  loss: 3.1696 (3.2880)  weight_decay: 0.0500 (0.0500)  time: 0.7099  data: 0.0004  max mem: 57114
Epoch: [87]  [ 80/156]  eta: 0:00:57  lr: 0.006906  min_lr: 0.006906  loss: 2.9066 (3.2562)  weight_decay: 0.0500 (0.0500)  time: 0.7083  data: 0.0004  max mem: 57114
Epoch: [87]  [ 90/156]  eta: 0:00:49  lr: 0.006904  min_lr: 0.006904  loss: 3.2455 (3.2505)  weight_decay: 0.0500 (0.0500)  time: 0.7124  data: 0.0004  max mem: 57114
Epoch: [87]  [100/156]  eta: 0:00:41  lr: 0.006902  min_lr: 0.006902  loss: 3.4065 (3.2699)  weight_decay: 0.0500 (0.0500)  time: 0.7083  data: 0.0004  max mem: 57114
Epoch: [87]  [110/156]  eta: 0:00:34  lr: 0.006900  min_lr: 0.006900  loss: 3.5731 (3.2878)  weight_decay: 0.0500 (0.0500)  time: 0.7051  data: 0.0004  max mem: 57114
Epoch: [87]  [120/156]  eta: 0:00:26  lr: 0.006898  min_lr: 0.006898  loss: 3.4015 (3.2741)  weight_decay: 0.0500 (0.0500)  time: 0.7057  data: 0.0004  max mem: 57114
Epoch: [87]  [130/156]  eta: 0:00:19  lr: 0.006896  min_lr: 0.006896  loss: 3.1994 (3.2662)  weight_decay: 0.0500 (0.0500)  time: 0.6967  data: 0.0009  max mem: 57114
Epoch: [87]  [140/156]  eta: 0:00:11  lr: 0.006894  min_lr: 0.006894  loss: 3.4719 (3.2810)  weight_decay: 0.0500 (0.0500)  time: 0.6876  data: 0.0007  max mem: 57114
Epoch: [87]  [150/156]  eta: 0:00:04  lr: 0.006893  min_lr: 0.006893  loss: 3.4954 (3.2892)  weight_decay: 0.0500 (0.0500)  time: 0.6887  data: 0.0001  max mem: 57114
Epoch: [87]  [155/156]  eta: 0:00:00  lr: 0.006892  min_lr: 0.006892  loss: 3.4377 (3.2858)  weight_decay: 0.0500 (0.0500)  time: 0.6866  data: 0.0001  max mem: 57114
Epoch: [87] Total time: 0:01:53 (0.7295 s / it)
Averaged stats: lr: 0.006892  min_lr: 0.006892  loss: 3.4377 (3.3342)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8878 (0.8878)  acc1: 80.7292 (80.7292)  acc5: 97.3958 (97.3958)  time: 2.1799  data: 1.9237  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9657 (0.9726)  acc1: 80.2083 (77.1100)  acc5: 96.8750 (94.7570)  time: 0.5867  data: 0.3848  max mem: 57114
Test: Total time: 0:00:03 (0.6102 s / it)
* Acc@1 76.019 Acc@5 93.568 loss 1.083
Accuracy of the model on the 50000 test images: 76.0%
Max accuracy: 76.02%
Test:  [0/5]  eta: 0:00:09  loss: 8.4240 (8.4240)  acc1: 0.5208 (0.5208)  acc5: 1.5625 (1.5625)  time: 1.9783  data: 1.7348  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.3296 (8.2618)  acc1: 0.0000 (0.3836)  acc5: 1.0417 (1.1509)  time: 0.5462  data: 0.3470  max mem: 57114
Test: Total time: 0:00:02 (0.5572 s / it)
* Acc@1 0.168 Acc@5 0.881 loss 8.164
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.17%
Epoch: [88]  [  0/156]  eta: 0:06:51  lr: 0.006891  min_lr: 0.006891  loss: 3.5162 (3.5162)  weight_decay: 0.0500 (0.0500)  time: 2.6356  data: 1.9797  max mem: 57114
Epoch: [88]  [ 10/156]  eta: 0:02:10  lr: 0.006889  min_lr: 0.006889  loss: 3.3240 (3.3092)  weight_decay: 0.0500 (0.0500)  time: 0.8943  data: 0.1803  max mem: 57114
Epoch: [88]  [ 20/156]  eta: 0:01:50  lr: 0.006887  min_lr: 0.006887  loss: 3.4080 (3.3655)  weight_decay: 0.0500 (0.0500)  time: 0.7229  data: 0.0003  max mem: 57114
Epoch: [88]  [ 30/156]  eta: 0:01:39  lr: 0.006885  min_lr: 0.006885  loss: 3.4080 (3.3848)  weight_decay: 0.0500 (0.0500)  time: 0.7319  data: 0.0003  max mem: 57114
Epoch: [88]  [ 40/156]  eta: 0:01:30  lr: 0.006883  min_lr: 0.006883  loss: 3.3575 (3.3941)  weight_decay: 0.0500 (0.0500)  time: 0.7421  data: 0.0004  max mem: 57114
Epoch: [88]  [ 50/156]  eta: 0:01:21  lr: 0.006881  min_lr: 0.006881  loss: 3.6209 (3.4555)  weight_decay: 0.0500 (0.0500)  time: 0.7422  data: 0.0004  max mem: 57114
Epoch: [88]  [ 60/156]  eta: 0:01:14  lr: 0.006879  min_lr: 0.006879  loss: 3.7725 (3.4324)  weight_decay: 0.0500 (0.0500)  time: 0.7561  data: 0.0004  max mem: 57114
Epoch: [88]  [ 70/156]  eta: 0:01:05  lr: 0.006877  min_lr: 0.006877  loss: 3.4847 (3.4326)  weight_decay: 0.0500 (0.0500)  time: 0.7394  data: 0.0004  max mem: 57114
Epoch: [88]  [ 80/156]  eta: 0:00:57  lr: 0.006875  min_lr: 0.006875  loss: 3.1978 (3.3915)  weight_decay: 0.0500 (0.0500)  time: 0.7087  data: 0.0004  max mem: 57114
Epoch: [88]  [ 90/156]  eta: 0:00:49  lr: 0.006873  min_lr: 0.006873  loss: 3.2129 (3.3788)  weight_decay: 0.0500 (0.0500)  time: 0.7102  data: 0.0004  max mem: 57114
Epoch: [88]  [100/156]  eta: 0:00:41  lr: 0.006871  min_lr: 0.006871  loss: 3.4032 (3.3684)  weight_decay: 0.0500 (0.0500)  time: 0.7034  data: 0.0004  max mem: 57114
Epoch: [88]  [110/156]  eta: 0:00:34  lr: 0.006869  min_lr: 0.006869  loss: 3.4032 (3.3584)  weight_decay: 0.0500 (0.0500)  time: 0.6983  data: 0.0004  max mem: 57114
Epoch: [88]  [120/156]  eta: 0:00:26  lr: 0.006867  min_lr: 0.006867  loss: 3.4302 (3.3593)  weight_decay: 0.0500 (0.0500)  time: 0.7047  data: 0.0004  max mem: 57114
Epoch: [88]  [130/156]  eta: 0:00:19  lr: 0.006865  min_lr: 0.006865  loss: 3.5730 (3.3799)  weight_decay: 0.0500 (0.0500)  time: 0.7098  data: 0.0005  max mem: 57114
Epoch: [88]  [140/156]  eta: 0:00:11  lr: 0.006863  min_lr: 0.006863  loss: 3.5294 (3.3752)  weight_decay: 0.0500 (0.0500)  time: 0.6932  data: 0.0004  max mem: 57114
Epoch: [88]  [150/156]  eta: 0:00:04  lr: 0.006861  min_lr: 0.006861  loss: 3.4107 (3.3731)  weight_decay: 0.0500 (0.0500)  time: 0.6812  data: 0.0001  max mem: 57114
Epoch: [88]  [155/156]  eta: 0:00:00  lr: 0.006860  min_lr: 0.006860  loss: 3.4107 (3.3644)  weight_decay: 0.0500 (0.0500)  time: 0.6857  data: 0.0001  max mem: 57114
Epoch: [88] Total time: 0:01:53 (0.7300 s / it)
Averaged stats: lr: 0.006860  min_lr: 0.006860  loss: 3.4107 (3.3474)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9475 (0.9475)  acc1: 79.6875 (79.6875)  acc5: 96.3542 (96.3542)  time: 2.0609  data: 1.8052  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0785 (1.0500)  acc1: 77.0833 (74.5524)  acc5: 93.7500 (93.3504)  time: 0.5629  data: 0.3611  max mem: 57114
Test: Total time: 0:00:02 (0.5837 s / it)
* Acc@1 74.257 Acc@5 92.445 loss 1.163
Accuracy of the model on the 50000 test images: 74.3%
Max accuracy: 76.02%
Test:  [0/5]  eta: 0:00:11  loss: 8.4950 (8.4950)  acc1: 1.0417 (1.0417)  acc5: 1.5625 (1.5625)  time: 2.3362  data: 2.0922  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.4219 (8.3371)  acc1: 0.0000 (0.5115)  acc5: 1.5625 (1.2788)  time: 0.6181  data: 0.4185  max mem: 57114
Test: Total time: 0:00:03 (0.6314 s / it)
* Acc@1 0.172 Acc@5 0.909 loss 8.235
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.17%
Epoch: [89]  [  0/156]  eta: 0:08:22  lr: 0.006860  min_lr: 0.006860  loss: 2.4674 (2.4674)  weight_decay: 0.0500 (0.0500)  time: 3.2204  data: 2.5712  max mem: 57114
Epoch: [89]  [ 10/156]  eta: 0:02:17  lr: 0.006858  min_lr: 0.006858  loss: 3.5643 (3.4154)  weight_decay: 0.0500 (0.0500)  time: 0.9388  data: 0.2340  max mem: 57114
Epoch: [89]  [ 20/156]  eta: 0:01:52  lr: 0.006856  min_lr: 0.006856  loss: 3.4821 (3.3844)  weight_decay: 0.0500 (0.0500)  time: 0.7109  data: 0.0004  max mem: 57114
Epoch: [89]  [ 30/156]  eta: 0:01:41  lr: 0.006854  min_lr: 0.006854  loss: 3.4252 (3.4163)  weight_decay: 0.0500 (0.0500)  time: 0.7301  data: 0.0004  max mem: 57114
Epoch: [89]  [ 40/156]  eta: 0:01:31  lr: 0.006852  min_lr: 0.006852  loss: 3.3995 (3.3269)  weight_decay: 0.0500 (0.0500)  time: 0.7420  data: 0.0004  max mem: 57114
Epoch: [89]  [ 50/156]  eta: 0:01:22  lr: 0.006850  min_lr: 0.006850  loss: 3.0389 (3.3170)  weight_decay: 0.0500 (0.0500)  time: 0.7329  data: 0.0004  max mem: 57114
Epoch: [89]  [ 60/156]  eta: 0:01:13  lr: 0.006848  min_lr: 0.006848  loss: 3.3878 (3.3320)  weight_decay: 0.0500 (0.0500)  time: 0.7219  data: 0.0004  max mem: 57114
Epoch: [89]  [ 70/156]  eta: 0:01:05  lr: 0.006846  min_lr: 0.006846  loss: 3.4037 (3.3065)  weight_decay: 0.0500 (0.0500)  time: 0.7149  data: 0.0004  max mem: 57114
Epoch: [89]  [ 80/156]  eta: 0:00:57  lr: 0.006844  min_lr: 0.006844  loss: 3.1794 (3.2633)  weight_decay: 0.0500 (0.0500)  time: 0.7075  data: 0.0004  max mem: 57114
Epoch: [89]  [ 90/156]  eta: 0:00:49  lr: 0.006842  min_lr: 0.006842  loss: 3.3455 (3.2955)  weight_decay: 0.0500 (0.0500)  time: 0.7020  data: 0.0005  max mem: 57114
Epoch: [89]  [100/156]  eta: 0:00:41  lr: 0.006840  min_lr: 0.006840  loss: 3.5809 (3.3015)  weight_decay: 0.0500 (0.0500)  time: 0.7034  data: 0.0004  max mem: 57114
Epoch: [89]  [110/156]  eta: 0:00:33  lr: 0.006838  min_lr: 0.006838  loss: 3.5809 (3.3288)  weight_decay: 0.0500 (0.0500)  time: 0.7022  data: 0.0003  max mem: 57114
Epoch: [89]  [120/156]  eta: 0:00:26  lr: 0.006836  min_lr: 0.006836  loss: 3.5982 (3.3520)  weight_decay: 0.0500 (0.0500)  time: 0.7083  data: 0.0003  max mem: 57114
Epoch: [89]  [130/156]  eta: 0:00:19  lr: 0.006834  min_lr: 0.006834  loss: 3.4805 (3.3483)  weight_decay: 0.0500 (0.0500)  time: 0.7016  data: 0.0008  max mem: 57114
Epoch: [89]  [140/156]  eta: 0:00:11  lr: 0.006832  min_lr: 0.006832  loss: 3.4603 (3.3626)  weight_decay: 0.0500 (0.0500)  time: 0.6893  data: 0.0007  max mem: 57114
Epoch: [89]  [150/156]  eta: 0:00:04  lr: 0.006830  min_lr: 0.006830  loss: 3.4572 (3.3645)  weight_decay: 0.0500 (0.0500)  time: 0.6892  data: 0.0001  max mem: 57114
Epoch: [89]  [155/156]  eta: 0:00:00  lr: 0.006829  min_lr: 0.006829  loss: 3.5007 (3.3688)  weight_decay: 0.0500 (0.0500)  time: 0.6871  data: 0.0001  max mem: 57114
Epoch: [89] Total time: 0:01:53 (0.7275 s / it)
Averaged stats: lr: 0.006829  min_lr: 0.006829  loss: 3.5007 (3.3418)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0958 (1.0958)  acc1: 81.2500 (81.2500)  acc5: 95.3125 (95.3125)  time: 2.0293  data: 1.7735  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2014 (1.1774)  acc1: 78.1250 (76.3427)  acc5: 95.3125 (93.3504)  time: 0.5565  data: 0.3548  max mem: 57114
Test: Total time: 0:00:02 (0.5811 s / it)
* Acc@1 75.657 Acc@5 93.187 loss 1.258
Accuracy of the model on the 50000 test images: 75.7%
Max accuracy: 76.02%
Test:  [0/5]  eta: 0:00:11  loss: 8.5391 (8.5391)  acc1: 1.0417 (1.0417)  acc5: 2.0833 (2.0833)  time: 2.3254  data: 2.0820  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.4888 (8.3821)  acc1: 0.5208 (0.6394)  acc5: 1.5625 (1.2788)  time: 0.6158  data: 0.4165  max mem: 57114
Test: Total time: 0:00:03 (0.6268 s / it)
* Acc@1 0.188 Acc@5 0.951 loss 8.279
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.19%
Epoch: [90]  [  0/156]  eta: 0:05:45  lr: 0.006829  min_lr: 0.006829  loss: 2.4796 (2.4796)  weight_decay: 0.0500 (0.0500)  time: 2.2136  data: 1.5578  max mem: 57114
Epoch: [90]  [ 10/156]  eta: 0:02:07  lr: 0.006827  min_lr: 0.006827  loss: 3.4346 (3.3406)  weight_decay: 0.0500 (0.0500)  time: 0.8740  data: 0.1419  max mem: 57114
Epoch: [90]  [ 20/156]  eta: 0:01:47  lr: 0.006825  min_lr: 0.006825  loss: 3.3110 (3.2883)  weight_decay: 0.0500 (0.0500)  time: 0.7214  data: 0.0003  max mem: 57114
Epoch: [90]  [ 30/156]  eta: 0:01:37  lr: 0.006822  min_lr: 0.006822  loss: 3.3902 (3.3772)  weight_decay: 0.0500 (0.0500)  time: 0.7145  data: 0.0004  max mem: 57114
Epoch: [90]  [ 40/156]  eta: 0:01:28  lr: 0.006820  min_lr: 0.006820  loss: 3.5559 (3.4115)  weight_decay: 0.0500 (0.0500)  time: 0.7276  data: 0.0004  max mem: 57114
Epoch: [90]  [ 50/156]  eta: 0:01:20  lr: 0.006818  min_lr: 0.006818  loss: 3.5393 (3.4500)  weight_decay: 0.0500 (0.0500)  time: 0.7374  data: 0.0004  max mem: 57114
Epoch: [90]  [ 60/156]  eta: 0:01:12  lr: 0.006816  min_lr: 0.006816  loss: 3.5393 (3.4386)  weight_decay: 0.0500 (0.0500)  time: 0.7319  data: 0.0004  max mem: 57114
Epoch: [90]  [ 70/156]  eta: 0:01:04  lr: 0.006814  min_lr: 0.006814  loss: 3.2070 (3.4107)  weight_decay: 0.0500 (0.0500)  time: 0.7142  data: 0.0005  max mem: 57114
Epoch: [90]  [ 80/156]  eta: 0:00:56  lr: 0.006812  min_lr: 0.006812  loss: 3.1956 (3.3858)  weight_decay: 0.0500 (0.0500)  time: 0.7063  data: 0.0004  max mem: 57114
Epoch: [90]  [ 90/156]  eta: 0:00:48  lr: 0.006810  min_lr: 0.006810  loss: 3.3299 (3.3844)  weight_decay: 0.0500 (0.0500)  time: 0.7079  data: 0.0004  max mem: 57114
Epoch: [90]  [100/156]  eta: 0:00:41  lr: 0.006808  min_lr: 0.006808  loss: 3.4970 (3.3799)  weight_decay: 0.0500 (0.0500)  time: 0.7112  data: 0.0004  max mem: 57114
Epoch: [90]  [110/156]  eta: 0:00:33  lr: 0.006806  min_lr: 0.006806  loss: 3.6161 (3.3971)  weight_decay: 0.0500 (0.0500)  time: 0.7065  data: 0.0004  max mem: 57114
Epoch: [90]  [120/156]  eta: 0:00:26  lr: 0.006804  min_lr: 0.006804  loss: 3.5940 (3.3918)  weight_decay: 0.0500 (0.0500)  time: 0.6984  data: 0.0004  max mem: 57114
Epoch: [90]  [130/156]  eta: 0:00:18  lr: 0.006802  min_lr: 0.006802  loss: 3.3344 (3.3753)  weight_decay: 0.0500 (0.0500)  time: 0.6933  data: 0.0009  max mem: 57114
Epoch: [90]  [140/156]  eta: 0:00:11  lr: 0.006800  min_lr: 0.006800  loss: 3.0691 (3.3652)  weight_decay: 0.0500 (0.0500)  time: 0.6896  data: 0.0008  max mem: 57114
Epoch: [90]  [150/156]  eta: 0:00:04  lr: 0.006798  min_lr: 0.006798  loss: 3.5251 (3.3783)  weight_decay: 0.0500 (0.0500)  time: 0.6928  data: 0.0001  max mem: 57114
Epoch: [90]  [155/156]  eta: 0:00:00  lr: 0.006797  min_lr: 0.006797  loss: 3.5424 (3.3832)  weight_decay: 0.0500 (0.0500)  time: 0.6877  data: 0.0001  max mem: 57114
Epoch: [90] Total time: 0:01:52 (0.7220 s / it)
Averaged stats: lr: 0.006797  min_lr: 0.006797  loss: 3.5424 (3.3372)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0259 (1.0259)  acc1: 78.6458 (78.6458)  acc5: 95.8333 (95.8333)  time: 2.0650  data: 1.8093  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0495 (1.0896)  acc1: 77.6042 (72.8900)  acc5: 94.7917 (92.3274)  time: 0.5637  data: 0.3619  max mem: 57114
Test: Total time: 0:00:02 (0.5856 s / it)
* Acc@1 74.351 Acc@5 92.399 loss 1.145
Accuracy of the model on the 50000 test images: 74.4%
Max accuracy: 76.02%
Test:  [0/5]  eta: 0:00:12  loss: 8.5887 (8.5887)  acc1: 1.0417 (1.0417)  acc5: 2.0833 (2.0833)  time: 2.4111  data: 2.1676  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.5592 (8.4319)  acc1: 0.5208 (0.6394)  acc5: 1.5625 (1.4066)  time: 0.6328  data: 0.4336  max mem: 57114
Test: Total time: 0:00:03 (0.6442 s / it)
* Acc@1 0.202 Acc@5 0.987 loss 8.328
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.20%
Epoch: [91]  [  0/156]  eta: 0:07:20  lr: 0.006797  min_lr: 0.006797  loss: 3.4553 (3.4553)  weight_decay: 0.0500 (0.0500)  time: 2.8221  data: 2.1739  max mem: 57114
Epoch: [91]  [ 10/156]  eta: 0:02:13  lr: 0.006795  min_lr: 0.006795  loss: 3.2555 (3.1751)  weight_decay: 0.0500 (0.0500)  time: 0.9128  data: 0.1979  max mem: 57114
Epoch: [91]  [ 20/156]  eta: 0:01:52  lr: 0.006793  min_lr: 0.006793  loss: 3.1643 (3.1632)  weight_decay: 0.0500 (0.0500)  time: 0.7241  data: 0.0003  max mem: 57114
Epoch: [91]  [ 30/156]  eta: 0:01:39  lr: 0.006790  min_lr: 0.006790  loss: 3.2747 (3.1393)  weight_decay: 0.0500 (0.0500)  time: 0.7173  data: 0.0003  max mem: 57114
Epoch: [91]  [ 40/156]  eta: 0:01:29  lr: 0.006788  min_lr: 0.006788  loss: 3.4876 (3.1919)  weight_decay: 0.0500 (0.0500)  time: 0.7246  data: 0.0004  max mem: 57114
Epoch: [91]  [ 50/156]  eta: 0:01:21  lr: 0.006786  min_lr: 0.006786  loss: 3.3844 (3.2358)  weight_decay: 0.0500 (0.0500)  time: 0.7421  data: 0.0004  max mem: 57114
Epoch: [91]  [ 60/156]  eta: 0:01:12  lr: 0.006784  min_lr: 0.006784  loss: 3.2967 (3.2357)  weight_decay: 0.0500 (0.0500)  time: 0.7244  data: 0.0004  max mem: 57114
Epoch: [91]  [ 70/156]  eta: 0:01:04  lr: 0.006782  min_lr: 0.006782  loss: 3.2967 (3.2324)  weight_decay: 0.0500 (0.0500)  time: 0.7122  data: 0.0003  max mem: 57114
Epoch: [91]  [ 80/156]  eta: 0:00:56  lr: 0.006780  min_lr: 0.006780  loss: 3.3543 (3.2517)  weight_decay: 0.0500 (0.0500)  time: 0.7116  data: 0.0004  max mem: 57114
Epoch: [91]  [ 90/156]  eta: 0:00:49  lr: 0.006778  min_lr: 0.006778  loss: 3.1231 (3.2473)  weight_decay: 0.0500 (0.0500)  time: 0.7078  data: 0.0004  max mem: 57114
Epoch: [91]  [100/156]  eta: 0:00:41  lr: 0.006776  min_lr: 0.006776  loss: 3.2493 (3.2595)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0004  max mem: 57114
Epoch: [91]  [110/156]  eta: 0:00:33  lr: 0.006774  min_lr: 0.006774  loss: 3.2493 (3.2599)  weight_decay: 0.0500 (0.0500)  time: 0.7024  data: 0.0004  max mem: 57114
Epoch: [91]  [120/156]  eta: 0:00:26  lr: 0.006772  min_lr: 0.006772  loss: 3.1689 (3.2514)  weight_decay: 0.0500 (0.0500)  time: 0.7115  data: 0.0004  max mem: 57114
Epoch: [91]  [130/156]  eta: 0:00:19  lr: 0.006770  min_lr: 0.006770  loss: 3.4324 (3.2703)  weight_decay: 0.0500 (0.0500)  time: 0.7115  data: 0.0009  max mem: 57114
Epoch: [91]  [140/156]  eta: 0:00:11  lr: 0.006768  min_lr: 0.006768  loss: 3.4398 (3.2720)  weight_decay: 0.0500 (0.0500)  time: 0.6844  data: 0.0007  max mem: 57114
Epoch: [91]  [150/156]  eta: 0:00:04  lr: 0.006766  min_lr: 0.006766  loss: 3.3963 (3.2823)  weight_decay: 0.0500 (0.0500)  time: 0.6794  data: 0.0001  max mem: 57114
Epoch: [91]  [155/156]  eta: 0:00:00  lr: 0.006765  min_lr: 0.006765  loss: 3.3963 (3.2895)  weight_decay: 0.0500 (0.0500)  time: 0.6822  data: 0.0001  max mem: 57114
Epoch: [91] Total time: 0:01:53 (0.7261 s / it)
Averaged stats: lr: 0.006765  min_lr: 0.006765  loss: 3.3963 (3.3334)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 0.8927 (0.8927)  acc1: 80.2083 (80.2083)  acc5: 95.8333 (95.8333)  time: 2.2228  data: 1.9670  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0464 (1.0961)  acc1: 80.2083 (76.0870)  acc5: 95.8333 (94.3734)  time: 0.5952  data: 0.3935  max mem: 57114
Test: Total time: 0:00:03 (0.6155 s / it)
* Acc@1 75.567 Acc@5 93.217 loss 1.188
Accuracy of the model on the 50000 test images: 75.6%
Max accuracy: 76.02%
Test:  [0/5]  eta: 0:00:11  loss: 8.6305 (8.6305)  acc1: 1.0417 (1.0417)  acc5: 2.0833 (2.0833)  time: 2.3268  data: 2.0832  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.6245 (8.4727)  acc1: 0.5208 (0.6394)  acc5: 1.5625 (1.5345)  time: 0.6160  data: 0.4167  max mem: 57114
Test: Total time: 0:00:03 (0.6483 s / it)
* Acc@1 0.230 Acc@5 1.019 loss 8.369
Accuracy of the model EMA on 50000 test images: 0.2%
Max EMA accuracy: 0.23%
Epoch: [92]  [  0/156]  eta: 0:06:44  lr: 0.006764  min_lr: 0.006764  loss: 3.4130 (3.4130)  weight_decay: 0.0500 (0.0500)  time: 2.5915  data: 1.9420  max mem: 57114
Epoch: [92]  [ 10/156]  eta: 0:02:10  lr: 0.006762  min_lr: 0.006762  loss: 3.4130 (3.1983)  weight_decay: 0.0500 (0.0500)  time: 0.8972  data: 0.1769  max mem: 57114
Epoch: [92]  [ 20/156]  eta: 0:01:51  lr: 0.006760  min_lr: 0.006760  loss: 3.3623 (3.3063)  weight_decay: 0.0500 (0.0500)  time: 0.7285  data: 0.0004  max mem: 57114
Epoch: [92]  [ 30/156]  eta: 0:01:38  lr: 0.006758  min_lr: 0.006758  loss: 3.4359 (3.3368)  weight_decay: 0.0500 (0.0500)  time: 0.7162  data: 0.0004  max mem: 57114
Epoch: [92]  [ 40/156]  eta: 0:01:28  lr: 0.006756  min_lr: 0.006756  loss: 3.4070 (3.3158)  weight_decay: 0.0500 (0.0500)  time: 0.7126  data: 0.0004  max mem: 57114
Epoch: [92]  [ 50/156]  eta: 0:01:19  lr: 0.006754  min_lr: 0.006754  loss: 3.3425 (3.3473)  weight_decay: 0.0500 (0.0500)  time: 0.7088  data: 0.0004  max mem: 57114
Epoch: [92]  [ 60/156]  eta: 0:01:12  lr: 0.006752  min_lr: 0.006752  loss: 3.5074 (3.3689)  weight_decay: 0.0500 (0.0500)  time: 0.7175  data: 0.0004  max mem: 57114
Epoch: [92]  [ 70/156]  eta: 0:01:04  lr: 0.006750  min_lr: 0.006750  loss: 3.5738 (3.3570)  weight_decay: 0.0500 (0.0500)  time: 0.7277  data: 0.0003  max mem: 57114
Epoch: [92]  [ 80/156]  eta: 0:00:56  lr: 0.006748  min_lr: 0.006748  loss: 3.5079 (3.3794)  weight_decay: 0.0500 (0.0500)  time: 0.7127  data: 0.0004  max mem: 57114
Epoch: [92]  [ 90/156]  eta: 0:00:48  lr: 0.006746  min_lr: 0.006746  loss: 3.5119 (3.3697)  weight_decay: 0.0500 (0.0500)  time: 0.7095  data: 0.0004  max mem: 57114
Epoch: [92]  [100/156]  eta: 0:00:41  lr: 0.006744  min_lr: 0.006744  loss: 3.5365 (3.3862)  weight_decay: 0.0500 (0.0500)  time: 0.7141  data: 0.0004  max mem: 57114
Epoch: [92]  [110/156]  eta: 0:00:33  lr: 0.006741  min_lr: 0.006741  loss: 3.5491 (3.3904)  weight_decay: 0.0500 (0.0500)  time: 0.7031  data: 0.0004  max mem: 57114
Epoch: [92]  [120/156]  eta: 0:00:26  lr: 0.006739  min_lr: 0.006739  loss: 3.6702 (3.4182)  weight_decay: 0.0500 (0.0500)  time: 0.6909  data: 0.0004  max mem: 57114
Epoch: [92]  [130/156]  eta: 0:00:18  lr: 0.006737  min_lr: 0.006737  loss: 3.5731 (3.4090)  weight_decay: 0.0500 (0.0500)  time: 0.6954  data: 0.0008  max mem: 57114
Epoch: [92]  [140/156]  eta: 0:00:11  lr: 0.006735  min_lr: 0.006735  loss: 3.4875 (3.4129)  weight_decay: 0.0500 (0.0500)  time: 0.6911  data: 0.0007  max mem: 57114
Epoch: [92]  [150/156]  eta: 0:00:04  lr: 0.006733  min_lr: 0.006733  loss: 3.3611 (3.4036)  weight_decay: 0.0500 (0.0500)  time: 0.6880  data: 0.0001  max mem: 57114
Epoch: [92]  [155/156]  eta: 0:00:00  lr: 0.006732  min_lr: 0.006732  loss: 3.3464 (3.4008)  weight_decay: 0.0500 (0.0500)  time: 0.6832  data: 0.0001  max mem: 57114
Epoch: [92] Total time: 0:01:52 (0.7215 s / it)
Averaged stats: lr: 0.006732  min_lr: 0.006732  loss: 3.3464 (3.3311)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9286 (0.9286)  acc1: 80.2083 (80.2083)  acc5: 95.8333 (95.8333)  time: 2.0503  data: 1.7944  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1111 (1.0430)  acc1: 79.1667 (76.0870)  acc5: 93.2292 (92.3274)  time: 0.5607  data: 0.3589  max mem: 57114
Test: Total time: 0:00:02 (0.5825 s / it)
* Acc@1 75.328 Acc@5 93.015 loss 1.133
Accuracy of the model on the 50000 test images: 75.3%
Max accuracy: 76.02%
Test:  [0/5]  eta: 0:00:11  loss: 8.6669 (8.6669)  acc1: 1.0417 (1.0417)  acc5: 2.0833 (2.0833)  time: 2.2963  data: 2.0528  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.6669 (8.5080)  acc1: 0.5208 (0.6394)  acc5: 1.5625 (1.5345)  time: 0.6099  data: 0.4107  max mem: 57114
Test: Total time: 0:00:03 (0.6227 s / it)
* Acc@1 0.256 Acc@5 1.055 loss 8.403
Accuracy of the model EMA on 50000 test images: 0.3%
Max EMA accuracy: 0.26%
Epoch: [93]  [  0/156]  eta: 0:08:13  lr: 0.006732  min_lr: 0.006732  loss: 3.0813 (3.0813)  weight_decay: 0.0500 (0.0500)  time: 3.1651  data: 2.5185  max mem: 57114
Epoch: [93]  [ 10/156]  eta: 0:02:15  lr: 0.006730  min_lr: 0.006730  loss: 3.5230 (3.4400)  weight_decay: 0.0500 (0.0500)  time: 0.9252  data: 0.2293  max mem: 57114
Epoch: [93]  [ 20/156]  eta: 0:01:53  lr: 0.006728  min_lr: 0.006728  loss: 3.4650 (3.3548)  weight_decay: 0.0500 (0.0500)  time: 0.7187  data: 0.0004  max mem: 57114
Epoch: [93]  [ 30/156]  eta: 0:01:39  lr: 0.006725  min_lr: 0.006725  loss: 3.3879 (3.3905)  weight_decay: 0.0500 (0.0500)  time: 0.7208  data: 0.0003  max mem: 57114
Epoch: [93]  [ 40/156]  eta: 0:01:30  lr: 0.006723  min_lr: 0.006723  loss: 3.3194 (3.3435)  weight_decay: 0.0500 (0.0500)  time: 0.7220  data: 0.0004  max mem: 57114
Epoch: [93]  [ 50/156]  eta: 0:01:21  lr: 0.006721  min_lr: 0.006721  loss: 3.3398 (3.3746)  weight_decay: 0.0500 (0.0500)  time: 0.7408  data: 0.0004  max mem: 57114
Epoch: [93]  [ 60/156]  eta: 0:01:13  lr: 0.006719  min_lr: 0.006719  loss: 3.3631 (3.3560)  weight_decay: 0.0500 (0.0500)  time: 0.7298  data: 0.0004  max mem: 57114
Epoch: [93]  [ 70/156]  eta: 0:01:04  lr: 0.006717  min_lr: 0.006717  loss: 3.3810 (3.3528)  weight_decay: 0.0500 (0.0500)  time: 0.7100  data: 0.0004  max mem: 57114
Epoch: [93]  [ 80/156]  eta: 0:00:56  lr: 0.006715  min_lr: 0.006715  loss: 3.3810 (3.3470)  weight_decay: 0.0500 (0.0500)  time: 0.7043  data: 0.0004  max mem: 57114
Epoch: [93]  [ 90/156]  eta: 0:00:49  lr: 0.006713  min_lr: 0.006713  loss: 3.4445 (3.3544)  weight_decay: 0.0500 (0.0500)  time: 0.7099  data: 0.0004  max mem: 57114
Epoch: [93]  [100/156]  eta: 0:00:41  lr: 0.006711  min_lr: 0.006711  loss: 3.6138 (3.3655)  weight_decay: 0.0500 (0.0500)  time: 0.7075  data: 0.0004  max mem: 57114
Epoch: [93]  [110/156]  eta: 0:00:33  lr: 0.006709  min_lr: 0.006709  loss: 3.6529 (3.3941)  weight_decay: 0.0500 (0.0500)  time: 0.6994  data: 0.0004  max mem: 57114
Epoch: [93]  [120/156]  eta: 0:00:26  lr: 0.006706  min_lr: 0.006706  loss: 3.6194 (3.3873)  weight_decay: 0.0500 (0.0500)  time: 0.7076  data: 0.0004  max mem: 57114
Epoch: [93]  [130/156]  eta: 0:00:19  lr: 0.006704  min_lr: 0.006704  loss: 3.3410 (3.3719)  weight_decay: 0.0500 (0.0500)  time: 0.7122  data: 0.0005  max mem: 57114
Epoch: [93]  [140/156]  eta: 0:00:11  lr: 0.006702  min_lr: 0.006702  loss: 3.3214 (3.3569)  weight_decay: 0.0500 (0.0500)  time: 0.6986  data: 0.0004  max mem: 57114
Epoch: [93]  [150/156]  eta: 0:00:04  lr: 0.006700  min_lr: 0.006700  loss: 3.3625 (3.3709)  weight_decay: 0.0500 (0.0500)  time: 0.6865  data: 0.0001  max mem: 57114
Epoch: [93]  [155/156]  eta: 0:00:00  lr: 0.006699  min_lr: 0.006699  loss: 3.3993 (3.3579)  weight_decay: 0.0500 (0.0500)  time: 0.6880  data: 0.0001  max mem: 57114
Epoch: [93] Total time: 0:01:53 (0.7277 s / it)
Averaged stats: lr: 0.006699  min_lr: 0.006699  loss: 3.3993 (3.3205)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8188 (0.8188)  acc1: 81.2500 (81.2500)  acc5: 97.9167 (97.9167)  time: 2.0806  data: 1.8247  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0167 (0.9900)  acc1: 79.1667 (76.4706)  acc5: 95.3125 (93.4783)  time: 0.5668  data: 0.3650  max mem: 57114
Test: Total time: 0:00:02 (0.5902 s / it)
* Acc@1 76.073 Acc@5 93.534 loss 1.061
Accuracy of the model on the 50000 test images: 76.1%
Max accuracy: 76.07%
Test:  [0/5]  eta: 0:00:10  loss: 8.6677 (8.6677)  acc1: 1.0417 (1.0417)  acc5: 2.0833 (2.0833)  time: 2.1167  data: 1.8732  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.6677 (8.5089)  acc1: 0.5208 (0.6394)  acc5: 1.5625 (1.5345)  time: 0.5740  data: 0.3748  max mem: 57114
Test: Total time: 0:00:02 (0.5863 s / it)
* Acc@1 0.268 Acc@5 1.101 loss 8.408
Accuracy of the model EMA on 50000 test images: 0.3%
Max EMA accuracy: 0.27%
Epoch: [94]  [  0/156]  eta: 0:06:30  lr: 0.006699  min_lr: 0.006699  loss: 3.5354 (3.5354)  weight_decay: 0.0500 (0.0500)  time: 2.5050  data: 1.8539  max mem: 57114
Epoch: [94]  [ 10/156]  eta: 0:02:12  lr: 0.006697  min_lr: 0.006697  loss: 3.4774 (3.1522)  weight_decay: 0.0500 (0.0500)  time: 0.9042  data: 0.1689  max mem: 57114
Epoch: [94]  [ 20/156]  eta: 0:01:53  lr: 0.006695  min_lr: 0.006695  loss: 3.2245 (3.1456)  weight_decay: 0.0500 (0.0500)  time: 0.7489  data: 0.0004  max mem: 57114
Epoch: [94]  [ 30/156]  eta: 0:01:40  lr: 0.006692  min_lr: 0.006692  loss: 2.9417 (3.0314)  weight_decay: 0.0500 (0.0500)  time: 0.7361  data: 0.0004  max mem: 57114
Epoch: [94]  [ 40/156]  eta: 0:01:30  lr: 0.006690  min_lr: 0.006690  loss: 2.9417 (3.0689)  weight_decay: 0.0500 (0.0500)  time: 0.7302  data: 0.0003  max mem: 57114
Epoch: [94]  [ 50/156]  eta: 0:01:22  lr: 0.006688  min_lr: 0.006688  loss: 3.3049 (3.1009)  weight_decay: 0.0500 (0.0500)  time: 0.7463  data: 0.0004  max mem: 57114
Epoch: [94]  [ 60/156]  eta: 0:01:14  lr: 0.006686  min_lr: 0.006686  loss: 3.3836 (3.1201)  weight_decay: 0.0500 (0.0500)  time: 0.7470  data: 0.0004  max mem: 57114
Epoch: [94]  [ 70/156]  eta: 0:01:05  lr: 0.006684  min_lr: 0.006684  loss: 3.4135 (3.1392)  weight_decay: 0.0500 (0.0500)  time: 0.7366  data: 0.0004  max mem: 57114
Epoch: [94]  [ 80/156]  eta: 0:00:57  lr: 0.006682  min_lr: 0.006682  loss: 3.4179 (3.1622)  weight_decay: 0.0500 (0.0500)  time: 0.7199  data: 0.0004  max mem: 57114
Epoch: [94]  [ 90/156]  eta: 0:00:49  lr: 0.006680  min_lr: 0.006680  loss: 3.3675 (3.1813)  weight_decay: 0.0500 (0.0500)  time: 0.7027  data: 0.0004  max mem: 57114
Epoch: [94]  [100/156]  eta: 0:00:41  lr: 0.006678  min_lr: 0.006678  loss: 3.4503 (3.1769)  weight_decay: 0.0500 (0.0500)  time: 0.6943  data: 0.0004  max mem: 57114
Epoch: [94]  [110/156]  eta: 0:00:34  lr: 0.006675  min_lr: 0.006675  loss: 2.9425 (3.1628)  weight_decay: 0.0500 (0.0500)  time: 0.6989  data: 0.0004  max mem: 57114
Epoch: [94]  [120/156]  eta: 0:00:26  lr: 0.006673  min_lr: 0.006673  loss: 3.1839 (3.1742)  weight_decay: 0.0500 (0.0500)  time: 0.7132  data: 0.0004  max mem: 57114
Epoch: [94]  [130/156]  eta: 0:00:19  lr: 0.006671  min_lr: 0.006671  loss: 3.4123 (3.2022)  weight_decay: 0.0500 (0.0500)  time: 0.7116  data: 0.0009  max mem: 57114
Epoch: [94]  [140/156]  eta: 0:00:11  lr: 0.006669  min_lr: 0.006669  loss: 3.4139 (3.2144)  weight_decay: 0.0500 (0.0500)  time: 0.6938  data: 0.0007  max mem: 57114
Epoch: [94]  [150/156]  eta: 0:00:04  lr: 0.006667  min_lr: 0.006667  loss: 3.3607 (3.2186)  weight_decay: 0.0500 (0.0500)  time: 0.6838  data: 0.0001  max mem: 57114
Epoch: [94]  [155/156]  eta: 0:00:00  lr: 0.006666  min_lr: 0.006666  loss: 3.3312 (3.2197)  weight_decay: 0.0500 (0.0500)  time: 0.6824  data: 0.0001  max mem: 57114
Epoch: [94] Total time: 0:01:53 (0.7306 s / it)
Averaged stats: lr: 0.006666  min_lr: 0.006666  loss: 3.3312 (3.3239)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0070 (1.0070)  acc1: 79.1667 (79.1667)  acc5: 95.3125 (95.3125)  time: 2.1201  data: 1.8583  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0550 (1.0800)  acc1: 79.1667 (76.0870)  acc5: 95.3125 (93.9898)  time: 0.5748  data: 0.3717  max mem: 57114
Test: Total time: 0:00:02 (0.5992 s / it)
* Acc@1 76.838 Acc@5 93.606 loss 1.130
Accuracy of the model on the 50000 test images: 76.8%
Max accuracy: 76.84%
Test:  [0/5]  eta: 0:00:10  loss: 8.6619 (8.6619)  acc1: 1.0417 (1.0417)  acc5: 2.6042 (2.6042)  time: 2.0772  data: 1.8336  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.6619 (8.4996)  acc1: 0.5208 (0.7673)  acc5: 1.5625 (1.6624)  time: 0.5660  data: 0.3668  max mem: 57114
Test: Total time: 0:00:02 (0.5796 s / it)
* Acc@1 0.286 Acc@5 1.189 loss 8.403
Accuracy of the model EMA on 50000 test images: 0.3%
Max EMA accuracy: 0.29%
Epoch: [95]  [  0/156]  eta: 0:06:36  lr: 0.006666  min_lr: 0.006666  loss: 2.7945 (2.7945)  weight_decay: 0.0500 (0.0500)  time: 2.5434  data: 1.8920  max mem: 57114
Epoch: [95]  [ 10/156]  eta: 0:02:09  lr: 0.006663  min_lr: 0.006663  loss: 3.1164 (3.1628)  weight_decay: 0.0500 (0.0500)  time: 0.8876  data: 0.1724  max mem: 57114
Epoch: [95]  [ 20/156]  eta: 0:01:51  lr: 0.006661  min_lr: 0.006661  loss: 3.2932 (3.2460)  weight_decay: 0.0500 (0.0500)  time: 0.7325  data: 0.0004  max mem: 57114
Epoch: [95]  [ 30/156]  eta: 0:01:40  lr: 0.006659  min_lr: 0.006659  loss: 3.4667 (3.3085)  weight_decay: 0.0500 (0.0500)  time: 0.7522  data: 0.0004  max mem: 57114
Epoch: [95]  [ 40/156]  eta: 0:01:31  lr: 0.006657  min_lr: 0.006657  loss: 3.2972 (3.2844)  weight_decay: 0.0500 (0.0500)  time: 0.7591  data: 0.0004  max mem: 57114
Epoch: [95]  [ 50/156]  eta: 0:01:22  lr: 0.006655  min_lr: 0.006655  loss: 3.2688 (3.2860)  weight_decay: 0.0500 (0.0500)  time: 0.7534  data: 0.0004  max mem: 57114
Epoch: [95]  [ 60/156]  eta: 0:01:14  lr: 0.006653  min_lr: 0.006653  loss: 3.4777 (3.3080)  weight_decay: 0.0500 (0.0500)  time: 0.7371  data: 0.0004  max mem: 57114
Epoch: [95]  [ 70/156]  eta: 0:01:05  lr: 0.006651  min_lr: 0.006651  loss: 3.4850 (3.3102)  weight_decay: 0.0500 (0.0500)  time: 0.7239  data: 0.0004  max mem: 57114
Epoch: [95]  [ 80/156]  eta: 0:00:57  lr: 0.006648  min_lr: 0.006648  loss: 3.5471 (3.3078)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0004  max mem: 57114
Epoch: [95]  [ 90/156]  eta: 0:00:49  lr: 0.006646  min_lr: 0.006646  loss: 3.4317 (3.3089)  weight_decay: 0.0500 (0.0500)  time: 0.7052  data: 0.0004  max mem: 57114
Epoch: [95]  [100/156]  eta: 0:00:41  lr: 0.006644  min_lr: 0.006644  loss: 3.4166 (3.3142)  weight_decay: 0.0500 (0.0500)  time: 0.7082  data: 0.0004  max mem: 57114
Epoch: [95]  [110/156]  eta: 0:00:34  lr: 0.006642  min_lr: 0.006642  loss: 3.3969 (3.2963)  weight_decay: 0.0500 (0.0500)  time: 0.7072  data: 0.0004  max mem: 57114
Epoch: [95]  [120/156]  eta: 0:00:26  lr: 0.006640  min_lr: 0.006640  loss: 3.4274 (3.3081)  weight_decay: 0.0500 (0.0500)  time: 0.7038  data: 0.0004  max mem: 57114
Epoch: [95]  [130/156]  eta: 0:00:19  lr: 0.006638  min_lr: 0.006638  loss: 3.4455 (3.3038)  weight_decay: 0.0500 (0.0500)  time: 0.6988  data: 0.0008  max mem: 57114
Epoch: [95]  [140/156]  eta: 0:00:11  lr: 0.006635  min_lr: 0.006635  loss: 3.3605 (3.3040)  weight_decay: 0.0500 (0.0500)  time: 0.6900  data: 0.0007  max mem: 57114
Epoch: [95]  [150/156]  eta: 0:00:04  lr: 0.006633  min_lr: 0.006633  loss: 3.3583 (3.3001)  weight_decay: 0.0500 (0.0500)  time: 0.6855  data: 0.0001  max mem: 57114
Epoch: [95]  [155/156]  eta: 0:00:00  lr: 0.006632  min_lr: 0.006632  loss: 3.3042 (3.2973)  weight_decay: 0.0500 (0.0500)  time: 0.6863  data: 0.0001  max mem: 57114
Epoch: [95] Total time: 0:01:54 (0.7309 s / it)
Averaged stats: lr: 0.006632  min_lr: 0.006632  loss: 3.3042 (3.3104)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0602 (1.0602)  acc1: 81.7708 (81.7708)  acc5: 95.8333 (95.8333)  time: 2.0922  data: 1.8363  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0602 (1.0604)  acc1: 81.7708 (77.8772)  acc5: 95.8333 (93.8619)  time: 0.5691  data: 0.3673  max mem: 57114
Test: Total time: 0:00:02 (0.5909 s / it)
* Acc@1 76.037 Acc@5 93.300 loss 1.154
Accuracy of the model on the 50000 test images: 76.0%
Max accuracy: 76.84%
Test:  [0/5]  eta: 0:00:11  loss: 8.6527 (8.6527)  acc1: 1.0417 (1.0417)  acc5: 2.6042 (2.6042)  time: 2.3105  data: 2.0671  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.6527 (8.4879)  acc1: 1.0417 (0.8951)  acc5: 1.5625 (1.6624)  time: 0.6128  data: 0.4135  max mem: 57114
Test: Total time: 0:00:03 (0.6256 s / it)
* Acc@1 0.310 Acc@5 1.247 loss 8.396
Accuracy of the model EMA on 50000 test images: 0.3%
Max EMA accuracy: 0.31%
Epoch: [96]  [  0/156]  eta: 0:06:59  lr: 0.006632  min_lr: 0.006632  loss: 3.3014 (3.3014)  weight_decay: 0.0500 (0.0500)  time: 2.6865  data: 2.0382  max mem: 57114
Epoch: [96]  [ 10/156]  eta: 0:02:11  lr: 0.006630  min_lr: 0.006630  loss: 3.4443 (3.3893)  weight_decay: 0.0500 (0.0500)  time: 0.8987  data: 0.1855  max mem: 57114
Epoch: [96]  [ 20/156]  eta: 0:01:50  lr: 0.006628  min_lr: 0.006628  loss: 3.5465 (3.3883)  weight_decay: 0.0500 (0.0500)  time: 0.7197  data: 0.0003  max mem: 57114
Epoch: [96]  [ 30/156]  eta: 0:01:38  lr: 0.006625  min_lr: 0.006625  loss: 3.6255 (3.4240)  weight_decay: 0.0500 (0.0500)  time: 0.7216  data: 0.0004  max mem: 57114
Epoch: [96]  [ 40/156]  eta: 0:01:29  lr: 0.006623  min_lr: 0.006623  loss: 3.5527 (3.3964)  weight_decay: 0.0500 (0.0500)  time: 0.7198  data: 0.0003  max mem: 57114
Epoch: [96]  [ 50/156]  eta: 0:01:20  lr: 0.006621  min_lr: 0.006621  loss: 3.4182 (3.3660)  weight_decay: 0.0500 (0.0500)  time: 0.7102  data: 0.0004  max mem: 57114
Epoch: [96]  [ 60/156]  eta: 0:01:12  lr: 0.006619  min_lr: 0.006619  loss: 3.2887 (3.3527)  weight_decay: 0.0500 (0.0500)  time: 0.7204  data: 0.0004  max mem: 57114
Epoch: [96]  [ 70/156]  eta: 0:01:04  lr: 0.006617  min_lr: 0.006617  loss: 3.3759 (3.3378)  weight_decay: 0.0500 (0.0500)  time: 0.7196  data: 0.0004  max mem: 57114
Epoch: [96]  [ 80/156]  eta: 0:00:56  lr: 0.006615  min_lr: 0.006615  loss: 3.0694 (3.3207)  weight_decay: 0.0500 (0.0500)  time: 0.7122  data: 0.0004  max mem: 57114
Epoch: [96]  [ 90/156]  eta: 0:00:48  lr: 0.006612  min_lr: 0.006612  loss: 3.0430 (3.2892)  weight_decay: 0.0500 (0.0500)  time: 0.7126  data: 0.0004  max mem: 57114
Epoch: [96]  [100/156]  eta: 0:00:41  lr: 0.006610  min_lr: 0.006610  loss: 3.3477 (3.3016)  weight_decay: 0.0500 (0.0500)  time: 0.7082  data: 0.0004  max mem: 57114
Epoch: [96]  [110/156]  eta: 0:00:33  lr: 0.006608  min_lr: 0.006608  loss: 3.3186 (3.2931)  weight_decay: 0.0500 (0.0500)  time: 0.7120  data: 0.0004  max mem: 57114
Epoch: [96]  [120/156]  eta: 0:00:26  lr: 0.006606  min_lr: 0.006606  loss: 3.4886 (3.3095)  weight_decay: 0.0500 (0.0500)  time: 0.7068  data: 0.0004  max mem: 57114
Epoch: [96]  [130/156]  eta: 0:00:18  lr: 0.006604  min_lr: 0.006604  loss: 3.4886 (3.2920)  weight_decay: 0.0500 (0.0500)  time: 0.6994  data: 0.0009  max mem: 57114
Epoch: [96]  [140/156]  eta: 0:00:11  lr: 0.006601  min_lr: 0.006601  loss: 3.4095 (3.3098)  weight_decay: 0.0500 (0.0500)  time: 0.6891  data: 0.0007  max mem: 57114
Epoch: [96]  [150/156]  eta: 0:00:04  lr: 0.006599  min_lr: 0.006599  loss: 3.4399 (3.3123)  weight_decay: 0.0500 (0.0500)  time: 0.6826  data: 0.0001  max mem: 57114
Epoch: [96]  [155/156]  eta: 0:00:00  lr: 0.006598  min_lr: 0.006598  loss: 3.3327 (3.2972)  weight_decay: 0.0500 (0.0500)  time: 0.6877  data: 0.0001  max mem: 57114
Epoch: [96] Total time: 0:01:52 (0.7238 s / it)
Averaged stats: lr: 0.006598  min_lr: 0.006598  loss: 3.3327 (3.3118)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9430 (0.9430)  acc1: 81.7708 (81.7708)  acc5: 96.8750 (96.8750)  time: 2.0677  data: 1.8121  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0664 (0.9891)  acc1: 78.1250 (77.4936)  acc5: 96.3542 (94.3734)  time: 0.5643  data: 0.3625  max mem: 57114
Test: Total time: 0:00:02 (0.5866 s / it)
* Acc@1 76.341 Acc@5 93.648 loss 1.113
Accuracy of the model on the 50000 test images: 76.3%
Max accuracy: 76.84%
Test:  [0/5]  eta: 0:00:11  loss: 8.6135 (8.6135)  acc1: 1.0417 (1.0417)  acc5: 2.6042 (2.6042)  time: 2.3389  data: 2.0954  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.6135 (8.4464)  acc1: 1.0417 (0.8951)  acc5: 1.5625 (1.9182)  time: 0.6184  data: 0.4192  max mem: 57114
Test: Total time: 0:00:03 (0.6302 s / it)
* Acc@1 0.348 Acc@5 1.309 loss 8.362
Accuracy of the model EMA on 50000 test images: 0.3%
Max EMA accuracy: 0.35%
Epoch: [97]  [  0/156]  eta: 0:06:21  lr: 0.006598  min_lr: 0.006598  loss: 3.1405 (3.1405)  weight_decay: 0.0500 (0.0500)  time: 2.4435  data: 1.7853  max mem: 57114
Epoch: [97]  [ 10/156]  eta: 0:02:09  lr: 0.006596  min_lr: 0.006596  loss: 3.4610 (3.3284)  weight_decay: 0.0500 (0.0500)  time: 0.8899  data: 0.1626  max mem: 57114
Epoch: [97]  [ 20/156]  eta: 0:01:50  lr: 0.006594  min_lr: 0.006594  loss: 3.4610 (3.4505)  weight_decay: 0.0500 (0.0500)  time: 0.7283  data: 0.0004  max mem: 57114
Epoch: [97]  [ 30/156]  eta: 0:01:38  lr: 0.006591  min_lr: 0.006591  loss: 3.4497 (3.3677)  weight_decay: 0.0500 (0.0500)  time: 0.7206  data: 0.0003  max mem: 57114
Epoch: [97]  [ 40/156]  eta: 0:01:29  lr: 0.006589  min_lr: 0.006589  loss: 3.4432 (3.3786)  weight_decay: 0.0500 (0.0500)  time: 0.7331  data: 0.0003  max mem: 57114
Epoch: [97]  [ 50/156]  eta: 0:01:20  lr: 0.006587  min_lr: 0.006587  loss: 3.5426 (3.4057)  weight_decay: 0.0500 (0.0500)  time: 0.7228  data: 0.0004  max mem: 57114
Epoch: [97]  [ 60/156]  eta: 0:01:12  lr: 0.006585  min_lr: 0.006585  loss: 3.5388 (3.3662)  weight_decay: 0.0500 (0.0500)  time: 0.7178  data: 0.0004  max mem: 57114
Epoch: [97]  [ 70/156]  eta: 0:01:04  lr: 0.006583  min_lr: 0.006583  loss: 3.3006 (3.3614)  weight_decay: 0.0500 (0.0500)  time: 0.7193  data: 0.0004  max mem: 57114
Epoch: [97]  [ 80/156]  eta: 0:00:56  lr: 0.006580  min_lr: 0.006580  loss: 3.5002 (3.3756)  weight_decay: 0.0500 (0.0500)  time: 0.7071  data: 0.0004  max mem: 57114
Epoch: [97]  [ 90/156]  eta: 0:00:48  lr: 0.006578  min_lr: 0.006578  loss: 3.5100 (3.3773)  weight_decay: 0.0500 (0.0500)  time: 0.7109  data: 0.0004  max mem: 57114
Epoch: [97]  [100/156]  eta: 0:00:41  lr: 0.006576  min_lr: 0.006576  loss: 3.3220 (3.3792)  weight_decay: 0.0500 (0.0500)  time: 0.6974  data: 0.0003  max mem: 57114
Epoch: [97]  [110/156]  eta: 0:00:33  lr: 0.006574  min_lr: 0.006574  loss: 3.2016 (3.3571)  weight_decay: 0.0500 (0.0500)  time: 0.6941  data: 0.0003  max mem: 57114
Epoch: [97]  [120/156]  eta: 0:00:26  lr: 0.006572  min_lr: 0.006572  loss: 3.3950 (3.3638)  weight_decay: 0.0500 (0.0500)  time: 0.7066  data: 0.0004  max mem: 57114
Epoch: [97]  [130/156]  eta: 0:00:18  lr: 0.006569  min_lr: 0.006569  loss: 3.5056 (3.3644)  weight_decay: 0.0500 (0.0500)  time: 0.7031  data: 0.0009  max mem: 57114
Epoch: [97]  [140/156]  eta: 0:00:11  lr: 0.006567  min_lr: 0.006567  loss: 3.3976 (3.3607)  weight_decay: 0.0500 (0.0500)  time: 0.6879  data: 0.0007  max mem: 57114
Epoch: [97]  [150/156]  eta: 0:00:04  lr: 0.006565  min_lr: 0.006565  loss: 3.5383 (3.3755)  weight_decay: 0.0500 (0.0500)  time: 0.6800  data: 0.0001  max mem: 57114
Epoch: [97]  [155/156]  eta: 0:00:00  lr: 0.006564  min_lr: 0.006564  loss: 3.5383 (3.3687)  weight_decay: 0.0500 (0.0500)  time: 0.6821  data: 0.0001  max mem: 57114
Epoch: [97] Total time: 0:01:52 (0.7213 s / it)
Averaged stats: lr: 0.006564  min_lr: 0.006564  loss: 3.5383 (3.3065)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8050 (0.8050)  acc1: 84.3750 (84.3750)  acc5: 96.8750 (96.8750)  time: 2.0416  data: 1.7857  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0612 (0.9871)  acc1: 79.6875 (77.7494)  acc5: 95.8333 (94.8849)  time: 0.5591  data: 0.3572  max mem: 57114
Test: Total time: 0:00:02 (0.5811 s / it)
* Acc@1 76.794 Acc@5 93.494 loss 1.086
Accuracy of the model on the 50000 test images: 76.8%
Max accuracy: 76.84%
Test:  [0/5]  eta: 0:00:11  loss: 8.5834 (8.5834)  acc1: 1.0417 (1.0417)  acc5: 3.1250 (3.1250)  time: 2.2967  data: 2.0533  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.5834 (8.4178)  acc1: 1.0417 (1.0230)  acc5: 2.0833 (2.0460)  time: 0.6099  data: 0.4107  max mem: 57114
Test: Total time: 0:00:03 (0.6242 s / it)
* Acc@1 0.398 Acc@5 1.399 loss 8.339
Accuracy of the model EMA on 50000 test images: 0.4%
Max EMA accuracy: 0.40%
Epoch: [98]  [  0/156]  eta: 0:06:29  lr: 0.006564  min_lr: 0.006564  loss: 3.7017 (3.7017)  weight_decay: 0.0500 (0.0500)  time: 2.4999  data: 1.8409  max mem: 57114
Epoch: [98]  [ 10/156]  eta: 0:02:07  lr: 0.006561  min_lr: 0.006561  loss: 3.6032 (3.4748)  weight_decay: 0.0500 (0.0500)  time: 0.8744  data: 0.1677  max mem: 57114
Epoch: [98]  [ 20/156]  eta: 0:01:47  lr: 0.006559  min_lr: 0.006559  loss: 3.2883 (3.2473)  weight_decay: 0.0500 (0.0500)  time: 0.7051  data: 0.0005  max mem: 57114
Epoch: [98]  [ 30/156]  eta: 0:01:36  lr: 0.006557  min_lr: 0.006557  loss: 3.2408 (3.2341)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0006  max mem: 57114
Epoch: [98]  [ 40/156]  eta: 0:01:28  lr: 0.006555  min_lr: 0.006555  loss: 3.2687 (3.1793)  weight_decay: 0.0500 (0.0500)  time: 0.7266  data: 0.0004  max mem: 57114
Epoch: [98]  [ 50/156]  eta: 0:01:20  lr: 0.006553  min_lr: 0.006553  loss: 3.2826 (3.2081)  weight_decay: 0.0500 (0.0500)  time: 0.7361  data: 0.0004  max mem: 57114
Epoch: [98]  [ 60/156]  eta: 0:01:12  lr: 0.006550  min_lr: 0.006550  loss: 3.2826 (3.2035)  weight_decay: 0.0500 (0.0500)  time: 0.7372  data: 0.0004  max mem: 57114
Epoch: [98]  [ 70/156]  eta: 0:01:04  lr: 0.006548  min_lr: 0.006548  loss: 3.5201 (3.2644)  weight_decay: 0.0500 (0.0500)  time: 0.7299  data: 0.0004  max mem: 57114
Epoch: [98]  [ 80/156]  eta: 0:00:56  lr: 0.006546  min_lr: 0.006546  loss: 3.5303 (3.2760)  weight_decay: 0.0500 (0.0500)  time: 0.7099  data: 0.0004  max mem: 57114
Epoch: [98]  [ 90/156]  eta: 0:00:48  lr: 0.006544  min_lr: 0.006544  loss: 3.4466 (3.2857)  weight_decay: 0.0500 (0.0500)  time: 0.6935  data: 0.0004  max mem: 57114
Epoch: [98]  [100/156]  eta: 0:00:41  lr: 0.006542  min_lr: 0.006542  loss: 3.3549 (3.2850)  weight_decay: 0.0500 (0.0500)  time: 0.7075  data: 0.0004  max mem: 57114
Epoch: [98]  [110/156]  eta: 0:00:33  lr: 0.006539  min_lr: 0.006539  loss: 3.0947 (3.2570)  weight_decay: 0.0500 (0.0500)  time: 0.7161  data: 0.0004  max mem: 57114
Epoch: [98]  [120/156]  eta: 0:00:26  lr: 0.006537  min_lr: 0.006537  loss: 3.0947 (3.2571)  weight_decay: 0.0500 (0.0500)  time: 0.7046  data: 0.0004  max mem: 57114
Epoch: [98]  [130/156]  eta: 0:00:18  lr: 0.006535  min_lr: 0.006535  loss: 3.2492 (3.2652)  weight_decay: 0.0500 (0.0500)  time: 0.6988  data: 0.0008  max mem: 57114
Epoch: [98]  [140/156]  eta: 0:00:11  lr: 0.006533  min_lr: 0.006533  loss: 3.2492 (3.2483)  weight_decay: 0.0500 (0.0500)  time: 0.6951  data: 0.0007  max mem: 57114
Epoch: [98]  [150/156]  eta: 0:00:04  lr: 0.006530  min_lr: 0.006530  loss: 2.9832 (3.2497)  weight_decay: 0.0500 (0.0500)  time: 0.6908  data: 0.0001  max mem: 57114
Epoch: [98]  [155/156]  eta: 0:00:00  lr: 0.006529  min_lr: 0.006529  loss: 3.4519 (3.2584)  weight_decay: 0.0500 (0.0500)  time: 0.6838  data: 0.0001  max mem: 57114
Epoch: [98] Total time: 0:01:52 (0.7233 s / it)
Averaged stats: lr: 0.006529  min_lr: 0.006529  loss: 3.4519 (3.3016)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9141 (0.9141)  acc1: 80.7292 (80.7292)  acc5: 96.8750 (96.8750)  time: 2.1563  data: 1.9002  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0650 (0.9859)  acc1: 73.4375 (75.5755)  acc5: 95.3125 (93.7340)  time: 0.5820  data: 0.3801  max mem: 57114
Test: Total time: 0:00:03 (0.6067 s / it)
* Acc@1 76.171 Acc@5 93.608 loss 1.076
Accuracy of the model on the 50000 test images: 76.2%
Max accuracy: 76.84%
Test:  [0/5]  eta: 0:00:11  loss: 8.5396 (8.5396)  acc1: 1.0417 (1.0417)  acc5: 3.1250 (3.1250)  time: 2.3735  data: 2.1301  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.5396 (8.3677)  acc1: 1.0417 (1.1509)  acc5: 3.1250 (2.3018)  time: 0.6253  data: 0.4261  max mem: 57114
Test: Total time: 0:00:03 (0.6395 s / it)
* Acc@1 0.450 Acc@5 1.481 loss 8.298
Accuracy of the model EMA on 50000 test images: 0.4%
Max EMA accuracy: 0.45%
Epoch: [99]  [  0/156]  eta: 0:07:21  lr: 0.006529  min_lr: 0.006529  loss: 3.4267 (3.4267)  weight_decay: 0.0500 (0.0500)  time: 2.8310  data: 2.1777  max mem: 57114
Epoch: [99]  [ 10/156]  eta: 0:02:13  lr: 0.006527  min_lr: 0.006527  loss: 3.4267 (3.1913)  weight_decay: 0.0500 (0.0500)  time: 0.9162  data: 0.1982  max mem: 57114
Epoch: [99]  [ 20/156]  eta: 0:01:52  lr: 0.006525  min_lr: 0.006525  loss: 3.3909 (3.1820)  weight_decay: 0.0500 (0.0500)  time: 0.7270  data: 0.0003  max mem: 57114
Epoch: [99]  [ 30/156]  eta: 0:01:40  lr: 0.006522  min_lr: 0.006522  loss: 2.9604 (3.1351)  weight_decay: 0.0500 (0.0500)  time: 0.7329  data: 0.0004  max mem: 57114
Epoch: [99]  [ 40/156]  eta: 0:01:30  lr: 0.006520  min_lr: 0.006520  loss: 3.1409 (3.2122)  weight_decay: 0.0500 (0.0500)  time: 0.7265  data: 0.0003  max mem: 57114
Epoch: [99]  [ 50/156]  eta: 0:01:21  lr: 0.006518  min_lr: 0.006518  loss: 3.6889 (3.2846)  weight_decay: 0.0500 (0.0500)  time: 0.7248  data: 0.0004  max mem: 57114
Epoch: [99]  [ 60/156]  eta: 0:01:13  lr: 0.006516  min_lr: 0.006516  loss: 3.5372 (3.2992)  weight_decay: 0.0500 (0.0500)  time: 0.7339  data: 0.0004  max mem: 57114
Epoch: [99]  [ 70/156]  eta: 0:01:05  lr: 0.006513  min_lr: 0.006513  loss: 3.5074 (3.3068)  weight_decay: 0.0500 (0.0500)  time: 0.7229  data: 0.0004  max mem: 57114
Epoch: [99]  [ 80/156]  eta: 0:00:56  lr: 0.006511  min_lr: 0.006511  loss: 3.5880 (3.3252)  weight_decay: 0.0500 (0.0500)  time: 0.7079  data: 0.0004  max mem: 57114
Epoch: [99]  [ 90/156]  eta: 0:00:49  lr: 0.006509  min_lr: 0.006509  loss: 3.5476 (3.3210)  weight_decay: 0.0500 (0.0500)  time: 0.7093  data: 0.0004  max mem: 57114
Epoch: [99]  [100/156]  eta: 0:00:41  lr: 0.006507  min_lr: 0.006507  loss: 3.5476 (3.3215)  weight_decay: 0.0500 (0.0500)  time: 0.7172  data: 0.0003  max mem: 57114
Epoch: [99]  [110/156]  eta: 0:00:34  lr: 0.006504  min_lr: 0.006504  loss: 3.4372 (3.3271)  weight_decay: 0.0500 (0.0500)  time: 0.7121  data: 0.0003  max mem: 57114
Epoch: [99]  [120/156]  eta: 0:00:26  lr: 0.006502  min_lr: 0.006502  loss: 3.3509 (3.3229)  weight_decay: 0.0500 (0.0500)  time: 0.7045  data: 0.0004  max mem: 57114
Epoch: [99]  [130/156]  eta: 0:00:19  lr: 0.006500  min_lr: 0.006500  loss: 3.4321 (3.3319)  weight_decay: 0.0500 (0.0500)  time: 0.7055  data: 0.0009  max mem: 57114
Epoch: [99]  [140/156]  eta: 0:00:11  lr: 0.006498  min_lr: 0.006498  loss: 3.5198 (3.3225)  weight_decay: 0.0500 (0.0500)  time: 0.6930  data: 0.0007  max mem: 57114
Epoch: [99]  [150/156]  eta: 0:00:04  lr: 0.006495  min_lr: 0.006495  loss: 3.1098 (3.3101)  weight_decay: 0.0500 (0.0500)  time: 0.6815  data: 0.0001  max mem: 57114
Epoch: [99]  [155/156]  eta: 0:00:00  lr: 0.006494  min_lr: 0.006494  loss: 3.3487 (3.3161)  weight_decay: 0.0500 (0.0500)  time: 0.6895  data: 0.0001  max mem: 57114
Epoch: [99] Total time: 0:01:53 (0.7289 s / it)
Averaged stats: lr: 0.006494  min_lr: 0.006494  loss: 3.3487 (3.2955)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9745 (0.9745)  acc1: 80.2083 (80.2083)  acc5: 96.3542 (96.3542)  time: 2.0671  data: 1.8111  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1142 (1.0643)  acc1: 77.6042 (75.7033)  acc5: 93.2292 (92.0716)  time: 0.5641  data: 0.3623  max mem: 57114
Test: Total time: 0:00:02 (0.5863 s / it)
* Acc@1 75.422 Acc@5 92.829 loss 1.201
Accuracy of the model on the 50000 test images: 75.4%
Max accuracy: 76.84%
Test:  [0/5]  eta: 0:00:11  loss: 8.4626 (8.4626)  acc1: 1.0417 (1.0417)  acc5: 3.1250 (3.1250)  time: 2.2750  data: 2.0316  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.4626 (8.2906)  acc1: 1.0417 (1.1509)  acc5: 3.1250 (2.3018)  time: 0.6056  data: 0.4064  max mem: 57114
Test: Total time: 0:00:03 (0.6184 s / it)
* Acc@1 0.523 Acc@5 1.582 loss 8.234
Accuracy of the model EMA on 50000 test images: 0.5%
Max EMA accuracy: 0.52%
Epoch: [100]  [  0/156]  eta: 0:06:02  lr: 0.006494  min_lr: 0.006494  loss: 3.0842 (3.0842)  weight_decay: 0.0500 (0.0500)  time: 2.3250  data: 1.6733  max mem: 57114
Epoch: [100]  [ 10/156]  eta: 0:02:05  lr: 0.006492  min_lr: 0.006492  loss: 3.4575 (3.2703)  weight_decay: 0.0500 (0.0500)  time: 0.8585  data: 0.1523  max mem: 57114
Epoch: [100]  [ 20/156]  eta: 0:01:47  lr: 0.006490  min_lr: 0.006490  loss: 3.4746 (3.3803)  weight_decay: 0.0500 (0.0500)  time: 0.7140  data: 0.0003  max mem: 57114
Epoch: [100]  [ 30/156]  eta: 0:01:36  lr: 0.006487  min_lr: 0.006487  loss: 3.5336 (3.4268)  weight_decay: 0.0500 (0.0500)  time: 0.7132  data: 0.0003  max mem: 57114
Epoch: [100]  [ 40/156]  eta: 0:01:27  lr: 0.006485  min_lr: 0.006485  loss: 3.3295 (3.3794)  weight_decay: 0.0500 (0.0500)  time: 0.7235  data: 0.0004  max mem: 57114
Epoch: [100]  [ 50/156]  eta: 0:01:19  lr: 0.006483  min_lr: 0.006483  loss: 3.0811 (3.2969)  weight_decay: 0.0500 (0.0500)  time: 0.7284  data: 0.0004  max mem: 57114
Epoch: [100]  [ 60/156]  eta: 0:01:11  lr: 0.006481  min_lr: 0.006481  loss: 3.1776 (3.3177)  weight_decay: 0.0500 (0.0500)  time: 0.7223  data: 0.0004  max mem: 57114
Epoch: [100]  [ 70/156]  eta: 0:01:03  lr: 0.006478  min_lr: 0.006478  loss: 3.4032 (3.3127)  weight_decay: 0.0500 (0.0500)  time: 0.7178  data: 0.0004  max mem: 57114
Epoch: [100]  [ 80/156]  eta: 0:00:56  lr: 0.006476  min_lr: 0.006476  loss: 3.4032 (3.3180)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0004  max mem: 57114
Epoch: [100]  [ 90/156]  eta: 0:00:48  lr: 0.006474  min_lr: 0.006474  loss: 3.1744 (3.2891)  weight_decay: 0.0500 (0.0500)  time: 0.7085  data: 0.0004  max mem: 57114
Epoch: [100]  [100/156]  eta: 0:00:40  lr: 0.006472  min_lr: 0.006472  loss: 3.3496 (3.2867)  weight_decay: 0.0500 (0.0500)  time: 0.7059  data: 0.0004  max mem: 57114
Epoch: [100]  [110/156]  eta: 0:00:33  lr: 0.006469  min_lr: 0.006469  loss: 3.4193 (3.2941)  weight_decay: 0.0500 (0.0500)  time: 0.7074  data: 0.0004  max mem: 57114
Epoch: [100]  [120/156]  eta: 0:00:26  lr: 0.006467  min_lr: 0.006467  loss: 3.4251 (3.3068)  weight_decay: 0.0500 (0.0500)  time: 0.7093  data: 0.0004  max mem: 57114
Epoch: [100]  [130/156]  eta: 0:00:18  lr: 0.006465  min_lr: 0.006465  loss: 3.4193 (3.2902)  weight_decay: 0.0500 (0.0500)  time: 0.7056  data: 0.0009  max mem: 57114
Epoch: [100]  [140/156]  eta: 0:00:11  lr: 0.006463  min_lr: 0.006463  loss: 3.4428 (3.2968)  weight_decay: 0.0500 (0.0500)  time: 0.6939  data: 0.0007  max mem: 57114
Epoch: [100]  [150/156]  eta: 0:00:04  lr: 0.006460  min_lr: 0.006460  loss: 3.5182 (3.3049)  weight_decay: 0.0500 (0.0500)  time: 0.6869  data: 0.0001  max mem: 57114
Epoch: [100]  [155/156]  eta: 0:00:00  lr: 0.006459  min_lr: 0.006459  loss: 3.3813 (3.3024)  weight_decay: 0.0500 (0.0500)  time: 0.6844  data: 0.0001  max mem: 57114
Epoch: [100] Total time: 0:01:52 (0.7213 s / it)
Averaged stats: lr: 0.006459  min_lr: 0.006459  loss: 3.3813 (3.2927)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8786 (0.8786)  acc1: 83.8542 (83.8542)  acc5: 96.3542 (96.3542)  time: 2.1034  data: 1.8468  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9532 (0.9645)  acc1: 78.6458 (76.4706)  acc5: 96.3542 (93.6061)  time: 0.5715  data: 0.3694  max mem: 57114
Test: Total time: 0:00:02 (0.5923 s / it)
* Acc@1 76.592 Acc@5 93.552 loss 1.075
Accuracy of the model on the 50000 test images: 76.6%
Max accuracy: 76.84%
Test:  [0/5]  eta: 0:00:11  loss: 8.3737 (8.3737)  acc1: 1.0417 (1.0417)  acc5: 4.1667 (4.1667)  time: 2.3032  data: 2.0599  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.3737 (8.2013)  acc1: 1.0417 (1.1509)  acc5: 3.1250 (2.5575)  time: 0.6113  data: 0.4121  max mem: 57114
Test: Total time: 0:00:03 (0.6227 s / it)
* Acc@1 0.583 Acc@5 1.702 loss 8.159
Accuracy of the model EMA on 50000 test images: 0.6%
Max EMA accuracy: 0.58%
Epoch: [101]  [  0/156]  eta: 0:07:07  lr: 0.006459  min_lr: 0.006459  loss: 3.1896 (3.1896)  weight_decay: 0.0500 (0.0500)  time: 2.7382  data: 2.0843  max mem: 57114
Epoch: [101]  [ 10/156]  eta: 0:02:10  lr: 0.006457  min_lr: 0.006457  loss: 3.2544 (3.0789)  weight_decay: 0.0500 (0.0500)  time: 0.8915  data: 0.1898  max mem: 57114
Epoch: [101]  [ 20/156]  eta: 0:01:50  lr: 0.006454  min_lr: 0.006454  loss: 3.4935 (3.2709)  weight_decay: 0.0500 (0.0500)  time: 0.7195  data: 0.0004  max mem: 57114
Epoch: [101]  [ 30/156]  eta: 0:01:39  lr: 0.006452  min_lr: 0.006452  loss: 3.4935 (3.2541)  weight_decay: 0.0500 (0.0500)  time: 0.7354  data: 0.0004  max mem: 57114
Epoch: [101]  [ 40/156]  eta: 0:01:29  lr: 0.006450  min_lr: 0.006450  loss: 3.5026 (3.2692)  weight_decay: 0.0500 (0.0500)  time: 0.7278  data: 0.0004  max mem: 57114
Epoch: [101]  [ 50/156]  eta: 0:01:20  lr: 0.006448  min_lr: 0.006448  loss: 3.4286 (3.2625)  weight_decay: 0.0500 (0.0500)  time: 0.7158  data: 0.0004  max mem: 57114
Epoch: [101]  [ 60/156]  eta: 0:01:12  lr: 0.006445  min_lr: 0.006445  loss: 3.3349 (3.2848)  weight_decay: 0.0500 (0.0500)  time: 0.7256  data: 0.0004  max mem: 57114
Epoch: [101]  [ 70/156]  eta: 0:01:04  lr: 0.006443  min_lr: 0.006443  loss: 3.3077 (3.2906)  weight_decay: 0.0500 (0.0500)  time: 0.7187  data: 0.0005  max mem: 57114
Epoch: [101]  [ 80/156]  eta: 0:00:56  lr: 0.006441  min_lr: 0.006441  loss: 3.3148 (3.3066)  weight_decay: 0.0500 (0.0500)  time: 0.6995  data: 0.0005  max mem: 57114
Epoch: [101]  [ 90/156]  eta: 0:00:48  lr: 0.006438  min_lr: 0.006438  loss: 3.4606 (3.3182)  weight_decay: 0.0500 (0.0500)  time: 0.7070  data: 0.0004  max mem: 57114
Epoch: [101]  [100/156]  eta: 0:00:41  lr: 0.006436  min_lr: 0.006436  loss: 3.4117 (3.3105)  weight_decay: 0.0500 (0.0500)  time: 0.7098  data: 0.0004  max mem: 57114
Epoch: [101]  [110/156]  eta: 0:00:33  lr: 0.006434  min_lr: 0.006434  loss: 3.4117 (3.3179)  weight_decay: 0.0500 (0.0500)  time: 0.7067  data: 0.0004  max mem: 57114
Epoch: [101]  [120/156]  eta: 0:00:26  lr: 0.006432  min_lr: 0.006432  loss: 3.5245 (3.3338)  weight_decay: 0.0500 (0.0500)  time: 0.7060  data: 0.0004  max mem: 57114
Epoch: [101]  [130/156]  eta: 0:00:18  lr: 0.006429  min_lr: 0.006429  loss: 3.5563 (3.3378)  weight_decay: 0.0500 (0.0500)  time: 0.6968  data: 0.0008  max mem: 57114
Epoch: [101]  [140/156]  eta: 0:00:11  lr: 0.006427  min_lr: 0.006427  loss: 3.4777 (3.3411)  weight_decay: 0.0500 (0.0500)  time: 0.6873  data: 0.0007  max mem: 57114
Epoch: [101]  [150/156]  eta: 0:00:04  lr: 0.006425  min_lr: 0.006425  loss: 3.4777 (3.3520)  weight_decay: 0.0500 (0.0500)  time: 0.6858  data: 0.0001  max mem: 57114
Epoch: [101]  [155/156]  eta: 0:00:00  lr: 0.006424  min_lr: 0.006424  loss: 3.3110 (3.3334)  weight_decay: 0.0500 (0.0500)  time: 0.6856  data: 0.0001  max mem: 57114
Epoch: [101] Total time: 0:01:52 (0.7241 s / it)
Averaged stats: lr: 0.006424  min_lr: 0.006424  loss: 3.3110 (3.2894)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8585 (0.8585)  acc1: 83.3333 (83.3333)  acc5: 95.8333 (95.8333)  time: 2.0001  data: 1.7443  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0666 (1.0375)  acc1: 77.0833 (75.9591)  acc5: 94.7917 (92.8389)  time: 0.5507  data: 0.3489  max mem: 57114
Test: Total time: 0:00:02 (0.5721 s / it)
* Acc@1 74.946 Acc@5 92.821 loss 1.124
Accuracy of the model on the 50000 test images: 74.9%
Max accuracy: 76.84%
Test:  [0/5]  eta: 0:00:11  loss: 8.2829 (8.2829)  acc1: 1.0417 (1.0417)  acc5: 4.1667 (4.1667)  time: 2.2745  data: 2.0309  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.2829 (8.1102)  acc1: 1.0417 (1.1509)  acc5: 3.1250 (2.5575)  time: 0.6057  data: 0.4063  max mem: 57114
Test: Total time: 0:00:03 (0.6186 s / it)
* Acc@1 0.661 Acc@5 1.868 loss 8.085
Accuracy of the model EMA on 50000 test images: 0.7%
Max EMA accuracy: 0.66%
Epoch: [102]  [  0/156]  eta: 0:07:37  lr: 0.006423  min_lr: 0.006423  loss: 3.6148 (3.6148)  weight_decay: 0.0500 (0.0500)  time: 2.9307  data: 2.2816  max mem: 57114
Epoch: [102]  [ 10/156]  eta: 0:02:16  lr: 0.006421  min_lr: 0.006421  loss: 3.2027 (3.1697)  weight_decay: 0.0500 (0.0500)  time: 0.9316  data: 0.2077  max mem: 57114
Epoch: [102]  [ 20/156]  eta: 0:01:53  lr: 0.006419  min_lr: 0.006419  loss: 2.9268 (2.9921)  weight_decay: 0.0500 (0.0500)  time: 0.7302  data: 0.0003  max mem: 57114
Epoch: [102]  [ 30/156]  eta: 0:01:39  lr: 0.006416  min_lr: 0.006416  loss: 3.1548 (3.0585)  weight_decay: 0.0500 (0.0500)  time: 0.7150  data: 0.0003  max mem: 57114
Epoch: [102]  [ 40/156]  eta: 0:01:29  lr: 0.006414  min_lr: 0.006414  loss: 3.1897 (3.0785)  weight_decay: 0.0500 (0.0500)  time: 0.7079  data: 0.0003  max mem: 57114
Epoch: [102]  [ 50/156]  eta: 0:01:20  lr: 0.006412  min_lr: 0.006412  loss: 3.1848 (3.1309)  weight_decay: 0.0500 (0.0500)  time: 0.7179  data: 0.0004  max mem: 57114
Epoch: [102]  [ 60/156]  eta: 0:01:12  lr: 0.006410  min_lr: 0.006410  loss: 3.4762 (3.1363)  weight_decay: 0.0500 (0.0500)  time: 0.7236  data: 0.0004  max mem: 57114
Epoch: [102]  [ 70/156]  eta: 0:01:04  lr: 0.006407  min_lr: 0.006407  loss: 3.3943 (3.1414)  weight_decay: 0.0500 (0.0500)  time: 0.7230  data: 0.0004  max mem: 57114
Epoch: [102]  [ 80/156]  eta: 0:00:56  lr: 0.006405  min_lr: 0.006405  loss: 3.0738 (3.1338)  weight_decay: 0.0500 (0.0500)  time: 0.7209  data: 0.0004  max mem: 57114
Epoch: [102]  [ 90/156]  eta: 0:00:49  lr: 0.006403  min_lr: 0.006403  loss: 3.2409 (3.1727)  weight_decay: 0.0500 (0.0500)  time: 0.7150  data: 0.0004  max mem: 57114
Epoch: [102]  [100/156]  eta: 0:00:41  lr: 0.006400  min_lr: 0.006400  loss: 3.3236 (3.1641)  weight_decay: 0.0500 (0.0500)  time: 0.7026  data: 0.0004  max mem: 57114
Epoch: [102]  [110/156]  eta: 0:00:33  lr: 0.006398  min_lr: 0.006398  loss: 3.2706 (3.1727)  weight_decay: 0.0500 (0.0500)  time: 0.7014  data: 0.0004  max mem: 57114
Epoch: [102]  [120/156]  eta: 0:00:26  lr: 0.006396  min_lr: 0.006396  loss: 3.2706 (3.1799)  weight_decay: 0.0500 (0.0500)  time: 0.7024  data: 0.0004  max mem: 57114
Epoch: [102]  [130/156]  eta: 0:00:18  lr: 0.006393  min_lr: 0.006393  loss: 3.3320 (3.1912)  weight_decay: 0.0500 (0.0500)  time: 0.6991  data: 0.0006  max mem: 57114
Epoch: [102]  [140/156]  eta: 0:00:11  lr: 0.006391  min_lr: 0.006391  loss: 3.4235 (3.1933)  weight_decay: 0.0500 (0.0500)  time: 0.6902  data: 0.0005  max mem: 57114
Epoch: [102]  [150/156]  eta: 0:00:04  lr: 0.006389  min_lr: 0.006389  loss: 3.1836 (3.1909)  weight_decay: 0.0500 (0.0500)  time: 0.6809  data: 0.0001  max mem: 57114
Epoch: [102]  [155/156]  eta: 0:00:00  lr: 0.006388  min_lr: 0.006388  loss: 3.4235 (3.1977)  weight_decay: 0.0500 (0.0500)  time: 0.6830  data: 0.0001  max mem: 57114
Epoch: [102] Total time: 0:01:53 (0.7245 s / it)
Averaged stats: lr: 0.006388  min_lr: 0.006388  loss: 3.4235 (3.2955)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9536 (0.9536)  acc1: 82.2917 (82.2917)  acc5: 97.3958 (97.3958)  time: 2.0626  data: 1.8072  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1188 (1.0703)  acc1: 77.6042 (76.3427)  acc5: 95.8333 (94.3734)  time: 0.5633  data: 0.3615  max mem: 57114
Test: Total time: 0:00:02 (0.5850 s / it)
* Acc@1 76.363 Acc@5 93.446 loss 1.151
Accuracy of the model on the 50000 test images: 76.4%
Max accuracy: 76.84%
Test:  [0/5]  eta: 0:00:11  loss: 8.1945 (8.1945)  acc1: 1.5625 (1.5625)  acc5: 4.1667 (4.1667)  time: 2.3866  data: 2.1431  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.1945 (8.0177)  acc1: 1.5625 (1.2788)  acc5: 3.1250 (2.8133)  time: 0.6280  data: 0.4287  max mem: 57114
Test: Total time: 0:00:03 (0.6406 s / it)
* Acc@1 0.733 Acc@5 2.032 loss 8.014
Accuracy of the model EMA on 50000 test images: 0.7%
Max EMA accuracy: 0.73%
Epoch: [103]  [  0/156]  eta: 0:06:22  lr: 0.006388  min_lr: 0.006388  loss: 2.3485 (2.3485)  weight_decay: 0.0500 (0.0500)  time: 2.4530  data: 1.8027  max mem: 57114
Epoch: [103]  [ 10/156]  eta: 0:02:11  lr: 0.006385  min_lr: 0.006385  loss: 3.4534 (3.3692)  weight_decay: 0.0500 (0.0500)  time: 0.9002  data: 0.1642  max mem: 57114
Epoch: [103]  [ 20/156]  eta: 0:01:49  lr: 0.006383  min_lr: 0.006383  loss: 3.4710 (3.3424)  weight_decay: 0.0500 (0.0500)  time: 0.7226  data: 0.0003  max mem: 57114
Epoch: [103]  [ 30/156]  eta: 0:01:38  lr: 0.006381  min_lr: 0.006381  loss: 3.3585 (3.2857)  weight_decay: 0.0500 (0.0500)  time: 0.7148  data: 0.0003  max mem: 57114
Epoch: [103]  [ 40/156]  eta: 0:01:28  lr: 0.006378  min_lr: 0.006378  loss: 3.2629 (3.2705)  weight_decay: 0.0500 (0.0500)  time: 0.7179  data: 0.0004  max mem: 57114
Epoch: [103]  [ 50/156]  eta: 0:01:20  lr: 0.006376  min_lr: 0.006376  loss: 3.1898 (3.2800)  weight_decay: 0.0500 (0.0500)  time: 0.7221  data: 0.0004  max mem: 57114
Epoch: [103]  [ 60/156]  eta: 0:01:12  lr: 0.006374  min_lr: 0.006374  loss: 3.3343 (3.2796)  weight_decay: 0.0500 (0.0500)  time: 0.7320  data: 0.0004  max mem: 57114
Epoch: [103]  [ 70/156]  eta: 0:01:04  lr: 0.006371  min_lr: 0.006371  loss: 3.5966 (3.3192)  weight_decay: 0.0500 (0.0500)  time: 0.7193  data: 0.0004  max mem: 57114
Epoch: [103]  [ 80/156]  eta: 0:00:56  lr: 0.006369  min_lr: 0.006369  loss: 3.6276 (3.3393)  weight_decay: 0.0500 (0.0500)  time: 0.7123  data: 0.0004  max mem: 57114
Epoch: [103]  [ 90/156]  eta: 0:00:48  lr: 0.006367  min_lr: 0.006367  loss: 3.4978 (3.3449)  weight_decay: 0.0500 (0.0500)  time: 0.7105  data: 0.0004  max mem: 57114
Epoch: [103]  [100/156]  eta: 0:00:41  lr: 0.006364  min_lr: 0.006364  loss: 3.3906 (3.3331)  weight_decay: 0.0500 (0.0500)  time: 0.7007  data: 0.0003  max mem: 57114
Epoch: [103]  [110/156]  eta: 0:00:33  lr: 0.006362  min_lr: 0.006362  loss: 3.2835 (3.3018)  weight_decay: 0.0500 (0.0500)  time: 0.6918  data: 0.0004  max mem: 57114
Epoch: [103]  [120/156]  eta: 0:00:26  lr: 0.006360  min_lr: 0.006360  loss: 3.0972 (3.2892)  weight_decay: 0.0500 (0.0500)  time: 0.7055  data: 0.0004  max mem: 57114
Epoch: [103]  [130/156]  eta: 0:00:18  lr: 0.006357  min_lr: 0.006357  loss: 3.2277 (3.2979)  weight_decay: 0.0500 (0.0500)  time: 0.7056  data: 0.0009  max mem: 57114
Epoch: [103]  [140/156]  eta: 0:00:11  lr: 0.006355  min_lr: 0.006355  loss: 3.6596 (3.3239)  weight_decay: 0.0500 (0.0500)  time: 0.6885  data: 0.0007  max mem: 57114
Epoch: [103]  [150/156]  eta: 0:00:04  lr: 0.006353  min_lr: 0.006353  loss: 3.5220 (3.3053)  weight_decay: 0.0500 (0.0500)  time: 0.6840  data: 0.0001  max mem: 57114
Epoch: [103]  [155/156]  eta: 0:00:00  lr: 0.006352  min_lr: 0.006352  loss: 3.3954 (3.3100)  weight_decay: 0.0500 (0.0500)  time: 0.6831  data: 0.0001  max mem: 57114
Epoch: [103] Total time: 0:01:52 (0.7221 s / it)
Averaged stats: lr: 0.006352  min_lr: 0.006352  loss: 3.3954 (3.2883)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8746 (0.8746)  acc1: 83.3333 (83.3333)  acc5: 96.3542 (96.3542)  time: 2.0476  data: 1.7914  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0551 (1.0379)  acc1: 75.5208 (75.9591)  acc5: 95.8333 (93.4783)  time: 0.5603  data: 0.3584  max mem: 57114
Test: Total time: 0:00:02 (0.5786 s / it)
* Acc@1 76.457 Acc@5 93.632 loss 1.109
Accuracy of the model on the 50000 test images: 76.5%
Max accuracy: 76.84%
Test:  [0/5]  eta: 0:00:11  loss: 8.0971 (8.0971)  acc1: 1.5625 (1.5625)  acc5: 4.6875 (4.6875)  time: 2.3024  data: 2.0588  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 8.0971 (7.9169)  acc1: 1.5625 (1.4066)  acc5: 3.1250 (3.4527)  time: 0.6111  data: 0.4118  max mem: 57114
Test: Total time: 0:00:03 (0.6206 s / it)
* Acc@1 0.825 Acc@5 2.322 loss 7.939
Accuracy of the model EMA on 50000 test images: 0.8%
Max EMA accuracy: 0.83%
Epoch: [104]  [  0/156]  eta: 0:06:58  lr: 0.006351  min_lr: 0.006351  loss: 3.4217 (3.4217)  weight_decay: 0.0500 (0.0500)  time: 2.6807  data: 2.0314  max mem: 57114
Epoch: [104]  [ 10/156]  eta: 0:02:07  lr: 0.006349  min_lr: 0.006349  loss: 3.3741 (3.2625)  weight_decay: 0.0500 (0.0500)  time: 0.8755  data: 0.1850  max mem: 57114
Epoch: [104]  [ 20/156]  eta: 0:01:49  lr: 0.006347  min_lr: 0.006347  loss: 3.2843 (3.2167)  weight_decay: 0.0500 (0.0500)  time: 0.7119  data: 0.0004  max mem: 57114
Epoch: [104]  [ 30/156]  eta: 0:01:38  lr: 0.006344  min_lr: 0.006344  loss: 2.9149 (3.1418)  weight_decay: 0.0500 (0.0500)  time: 0.7261  data: 0.0003  max mem: 57114
Epoch: [104]  [ 40/156]  eta: 0:01:29  lr: 0.006342  min_lr: 0.006342  loss: 3.0900 (3.1957)  weight_decay: 0.0500 (0.0500)  time: 0.7291  data: 0.0004  max mem: 57114
Epoch: [104]  [ 50/156]  eta: 0:01:20  lr: 0.006340  min_lr: 0.006340  loss: 3.3372 (3.1644)  weight_decay: 0.0500 (0.0500)  time: 0.7282  data: 0.0004  max mem: 57114
Epoch: [104]  [ 60/156]  eta: 0:01:12  lr: 0.006337  min_lr: 0.006337  loss: 3.3372 (3.2028)  weight_decay: 0.0500 (0.0500)  time: 0.7181  data: 0.0004  max mem: 57114
Epoch: [104]  [ 70/156]  eta: 0:01:04  lr: 0.006335  min_lr: 0.006335  loss: 3.2609 (3.1947)  weight_decay: 0.0500 (0.0500)  time: 0.7191  data: 0.0004  max mem: 57114
Epoch: [104]  [ 80/156]  eta: 0:00:56  lr: 0.006333  min_lr: 0.006333  loss: 3.0610 (3.1897)  weight_decay: 0.0500 (0.0500)  time: 0.7124  data: 0.0004  max mem: 57114
Epoch: [104]  [ 90/156]  eta: 0:00:48  lr: 0.006330  min_lr: 0.006330  loss: 3.3641 (3.2058)  weight_decay: 0.0500 (0.0500)  time: 0.7167  data: 0.0004  max mem: 57114
Epoch: [104]  [100/156]  eta: 0:00:41  lr: 0.006328  min_lr: 0.006328  loss: 3.3474 (3.2130)  weight_decay: 0.0500 (0.0500)  time: 0.7201  data: 0.0004  max mem: 57114
Epoch: [104]  [110/156]  eta: 0:00:33  lr: 0.006326  min_lr: 0.006326  loss: 3.3874 (3.2205)  weight_decay: 0.0500 (0.0500)  time: 0.7081  data: 0.0004  max mem: 57114
Epoch: [104]  [120/156]  eta: 0:00:26  lr: 0.006323  min_lr: 0.006323  loss: 3.4106 (3.2153)  weight_decay: 0.0500 (0.0500)  time: 0.7093  data: 0.0003  max mem: 57114
Epoch: [104]  [130/156]  eta: 0:00:18  lr: 0.006321  min_lr: 0.006321  loss: 3.1572 (3.2054)  weight_decay: 0.0500 (0.0500)  time: 0.7014  data: 0.0008  max mem: 57114
Epoch: [104]  [140/156]  eta: 0:00:11  lr: 0.006319  min_lr: 0.006319  loss: 3.1905 (3.2135)  weight_decay: 0.0500 (0.0500)  time: 0.6877  data: 0.0007  max mem: 57114
Epoch: [104]  [150/156]  eta: 0:00:04  lr: 0.006316  min_lr: 0.006316  loss: 3.4174 (3.2195)  weight_decay: 0.0500 (0.0500)  time: 0.6837  data: 0.0001  max mem: 57114
Epoch: [104]  [155/156]  eta: 0:00:00  lr: 0.006315  min_lr: 0.006315  loss: 3.4560 (3.2249)  weight_decay: 0.0500 (0.0500)  time: 0.6897  data: 0.0001  max mem: 57114
Epoch: [104] Total time: 0:01:53 (0.7256 s / it)
Averaged stats: lr: 0.006315  min_lr: 0.006315  loss: 3.4560 (3.2794)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9425 (0.9425)  acc1: 82.2917 (82.2917)  acc5: 96.3542 (96.3542)  time: 2.0760  data: 1.8205  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0973 (1.0892)  acc1: 80.2083 (76.8542)  acc5: 96.3542 (93.2225)  time: 0.5658  data: 0.3642  max mem: 57114
Test: Total time: 0:00:02 (0.5897 s / it)
* Acc@1 76.037 Acc@5 93.203 loss 1.152
Accuracy of the model on the 50000 test images: 76.0%
Max accuracy: 76.84%
Test:  [0/5]  eta: 0:00:11  loss: 7.9901 (7.9901)  acc1: 3.1250 (3.1250)  acc5: 4.6875 (4.6875)  time: 2.3381  data: 2.0947  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.9901 (7.8051)  acc1: 2.0833 (1.9182)  acc5: 3.6458 (3.8363)  time: 0.6182  data: 0.4190  max mem: 57114
Test: Total time: 0:00:03 (0.6353 s / it)
* Acc@1 0.957 Acc@5 2.619 loss 7.861
Accuracy of the model EMA on 50000 test images: 1.0%
Max EMA accuracy: 0.96%
Epoch: [105]  [  0/156]  eta: 0:07:20  lr: 0.006315  min_lr: 0.006315  loss: 2.1503 (2.1503)  weight_decay: 0.0500 (0.0500)  time: 2.8211  data: 2.1625  max mem: 57114
Epoch: [105]  [ 10/156]  eta: 0:02:13  lr: 0.006313  min_lr: 0.006313  loss: 3.2903 (3.1135)  weight_decay: 0.0500 (0.0500)  time: 0.9125  data: 0.1968  max mem: 57114
Epoch: [105]  [ 20/156]  eta: 0:01:51  lr: 0.006310  min_lr: 0.006310  loss: 3.2834 (3.0990)  weight_decay: 0.0500 (0.0500)  time: 0.7169  data: 0.0003  max mem: 57114
Epoch: [105]  [ 30/156]  eta: 0:01:39  lr: 0.006308  min_lr: 0.006308  loss: 3.2782 (3.1636)  weight_decay: 0.0500 (0.0500)  time: 0.7203  data: 0.0003  max mem: 57114
Epoch: [105]  [ 40/156]  eta: 0:01:30  lr: 0.006306  min_lr: 0.006306  loss: 3.1726 (3.1594)  weight_decay: 0.0500 (0.0500)  time: 0.7339  data: 0.0003  max mem: 57114
Epoch: [105]  [ 50/156]  eta: 0:01:20  lr: 0.006303  min_lr: 0.006303  loss: 3.1829 (3.1954)  weight_decay: 0.0500 (0.0500)  time: 0.7163  data: 0.0004  max mem: 57114
Epoch: [105]  [ 60/156]  eta: 0:01:12  lr: 0.006301  min_lr: 0.006301  loss: 3.4861 (3.2450)  weight_decay: 0.0500 (0.0500)  time: 0.7039  data: 0.0004  max mem: 57114
Epoch: [105]  [ 70/156]  eta: 0:01:04  lr: 0.006298  min_lr: 0.006298  loss: 3.4443 (3.2456)  weight_decay: 0.0500 (0.0500)  time: 0.7261  data: 0.0003  max mem: 57114
Epoch: [105]  [ 80/156]  eta: 0:00:56  lr: 0.006296  min_lr: 0.006296  loss: 3.2428 (3.2367)  weight_decay: 0.0500 (0.0500)  time: 0.7212  data: 0.0004  max mem: 57114
Epoch: [105]  [ 90/156]  eta: 0:00:48  lr: 0.006294  min_lr: 0.006294  loss: 3.2428 (3.2345)  weight_decay: 0.0500 (0.0500)  time: 0.7108  data: 0.0004  max mem: 57114
Epoch: [105]  [100/156]  eta: 0:00:41  lr: 0.006291  min_lr: 0.006291  loss: 3.2994 (3.2444)  weight_decay: 0.0500 (0.0500)  time: 0.7099  data: 0.0004  max mem: 57114
Epoch: [105]  [110/156]  eta: 0:00:33  lr: 0.006289  min_lr: 0.006289  loss: 3.3244 (3.2421)  weight_decay: 0.0500 (0.0500)  time: 0.7014  data: 0.0004  max mem: 57114
Epoch: [105]  [120/156]  eta: 0:00:26  lr: 0.006287  min_lr: 0.006287  loss: 3.4886 (3.2680)  weight_decay: 0.0500 (0.0500)  time: 0.7003  data: 0.0004  max mem: 57114
Epoch: [105]  [130/156]  eta: 0:00:18  lr: 0.006284  min_lr: 0.006284  loss: 3.5186 (3.2793)  weight_decay: 0.0500 (0.0500)  time: 0.7036  data: 0.0009  max mem: 57114
Epoch: [105]  [140/156]  eta: 0:00:11  lr: 0.006282  min_lr: 0.006282  loss: 3.4790 (3.2816)  weight_decay: 0.0500 (0.0500)  time: 0.6957  data: 0.0007  max mem: 57114
Epoch: [105]  [150/156]  eta: 0:00:04  lr: 0.006280  min_lr: 0.006280  loss: 3.2522 (3.2720)  weight_decay: 0.0500 (0.0500)  time: 0.6805  data: 0.0001  max mem: 57114
Epoch: [105]  [155/156]  eta: 0:00:00  lr: 0.006278  min_lr: 0.006278  loss: 3.0658 (3.2625)  weight_decay: 0.0500 (0.0500)  time: 0.6841  data: 0.0001  max mem: 57114
Epoch: [105] Total time: 0:01:53 (0.7248 s / it)
Averaged stats: lr: 0.006278  min_lr: 0.006278  loss: 3.0658 (3.2804)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8134 (0.8134)  acc1: 85.9375 (85.9375)  acc5: 96.8750 (96.8750)  time: 2.0082  data: 1.7524  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9374 (0.8985)  acc1: 79.6875 (78.7724)  acc5: 95.8333 (93.8619)  time: 0.5524  data: 0.3506  max mem: 57114
Test: Total time: 0:00:02 (0.5727 s / it)
* Acc@1 76.590 Acc@5 93.700 loss 1.032
Accuracy of the model on the 50000 test images: 76.6%
Max accuracy: 76.84%
Test:  [0/5]  eta: 0:00:11  loss: 7.8365 (7.8365)  acc1: 3.6458 (3.6458)  acc5: 5.2083 (5.2083)  time: 2.2916  data: 2.0481  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.8365 (7.6530)  acc1: 2.0833 (2.0460)  acc5: 4.1667 (4.7315)  time: 0.6090  data: 0.4097  max mem: 57114
Test: Total time: 0:00:03 (0.6186 s / it)
* Acc@1 1.109 Acc@5 3.033 loss 7.752
Accuracy of the model EMA on 50000 test images: 1.1%
Max EMA accuracy: 1.11%
Epoch: [106]  [  0/156]  eta: 0:06:35  lr: 0.006278  min_lr: 0.006278  loss: 3.4170 (3.4170)  weight_decay: 0.0500 (0.0500)  time: 2.5372  data: 1.8812  max mem: 57114
Epoch: [106]  [ 10/156]  eta: 0:02:09  lr: 0.006276  min_lr: 0.006276  loss: 3.3510 (3.1935)  weight_decay: 0.0500 (0.0500)  time: 0.8878  data: 0.1713  max mem: 57114
Epoch: [106]  [ 20/156]  eta: 0:01:48  lr: 0.006273  min_lr: 0.006273  loss: 3.3145 (3.1094)  weight_decay: 0.0500 (0.0500)  time: 0.7123  data: 0.0003  max mem: 57114
Epoch: [106]  [ 30/156]  eta: 0:01:37  lr: 0.006271  min_lr: 0.006271  loss: 3.3145 (3.1766)  weight_decay: 0.0500 (0.0500)  time: 0.7162  data: 0.0004  max mem: 57114
Epoch: [106]  [ 40/156]  eta: 0:01:28  lr: 0.006269  min_lr: 0.006269  loss: 3.3736 (3.2061)  weight_decay: 0.0500 (0.0500)  time: 0.7303  data: 0.0004  max mem: 57114
Epoch: [106]  [ 50/156]  eta: 0:01:20  lr: 0.006266  min_lr: 0.006266  loss: 3.4092 (3.2741)  weight_decay: 0.0500 (0.0500)  time: 0.7209  data: 0.0004  max mem: 57114
Epoch: [106]  [ 60/156]  eta: 0:01:12  lr: 0.006264  min_lr: 0.006264  loss: 3.4455 (3.2418)  weight_decay: 0.0500 (0.0500)  time: 0.7240  data: 0.0004  max mem: 57114
Epoch: [106]  [ 70/156]  eta: 0:01:04  lr: 0.006262  min_lr: 0.006262  loss: 3.2877 (3.2467)  weight_decay: 0.0500 (0.0500)  time: 0.7212  data: 0.0004  max mem: 57114
Epoch: [106]  [ 80/156]  eta: 0:00:56  lr: 0.006259  min_lr: 0.006259  loss: 3.3360 (3.2462)  weight_decay: 0.0500 (0.0500)  time: 0.7105  data: 0.0004  max mem: 57114
Epoch: [106]  [ 90/156]  eta: 0:00:48  lr: 0.006257  min_lr: 0.006257  loss: 3.3562 (3.2428)  weight_decay: 0.0500 (0.0500)  time: 0.7068  data: 0.0004  max mem: 57114
Epoch: [106]  [100/156]  eta: 0:00:41  lr: 0.006254  min_lr: 0.006254  loss: 3.2761 (3.2380)  weight_decay: 0.0500 (0.0500)  time: 0.7051  data: 0.0004  max mem: 57114
Epoch: [106]  [110/156]  eta: 0:00:33  lr: 0.006252  min_lr: 0.006252  loss: 3.4701 (3.2531)  weight_decay: 0.0500 (0.0500)  time: 0.7029  data: 0.0004  max mem: 57114
Epoch: [106]  [120/156]  eta: 0:00:26  lr: 0.006250  min_lr: 0.006250  loss: 3.4539 (3.2462)  weight_decay: 0.0500 (0.0500)  time: 0.7008  data: 0.0004  max mem: 57114
Epoch: [106]  [130/156]  eta: 0:00:18  lr: 0.006247  min_lr: 0.006247  loss: 3.4307 (3.2435)  weight_decay: 0.0500 (0.0500)  time: 0.7028  data: 0.0008  max mem: 57114
Epoch: [106]  [140/156]  eta: 0:00:11  lr: 0.006245  min_lr: 0.006245  loss: 3.2321 (3.2412)  weight_decay: 0.0500 (0.0500)  time: 0.6912  data: 0.0007  max mem: 57114
Epoch: [106]  [150/156]  eta: 0:00:04  lr: 0.006243  min_lr: 0.006243  loss: 3.2321 (3.2458)  weight_decay: 0.0500 (0.0500)  time: 0.6826  data: 0.0001  max mem: 57114
Epoch: [106]  [155/156]  eta: 0:00:00  lr: 0.006241  min_lr: 0.006241  loss: 3.2088 (3.2481)  weight_decay: 0.0500 (0.0500)  time: 0.6841  data: 0.0001  max mem: 57114
Epoch: [106] Total time: 0:01:52 (0.7216 s / it)
Averaged stats: lr: 0.006241  min_lr: 0.006241  loss: 3.2088 (3.2746)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9000 (0.9000)  acc1: 81.7708 (81.7708)  acc5: 96.3542 (96.3542)  time: 2.0307  data: 1.7749  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0351 (1.0091)  acc1: 79.6875 (78.0051)  acc5: 94.7917 (93.8619)  time: 0.5571  data: 0.3551  max mem: 57114
Test: Total time: 0:00:02 (0.5791 s / it)
* Acc@1 76.281 Acc@5 93.336 loss 1.108
Accuracy of the model on the 50000 test images: 76.3%
Max accuracy: 76.84%
Test:  [0/5]  eta: 0:00:11  loss: 7.6551 (7.6551)  acc1: 3.6458 (3.6458)  acc5: 5.2083 (5.2083)  time: 2.2604  data: 2.0170  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.6551 (7.4754)  acc1: 2.0833 (2.0460)  acc5: 5.2083 (5.6266)  time: 0.6027  data: 0.4035  max mem: 57114
Test: Total time: 0:00:03 (0.6196 s / it)
* Acc@1 1.301 Acc@5 3.581 loss 7.624
Accuracy of the model EMA on 50000 test images: 1.3%
Max EMA accuracy: 1.30%
Epoch: [107]  [  0/156]  eta: 0:06:05  lr: 0.006241  min_lr: 0.006241  loss: 3.7543 (3.7543)  weight_decay: 0.0500 (0.0500)  time: 2.3418  data: 1.6898  max mem: 57114
Epoch: [107]  [ 10/156]  eta: 0:02:05  lr: 0.006239  min_lr: 0.006239  loss: 3.5658 (3.4602)  weight_decay: 0.0500 (0.0500)  time: 0.8600  data: 0.1539  max mem: 57114
Epoch: [107]  [ 20/156]  eta: 0:01:47  lr: 0.006236  min_lr: 0.006236  loss: 3.5086 (3.4582)  weight_decay: 0.0500 (0.0500)  time: 0.7097  data: 0.0003  max mem: 57114
Epoch: [107]  [ 30/156]  eta: 0:01:36  lr: 0.006234  min_lr: 0.006234  loss: 3.2714 (3.3086)  weight_decay: 0.0500 (0.0500)  time: 0.7128  data: 0.0004  max mem: 57114
Epoch: [107]  [ 40/156]  eta: 0:01:27  lr: 0.006232  min_lr: 0.006232  loss: 3.1396 (3.2780)  weight_decay: 0.0500 (0.0500)  time: 0.7238  data: 0.0004  max mem: 57114
Epoch: [107]  [ 50/156]  eta: 0:01:19  lr: 0.006229  min_lr: 0.006229  loss: 3.4153 (3.2867)  weight_decay: 0.0500 (0.0500)  time: 0.7280  data: 0.0004  max mem: 57114
Epoch: [107]  [ 60/156]  eta: 0:01:11  lr: 0.006227  min_lr: 0.006227  loss: 3.4746 (3.3208)  weight_decay: 0.0500 (0.0500)  time: 0.7288  data: 0.0004  max mem: 57114
Epoch: [107]  [ 70/156]  eta: 0:01:03  lr: 0.006224  min_lr: 0.006224  loss: 3.5757 (3.3420)  weight_decay: 0.0500 (0.0500)  time: 0.7212  data: 0.0004  max mem: 57114
Epoch: [107]  [ 80/156]  eta: 0:00:56  lr: 0.006222  min_lr: 0.006222  loss: 3.5882 (3.3740)  weight_decay: 0.0500 (0.0500)  time: 0.7104  data: 0.0004  max mem: 57114
Epoch: [107]  [ 90/156]  eta: 0:00:48  lr: 0.006220  min_lr: 0.006220  loss: 3.4255 (3.3549)  weight_decay: 0.0500 (0.0500)  time: 0.7052  data: 0.0004  max mem: 57114
Epoch: [107]  [100/156]  eta: 0:00:40  lr: 0.006217  min_lr: 0.006217  loss: 3.1157 (3.3144)  weight_decay: 0.0500 (0.0500)  time: 0.7073  data: 0.0004  max mem: 57114
Epoch: [107]  [110/156]  eta: 0:00:33  lr: 0.006215  min_lr: 0.006215  loss: 3.1157 (3.3078)  weight_decay: 0.0500 (0.0500)  time: 0.7177  data: 0.0004  max mem: 57114
Epoch: [107]  [120/156]  eta: 0:00:26  lr: 0.006212  min_lr: 0.006212  loss: 3.4430 (3.2929)  weight_decay: 0.0500 (0.0500)  time: 0.7175  data: 0.0004  max mem: 57114
Epoch: [107]  [130/156]  eta: 0:00:18  lr: 0.006210  min_lr: 0.006210  loss: 3.3331 (3.2889)  weight_decay: 0.0500 (0.0500)  time: 0.6993  data: 0.0010  max mem: 57114
Epoch: [107]  [140/156]  eta: 0:00:11  lr: 0.006208  min_lr: 0.006208  loss: 3.4476 (3.3051)  weight_decay: 0.0500 (0.0500)  time: 0.6805  data: 0.0009  max mem: 57114
Epoch: [107]  [150/156]  eta: 0:00:04  lr: 0.006205  min_lr: 0.006205  loss: 3.4476 (3.3148)  weight_decay: 0.0500 (0.0500)  time: 0.6793  data: 0.0001  max mem: 57114
Epoch: [107]  [155/156]  eta: 0:00:00  lr: 0.006204  min_lr: 0.006204  loss: 3.4326 (3.3169)  weight_decay: 0.0500 (0.0500)  time: 0.6847  data: 0.0001  max mem: 57114
Epoch: [107] Total time: 0:01:52 (0.7212 s / it)
Averaged stats: lr: 0.006204  min_lr: 0.006204  loss: 3.4326 (3.2650)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0247 (1.0247)  acc1: 83.8542 (83.8542)  acc5: 96.8750 (96.8750)  time: 2.0200  data: 1.7642  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1218 (1.0633)  acc1: 76.5625 (77.8772)  acc5: 95.3125 (93.8619)  time: 0.5547  data: 0.3529  max mem: 57114
Test: Total time: 0:00:02 (0.5757 s / it)
* Acc@1 75.975 Acc@5 93.370 loss 1.170
Accuracy of the model on the 50000 test images: 76.0%
Max accuracy: 76.84%
Test:  [0/5]  eta: 0:00:11  loss: 7.4317 (7.4317)  acc1: 3.6458 (3.6458)  acc5: 6.2500 (6.2500)  time: 2.2867  data: 2.0431  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.4317 (7.2645)  acc1: 2.0833 (2.0460)  acc5: 6.2500 (6.2660)  time: 0.6080  data: 0.4087  max mem: 57114
Test: Total time: 0:00:03 (0.6225 s / it)
* Acc@1 1.561 Acc@5 4.204 loss 7.471
Accuracy of the model EMA on 50000 test images: 1.6%
Max EMA accuracy: 1.56%
Epoch: [108]  [  0/156]  eta: 0:07:35  lr: 0.006204  min_lr: 0.006204  loss: 4.0741 (4.0741)  weight_decay: 0.0500 (0.0500)  time: 2.9190  data: 2.2690  max mem: 57114
Epoch: [108]  [ 10/156]  eta: 0:02:15  lr: 0.006201  min_lr: 0.006201  loss: 3.2684 (3.1007)  weight_decay: 0.0500 (0.0500)  time: 0.9268  data: 0.2066  max mem: 57114
Epoch: [108]  [ 20/156]  eta: 0:01:53  lr: 0.006199  min_lr: 0.006199  loss: 3.3218 (3.2556)  weight_decay: 0.0500 (0.0500)  time: 0.7296  data: 0.0003  max mem: 57114
Epoch: [108]  [ 30/156]  eta: 0:01:41  lr: 0.006197  min_lr: 0.006197  loss: 3.4960 (3.2693)  weight_decay: 0.0500 (0.0500)  time: 0.7392  data: 0.0003  max mem: 57114
Epoch: [108]  [ 40/156]  eta: 0:01:30  lr: 0.006194  min_lr: 0.006194  loss: 3.3199 (3.2434)  weight_decay: 0.0500 (0.0500)  time: 0.7277  data: 0.0004  max mem: 57114
Epoch: [108]  [ 50/156]  eta: 0:01:21  lr: 0.006192  min_lr: 0.006192  loss: 3.2749 (3.2470)  weight_decay: 0.0500 (0.0500)  time: 0.7192  data: 0.0004  max mem: 57114
Epoch: [108]  [ 60/156]  eta: 0:01:13  lr: 0.006189  min_lr: 0.006189  loss: 3.1098 (3.2104)  weight_decay: 0.0500 (0.0500)  time: 0.7206  data: 0.0004  max mem: 57114
Epoch: [108]  [ 70/156]  eta: 0:01:05  lr: 0.006187  min_lr: 0.006187  loss: 3.4582 (3.2507)  weight_decay: 0.0500 (0.0500)  time: 0.7201  data: 0.0004  max mem: 57114
Epoch: [108]  [ 80/156]  eta: 0:00:57  lr: 0.006185  min_lr: 0.006185  loss: 3.4654 (3.2283)  weight_decay: 0.0500 (0.0500)  time: 0.7172  data: 0.0004  max mem: 57114
Epoch: [108]  [ 90/156]  eta: 0:00:49  lr: 0.006182  min_lr: 0.006182  loss: 2.9808 (3.2128)  weight_decay: 0.0500 (0.0500)  time: 0.7112  data: 0.0004  max mem: 57114
Epoch: [108]  [100/156]  eta: 0:00:41  lr: 0.006180  min_lr: 0.006180  loss: 3.2976 (3.2323)  weight_decay: 0.0500 (0.0500)  time: 0.7122  data: 0.0003  max mem: 57114
Epoch: [108]  [110/156]  eta: 0:00:34  lr: 0.006177  min_lr: 0.006177  loss: 3.4667 (3.2443)  weight_decay: 0.0500 (0.0500)  time: 0.7107  data: 0.0003  max mem: 57114
Epoch: [108]  [120/156]  eta: 0:00:26  lr: 0.006175  min_lr: 0.006175  loss: 3.4051 (3.2340)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0004  max mem: 57114
Epoch: [108]  [130/156]  eta: 0:00:19  lr: 0.006173  min_lr: 0.006173  loss: 3.4143 (3.2402)  weight_decay: 0.0500 (0.0500)  time: 0.7002  data: 0.0008  max mem: 57114
Epoch: [108]  [140/156]  eta: 0:00:11  lr: 0.006170  min_lr: 0.006170  loss: 3.3509 (3.2252)  weight_decay: 0.0500 (0.0500)  time: 0.6884  data: 0.0007  max mem: 57114
Epoch: [108]  [150/156]  eta: 0:00:04  lr: 0.006168  min_lr: 0.006168  loss: 3.3579 (3.2318)  weight_decay: 0.0500 (0.0500)  time: 0.6825  data: 0.0002  max mem: 57114
Epoch: [108]  [155/156]  eta: 0:00:00  lr: 0.006166  min_lr: 0.006166  loss: 3.3675 (3.2400)  weight_decay: 0.0500 (0.0500)  time: 0.6906  data: 0.0002  max mem: 57114
Epoch: [108] Total time: 0:01:53 (0.7290 s / it)
Averaged stats: lr: 0.006166  min_lr: 0.006166  loss: 3.3675 (3.2764)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8376 (0.8376)  acc1: 84.8958 (84.8958)  acc5: 96.8750 (96.8750)  time: 2.0755  data: 1.8196  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9625 (0.9623)  acc1: 81.2500 (77.8772)  acc5: 94.7917 (93.9898)  time: 0.5658  data: 0.3640  max mem: 57114
Test: Total time: 0:00:02 (0.5879 s / it)
* Acc@1 76.557 Acc@5 93.590 loss 1.074
Accuracy of the model on the 50000 test images: 76.6%
Max accuracy: 76.84%
Test:  [0/5]  eta: 0:00:11  loss: 7.1809 (7.1809)  acc1: 4.1667 (4.1667)  acc5: 7.8125 (7.8125)  time: 2.2613  data: 2.0178  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 7.1809 (7.0319)  acc1: 2.6042 (2.8133)  acc5: 7.8125 (7.8005)  time: 0.6029  data: 0.4036  max mem: 57114
Test: Total time: 0:00:03 (0.6146 s / it)
* Acc@1 1.886 Acc@5 5.075 loss 7.300
Accuracy of the model EMA on 50000 test images: 1.9%
Max EMA accuracy: 1.89%
Epoch: [109]  [  0/156]  eta: 0:05:27  lr: 0.006166  min_lr: 0.006166  loss: 3.4467 (3.4467)  weight_decay: 0.0500 (0.0500)  time: 2.1002  data: 1.4464  max mem: 57114
Epoch: [109]  [ 10/156]  eta: 0:02:03  lr: 0.006164  min_lr: 0.006164  loss: 3.3674 (3.0996)  weight_decay: 0.0500 (0.0500)  time: 0.8465  data: 0.1317  max mem: 57114
Epoch: [109]  [ 20/156]  eta: 0:01:47  lr: 0.006161  min_lr: 0.006161  loss: 2.9836 (3.0884)  weight_decay: 0.0500 (0.0500)  time: 0.7265  data: 0.0003  max mem: 57114
Epoch: [109]  [ 30/156]  eta: 0:01:37  lr: 0.006159  min_lr: 0.006159  loss: 2.9836 (3.0470)  weight_decay: 0.0500 (0.0500)  time: 0.7297  data: 0.0004  max mem: 57114
Epoch: [109]  [ 40/156]  eta: 0:01:28  lr: 0.006157  min_lr: 0.006157  loss: 3.2765 (3.1068)  weight_decay: 0.0500 (0.0500)  time: 0.7270  data: 0.0003  max mem: 57114
Epoch: [109]  [ 50/156]  eta: 0:01:19  lr: 0.006154  min_lr: 0.006154  loss: 3.3772 (3.1092)  weight_decay: 0.0500 (0.0500)  time: 0.7113  data: 0.0004  max mem: 57114
Epoch: [109]  [ 60/156]  eta: 0:01:11  lr: 0.006152  min_lr: 0.006152  loss: 2.9480 (3.1225)  weight_decay: 0.0500 (0.0500)  time: 0.7010  data: 0.0004  max mem: 57114
Epoch: [109]  [ 70/156]  eta: 0:01:03  lr: 0.006149  min_lr: 0.006149  loss: 3.3441 (3.1457)  weight_decay: 0.0500 (0.0500)  time: 0.7136  data: 0.0004  max mem: 57114
Epoch: [109]  [ 80/156]  eta: 0:00:55  lr: 0.006147  min_lr: 0.006147  loss: 3.4035 (3.1771)  weight_decay: 0.0500 (0.0500)  time: 0.7231  data: 0.0004  max mem: 57114
Epoch: [109]  [ 90/156]  eta: 0:00:48  lr: 0.006144  min_lr: 0.006144  loss: 3.3627 (3.1813)  weight_decay: 0.0500 (0.0500)  time: 0.7155  data: 0.0004  max mem: 57114
Epoch: [109]  [100/156]  eta: 0:00:40  lr: 0.006142  min_lr: 0.006142  loss: 3.3998 (3.1941)  weight_decay: 0.0500 (0.0500)  time: 0.7087  data: 0.0004  max mem: 57114
Epoch: [109]  [110/156]  eta: 0:00:33  lr: 0.006140  min_lr: 0.006140  loss: 3.4230 (3.2033)  weight_decay: 0.0500 (0.0500)  time: 0.7087  data: 0.0004  max mem: 57114
Epoch: [109]  [120/156]  eta: 0:00:26  lr: 0.006137  min_lr: 0.006137  loss: 3.4395 (3.2232)  weight_decay: 0.0500 (0.0500)  time: 0.7005  data: 0.0004  max mem: 57114
Epoch: [109]  [130/156]  eta: 0:00:18  lr: 0.006135  min_lr: 0.006135  loss: 3.5865 (3.2263)  weight_decay: 0.0500 (0.0500)  time: 0.6975  data: 0.0009  max mem: 57114
Epoch: [109]  [140/156]  eta: 0:00:11  lr: 0.006132  min_lr: 0.006132  loss: 3.4915 (3.2189)  weight_decay: 0.0500 (0.0500)  time: 0.6930  data: 0.0007  max mem: 57114
Epoch: [109]  [150/156]  eta: 0:00:04  lr: 0.006130  min_lr: 0.006130  loss: 3.3825 (3.2301)  weight_decay: 0.0500 (0.0500)  time: 0.6856  data: 0.0001  max mem: 57114
Epoch: [109]  [155/156]  eta: 0:00:00  lr: 0.006129  min_lr: 0.006129  loss: 3.4915 (3.2404)  weight_decay: 0.0500 (0.0500)  time: 0.6857  data: 0.0001  max mem: 57114
Epoch: [109] Total time: 0:01:52 (0.7205 s / it)
Averaged stats: lr: 0.006129  min_lr: 0.006129  loss: 3.4915 (3.2662)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0960 (1.0960)  acc1: 81.2500 (81.2500)  acc5: 96.3542 (96.3542)  time: 2.0373  data: 1.7813  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2023 (1.1841)  acc1: 81.2500 (76.9821)  acc5: 95.3125 (93.4783)  time: 0.5582  data: 0.3563  max mem: 57114
Test: Total time: 0:00:02 (0.5781 s / it)
* Acc@1 76.636 Acc@5 93.682 loss 1.289
Accuracy of the model on the 50000 test images: 76.6%
Max accuracy: 76.84%
Test:  [0/5]  eta: 0:00:11  loss: 6.9181 (6.9181)  acc1: 5.7292 (5.7292)  acc5: 10.9375 (10.9375)  time: 2.2923  data: 2.0487  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.9365 (6.7919)  acc1: 4.6875 (3.8363)  acc5: 10.9375 (10.1023)  time: 0.6091  data: 0.4098  max mem: 57114
Test: Total time: 0:00:03 (0.6208 s / it)
* Acc@1 2.370 Acc@5 6.278 loss 7.118
Accuracy of the model EMA on 50000 test images: 2.4%
Max EMA accuracy: 2.37%
Epoch: [110]  [  0/156]  eta: 0:07:29  lr: 0.006128  min_lr: 0.006128  loss: 4.0171 (4.0171)  weight_decay: 0.0500 (0.0500)  time: 2.8836  data: 2.2278  max mem: 57114
Epoch: [110]  [ 10/156]  eta: 0:02:14  lr: 0.006126  min_lr: 0.006126  loss: 3.3014 (3.3183)  weight_decay: 0.0500 (0.0500)  time: 0.9207  data: 0.2028  max mem: 57114
Epoch: [110]  [ 20/156]  eta: 0:01:51  lr: 0.006123  min_lr: 0.006123  loss: 3.1693 (3.2126)  weight_decay: 0.0500 (0.0500)  time: 0.7172  data: 0.0003  max mem: 57114
Epoch: [110]  [ 30/156]  eta: 0:01:38  lr: 0.006121  min_lr: 0.006121  loss: 3.0347 (3.1072)  weight_decay: 0.0500 (0.0500)  time: 0.7079  data: 0.0003  max mem: 57114
Epoch: [110]  [ 40/156]  eta: 0:01:29  lr: 0.006119  min_lr: 0.006119  loss: 3.3093 (3.1737)  weight_decay: 0.0500 (0.0500)  time: 0.7225  data: 0.0004  max mem: 57114
Epoch: [110]  [ 50/156]  eta: 0:01:21  lr: 0.006116  min_lr: 0.006116  loss: 3.4063 (3.2118)  weight_decay: 0.0500 (0.0500)  time: 0.7348  data: 0.0004  max mem: 57114
Epoch: [110]  [ 60/156]  eta: 0:01:12  lr: 0.006114  min_lr: 0.006114  loss: 3.4063 (3.2113)  weight_decay: 0.0500 (0.0500)  time: 0.7336  data: 0.0004  max mem: 57114
Epoch: [110]  [ 70/156]  eta: 0:01:04  lr: 0.006111  min_lr: 0.006111  loss: 3.1780 (3.1956)  weight_decay: 0.0500 (0.0500)  time: 0.7311  data: 0.0004  max mem: 57114
Epoch: [110]  [ 80/156]  eta: 0:00:56  lr: 0.006109  min_lr: 0.006109  loss: 3.2535 (3.2020)  weight_decay: 0.0500 (0.0500)  time: 0.7199  data: 0.0004  max mem: 57114
Epoch: [110]  [ 90/156]  eta: 0:00:49  lr: 0.006106  min_lr: 0.006106  loss: 3.4662 (3.2304)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0004  max mem: 57114
Epoch: [110]  [100/156]  eta: 0:00:41  lr: 0.006104  min_lr: 0.006104  loss: 3.4662 (3.2448)  weight_decay: 0.0500 (0.0500)  time: 0.7068  data: 0.0004  max mem: 57114
Epoch: [110]  [110/156]  eta: 0:00:33  lr: 0.006102  min_lr: 0.006102  loss: 3.4560 (3.2338)  weight_decay: 0.0500 (0.0500)  time: 0.6965  data: 0.0004  max mem: 57114
Epoch: [110]  [120/156]  eta: 0:00:26  lr: 0.006099  min_lr: 0.006099  loss: 3.2841 (3.2244)  weight_decay: 0.0500 (0.0500)  time: 0.7022  data: 0.0004  max mem: 57114
Epoch: [110]  [130/156]  eta: 0:00:19  lr: 0.006097  min_lr: 0.006097  loss: 3.5418 (3.2450)  weight_decay: 0.0500 (0.0500)  time: 0.7087  data: 0.0008  max mem: 57114
Epoch: [110]  [140/156]  eta: 0:00:11  lr: 0.006094  min_lr: 0.006094  loss: 3.3390 (3.2435)  weight_decay: 0.0500 (0.0500)  time: 0.6952  data: 0.0007  max mem: 57114
Epoch: [110]  [150/156]  eta: 0:00:04  lr: 0.006092  min_lr: 0.006092  loss: 3.3390 (3.2464)  weight_decay: 0.0500 (0.0500)  time: 0.6763  data: 0.0001  max mem: 57114
Epoch: [110]  [155/156]  eta: 0:00:00  lr: 0.006090  min_lr: 0.006090  loss: 3.3104 (3.2460)  weight_decay: 0.0500 (0.0500)  time: 0.6758  data: 0.0001  max mem: 57114
Epoch: [110] Total time: 0:01:53 (0.7260 s / it)
Averaged stats: lr: 0.006090  min_lr: 0.006090  loss: 3.3104 (3.2679)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9250 (0.9250)  acc1: 83.3333 (83.3333)  acc5: 94.7917 (94.7917)  time: 2.0253  data: 1.7677  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0178 (1.0100)  acc1: 80.2083 (76.5985)  acc5: 94.7917 (93.6061)  time: 0.5558  data: 0.3536  max mem: 57114
Test: Total time: 0:00:02 (0.5814 s / it)
* Acc@1 76.770 Acc@5 93.696 loss 1.103
Accuracy of the model on the 50000 test images: 76.8%
Max accuracy: 76.84%
Test:  [0/5]  eta: 0:00:11  loss: 6.6384 (6.6384)  acc1: 6.2500 (6.2500)  acc5: 13.5417 (13.5417)  time: 2.2612  data: 2.0179  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.7214 (6.5367)  acc1: 6.2500 (5.2430)  acc5: 13.5417 (12.0205)  time: 0.6035  data: 0.4038  max mem: 57114
Test: Total time: 0:00:03 (0.6153 s / it)
* Acc@1 2.985 Acc@5 7.864 loss 6.916
Accuracy of the model EMA on 50000 test images: 3.0%
Max EMA accuracy: 2.99%
Epoch: [111]  [  0/156]  eta: 0:07:08  lr: 0.006090  min_lr: 0.006090  loss: 3.5641 (3.5641)  weight_decay: 0.0500 (0.0500)  time: 2.7486  data: 2.0974  max mem: 57114
Epoch: [111]  [ 10/156]  eta: 0:02:11  lr: 0.006088  min_lr: 0.006088  loss: 3.4495 (3.3728)  weight_decay: 0.0500 (0.0500)  time: 0.8983  data: 0.1909  max mem: 57114
Epoch: [111]  [ 20/156]  eta: 0:01:51  lr: 0.006085  min_lr: 0.006085  loss: 3.3310 (3.2978)  weight_decay: 0.0500 (0.0500)  time: 0.7208  data: 0.0003  max mem: 57114
Epoch: [111]  [ 30/156]  eta: 0:01:39  lr: 0.006083  min_lr: 0.006083  loss: 3.2400 (3.2314)  weight_decay: 0.0500 (0.0500)  time: 0.7250  data: 0.0003  max mem: 57114
Epoch: [111]  [ 40/156]  eta: 0:01:30  lr: 0.006080  min_lr: 0.006080  loss: 3.2643 (3.2691)  weight_decay: 0.0500 (0.0500)  time: 0.7379  data: 0.0004  max mem: 57114
Epoch: [111]  [ 50/156]  eta: 0:01:21  lr: 0.006078  min_lr: 0.006078  loss: 3.3925 (3.2132)  weight_decay: 0.0500 (0.0500)  time: 0.7394  data: 0.0004  max mem: 57114
Epoch: [111]  [ 60/156]  eta: 0:01:12  lr: 0.006075  min_lr: 0.006075  loss: 3.3274 (3.2141)  weight_decay: 0.0500 (0.0500)  time: 0.7151  data: 0.0004  max mem: 57114
Epoch: [111]  [ 70/156]  eta: 0:01:04  lr: 0.006073  min_lr: 0.006073  loss: 3.3107 (3.1851)  weight_decay: 0.0500 (0.0500)  time: 0.7121  data: 0.0004  max mem: 57114
Epoch: [111]  [ 80/156]  eta: 0:00:56  lr: 0.006071  min_lr: 0.006071  loss: 3.1794 (3.1902)  weight_decay: 0.0500 (0.0500)  time: 0.7177  data: 0.0004  max mem: 57114
Epoch: [111]  [ 90/156]  eta: 0:00:49  lr: 0.006068  min_lr: 0.006068  loss: 3.2359 (3.2110)  weight_decay: 0.0500 (0.0500)  time: 0.7201  data: 0.0004  max mem: 57114
Epoch: [111]  [100/156]  eta: 0:00:41  lr: 0.006066  min_lr: 0.006066  loss: 3.3875 (3.2278)  weight_decay: 0.0500 (0.0500)  time: 0.7093  data: 0.0004  max mem: 57114
Epoch: [111]  [110/156]  eta: 0:00:33  lr: 0.006063  min_lr: 0.006063  loss: 3.3527 (3.2351)  weight_decay: 0.0500 (0.0500)  time: 0.7059  data: 0.0004  max mem: 57114
Epoch: [111]  [120/156]  eta: 0:00:26  lr: 0.006061  min_lr: 0.006061  loss: 3.3012 (3.2332)  weight_decay: 0.0500 (0.0500)  time: 0.7052  data: 0.0004  max mem: 57114
Epoch: [111]  [130/156]  eta: 0:00:19  lr: 0.006058  min_lr: 0.006058  loss: 3.2536 (3.2290)  weight_decay: 0.0500 (0.0500)  time: 0.7002  data: 0.0008  max mem: 57114
Epoch: [111]  [140/156]  eta: 0:00:11  lr: 0.006056  min_lr: 0.006056  loss: 3.0099 (3.2103)  weight_decay: 0.0500 (0.0500)  time: 0.6954  data: 0.0007  max mem: 57114
Epoch: [111]  [150/156]  eta: 0:00:04  lr: 0.006053  min_lr: 0.006053  loss: 3.1702 (3.2224)  weight_decay: 0.0500 (0.0500)  time: 0.6838  data: 0.0001  max mem: 57114
Epoch: [111]  [155/156]  eta: 0:00:00  lr: 0.006052  min_lr: 0.006052  loss: 3.3150 (3.2187)  weight_decay: 0.0500 (0.0500)  time: 0.6851  data: 0.0001  max mem: 57114
Epoch: [111] Total time: 0:01:53 (0.7271 s / it)
Averaged stats: lr: 0.006052  min_lr: 0.006052  loss: 3.3150 (3.2492)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8154 (0.8154)  acc1: 82.8125 (82.8125)  acc5: 96.3542 (96.3542)  time: 2.0935  data: 1.8375  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9915 (0.9163)  acc1: 78.6458 (77.3657)  acc5: 95.8333 (94.1177)  time: 0.5694  data: 0.3676  max mem: 57114
Test: Total time: 0:00:02 (0.5942 s / it)
* Acc@1 77.340 Acc@5 93.938 loss 1.014
Accuracy of the model on the 50000 test images: 77.3%
Max accuracy: 77.34%
Test:  [0/5]  eta: 0:00:10  loss: 6.3525 (6.3525)  acc1: 8.3333 (8.3333)  acc5: 16.6667 (16.6667)  time: 2.1238  data: 1.8805  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.4911 (6.2716)  acc1: 8.3333 (6.6496)  acc5: 16.6667 (13.8107)  time: 0.5753  data: 0.3762  max mem: 57114
Test: Total time: 0:00:02 (0.5867 s / it)
* Acc@1 3.834 Acc@5 9.797 loss 6.693
Accuracy of the model EMA on 50000 test images: 3.8%
Max EMA accuracy: 3.83%
Epoch: [112]  [  0/156]  eta: 0:05:31  lr: 0.006052  min_lr: 0.006052  loss: 2.9165 (2.9165)  weight_decay: 0.0500 (0.0500)  time: 2.1270  data: 1.4758  max mem: 57114
Epoch: [112]  [ 10/156]  eta: 0:02:08  lr: 0.006049  min_lr: 0.006049  loss: 3.4394 (3.1793)  weight_decay: 0.0500 (0.0500)  time: 0.8788  data: 0.1344  max mem: 57114
Epoch: [112]  [ 20/156]  eta: 0:01:50  lr: 0.006047  min_lr: 0.006047  loss: 3.4227 (3.2027)  weight_decay: 0.0500 (0.0500)  time: 0.7489  data: 0.0003  max mem: 57114
Epoch: [112]  [ 30/156]  eta: 0:01:39  lr: 0.006044  min_lr: 0.006044  loss: 3.1249 (3.0978)  weight_decay: 0.0500 (0.0500)  time: 0.7448  data: 0.0004  max mem: 57114
Epoch: [112]  [ 40/156]  eta: 0:01:30  lr: 0.006042  min_lr: 0.006042  loss: 3.0628 (3.1285)  weight_decay: 0.0500 (0.0500)  time: 0.7448  data: 0.0004  max mem: 57114
Epoch: [112]  [ 50/156]  eta: 0:01:22  lr: 0.006039  min_lr: 0.006039  loss: 3.3909 (3.1918)  weight_decay: 0.0500 (0.0500)  time: 0.7463  data: 0.0004  max mem: 57114
Epoch: [112]  [ 60/156]  eta: 0:01:13  lr: 0.006037  min_lr: 0.006037  loss: 3.4074 (3.1713)  weight_decay: 0.0500 (0.0500)  time: 0.7321  data: 0.0004  max mem: 57114
Epoch: [112]  [ 70/156]  eta: 0:01:04  lr: 0.006035  min_lr: 0.006035  loss: 3.4432 (3.2065)  weight_decay: 0.0500 (0.0500)  time: 0.7077  data: 0.0004  max mem: 57114
Epoch: [112]  [ 80/156]  eta: 0:00:57  lr: 0.006032  min_lr: 0.006032  loss: 3.4358 (3.1997)  weight_decay: 0.0500 (0.0500)  time: 0.7112  data: 0.0004  max mem: 57114
Epoch: [112]  [ 90/156]  eta: 0:00:49  lr: 0.006030  min_lr: 0.006030  loss: 3.1199 (3.1869)  weight_decay: 0.0500 (0.0500)  time: 0.7169  data: 0.0004  max mem: 57114
Epoch: [112]  [100/156]  eta: 0:00:41  lr: 0.006027  min_lr: 0.006027  loss: 3.4936 (3.2261)  weight_decay: 0.0500 (0.0500)  time: 0.7139  data: 0.0004  max mem: 57114
Epoch: [112]  [110/156]  eta: 0:00:34  lr: 0.006025  min_lr: 0.006025  loss: 3.4745 (3.2280)  weight_decay: 0.0500 (0.0500)  time: 0.7094  data: 0.0004  max mem: 57114
Epoch: [112]  [120/156]  eta: 0:00:26  lr: 0.006022  min_lr: 0.006022  loss: 3.4150 (3.2411)  weight_decay: 0.0500 (0.0500)  time: 0.7064  data: 0.0004  max mem: 57114
Epoch: [112]  [130/156]  eta: 0:00:19  lr: 0.006020  min_lr: 0.006020  loss: 3.4245 (3.2468)  weight_decay: 0.0500 (0.0500)  time: 0.7054  data: 0.0009  max mem: 57114
Epoch: [112]  [140/156]  eta: 0:00:11  lr: 0.006017  min_lr: 0.006017  loss: 3.3044 (3.2492)  weight_decay: 0.0500 (0.0500)  time: 0.6925  data: 0.0007  max mem: 57114
Epoch: [112]  [150/156]  eta: 0:00:04  lr: 0.006015  min_lr: 0.006015  loss: 3.3838 (3.2507)  weight_decay: 0.0500 (0.0500)  time: 0.6830  data: 0.0001  max mem: 57114
Epoch: [112]  [155/156]  eta: 0:00:00  lr: 0.006013  min_lr: 0.006013  loss: 3.3130 (3.2520)  weight_decay: 0.0500 (0.0500)  time: 0.6874  data: 0.0001  max mem: 57114
Epoch: [112] Total time: 0:01:53 (0.7293 s / it)
Averaged stats: lr: 0.006013  min_lr: 0.006013  loss: 3.3130 (3.2466)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9929 (0.9929)  acc1: 84.3750 (84.3750)  acc5: 96.8750 (96.8750)  time: 2.1379  data: 1.8823  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0938 (1.0863)  acc1: 79.1667 (78.1330)  acc5: 94.7917 (93.6061)  time: 0.5782  data: 0.3765  max mem: 57114
Test: Total time: 0:00:03 (0.6028 s / it)
* Acc@1 76.367 Acc@5 93.636 loss 1.190
Accuracy of the model on the 50000 test images: 76.4%
Max accuracy: 77.34%
Test:  [0/5]  eta: 0:00:11  loss: 6.0599 (6.0599)  acc1: 8.8542 (8.8542)  acc5: 20.8333 (20.8333)  time: 2.3957  data: 2.1520  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.2618 (6.0086)  acc1: 8.8542 (7.6726)  acc5: 19.7917 (16.4962)  time: 0.6298  data: 0.4305  max mem: 57114
Test: Total time: 0:00:03 (0.6418 s / it)
* Acc@1 4.965 Acc@5 12.108 loss 6.458
Accuracy of the model EMA on 50000 test images: 5.0%
Max EMA accuracy: 4.97%
Epoch: [113]  [  0/156]  eta: 0:06:54  lr: 0.006013  min_lr: 0.006013  loss: 2.0871 (2.0871)  weight_decay: 0.0500 (0.0500)  time: 2.6602  data: 2.0111  max mem: 57114
Epoch: [113]  [ 10/156]  eta: 0:02:11  lr: 0.006011  min_lr: 0.006011  loss: 3.3408 (3.2755)  weight_decay: 0.0500 (0.0500)  time: 0.8983  data: 0.1832  max mem: 57114
Epoch: [113]  [ 20/156]  eta: 0:01:50  lr: 0.006008  min_lr: 0.006008  loss: 3.2733 (3.2021)  weight_decay: 0.0500 (0.0500)  time: 0.7180  data: 0.0004  max mem: 57114
Epoch: [113]  [ 30/156]  eta: 0:01:38  lr: 0.006006  min_lr: 0.006006  loss: 3.2631 (3.2493)  weight_decay: 0.0500 (0.0500)  time: 0.7133  data: 0.0004  max mem: 57114
Epoch: [113]  [ 40/156]  eta: 0:01:28  lr: 0.006003  min_lr: 0.006003  loss: 3.2476 (3.2045)  weight_decay: 0.0500 (0.0500)  time: 0.7189  data: 0.0005  max mem: 57114
Epoch: [113]  [ 50/156]  eta: 0:01:20  lr: 0.006001  min_lr: 0.006001  loss: 3.3565 (3.2181)  weight_decay: 0.0500 (0.0500)  time: 0.7203  data: 0.0005  max mem: 57114
Epoch: [113]  [ 60/156]  eta: 0:01:12  lr: 0.005998  min_lr: 0.005998  loss: 3.3259 (3.1582)  weight_decay: 0.0500 (0.0500)  time: 0.7199  data: 0.0004  max mem: 57114
Epoch: [113]  [ 70/156]  eta: 0:01:04  lr: 0.005996  min_lr: 0.005996  loss: 2.7268 (3.1291)  weight_decay: 0.0500 (0.0500)  time: 0.7187  data: 0.0005  max mem: 57114
Epoch: [113]  [ 80/156]  eta: 0:00:56  lr: 0.005993  min_lr: 0.005993  loss: 3.3484 (3.1628)  weight_decay: 0.0500 (0.0500)  time: 0.7225  data: 0.0005  max mem: 57114
Epoch: [113]  [ 90/156]  eta: 0:00:48  lr: 0.005991  min_lr: 0.005991  loss: 3.4658 (3.1751)  weight_decay: 0.0500 (0.0500)  time: 0.7161  data: 0.0006  max mem: 57114
Epoch: [113]  [100/156]  eta: 0:00:41  lr: 0.005988  min_lr: 0.005988  loss: 3.2450 (3.1605)  weight_decay: 0.0500 (0.0500)  time: 0.7036  data: 0.0007  max mem: 57114
Epoch: [113]  [110/156]  eta: 0:00:33  lr: 0.005986  min_lr: 0.005986  loss: 3.2450 (3.1754)  weight_decay: 0.0500 (0.0500)  time: 0.7139  data: 0.0007  max mem: 57114
Epoch: [113]  [120/156]  eta: 0:00:26  lr: 0.005983  min_lr: 0.005983  loss: 3.3250 (3.1766)  weight_decay: 0.0500 (0.0500)  time: 0.7215  data: 0.0007  max mem: 57114
Epoch: [113]  [130/156]  eta: 0:00:18  lr: 0.005981  min_lr: 0.005981  loss: 3.3243 (3.1850)  weight_decay: 0.0500 (0.0500)  time: 0.7070  data: 0.0018  max mem: 57114
Epoch: [113]  [140/156]  eta: 0:00:11  lr: 0.005978  min_lr: 0.005978  loss: 3.3692 (3.1869)  weight_decay: 0.0500 (0.0500)  time: 0.6870  data: 0.0015  max mem: 57114
Epoch: [113]  [150/156]  eta: 0:00:04  lr: 0.005976  min_lr: 0.005976  loss: 3.3692 (3.1895)  weight_decay: 0.0500 (0.0500)  time: 0.6798  data: 0.0001  max mem: 57114
Epoch: [113]  [155/156]  eta: 0:00:00  lr: 0.005975  min_lr: 0.005975  loss: 3.3671 (3.1859)  weight_decay: 0.0500 (0.0500)  time: 0.6814  data: 0.0001  max mem: 57114
Epoch: [113] Total time: 0:01:53 (0.7245 s / it)
Averaged stats: lr: 0.005975  min_lr: 0.005975  loss: 3.3671 (3.2468)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7906 (0.7906)  acc1: 82.2917 (82.2917)  acc5: 95.8333 (95.8333)  time: 2.1734  data: 1.9164  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8281 (0.8507)  acc1: 81.2500 (77.8772)  acc5: 95.8333 (94.5013)  time: 0.5861  data: 0.3835  max mem: 57114
Test: Total time: 0:00:03 (0.6100 s / it)
* Acc@1 77.508 Acc@5 93.968 loss 0.982
Accuracy of the model on the 50000 test images: 77.5%
Max accuracy: 77.51%
Test:  [0/5]  eta: 0:00:10  loss: 5.7622 (5.7622)  acc1: 11.9792 (11.9792)  acc5: 22.9167 (22.9167)  time: 2.1121  data: 1.8684  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 6.0325 (5.7464)  acc1: 11.9792 (9.7187)  acc5: 21.8750 (19.0537)  time: 0.5730  data: 0.3737  max mem: 57114
Test: Total time: 0:00:02 (0.5848 s / it)
* Acc@1 6.252 Acc@5 14.796 loss 6.214
Accuracy of the model EMA on 50000 test images: 6.3%
Max EMA accuracy: 6.25%
Epoch: [114]  [  0/156]  eta: 0:06:02  lr: 0.005974  min_lr: 0.005974  loss: 2.5972 (2.5972)  weight_decay: 0.0500 (0.0500)  time: 2.3221  data: 1.6727  max mem: 57114
Epoch: [114]  [ 10/156]  eta: 0:02:06  lr: 0.005972  min_lr: 0.005972  loss: 3.4731 (3.2128)  weight_decay: 0.0500 (0.0500)  time: 0.8661  data: 0.1527  max mem: 57114
Epoch: [114]  [ 20/156]  eta: 0:01:49  lr: 0.005969  min_lr: 0.005969  loss: 3.4647 (3.2572)  weight_decay: 0.0500 (0.0500)  time: 0.7328  data: 0.0007  max mem: 57114
Epoch: [114]  [ 30/156]  eta: 0:01:38  lr: 0.005967  min_lr: 0.005967  loss: 3.4647 (3.2546)  weight_decay: 0.0500 (0.0500)  time: 0.7356  data: 0.0007  max mem: 57114
Epoch: [114]  [ 40/156]  eta: 0:01:30  lr: 0.005964  min_lr: 0.005964  loss: 3.3346 (3.2102)  weight_decay: 0.0500 (0.0500)  time: 0.7505  data: 0.0006  max mem: 57114
Epoch: [114]  [ 50/156]  eta: 0:01:21  lr: 0.005962  min_lr: 0.005962  loss: 3.2401 (3.2165)  weight_decay: 0.0500 (0.0500)  time: 0.7414  data: 0.0005  max mem: 57114
Epoch: [114]  [ 60/156]  eta: 0:01:12  lr: 0.005959  min_lr: 0.005959  loss: 3.4305 (3.2349)  weight_decay: 0.0500 (0.0500)  time: 0.7145  data: 0.0004  max mem: 57114
Epoch: [114]  [ 70/156]  eta: 0:01:04  lr: 0.005957  min_lr: 0.005957  loss: 3.3289 (3.2150)  weight_decay: 0.0500 (0.0500)  time: 0.7074  data: 0.0004  max mem: 57114
Epoch: [114]  [ 80/156]  eta: 0:00:56  lr: 0.005954  min_lr: 0.005954  loss: 2.9937 (3.2167)  weight_decay: 0.0500 (0.0500)  time: 0.7118  data: 0.0004  max mem: 57114
Epoch: [114]  [ 90/156]  eta: 0:00:49  lr: 0.005952  min_lr: 0.005952  loss: 3.3875 (3.2285)  weight_decay: 0.0500 (0.0500)  time: 0.7229  data: 0.0004  max mem: 57114
Epoch: [114]  [100/156]  eta: 0:00:41  lr: 0.005949  min_lr: 0.005949  loss: 3.3819 (3.1975)  weight_decay: 0.0500 (0.0500)  time: 0.7156  data: 0.0004  max mem: 57114
Epoch: [114]  [110/156]  eta: 0:00:33  lr: 0.005947  min_lr: 0.005947  loss: 3.3740 (3.2019)  weight_decay: 0.0500 (0.0500)  time: 0.7040  data: 0.0004  max mem: 57114
Epoch: [114]  [120/156]  eta: 0:00:26  lr: 0.005944  min_lr: 0.005944  loss: 3.4010 (3.2052)  weight_decay: 0.0500 (0.0500)  time: 0.6986  data: 0.0004  max mem: 57114
Epoch: [114]  [130/156]  eta: 0:00:18  lr: 0.005942  min_lr: 0.005942  loss: 3.4326 (3.2306)  weight_decay: 0.0500 (0.0500)  time: 0.6952  data: 0.0005  max mem: 57114
Epoch: [114]  [140/156]  eta: 0:00:11  lr: 0.005939  min_lr: 0.005939  loss: 3.5456 (3.2361)  weight_decay: 0.0500 (0.0500)  time: 0.6809  data: 0.0004  max mem: 57114
Epoch: [114]  [150/156]  eta: 0:00:04  lr: 0.005937  min_lr: 0.005937  loss: 3.1859 (3.2221)  weight_decay: 0.0500 (0.0500)  time: 0.6850  data: 0.0001  max mem: 57114
Epoch: [114]  [155/156]  eta: 0:00:00  lr: 0.005935  min_lr: 0.005935  loss: 3.2581 (3.2256)  weight_decay: 0.0500 (0.0500)  time: 0.6844  data: 0.0001  max mem: 57114
Epoch: [114] Total time: 0:01:53 (0.7252 s / it)
Averaged stats: lr: 0.005935  min_lr: 0.005935  loss: 3.2581 (3.2477)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8308 (0.8308)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.0722  data: 1.8161  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9656 (1.0002)  acc1: 81.7708 (79.2839)  acc5: 95.3125 (94.2455)  time: 0.5650  data: 0.3633  max mem: 57114
Test: Total time: 0:00:02 (0.5902 s / it)
* Acc@1 76.906 Acc@5 93.746 loss 1.114
Accuracy of the model on the 50000 test images: 76.9%
Max accuracy: 77.51%
Test:  [0/5]  eta: 0:00:11  loss: 5.4480 (5.4480)  acc1: 14.5833 (14.5833)  acc5: 25.0000 (25.0000)  time: 2.3230  data: 2.0796  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 5.7997 (5.4779)  acc1: 14.0625 (11.7647)  acc5: 24.4792 (22.3785)  time: 0.6153  data: 0.4160  max mem: 57114
Test: Total time: 0:00:03 (0.6267 s / it)
* Acc@1 7.801 Acc@5 17.941 loss 5.950
Accuracy of the model EMA on 50000 test images: 7.8%
Max EMA accuracy: 7.80%
Epoch: [115]  [  0/156]  eta: 0:07:35  lr: 0.005935  min_lr: 0.005935  loss: 3.2565 (3.2565)  weight_decay: 0.0500 (0.0500)  time: 2.9200  data: 2.2640  max mem: 57114
Epoch: [115]  [ 10/156]  eta: 0:02:13  lr: 0.005933  min_lr: 0.005933  loss: 3.2049 (3.1185)  weight_decay: 0.0500 (0.0500)  time: 0.9172  data: 0.2061  max mem: 57114
Epoch: [115]  [ 20/156]  eta: 0:01:52  lr: 0.005930  min_lr: 0.005930  loss: 3.1701 (3.1634)  weight_decay: 0.0500 (0.0500)  time: 0.7206  data: 0.0004  max mem: 57114
Epoch: [115]  [ 30/156]  eta: 0:01:39  lr: 0.005928  min_lr: 0.005928  loss: 3.2681 (3.2188)  weight_decay: 0.0500 (0.0500)  time: 0.7246  data: 0.0005  max mem: 57114
Epoch: [115]  [ 40/156]  eta: 0:01:29  lr: 0.005925  min_lr: 0.005925  loss: 3.2681 (3.1585)  weight_decay: 0.0500 (0.0500)  time: 0.7219  data: 0.0005  max mem: 57114
Epoch: [115]  [ 50/156]  eta: 0:01:20  lr: 0.005923  min_lr: 0.005923  loss: 3.1684 (3.1353)  weight_decay: 0.0500 (0.0500)  time: 0.7184  data: 0.0005  max mem: 57114
Epoch: [115]  [ 60/156]  eta: 0:01:12  lr: 0.005920  min_lr: 0.005920  loss: 3.2340 (3.1515)  weight_decay: 0.0500 (0.0500)  time: 0.7228  data: 0.0004  max mem: 57114
Epoch: [115]  [ 70/156]  eta: 0:01:04  lr: 0.005917  min_lr: 0.005917  loss: 3.2969 (3.1732)  weight_decay: 0.0500 (0.0500)  time: 0.7332  data: 0.0006  max mem: 57114
Epoch: [115]  [ 80/156]  eta: 0:00:56  lr: 0.005915  min_lr: 0.005915  loss: 3.4511 (3.1893)  weight_decay: 0.0500 (0.0500)  time: 0.7229  data: 0.0007  max mem: 57114
Epoch: [115]  [ 90/156]  eta: 0:00:49  lr: 0.005912  min_lr: 0.005912  loss: 3.3029 (3.1735)  weight_decay: 0.0500 (0.0500)  time: 0.7026  data: 0.0007  max mem: 57114
Epoch: [115]  [100/156]  eta: 0:00:41  lr: 0.005910  min_lr: 0.005910  loss: 3.2135 (3.1877)  weight_decay: 0.0500 (0.0500)  time: 0.7016  data: 0.0007  max mem: 57114
Epoch: [115]  [110/156]  eta: 0:00:33  lr: 0.005907  min_lr: 0.005907  loss: 3.3497 (3.1928)  weight_decay: 0.0500 (0.0500)  time: 0.6961  data: 0.0006  max mem: 57114
Epoch: [115]  [120/156]  eta: 0:00:26  lr: 0.005905  min_lr: 0.005905  loss: 3.4831 (3.1915)  weight_decay: 0.0500 (0.0500)  time: 0.7036  data: 0.0007  max mem: 57114
Epoch: [115]  [130/156]  eta: 0:00:19  lr: 0.005902  min_lr: 0.005902  loss: 3.4831 (3.1980)  weight_decay: 0.0500 (0.0500)  time: 0.7077  data: 0.0007  max mem: 57114
Epoch: [115]  [140/156]  eta: 0:00:11  lr: 0.005900  min_lr: 0.005900  loss: 3.5050 (3.2124)  weight_decay: 0.0500 (0.0500)  time: 0.6909  data: 0.0004  max mem: 57114
Epoch: [115]  [150/156]  eta: 0:00:04  lr: 0.005897  min_lr: 0.005897  loss: 3.5050 (3.2229)  weight_decay: 0.0500 (0.0500)  time: 0.6880  data: 0.0001  max mem: 57114
Epoch: [115]  [155/156]  eta: 0:00:00  lr: 0.005896  min_lr: 0.005896  loss: 3.5010 (3.2360)  weight_decay: 0.0500 (0.0500)  time: 0.6908  data: 0.0001  max mem: 57114
Epoch: [115] Total time: 0:01:53 (0.7268 s / it)
Averaged stats: lr: 0.005896  min_lr: 0.005896  loss: 3.5010 (3.2404)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 0.9813 (0.9813)  acc1: 84.3750 (84.3750)  acc5: 96.8750 (96.8750)  time: 2.2500  data: 1.9939  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0555 (1.0774)  acc1: 83.8542 (78.6445)  acc5: 96.3542 (94.3734)  time: 0.6012  data: 0.3989  max mem: 57114
Test: Total time: 0:00:03 (0.6224 s / it)
* Acc@1 76.966 Acc@5 93.740 loss 1.179
Accuracy of the model on the 50000 test images: 77.0%
Max accuracy: 77.51%
Test:  [0/5]  eta: 0:00:11  loss: 5.1427 (5.1427)  acc1: 17.1875 (17.1875)  acc5: 29.1667 (29.1667)  time: 2.2287  data: 1.9853  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 5.5735 (5.2154)  acc1: 15.1042 (13.8107)  acc5: 27.6042 (26.0870)  time: 0.5965  data: 0.3972  max mem: 57114
Test: Total time: 0:00:03 (0.6107 s / it)
* Acc@1 9.721 Acc@5 21.260 loss 5.685
Accuracy of the model EMA on 50000 test images: 9.7%
Max EMA accuracy: 9.72%
Epoch: [116]  [  0/156]  eta: 0:07:48  lr: 0.005896  min_lr: 0.005896  loss: 3.7110 (3.7110)  weight_decay: 0.0500 (0.0500)  time: 3.0048  data: 2.3557  max mem: 57114
Epoch: [116]  [ 10/156]  eta: 0:02:15  lr: 0.005893  min_lr: 0.005893  loss: 3.4444 (3.3539)  weight_decay: 0.0500 (0.0500)  time: 0.9284  data: 0.2144  max mem: 57114
Epoch: [116]  [ 20/156]  eta: 0:01:52  lr: 0.005891  min_lr: 0.005891  loss: 3.3831 (3.2869)  weight_decay: 0.0500 (0.0500)  time: 0.7154  data: 0.0003  max mem: 57114
Epoch: [116]  [ 30/156]  eta: 0:01:39  lr: 0.005888  min_lr: 0.005888  loss: 3.3002 (3.2643)  weight_decay: 0.0500 (0.0500)  time: 0.7164  data: 0.0004  max mem: 57114
Epoch: [116]  [ 40/156]  eta: 0:01:30  lr: 0.005886  min_lr: 0.005886  loss: 3.2821 (3.2377)  weight_decay: 0.0500 (0.0500)  time: 0.7278  data: 0.0004  max mem: 57114
Epoch: [116]  [ 50/156]  eta: 0:01:21  lr: 0.005883  min_lr: 0.005883  loss: 3.2932 (3.2449)  weight_decay: 0.0500 (0.0500)  time: 0.7290  data: 0.0004  max mem: 57114
Epoch: [116]  [ 60/156]  eta: 0:01:13  lr: 0.005881  min_lr: 0.005881  loss: 3.3420 (3.2478)  weight_decay: 0.0500 (0.0500)  time: 0.7336  data: 0.0004  max mem: 57114
Epoch: [116]  [ 70/156]  eta: 0:01:04  lr: 0.005878  min_lr: 0.005878  loss: 3.3667 (3.2350)  weight_decay: 0.0500 (0.0500)  time: 0.7167  data: 0.0004  max mem: 57114
Epoch: [116]  [ 80/156]  eta: 0:00:56  lr: 0.005875  min_lr: 0.005875  loss: 3.3829 (3.2515)  weight_decay: 0.0500 (0.0500)  time: 0.6996  data: 0.0004  max mem: 57114
Epoch: [116]  [ 90/156]  eta: 0:00:48  lr: 0.005873  min_lr: 0.005873  loss: 3.3128 (3.2491)  weight_decay: 0.0500 (0.0500)  time: 0.6936  data: 0.0004  max mem: 57114
Epoch: [116]  [100/156]  eta: 0:00:41  lr: 0.005870  min_lr: 0.005870  loss: 3.3099 (3.2601)  weight_decay: 0.0500 (0.0500)  time: 0.6973  data: 0.0004  max mem: 57114
Epoch: [116]  [110/156]  eta: 0:00:33  lr: 0.005868  min_lr: 0.005868  loss: 3.3099 (3.2663)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0004  max mem: 57114
Epoch: [116]  [120/156]  eta: 0:00:26  lr: 0.005865  min_lr: 0.005865  loss: 3.2868 (3.2789)  weight_decay: 0.0500 (0.0500)  time: 0.7066  data: 0.0004  max mem: 57114
Epoch: [116]  [130/156]  eta: 0:00:18  lr: 0.005863  min_lr: 0.005863  loss: 3.3873 (3.2827)  weight_decay: 0.0500 (0.0500)  time: 0.7056  data: 0.0009  max mem: 57114
Epoch: [116]  [140/156]  eta: 0:00:11  lr: 0.005860  min_lr: 0.005860  loss: 3.4318 (3.2842)  weight_decay: 0.0500 (0.0500)  time: 0.6867  data: 0.0007  max mem: 57114
Epoch: [116]  [150/156]  eta: 0:00:04  lr: 0.005858  min_lr: 0.005858  loss: 3.4136 (3.2844)  weight_decay: 0.0500 (0.0500)  time: 0.6775  data: 0.0001  max mem: 57114
Epoch: [116]  [155/156]  eta: 0:00:00  lr: 0.005856  min_lr: 0.005856  loss: 3.4136 (3.2837)  weight_decay: 0.0500 (0.0500)  time: 0.6764  data: 0.0001  max mem: 57114
Epoch: [116] Total time: 0:01:52 (0.7236 s / it)
Averaged stats: lr: 0.005856  min_lr: 0.005856  loss: 3.4136 (3.2386)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8826 (0.8826)  acc1: 82.2917 (82.2917)  acc5: 96.3542 (96.3542)  time: 2.0410  data: 1.7847  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0555 (1.0212)  acc1: 79.6875 (78.6445)  acc5: 96.3542 (93.7340)  time: 0.5590  data: 0.3570  max mem: 57114
Test: Total time: 0:00:02 (0.5820 s / it)
* Acc@1 77.272 Acc@5 93.886 loss 1.117
Accuracy of the model on the 50000 test images: 77.3%
Max accuracy: 77.51%
Test:  [0/5]  eta: 0:00:11  loss: 4.8327 (4.8327)  acc1: 19.2708 (19.2708)  acc5: 36.4583 (36.4583)  time: 2.3368  data: 2.0933  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 5.3535 (4.9554)  acc1: 17.1875 (15.2174)  acc5: 28.6458 (30.1790)  time: 0.6180  data: 0.4188  max mem: 57114
Test: Total time: 0:00:03 (0.6307 s / it)
* Acc@1 11.743 Acc@5 24.838 loss 5.417
Accuracy of the model EMA on 50000 test images: 11.7%
Max EMA accuracy: 11.74%
Epoch: [117]  [  0/156]  eta: 0:08:03  lr: 0.005856  min_lr: 0.005856  loss: 3.6707 (3.6707)  weight_decay: 0.0500 (0.0500)  time: 3.0962  data: 2.4397  max mem: 57114
Epoch: [117]  [ 10/156]  eta: 0:02:18  lr: 0.005854  min_lr: 0.005854  loss: 3.3780 (3.4774)  weight_decay: 0.0500 (0.0500)  time: 0.9503  data: 0.2222  max mem: 57114
Epoch: [117]  [ 20/156]  eta: 0:01:53  lr: 0.005851  min_lr: 0.005851  loss: 3.4520 (3.4969)  weight_decay: 0.0500 (0.0500)  time: 0.7248  data: 0.0005  max mem: 57114
Epoch: [117]  [ 30/156]  eta: 0:01:41  lr: 0.005848  min_lr: 0.005848  loss: 3.4314 (3.4403)  weight_decay: 0.0500 (0.0500)  time: 0.7284  data: 0.0004  max mem: 57114
Epoch: [117]  [ 40/156]  eta: 0:01:31  lr: 0.005846  min_lr: 0.005846  loss: 3.2226 (3.3461)  weight_decay: 0.0500 (0.0500)  time: 0.7364  data: 0.0004  max mem: 57114
Epoch: [117]  [ 50/156]  eta: 0:01:21  lr: 0.005843  min_lr: 0.005843  loss: 3.2125 (3.3526)  weight_decay: 0.0500 (0.0500)  time: 0.7205  data: 0.0004  max mem: 57114
Epoch: [117]  [ 60/156]  eta: 0:01:13  lr: 0.005841  min_lr: 0.005841  loss: 3.3107 (3.3364)  weight_decay: 0.0500 (0.0500)  time: 0.7147  data: 0.0004  max mem: 57114
Epoch: [117]  [ 70/156]  eta: 0:01:05  lr: 0.005838  min_lr: 0.005838  loss: 3.4338 (3.3330)  weight_decay: 0.0500 (0.0500)  time: 0.7155  data: 0.0005  max mem: 57114
Epoch: [117]  [ 80/156]  eta: 0:00:56  lr: 0.005836  min_lr: 0.005836  loss: 3.2639 (3.3058)  weight_decay: 0.0500 (0.0500)  time: 0.7024  data: 0.0004  max mem: 57114
Epoch: [117]  [ 90/156]  eta: 0:00:49  lr: 0.005833  min_lr: 0.005833  loss: 3.2560 (3.2943)  weight_decay: 0.0500 (0.0500)  time: 0.7085  data: 0.0004  max mem: 57114
Epoch: [117]  [100/156]  eta: 0:00:41  lr: 0.005831  min_lr: 0.005831  loss: 3.1563 (3.2747)  weight_decay: 0.0500 (0.0500)  time: 0.7108  data: 0.0004  max mem: 57114
Epoch: [117]  [110/156]  eta: 0:00:34  lr: 0.005828  min_lr: 0.005828  loss: 3.1563 (3.2588)  weight_decay: 0.0500 (0.0500)  time: 0.7120  data: 0.0004  max mem: 57114
Epoch: [117]  [120/156]  eta: 0:00:26  lr: 0.005825  min_lr: 0.005825  loss: 3.2732 (3.2545)  weight_decay: 0.0500 (0.0500)  time: 0.7202  data: 0.0003  max mem: 57114
Epoch: [117]  [130/156]  eta: 0:00:19  lr: 0.005823  min_lr: 0.005823  loss: 3.3021 (3.2591)  weight_decay: 0.0500 (0.0500)  time: 0.7055  data: 0.0006  max mem: 57114
Epoch: [117]  [140/156]  eta: 0:00:11  lr: 0.005820  min_lr: 0.005820  loss: 3.4731 (3.2747)  weight_decay: 0.0500 (0.0500)  time: 0.6884  data: 0.0005  max mem: 57114
Epoch: [117]  [150/156]  eta: 0:00:04  lr: 0.005818  min_lr: 0.005818  loss: 3.4512 (3.2693)  weight_decay: 0.0500 (0.0500)  time: 0.6806  data: 0.0001  max mem: 57114
Epoch: [117]  [155/156]  eta: 0:00:00  lr: 0.005816  min_lr: 0.005816  loss: 3.3563 (3.2679)  weight_decay: 0.0500 (0.0500)  time: 0.6802  data: 0.0001  max mem: 57114
Epoch: [117] Total time: 0:01:53 (0.7285 s / it)
Averaged stats: lr: 0.005816  min_lr: 0.005816  loss: 3.3563 (3.2324)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0545 (1.0545)  acc1: 81.2500 (81.2500)  acc5: 96.8750 (96.8750)  time: 2.0881  data: 1.8323  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1459 (1.1415)  acc1: 80.2083 (77.8772)  acc5: 96.8750 (95.0128)  time: 0.5683  data: 0.3665  max mem: 57114
Test: Total time: 0:00:02 (0.5913 s / it)
* Acc@1 76.712 Acc@5 93.380 loss 1.250
Accuracy of the model on the 50000 test images: 76.7%
Max accuracy: 77.51%
Test:  [0/5]  eta: 0:00:11  loss: 4.5118 (4.5118)  acc1: 20.8333 (20.8333)  acc5: 40.6250 (40.6250)  time: 2.2887  data: 2.0453  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 5.0047 (4.6880)  acc1: 19.2708 (17.6471)  acc5: 32.2917 (34.1432)  time: 0.6084  data: 0.4091  max mem: 57114
Test: Total time: 0:00:03 (0.6205 s / it)
* Acc@1 14.148 Acc@5 28.748 loss 5.140
Accuracy of the model EMA on 50000 test images: 14.1%
Max EMA accuracy: 14.15%
Epoch: [118]  [  0/156]  eta: 0:07:11  lr: 0.005816  min_lr: 0.005816  loss: 3.2462 (3.2462)  weight_decay: 0.0500 (0.0500)  time: 2.7638  data: 2.1106  max mem: 57114
Epoch: [118]  [ 10/156]  eta: 0:02:12  lr: 0.005814  min_lr: 0.005814  loss: 3.6729 (3.3563)  weight_decay: 0.0500 (0.0500)  time: 0.9100  data: 0.1921  max mem: 57114
Epoch: [118]  [ 20/156]  eta: 0:01:52  lr: 0.005811  min_lr: 0.005811  loss: 3.2051 (3.2230)  weight_decay: 0.0500 (0.0500)  time: 0.7296  data: 0.0003  max mem: 57114
Epoch: [118]  [ 30/156]  eta: 0:01:40  lr: 0.005809  min_lr: 0.005809  loss: 3.2051 (3.1991)  weight_decay: 0.0500 (0.0500)  time: 0.7411  data: 0.0003  max mem: 57114
Epoch: [118]  [ 40/156]  eta: 0:01:30  lr: 0.005806  min_lr: 0.005806  loss: 3.1968 (3.1943)  weight_decay: 0.0500 (0.0500)  time: 0.7349  data: 0.0004  max mem: 57114
Epoch: [118]  [ 50/156]  eta: 0:01:21  lr: 0.005803  min_lr: 0.005803  loss: 3.0053 (3.1655)  weight_decay: 0.0500 (0.0500)  time: 0.7175  data: 0.0003  max mem: 57114
Epoch: [118]  [ 60/156]  eta: 0:01:13  lr: 0.005801  min_lr: 0.005801  loss: 2.9455 (3.1490)  weight_decay: 0.0500 (0.0500)  time: 0.7191  data: 0.0003  max mem: 57114
Epoch: [118]  [ 70/156]  eta: 0:01:04  lr: 0.005798  min_lr: 0.005798  loss: 3.2647 (3.2039)  weight_decay: 0.0500 (0.0500)  time: 0.7212  data: 0.0003  max mem: 57114
Epoch: [118]  [ 80/156]  eta: 0:00:57  lr: 0.005796  min_lr: 0.005796  loss: 3.4230 (3.2066)  weight_decay: 0.0500 (0.0500)  time: 0.7173  data: 0.0004  max mem: 57114
Epoch: [118]  [ 90/156]  eta: 0:00:49  lr: 0.005793  min_lr: 0.005793  loss: 3.2841 (3.2224)  weight_decay: 0.0500 (0.0500)  time: 0.7131  data: 0.0004  max mem: 57114
Epoch: [118]  [100/156]  eta: 0:00:41  lr: 0.005791  min_lr: 0.005791  loss: 3.3627 (3.2350)  weight_decay: 0.0500 (0.0500)  time: 0.7074  data: 0.0004  max mem: 57114
Epoch: [118]  [110/156]  eta: 0:00:33  lr: 0.005788  min_lr: 0.005788  loss: 3.4687 (3.2534)  weight_decay: 0.0500 (0.0500)  time: 0.7062  data: 0.0004  max mem: 57114
Epoch: [118]  [120/156]  eta: 0:00:26  lr: 0.005785  min_lr: 0.005785  loss: 3.3151 (3.2518)  weight_decay: 0.0500 (0.0500)  time: 0.7100  data: 0.0004  max mem: 57114
Epoch: [118]  [130/156]  eta: 0:00:19  lr: 0.005783  min_lr: 0.005783  loss: 3.1067 (3.2280)  weight_decay: 0.0500 (0.0500)  time: 0.7065  data: 0.0008  max mem: 57114
Epoch: [118]  [140/156]  eta: 0:00:11  lr: 0.005780  min_lr: 0.005780  loss: 2.7950 (3.2122)  weight_decay: 0.0500 (0.0500)  time: 0.6920  data: 0.0007  max mem: 57114
Epoch: [118]  [150/156]  eta: 0:00:04  lr: 0.005778  min_lr: 0.005778  loss: 3.3387 (3.2272)  weight_decay: 0.0500 (0.0500)  time: 0.6837  data: 0.0001  max mem: 57114
Epoch: [118]  [155/156]  eta: 0:00:00  lr: 0.005776  min_lr: 0.005776  loss: 3.3387 (3.2225)  weight_decay: 0.0500 (0.0500)  time: 0.6813  data: 0.0001  max mem: 57114
Epoch: [118] Total time: 0:01:53 (0.7277 s / it)
Averaged stats: lr: 0.005776  min_lr: 0.005776  loss: 3.3387 (3.2341)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8750 (0.8750)  acc1: 85.9375 (85.9375)  acc5: 97.3958 (97.3958)  time: 2.0215  data: 1.7658  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0395 (0.9685)  acc1: 80.2083 (80.0512)  acc5: 96.8750 (95.0128)  time: 0.5549  data: 0.3532  max mem: 57114
Test: Total time: 0:00:02 (0.5752 s / it)
* Acc@1 77.190 Acc@5 93.902 loss 1.103
Accuracy of the model on the 50000 test images: 77.2%
Max accuracy: 77.51%
Test:  [0/5]  eta: 0:00:11  loss: 4.2051 (4.2051)  acc1: 22.9167 (22.9167)  acc5: 45.8333 (45.8333)  time: 2.2770  data: 2.0337  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 4.6836 (4.4350)  acc1: 21.3542 (19.8210)  acc5: 37.5000 (39.0026)  time: 0.6060  data: 0.4068  max mem: 57114
Test: Total time: 0:00:03 (0.6198 s / it)
* Acc@1 16.496 Acc@5 32.695 loss 4.874
Accuracy of the model EMA on 50000 test images: 16.5%
Max EMA accuracy: 16.50%
Epoch: [119]  [  0/156]  eta: 0:07:44  lr: 0.005776  min_lr: 0.005776  loss: 2.4153 (2.4153)  weight_decay: 0.0500 (0.0500)  time: 2.9747  data: 2.3094  max mem: 57114
Epoch: [119]  [ 10/156]  eta: 0:02:17  lr: 0.005774  min_lr: 0.005774  loss: 3.1715 (3.0960)  weight_decay: 0.0500 (0.0500)  time: 0.9402  data: 0.2102  max mem: 57114
Epoch: [119]  [ 20/156]  eta: 0:01:53  lr: 0.005771  min_lr: 0.005771  loss: 3.1715 (3.1574)  weight_decay: 0.0500 (0.0500)  time: 0.7287  data: 0.0003  max mem: 57114
Epoch: [119]  [ 30/156]  eta: 0:01:40  lr: 0.005768  min_lr: 0.005768  loss: 3.1557 (3.1371)  weight_decay: 0.0500 (0.0500)  time: 0.7214  data: 0.0003  max mem: 57114
Epoch: [119]  [ 40/156]  eta: 0:01:30  lr: 0.005766  min_lr: 0.005766  loss: 3.1682 (3.1432)  weight_decay: 0.0500 (0.0500)  time: 0.7225  data: 0.0004  max mem: 57114
Epoch: [119]  [ 50/156]  eta: 0:01:21  lr: 0.005763  min_lr: 0.005763  loss: 3.3127 (3.1638)  weight_decay: 0.0500 (0.0500)  time: 0.7179  data: 0.0003  max mem: 57114
Epoch: [119]  [ 60/156]  eta: 0:01:13  lr: 0.005761  min_lr: 0.005761  loss: 3.3646 (3.1442)  weight_decay: 0.0500 (0.0500)  time: 0.7231  data: 0.0003  max mem: 57114
Epoch: [119]  [ 70/156]  eta: 0:01:04  lr: 0.005758  min_lr: 0.005758  loss: 3.0004 (3.1360)  weight_decay: 0.0500 (0.0500)  time: 0.7227  data: 0.0004  max mem: 57114
Epoch: [119]  [ 80/156]  eta: 0:00:56  lr: 0.005755  min_lr: 0.005755  loss: 3.3776 (3.1677)  weight_decay: 0.0500 (0.0500)  time: 0.7144  data: 0.0004  max mem: 57114
Epoch: [119]  [ 90/156]  eta: 0:00:49  lr: 0.005753  min_lr: 0.005753  loss: 3.3776 (3.1755)  weight_decay: 0.0500 (0.0500)  time: 0.7162  data: 0.0004  max mem: 57114
Epoch: [119]  [100/156]  eta: 0:00:41  lr: 0.005750  min_lr: 0.005750  loss: 3.3168 (3.1834)  weight_decay: 0.0500 (0.0500)  time: 0.7037  data: 0.0004  max mem: 57114
Epoch: [119]  [110/156]  eta: 0:00:33  lr: 0.005748  min_lr: 0.005748  loss: 3.1922 (3.1842)  weight_decay: 0.0500 (0.0500)  time: 0.6989  data: 0.0003  max mem: 57114
Epoch: [119]  [120/156]  eta: 0:00:26  lr: 0.005745  min_lr: 0.005745  loss: 3.2854 (3.1834)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0004  max mem: 57114
Epoch: [119]  [130/156]  eta: 0:00:19  lr: 0.005743  min_lr: 0.005743  loss: 3.4111 (3.2110)  weight_decay: 0.0500 (0.0500)  time: 0.7026  data: 0.0009  max mem: 57114
Epoch: [119]  [140/156]  eta: 0:00:11  lr: 0.005740  min_lr: 0.005740  loss: 3.3827 (3.2033)  weight_decay: 0.0500 (0.0500)  time: 0.6933  data: 0.0007  max mem: 57114
Epoch: [119]  [150/156]  eta: 0:00:04  lr: 0.005737  min_lr: 0.005737  loss: 2.9390 (3.1908)  weight_decay: 0.0500 (0.0500)  time: 0.6890  data: 0.0001  max mem: 57114
Epoch: [119]  [155/156]  eta: 0:00:00  lr: 0.005736  min_lr: 0.005736  loss: 2.9390 (3.1860)  weight_decay: 0.0500 (0.0500)  time: 0.6882  data: 0.0001  max mem: 57114
Epoch: [119] Total time: 0:01:53 (0.7274 s / it)
Averaged stats: lr: 0.005736  min_lr: 0.005736  loss: 2.9390 (3.2214)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9294 (0.9294)  acc1: 81.7708 (81.7708)  acc5: 94.7917 (94.7917)  time: 2.0440  data: 1.7865  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9969 (1.0052)  acc1: 80.2083 (77.2379)  acc5: 94.7917 (94.1177)  time: 0.5597  data: 0.3574  max mem: 57114
Test: Total time: 0:00:02 (0.5799 s / it)
* Acc@1 77.707 Acc@5 94.046 loss 1.066
Accuracy of the model on the 50000 test images: 77.7%
Max accuracy: 77.71%
Test:  [0/5]  eta: 0:00:10  loss: 3.9186 (3.9186)  acc1: 26.5625 (26.5625)  acc5: 50.5208 (50.5208)  time: 2.0721  data: 1.8285  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 4.3908 (4.1958)  acc1: 22.3958 (22.5064)  acc5: 42.7083 (42.5831)  time: 0.5650  data: 0.3658  max mem: 57114
Test: Total time: 0:00:02 (0.5776 s / it)
* Acc@1 19.006 Acc@5 36.423 loss 4.624
Accuracy of the model EMA on 50000 test images: 19.0%
Max EMA accuracy: 19.01%
Epoch: [120]  [  0/156]  eta: 0:06:17  lr: 0.005736  min_lr: 0.005736  loss: 3.3410 (3.3410)  weight_decay: 0.0500 (0.0500)  time: 2.4227  data: 1.7713  max mem: 57114
Epoch: [120]  [ 10/156]  eta: 0:02:07  lr: 0.005733  min_lr: 0.005733  loss: 3.3410 (3.2433)  weight_decay: 0.0500 (0.0500)  time: 0.8725  data: 0.1613  max mem: 57114
Epoch: [120]  [ 20/156]  eta: 0:01:49  lr: 0.005731  min_lr: 0.005731  loss: 3.1090 (3.1411)  weight_decay: 0.0500 (0.0500)  time: 0.7248  data: 0.0003  max mem: 57114
Epoch: [120]  [ 30/156]  eta: 0:01:38  lr: 0.005728  min_lr: 0.005728  loss: 3.1321 (3.1581)  weight_decay: 0.0500 (0.0500)  time: 0.7268  data: 0.0004  max mem: 57114
Epoch: [120]  [ 40/156]  eta: 0:01:29  lr: 0.005725  min_lr: 0.005725  loss: 3.1106 (3.1162)  weight_decay: 0.0500 (0.0500)  time: 0.7383  data: 0.0004  max mem: 57114
Epoch: [120]  [ 50/156]  eta: 0:01:21  lr: 0.005723  min_lr: 0.005723  loss: 3.2023 (3.1544)  weight_decay: 0.0500 (0.0500)  time: 0.7509  data: 0.0004  max mem: 57114
Epoch: [120]  [ 60/156]  eta: 0:01:13  lr: 0.005720  min_lr: 0.005720  loss: 3.3103 (3.1801)  weight_decay: 0.0500 (0.0500)  time: 0.7449  data: 0.0004  max mem: 57114
Epoch: [120]  [ 70/156]  eta: 0:01:04  lr: 0.005718  min_lr: 0.005718  loss: 3.2202 (3.1780)  weight_decay: 0.0500 (0.0500)  time: 0.7187  data: 0.0004  max mem: 57114
Epoch: [120]  [ 80/156]  eta: 0:00:57  lr: 0.005715  min_lr: 0.005715  loss: 3.3135 (3.2013)  weight_decay: 0.0500 (0.0500)  time: 0.7106  data: 0.0004  max mem: 57114
Epoch: [120]  [ 90/156]  eta: 0:00:49  lr: 0.005712  min_lr: 0.005712  loss: 3.3770 (3.2111)  weight_decay: 0.0500 (0.0500)  time: 0.7169  data: 0.0003  max mem: 57114
Epoch: [120]  [100/156]  eta: 0:00:41  lr: 0.005710  min_lr: 0.005710  loss: 3.2196 (3.2002)  weight_decay: 0.0500 (0.0500)  time: 0.7110  data: 0.0004  max mem: 57114
Epoch: [120]  [110/156]  eta: 0:00:34  lr: 0.005707  min_lr: 0.005707  loss: 3.1503 (3.1852)  weight_decay: 0.0500 (0.0500)  time: 0.7113  data: 0.0004  max mem: 57114
Epoch: [120]  [120/156]  eta: 0:00:26  lr: 0.005705  min_lr: 0.005705  loss: 3.2504 (3.1752)  weight_decay: 0.0500 (0.0500)  time: 0.7076  data: 0.0004  max mem: 57114
Epoch: [120]  [130/156]  eta: 0:00:19  lr: 0.005702  min_lr: 0.005702  loss: 3.1842 (3.1541)  weight_decay: 0.0500 (0.0500)  time: 0.6997  data: 0.0009  max mem: 57114
Epoch: [120]  [140/156]  eta: 0:00:11  lr: 0.005699  min_lr: 0.005699  loss: 3.1842 (3.1552)  weight_decay: 0.0500 (0.0500)  time: 0.6879  data: 0.0007  max mem: 57114
Epoch: [120]  [150/156]  eta: 0:00:04  lr: 0.005697  min_lr: 0.005697  loss: 3.3468 (3.1717)  weight_decay: 0.0500 (0.0500)  time: 0.6814  data: 0.0001  max mem: 57114
Epoch: [120]  [155/156]  eta: 0:00:00  lr: 0.005696  min_lr: 0.005696  loss: 3.4301 (3.1715)  weight_decay: 0.0500 (0.0500)  time: 0.6813  data: 0.0001  max mem: 57114
Epoch: [120] Total time: 0:01:53 (0.7276 s / it)
Averaged stats: lr: 0.005696  min_lr: 0.005696  loss: 3.4301 (3.2249)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9290 (0.9290)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.0551  data: 1.7995  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0135 (1.0649)  acc1: 81.2500 (78.3887)  acc5: 95.8333 (94.6292)  time: 0.5617  data: 0.3600  max mem: 57114
Test: Total time: 0:00:02 (0.5838 s / it)
* Acc@1 77.851 Acc@5 94.224 loss 1.101
Accuracy of the model on the 50000 test images: 77.9%
Max accuracy: 77.85%
Test:  [0/5]  eta: 0:00:10  loss: 3.6501 (3.6501)  acc1: 31.2500 (31.2500)  acc5: 53.1250 (53.1250)  time: 2.0186  data: 1.7751  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 4.1155 (3.9654)  acc1: 25.0000 (25.5754)  acc5: 47.3958 (46.4194)  time: 0.5544  data: 0.3551  max mem: 57114
Test: Total time: 0:00:02 (0.5653 s / it)
* Acc@1 21.429 Acc@5 40.123 loss 4.382
Accuracy of the model EMA on 50000 test images: 21.4%
Max EMA accuracy: 21.43%
Epoch: [121]  [  0/156]  eta: 0:07:03  lr: 0.005695  min_lr: 0.005695  loss: 3.4073 (3.4073)  weight_decay: 0.0500 (0.0500)  time: 2.7163  data: 1.9855  max mem: 57114
Epoch: [121]  [ 10/156]  eta: 0:02:11  lr: 0.005693  min_lr: 0.005693  loss: 3.5053 (3.3765)  weight_decay: 0.0500 (0.0500)  time: 0.9027  data: 0.1808  max mem: 57114
Epoch: [121]  [ 20/156]  eta: 0:01:53  lr: 0.005690  min_lr: 0.005690  loss: 3.5053 (3.4763)  weight_decay: 0.0500 (0.0500)  time: 0.7367  data: 0.0004  max mem: 57114
Epoch: [121]  [ 30/156]  eta: 0:01:41  lr: 0.005687  min_lr: 0.005687  loss: 3.4155 (3.3870)  weight_decay: 0.0500 (0.0500)  time: 0.7474  data: 0.0004  max mem: 57114
Epoch: [121]  [ 40/156]  eta: 0:01:31  lr: 0.005685  min_lr: 0.005685  loss: 3.2233 (3.3280)  weight_decay: 0.0500 (0.0500)  time: 0.7376  data: 0.0004  max mem: 57114
Epoch: [121]  [ 50/156]  eta: 0:01:22  lr: 0.005682  min_lr: 0.005682  loss: 3.2309 (3.3297)  weight_decay: 0.0500 (0.0500)  time: 0.7378  data: 0.0004  max mem: 57114
Epoch: [121]  [ 60/156]  eta: 0:01:13  lr: 0.005680  min_lr: 0.005680  loss: 3.2993 (3.3032)  weight_decay: 0.0500 (0.0500)  time: 0.7287  data: 0.0004  max mem: 57114
Epoch: [121]  [ 70/156]  eta: 0:01:05  lr: 0.005677  min_lr: 0.005677  loss: 3.1367 (3.2622)  weight_decay: 0.0500 (0.0500)  time: 0.7225  data: 0.0004  max mem: 57114
Epoch: [121]  [ 80/156]  eta: 0:00:57  lr: 0.005674  min_lr: 0.005674  loss: 3.1367 (3.2320)  weight_decay: 0.0500 (0.0500)  time: 0.7252  data: 0.0004  max mem: 57114
Epoch: [121]  [ 90/156]  eta: 0:00:49  lr: 0.005672  min_lr: 0.005672  loss: 3.1858 (3.2303)  weight_decay: 0.0500 (0.0500)  time: 0.7133  data: 0.0004  max mem: 57114
Epoch: [121]  [100/156]  eta: 0:00:41  lr: 0.005669  min_lr: 0.005669  loss: 3.3783 (3.2493)  weight_decay: 0.0500 (0.0500)  time: 0.7130  data: 0.0003  max mem: 57114
Epoch: [121]  [110/156]  eta: 0:00:34  lr: 0.005667  min_lr: 0.005667  loss: 3.4843 (3.2356)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0004  max mem: 57114
Epoch: [121]  [120/156]  eta: 0:00:26  lr: 0.005664  min_lr: 0.005664  loss: 3.3207 (3.2321)  weight_decay: 0.0500 (0.0500)  time: 0.7051  data: 0.0004  max mem: 57114
Epoch: [121]  [130/156]  eta: 0:00:19  lr: 0.005661  min_lr: 0.005661  loss: 3.3968 (3.2308)  weight_decay: 0.0500 (0.0500)  time: 0.7029  data: 0.0009  max mem: 57114
Epoch: [121]  [140/156]  eta: 0:00:11  lr: 0.005659  min_lr: 0.005659  loss: 3.3968 (3.2369)  weight_decay: 0.0500 (0.0500)  time: 0.6914  data: 0.0007  max mem: 57114
Epoch: [121]  [150/156]  eta: 0:00:04  lr: 0.005656  min_lr: 0.005656  loss: 3.3293 (3.2319)  weight_decay: 0.0500 (0.0500)  time: 0.6866  data: 0.0001  max mem: 57114
Epoch: [121]  [155/156]  eta: 0:00:00  lr: 0.005655  min_lr: 0.005655  loss: 3.3293 (3.2223)  weight_decay: 0.0500 (0.0500)  time: 0.6831  data: 0.0001  max mem: 57114
Epoch: [121] Total time: 0:01:54 (0.7309 s / it)
Averaged stats: lr: 0.005655  min_lr: 0.005655  loss: 3.3293 (3.2155)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7952 (0.7952)  acc1: 85.4167 (85.4167)  acc5: 97.9167 (97.9167)  time: 2.1141  data: 1.8583  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9399 (0.9621)  acc1: 79.1667 (77.8772)  acc5: 96.8750 (95.1407)  time: 0.5736  data: 0.3717  max mem: 57114
Test: Total time: 0:00:02 (0.5959 s / it)
* Acc@1 77.753 Acc@5 94.132 loss 1.044
Accuracy of the model on the 50000 test images: 77.8%
Max accuracy: 77.85%
Test:  [0/5]  eta: 0:00:11  loss: 3.3996 (3.3996)  acc1: 34.3750 (34.3750)  acc5: 56.7708 (56.7708)  time: 2.3639  data: 2.1205  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 3.8537 (3.7477)  acc1: 26.5625 (28.0051)  acc5: 50.5208 (49.2327)  time: 0.6234  data: 0.4242  max mem: 57114
Test: Total time: 0:00:03 (0.6410 s / it)
* Acc@1 23.967 Acc@5 43.848 loss 4.152
Accuracy of the model EMA on 50000 test images: 24.0%
Max EMA accuracy: 23.97%
Epoch: [122]  [  0/156]  eta: 0:07:13  lr: 0.005655  min_lr: 0.005655  loss: 3.6342 (3.6342)  weight_decay: 0.0500 (0.0500)  time: 2.7794  data: 2.1299  max mem: 57114
Epoch: [122]  [ 10/156]  eta: 0:02:14  lr: 0.005652  min_lr: 0.005652  loss: 2.8658 (2.8999)  weight_decay: 0.0500 (0.0500)  time: 0.9198  data: 0.1939  max mem: 57114
Epoch: [122]  [ 20/156]  eta: 0:01:52  lr: 0.005649  min_lr: 0.005649  loss: 3.1139 (3.0632)  weight_decay: 0.0500 (0.0500)  time: 0.7306  data: 0.0003  max mem: 57114
Epoch: [122]  [ 30/156]  eta: 0:01:39  lr: 0.005647  min_lr: 0.005647  loss: 3.2906 (3.1477)  weight_decay: 0.0500 (0.0500)  time: 0.7224  data: 0.0003  max mem: 57114
Epoch: [122]  [ 40/156]  eta: 0:01:29  lr: 0.005644  min_lr: 0.005644  loss: 3.2966 (3.2021)  weight_decay: 0.0500 (0.0500)  time: 0.7209  data: 0.0004  max mem: 57114
Epoch: [122]  [ 50/156]  eta: 0:01:21  lr: 0.005641  min_lr: 0.005641  loss: 3.5280 (3.2364)  weight_decay: 0.0500 (0.0500)  time: 0.7221  data: 0.0005  max mem: 57114
Epoch: [122]  [ 60/156]  eta: 0:01:12  lr: 0.005639  min_lr: 0.005639  loss: 3.2670 (3.1973)  weight_decay: 0.0500 (0.0500)  time: 0.7192  data: 0.0005  max mem: 57114
Epoch: [122]  [ 70/156]  eta: 0:01:04  lr: 0.005636  min_lr: 0.005636  loss: 3.2278 (3.2315)  weight_decay: 0.0500 (0.0500)  time: 0.7221  data: 0.0004  max mem: 57114
Epoch: [122]  [ 80/156]  eta: 0:00:56  lr: 0.005634  min_lr: 0.005634  loss: 3.4736 (3.2597)  weight_decay: 0.0500 (0.0500)  time: 0.7247  data: 0.0004  max mem: 57114
Epoch: [122]  [ 90/156]  eta: 0:00:49  lr: 0.005631  min_lr: 0.005631  loss: 3.3408 (3.2402)  weight_decay: 0.0500 (0.0500)  time: 0.7117  data: 0.0005  max mem: 57114
Epoch: [122]  [100/156]  eta: 0:00:41  lr: 0.005628  min_lr: 0.005628  loss: 3.0534 (3.2015)  weight_decay: 0.0500 (0.0500)  time: 0.6990  data: 0.0005  max mem: 57114
Epoch: [122]  [110/156]  eta: 0:00:33  lr: 0.005626  min_lr: 0.005626  loss: 3.0534 (3.1946)  weight_decay: 0.0500 (0.0500)  time: 0.7027  data: 0.0004  max mem: 57114
Epoch: [122]  [120/156]  eta: 0:00:26  lr: 0.005623  min_lr: 0.005623  loss: 3.1813 (3.1984)  weight_decay: 0.0500 (0.0500)  time: 0.7070  data: 0.0005  max mem: 57114
Epoch: [122]  [130/156]  eta: 0:00:19  lr: 0.005620  min_lr: 0.005620  loss: 3.3511 (3.1947)  weight_decay: 0.0500 (0.0500)  time: 0.7017  data: 0.0014  max mem: 57114
Epoch: [122]  [140/156]  eta: 0:00:11  lr: 0.005618  min_lr: 0.005618  loss: 3.4552 (3.2089)  weight_decay: 0.0500 (0.0500)  time: 0.6913  data: 0.0012  max mem: 57114
Epoch: [122]  [150/156]  eta: 0:00:04  lr: 0.005615  min_lr: 0.005615  loss: 3.5369 (3.2181)  weight_decay: 0.0500 (0.0500)  time: 0.6878  data: 0.0001  max mem: 57114
Epoch: [122]  [155/156]  eta: 0:00:00  lr: 0.005614  min_lr: 0.005614  loss: 3.4971 (3.2339)  weight_decay: 0.0500 (0.0500)  time: 0.6856  data: 0.0001  max mem: 57114
Epoch: [122] Total time: 0:01:53 (0.7260 s / it)
Averaged stats: lr: 0.005614  min_lr: 0.005614  loss: 3.4971 (3.2154)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8712 (0.8712)  acc1: 83.3333 (83.3333)  acc5: 98.4375 (98.4375)  time: 2.0781  data: 1.8222  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9839 (1.0192)  acc1: 79.1667 (78.1330)  acc5: 97.3958 (94.8849)  time: 0.5663  data: 0.3645  max mem: 57114
Test: Total time: 0:00:02 (0.5883 s / it)
* Acc@1 77.352 Acc@5 93.856 loss 1.101
Accuracy of the model on the 50000 test images: 77.4%
Max accuracy: 77.85%
Test:  [0/5]  eta: 0:00:11  loss: 3.1717 (3.1717)  acc1: 36.9792 (36.9792)  acc5: 63.5417 (63.5417)  time: 2.3606  data: 2.1173  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 3.6089 (3.5412)  acc1: 29.1667 (30.1790)  acc5: 57.2917 (53.7084)  time: 0.6228  data: 0.4236  max mem: 57114
Test: Total time: 0:00:03 (0.6365 s / it)
* Acc@1 26.381 Acc@5 47.305 loss 3.934
Accuracy of the model EMA on 50000 test images: 26.4%
Max EMA accuracy: 26.38%
Epoch: [123]  [  0/156]  eta: 0:07:07  lr: 0.005614  min_lr: 0.005614  loss: 3.3880 (3.3880)  weight_decay: 0.0500 (0.0500)  time: 2.7384  data: 2.0855  max mem: 57114
Epoch: [123]  [ 10/156]  eta: 0:02:14  lr: 0.005611  min_lr: 0.005611  loss: 3.1692 (3.1463)  weight_decay: 0.0500 (0.0500)  time: 0.9188  data: 0.1899  max mem: 57114
Epoch: [123]  [ 20/156]  eta: 0:01:51  lr: 0.005608  min_lr: 0.005608  loss: 3.1692 (3.2161)  weight_decay: 0.0500 (0.0500)  time: 0.7239  data: 0.0004  max mem: 57114
Epoch: [123]  [ 30/156]  eta: 0:01:39  lr: 0.005606  min_lr: 0.005606  loss: 3.5388 (3.2886)  weight_decay: 0.0500 (0.0500)  time: 0.7183  data: 0.0004  max mem: 57114
Epoch: [123]  [ 40/156]  eta: 0:01:30  lr: 0.005603  min_lr: 0.005603  loss: 3.4779 (3.3338)  weight_decay: 0.0500 (0.0500)  time: 0.7302  data: 0.0004  max mem: 57114
Epoch: [123]  [ 50/156]  eta: 0:01:21  lr: 0.005600  min_lr: 0.005600  loss: 3.4178 (3.3304)  weight_decay: 0.0500 (0.0500)  time: 0.7333  data: 0.0004  max mem: 57114
Epoch: [123]  [ 60/156]  eta: 0:01:13  lr: 0.005598  min_lr: 0.005598  loss: 3.3607 (3.3062)  weight_decay: 0.0500 (0.0500)  time: 0.7312  data: 0.0004  max mem: 57114
Epoch: [123]  [ 70/156]  eta: 0:01:04  lr: 0.005595  min_lr: 0.005595  loss: 3.0917 (3.2373)  weight_decay: 0.0500 (0.0500)  time: 0.7219  data: 0.0004  max mem: 57114
Epoch: [123]  [ 80/156]  eta: 0:00:56  lr: 0.005592  min_lr: 0.005592  loss: 2.9275 (3.2290)  weight_decay: 0.0500 (0.0500)  time: 0.7043  data: 0.0004  max mem: 57114
Epoch: [123]  [ 90/156]  eta: 0:00:49  lr: 0.005590  min_lr: 0.005590  loss: 3.3001 (3.2250)  weight_decay: 0.0500 (0.0500)  time: 0.7022  data: 0.0004  max mem: 57114
Epoch: [123]  [100/156]  eta: 0:00:41  lr: 0.005587  min_lr: 0.005587  loss: 3.2500 (3.2105)  weight_decay: 0.0500 (0.0500)  time: 0.7074  data: 0.0004  max mem: 57114
Epoch: [123]  [110/156]  eta: 0:00:33  lr: 0.005585  min_lr: 0.005585  loss: 3.2199 (3.2066)  weight_decay: 0.0500 (0.0500)  time: 0.7072  data: 0.0004  max mem: 57114
Epoch: [123]  [120/156]  eta: 0:00:26  lr: 0.005582  min_lr: 0.005582  loss: 3.2487 (3.2104)  weight_decay: 0.0500 (0.0500)  time: 0.7186  data: 0.0004  max mem: 57114
Epoch: [123]  [130/156]  eta: 0:00:19  lr: 0.005579  min_lr: 0.005579  loss: 3.2487 (3.2067)  weight_decay: 0.0500 (0.0500)  time: 0.7073  data: 0.0008  max mem: 57114
Epoch: [123]  [140/156]  eta: 0:00:11  lr: 0.005577  min_lr: 0.005577  loss: 3.1682 (3.2054)  weight_decay: 0.0500 (0.0500)  time: 0.6849  data: 0.0007  max mem: 57114
Epoch: [123]  [150/156]  eta: 0:00:04  lr: 0.005574  min_lr: 0.005574  loss: 3.1788 (3.2001)  weight_decay: 0.0500 (0.0500)  time: 0.6849  data: 0.0001  max mem: 57114
Epoch: [123]  [155/156]  eta: 0:00:00  lr: 0.005573  min_lr: 0.005573  loss: 3.2090 (3.2066)  weight_decay: 0.0500 (0.0500)  time: 0.6889  data: 0.0001  max mem: 57114
Epoch: [123] Total time: 0:01:53 (0.7272 s / it)
Averaged stats: lr: 0.005573  min_lr: 0.005573  loss: 3.2090 (3.2146)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9214 (0.9214)  acc1: 81.7708 (81.7708)  acc5: 97.3958 (97.3958)  time: 2.1101  data: 1.8538  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9833 (0.9550)  acc1: 80.7292 (78.2609)  acc5: 95.3125 (94.8849)  time: 0.5728  data: 0.3708  max mem: 57114
Test: Total time: 0:00:02 (0.5971 s / it)
* Acc@1 77.256 Acc@5 93.966 loss 1.066
Accuracy of the model on the 50000 test images: 77.3%
Max accuracy: 77.85%
Test:  [0/5]  eta: 0:00:11  loss: 2.9552 (2.9552)  acc1: 40.1042 (40.1042)  acc5: 67.7083 (67.7083)  time: 2.2993  data: 2.0559  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 3.3822 (3.3423)  acc1: 31.7708 (32.7366)  acc5: 60.4167 (56.9054)  time: 0.6106  data: 0.4113  max mem: 57114
Test: Total time: 0:00:03 (0.6238 s / it)
* Acc@1 28.916 Acc@5 50.569 loss 3.726
Accuracy of the model EMA on 50000 test images: 28.9%
Max EMA accuracy: 28.92%
Epoch: [124]  [  0/156]  eta: 0:07:13  lr: 0.005572  min_lr: 0.005572  loss: 3.4174 (3.4174)  weight_decay: 0.0500 (0.0500)  time: 2.7792  data: 2.1303  max mem: 57114
Epoch: [124]  [ 10/156]  eta: 0:02:10  lr: 0.005570  min_lr: 0.005570  loss: 3.3403 (3.2297)  weight_decay: 0.0500 (0.0500)  time: 0.8963  data: 0.1939  max mem: 57114
Epoch: [124]  [ 20/156]  eta: 0:01:50  lr: 0.005567  min_lr: 0.005567  loss: 3.3403 (3.2729)  weight_decay: 0.0500 (0.0500)  time: 0.7177  data: 0.0003  max mem: 57114
Epoch: [124]  [ 30/156]  eta: 0:01:39  lr: 0.005564  min_lr: 0.005564  loss: 3.1523 (3.1647)  weight_decay: 0.0500 (0.0500)  time: 0.7311  data: 0.0003  max mem: 57114
Epoch: [124]  [ 40/156]  eta: 0:01:29  lr: 0.005562  min_lr: 0.005562  loss: 3.0133 (3.1386)  weight_decay: 0.0500 (0.0500)  time: 0.7255  data: 0.0004  max mem: 57114
Epoch: [124]  [ 50/156]  eta: 0:01:20  lr: 0.005559  min_lr: 0.005559  loss: 3.0620 (3.0881)  weight_decay: 0.0500 (0.0500)  time: 0.7141  data: 0.0004  max mem: 57114
Epoch: [124]  [ 60/156]  eta: 0:01:12  lr: 0.005557  min_lr: 0.005557  loss: 3.0784 (3.1241)  weight_decay: 0.0500 (0.0500)  time: 0.7161  data: 0.0004  max mem: 57114
Epoch: [124]  [ 70/156]  eta: 0:01:04  lr: 0.005554  min_lr: 0.005554  loss: 3.4743 (3.1625)  weight_decay: 0.0500 (0.0500)  time: 0.7149  data: 0.0004  max mem: 57114
Epoch: [124]  [ 80/156]  eta: 0:00:56  lr: 0.005551  min_lr: 0.005551  loss: 3.5700 (3.2197)  weight_decay: 0.0500 (0.0500)  time: 0.7168  data: 0.0004  max mem: 57114
Epoch: [124]  [ 90/156]  eta: 0:00:48  lr: 0.005549  min_lr: 0.005549  loss: 3.4086 (3.2118)  weight_decay: 0.0500 (0.0500)  time: 0.7127  data: 0.0004  max mem: 57114
Epoch: [124]  [100/156]  eta: 0:00:41  lr: 0.005546  min_lr: 0.005546  loss: 3.3190 (3.2277)  weight_decay: 0.0500 (0.0500)  time: 0.7048  data: 0.0004  max mem: 57114
Epoch: [124]  [110/156]  eta: 0:00:33  lr: 0.005543  min_lr: 0.005543  loss: 3.5100 (3.2380)  weight_decay: 0.0500 (0.0500)  time: 0.7048  data: 0.0004  max mem: 57114
Epoch: [124]  [120/156]  eta: 0:00:26  lr: 0.005541  min_lr: 0.005541  loss: 3.3399 (3.2351)  weight_decay: 0.0500 (0.0500)  time: 0.7016  data: 0.0004  max mem: 57114
Epoch: [124]  [130/156]  eta: 0:00:18  lr: 0.005538  min_lr: 0.005538  loss: 3.2984 (3.2361)  weight_decay: 0.0500 (0.0500)  time: 0.6972  data: 0.0009  max mem: 57114
Epoch: [124]  [140/156]  eta: 0:00:11  lr: 0.005535  min_lr: 0.005535  loss: 3.4028 (3.2461)  weight_decay: 0.0500 (0.0500)  time: 0.6889  data: 0.0007  max mem: 57114
Epoch: [124]  [150/156]  eta: 0:00:04  lr: 0.005533  min_lr: 0.005533  loss: 3.4326 (3.2455)  weight_decay: 0.0500 (0.0500)  time: 0.6868  data: 0.0001  max mem: 57114
Epoch: [124]  [155/156]  eta: 0:00:00  lr: 0.005531  min_lr: 0.005531  loss: 3.3589 (3.2441)  weight_decay: 0.0500 (0.0500)  time: 0.6863  data: 0.0001  max mem: 57114
Epoch: [124] Total time: 0:01:52 (0.7234 s / it)
Averaged stats: lr: 0.005531  min_lr: 0.005531  loss: 3.3589 (3.2101)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 0.9680 (0.9680)  acc1: 86.9792 (86.9792)  acc5: 97.3958 (97.3958)  time: 2.2047  data: 1.9492  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1202 (1.1449)  acc1: 78.6458 (79.7954)  acc5: 97.3958 (94.6292)  time: 0.5915  data: 0.3899  max mem: 57114
Test: Total time: 0:00:03 (0.6115 s / it)
* Acc@1 77.829 Acc@5 94.142 loss 1.259
Accuracy of the model on the 50000 test images: 77.8%
Max accuracy: 77.85%
Test:  [0/5]  eta: 0:00:12  loss: 2.7563 (2.7563)  acc1: 43.2292 (43.2292)  acc5: 71.8750 (71.8750)  time: 2.4822  data: 2.2387  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 3.1783 (3.1542)  acc1: 34.8958 (35.2941)  acc5: 63.5417 (59.8465)  time: 0.6471  data: 0.4478  max mem: 57114
Test: Total time: 0:00:03 (0.6594 s / it)
* Acc@1 31.314 Acc@5 53.736 loss 3.529
Accuracy of the model EMA on 50000 test images: 31.3%
Max EMA accuracy: 31.31%
Epoch: [125]  [  0/156]  eta: 0:07:26  lr: 0.005531  min_lr: 0.005531  loss: 3.0864 (3.0864)  weight_decay: 0.0500 (0.0500)  time: 2.8619  data: 2.2075  max mem: 57114
Epoch: [125]  [ 10/156]  eta: 0:02:13  lr: 0.005528  min_lr: 0.005528  loss: 3.0864 (3.0006)  weight_decay: 0.0500 (0.0500)  time: 0.9121  data: 0.2009  max mem: 57114
Epoch: [125]  [ 20/156]  eta: 0:01:51  lr: 0.005526  min_lr: 0.005526  loss: 3.2645 (3.1157)  weight_decay: 0.0500 (0.0500)  time: 0.7162  data: 0.0003  max mem: 57114
Epoch: [125]  [ 30/156]  eta: 0:01:39  lr: 0.005523  min_lr: 0.005523  loss: 3.2645 (3.1341)  weight_decay: 0.0500 (0.0500)  time: 0.7258  data: 0.0003  max mem: 57114
Epoch: [125]  [ 40/156]  eta: 0:01:29  lr: 0.005520  min_lr: 0.005520  loss: 3.3568 (3.2049)  weight_decay: 0.0500 (0.0500)  time: 0.7229  data: 0.0004  max mem: 57114
Epoch: [125]  [ 50/156]  eta: 0:01:21  lr: 0.005518  min_lr: 0.005518  loss: 3.3219 (3.1898)  weight_decay: 0.0500 (0.0500)  time: 0.7284  data: 0.0004  max mem: 57114
Epoch: [125]  [ 60/156]  eta: 0:01:12  lr: 0.005515  min_lr: 0.005515  loss: 3.2714 (3.1875)  weight_decay: 0.0500 (0.0500)  time: 0.7338  data: 0.0004  max mem: 57114
Epoch: [125]  [ 70/156]  eta: 0:01:05  lr: 0.005512  min_lr: 0.005512  loss: 3.4745 (3.2480)  weight_decay: 0.0500 (0.0500)  time: 0.7283  data: 0.0004  max mem: 57114
Epoch: [125]  [ 80/156]  eta: 0:00:56  lr: 0.005510  min_lr: 0.005510  loss: 3.5404 (3.2588)  weight_decay: 0.0500 (0.0500)  time: 0.7189  data: 0.0004  max mem: 57114
Epoch: [125]  [ 90/156]  eta: 0:00:49  lr: 0.005507  min_lr: 0.005507  loss: 3.3392 (3.2371)  weight_decay: 0.0500 (0.0500)  time: 0.7065  data: 0.0004  max mem: 57114
Epoch: [125]  [100/156]  eta: 0:00:41  lr: 0.005504  min_lr: 0.005504  loss: 3.2377 (3.2283)  weight_decay: 0.0500 (0.0500)  time: 0.7113  data: 0.0004  max mem: 57114
Epoch: [125]  [110/156]  eta: 0:00:33  lr: 0.005502  min_lr: 0.005502  loss: 3.3526 (3.2420)  weight_decay: 0.0500 (0.0500)  time: 0.7081  data: 0.0004  max mem: 57114
Epoch: [125]  [120/156]  eta: 0:00:26  lr: 0.005499  min_lr: 0.005499  loss: 3.3812 (3.2440)  weight_decay: 0.0500 (0.0500)  time: 0.7126  data: 0.0004  max mem: 57114
Epoch: [125]  [130/156]  eta: 0:00:19  lr: 0.005496  min_lr: 0.005496  loss: 3.3812 (3.2505)  weight_decay: 0.0500 (0.0500)  time: 0.7069  data: 0.0008  max mem: 57114
Epoch: [125]  [140/156]  eta: 0:00:11  lr: 0.005494  min_lr: 0.005494  loss: 3.4270 (3.2470)  weight_decay: 0.0500 (0.0500)  time: 0.6917  data: 0.0007  max mem: 57114
Epoch: [125]  [150/156]  eta: 0:00:04  lr: 0.005491  min_lr: 0.005491  loss: 3.1040 (3.2312)  weight_decay: 0.0500 (0.0500)  time: 0.6883  data: 0.0001  max mem: 57114
Epoch: [125]  [155/156]  eta: 0:00:00  lr: 0.005490  min_lr: 0.005490  loss: 3.0640 (3.2227)  weight_decay: 0.0500 (0.0500)  time: 0.6831  data: 0.0001  max mem: 57114
Epoch: [125] Total time: 0:01:53 (0.7282 s / it)
Averaged stats: lr: 0.005490  min_lr: 0.005490  loss: 3.0640 (3.2055)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7994 (0.7994)  acc1: 83.8542 (83.8542)  acc5: 97.3958 (97.3958)  time: 2.0773  data: 1.8215  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8952 (0.9035)  acc1: 79.6875 (80.0512)  acc5: 96.3542 (93.8619)  time: 0.5661  data: 0.3644  max mem: 57114
Test: Total time: 0:00:02 (0.5921 s / it)
* Acc@1 78.203 Acc@5 94.443 loss 1.010
Accuracy of the model on the 50000 test images: 78.2%
Max accuracy: 78.20%
Test:  [0/5]  eta: 0:00:10  loss: 2.5731 (2.5731)  acc1: 44.7917 (44.7917)  acc5: 74.4792 (74.4792)  time: 2.0528  data: 1.8093  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 2.9978 (2.9804)  acc1: 39.0625 (38.2353)  acc5: 65.6250 (62.1483)  time: 0.5612  data: 0.3619  max mem: 57114
Test: Total time: 0:00:02 (0.5732 s / it)
* Acc@1 33.758 Acc@5 56.624 loss 3.350
Accuracy of the model EMA on 50000 test images: 33.8%
Max EMA accuracy: 33.76%
Epoch: [126]  [  0/156]  eta: 0:08:23  lr: 0.005489  min_lr: 0.005489  loss: 2.7455 (2.7455)  weight_decay: 0.0500 (0.0500)  time: 3.2281  data: 2.5795  max mem: 57114
Epoch: [126]  [ 10/156]  eta: 0:02:16  lr: 0.005487  min_lr: 0.005487  loss: 3.4330 (3.1263)  weight_decay: 0.0500 (0.0500)  time: 0.9367  data: 0.2347  max mem: 57114
Epoch: [126]  [ 20/156]  eta: 0:01:53  lr: 0.005484  min_lr: 0.005484  loss: 3.3097 (3.1746)  weight_decay: 0.0500 (0.0500)  time: 0.7140  data: 0.0003  max mem: 57114
Epoch: [126]  [ 30/156]  eta: 0:01:41  lr: 0.005481  min_lr: 0.005481  loss: 3.1832 (3.1534)  weight_decay: 0.0500 (0.0500)  time: 0.7363  data: 0.0003  max mem: 57114
Epoch: [126]  [ 40/156]  eta: 0:01:31  lr: 0.005479  min_lr: 0.005479  loss: 3.2290 (3.1666)  weight_decay: 0.0500 (0.0500)  time: 0.7444  data: 0.0004  max mem: 57114
Epoch: [126]  [ 50/156]  eta: 0:01:22  lr: 0.005476  min_lr: 0.005476  loss: 3.3707 (3.2053)  weight_decay: 0.0500 (0.0500)  time: 0.7374  data: 0.0004  max mem: 57114
Epoch: [126]  [ 60/156]  eta: 0:01:14  lr: 0.005473  min_lr: 0.005473  loss: 3.3211 (3.1621)  weight_decay: 0.0500 (0.0500)  time: 0.7480  data: 0.0004  max mem: 57114
Epoch: [126]  [ 70/156]  eta: 0:01:05  lr: 0.005471  min_lr: 0.005471  loss: 3.2061 (3.1963)  weight_decay: 0.0500 (0.0500)  time: 0.7324  data: 0.0004  max mem: 57114
Epoch: [126]  [ 80/156]  eta: 0:00:57  lr: 0.005468  min_lr: 0.005468  loss: 3.3601 (3.2029)  weight_decay: 0.0500 (0.0500)  time: 0.7072  data: 0.0004  max mem: 57114
Epoch: [126]  [ 90/156]  eta: 0:00:49  lr: 0.005465  min_lr: 0.005465  loss: 3.4123 (3.2075)  weight_decay: 0.0500 (0.0500)  time: 0.7018  data: 0.0004  max mem: 57114
Epoch: [126]  [100/156]  eta: 0:00:41  lr: 0.005463  min_lr: 0.005463  loss: 3.2928 (3.2002)  weight_decay: 0.0500 (0.0500)  time: 0.7046  data: 0.0003  max mem: 57114
Epoch: [126]  [110/156]  eta: 0:00:34  lr: 0.005460  min_lr: 0.005460  loss: 3.4460 (3.2234)  weight_decay: 0.0500 (0.0500)  time: 0.7171  data: 0.0003  max mem: 57114
Epoch: [126]  [120/156]  eta: 0:00:26  lr: 0.005457  min_lr: 0.005457  loss: 3.4077 (3.2039)  weight_decay: 0.0500 (0.0500)  time: 0.7082  data: 0.0004  max mem: 57114
Epoch: [126]  [130/156]  eta: 0:00:19  lr: 0.005455  min_lr: 0.005455  loss: 3.1595 (3.2059)  weight_decay: 0.0500 (0.0500)  time: 0.6943  data: 0.0009  max mem: 57114
Epoch: [126]  [140/156]  eta: 0:00:11  lr: 0.005452  min_lr: 0.005452  loss: 3.2204 (3.1923)  weight_decay: 0.0500 (0.0500)  time: 0.6888  data: 0.0007  max mem: 57114
Epoch: [126]  [150/156]  eta: 0:00:04  lr: 0.005449  min_lr: 0.005449  loss: 3.2204 (3.1926)  weight_decay: 0.0500 (0.0500)  time: 0.6892  data: 0.0001  max mem: 57114
Epoch: [126]  [155/156]  eta: 0:00:00  lr: 0.005448  min_lr: 0.005448  loss: 3.1203 (3.1886)  weight_decay: 0.0500 (0.0500)  time: 0.6927  data: 0.0001  max mem: 57114
Epoch: [126] Total time: 0:01:54 (0.7331 s / it)
Averaged stats: lr: 0.005448  min_lr: 0.005448  loss: 3.1203 (3.2026)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8597 (0.8597)  acc1: 81.7708 (81.7708)  acc5: 96.8750 (96.8750)  time: 2.1080  data: 1.8521  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9322 (0.9139)  acc1: 79.6875 (77.7494)  acc5: 95.3125 (94.7570)  time: 0.5722  data: 0.3705  max mem: 57114
Test: Total time: 0:00:02 (0.5911 s / it)
* Acc@1 78.207 Acc@5 94.463 loss 0.995
Accuracy of the model on the 50000 test images: 78.2%
Max accuracy: 78.21%
Test:  [0/5]  eta: 0:00:10  loss: 2.4107 (2.4107)  acc1: 47.9167 (47.9167)  acc5: 75.5208 (75.5208)  time: 2.0265  data: 1.7829  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 2.8385 (2.8238)  acc1: 40.1042 (39.8977)  acc5: 68.7500 (64.8338)  time: 0.5562  data: 0.3568  max mem: 57114
Test: Total time: 0:00:02 (0.5687 s / it)
* Acc@1 35.955 Acc@5 59.275 loss 3.186
Accuracy of the model EMA on 50000 test images: 36.0%
Max EMA accuracy: 35.96%
Epoch: [127]  [  0/156]  eta: 0:07:13  lr: 0.005448  min_lr: 0.005448  loss: 3.1202 (3.1202)  weight_decay: 0.0500 (0.0500)  time: 2.7809  data: 2.1179  max mem: 57114
Epoch: [127]  [ 10/156]  eta: 0:02:12  lr: 0.005445  min_lr: 0.005445  loss: 3.3376 (3.1800)  weight_decay: 0.0500 (0.0500)  time: 0.9088  data: 0.1929  max mem: 57114
Epoch: [127]  [ 20/156]  eta: 0:01:51  lr: 0.005442  min_lr: 0.005442  loss: 3.2205 (3.0701)  weight_decay: 0.0500 (0.0500)  time: 0.7230  data: 0.0004  max mem: 57114
Epoch: [127]  [ 30/156]  eta: 0:01:39  lr: 0.005440  min_lr: 0.005440  loss: 3.2422 (3.1039)  weight_decay: 0.0500 (0.0500)  time: 0.7255  data: 0.0003  max mem: 57114
Epoch: [127]  [ 40/156]  eta: 0:01:30  lr: 0.005437  min_lr: 0.005437  loss: 3.2987 (3.0458)  weight_decay: 0.0500 (0.0500)  time: 0.7329  data: 0.0004  max mem: 57114
Epoch: [127]  [ 50/156]  eta: 0:01:21  lr: 0.005434  min_lr: 0.005434  loss: 3.1676 (3.1070)  weight_decay: 0.0500 (0.0500)  time: 0.7434  data: 0.0004  max mem: 57114
Epoch: [127]  [ 60/156]  eta: 0:01:13  lr: 0.005432  min_lr: 0.005432  loss: 3.2970 (3.1231)  weight_decay: 0.0500 (0.0500)  time: 0.7442  data: 0.0004  max mem: 57114
Epoch: [127]  [ 70/156]  eta: 0:01:05  lr: 0.005429  min_lr: 0.005429  loss: 3.4689 (3.1573)  weight_decay: 0.0500 (0.0500)  time: 0.7407  data: 0.0004  max mem: 57114
Epoch: [127]  [ 80/156]  eta: 0:00:57  lr: 0.005426  min_lr: 0.005426  loss: 3.2919 (3.1640)  weight_decay: 0.0500 (0.0500)  time: 0.7312  data: 0.0004  max mem: 57114
Epoch: [127]  [ 90/156]  eta: 0:00:49  lr: 0.005424  min_lr: 0.005424  loss: 3.2601 (3.1852)  weight_decay: 0.0500 (0.0500)  time: 0.7107  data: 0.0004  max mem: 57114
Epoch: [127]  [100/156]  eta: 0:00:41  lr: 0.005421  min_lr: 0.005421  loss: 3.0685 (3.1703)  weight_decay: 0.0500 (0.0500)  time: 0.6963  data: 0.0004  max mem: 57114
Epoch: [127]  [110/156]  eta: 0:00:34  lr: 0.005418  min_lr: 0.005418  loss: 3.2083 (3.1865)  weight_decay: 0.0500 (0.0500)  time: 0.7023  data: 0.0004  max mem: 57114
Epoch: [127]  [120/156]  eta: 0:00:26  lr: 0.005416  min_lr: 0.005416  loss: 3.5054 (3.1970)  weight_decay: 0.0500 (0.0500)  time: 0.7109  data: 0.0004  max mem: 57114
Epoch: [127]  [130/156]  eta: 0:00:19  lr: 0.005413  min_lr: 0.005413  loss: 3.3679 (3.2009)  weight_decay: 0.0500 (0.0500)  time: 0.6995  data: 0.0009  max mem: 57114
Epoch: [127]  [140/156]  eta: 0:00:11  lr: 0.005410  min_lr: 0.005410  loss: 3.2107 (3.2006)  weight_decay: 0.0500 (0.0500)  time: 0.6866  data: 0.0007  max mem: 57114
Epoch: [127]  [150/156]  eta: 0:00:04  lr: 0.005407  min_lr: 0.005407  loss: 3.2107 (3.1920)  weight_decay: 0.0500 (0.0500)  time: 0.6849  data: 0.0001  max mem: 57114
Epoch: [127]  [155/156]  eta: 0:00:00  lr: 0.005406  min_lr: 0.005406  loss: 3.3517 (3.2007)  weight_decay: 0.0500 (0.0500)  time: 0.6836  data: 0.0001  max mem: 57114
Epoch: [127] Total time: 0:01:53 (0.7303 s / it)
Averaged stats: lr: 0.005406  min_lr: 0.005406  loss: 3.3517 (3.1962)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8597 (0.8597)  acc1: 83.3333 (83.3333)  acc5: 98.9583 (98.9583)  time: 2.0389  data: 1.7834  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0050 (0.9792)  acc1: 81.7708 (79.0281)  acc5: 95.8333 (94.6292)  time: 0.5585  data: 0.3567  max mem: 57114
Test: Total time: 0:00:02 (0.5788 s / it)
* Acc@1 78.379 Acc@5 94.479 loss 1.088
Accuracy of the model on the 50000 test images: 78.4%
Max accuracy: 78.38%
Test:  [0/5]  eta: 0:00:11  loss: 2.2632 (2.2632)  acc1: 50.5208 (50.5208)  acc5: 78.1250 (78.1250)  time: 2.2054  data: 1.9619  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 2.6951 (2.6783)  acc1: 42.1875 (42.0716)  acc5: 71.3542 (67.0077)  time: 0.5917  data: 0.3925  max mem: 57114
Test: Total time: 0:00:03 (0.6061 s / it)
* Acc@1 37.990 Acc@5 61.813 loss 3.034
Accuracy of the model EMA on 50000 test images: 38.0%
Max EMA accuracy: 37.99%
Epoch: [128]  [  0/156]  eta: 0:07:33  lr: 0.005406  min_lr: 0.005406  loss: 3.4155 (3.4155)  weight_decay: 0.0500 (0.0500)  time: 2.9102  data: 2.2600  max mem: 57114
Epoch: [128]  [ 10/156]  eta: 0:02:16  lr: 0.005403  min_lr: 0.005403  loss: 2.9810 (3.1389)  weight_decay: 0.0500 (0.0500)  time: 0.9350  data: 0.2057  max mem: 57114
Epoch: [128]  [ 20/156]  eta: 0:01:54  lr: 0.005400  min_lr: 0.005400  loss: 3.0561 (3.1474)  weight_decay: 0.0500 (0.0500)  time: 0.7408  data: 0.0003  max mem: 57114
Epoch: [128]  [ 30/156]  eta: 0:01:41  lr: 0.005398  min_lr: 0.005398  loss: 3.0944 (3.1231)  weight_decay: 0.0500 (0.0500)  time: 0.7322  data: 0.0003  max mem: 57114
Epoch: [128]  [ 40/156]  eta: 0:01:31  lr: 0.005395  min_lr: 0.005395  loss: 3.0944 (3.1388)  weight_decay: 0.0500 (0.0500)  time: 0.7372  data: 0.0004  max mem: 57114
Epoch: [128]  [ 50/156]  eta: 0:01:22  lr: 0.005392  min_lr: 0.005392  loss: 3.2982 (3.1738)  weight_decay: 0.0500 (0.0500)  time: 0.7427  data: 0.0004  max mem: 57114
Epoch: [128]  [ 60/156]  eta: 0:01:14  lr: 0.005390  min_lr: 0.005390  loss: 3.3593 (3.1693)  weight_decay: 0.0500 (0.0500)  time: 0.7372  data: 0.0004  max mem: 57114
Epoch: [128]  [ 70/156]  eta: 0:01:05  lr: 0.005387  min_lr: 0.005387  loss: 3.3170 (3.1987)  weight_decay: 0.0500 (0.0500)  time: 0.7334  data: 0.0004  max mem: 57114
Epoch: [128]  [ 80/156]  eta: 0:00:57  lr: 0.005384  min_lr: 0.005384  loss: 3.3460 (3.2176)  weight_decay: 0.0500 (0.0500)  time: 0.7209  data: 0.0004  max mem: 57114
Epoch: [128]  [ 90/156]  eta: 0:00:49  lr: 0.005382  min_lr: 0.005382  loss: 3.3033 (3.2097)  weight_decay: 0.0500 (0.0500)  time: 0.7105  data: 0.0004  max mem: 57114
Epoch: [128]  [100/156]  eta: 0:00:42  lr: 0.005379  min_lr: 0.005379  loss: 3.3596 (3.2222)  weight_decay: 0.0500 (0.0500)  time: 0.7061  data: 0.0004  max mem: 57114
Epoch: [128]  [110/156]  eta: 0:00:34  lr: 0.005376  min_lr: 0.005376  loss: 3.5529 (3.2437)  weight_decay: 0.0500 (0.0500)  time: 0.7132  data: 0.0004  max mem: 57114
Epoch: [128]  [120/156]  eta: 0:00:26  lr: 0.005373  min_lr: 0.005373  loss: 3.2476 (3.2343)  weight_decay: 0.0500 (0.0500)  time: 0.7175  data: 0.0004  max mem: 57114
Epoch: [128]  [130/156]  eta: 0:00:19  lr: 0.005371  min_lr: 0.005371  loss: 3.3170 (3.2531)  weight_decay: 0.0500 (0.0500)  time: 0.7065  data: 0.0008  max mem: 57114
Epoch: [128]  [140/156]  eta: 0:00:11  lr: 0.005368  min_lr: 0.005368  loss: 3.3911 (3.2449)  weight_decay: 0.0500 (0.0500)  time: 0.6877  data: 0.0007  max mem: 57114
Epoch: [128]  [150/156]  eta: 0:00:04  lr: 0.005365  min_lr: 0.005365  loss: 3.1131 (3.2274)  weight_decay: 0.0500 (0.0500)  time: 0.6854  data: 0.0001  max mem: 57114
Epoch: [128]  [155/156]  eta: 0:00:00  lr: 0.005364  min_lr: 0.005364  loss: 3.1131 (3.2194)  weight_decay: 0.0500 (0.0500)  time: 0.6879  data: 0.0001  max mem: 57114
Epoch: [128] Total time: 0:01:54 (0.7342 s / it)
Averaged stats: lr: 0.005364  min_lr: 0.005364  loss: 3.1131 (3.1960)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8612 (0.8612)  acc1: 83.8542 (83.8542)  acc5: 96.8750 (96.8750)  time: 2.0788  data: 1.8225  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0472 (0.9972)  acc1: 80.7292 (79.1560)  acc5: 96.3542 (94.8849)  time: 0.5665  data: 0.3646  max mem: 57114
Test: Total time: 0:00:02 (0.5909 s / it)
* Acc@1 77.899 Acc@5 94.246 loss 1.072
Accuracy of the model on the 50000 test images: 77.9%
Max accuracy: 78.38%
Test:  [0/5]  eta: 0:00:11  loss: 2.1291 (2.1291)  acc1: 54.6875 (54.6875)  acc5: 79.1667 (79.1667)  time: 2.3852  data: 2.1417  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 2.5642 (2.5435)  acc1: 45.8333 (45.1407)  acc5: 73.4375 (68.6701)  time: 0.6277  data: 0.4284  max mem: 57114
Test: Total time: 0:00:03 (0.6399 s / it)
* Acc@1 40.185 Acc@5 64.146 loss 2.892
Accuracy of the model EMA on 50000 test images: 40.2%
Max EMA accuracy: 40.19%
Epoch: [129]  [  0/156]  eta: 0:07:07  lr: 0.005364  min_lr: 0.005364  loss: 3.8080 (3.8080)  weight_decay: 0.0500 (0.0500)  time: 2.7385  data: 2.0862  max mem: 57114
Epoch: [129]  [ 10/156]  eta: 0:02:14  lr: 0.005361  min_lr: 0.005361  loss: 3.2138 (3.1780)  weight_decay: 0.0500 (0.0500)  time: 0.9202  data: 0.1899  max mem: 57114
Epoch: [129]  [ 20/156]  eta: 0:01:51  lr: 0.005358  min_lr: 0.005358  loss: 3.1525 (3.1559)  weight_decay: 0.0500 (0.0500)  time: 0.7235  data: 0.0003  max mem: 57114
Epoch: [129]  [ 30/156]  eta: 0:01:39  lr: 0.005356  min_lr: 0.005356  loss: 3.1772 (3.1789)  weight_decay: 0.0500 (0.0500)  time: 0.7136  data: 0.0003  max mem: 57114
Epoch: [129]  [ 40/156]  eta: 0:01:29  lr: 0.005353  min_lr: 0.005353  loss: 3.2351 (3.1768)  weight_decay: 0.0500 (0.0500)  time: 0.7170  data: 0.0003  max mem: 57114
Epoch: [129]  [ 50/156]  eta: 0:01:20  lr: 0.005350  min_lr: 0.005350  loss: 3.3275 (3.1978)  weight_decay: 0.0500 (0.0500)  time: 0.7245  data: 0.0004  max mem: 57114
Epoch: [129]  [ 60/156]  eta: 0:01:12  lr: 0.005347  min_lr: 0.005347  loss: 3.0467 (3.1208)  weight_decay: 0.0500 (0.0500)  time: 0.7253  data: 0.0004  max mem: 57114
Epoch: [129]  [ 70/156]  eta: 0:01:04  lr: 0.005345  min_lr: 0.005345  loss: 2.8959 (3.1326)  weight_decay: 0.0500 (0.0500)  time: 0.7159  data: 0.0004  max mem: 57114
Epoch: [129]  [ 80/156]  eta: 0:00:56  lr: 0.005342  min_lr: 0.005342  loss: 3.1724 (3.1410)  weight_decay: 0.0500 (0.0500)  time: 0.7121  data: 0.0004  max mem: 57114
Epoch: [129]  [ 90/156]  eta: 0:00:48  lr: 0.005339  min_lr: 0.005339  loss: 3.2980 (3.1418)  weight_decay: 0.0500 (0.0500)  time: 0.7033  data: 0.0004  max mem: 57114
Epoch: [129]  [100/156]  eta: 0:00:41  lr: 0.005337  min_lr: 0.005337  loss: 3.2193 (3.1415)  weight_decay: 0.0500 (0.0500)  time: 0.7045  data: 0.0004  max mem: 57114
Epoch: [129]  [110/156]  eta: 0:00:33  lr: 0.005334  min_lr: 0.005334  loss: 3.2340 (3.1422)  weight_decay: 0.0500 (0.0500)  time: 0.7058  data: 0.0004  max mem: 57114
Epoch: [129]  [120/156]  eta: 0:00:26  lr: 0.005331  min_lr: 0.005331  loss: 3.4618 (3.1592)  weight_decay: 0.0500 (0.0500)  time: 0.7148  data: 0.0004  max mem: 57114
Epoch: [129]  [130/156]  eta: 0:00:18  lr: 0.005329  min_lr: 0.005329  loss: 3.2008 (3.1390)  weight_decay: 0.0500 (0.0500)  time: 0.7111  data: 0.0009  max mem: 57114
Epoch: [129]  [140/156]  eta: 0:00:11  lr: 0.005326  min_lr: 0.005326  loss: 3.0633 (3.1498)  weight_decay: 0.0500 (0.0500)  time: 0.6904  data: 0.0007  max mem: 57114
Epoch: [129]  [150/156]  eta: 0:00:04  lr: 0.005323  min_lr: 0.005323  loss: 3.2466 (3.1509)  weight_decay: 0.0500 (0.0500)  time: 0.6883  data: 0.0001  max mem: 57114
Epoch: [129]  [155/156]  eta: 0:00:00  lr: 0.005322  min_lr: 0.005322  loss: 3.2466 (3.1561)  weight_decay: 0.0500 (0.0500)  time: 0.6851  data: 0.0001  max mem: 57114
Epoch: [129] Total time: 0:01:53 (0.7246 s / it)
Averaged stats: lr: 0.005322  min_lr: 0.005322  loss: 3.2466 (3.1951)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 0.9103 (0.9103)  acc1: 87.5000 (87.5000)  acc5: 97.9167 (97.9167)  time: 2.2171  data: 1.9576  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0562 (1.0549)  acc1: 81.7708 (80.4348)  acc5: 96.3542 (94.8849)  time: 0.5944  data: 0.3917  max mem: 57114
Test: Total time: 0:00:03 (0.6183 s / it)
* Acc@1 78.249 Acc@5 94.230 loss 1.146
Accuracy of the model on the 50000 test images: 78.2%
Max accuracy: 78.38%
Test:  [0/5]  eta: 0:00:11  loss: 2.0128 (2.0128)  acc1: 57.8125 (57.8125)  acc5: 80.2083 (80.2083)  time: 2.3180  data: 2.0745  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 2.4495 (2.4234)  acc1: 47.3958 (47.0588)  acc5: 74.4792 (70.3325)  time: 0.6143  data: 0.4150  max mem: 57114
Test: Total time: 0:00:03 (0.6287 s / it)
* Acc@1 42.074 Acc@5 66.312 loss 2.763
Accuracy of the model EMA on 50000 test images: 42.1%
Max EMA accuracy: 42.07%
Epoch: [130]  [  0/156]  eta: 0:06:10  lr: 0.005321  min_lr: 0.005321  loss: 3.6640 (3.6640)  weight_decay: 0.0500 (0.0500)  time: 2.3742  data: 1.7192  max mem: 57114
Epoch: [130]  [ 10/156]  eta: 0:02:05  lr: 0.005319  min_lr: 0.005319  loss: 3.5940 (3.3331)  weight_decay: 0.0500 (0.0500)  time: 0.8599  data: 0.1565  max mem: 57114
Epoch: [130]  [ 20/156]  eta: 0:01:49  lr: 0.005316  min_lr: 0.005316  loss: 3.4676 (3.3104)  weight_decay: 0.0500 (0.0500)  time: 0.7298  data: 0.0003  max mem: 57114
Epoch: [130]  [ 30/156]  eta: 0:01:38  lr: 0.005313  min_lr: 0.005313  loss: 3.3589 (3.2398)  weight_decay: 0.0500 (0.0500)  time: 0.7389  data: 0.0003  max mem: 57114
Epoch: [130]  [ 40/156]  eta: 0:01:29  lr: 0.005311  min_lr: 0.005311  loss: 3.3071 (3.2324)  weight_decay: 0.0500 (0.0500)  time: 0.7258  data: 0.0003  max mem: 57114
Epoch: [130]  [ 50/156]  eta: 0:01:20  lr: 0.005308  min_lr: 0.005308  loss: 3.1167 (3.2077)  weight_decay: 0.0500 (0.0500)  time: 0.7314  data: 0.0004  max mem: 57114
Epoch: [130]  [ 60/156]  eta: 0:01:12  lr: 0.005305  min_lr: 0.005305  loss: 3.1600 (3.2302)  weight_decay: 0.0500 (0.0500)  time: 0.7247  data: 0.0004  max mem: 57114
Epoch: [130]  [ 70/156]  eta: 0:01:04  lr: 0.005302  min_lr: 0.005302  loss: 3.0815 (3.1812)  weight_decay: 0.0500 (0.0500)  time: 0.7029  data: 0.0004  max mem: 57114
Epoch: [130]  [ 80/156]  eta: 0:00:56  lr: 0.005300  min_lr: 0.005300  loss: 2.8395 (3.1487)  weight_decay: 0.0500 (0.0500)  time: 0.6972  data: 0.0004  max mem: 57114
Epoch: [130]  [ 90/156]  eta: 0:00:48  lr: 0.005297  min_lr: 0.005297  loss: 3.1462 (3.1782)  weight_decay: 0.0500 (0.0500)  time: 0.7106  data: 0.0003  max mem: 57114
Epoch: [130]  [100/156]  eta: 0:00:41  lr: 0.005294  min_lr: 0.005294  loss: 3.3612 (3.1846)  weight_decay: 0.0500 (0.0500)  time: 0.7239  data: 0.0004  max mem: 57114
Epoch: [130]  [110/156]  eta: 0:00:33  lr: 0.005292  min_lr: 0.005292  loss: 3.3511 (3.1963)  weight_decay: 0.0500 (0.0500)  time: 0.7193  data: 0.0004  max mem: 57114
Epoch: [130]  [120/156]  eta: 0:00:26  lr: 0.005289  min_lr: 0.005289  loss: 3.3471 (3.1848)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0004  max mem: 57114
Epoch: [130]  [130/156]  eta: 0:00:18  lr: 0.005286  min_lr: 0.005286  loss: 3.1380 (3.1795)  weight_decay: 0.0500 (0.0500)  time: 0.7003  data: 0.0008  max mem: 57114
Epoch: [130]  [140/156]  eta: 0:00:11  lr: 0.005283  min_lr: 0.005283  loss: 3.4206 (3.1939)  weight_decay: 0.0500 (0.0500)  time: 0.6851  data: 0.0007  max mem: 57114
Epoch: [130]  [150/156]  eta: 0:00:04  lr: 0.005281  min_lr: 0.005281  loss: 3.4505 (3.2040)  weight_decay: 0.0500 (0.0500)  time: 0.6785  data: 0.0001  max mem: 57114
Epoch: [130]  [155/156]  eta: 0:00:00  lr: 0.005279  min_lr: 0.005279  loss: 3.4262 (3.2019)  weight_decay: 0.0500 (0.0500)  time: 0.6829  data: 0.0001  max mem: 57114
Epoch: [130] Total time: 0:01:52 (0.7238 s / it)
Averaged stats: lr: 0.005279  min_lr: 0.005279  loss: 3.4262 (3.1854)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8130 (0.8130)  acc1: 84.3750 (84.3750)  acc5: 97.9167 (97.9167)  time: 2.0427  data: 1.7868  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9661 (0.9490)  acc1: 79.6875 (78.5166)  acc5: 96.3542 (94.3734)  time: 0.5592  data: 0.3574  max mem: 57114
Test: Total time: 0:00:02 (0.5834 s / it)
* Acc@1 77.709 Acc@5 93.978 loss 1.038
Accuracy of the model on the 50000 test images: 77.7%
Max accuracy: 78.38%
Test:  [0/5]  eta: 0:00:11  loss: 1.9109 (1.9109)  acc1: 58.8542 (58.8542)  acc5: 80.2083 (80.2083)  time: 2.3074  data: 2.0638  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 2.3469 (2.3149)  acc1: 48.4375 (47.9540)  acc5: 76.5625 (71.9949)  time: 0.6122  data: 0.4128  max mem: 57114
Test: Total time: 0:00:03 (0.6239 s / it)
* Acc@1 43.914 Acc@5 68.264 loss 2.647
Accuracy of the model EMA on 50000 test images: 43.9%
Max EMA accuracy: 43.91%
Epoch: [131]  [  0/156]  eta: 0:06:48  lr: 0.005279  min_lr: 0.005279  loss: 3.0472 (3.0472)  weight_decay: 0.0500 (0.0500)  time: 2.6203  data: 1.8317  max mem: 57114
Epoch: [131]  [ 10/156]  eta: 0:02:09  lr: 0.005276  min_lr: 0.005276  loss: 3.0303 (2.9676)  weight_decay: 0.0500 (0.0500)  time: 0.8869  data: 0.1668  max mem: 57114
Epoch: [131]  [ 20/156]  eta: 0:01:50  lr: 0.005274  min_lr: 0.005274  loss: 3.2569 (3.1148)  weight_decay: 0.0500 (0.0500)  time: 0.7243  data: 0.0003  max mem: 57114
Epoch: [131]  [ 30/156]  eta: 0:01:39  lr: 0.005271  min_lr: 0.005271  loss: 3.3658 (3.1377)  weight_decay: 0.0500 (0.0500)  time: 0.7315  data: 0.0003  max mem: 57114
Epoch: [131]  [ 40/156]  eta: 0:01:29  lr: 0.005268  min_lr: 0.005268  loss: 3.2489 (3.1813)  weight_decay: 0.0500 (0.0500)  time: 0.7240  data: 0.0004  max mem: 57114
Epoch: [131]  [ 50/156]  eta: 0:01:20  lr: 0.005265  min_lr: 0.005265  loss: 3.2577 (3.2089)  weight_decay: 0.0500 (0.0500)  time: 0.7141  data: 0.0004  max mem: 57114
Epoch: [131]  [ 60/156]  eta: 0:01:12  lr: 0.005263  min_lr: 0.005263  loss: 3.2432 (3.2003)  weight_decay: 0.0500 (0.0500)  time: 0.7153  data: 0.0004  max mem: 57114
Epoch: [131]  [ 70/156]  eta: 0:01:04  lr: 0.005260  min_lr: 0.005260  loss: 3.2523 (3.1842)  weight_decay: 0.0500 (0.0500)  time: 0.7210  data: 0.0004  max mem: 57114
Epoch: [131]  [ 80/156]  eta: 0:00:56  lr: 0.005257  min_lr: 0.005257  loss: 3.2767 (3.1947)  weight_decay: 0.0500 (0.0500)  time: 0.7162  data: 0.0003  max mem: 57114
Epoch: [131]  [ 90/156]  eta: 0:00:48  lr: 0.005254  min_lr: 0.005254  loss: 3.3351 (3.2033)  weight_decay: 0.0500 (0.0500)  time: 0.7144  data: 0.0003  max mem: 57114
Epoch: [131]  [100/156]  eta: 0:00:41  lr: 0.005252  min_lr: 0.005252  loss: 3.4480 (3.2002)  weight_decay: 0.0500 (0.0500)  time: 0.7132  data: 0.0003  max mem: 57114
Epoch: [131]  [110/156]  eta: 0:00:33  lr: 0.005249  min_lr: 0.005249  loss: 3.3529 (3.2138)  weight_decay: 0.0500 (0.0500)  time: 0.7083  data: 0.0004  max mem: 57114
Epoch: [131]  [120/156]  eta: 0:00:26  lr: 0.005246  min_lr: 0.005246  loss: 3.1567 (3.1965)  weight_decay: 0.0500 (0.0500)  time: 0.7037  data: 0.0004  max mem: 57114
Epoch: [131]  [130/156]  eta: 0:00:18  lr: 0.005244  min_lr: 0.005244  loss: 3.1326 (3.2034)  weight_decay: 0.0500 (0.0500)  time: 0.6952  data: 0.0009  max mem: 57114
Epoch: [131]  [140/156]  eta: 0:00:11  lr: 0.005241  min_lr: 0.005241  loss: 3.3101 (3.2088)  weight_decay: 0.0500 (0.0500)  time: 0.6839  data: 0.0008  max mem: 57114
Epoch: [131]  [150/156]  eta: 0:00:04  lr: 0.005238  min_lr: 0.005238  loss: 3.4211 (3.2213)  weight_decay: 0.0500 (0.0500)  time: 0.6862  data: 0.0001  max mem: 57114
Epoch: [131]  [155/156]  eta: 0:00:00  lr: 0.005237  min_lr: 0.005237  loss: 3.4076 (3.2197)  weight_decay: 0.0500 (0.0500)  time: 0.6886  data: 0.0001  max mem: 57114
Epoch: [131] Total time: 0:01:52 (0.7240 s / it)
Averaged stats: lr: 0.005237  min_lr: 0.005237  loss: 3.4076 (3.1837)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8272 (0.8272)  acc1: 84.8958 (84.8958)  acc5: 96.3542 (96.3542)  time: 2.1725  data: 1.9170  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8848 (0.9353)  acc1: 80.7292 (77.2379)  acc5: 96.3542 (94.7570)  time: 0.5852  data: 0.3835  max mem: 57114
Test: Total time: 0:00:03 (0.6060 s / it)
* Acc@1 77.939 Acc@5 94.122 loss 1.020
Accuracy of the model on the 50000 test images: 77.9%
Max accuracy: 78.38%
Test:  [0/5]  eta: 0:00:11  loss: 1.8143 (1.8143)  acc1: 60.9375 (60.9375)  acc5: 81.2500 (81.2500)  time: 2.2638  data: 2.0202  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 2.2482 (2.2103)  acc1: 52.6042 (49.8721)  acc5: 78.1250 (73.6573)  time: 0.6034  data: 0.4041  max mem: 57114
Test: Total time: 0:00:03 (0.6154 s / it)
* Acc@1 45.776 Acc@5 70.097 loss 2.535
Accuracy of the model EMA on 50000 test images: 45.8%
Max EMA accuracy: 45.78%
Epoch: [132]  [  0/156]  eta: 0:06:38  lr: 0.005236  min_lr: 0.005236  loss: 3.5424 (3.5424)  weight_decay: 0.0500 (0.0500)  time: 2.5521  data: 1.9048  max mem: 57114
Epoch: [132]  [ 10/156]  eta: 0:02:09  lr: 0.005234  min_lr: 0.005234  loss: 3.5279 (3.3280)  weight_decay: 0.0500 (0.0500)  time: 0.8853  data: 0.1734  max mem: 57114
Epoch: [132]  [ 20/156]  eta: 0:01:51  lr: 0.005231  min_lr: 0.005231  loss: 3.3321 (3.3592)  weight_decay: 0.0500 (0.0500)  time: 0.7308  data: 0.0003  max mem: 57114
Epoch: [132]  [ 30/156]  eta: 0:01:38  lr: 0.005228  min_lr: 0.005228  loss: 3.3321 (3.3268)  weight_decay: 0.0500 (0.0500)  time: 0.7205  data: 0.0004  max mem: 57114
Epoch: [132]  [ 40/156]  eta: 0:01:28  lr: 0.005225  min_lr: 0.005225  loss: 3.2188 (3.2401)  weight_decay: 0.0500 (0.0500)  time: 0.7082  data: 0.0004  max mem: 57114
Epoch: [132]  [ 50/156]  eta: 0:01:19  lr: 0.005223  min_lr: 0.005223  loss: 3.1271 (3.2002)  weight_decay: 0.0500 (0.0500)  time: 0.7141  data: 0.0004  max mem: 57114
Epoch: [132]  [ 60/156]  eta: 0:01:12  lr: 0.005220  min_lr: 0.005220  loss: 3.2274 (3.2229)  weight_decay: 0.0500 (0.0500)  time: 0.7226  data: 0.0004  max mem: 57114
Epoch: [132]  [ 70/156]  eta: 0:01:04  lr: 0.005217  min_lr: 0.005217  loss: 3.3493 (3.2256)  weight_decay: 0.0500 (0.0500)  time: 0.7385  data: 0.0004  max mem: 57114
Epoch: [132]  [ 80/156]  eta: 0:00:56  lr: 0.005215  min_lr: 0.005215  loss: 3.1087 (3.2035)  weight_decay: 0.0500 (0.0500)  time: 0.7255  data: 0.0004  max mem: 57114
Epoch: [132]  [ 90/156]  eta: 0:00:48  lr: 0.005212  min_lr: 0.005212  loss: 3.2247 (3.2181)  weight_decay: 0.0500 (0.0500)  time: 0.7050  data: 0.0004  max mem: 57114
Epoch: [132]  [100/156]  eta: 0:00:41  lr: 0.005209  min_lr: 0.005209  loss: 3.2340 (3.2143)  weight_decay: 0.0500 (0.0500)  time: 0.7030  data: 0.0004  max mem: 57114
Epoch: [132]  [110/156]  eta: 0:00:33  lr: 0.005206  min_lr: 0.005206  loss: 3.2340 (3.2197)  weight_decay: 0.0500 (0.0500)  time: 0.7044  data: 0.0004  max mem: 57114
Epoch: [132]  [120/156]  eta: 0:00:26  lr: 0.005204  min_lr: 0.005204  loss: 3.3699 (3.2154)  weight_decay: 0.0500 (0.0500)  time: 0.7012  data: 0.0004  max mem: 57114
Epoch: [132]  [130/156]  eta: 0:00:18  lr: 0.005201  min_lr: 0.005201  loss: 3.2870 (3.2125)  weight_decay: 0.0500 (0.0500)  time: 0.7040  data: 0.0009  max mem: 57114
Epoch: [132]  [140/156]  eta: 0:00:11  lr: 0.005198  min_lr: 0.005198  loss: 3.2779 (3.2123)  weight_decay: 0.0500 (0.0500)  time: 0.6957  data: 0.0007  max mem: 57114
Epoch: [132]  [150/156]  eta: 0:00:04  lr: 0.005195  min_lr: 0.005195  loss: 3.1535 (3.2006)  weight_decay: 0.0500 (0.0500)  time: 0.6811  data: 0.0001  max mem: 57114
Epoch: [132]  [155/156]  eta: 0:00:00  lr: 0.005194  min_lr: 0.005194  loss: 3.1535 (3.2043)  weight_decay: 0.0500 (0.0500)  time: 0.6795  data: 0.0001  max mem: 57114
Epoch: [132] Total time: 0:01:52 (0.7232 s / it)
Averaged stats: lr: 0.005194  min_lr: 0.005194  loss: 3.1535 (3.1843)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9324 (0.9324)  acc1: 87.5000 (87.5000)  acc5: 96.3542 (96.3542)  time: 2.0951  data: 1.8394  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1560 (1.0929)  acc1: 78.6458 (78.2609)  acc5: 96.3542 (94.7570)  time: 0.5697  data: 0.3680  max mem: 57114
Test: Total time: 0:00:02 (0.5923 s / it)
* Acc@1 77.815 Acc@5 94.088 loss 1.167
Accuracy of the model on the 50000 test images: 77.8%
Max accuracy: 78.38%
Test:  [0/5]  eta: 0:00:11  loss: 1.7272 (1.7272)  acc1: 61.4583 (61.4583)  acc5: 81.7708 (81.7708)  time: 2.2680  data: 2.0246  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 2.1580 (2.1143)  acc1: 52.6042 (51.1509)  acc5: 78.6458 (74.6803)  time: 0.6043  data: 0.4050  max mem: 57114
Test: Total time: 0:00:03 (0.6151 s / it)
* Acc@1 47.398 Acc@5 71.757 loss 2.433
Accuracy of the model EMA on 50000 test images: 47.4%
Max EMA accuracy: 47.40%
Epoch: [133]  [  0/156]  eta: 0:06:35  lr: 0.005194  min_lr: 0.005194  loss: 3.4347 (3.4347)  weight_decay: 0.0500 (0.0500)  time: 2.5343  data: 1.8855  max mem: 57114
Epoch: [133]  [ 10/156]  eta: 0:02:07  lr: 0.005191  min_lr: 0.005191  loss: 3.3050 (3.0611)  weight_decay: 0.0500 (0.0500)  time: 0.8755  data: 0.1717  max mem: 57114
Epoch: [133]  [ 20/156]  eta: 0:01:49  lr: 0.005188  min_lr: 0.005188  loss: 3.3050 (3.1513)  weight_decay: 0.0500 (0.0500)  time: 0.7180  data: 0.0003  max mem: 57114
Epoch: [133]  [ 30/156]  eta: 0:01:36  lr: 0.005185  min_lr: 0.005185  loss: 3.3669 (3.1413)  weight_decay: 0.0500 (0.0500)  time: 0.7102  data: 0.0003  max mem: 57114
Epoch: [133]  [ 40/156]  eta: 0:01:28  lr: 0.005183  min_lr: 0.005183  loss: 3.4067 (3.2154)  weight_decay: 0.0500 (0.0500)  time: 0.7264  data: 0.0004  max mem: 57114
Epoch: [133]  [ 50/156]  eta: 0:01:20  lr: 0.005180  min_lr: 0.005180  loss: 3.4067 (3.1493)  weight_decay: 0.0500 (0.0500)  time: 0.7517  data: 0.0004  max mem: 57114
Epoch: [133]  [ 60/156]  eta: 0:01:12  lr: 0.005177  min_lr: 0.005177  loss: 3.1446 (3.1364)  weight_decay: 0.0500 (0.0500)  time: 0.7313  data: 0.0004  max mem: 57114
Epoch: [133]  [ 70/156]  eta: 0:01:04  lr: 0.005174  min_lr: 0.005174  loss: 3.3702 (3.1616)  weight_decay: 0.0500 (0.0500)  time: 0.7150  data: 0.0004  max mem: 57114
Epoch: [133]  [ 80/156]  eta: 0:00:56  lr: 0.005172  min_lr: 0.005172  loss: 3.3702 (3.1627)  weight_decay: 0.0500 (0.0500)  time: 0.7087  data: 0.0004  max mem: 57114
Epoch: [133]  [ 90/156]  eta: 0:00:48  lr: 0.005169  min_lr: 0.005169  loss: 3.3586 (3.1796)  weight_decay: 0.0500 (0.0500)  time: 0.7044  data: 0.0004  max mem: 57114
Epoch: [133]  [100/156]  eta: 0:00:41  lr: 0.005166  min_lr: 0.005166  loss: 3.3371 (3.1813)  weight_decay: 0.0500 (0.0500)  time: 0.7072  data: 0.0004  max mem: 57114
Epoch: [133]  [110/156]  eta: 0:00:33  lr: 0.005163  min_lr: 0.005163  loss: 3.1380 (3.1697)  weight_decay: 0.0500 (0.0500)  time: 0.7082  data: 0.0004  max mem: 57114
Epoch: [133]  [120/156]  eta: 0:00:26  lr: 0.005161  min_lr: 0.005161  loss: 2.9533 (3.1463)  weight_decay: 0.0500 (0.0500)  time: 0.7057  data: 0.0004  max mem: 57114
Epoch: [133]  [130/156]  eta: 0:00:18  lr: 0.005158  min_lr: 0.005158  loss: 3.2009 (3.1602)  weight_decay: 0.0500 (0.0500)  time: 0.7059  data: 0.0009  max mem: 57114
Epoch: [133]  [140/156]  eta: 0:00:11  lr: 0.005155  min_lr: 0.005155  loss: 3.2728 (3.1429)  weight_decay: 0.0500 (0.0500)  time: 0.6962  data: 0.0007  max mem: 57114
Epoch: [133]  [150/156]  eta: 0:00:04  lr: 0.005152  min_lr: 0.005152  loss: 3.2141 (3.1499)  weight_decay: 0.0500 (0.0500)  time: 0.6840  data: 0.0001  max mem: 57114
Epoch: [133]  [155/156]  eta: 0:00:00  lr: 0.005151  min_lr: 0.005151  loss: 3.2445 (3.1508)  weight_decay: 0.0500 (0.0500)  time: 0.6839  data: 0.0001  max mem: 57114
Epoch: [133] Total time: 0:01:53 (0.7244 s / it)
Averaged stats: lr: 0.005151  min_lr: 0.005151  loss: 3.2445 (3.1725)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8149 (0.8149)  acc1: 85.9375 (85.9375)  acc5: 97.9167 (97.9167)  time: 2.0730  data: 1.8157  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9726 (0.9487)  acc1: 77.6042 (77.7494)  acc5: 95.8333 (94.8849)  time: 0.5655  data: 0.3632  max mem: 57114
Test: Total time: 0:00:02 (0.5870 s / it)
* Acc@1 78.395 Acc@5 94.523 loss 1.050
Accuracy of the model on the 50000 test images: 78.4%
Max accuracy: 78.39%
Test:  [0/5]  eta: 0:00:10  loss: 1.6493 (1.6493)  acc1: 63.5417 (63.5417)  acc5: 83.3333 (83.3333)  time: 2.0521  data: 1.8086  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 2.0742 (2.0252)  acc1: 54.1667 (53.0691)  acc5: 79.6875 (75.7033)  time: 0.5610  data: 0.3618  max mem: 57114
Test: Total time: 0:00:02 (0.5730 s / it)
* Acc@1 48.921 Acc@5 73.274 loss 2.338
Accuracy of the model EMA on 50000 test images: 48.9%
Max EMA accuracy: 48.92%
Epoch: [134]  [  0/156]  eta: 0:06:24  lr: 0.005151  min_lr: 0.005151  loss: 3.1865 (3.1865)  weight_decay: 0.0500 (0.0500)  time: 2.4627  data: 1.8104  max mem: 57114
Epoch: [134]  [ 10/156]  eta: 0:02:09  lr: 0.005148  min_lr: 0.005148  loss: 3.0441 (3.0695)  weight_decay: 0.0500 (0.0500)  time: 0.8875  data: 0.1648  max mem: 57114
Epoch: [134]  [ 20/156]  eta: 0:01:51  lr: 0.005145  min_lr: 0.005145  loss: 3.2615 (3.1731)  weight_decay: 0.0500 (0.0500)  time: 0.7416  data: 0.0003  max mem: 57114
Epoch: [134]  [ 30/156]  eta: 0:01:40  lr: 0.005142  min_lr: 0.005142  loss: 3.3083 (3.1196)  weight_decay: 0.0500 (0.0500)  time: 0.7492  data: 0.0003  max mem: 57114
Epoch: [134]  [ 40/156]  eta: 0:01:31  lr: 0.005140  min_lr: 0.005140  loss: 2.9147 (3.0601)  weight_decay: 0.0500 (0.0500)  time: 0.7500  data: 0.0003  max mem: 57114
Epoch: [134]  [ 50/156]  eta: 0:01:22  lr: 0.005137  min_lr: 0.005137  loss: 3.2010 (3.1193)  weight_decay: 0.0500 (0.0500)  time: 0.7428  data: 0.0004  max mem: 57114
Epoch: [134]  [ 60/156]  eta: 0:01:13  lr: 0.005134  min_lr: 0.005134  loss: 3.3568 (3.1076)  weight_decay: 0.0500 (0.0500)  time: 0.7268  data: 0.0004  max mem: 57114
Epoch: [134]  [ 70/156]  eta: 0:01:05  lr: 0.005131  min_lr: 0.005131  loss: 3.1972 (3.1041)  weight_decay: 0.0500 (0.0500)  time: 0.7226  data: 0.0004  max mem: 57114
Epoch: [134]  [ 80/156]  eta: 0:00:57  lr: 0.005129  min_lr: 0.005129  loss: 3.1893 (3.1111)  weight_decay: 0.0500 (0.0500)  time: 0.7188  data: 0.0004  max mem: 57114
Epoch: [134]  [ 90/156]  eta: 0:00:49  lr: 0.005126  min_lr: 0.005126  loss: 3.0992 (3.1167)  weight_decay: 0.0500 (0.0500)  time: 0.7126  data: 0.0004  max mem: 57114
Epoch: [134]  [100/156]  eta: 0:00:41  lr: 0.005123  min_lr: 0.005123  loss: 3.2361 (3.1335)  weight_decay: 0.0500 (0.0500)  time: 0.7098  data: 0.0004  max mem: 57114
Epoch: [134]  [110/156]  eta: 0:00:34  lr: 0.005120  min_lr: 0.005120  loss: 3.2361 (3.1238)  weight_decay: 0.0500 (0.0500)  time: 0.7094  data: 0.0003  max mem: 57114
Epoch: [134]  [120/156]  eta: 0:00:26  lr: 0.005118  min_lr: 0.005118  loss: 2.9414 (3.1058)  weight_decay: 0.0500 (0.0500)  time: 0.7104  data: 0.0004  max mem: 57114
Epoch: [134]  [130/156]  eta: 0:00:19  lr: 0.005115  min_lr: 0.005115  loss: 3.1897 (3.1040)  weight_decay: 0.0500 (0.0500)  time: 0.7051  data: 0.0009  max mem: 57114
Epoch: [134]  [140/156]  eta: 0:00:11  lr: 0.005112  min_lr: 0.005112  loss: 3.1191 (3.1093)  weight_decay: 0.0500 (0.0500)  time: 0.6888  data: 0.0007  max mem: 57114
Epoch: [134]  [150/156]  eta: 0:00:04  lr: 0.005109  min_lr: 0.005109  loss: 3.1191 (3.1106)  weight_decay: 0.0500 (0.0500)  time: 0.6817  data: 0.0001  max mem: 57114
Epoch: [134]  [155/156]  eta: 0:00:00  lr: 0.005108  min_lr: 0.005108  loss: 3.2053 (3.1129)  weight_decay: 0.0500 (0.0500)  time: 0.6849  data: 0.0001  max mem: 57114
Epoch: [134] Total time: 0:01:54 (0.7308 s / it)
Averaged stats: lr: 0.005108  min_lr: 0.005108  loss: 3.2053 (3.1683)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9446 (0.9446)  acc1: 86.9792 (86.9792)  acc5: 96.8750 (96.8750)  time: 2.0994  data: 1.8435  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0467 (1.0848)  acc1: 82.8125 (78.5166)  acc5: 96.8750 (94.8849)  time: 0.5707  data: 0.3688  max mem: 57114
Test: Total time: 0:00:02 (0.5971 s / it)
* Acc@1 77.490 Acc@5 94.034 loss 1.196
Accuracy of the model on the 50000 test images: 77.5%
Max accuracy: 78.39%
Test:  [0/5]  eta: 0:00:11  loss: 1.5758 (1.5758)  acc1: 64.0625 (64.0625)  acc5: 84.3750 (84.3750)  time: 2.2990  data: 2.0555  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.9957 (1.9416)  acc1: 54.6875 (54.0921)  acc5: 80.7292 (77.1100)  time: 0.6104  data: 0.4112  max mem: 57114
Test: Total time: 0:00:03 (0.6227 s / it)
* Acc@1 50.508 Acc@5 74.756 loss 2.249
Accuracy of the model EMA on 50000 test images: 50.5%
Max EMA accuracy: 50.51%
Epoch: [135]  [  0/156]  eta: 0:06:58  lr: 0.005108  min_lr: 0.005108  loss: 3.3182 (3.3182)  weight_decay: 0.0500 (0.0500)  time: 2.6812  data: 2.0323  max mem: 57114
Epoch: [135]  [ 10/156]  eta: 0:02:09  lr: 0.005105  min_lr: 0.005105  loss: 3.2166 (3.0713)  weight_decay: 0.0500 (0.0500)  time: 0.8874  data: 0.1850  max mem: 57114
Epoch: [135]  [ 20/156]  eta: 0:01:49  lr: 0.005102  min_lr: 0.005102  loss: 2.9882 (2.9870)  weight_decay: 0.0500 (0.0500)  time: 0.7138  data: 0.0003  max mem: 57114
Epoch: [135]  [ 30/156]  eta: 0:01:38  lr: 0.005099  min_lr: 0.005099  loss: 3.2408 (3.0788)  weight_decay: 0.0500 (0.0500)  time: 0.7193  data: 0.0004  max mem: 57114
Epoch: [135]  [ 40/156]  eta: 0:01:29  lr: 0.005097  min_lr: 0.005097  loss: 3.2507 (3.0663)  weight_decay: 0.0500 (0.0500)  time: 0.7268  data: 0.0003  max mem: 57114
Epoch: [135]  [ 50/156]  eta: 0:01:20  lr: 0.005094  min_lr: 0.005094  loss: 3.0938 (3.0462)  weight_decay: 0.0500 (0.0500)  time: 0.7270  data: 0.0004  max mem: 57114
Epoch: [135]  [ 60/156]  eta: 0:01:12  lr: 0.005091  min_lr: 0.005091  loss: 3.0938 (3.0538)  weight_decay: 0.0500 (0.0500)  time: 0.7156  data: 0.0004  max mem: 57114
Epoch: [135]  [ 70/156]  eta: 0:01:03  lr: 0.005088  min_lr: 0.005088  loss: 3.0834 (3.0575)  weight_decay: 0.0500 (0.0500)  time: 0.7073  data: 0.0004  max mem: 57114
Epoch: [135]  [ 80/156]  eta: 0:00:56  lr: 0.005086  min_lr: 0.005086  loss: 3.0233 (3.0537)  weight_decay: 0.0500 (0.0500)  time: 0.7013  data: 0.0003  max mem: 57114
Epoch: [135]  [ 90/156]  eta: 0:00:48  lr: 0.005083  min_lr: 0.005083  loss: 3.3446 (3.0953)  weight_decay: 0.0500 (0.0500)  time: 0.7058  data: 0.0004  max mem: 57114
Epoch: [135]  [100/156]  eta: 0:00:41  lr: 0.005080  min_lr: 0.005080  loss: 3.3446 (3.1041)  weight_decay: 0.0500 (0.0500)  time: 0.7104  data: 0.0004  max mem: 57114
Epoch: [135]  [110/156]  eta: 0:00:33  lr: 0.005077  min_lr: 0.005077  loss: 3.1880 (3.0984)  weight_decay: 0.0500 (0.0500)  time: 0.7032  data: 0.0004  max mem: 57114
Epoch: [135]  [120/156]  eta: 0:00:26  lr: 0.005074  min_lr: 0.005074  loss: 3.1001 (3.0935)  weight_decay: 0.0500 (0.0500)  time: 0.7030  data: 0.0004  max mem: 57114
Epoch: [135]  [130/156]  eta: 0:00:18  lr: 0.005072  min_lr: 0.005072  loss: 3.2592 (3.1087)  weight_decay: 0.0500 (0.0500)  time: 0.6996  data: 0.0008  max mem: 57114
Epoch: [135]  [140/156]  eta: 0:00:11  lr: 0.005069  min_lr: 0.005069  loss: 3.4050 (3.1248)  weight_decay: 0.0500 (0.0500)  time: 0.6882  data: 0.0007  max mem: 57114
Epoch: [135]  [150/156]  eta: 0:00:04  lr: 0.005066  min_lr: 0.005066  loss: 3.5652 (3.1456)  weight_decay: 0.0500 (0.0500)  time: 0.6837  data: 0.0001  max mem: 57114
Epoch: [135]  [155/156]  eta: 0:00:00  lr: 0.005065  min_lr: 0.005065  loss: 3.4144 (3.1484)  weight_decay: 0.0500 (0.0500)  time: 0.6842  data: 0.0001  max mem: 57114
Epoch: [135] Total time: 0:01:52 (0.7211 s / it)
Averaged stats: lr: 0.005065  min_lr: 0.005065  loss: 3.4144 (3.1724)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0967 (1.0967)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.0579  data: 1.8018  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2342 (1.1735)  acc1: 80.2083 (78.9003)  acc5: 96.3542 (94.7570)  time: 0.5623  data: 0.3604  max mem: 57114
Test: Total time: 0:00:02 (0.5807 s / it)
* Acc@1 77.857 Acc@5 94.160 loss 1.289
Accuracy of the model on the 50000 test images: 77.9%
Max accuracy: 78.39%
Test:  [0/5]  eta: 0:00:11  loss: 1.5087 (1.5087)  acc1: 65.1042 (65.1042)  acc5: 84.8958 (84.8958)  time: 2.2647  data: 2.0213  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.9223 (1.8654)  acc1: 56.2500 (55.1151)  acc5: 82.2917 (78.3888)  time: 0.6036  data: 0.4044  max mem: 57114
Test: Total time: 0:00:03 (0.6164 s / it)
* Acc@1 51.964 Acc@5 76.069 loss 2.167
Accuracy of the model EMA on 50000 test images: 52.0%
Max EMA accuracy: 51.96%
Epoch: [136]  [  0/156]  eta: 0:07:09  lr: 0.005065  min_lr: 0.005065  loss: 3.2038 (3.2038)  weight_decay: 0.0500 (0.0500)  time: 2.7547  data: 2.1046  max mem: 57114
Epoch: [136]  [ 10/156]  eta: 0:02:11  lr: 0.005062  min_lr: 0.005062  loss: 3.2900 (3.1577)  weight_decay: 0.0500 (0.0500)  time: 0.8987  data: 0.1917  max mem: 57114
Epoch: [136]  [ 20/156]  eta: 0:01:52  lr: 0.005059  min_lr: 0.005059  loss: 3.1134 (3.1357)  weight_decay: 0.0500 (0.0500)  time: 0.7301  data: 0.0004  max mem: 57114
Epoch: [136]  [ 30/156]  eta: 0:01:40  lr: 0.005056  min_lr: 0.005056  loss: 3.1834 (3.1805)  weight_decay: 0.0500 (0.0500)  time: 0.7367  data: 0.0004  max mem: 57114
Epoch: [136]  [ 40/156]  eta: 0:01:29  lr: 0.005053  min_lr: 0.005053  loss: 3.2660 (3.1754)  weight_decay: 0.0500 (0.0500)  time: 0.7220  data: 0.0004  max mem: 57114
Epoch: [136]  [ 50/156]  eta: 0:01:20  lr: 0.005051  min_lr: 0.005051  loss: 3.1685 (3.1730)  weight_decay: 0.0500 (0.0500)  time: 0.7091  data: 0.0004  max mem: 57114
Epoch: [136]  [ 60/156]  eta: 0:01:12  lr: 0.005048  min_lr: 0.005048  loss: 3.2181 (3.1785)  weight_decay: 0.0500 (0.0500)  time: 0.7218  data: 0.0004  max mem: 57114
Epoch: [136]  [ 70/156]  eta: 0:01:04  lr: 0.005045  min_lr: 0.005045  loss: 3.2752 (3.2067)  weight_decay: 0.0500 (0.0500)  time: 0.7275  data: 0.0004  max mem: 57114
Epoch: [136]  [ 80/156]  eta: 0:00:56  lr: 0.005042  min_lr: 0.005042  loss: 3.2589 (3.1725)  weight_decay: 0.0500 (0.0500)  time: 0.7145  data: 0.0004  max mem: 57114
Epoch: [136]  [ 90/156]  eta: 0:00:49  lr: 0.005040  min_lr: 0.005040  loss: 2.9832 (3.1449)  weight_decay: 0.0500 (0.0500)  time: 0.7121  data: 0.0004  max mem: 57114
Epoch: [136]  [100/156]  eta: 0:00:41  lr: 0.005037  min_lr: 0.005037  loss: 2.7559 (3.0960)  weight_decay: 0.0500 (0.0500)  time: 0.7062  data: 0.0004  max mem: 57114
Epoch: [136]  [110/156]  eta: 0:00:33  lr: 0.005034  min_lr: 0.005034  loss: 2.8356 (3.1075)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0004  max mem: 57114
Epoch: [136]  [120/156]  eta: 0:00:26  lr: 0.005031  min_lr: 0.005031  loss: 3.2979 (3.1146)  weight_decay: 0.0500 (0.0500)  time: 0.7043  data: 0.0004  max mem: 57114
Epoch: [136]  [130/156]  eta: 0:00:18  lr: 0.005028  min_lr: 0.005028  loss: 3.2806 (3.1241)  weight_decay: 0.0500 (0.0500)  time: 0.6956  data: 0.0009  max mem: 57114
Epoch: [136]  [140/156]  eta: 0:00:11  lr: 0.005026  min_lr: 0.005026  loss: 3.2535 (3.1378)  weight_decay: 0.0500 (0.0500)  time: 0.6938  data: 0.0007  max mem: 57114
Epoch: [136]  [150/156]  eta: 0:00:04  lr: 0.005023  min_lr: 0.005023  loss: 3.3308 (3.1525)  weight_decay: 0.0500 (0.0500)  time: 0.6898  data: 0.0001  max mem: 57114
Epoch: [136]  [155/156]  eta: 0:00:00  lr: 0.005021  min_lr: 0.005021  loss: 3.3948 (3.1553)  weight_decay: 0.0500 (0.0500)  time: 0.6854  data: 0.0001  max mem: 57114
Epoch: [136] Total time: 0:01:53 (0.7260 s / it)
Averaged stats: lr: 0.005021  min_lr: 0.005021  loss: 3.3948 (3.1667)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9760 (0.9760)  acc1: 85.4167 (85.4167)  acc5: 96.3542 (96.3542)  time: 2.0904  data: 1.8352  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0846 (1.1006)  acc1: 83.3333 (79.1560)  acc5: 96.3542 (94.1177)  time: 0.5687  data: 0.3671  max mem: 57114
Test: Total time: 0:00:02 (0.5914 s / it)
* Acc@1 77.396 Acc@5 93.934 loss 1.147
Accuracy of the model on the 50000 test images: 77.4%
Max accuracy: 78.39%
Test:  [0/5]  eta: 0:00:11  loss: 1.4490 (1.4490)  acc1: 66.1458 (66.1458)  acc5: 85.4167 (85.4167)  time: 2.2829  data: 2.0395  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.8539 (1.7962)  acc1: 58.3333 (56.7775)  acc5: 83.3333 (79.5396)  time: 0.6072  data: 0.4080  max mem: 57114
Test: Total time: 0:00:03 (0.6219 s / it)
* Acc@1 53.275 Acc@5 77.170 loss 2.093
Accuracy of the model EMA on 50000 test images: 53.3%
Max EMA accuracy: 53.27%
Epoch: [137]  [  0/156]  eta: 0:07:04  lr: 0.005021  min_lr: 0.005021  loss: 3.7286 (3.7286)  weight_decay: 0.0500 (0.0500)  time: 2.7194  data: 2.0609  max mem: 57114
Epoch: [137]  [ 10/156]  eta: 0:02:14  lr: 0.005018  min_lr: 0.005018  loss: 2.9422 (3.0029)  weight_decay: 0.0500 (0.0500)  time: 0.9226  data: 0.1876  max mem: 57114
Epoch: [137]  [ 20/156]  eta: 0:01:51  lr: 0.005016  min_lr: 0.005016  loss: 3.1090 (3.1318)  weight_decay: 0.0500 (0.0500)  time: 0.7279  data: 0.0003  max mem: 57114
Epoch: [137]  [ 30/156]  eta: 0:01:39  lr: 0.005013  min_lr: 0.005013  loss: 3.2235 (3.1746)  weight_decay: 0.0500 (0.0500)  time: 0.7162  data: 0.0003  max mem: 57114
Epoch: [137]  [ 40/156]  eta: 0:01:30  lr: 0.005010  min_lr: 0.005010  loss: 2.9876 (3.0975)  weight_decay: 0.0500 (0.0500)  time: 0.7369  data: 0.0004  max mem: 57114
Epoch: [137]  [ 50/156]  eta: 0:01:21  lr: 0.005007  min_lr: 0.005007  loss: 3.1264 (3.1148)  weight_decay: 0.0500 (0.0500)  time: 0.7391  data: 0.0004  max mem: 57114
Epoch: [137]  [ 60/156]  eta: 0:01:13  lr: 0.005004  min_lr: 0.005004  loss: 3.3410 (3.1101)  weight_decay: 0.0500 (0.0500)  time: 0.7331  data: 0.0004  max mem: 57114
Epoch: [137]  [ 70/156]  eta: 0:01:05  lr: 0.005002  min_lr: 0.005002  loss: 2.9608 (3.0936)  weight_decay: 0.0500 (0.0500)  time: 0.7211  data: 0.0004  max mem: 57114
Epoch: [137]  [ 80/156]  eta: 0:00:56  lr: 0.004999  min_lr: 0.004999  loss: 3.1010 (3.1183)  weight_decay: 0.0500 (0.0500)  time: 0.7028  data: 0.0004  max mem: 57114
Epoch: [137]  [ 90/156]  eta: 0:00:49  lr: 0.004996  min_lr: 0.004996  loss: 3.1936 (3.0943)  weight_decay: 0.0500 (0.0500)  time: 0.7080  data: 0.0004  max mem: 57114
Epoch: [137]  [100/156]  eta: 0:00:41  lr: 0.004993  min_lr: 0.004993  loss: 3.2825 (3.1074)  weight_decay: 0.0500 (0.0500)  time: 0.7200  data: 0.0004  max mem: 57114
Epoch: [137]  [110/156]  eta: 0:00:33  lr: 0.004991  min_lr: 0.004991  loss: 3.3362 (3.1155)  weight_decay: 0.0500 (0.0500)  time: 0.7094  data: 0.0004  max mem: 57114
Epoch: [137]  [120/156]  eta: 0:00:26  lr: 0.004988  min_lr: 0.004988  loss: 3.1131 (3.1084)  weight_decay: 0.0500 (0.0500)  time: 0.7006  data: 0.0004  max mem: 57114
Epoch: [137]  [130/156]  eta: 0:00:19  lr: 0.004985  min_lr: 0.004985  loss: 2.9640 (3.0987)  weight_decay: 0.0500 (0.0500)  time: 0.7021  data: 0.0008  max mem: 57114
Epoch: [137]  [140/156]  eta: 0:00:11  lr: 0.004982  min_lr: 0.004982  loss: 3.0072 (3.1002)  weight_decay: 0.0500 (0.0500)  time: 0.6825  data: 0.0007  max mem: 57114
Epoch: [137]  [150/156]  eta: 0:00:04  lr: 0.004979  min_lr: 0.004979  loss: 3.3923 (3.1097)  weight_decay: 0.0500 (0.0500)  time: 0.6792  data: 0.0001  max mem: 57114
Epoch: [137]  [155/156]  eta: 0:00:00  lr: 0.004978  min_lr: 0.004978  loss: 3.4201 (3.1192)  weight_decay: 0.0500 (0.0500)  time: 0.6841  data: 0.0001  max mem: 57114
Epoch: [137] Total time: 0:01:53 (0.7272 s / it)
Averaged stats: lr: 0.004978  min_lr: 0.004978  loss: 3.4201 (3.1603)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9355 (0.9355)  acc1: 86.4583 (86.4583)  acc5: 98.9583 (98.9583)  time: 2.0631  data: 1.8061  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1010 (1.0717)  acc1: 82.2917 (80.0512)  acc5: 96.3542 (95.6522)  time: 0.5633  data: 0.3613  max mem: 57114
Test: Total time: 0:00:02 (0.5833 s / it)
* Acc@1 77.713 Acc@5 94.222 loss 1.189
Accuracy of the model on the 50000 test images: 77.7%
Max accuracy: 78.39%
Test:  [0/5]  eta: 0:00:11  loss: 1.3943 (1.3943)  acc1: 68.2292 (68.2292)  acc5: 85.9375 (85.9375)  time: 2.2397  data: 1.9961  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.7893 (1.7314)  acc1: 58.8542 (58.0563)  acc5: 84.3750 (80.5627)  time: 0.5987  data: 0.3993  max mem: 57114
Test: Total time: 0:00:03 (0.6147 s / it)
* Acc@1 54.516 Acc@5 78.227 loss 2.023
Accuracy of the model EMA on 50000 test images: 54.5%
Max EMA accuracy: 54.52%
Epoch: [138]  [  0/156]  eta: 0:07:42  lr: 0.004978  min_lr: 0.004978  loss: 2.3290 (2.3290)  weight_decay: 0.0500 (0.0500)  time: 2.9620  data: 2.3059  max mem: 57114
Epoch: [138]  [ 10/156]  eta: 0:02:18  lr: 0.004975  min_lr: 0.004975  loss: 2.7826 (2.8718)  weight_decay: 0.0500 (0.0500)  time: 0.9456  data: 0.2099  max mem: 57114
Epoch: [138]  [ 20/156]  eta: 0:01:54  lr: 0.004972  min_lr: 0.004972  loss: 2.8015 (2.9164)  weight_decay: 0.0500 (0.0500)  time: 0.7340  data: 0.0003  max mem: 57114
Epoch: [138]  [ 30/156]  eta: 0:01:41  lr: 0.004969  min_lr: 0.004969  loss: 3.0182 (2.9241)  weight_decay: 0.0500 (0.0500)  time: 0.7293  data: 0.0003  max mem: 57114
Epoch: [138]  [ 40/156]  eta: 0:01:31  lr: 0.004967  min_lr: 0.004967  loss: 3.2356 (3.0282)  weight_decay: 0.0500 (0.0500)  time: 0.7387  data: 0.0004  max mem: 57114
Epoch: [138]  [ 50/156]  eta: 0:01:22  lr: 0.004964  min_lr: 0.004964  loss: 3.2831 (3.0616)  weight_decay: 0.0500 (0.0500)  time: 0.7261  data: 0.0004  max mem: 57114
Epoch: [138]  [ 60/156]  eta: 0:01:13  lr: 0.004961  min_lr: 0.004961  loss: 3.3583 (3.0875)  weight_decay: 0.0500 (0.0500)  time: 0.7153  data: 0.0004  max mem: 57114
Epoch: [138]  [ 70/156]  eta: 0:01:05  lr: 0.004958  min_lr: 0.004958  loss: 3.0434 (3.0392)  weight_decay: 0.0500 (0.0500)  time: 0.7113  data: 0.0004  max mem: 57114
Epoch: [138]  [ 80/156]  eta: 0:00:57  lr: 0.004955  min_lr: 0.004955  loss: 2.7563 (3.0290)  weight_decay: 0.0500 (0.0500)  time: 0.7108  data: 0.0004  max mem: 57114
Epoch: [138]  [ 90/156]  eta: 0:00:49  lr: 0.004953  min_lr: 0.004953  loss: 2.8366 (3.0109)  weight_decay: 0.0500 (0.0500)  time: 0.7230  data: 0.0004  max mem: 57114
Epoch: [138]  [100/156]  eta: 0:00:41  lr: 0.004950  min_lr: 0.004950  loss: 3.1501 (3.0295)  weight_decay: 0.0500 (0.0500)  time: 0.7151  data: 0.0004  max mem: 57114
Epoch: [138]  [110/156]  eta: 0:00:34  lr: 0.004947  min_lr: 0.004947  loss: 3.3999 (3.0574)  weight_decay: 0.0500 (0.0500)  time: 0.7076  data: 0.0003  max mem: 57114
Epoch: [138]  [120/156]  eta: 0:00:26  lr: 0.004944  min_lr: 0.004944  loss: 3.5187 (3.0723)  weight_decay: 0.0500 (0.0500)  time: 0.7109  data: 0.0003  max mem: 57114
Epoch: [138]  [130/156]  eta: 0:00:19  lr: 0.004941  min_lr: 0.004941  loss: 3.2502 (3.0713)  weight_decay: 0.0500 (0.0500)  time: 0.7008  data: 0.0008  max mem: 57114
Epoch: [138]  [140/156]  eta: 0:00:11  lr: 0.004939  min_lr: 0.004939  loss: 2.9372 (3.0735)  weight_decay: 0.0500 (0.0500)  time: 0.6890  data: 0.0007  max mem: 57114
Epoch: [138]  [150/156]  eta: 0:00:04  lr: 0.004936  min_lr: 0.004936  loss: 3.2187 (3.0869)  weight_decay: 0.0500 (0.0500)  time: 0.6870  data: 0.0001  max mem: 57114
Epoch: [138]  [155/156]  eta: 0:00:00  lr: 0.004934  min_lr: 0.004934  loss: 3.2187 (3.0875)  weight_decay: 0.0500 (0.0500)  time: 0.6888  data: 0.0001  max mem: 57114
Epoch: [138] Total time: 0:01:53 (0.7300 s / it)
Averaged stats: lr: 0.004934  min_lr: 0.004934  loss: 3.2187 (3.1590)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9046 (0.9046)  acc1: 82.8125 (82.8125)  acc5: 96.8750 (96.8750)  time: 2.1454  data: 1.8884  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9882 (0.9683)  acc1: 78.6458 (76.7263)  acc5: 96.8750 (95.0128)  time: 0.5798  data: 0.3778  max mem: 57114
Test: Total time: 0:00:03 (0.6038 s / it)
* Acc@1 78.543 Acc@5 94.507 loss 1.066
Accuracy of the model on the 50000 test images: 78.5%
Max accuracy: 78.54%
Test:  [0/5]  eta: 0:00:10  loss: 1.3428 (1.3428)  acc1: 68.2292 (68.2292)  acc5: 87.5000 (87.5000)  time: 2.0925  data: 1.8489  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.7270 (1.6709)  acc1: 60.4167 (58.8235)  acc5: 85.4167 (81.4578)  time: 0.5692  data: 0.3699  max mem: 57114
Test: Total time: 0:00:02 (0.5824 s / it)
* Acc@1 55.645 Acc@5 79.196 loss 1.958
Accuracy of the model EMA on 50000 test images: 55.6%
Max EMA accuracy: 55.64%
Epoch: [139]  [  0/156]  eta: 0:08:53  lr: 0.004934  min_lr: 0.004934  loss: 2.4697 (2.4697)  weight_decay: 0.0500 (0.0500)  time: 3.4174  data: 2.7700  max mem: 57114
Epoch: [139]  [ 10/156]  eta: 0:02:24  lr: 0.004931  min_lr: 0.004931  loss: 3.3311 (3.2610)  weight_decay: 0.0500 (0.0500)  time: 0.9865  data: 0.2521  max mem: 57114
Epoch: [139]  [ 20/156]  eta: 0:01:59  lr: 0.004929  min_lr: 0.004929  loss: 3.2867 (3.1863)  weight_decay: 0.0500 (0.0500)  time: 0.7496  data: 0.0003  max mem: 57114
Epoch: [139]  [ 30/156]  eta: 0:01:44  lr: 0.004926  min_lr: 0.004926  loss: 3.2867 (3.2092)  weight_decay: 0.0500 (0.0500)  time: 0.7394  data: 0.0003  max mem: 57114
Epoch: [139]  [ 40/156]  eta: 0:01:33  lr: 0.004923  min_lr: 0.004923  loss: 3.2994 (3.2080)  weight_decay: 0.0500 (0.0500)  time: 0.7346  data: 0.0004  max mem: 57114
Epoch: [139]  [ 50/156]  eta: 0:01:24  lr: 0.004920  min_lr: 0.004920  loss: 3.1002 (3.1920)  weight_decay: 0.0500 (0.0500)  time: 0.7454  data: 0.0004  max mem: 57114
Epoch: [139]  [ 60/156]  eta: 0:01:15  lr: 0.004917  min_lr: 0.004917  loss: 3.0099 (3.1572)  weight_decay: 0.0500 (0.0500)  time: 0.7287  data: 0.0004  max mem: 57114
Epoch: [139]  [ 70/156]  eta: 0:01:06  lr: 0.004915  min_lr: 0.004915  loss: 3.0867 (3.1491)  weight_decay: 0.0500 (0.0500)  time: 0.7254  data: 0.0004  max mem: 57114
Epoch: [139]  [ 80/156]  eta: 0:00:58  lr: 0.004912  min_lr: 0.004912  loss: 3.3375 (3.1873)  weight_decay: 0.0500 (0.0500)  time: 0.7381  data: 0.0004  max mem: 57114
Epoch: [139]  [ 90/156]  eta: 0:00:50  lr: 0.004909  min_lr: 0.004909  loss: 3.3375 (3.1716)  weight_decay: 0.0500 (0.0500)  time: 0.7255  data: 0.0004  max mem: 57114
Epoch: [139]  [100/156]  eta: 0:00:42  lr: 0.004906  min_lr: 0.004906  loss: 2.9355 (3.1582)  weight_decay: 0.0500 (0.0500)  time: 0.7096  data: 0.0003  max mem: 57114
Epoch: [139]  [110/156]  eta: 0:00:34  lr: 0.004903  min_lr: 0.004903  loss: 2.9355 (3.1399)  weight_decay: 0.0500 (0.0500)  time: 0.7082  data: 0.0003  max mem: 57114
Epoch: [139]  [120/156]  eta: 0:00:27  lr: 0.004901  min_lr: 0.004901  loss: 3.1300 (3.1413)  weight_decay: 0.0500 (0.0500)  time: 0.7084  data: 0.0003  max mem: 57114
Epoch: [139]  [130/156]  eta: 0:00:19  lr: 0.004898  min_lr: 0.004898  loss: 3.3269 (3.1539)  weight_decay: 0.0500 (0.0500)  time: 0.6989  data: 0.0008  max mem: 57114
Epoch: [139]  [140/156]  eta: 0:00:11  lr: 0.004895  min_lr: 0.004895  loss: 3.3807 (3.1585)  weight_decay: 0.0500 (0.0500)  time: 0.6866  data: 0.0007  max mem: 57114
Epoch: [139]  [150/156]  eta: 0:00:04  lr: 0.004892  min_lr: 0.004892  loss: 3.4227 (3.1704)  weight_decay: 0.0500 (0.0500)  time: 0.6863  data: 0.0001  max mem: 57114
Epoch: [139]  [155/156]  eta: 0:00:00  lr: 0.004891  min_lr: 0.004891  loss: 3.1934 (3.1622)  weight_decay: 0.0500 (0.0500)  time: 0.6901  data: 0.0001  max mem: 57114
Epoch: [139] Total time: 0:01:55 (0.7389 s / it)
Averaged stats: lr: 0.004891  min_lr: 0.004891  loss: 3.1934 (3.1544)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8747 (0.8747)  acc1: 85.9375 (85.9375)  acc5: 98.4375 (98.4375)  time: 2.0106  data: 1.7553  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0756 (1.0140)  acc1: 80.2083 (79.1560)  acc5: 96.3542 (95.2685)  time: 0.5529  data: 0.3511  max mem: 57114
Test: Total time: 0:00:02 (0.5749 s / it)
* Acc@1 78.666 Acc@5 94.639 loss 1.090
Accuracy of the model on the 50000 test images: 78.7%
Max accuracy: 78.67%
Test:  [0/5]  eta: 0:00:09  loss: 1.2929 (1.2929)  acc1: 68.2292 (68.2292)  acc5: 88.5417 (88.5417)  time: 1.9833  data: 1.7398  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.6681 (1.6135)  acc1: 61.4583 (59.7187)  acc5: 86.9792 (82.4808)  time: 0.5473  data: 0.3480  max mem: 57114
Test: Total time: 0:00:02 (0.5585 s / it)
* Acc@1 56.728 Acc@5 80.141 loss 1.896
Accuracy of the model EMA on 50000 test images: 56.7%
Max EMA accuracy: 56.73%
Epoch: [140]  [  0/156]  eta: 0:07:34  lr: 0.004890  min_lr: 0.004890  loss: 3.7223 (3.7223)  weight_decay: 0.0500 (0.0500)  time: 2.9154  data: 2.2645  max mem: 57114
Epoch: [140]  [ 10/156]  eta: 0:02:15  lr: 0.004888  min_lr: 0.004888  loss: 3.3313 (3.1502)  weight_decay: 0.0500 (0.0500)  time: 0.9306  data: 0.2062  max mem: 57114
Epoch: [140]  [ 20/156]  eta: 0:01:54  lr: 0.004885  min_lr: 0.004885  loss: 2.9456 (3.1035)  weight_decay: 0.0500 (0.0500)  time: 0.7390  data: 0.0003  max mem: 57114
Epoch: [140]  [ 30/156]  eta: 0:01:40  lr: 0.004882  min_lr: 0.004882  loss: 3.1327 (3.1762)  weight_decay: 0.0500 (0.0500)  time: 0.7206  data: 0.0003  max mem: 57114
Epoch: [140]  [ 40/156]  eta: 0:01:31  lr: 0.004879  min_lr: 0.004879  loss: 3.3149 (3.1830)  weight_decay: 0.0500 (0.0500)  time: 0.7272  data: 0.0004  max mem: 57114
Epoch: [140]  [ 50/156]  eta: 0:01:22  lr: 0.004876  min_lr: 0.004876  loss: 3.1823 (3.1677)  weight_decay: 0.0500 (0.0500)  time: 0.7608  data: 0.0004  max mem: 57114
Epoch: [140]  [ 60/156]  eta: 0:01:14  lr: 0.004874  min_lr: 0.004874  loss: 3.3131 (3.1898)  weight_decay: 0.0500 (0.0500)  time: 0.7470  data: 0.0004  max mem: 57114
Epoch: [140]  [ 70/156]  eta: 0:01:05  lr: 0.004871  min_lr: 0.004871  loss: 3.1252 (3.1370)  weight_decay: 0.0500 (0.0500)  time: 0.7212  data: 0.0004  max mem: 57114
Epoch: [140]  [ 80/156]  eta: 0:00:57  lr: 0.004868  min_lr: 0.004868  loss: 3.0799 (3.1572)  weight_decay: 0.0500 (0.0500)  time: 0.7203  data: 0.0004  max mem: 57114
Epoch: [140]  [ 90/156]  eta: 0:00:49  lr: 0.004865  min_lr: 0.004865  loss: 3.4558 (3.1785)  weight_decay: 0.0500 (0.0500)  time: 0.7169  data: 0.0004  max mem: 57114
Epoch: [140]  [100/156]  eta: 0:00:41  lr: 0.004862  min_lr: 0.004862  loss: 3.3252 (3.1732)  weight_decay: 0.0500 (0.0500)  time: 0.7057  data: 0.0004  max mem: 57114
Epoch: [140]  [110/156]  eta: 0:00:34  lr: 0.004860  min_lr: 0.004860  loss: 3.2178 (3.1611)  weight_decay: 0.0500 (0.0500)  time: 0.7116  data: 0.0004  max mem: 57114
Epoch: [140]  [120/156]  eta: 0:00:26  lr: 0.004857  min_lr: 0.004857  loss: 3.1725 (3.1553)  weight_decay: 0.0500 (0.0500)  time: 0.7144  data: 0.0003  max mem: 57114
Epoch: [140]  [130/156]  eta: 0:00:19  lr: 0.004854  min_lr: 0.004854  loss: 3.0720 (3.1461)  weight_decay: 0.0500 (0.0500)  time: 0.7058  data: 0.0004  max mem: 57114
Epoch: [140]  [140/156]  eta: 0:00:11  lr: 0.004851  min_lr: 0.004851  loss: 3.0945 (3.1602)  weight_decay: 0.0500 (0.0500)  time: 0.6922  data: 0.0003  max mem: 57114
Epoch: [140]  [150/156]  eta: 0:00:04  lr: 0.004848  min_lr: 0.004848  loss: 3.4423 (3.1728)  weight_decay: 0.0500 (0.0500)  time: 0.6802  data: 0.0001  max mem: 57114
Epoch: [140]  [155/156]  eta: 0:00:00  lr: 0.004847  min_lr: 0.004847  loss: 3.3788 (3.1722)  weight_decay: 0.0500 (0.0500)  time: 0.6814  data: 0.0001  max mem: 57114
Epoch: [140] Total time: 0:01:54 (0.7329 s / it)
Averaged stats: lr: 0.004847  min_lr: 0.004847  loss: 3.3788 (3.1537)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8120 (0.8120)  acc1: 88.5417 (88.5417)  acc5: 97.9167 (97.9167)  time: 2.0761  data: 1.8205  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0669 (1.0525)  acc1: 79.6875 (79.5396)  acc5: 93.2292 (94.7570)  time: 0.5659  data: 0.3642  max mem: 57114
Test: Total time: 0:00:02 (0.5906 s / it)
* Acc@1 78.461 Acc@5 94.549 loss 1.129
Accuracy of the model on the 50000 test images: 78.5%
Max accuracy: 78.67%
Test:  [0/5]  eta: 0:00:11  loss: 1.2477 (1.2477)  acc1: 69.7917 (69.7917)  acc5: 89.5833 (89.5833)  time: 2.2868  data: 2.0434  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.6131 (1.5599)  acc1: 61.9792 (60.9974)  acc5: 87.5000 (83.2481)  time: 0.6080  data: 0.4088  max mem: 57114
Test: Total time: 0:00:03 (0.6185 s / it)
* Acc@1 57.737 Acc@5 80.978 loss 1.837
Accuracy of the model EMA on 50000 test images: 57.7%
Max EMA accuracy: 57.74%
Epoch: [141]  [  0/156]  eta: 0:07:44  lr: 0.004847  min_lr: 0.004847  loss: 3.3103 (3.3103)  weight_decay: 0.0500 (0.0500)  time: 2.9744  data: 2.3176  max mem: 57114
Epoch: [141]  [ 10/156]  eta: 0:02:17  lr: 0.004844  min_lr: 0.004844  loss: 3.4335 (3.3233)  weight_decay: 0.0500 (0.0500)  time: 0.9423  data: 0.2110  max mem: 57114
Epoch: [141]  [ 20/156]  eta: 0:01:54  lr: 0.004841  min_lr: 0.004841  loss: 3.3538 (3.1937)  weight_decay: 0.0500 (0.0500)  time: 0.7319  data: 0.0003  max mem: 57114
Epoch: [141]  [ 30/156]  eta: 0:01:41  lr: 0.004838  min_lr: 0.004838  loss: 2.9343 (3.1084)  weight_decay: 0.0500 (0.0500)  time: 0.7282  data: 0.0003  max mem: 57114
Epoch: [141]  [ 40/156]  eta: 0:01:31  lr: 0.004835  min_lr: 0.004835  loss: 2.8492 (3.0489)  weight_decay: 0.0500 (0.0500)  time: 0.7289  data: 0.0004  max mem: 57114
Epoch: [141]  [ 50/156]  eta: 0:01:22  lr: 0.004833  min_lr: 0.004833  loss: 3.2254 (3.0771)  weight_decay: 0.0500 (0.0500)  time: 0.7306  data: 0.0004  max mem: 57114
Epoch: [141]  [ 60/156]  eta: 0:01:13  lr: 0.004830  min_lr: 0.004830  loss: 3.2284 (3.0878)  weight_decay: 0.0500 (0.0500)  time: 0.7204  data: 0.0004  max mem: 57114
Epoch: [141]  [ 70/156]  eta: 0:01:05  lr: 0.004827  min_lr: 0.004827  loss: 3.2607 (3.1167)  weight_decay: 0.0500 (0.0500)  time: 0.7121  data: 0.0004  max mem: 57114
Epoch: [141]  [ 80/156]  eta: 0:00:57  lr: 0.004824  min_lr: 0.004824  loss: 3.4117 (3.1176)  weight_decay: 0.0500 (0.0500)  time: 0.7127  data: 0.0004  max mem: 57114
Epoch: [141]  [ 90/156]  eta: 0:00:49  lr: 0.004821  min_lr: 0.004821  loss: 3.3906 (3.1194)  weight_decay: 0.0500 (0.0500)  time: 0.7122  data: 0.0004  max mem: 57114
Epoch: [141]  [100/156]  eta: 0:00:41  lr: 0.004819  min_lr: 0.004819  loss: 3.0630 (3.1161)  weight_decay: 0.0500 (0.0500)  time: 0.7173  data: 0.0004  max mem: 57114
Epoch: [141]  [110/156]  eta: 0:00:34  lr: 0.004816  min_lr: 0.004816  loss: 3.2115 (3.1344)  weight_decay: 0.0500 (0.0500)  time: 0.7128  data: 0.0004  max mem: 57114
Epoch: [141]  [120/156]  eta: 0:00:26  lr: 0.004813  min_lr: 0.004813  loss: 3.1591 (3.1081)  weight_decay: 0.0500 (0.0500)  time: 0.6960  data: 0.0004  max mem: 57114
Epoch: [141]  [130/156]  eta: 0:00:19  lr: 0.004810  min_lr: 0.004810  loss: 2.9633 (3.1118)  weight_decay: 0.0500 (0.0500)  time: 0.6925  data: 0.0008  max mem: 57114
Epoch: [141]  [140/156]  eta: 0:00:11  lr: 0.004807  min_lr: 0.004807  loss: 3.2846 (3.1191)  weight_decay: 0.0500 (0.0500)  time: 0.6971  data: 0.0007  max mem: 57114
Epoch: [141]  [150/156]  eta: 0:00:04  lr: 0.004804  min_lr: 0.004804  loss: 3.2084 (3.1124)  weight_decay: 0.0500 (0.0500)  time: 0.6891  data: 0.0001  max mem: 57114
Epoch: [141]  [155/156]  eta: 0:00:00  lr: 0.004803  min_lr: 0.004803  loss: 3.2604 (3.1227)  weight_decay: 0.0500 (0.0500)  time: 0.6912  data: 0.0001  max mem: 57114
Epoch: [141] Total time: 0:01:53 (0.7288 s / it)
Averaged stats: lr: 0.004803  min_lr: 0.004803  loss: 3.2604 (3.1480)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9489 (0.9489)  acc1: 84.3750 (84.3750)  acc5: 96.8750 (96.8750)  time: 2.1362  data: 1.8807  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0704 (1.0603)  acc1: 80.2083 (78.7724)  acc5: 96.8750 (94.7570)  time: 0.5779  data: 0.3762  max mem: 57114
Test: Total time: 0:00:03 (0.6020 s / it)
* Acc@1 78.077 Acc@5 94.060 loss 1.140
Accuracy of the model on the 50000 test images: 78.1%
Max accuracy: 78.67%
Test:  [0/5]  eta: 0:00:11  loss: 1.2054 (1.2054)  acc1: 69.7917 (69.7917)  acc5: 89.5833 (89.5833)  time: 2.2810  data: 2.0374  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.5613 (1.5106)  acc1: 63.5417 (61.8926)  acc5: 87.5000 (83.6317)  time: 0.6068  data: 0.4076  max mem: 57114
Test: Total time: 0:00:03 (0.6189 s / it)
* Acc@1 58.742 Acc@5 81.817 loss 1.783
Accuracy of the model EMA on 50000 test images: 58.7%
Max EMA accuracy: 58.74%
Epoch: [142]  [  0/156]  eta: 0:07:35  lr: 0.004803  min_lr: 0.004803  loss: 2.4096 (2.4096)  weight_decay: 0.0500 (0.0500)  time: 2.9199  data: 2.2644  max mem: 57114
Epoch: [142]  [ 10/156]  eta: 0:02:14  lr: 0.004800  min_lr: 0.004800  loss: 2.8853 (2.8987)  weight_decay: 0.0500 (0.0500)  time: 0.9231  data: 0.2061  max mem: 57114
Epoch: [142]  [ 20/156]  eta: 0:01:53  lr: 0.004797  min_lr: 0.004797  loss: 3.2543 (3.0376)  weight_decay: 0.0500 (0.0500)  time: 0.7290  data: 0.0003  max mem: 57114
Epoch: [142]  [ 30/156]  eta: 0:01:40  lr: 0.004794  min_lr: 0.004794  loss: 3.3267 (3.1245)  weight_decay: 0.0500 (0.0500)  time: 0.7287  data: 0.0003  max mem: 57114
Epoch: [142]  [ 40/156]  eta: 0:01:30  lr: 0.004791  min_lr: 0.004791  loss: 3.4527 (3.0967)  weight_decay: 0.0500 (0.0500)  time: 0.7237  data: 0.0004  max mem: 57114
Epoch: [142]  [ 50/156]  eta: 0:01:21  lr: 0.004789  min_lr: 0.004789  loss: 3.4360 (3.1047)  weight_decay: 0.0500 (0.0500)  time: 0.7218  data: 0.0004  max mem: 57114
Epoch: [142]  [ 60/156]  eta: 0:01:12  lr: 0.004786  min_lr: 0.004786  loss: 3.2130 (3.0980)  weight_decay: 0.0500 (0.0500)  time: 0.7170  data: 0.0004  max mem: 57114
Epoch: [142]  [ 70/156]  eta: 0:01:04  lr: 0.004783  min_lr: 0.004783  loss: 3.2481 (3.1143)  weight_decay: 0.0500 (0.0500)  time: 0.7181  data: 0.0004  max mem: 57114
Epoch: [142]  [ 80/156]  eta: 0:00:56  lr: 0.004780  min_lr: 0.004780  loss: 3.1901 (3.1053)  weight_decay: 0.0500 (0.0500)  time: 0.7169  data: 0.0004  max mem: 57114
Epoch: [142]  [ 90/156]  eta: 0:00:49  lr: 0.004777  min_lr: 0.004777  loss: 3.1394 (3.1074)  weight_decay: 0.0500 (0.0500)  time: 0.7094  data: 0.0004  max mem: 57114
Epoch: [142]  [100/156]  eta: 0:00:41  lr: 0.004775  min_lr: 0.004775  loss: 3.2408 (3.1300)  weight_decay: 0.0500 (0.0500)  time: 0.7076  data: 0.0004  max mem: 57114
Epoch: [142]  [110/156]  eta: 0:00:33  lr: 0.004772  min_lr: 0.004772  loss: 3.3858 (3.1594)  weight_decay: 0.0500 (0.0500)  time: 0.7073  data: 0.0004  max mem: 57114
Epoch: [142]  [120/156]  eta: 0:00:26  lr: 0.004769  min_lr: 0.004769  loss: 3.3773 (3.1713)  weight_decay: 0.0500 (0.0500)  time: 0.7006  data: 0.0004  max mem: 57114
Epoch: [142]  [130/156]  eta: 0:00:19  lr: 0.004766  min_lr: 0.004766  loss: 3.3773 (3.1907)  weight_decay: 0.0500 (0.0500)  time: 0.6955  data: 0.0008  max mem: 57114
Epoch: [142]  [140/156]  eta: 0:00:11  lr: 0.004763  min_lr: 0.004763  loss: 3.1679 (3.1750)  weight_decay: 0.0500 (0.0500)  time: 0.6918  data: 0.0007  max mem: 57114
Epoch: [142]  [150/156]  eta: 0:00:04  lr: 0.004760  min_lr: 0.004760  loss: 2.9572 (3.1634)  weight_decay: 0.0500 (0.0500)  time: 0.6870  data: 0.0001  max mem: 57114
Epoch: [142]  [155/156]  eta: 0:00:00  lr: 0.004759  min_lr: 0.004759  loss: 3.0725 (3.1638)  weight_decay: 0.0500 (0.0500)  time: 0.6892  data: 0.0001  max mem: 57114
Epoch: [142] Total time: 0:01:53 (0.7266 s / it)
Averaged stats: lr: 0.004759  min_lr: 0.004759  loss: 3.0725 (3.1386)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8627 (0.8627)  acc1: 83.3333 (83.3333)  acc5: 97.3958 (97.3958)  time: 2.0905  data: 1.8343  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9578 (0.9627)  acc1: 80.7292 (79.2839)  acc5: 95.8333 (94.2455)  time: 0.5688  data: 0.3669  max mem: 57114
Test: Total time: 0:00:02 (0.5927 s / it)
* Acc@1 78.605 Acc@5 94.557 loss 1.042
Accuracy of the model on the 50000 test images: 78.6%
Max accuracy: 78.67%
Test:  [0/5]  eta: 0:00:11  loss: 1.1662 (1.1662)  acc1: 71.8750 (71.8750)  acc5: 90.1042 (90.1042)  time: 2.2931  data: 2.0498  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.5130 (1.4648)  acc1: 64.0625 (63.1714)  acc5: 88.0208 (84.0153)  time: 0.6092  data: 0.4100  max mem: 57114
Test: Total time: 0:00:03 (0.6209 s / it)
* Acc@1 59.649 Acc@5 82.651 loss 1.733
Accuracy of the model EMA on 50000 test images: 59.6%
Max EMA accuracy: 59.65%
Epoch: [143]  [  0/156]  eta: 0:07:00  lr: 0.004759  min_lr: 0.004759  loss: 3.0555 (3.0555)  weight_decay: 0.0500 (0.0500)  time: 2.6951  data: 2.0476  max mem: 57114
Epoch: [143]  [ 10/156]  eta: 0:02:12  lr: 0.004756  min_lr: 0.004756  loss: 3.4129 (3.3331)  weight_decay: 0.0500 (0.0500)  time: 0.9102  data: 0.1864  max mem: 57114
Epoch: [143]  [ 20/156]  eta: 0:01:52  lr: 0.004753  min_lr: 0.004753  loss: 3.4129 (3.2863)  weight_decay: 0.0500 (0.0500)  time: 0.7300  data: 0.0004  max mem: 57114
Epoch: [143]  [ 30/156]  eta: 0:01:39  lr: 0.004750  min_lr: 0.004750  loss: 3.4384 (3.3101)  weight_decay: 0.0500 (0.0500)  time: 0.7255  data: 0.0005  max mem: 57114
Epoch: [143]  [ 40/156]  eta: 0:01:29  lr: 0.004747  min_lr: 0.004747  loss: 3.3786 (3.2737)  weight_decay: 0.0500 (0.0500)  time: 0.7175  data: 0.0004  max mem: 57114
Epoch: [143]  [ 50/156]  eta: 0:01:20  lr: 0.004745  min_lr: 0.004745  loss: 3.1940 (3.2560)  weight_decay: 0.0500 (0.0500)  time: 0.7125  data: 0.0004  max mem: 57114
Epoch: [143]  [ 60/156]  eta: 0:01:12  lr: 0.004742  min_lr: 0.004742  loss: 3.2380 (3.2449)  weight_decay: 0.0500 (0.0500)  time: 0.7243  data: 0.0004  max mem: 57114
Epoch: [143]  [ 70/156]  eta: 0:01:04  lr: 0.004739  min_lr: 0.004739  loss: 3.2380 (3.2421)  weight_decay: 0.0500 (0.0500)  time: 0.7312  data: 0.0004  max mem: 57114
Epoch: [143]  [ 80/156]  eta: 0:00:56  lr: 0.004736  min_lr: 0.004736  loss: 3.3776 (3.2533)  weight_decay: 0.0500 (0.0500)  time: 0.7145  data: 0.0004  max mem: 57114
Epoch: [143]  [ 90/156]  eta: 0:00:48  lr: 0.004733  min_lr: 0.004733  loss: 3.3476 (3.2541)  weight_decay: 0.0500 (0.0500)  time: 0.7077  data: 0.0003  max mem: 57114
Epoch: [143]  [100/156]  eta: 0:00:41  lr: 0.004730  min_lr: 0.004730  loss: 3.3336 (3.2371)  weight_decay: 0.0500 (0.0500)  time: 0.7091  data: 0.0004  max mem: 57114
Epoch: [143]  [110/156]  eta: 0:00:33  lr: 0.004728  min_lr: 0.004728  loss: 3.3226 (3.2423)  weight_decay: 0.0500 (0.0500)  time: 0.7053  data: 0.0004  max mem: 57114
Epoch: [143]  [120/156]  eta: 0:00:26  lr: 0.004725  min_lr: 0.004725  loss: 3.1446 (3.2189)  weight_decay: 0.0500 (0.0500)  time: 0.7076  data: 0.0004  max mem: 57114
Epoch: [143]  [130/156]  eta: 0:00:18  lr: 0.004722  min_lr: 0.004722  loss: 3.1446 (3.2249)  weight_decay: 0.0500 (0.0500)  time: 0.7015  data: 0.0008  max mem: 57114
Epoch: [143]  [140/156]  eta: 0:00:11  lr: 0.004719  min_lr: 0.004719  loss: 3.4025 (3.2246)  weight_decay: 0.0500 (0.0500)  time: 0.6951  data: 0.0007  max mem: 57114
Epoch: [143]  [150/156]  eta: 0:00:04  lr: 0.004716  min_lr: 0.004716  loss: 3.3373 (3.2261)  weight_decay: 0.0500 (0.0500)  time: 0.6899  data: 0.0001  max mem: 57114
Epoch: [143]  [155/156]  eta: 0:00:00  lr: 0.004715  min_lr: 0.004715  loss: 3.3773 (3.2348)  weight_decay: 0.0500 (0.0500)  time: 0.6871  data: 0.0001  max mem: 57114
Epoch: [143] Total time: 0:01:53 (0.7261 s / it)
Averaged stats: lr: 0.004715  min_lr: 0.004715  loss: 3.3773 (3.1380)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0398 (1.0398)  acc1: 84.8958 (84.8958)  acc5: 97.9167 (97.9167)  time: 2.0836  data: 1.8276  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0927 (1.1011)  acc1: 82.2917 (79.7954)  acc5: 96.8750 (95.5243)  time: 0.5674  data: 0.3656  max mem: 57114
Test: Total time: 0:00:02 (0.5909 s / it)
* Acc@1 78.878 Acc@5 94.569 loss 1.196
Accuracy of the model on the 50000 test images: 78.9%
Max accuracy: 78.88%
Test:  [0/5]  eta: 0:00:10  loss: 1.1296 (1.1296)  acc1: 72.3958 (72.3958)  acc5: 90.6250 (90.6250)  time: 2.0217  data: 1.7783  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.4672 (1.4218)  acc1: 66.1458 (64.3223)  acc5: 88.0208 (84.1432)  time: 0.5551  data: 0.3558  max mem: 57114
Test: Total time: 0:00:02 (0.5674 s / it)
* Acc@1 60.542 Acc@5 83.312 loss 1.686
Accuracy of the model EMA on 50000 test images: 60.5%
Max EMA accuracy: 60.54%
Epoch: [144]  [  0/156]  eta: 0:08:51  lr: 0.004715  min_lr: 0.004715  loss: 2.6195 (2.6195)  weight_decay: 0.0500 (0.0500)  time: 3.4081  data: 2.7578  max mem: 57114
Epoch: [144]  [ 10/156]  eta: 0:02:25  lr: 0.004712  min_lr: 0.004712  loss: 3.1916 (3.0126)  weight_decay: 0.0500 (0.0500)  time: 0.9963  data: 0.2509  max mem: 57114
Epoch: [144]  [ 20/156]  eta: 0:01:59  lr: 0.004709  min_lr: 0.004709  loss: 3.3182 (3.1020)  weight_decay: 0.0500 (0.0500)  time: 0.7487  data: 0.0003  max mem: 57114
Epoch: [144]  [ 30/156]  eta: 0:01:44  lr: 0.004706  min_lr: 0.004706  loss: 3.2857 (3.0535)  weight_decay: 0.0500 (0.0500)  time: 0.7384  data: 0.0003  max mem: 57114
Epoch: [144]  [ 40/156]  eta: 0:01:33  lr: 0.004703  min_lr: 0.004703  loss: 3.2332 (3.0788)  weight_decay: 0.0500 (0.0500)  time: 0.7369  data: 0.0004  max mem: 57114
Epoch: [144]  [ 50/156]  eta: 0:01:24  lr: 0.004700  min_lr: 0.004700  loss: 3.3658 (3.1595)  weight_decay: 0.0500 (0.0500)  time: 0.7364  data: 0.0004  max mem: 57114
Epoch: [144]  [ 60/156]  eta: 0:01:14  lr: 0.004698  min_lr: 0.004698  loss: 3.3593 (3.1639)  weight_decay: 0.0500 (0.0500)  time: 0.7223  data: 0.0004  max mem: 57114
Epoch: [144]  [ 70/156]  eta: 0:01:06  lr: 0.004695  min_lr: 0.004695  loss: 3.1473 (3.1565)  weight_decay: 0.0500 (0.0500)  time: 0.7229  data: 0.0004  max mem: 57114
Epoch: [144]  [ 80/156]  eta: 0:00:58  lr: 0.004692  min_lr: 0.004692  loss: 2.9571 (3.1281)  weight_decay: 0.0500 (0.0500)  time: 0.7199  data: 0.0004  max mem: 57114
Epoch: [144]  [ 90/156]  eta: 0:00:50  lr: 0.004689  min_lr: 0.004689  loss: 3.1817 (3.1319)  weight_decay: 0.0500 (0.0500)  time: 0.7059  data: 0.0004  max mem: 57114
Epoch: [144]  [100/156]  eta: 0:00:42  lr: 0.004686  min_lr: 0.004686  loss: 3.1817 (3.1244)  weight_decay: 0.0500 (0.0500)  time: 0.6981  data: 0.0003  max mem: 57114
Epoch: [144]  [110/156]  eta: 0:00:34  lr: 0.004683  min_lr: 0.004683  loss: 3.2904 (3.1330)  weight_decay: 0.0500 (0.0500)  time: 0.6984  data: 0.0003  max mem: 57114
Epoch: [144]  [120/156]  eta: 0:00:26  lr: 0.004681  min_lr: 0.004681  loss: 3.3853 (3.1293)  weight_decay: 0.0500 (0.0500)  time: 0.7076  data: 0.0004  max mem: 57114
Epoch: [144]  [130/156]  eta: 0:00:19  lr: 0.004678  min_lr: 0.004678  loss: 2.8310 (3.0982)  weight_decay: 0.0500 (0.0500)  time: 0.6999  data: 0.0008  max mem: 57114
Epoch: [144]  [140/156]  eta: 0:00:11  lr: 0.004675  min_lr: 0.004675  loss: 2.9535 (3.1090)  weight_decay: 0.0500 (0.0500)  time: 0.6885  data: 0.0007  max mem: 57114
Epoch: [144]  [150/156]  eta: 0:00:04  lr: 0.004672  min_lr: 0.004672  loss: 3.2450 (3.1111)  weight_decay: 0.0500 (0.0500)  time: 0.6898  data: 0.0001  max mem: 57114
Epoch: [144]  [155/156]  eta: 0:00:00  lr: 0.004671  min_lr: 0.004671  loss: 3.2223 (3.1076)  weight_decay: 0.0500 (0.0500)  time: 0.6880  data: 0.0001  max mem: 57114
Epoch: [144] Total time: 0:01:54 (0.7342 s / it)
Averaged stats: lr: 0.004671  min_lr: 0.004671  loss: 3.2223 (3.1302)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7568 (0.7568)  acc1: 88.5417 (88.5417)  acc5: 97.3958 (97.3958)  time: 2.1194  data: 1.8636  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9657 (0.9194)  acc1: 79.6875 (80.0512)  acc5: 96.3542 (94.8849)  time: 0.5746  data: 0.3728  max mem: 57114
Test: Total time: 0:00:02 (0.5974 s / it)
* Acc@1 79.124 Acc@5 94.829 loss 1.019
Accuracy of the model on the 50000 test images: 79.1%
Max accuracy: 79.12%
Test:  [0/5]  eta: 0:00:09  loss: 1.0968 (1.0968)  acc1: 73.4375 (73.4375)  acc5: 90.6250 (90.6250)  time: 1.9434  data: 1.7001  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.4250 (1.3829)  acc1: 66.6667 (64.5780)  acc5: 89.0625 (84.9105)  time: 0.5394  data: 0.3401  max mem: 57114
Test: Total time: 0:00:02 (0.5500 s / it)
* Acc@1 61.345 Acc@5 83.951 loss 1.643
Accuracy of the model EMA on 50000 test images: 61.3%
Max EMA accuracy: 61.35%
Epoch: [145]  [  0/156]  eta: 0:09:08  lr: 0.004670  min_lr: 0.004670  loss: 3.4100 (3.4100)  weight_decay: 0.0500 (0.0500)  time: 3.5160  data: 2.8687  max mem: 57114
Epoch: [145]  [ 10/156]  eta: 0:02:22  lr: 0.004668  min_lr: 0.004668  loss: 3.4100 (3.3075)  weight_decay: 0.0500 (0.0500)  time: 0.9773  data: 0.2611  max mem: 57114
Epoch: [145]  [ 20/156]  eta: 0:01:56  lr: 0.004665  min_lr: 0.004665  loss: 3.3183 (3.3485)  weight_decay: 0.0500 (0.0500)  time: 0.7268  data: 0.0003  max mem: 57114
Epoch: [145]  [ 30/156]  eta: 0:01:42  lr: 0.004662  min_lr: 0.004662  loss: 3.1847 (3.2489)  weight_decay: 0.0500 (0.0500)  time: 0.7243  data: 0.0004  max mem: 57114
Epoch: [145]  [ 40/156]  eta: 0:01:32  lr: 0.004659  min_lr: 0.004659  loss: 3.3433 (3.2610)  weight_decay: 0.0500 (0.0500)  time: 0.7272  data: 0.0004  max mem: 57114
Epoch: [145]  [ 50/156]  eta: 0:01:23  lr: 0.004656  min_lr: 0.004656  loss: 3.3471 (3.2187)  weight_decay: 0.0500 (0.0500)  time: 0.7476  data: 0.0004  max mem: 57114
Epoch: [145]  [ 60/156]  eta: 0:01:14  lr: 0.004653  min_lr: 0.004653  loss: 3.0658 (3.1809)  weight_decay: 0.0500 (0.0500)  time: 0.7375  data: 0.0004  max mem: 57114
Epoch: [145]  [ 70/156]  eta: 0:01:06  lr: 0.004651  min_lr: 0.004651  loss: 2.6775 (3.0947)  weight_decay: 0.0500 (0.0500)  time: 0.7201  data: 0.0004  max mem: 57114
Epoch: [145]  [ 80/156]  eta: 0:00:57  lr: 0.004648  min_lr: 0.004648  loss: 2.7215 (3.0874)  weight_decay: 0.0500 (0.0500)  time: 0.7155  data: 0.0004  max mem: 57114
Epoch: [145]  [ 90/156]  eta: 0:00:49  lr: 0.004645  min_lr: 0.004645  loss: 3.1032 (3.0770)  weight_decay: 0.0500 (0.0500)  time: 0.7070  data: 0.0003  max mem: 57114
Epoch: [145]  [100/156]  eta: 0:00:41  lr: 0.004642  min_lr: 0.004642  loss: 3.2003 (3.0851)  weight_decay: 0.0500 (0.0500)  time: 0.7001  data: 0.0004  max mem: 57114
Epoch: [145]  [110/156]  eta: 0:00:34  lr: 0.004639  min_lr: 0.004639  loss: 3.2414 (3.0863)  weight_decay: 0.0500 (0.0500)  time: 0.7046  data: 0.0004  max mem: 57114
Epoch: [145]  [120/156]  eta: 0:00:26  lr: 0.004636  min_lr: 0.004636  loss: 3.1414 (3.0812)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0004  max mem: 57114
Epoch: [145]  [130/156]  eta: 0:00:19  lr: 0.004634  min_lr: 0.004634  loss: 3.2502 (3.0955)  weight_decay: 0.0500 (0.0500)  time: 0.6985  data: 0.0006  max mem: 57114
Epoch: [145]  [140/156]  eta: 0:00:11  lr: 0.004631  min_lr: 0.004631  loss: 3.3870 (3.1077)  weight_decay: 0.0500 (0.0500)  time: 0.6898  data: 0.0005  max mem: 57114
Epoch: [145]  [150/156]  eta: 0:00:04  lr: 0.004628  min_lr: 0.004628  loss: 3.2476 (3.1024)  weight_decay: 0.0500 (0.0500)  time: 0.6841  data: 0.0001  max mem: 57114
Epoch: [145]  [155/156]  eta: 0:00:00  lr: 0.004626  min_lr: 0.004626  loss: 3.0972 (3.0957)  weight_decay: 0.0500 (0.0500)  time: 0.6824  data: 0.0001  max mem: 57114
Epoch: [145] Total time: 0:01:54 (0.7319 s / it)
Averaged stats: lr: 0.004626  min_lr: 0.004626  loss: 3.0972 (3.1261)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8940 (0.8940)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.1028  data: 1.8472  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9510 (1.0006)  acc1: 82.2917 (79.7954)  acc5: 96.3542 (94.6292)  time: 0.5712  data: 0.3695  max mem: 57114
Test: Total time: 0:00:02 (0.5917 s / it)
* Acc@1 78.359 Acc@5 94.369 loss 1.093
Accuracy of the model on the 50000 test images: 78.4%
Max accuracy: 79.12%
Test:  [0/5]  eta: 0:00:11  loss: 1.0667 (1.0667)  acc1: 74.4792 (74.4792)  acc5: 91.6667 (91.6667)  time: 2.3294  data: 2.0857  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.3850 (1.3464)  acc1: 67.1875 (65.7289)  acc5: 89.5833 (85.5499)  time: 0.6165  data: 0.4172  max mem: 57114
Test: Total time: 0:00:03 (0.6290 s / it)
* Acc@1 62.156 Acc@5 84.561 loss 1.602
Accuracy of the model EMA on 50000 test images: 62.2%
Max EMA accuracy: 62.16%
Epoch: [146]  [  0/156]  eta: 0:07:12  lr: 0.004626  min_lr: 0.004626  loss: 3.4905 (3.4905)  weight_decay: 0.0500 (0.0500)  time: 2.7696  data: 2.1191  max mem: 57114
Epoch: [146]  [ 10/156]  eta: 0:02:13  lr: 0.004623  min_lr: 0.004623  loss: 3.4399 (3.3732)  weight_decay: 0.0500 (0.0500)  time: 0.9132  data: 0.1929  max mem: 57114
Epoch: [146]  [ 20/156]  eta: 0:01:51  lr: 0.004620  min_lr: 0.004620  loss: 3.2715 (3.1972)  weight_decay: 0.0500 (0.0500)  time: 0.7249  data: 0.0003  max mem: 57114
Epoch: [146]  [ 30/156]  eta: 0:01:38  lr: 0.004618  min_lr: 0.004618  loss: 3.3243 (3.2619)  weight_decay: 0.0500 (0.0500)  time: 0.7147  data: 0.0003  max mem: 57114
Epoch: [146]  [ 40/156]  eta: 0:01:29  lr: 0.004615  min_lr: 0.004615  loss: 3.3996 (3.2326)  weight_decay: 0.0500 (0.0500)  time: 0.7130  data: 0.0004  max mem: 57114
Epoch: [146]  [ 50/156]  eta: 0:01:20  lr: 0.004612  min_lr: 0.004612  loss: 3.0898 (3.1627)  weight_decay: 0.0500 (0.0500)  time: 0.7223  data: 0.0003  max mem: 57114
Epoch: [146]  [ 60/156]  eta: 0:01:12  lr: 0.004609  min_lr: 0.004609  loss: 2.9559 (3.1771)  weight_decay: 0.0500 (0.0500)  time: 0.7235  data: 0.0004  max mem: 57114
Epoch: [146]  [ 70/156]  eta: 0:01:04  lr: 0.004606  min_lr: 0.004606  loss: 3.2465 (3.1714)  weight_decay: 0.0500 (0.0500)  time: 0.7132  data: 0.0004  max mem: 57114
Epoch: [146]  [ 80/156]  eta: 0:00:56  lr: 0.004603  min_lr: 0.004603  loss: 3.2710 (3.1693)  weight_decay: 0.0500 (0.0500)  time: 0.7101  data: 0.0004  max mem: 57114
Epoch: [146]  [ 90/156]  eta: 0:00:48  lr: 0.004601  min_lr: 0.004601  loss: 3.1778 (3.1454)  weight_decay: 0.0500 (0.0500)  time: 0.7080  data: 0.0004  max mem: 57114
Epoch: [146]  [100/156]  eta: 0:00:41  lr: 0.004598  min_lr: 0.004598  loss: 2.7320 (3.1193)  weight_decay: 0.0500 (0.0500)  time: 0.7051  data: 0.0004  max mem: 57114
Epoch: [146]  [110/156]  eta: 0:00:33  lr: 0.004595  min_lr: 0.004595  loss: 3.0096 (3.1134)  weight_decay: 0.0500 (0.0500)  time: 0.7110  data: 0.0003  max mem: 57114
Epoch: [146]  [120/156]  eta: 0:00:26  lr: 0.004592  min_lr: 0.004592  loss: 3.2725 (3.1217)  weight_decay: 0.0500 (0.0500)  time: 0.7109  data: 0.0003  max mem: 57114
Epoch: [146]  [130/156]  eta: 0:00:18  lr: 0.004589  min_lr: 0.004589  loss: 3.4049 (3.1182)  weight_decay: 0.0500 (0.0500)  time: 0.7039  data: 0.0008  max mem: 57114
Epoch: [146]  [140/156]  eta: 0:00:11  lr: 0.004586  min_lr: 0.004586  loss: 3.2805 (3.1281)  weight_decay: 0.0500 (0.0500)  time: 0.6924  data: 0.0007  max mem: 57114
Epoch: [146]  [150/156]  eta: 0:00:04  lr: 0.004584  min_lr: 0.004584  loss: 3.1616 (3.1187)  weight_decay: 0.0500 (0.0500)  time: 0.6850  data: 0.0001  max mem: 57114
Epoch: [146]  [155/156]  eta: 0:00:00  lr: 0.004582  min_lr: 0.004582  loss: 3.0374 (3.1197)  weight_decay: 0.0500 (0.0500)  time: 0.6802  data: 0.0001  max mem: 57114
Epoch: [146] Total time: 0:01:52 (0.7242 s / it)
Averaged stats: lr: 0.004582  min_lr: 0.004582  loss: 3.0374 (3.1205)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9021 (0.9021)  acc1: 81.7708 (81.7708)  acc5: 96.3542 (96.3542)  time: 2.1534  data: 1.8977  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9230 (0.9576)  acc1: 81.2500 (79.0281)  acc5: 96.3542 (94.6292)  time: 0.5814  data: 0.3796  max mem: 57114
Test: Total time: 0:00:03 (0.6039 s / it)
* Acc@1 78.804 Acc@5 94.693 loss 1.025
Accuracy of the model on the 50000 test images: 78.8%
Max accuracy: 79.12%
Test:  [0/5]  eta: 0:00:11  loss: 1.0388 (1.0388)  acc1: 75.5208 (75.5208)  acc5: 92.1875 (92.1875)  time: 2.3060  data: 2.0625  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.3454 (1.3108)  acc1: 67.7083 (66.3683)  acc5: 89.5833 (85.9335)  time: 0.6119  data: 0.4126  max mem: 57114
Test: Total time: 0:00:03 (0.6236 s / it)
* Acc@1 62.902 Acc@5 85.126 loss 1.563
Accuracy of the model EMA on 50000 test images: 62.9%
Max EMA accuracy: 62.90%
Epoch: [147]  [  0/156]  eta: 0:05:57  lr: 0.004582  min_lr: 0.004582  loss: 3.3232 (3.3232)  weight_decay: 0.0500 (0.0500)  time: 2.2899  data: 1.6370  max mem: 57114
Epoch: [147]  [ 10/156]  eta: 0:02:04  lr: 0.004579  min_lr: 0.004579  loss: 3.2422 (3.0401)  weight_decay: 0.0500 (0.0500)  time: 0.8527  data: 0.1491  max mem: 57114
Epoch: [147]  [ 20/156]  eta: 0:01:47  lr: 0.004576  min_lr: 0.004576  loss: 3.2422 (3.1094)  weight_decay: 0.0500 (0.0500)  time: 0.7171  data: 0.0003  max mem: 57114
Epoch: [147]  [ 30/156]  eta: 0:01:37  lr: 0.004573  min_lr: 0.004573  loss: 3.3630 (3.1208)  weight_decay: 0.0500 (0.0500)  time: 0.7268  data: 0.0003  max mem: 57114
Epoch: [147]  [ 40/156]  eta: 0:01:28  lr: 0.004570  min_lr: 0.004570  loss: 3.1622 (3.0687)  weight_decay: 0.0500 (0.0500)  time: 0.7263  data: 0.0003  max mem: 57114
Epoch: [147]  [ 50/156]  eta: 0:01:19  lr: 0.004568  min_lr: 0.004568  loss: 3.2055 (3.1095)  weight_decay: 0.0500 (0.0500)  time: 0.7220  data: 0.0003  max mem: 57114
Epoch: [147]  [ 60/156]  eta: 0:01:11  lr: 0.004565  min_lr: 0.004565  loss: 3.1824 (3.0794)  weight_decay: 0.0500 (0.0500)  time: 0.7207  data: 0.0003  max mem: 57114
Epoch: [147]  [ 70/156]  eta: 0:01:03  lr: 0.004562  min_lr: 0.004562  loss: 3.1918 (3.0906)  weight_decay: 0.0500 (0.0500)  time: 0.7212  data: 0.0003  max mem: 57114
Epoch: [147]  [ 80/156]  eta: 0:00:56  lr: 0.004559  min_lr: 0.004559  loss: 3.3017 (3.1217)  weight_decay: 0.0500 (0.0500)  time: 0.7091  data: 0.0004  max mem: 57114
Epoch: [147]  [ 90/156]  eta: 0:00:48  lr: 0.004556  min_lr: 0.004556  loss: 3.2772 (3.1143)  weight_decay: 0.0500 (0.0500)  time: 0.7044  data: 0.0004  max mem: 57114
Epoch: [147]  [100/156]  eta: 0:00:40  lr: 0.004553  min_lr: 0.004553  loss: 3.0074 (3.0982)  weight_decay: 0.0500 (0.0500)  time: 0.7036  data: 0.0004  max mem: 57114
Epoch: [147]  [110/156]  eta: 0:00:33  lr: 0.004550  min_lr: 0.004550  loss: 3.0258 (3.0864)  weight_decay: 0.0500 (0.0500)  time: 0.7029  data: 0.0004  max mem: 57114
Epoch: [147]  [120/156]  eta: 0:00:26  lr: 0.004548  min_lr: 0.004548  loss: 2.9980 (3.0755)  weight_decay: 0.0500 (0.0500)  time: 0.7051  data: 0.0004  max mem: 57114
Epoch: [147]  [130/156]  eta: 0:00:18  lr: 0.004545  min_lr: 0.004545  loss: 2.7147 (3.0520)  weight_decay: 0.0500 (0.0500)  time: 0.7001  data: 0.0008  max mem: 57114
Epoch: [147]  [140/156]  eta: 0:00:11  lr: 0.004542  min_lr: 0.004542  loss: 2.8897 (3.0544)  weight_decay: 0.0500 (0.0500)  time: 0.6928  data: 0.0006  max mem: 57114
Epoch: [147]  [150/156]  eta: 0:00:04  lr: 0.004539  min_lr: 0.004539  loss: 3.2841 (3.0603)  weight_decay: 0.0500 (0.0500)  time: 0.6788  data: 0.0001  max mem: 57114
Epoch: [147]  [155/156]  eta: 0:00:00  lr: 0.004538  min_lr: 0.004538  loss: 3.2841 (3.0676)  weight_decay: 0.0500 (0.0500)  time: 0.6753  data: 0.0001  max mem: 57114
Epoch: [147] Total time: 0:01:52 (0.7191 s / it)
Averaged stats: lr: 0.004538  min_lr: 0.004538  loss: 3.2841 (3.1258)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8712 (0.8712)  acc1: 83.3333 (83.3333)  acc5: 96.8750 (96.8750)  time: 2.0166  data: 1.7610  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0081 (0.9692)  acc1: 81.2500 (79.6675)  acc5: 95.8333 (94.7570)  time: 0.5539  data: 0.3523  max mem: 57114
Test: Total time: 0:00:02 (0.5759 s / it)
* Acc@1 78.896 Acc@5 94.679 loss 1.079
Accuracy of the model on the 50000 test images: 78.9%
Max accuracy: 79.12%
Test:  [0/5]  eta: 0:00:11  loss: 1.0134 (1.0134)  acc1: 76.5625 (76.5625)  acc5: 92.7083 (92.7083)  time: 2.2864  data: 2.0428  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.3080 (1.2778)  acc1: 69.2708 (67.3913)  acc5: 90.1042 (86.3171)  time: 0.6080  data: 0.4087  max mem: 57114
Test: Total time: 0:00:03 (0.6185 s / it)
* Acc@1 63.639 Acc@5 85.632 loss 1.526
Accuracy of the model EMA on 50000 test images: 63.6%
Max EMA accuracy: 63.64%
Epoch: [148]  [  0/156]  eta: 0:06:59  lr: 0.004537  min_lr: 0.004537  loss: 3.0239 (3.0239)  weight_decay: 0.0500 (0.0500)  time: 2.6876  data: 2.0393  max mem: 57114
Epoch: [148]  [ 10/156]  eta: 0:02:13  lr: 0.004535  min_lr: 0.004535  loss: 3.2050 (3.0766)  weight_decay: 0.0500 (0.0500)  time: 0.9165  data: 0.1857  max mem: 57114
Epoch: [148]  [ 20/156]  eta: 0:01:50  lr: 0.004532  min_lr: 0.004532  loss: 3.2915 (3.1386)  weight_decay: 0.0500 (0.0500)  time: 0.7202  data: 0.0003  max mem: 57114
Epoch: [148]  [ 30/156]  eta: 0:01:39  lr: 0.004529  min_lr: 0.004529  loss: 3.2281 (3.0362)  weight_decay: 0.0500 (0.0500)  time: 0.7182  data: 0.0004  max mem: 57114
Epoch: [148]  [ 40/156]  eta: 0:01:29  lr: 0.004526  min_lr: 0.004526  loss: 2.9398 (3.0590)  weight_decay: 0.0500 (0.0500)  time: 0.7302  data: 0.0004  max mem: 57114
Epoch: [148]  [ 50/156]  eta: 0:01:20  lr: 0.004523  min_lr: 0.004523  loss: 3.1205 (3.0550)  weight_decay: 0.0500 (0.0500)  time: 0.7252  data: 0.0004  max mem: 57114
Epoch: [148]  [ 60/156]  eta: 0:01:12  lr: 0.004520  min_lr: 0.004520  loss: 3.3162 (3.0769)  weight_decay: 0.0500 (0.0500)  time: 0.7247  data: 0.0004  max mem: 57114
Epoch: [148]  [ 70/156]  eta: 0:01:04  lr: 0.004517  min_lr: 0.004517  loss: 3.2444 (3.0910)  weight_decay: 0.0500 (0.0500)  time: 0.7154  data: 0.0003  max mem: 57114
Epoch: [148]  [ 80/156]  eta: 0:00:56  lr: 0.004515  min_lr: 0.004515  loss: 3.1591 (3.0763)  weight_decay: 0.0500 (0.0500)  time: 0.7023  data: 0.0003  max mem: 57114
Epoch: [148]  [ 90/156]  eta: 0:00:48  lr: 0.004512  min_lr: 0.004512  loss: 3.1591 (3.0844)  weight_decay: 0.0500 (0.0500)  time: 0.7027  data: 0.0003  max mem: 57114
Epoch: [148]  [100/156]  eta: 0:00:41  lr: 0.004509  min_lr: 0.004509  loss: 3.1704 (3.0623)  weight_decay: 0.0500 (0.0500)  time: 0.7111  data: 0.0004  max mem: 57114
Epoch: [148]  [110/156]  eta: 0:00:33  lr: 0.004506  min_lr: 0.004506  loss: 3.2904 (3.0832)  weight_decay: 0.0500 (0.0500)  time: 0.7127  data: 0.0003  max mem: 57114
Epoch: [148]  [120/156]  eta: 0:00:26  lr: 0.004503  min_lr: 0.004503  loss: 3.3698 (3.0822)  weight_decay: 0.0500 (0.0500)  time: 0.7074  data: 0.0004  max mem: 57114
Epoch: [148]  [130/156]  eta: 0:00:18  lr: 0.004500  min_lr: 0.004500  loss: 3.2531 (3.0860)  weight_decay: 0.0500 (0.0500)  time: 0.6994  data: 0.0009  max mem: 57114
Epoch: [148]  [140/156]  eta: 0:00:11  lr: 0.004497  min_lr: 0.004497  loss: 3.2754 (3.0984)  weight_decay: 0.0500 (0.0500)  time: 0.6879  data: 0.0007  max mem: 57114
Epoch: [148]  [150/156]  eta: 0:00:04  lr: 0.004495  min_lr: 0.004495  loss: 3.2202 (3.1093)  weight_decay: 0.0500 (0.0500)  time: 0.6817  data: 0.0001  max mem: 57114
Epoch: [148]  [155/156]  eta: 0:00:00  lr: 0.004493  min_lr: 0.004493  loss: 3.1817 (3.1064)  weight_decay: 0.0500 (0.0500)  time: 0.6859  data: 0.0001  max mem: 57114
Epoch: [148] Total time: 0:01:53 (0.7261 s / it)
Averaged stats: lr: 0.004493  min_lr: 0.004493  loss: 3.1817 (3.1279)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8114 (0.8114)  acc1: 85.9375 (85.9375)  acc5: 99.4792 (99.4792)  time: 2.0752  data: 1.8192  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0277 (0.9525)  acc1: 82.8125 (81.2020)  acc5: 97.3958 (96.5473)  time: 0.5657  data: 0.3639  max mem: 57114
Test: Total time: 0:00:02 (0.5864 s / it)
* Acc@1 79.384 Acc@5 94.867 loss 1.067
Accuracy of the model on the 50000 test images: 79.4%
Max accuracy: 79.38%
Test:  [0/5]  eta: 0:00:10  loss: 0.9907 (0.9907)  acc1: 77.0833 (77.0833)  acc5: 92.7083 (92.7083)  time: 2.0184  data: 1.7749  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2729 (1.2470)  acc1: 69.2708 (67.7749)  acc5: 90.6250 (86.8286)  time: 0.5543  data: 0.3551  max mem: 57114
Test: Total time: 0:00:02 (0.5686 s / it)
* Acc@1 64.286 Acc@5 86.125 loss 1.491
Accuracy of the model EMA on 50000 test images: 64.3%
Max EMA accuracy: 64.29%
Epoch: [149]  [  0/156]  eta: 0:07:15  lr: 0.004493  min_lr: 0.004493  loss: 2.7677 (2.7677)  weight_decay: 0.0500 (0.0500)  time: 2.7896  data: 2.1387  max mem: 57114
Epoch: [149]  [ 10/156]  eta: 0:02:14  lr: 0.004490  min_lr: 0.004490  loss: 3.1821 (3.0584)  weight_decay: 0.0500 (0.0500)  time: 0.9192  data: 0.1947  max mem: 57114
Epoch: [149]  [ 20/156]  eta: 0:01:52  lr: 0.004487  min_lr: 0.004487  loss: 3.1821 (3.1679)  weight_decay: 0.0500 (0.0500)  time: 0.7319  data: 0.0003  max mem: 57114
Epoch: [149]  [ 30/156]  eta: 0:01:41  lr: 0.004484  min_lr: 0.004484  loss: 3.1439 (3.1599)  weight_decay: 0.0500 (0.0500)  time: 0.7396  data: 0.0003  max mem: 57114
Epoch: [149]  [ 40/156]  eta: 0:01:30  lr: 0.004481  min_lr: 0.004481  loss: 3.0870 (3.1089)  weight_decay: 0.0500 (0.0500)  time: 0.7309  data: 0.0003  max mem: 57114
Epoch: [149]  [ 50/156]  eta: 0:01:21  lr: 0.004479  min_lr: 0.004479  loss: 3.1570 (3.0855)  weight_decay: 0.0500 (0.0500)  time: 0.7179  data: 0.0003  max mem: 57114
Epoch: [149]  [ 60/156]  eta: 0:01:13  lr: 0.004476  min_lr: 0.004476  loss: 3.2388 (3.0814)  weight_decay: 0.0500 (0.0500)  time: 0.7351  data: 0.0005  max mem: 57114
Epoch: [149]  [ 70/156]  eta: 0:01:05  lr: 0.004473  min_lr: 0.004473  loss: 3.3638 (3.1187)  weight_decay: 0.0500 (0.0500)  time: 0.7251  data: 0.0006  max mem: 57114
Epoch: [149]  [ 80/156]  eta: 0:00:57  lr: 0.004470  min_lr: 0.004470  loss: 3.3755 (3.1272)  weight_decay: 0.0500 (0.0500)  time: 0.7131  data: 0.0004  max mem: 57114
Epoch: [149]  [ 90/156]  eta: 0:00:49  lr: 0.004467  min_lr: 0.004467  loss: 3.1422 (3.1140)  weight_decay: 0.0500 (0.0500)  time: 0.7215  data: 0.0004  max mem: 57114
Epoch: [149]  [100/156]  eta: 0:00:41  lr: 0.004464  min_lr: 0.004464  loss: 3.1422 (3.1046)  weight_decay: 0.0500 (0.0500)  time: 0.7087  data: 0.0004  max mem: 57114
Epoch: [149]  [110/156]  eta: 0:00:34  lr: 0.004461  min_lr: 0.004461  loss: 3.1780 (3.1101)  weight_decay: 0.0500 (0.0500)  time: 0.7060  data: 0.0004  max mem: 57114
Epoch: [149]  [120/156]  eta: 0:00:26  lr: 0.004459  min_lr: 0.004459  loss: 3.3198 (3.1094)  weight_decay: 0.0500 (0.0500)  time: 0.7085  data: 0.0003  max mem: 57114
Epoch: [149]  [130/156]  eta: 0:00:19  lr: 0.004456  min_lr: 0.004456  loss: 3.0770 (3.0907)  weight_decay: 0.0500 (0.0500)  time: 0.6976  data: 0.0007  max mem: 57114
Epoch: [149]  [140/156]  eta: 0:00:11  lr: 0.004453  min_lr: 0.004453  loss: 3.0770 (3.0824)  weight_decay: 0.0500 (0.0500)  time: 0.6840  data: 0.0006  max mem: 57114
Epoch: [149]  [150/156]  eta: 0:00:04  lr: 0.004450  min_lr: 0.004450  loss: 3.2259 (3.0889)  weight_decay: 0.0500 (0.0500)  time: 0.6819  data: 0.0001  max mem: 57114
Epoch: [149]  [155/156]  eta: 0:00:00  lr: 0.004449  min_lr: 0.004449  loss: 3.1307 (3.0773)  weight_decay: 0.0500 (0.0500)  time: 0.6835  data: 0.0001  max mem: 57114
Epoch: [149] Total time: 0:01:53 (0.7289 s / it)
Averaged stats: lr: 0.004449  min_lr: 0.004449  loss: 3.1307 (3.1106)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6222 (0.6222)  acc1: 88.5417 (88.5417)  acc5: 99.4792 (99.4792)  time: 2.0484  data: 1.7929  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7895 (0.7724)  acc1: 84.3750 (80.9463)  acc5: 96.8750 (96.1637)  time: 0.5603  data: 0.3586  max mem: 57114
Test: Total time: 0:00:02 (0.5813 s / it)
* Acc@1 79.638 Acc@5 95.117 loss 0.891
Accuracy of the model on the 50000 test images: 79.6%
Max accuracy: 79.64%
Test:  [0/5]  eta: 0:00:10  loss: 0.9690 (0.9690)  acc1: 77.0833 (77.0833)  acc5: 92.7083 (92.7083)  time: 2.1488  data: 1.9052  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2388 (1.2177)  acc1: 69.2708 (68.2864)  acc5: 91.6667 (87.4680)  time: 0.5804  data: 0.3811  max mem: 57114
Test: Total time: 0:00:02 (0.5928 s / it)
* Acc@1 64.918 Acc@5 86.583 loss 1.458
Accuracy of the model EMA on 50000 test images: 64.9%
Max EMA accuracy: 64.92%
Epoch: [150]  [  0/156]  eta: 0:06:49  lr: 0.004448  min_lr: 0.004448  loss: 3.4153 (3.4153)  weight_decay: 0.0500 (0.0500)  time: 2.6261  data: 1.9763  max mem: 57114
Epoch: [150]  [ 10/156]  eta: 0:02:09  lr: 0.004445  min_lr: 0.004445  loss: 3.1161 (2.8818)  weight_decay: 0.0500 (0.0500)  time: 0.8877  data: 0.1799  max mem: 57114
Epoch: [150]  [ 20/156]  eta: 0:01:51  lr: 0.004443  min_lr: 0.004443  loss: 2.9162 (2.8640)  weight_decay: 0.0500 (0.0500)  time: 0.7272  data: 0.0003  max mem: 57114
Epoch: [150]  [ 30/156]  eta: 0:01:38  lr: 0.004440  min_lr: 0.004440  loss: 3.1409 (2.9070)  weight_decay: 0.0500 (0.0500)  time: 0.7238  data: 0.0004  max mem: 57114
Epoch: [150]  [ 40/156]  eta: 0:01:29  lr: 0.004437  min_lr: 0.004437  loss: 3.1918 (2.9421)  weight_decay: 0.0500 (0.0500)  time: 0.7175  data: 0.0004  max mem: 57114
Epoch: [150]  [ 50/156]  eta: 0:01:20  lr: 0.004434  min_lr: 0.004434  loss: 3.1918 (2.9982)  weight_decay: 0.0500 (0.0500)  time: 0.7229  data: 0.0004  max mem: 57114
Epoch: [150]  [ 60/156]  eta: 0:01:12  lr: 0.004431  min_lr: 0.004431  loss: 3.3879 (3.0544)  weight_decay: 0.0500 (0.0500)  time: 0.7373  data: 0.0004  max mem: 57114
Epoch: [150]  [ 70/156]  eta: 0:01:05  lr: 0.004428  min_lr: 0.004428  loss: 3.3706 (3.0738)  weight_decay: 0.0500 (0.0500)  time: 0.7488  data: 0.0004  max mem: 57114
Epoch: [150]  [ 80/156]  eta: 0:00:57  lr: 0.004425  min_lr: 0.004425  loss: 3.2811 (3.0903)  weight_decay: 0.0500 (0.0500)  time: 0.7293  data: 0.0004  max mem: 57114
Epoch: [150]  [ 90/156]  eta: 0:00:49  lr: 0.004423  min_lr: 0.004423  loss: 3.2811 (3.0836)  weight_decay: 0.0500 (0.0500)  time: 0.7106  data: 0.0004  max mem: 57114
Epoch: [150]  [100/156]  eta: 0:00:41  lr: 0.004420  min_lr: 0.004420  loss: 3.0087 (3.0481)  weight_decay: 0.0500 (0.0500)  time: 0.7136  data: 0.0003  max mem: 57114
Epoch: [150]  [110/156]  eta: 0:00:34  lr: 0.004417  min_lr: 0.004417  loss: 2.8160 (3.0454)  weight_decay: 0.0500 (0.0500)  time: 0.7117  data: 0.0003  max mem: 57114
Epoch: [150]  [120/156]  eta: 0:00:26  lr: 0.004414  min_lr: 0.004414  loss: 3.1493 (3.0502)  weight_decay: 0.0500 (0.0500)  time: 0.6993  data: 0.0003  max mem: 57114
Epoch: [150]  [130/156]  eta: 0:00:19  lr: 0.004411  min_lr: 0.004411  loss: 3.1574 (3.0599)  weight_decay: 0.0500 (0.0500)  time: 0.6967  data: 0.0007  max mem: 57114
Epoch: [150]  [140/156]  eta: 0:00:11  lr: 0.004408  min_lr: 0.004408  loss: 3.2648 (3.0569)  weight_decay: 0.0500 (0.0500)  time: 0.6882  data: 0.0006  max mem: 57114
Epoch: [150]  [150/156]  eta: 0:00:04  lr: 0.004405  min_lr: 0.004405  loss: 3.2648 (3.0741)  weight_decay: 0.0500 (0.0500)  time: 0.6824  data: 0.0001  max mem: 57114
Epoch: [150]  [155/156]  eta: 0:00:00  lr: 0.004404  min_lr: 0.004404  loss: 3.1900 (3.0796)  weight_decay: 0.0500 (0.0500)  time: 0.6842  data: 0.0001  max mem: 57114
Epoch: [150] Total time: 0:01:53 (0.7273 s / it)
Averaged stats: lr: 0.004404  min_lr: 0.004404  loss: 3.1900 (3.1069)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7879 (0.7879)  acc1: 84.3750 (84.3750)  acc5: 98.4375 (98.4375)  time: 2.0447  data: 1.7891  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9576 (0.9562)  acc1: 82.8125 (80.4348)  acc5: 95.8333 (94.7570)  time: 0.5596  data: 0.3579  max mem: 57114
Test: Total time: 0:00:02 (0.5835 s / it)
* Acc@1 79.170 Acc@5 94.691 loss 1.052
Accuracy of the model on the 50000 test images: 79.2%
Max accuracy: 79.64%
Test:  [0/5]  eta: 0:00:11  loss: 0.9507 (0.9507)  acc1: 77.6042 (77.6042)  acc5: 92.7083 (92.7083)  time: 2.3851  data: 2.1416  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.2080 (1.1915)  acc1: 70.3125 (69.1816)  acc5: 91.6667 (87.5959)  time: 0.6280  data: 0.4285  max mem: 57114
Test: Total time: 0:00:03 (0.6432 s / it)
* Acc@1 65.525 Acc@5 86.959 loss 1.429
Accuracy of the model EMA on 50000 test images: 65.5%
Max EMA accuracy: 65.53%
Epoch: [151]  [  0/156]  eta: 0:07:53  lr: 0.004404  min_lr: 0.004404  loss: 3.0349 (3.0349)  weight_decay: 0.0500 (0.0500)  time: 3.0326  data: 2.3851  max mem: 57114
Epoch: [151]  [ 10/156]  eta: 0:02:15  lr: 0.004401  min_lr: 0.004401  loss: 3.2505 (3.1618)  weight_decay: 0.0500 (0.0500)  time: 0.9299  data: 0.2171  max mem: 57114
Epoch: [151]  [ 20/156]  eta: 0:01:52  lr: 0.004398  min_lr: 0.004398  loss: 3.2226 (3.1840)  weight_decay: 0.0500 (0.0500)  time: 0.7184  data: 0.0003  max mem: 57114
Epoch: [151]  [ 30/156]  eta: 0:01:40  lr: 0.004395  min_lr: 0.004395  loss: 3.1761 (3.1598)  weight_decay: 0.0500 (0.0500)  time: 0.7233  data: 0.0003  max mem: 57114
Epoch: [151]  [ 40/156]  eta: 0:01:30  lr: 0.004392  min_lr: 0.004392  loss: 3.2095 (3.1906)  weight_decay: 0.0500 (0.0500)  time: 0.7232  data: 0.0004  max mem: 57114
Epoch: [151]  [ 50/156]  eta: 0:01:21  lr: 0.004389  min_lr: 0.004389  loss: 3.4740 (3.2007)  weight_decay: 0.0500 (0.0500)  time: 0.7248  data: 0.0004  max mem: 57114
Epoch: [151]  [ 60/156]  eta: 0:01:13  lr: 0.004387  min_lr: 0.004387  loss: 3.4466 (3.2161)  weight_decay: 0.0500 (0.0500)  time: 0.7274  data: 0.0004  max mem: 57114
Epoch: [151]  [ 70/156]  eta: 0:01:04  lr: 0.004384  min_lr: 0.004384  loss: 3.3334 (3.2168)  weight_decay: 0.0500 (0.0500)  time: 0.7211  data: 0.0004  max mem: 57114
Epoch: [151]  [ 80/156]  eta: 0:00:57  lr: 0.004381  min_lr: 0.004381  loss: 3.0838 (3.1976)  weight_decay: 0.0500 (0.0500)  time: 0.7176  data: 0.0004  max mem: 57114
Epoch: [151]  [ 90/156]  eta: 0:00:49  lr: 0.004378  min_lr: 0.004378  loss: 3.1199 (3.1833)  weight_decay: 0.0500 (0.0500)  time: 0.7091  data: 0.0004  max mem: 57114
Epoch: [151]  [100/156]  eta: 0:00:41  lr: 0.004375  min_lr: 0.004375  loss: 2.9111 (3.1478)  weight_decay: 0.0500 (0.0500)  time: 0.7040  data: 0.0004  max mem: 57114
Epoch: [151]  [110/156]  eta: 0:00:33  lr: 0.004372  min_lr: 0.004372  loss: 3.2341 (3.1616)  weight_decay: 0.0500 (0.0500)  time: 0.7059  data: 0.0003  max mem: 57114
Epoch: [151]  [120/156]  eta: 0:00:26  lr: 0.004369  min_lr: 0.004369  loss: 3.3910 (3.1681)  weight_decay: 0.0500 (0.0500)  time: 0.7106  data: 0.0003  max mem: 57114
Epoch: [151]  [130/156]  eta: 0:00:19  lr: 0.004366  min_lr: 0.004366  loss: 3.3630 (3.1619)  weight_decay: 0.0500 (0.0500)  time: 0.7026  data: 0.0009  max mem: 57114
Epoch: [151]  [140/156]  eta: 0:00:11  lr: 0.004364  min_lr: 0.004364  loss: 3.1003 (3.1583)  weight_decay: 0.0500 (0.0500)  time: 0.6929  data: 0.0007  max mem: 57114
Epoch: [151]  [150/156]  eta: 0:00:04  lr: 0.004361  min_lr: 0.004361  loss: 3.0285 (3.1433)  weight_decay: 0.0500 (0.0500)  time: 0.6882  data: 0.0001  max mem: 57114
Epoch: [151]  [155/156]  eta: 0:00:00  lr: 0.004359  min_lr: 0.004359  loss: 3.1003 (3.1477)  weight_decay: 0.0500 (0.0500)  time: 0.6845  data: 0.0001  max mem: 57114
Epoch: [151] Total time: 0:01:53 (0.7274 s / it)
Averaged stats: lr: 0.004359  min_lr: 0.004359  loss: 3.1003 (3.0998)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7673 (0.7673)  acc1: 86.9792 (86.9792)  acc5: 97.9167 (97.9167)  time: 2.1844  data: 1.9286  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0297 (0.9245)  acc1: 80.2083 (80.0512)  acc5: 96.3542 (95.3964)  time: 0.5889  data: 0.3860  max mem: 57114
Test: Total time: 0:00:03 (0.6125 s / it)
* Acc@1 79.034 Acc@5 94.849 loss 1.022
Accuracy of the model on the 50000 test images: 79.0%
Max accuracy: 79.64%
Test:  [0/5]  eta: 0:00:11  loss: 0.9323 (0.9323)  acc1: 77.6042 (77.6042)  acc5: 93.2292 (93.2292)  time: 2.3782  data: 2.1348  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1777 (1.1651)  acc1: 70.8333 (69.4373)  acc5: 91.6667 (87.7238)  time: 0.6263  data: 0.4270  max mem: 57114
Test: Total time: 0:00:03 (0.6360 s / it)
* Acc@1 66.095 Acc@5 87.388 loss 1.399
Accuracy of the model EMA on 50000 test images: 66.1%
Max EMA accuracy: 66.09%
Epoch: [152]  [  0/156]  eta: 0:07:01  lr: 0.004359  min_lr: 0.004359  loss: 3.2770 (3.2770)  weight_decay: 0.0500 (0.0500)  time: 2.7037  data: 2.0464  max mem: 57114
Epoch: [152]  [ 10/156]  eta: 0:02:09  lr: 0.004356  min_lr: 0.004356  loss: 3.2770 (3.1858)  weight_decay: 0.0500 (0.0500)  time: 0.8849  data: 0.1863  max mem: 57114
Epoch: [152]  [ 20/156]  eta: 0:01:50  lr: 0.004353  min_lr: 0.004353  loss: 3.1422 (3.0597)  weight_decay: 0.0500 (0.0500)  time: 0.7142  data: 0.0004  max mem: 57114
Epoch: [152]  [ 30/156]  eta: 0:01:38  lr: 0.004350  min_lr: 0.004350  loss: 3.0356 (3.0692)  weight_decay: 0.0500 (0.0500)  time: 0.7278  data: 0.0004  max mem: 57114
Epoch: [152]  [ 40/156]  eta: 0:01:28  lr: 0.004348  min_lr: 0.004348  loss: 3.0181 (3.0611)  weight_decay: 0.0500 (0.0500)  time: 0.7220  data: 0.0004  max mem: 57114
Epoch: [152]  [ 50/156]  eta: 0:01:20  lr: 0.004345  min_lr: 0.004345  loss: 3.0181 (3.0527)  weight_decay: 0.0500 (0.0500)  time: 0.7292  data: 0.0004  max mem: 57114
Epoch: [152]  [ 60/156]  eta: 0:01:12  lr: 0.004342  min_lr: 0.004342  loss: 3.0464 (3.0615)  weight_decay: 0.0500 (0.0500)  time: 0.7344  data: 0.0004  max mem: 57114
Epoch: [152]  [ 70/156]  eta: 0:01:04  lr: 0.004339  min_lr: 0.004339  loss: 2.9665 (3.0345)  weight_decay: 0.0500 (0.0500)  time: 0.7158  data: 0.0004  max mem: 57114
Epoch: [152]  [ 80/156]  eta: 0:00:56  lr: 0.004336  min_lr: 0.004336  loss: 2.8148 (3.0220)  weight_decay: 0.0500 (0.0500)  time: 0.7017  data: 0.0004  max mem: 57114
Epoch: [152]  [ 90/156]  eta: 0:00:48  lr: 0.004333  min_lr: 0.004333  loss: 3.0657 (3.0196)  weight_decay: 0.0500 (0.0500)  time: 0.7003  data: 0.0003  max mem: 57114
Epoch: [152]  [100/156]  eta: 0:00:41  lr: 0.004330  min_lr: 0.004330  loss: 3.1491 (3.0176)  weight_decay: 0.0500 (0.0500)  time: 0.7044  data: 0.0003  max mem: 57114
Epoch: [152]  [110/156]  eta: 0:00:33  lr: 0.004327  min_lr: 0.004327  loss: 3.0320 (3.0094)  weight_decay: 0.0500 (0.0500)  time: 0.7113  data: 0.0003  max mem: 57114
Epoch: [152]  [120/156]  eta: 0:00:26  lr: 0.004325  min_lr: 0.004325  loss: 3.2070 (3.0261)  weight_decay: 0.0500 (0.0500)  time: 0.7118  data: 0.0004  max mem: 57114
Epoch: [152]  [130/156]  eta: 0:00:18  lr: 0.004322  min_lr: 0.004322  loss: 3.1923 (3.0253)  weight_decay: 0.0500 (0.0500)  time: 0.7009  data: 0.0008  max mem: 57114
Epoch: [152]  [140/156]  eta: 0:00:11  lr: 0.004319  min_lr: 0.004319  loss: 3.0490 (3.0154)  weight_decay: 0.0500 (0.0500)  time: 0.6909  data: 0.0007  max mem: 57114
Epoch: [152]  [150/156]  eta: 0:00:04  lr: 0.004316  min_lr: 0.004316  loss: 3.0490 (3.0335)  weight_decay: 0.0500 (0.0500)  time: 0.6810  data: 0.0001  max mem: 57114
Epoch: [152]  [155/156]  eta: 0:00:00  lr: 0.004315  min_lr: 0.004315  loss: 3.4591 (3.0506)  weight_decay: 0.0500 (0.0500)  time: 0.6867  data: 0.0001  max mem: 57114
Epoch: [152] Total time: 0:01:52 (0.7237 s / it)
Averaged stats: lr: 0.004315  min_lr: 0.004315  loss: 3.4591 (3.0984)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9663 (0.9663)  acc1: 85.4167 (85.4167)  acc5: 97.3958 (97.3958)  time: 2.1707  data: 1.9149  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0904 (1.0302)  acc1: 80.2083 (80.6905)  acc5: 96.3542 (95.6522)  time: 0.5851  data: 0.3831  max mem: 57114
Test: Total time: 0:00:03 (0.6072 s / it)
* Acc@1 78.844 Acc@5 94.731 loss 1.156
Accuracy of the model on the 50000 test images: 78.8%
Max accuracy: 79.64%
Test:  [0/5]  eta: 0:00:11  loss: 0.9156 (0.9156)  acc1: 78.1250 (78.1250)  acc5: 93.2292 (93.2292)  time: 2.3028  data: 2.0592  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1502 (1.1409)  acc1: 71.8750 (69.9489)  acc5: 91.6667 (88.2353)  time: 0.6112  data: 0.4119  max mem: 57114
Test: Total time: 0:00:03 (0.6224 s / it)
* Acc@1 66.600 Acc@5 87.728 loss 1.372
Accuracy of the model EMA on 50000 test images: 66.6%
Max EMA accuracy: 66.60%
Epoch: [153]  [  0/156]  eta: 0:07:16  lr: 0.004314  min_lr: 0.004314  loss: 2.9710 (2.9710)  weight_decay: 0.0500 (0.0500)  time: 2.7977  data: 2.1318  max mem: 57114
Epoch: [153]  [ 10/156]  eta: 0:02:13  lr: 0.004311  min_lr: 0.004311  loss: 3.1404 (2.9449)  weight_decay: 0.0500 (0.0500)  time: 0.9177  data: 0.1941  max mem: 57114
Epoch: [153]  [ 20/156]  eta: 0:01:52  lr: 0.004309  min_lr: 0.004309  loss: 3.2341 (3.0885)  weight_decay: 0.0500 (0.0500)  time: 0.7276  data: 0.0003  max mem: 57114
Epoch: [153]  [ 30/156]  eta: 0:01:39  lr: 0.004306  min_lr: 0.004306  loss: 3.2607 (3.1011)  weight_decay: 0.0500 (0.0500)  time: 0.7210  data: 0.0004  max mem: 57114
Epoch: [153]  [ 40/156]  eta: 0:01:30  lr: 0.004303  min_lr: 0.004303  loss: 3.1171 (3.0996)  weight_decay: 0.0500 (0.0500)  time: 0.7273  data: 0.0004  max mem: 57114
Epoch: [153]  [ 50/156]  eta: 0:01:20  lr: 0.004300  min_lr: 0.004300  loss: 3.1171 (3.0975)  weight_decay: 0.0500 (0.0500)  time: 0.7214  data: 0.0004  max mem: 57114
Epoch: [153]  [ 60/156]  eta: 0:01:12  lr: 0.004297  min_lr: 0.004297  loss: 3.1782 (3.1016)  weight_decay: 0.0500 (0.0500)  time: 0.7146  data: 0.0004  max mem: 57114
Epoch: [153]  [ 70/156]  eta: 0:01:04  lr: 0.004294  min_lr: 0.004294  loss: 3.2102 (3.1211)  weight_decay: 0.0500 (0.0500)  time: 0.7192  data: 0.0004  max mem: 57114
Epoch: [153]  [ 80/156]  eta: 0:00:56  lr: 0.004291  min_lr: 0.004291  loss: 3.2102 (3.1154)  weight_decay: 0.0500 (0.0500)  time: 0.7065  data: 0.0004  max mem: 57114
Epoch: [153]  [ 90/156]  eta: 0:00:48  lr: 0.004288  min_lr: 0.004288  loss: 3.1995 (3.1313)  weight_decay: 0.0500 (0.0500)  time: 0.6980  data: 0.0004  max mem: 57114
Epoch: [153]  [100/156]  eta: 0:00:41  lr: 0.004286  min_lr: 0.004286  loss: 3.2563 (3.1444)  weight_decay: 0.0500 (0.0500)  time: 0.7062  data: 0.0004  max mem: 57114
Epoch: [153]  [110/156]  eta: 0:00:33  lr: 0.004283  min_lr: 0.004283  loss: 3.2032 (3.1316)  weight_decay: 0.0500 (0.0500)  time: 0.7103  data: 0.0004  max mem: 57114
Epoch: [153]  [120/156]  eta: 0:00:26  lr: 0.004280  min_lr: 0.004280  loss: 3.2032 (3.1356)  weight_decay: 0.0500 (0.0500)  time: 0.7070  data: 0.0004  max mem: 57114
Epoch: [153]  [130/156]  eta: 0:00:18  lr: 0.004277  min_lr: 0.004277  loss: 3.1366 (3.1288)  weight_decay: 0.0500 (0.0500)  time: 0.6982  data: 0.0010  max mem: 57114
Epoch: [153]  [140/156]  eta: 0:00:11  lr: 0.004274  min_lr: 0.004274  loss: 3.2972 (3.1467)  weight_decay: 0.0500 (0.0500)  time: 0.6874  data: 0.0008  max mem: 57114
Epoch: [153]  [150/156]  eta: 0:00:04  lr: 0.004271  min_lr: 0.004271  loss: 3.3365 (3.1511)  weight_decay: 0.0500 (0.0500)  time: 0.6857  data: 0.0001  max mem: 57114
Epoch: [153]  [155/156]  eta: 0:00:00  lr: 0.004270  min_lr: 0.004270  loss: 3.3018 (3.1457)  weight_decay: 0.0500 (0.0500)  time: 0.6853  data: 0.0001  max mem: 57114
Epoch: [153] Total time: 0:01:52 (0.7238 s / it)
Averaged stats: lr: 0.004270  min_lr: 0.004270  loss: 3.3018 (3.0928)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 0.8714 (0.8714)  acc1: 87.5000 (87.5000)  acc5: 97.3958 (97.3958)  time: 2.2330  data: 1.9767  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9757 (0.9677)  acc1: 80.2083 (80.4348)  acc5: 97.3958 (95.5243)  time: 0.5973  data: 0.3954  max mem: 57114
Test: Total time: 0:00:03 (0.6187 s / it)
* Acc@1 79.232 Acc@5 95.009 loss 1.083
Accuracy of the model on the 50000 test images: 79.2%
Max accuracy: 79.64%
Test:  [0/5]  eta: 0:00:11  loss: 0.8993 (0.8993)  acc1: 78.1250 (78.1250)  acc5: 93.2292 (93.2292)  time: 2.2856  data: 2.0421  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1235 (1.1171)  acc1: 71.8750 (70.2046)  acc5: 91.6667 (88.3632)  time: 0.6078  data: 0.4085  max mem: 57114
Test: Total time: 0:00:03 (0.6241 s / it)
* Acc@1 67.056 Acc@5 88.045 loss 1.345
Accuracy of the model EMA on 50000 test images: 67.1%
Max EMA accuracy: 67.06%
Epoch: [154]  [  0/156]  eta: 0:06:38  lr: 0.004270  min_lr: 0.004270  loss: 3.5936 (3.5936)  weight_decay: 0.0500 (0.0500)  time: 2.5569  data: 1.9073  max mem: 57114
Epoch: [154]  [ 10/156]  eta: 0:02:08  lr: 0.004267  min_lr: 0.004267  loss: 2.6284 (2.8113)  weight_decay: 0.0500 (0.0500)  time: 0.8773  data: 0.1737  max mem: 57114
Epoch: [154]  [ 20/156]  eta: 0:01:50  lr: 0.004264  min_lr: 0.004264  loss: 2.9721 (2.9647)  weight_decay: 0.0500 (0.0500)  time: 0.7266  data: 0.0003  max mem: 57114
Epoch: [154]  [ 30/156]  eta: 0:01:39  lr: 0.004261  min_lr: 0.004261  loss: 3.2273 (3.0530)  weight_decay: 0.0500 (0.0500)  time: 0.7444  data: 0.0003  max mem: 57114
Epoch: [154]  [ 40/156]  eta: 0:01:29  lr: 0.004258  min_lr: 0.004258  loss: 3.2273 (3.0192)  weight_decay: 0.0500 (0.0500)  time: 0.7327  data: 0.0004  max mem: 57114
Epoch: [154]  [ 50/156]  eta: 0:01:20  lr: 0.004255  min_lr: 0.004255  loss: 2.9240 (3.0215)  weight_decay: 0.0500 (0.0500)  time: 0.7199  data: 0.0004  max mem: 57114
Epoch: [154]  [ 60/156]  eta: 0:01:12  lr: 0.004252  min_lr: 0.004252  loss: 2.9736 (3.0166)  weight_decay: 0.0500 (0.0500)  time: 0.7233  data: 0.0004  max mem: 57114
Epoch: [154]  [ 70/156]  eta: 0:01:04  lr: 0.004249  min_lr: 0.004249  loss: 3.2635 (3.0681)  weight_decay: 0.0500 (0.0500)  time: 0.7203  data: 0.0004  max mem: 57114
Epoch: [154]  [ 80/156]  eta: 0:00:56  lr: 0.004247  min_lr: 0.004247  loss: 3.2402 (3.0760)  weight_decay: 0.0500 (0.0500)  time: 0.7139  data: 0.0004  max mem: 57114
Epoch: [154]  [ 90/156]  eta: 0:00:49  lr: 0.004244  min_lr: 0.004244  loss: 3.1331 (3.0772)  weight_decay: 0.0500 (0.0500)  time: 0.7211  data: 0.0004  max mem: 57114
Epoch: [154]  [100/156]  eta: 0:00:41  lr: 0.004241  min_lr: 0.004241  loss: 3.1242 (3.0573)  weight_decay: 0.0500 (0.0500)  time: 0.7242  data: 0.0003  max mem: 57114
Epoch: [154]  [110/156]  eta: 0:00:33  lr: 0.004238  min_lr: 0.004238  loss: 3.1749 (3.0674)  weight_decay: 0.0500 (0.0500)  time: 0.7121  data: 0.0004  max mem: 57114
Epoch: [154]  [120/156]  eta: 0:00:26  lr: 0.004235  min_lr: 0.004235  loss: 3.1842 (3.0806)  weight_decay: 0.0500 (0.0500)  time: 0.7035  data: 0.0004  max mem: 57114
Epoch: [154]  [130/156]  eta: 0:00:19  lr: 0.004232  min_lr: 0.004232  loss: 3.2143 (3.0834)  weight_decay: 0.0500 (0.0500)  time: 0.7039  data: 0.0009  max mem: 57114
Epoch: [154]  [140/156]  eta: 0:00:11  lr: 0.004229  min_lr: 0.004229  loss: 3.3321 (3.1072)  weight_decay: 0.0500 (0.0500)  time: 0.6968  data: 0.0008  max mem: 57114
Epoch: [154]  [150/156]  eta: 0:00:04  lr: 0.004226  min_lr: 0.004226  loss: 3.2878 (3.0969)  weight_decay: 0.0500 (0.0500)  time: 0.6877  data: 0.0001  max mem: 57114
Epoch: [154]  [155/156]  eta: 0:00:00  lr: 0.004225  min_lr: 0.004225  loss: 3.2372 (3.1024)  weight_decay: 0.0500 (0.0500)  time: 0.6884  data: 0.0001  max mem: 57114
Epoch: [154] Total time: 0:01:53 (0.7286 s / it)
Averaged stats: lr: 0.004225  min_lr: 0.004225  loss: 3.2372 (3.0908)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8690 (0.8690)  acc1: 85.9375 (85.9375)  acc5: 97.9167 (97.9167)  time: 2.0550  data: 1.7991  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9770 (0.9445)  acc1: 81.7708 (80.0512)  acc5: 96.8750 (95.2685)  time: 0.5617  data: 0.3599  max mem: 57114
Test: Total time: 0:00:02 (0.5835 s / it)
* Acc@1 79.168 Acc@5 94.905 loss 1.069
Accuracy of the model on the 50000 test images: 79.2%
Max accuracy: 79.64%
Test:  [0/5]  eta: 0:00:11  loss: 0.8838 (0.8838)  acc1: 78.6458 (78.6458)  acc5: 93.2292 (93.2292)  time: 2.2843  data: 2.0408  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0992 (1.0954)  acc1: 71.8750 (70.7161)  acc5: 91.6667 (88.6189)  time: 0.6075  data: 0.4082  max mem: 57114
Test: Total time: 0:00:03 (0.6222 s / it)
* Acc@1 67.539 Acc@5 88.405 loss 1.321
Accuracy of the model EMA on 50000 test images: 67.5%
Max EMA accuracy: 67.54%
Epoch: [155]  [  0/156]  eta: 0:08:35  lr: 0.004225  min_lr: 0.004225  loss: 3.3469 (3.3469)  weight_decay: 0.0500 (0.0500)  time: 3.3022  data: 2.6465  max mem: 57114
Epoch: [155]  [ 10/156]  eta: 0:02:20  lr: 0.004222  min_lr: 0.004222  loss: 3.1917 (3.1237)  weight_decay: 0.0500 (0.0500)  time: 0.9634  data: 0.2409  max mem: 57114
Epoch: [155]  [ 20/156]  eta: 0:01:55  lr: 0.004219  min_lr: 0.004219  loss: 3.2364 (3.1850)  weight_decay: 0.0500 (0.0500)  time: 0.7245  data: 0.0003  max mem: 57114
Epoch: [155]  [ 30/156]  eta: 0:01:41  lr: 0.004216  min_lr: 0.004216  loss: 3.2364 (3.1401)  weight_decay: 0.0500 (0.0500)  time: 0.7226  data: 0.0003  max mem: 57114
Epoch: [155]  [ 40/156]  eta: 0:01:31  lr: 0.004213  min_lr: 0.004213  loss: 2.9351 (3.0764)  weight_decay: 0.0500 (0.0500)  time: 0.7250  data: 0.0003  max mem: 57114
Epoch: [155]  [ 50/156]  eta: 0:01:22  lr: 0.004210  min_lr: 0.004210  loss: 3.0090 (3.0571)  weight_decay: 0.0500 (0.0500)  time: 0.7271  data: 0.0004  max mem: 57114
Epoch: [155]  [ 60/156]  eta: 0:01:13  lr: 0.004208  min_lr: 0.004208  loss: 3.0489 (3.0349)  weight_decay: 0.0500 (0.0500)  time: 0.7253  data: 0.0004  max mem: 57114
Epoch: [155]  [ 70/156]  eta: 0:01:05  lr: 0.004205  min_lr: 0.004205  loss: 3.1368 (3.0409)  weight_decay: 0.0500 (0.0500)  time: 0.7254  data: 0.0004  max mem: 57114
Epoch: [155]  [ 80/156]  eta: 0:00:57  lr: 0.004202  min_lr: 0.004202  loss: 3.1368 (3.0455)  weight_decay: 0.0500 (0.0500)  time: 0.7194  data: 0.0003  max mem: 57114
Epoch: [155]  [ 90/156]  eta: 0:00:49  lr: 0.004199  min_lr: 0.004199  loss: 3.1672 (3.0663)  weight_decay: 0.0500 (0.0500)  time: 0.7107  data: 0.0004  max mem: 57114
Epoch: [155]  [100/156]  eta: 0:00:41  lr: 0.004196  min_lr: 0.004196  loss: 3.2385 (3.0869)  weight_decay: 0.0500 (0.0500)  time: 0.7098  data: 0.0004  max mem: 57114
Epoch: [155]  [110/156]  eta: 0:00:34  lr: 0.004193  min_lr: 0.004193  loss: 3.1082 (3.0725)  weight_decay: 0.0500 (0.0500)  time: 0.7069  data: 0.0004  max mem: 57114
Epoch: [155]  [120/156]  eta: 0:00:26  lr: 0.004190  min_lr: 0.004190  loss: 3.0384 (3.0810)  weight_decay: 0.0500 (0.0500)  time: 0.7085  data: 0.0004  max mem: 57114
Epoch: [155]  [130/156]  eta: 0:00:19  lr: 0.004187  min_lr: 0.004187  loss: 3.3302 (3.0979)  weight_decay: 0.0500 (0.0500)  time: 0.7068  data: 0.0009  max mem: 57114
Epoch: [155]  [140/156]  eta: 0:00:11  lr: 0.004185  min_lr: 0.004185  loss: 3.1854 (3.1021)  weight_decay: 0.0500 (0.0500)  time: 0.6920  data: 0.0007  max mem: 57114
Epoch: [155]  [150/156]  eta: 0:00:04  lr: 0.004182  min_lr: 0.004182  loss: 3.1536 (3.1081)  weight_decay: 0.0500 (0.0500)  time: 0.6824  data: 0.0001  max mem: 57114
Epoch: [155]  [155/156]  eta: 0:00:00  lr: 0.004180  min_lr: 0.004180  loss: 3.1536 (3.1032)  weight_decay: 0.0500 (0.0500)  time: 0.6769  data: 0.0001  max mem: 57114
Epoch: [155] Total time: 0:01:53 (0.7303 s / it)
Averaged stats: lr: 0.004180  min_lr: 0.004180  loss: 3.1536 (3.0958)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8031 (0.8031)  acc1: 85.4167 (85.4167)  acc5: 98.4375 (98.4375)  time: 2.0900  data: 1.8335  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9218 (0.9377)  acc1: 83.3333 (80.8184)  acc5: 96.3542 (95.3964)  time: 0.5688  data: 0.3668  max mem: 57114
Test: Total time: 0:00:02 (0.5906 s / it)
* Acc@1 79.206 Acc@5 94.903 loss 1.038
Accuracy of the model on the 50000 test images: 79.2%
Max accuracy: 79.64%
Test:  [0/5]  eta: 0:00:11  loss: 0.8689 (0.8689)  acc1: 78.6458 (78.6458)  acc5: 93.2292 (93.2292)  time: 2.2752  data: 2.0315  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0765 (1.0746)  acc1: 71.8750 (70.9719)  acc5: 91.6667 (88.6189)  time: 0.6062  data: 0.4065  max mem: 57114
Test: Total time: 0:00:03 (0.6187 s / it)
* Acc@1 68.051 Acc@5 88.673 loss 1.297
Accuracy of the model EMA on 50000 test images: 68.1%
Max EMA accuracy: 68.05%
Epoch: [156]  [  0/156]  eta: 0:07:27  lr: 0.004180  min_lr: 0.004180  loss: 1.9047 (1.9047)  weight_decay: 0.0500 (0.0500)  time: 2.8707  data: 2.2237  max mem: 57114
Epoch: [156]  [ 10/156]  eta: 0:02:11  lr: 0.004177  min_lr: 0.004177  loss: 3.3082 (3.1220)  weight_decay: 0.0500 (0.0500)  time: 0.9025  data: 0.2024  max mem: 57114
Epoch: [156]  [ 20/156]  eta: 0:01:51  lr: 0.004174  min_lr: 0.004174  loss: 3.2832 (3.0555)  weight_decay: 0.0500 (0.0500)  time: 0.7161  data: 0.0003  max mem: 57114
Epoch: [156]  [ 30/156]  eta: 0:01:39  lr: 0.004171  min_lr: 0.004171  loss: 3.0317 (3.0717)  weight_decay: 0.0500 (0.0500)  time: 0.7319  data: 0.0003  max mem: 57114
Epoch: [156]  [ 40/156]  eta: 0:01:29  lr: 0.004168  min_lr: 0.004168  loss: 3.0317 (3.0753)  weight_decay: 0.0500 (0.0500)  time: 0.7285  data: 0.0003  max mem: 57114
Epoch: [156]  [ 50/156]  eta: 0:01:21  lr: 0.004166  min_lr: 0.004166  loss: 2.8185 (2.9944)  weight_decay: 0.0500 (0.0500)  time: 0.7284  data: 0.0004  max mem: 57114
Epoch: [156]  [ 60/156]  eta: 0:01:12  lr: 0.004163  min_lr: 0.004163  loss: 2.9229 (3.0319)  weight_decay: 0.0500 (0.0500)  time: 0.7289  data: 0.0005  max mem: 57114
Epoch: [156]  [ 70/156]  eta: 0:01:04  lr: 0.004160  min_lr: 0.004160  loss: 3.3315 (3.0402)  weight_decay: 0.0500 (0.0500)  time: 0.7220  data: 0.0005  max mem: 57114
Epoch: [156]  [ 80/156]  eta: 0:00:56  lr: 0.004157  min_lr: 0.004157  loss: 3.3122 (3.0505)  weight_decay: 0.0500 (0.0500)  time: 0.7108  data: 0.0004  max mem: 57114
Epoch: [156]  [ 90/156]  eta: 0:00:49  lr: 0.004154  min_lr: 0.004154  loss: 3.1431 (3.0283)  weight_decay: 0.0500 (0.0500)  time: 0.7031  data: 0.0003  max mem: 57114
Epoch: [156]  [100/156]  eta: 0:00:41  lr: 0.004151  min_lr: 0.004151  loss: 3.0823 (3.0294)  weight_decay: 0.0500 (0.0500)  time: 0.7106  data: 0.0003  max mem: 57114
Epoch: [156]  [110/156]  eta: 0:00:34  lr: 0.004148  min_lr: 0.004148  loss: 2.9590 (3.0173)  weight_decay: 0.0500 (0.0500)  time: 0.7207  data: 0.0003  max mem: 57114
Epoch: [156]  [120/156]  eta: 0:00:26  lr: 0.004145  min_lr: 0.004145  loss: 2.8348 (3.0034)  weight_decay: 0.0500 (0.0500)  time: 0.7136  data: 0.0004  max mem: 57114
Epoch: [156]  [130/156]  eta: 0:00:19  lr: 0.004143  min_lr: 0.004143  loss: 3.0060 (3.0135)  weight_decay: 0.0500 (0.0500)  time: 0.6983  data: 0.0008  max mem: 57114
Epoch: [156]  [140/156]  eta: 0:00:11  lr: 0.004140  min_lr: 0.004140  loss: 3.1855 (3.0336)  weight_decay: 0.0500 (0.0500)  time: 0.6955  data: 0.0007  max mem: 57114
Epoch: [156]  [150/156]  eta: 0:00:04  lr: 0.004137  min_lr: 0.004137  loss: 3.0988 (3.0272)  weight_decay: 0.0500 (0.0500)  time: 0.6896  data: 0.0001  max mem: 57114
Epoch: [156]  [155/156]  eta: 0:00:00  lr: 0.004135  min_lr: 0.004135  loss: 3.0869 (3.0319)  weight_decay: 0.0500 (0.0500)  time: 0.6842  data: 0.0001  max mem: 57114
Epoch: [156] Total time: 0:01:53 (0.7278 s / it)
Averaged stats: lr: 0.004135  min_lr: 0.004135  loss: 3.0869 (3.0856)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8460 (0.8460)  acc1: 86.9792 (86.9792)  acc5: 98.4375 (98.4375)  time: 2.0217  data: 1.7659  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9228 (0.9407)  acc1: 81.7708 (81.5857)  acc5: 96.3542 (95.2685)  time: 0.5550  data: 0.3532  max mem: 57114
Test: Total time: 0:00:02 (0.5853 s / it)
* Acc@1 79.715 Acc@5 94.935 loss 1.016
Accuracy of the model on the 50000 test images: 79.7%
Max accuracy: 79.72%
Test:  [0/5]  eta: 0:00:09  loss: 0.8545 (0.8545)  acc1: 78.6458 (78.6458)  acc5: 93.2292 (93.2292)  time: 1.9949  data: 1.7514  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0555 (1.0547)  acc1: 71.8750 (71.2276)  acc5: 91.6667 (88.8747)  time: 0.5496  data: 0.3504  max mem: 57114
Test: Total time: 0:00:02 (0.5611 s / it)
* Acc@1 68.452 Acc@5 88.935 loss 1.275
Accuracy of the model EMA on 50000 test images: 68.5%
Max EMA accuracy: 68.45%
Epoch: [157]  [  0/156]  eta: 0:06:43  lr: 0.004135  min_lr: 0.004135  loss: 3.2011 (3.2011)  weight_decay: 0.0500 (0.0500)  time: 2.5853  data: 1.9332  max mem: 57114
Epoch: [157]  [ 10/156]  eta: 0:02:09  lr: 0.004132  min_lr: 0.004132  loss: 3.1985 (2.9729)  weight_decay: 0.0500 (0.0500)  time: 0.8882  data: 0.1760  max mem: 57114
Epoch: [157]  [ 20/156]  eta: 0:01:52  lr: 0.004129  min_lr: 0.004129  loss: 3.1985 (3.0148)  weight_decay: 0.0500 (0.0500)  time: 0.7403  data: 0.0004  max mem: 57114
Epoch: [157]  [ 30/156]  eta: 0:01:40  lr: 0.004126  min_lr: 0.004126  loss: 3.2139 (3.0333)  weight_decay: 0.0500 (0.0500)  time: 0.7474  data: 0.0004  max mem: 57114
Epoch: [157]  [ 40/156]  eta: 0:01:31  lr: 0.004124  min_lr: 0.004124  loss: 3.2469 (3.0764)  weight_decay: 0.0500 (0.0500)  time: 0.7521  data: 0.0004  max mem: 57114
Epoch: [157]  [ 50/156]  eta: 0:01:22  lr: 0.004121  min_lr: 0.004121  loss: 3.3261 (3.1154)  weight_decay: 0.0500 (0.0500)  time: 0.7604  data: 0.0004  max mem: 57114
Epoch: [157]  [ 60/156]  eta: 0:01:14  lr: 0.004118  min_lr: 0.004118  loss: 3.3151 (3.1304)  weight_decay: 0.0500 (0.0500)  time: 0.7364  data: 0.0004  max mem: 57114
Epoch: [157]  [ 70/156]  eta: 0:01:05  lr: 0.004115  min_lr: 0.004115  loss: 3.2128 (3.0953)  weight_decay: 0.0500 (0.0500)  time: 0.7184  data: 0.0004  max mem: 57114
Epoch: [157]  [ 80/156]  eta: 0:00:57  lr: 0.004112  min_lr: 0.004112  loss: 3.2128 (3.1112)  weight_decay: 0.0500 (0.0500)  time: 0.7158  data: 0.0004  max mem: 57114
Epoch: [157]  [ 90/156]  eta: 0:00:49  lr: 0.004109  min_lr: 0.004109  loss: 3.2555 (3.1122)  weight_decay: 0.0500 (0.0500)  time: 0.7105  data: 0.0004  max mem: 57114
Epoch: [157]  [100/156]  eta: 0:00:41  lr: 0.004106  min_lr: 0.004106  loss: 3.2320 (3.1149)  weight_decay: 0.0500 (0.0500)  time: 0.7120  data: 0.0004  max mem: 57114
Epoch: [157]  [110/156]  eta: 0:00:34  lr: 0.004103  min_lr: 0.004103  loss: 3.1840 (3.1123)  weight_decay: 0.0500 (0.0500)  time: 0.7165  data: 0.0004  max mem: 57114
Epoch: [157]  [120/156]  eta: 0:00:26  lr: 0.004101  min_lr: 0.004101  loss: 3.0093 (3.0939)  weight_decay: 0.0500 (0.0500)  time: 0.7107  data: 0.0004  max mem: 57114
Epoch: [157]  [130/156]  eta: 0:00:19  lr: 0.004098  min_lr: 0.004098  loss: 2.9843 (3.1014)  weight_decay: 0.0500 (0.0500)  time: 0.7021  data: 0.0009  max mem: 57114
Epoch: [157]  [140/156]  eta: 0:00:11  lr: 0.004095  min_lr: 0.004095  loss: 3.1430 (3.1058)  weight_decay: 0.0500 (0.0500)  time: 0.6886  data: 0.0007  max mem: 57114
Epoch: [157]  [150/156]  eta: 0:00:04  lr: 0.004092  min_lr: 0.004092  loss: 3.1430 (3.1043)  weight_decay: 0.0500 (0.0500)  time: 0.6848  data: 0.0001  max mem: 57114
Epoch: [157]  [155/156]  eta: 0:00:00  lr: 0.004091  min_lr: 0.004091  loss: 3.2528 (3.1163)  weight_decay: 0.0500 (0.0500)  time: 0.6845  data: 0.0001  max mem: 57114
Epoch: [157] Total time: 0:01:54 (0.7328 s / it)
Averaged stats: lr: 0.004091  min_lr: 0.004091  loss: 3.2528 (3.0809)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9783 (0.9783)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.0465  data: 1.7906  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0878 (1.0879)  acc1: 83.3333 (79.4118)  acc5: 96.3542 (95.0128)  time: 0.5600  data: 0.3582  max mem: 57114
Test: Total time: 0:00:02 (0.5827 s / it)
* Acc@1 79.044 Acc@5 94.871 loss 1.169
Accuracy of the model on the 50000 test images: 79.0%
Max accuracy: 79.72%
Test:  [0/5]  eta: 0:00:11  loss: 0.8422 (0.8422)  acc1: 78.6458 (78.6458)  acc5: 93.2292 (93.2292)  time: 2.2824  data: 2.0389  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0368 (1.0373)  acc1: 73.4375 (71.8670)  acc5: 91.6667 (89.1304)  time: 0.6071  data: 0.4079  max mem: 57114
Test: Total time: 0:00:03 (0.6216 s / it)
* Acc@1 68.882 Acc@5 89.166 loss 1.254
Accuracy of the model EMA on 50000 test images: 68.9%
Max EMA accuracy: 68.88%
Epoch: [158]  [  0/156]  eta: 0:07:56  lr: 0.004090  min_lr: 0.004090  loss: 2.7943 (2.7943)  weight_decay: 0.0500 (0.0500)  time: 3.0563  data: 2.4023  max mem: 57114
Epoch: [158]  [ 10/156]  eta: 0:02:20  lr: 0.004087  min_lr: 0.004087  loss: 3.5150 (3.2083)  weight_decay: 0.0500 (0.0500)  time: 0.9652  data: 0.2186  max mem: 57114
Epoch: [158]  [ 20/156]  eta: 0:01:55  lr: 0.004084  min_lr: 0.004084  loss: 3.3945 (3.1415)  weight_decay: 0.0500 (0.0500)  time: 0.7378  data: 0.0003  max mem: 57114
Epoch: [158]  [ 30/156]  eta: 0:01:41  lr: 0.004082  min_lr: 0.004082  loss: 3.0699 (3.1278)  weight_decay: 0.0500 (0.0500)  time: 0.7179  data: 0.0004  max mem: 57114
Epoch: [158]  [ 40/156]  eta: 0:01:30  lr: 0.004079  min_lr: 0.004079  loss: 3.0420 (3.1099)  weight_decay: 0.0500 (0.0500)  time: 0.7171  data: 0.0004  max mem: 57114
Epoch: [158]  [ 50/156]  eta: 0:01:21  lr: 0.004076  min_lr: 0.004076  loss: 3.1128 (3.1209)  weight_decay: 0.0500 (0.0500)  time: 0.7198  data: 0.0004  max mem: 57114
Epoch: [158]  [ 60/156]  eta: 0:01:13  lr: 0.004073  min_lr: 0.004073  loss: 3.1128 (3.1059)  weight_decay: 0.0500 (0.0500)  time: 0.7304  data: 0.0004  max mem: 57114
Epoch: [158]  [ 70/156]  eta: 0:01:05  lr: 0.004070  min_lr: 0.004070  loss: 3.1702 (3.1242)  weight_decay: 0.0500 (0.0500)  time: 0.7225  data: 0.0004  max mem: 57114
Epoch: [158]  [ 80/156]  eta: 0:00:57  lr: 0.004067  min_lr: 0.004067  loss: 3.1702 (3.1300)  weight_decay: 0.0500 (0.0500)  time: 0.7110  data: 0.0004  max mem: 57114
Epoch: [158]  [ 90/156]  eta: 0:00:49  lr: 0.004064  min_lr: 0.004064  loss: 3.3005 (3.1668)  weight_decay: 0.0500 (0.0500)  time: 0.7164  data: 0.0004  max mem: 57114
Epoch: [158]  [100/156]  eta: 0:00:41  lr: 0.004061  min_lr: 0.004061  loss: 3.4865 (3.1823)  weight_decay: 0.0500 (0.0500)  time: 0.7126  data: 0.0003  max mem: 57114
Epoch: [158]  [110/156]  eta: 0:00:34  lr: 0.004059  min_lr: 0.004059  loss: 3.2573 (3.1837)  weight_decay: 0.0500 (0.0500)  time: 0.7043  data: 0.0004  max mem: 57114
Epoch: [158]  [120/156]  eta: 0:00:26  lr: 0.004056  min_lr: 0.004056  loss: 3.1345 (3.1813)  weight_decay: 0.0500 (0.0500)  time: 0.7024  data: 0.0004  max mem: 57114
Epoch: [158]  [130/156]  eta: 0:00:19  lr: 0.004053  min_lr: 0.004053  loss: 3.2044 (3.1910)  weight_decay: 0.0500 (0.0500)  time: 0.6999  data: 0.0008  max mem: 57114
Epoch: [158]  [140/156]  eta: 0:00:11  lr: 0.004050  min_lr: 0.004050  loss: 2.9071 (3.1428)  weight_decay: 0.0500 (0.0500)  time: 0.6912  data: 0.0007  max mem: 57114
Epoch: [158]  [150/156]  eta: 0:00:04  lr: 0.004047  min_lr: 0.004047  loss: 2.7957 (3.1356)  weight_decay: 0.0500 (0.0500)  time: 0.6883  data: 0.0001  max mem: 57114
Epoch: [158]  [155/156]  eta: 0:00:00  lr: 0.004046  min_lr: 0.004046  loss: 2.9930 (3.1321)  weight_decay: 0.0500 (0.0500)  time: 0.6894  data: 0.0001  max mem: 57114
Epoch: [158] Total time: 0:01:53 (0.7293 s / it)
Averaged stats: lr: 0.004046  min_lr: 0.004046  loss: 2.9930 (3.0811)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8046 (0.8046)  acc1: 81.7708 (81.7708)  acc5: 98.4375 (98.4375)  time: 2.0403  data: 1.7844  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8999 (0.8796)  acc1: 81.7708 (79.6675)  acc5: 96.3542 (95.1407)  time: 0.5587  data: 0.3569  max mem: 57114
Test: Total time: 0:00:02 (0.5812 s / it)
* Acc@1 79.574 Acc@5 95.079 loss 0.963
Accuracy of the model on the 50000 test images: 79.6%
Max accuracy: 79.72%
Test:  [0/5]  eta: 0:00:11  loss: 0.8303 (0.8303)  acc1: 78.6458 (78.6458)  acc5: 93.7500 (93.7500)  time: 2.3338  data: 2.0903  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0191 (1.0200)  acc1: 74.4792 (72.2506)  acc5: 91.6667 (89.5141)  time: 0.6174  data: 0.4181  max mem: 57114
Test: Total time: 0:00:03 (0.6312 s / it)
* Acc@1 69.238 Acc@5 89.362 loss 1.235
Accuracy of the model EMA on 50000 test images: 69.2%
Max EMA accuracy: 69.24%
Epoch: [159]  [  0/156]  eta: 0:05:37  lr: 0.004045  min_lr: 0.004045  loss: 3.0553 (3.0553)  weight_decay: 0.0500 (0.0500)  time: 2.1643  data: 1.5121  max mem: 57114
Epoch: [159]  [ 10/156]  eta: 0:02:06  lr: 0.004042  min_lr: 0.004042  loss: 3.1676 (3.0739)  weight_decay: 0.0500 (0.0500)  time: 0.8684  data: 0.1377  max mem: 57114
Epoch: [159]  [ 20/156]  eta: 0:01:48  lr: 0.004040  min_lr: 0.004040  loss: 3.2425 (3.1086)  weight_decay: 0.0500 (0.0500)  time: 0.7317  data: 0.0003  max mem: 57114
Epoch: [159]  [ 30/156]  eta: 0:01:37  lr: 0.004037  min_lr: 0.004037  loss: 2.9713 (2.9827)  weight_decay: 0.0500 (0.0500)  time: 0.7231  data: 0.0003  max mem: 57114
Epoch: [159]  [ 40/156]  eta: 0:01:28  lr: 0.004034  min_lr: 0.004034  loss: 2.8845 (2.9800)  weight_decay: 0.0500 (0.0500)  time: 0.7313  data: 0.0003  max mem: 57114
Epoch: [159]  [ 50/156]  eta: 0:01:20  lr: 0.004031  min_lr: 0.004031  loss: 3.1839 (3.0212)  weight_decay: 0.0500 (0.0500)  time: 0.7361  data: 0.0004  max mem: 57114
Epoch: [159]  [ 60/156]  eta: 0:01:12  lr: 0.004028  min_lr: 0.004028  loss: 3.2642 (3.0782)  weight_decay: 0.0500 (0.0500)  time: 0.7264  data: 0.0004  max mem: 57114
Epoch: [159]  [ 70/156]  eta: 0:01:04  lr: 0.004025  min_lr: 0.004025  loss: 3.3791 (3.0769)  weight_decay: 0.0500 (0.0500)  time: 0.7215  data: 0.0004  max mem: 57114
Epoch: [159]  [ 80/156]  eta: 0:00:56  lr: 0.004022  min_lr: 0.004022  loss: 2.9613 (3.0602)  weight_decay: 0.0500 (0.0500)  time: 0.7199  data: 0.0004  max mem: 57114
Epoch: [159]  [ 90/156]  eta: 0:00:48  lr: 0.004019  min_lr: 0.004019  loss: 3.1606 (3.0528)  weight_decay: 0.0500 (0.0500)  time: 0.7191  data: 0.0003  max mem: 57114
Epoch: [159]  [100/156]  eta: 0:00:41  lr: 0.004017  min_lr: 0.004017  loss: 3.1762 (3.0345)  weight_decay: 0.0500 (0.0500)  time: 0.7057  data: 0.0003  max mem: 57114
Epoch: [159]  [110/156]  eta: 0:00:33  lr: 0.004014  min_lr: 0.004014  loss: 3.0089 (3.0347)  weight_decay: 0.0500 (0.0500)  time: 0.7025  data: 0.0004  max mem: 57114
Epoch: [159]  [120/156]  eta: 0:00:26  lr: 0.004011  min_lr: 0.004011  loss: 3.0297 (3.0401)  weight_decay: 0.0500 (0.0500)  time: 0.7098  data: 0.0003  max mem: 57114
Epoch: [159]  [130/156]  eta: 0:00:18  lr: 0.004008  min_lr: 0.004008  loss: 3.0680 (3.0297)  weight_decay: 0.0500 (0.0500)  time: 0.7018  data: 0.0008  max mem: 57114
Epoch: [159]  [140/156]  eta: 0:00:11  lr: 0.004005  min_lr: 0.004005  loss: 3.2989 (3.0513)  weight_decay: 0.0500 (0.0500)  time: 0.6870  data: 0.0007  max mem: 57114
Epoch: [159]  [150/156]  eta: 0:00:04  lr: 0.004002  min_lr: 0.004002  loss: 3.3260 (3.0600)  weight_decay: 0.0500 (0.0500)  time: 0.6846  data: 0.0001  max mem: 57114
Epoch: [159]  [155/156]  eta: 0:00:00  lr: 0.004001  min_lr: 0.004001  loss: 3.3260 (3.0732)  weight_decay: 0.0500 (0.0500)  time: 0.6873  data: 0.0001  max mem: 57114
Epoch: [159] Total time: 0:01:53 (0.7252 s / it)
Averaged stats: lr: 0.004001  min_lr: 0.004001  loss: 3.3260 (3.0633)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 1.0064 (1.0064)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.0919  data: 1.8362  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1210 (1.0812)  acc1: 80.7292 (79.4118)  acc5: 97.3958 (96.0358)  time: 0.5691  data: 0.3673  max mem: 57114
Test: Total time: 0:00:02 (0.5895 s / it)
* Acc@1 79.328 Acc@5 95.031 loss 1.162
Accuracy of the model on the 50000 test images: 79.3%
Max accuracy: 79.72%
Test:  [0/5]  eta: 0:00:11  loss: 0.8191 (0.8191)  acc1: 79.1667 (79.1667)  acc5: 93.7500 (93.7500)  time: 2.3521  data: 2.1085  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0028 (1.0040)  acc1: 74.4792 (73.0179)  acc5: 91.6667 (89.6419)  time: 0.6210  data: 0.4218  max mem: 57114
Test: Total time: 0:00:03 (0.6342 s / it)
* Acc@1 69.561 Acc@5 89.558 loss 1.216
Accuracy of the model EMA on 50000 test images: 69.6%
Max EMA accuracy: 69.56%
Epoch: [160]  [  0/156]  eta: 0:06:38  lr: 0.004001  min_lr: 0.004001  loss: 2.9841 (2.9841)  weight_decay: 0.0500 (0.0500)  time: 2.5524  data: 1.9036  max mem: 57114
Epoch: [160]  [ 10/156]  eta: 0:02:10  lr: 0.003998  min_lr: 0.003998  loss: 3.2237 (3.0936)  weight_decay: 0.0500 (0.0500)  time: 0.8924  data: 0.1733  max mem: 57114
Epoch: [160]  [ 20/156]  eta: 0:01:49  lr: 0.003995  min_lr: 0.003995  loss: 3.2899 (3.2345)  weight_decay: 0.0500 (0.0500)  time: 0.7192  data: 0.0003  max mem: 57114
Epoch: [160]  [ 30/156]  eta: 0:01:38  lr: 0.003992  min_lr: 0.003992  loss: 3.1426 (3.0974)  weight_decay: 0.0500 (0.0500)  time: 0.7264  data: 0.0003  max mem: 57114
Epoch: [160]  [ 40/156]  eta: 0:01:29  lr: 0.003989  min_lr: 0.003989  loss: 3.1591 (3.1179)  weight_decay: 0.0500 (0.0500)  time: 0.7344  data: 0.0004  max mem: 57114
Epoch: [160]  [ 50/156]  eta: 0:01:20  lr: 0.003986  min_lr: 0.003986  loss: 3.2737 (3.1112)  weight_decay: 0.0500 (0.0500)  time: 0.7262  data: 0.0004  max mem: 57114
Epoch: [160]  [ 60/156]  eta: 0:01:12  lr: 0.003983  min_lr: 0.003983  loss: 3.1823 (3.0739)  weight_decay: 0.0500 (0.0500)  time: 0.7279  data: 0.0004  max mem: 57114
Epoch: [160]  [ 70/156]  eta: 0:01:04  lr: 0.003980  min_lr: 0.003980  loss: 3.2181 (3.0853)  weight_decay: 0.0500 (0.0500)  time: 0.7185  data: 0.0004  max mem: 57114
Epoch: [160]  [ 80/156]  eta: 0:00:56  lr: 0.003977  min_lr: 0.003977  loss: 3.3113 (3.1009)  weight_decay: 0.0500 (0.0500)  time: 0.7111  data: 0.0004  max mem: 57114
Epoch: [160]  [ 90/156]  eta: 0:00:48  lr: 0.003975  min_lr: 0.003975  loss: 3.1714 (3.0908)  weight_decay: 0.0500 (0.0500)  time: 0.7135  data: 0.0004  max mem: 57114
Epoch: [160]  [100/156]  eta: 0:00:41  lr: 0.003972  min_lr: 0.003972  loss: 2.9496 (3.0544)  weight_decay: 0.0500 (0.0500)  time: 0.7104  data: 0.0004  max mem: 57114
Epoch: [160]  [110/156]  eta: 0:00:33  lr: 0.003969  min_lr: 0.003969  loss: 2.9082 (3.0492)  weight_decay: 0.0500 (0.0500)  time: 0.7045  data: 0.0004  max mem: 57114
Epoch: [160]  [120/156]  eta: 0:00:26  lr: 0.003966  min_lr: 0.003966  loss: 3.0475 (3.0536)  weight_decay: 0.0500 (0.0500)  time: 0.7043  data: 0.0004  max mem: 57114
Epoch: [160]  [130/156]  eta: 0:00:18  lr: 0.003963  min_lr: 0.003963  loss: 3.1143 (3.0613)  weight_decay: 0.0500 (0.0500)  time: 0.7033  data: 0.0008  max mem: 57114
Epoch: [160]  [140/156]  eta: 0:00:11  lr: 0.003960  min_lr: 0.003960  loss: 3.1759 (3.0621)  weight_decay: 0.0500 (0.0500)  time: 0.6941  data: 0.0007  max mem: 57114
Epoch: [160]  [150/156]  eta: 0:00:04  lr: 0.003957  min_lr: 0.003957  loss: 3.0588 (3.0486)  weight_decay: 0.0500 (0.0500)  time: 0.6929  data: 0.0001  max mem: 57114
Epoch: [160]  [155/156]  eta: 0:00:00  lr: 0.003956  min_lr: 0.003956  loss: 2.8063 (3.0370)  weight_decay: 0.0500 (0.0500)  time: 0.6908  data: 0.0001  max mem: 57114
Epoch: [160] Total time: 0:01:53 (0.7258 s / it)
Averaged stats: lr: 0.003956  min_lr: 0.003956  loss: 2.8063 (3.0712)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6980 (0.6980)  acc1: 86.9792 (86.9792)  acc5: 97.9167 (97.9167)  time: 2.0455  data: 1.7899  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8502 (0.8395)  acc1: 82.2917 (80.1790)  acc5: 96.3542 (95.2685)  time: 0.5597  data: 0.3580  max mem: 57114
Test: Total time: 0:00:02 (0.5836 s / it)
* Acc@1 79.200 Acc@5 94.891 loss 0.947
Accuracy of the model on the 50000 test images: 79.2%
Max accuracy: 79.72%
Test:  [0/5]  eta: 0:00:11  loss: 0.8080 (0.8080)  acc1: 79.6875 (79.6875)  acc5: 94.2708 (94.2708)  time: 2.3083  data: 2.0647  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9878 (0.9885)  acc1: 74.4792 (73.2737)  acc5: 91.6667 (89.8977)  time: 0.6123  data: 0.4130  max mem: 57114
Test: Total time: 0:00:03 (0.6235 s / it)
* Acc@1 69.935 Acc@5 89.808 loss 1.199
Accuracy of the model EMA on 50000 test images: 69.9%
Max EMA accuracy: 69.93%
Epoch: [161]  [  0/156]  eta: 0:07:38  lr: 0.003956  min_lr: 0.003956  loss: 2.3420 (2.3420)  weight_decay: 0.0500 (0.0500)  time: 2.9361  data: 2.2914  max mem: 57114
Epoch: [161]  [ 10/156]  eta: 0:02:16  lr: 0.003953  min_lr: 0.003953  loss: 2.9392 (2.8553)  weight_decay: 0.0500 (0.0500)  time: 0.9316  data: 0.2086  max mem: 57114
Epoch: [161]  [ 20/156]  eta: 0:01:53  lr: 0.003950  min_lr: 0.003950  loss: 2.9507 (2.9074)  weight_decay: 0.0500 (0.0500)  time: 0.7307  data: 0.0003  max mem: 57114
Epoch: [161]  [ 30/156]  eta: 0:01:41  lr: 0.003947  min_lr: 0.003947  loss: 2.9768 (2.9454)  weight_decay: 0.0500 (0.0500)  time: 0.7349  data: 0.0003  max mem: 57114
Epoch: [161]  [ 40/156]  eta: 0:01:31  lr: 0.003944  min_lr: 0.003944  loss: 3.0421 (2.9443)  weight_decay: 0.0500 (0.0500)  time: 0.7356  data: 0.0004  max mem: 57114
Epoch: [161]  [ 50/156]  eta: 0:01:21  lr: 0.003941  min_lr: 0.003941  loss: 3.1201 (2.9938)  weight_decay: 0.0500 (0.0500)  time: 0.7171  data: 0.0004  max mem: 57114
Epoch: [161]  [ 60/156]  eta: 0:01:13  lr: 0.003938  min_lr: 0.003938  loss: 3.1969 (3.0111)  weight_decay: 0.0500 (0.0500)  time: 0.7175  data: 0.0004  max mem: 57114
Epoch: [161]  [ 70/156]  eta: 0:01:05  lr: 0.003935  min_lr: 0.003935  loss: 3.0572 (2.9842)  weight_decay: 0.0500 (0.0500)  time: 0.7208  data: 0.0004  max mem: 57114
Epoch: [161]  [ 80/156]  eta: 0:00:57  lr: 0.003933  min_lr: 0.003933  loss: 2.8922 (3.0018)  weight_decay: 0.0500 (0.0500)  time: 0.7119  data: 0.0004  max mem: 57114
Epoch: [161]  [ 90/156]  eta: 0:00:49  lr: 0.003930  min_lr: 0.003930  loss: 3.2160 (3.0339)  weight_decay: 0.0500 (0.0500)  time: 0.7133  data: 0.0004  max mem: 57114
Epoch: [161]  [100/156]  eta: 0:00:41  lr: 0.003927  min_lr: 0.003927  loss: 3.2566 (3.0597)  weight_decay: 0.0500 (0.0500)  time: 0.7056  data: 0.0004  max mem: 57114
Epoch: [161]  [110/156]  eta: 0:00:34  lr: 0.003924  min_lr: 0.003924  loss: 3.3022 (3.0808)  weight_decay: 0.0500 (0.0500)  time: 0.7081  data: 0.0003  max mem: 57114
Epoch: [161]  [120/156]  eta: 0:00:26  lr: 0.003921  min_lr: 0.003921  loss: 3.2310 (3.0907)  weight_decay: 0.0500 (0.0500)  time: 0.7105  data: 0.0004  max mem: 57114
Epoch: [161]  [130/156]  eta: 0:00:19  lr: 0.003918  min_lr: 0.003918  loss: 3.1533 (3.0819)  weight_decay: 0.0500 (0.0500)  time: 0.7005  data: 0.0009  max mem: 57114
Epoch: [161]  [140/156]  eta: 0:00:11  lr: 0.003915  min_lr: 0.003915  loss: 3.1168 (3.0959)  weight_decay: 0.0500 (0.0500)  time: 0.6931  data: 0.0007  max mem: 57114
Epoch: [161]  [150/156]  eta: 0:00:04  lr: 0.003912  min_lr: 0.003912  loss: 3.0798 (3.0838)  weight_decay: 0.0500 (0.0500)  time: 0.6894  data: 0.0001  max mem: 57114
Epoch: [161]  [155/156]  eta: 0:00:00  lr: 0.003911  min_lr: 0.003911  loss: 3.0451 (3.0795)  weight_decay: 0.0500 (0.0500)  time: 0.6895  data: 0.0001  max mem: 57114
Epoch: [161] Total time: 0:01:53 (0.7288 s / it)
Averaged stats: lr: 0.003911  min_lr: 0.003911  loss: 3.0451 (3.0722)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9175 (0.9175)  acc1: 84.3750 (84.3750)  acc5: 96.3542 (96.3542)  time: 2.1013  data: 1.8452  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0433 (0.9812)  acc1: 78.1250 (78.6445)  acc5: 96.3542 (95.5243)  time: 0.5709  data: 0.3691  max mem: 57114
Test: Total time: 0:00:02 (0.5938 s / it)
* Acc@1 78.872 Acc@5 94.901 loss 1.086
Accuracy of the model on the 50000 test images: 78.9%
Max accuracy: 79.72%
Test:  [0/5]  eta: 0:00:11  loss: 0.7979 (0.7979)  acc1: 80.7292 (80.7292)  acc5: 94.2708 (94.2708)  time: 2.3052  data: 2.0618  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9743 (0.9743)  acc1: 75.5208 (73.6573)  acc5: 92.1875 (90.1535)  time: 0.6117  data: 0.4124  max mem: 57114
Test: Total time: 0:00:03 (0.6256 s / it)
* Acc@1 70.257 Acc@5 89.992 loss 1.182
Accuracy of the model EMA on 50000 test images: 70.3%
Max EMA accuracy: 70.26%
Epoch: [162]  [  0/156]  eta: 0:08:41  lr: 0.003911  min_lr: 0.003911  loss: 3.4539 (3.4539)  weight_decay: 0.0500 (0.0500)  time: 3.3422  data: 2.6958  max mem: 57114
Epoch: [162]  [ 10/156]  eta: 0:02:17  lr: 0.003908  min_lr: 0.003908  loss: 3.3084 (3.1515)  weight_decay: 0.0500 (0.0500)  time: 0.9411  data: 0.2453  max mem: 57114
Epoch: [162]  [ 20/156]  eta: 0:01:53  lr: 0.003905  min_lr: 0.003905  loss: 3.0992 (3.0946)  weight_decay: 0.0500 (0.0500)  time: 0.7097  data: 0.0003  max mem: 57114
Epoch: [162]  [ 30/156]  eta: 0:01:40  lr: 0.003902  min_lr: 0.003902  loss: 3.1726 (3.1283)  weight_decay: 0.0500 (0.0500)  time: 0.7209  data: 0.0003  max mem: 57114
Epoch: [162]  [ 40/156]  eta: 0:01:31  lr: 0.003899  min_lr: 0.003899  loss: 3.1726 (3.1002)  weight_decay: 0.0500 (0.0500)  time: 0.7323  data: 0.0004  max mem: 57114
Epoch: [162]  [ 50/156]  eta: 0:01:21  lr: 0.003896  min_lr: 0.003896  loss: 2.9695 (3.0808)  weight_decay: 0.0500 (0.0500)  time: 0.7308  data: 0.0004  max mem: 57114
Epoch: [162]  [ 60/156]  eta: 0:01:13  lr: 0.003894  min_lr: 0.003894  loss: 3.0992 (3.0728)  weight_decay: 0.0500 (0.0500)  time: 0.7286  data: 0.0004  max mem: 57114
Epoch: [162]  [ 70/156]  eta: 0:01:05  lr: 0.003891  min_lr: 0.003891  loss: 3.0992 (3.0630)  weight_decay: 0.0500 (0.0500)  time: 0.7200  data: 0.0004  max mem: 57114
Epoch: [162]  [ 80/156]  eta: 0:00:57  lr: 0.003888  min_lr: 0.003888  loss: 3.2277 (3.0844)  weight_decay: 0.0500 (0.0500)  time: 0.7054  data: 0.0004  max mem: 57114
Epoch: [162]  [ 90/156]  eta: 0:00:49  lr: 0.003885  min_lr: 0.003885  loss: 3.2538 (3.0916)  weight_decay: 0.0500 (0.0500)  time: 0.7011  data: 0.0004  max mem: 57114
Epoch: [162]  [100/156]  eta: 0:00:41  lr: 0.003882  min_lr: 0.003882  loss: 3.1687 (3.0847)  weight_decay: 0.0500 (0.0500)  time: 0.7035  data: 0.0004  max mem: 57114
Epoch: [162]  [110/156]  eta: 0:00:33  lr: 0.003879  min_lr: 0.003879  loss: 3.0422 (3.0653)  weight_decay: 0.0500 (0.0500)  time: 0.7116  data: 0.0003  max mem: 57114
Epoch: [162]  [120/156]  eta: 0:00:26  lr: 0.003876  min_lr: 0.003876  loss: 2.9877 (3.0577)  weight_decay: 0.0500 (0.0500)  time: 0.7078  data: 0.0003  max mem: 57114
Epoch: [162]  [130/156]  eta: 0:00:19  lr: 0.003873  min_lr: 0.003873  loss: 3.0708 (3.0620)  weight_decay: 0.0500 (0.0500)  time: 0.7005  data: 0.0008  max mem: 57114
Epoch: [162]  [140/156]  eta: 0:00:11  lr: 0.003871  min_lr: 0.003871  loss: 3.1583 (3.0542)  weight_decay: 0.0500 (0.0500)  time: 0.6925  data: 0.0007  max mem: 57114
Epoch: [162]  [150/156]  eta: 0:00:04  lr: 0.003868  min_lr: 0.003868  loss: 2.8641 (3.0417)  weight_decay: 0.0500 (0.0500)  time: 0.6864  data: 0.0001  max mem: 57114
Epoch: [162]  [155/156]  eta: 0:00:00  lr: 0.003866  min_lr: 0.003866  loss: 2.8803 (3.0484)  weight_decay: 0.0500 (0.0500)  time: 0.6856  data: 0.0001  max mem: 57114
Epoch: [162] Total time: 0:01:53 (0.7275 s / it)
Averaged stats: lr: 0.003866  min_lr: 0.003866  loss: 2.8803 (3.0571)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8350 (0.8350)  acc1: 85.4167 (85.4167)  acc5: 97.3958 (97.3958)  time: 2.0725  data: 1.8151  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9400 (0.9239)  acc1: 82.2917 (80.6905)  acc5: 95.8333 (94.5013)  time: 0.5652  data: 0.3631  max mem: 57114
Test: Total time: 0:00:02 (0.5877 s / it)
* Acc@1 79.186 Acc@5 94.731 loss 1.007
Accuracy of the model on the 50000 test images: 79.2%
Max accuracy: 79.72%
Test:  [0/5]  eta: 0:00:11  loss: 0.7894 (0.7894)  acc1: 80.7292 (80.7292)  acc5: 94.2708 (94.2708)  time: 2.2642  data: 2.0205  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9612 (0.9610)  acc1: 76.5625 (73.9130)  acc5: 92.1875 (90.4092)  time: 0.6035  data: 0.4042  max mem: 57114
Test: Total time: 0:00:03 (0.6153 s / it)
* Acc@1 70.552 Acc@5 90.173 loss 1.166
Accuracy of the model EMA on 50000 test images: 70.6%
Max EMA accuracy: 70.55%
Epoch: [163]  [  0/156]  eta: 0:07:42  lr: 0.003866  min_lr: 0.003866  loss: 1.7212 (1.7212)  weight_decay: 0.0500 (0.0500)  time: 2.9658  data: 2.3140  max mem: 57114
Epoch: [163]  [ 10/156]  eta: 0:02:15  lr: 0.003863  min_lr: 0.003863  loss: 3.4656 (3.3029)  weight_decay: 0.0500 (0.0500)  time: 0.9289  data: 0.2106  max mem: 57114
Epoch: [163]  [ 20/156]  eta: 0:01:52  lr: 0.003860  min_lr: 0.003860  loss: 3.4251 (3.3004)  weight_decay: 0.0500 (0.0500)  time: 0.7179  data: 0.0003  max mem: 57114
Epoch: [163]  [ 30/156]  eta: 0:01:40  lr: 0.003857  min_lr: 0.003857  loss: 3.2014 (3.1407)  weight_decay: 0.0500 (0.0500)  time: 0.7286  data: 0.0004  max mem: 57114
Epoch: [163]  [ 40/156]  eta: 0:01:30  lr: 0.003854  min_lr: 0.003854  loss: 2.7340 (3.0487)  weight_decay: 0.0500 (0.0500)  time: 0.7345  data: 0.0004  max mem: 57114
Epoch: [163]  [ 50/156]  eta: 0:01:21  lr: 0.003852  min_lr: 0.003852  loss: 2.8867 (3.0531)  weight_decay: 0.0500 (0.0500)  time: 0.7200  data: 0.0003  max mem: 57114
Epoch: [163]  [ 60/156]  eta: 0:01:13  lr: 0.003849  min_lr: 0.003849  loss: 3.1401 (3.0445)  weight_decay: 0.0500 (0.0500)  time: 0.7237  data: 0.0003  max mem: 57114
Epoch: [163]  [ 70/156]  eta: 0:01:04  lr: 0.003846  min_lr: 0.003846  loss: 3.0643 (3.0159)  weight_decay: 0.0500 (0.0500)  time: 0.7180  data: 0.0004  max mem: 57114
Epoch: [163]  [ 80/156]  eta: 0:00:56  lr: 0.003843  min_lr: 0.003843  loss: 2.6927 (3.0102)  weight_decay: 0.0500 (0.0500)  time: 0.7063  data: 0.0004  max mem: 57114
Epoch: [163]  [ 90/156]  eta: 0:00:49  lr: 0.003840  min_lr: 0.003840  loss: 2.9872 (2.9883)  weight_decay: 0.0500 (0.0500)  time: 0.7084  data: 0.0004  max mem: 57114
Epoch: [163]  [100/156]  eta: 0:00:41  lr: 0.003837  min_lr: 0.003837  loss: 2.9752 (2.9607)  weight_decay: 0.0500 (0.0500)  time: 0.7125  data: 0.0004  max mem: 57114
Epoch: [163]  [110/156]  eta: 0:00:33  lr: 0.003834  min_lr: 0.003834  loss: 2.7401 (2.9437)  weight_decay: 0.0500 (0.0500)  time: 0.7039  data: 0.0003  max mem: 57114
Epoch: [163]  [120/156]  eta: 0:00:26  lr: 0.003831  min_lr: 0.003831  loss: 3.0302 (2.9585)  weight_decay: 0.0500 (0.0500)  time: 0.7015  data: 0.0004  max mem: 57114
Epoch: [163]  [130/156]  eta: 0:00:19  lr: 0.003829  min_lr: 0.003829  loss: 3.1653 (2.9558)  weight_decay: 0.0500 (0.0500)  time: 0.6994  data: 0.0008  max mem: 57114
Epoch: [163]  [140/156]  eta: 0:00:11  lr: 0.003826  min_lr: 0.003826  loss: 3.1653 (2.9690)  weight_decay: 0.0500 (0.0500)  time: 0.6865  data: 0.0006  max mem: 57114
Epoch: [163]  [150/156]  eta: 0:00:04  lr: 0.003823  min_lr: 0.003823  loss: 2.9717 (2.9717)  weight_decay: 0.0500 (0.0500)  time: 0.6850  data: 0.0001  max mem: 57114
Epoch: [163]  [155/156]  eta: 0:00:00  lr: 0.003821  min_lr: 0.003821  loss: 2.9717 (2.9805)  weight_decay: 0.0500 (0.0500)  time: 0.6875  data: 0.0001  max mem: 57114
Epoch: [163] Total time: 0:01:53 (0.7260 s / it)
Averaged stats: lr: 0.003821  min_lr: 0.003821  loss: 2.9717 (3.0468)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8472 (0.8472)  acc1: 84.3750 (84.3750)  acc5: 98.4375 (98.4375)  time: 2.1605  data: 1.9047  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0390 (1.0074)  acc1: 81.7708 (79.9233)  acc5: 95.8333 (95.1407)  time: 0.5828  data: 0.3810  max mem: 57114
Test: Total time: 0:00:03 (0.6059 s / it)
* Acc@1 79.907 Acc@5 95.137 loss 1.083
Accuracy of the model on the 50000 test images: 79.9%
Max accuracy: 79.91%
Test:  [0/5]  eta: 0:00:09  loss: 0.7807 (0.7807)  acc1: 81.2500 (81.2500)  acc5: 94.2708 (94.2708)  time: 1.9569  data: 1.7134  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9487 (0.9476)  acc1: 76.5625 (74.5524)  acc5: 92.1875 (90.4092)  time: 0.5420  data: 0.3428  max mem: 57114
Test: Total time: 0:00:02 (0.5527 s / it)
* Acc@1 70.866 Acc@5 90.353 loss 1.151
Accuracy of the model EMA on 50000 test images: 70.9%
Max EMA accuracy: 70.87%
Epoch: [164]  [  0/156]  eta: 0:08:32  lr: 0.003821  min_lr: 0.003821  loss: 2.8352 (2.8352)  weight_decay: 0.0500 (0.0500)  time: 3.2850  data: 2.6355  max mem: 57114
Epoch: [164]  [ 10/156]  eta: 0:02:20  lr: 0.003818  min_lr: 0.003818  loss: 3.2044 (3.1646)  weight_decay: 0.0500 (0.0500)  time: 0.9612  data: 0.2399  max mem: 57114
Epoch: [164]  [ 20/156]  eta: 0:01:55  lr: 0.003815  min_lr: 0.003815  loss: 3.1511 (3.0449)  weight_decay: 0.0500 (0.0500)  time: 0.7263  data: 0.0004  max mem: 57114
Epoch: [164]  [ 30/156]  eta: 0:01:42  lr: 0.003812  min_lr: 0.003812  loss: 3.2901 (3.1692)  weight_decay: 0.0500 (0.0500)  time: 0.7277  data: 0.0004  max mem: 57114
Epoch: [164]  [ 40/156]  eta: 0:01:32  lr: 0.003810  min_lr: 0.003810  loss: 3.2901 (3.1366)  weight_decay: 0.0500 (0.0500)  time: 0.7395  data: 0.0004  max mem: 57114
Epoch: [164]  [ 50/156]  eta: 0:01:23  lr: 0.003807  min_lr: 0.003807  loss: 3.0422 (3.1157)  weight_decay: 0.0500 (0.0500)  time: 0.7476  data: 0.0004  max mem: 57114
Epoch: [164]  [ 60/156]  eta: 0:01:14  lr: 0.003804  min_lr: 0.003804  loss: 2.9696 (3.0734)  weight_decay: 0.0500 (0.0500)  time: 0.7363  data: 0.0004  max mem: 57114
Epoch: [164]  [ 70/156]  eta: 0:01:06  lr: 0.003801  min_lr: 0.003801  loss: 2.9696 (3.0782)  weight_decay: 0.0500 (0.0500)  time: 0.7273  data: 0.0004  max mem: 57114
Epoch: [164]  [ 80/156]  eta: 0:00:57  lr: 0.003798  min_lr: 0.003798  loss: 3.1425 (3.0755)  weight_decay: 0.0500 (0.0500)  time: 0.7213  data: 0.0004  max mem: 57114
Epoch: [164]  [ 90/156]  eta: 0:00:49  lr: 0.003795  min_lr: 0.003795  loss: 3.1759 (3.0622)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0004  max mem: 57114
Epoch: [164]  [100/156]  eta: 0:00:42  lr: 0.003792  min_lr: 0.003792  loss: 2.9022 (3.0442)  weight_decay: 0.0500 (0.0500)  time: 0.7024  data: 0.0003  max mem: 57114
Epoch: [164]  [110/156]  eta: 0:00:34  lr: 0.003789  min_lr: 0.003789  loss: 2.9338 (3.0425)  weight_decay: 0.0500 (0.0500)  time: 0.7012  data: 0.0003  max mem: 57114
Epoch: [164]  [120/156]  eta: 0:00:26  lr: 0.003787  min_lr: 0.003787  loss: 3.2043 (3.0502)  weight_decay: 0.0500 (0.0500)  time: 0.7068  data: 0.0004  max mem: 57114
Epoch: [164]  [130/156]  eta: 0:00:19  lr: 0.003784  min_lr: 0.003784  loss: 3.2191 (3.0592)  weight_decay: 0.0500 (0.0500)  time: 0.7032  data: 0.0008  max mem: 57114
Epoch: [164]  [140/156]  eta: 0:00:11  lr: 0.003781  min_lr: 0.003781  loss: 3.1717 (3.0649)  weight_decay: 0.0500 (0.0500)  time: 0.6899  data: 0.0007  max mem: 57114
Epoch: [164]  [150/156]  eta: 0:00:04  lr: 0.003778  min_lr: 0.003778  loss: 3.2965 (3.0709)  weight_decay: 0.0500 (0.0500)  time: 0.6859  data: 0.0001  max mem: 57114
Epoch: [164]  [155/156]  eta: 0:00:00  lr: 0.003777  min_lr: 0.003777  loss: 3.1985 (3.0736)  weight_decay: 0.0500 (0.0500)  time: 0.6900  data: 0.0001  max mem: 57114
Epoch: [164] Total time: 0:01:54 (0.7335 s / it)
Averaged stats: lr: 0.003777  min_lr: 0.003777  loss: 3.1985 (3.0588)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9332 (0.9332)  acc1: 85.9375 (85.9375)  acc5: 97.9167 (97.9167)  time: 2.1869  data: 1.9313  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.1245 (1.0692)  acc1: 80.7292 (79.6675)  acc5: 96.8750 (95.5243)  time: 0.5880  data: 0.3863  max mem: 57114
Test: Total time: 0:00:03 (0.6114 s / it)
* Acc@1 79.384 Acc@5 95.019 loss 1.138
Accuracy of the model on the 50000 test images: 79.4%
Max accuracy: 79.91%
Test:  [0/5]  eta: 0:00:11  loss: 0.7721 (0.7721)  acc1: 81.7708 (81.7708)  acc5: 94.2708 (94.2708)  time: 2.3116  data: 2.0638  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9368 (0.9349)  acc1: 76.5625 (74.8082)  acc5: 92.7083 (90.5371)  time: 0.6130  data: 0.4128  max mem: 57114
Test: Total time: 0:00:03 (0.6228 s / it)
* Acc@1 71.140 Acc@5 90.525 loss 1.136
Accuracy of the model EMA on 50000 test images: 71.1%
Max EMA accuracy: 71.14%
Epoch: [165]  [  0/156]  eta: 0:06:57  lr: 0.003776  min_lr: 0.003776  loss: 3.1592 (3.1592)  weight_decay: 0.0500 (0.0500)  time: 2.6785  data: 2.0282  max mem: 57114
Epoch: [165]  [ 10/156]  eta: 0:02:10  lr: 0.003773  min_lr: 0.003773  loss: 2.9079 (2.9033)  weight_decay: 0.0500 (0.0500)  time: 0.8906  data: 0.1846  max mem: 57114
Epoch: [165]  [ 20/156]  eta: 0:01:49  lr: 0.003771  min_lr: 0.003771  loss: 3.2626 (3.0439)  weight_decay: 0.0500 (0.0500)  time: 0.7100  data: 0.0003  max mem: 57114
Epoch: [165]  [ 30/156]  eta: 0:01:38  lr: 0.003768  min_lr: 0.003768  loss: 2.9879 (2.9332)  weight_decay: 0.0500 (0.0500)  time: 0.7229  data: 0.0003  max mem: 57114
Epoch: [165]  [ 40/156]  eta: 0:01:29  lr: 0.003765  min_lr: 0.003765  loss: 2.7951 (2.9349)  weight_decay: 0.0500 (0.0500)  time: 0.7318  data: 0.0003  max mem: 57114
Epoch: [165]  [ 50/156]  eta: 0:01:20  lr: 0.003762  min_lr: 0.003762  loss: 2.8498 (2.9527)  weight_decay: 0.0500 (0.0500)  time: 0.7306  data: 0.0003  max mem: 57114
Epoch: [165]  [ 60/156]  eta: 0:01:12  lr: 0.003759  min_lr: 0.003759  loss: 2.9895 (2.9703)  weight_decay: 0.0500 (0.0500)  time: 0.7305  data: 0.0003  max mem: 57114
Epoch: [165]  [ 70/156]  eta: 0:01:04  lr: 0.003756  min_lr: 0.003756  loss: 3.0166 (2.9862)  weight_decay: 0.0500 (0.0500)  time: 0.7224  data: 0.0004  max mem: 57114
Epoch: [165]  [ 80/156]  eta: 0:00:56  lr: 0.003753  min_lr: 0.003753  loss: 3.1074 (2.9891)  weight_decay: 0.0500 (0.0500)  time: 0.7035  data: 0.0004  max mem: 57114
Epoch: [165]  [ 90/156]  eta: 0:00:48  lr: 0.003750  min_lr: 0.003750  loss: 3.1074 (2.9920)  weight_decay: 0.0500 (0.0500)  time: 0.7012  data: 0.0004  max mem: 57114
Epoch: [165]  [100/156]  eta: 0:00:41  lr: 0.003748  min_lr: 0.003748  loss: 3.1400 (2.9945)  weight_decay: 0.0500 (0.0500)  time: 0.7087  data: 0.0003  max mem: 57114
Epoch: [165]  [110/156]  eta: 0:00:33  lr: 0.003745  min_lr: 0.003745  loss: 3.1472 (3.0152)  weight_decay: 0.0500 (0.0500)  time: 0.7040  data: 0.0003  max mem: 57114
Epoch: [165]  [120/156]  eta: 0:00:26  lr: 0.003742  min_lr: 0.003742  loss: 3.3019 (3.0252)  weight_decay: 0.0500 (0.0500)  time: 0.7159  data: 0.0004  max mem: 57114
Epoch: [165]  [130/156]  eta: 0:00:18  lr: 0.003739  min_lr: 0.003739  loss: 3.2815 (3.0455)  weight_decay: 0.0500 (0.0500)  time: 0.7110  data: 0.0010  max mem: 57114
Epoch: [165]  [140/156]  eta: 0:00:11  lr: 0.003736  min_lr: 0.003736  loss: 3.1316 (3.0275)  weight_decay: 0.0500 (0.0500)  time: 0.6889  data: 0.0009  max mem: 57114
Epoch: [165]  [150/156]  eta: 0:00:04  lr: 0.003733  min_lr: 0.003733  loss: 2.8092 (3.0048)  weight_decay: 0.0500 (0.0500)  time: 0.6823  data: 0.0001  max mem: 57114
Epoch: [165]  [155/156]  eta: 0:00:00  lr: 0.003732  min_lr: 0.003732  loss: 2.8092 (3.0107)  weight_decay: 0.0500 (0.0500)  time: 0.6798  data: 0.0001  max mem: 57114
Epoch: [165] Total time: 0:01:52 (0.7243 s / it)
Averaged stats: lr: 0.003732  min_lr: 0.003732  loss: 2.8092 (3.0415)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7816 (0.7816)  acc1: 85.4167 (85.4167)  acc5: 97.9167 (97.9167)  time: 2.0402  data: 1.7842  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9415 (0.8814)  acc1: 80.2083 (81.2020)  acc5: 95.8333 (95.5243)  time: 0.5588  data: 0.3569  max mem: 57114
Test: Total time: 0:00:02 (0.5820 s / it)
* Acc@1 80.065 Acc@5 95.434 loss 0.971
Accuracy of the model on the 50000 test images: 80.1%
Max accuracy: 80.07%
Test:  [0/5]  eta: 0:00:09  loss: 0.7632 (0.7632)  acc1: 81.7708 (81.7708)  acc5: 94.7917 (94.7917)  time: 1.9782  data: 1.7347  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9258 (0.9226)  acc1: 76.5625 (75.1918)  acc5: 93.2292 (90.7928)  time: 0.5463  data: 0.3470  max mem: 57114
Test: Total time: 0:00:02 (0.5573 s / it)
* Acc@1 71.473 Acc@5 90.709 loss 1.122
Accuracy of the model EMA on 50000 test images: 71.5%
Max EMA accuracy: 71.47%
Epoch: [166]  [  0/156]  eta: 0:07:29  lr: 0.003731  min_lr: 0.003731  loss: 3.4418 (3.4418)  weight_decay: 0.0500 (0.0500)  time: 2.8802  data: 2.2135  max mem: 57114
Epoch: [166]  [ 10/156]  eta: 0:02:16  lr: 0.003729  min_lr: 0.003729  loss: 3.0028 (2.8489)  weight_decay: 0.0500 (0.0500)  time: 0.9357  data: 0.2017  max mem: 57114
Epoch: [166]  [ 20/156]  eta: 0:01:53  lr: 0.003726  min_lr: 0.003726  loss: 3.0028 (2.9377)  weight_decay: 0.0500 (0.0500)  time: 0.7311  data: 0.0004  max mem: 57114
Epoch: [166]  [ 30/156]  eta: 0:01:42  lr: 0.003723  min_lr: 0.003723  loss: 3.0941 (2.9578)  weight_decay: 0.0500 (0.0500)  time: 0.7403  data: 0.0004  max mem: 57114
Epoch: [166]  [ 40/156]  eta: 0:01:31  lr: 0.003720  min_lr: 0.003720  loss: 3.0047 (2.9520)  weight_decay: 0.0500 (0.0500)  time: 0.7453  data: 0.0004  max mem: 57114
Epoch: [166]  [ 50/156]  eta: 0:01:22  lr: 0.003717  min_lr: 0.003717  loss: 3.0941 (2.9664)  weight_decay: 0.0500 (0.0500)  time: 0.7411  data: 0.0004  max mem: 57114
Epoch: [166]  [ 60/156]  eta: 0:01:14  lr: 0.003714  min_lr: 0.003714  loss: 2.7891 (2.9288)  weight_decay: 0.0500 (0.0500)  time: 0.7354  data: 0.0004  max mem: 57114
Epoch: [166]  [ 70/156]  eta: 0:01:05  lr: 0.003711  min_lr: 0.003711  loss: 3.0287 (2.9490)  weight_decay: 0.0500 (0.0500)  time: 0.7196  data: 0.0004  max mem: 57114
Epoch: [166]  [ 80/156]  eta: 0:00:57  lr: 0.003709  min_lr: 0.003709  loss: 3.2241 (2.9443)  weight_decay: 0.0500 (0.0500)  time: 0.7063  data: 0.0004  max mem: 57114
Epoch: [166]  [ 90/156]  eta: 0:00:49  lr: 0.003706  min_lr: 0.003706  loss: 2.9120 (2.9370)  weight_decay: 0.0500 (0.0500)  time: 0.7061  data: 0.0004  max mem: 57114
Epoch: [166]  [100/156]  eta: 0:00:41  lr: 0.003703  min_lr: 0.003703  loss: 2.8628 (2.9348)  weight_decay: 0.0500 (0.0500)  time: 0.7122  data: 0.0004  max mem: 57114
Epoch: [166]  [110/156]  eta: 0:00:34  lr: 0.003700  min_lr: 0.003700  loss: 3.0879 (2.9463)  weight_decay: 0.0500 (0.0500)  time: 0.7032  data: 0.0004  max mem: 57114
Epoch: [166]  [120/156]  eta: 0:00:26  lr: 0.003697  min_lr: 0.003697  loss: 3.1018 (2.9476)  weight_decay: 0.0500 (0.0500)  time: 0.7051  data: 0.0004  max mem: 57114
Epoch: [166]  [130/156]  eta: 0:00:19  lr: 0.003694  min_lr: 0.003694  loss: 3.1447 (2.9648)  weight_decay: 0.0500 (0.0500)  time: 0.7050  data: 0.0009  max mem: 57114
Epoch: [166]  [140/156]  eta: 0:00:11  lr: 0.003691  min_lr: 0.003691  loss: 3.1801 (2.9733)  weight_decay: 0.0500 (0.0500)  time: 0.6930  data: 0.0007  max mem: 57114
Epoch: [166]  [150/156]  eta: 0:00:04  lr: 0.003688  min_lr: 0.003688  loss: 3.1348 (2.9704)  weight_decay: 0.0500 (0.0500)  time: 0.6874  data: 0.0001  max mem: 57114
Epoch: [166]  [155/156]  eta: 0:00:00  lr: 0.003687  min_lr: 0.003687  loss: 2.8741 (2.9676)  weight_decay: 0.0500 (0.0500)  time: 0.6856  data: 0.0001  max mem: 57114
Epoch: [166] Total time: 0:01:54 (0.7317 s / it)
Averaged stats: lr: 0.003687  min_lr: 0.003687  loss: 2.8741 (3.0458)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7129 (0.7129)  acc1: 88.0208 (88.0208)  acc5: 98.4375 (98.4375)  time: 2.1475  data: 1.8916  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9037 (0.8391)  acc1: 79.6875 (80.6905)  acc5: 95.8333 (95.9079)  time: 0.5804  data: 0.3785  max mem: 57114
Test: Total time: 0:00:03 (0.6027 s / it)
* Acc@1 80.155 Acc@5 95.277 loss 0.915
Accuracy of the model on the 50000 test images: 80.2%
Max accuracy: 80.16%
Test:  [0/5]  eta: 0:00:10  loss: 0.7542 (0.7542)  acc1: 82.2917 (82.2917)  acc5: 94.7917 (94.7917)  time: 2.0547  data: 1.8113  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9146 (0.9106)  acc1: 76.5625 (75.3197)  acc5: 93.2292 (91.0486)  time: 0.5616  data: 0.3623  max mem: 57114
Test: Total time: 0:00:02 (0.5715 s / it)
* Acc@1 71.783 Acc@5 90.881 loss 1.109
Accuracy of the model EMA on 50000 test images: 71.8%
Max EMA accuracy: 71.78%
Epoch: [167]  [  0/156]  eta: 0:09:17  lr: 0.003687  min_lr: 0.003687  loss: 3.1833 (3.1833)  weight_decay: 0.0500 (0.0500)  time: 3.5737  data: 2.9212  max mem: 57114
Epoch: [167]  [ 10/156]  eta: 0:02:22  lr: 0.003684  min_lr: 0.003684  loss: 3.1833 (2.9719)  weight_decay: 0.0500 (0.0500)  time: 0.9732  data: 0.2658  max mem: 57114
Epoch: [167]  [ 20/156]  eta: 0:01:58  lr: 0.003681  min_lr: 0.003681  loss: 3.1186 (2.9757)  weight_decay: 0.0500 (0.0500)  time: 0.7331  data: 0.0003  max mem: 57114
Epoch: [167]  [ 30/156]  eta: 0:01:43  lr: 0.003678  min_lr: 0.003678  loss: 3.1084 (2.9598)  weight_decay: 0.0500 (0.0500)  time: 0.7373  data: 0.0003  max mem: 57114
Epoch: [167]  [ 40/156]  eta: 0:01:32  lr: 0.003675  min_lr: 0.003675  loss: 3.0861 (2.9794)  weight_decay: 0.0500 (0.0500)  time: 0.7298  data: 0.0003  max mem: 57114
Epoch: [167]  [ 50/156]  eta: 0:01:23  lr: 0.003672  min_lr: 0.003672  loss: 3.0666 (2.9735)  weight_decay: 0.0500 (0.0500)  time: 0.7368  data: 0.0004  max mem: 57114
Epoch: [167]  [ 60/156]  eta: 0:01:14  lr: 0.003669  min_lr: 0.003669  loss: 3.1841 (2.9980)  weight_decay: 0.0500 (0.0500)  time: 0.7408  data: 0.0004  max mem: 57114
Epoch: [167]  [ 70/156]  eta: 0:01:06  lr: 0.003667  min_lr: 0.003667  loss: 3.0528 (2.9654)  weight_decay: 0.0500 (0.0500)  time: 0.7311  data: 0.0004  max mem: 57114
Epoch: [167]  [ 80/156]  eta: 0:00:57  lr: 0.003664  min_lr: 0.003664  loss: 2.8583 (2.9734)  weight_decay: 0.0500 (0.0500)  time: 0.7068  data: 0.0004  max mem: 57114
Epoch: [167]  [ 90/156]  eta: 0:00:49  lr: 0.003661  min_lr: 0.003661  loss: 2.9390 (2.9734)  weight_decay: 0.0500 (0.0500)  time: 0.7012  data: 0.0004  max mem: 57114
Epoch: [167]  [100/156]  eta: 0:00:42  lr: 0.003658  min_lr: 0.003658  loss: 2.9390 (2.9742)  weight_decay: 0.0500 (0.0500)  time: 0.7026  data: 0.0004  max mem: 57114
Epoch: [167]  [110/156]  eta: 0:00:34  lr: 0.003655  min_lr: 0.003655  loss: 2.8563 (2.9570)  weight_decay: 0.0500 (0.0500)  time: 0.7034  data: 0.0004  max mem: 57114
Epoch: [167]  [120/156]  eta: 0:00:26  lr: 0.003652  min_lr: 0.003652  loss: 3.2142 (2.9776)  weight_decay: 0.0500 (0.0500)  time: 0.7091  data: 0.0003  max mem: 57114
Epoch: [167]  [130/156]  eta: 0:00:19  lr: 0.003649  min_lr: 0.003649  loss: 3.2547 (2.9891)  weight_decay: 0.0500 (0.0500)  time: 0.7036  data: 0.0005  max mem: 57114
Epoch: [167]  [140/156]  eta: 0:00:11  lr: 0.003647  min_lr: 0.003647  loss: 3.2331 (2.9912)  weight_decay: 0.0500 (0.0500)  time: 0.6932  data: 0.0004  max mem: 57114
Epoch: [167]  [150/156]  eta: 0:00:04  lr: 0.003644  min_lr: 0.003644  loss: 3.1454 (2.9925)  weight_decay: 0.0500 (0.0500)  time: 0.6886  data: 0.0001  max mem: 57114
Epoch: [167]  [155/156]  eta: 0:00:00  lr: 0.003642  min_lr: 0.003642  loss: 2.9480 (2.9887)  weight_decay: 0.0500 (0.0500)  time: 0.6899  data: 0.0001  max mem: 57114
Epoch: [167] Total time: 0:01:54 (0.7340 s / it)
Averaged stats: lr: 0.003642  min_lr: 0.003642  loss: 2.9480 (3.0403)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8098 (0.8098)  acc1: 86.4583 (86.4583)  acc5: 97.9167 (97.9167)  time: 2.1645  data: 1.9090  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9189 (0.9172)  acc1: 79.6875 (79.7954)  acc5: 96.3542 (96.5473)  time: 0.5836  data: 0.3819  max mem: 57114
Test: Total time: 0:00:03 (0.6068 s / it)
* Acc@1 80.133 Acc@5 95.402 loss 1.007
Accuracy of the model on the 50000 test images: 80.1%
Max accuracy: 80.16%
Test:  [0/5]  eta: 0:00:11  loss: 0.7463 (0.7463)  acc1: 82.2917 (82.2917)  acc5: 94.7917 (94.7917)  time: 2.2719  data: 2.0283  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9048 (0.8998)  acc1: 76.5625 (75.5755)  acc5: 93.7500 (91.1765)  time: 0.6051  data: 0.4058  max mem: 57114
Test: Total time: 0:00:03 (0.6150 s / it)
* Acc@1 72.049 Acc@5 91.009 loss 1.095
Accuracy of the model EMA on 50000 test images: 72.0%
Max EMA accuracy: 72.05%
Epoch: [168]  [  0/156]  eta: 0:06:49  lr: 0.003642  min_lr: 0.003642  loss: 2.8310 (2.8310)  weight_decay: 0.0500 (0.0500)  time: 2.6254  data: 1.9742  max mem: 57114
Epoch: [168]  [ 10/156]  eta: 0:02:09  lr: 0.003639  min_lr: 0.003639  loss: 3.2125 (3.0821)  weight_decay: 0.0500 (0.0500)  time: 0.8898  data: 0.1798  max mem: 57114
Epoch: [168]  [ 20/156]  eta: 0:01:49  lr: 0.003636  min_lr: 0.003636  loss: 3.1940 (2.9825)  weight_decay: 0.0500 (0.0500)  time: 0.7170  data: 0.0003  max mem: 57114
Epoch: [168]  [ 30/156]  eta: 0:01:39  lr: 0.003633  min_lr: 0.003633  loss: 3.0485 (2.9553)  weight_decay: 0.0500 (0.0500)  time: 0.7309  data: 0.0004  max mem: 57114
Epoch: [168]  [ 40/156]  eta: 0:01:29  lr: 0.003631  min_lr: 0.003631  loss: 3.0469 (2.9856)  weight_decay: 0.0500 (0.0500)  time: 0.7305  data: 0.0004  max mem: 57114
Epoch: [168]  [ 50/156]  eta: 0:01:20  lr: 0.003628  min_lr: 0.003628  loss: 3.2583 (3.0262)  weight_decay: 0.0500 (0.0500)  time: 0.7192  data: 0.0004  max mem: 57114
Epoch: [168]  [ 60/156]  eta: 0:01:12  lr: 0.003625  min_lr: 0.003625  loss: 3.2886 (3.0602)  weight_decay: 0.0500 (0.0500)  time: 0.7283  data: 0.0004  max mem: 57114
Epoch: [168]  [ 70/156]  eta: 0:01:04  lr: 0.003622  min_lr: 0.003622  loss: 3.2797 (3.0668)  weight_decay: 0.0500 (0.0500)  time: 0.7220  data: 0.0003  max mem: 57114
Epoch: [168]  [ 80/156]  eta: 0:00:56  lr: 0.003619  min_lr: 0.003619  loss: 3.2854 (3.0994)  weight_decay: 0.0500 (0.0500)  time: 0.6994  data: 0.0004  max mem: 57114
Epoch: [168]  [ 90/156]  eta: 0:00:48  lr: 0.003616  min_lr: 0.003616  loss: 3.0818 (3.0846)  weight_decay: 0.0500 (0.0500)  time: 0.6993  data: 0.0004  max mem: 57114
Epoch: [168]  [100/156]  eta: 0:00:41  lr: 0.003613  min_lr: 0.003613  loss: 3.0544 (3.0891)  weight_decay: 0.0500 (0.0500)  time: 0.7053  data: 0.0004  max mem: 57114
Epoch: [168]  [110/156]  eta: 0:00:33  lr: 0.003610  min_lr: 0.003610  loss: 3.0220 (3.0690)  weight_decay: 0.0500 (0.0500)  time: 0.7075  data: 0.0004  max mem: 57114
Epoch: [168]  [120/156]  eta: 0:00:26  lr: 0.003608  min_lr: 0.003608  loss: 3.0240 (3.0709)  weight_decay: 0.0500 (0.0500)  time: 0.7095  data: 0.0004  max mem: 57114
Epoch: [168]  [130/156]  eta: 0:00:18  lr: 0.003605  min_lr: 0.003605  loss: 3.1233 (3.0685)  weight_decay: 0.0500 (0.0500)  time: 0.7044  data: 0.0004  max mem: 57114
Epoch: [168]  [140/156]  eta: 0:00:11  lr: 0.003602  min_lr: 0.003602  loss: 3.1529 (3.0764)  weight_decay: 0.0500 (0.0500)  time: 0.6894  data: 0.0003  max mem: 57114
Epoch: [168]  [150/156]  eta: 0:00:04  lr: 0.003599  min_lr: 0.003599  loss: 3.1554 (3.0674)  weight_decay: 0.0500 (0.0500)  time: 0.6774  data: 0.0001  max mem: 57114
Epoch: [168]  [155/156]  eta: 0:00:00  lr: 0.003598  min_lr: 0.003598  loss: 3.1554 (3.0697)  weight_decay: 0.0500 (0.0500)  time: 0.6784  data: 0.0001  max mem: 57114
Epoch: [168] Total time: 0:01:52 (0.7226 s / it)
Averaged stats: lr: 0.003598  min_lr: 0.003598  loss: 3.1554 (3.0364)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7796 (0.7796)  acc1: 86.9792 (86.9792)  acc5: 98.4375 (98.4375)  time: 2.0250  data: 1.7690  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9344 (0.9405)  acc1: 79.6875 (79.1560)  acc5: 96.8750 (95.3964)  time: 0.5557  data: 0.3539  max mem: 57114
Test: Total time: 0:00:02 (0.5787 s / it)
* Acc@1 79.915 Acc@5 95.093 loss 1.003
Accuracy of the model on the 50000 test images: 79.9%
Max accuracy: 80.16%
Test:  [0/5]  eta: 0:00:11  loss: 0.7389 (0.7389)  acc1: 83.3333 (83.3333)  acc5: 94.7917 (94.7917)  time: 2.3254  data: 2.0820  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8952 (0.8890)  acc1: 77.0833 (75.8312)  acc5: 93.7500 (91.1765)  time: 0.6157  data: 0.4165  max mem: 57114
Test: Total time: 0:00:03 (0.6257 s / it)
* Acc@1 72.309 Acc@5 91.145 loss 1.082
Accuracy of the model EMA on 50000 test images: 72.3%
Max EMA accuracy: 72.31%
Epoch: [169]  [  0/156]  eta: 0:06:33  lr: 0.003597  min_lr: 0.003597  loss: 3.3579 (3.3579)  weight_decay: 0.0500 (0.0500)  time: 2.5225  data: 1.8704  max mem: 57114
Epoch: [169]  [ 10/156]  eta: 0:02:10  lr: 0.003594  min_lr: 0.003594  loss: 3.3579 (3.1020)  weight_decay: 0.0500 (0.0500)  time: 0.8909  data: 0.1703  max mem: 57114
Epoch: [169]  [ 20/156]  eta: 0:01:50  lr: 0.003592  min_lr: 0.003592  loss: 3.2373 (3.0860)  weight_decay: 0.0500 (0.0500)  time: 0.7239  data: 0.0003  max mem: 57114
Epoch: [169]  [ 30/156]  eta: 0:01:39  lr: 0.003589  min_lr: 0.003589  loss: 3.0271 (3.0986)  weight_decay: 0.0500 (0.0500)  time: 0.7304  data: 0.0004  max mem: 57114
Epoch: [169]  [ 40/156]  eta: 0:01:29  lr: 0.003586  min_lr: 0.003586  loss: 2.9873 (3.0371)  weight_decay: 0.0500 (0.0500)  time: 0.7304  data: 0.0004  max mem: 57114
Epoch: [169]  [ 50/156]  eta: 0:01:20  lr: 0.003583  min_lr: 0.003583  loss: 2.8853 (3.0225)  weight_decay: 0.0500 (0.0500)  time: 0.7161  data: 0.0004  max mem: 57114
Epoch: [169]  [ 60/156]  eta: 0:01:12  lr: 0.003580  min_lr: 0.003580  loss: 3.1682 (3.0559)  weight_decay: 0.0500 (0.0500)  time: 0.7175  data: 0.0003  max mem: 57114
Epoch: [169]  [ 70/156]  eta: 0:01:04  lr: 0.003577  min_lr: 0.003577  loss: 3.1682 (3.0414)  weight_decay: 0.0500 (0.0500)  time: 0.7103  data: 0.0003  max mem: 57114
Epoch: [169]  [ 80/156]  eta: 0:00:56  lr: 0.003574  min_lr: 0.003574  loss: 3.0584 (3.0476)  weight_decay: 0.0500 (0.0500)  time: 0.7057  data: 0.0003  max mem: 57114
Epoch: [169]  [ 90/156]  eta: 0:00:48  lr: 0.003572  min_lr: 0.003572  loss: 3.0584 (3.0496)  weight_decay: 0.0500 (0.0500)  time: 0.7117  data: 0.0003  max mem: 57114
Epoch: [169]  [100/156]  eta: 0:00:41  lr: 0.003569  min_lr: 0.003569  loss: 3.1250 (3.0632)  weight_decay: 0.0500 (0.0500)  time: 0.7081  data: 0.0004  max mem: 57114
Epoch: [169]  [110/156]  eta: 0:00:33  lr: 0.003566  min_lr: 0.003566  loss: 3.1250 (3.0492)  weight_decay: 0.0500 (0.0500)  time: 0.7108  data: 0.0004  max mem: 57114
Epoch: [169]  [120/156]  eta: 0:00:26  lr: 0.003563  min_lr: 0.003563  loss: 2.9896 (3.0295)  weight_decay: 0.0500 (0.0500)  time: 0.7079  data: 0.0004  max mem: 57114
Epoch: [169]  [130/156]  eta: 0:00:18  lr: 0.003560  min_lr: 0.003560  loss: 2.8497 (3.0184)  weight_decay: 0.0500 (0.0500)  time: 0.6973  data: 0.0010  max mem: 57114
Epoch: [169]  [140/156]  eta: 0:00:11  lr: 0.003557  min_lr: 0.003557  loss: 3.0073 (3.0166)  weight_decay: 0.0500 (0.0500)  time: 0.6893  data: 0.0008  max mem: 57114
Epoch: [169]  [150/156]  eta: 0:00:04  lr: 0.003554  min_lr: 0.003554  loss: 3.2283 (3.0319)  weight_decay: 0.0500 (0.0500)  time: 0.6848  data: 0.0001  max mem: 57114
Epoch: [169]  [155/156]  eta: 0:00:00  lr: 0.003553  min_lr: 0.003553  loss: 3.2395 (3.0288)  weight_decay: 0.0500 (0.0500)  time: 0.6861  data: 0.0001  max mem: 57114
Epoch: [169] Total time: 0:01:52 (0.7230 s / it)
Averaged stats: lr: 0.003553  min_lr: 0.003553  loss: 3.2395 (3.0304)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8110 (0.8110)  acc1: 85.9375 (85.9375)  acc5: 97.9167 (97.9167)  time: 2.1649  data: 1.9087  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9468 (0.8662)  acc1: 82.2917 (80.8184)  acc5: 97.9167 (96.0358)  time: 0.5838  data: 0.3818  max mem: 57114
Test: Total time: 0:00:03 (0.6096 s / it)
* Acc@1 80.135 Acc@5 95.279 loss 0.969
Accuracy of the model on the 50000 test images: 80.1%
Max accuracy: 80.16%
Test:  [0/5]  eta: 0:00:11  loss: 0.7318 (0.7318)  acc1: 83.3333 (83.3333)  acc5: 94.7917 (94.7917)  time: 2.3413  data: 2.0978  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8862 (0.8784)  acc1: 77.6042 (76.0870)  acc5: 93.7500 (91.4322)  time: 0.6190  data: 0.4197  max mem: 57114
Test: Total time: 0:00:03 (0.6305 s / it)
* Acc@1 72.530 Acc@5 91.284 loss 1.070
Accuracy of the model EMA on 50000 test images: 72.5%
Max EMA accuracy: 72.53%
Epoch: [170]  [  0/156]  eta: 0:07:44  lr: 0.003553  min_lr: 0.003553  loss: 2.8674 (2.8674)  weight_decay: 0.0500 (0.0500)  time: 2.9745  data: 2.3251  max mem: 57114
Epoch: [170]  [ 10/156]  eta: 0:02:16  lr: 0.003550  min_lr: 0.003550  loss: 3.1418 (3.0585)  weight_decay: 0.0500 (0.0500)  time: 0.9355  data: 0.2116  max mem: 57114
Epoch: [170]  [ 20/156]  eta: 0:01:53  lr: 0.003547  min_lr: 0.003547  loss: 3.1588 (3.1487)  weight_decay: 0.0500 (0.0500)  time: 0.7279  data: 0.0003  max mem: 57114
Epoch: [170]  [ 30/156]  eta: 0:01:41  lr: 0.003544  min_lr: 0.003544  loss: 3.1869 (3.1433)  weight_decay: 0.0500 (0.0500)  time: 0.7302  data: 0.0003  max mem: 57114
Epoch: [170]  [ 40/156]  eta: 0:01:31  lr: 0.003541  min_lr: 0.003541  loss: 3.1250 (3.1078)  weight_decay: 0.0500 (0.0500)  time: 0.7339  data: 0.0004  max mem: 57114
Epoch: [170]  [ 50/156]  eta: 0:01:21  lr: 0.003538  min_lr: 0.003538  loss: 3.2284 (3.1151)  weight_decay: 0.0500 (0.0500)  time: 0.7237  data: 0.0004  max mem: 57114
Epoch: [170]  [ 60/156]  eta: 0:01:13  lr: 0.003536  min_lr: 0.003536  loss: 3.1536 (3.0970)  weight_decay: 0.0500 (0.0500)  time: 0.7126  data: 0.0004  max mem: 57114
Epoch: [170]  [ 70/156]  eta: 0:01:04  lr: 0.003533  min_lr: 0.003533  loss: 3.1536 (3.0804)  weight_decay: 0.0500 (0.0500)  time: 0.7081  data: 0.0004  max mem: 57114
Epoch: [170]  [ 80/156]  eta: 0:00:56  lr: 0.003530  min_lr: 0.003530  loss: 3.2029 (3.0741)  weight_decay: 0.0500 (0.0500)  time: 0.7124  data: 0.0003  max mem: 57114
Epoch: [170]  [ 90/156]  eta: 0:00:49  lr: 0.003527  min_lr: 0.003527  loss: 2.9983 (3.0514)  weight_decay: 0.0500 (0.0500)  time: 0.7118  data: 0.0003  max mem: 57114
Epoch: [170]  [100/156]  eta: 0:00:41  lr: 0.003524  min_lr: 0.003524  loss: 2.9395 (3.0440)  weight_decay: 0.0500 (0.0500)  time: 0.7070  data: 0.0004  max mem: 57114
Epoch: [170]  [110/156]  eta: 0:00:33  lr: 0.003521  min_lr: 0.003521  loss: 2.9395 (3.0303)  weight_decay: 0.0500 (0.0500)  time: 0.7116  data: 0.0004  max mem: 57114
Epoch: [170]  [120/156]  eta: 0:00:26  lr: 0.003518  min_lr: 0.003518  loss: 2.9354 (3.0102)  weight_decay: 0.0500 (0.0500)  time: 0.7089  data: 0.0004  max mem: 57114
Epoch: [170]  [130/156]  eta: 0:00:19  lr: 0.003516  min_lr: 0.003516  loss: 3.0578 (3.0158)  weight_decay: 0.0500 (0.0500)  time: 0.6963  data: 0.0008  max mem: 57114
Epoch: [170]  [140/156]  eta: 0:00:11  lr: 0.003513  min_lr: 0.003513  loss: 3.1586 (3.0365)  weight_decay: 0.0500 (0.0500)  time: 0.6884  data: 0.0007  max mem: 57114
Epoch: [170]  [150/156]  eta: 0:00:04  lr: 0.003510  min_lr: 0.003510  loss: 3.3746 (3.0483)  weight_decay: 0.0500 (0.0500)  time: 0.6859  data: 0.0001  max mem: 57114
Epoch: [170]  [155/156]  eta: 0:00:00  lr: 0.003508  min_lr: 0.003508  loss: 3.3194 (3.0544)  weight_decay: 0.0500 (0.0500)  time: 0.6935  data: 0.0001  max mem: 57114
Epoch: [170] Total time: 0:01:53 (0.7276 s / it)
Averaged stats: lr: 0.003508  min_lr: 0.003508  loss: 3.3194 (3.0197)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9841 (0.9841)  acc1: 85.4167 (85.4167)  acc5: 97.3958 (97.3958)  time: 2.0593  data: 1.8030  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0706 (1.0858)  acc1: 79.1667 (80.4348)  acc5: 96.3542 (95.1407)  time: 0.5625  data: 0.3607  max mem: 57114
Test: Total time: 0:00:02 (0.5860 s / it)
* Acc@1 79.574 Acc@5 95.005 loss 1.177
Accuracy of the model on the 50000 test images: 79.6%
Max accuracy: 80.16%
Test:  [0/5]  eta: 0:00:12  loss: 0.7257 (0.7257)  acc1: 82.8125 (82.8125)  acc5: 94.7917 (94.7917)  time: 2.4192  data: 2.1757  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8777 (0.8688)  acc1: 78.1250 (76.0870)  acc5: 93.7500 (91.4322)  time: 0.6346  data: 0.4352  max mem: 57114
Test: Total time: 0:00:03 (0.6461 s / it)
* Acc@1 72.766 Acc@5 91.384 loss 1.059
Accuracy of the model EMA on 50000 test images: 72.8%
Max EMA accuracy: 72.77%
Epoch: [171]  [  0/156]  eta: 0:07:04  lr: 0.003508  min_lr: 0.003508  loss: 2.8334 (2.8334)  weight_decay: 0.0500 (0.0500)  time: 2.7195  data: 2.0635  max mem: 57114
Epoch: [171]  [ 10/156]  eta: 0:02:14  lr: 0.003505  min_lr: 0.003505  loss: 3.0729 (3.0551)  weight_decay: 0.0500 (0.0500)  time: 0.9217  data: 0.1879  max mem: 57114
Epoch: [171]  [ 20/156]  eta: 0:01:52  lr: 0.003502  min_lr: 0.003502  loss: 3.1145 (3.0793)  weight_decay: 0.0500 (0.0500)  time: 0.7293  data: 0.0003  max mem: 57114
Epoch: [171]  [ 30/156]  eta: 0:01:39  lr: 0.003500  min_lr: 0.003500  loss: 3.3326 (3.1664)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0004  max mem: 57114
Epoch: [171]  [ 40/156]  eta: 0:01:29  lr: 0.003497  min_lr: 0.003497  loss: 3.0507 (3.1129)  weight_decay: 0.0500 (0.0500)  time: 0.7128  data: 0.0004  max mem: 57114
Epoch: [171]  [ 50/156]  eta: 0:01:20  lr: 0.003494  min_lr: 0.003494  loss: 3.0507 (3.1197)  weight_decay: 0.0500 (0.0500)  time: 0.7112  data: 0.0004  max mem: 57114
Epoch: [171]  [ 60/156]  eta: 0:01:12  lr: 0.003491  min_lr: 0.003491  loss: 3.1882 (3.0917)  weight_decay: 0.0500 (0.0500)  time: 0.7215  data: 0.0004  max mem: 57114
Epoch: [171]  [ 70/156]  eta: 0:01:04  lr: 0.003488  min_lr: 0.003488  loss: 3.1882 (3.0817)  weight_decay: 0.0500 (0.0500)  time: 0.7263  data: 0.0004  max mem: 57114
Epoch: [171]  [ 80/156]  eta: 0:00:56  lr: 0.003485  min_lr: 0.003485  loss: 3.0997 (3.0580)  weight_decay: 0.0500 (0.0500)  time: 0.7119  data: 0.0003  max mem: 57114
Epoch: [171]  [ 90/156]  eta: 0:00:48  lr: 0.003482  min_lr: 0.003482  loss: 2.8502 (3.0406)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0003  max mem: 57114
Epoch: [171]  [100/156]  eta: 0:00:41  lr: 0.003480  min_lr: 0.003480  loss: 3.0861 (3.0364)  weight_decay: 0.0500 (0.0500)  time: 0.7062  data: 0.0004  max mem: 57114
Epoch: [171]  [110/156]  eta: 0:00:33  lr: 0.003477  min_lr: 0.003477  loss: 3.0861 (3.0513)  weight_decay: 0.0500 (0.0500)  time: 0.6972  data: 0.0003  max mem: 57114
Epoch: [171]  [120/156]  eta: 0:00:26  lr: 0.003474  min_lr: 0.003474  loss: 3.1578 (3.0543)  weight_decay: 0.0500 (0.0500)  time: 0.7018  data: 0.0004  max mem: 57114
Epoch: [171]  [130/156]  eta: 0:00:18  lr: 0.003471  min_lr: 0.003471  loss: 3.1578 (3.0645)  weight_decay: 0.0500 (0.0500)  time: 0.7073  data: 0.0009  max mem: 57114
Epoch: [171]  [140/156]  eta: 0:00:11  lr: 0.003468  min_lr: 0.003468  loss: 3.0237 (3.0429)  weight_decay: 0.0500 (0.0500)  time: 0.6942  data: 0.0007  max mem: 57114
Epoch: [171]  [150/156]  eta: 0:00:04  lr: 0.003465  min_lr: 0.003465  loss: 3.0651 (3.0588)  weight_decay: 0.0500 (0.0500)  time: 0.6860  data: 0.0001  max mem: 57114
Epoch: [171]  [155/156]  eta: 0:00:00  lr: 0.003464  min_lr: 0.003464  loss: 3.0651 (3.0519)  weight_decay: 0.0500 (0.0500)  time: 0.6866  data: 0.0001  max mem: 57114
Epoch: [171] Total time: 0:01:52 (0.7240 s / it)
Averaged stats: lr: 0.003464  min_lr: 0.003464  loss: 3.0651 (3.0197)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8448 (0.8448)  acc1: 84.8958 (84.8958)  acc5: 96.8750 (96.8750)  time: 2.0502  data: 1.7948  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9320 (0.9043)  acc1: 81.2500 (79.7954)  acc5: 96.8750 (95.3964)  time: 0.5607  data: 0.3590  max mem: 57114
Test: Total time: 0:00:02 (0.5832 s / it)
* Acc@1 80.439 Acc@5 95.412 loss 0.989
Accuracy of the model on the 50000 test images: 80.4%
Max accuracy: 80.44%
Test:  [0/5]  eta: 0:00:10  loss: 0.7189 (0.7189)  acc1: 82.8125 (82.8125)  acc5: 95.3125 (95.3125)  time: 2.0255  data: 1.7820  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8693 (0.8592)  acc1: 78.1250 (76.0870)  acc5: 94.2708 (91.5601)  time: 0.5557  data: 0.3565  max mem: 57114
Test: Total time: 0:00:02 (0.5673 s / it)
* Acc@1 72.996 Acc@5 91.508 loss 1.047
Accuracy of the model EMA on 50000 test images: 73.0%
Max EMA accuracy: 73.00%
Epoch: [172]  [  0/156]  eta: 0:08:51  lr: 0.003464  min_lr: 0.003464  loss: 1.8039 (1.8039)  weight_decay: 0.0500 (0.0500)  time: 3.4098  data: 2.7574  max mem: 57114
Epoch: [172]  [ 10/156]  eta: 0:02:22  lr: 0.003461  min_lr: 0.003461  loss: 2.8699 (2.7984)  weight_decay: 0.0500 (0.0500)  time: 0.9788  data: 0.2510  max mem: 57114
Epoch: [172]  [ 20/156]  eta: 0:01:56  lr: 0.003458  min_lr: 0.003458  loss: 3.0886 (2.9682)  weight_decay: 0.0500 (0.0500)  time: 0.7287  data: 0.0004  max mem: 57114
Epoch: [172]  [ 30/156]  eta: 0:01:41  lr: 0.003455  min_lr: 0.003455  loss: 3.1971 (3.0368)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0004  max mem: 57114
Epoch: [172]  [ 40/156]  eta: 0:01:32  lr: 0.003452  min_lr: 0.003452  loss: 3.1777 (3.0405)  weight_decay: 0.0500 (0.0500)  time: 0.7267  data: 0.0004  max mem: 57114
Epoch: [172]  [ 50/156]  eta: 0:01:22  lr: 0.003449  min_lr: 0.003449  loss: 3.1777 (3.0591)  weight_decay: 0.0500 (0.0500)  time: 0.7343  data: 0.0005  max mem: 57114
Epoch: [172]  [ 60/156]  eta: 0:01:14  lr: 0.003447  min_lr: 0.003447  loss: 2.9796 (3.0205)  weight_decay: 0.0500 (0.0500)  time: 0.7301  data: 0.0004  max mem: 57114
Epoch: [172]  [ 70/156]  eta: 0:01:05  lr: 0.003444  min_lr: 0.003444  loss: 2.9945 (3.0388)  weight_decay: 0.0500 (0.0500)  time: 0.7414  data: 0.0004  max mem: 57114
Epoch: [172]  [ 80/156]  eta: 0:00:57  lr: 0.003441  min_lr: 0.003441  loss: 3.0990 (3.0288)  weight_decay: 0.0500 (0.0500)  time: 0.7212  data: 0.0004  max mem: 57114
Epoch: [172]  [ 90/156]  eta: 0:00:49  lr: 0.003438  min_lr: 0.003438  loss: 3.1634 (3.0392)  weight_decay: 0.0500 (0.0500)  time: 0.7102  data: 0.0004  max mem: 57114
Epoch: [172]  [100/156]  eta: 0:00:42  lr: 0.003435  min_lr: 0.003435  loss: 3.1634 (3.0293)  weight_decay: 0.0500 (0.0500)  time: 0.7169  data: 0.0005  max mem: 57114
Epoch: [172]  [110/156]  eta: 0:00:34  lr: 0.003432  min_lr: 0.003432  loss: 2.9890 (3.0298)  weight_decay: 0.0500 (0.0500)  time: 0.7123  data: 0.0005  max mem: 57114
Epoch: [172]  [120/156]  eta: 0:00:26  lr: 0.003429  min_lr: 0.003429  loss: 3.2012 (3.0400)  weight_decay: 0.0500 (0.0500)  time: 0.7031  data: 0.0004  max mem: 57114
Epoch: [172]  [130/156]  eta: 0:00:19  lr: 0.003427  min_lr: 0.003427  loss: 3.2615 (3.0380)  weight_decay: 0.0500 (0.0500)  time: 0.6983  data: 0.0008  max mem: 57114
Epoch: [172]  [140/156]  eta: 0:00:11  lr: 0.003424  min_lr: 0.003424  loss: 3.1555 (3.0416)  weight_decay: 0.0500 (0.0500)  time: 0.6861  data: 0.0007  max mem: 57114
Epoch: [172]  [150/156]  eta: 0:00:04  lr: 0.003421  min_lr: 0.003421  loss: 3.0872 (3.0325)  weight_decay: 0.0500 (0.0500)  time: 0.6786  data: 0.0001  max mem: 57114
Epoch: [172]  [155/156]  eta: 0:00:00  lr: 0.003419  min_lr: 0.003419  loss: 3.0014 (3.0351)  weight_decay: 0.0500 (0.0500)  time: 0.6853  data: 0.0001  max mem: 57114
Epoch: [172] Total time: 0:01:54 (0.7326 s / it)
Averaged stats: lr: 0.003419  min_lr: 0.003419  loss: 3.0014 (3.0173)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6568 (0.6568)  acc1: 88.0208 (88.0208)  acc5: 98.9583 (98.9583)  time: 2.1217  data: 1.8656  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9728 (0.8621)  acc1: 83.3333 (80.6905)  acc5: 95.8333 (95.5243)  time: 0.5750  data: 0.3732  max mem: 57114
Test: Total time: 0:00:02 (0.5973 s / it)
* Acc@1 79.570 Acc@5 94.959 loss 0.981
Accuracy of the model on the 50000 test images: 79.6%
Max accuracy: 80.44%
Test:  [0/5]  eta: 0:00:11  loss: 0.7127 (0.7127)  acc1: 82.8125 (82.8125)  acc5: 95.3125 (95.3125)  time: 2.2863  data: 2.0429  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8617 (0.8502)  acc1: 78.1250 (76.3427)  acc5: 94.2708 (91.5601)  time: 0.6082  data: 0.4087  max mem: 57114
Test: Total time: 0:00:03 (0.6195 s / it)
* Acc@1 73.272 Acc@5 91.628 loss 1.036
Accuracy of the model EMA on 50000 test images: 73.3%
Max EMA accuracy: 73.27%
Epoch: [173]  [  0/156]  eta: 0:09:11  lr: 0.003419  min_lr: 0.003419  loss: 2.8071 (2.8071)  weight_decay: 0.0500 (0.0500)  time: 3.5348  data: 2.8857  max mem: 57114
Epoch: [173]  [ 10/156]  eta: 0:02:24  lr: 0.003416  min_lr: 0.003416  loss: 3.1011 (3.0468)  weight_decay: 0.0500 (0.0500)  time: 0.9901  data: 0.2626  max mem: 57114
Epoch: [173]  [ 20/156]  eta: 0:01:57  lr: 0.003414  min_lr: 0.003414  loss: 3.0553 (2.9491)  weight_decay: 0.0500 (0.0500)  time: 0.7308  data: 0.0003  max mem: 57114
Epoch: [173]  [ 30/156]  eta: 0:01:42  lr: 0.003411  min_lr: 0.003411  loss: 2.8406 (2.9548)  weight_decay: 0.0500 (0.0500)  time: 0.7215  data: 0.0003  max mem: 57114
Epoch: [173]  [ 40/156]  eta: 0:01:31  lr: 0.003408  min_lr: 0.003408  loss: 3.0470 (3.0003)  weight_decay: 0.0500 (0.0500)  time: 0.7165  data: 0.0003  max mem: 57114
Epoch: [173]  [ 50/156]  eta: 0:01:22  lr: 0.003405  min_lr: 0.003405  loss: 2.9736 (2.9585)  weight_decay: 0.0500 (0.0500)  time: 0.7190  data: 0.0004  max mem: 57114
Epoch: [173]  [ 60/156]  eta: 0:01:14  lr: 0.003402  min_lr: 0.003402  loss: 3.1537 (3.0043)  weight_decay: 0.0500 (0.0500)  time: 0.7309  data: 0.0004  max mem: 57114
Epoch: [173]  [ 70/156]  eta: 0:01:05  lr: 0.003399  min_lr: 0.003399  loss: 3.1537 (2.9893)  weight_decay: 0.0500 (0.0500)  time: 0.7289  data: 0.0004  max mem: 57114
Epoch: [173]  [ 80/156]  eta: 0:00:57  lr: 0.003396  min_lr: 0.003396  loss: 3.0834 (2.9929)  weight_decay: 0.0500 (0.0500)  time: 0.7214  data: 0.0004  max mem: 57114
Epoch: [173]  [ 90/156]  eta: 0:00:49  lr: 0.003394  min_lr: 0.003394  loss: 3.0834 (2.9889)  weight_decay: 0.0500 (0.0500)  time: 0.7166  data: 0.0004  max mem: 57114
Epoch: [173]  [100/156]  eta: 0:00:41  lr: 0.003391  min_lr: 0.003391  loss: 3.1455 (3.0073)  weight_decay: 0.0500 (0.0500)  time: 0.7016  data: 0.0004  max mem: 57114
Epoch: [173]  [110/156]  eta: 0:00:34  lr: 0.003388  min_lr: 0.003388  loss: 3.2436 (3.0183)  weight_decay: 0.0500 (0.0500)  time: 0.7030  data: 0.0005  max mem: 57114
Epoch: [173]  [120/156]  eta: 0:00:26  lr: 0.003385  min_lr: 0.003385  loss: 3.0754 (3.0251)  weight_decay: 0.0500 (0.0500)  time: 0.7091  data: 0.0004  max mem: 57114
Epoch: [173]  [130/156]  eta: 0:00:19  lr: 0.003382  min_lr: 0.003382  loss: 3.1302 (3.0402)  weight_decay: 0.0500 (0.0500)  time: 0.6953  data: 0.0008  max mem: 57114
Epoch: [173]  [140/156]  eta: 0:00:11  lr: 0.003379  min_lr: 0.003379  loss: 3.3607 (3.0554)  weight_decay: 0.0500 (0.0500)  time: 0.6844  data: 0.0007  max mem: 57114
Epoch: [173]  [150/156]  eta: 0:00:04  lr: 0.003377  min_lr: 0.003377  loss: 3.2492 (3.0472)  weight_decay: 0.0500 (0.0500)  time: 0.6838  data: 0.0001  max mem: 57114
Epoch: [173]  [155/156]  eta: 0:00:00  lr: 0.003375  min_lr: 0.003375  loss: 3.0598 (3.0392)  weight_decay: 0.0500 (0.0500)  time: 0.6837  data: 0.0001  max mem: 57114
Epoch: [173] Total time: 0:01:54 (0.7309 s / it)
Averaged stats: lr: 0.003375  min_lr: 0.003375  loss: 3.0598 (3.0152)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7167 (0.7167)  acc1: 88.0208 (88.0208)  acc5: 98.4375 (98.4375)  time: 2.0970  data: 1.8407  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9425 (0.9099)  acc1: 82.2917 (79.7954)  acc5: 96.3542 (95.5243)  time: 0.5701  data: 0.3682  max mem: 57114
Test: Total time: 0:00:02 (0.5926 s / it)
* Acc@1 80.055 Acc@5 95.231 loss 0.975
Accuracy of the model on the 50000 test images: 80.1%
Max accuracy: 80.44%
Test:  [0/5]  eta: 0:00:11  loss: 0.7070 (0.7070)  acc1: 83.3333 (83.3333)  acc5: 95.3125 (95.3125)  time: 2.3390  data: 2.0954  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8546 (0.8419)  acc1: 78.1250 (76.5985)  acc5: 94.7917 (91.6880)  time: 0.6184  data: 0.4192  max mem: 57114
Test: Total time: 0:00:03 (0.6313 s / it)
* Acc@1 73.430 Acc@5 91.722 loss 1.026
Accuracy of the model EMA on 50000 test images: 73.4%
Max EMA accuracy: 73.43%
Epoch: [174]  [  0/156]  eta: 0:06:29  lr: 0.003375  min_lr: 0.003375  loss: 2.8825 (2.8825)  weight_decay: 0.0500 (0.0500)  time: 2.4936  data: 1.8433  max mem: 57114
Epoch: [174]  [ 10/156]  eta: 0:02:06  lr: 0.003372  min_lr: 0.003372  loss: 3.2770 (3.1858)  weight_decay: 0.0500 (0.0500)  time: 0.8679  data: 0.1678  max mem: 57114
Epoch: [174]  [ 20/156]  eta: 0:01:48  lr: 0.003369  min_lr: 0.003369  loss: 3.2796 (3.1915)  weight_decay: 0.0500 (0.0500)  time: 0.7148  data: 0.0003  max mem: 57114
Epoch: [174]  [ 30/156]  eta: 0:01:37  lr: 0.003366  min_lr: 0.003366  loss: 3.0063 (3.1082)  weight_decay: 0.0500 (0.0500)  time: 0.7253  data: 0.0003  max mem: 57114
Epoch: [174]  [ 40/156]  eta: 0:01:28  lr: 0.003363  min_lr: 0.003363  loss: 2.9361 (3.0619)  weight_decay: 0.0500 (0.0500)  time: 0.7264  data: 0.0004  max mem: 57114
Epoch: [174]  [ 50/156]  eta: 0:01:20  lr: 0.003361  min_lr: 0.003361  loss: 3.1434 (3.0688)  weight_decay: 0.0500 (0.0500)  time: 0.7339  data: 0.0004  max mem: 57114
Epoch: [174]  [ 60/156]  eta: 0:01:12  lr: 0.003358  min_lr: 0.003358  loss: 3.2407 (3.0773)  weight_decay: 0.0500 (0.0500)  time: 0.7365  data: 0.0004  max mem: 57114
Epoch: [174]  [ 70/156]  eta: 0:01:04  lr: 0.003355  min_lr: 0.003355  loss: 2.9681 (3.0330)  weight_decay: 0.0500 (0.0500)  time: 0.7133  data: 0.0003  max mem: 57114
Epoch: [174]  [ 80/156]  eta: 0:00:56  lr: 0.003352  min_lr: 0.003352  loss: 2.8364 (3.0147)  weight_decay: 0.0500 (0.0500)  time: 0.7052  data: 0.0004  max mem: 57114
Epoch: [174]  [ 90/156]  eta: 0:00:48  lr: 0.003349  min_lr: 0.003349  loss: 2.8364 (2.9915)  weight_decay: 0.0500 (0.0500)  time: 0.7094  data: 0.0004  max mem: 57114
Epoch: [174]  [100/156]  eta: 0:00:41  lr: 0.003346  min_lr: 0.003346  loss: 2.6797 (2.9736)  weight_decay: 0.0500 (0.0500)  time: 0.7152  data: 0.0003  max mem: 57114
Epoch: [174]  [110/156]  eta: 0:00:33  lr: 0.003344  min_lr: 0.003344  loss: 3.1858 (2.9964)  weight_decay: 0.0500 (0.0500)  time: 0.7103  data: 0.0004  max mem: 57114
Epoch: [174]  [120/156]  eta: 0:00:26  lr: 0.003341  min_lr: 0.003341  loss: 3.2130 (3.0202)  weight_decay: 0.0500 (0.0500)  time: 0.7080  data: 0.0004  max mem: 57114
Epoch: [174]  [130/156]  eta: 0:00:18  lr: 0.003338  min_lr: 0.003338  loss: 3.1855 (3.0230)  weight_decay: 0.0500 (0.0500)  time: 0.7050  data: 0.0009  max mem: 57114
Epoch: [174]  [140/156]  eta: 0:00:11  lr: 0.003335  min_lr: 0.003335  loss: 3.0712 (3.0224)  weight_decay: 0.0500 (0.0500)  time: 0.6890  data: 0.0007  max mem: 57114
Epoch: [174]  [150/156]  eta: 0:00:04  lr: 0.003332  min_lr: 0.003332  loss: 3.1293 (3.0166)  weight_decay: 0.0500 (0.0500)  time: 0.6847  data: 0.0001  max mem: 57114
Epoch: [174]  [155/156]  eta: 0:00:00  lr: 0.003331  min_lr: 0.003331  loss: 3.1293 (3.0157)  weight_decay: 0.0500 (0.0500)  time: 0.6804  data: 0.0001  max mem: 57114
Epoch: [174] Total time: 0:01:52 (0.7236 s / it)
Averaged stats: lr: 0.003331  min_lr: 0.003331  loss: 3.1293 (3.0042)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7401 (0.7401)  acc1: 88.5417 (88.5417)  acc5: 98.9583 (98.9583)  time: 2.0382  data: 1.7820  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9828 (0.8961)  acc1: 79.6875 (80.9463)  acc5: 96.3542 (96.0358)  time: 0.5584  data: 0.3565  max mem: 57114
Test: Total time: 0:00:02 (0.5792 s / it)
* Acc@1 80.393 Acc@5 95.402 loss 0.967
Accuracy of the model on the 50000 test images: 80.4%
Max accuracy: 80.44%
Test:  [0/5]  eta: 0:00:11  loss: 0.7018 (0.7018)  acc1: 83.3333 (83.3333)  acc5: 95.3125 (95.3125)  time: 2.3414  data: 2.0977  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8479 (0.8342)  acc1: 78.1250 (76.7263)  acc5: 94.7917 (91.6880)  time: 0.6190  data: 0.4196  max mem: 57114
Test: Total time: 0:00:03 (0.6322 s / it)
* Acc@1 73.599 Acc@5 91.824 loss 1.016
Accuracy of the model EMA on 50000 test images: 73.6%
Max EMA accuracy: 73.60%
Epoch: [175]  [  0/156]  eta: 0:07:16  lr: 0.003331  min_lr: 0.003331  loss: 3.0630 (3.0630)  weight_decay: 0.0500 (0.0500)  time: 2.7989  data: 2.1511  max mem: 57114
Epoch: [175]  [ 10/156]  eta: 0:02:14  lr: 0.003328  min_lr: 0.003328  loss: 3.0134 (2.8622)  weight_decay: 0.0500 (0.0500)  time: 0.9213  data: 0.1958  max mem: 57114
Epoch: [175]  [ 20/156]  eta: 0:01:52  lr: 0.003325  min_lr: 0.003325  loss: 3.0597 (2.9750)  weight_decay: 0.0500 (0.0500)  time: 0.7264  data: 0.0003  max mem: 57114
Epoch: [175]  [ 30/156]  eta: 0:01:40  lr: 0.003322  min_lr: 0.003322  loss: 3.2069 (3.0256)  weight_decay: 0.0500 (0.0500)  time: 0.7279  data: 0.0003  max mem: 57114
Epoch: [175]  [ 40/156]  eta: 0:01:29  lr: 0.003319  min_lr: 0.003319  loss: 3.1317 (2.9857)  weight_decay: 0.0500 (0.0500)  time: 0.7191  data: 0.0004  max mem: 57114
Epoch: [175]  [ 50/156]  eta: 0:01:20  lr: 0.003316  min_lr: 0.003316  loss: 2.7596 (3.0012)  weight_decay: 0.0500 (0.0500)  time: 0.7095  data: 0.0004  max mem: 57114
Epoch: [175]  [ 60/156]  eta: 0:01:12  lr: 0.003314  min_lr: 0.003314  loss: 3.2091 (2.9930)  weight_decay: 0.0500 (0.0500)  time: 0.7193  data: 0.0004  max mem: 57114
Epoch: [175]  [ 70/156]  eta: 0:01:04  lr: 0.003311  min_lr: 0.003311  loss: 3.0359 (2.9748)  weight_decay: 0.0500 (0.0500)  time: 0.7148  data: 0.0004  max mem: 57114
Epoch: [175]  [ 80/156]  eta: 0:00:56  lr: 0.003308  min_lr: 0.003308  loss: 3.0359 (2.9951)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0004  max mem: 57114
Epoch: [175]  [ 90/156]  eta: 0:00:48  lr: 0.003305  min_lr: 0.003305  loss: 3.1828 (3.0047)  weight_decay: 0.0500 (0.0500)  time: 0.7096  data: 0.0003  max mem: 57114
Epoch: [175]  [100/156]  eta: 0:00:41  lr: 0.003302  min_lr: 0.003302  loss: 3.1464 (3.0142)  weight_decay: 0.0500 (0.0500)  time: 0.7093  data: 0.0004  max mem: 57114
Epoch: [175]  [110/156]  eta: 0:00:33  lr: 0.003299  min_lr: 0.003299  loss: 3.0667 (2.9980)  weight_decay: 0.0500 (0.0500)  time: 0.7061  data: 0.0005  max mem: 57114
Epoch: [175]  [120/156]  eta: 0:00:26  lr: 0.003297  min_lr: 0.003297  loss: 3.0882 (2.9935)  weight_decay: 0.0500 (0.0500)  time: 0.7159  data: 0.0004  max mem: 57114
Epoch: [175]  [130/156]  eta: 0:00:18  lr: 0.003294  min_lr: 0.003294  loss: 3.0451 (2.9869)  weight_decay: 0.0500 (0.0500)  time: 0.7062  data: 0.0008  max mem: 57114
Epoch: [175]  [140/156]  eta: 0:00:11  lr: 0.003291  min_lr: 0.003291  loss: 3.0451 (2.9921)  weight_decay: 0.0500 (0.0500)  time: 0.6878  data: 0.0007  max mem: 57114
Epoch: [175]  [150/156]  eta: 0:00:04  lr: 0.003288  min_lr: 0.003288  loss: 2.9817 (2.9846)  weight_decay: 0.0500 (0.0500)  time: 0.6829  data: 0.0001  max mem: 57114
Epoch: [175]  [155/156]  eta: 0:00:00  lr: 0.003287  min_lr: 0.003287  loss: 2.9992 (2.9873)  weight_decay: 0.0500 (0.0500)  time: 0.6844  data: 0.0001  max mem: 57114
Epoch: [175] Total time: 0:01:53 (0.7250 s / it)
Averaged stats: lr: 0.003287  min_lr: 0.003287  loss: 2.9992 (2.9992)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7791 (0.7791)  acc1: 86.9792 (86.9792)  acc5: 97.3958 (97.3958)  time: 2.0934  data: 1.8379  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9032 (0.8556)  acc1: 82.2917 (81.8414)  acc5: 97.3958 (96.2916)  time: 0.5693  data: 0.3676  max mem: 57114
Test: Total time: 0:00:02 (0.5913 s / it)
* Acc@1 80.305 Acc@5 95.322 loss 0.952
Accuracy of the model on the 50000 test images: 80.3%
Max accuracy: 80.44%
Test:  [0/5]  eta: 0:00:11  loss: 0.6962 (0.6962)  acc1: 83.3333 (83.3333)  acc5: 95.3125 (95.3125)  time: 2.2630  data: 2.0195  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8412 (0.8265)  acc1: 78.1250 (76.5985)  acc5: 94.7917 (91.6880)  time: 0.6032  data: 0.4040  max mem: 57114
Test: Total time: 0:00:03 (0.6155 s / it)
* Acc@1 73.781 Acc@5 91.922 loss 1.006
Accuracy of the model EMA on 50000 test images: 73.8%
Max EMA accuracy: 73.78%
Epoch: [176]  [  0/156]  eta: 0:06:55  lr: 0.003286  min_lr: 0.003286  loss: 1.9077 (1.9077)  weight_decay: 0.0500 (0.0500)  time: 2.6619  data: 2.0110  max mem: 57114
Epoch: [176]  [ 10/156]  eta: 0:02:14  lr: 0.003284  min_lr: 0.003284  loss: 2.7810 (2.7117)  weight_decay: 0.0500 (0.0500)  time: 0.9180  data: 0.1831  max mem: 57114
Epoch: [176]  [ 20/156]  eta: 0:01:51  lr: 0.003281  min_lr: 0.003281  loss: 2.8669 (2.8443)  weight_decay: 0.0500 (0.0500)  time: 0.7311  data: 0.0003  max mem: 57114
Epoch: [176]  [ 30/156]  eta: 0:01:40  lr: 0.003278  min_lr: 0.003278  loss: 3.0628 (2.9230)  weight_decay: 0.0500 (0.0500)  time: 0.7280  data: 0.0003  max mem: 57114
Epoch: [176]  [ 40/156]  eta: 0:01:29  lr: 0.003275  min_lr: 0.003275  loss: 3.2039 (2.9500)  weight_decay: 0.0500 (0.0500)  time: 0.7263  data: 0.0003  max mem: 57114
Epoch: [176]  [ 50/156]  eta: 0:01:20  lr: 0.003272  min_lr: 0.003272  loss: 2.9913 (2.9456)  weight_decay: 0.0500 (0.0500)  time: 0.7060  data: 0.0004  max mem: 57114
Epoch: [176]  [ 60/156]  eta: 0:01:12  lr: 0.003269  min_lr: 0.003269  loss: 2.9160 (2.9281)  weight_decay: 0.0500 (0.0500)  time: 0.7167  data: 0.0004  max mem: 57114
Epoch: [176]  [ 70/156]  eta: 0:01:04  lr: 0.003267  min_lr: 0.003267  loss: 2.9784 (2.9417)  weight_decay: 0.0500 (0.0500)  time: 0.7205  data: 0.0004  max mem: 57114
Epoch: [176]  [ 80/156]  eta: 0:00:56  lr: 0.003264  min_lr: 0.003264  loss: 3.1099 (2.9669)  weight_decay: 0.0500 (0.0500)  time: 0.7116  data: 0.0004  max mem: 57114
Epoch: [176]  [ 90/156]  eta: 0:00:48  lr: 0.003261  min_lr: 0.003261  loss: 3.2209 (2.9719)  weight_decay: 0.0500 (0.0500)  time: 0.7123  data: 0.0004  max mem: 57114
Epoch: [176]  [100/156]  eta: 0:00:41  lr: 0.003258  min_lr: 0.003258  loss: 3.2394 (2.9915)  weight_decay: 0.0500 (0.0500)  time: 0.7075  data: 0.0004  max mem: 57114
Epoch: [176]  [110/156]  eta: 0:00:33  lr: 0.003255  min_lr: 0.003255  loss: 3.0928 (2.9760)  weight_decay: 0.0500 (0.0500)  time: 0.7064  data: 0.0004  max mem: 57114
Epoch: [176]  [120/156]  eta: 0:00:26  lr: 0.003252  min_lr: 0.003252  loss: 2.9059 (2.9630)  weight_decay: 0.0500 (0.0500)  time: 0.6984  data: 0.0004  max mem: 57114
Epoch: [176]  [130/156]  eta: 0:00:18  lr: 0.003250  min_lr: 0.003250  loss: 3.0220 (2.9778)  weight_decay: 0.0500 (0.0500)  time: 0.6939  data: 0.0009  max mem: 57114
Epoch: [176]  [140/156]  eta: 0:00:11  lr: 0.003247  min_lr: 0.003247  loss: 3.0051 (2.9784)  weight_decay: 0.0500 (0.0500)  time: 0.6899  data: 0.0007  max mem: 57114
Epoch: [176]  [150/156]  eta: 0:00:04  lr: 0.003244  min_lr: 0.003244  loss: 2.9380 (2.9785)  weight_decay: 0.0500 (0.0500)  time: 0.6891  data: 0.0001  max mem: 57114
Epoch: [176]  [155/156]  eta: 0:00:00  lr: 0.003243  min_lr: 0.003243  loss: 3.0051 (2.9821)  weight_decay: 0.0500 (0.0500)  time: 0.6882  data: 0.0001  max mem: 57114
Epoch: [176] Total time: 0:01:52 (0.7243 s / it)
Averaged stats: lr: 0.003243  min_lr: 0.003243  loss: 3.0051 (2.9956)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8918 (0.8918)  acc1: 83.3333 (83.3333)  acc5: 97.9167 (97.9167)  time: 2.1975  data: 1.9413  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0214 (0.9949)  acc1: 79.1667 (79.4118)  acc5: 96.8750 (95.2685)  time: 0.5902  data: 0.3883  max mem: 57114
Test: Total time: 0:00:03 (0.6155 s / it)
* Acc@1 80.483 Acc@5 95.406 loss 1.057
Accuracy of the model on the 50000 test images: 80.5%
Max accuracy: 80.48%
Test:  [0/5]  eta: 0:00:10  loss: 0.6906 (0.6906)  acc1: 82.8125 (82.8125)  acc5: 95.3125 (95.3125)  time: 2.0001  data: 1.7567  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8345 (0.8189)  acc1: 78.1250 (76.5985)  acc5: 94.7917 (91.6880)  time: 0.5507  data: 0.3514  max mem: 57114
Test: Total time: 0:00:02 (0.5607 s / it)
* Acc@1 73.979 Acc@5 92.002 loss 0.997
Accuracy of the model EMA on 50000 test images: 74.0%
Max EMA accuracy: 73.98%
Epoch: [177]  [  0/156]  eta: 0:07:21  lr: 0.003242  min_lr: 0.003242  loss: 3.1074 (3.1074)  weight_decay: 0.0500 (0.0500)  time: 2.8287  data: 2.1781  max mem: 57114
Epoch: [177]  [ 10/156]  eta: 0:02:14  lr: 0.003239  min_lr: 0.003239  loss: 2.5245 (2.7406)  weight_decay: 0.0500 (0.0500)  time: 0.9182  data: 0.1983  max mem: 57114
Epoch: [177]  [ 20/156]  eta: 0:01:52  lr: 0.003237  min_lr: 0.003237  loss: 3.0626 (2.9111)  weight_decay: 0.0500 (0.0500)  time: 0.7249  data: 0.0004  max mem: 57114
Epoch: [177]  [ 30/156]  eta: 0:01:40  lr: 0.003234  min_lr: 0.003234  loss: 3.0713 (2.9133)  weight_decay: 0.0500 (0.0500)  time: 0.7269  data: 0.0005  max mem: 57114
Epoch: [177]  [ 40/156]  eta: 0:01:30  lr: 0.003231  min_lr: 0.003231  loss: 2.8920 (2.9180)  weight_decay: 0.0500 (0.0500)  time: 0.7363  data: 0.0006  max mem: 57114
Epoch: [177]  [ 50/156]  eta: 0:01:21  lr: 0.003228  min_lr: 0.003228  loss: 3.0970 (2.9737)  weight_decay: 0.0500 (0.0500)  time: 0.7360  data: 0.0007  max mem: 57114
Epoch: [177]  [ 60/156]  eta: 0:01:14  lr: 0.003225  min_lr: 0.003225  loss: 3.1615 (2.9770)  weight_decay: 0.0500 (0.0500)  time: 0.7532  data: 0.0008  max mem: 57114
Epoch: [177]  [ 70/156]  eta: 0:01:05  lr: 0.003222  min_lr: 0.003222  loss: 3.0303 (2.9472)  weight_decay: 0.0500 (0.0500)  time: 0.7489  data: 0.0008  max mem: 57114
Epoch: [177]  [ 80/156]  eta: 0:00:57  lr: 0.003220  min_lr: 0.003220  loss: 3.1106 (2.9560)  weight_decay: 0.0500 (0.0500)  time: 0.7164  data: 0.0007  max mem: 57114
Epoch: [177]  [ 90/156]  eta: 0:00:49  lr: 0.003217  min_lr: 0.003217  loss: 3.2106 (2.9735)  weight_decay: 0.0500 (0.0500)  time: 0.6983  data: 0.0007  max mem: 57114
Epoch: [177]  [100/156]  eta: 0:00:41  lr: 0.003214  min_lr: 0.003214  loss: 3.0666 (2.9550)  weight_decay: 0.0500 (0.0500)  time: 0.6938  data: 0.0006  max mem: 57114
Epoch: [177]  [110/156]  eta: 0:00:34  lr: 0.003211  min_lr: 0.003211  loss: 2.9995 (2.9691)  weight_decay: 0.0500 (0.0500)  time: 0.7068  data: 0.0004  max mem: 57114
Epoch: [177]  [120/156]  eta: 0:00:26  lr: 0.003208  min_lr: 0.003208  loss: 3.0119 (2.9715)  weight_decay: 0.0500 (0.0500)  time: 0.7137  data: 0.0004  max mem: 57114
Epoch: [177]  [130/156]  eta: 0:00:19  lr: 0.003206  min_lr: 0.003206  loss: 3.0014 (2.9557)  weight_decay: 0.0500 (0.0500)  time: 0.7104  data: 0.0010  max mem: 57114
Epoch: [177]  [140/156]  eta: 0:00:11  lr: 0.003203  min_lr: 0.003203  loss: 3.0014 (2.9568)  weight_decay: 0.0500 (0.0500)  time: 0.6958  data: 0.0008  max mem: 57114
Epoch: [177]  [150/156]  eta: 0:00:04  lr: 0.003200  min_lr: 0.003200  loss: 3.1209 (2.9587)  weight_decay: 0.0500 (0.0500)  time: 0.6819  data: 0.0001  max mem: 57114
Epoch: [177]  [155/156]  eta: 0:00:00  lr: 0.003199  min_lr: 0.003199  loss: 2.9150 (2.9548)  weight_decay: 0.0500 (0.0500)  time: 0.6774  data: 0.0001  max mem: 57114
Epoch: [177] Total time: 0:01:53 (0.7306 s / it)
Averaged stats: lr: 0.003199  min_lr: 0.003199  loss: 2.9150 (2.9926)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6839 (0.6839)  acc1: 88.5417 (88.5417)  acc5: 97.9167 (97.9167)  time: 2.1390  data: 1.8831  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9398 (0.8365)  acc1: 80.2083 (81.3299)  acc5: 97.3958 (96.0358)  time: 0.5786  data: 0.3767  max mem: 57114
Test: Total time: 0:00:03 (0.6022 s / it)
* Acc@1 80.353 Acc@5 95.508 loss 0.927
Accuracy of the model on the 50000 test images: 80.4%
Max accuracy: 80.48%
Test:  [0/5]  eta: 0:00:11  loss: 0.6858 (0.6858)  acc1: 82.8125 (82.8125)  acc5: 95.3125 (95.3125)  time: 2.2828  data: 2.0393  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8280 (0.8120)  acc1: 78.6458 (76.7263)  acc5: 94.7917 (91.6880)  time: 0.6072  data: 0.4079  max mem: 57114
Test: Total time: 0:00:03 (0.6192 s / it)
* Acc@1 74.153 Acc@5 92.080 loss 0.988
Accuracy of the model EMA on 50000 test images: 74.2%
Max EMA accuracy: 74.15%
Epoch: [178]  [  0/156]  eta: 0:08:22  lr: 0.003198  min_lr: 0.003198  loss: 3.6086 (3.6086)  weight_decay: 0.0500 (0.0500)  time: 3.2194  data: 2.5636  max mem: 57114
Epoch: [178]  [ 10/156]  eta: 0:02:18  lr: 0.003195  min_lr: 0.003195  loss: 3.0635 (3.0821)  weight_decay: 0.0500 (0.0500)  time: 0.9503  data: 0.2334  max mem: 57114
Epoch: [178]  [ 20/156]  eta: 0:01:54  lr: 0.003193  min_lr: 0.003193  loss: 3.0278 (3.0377)  weight_decay: 0.0500 (0.0500)  time: 0.7240  data: 0.0004  max mem: 57114
Epoch: [178]  [ 30/156]  eta: 0:01:41  lr: 0.003190  min_lr: 0.003190  loss: 3.0253 (3.0065)  weight_decay: 0.0500 (0.0500)  time: 0.7296  data: 0.0004  max mem: 57114
Epoch: [178]  [ 40/156]  eta: 0:01:32  lr: 0.003187  min_lr: 0.003187  loss: 3.1269 (3.0501)  weight_decay: 0.0500 (0.0500)  time: 0.7437  data: 0.0003  max mem: 57114
Epoch: [178]  [ 50/156]  eta: 0:01:22  lr: 0.003184  min_lr: 0.003184  loss: 3.2104 (3.0450)  weight_decay: 0.0500 (0.0500)  time: 0.7385  data: 0.0004  max mem: 57114
Epoch: [178]  [ 60/156]  eta: 0:01:13  lr: 0.003181  min_lr: 0.003181  loss: 3.1909 (3.0538)  weight_decay: 0.0500 (0.0500)  time: 0.7129  data: 0.0004  max mem: 57114
Epoch: [178]  [ 70/156]  eta: 0:01:05  lr: 0.003179  min_lr: 0.003179  loss: 3.1838 (3.0480)  weight_decay: 0.0500 (0.0500)  time: 0.7101  data: 0.0004  max mem: 57114
Epoch: [178]  [ 80/156]  eta: 0:00:57  lr: 0.003176  min_lr: 0.003176  loss: 3.2744 (3.0703)  weight_decay: 0.0500 (0.0500)  time: 0.7060  data: 0.0004  max mem: 57114
Epoch: [178]  [ 90/156]  eta: 0:00:49  lr: 0.003173  min_lr: 0.003173  loss: 3.2847 (3.0855)  weight_decay: 0.0500 (0.0500)  time: 0.6994  data: 0.0004  max mem: 57114
Epoch: [178]  [100/156]  eta: 0:00:41  lr: 0.003170  min_lr: 0.003170  loss: 3.2565 (3.0821)  weight_decay: 0.0500 (0.0500)  time: 0.7093  data: 0.0004  max mem: 57114
Epoch: [178]  [110/156]  eta: 0:00:34  lr: 0.003167  min_lr: 0.003167  loss: 3.1840 (3.0849)  weight_decay: 0.0500 (0.0500)  time: 0.7147  data: 0.0004  max mem: 57114
Epoch: [178]  [120/156]  eta: 0:00:26  lr: 0.003164  min_lr: 0.003164  loss: 3.0880 (3.0683)  weight_decay: 0.0500 (0.0500)  time: 0.7099  data: 0.0004  max mem: 57114
Epoch: [178]  [130/156]  eta: 0:00:19  lr: 0.003162  min_lr: 0.003162  loss: 2.9327 (3.0486)  weight_decay: 0.0500 (0.0500)  time: 0.7011  data: 0.0008  max mem: 57114
Epoch: [178]  [140/156]  eta: 0:00:11  lr: 0.003159  min_lr: 0.003159  loss: 2.8341 (3.0312)  weight_decay: 0.0500 (0.0500)  time: 0.6917  data: 0.0007  max mem: 57114
Epoch: [178]  [150/156]  eta: 0:00:04  lr: 0.003156  min_lr: 0.003156  loss: 2.8482 (3.0294)  weight_decay: 0.0500 (0.0500)  time: 0.6794  data: 0.0001  max mem: 57114
Epoch: [178]  [155/156]  eta: 0:00:00  lr: 0.003155  min_lr: 0.003155  loss: 2.9926 (3.0319)  weight_decay: 0.0500 (0.0500)  time: 0.6808  data: 0.0001  max mem: 57114
Epoch: [178] Total time: 0:01:53 (0.7288 s / it)
Averaged stats: lr: 0.003155  min_lr: 0.003155  loss: 2.9926 (2.9841)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7293 (0.7293)  acc1: 88.0208 (88.0208)  acc5: 98.9583 (98.9583)  time: 2.0703  data: 1.8142  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0313 (0.8931)  acc1: 81.2500 (81.4578)  acc5: 95.3125 (95.5243)  time: 0.5648  data: 0.3629  max mem: 57114
Test: Total time: 0:00:02 (0.5868 s / it)
* Acc@1 80.209 Acc@5 95.432 loss 0.980
Accuracy of the model on the 50000 test images: 80.2%
Max accuracy: 80.48%
Test:  [0/5]  eta: 0:00:11  loss: 0.6806 (0.6806)  acc1: 82.8125 (82.8125)  acc5: 95.3125 (95.3125)  time: 2.2940  data: 2.0504  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8217 (0.8049)  acc1: 78.6458 (76.9821)  acc5: 94.7917 (91.6880)  time: 0.6095  data: 0.4102  max mem: 57114
Test: Total time: 0:00:03 (0.6214 s / it)
* Acc@1 74.299 Acc@5 92.162 loss 0.980
Accuracy of the model EMA on 50000 test images: 74.3%
Max EMA accuracy: 74.30%
Epoch: [179]  [  0/156]  eta: 0:06:21  lr: 0.003154  min_lr: 0.003154  loss: 3.5874 (3.5874)  weight_decay: 0.0500 (0.0500)  time: 2.4483  data: 1.7991  max mem: 57114
Epoch: [179]  [ 10/156]  eta: 0:02:08  lr: 0.003152  min_lr: 0.003152  loss: 3.1798 (3.0596)  weight_decay: 0.0500 (0.0500)  time: 0.8799  data: 0.1639  max mem: 57114
Epoch: [179]  [ 20/156]  eta: 0:01:49  lr: 0.003149  min_lr: 0.003149  loss: 2.9158 (2.9264)  weight_decay: 0.0500 (0.0500)  time: 0.7205  data: 0.0003  max mem: 57114
Epoch: [179]  [ 30/156]  eta: 0:01:37  lr: 0.003146  min_lr: 0.003146  loss: 2.9158 (2.9310)  weight_decay: 0.0500 (0.0500)  time: 0.7196  data: 0.0004  max mem: 57114
Epoch: [179]  [ 40/156]  eta: 0:01:28  lr: 0.003143  min_lr: 0.003143  loss: 3.0826 (2.9890)  weight_decay: 0.0500 (0.0500)  time: 0.7270  data: 0.0004  max mem: 57114
Epoch: [179]  [ 50/156]  eta: 0:01:20  lr: 0.003140  min_lr: 0.003140  loss: 3.2097 (3.0217)  weight_decay: 0.0500 (0.0500)  time: 0.7346  data: 0.0004  max mem: 57114
Epoch: [179]  [ 60/156]  eta: 0:01:12  lr: 0.003137  min_lr: 0.003137  loss: 3.2097 (2.9888)  weight_decay: 0.0500 (0.0500)  time: 0.7193  data: 0.0004  max mem: 57114
Epoch: [179]  [ 70/156]  eta: 0:01:04  lr: 0.003135  min_lr: 0.003135  loss: 2.7664 (2.9844)  weight_decay: 0.0500 (0.0500)  time: 0.7075  data: 0.0004  max mem: 57114
Epoch: [179]  [ 80/156]  eta: 0:00:56  lr: 0.003132  min_lr: 0.003132  loss: 3.0625 (2.9974)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0004  max mem: 57114
Epoch: [179]  [ 90/156]  eta: 0:00:48  lr: 0.003129  min_lr: 0.003129  loss: 3.1535 (2.9863)  weight_decay: 0.0500 (0.0500)  time: 0.7112  data: 0.0004  max mem: 57114
Epoch: [179]  [100/156]  eta: 0:00:41  lr: 0.003126  min_lr: 0.003126  loss: 3.1340 (2.9784)  weight_decay: 0.0500 (0.0500)  time: 0.7089  data: 0.0003  max mem: 57114
Epoch: [179]  [110/156]  eta: 0:00:33  lr: 0.003123  min_lr: 0.003123  loss: 3.1340 (2.9759)  weight_decay: 0.0500 (0.0500)  time: 0.7149  data: 0.0003  max mem: 57114
Epoch: [179]  [120/156]  eta: 0:00:26  lr: 0.003121  min_lr: 0.003121  loss: 2.9915 (2.9712)  weight_decay: 0.0500 (0.0500)  time: 0.7124  data: 0.0003  max mem: 57114
Epoch: [179]  [130/156]  eta: 0:00:18  lr: 0.003118  min_lr: 0.003118  loss: 2.9075 (2.9651)  weight_decay: 0.0500 (0.0500)  time: 0.6960  data: 0.0008  max mem: 57114
Epoch: [179]  [140/156]  eta: 0:00:11  lr: 0.003115  min_lr: 0.003115  loss: 2.9075 (2.9639)  weight_decay: 0.0500 (0.0500)  time: 0.6848  data: 0.0007  max mem: 57114
Epoch: [179]  [150/156]  eta: 0:00:04  lr: 0.003112  min_lr: 0.003112  loss: 3.0726 (2.9617)  weight_decay: 0.0500 (0.0500)  time: 0.6818  data: 0.0001  max mem: 57114
Epoch: [179]  [155/156]  eta: 0:00:00  lr: 0.003111  min_lr: 0.003111  loss: 3.1045 (2.9730)  weight_decay: 0.0500 (0.0500)  time: 0.6827  data: 0.0001  max mem: 57114
Epoch: [179] Total time: 0:01:52 (0.7228 s / it)
Averaged stats: lr: 0.003111  min_lr: 0.003111  loss: 3.1045 (2.9811)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7158 (0.7158)  acc1: 88.0208 (88.0208)  acc5: 97.3958 (97.3958)  time: 2.0945  data: 1.8391  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8807 (0.8380)  acc1: 82.2917 (81.7136)  acc5: 97.3958 (96.4194)  time: 0.5696  data: 0.3679  max mem: 57114
Test: Total time: 0:00:02 (0.5919 s / it)
* Acc@1 80.507 Acc@5 95.448 loss 0.941
Accuracy of the model on the 50000 test images: 80.5%
Max accuracy: 80.51%
Test:  [0/5]  eta: 0:00:10  loss: 0.6757 (0.6757)  acc1: 82.2917 (82.2917)  acc5: 95.3125 (95.3125)  time: 2.0932  data: 1.8498  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8161 (0.7981)  acc1: 78.6458 (76.9821)  acc5: 94.7917 (91.8159)  time: 0.5696  data: 0.3702  max mem: 57114
Test: Total time: 0:00:02 (0.5843 s / it)
* Acc@1 74.500 Acc@5 92.259 loss 0.971
Accuracy of the model EMA on 50000 test images: 74.5%
Max EMA accuracy: 74.50%
Epoch: [180]  [  0/156]  eta: 0:08:25  lr: 0.003111  min_lr: 0.003111  loss: 2.5864 (2.5864)  weight_decay: 0.0500 (0.0500)  time: 3.2373  data: 2.5907  max mem: 57114
Epoch: [180]  [ 10/156]  eta: 0:02:22  lr: 0.003108  min_lr: 0.003108  loss: 3.2952 (3.2368)  weight_decay: 0.0500 (0.0500)  time: 0.9737  data: 0.2358  max mem: 57114
Epoch: [180]  [ 20/156]  eta: 0:01:55  lr: 0.003105  min_lr: 0.003105  loss: 3.1887 (2.9850)  weight_decay: 0.0500 (0.0500)  time: 0.7316  data: 0.0003  max mem: 57114
Epoch: [180]  [ 30/156]  eta: 0:01:42  lr: 0.003102  min_lr: 0.003102  loss: 3.0194 (2.9862)  weight_decay: 0.0500 (0.0500)  time: 0.7285  data: 0.0003  max mem: 57114
Epoch: [180]  [ 40/156]  eta: 0:01:31  lr: 0.003099  min_lr: 0.003099  loss: 3.0194 (2.9656)  weight_decay: 0.0500 (0.0500)  time: 0.7290  data: 0.0004  max mem: 57114
Epoch: [180]  [ 50/156]  eta: 0:01:22  lr: 0.003097  min_lr: 0.003097  loss: 2.9987 (2.9606)  weight_decay: 0.0500 (0.0500)  time: 0.7302  data: 0.0004  max mem: 57114
Epoch: [180]  [ 60/156]  eta: 0:01:14  lr: 0.003094  min_lr: 0.003094  loss: 2.9987 (2.9151)  weight_decay: 0.0500 (0.0500)  time: 0.7495  data: 0.0004  max mem: 57114
Epoch: [180]  [ 70/156]  eta: 0:01:06  lr: 0.003091  min_lr: 0.003091  loss: 3.1776 (2.9666)  weight_decay: 0.0500 (0.0500)  time: 0.7385  data: 0.0004  max mem: 57114
Epoch: [180]  [ 80/156]  eta: 0:00:57  lr: 0.003088  min_lr: 0.003088  loss: 3.2019 (2.9545)  weight_decay: 0.0500 (0.0500)  time: 0.7115  data: 0.0004  max mem: 57114
Epoch: [180]  [ 90/156]  eta: 0:00:49  lr: 0.003085  min_lr: 0.003085  loss: 3.0777 (2.9635)  weight_decay: 0.0500 (0.0500)  time: 0.7053  data: 0.0004  max mem: 57114
Epoch: [180]  [100/156]  eta: 0:00:42  lr: 0.003083  min_lr: 0.003083  loss: 3.0777 (2.9786)  weight_decay: 0.0500 (0.0500)  time: 0.7082  data: 0.0004  max mem: 57114
Epoch: [180]  [110/156]  eta: 0:00:34  lr: 0.003080  min_lr: 0.003080  loss: 3.0675 (2.9774)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0004  max mem: 57114
Epoch: [180]  [120/156]  eta: 0:00:26  lr: 0.003077  min_lr: 0.003077  loss: 3.0675 (2.9662)  weight_decay: 0.0500 (0.0500)  time: 0.7155  data: 0.0003  max mem: 57114
Epoch: [180]  [130/156]  eta: 0:00:19  lr: 0.003074  min_lr: 0.003074  loss: 2.9994 (2.9651)  weight_decay: 0.0500 (0.0500)  time: 0.7035  data: 0.0008  max mem: 57114
Epoch: [180]  [140/156]  eta: 0:00:11  lr: 0.003071  min_lr: 0.003071  loss: 2.8936 (2.9586)  weight_decay: 0.0500 (0.0500)  time: 0.6894  data: 0.0007  max mem: 57114
Epoch: [180]  [150/156]  eta: 0:00:04  lr: 0.003069  min_lr: 0.003069  loss: 3.0313 (2.9777)  weight_decay: 0.0500 (0.0500)  time: 0.6819  data: 0.0001  max mem: 57114
Epoch: [180]  [155/156]  eta: 0:00:00  lr: 0.003067  min_lr: 0.003067  loss: 3.0313 (2.9719)  weight_decay: 0.0500 (0.0500)  time: 0.6834  data: 0.0001  max mem: 57114
Epoch: [180] Total time: 0:01:54 (0.7338 s / it)
Averaged stats: lr: 0.003067  min_lr: 0.003067  loss: 3.0313 (2.9862)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7484 (0.7484)  acc1: 86.9792 (86.9792)  acc5: 98.9583 (98.9583)  time: 2.1210  data: 1.8648  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9837 (0.8850)  acc1: 79.1667 (80.6905)  acc5: 95.8333 (96.2916)  time: 0.5750  data: 0.3731  max mem: 57114
Test: Total time: 0:00:02 (0.5951 s / it)
* Acc@1 80.850 Acc@5 95.608 loss 0.965
Accuracy of the model on the 50000 test images: 80.9%
Max accuracy: 80.85%
Test:  [0/5]  eta: 0:00:10  loss: 0.6708 (0.6708)  acc1: 82.2917 (82.2917)  acc5: 95.3125 (95.3125)  time: 2.1817  data: 1.9381  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8106 (0.7915)  acc1: 78.6458 (76.9821)  acc5: 94.7917 (91.9437)  time: 0.5870  data: 0.3877  max mem: 57114
Test: Total time: 0:00:02 (0.5998 s / it)
* Acc@1 74.656 Acc@5 92.357 loss 0.963
Accuracy of the model EMA on 50000 test images: 74.7%
Max EMA accuracy: 74.66%
Epoch: [181]  [  0/156]  eta: 0:07:15  lr: 0.003067  min_lr: 0.003067  loss: 3.3196 (3.3196)  weight_decay: 0.0500 (0.0500)  time: 2.7892  data: 2.1345  max mem: 57114
Epoch: [181]  [ 10/156]  eta: 0:02:10  lr: 0.003064  min_lr: 0.003064  loss: 2.8718 (2.7324)  weight_decay: 0.0500 (0.0500)  time: 0.8962  data: 0.1945  max mem: 57114
Epoch: [181]  [ 20/156]  eta: 0:01:52  lr: 0.003061  min_lr: 0.003061  loss: 2.9916 (2.9211)  weight_decay: 0.0500 (0.0500)  time: 0.7300  data: 0.0004  max mem: 57114
Epoch: [181]  [ 30/156]  eta: 0:01:41  lr: 0.003058  min_lr: 0.003058  loss: 3.1307 (3.0140)  weight_decay: 0.0500 (0.0500)  time: 0.7593  data: 0.0004  max mem: 57114
Epoch: [181]  [ 40/156]  eta: 0:01:31  lr: 0.003056  min_lr: 0.003056  loss: 3.1796 (3.0272)  weight_decay: 0.0500 (0.0500)  time: 0.7446  data: 0.0005  max mem: 57114
Epoch: [181]  [ 50/156]  eta: 0:01:21  lr: 0.003053  min_lr: 0.003053  loss: 2.9838 (2.9732)  weight_decay: 0.0500 (0.0500)  time: 0.7200  data: 0.0006  max mem: 57114
Epoch: [181]  [ 60/156]  eta: 0:01:13  lr: 0.003050  min_lr: 0.003050  loss: 2.8833 (2.9446)  weight_decay: 0.0500 (0.0500)  time: 0.7230  data: 0.0005  max mem: 57114
Epoch: [181]  [ 70/156]  eta: 0:01:05  lr: 0.003047  min_lr: 0.003047  loss: 2.9407 (2.9488)  weight_decay: 0.0500 (0.0500)  time: 0.7202  data: 0.0005  max mem: 57114
Epoch: [181]  [ 80/156]  eta: 0:00:57  lr: 0.003044  min_lr: 0.003044  loss: 3.0188 (2.9426)  weight_decay: 0.0500 (0.0500)  time: 0.7201  data: 0.0004  max mem: 57114
Epoch: [181]  [ 90/156]  eta: 0:00:49  lr: 0.003042  min_lr: 0.003042  loss: 3.0204 (2.9422)  weight_decay: 0.0500 (0.0500)  time: 0.7158  data: 0.0004  max mem: 57114
Epoch: [181]  [100/156]  eta: 0:00:41  lr: 0.003039  min_lr: 0.003039  loss: 2.9131 (2.9329)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0004  max mem: 57114
Epoch: [181]  [110/156]  eta: 0:00:34  lr: 0.003036  min_lr: 0.003036  loss: 2.9091 (2.9327)  weight_decay: 0.0500 (0.0500)  time: 0.7122  data: 0.0005  max mem: 57114
Epoch: [181]  [120/156]  eta: 0:00:26  lr: 0.003033  min_lr: 0.003033  loss: 2.9965 (2.9338)  weight_decay: 0.0500 (0.0500)  time: 0.7018  data: 0.0007  max mem: 57114
Epoch: [181]  [130/156]  eta: 0:00:19  lr: 0.003031  min_lr: 0.003031  loss: 3.1758 (2.9489)  weight_decay: 0.0500 (0.0500)  time: 0.6969  data: 0.0016  max mem: 57114
Epoch: [181]  [140/156]  eta: 0:00:11  lr: 0.003028  min_lr: 0.003028  loss: 3.1949 (2.9536)  weight_decay: 0.0500 (0.0500)  time: 0.6836  data: 0.0013  max mem: 57114
Epoch: [181]  [150/156]  eta: 0:00:04  lr: 0.003025  min_lr: 0.003025  loss: 2.9459 (2.9480)  weight_decay: 0.0500 (0.0500)  time: 0.6812  data: 0.0001  max mem: 57114
Epoch: [181]  [155/156]  eta: 0:00:00  lr: 0.003024  min_lr: 0.003024  loss: 2.9459 (2.9443)  weight_decay: 0.0500 (0.0500)  time: 0.6899  data: 0.0001  max mem: 57114
Epoch: [181] Total time: 0:01:53 (0.7297 s / it)
Averaged stats: lr: 0.003024  min_lr: 0.003024  loss: 2.9459 (2.9686)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6820 (0.6820)  acc1: 89.5833 (89.5833)  acc5: 98.4375 (98.4375)  time: 2.0738  data: 1.8180  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8860 (0.8081)  acc1: 82.8125 (83.6317)  acc5: 97.9167 (97.1867)  time: 0.5654  data: 0.3637  max mem: 57114
Test: Total time: 0:00:02 (0.5858 s / it)
* Acc@1 80.900 Acc@5 95.670 loss 0.906
Accuracy of the model on the 50000 test images: 80.9%
Max accuracy: 80.90%
Test:  [0/5]  eta: 0:00:09  loss: 0.6665 (0.6665)  acc1: 82.2917 (82.2917)  acc5: 95.8333 (95.8333)  time: 1.9918  data: 1.7482  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8054 (0.7854)  acc1: 78.6458 (76.9821)  acc5: 94.7917 (92.0716)  time: 0.5490  data: 0.3497  max mem: 57114
Test: Total time: 0:00:02 (0.5607 s / it)
* Acc@1 74.836 Acc@5 92.447 loss 0.955
Accuracy of the model EMA on 50000 test images: 74.8%
Max EMA accuracy: 74.84%
Epoch: [182]  [  0/156]  eta: 0:06:27  lr: 0.003023  min_lr: 0.003023  loss: 2.7551 (2.7551)  weight_decay: 0.0500 (0.0500)  time: 2.4840  data: 1.8318  max mem: 57114
Epoch: [182]  [ 10/156]  eta: 0:02:11  lr: 0.003020  min_lr: 0.003020  loss: 3.1409 (2.9480)  weight_decay: 0.0500 (0.0500)  time: 0.9000  data: 0.1668  max mem: 57114
Epoch: [182]  [ 20/156]  eta: 0:01:50  lr: 0.003018  min_lr: 0.003018  loss: 3.0922 (2.9746)  weight_decay: 0.0500 (0.0500)  time: 0.7326  data: 0.0003  max mem: 57114
Epoch: [182]  [ 30/156]  eta: 0:01:38  lr: 0.003015  min_lr: 0.003015  loss: 3.0688 (2.9652)  weight_decay: 0.0500 (0.0500)  time: 0.7198  data: 0.0003  max mem: 57114
Epoch: [182]  [ 40/156]  eta: 0:01:29  lr: 0.003012  min_lr: 0.003012  loss: 3.0014 (2.9555)  weight_decay: 0.0500 (0.0500)  time: 0.7228  data: 0.0003  max mem: 57114
Epoch: [182]  [ 50/156]  eta: 0:01:21  lr: 0.003009  min_lr: 0.003009  loss: 2.8850 (2.9304)  weight_decay: 0.0500 (0.0500)  time: 0.7465  data: 0.0004  max mem: 57114
Epoch: [182]  [ 60/156]  eta: 0:01:13  lr: 0.003007  min_lr: 0.003007  loss: 2.8850 (2.9335)  weight_decay: 0.0500 (0.0500)  time: 0.7437  data: 0.0004  max mem: 57114
Epoch: [182]  [ 70/156]  eta: 0:01:05  lr: 0.003004  min_lr: 0.003004  loss: 3.0219 (2.9516)  weight_decay: 0.0500 (0.0500)  time: 0.7279  data: 0.0004  max mem: 57114
Epoch: [182]  [ 80/156]  eta: 0:00:57  lr: 0.003001  min_lr: 0.003001  loss: 3.1244 (2.9601)  weight_decay: 0.0500 (0.0500)  time: 0.7163  data: 0.0004  max mem: 57114
Epoch: [182]  [ 90/156]  eta: 0:00:49  lr: 0.002998  min_lr: 0.002998  loss: 3.0554 (2.9464)  weight_decay: 0.0500 (0.0500)  time: 0.7014  data: 0.0003  max mem: 57114
Epoch: [182]  [100/156]  eta: 0:00:41  lr: 0.002995  min_lr: 0.002995  loss: 2.9983 (2.9525)  weight_decay: 0.0500 (0.0500)  time: 0.6960  data: 0.0003  max mem: 57114
Epoch: [182]  [110/156]  eta: 0:00:33  lr: 0.002993  min_lr: 0.002993  loss: 3.0444 (2.9497)  weight_decay: 0.0500 (0.0500)  time: 0.6968  data: 0.0003  max mem: 57114
Epoch: [182]  [120/156]  eta: 0:00:26  lr: 0.002990  min_lr: 0.002990  loss: 3.0490 (2.9657)  weight_decay: 0.0500 (0.0500)  time: 0.7087  data: 0.0004  max mem: 57114
Epoch: [182]  [130/156]  eta: 0:00:19  lr: 0.002987  min_lr: 0.002987  loss: 3.0128 (2.9591)  weight_decay: 0.0500 (0.0500)  time: 0.7083  data: 0.0009  max mem: 57114
Epoch: [182]  [140/156]  eta: 0:00:11  lr: 0.002984  min_lr: 0.002984  loss: 2.9560 (2.9577)  weight_decay: 0.0500 (0.0500)  time: 0.6957  data: 0.0007  max mem: 57114
Epoch: [182]  [150/156]  eta: 0:00:04  lr: 0.002981  min_lr: 0.002981  loss: 2.9647 (2.9650)  weight_decay: 0.0500 (0.0500)  time: 0.6840  data: 0.0001  max mem: 57114
Epoch: [182]  [155/156]  eta: 0:00:00  lr: 0.002980  min_lr: 0.002980  loss: 3.0970 (2.9728)  weight_decay: 0.0500 (0.0500)  time: 0.6846  data: 0.0001  max mem: 57114
Epoch: [182] Total time: 0:01:53 (0.7264 s / it)
Averaged stats: lr: 0.002980  min_lr: 0.002980  loss: 3.0970 (2.9636)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.9241 (0.9241)  acc1: 84.3750 (84.3750)  acc5: 97.9167 (97.9167)  time: 2.0798  data: 1.8238  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0312 (1.0499)  acc1: 82.8125 (79.7954)  acc5: 95.8333 (95.0128)  time: 0.5667  data: 0.3648  max mem: 57114
Test: Total time: 0:00:02 (0.5884 s / it)
* Acc@1 79.933 Acc@5 95.368 loss 1.100
Accuracy of the model on the 50000 test images: 79.9%
Max accuracy: 80.90%
Test:  [0/5]  eta: 0:00:11  loss: 0.6624 (0.6624)  acc1: 82.2917 (82.2917)  acc5: 95.8333 (95.8333)  time: 2.3208  data: 2.0772  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8001 (0.7796)  acc1: 78.6458 (77.1100)  acc5: 94.7917 (92.3274)  time: 0.6148  data: 0.4155  max mem: 57114
Test: Total time: 0:00:03 (0.6272 s / it)
* Acc@1 74.922 Acc@5 92.543 loss 0.948
Accuracy of the model EMA on 50000 test images: 74.9%
Max EMA accuracy: 74.92%
Epoch: [183]  [  0/156]  eta: 0:07:50  lr: 0.002980  min_lr: 0.002980  loss: 1.9212 (1.9212)  weight_decay: 0.0500 (0.0500)  time: 3.0177  data: 2.3690  max mem: 57114
Epoch: [183]  [ 10/156]  eta: 0:02:14  lr: 0.002977  min_lr: 0.002977  loss: 3.1364 (2.9980)  weight_decay: 0.0500 (0.0500)  time: 0.9210  data: 0.2156  max mem: 57114
Epoch: [183]  [ 20/156]  eta: 0:01:52  lr: 0.002974  min_lr: 0.002974  loss: 3.0747 (2.9987)  weight_decay: 0.0500 (0.0500)  time: 0.7143  data: 0.0003  max mem: 57114
Epoch: [183]  [ 30/156]  eta: 0:01:41  lr: 0.002971  min_lr: 0.002971  loss: 3.0747 (2.9673)  weight_decay: 0.0500 (0.0500)  time: 0.7363  data: 0.0003  max mem: 57114
Epoch: [183]  [ 40/156]  eta: 0:01:31  lr: 0.002969  min_lr: 0.002969  loss: 3.1236 (2.9883)  weight_decay: 0.0500 (0.0500)  time: 0.7452  data: 0.0003  max mem: 57114
Epoch: [183]  [ 50/156]  eta: 0:01:22  lr: 0.002966  min_lr: 0.002966  loss: 3.1277 (2.9944)  weight_decay: 0.0500 (0.0500)  time: 0.7313  data: 0.0004  max mem: 57114
Epoch: [183]  [ 60/156]  eta: 0:01:13  lr: 0.002963  min_lr: 0.002963  loss: 3.0446 (2.9849)  weight_decay: 0.0500 (0.0500)  time: 0.7100  data: 0.0004  max mem: 57114
Epoch: [183]  [ 70/156]  eta: 0:01:04  lr: 0.002960  min_lr: 0.002960  loss: 3.0446 (3.0119)  weight_decay: 0.0500 (0.0500)  time: 0.6999  data: 0.0004  max mem: 57114
Epoch: [183]  [ 80/156]  eta: 0:00:56  lr: 0.002958  min_lr: 0.002958  loss: 3.0168 (3.0106)  weight_decay: 0.0500 (0.0500)  time: 0.7067  data: 0.0003  max mem: 57114
Epoch: [183]  [ 90/156]  eta: 0:00:48  lr: 0.002955  min_lr: 0.002955  loss: 3.0919 (3.0206)  weight_decay: 0.0500 (0.0500)  time: 0.7027  data: 0.0004  max mem: 57114
Epoch: [183]  [100/156]  eta: 0:00:41  lr: 0.002952  min_lr: 0.002952  loss: 3.1501 (3.0087)  weight_decay: 0.0500 (0.0500)  time: 0.7075  data: 0.0004  max mem: 57114
Epoch: [183]  [110/156]  eta: 0:00:33  lr: 0.002949  min_lr: 0.002949  loss: 3.1487 (3.0180)  weight_decay: 0.0500 (0.0500)  time: 0.7152  data: 0.0003  max mem: 57114
Epoch: [183]  [120/156]  eta: 0:00:26  lr: 0.002946  min_lr: 0.002946  loss: 2.9576 (3.0099)  weight_decay: 0.0500 (0.0500)  time: 0.7085  data: 0.0004  max mem: 57114
Epoch: [183]  [130/156]  eta: 0:00:19  lr: 0.002944  min_lr: 0.002944  loss: 2.9448 (3.0047)  weight_decay: 0.0500 (0.0500)  time: 0.6980  data: 0.0009  max mem: 57114
Epoch: [183]  [140/156]  eta: 0:00:11  lr: 0.002941  min_lr: 0.002941  loss: 2.9432 (3.0027)  weight_decay: 0.0500 (0.0500)  time: 0.6873  data: 0.0007  max mem: 57114
Epoch: [183]  [150/156]  eta: 0:00:04  lr: 0.002938  min_lr: 0.002938  loss: 3.0750 (3.0169)  weight_decay: 0.0500 (0.0500)  time: 0.6810  data: 0.0001  max mem: 57114
Epoch: [183]  [155/156]  eta: 0:00:00  lr: 0.002937  min_lr: 0.002937  loss: 3.1759 (3.0200)  weight_decay: 0.0500 (0.0500)  time: 0.6824  data: 0.0001  max mem: 57114
Epoch: [183] Total time: 0:01:53 (0.7257 s / it)
Averaged stats: lr: 0.002937  min_lr: 0.002937  loss: 3.1759 (2.9645)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8423 (0.8423)  acc1: 86.4583 (86.4583)  acc5: 98.4375 (98.4375)  time: 2.0730  data: 1.8173  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0144 (0.9956)  acc1: 79.6875 (79.4118)  acc5: 96.3542 (95.5243)  time: 0.5655  data: 0.3635  max mem: 57114
Test: Total time: 0:00:02 (0.5880 s / it)
* Acc@1 80.327 Acc@5 95.336 loss 1.071
Accuracy of the model on the 50000 test images: 80.3%
Max accuracy: 80.90%
Test:  [0/5]  eta: 0:00:11  loss: 0.6581 (0.6581)  acc1: 82.2917 (82.2917)  acc5: 95.8333 (95.8333)  time: 2.2826  data: 2.0392  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7949 (0.7738)  acc1: 78.6458 (77.1100)  acc5: 94.7917 (92.4552)  time: 0.6072  data: 0.4079  max mem: 57114
Test: Total time: 0:00:03 (0.6193 s / it)
* Acc@1 75.048 Acc@5 92.615 loss 0.941
Accuracy of the model EMA on 50000 test images: 75.0%
Max EMA accuracy: 75.05%
Epoch: [184]  [  0/156]  eta: 0:06:28  lr: 0.002936  min_lr: 0.002936  loss: 3.3497 (3.3497)  weight_decay: 0.0500 (0.0500)  time: 2.4913  data: 1.8387  max mem: 57114
Epoch: [184]  [ 10/156]  eta: 0:02:09  lr: 0.002934  min_lr: 0.002934  loss: 3.1756 (2.8088)  weight_decay: 0.0500 (0.0500)  time: 0.8901  data: 0.1674  max mem: 57114
Epoch: [184]  [ 20/156]  eta: 0:01:49  lr: 0.002931  min_lr: 0.002931  loss: 3.1602 (2.9423)  weight_decay: 0.0500 (0.0500)  time: 0.7193  data: 0.0003  max mem: 57114
Epoch: [184]  [ 30/156]  eta: 0:01:38  lr: 0.002928  min_lr: 0.002928  loss: 3.1724 (3.0211)  weight_decay: 0.0500 (0.0500)  time: 0.7160  data: 0.0003  max mem: 57114
Epoch: [184]  [ 40/156]  eta: 0:01:28  lr: 0.002925  min_lr: 0.002925  loss: 3.1473 (2.9647)  weight_decay: 0.0500 (0.0500)  time: 0.7202  data: 0.0004  max mem: 57114
Epoch: [184]  [ 50/156]  eta: 0:01:20  lr: 0.002923  min_lr: 0.002923  loss: 3.0215 (2.9685)  weight_decay: 0.0500 (0.0500)  time: 0.7237  data: 0.0004  max mem: 57114
Epoch: [184]  [ 60/156]  eta: 0:01:11  lr: 0.002920  min_lr: 0.002920  loss: 2.9488 (2.9278)  weight_decay: 0.0500 (0.0500)  time: 0.7208  data: 0.0004  max mem: 57114
Epoch: [184]  [ 70/156]  eta: 0:01:04  lr: 0.002917  min_lr: 0.002917  loss: 2.9187 (2.8974)  weight_decay: 0.0500 (0.0500)  time: 0.7199  data: 0.0004  max mem: 57114
Epoch: [184]  [ 80/156]  eta: 0:00:56  lr: 0.002914  min_lr: 0.002914  loss: 2.8860 (2.8991)  weight_decay: 0.0500 (0.0500)  time: 0.7241  data: 0.0004  max mem: 57114
Epoch: [184]  [ 90/156]  eta: 0:00:48  lr: 0.002912  min_lr: 0.002912  loss: 3.0111 (2.9130)  weight_decay: 0.0500 (0.0500)  time: 0.7170  data: 0.0004  max mem: 57114
Epoch: [184]  [100/156]  eta: 0:00:41  lr: 0.002909  min_lr: 0.002909  loss: 3.1108 (2.9262)  weight_decay: 0.0500 (0.0500)  time: 0.7087  data: 0.0004  max mem: 57114
Epoch: [184]  [110/156]  eta: 0:00:33  lr: 0.002906  min_lr: 0.002906  loss: 3.0966 (2.9199)  weight_decay: 0.0500 (0.0500)  time: 0.6979  data: 0.0004  max mem: 57114
Epoch: [184]  [120/156]  eta: 0:00:26  lr: 0.002903  min_lr: 0.002903  loss: 3.0201 (2.9196)  weight_decay: 0.0500 (0.0500)  time: 0.7025  data: 0.0004  max mem: 57114
Epoch: [184]  [130/156]  eta: 0:00:18  lr: 0.002900  min_lr: 0.002900  loss: 3.0978 (2.9268)  weight_decay: 0.0500 (0.0500)  time: 0.7008  data: 0.0009  max mem: 57114
Epoch: [184]  [140/156]  eta: 0:00:11  lr: 0.002898  min_lr: 0.002898  loss: 3.0710 (2.9223)  weight_decay: 0.0500 (0.0500)  time: 0.6903  data: 0.0007  max mem: 57114
Epoch: [184]  [150/156]  eta: 0:00:04  lr: 0.002895  min_lr: 0.002895  loss: 3.0710 (2.9339)  weight_decay: 0.0500 (0.0500)  time: 0.6861  data: 0.0001  max mem: 57114
Epoch: [184]  [155/156]  eta: 0:00:00  lr: 0.002894  min_lr: 0.002894  loss: 3.0348 (2.9386)  weight_decay: 0.0500 (0.0500)  time: 0.6904  data: 0.0001  max mem: 57114
Epoch: [184] Total time: 0:01:52 (0.7235 s / it)
Averaged stats: lr: 0.002894  min_lr: 0.002894  loss: 3.0348 (2.9601)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7235 (0.7235)  acc1: 89.0625 (89.0625)  acc5: 98.4375 (98.4375)  time: 2.1639  data: 1.9084  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9237 (0.8652)  acc1: 81.2500 (82.6087)  acc5: 96.8750 (96.1637)  time: 0.5835  data: 0.3817  max mem: 57114
Test: Total time: 0:00:03 (0.6056 s / it)
* Acc@1 80.760 Acc@5 95.532 loss 0.969
Accuracy of the model on the 50000 test images: 80.8%
Max accuracy: 80.90%
Test:  [0/5]  eta: 0:00:11  loss: 0.6543 (0.6543)  acc1: 81.7708 (81.7708)  acc5: 95.8333 (95.8333)  time: 2.2883  data: 2.0448  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7901 (0.7683)  acc1: 79.1667 (77.2379)  acc5: 94.7917 (92.4552)  time: 0.6084  data: 0.4090  max mem: 57114
Test: Total time: 0:00:03 (0.6249 s / it)
* Acc@1 75.170 Acc@5 92.679 loss 0.934
Accuracy of the model EMA on 50000 test images: 75.2%
Max EMA accuracy: 75.17%
Epoch: [185]  [  0/156]  eta: 0:08:52  lr: 0.002893  min_lr: 0.002893  loss: 3.1643 (3.1643)  weight_decay: 0.0500 (0.0500)  time: 3.4139  data: 2.7642  max mem: 57114
Epoch: [185]  [ 10/156]  eta: 0:02:22  lr: 0.002891  min_lr: 0.002891  loss: 3.3065 (3.2360)  weight_decay: 0.0500 (0.0500)  time: 0.9787  data: 0.2515  max mem: 57114
Epoch: [185]  [ 20/156]  eta: 0:01:54  lr: 0.002888  min_lr: 0.002888  loss: 3.1169 (3.0844)  weight_decay: 0.0500 (0.0500)  time: 0.7163  data: 0.0003  max mem: 57114
Epoch: [185]  [ 30/156]  eta: 0:01:40  lr: 0.002885  min_lr: 0.002885  loss: 2.9225 (3.0114)  weight_decay: 0.0500 (0.0500)  time: 0.7041  data: 0.0004  max mem: 57114
Epoch: [185]  [ 40/156]  eta: 0:01:30  lr: 0.002882  min_lr: 0.002882  loss: 2.8873 (2.9600)  weight_decay: 0.0500 (0.0500)  time: 0.7135  data: 0.0004  max mem: 57114
Epoch: [185]  [ 50/156]  eta: 0:01:21  lr: 0.002879  min_lr: 0.002879  loss: 2.7358 (2.9250)  weight_decay: 0.0500 (0.0500)  time: 0.7214  data: 0.0004  max mem: 57114
Epoch: [185]  [ 60/156]  eta: 0:01:13  lr: 0.002877  min_lr: 0.002877  loss: 3.0355 (2.9441)  weight_decay: 0.0500 (0.0500)  time: 0.7312  data: 0.0004  max mem: 57114
Epoch: [185]  [ 70/156]  eta: 0:01:05  lr: 0.002874  min_lr: 0.002874  loss: 2.8922 (2.9048)  weight_decay: 0.0500 (0.0500)  time: 0.7245  data: 0.0004  max mem: 57114
Epoch: [185]  [ 80/156]  eta: 0:00:57  lr: 0.002871  min_lr: 0.002871  loss: 2.8110 (2.9019)  weight_decay: 0.0500 (0.0500)  time: 0.7110  data: 0.0003  max mem: 57114
Epoch: [185]  [ 90/156]  eta: 0:00:49  lr: 0.002868  min_lr: 0.002868  loss: 2.8098 (2.8816)  weight_decay: 0.0500 (0.0500)  time: 0.7046  data: 0.0003  max mem: 57114
Epoch: [185]  [100/156]  eta: 0:00:41  lr: 0.002866  min_lr: 0.002866  loss: 2.9726 (2.8975)  weight_decay: 0.0500 (0.0500)  time: 0.6967  data: 0.0004  max mem: 57114
Epoch: [185]  [110/156]  eta: 0:00:34  lr: 0.002863  min_lr: 0.002863  loss: 3.0754 (2.9125)  weight_decay: 0.0500 (0.0500)  time: 0.7117  data: 0.0004  max mem: 57114
Epoch: [185]  [120/156]  eta: 0:00:26  lr: 0.002860  min_lr: 0.002860  loss: 3.0801 (2.9258)  weight_decay: 0.0500 (0.0500)  time: 0.7276  data: 0.0004  max mem: 57114
Epoch: [185]  [130/156]  eta: 0:00:19  lr: 0.002857  min_lr: 0.002857  loss: 3.1468 (2.9339)  weight_decay: 0.0500 (0.0500)  time: 0.7125  data: 0.0009  max mem: 57114
Epoch: [185]  [140/156]  eta: 0:00:11  lr: 0.002855  min_lr: 0.002855  loss: 3.1109 (2.9268)  weight_decay: 0.0500 (0.0500)  time: 0.6906  data: 0.0007  max mem: 57114
Epoch: [185]  [150/156]  eta: 0:00:04  lr: 0.002852  min_lr: 0.002852  loss: 3.0620 (2.9254)  weight_decay: 0.0500 (0.0500)  time: 0.6843  data: 0.0001  max mem: 57114
Epoch: [185]  [155/156]  eta: 0:00:00  lr: 0.002851  min_lr: 0.002851  loss: 2.8423 (2.9243)  weight_decay: 0.0500 (0.0500)  time: 0.6861  data: 0.0001  max mem: 57114
Epoch: [185] Total time: 0:01:53 (0.7295 s / it)
Averaged stats: lr: 0.002851  min_lr: 0.002851  loss: 2.8423 (2.9536)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6468 (0.6468)  acc1: 86.4583 (86.4583)  acc5: 98.4375 (98.4375)  time: 2.0142  data: 1.7585  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8048 (0.7636)  acc1: 80.7292 (80.6905)  acc5: 97.3958 (96.9309)  time: 0.5536  data: 0.3518  max mem: 57114
Test: Total time: 0:00:02 (0.5766 s / it)
* Acc@1 81.270 Acc@5 95.898 loss 0.843
Accuracy of the model on the 50000 test images: 81.3%
Max accuracy: 81.27%
Test:  [0/5]  eta: 0:00:10  loss: 0.6507 (0.6507)  acc1: 82.2917 (82.2917)  acc5: 95.8333 (95.8333)  time: 2.0234  data: 1.7800  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7855 (0.7631)  acc1: 79.1667 (77.3657)  acc5: 94.7917 (92.5831)  time: 0.5553  data: 0.3561  max mem: 57114
Test: Total time: 0:00:02 (0.5682 s / it)
* Acc@1 75.294 Acc@5 92.767 loss 0.928
Accuracy of the model EMA on 50000 test images: 75.3%
Max EMA accuracy: 75.29%
Epoch: [186]  [  0/156]  eta: 0:07:33  lr: 0.002850  min_lr: 0.002850  loss: 3.3841 (3.3841)  weight_decay: 0.0500 (0.0500)  time: 2.9091  data: 2.2553  max mem: 57114
Epoch: [186]  [ 10/156]  eta: 0:02:12  lr: 0.002847  min_lr: 0.002847  loss: 3.2019 (3.0407)  weight_decay: 0.0500 (0.0500)  time: 0.9084  data: 0.2053  max mem: 57114
Epoch: [186]  [ 20/156]  eta: 0:01:53  lr: 0.002845  min_lr: 0.002845  loss: 3.1101 (3.0325)  weight_decay: 0.0500 (0.0500)  time: 0.7330  data: 0.0003  max mem: 57114
Epoch: [186]  [ 30/156]  eta: 0:01:42  lr: 0.002842  min_lr: 0.002842  loss: 2.9927 (2.9983)  weight_decay: 0.0500 (0.0500)  time: 0.7605  data: 0.0004  max mem: 57114
Epoch: [186]  [ 40/156]  eta: 0:01:31  lr: 0.002839  min_lr: 0.002839  loss: 2.9927 (2.9794)  weight_decay: 0.0500 (0.0500)  time: 0.7440  data: 0.0004  max mem: 57114
Epoch: [186]  [ 50/156]  eta: 0:01:22  lr: 0.002836  min_lr: 0.002836  loss: 3.1183 (2.9975)  weight_decay: 0.0500 (0.0500)  time: 0.7278  data: 0.0004  max mem: 57114
Epoch: [186]  [ 60/156]  eta: 0:01:14  lr: 0.002834  min_lr: 0.002834  loss: 2.9654 (2.9601)  weight_decay: 0.0500 (0.0500)  time: 0.7294  data: 0.0004  max mem: 57114
Epoch: [186]  [ 70/156]  eta: 0:01:05  lr: 0.002831  min_lr: 0.002831  loss: 2.9654 (2.9889)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0004  max mem: 57114
Epoch: [186]  [ 80/156]  eta: 0:00:57  lr: 0.002828  min_lr: 0.002828  loss: 3.0724 (2.9752)  weight_decay: 0.0500 (0.0500)  time: 0.7012  data: 0.0004  max mem: 57114
Epoch: [186]  [ 90/156]  eta: 0:00:49  lr: 0.002825  min_lr: 0.002825  loss: 2.9349 (2.9595)  weight_decay: 0.0500 (0.0500)  time: 0.7099  data: 0.0004  max mem: 57114
Epoch: [186]  [100/156]  eta: 0:00:41  lr: 0.002823  min_lr: 0.002823  loss: 3.0237 (2.9556)  weight_decay: 0.0500 (0.0500)  time: 0.7133  data: 0.0004  max mem: 57114
Epoch: [186]  [110/156]  eta: 0:00:34  lr: 0.002820  min_lr: 0.002820  loss: 3.0237 (2.9519)  weight_decay: 0.0500 (0.0500)  time: 0.7076  data: 0.0004  max mem: 57114
Epoch: [186]  [120/156]  eta: 0:00:26  lr: 0.002817  min_lr: 0.002817  loss: 2.9905 (2.9452)  weight_decay: 0.0500 (0.0500)  time: 0.7008  data: 0.0004  max mem: 57114
Epoch: [186]  [130/156]  eta: 0:00:19  lr: 0.002814  min_lr: 0.002814  loss: 2.9905 (2.9576)  weight_decay: 0.0500 (0.0500)  time: 0.6977  data: 0.0009  max mem: 57114
Epoch: [186]  [140/156]  eta: 0:00:11  lr: 0.002812  min_lr: 0.002812  loss: 3.1214 (2.9591)  weight_decay: 0.0500 (0.0500)  time: 0.6906  data: 0.0007  max mem: 57114
Epoch: [186]  [150/156]  eta: 0:00:04  lr: 0.002809  min_lr: 0.002809  loss: 3.0430 (2.9593)  weight_decay: 0.0500 (0.0500)  time: 0.6849  data: 0.0001  max mem: 57114
Epoch: [186]  [155/156]  eta: 0:00:00  lr: 0.002808  min_lr: 0.002808  loss: 3.0430 (2.9633)  weight_decay: 0.0500 (0.0500)  time: 0.6843  data: 0.0001  max mem: 57114
Epoch: [186] Total time: 0:01:53 (0.7294 s / it)
Averaged stats: lr: 0.002808  min_lr: 0.002808  loss: 3.0430 (2.9614)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8306 (0.8306)  acc1: 88.0208 (88.0208)  acc5: 98.4375 (98.4375)  time: 2.0146  data: 1.7590  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9203 (0.9401)  acc1: 84.8958 (81.8414)  acc5: 96.8750 (95.9079)  time: 0.5537  data: 0.3519  max mem: 57114
Test: Total time: 0:00:02 (0.5755 s / it)
* Acc@1 80.369 Acc@5 95.484 loss 1.046
Accuracy of the model on the 50000 test images: 80.4%
Max accuracy: 81.27%
Test:  [0/5]  eta: 0:00:11  loss: 0.6473 (0.6473)  acc1: 82.2917 (82.2917)  acc5: 95.8333 (95.8333)  time: 2.2502  data: 2.0067  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7807 (0.7581)  acc1: 79.6875 (77.4936)  acc5: 94.7917 (92.5831)  time: 0.6007  data: 0.4014  max mem: 57114
Test: Total time: 0:00:03 (0.6150 s / it)
* Acc@1 75.456 Acc@5 92.809 loss 0.922
Accuracy of the model EMA on 50000 test images: 75.5%
Max EMA accuracy: 75.46%
Epoch: [187]  [  0/156]  eta: 0:07:37  lr: 0.002807  min_lr: 0.002807  loss: 1.8706 (1.8706)  weight_decay: 0.0500 (0.0500)  time: 2.9348  data: 2.2825  max mem: 57114
Epoch: [187]  [ 10/156]  eta: 0:02:12  lr: 0.002805  min_lr: 0.002805  loss: 2.7937 (2.7398)  weight_decay: 0.0500 (0.0500)  time: 0.9083  data: 0.2078  max mem: 57114
Epoch: [187]  [ 20/156]  eta: 0:01:52  lr: 0.002802  min_lr: 0.002802  loss: 2.7937 (2.7745)  weight_decay: 0.0500 (0.0500)  time: 0.7200  data: 0.0004  max mem: 57114
Epoch: [187]  [ 30/156]  eta: 0:01:39  lr: 0.002799  min_lr: 0.002799  loss: 3.0181 (2.8198)  weight_decay: 0.0500 (0.0500)  time: 0.7270  data: 0.0003  max mem: 57114
Epoch: [187]  [ 40/156]  eta: 0:01:30  lr: 0.002796  min_lr: 0.002796  loss: 3.0543 (2.8490)  weight_decay: 0.0500 (0.0500)  time: 0.7249  data: 0.0004  max mem: 57114
Epoch: [187]  [ 50/156]  eta: 0:01:21  lr: 0.002794  min_lr: 0.002794  loss: 3.0464 (2.8771)  weight_decay: 0.0500 (0.0500)  time: 0.7234  data: 0.0004  max mem: 57114
Epoch: [187]  [ 60/156]  eta: 0:01:12  lr: 0.002791  min_lr: 0.002791  loss: 3.0379 (2.8769)  weight_decay: 0.0500 (0.0500)  time: 0.7195  data: 0.0003  max mem: 57114
Epoch: [187]  [ 70/156]  eta: 0:01:04  lr: 0.002788  min_lr: 0.002788  loss: 3.0655 (2.8900)  weight_decay: 0.0500 (0.0500)  time: 0.7208  data: 0.0004  max mem: 57114
Epoch: [187]  [ 80/156]  eta: 0:00:56  lr: 0.002785  min_lr: 0.002785  loss: 3.1737 (2.9076)  weight_decay: 0.0500 (0.0500)  time: 0.7069  data: 0.0004  max mem: 57114
Epoch: [187]  [ 90/156]  eta: 0:00:48  lr: 0.002783  min_lr: 0.002783  loss: 3.1293 (2.9045)  weight_decay: 0.0500 (0.0500)  time: 0.7027  data: 0.0004  max mem: 57114
Epoch: [187]  [100/156]  eta: 0:00:41  lr: 0.002780  min_lr: 0.002780  loss: 3.0176 (2.9104)  weight_decay: 0.0500 (0.0500)  time: 0.7071  data: 0.0004  max mem: 57114
Epoch: [187]  [110/156]  eta: 0:00:33  lr: 0.002777  min_lr: 0.002777  loss: 3.0556 (2.9395)  weight_decay: 0.0500 (0.0500)  time: 0.7011  data: 0.0004  max mem: 57114
Epoch: [187]  [120/156]  eta: 0:00:26  lr: 0.002774  min_lr: 0.002774  loss: 3.2161 (2.9618)  weight_decay: 0.0500 (0.0500)  time: 0.6981  data: 0.0004  max mem: 57114
Epoch: [187]  [130/156]  eta: 0:00:18  lr: 0.002772  min_lr: 0.002772  loss: 3.1946 (2.9597)  weight_decay: 0.0500 (0.0500)  time: 0.7021  data: 0.0005  max mem: 57114
Epoch: [187]  [140/156]  eta: 0:00:11  lr: 0.002769  min_lr: 0.002769  loss: 2.9646 (2.9620)  weight_decay: 0.0500 (0.0500)  time: 0.6964  data: 0.0003  max mem: 57114
Epoch: [187]  [150/156]  eta: 0:00:04  lr: 0.002766  min_lr: 0.002766  loss: 2.8989 (2.9584)  weight_decay: 0.0500 (0.0500)  time: 0.6830  data: 0.0001  max mem: 57114
Epoch: [187]  [155/156]  eta: 0:00:00  lr: 0.002765  min_lr: 0.002765  loss: 2.9089 (2.9631)  weight_decay: 0.0500 (0.0500)  time: 0.6834  data: 0.0001  max mem: 57114
Epoch: [187] Total time: 0:01:52 (0.7243 s / it)
Averaged stats: lr: 0.002765  min_lr: 0.002765  loss: 2.9089 (2.9455)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8471 (0.8471)  acc1: 87.5000 (87.5000)  acc5: 98.9583 (98.9583)  time: 2.0512  data: 1.7957  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9747 (0.9420)  acc1: 82.8125 (81.7136)  acc5: 97.3958 (96.1637)  time: 0.5609  data: 0.3592  max mem: 57114
Test: Total time: 0:00:02 (0.5834 s / it)
* Acc@1 81.328 Acc@5 95.872 loss 1.037
Accuracy of the model on the 50000 test images: 81.3%
Max accuracy: 81.33%
Test:  [0/5]  eta: 0:00:10  loss: 0.6441 (0.6441)  acc1: 82.8125 (82.8125)  acc5: 96.3542 (96.3542)  time: 2.0349  data: 1.7915  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7761 (0.7533)  acc1: 79.6875 (77.6215)  acc5: 94.7917 (92.7110)  time: 0.5576  data: 0.3584  max mem: 57114
Test: Total time: 0:00:02 (0.5693 s / it)
* Acc@1 75.553 Acc@5 92.879 loss 0.916
Accuracy of the model EMA on 50000 test images: 75.6%
Max EMA accuracy: 75.55%
Epoch: [188]  [  0/156]  eta: 0:07:27  lr: 0.002765  min_lr: 0.002765  loss: 3.4149 (3.4149)  weight_decay: 0.0500 (0.0500)  time: 2.8686  data: 2.2085  max mem: 57114
Epoch: [188]  [ 10/156]  eta: 0:02:16  lr: 0.002762  min_lr: 0.002762  loss: 3.1533 (3.0511)  weight_decay: 0.0500 (0.0500)  time: 0.9367  data: 0.2011  max mem: 57114
Epoch: [188]  [ 20/156]  eta: 0:01:52  lr: 0.002759  min_lr: 0.002759  loss: 3.1285 (3.0155)  weight_decay: 0.0500 (0.0500)  time: 0.7246  data: 0.0003  max mem: 57114
Epoch: [188]  [ 30/156]  eta: 0:01:41  lr: 0.002756  min_lr: 0.002756  loss: 3.1248 (2.9652)  weight_decay: 0.0500 (0.0500)  time: 0.7289  data: 0.0003  max mem: 57114
Epoch: [188]  [ 40/156]  eta: 0:01:30  lr: 0.002754  min_lr: 0.002754  loss: 2.9775 (2.9518)  weight_decay: 0.0500 (0.0500)  time: 0.7382  data: 0.0003  max mem: 57114
Epoch: [188]  [ 50/156]  eta: 0:01:22  lr: 0.002751  min_lr: 0.002751  loss: 2.9259 (2.9609)  weight_decay: 0.0500 (0.0500)  time: 0.7302  data: 0.0004  max mem: 57114
Epoch: [188]  [ 60/156]  eta: 0:01:13  lr: 0.002748  min_lr: 0.002748  loss: 3.0002 (2.9289)  weight_decay: 0.0500 (0.0500)  time: 0.7354  data: 0.0004  max mem: 57114
Epoch: [188]  [ 70/156]  eta: 0:01:05  lr: 0.002745  min_lr: 0.002745  loss: 3.0678 (2.9543)  weight_decay: 0.0500 (0.0500)  time: 0.7363  data: 0.0004  max mem: 57114
Epoch: [188]  [ 80/156]  eta: 0:00:57  lr: 0.002743  min_lr: 0.002743  loss: 3.0868 (2.9360)  weight_decay: 0.0500 (0.0500)  time: 0.7336  data: 0.0004  max mem: 57114
Epoch: [188]  [ 90/156]  eta: 0:00:49  lr: 0.002740  min_lr: 0.002740  loss: 2.9887 (2.9403)  weight_decay: 0.0500 (0.0500)  time: 0.7260  data: 0.0004  max mem: 57114
Epoch: [188]  [100/156]  eta: 0:00:41  lr: 0.002737  min_lr: 0.002737  loss: 3.1671 (2.9481)  weight_decay: 0.0500 (0.0500)  time: 0.7100  data: 0.0003  max mem: 57114
Epoch: [188]  [110/156]  eta: 0:00:34  lr: 0.002735  min_lr: 0.002735  loss: 3.1340 (2.9461)  weight_decay: 0.0500 (0.0500)  time: 0.7007  data: 0.0003  max mem: 57114
Epoch: [188]  [120/156]  eta: 0:00:26  lr: 0.002732  min_lr: 0.002732  loss: 2.9396 (2.9198)  weight_decay: 0.0500 (0.0500)  time: 0.7062  data: 0.0004  max mem: 57114
Epoch: [188]  [130/156]  eta: 0:00:19  lr: 0.002729  min_lr: 0.002729  loss: 2.9485 (2.9233)  weight_decay: 0.0500 (0.0500)  time: 0.7014  data: 0.0008  max mem: 57114
Epoch: [188]  [140/156]  eta: 0:00:11  lr: 0.002726  min_lr: 0.002726  loss: 3.1210 (2.9260)  weight_decay: 0.0500 (0.0500)  time: 0.6870  data: 0.0007  max mem: 57114
Epoch: [188]  [150/156]  eta: 0:00:04  lr: 0.002724  min_lr: 0.002724  loss: 3.1210 (2.9356)  weight_decay: 0.0500 (0.0500)  time: 0.6829  data: 0.0001  max mem: 57114
Epoch: [188]  [155/156]  eta: 0:00:00  lr: 0.002722  min_lr: 0.002722  loss: 3.1252 (2.9425)  weight_decay: 0.0500 (0.0500)  time: 0.6842  data: 0.0001  max mem: 57114
Epoch: [188] Total time: 0:01:54 (0.7317 s / it)
Averaged stats: lr: 0.002722  min_lr: 0.002722  loss: 3.1252 (2.9316)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8089 (0.8089)  acc1: 88.5417 (88.5417)  acc5: 98.9583 (98.9583)  time: 2.0810  data: 1.8251  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9888 (0.9579)  acc1: 83.3333 (82.8645)  acc5: 97.3958 (96.5473)  time: 0.5669  data: 0.3651  max mem: 57114
Test: Total time: 0:00:02 (0.5909 s / it)
* Acc@1 80.890 Acc@5 95.548 loss 1.050
Accuracy of the model on the 50000 test images: 80.9%
Max accuracy: 81.33%
Test:  [0/5]  eta: 0:00:11  loss: 0.6408 (0.6408)  acc1: 82.8125 (82.8125)  acc5: 96.3542 (96.3542)  time: 2.3211  data: 2.0776  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7718 (0.7486)  acc1: 79.6875 (77.6215)  acc5: 94.7917 (92.8389)  time: 0.6149  data: 0.4156  max mem: 57114
Test: Total time: 0:00:03 (0.6275 s / it)
* Acc@1 75.647 Acc@5 92.945 loss 0.910
Accuracy of the model EMA on 50000 test images: 75.6%
Max EMA accuracy: 75.65%
Epoch: [189]  [  0/156]  eta: 0:06:04  lr: 0.002722  min_lr: 0.002722  loss: 1.6168 (1.6168)  weight_decay: 0.0500 (0.0500)  time: 2.3356  data: 1.6895  max mem: 57114
Epoch: [189]  [ 10/156]  eta: 0:02:04  lr: 0.002719  min_lr: 0.002719  loss: 3.3567 (2.9256)  weight_decay: 0.0500 (0.0500)  time: 0.8560  data: 0.1539  max mem: 57114
Epoch: [189]  [ 20/156]  eta: 0:01:47  lr: 0.002717  min_lr: 0.002717  loss: 3.1696 (2.9788)  weight_decay: 0.0500 (0.0500)  time: 0.7094  data: 0.0003  max mem: 57114
Epoch: [189]  [ 30/156]  eta: 0:01:36  lr: 0.002714  min_lr: 0.002714  loss: 3.0126 (3.0033)  weight_decay: 0.0500 (0.0500)  time: 0.7182  data: 0.0003  max mem: 57114
Epoch: [189]  [ 40/156]  eta: 0:01:27  lr: 0.002711  min_lr: 0.002711  loss: 2.8315 (2.9150)  weight_decay: 0.0500 (0.0500)  time: 0.7272  data: 0.0004  max mem: 57114
Epoch: [189]  [ 50/156]  eta: 0:01:19  lr: 0.002708  min_lr: 0.002708  loss: 2.6264 (2.9079)  weight_decay: 0.0500 (0.0500)  time: 0.7293  data: 0.0004  max mem: 57114
Epoch: [189]  [ 60/156]  eta: 0:01:11  lr: 0.002706  min_lr: 0.002706  loss: 2.9806 (2.9261)  weight_decay: 0.0500 (0.0500)  time: 0.7094  data: 0.0004  max mem: 57114
Epoch: [189]  [ 70/156]  eta: 0:01:03  lr: 0.002703  min_lr: 0.002703  loss: 3.0577 (2.9156)  weight_decay: 0.0500 (0.0500)  time: 0.7107  data: 0.0004  max mem: 57114
Epoch: [189]  [ 80/156]  eta: 0:00:56  lr: 0.002700  min_lr: 0.002700  loss: 2.8834 (2.9305)  weight_decay: 0.0500 (0.0500)  time: 0.7219  data: 0.0004  max mem: 57114
Epoch: [189]  [ 90/156]  eta: 0:00:48  lr: 0.002697  min_lr: 0.002697  loss: 3.0158 (2.9350)  weight_decay: 0.0500 (0.0500)  time: 0.7061  data: 0.0003  max mem: 57114
Epoch: [189]  [100/156]  eta: 0:00:40  lr: 0.002695  min_lr: 0.002695  loss: 2.9662 (2.9223)  weight_decay: 0.0500 (0.0500)  time: 0.7031  data: 0.0004  max mem: 57114
Epoch: [189]  [110/156]  eta: 0:00:33  lr: 0.002692  min_lr: 0.002692  loss: 2.8646 (2.9141)  weight_decay: 0.0500 (0.0500)  time: 0.7062  data: 0.0004  max mem: 57114
Epoch: [189]  [120/156]  eta: 0:00:26  lr: 0.002689  min_lr: 0.002689  loss: 2.8164 (2.9062)  weight_decay: 0.0500 (0.0500)  time: 0.7120  data: 0.0004  max mem: 57114
Epoch: [189]  [130/156]  eta: 0:00:18  lr: 0.002687  min_lr: 0.002687  loss: 2.8608 (2.8951)  weight_decay: 0.0500 (0.0500)  time: 0.6993  data: 0.0009  max mem: 57114
Epoch: [189]  [140/156]  eta: 0:00:11  lr: 0.002684  min_lr: 0.002684  loss: 2.7419 (2.8843)  weight_decay: 0.0500 (0.0500)  time: 0.6859  data: 0.0007  max mem: 57114
Epoch: [189]  [150/156]  eta: 0:00:04  lr: 0.002681  min_lr: 0.002681  loss: 2.8198 (2.8930)  weight_decay: 0.0500 (0.0500)  time: 0.6859  data: 0.0001  max mem: 57114
Epoch: [189]  [155/156]  eta: 0:00:00  lr: 0.002680  min_lr: 0.002680  loss: 3.1009 (2.9056)  weight_decay: 0.0500 (0.0500)  time: 0.6804  data: 0.0001  max mem: 57114
Epoch: [189] Total time: 0:01:52 (0.7195 s / it)
Averaged stats: lr: 0.002680  min_lr: 0.002680  loss: 3.1009 (2.9340)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8119 (0.8119)  acc1: 88.5417 (88.5417)  acc5: 98.9583 (98.9583)  time: 2.1671  data: 1.9109  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0093 (0.9483)  acc1: 80.7292 (81.0742)  acc5: 96.8750 (96.0358)  time: 0.5842  data: 0.3823  max mem: 57114
Test: Total time: 0:00:03 (0.6082 s / it)
* Acc@1 80.896 Acc@5 95.824 loss 1.027
Accuracy of the model on the 50000 test images: 80.9%
Max accuracy: 81.33%
Test:  [0/5]  eta: 0:00:11  loss: 0.6376 (0.6376)  acc1: 83.3333 (83.3333)  acc5: 96.3542 (96.3542)  time: 2.3419  data: 2.0984  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7678 (0.7442)  acc1: 80.2083 (77.8772)  acc5: 94.7917 (92.9668)  time: 0.6190  data: 0.4198  max mem: 57114
Test: Total time: 0:00:03 (0.6288 s / it)
* Acc@1 75.751 Acc@5 93.001 loss 0.905
Accuracy of the model EMA on 50000 test images: 75.8%
Max EMA accuracy: 75.75%
Epoch: [190]  [  0/156]  eta: 0:07:48  lr: 0.002680  min_lr: 0.002680  loss: 2.9301 (2.9301)  weight_decay: 0.0500 (0.0500)  time: 3.0052  data: 2.3503  max mem: 57114
Epoch: [190]  [ 10/156]  eta: 0:02:15  lr: 0.002677  min_lr: 0.002677  loss: 3.0585 (2.9608)  weight_decay: 0.0500 (0.0500)  time: 0.9251  data: 0.2139  max mem: 57114
Epoch: [190]  [ 20/156]  eta: 0:01:53  lr: 0.002674  min_lr: 0.002674  loss: 2.9337 (2.8705)  weight_decay: 0.0500 (0.0500)  time: 0.7259  data: 0.0003  max mem: 57114
Epoch: [190]  [ 30/156]  eta: 0:01:40  lr: 0.002671  min_lr: 0.002671  loss: 2.9337 (2.8998)  weight_decay: 0.0500 (0.0500)  time: 0.7304  data: 0.0003  max mem: 57114
Epoch: [190]  [ 40/156]  eta: 0:01:30  lr: 0.002669  min_lr: 0.002669  loss: 3.0810 (2.9468)  weight_decay: 0.0500 (0.0500)  time: 0.7206  data: 0.0005  max mem: 57114
Epoch: [190]  [ 50/156]  eta: 0:01:21  lr: 0.002666  min_lr: 0.002666  loss: 3.0810 (2.9582)  weight_decay: 0.0500 (0.0500)  time: 0.7177  data: 0.0005  max mem: 57114
Epoch: [190]  [ 60/156]  eta: 0:01:12  lr: 0.002663  min_lr: 0.002663  loss: 3.1615 (2.9807)  weight_decay: 0.0500 (0.0500)  time: 0.7089  data: 0.0004  max mem: 57114
Epoch: [190]  [ 70/156]  eta: 0:01:04  lr: 0.002661  min_lr: 0.002661  loss: 2.9762 (2.9652)  weight_decay: 0.0500 (0.0500)  time: 0.7102  data: 0.0004  max mem: 57114
Epoch: [190]  [ 80/156]  eta: 0:00:56  lr: 0.002658  min_lr: 0.002658  loss: 2.9762 (2.9767)  weight_decay: 0.0500 (0.0500)  time: 0.7098  data: 0.0004  max mem: 57114
Epoch: [190]  [ 90/156]  eta: 0:00:48  lr: 0.002655  min_lr: 0.002655  loss: 2.9255 (2.9711)  weight_decay: 0.0500 (0.0500)  time: 0.7042  data: 0.0004  max mem: 57114
Epoch: [190]  [100/156]  eta: 0:00:41  lr: 0.002652  min_lr: 0.002652  loss: 2.8812 (2.9519)  weight_decay: 0.0500 (0.0500)  time: 0.7111  data: 0.0004  max mem: 57114
Epoch: [190]  [110/156]  eta: 0:00:33  lr: 0.002650  min_lr: 0.002650  loss: 2.8812 (2.9487)  weight_decay: 0.0500 (0.0500)  time: 0.7122  data: 0.0003  max mem: 57114
Epoch: [190]  [120/156]  eta: 0:00:26  lr: 0.002647  min_lr: 0.002647  loss: 2.8791 (2.9469)  weight_decay: 0.0500 (0.0500)  time: 0.6988  data: 0.0004  max mem: 57114
Epoch: [190]  [130/156]  eta: 0:00:18  lr: 0.002644  min_lr: 0.002644  loss: 2.9948 (2.9470)  weight_decay: 0.0500 (0.0500)  time: 0.6878  data: 0.0009  max mem: 57114
Epoch: [190]  [140/156]  eta: 0:00:11  lr: 0.002642  min_lr: 0.002642  loss: 2.9572 (2.9230)  weight_decay: 0.0500 (0.0500)  time: 0.6853  data: 0.0007  max mem: 57114
Epoch: [190]  [150/156]  eta: 0:00:04  lr: 0.002639  min_lr: 0.002639  loss: 2.8177 (2.9315)  weight_decay: 0.0500 (0.0500)  time: 0.6813  data: 0.0001  max mem: 57114
Epoch: [190]  [155/156]  eta: 0:00:00  lr: 0.002638  min_lr: 0.002638  loss: 3.0295 (2.9385)  weight_decay: 0.0500 (0.0500)  time: 0.6824  data: 0.0001  max mem: 57114
Epoch: [190] Total time: 0:01:52 (0.7232 s / it)
Averaged stats: lr: 0.002638  min_lr: 0.002638  loss: 3.0295 (2.9323)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5760 (0.5760)  acc1: 89.0625 (89.0625)  acc5: 98.9583 (98.9583)  time: 2.0705  data: 1.8149  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8151 (0.7596)  acc1: 80.7292 (81.8414)  acc5: 96.8750 (96.1637)  time: 0.5648  data: 0.3630  max mem: 57114
Test: Total time: 0:00:02 (0.5871 s / it)
* Acc@1 81.212 Acc@5 95.638 loss 0.860
Accuracy of the model on the 50000 test images: 81.2%
Max accuracy: 81.33%
Test:  [0/5]  eta: 0:00:11  loss: 0.6346 (0.6346)  acc1: 83.3333 (83.3333)  acc5: 96.3542 (96.3542)  time: 2.3527  data: 2.1092  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7636 (0.7399)  acc1: 80.2083 (77.8772)  acc5: 94.7917 (93.0946)  time: 0.6212  data: 0.4219  max mem: 57114
Test: Total time: 0:00:03 (0.6372 s / it)
* Acc@1 75.841 Acc@5 93.065 loss 0.899
Accuracy of the model EMA on 50000 test images: 75.8%
Max EMA accuracy: 75.84%
Epoch: [191]  [  0/156]  eta: 0:07:51  lr: 0.002637  min_lr: 0.002637  loss: 3.2129 (3.2129)  weight_decay: 0.0500 (0.0500)  time: 3.0211  data: 2.3730  max mem: 57114
Epoch: [191]  [ 10/156]  eta: 0:02:18  lr: 0.002635  min_lr: 0.002635  loss: 3.1058 (3.0430)  weight_decay: 0.0500 (0.0500)  time: 0.9486  data: 0.2161  max mem: 57114
Epoch: [191]  [ 20/156]  eta: 0:01:55  lr: 0.002632  min_lr: 0.002632  loss: 3.1058 (2.9967)  weight_decay: 0.0500 (0.0500)  time: 0.7387  data: 0.0004  max mem: 57114
Epoch: [191]  [ 30/156]  eta: 0:01:41  lr: 0.002629  min_lr: 0.002629  loss: 3.1270 (3.0117)  weight_decay: 0.0500 (0.0500)  time: 0.7224  data: 0.0004  max mem: 57114
Epoch: [191]  [ 40/156]  eta: 0:01:30  lr: 0.002626  min_lr: 0.002626  loss: 3.1244 (3.0103)  weight_decay: 0.0500 (0.0500)  time: 0.7083  data: 0.0003  max mem: 57114
Epoch: [191]  [ 50/156]  eta: 0:01:21  lr: 0.002624  min_lr: 0.002624  loss: 3.0968 (2.9928)  weight_decay: 0.0500 (0.0500)  time: 0.7154  data: 0.0004  max mem: 57114
Epoch: [191]  [ 60/156]  eta: 0:01:13  lr: 0.002621  min_lr: 0.002621  loss: 3.0598 (2.9826)  weight_decay: 0.0500 (0.0500)  time: 0.7301  data: 0.0004  max mem: 57114
Epoch: [191]  [ 70/156]  eta: 0:01:04  lr: 0.002618  min_lr: 0.002618  loss: 3.0581 (2.9779)  weight_decay: 0.0500 (0.0500)  time: 0.7213  data: 0.0004  max mem: 57114
Epoch: [191]  [ 80/156]  eta: 0:00:57  lr: 0.002616  min_lr: 0.002616  loss: 3.0581 (2.9677)  weight_decay: 0.0500 (0.0500)  time: 0.7154  data: 0.0004  max mem: 57114
Epoch: [191]  [ 90/156]  eta: 0:00:49  lr: 0.002613  min_lr: 0.002613  loss: 2.7591 (2.9416)  weight_decay: 0.0500 (0.0500)  time: 0.7173  data: 0.0004  max mem: 57114
Epoch: [191]  [100/156]  eta: 0:00:41  lr: 0.002610  min_lr: 0.002610  loss: 2.7591 (2.9313)  weight_decay: 0.0500 (0.0500)  time: 0.7079  data: 0.0003  max mem: 57114
Epoch: [191]  [110/156]  eta: 0:00:34  lr: 0.002608  min_lr: 0.002608  loss: 3.0424 (2.9393)  weight_decay: 0.0500 (0.0500)  time: 0.7049  data: 0.0004  max mem: 57114
Epoch: [191]  [120/156]  eta: 0:00:26  lr: 0.002605  min_lr: 0.002605  loss: 3.0523 (2.9363)  weight_decay: 0.0500 (0.0500)  time: 0.7052  data: 0.0004  max mem: 57114
Epoch: [191]  [130/156]  eta: 0:00:19  lr: 0.002602  min_lr: 0.002602  loss: 3.0098 (2.9270)  weight_decay: 0.0500 (0.0500)  time: 0.7030  data: 0.0009  max mem: 57114
Epoch: [191]  [140/156]  eta: 0:00:11  lr: 0.002599  min_lr: 0.002599  loss: 2.8368 (2.9265)  weight_decay: 0.0500 (0.0500)  time: 0.6920  data: 0.0007  max mem: 57114
Epoch: [191]  [150/156]  eta: 0:00:04  lr: 0.002597  min_lr: 0.002597  loss: 3.1057 (2.9330)  weight_decay: 0.0500 (0.0500)  time: 0.6876  data: 0.0001  max mem: 57114
Epoch: [191]  [155/156]  eta: 0:00:00  lr: 0.002595  min_lr: 0.002595  loss: 3.1177 (2.9450)  weight_decay: 0.0500 (0.0500)  time: 0.6851  data: 0.0001  max mem: 57114
Epoch: [191] Total time: 0:01:53 (0.7279 s / it)
Averaged stats: lr: 0.002595  min_lr: 0.002595  loss: 3.1177 (2.9275)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8171 (0.8171)  acc1: 87.5000 (87.5000)  acc5: 98.4375 (98.4375)  time: 2.0279  data: 1.7715  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9928 (0.9386)  acc1: 82.2917 (82.2251)  acc5: 98.4375 (96.4194)  time: 0.5563  data: 0.3543  max mem: 57114
Test: Total time: 0:00:02 (0.5786 s / it)
* Acc@1 81.234 Acc@5 95.746 loss 1.022
Accuracy of the model on the 50000 test images: 81.2%
Max accuracy: 81.33%
Test:  [0/5]  eta: 0:00:11  loss: 0.6319 (0.6319)  acc1: 83.3333 (83.3333)  acc5: 96.3542 (96.3542)  time: 2.2611  data: 2.0175  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7601 (0.7360)  acc1: 80.7292 (78.1330)  acc5: 94.7917 (93.0946)  time: 0.6029  data: 0.4036  max mem: 57114
Test: Total time: 0:00:03 (0.6145 s / it)
* Acc@1 75.949 Acc@5 93.135 loss 0.894
Accuracy of the model EMA on 50000 test images: 75.9%
Max EMA accuracy: 75.95%
Epoch: [192]  [  0/156]  eta: 0:07:22  lr: 0.002595  min_lr: 0.002595  loss: 3.5172 (3.5172)  weight_decay: 0.0500 (0.0500)  time: 2.8347  data: 2.1864  max mem: 57114
Epoch: [192]  [ 10/156]  eta: 0:02:13  lr: 0.002592  min_lr: 0.002592  loss: 3.1484 (3.0348)  weight_decay: 0.0500 (0.0500)  time: 0.9175  data: 0.1990  max mem: 57114
Epoch: [192]  [ 20/156]  eta: 0:01:52  lr: 0.002590  min_lr: 0.002590  loss: 3.1484 (3.0754)  weight_decay: 0.0500 (0.0500)  time: 0.7282  data: 0.0003  max mem: 57114
Epoch: [192]  [ 30/156]  eta: 0:01:40  lr: 0.002587  min_lr: 0.002587  loss: 3.1216 (2.9568)  weight_decay: 0.0500 (0.0500)  time: 0.7315  data: 0.0004  max mem: 57114
Epoch: [192]  [ 40/156]  eta: 0:01:30  lr: 0.002584  min_lr: 0.002584  loss: 3.1216 (2.9642)  weight_decay: 0.0500 (0.0500)  time: 0.7320  data: 0.0003  max mem: 57114
Epoch: [192]  [ 50/156]  eta: 0:01:21  lr: 0.002582  min_lr: 0.002582  loss: 3.1801 (2.9703)  weight_decay: 0.0500 (0.0500)  time: 0.7343  data: 0.0004  max mem: 57114
Epoch: [192]  [ 60/156]  eta: 0:01:13  lr: 0.002579  min_lr: 0.002579  loss: 3.0599 (2.9729)  weight_decay: 0.0500 (0.0500)  time: 0.7258  data: 0.0004  max mem: 57114
Epoch: [192]  [ 70/156]  eta: 0:01:05  lr: 0.002576  min_lr: 0.002576  loss: 3.0599 (2.9610)  weight_decay: 0.0500 (0.0500)  time: 0.7147  data: 0.0004  max mem: 57114
Epoch: [192]  [ 80/156]  eta: 0:00:57  lr: 0.002574  min_lr: 0.002574  loss: 3.0084 (2.9523)  weight_decay: 0.0500 (0.0500)  time: 0.7100  data: 0.0004  max mem: 57114
Epoch: [192]  [ 90/156]  eta: 0:00:49  lr: 0.002571  min_lr: 0.002571  loss: 2.8182 (2.9372)  weight_decay: 0.0500 (0.0500)  time: 0.7052  data: 0.0004  max mem: 57114
Epoch: [192]  [100/156]  eta: 0:00:41  lr: 0.002568  min_lr: 0.002568  loss: 2.9116 (2.9322)  weight_decay: 0.0500 (0.0500)  time: 0.7079  data: 0.0003  max mem: 57114
Epoch: [192]  [110/156]  eta: 0:00:33  lr: 0.002566  min_lr: 0.002566  loss: 3.0430 (2.9499)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0004  max mem: 57114
Epoch: [192]  [120/156]  eta: 0:00:26  lr: 0.002563  min_lr: 0.002563  loss: 2.9443 (2.9358)  weight_decay: 0.0500 (0.0500)  time: 0.7127  data: 0.0004  max mem: 57114
Epoch: [192]  [130/156]  eta: 0:00:19  lr: 0.002560  min_lr: 0.002560  loss: 2.8208 (2.9299)  weight_decay: 0.0500 (0.0500)  time: 0.7051  data: 0.0009  max mem: 57114
Epoch: [192]  [140/156]  eta: 0:00:11  lr: 0.002558  min_lr: 0.002558  loss: 2.9404 (2.9268)  weight_decay: 0.0500 (0.0500)  time: 0.6861  data: 0.0007  max mem: 57114
Epoch: [192]  [150/156]  eta: 0:00:04  lr: 0.002555  min_lr: 0.002555  loss: 3.1219 (2.9449)  weight_decay: 0.0500 (0.0500)  time: 0.6823  data: 0.0001  max mem: 57114
Epoch: [192]  [155/156]  eta: 0:00:00  lr: 0.002554  min_lr: 0.002554  loss: 3.0483 (2.9398)  weight_decay: 0.0500 (0.0500)  time: 0.6843  data: 0.0001  max mem: 57114
Epoch: [192] Total time: 0:01:53 (0.7277 s / it)
Averaged stats: lr: 0.002554  min_lr: 0.002554  loss: 3.0483 (2.9368)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6746 (0.6746)  acc1: 86.4583 (86.4583)  acc5: 98.4375 (98.4375)  time: 2.1485  data: 1.8923  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9711 (0.8065)  acc1: 80.2083 (81.2020)  acc5: 96.8750 (96.5473)  time: 0.5804  data: 0.3785  max mem: 57114
Test: Total time: 0:00:03 (0.6042 s / it)
* Acc@1 81.522 Acc@5 96.006 loss 0.888
Accuracy of the model on the 50000 test images: 81.5%
Max accuracy: 81.52%
Test:  [0/5]  eta: 0:00:09  loss: 0.6292 (0.6292)  acc1: 83.3333 (83.3333)  acc5: 96.3542 (96.3542)  time: 1.9958  data: 1.7524  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7566 (0.7321)  acc1: 80.7292 (78.0051)  acc5: 94.2708 (93.0946)  time: 0.5498  data: 0.3506  max mem: 57114
Test: Total time: 0:00:02 (0.5615 s / it)
* Acc@1 76.025 Acc@5 93.193 loss 0.889
Accuracy of the model EMA on 50000 test images: 76.0%
Max EMA accuracy: 76.03%
Epoch: [193]  [  0/156]  eta: 0:07:05  lr: 0.002553  min_lr: 0.002553  loss: 2.9424 (2.9424)  weight_decay: 0.0500 (0.0500)  time: 2.7300  data: 2.0727  max mem: 57114
Epoch: [193]  [ 10/156]  eta: 0:02:12  lr: 0.002551  min_lr: 0.002551  loss: 3.0541 (3.0533)  weight_decay: 0.0500 (0.0500)  time: 0.9103  data: 0.1887  max mem: 57114
Epoch: [193]  [ 20/156]  eta: 0:01:54  lr: 0.002548  min_lr: 0.002548  loss: 3.1877 (3.0662)  weight_decay: 0.0500 (0.0500)  time: 0.7446  data: 0.0003  max mem: 57114
Epoch: [193]  [ 30/156]  eta: 0:01:41  lr: 0.002545  min_lr: 0.002545  loss: 3.1827 (3.0333)  weight_decay: 0.0500 (0.0500)  time: 0.7500  data: 0.0004  max mem: 57114
Epoch: [193]  [ 40/156]  eta: 0:01:31  lr: 0.002543  min_lr: 0.002543  loss: 3.0496 (2.9775)  weight_decay: 0.0500 (0.0500)  time: 0.7420  data: 0.0003  max mem: 57114
Epoch: [193]  [ 50/156]  eta: 0:01:22  lr: 0.002540  min_lr: 0.002540  loss: 3.0496 (2.9814)  weight_decay: 0.0500 (0.0500)  time: 0.7317  data: 0.0004  max mem: 57114
Epoch: [193]  [ 60/156]  eta: 0:01:13  lr: 0.002537  min_lr: 0.002537  loss: 3.0594 (2.9651)  weight_decay: 0.0500 (0.0500)  time: 0.7160  data: 0.0004  max mem: 57114
Epoch: [193]  [ 70/156]  eta: 0:01:05  lr: 0.002534  min_lr: 0.002534  loss: 3.0594 (2.9816)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0004  max mem: 57114
Epoch: [193]  [ 80/156]  eta: 0:00:57  lr: 0.002532  min_lr: 0.002532  loss: 3.0897 (2.9815)  weight_decay: 0.0500 (0.0500)  time: 0.7225  data: 0.0003  max mem: 57114
Epoch: [193]  [ 90/156]  eta: 0:00:49  lr: 0.002529  min_lr: 0.002529  loss: 2.9258 (2.9811)  weight_decay: 0.0500 (0.0500)  time: 0.7228  data: 0.0003  max mem: 57114
Epoch: [193]  [100/156]  eta: 0:00:41  lr: 0.002526  min_lr: 0.002526  loss: 3.1482 (2.9946)  weight_decay: 0.0500 (0.0500)  time: 0.7109  data: 0.0004  max mem: 57114
Epoch: [193]  [110/156]  eta: 0:00:34  lr: 0.002524  min_lr: 0.002524  loss: 3.1410 (2.9861)  weight_decay: 0.0500 (0.0500)  time: 0.7114  data: 0.0004  max mem: 57114
Epoch: [193]  [120/156]  eta: 0:00:26  lr: 0.002521  min_lr: 0.002521  loss: 2.7180 (2.9464)  weight_decay: 0.0500 (0.0500)  time: 0.7009  data: 0.0004  max mem: 57114
Epoch: [193]  [130/156]  eta: 0:00:19  lr: 0.002518  min_lr: 0.002518  loss: 2.7349 (2.9461)  weight_decay: 0.0500 (0.0500)  time: 0.6903  data: 0.0008  max mem: 57114
Epoch: [193]  [140/156]  eta: 0:00:11  lr: 0.002516  min_lr: 0.002516  loss: 2.9160 (2.9448)  weight_decay: 0.0500 (0.0500)  time: 0.6890  data: 0.0007  max mem: 57114
Epoch: [193]  [150/156]  eta: 0:00:04  lr: 0.002513  min_lr: 0.002513  loss: 2.9906 (2.9573)  weight_decay: 0.0500 (0.0500)  time: 0.6846  data: 0.0001  max mem: 57114
Epoch: [193]  [155/156]  eta: 0:00:00  lr: 0.002512  min_lr: 0.002512  loss: 3.1041 (2.9614)  weight_decay: 0.0500 (0.0500)  time: 0.6831  data: 0.0001  max mem: 57114
Epoch: [193] Total time: 0:01:53 (0.7299 s / it)
Averaged stats: lr: 0.002512  min_lr: 0.002512  loss: 3.1041 (2.9178)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7891 (0.7891)  acc1: 88.0208 (88.0208)  acc5: 98.9583 (98.9583)  time: 2.1399  data: 1.8833  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9602 (0.8992)  acc1: 81.2500 (81.3299)  acc5: 96.8750 (96.5473)  time: 0.5787  data: 0.3767  max mem: 57114
Test: Total time: 0:00:02 (0.5982 s / it)
* Acc@1 81.420 Acc@5 95.784 loss 0.966
Accuracy of the model on the 50000 test images: 81.4%
Max accuracy: 81.52%
Test:  [0/5]  eta: 0:00:11  loss: 0.6264 (0.6264)  acc1: 83.3333 (83.3333)  acc5: 96.8750 (96.8750)  time: 2.3493  data: 2.1057  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7531 (0.7283)  acc1: 80.7292 (78.1330)  acc5: 94.2708 (93.2225)  time: 0.6205  data: 0.4212  max mem: 57114
Test: Total time: 0:00:03 (0.6321 s / it)
* Acc@1 76.113 Acc@5 93.252 loss 0.885
Accuracy of the model EMA on 50000 test images: 76.1%
Max EMA accuracy: 76.11%
Epoch: [194]  [  0/156]  eta: 0:07:05  lr: 0.002512  min_lr: 0.002512  loss: 2.7745 (2.7745)  weight_decay: 0.0500 (0.0500)  time: 2.7295  data: 2.0796  max mem: 57114
Epoch: [194]  [ 10/156]  eta: 0:02:10  lr: 0.002509  min_lr: 0.002509  loss: 2.9269 (2.8751)  weight_decay: 0.0500 (0.0500)  time: 0.8956  data: 0.1893  max mem: 57114
Epoch: [194]  [ 20/156]  eta: 0:01:50  lr: 0.002506  min_lr: 0.002506  loss: 2.8822 (2.8190)  weight_decay: 0.0500 (0.0500)  time: 0.7198  data: 0.0003  max mem: 57114
Epoch: [194]  [ 30/156]  eta: 0:01:39  lr: 0.002504  min_lr: 0.002504  loss: 2.8246 (2.8060)  weight_decay: 0.0500 (0.0500)  time: 0.7302  data: 0.0003  max mem: 57114
Epoch: [194]  [ 40/156]  eta: 0:01:29  lr: 0.002501  min_lr: 0.002501  loss: 3.0370 (2.8592)  weight_decay: 0.0500 (0.0500)  time: 0.7324  data: 0.0003  max mem: 57114
Epoch: [194]  [ 50/156]  eta: 0:01:20  lr: 0.002498  min_lr: 0.002498  loss: 3.0739 (2.8707)  weight_decay: 0.0500 (0.0500)  time: 0.7252  data: 0.0004  max mem: 57114
Epoch: [194]  [ 60/156]  eta: 0:01:12  lr: 0.002496  min_lr: 0.002496  loss: 3.0984 (2.8994)  weight_decay: 0.0500 (0.0500)  time: 0.7229  data: 0.0004  max mem: 57114
Epoch: [194]  [ 70/156]  eta: 0:01:04  lr: 0.002493  min_lr: 0.002493  loss: 3.0984 (2.8972)  weight_decay: 0.0500 (0.0500)  time: 0.7191  data: 0.0004  max mem: 57114
Epoch: [194]  [ 80/156]  eta: 0:00:56  lr: 0.002490  min_lr: 0.002490  loss: 3.0049 (2.9028)  weight_decay: 0.0500 (0.0500)  time: 0.7084  data: 0.0004  max mem: 57114
Epoch: [194]  [ 90/156]  eta: 0:00:48  lr: 0.002488  min_lr: 0.002488  loss: 2.9636 (2.8937)  weight_decay: 0.0500 (0.0500)  time: 0.7093  data: 0.0004  max mem: 57114
Epoch: [194]  [100/156]  eta: 0:00:41  lr: 0.002485  min_lr: 0.002485  loss: 2.9389 (2.9067)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0004  max mem: 57114
Epoch: [194]  [110/156]  eta: 0:00:33  lr: 0.002482  min_lr: 0.002482  loss: 3.0241 (2.8973)  weight_decay: 0.0500 (0.0500)  time: 0.7006  data: 0.0004  max mem: 57114
Epoch: [194]  [120/156]  eta: 0:00:26  lr: 0.002480  min_lr: 0.002480  loss: 3.0241 (2.9009)  weight_decay: 0.0500 (0.0500)  time: 0.6980  data: 0.0004  max mem: 57114
Epoch: [194]  [130/156]  eta: 0:00:18  lr: 0.002477  min_lr: 0.002477  loss: 3.0424 (2.8988)  weight_decay: 0.0500 (0.0500)  time: 0.6947  data: 0.0008  max mem: 57114
Epoch: [194]  [140/156]  eta: 0:00:11  lr: 0.002474  min_lr: 0.002474  loss: 2.9280 (2.8848)  weight_decay: 0.0500 (0.0500)  time: 0.6883  data: 0.0007  max mem: 57114
Epoch: [194]  [150/156]  eta: 0:00:04  lr: 0.002472  min_lr: 0.002472  loss: 2.5248 (2.8734)  weight_decay: 0.0500 (0.0500)  time: 0.6817  data: 0.0001  max mem: 57114
Epoch: [194]  [155/156]  eta: 0:00:00  lr: 0.002470  min_lr: 0.002470  loss: 2.8091 (2.8860)  weight_decay: 0.0500 (0.0500)  time: 0.6836  data: 0.0001  max mem: 57114
Epoch: [194] Total time: 0:01:52 (0.7238 s / it)
Averaged stats: lr: 0.002470  min_lr: 0.002470  loss: 2.8091 (2.9121)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7127 (0.7127)  acc1: 90.1042 (90.1042)  acc5: 98.9583 (98.9583)  time: 2.0438  data: 1.7877  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9491 (0.8650)  acc1: 83.3333 (82.8645)  acc5: 97.3958 (96.2916)  time: 0.5595  data: 0.3576  max mem: 57114
Test: Total time: 0:00:02 (0.5845 s / it)
* Acc@1 81.432 Acc@5 95.806 loss 0.958
Accuracy of the model on the 50000 test images: 81.4%
Max accuracy: 81.52%
Test:  [0/5]  eta: 0:00:11  loss: 0.6235 (0.6235)  acc1: 83.3333 (83.3333)  acc5: 96.8750 (96.8750)  time: 2.2930  data: 2.0494  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7497 (0.7246)  acc1: 80.7292 (78.1330)  acc5: 94.2708 (93.2225)  time: 0.6093  data: 0.4100  max mem: 57114
Test: Total time: 0:00:03 (0.6238 s / it)
* Acc@1 76.171 Acc@5 93.310 loss 0.880
Accuracy of the model EMA on 50000 test images: 76.2%
Max EMA accuracy: 76.17%
Epoch: [195]  [  0/156]  eta: 0:09:00  lr: 0.002470  min_lr: 0.002470  loss: 3.2102 (3.2102)  weight_decay: 0.0500 (0.0500)  time: 3.4640  data: 2.8160  max mem: 57114
Epoch: [195]  [ 10/156]  eta: 0:02:21  lr: 0.002467  min_lr: 0.002467  loss: 3.0470 (3.0290)  weight_decay: 0.0500 (0.0500)  time: 0.9674  data: 0.2562  max mem: 57114
Epoch: [195]  [ 20/156]  eta: 0:01:55  lr: 0.002465  min_lr: 0.002465  loss: 3.0690 (3.0217)  weight_decay: 0.0500 (0.0500)  time: 0.7157  data: 0.0003  max mem: 57114
Epoch: [195]  [ 30/156]  eta: 0:01:41  lr: 0.002462  min_lr: 0.002462  loss: 2.8441 (2.9027)  weight_decay: 0.0500 (0.0500)  time: 0.7133  data: 0.0003  max mem: 57114
Epoch: [195]  [ 40/156]  eta: 0:01:30  lr: 0.002459  min_lr: 0.002459  loss: 2.8441 (2.9184)  weight_decay: 0.0500 (0.0500)  time: 0.7141  data: 0.0004  max mem: 57114
Epoch: [195]  [ 50/156]  eta: 0:01:21  lr: 0.002457  min_lr: 0.002457  loss: 3.0509 (2.9395)  weight_decay: 0.0500 (0.0500)  time: 0.7168  data: 0.0004  max mem: 57114
Epoch: [195]  [ 60/156]  eta: 0:01:13  lr: 0.002454  min_lr: 0.002454  loss: 3.0346 (2.9486)  weight_decay: 0.0500 (0.0500)  time: 0.7201  data: 0.0004  max mem: 57114
Epoch: [195]  [ 70/156]  eta: 0:01:05  lr: 0.002451  min_lr: 0.002451  loss: 3.1718 (2.9790)  weight_decay: 0.0500 (0.0500)  time: 0.7237  data: 0.0004  max mem: 57114
Epoch: [195]  [ 80/156]  eta: 0:00:57  lr: 0.002449  min_lr: 0.002449  loss: 3.0486 (2.9664)  weight_decay: 0.0500 (0.0500)  time: 0.7206  data: 0.0003  max mem: 57114
Epoch: [195]  [ 90/156]  eta: 0:00:49  lr: 0.002446  min_lr: 0.002446  loss: 2.7741 (2.9410)  weight_decay: 0.0500 (0.0500)  time: 0.7088  data: 0.0004  max mem: 57114
Epoch: [195]  [100/156]  eta: 0:00:41  lr: 0.002443  min_lr: 0.002443  loss: 2.6596 (2.9198)  weight_decay: 0.0500 (0.0500)  time: 0.7037  data: 0.0004  max mem: 57114
Epoch: [195]  [110/156]  eta: 0:00:33  lr: 0.002441  min_lr: 0.002441  loss: 2.5992 (2.8909)  weight_decay: 0.0500 (0.0500)  time: 0.7000  data: 0.0004  max mem: 57114
Epoch: [195]  [120/156]  eta: 0:00:26  lr: 0.002438  min_lr: 0.002438  loss: 2.6265 (2.8919)  weight_decay: 0.0500 (0.0500)  time: 0.7047  data: 0.0004  max mem: 57114
Epoch: [195]  [130/156]  eta: 0:00:19  lr: 0.002435  min_lr: 0.002435  loss: 3.0298 (2.8959)  weight_decay: 0.0500 (0.0500)  time: 0.7051  data: 0.0008  max mem: 57114
Epoch: [195]  [140/156]  eta: 0:00:11  lr: 0.002433  min_lr: 0.002433  loss: 2.8706 (2.8844)  weight_decay: 0.0500 (0.0500)  time: 0.6943  data: 0.0007  max mem: 57114
Epoch: [195]  [150/156]  eta: 0:00:04  lr: 0.002430  min_lr: 0.002430  loss: 2.8632 (2.8782)  weight_decay: 0.0500 (0.0500)  time: 0.6887  data: 0.0001  max mem: 57114
Epoch: [195]  [155/156]  eta: 0:00:00  lr: 0.002429  min_lr: 0.002429  loss: 2.7916 (2.8740)  weight_decay: 0.0500 (0.0500)  time: 0.6852  data: 0.0001  max mem: 57114
Epoch: [195] Total time: 0:01:53 (0.7276 s / it)
Averaged stats: lr: 0.002429  min_lr: 0.002429  loss: 2.7916 (2.9055)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5383 (0.5383)  acc1: 88.0208 (88.0208)  acc5: 98.9583 (98.9583)  time: 2.0225  data: 1.7667  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7504 (0.6948)  acc1: 83.8542 (82.7366)  acc5: 96.3542 (96.2916)  time: 0.5554  data: 0.3534  max mem: 57114
Test: Total time: 0:00:02 (0.5757 s / it)
* Acc@1 81.877 Acc@5 95.962 loss 0.782
Accuracy of the model on the 50000 test images: 81.9%
Max accuracy: 81.88%
Test:  [0/5]  eta: 0:00:10  loss: 0.6208 (0.6208)  acc1: 83.8542 (83.8542)  acc5: 96.8750 (96.8750)  time: 2.0773  data: 1.8338  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7463 (0.7209)  acc1: 80.2083 (78.2609)  acc5: 94.2708 (93.2225)  time: 0.5662  data: 0.3669  max mem: 57114
Test: Total time: 0:00:02 (0.5771 s / it)
* Acc@1 76.287 Acc@5 93.356 loss 0.876
Accuracy of the model EMA on 50000 test images: 76.3%
Max EMA accuracy: 76.29%
Epoch: [196]  [  0/156]  eta: 0:06:28  lr: 0.002429  min_lr: 0.002429  loss: 3.0527 (3.0527)  weight_decay: 0.0500 (0.0500)  time: 2.4901  data: 1.8423  max mem: 57114
Epoch: [196]  [ 10/156]  eta: 0:02:05  lr: 0.002426  min_lr: 0.002426  loss: 3.0527 (2.9372)  weight_decay: 0.0500 (0.0500)  time: 0.8624  data: 0.1678  max mem: 57114
Epoch: [196]  [ 20/156]  eta: 0:01:48  lr: 0.002423  min_lr: 0.002423  loss: 3.1208 (2.9783)  weight_decay: 0.0500 (0.0500)  time: 0.7125  data: 0.0003  max mem: 57114
Epoch: [196]  [ 30/156]  eta: 0:01:37  lr: 0.002421  min_lr: 0.002421  loss: 3.1034 (2.9392)  weight_decay: 0.0500 (0.0500)  time: 0.7195  data: 0.0003  max mem: 57114
Epoch: [196]  [ 40/156]  eta: 0:01:28  lr: 0.002418  min_lr: 0.002418  loss: 3.0120 (2.9238)  weight_decay: 0.0500 (0.0500)  time: 0.7299  data: 0.0004  max mem: 57114
Epoch: [196]  [ 50/156]  eta: 0:01:20  lr: 0.002415  min_lr: 0.002415  loss: 3.0814 (2.9454)  weight_decay: 0.0500 (0.0500)  time: 0.7455  data: 0.0004  max mem: 57114
Epoch: [196]  [ 60/156]  eta: 0:01:12  lr: 0.002413  min_lr: 0.002413  loss: 3.1084 (2.9261)  weight_decay: 0.0500 (0.0500)  time: 0.7360  data: 0.0004  max mem: 57114
Epoch: [196]  [ 70/156]  eta: 0:01:04  lr: 0.002410  min_lr: 0.002410  loss: 2.9302 (2.9142)  weight_decay: 0.0500 (0.0500)  time: 0.7227  data: 0.0005  max mem: 57114
Epoch: [196]  [ 80/156]  eta: 0:00:56  lr: 0.002407  min_lr: 0.002407  loss: 3.1020 (2.9528)  weight_decay: 0.0500 (0.0500)  time: 0.7265  data: 0.0004  max mem: 57114
Epoch: [196]  [ 90/156]  eta: 0:00:49  lr: 0.002405  min_lr: 0.002405  loss: 3.1294 (2.9565)  weight_decay: 0.0500 (0.0500)  time: 0.7170  data: 0.0004  max mem: 57114
Epoch: [196]  [100/156]  eta: 0:00:41  lr: 0.002402  min_lr: 0.002402  loss: 3.0553 (2.9403)  weight_decay: 0.0500 (0.0500)  time: 0.7013  data: 0.0003  max mem: 57114
Epoch: [196]  [110/156]  eta: 0:00:33  lr: 0.002400  min_lr: 0.002400  loss: 2.4005 (2.8952)  weight_decay: 0.0500 (0.0500)  time: 0.6945  data: 0.0004  max mem: 57114
Epoch: [196]  [120/156]  eta: 0:00:26  lr: 0.002397  min_lr: 0.002397  loss: 2.4685 (2.8938)  weight_decay: 0.0500 (0.0500)  time: 0.6993  data: 0.0004  max mem: 57114
Epoch: [196]  [130/156]  eta: 0:00:18  lr: 0.002394  min_lr: 0.002394  loss: 2.8681 (2.8905)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0009  max mem: 57114
Epoch: [196]  [140/156]  eta: 0:00:11  lr: 0.002392  min_lr: 0.002392  loss: 2.9842 (2.8991)  weight_decay: 0.0500 (0.0500)  time: 0.6899  data: 0.0007  max mem: 57114
Epoch: [196]  [150/156]  eta: 0:00:04  lr: 0.002389  min_lr: 0.002389  loss: 3.0327 (2.9047)  weight_decay: 0.0500 (0.0500)  time: 0.6769  data: 0.0001  max mem: 57114
Epoch: [196]  [155/156]  eta: 0:00:00  lr: 0.002388  min_lr: 0.002388  loss: 2.9874 (2.9053)  weight_decay: 0.0500 (0.0500)  time: 0.6785  data: 0.0001  max mem: 57114
Epoch: [196] Total time: 0:01:52 (0.7242 s / it)
Averaged stats: lr: 0.002388  min_lr: 0.002388  loss: 2.9874 (2.9020)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7510 (0.7510)  acc1: 86.9792 (86.9792)  acc5: 98.4375 (98.4375)  time: 2.0660  data: 1.8104  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9748 (0.8688)  acc1: 81.2500 (81.9693)  acc5: 96.8750 (96.2916)  time: 0.5639  data: 0.3622  max mem: 57114
Test: Total time: 0:00:02 (0.5831 s / it)
* Acc@1 81.566 Acc@5 95.950 loss 0.935
Accuracy of the model on the 50000 test images: 81.6%
Max accuracy: 81.88%
Test:  [0/5]  eta: 0:00:11  loss: 0.6180 (0.6180)  acc1: 84.3750 (84.3750)  acc5: 96.8750 (96.8750)  time: 2.2461  data: 2.0026  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7433 (0.7174)  acc1: 79.6875 (78.2609)  acc5: 94.7917 (93.3504)  time: 0.5999  data: 0.4006  max mem: 57114
Test: Total time: 0:00:03 (0.6098 s / it)
* Acc@1 76.357 Acc@5 93.400 loss 0.871
Accuracy of the model EMA on 50000 test images: 76.4%
Max EMA accuracy: 76.36%
Epoch: [197]  [  0/156]  eta: 0:05:54  lr: 0.002387  min_lr: 0.002387  loss: 2.7355 (2.7355)  weight_decay: 0.0500 (0.0500)  time: 2.2705  data: 1.5962  max mem: 57114
Epoch: [197]  [ 10/156]  eta: 0:02:05  lr: 0.002385  min_lr: 0.002385  loss: 2.8965 (2.8591)  weight_decay: 0.0500 (0.0500)  time: 0.8584  data: 0.1454  max mem: 57114
Epoch: [197]  [ 20/156]  eta: 0:01:47  lr: 0.002382  min_lr: 0.002382  loss: 3.0176 (2.8965)  weight_decay: 0.0500 (0.0500)  time: 0.7135  data: 0.0003  max mem: 57114
Epoch: [197]  [ 30/156]  eta: 0:01:36  lr: 0.002380  min_lr: 0.002380  loss: 3.0778 (2.9096)  weight_decay: 0.0500 (0.0500)  time: 0.7158  data: 0.0003  max mem: 57114
Epoch: [197]  [ 40/156]  eta: 0:01:27  lr: 0.002377  min_lr: 0.002377  loss: 3.1921 (2.9584)  weight_decay: 0.0500 (0.0500)  time: 0.7178  data: 0.0004  max mem: 57114
Epoch: [197]  [ 50/156]  eta: 0:01:19  lr: 0.002374  min_lr: 0.002374  loss: 3.1876 (2.9541)  weight_decay: 0.0500 (0.0500)  time: 0.7162  data: 0.0004  max mem: 57114
Epoch: [197]  [ 60/156]  eta: 0:01:11  lr: 0.002372  min_lr: 0.002372  loss: 3.1111 (2.9693)  weight_decay: 0.0500 (0.0500)  time: 0.7159  data: 0.0004  max mem: 57114
Epoch: [197]  [ 70/156]  eta: 0:01:03  lr: 0.002369  min_lr: 0.002369  loss: 3.1111 (2.9778)  weight_decay: 0.0500 (0.0500)  time: 0.7125  data: 0.0004  max mem: 57114
Epoch: [197]  [ 80/156]  eta: 0:00:55  lr: 0.002366  min_lr: 0.002366  loss: 3.0948 (2.9910)  weight_decay: 0.0500 (0.0500)  time: 0.7064  data: 0.0004  max mem: 57114
Epoch: [197]  [ 90/156]  eta: 0:00:48  lr: 0.002364  min_lr: 0.002364  loss: 3.1648 (3.0092)  weight_decay: 0.0500 (0.0500)  time: 0.7052  data: 0.0004  max mem: 57114
Epoch: [197]  [100/156]  eta: 0:00:40  lr: 0.002361  min_lr: 0.002361  loss: 3.1648 (3.0144)  weight_decay: 0.0500 (0.0500)  time: 0.7101  data: 0.0004  max mem: 57114
Epoch: [197]  [110/156]  eta: 0:00:33  lr: 0.002359  min_lr: 0.002359  loss: 3.0876 (3.0001)  weight_decay: 0.0500 (0.0500)  time: 0.7141  data: 0.0004  max mem: 57114
Epoch: [197]  [120/156]  eta: 0:00:26  lr: 0.002356  min_lr: 0.002356  loss: 3.0353 (2.9988)  weight_decay: 0.0500 (0.0500)  time: 0.7130  data: 0.0004  max mem: 57114
Epoch: [197]  [130/156]  eta: 0:00:18  lr: 0.002353  min_lr: 0.002353  loss: 3.0530 (3.0081)  weight_decay: 0.0500 (0.0500)  time: 0.7035  data: 0.0009  max mem: 57114
Epoch: [197]  [140/156]  eta: 0:00:11  lr: 0.002351  min_lr: 0.002351  loss: 3.1119 (3.0153)  weight_decay: 0.0500 (0.0500)  time: 0.6901  data: 0.0007  max mem: 57114
Epoch: [197]  [150/156]  eta: 0:00:04  lr: 0.002348  min_lr: 0.002348  loss: 3.0556 (3.0019)  weight_decay: 0.0500 (0.0500)  time: 0.6824  data: 0.0001  max mem: 57114
Epoch: [197]  [155/156]  eta: 0:00:00  lr: 0.002347  min_lr: 0.002347  loss: 3.0879 (3.0037)  weight_decay: 0.0500 (0.0500)  time: 0.6802  data: 0.0001  max mem: 57114
Epoch: [197] Total time: 0:01:52 (0.7190 s / it)
Averaged stats: lr: 0.002347  min_lr: 0.002347  loss: 3.0879 (2.8980)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7143 (0.7143)  acc1: 86.9792 (86.9792)  acc5: 98.9583 (98.9583)  time: 2.0536  data: 1.7976  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9401 (0.8636)  acc1: 82.8125 (81.4578)  acc5: 96.8750 (96.5473)  time: 0.5613  data: 0.3596  max mem: 57114
Test: Total time: 0:00:02 (0.5852 s / it)
* Acc@1 81.849 Acc@5 95.974 loss 0.926
Accuracy of the model on the 50000 test images: 81.8%
Max accuracy: 81.88%
Test:  [0/5]  eta: 0:00:11  loss: 0.6155 (0.6155)  acc1: 84.3750 (84.3750)  acc5: 96.3542 (96.3542)  time: 2.2523  data: 2.0087  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7405 (0.7141)  acc1: 79.6875 (78.2609)  acc5: 94.7917 (93.0946)  time: 0.6011  data: 0.4018  max mem: 57114
Test: Total time: 0:00:03 (0.6128 s / it)
* Acc@1 76.415 Acc@5 93.428 loss 0.867
Accuracy of the model EMA on 50000 test images: 76.4%
Max EMA accuracy: 76.41%
Epoch: [198]  [  0/156]  eta: 0:06:06  lr: 0.002346  min_lr: 0.002346  loss: 2.9177 (2.9177)  weight_decay: 0.0500 (0.0500)  time: 2.3526  data: 1.7030  max mem: 57114
Epoch: [198]  [ 10/156]  eta: 0:02:10  lr: 0.002344  min_lr: 0.002344  loss: 2.8825 (2.8014)  weight_decay: 0.0500 (0.0500)  time: 0.8927  data: 0.1551  max mem: 57114
Epoch: [198]  [ 20/156]  eta: 0:01:50  lr: 0.002341  min_lr: 0.002341  loss: 2.8825 (2.8690)  weight_decay: 0.0500 (0.0500)  time: 0.7344  data: 0.0003  max mem: 57114
Epoch: [198]  [ 30/156]  eta: 0:01:38  lr: 0.002339  min_lr: 0.002339  loss: 2.9950 (2.9114)  weight_decay: 0.0500 (0.0500)  time: 0.7182  data: 0.0004  max mem: 57114
Epoch: [198]  [ 40/156]  eta: 0:01:28  lr: 0.002336  min_lr: 0.002336  loss: 3.0232 (2.9233)  weight_decay: 0.0500 (0.0500)  time: 0.7173  data: 0.0004  max mem: 57114
Epoch: [198]  [ 50/156]  eta: 0:01:19  lr: 0.002333  min_lr: 0.002333  loss: 3.1520 (2.9292)  weight_decay: 0.0500 (0.0500)  time: 0.7129  data: 0.0004  max mem: 57114
Epoch: [198]  [ 60/156]  eta: 0:01:11  lr: 0.002331  min_lr: 0.002331  loss: 3.1388 (2.9044)  weight_decay: 0.0500 (0.0500)  time: 0.7109  data: 0.0004  max mem: 57114
Epoch: [198]  [ 70/156]  eta: 0:01:04  lr: 0.002328  min_lr: 0.002328  loss: 2.9498 (2.8804)  weight_decay: 0.0500 (0.0500)  time: 0.7205  data: 0.0004  max mem: 57114
Epoch: [198]  [ 80/156]  eta: 0:00:56  lr: 0.002326  min_lr: 0.002326  loss: 2.9498 (2.8705)  weight_decay: 0.0500 (0.0500)  time: 0.7220  data: 0.0004  max mem: 57114
Epoch: [198]  [ 90/156]  eta: 0:00:48  lr: 0.002323  min_lr: 0.002323  loss: 2.9194 (2.8733)  weight_decay: 0.0500 (0.0500)  time: 0.7143  data: 0.0004  max mem: 57114
Epoch: [198]  [100/156]  eta: 0:00:41  lr: 0.002320  min_lr: 0.002320  loss: 2.9231 (2.8735)  weight_decay: 0.0500 (0.0500)  time: 0.7042  data: 0.0004  max mem: 57114
Epoch: [198]  [110/156]  eta: 0:00:33  lr: 0.002318  min_lr: 0.002318  loss: 2.9609 (2.8746)  weight_decay: 0.0500 (0.0500)  time: 0.7035  data: 0.0004  max mem: 57114
Epoch: [198]  [120/156]  eta: 0:00:26  lr: 0.002315  min_lr: 0.002315  loss: 2.8412 (2.8652)  weight_decay: 0.0500 (0.0500)  time: 0.7014  data: 0.0004  max mem: 57114
Epoch: [198]  [130/156]  eta: 0:00:18  lr: 0.002312  min_lr: 0.002312  loss: 3.0300 (2.8705)  weight_decay: 0.0500 (0.0500)  time: 0.6972  data: 0.0005  max mem: 57114
Epoch: [198]  [140/156]  eta: 0:00:11  lr: 0.002310  min_lr: 0.002310  loss: 3.0766 (2.8682)  weight_decay: 0.0500 (0.0500)  time: 0.6932  data: 0.0003  max mem: 57114
Epoch: [198]  [150/156]  eta: 0:00:04  lr: 0.002307  min_lr: 0.002307  loss: 2.8819 (2.8754)  weight_decay: 0.0500 (0.0500)  time: 0.6873  data: 0.0001  max mem: 57114
Epoch: [198]  [155/156]  eta: 0:00:00  lr: 0.002306  min_lr: 0.002306  loss: 3.1004 (2.8887)  weight_decay: 0.0500 (0.0500)  time: 0.6838  data: 0.0001  max mem: 57114
Epoch: [198] Total time: 0:01:52 (0.7222 s / it)
Averaged stats: lr: 0.002306  min_lr: 0.002306  loss: 3.1004 (2.8911)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.8073 (0.8073)  acc1: 90.6250 (90.6250)  acc5: 98.4375 (98.4375)  time: 2.1079  data: 1.8521  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 1.0319 (0.9654)  acc1: 83.8542 (82.9923)  acc5: 96.8750 (96.1637)  time: 0.5722  data: 0.3705  max mem: 57114
Test: Total time: 0:00:02 (0.5953 s / it)
* Acc@1 81.839 Acc@5 95.868 loss 1.035
Accuracy of the model on the 50000 test images: 81.8%
Max accuracy: 81.88%
Test:  [0/5]  eta: 0:00:11  loss: 0.6130 (0.6130)  acc1: 84.3750 (84.3750)  acc5: 96.3542 (96.3542)  time: 2.3075  data: 2.0639  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7379 (0.7112)  acc1: 79.6875 (78.2609)  acc5: 94.7917 (93.0946)  time: 0.6122  data: 0.4129  max mem: 57114
Test: Total time: 0:00:03 (0.6255 s / it)
* Acc@1 76.507 Acc@5 93.460 loss 0.863
Accuracy of the model EMA on 50000 test images: 76.5%
Max EMA accuracy: 76.51%
Epoch: [199]  [  0/156]  eta: 0:08:13  lr: 0.002306  min_lr: 0.002306  loss: 3.0608 (3.0608)  weight_decay: 0.0500 (0.0500)  time: 3.1616  data: 2.5098  max mem: 57114
Epoch: [199]  [ 10/156]  eta: 0:02:18  lr: 0.002303  min_lr: 0.002303  loss: 3.3678 (3.0794)  weight_decay: 0.0500 (0.0500)  time: 0.9468  data: 0.2284  max mem: 57114
Epoch: [199]  [ 20/156]  eta: 0:01:54  lr: 0.002301  min_lr: 0.002301  loss: 3.1734 (3.0585)  weight_decay: 0.0500 (0.0500)  time: 0.7282  data: 0.0003  max mem: 57114
Epoch: [199]  [ 30/156]  eta: 0:01:40  lr: 0.002298  min_lr: 0.002298  loss: 3.0988 (3.0199)  weight_decay: 0.0500 (0.0500)  time: 0.7198  data: 0.0003  max mem: 57114
Epoch: [199]  [ 40/156]  eta: 0:01:30  lr: 0.002295  min_lr: 0.002295  loss: 3.0067 (3.0226)  weight_decay: 0.0500 (0.0500)  time: 0.7182  data: 0.0003  max mem: 57114
Epoch: [199]  [ 50/156]  eta: 0:01:21  lr: 0.002293  min_lr: 0.002293  loss: 3.0067 (3.0037)  weight_decay: 0.0500 (0.0500)  time: 0.7251  data: 0.0004  max mem: 57114
Epoch: [199]  [ 60/156]  eta: 0:01:13  lr: 0.002290  min_lr: 0.002290  loss: 3.0141 (3.0114)  weight_decay: 0.0500 (0.0500)  time: 0.7307  data: 0.0004  max mem: 57114
Epoch: [199]  [ 70/156]  eta: 0:01:05  lr: 0.002288  min_lr: 0.002288  loss: 3.0639 (3.0066)  weight_decay: 0.0500 (0.0500)  time: 0.7216  data: 0.0004  max mem: 57114
Epoch: [199]  [ 80/156]  eta: 0:00:57  lr: 0.002285  min_lr: 0.002285  loss: 2.9728 (2.9919)  weight_decay: 0.0500 (0.0500)  time: 0.7036  data: 0.0004  max mem: 57114
Epoch: [199]  [ 90/156]  eta: 0:00:49  lr: 0.002282  min_lr: 0.002282  loss: 3.0590 (2.9990)  weight_decay: 0.0500 (0.0500)  time: 0.7080  data: 0.0004  max mem: 57114
Epoch: [199]  [100/156]  eta: 0:00:41  lr: 0.002280  min_lr: 0.002280  loss: 3.1377 (3.0062)  weight_decay: 0.0500 (0.0500)  time: 0.7105  data: 0.0004  max mem: 57114
Epoch: [199]  [110/156]  eta: 0:00:33  lr: 0.002277  min_lr: 0.002277  loss: 3.1296 (3.0014)  weight_decay: 0.0500 (0.0500)  time: 0.6996  data: 0.0004  max mem: 57114
Epoch: [199]  [120/156]  eta: 0:00:26  lr: 0.002275  min_lr: 0.002275  loss: 3.0393 (2.9828)  weight_decay: 0.0500 (0.0500)  time: 0.6979  data: 0.0004  max mem: 57114
Epoch: [199]  [130/156]  eta: 0:00:19  lr: 0.002272  min_lr: 0.002272  loss: 2.9767 (2.9802)  weight_decay: 0.0500 (0.0500)  time: 0.7044  data: 0.0009  max mem: 57114
Epoch: [199]  [140/156]  eta: 0:00:11  lr: 0.002269  min_lr: 0.002269  loss: 2.9767 (2.9675)  weight_decay: 0.0500 (0.0500)  time: 0.6928  data: 0.0007  max mem: 57114
Epoch: [199]  [150/156]  eta: 0:00:04  lr: 0.002267  min_lr: 0.002267  loss: 2.7153 (2.9498)  weight_decay: 0.0500 (0.0500)  time: 0.6854  data: 0.0001  max mem: 57114
Epoch: [199]  [155/156]  eta: 0:00:00  lr: 0.002265  min_lr: 0.002265  loss: 2.8253 (2.9476)  weight_decay: 0.0500 (0.0500)  time: 0.6906  data: 0.0001  max mem: 57114
Epoch: [199] Total time: 0:01:53 (0.7277 s / it)
Averaged stats: lr: 0.002265  min_lr: 0.002265  loss: 2.8253 (2.8888)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5975 (0.5975)  acc1: 89.0625 (89.0625)  acc5: 98.9583 (98.9583)  time: 2.0919  data: 1.8359  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8204 (0.7653)  acc1: 84.3750 (83.6317)  acc5: 98.4375 (96.8031)  time: 0.5691  data: 0.3672  max mem: 57114
Test: Total time: 0:00:02 (0.5908 s / it)
* Acc@1 81.616 Acc@5 95.984 loss 0.864
Accuracy of the model on the 50000 test images: 81.6%
Max accuracy: 81.88%
Test:  [0/5]  eta: 0:00:11  loss: 0.6106 (0.6106)  acc1: 83.8542 (83.8542)  acc5: 96.8750 (96.8750)  time: 2.2918  data: 2.0483  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7354 (0.7082)  acc1: 79.6875 (78.1330)  acc5: 95.3125 (93.4783)  time: 0.6090  data: 0.4097  max mem: 57114
Test: Total time: 0:00:03 (0.6190 s / it)
* Acc@1 76.576 Acc@5 93.506 loss 0.860
Accuracy of the model EMA on 50000 test images: 76.6%
Max EMA accuracy: 76.58%
Epoch: [200]  [  0/156]  eta: 0:07:06  lr: 0.002265  min_lr: 0.002265  loss: 2.2504 (2.2504)  weight_decay: 0.0500 (0.0500)  time: 2.7314  data: 2.0789  max mem: 57114
Epoch: [200]  [ 10/156]  eta: 0:02:09  lr: 0.002263  min_lr: 0.002263  loss: 2.8075 (2.7156)  weight_decay: 0.0500 (0.0500)  time: 0.8889  data: 0.1893  max mem: 57114
Epoch: [200]  [ 20/156]  eta: 0:01:49  lr: 0.002260  min_lr: 0.002260  loss: 2.8606 (2.7959)  weight_decay: 0.0500 (0.0500)  time: 0.7102  data: 0.0003  max mem: 57114
Epoch: [200]  [ 30/156]  eta: 0:01:38  lr: 0.002257  min_lr: 0.002257  loss: 2.9564 (2.8376)  weight_decay: 0.0500 (0.0500)  time: 0.7182  data: 0.0003  max mem: 57114
Epoch: [200]  [ 40/156]  eta: 0:01:29  lr: 0.002255  min_lr: 0.002255  loss: 2.9796 (2.8750)  weight_decay: 0.0500 (0.0500)  time: 0.7272  data: 0.0003  max mem: 57114
Epoch: [200]  [ 50/156]  eta: 0:01:20  lr: 0.002252  min_lr: 0.002252  loss: 3.0076 (2.8945)  weight_decay: 0.0500 (0.0500)  time: 0.7334  data: 0.0003  max mem: 57114
Epoch: [200]  [ 60/156]  eta: 0:01:12  lr: 0.002250  min_lr: 0.002250  loss: 3.0519 (2.8776)  weight_decay: 0.0500 (0.0500)  time: 0.7282  data: 0.0004  max mem: 57114
Epoch: [200]  [ 70/156]  eta: 0:01:04  lr: 0.002247  min_lr: 0.002247  loss: 3.0162 (2.8736)  weight_decay: 0.0500 (0.0500)  time: 0.7075  data: 0.0004  max mem: 57114
Epoch: [200]  [ 80/156]  eta: 0:00:56  lr: 0.002244  min_lr: 0.002244  loss: 3.0293 (2.8897)  weight_decay: 0.0500 (0.0500)  time: 0.7006  data: 0.0004  max mem: 57114
Epoch: [200]  [ 90/156]  eta: 0:00:48  lr: 0.002242  min_lr: 0.002242  loss: 3.0600 (2.8910)  weight_decay: 0.0500 (0.0500)  time: 0.7056  data: 0.0004  max mem: 57114
Epoch: [200]  [100/156]  eta: 0:00:41  lr: 0.002239  min_lr: 0.002239  loss: 2.6617 (2.8588)  weight_decay: 0.0500 (0.0500)  time: 0.6989  data: 0.0004  max mem: 57114
Epoch: [200]  [110/156]  eta: 0:00:33  lr: 0.002237  min_lr: 0.002237  loss: 2.8961 (2.8605)  weight_decay: 0.0500 (0.0500)  time: 0.6998  data: 0.0003  max mem: 57114
Epoch: [200]  [120/156]  eta: 0:00:26  lr: 0.002234  min_lr: 0.002234  loss: 2.8925 (2.8471)  weight_decay: 0.0500 (0.0500)  time: 0.7150  data: 0.0003  max mem: 57114
Epoch: [200]  [130/156]  eta: 0:00:18  lr: 0.002232  min_lr: 0.002232  loss: 2.8475 (2.8606)  weight_decay: 0.0500 (0.0500)  time: 0.7123  data: 0.0005  max mem: 57114
Epoch: [200]  [140/156]  eta: 0:00:11  lr: 0.002229  min_lr: 0.002229  loss: 3.1995 (2.8892)  weight_decay: 0.0500 (0.0500)  time: 0.6899  data: 0.0004  max mem: 57114
Epoch: [200]  [150/156]  eta: 0:00:04  lr: 0.002226  min_lr: 0.002226  loss: 3.0945 (2.8797)  weight_decay: 0.0500 (0.0500)  time: 0.6813  data: 0.0001  max mem: 57114
Epoch: [200]  [155/156]  eta: 0:00:00  lr: 0.002225  min_lr: 0.002225  loss: 2.9625 (2.8811)  weight_decay: 0.0500 (0.0500)  time: 0.6866  data: 0.0001  max mem: 57114
Epoch: [200] Total time: 0:01:52 (0.7229 s / it)
Averaged stats: lr: 0.002225  min_lr: 0.002225  loss: 2.9625 (2.8840)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6576 (0.6576)  acc1: 89.0625 (89.0625)  acc5: 98.9583 (98.9583)  time: 2.1290  data: 1.8728  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8788 (0.7900)  acc1: 84.3750 (82.9923)  acc5: 98.9583 (97.5703)  time: 0.5767  data: 0.3746  max mem: 57114
Test: Total time: 0:00:02 (0.5987 s / it)
* Acc@1 81.977 Acc@5 96.000 loss 0.893
Accuracy of the model on the 50000 test images: 82.0%
Max accuracy: 81.98%
Test:  [0/5]  eta: 0:00:10  loss: 0.6085 (0.6085)  acc1: 84.3750 (84.3750)  acc5: 96.8750 (96.8750)  time: 2.0602  data: 1.8169  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7336 (0.7055)  acc1: 79.6875 (78.2609)  acc5: 95.3125 (93.6061)  time: 0.5626  data: 0.3635  max mem: 57114
Test: Total time: 0:00:02 (0.5762 s / it)
* Acc@1 76.624 Acc@5 93.544 loss 0.856
Accuracy of the model EMA on 50000 test images: 76.6%
Max EMA accuracy: 76.62%
Epoch: [201]  [  0/156]  eta: 0:07:32  lr: 0.002225  min_lr: 0.002225  loss: 2.8912 (2.8912)  weight_decay: 0.0500 (0.0500)  time: 2.9033  data: 2.2354  max mem: 57114
Epoch: [201]  [ 10/156]  eta: 0:02:12  lr: 0.002222  min_lr: 0.002222  loss: 3.1337 (2.9810)  weight_decay: 0.0500 (0.0500)  time: 0.9059  data: 0.2035  max mem: 57114
Epoch: [201]  [ 20/156]  eta: 0:01:51  lr: 0.002220  min_lr: 0.002220  loss: 3.0247 (2.9404)  weight_decay: 0.0500 (0.0500)  time: 0.7177  data: 0.0003  max mem: 57114
Epoch: [201]  [ 30/156]  eta: 0:01:40  lr: 0.002217  min_lr: 0.002217  loss: 3.0003 (2.9756)  weight_decay: 0.0500 (0.0500)  time: 0.7385  data: 0.0003  max mem: 57114
Epoch: [201]  [ 40/156]  eta: 0:01:31  lr: 0.002215  min_lr: 0.002215  loss: 3.0165 (2.9291)  weight_decay: 0.0500 (0.0500)  time: 0.7477  data: 0.0003  max mem: 57114
Epoch: [201]  [ 50/156]  eta: 0:01:21  lr: 0.002212  min_lr: 0.002212  loss: 3.0435 (2.9196)  weight_decay: 0.0500 (0.0500)  time: 0.7253  data: 0.0003  max mem: 57114
Epoch: [201]  [ 60/156]  eta: 0:01:13  lr: 0.002209  min_lr: 0.002209  loss: 2.9208 (2.9063)  weight_decay: 0.0500 (0.0500)  time: 0.7267  data: 0.0004  max mem: 57114
Epoch: [201]  [ 70/156]  eta: 0:01:05  lr: 0.002207  min_lr: 0.002207  loss: 2.9598 (2.9391)  weight_decay: 0.0500 (0.0500)  time: 0.7327  data: 0.0004  max mem: 57114
Epoch: [201]  [ 80/156]  eta: 0:00:57  lr: 0.002204  min_lr: 0.002204  loss: 3.1282 (2.9394)  weight_decay: 0.0500 (0.0500)  time: 0.7131  data: 0.0004  max mem: 57114
Epoch: [201]  [ 90/156]  eta: 0:00:49  lr: 0.002202  min_lr: 0.002202  loss: 2.9318 (2.9186)  weight_decay: 0.0500 (0.0500)  time: 0.7066  data: 0.0004  max mem: 57114
Epoch: [201]  [100/156]  eta: 0:00:41  lr: 0.002199  min_lr: 0.002199  loss: 2.8449 (2.9142)  weight_decay: 0.0500 (0.0500)  time: 0.7140  data: 0.0004  max mem: 57114
Epoch: [201]  [110/156]  eta: 0:00:34  lr: 0.002197  min_lr: 0.002197  loss: 3.0445 (2.9244)  weight_decay: 0.0500 (0.0500)  time: 0.7218  data: 0.0003  max mem: 57114
Epoch: [201]  [120/156]  eta: 0:00:26  lr: 0.002194  min_lr: 0.002194  loss: 3.0445 (2.9306)  weight_decay: 0.0500 (0.0500)  time: 0.7111  data: 0.0003  max mem: 57114
Epoch: [201]  [130/156]  eta: 0:00:19  lr: 0.002191  min_lr: 0.002191  loss: 3.1183 (2.9280)  weight_decay: 0.0500 (0.0500)  time: 0.6997  data: 0.0008  max mem: 57114
Epoch: [201]  [140/156]  eta: 0:00:11  lr: 0.002189  min_lr: 0.002189  loss: 2.9625 (2.9180)  weight_decay: 0.0500 (0.0500)  time: 0.6874  data: 0.0007  max mem: 57114
Epoch: [201]  [150/156]  eta: 0:00:04  lr: 0.002186  min_lr: 0.002186  loss: 2.9533 (2.9182)  weight_decay: 0.0500 (0.0500)  time: 0.6844  data: 0.0001  max mem: 57114
Epoch: [201]  [155/156]  eta: 0:00:00  lr: 0.002185  min_lr: 0.002185  loss: 2.8550 (2.9094)  weight_decay: 0.0500 (0.0500)  time: 0.6856  data: 0.0001  max mem: 57114
Epoch: [201] Total time: 0:01:53 (0.7299 s / it)
Averaged stats: lr: 0.002185  min_lr: 0.002185  loss: 2.8550 (2.8841)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7064 (0.7064)  acc1: 88.5417 (88.5417)  acc5: 98.9583 (98.9583)  time: 2.0233  data: 1.7647  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9368 (0.8301)  acc1: 80.7292 (81.9693)  acc5: 97.3958 (96.5473)  time: 0.5553  data: 0.3530  max mem: 57114
Test: Total time: 0:00:02 (0.5783 s / it)
* Acc@1 81.951 Acc@5 96.092 loss 0.907
Accuracy of the model on the 50000 test images: 82.0%
Max accuracy: 81.98%
Test:  [0/5]  eta: 0:00:11  loss: 0.6067 (0.6067)  acc1: 84.3750 (84.3750)  acc5: 96.8750 (96.8750)  time: 2.2651  data: 2.0216  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7316 (0.7030)  acc1: 79.6875 (78.2609)  acc5: 95.3125 (93.6061)  time: 0.6037  data: 0.4044  max mem: 57114
Test: Total time: 0:00:03 (0.6143 s / it)
* Acc@1 76.680 Acc@5 93.580 loss 0.853
Accuracy of the model EMA on 50000 test images: 76.7%
Max EMA accuracy: 76.68%
Epoch: [202]  [  0/156]  eta: 0:08:16  lr: 0.002185  min_lr: 0.002185  loss: 2.6517 (2.6517)  weight_decay: 0.0500 (0.0500)  time: 3.1844  data: 2.5384  max mem: 57114
Epoch: [202]  [ 10/156]  eta: 0:02:20  lr: 0.002182  min_lr: 0.002182  loss: 3.1549 (3.0522)  weight_decay: 0.0500 (0.0500)  time: 0.9596  data: 0.2311  max mem: 57114
Epoch: [202]  [ 20/156]  eta: 0:01:55  lr: 0.002180  min_lr: 0.002180  loss: 3.0674 (2.9805)  weight_decay: 0.0500 (0.0500)  time: 0.7291  data: 0.0004  max mem: 57114
Epoch: [202]  [ 30/156]  eta: 0:01:42  lr: 0.002177  min_lr: 0.002177  loss: 3.0674 (2.9238)  weight_decay: 0.0500 (0.0500)  time: 0.7277  data: 0.0004  max mem: 57114
Epoch: [202]  [ 40/156]  eta: 0:01:31  lr: 0.002175  min_lr: 0.002175  loss: 2.5730 (2.8619)  weight_decay: 0.0500 (0.0500)  time: 0.7289  data: 0.0003  max mem: 57114
Epoch: [202]  [ 50/156]  eta: 0:01:22  lr: 0.002172  min_lr: 0.002172  loss: 2.9037 (2.8618)  weight_decay: 0.0500 (0.0500)  time: 0.7196  data: 0.0004  max mem: 57114
Epoch: [202]  [ 60/156]  eta: 0:01:13  lr: 0.002169  min_lr: 0.002169  loss: 3.0188 (2.8902)  weight_decay: 0.0500 (0.0500)  time: 0.7093  data: 0.0004  max mem: 57114
Epoch: [202]  [ 70/156]  eta: 0:01:05  lr: 0.002167  min_lr: 0.002167  loss: 3.1123 (2.9212)  weight_decay: 0.0500 (0.0500)  time: 0.7110  data: 0.0003  max mem: 57114
Epoch: [202]  [ 80/156]  eta: 0:00:56  lr: 0.002164  min_lr: 0.002164  loss: 3.1864 (2.9333)  weight_decay: 0.0500 (0.0500)  time: 0.7044  data: 0.0004  max mem: 57114
Epoch: [202]  [ 90/156]  eta: 0:00:49  lr: 0.002162  min_lr: 0.002162  loss: 2.9442 (2.9070)  weight_decay: 0.0500 (0.0500)  time: 0.7027  data: 0.0004  max mem: 57114
Epoch: [202]  [100/156]  eta: 0:00:41  lr: 0.002159  min_lr: 0.002159  loss: 2.6525 (2.8851)  weight_decay: 0.0500 (0.0500)  time: 0.7197  data: 0.0004  max mem: 57114
Epoch: [202]  [110/156]  eta: 0:00:34  lr: 0.002157  min_lr: 0.002157  loss: 2.5961 (2.8804)  weight_decay: 0.0500 (0.0500)  time: 0.7148  data: 0.0003  max mem: 57114
Epoch: [202]  [120/156]  eta: 0:00:26  lr: 0.002154  min_lr: 0.002154  loss: 2.7600 (2.8744)  weight_decay: 0.0500 (0.0500)  time: 0.7003  data: 0.0003  max mem: 57114
Epoch: [202]  [130/156]  eta: 0:00:19  lr: 0.002152  min_lr: 0.002152  loss: 2.9744 (2.8843)  weight_decay: 0.0500 (0.0500)  time: 0.6901  data: 0.0004  max mem: 57114
Epoch: [202]  [140/156]  eta: 0:00:11  lr: 0.002149  min_lr: 0.002149  loss: 3.0788 (2.8942)  weight_decay: 0.0500 (0.0500)  time: 0.6850  data: 0.0003  max mem: 57114
Epoch: [202]  [150/156]  eta: 0:00:04  lr: 0.002146  min_lr: 0.002146  loss: 3.0705 (2.8849)  weight_decay: 0.0500 (0.0500)  time: 0.6838  data: 0.0001  max mem: 57114
Epoch: [202]  [155/156]  eta: 0:00:00  lr: 0.002145  min_lr: 0.002145  loss: 3.0556 (2.8901)  weight_decay: 0.0500 (0.0500)  time: 0.6885  data: 0.0001  max mem: 57114
Epoch: [202] Total time: 0:01:53 (0.7267 s / it)
Averaged stats: lr: 0.002145  min_lr: 0.002145  loss: 3.0556 (2.8757)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6629 (0.6629)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.1883  data: 1.9318  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8884 (0.8298)  acc1: 81.7708 (82.6087)  acc5: 97.3958 (96.6752)  time: 0.5884  data: 0.3864  max mem: 57114
Test: Total time: 0:00:03 (0.6095 s / it)
* Acc@1 81.781 Acc@5 96.010 loss 0.917
Accuracy of the model on the 50000 test images: 81.8%
Max accuracy: 81.98%
Test:  [0/5]  eta: 0:00:12  loss: 0.6048 (0.6048)  acc1: 84.3750 (84.3750)  acc5: 96.8750 (96.8750)  time: 2.4120  data: 2.1684  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7295 (0.7006)  acc1: 79.6875 (78.2609)  acc5: 95.3125 (93.6061)  time: 0.6331  data: 0.4338  max mem: 57114
Test: Total time: 0:00:03 (0.6446 s / it)
* Acc@1 76.720 Acc@5 93.602 loss 0.850
Accuracy of the model EMA on 50000 test images: 76.7%
Max EMA accuracy: 76.72%
Epoch: [203]  [  0/156]  eta: 0:08:07  lr: 0.002145  min_lr: 0.002145  loss: 3.2289 (3.2289)  weight_decay: 0.0500 (0.0500)  time: 3.1255  data: 2.4649  max mem: 57114
Epoch: [203]  [ 10/156]  eta: 0:02:19  lr: 0.002142  min_lr: 0.002142  loss: 3.0699 (2.9277)  weight_decay: 0.0500 (0.0500)  time: 0.9534  data: 0.2243  max mem: 57114
Epoch: [203]  [ 20/156]  eta: 0:01:54  lr: 0.002140  min_lr: 0.002140  loss: 3.0299 (2.9520)  weight_decay: 0.0500 (0.0500)  time: 0.7311  data: 0.0003  max mem: 57114
Epoch: [203]  [ 30/156]  eta: 0:01:41  lr: 0.002137  min_lr: 0.002137  loss: 2.9963 (2.9025)  weight_decay: 0.0500 (0.0500)  time: 0.7229  data: 0.0003  max mem: 57114
Epoch: [203]  [ 40/156]  eta: 0:01:30  lr: 0.002135  min_lr: 0.002135  loss: 2.9963 (2.9342)  weight_decay: 0.0500 (0.0500)  time: 0.7150  data: 0.0003  max mem: 57114
Epoch: [203]  [ 50/156]  eta: 0:01:21  lr: 0.002132  min_lr: 0.002132  loss: 3.0563 (2.9230)  weight_decay: 0.0500 (0.0500)  time: 0.7153  data: 0.0003  max mem: 57114
Epoch: [203]  [ 60/156]  eta: 0:01:13  lr: 0.002130  min_lr: 0.002130  loss: 3.0869 (2.9365)  weight_decay: 0.0500 (0.0500)  time: 0.7288  data: 0.0003  max mem: 57114
Epoch: [203]  [ 70/156]  eta: 0:01:04  lr: 0.002127  min_lr: 0.002127  loss: 2.8190 (2.8928)  weight_decay: 0.0500 (0.0500)  time: 0.7189  data: 0.0004  max mem: 57114
Epoch: [203]  [ 80/156]  eta: 0:00:57  lr: 0.002125  min_lr: 0.002125  loss: 2.8091 (2.8912)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0003  max mem: 57114
Epoch: [203]  [ 90/156]  eta: 0:00:49  lr: 0.002122  min_lr: 0.002122  loss: 2.8991 (2.8988)  weight_decay: 0.0500 (0.0500)  time: 0.7178  data: 0.0004  max mem: 57114
Epoch: [203]  [100/156]  eta: 0:00:41  lr: 0.002119  min_lr: 0.002119  loss: 3.0220 (2.9134)  weight_decay: 0.0500 (0.0500)  time: 0.7043  data: 0.0003  max mem: 57114
Epoch: [203]  [110/156]  eta: 0:00:34  lr: 0.002117  min_lr: 0.002117  loss: 3.0428 (2.9069)  weight_decay: 0.0500 (0.0500)  time: 0.7086  data: 0.0003  max mem: 57114
Epoch: [203]  [120/156]  eta: 0:00:26  lr: 0.002114  min_lr: 0.002114  loss: 2.9522 (2.9094)  weight_decay: 0.0500 (0.0500)  time: 0.7095  data: 0.0003  max mem: 57114
Epoch: [203]  [130/156]  eta: 0:00:19  lr: 0.002112  min_lr: 0.002112  loss: 2.9522 (2.9258)  weight_decay: 0.0500 (0.0500)  time: 0.6989  data: 0.0008  max mem: 57114
Epoch: [203]  [140/156]  eta: 0:00:11  lr: 0.002109  min_lr: 0.002109  loss: 3.0045 (2.9147)  weight_decay: 0.0500 (0.0500)  time: 0.6887  data: 0.0007  max mem: 57114
Epoch: [203]  [150/156]  eta: 0:00:04  lr: 0.002107  min_lr: 0.002107  loss: 3.0045 (2.9123)  weight_decay: 0.0500 (0.0500)  time: 0.6806  data: 0.0001  max mem: 57114
Epoch: [203]  [155/156]  eta: 0:00:00  lr: 0.002106  min_lr: 0.002106  loss: 2.8478 (2.9105)  weight_decay: 0.0500 (0.0500)  time: 0.6816  data: 0.0001  max mem: 57114
Epoch: [203] Total time: 0:01:53 (0.7277 s / it)
Averaged stats: lr: 0.002106  min_lr: 0.002106  loss: 2.8478 (2.8670)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6582 (0.6582)  acc1: 89.0625 (89.0625)  acc5: 100.0000 (100.0000)  time: 2.1707  data: 1.9150  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8506 (0.8268)  acc1: 84.3750 (82.6087)  acc5: 97.3958 (96.6752)  time: 0.5848  data: 0.3831  max mem: 57114
Test: Total time: 0:00:03 (0.6094 s / it)
* Acc@1 82.045 Acc@5 95.968 loss 0.916
Accuracy of the model on the 50000 test images: 82.0%
Max accuracy: 82.05%
Test:  [0/5]  eta: 0:00:10  loss: 0.6028 (0.6028)  acc1: 84.3750 (84.3750)  acc5: 96.8750 (96.8750)  time: 2.0350  data: 1.7915  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7275 (0.6982)  acc1: 79.6875 (78.2609)  acc5: 95.3125 (93.7340)  time: 0.5576  data: 0.3584  max mem: 57114
Test: Total time: 0:00:02 (0.5673 s / it)
* Acc@1 76.784 Acc@5 93.630 loss 0.847
Accuracy of the model EMA on 50000 test images: 76.8%
Max EMA accuracy: 76.78%
Epoch: [204]  [  0/156]  eta: 0:08:34  lr: 0.002105  min_lr: 0.002105  loss: 3.3828 (3.3828)  weight_decay: 0.0500 (0.0500)  time: 3.2959  data: 2.6469  max mem: 57114
Epoch: [204]  [ 10/156]  eta: 0:02:21  lr: 0.002103  min_lr: 0.002103  loss: 3.1751 (2.9801)  weight_decay: 0.0500 (0.0500)  time: 0.9710  data: 0.2409  max mem: 57114
Epoch: [204]  [ 20/156]  eta: 0:01:57  lr: 0.002100  min_lr: 0.002100  loss: 2.9862 (2.9111)  weight_decay: 0.0500 (0.0500)  time: 0.7418  data: 0.0003  max mem: 57114
Epoch: [204]  [ 30/156]  eta: 0:01:42  lr: 0.002098  min_lr: 0.002098  loss: 2.7160 (2.8332)  weight_decay: 0.0500 (0.0500)  time: 0.7301  data: 0.0003  max mem: 57114
Epoch: [204]  [ 40/156]  eta: 0:01:32  lr: 0.002095  min_lr: 0.002095  loss: 2.9142 (2.8493)  weight_decay: 0.0500 (0.0500)  time: 0.7308  data: 0.0003  max mem: 57114
Epoch: [204]  [ 50/156]  eta: 0:01:22  lr: 0.002093  min_lr: 0.002093  loss: 3.0503 (2.8599)  weight_decay: 0.0500 (0.0500)  time: 0.7226  data: 0.0004  max mem: 57114
Epoch: [204]  [ 60/156]  eta: 0:01:14  lr: 0.002090  min_lr: 0.002090  loss: 2.8788 (2.8463)  weight_decay: 0.0500 (0.0500)  time: 0.7167  data: 0.0003  max mem: 57114
Epoch: [204]  [ 70/156]  eta: 0:01:05  lr: 0.002088  min_lr: 0.002088  loss: 2.8124 (2.8430)  weight_decay: 0.0500 (0.0500)  time: 0.7252  data: 0.0004  max mem: 57114
Epoch: [204]  [ 80/156]  eta: 0:00:57  lr: 0.002085  min_lr: 0.002085  loss: 2.7394 (2.8466)  weight_decay: 0.0500 (0.0500)  time: 0.7218  data: 0.0004  max mem: 57114
Epoch: [204]  [ 90/156]  eta: 0:00:49  lr: 0.002083  min_lr: 0.002083  loss: 2.9333 (2.8600)  weight_decay: 0.0500 (0.0500)  time: 0.7124  data: 0.0004  max mem: 57114
Epoch: [204]  [100/156]  eta: 0:00:41  lr: 0.002080  min_lr: 0.002080  loss: 2.9810 (2.8702)  weight_decay: 0.0500 (0.0500)  time: 0.6977  data: 0.0004  max mem: 57114
Epoch: [204]  [110/156]  eta: 0:00:34  lr: 0.002077  min_lr: 0.002077  loss: 2.9876 (2.8785)  weight_decay: 0.0500 (0.0500)  time: 0.7021  data: 0.0004  max mem: 57114
Epoch: [204]  [120/156]  eta: 0:00:26  lr: 0.002075  min_lr: 0.002075  loss: 3.0552 (2.8857)  weight_decay: 0.0500 (0.0500)  time: 0.6995  data: 0.0004  max mem: 57114
Epoch: [204]  [130/156]  eta: 0:00:19  lr: 0.002072  min_lr: 0.002072  loss: 3.1503 (2.8891)  weight_decay: 0.0500 (0.0500)  time: 0.6930  data: 0.0009  max mem: 57114
Epoch: [204]  [140/156]  eta: 0:00:11  lr: 0.002070  min_lr: 0.002070  loss: 2.9157 (2.8769)  weight_decay: 0.0500 (0.0500)  time: 0.6867  data: 0.0008  max mem: 57114
Epoch: [204]  [150/156]  eta: 0:00:04  lr: 0.002067  min_lr: 0.002067  loss: 2.9157 (2.8713)  weight_decay: 0.0500 (0.0500)  time: 0.6868  data: 0.0001  max mem: 57114
Epoch: [204]  [155/156]  eta: 0:00:00  lr: 0.002066  min_lr: 0.002066  loss: 2.9157 (2.8613)  weight_decay: 0.0500 (0.0500)  time: 0.6860  data: 0.0001  max mem: 57114
Epoch: [204] Total time: 0:01:53 (0.7297 s / it)
Averaged stats: lr: 0.002066  min_lr: 0.002066  loss: 2.9157 (2.8555)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6539 (0.6539)  acc1: 88.5417 (88.5417)  acc5: 98.9583 (98.9583)  time: 2.0715  data: 1.8155  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8525 (0.7874)  acc1: 82.8125 (82.4808)  acc5: 97.3958 (96.8031)  time: 0.5650  data: 0.3632  max mem: 57114
Test: Total time: 0:00:02 (0.5885 s / it)
* Acc@1 82.415 Acc@5 96.124 loss 0.841
Accuracy of the model on the 50000 test images: 82.4%
Max accuracy: 82.41%
Test:  [0/5]  eta: 0:00:10  loss: 0.6008 (0.6008)  acc1: 84.3750 (84.3750)  acc5: 96.8750 (96.8750)  time: 2.0420  data: 1.7987  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7253 (0.6960)  acc1: 79.6875 (78.2609)  acc5: 95.3125 (93.7340)  time: 0.5589  data: 0.3598  max mem: 57114
Test: Total time: 0:00:02 (0.5717 s / it)
* Acc@1 76.848 Acc@5 93.674 loss 0.844
Accuracy of the model EMA on 50000 test images: 76.8%
Max EMA accuracy: 76.85%
Epoch: [205]  [  0/156]  eta: 0:09:01  lr: 0.002066  min_lr: 0.002066  loss: 3.0317 (3.0317)  weight_decay: 0.0500 (0.0500)  time: 3.4706  data: 2.8180  max mem: 57114
Epoch: [205]  [ 10/156]  eta: 0:02:23  lr: 0.002063  min_lr: 0.002063  loss: 2.7307 (2.7127)  weight_decay: 0.0500 (0.0500)  time: 0.9816  data: 0.2564  max mem: 57114
Epoch: [205]  [ 20/156]  eta: 0:01:57  lr: 0.002061  min_lr: 0.002061  loss: 2.6662 (2.7246)  weight_decay: 0.0500 (0.0500)  time: 0.7316  data: 0.0003  max mem: 57114
Epoch: [205]  [ 30/156]  eta: 0:01:42  lr: 0.002058  min_lr: 0.002058  loss: 2.8829 (2.7725)  weight_decay: 0.0500 (0.0500)  time: 0.7242  data: 0.0003  max mem: 57114
Epoch: [205]  [ 40/156]  eta: 0:01:32  lr: 0.002056  min_lr: 0.002056  loss: 2.8829 (2.7704)  weight_decay: 0.0500 (0.0500)  time: 0.7319  data: 0.0004  max mem: 57114
Epoch: [205]  [ 50/156]  eta: 0:01:23  lr: 0.002053  min_lr: 0.002053  loss: 2.8606 (2.7623)  weight_decay: 0.0500 (0.0500)  time: 0.7408  data: 0.0004  max mem: 57114
Epoch: [205]  [ 60/156]  eta: 0:01:14  lr: 0.002051  min_lr: 0.002051  loss: 2.6102 (2.7556)  weight_decay: 0.0500 (0.0500)  time: 0.7146  data: 0.0003  max mem: 57114
Epoch: [205]  [ 70/156]  eta: 0:01:05  lr: 0.002048  min_lr: 0.002048  loss: 2.6113 (2.7602)  weight_decay: 0.0500 (0.0500)  time: 0.7056  data: 0.0003  max mem: 57114
Epoch: [205]  [ 80/156]  eta: 0:00:57  lr: 0.002046  min_lr: 0.002046  loss: 2.9097 (2.7888)  weight_decay: 0.0500 (0.0500)  time: 0.7188  data: 0.0004  max mem: 57114
Epoch: [205]  [ 90/156]  eta: 0:00:49  lr: 0.002043  min_lr: 0.002043  loss: 2.9515 (2.8096)  weight_decay: 0.0500 (0.0500)  time: 0.7100  data: 0.0004  max mem: 57114
Epoch: [205]  [100/156]  eta: 0:00:41  lr: 0.002041  min_lr: 0.002041  loss: 3.0389 (2.8234)  weight_decay: 0.0500 (0.0500)  time: 0.7060  data: 0.0004  max mem: 57114
Epoch: [205]  [110/156]  eta: 0:00:34  lr: 0.002038  min_lr: 0.002038  loss: 2.8946 (2.8172)  weight_decay: 0.0500 (0.0500)  time: 0.7121  data: 0.0004  max mem: 57114
Epoch: [205]  [120/156]  eta: 0:00:26  lr: 0.002036  min_lr: 0.002036  loss: 2.9026 (2.8285)  weight_decay: 0.0500 (0.0500)  time: 0.7020  data: 0.0003  max mem: 57114
Epoch: [205]  [130/156]  eta: 0:00:19  lr: 0.002033  min_lr: 0.002033  loss: 3.0006 (2.8440)  weight_decay: 0.0500 (0.0500)  time: 0.6902  data: 0.0008  max mem: 57114
Epoch: [205]  [140/156]  eta: 0:00:11  lr: 0.002031  min_lr: 0.002031  loss: 3.0377 (2.8579)  weight_decay: 0.0500 (0.0500)  time: 0.6854  data: 0.0008  max mem: 57114
Epoch: [205]  [150/156]  eta: 0:00:04  lr: 0.002028  min_lr: 0.002028  loss: 3.0946 (2.8685)  weight_decay: 0.0500 (0.0500)  time: 0.6838  data: 0.0001  max mem: 57114
Epoch: [205]  [155/156]  eta: 0:00:00  lr: 0.002027  min_lr: 0.002027  loss: 2.9352 (2.8619)  weight_decay: 0.0500 (0.0500)  time: 0.6865  data: 0.0001  max mem: 57114
Epoch: [205] Total time: 0:01:53 (0.7301 s / it)
Averaged stats: lr: 0.002027  min_lr: 0.002027  loss: 2.9352 (2.8583)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6436 (0.6436)  acc1: 89.5833 (89.5833)  acc5: 98.9583 (98.9583)  time: 2.1166  data: 1.8605  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9419 (0.8389)  acc1: 82.2917 (82.7366)  acc5: 96.8750 (96.8031)  time: 0.5741  data: 0.3722  max mem: 57114
Test: Total time: 0:00:02 (0.5967 s / it)
* Acc@1 82.049 Acc@5 96.110 loss 0.926
Accuracy of the model on the 50000 test images: 82.0%
Max accuracy: 82.41%
Test:  [0/5]  eta: 0:00:12  loss: 0.5991 (0.5991)  acc1: 84.3750 (84.3750)  acc5: 96.8750 (96.8750)  time: 2.4856  data: 2.2421  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7233 (0.6940)  acc1: 79.6875 (78.2609)  acc5: 95.8333 (93.8619)  time: 0.6485  data: 0.4487  max mem: 57114
Test: Total time: 0:00:03 (0.6622 s / it)
* Acc@1 76.914 Acc@5 93.694 loss 0.841
Accuracy of the model EMA on 50000 test images: 76.9%
Max EMA accuracy: 76.91%
Epoch: [206]  [  0/156]  eta: 0:08:26  lr: 0.002027  min_lr: 0.002027  loss: 3.0778 (3.0778)  weight_decay: 0.0500 (0.0500)  time: 3.2439  data: 2.5953  max mem: 57114
Epoch: [206]  [ 10/156]  eta: 0:02:18  lr: 0.002024  min_lr: 0.002024  loss: 3.0778 (2.9486)  weight_decay: 0.0500 (0.0500)  time: 0.9468  data: 0.2364  max mem: 57114
Epoch: [206]  [ 20/156]  eta: 0:01:55  lr: 0.002022  min_lr: 0.002022  loss: 3.0514 (2.9435)  weight_decay: 0.0500 (0.0500)  time: 0.7292  data: 0.0004  max mem: 57114
Epoch: [206]  [ 30/156]  eta: 0:01:41  lr: 0.002019  min_lr: 0.002019  loss: 3.0537 (2.9865)  weight_decay: 0.0500 (0.0500)  time: 0.7262  data: 0.0004  max mem: 57114
Epoch: [206]  [ 40/156]  eta: 0:01:30  lr: 0.002017  min_lr: 0.002017  loss: 3.0949 (2.9787)  weight_decay: 0.0500 (0.0500)  time: 0.7117  data: 0.0004  max mem: 57114
Epoch: [206]  [ 50/156]  eta: 0:01:21  lr: 0.002014  min_lr: 0.002014  loss: 3.0979 (2.9714)  weight_decay: 0.0500 (0.0500)  time: 0.7129  data: 0.0004  max mem: 57114
Epoch: [206]  [ 60/156]  eta: 0:01:13  lr: 0.002012  min_lr: 0.002012  loss: 3.0845 (2.9682)  weight_decay: 0.0500 (0.0500)  time: 0.7175  data: 0.0004  max mem: 57114
Epoch: [206]  [ 70/156]  eta: 0:01:04  lr: 0.002009  min_lr: 0.002009  loss: 2.9499 (2.9503)  weight_decay: 0.0500 (0.0500)  time: 0.7094  data: 0.0004  max mem: 57114
Epoch: [206]  [ 80/156]  eta: 0:00:56  lr: 0.002007  min_lr: 0.002007  loss: 2.8573 (2.9522)  weight_decay: 0.0500 (0.0500)  time: 0.7053  data: 0.0004  max mem: 57114
Epoch: [206]  [ 90/156]  eta: 0:00:48  lr: 0.002004  min_lr: 0.002004  loss: 2.9706 (2.9434)  weight_decay: 0.0500 (0.0500)  time: 0.7078  data: 0.0004  max mem: 57114
Epoch: [206]  [100/156]  eta: 0:00:41  lr: 0.002002  min_lr: 0.002002  loss: 2.8239 (2.9117)  weight_decay: 0.0500 (0.0500)  time: 0.7083  data: 0.0003  max mem: 57114
Epoch: [206]  [110/156]  eta: 0:00:33  lr: 0.001999  min_lr: 0.001999  loss: 2.8878 (2.9139)  weight_decay: 0.0500 (0.0500)  time: 0.7026  data: 0.0003  max mem: 57114
Epoch: [206]  [120/156]  eta: 0:00:26  lr: 0.001997  min_lr: 0.001997  loss: 2.9648 (2.9099)  weight_decay: 0.0500 (0.0500)  time: 0.7059  data: 0.0003  max mem: 57114
Epoch: [206]  [130/156]  eta: 0:00:19  lr: 0.001994  min_lr: 0.001994  loss: 3.0898 (2.9113)  weight_decay: 0.0500 (0.0500)  time: 0.7070  data: 0.0008  max mem: 57114
Epoch: [206]  [140/156]  eta: 0:00:11  lr: 0.001992  min_lr: 0.001992  loss: 3.0875 (2.9075)  weight_decay: 0.0500 (0.0500)  time: 0.6919  data: 0.0007  max mem: 57114
Epoch: [206]  [150/156]  eta: 0:00:04  lr: 0.001989  min_lr: 0.001989  loss: 2.9916 (2.8966)  weight_decay: 0.0500 (0.0500)  time: 0.6889  data: 0.0001  max mem: 57114
Epoch: [206]  [155/156]  eta: 0:00:00  lr: 0.001988  min_lr: 0.001988  loss: 2.7660 (2.8930)  weight_decay: 0.0500 (0.0500)  time: 0.6850  data: 0.0001  max mem: 57114
Epoch: [206] Total time: 0:01:53 (0.7263 s / it)
Averaged stats: lr: 0.001988  min_lr: 0.001988  loss: 2.7660 (2.8540)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6593 (0.6593)  acc1: 86.9792 (86.9792)  acc5: 99.4792 (99.4792)  time: 2.0840  data: 1.8280  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9029 (0.7996)  acc1: 83.8542 (82.7366)  acc5: 97.9167 (97.3146)  time: 0.5676  data: 0.3657  max mem: 57114
Test: Total time: 0:00:02 (0.5901 s / it)
* Acc@1 82.447 Acc@5 96.350 loss 0.893
Accuracy of the model on the 50000 test images: 82.4%
Max accuracy: 82.45%
Test:  [0/5]  eta: 0:00:10  loss: 0.5978 (0.5978)  acc1: 84.3750 (84.3750)  acc5: 96.8750 (96.8750)  time: 2.1061  data: 1.8626  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7210 (0.6919)  acc1: 79.6875 (78.3887)  acc5: 95.8333 (93.8619)  time: 0.5718  data: 0.3726  max mem: 57114
Test: Total time: 0:00:02 (0.5814 s / it)
* Acc@1 76.956 Acc@5 93.708 loss 0.838
Accuracy of the model EMA on 50000 test images: 77.0%
Max EMA accuracy: 76.96%
Epoch: [207]  [  0/156]  eta: 0:08:25  lr: 0.001988  min_lr: 0.001988  loss: 3.2162 (3.2162)  weight_decay: 0.0500 (0.0500)  time: 3.2414  data: 2.5870  max mem: 57114
Epoch: [207]  [ 10/156]  eta: 0:02:17  lr: 0.001985  min_lr: 0.001985  loss: 3.0816 (3.0095)  weight_decay: 0.0500 (0.0500)  time: 0.9401  data: 0.2354  max mem: 57114
Epoch: [207]  [ 20/156]  eta: 0:01:54  lr: 0.001983  min_lr: 0.001983  loss: 3.0271 (2.9609)  weight_decay: 0.0500 (0.0500)  time: 0.7248  data: 0.0003  max mem: 57114
Epoch: [207]  [ 30/156]  eta: 0:01:42  lr: 0.001980  min_lr: 0.001980  loss: 3.0578 (3.0083)  weight_decay: 0.0500 (0.0500)  time: 0.7425  data: 0.0004  max mem: 57114
Epoch: [207]  [ 40/156]  eta: 0:01:32  lr: 0.001978  min_lr: 0.001978  loss: 3.1191 (2.9845)  weight_decay: 0.0500 (0.0500)  time: 0.7453  data: 0.0004  max mem: 57114
Epoch: [207]  [ 50/156]  eta: 0:01:22  lr: 0.001975  min_lr: 0.001975  loss: 3.1153 (2.9966)  weight_decay: 0.0500 (0.0500)  time: 0.7361  data: 0.0004  max mem: 57114
Epoch: [207]  [ 60/156]  eta: 0:01:14  lr: 0.001973  min_lr: 0.001973  loss: 3.0331 (2.9686)  weight_decay: 0.0500 (0.0500)  time: 0.7281  data: 0.0004  max mem: 57114
Epoch: [207]  [ 70/156]  eta: 0:01:06  lr: 0.001970  min_lr: 0.001970  loss: 2.9175 (2.9576)  weight_decay: 0.0500 (0.0500)  time: 0.7293  data: 0.0004  max mem: 57114
Epoch: [207]  [ 80/156]  eta: 0:00:57  lr: 0.001968  min_lr: 0.001968  loss: 2.7604 (2.9210)  weight_decay: 0.0500 (0.0500)  time: 0.7123  data: 0.0004  max mem: 57114
Epoch: [207]  [ 90/156]  eta: 0:00:49  lr: 0.001965  min_lr: 0.001965  loss: 2.8384 (2.9170)  weight_decay: 0.0500 (0.0500)  time: 0.6956  data: 0.0004  max mem: 57114
Epoch: [207]  [100/156]  eta: 0:00:41  lr: 0.001963  min_lr: 0.001963  loss: 2.9829 (2.9128)  weight_decay: 0.0500 (0.0500)  time: 0.7069  data: 0.0004  max mem: 57114
Epoch: [207]  [110/156]  eta: 0:00:34  lr: 0.001961  min_lr: 0.001961  loss: 2.9577 (2.8994)  weight_decay: 0.0500 (0.0500)  time: 0.7149  data: 0.0004  max mem: 57114
Epoch: [207]  [120/156]  eta: 0:00:26  lr: 0.001958  min_lr: 0.001958  loss: 2.9442 (2.9067)  weight_decay: 0.0500 (0.0500)  time: 0.7081  data: 0.0004  max mem: 57114
Epoch: [207]  [130/156]  eta: 0:00:19  lr: 0.001956  min_lr: 0.001956  loss: 2.9442 (2.9045)  weight_decay: 0.0500 (0.0500)  time: 0.7021  data: 0.0009  max mem: 57114
Epoch: [207]  [140/156]  eta: 0:00:11  lr: 0.001953  min_lr: 0.001953  loss: 2.9117 (2.8904)  weight_decay: 0.0500 (0.0500)  time: 0.6895  data: 0.0007  max mem: 57114
Epoch: [207]  [150/156]  eta: 0:00:04  lr: 0.001951  min_lr: 0.001951  loss: 3.0935 (2.9118)  weight_decay: 0.0500 (0.0500)  time: 0.6833  data: 0.0001  max mem: 57114
Epoch: [207]  [155/156]  eta: 0:00:00  lr: 0.001949  min_lr: 0.001949  loss: 3.0935 (2.9017)  weight_decay: 0.0500 (0.0500)  time: 0.6866  data: 0.0001  max mem: 57114
Epoch: [207] Total time: 0:01:54 (0.7320 s / it)
Averaged stats: lr: 0.001949  min_lr: 0.001949  loss: 3.0935 (2.8436)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 0.6591 (0.6591)  acc1: 90.1042 (90.1042)  acc5: 98.9583 (98.9583)  time: 2.2670  data: 2.0112  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8585 (0.8118)  acc1: 81.2500 (81.7136)  acc5: 96.8750 (96.5473)  time: 0.6040  data: 0.4023  max mem: 57114
Test: Total time: 0:00:03 (0.6272 s / it)
* Acc@1 82.171 Acc@5 96.164 loss 0.896
Accuracy of the model on the 50000 test images: 82.2%
Max accuracy: 82.45%
Test:  [0/5]  eta: 0:00:12  loss: 0.5962 (0.5962)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.5014  data: 2.2576  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7190 (0.6899)  acc1: 79.6875 (78.3887)  acc5: 95.8333 (93.9898)  time: 0.6511  data: 0.4517  max mem: 57114
Test: Total time: 0:00:03 (0.6657 s / it)
* Acc@1 76.988 Acc@5 93.742 loss 0.836
Accuracy of the model EMA on 50000 test images: 77.0%
Max EMA accuracy: 76.99%
Epoch: [208]  [  0/156]  eta: 0:06:30  lr: 0.001949  min_lr: 0.001949  loss: 2.3653 (2.3653)  weight_decay: 0.0500 (0.0500)  time: 2.5022  data: 1.8492  max mem: 57114
Epoch: [208]  [ 10/156]  eta: 0:02:09  lr: 0.001947  min_lr: 0.001947  loss: 2.6498 (2.7259)  weight_decay: 0.0500 (0.0500)  time: 0.8863  data: 0.1684  max mem: 57114
Epoch: [208]  [ 20/156]  eta: 0:01:50  lr: 0.001944  min_lr: 0.001944  loss: 2.8921 (2.7735)  weight_decay: 0.0500 (0.0500)  time: 0.7245  data: 0.0003  max mem: 57114
Epoch: [208]  [ 30/156]  eta: 0:01:38  lr: 0.001942  min_lr: 0.001942  loss: 2.9192 (2.8188)  weight_decay: 0.0500 (0.0500)  time: 0.7241  data: 0.0003  max mem: 57114
Epoch: [208]  [ 40/156]  eta: 0:01:29  lr: 0.001939  min_lr: 0.001939  loss: 2.9192 (2.7795)  weight_decay: 0.0500 (0.0500)  time: 0.7247  data: 0.0004  max mem: 57114
Epoch: [208]  [ 50/156]  eta: 0:01:20  lr: 0.001937  min_lr: 0.001937  loss: 2.5062 (2.7563)  weight_decay: 0.0500 (0.0500)  time: 0.7299  data: 0.0004  max mem: 57114
Epoch: [208]  [ 60/156]  eta: 0:01:12  lr: 0.001934  min_lr: 0.001934  loss: 2.7591 (2.7825)  weight_decay: 0.0500 (0.0500)  time: 0.7279  data: 0.0004  max mem: 57114
Epoch: [208]  [ 70/156]  eta: 0:01:04  lr: 0.001932  min_lr: 0.001932  loss: 2.7995 (2.7872)  weight_decay: 0.0500 (0.0500)  time: 0.7124  data: 0.0004  max mem: 57114
Epoch: [208]  [ 80/156]  eta: 0:00:56  lr: 0.001929  min_lr: 0.001929  loss: 2.7038 (2.7863)  weight_decay: 0.0500 (0.0500)  time: 0.7076  data: 0.0004  max mem: 57114
Epoch: [208]  [ 90/156]  eta: 0:00:48  lr: 0.001927  min_lr: 0.001927  loss: 2.7950 (2.7907)  weight_decay: 0.0500 (0.0500)  time: 0.7110  data: 0.0004  max mem: 57114
Epoch: [208]  [100/156]  eta: 0:00:41  lr: 0.001925  min_lr: 0.001925  loss: 2.8195 (2.7903)  weight_decay: 0.0500 (0.0500)  time: 0.7033  data: 0.0004  max mem: 57114
Epoch: [208]  [110/156]  eta: 0:00:33  lr: 0.001922  min_lr: 0.001922  loss: 2.8155 (2.7735)  weight_decay: 0.0500 (0.0500)  time: 0.6991  data: 0.0004  max mem: 57114
Epoch: [208]  [120/156]  eta: 0:00:26  lr: 0.001920  min_lr: 0.001920  loss: 2.9433 (2.7944)  weight_decay: 0.0500 (0.0500)  time: 0.7022  data: 0.0004  max mem: 57114
Epoch: [208]  [130/156]  eta: 0:00:18  lr: 0.001917  min_lr: 0.001917  loss: 2.9844 (2.8112)  weight_decay: 0.0500 (0.0500)  time: 0.6950  data: 0.0009  max mem: 57114
Epoch: [208]  [140/156]  eta: 0:00:11  lr: 0.001915  min_lr: 0.001915  loss: 2.9073 (2.8151)  weight_decay: 0.0500 (0.0500)  time: 0.6819  data: 0.0007  max mem: 57114
Epoch: [208]  [150/156]  eta: 0:00:04  lr: 0.001912  min_lr: 0.001912  loss: 2.7715 (2.8060)  weight_decay: 0.0500 (0.0500)  time: 0.6817  data: 0.0001  max mem: 57114
Epoch: [208]  [155/156]  eta: 0:00:00  lr: 0.001911  min_lr: 0.001911  loss: 2.5928 (2.8007)  weight_decay: 0.0500 (0.0500)  time: 0.6808  data: 0.0001  max mem: 57114
Epoch: [208] Total time: 0:01:52 (0.7214 s / it)
Averaged stats: lr: 0.001911  min_lr: 0.001911  loss: 2.5928 (2.8373)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6579 (0.6579)  acc1: 86.9792 (86.9792)  acc5: 98.9583 (98.9583)  time: 2.1693  data: 1.9134  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8435 (0.7695)  acc1: 82.8125 (82.8645)  acc5: 96.8750 (96.9309)  time: 0.5845  data: 0.3827  max mem: 57114
Test: Total time: 0:00:03 (0.6103 s / it)
* Acc@1 82.151 Acc@5 96.112 loss 0.859
Accuracy of the model on the 50000 test images: 82.2%
Max accuracy: 82.45%
Test:  [0/5]  eta: 0:00:11  loss: 0.5953 (0.5953)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.3392  data: 2.0958  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7171 (0.6882)  acc1: 80.2083 (78.5166)  acc5: 96.3542 (94.1177)  time: 0.6185  data: 0.4192  max mem: 57114
Test: Total time: 0:00:03 (0.6324 s / it)
* Acc@1 77.040 Acc@5 93.756 loss 0.833
Accuracy of the model EMA on 50000 test images: 77.0%
Max EMA accuracy: 77.04%
Epoch: [209]  [  0/156]  eta: 0:05:55  lr: 0.001911  min_lr: 0.001911  loss: 3.2205 (3.2205)  weight_decay: 0.0500 (0.0500)  time: 2.2790  data: 1.6312  max mem: 57114
Epoch: [209]  [ 10/156]  eta: 0:02:06  lr: 0.001908  min_lr: 0.001908  loss: 2.9805 (2.7906)  weight_decay: 0.0500 (0.0500)  time: 0.8667  data: 0.1485  max mem: 57114
Epoch: [209]  [ 20/156]  eta: 0:01:49  lr: 0.001906  min_lr: 0.001906  loss: 2.6608 (2.7588)  weight_decay: 0.0500 (0.0500)  time: 0.7304  data: 0.0003  max mem: 57114
Epoch: [209]  [ 30/156]  eta: 0:01:38  lr: 0.001903  min_lr: 0.001903  loss: 2.8798 (2.7739)  weight_decay: 0.0500 (0.0500)  time: 0.7300  data: 0.0004  max mem: 57114
Epoch: [209]  [ 40/156]  eta: 0:01:28  lr: 0.001901  min_lr: 0.001901  loss: 3.1848 (2.8489)  weight_decay: 0.0500 (0.0500)  time: 0.7244  data: 0.0004  max mem: 57114
Epoch: [209]  [ 50/156]  eta: 0:01:20  lr: 0.001899  min_lr: 0.001899  loss: 2.9932 (2.8273)  weight_decay: 0.0500 (0.0500)  time: 0.7211  data: 0.0004  max mem: 57114
Epoch: [209]  [ 60/156]  eta: 0:01:12  lr: 0.001896  min_lr: 0.001896  loss: 2.8371 (2.8232)  weight_decay: 0.0500 (0.0500)  time: 0.7236  data: 0.0004  max mem: 57114
Epoch: [209]  [ 70/156]  eta: 0:01:04  lr: 0.001894  min_lr: 0.001894  loss: 2.8855 (2.8408)  weight_decay: 0.0500 (0.0500)  time: 0.7234  data: 0.0004  max mem: 57114
Epoch: [209]  [ 80/156]  eta: 0:00:56  lr: 0.001891  min_lr: 0.001891  loss: 3.0248 (2.8516)  weight_decay: 0.0500 (0.0500)  time: 0.7129  data: 0.0004  max mem: 57114
Epoch: [209]  [ 90/156]  eta: 0:00:48  lr: 0.001889  min_lr: 0.001889  loss: 2.8449 (2.8273)  weight_decay: 0.0500 (0.0500)  time: 0.7110  data: 0.0004  max mem: 57114
Epoch: [209]  [100/156]  eta: 0:00:41  lr: 0.001886  min_lr: 0.001886  loss: 2.6148 (2.8060)  weight_decay: 0.0500 (0.0500)  time: 0.7042  data: 0.0004  max mem: 57114
Epoch: [209]  [110/156]  eta: 0:00:33  lr: 0.001884  min_lr: 0.001884  loss: 2.8853 (2.8265)  weight_decay: 0.0500 (0.0500)  time: 0.7009  data: 0.0004  max mem: 57114
Epoch: [209]  [120/156]  eta: 0:00:26  lr: 0.001881  min_lr: 0.001881  loss: 3.0647 (2.8343)  weight_decay: 0.0500 (0.0500)  time: 0.6978  data: 0.0004  max mem: 57114
Epoch: [209]  [130/156]  eta: 0:00:18  lr: 0.001879  min_lr: 0.001879  loss: 3.0032 (2.8595)  weight_decay: 0.0500 (0.0500)  time: 0.6963  data: 0.0010  max mem: 57114
Epoch: [209]  [140/156]  eta: 0:00:11  lr: 0.001877  min_lr: 0.001877  loss: 2.9019 (2.8498)  weight_decay: 0.0500 (0.0500)  time: 0.6931  data: 0.0008  max mem: 57114
Epoch: [209]  [150/156]  eta: 0:00:04  lr: 0.001874  min_lr: 0.001874  loss: 2.7419 (2.8375)  weight_decay: 0.0500 (0.0500)  time: 0.6877  data: 0.0001  max mem: 57114
Epoch: [209]  [155/156]  eta: 0:00:00  lr: 0.001873  min_lr: 0.001873  loss: 2.7713 (2.8473)  weight_decay: 0.0500 (0.0500)  time: 0.6893  data: 0.0001  max mem: 57114
Epoch: [209] Total time: 0:01:52 (0.7225 s / it)
Averaged stats: lr: 0.001873  min_lr: 0.001873  loss: 2.7713 (2.8363)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7393 (0.7393)  acc1: 89.5833 (89.5833)  acc5: 98.9583 (98.9583)  time: 2.1062  data: 1.8476  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9426 (0.8733)  acc1: 81.7708 (82.0972)  acc5: 97.9167 (97.1867)  time: 0.5720  data: 0.3696  max mem: 57114
Test: Total time: 0:00:02 (0.5960 s / it)
* Acc@1 82.363 Acc@5 96.250 loss 0.954
Accuracy of the model on the 50000 test images: 82.4%
Max accuracy: 82.45%
Test:  [0/5]  eta: 0:00:11  loss: 0.5939 (0.5939)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.2756  data: 2.0321  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7154 (0.6866)  acc1: 80.7292 (78.6445)  acc5: 96.3542 (94.1177)  time: 0.6058  data: 0.4065  max mem: 57114
Test: Total time: 0:00:03 (0.6192 s / it)
* Acc@1 77.108 Acc@5 93.770 loss 0.831
Accuracy of the model EMA on 50000 test images: 77.1%
Max EMA accuracy: 77.11%
Epoch: [210]  [  0/156]  eta: 0:06:49  lr: 0.001873  min_lr: 0.001873  loss: 1.8876 (1.8876)  weight_decay: 0.0500 (0.0500)  time: 2.6278  data: 1.9814  max mem: 57114
Epoch: [210]  [ 10/156]  eta: 0:02:07  lr: 0.001870  min_lr: 0.001870  loss: 2.9087 (2.8264)  weight_decay: 0.0500 (0.0500)  time: 0.8740  data: 0.1804  max mem: 57114
Epoch: [210]  [ 20/156]  eta: 0:01:48  lr: 0.001868  min_lr: 0.001868  loss: 2.9087 (2.8784)  weight_decay: 0.0500 (0.0500)  time: 0.7095  data: 0.0004  max mem: 57114
Epoch: [210]  [ 30/156]  eta: 0:01:36  lr: 0.001865  min_lr: 0.001865  loss: 2.7975 (2.8876)  weight_decay: 0.0500 (0.0500)  time: 0.7113  data: 0.0004  max mem: 57114
Epoch: [210]  [ 40/156]  eta: 0:01:28  lr: 0.001863  min_lr: 0.001863  loss: 2.9561 (2.8899)  weight_decay: 0.0500 (0.0500)  time: 0.7215  data: 0.0004  max mem: 57114
Epoch: [210]  [ 50/156]  eta: 0:01:19  lr: 0.001860  min_lr: 0.001860  loss: 3.0177 (2.8810)  weight_decay: 0.0500 (0.0500)  time: 0.7230  data: 0.0004  max mem: 57114
Epoch: [210]  [ 60/156]  eta: 0:01:11  lr: 0.001858  min_lr: 0.001858  loss: 3.0304 (2.8936)  weight_decay: 0.0500 (0.0500)  time: 0.7191  data: 0.0004  max mem: 57114
Epoch: [210]  [ 70/156]  eta: 0:01:04  lr: 0.001856  min_lr: 0.001856  loss: 3.0808 (2.9262)  weight_decay: 0.0500 (0.0500)  time: 0.7275  data: 0.0004  max mem: 57114
Epoch: [210]  [ 80/156]  eta: 0:00:56  lr: 0.001853  min_lr: 0.001853  loss: 3.0093 (2.8901)  weight_decay: 0.0500 (0.0500)  time: 0.7163  data: 0.0004  max mem: 57114
Epoch: [210]  [ 90/156]  eta: 0:00:48  lr: 0.001851  min_lr: 0.001851  loss: 2.9949 (2.9149)  weight_decay: 0.0500 (0.0500)  time: 0.6992  data: 0.0004  max mem: 57114
Epoch: [210]  [100/156]  eta: 0:00:40  lr: 0.001848  min_lr: 0.001848  loss: 3.1043 (2.9181)  weight_decay: 0.0500 (0.0500)  time: 0.6958  data: 0.0004  max mem: 57114
Epoch: [210]  [110/156]  eta: 0:00:33  lr: 0.001846  min_lr: 0.001846  loss: 3.0506 (2.9087)  weight_decay: 0.0500 (0.0500)  time: 0.7088  data: 0.0004  max mem: 57114
Epoch: [210]  [120/156]  eta: 0:00:26  lr: 0.001843  min_lr: 0.001843  loss: 2.8883 (2.8875)  weight_decay: 0.0500 (0.0500)  time: 0.7120  data: 0.0004  max mem: 57114
Epoch: [210]  [130/156]  eta: 0:00:18  lr: 0.001841  min_lr: 0.001841  loss: 2.8476 (2.8786)  weight_decay: 0.0500 (0.0500)  time: 0.6991  data: 0.0009  max mem: 57114
Epoch: [210]  [140/156]  eta: 0:00:11  lr: 0.001839  min_lr: 0.001839  loss: 3.0155 (2.8684)  weight_decay: 0.0500 (0.0500)  time: 0.6918  data: 0.0007  max mem: 57114
Epoch: [210]  [150/156]  eta: 0:00:04  lr: 0.001836  min_lr: 0.001836  loss: 2.9874 (2.8691)  weight_decay: 0.0500 (0.0500)  time: 0.6874  data: 0.0001  max mem: 57114
Epoch: [210]  [155/156]  eta: 0:00:00  lr: 0.001835  min_lr: 0.001835  loss: 2.9272 (2.8727)  weight_decay: 0.0500 (0.0500)  time: 0.6844  data: 0.0001  max mem: 57114
Epoch: [210] Total time: 0:01:52 (0.7212 s / it)
Averaged stats: lr: 0.001835  min_lr: 0.001835  loss: 2.9272 (2.8311)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7164 (0.7164)  acc1: 88.0208 (88.0208)  acc5: 98.9583 (98.9583)  time: 2.0817  data: 1.8263  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9071 (0.8340)  acc1: 82.8125 (83.7596)  acc5: 98.4375 (96.9309)  time: 0.5670  data: 0.3653  max mem: 57114
Test: Total time: 0:00:02 (0.5909 s / it)
* Acc@1 82.451 Acc@5 96.344 loss 0.933
Accuracy of the model on the 50000 test images: 82.5%
Max accuracy: 82.45%
Test:  [0/5]  eta: 0:00:10  loss: 0.5924 (0.5924)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.0267  data: 1.7832  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7134 (0.6848)  acc1: 80.7292 (78.7724)  acc5: 96.3542 (94.1177)  time: 0.5560  data: 0.3567  max mem: 57114
Test: Total time: 0:00:02 (0.5687 s / it)
* Acc@1 77.150 Acc@5 93.788 loss 0.829
Accuracy of the model EMA on 50000 test images: 77.1%
Max EMA accuracy: 77.15%
Epoch: [211]  [  0/156]  eta: 0:06:46  lr: 0.001835  min_lr: 0.001835  loss: 2.2324 (2.2324)  weight_decay: 0.0500 (0.0500)  time: 2.6040  data: 1.9459  max mem: 57114
Epoch: [211]  [ 10/156]  eta: 0:02:12  lr: 0.001832  min_lr: 0.001832  loss: 2.8865 (2.7494)  weight_decay: 0.0500 (0.0500)  time: 0.9070  data: 0.1771  max mem: 57114
Epoch: [211]  [ 20/156]  eta: 0:01:51  lr: 0.001830  min_lr: 0.001830  loss: 2.9698 (2.8458)  weight_decay: 0.0500 (0.0500)  time: 0.7317  data: 0.0003  max mem: 57114
Epoch: [211]  [ 30/156]  eta: 0:01:40  lr: 0.001828  min_lr: 0.001828  loss: 2.9698 (2.8725)  weight_decay: 0.0500 (0.0500)  time: 0.7318  data: 0.0004  max mem: 57114
Epoch: [211]  [ 40/156]  eta: 0:01:30  lr: 0.001825  min_lr: 0.001825  loss: 2.7753 (2.8481)  weight_decay: 0.0500 (0.0500)  time: 0.7320  data: 0.0004  max mem: 57114
Epoch: [211]  [ 50/156]  eta: 0:01:21  lr: 0.001823  min_lr: 0.001823  loss: 2.7357 (2.8277)  weight_decay: 0.0500 (0.0500)  time: 0.7367  data: 0.0004  max mem: 57114
Epoch: [211]  [ 60/156]  eta: 0:01:13  lr: 0.001820  min_lr: 0.001820  loss: 2.9037 (2.8149)  weight_decay: 0.0500 (0.0500)  time: 0.7411  data: 0.0004  max mem: 57114
Epoch: [211]  [ 70/156]  eta: 0:01:05  lr: 0.001818  min_lr: 0.001818  loss: 2.6284 (2.7782)  weight_decay: 0.0500 (0.0500)  time: 0.7327  data: 0.0004  max mem: 57114
Epoch: [211]  [ 80/156]  eta: 0:00:57  lr: 0.001815  min_lr: 0.001815  loss: 2.7352 (2.7930)  weight_decay: 0.0500 (0.0500)  time: 0.7195  data: 0.0004  max mem: 57114
Epoch: [211]  [ 90/156]  eta: 0:00:49  lr: 0.001813  min_lr: 0.001813  loss: 2.8236 (2.7991)  weight_decay: 0.0500 (0.0500)  time: 0.7054  data: 0.0004  max mem: 57114
Epoch: [211]  [100/156]  eta: 0:00:41  lr: 0.001811  min_lr: 0.001811  loss: 2.8873 (2.8024)  weight_decay: 0.0500 (0.0500)  time: 0.7038  data: 0.0004  max mem: 57114
Epoch: [211]  [110/156]  eta: 0:00:34  lr: 0.001808  min_lr: 0.001808  loss: 3.1638 (2.8285)  weight_decay: 0.0500 (0.0500)  time: 0.7083  data: 0.0004  max mem: 57114
Epoch: [211]  [120/156]  eta: 0:00:26  lr: 0.001806  min_lr: 0.001806  loss: 3.0965 (2.8288)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0004  max mem: 57114
Epoch: [211]  [130/156]  eta: 0:00:19  lr: 0.001803  min_lr: 0.001803  loss: 2.9411 (2.8417)  weight_decay: 0.0500 (0.0500)  time: 0.7021  data: 0.0009  max mem: 57114
Epoch: [211]  [140/156]  eta: 0:00:11  lr: 0.001801  min_lr: 0.001801  loss: 2.9726 (2.8514)  weight_decay: 0.0500 (0.0500)  time: 0.6949  data: 0.0007  max mem: 57114
Epoch: [211]  [150/156]  eta: 0:00:04  lr: 0.001799  min_lr: 0.001799  loss: 2.9880 (2.8600)  weight_decay: 0.0500 (0.0500)  time: 0.6873  data: 0.0001  max mem: 57114
Epoch: [211]  [155/156]  eta: 0:00:00  lr: 0.001797  min_lr: 0.001797  loss: 2.9726 (2.8473)  weight_decay: 0.0500 (0.0500)  time: 0.6833  data: 0.0001  max mem: 57114
Epoch: [211] Total time: 0:01:53 (0.7295 s / it)
Averaged stats: lr: 0.001797  min_lr: 0.001797  loss: 2.9726 (2.8316)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6444 (0.6444)  acc1: 87.5000 (87.5000)  acc5: 98.4375 (98.4375)  time: 2.0453  data: 1.7895  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8285 (0.7840)  acc1: 83.8542 (82.7366)  acc5: 98.4375 (96.6752)  time: 0.5597  data: 0.3580  max mem: 57114
Test: Total time: 0:00:02 (0.5842 s / it)
* Acc@1 82.235 Acc@5 96.120 loss 0.872
Accuracy of the model on the 50000 test images: 82.2%
Max accuracy: 82.45%
Test:  [0/5]  eta: 0:00:11  loss: 0.5910 (0.5910)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.2922  data: 2.0487  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7116 (0.6832)  acc1: 80.7292 (78.7724)  acc5: 96.3542 (94.1177)  time: 0.6091  data: 0.4098  max mem: 57114
Test: Total time: 0:00:03 (0.6206 s / it)
* Acc@1 77.202 Acc@5 93.826 loss 0.827
Accuracy of the model EMA on 50000 test images: 77.2%
Max EMA accuracy: 77.20%
Epoch: [212]  [  0/156]  eta: 0:06:25  lr: 0.001797  min_lr: 0.001797  loss: 2.5425 (2.5425)  weight_decay: 0.0500 (0.0500)  time: 2.4691  data: 1.8108  max mem: 57114
Epoch: [212]  [ 10/156]  eta: 0:02:06  lr: 0.001795  min_lr: 0.001795  loss: 2.9265 (2.8999)  weight_decay: 0.0500 (0.0500)  time: 0.8650  data: 0.1649  max mem: 57114
Epoch: [212]  [ 20/156]  eta: 0:01:48  lr: 0.001792  min_lr: 0.001792  loss: 2.9320 (2.8968)  weight_decay: 0.0500 (0.0500)  time: 0.7108  data: 0.0003  max mem: 57114
Epoch: [212]  [ 30/156]  eta: 0:01:36  lr: 0.001790  min_lr: 0.001790  loss: 2.8940 (2.8384)  weight_decay: 0.0500 (0.0500)  time: 0.7155  data: 0.0004  max mem: 57114
Epoch: [212]  [ 40/156]  eta: 0:01:27  lr: 0.001788  min_lr: 0.001788  loss: 2.9729 (2.8943)  weight_decay: 0.0500 (0.0500)  time: 0.7089  data: 0.0004  max mem: 57114
Epoch: [212]  [ 50/156]  eta: 0:01:19  lr: 0.001785  min_lr: 0.001785  loss: 3.0534 (2.8944)  weight_decay: 0.0500 (0.0500)  time: 0.7126  data: 0.0004  max mem: 57114
Epoch: [212]  [ 60/156]  eta: 0:01:11  lr: 0.001783  min_lr: 0.001783  loss: 2.9359 (2.8789)  weight_decay: 0.0500 (0.0500)  time: 0.7256  data: 0.0004  max mem: 57114
Epoch: [212]  [ 70/156]  eta: 0:01:03  lr: 0.001780  min_lr: 0.001780  loss: 2.7168 (2.8216)  weight_decay: 0.0500 (0.0500)  time: 0.7286  data: 0.0003  max mem: 57114
Epoch: [212]  [ 80/156]  eta: 0:00:56  lr: 0.001778  min_lr: 0.001778  loss: 2.6998 (2.8020)  weight_decay: 0.0500 (0.0500)  time: 0.7192  data: 0.0004  max mem: 57114
Epoch: [212]  [ 90/156]  eta: 0:00:48  lr: 0.001776  min_lr: 0.001776  loss: 2.8671 (2.8140)  weight_decay: 0.0500 (0.0500)  time: 0.7067  data: 0.0004  max mem: 57114
Epoch: [212]  [100/156]  eta: 0:00:40  lr: 0.001773  min_lr: 0.001773  loss: 2.9251 (2.8133)  weight_decay: 0.0500 (0.0500)  time: 0.7072  data: 0.0004  max mem: 57114
Epoch: [212]  [110/156]  eta: 0:00:33  lr: 0.001771  min_lr: 0.001771  loss: 2.9506 (2.8076)  weight_decay: 0.0500 (0.0500)  time: 0.7083  data: 0.0004  max mem: 57114
Epoch: [212]  [120/156]  eta: 0:00:26  lr: 0.001768  min_lr: 0.001768  loss: 3.0144 (2.8126)  weight_decay: 0.0500 (0.0500)  time: 0.6975  data: 0.0003  max mem: 57114
Epoch: [212]  [130/156]  eta: 0:00:18  lr: 0.001766  min_lr: 0.001766  loss: 2.9154 (2.8157)  weight_decay: 0.0500 (0.0500)  time: 0.6976  data: 0.0008  max mem: 57114
Epoch: [212]  [140/156]  eta: 0:00:11  lr: 0.001764  min_lr: 0.001764  loss: 2.7521 (2.8109)  weight_decay: 0.0500 (0.0500)  time: 0.6971  data: 0.0007  max mem: 57114
Epoch: [212]  [150/156]  eta: 0:00:04  lr: 0.001761  min_lr: 0.001761  loss: 2.7204 (2.7982)  weight_decay: 0.0500 (0.0500)  time: 0.6869  data: 0.0001  max mem: 57114
Epoch: [212]  [155/156]  eta: 0:00:00  lr: 0.001760  min_lr: 0.001760  loss: 2.6853 (2.8029)  weight_decay: 0.0500 (0.0500)  time: 0.6863  data: 0.0001  max mem: 57114
Epoch: [212] Total time: 0:01:52 (0.7208 s / it)
Averaged stats: lr: 0.001760  min_lr: 0.001760  loss: 2.6853 (2.8191)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5953 (0.5953)  acc1: 91.1458 (91.1458)  acc5: 98.9583 (98.9583)  time: 2.0655  data: 1.8098  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8914 (0.7964)  acc1: 81.2500 (82.8645)  acc5: 97.3958 (96.6752)  time: 0.5638  data: 0.3620  max mem: 57114
Test: Total time: 0:00:02 (0.5850 s / it)
* Acc@1 82.491 Acc@5 96.144 loss 0.867
Accuracy of the model on the 50000 test images: 82.5%
Max accuracy: 82.49%
Test:  [0/5]  eta: 0:00:10  loss: 0.5903 (0.5903)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.0923  data: 1.8488  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7099 (0.6819)  acc1: 80.7292 (78.9003)  acc5: 96.3542 (94.1177)  time: 0.5691  data: 0.3698  max mem: 57114
Test: Total time: 0:00:02 (0.5818 s / it)
* Acc@1 77.246 Acc@5 93.858 loss 0.825
Accuracy of the model EMA on 50000 test images: 77.2%
Max EMA accuracy: 77.25%
Epoch: [213]  [  0/156]  eta: 0:08:00  lr: 0.001760  min_lr: 0.001760  loss: 2.7628 (2.7628)  weight_decay: 0.0500 (0.0500)  time: 3.0796  data: 2.4280  max mem: 57114
Epoch: [213]  [ 10/156]  eta: 0:02:17  lr: 0.001757  min_lr: 0.001757  loss: 2.8211 (2.8692)  weight_decay: 0.0500 (0.0500)  time: 0.9402  data: 0.2210  max mem: 57114
Epoch: [213]  [ 20/156]  eta: 0:01:54  lr: 0.001755  min_lr: 0.001755  loss: 3.0785 (2.9435)  weight_decay: 0.0500 (0.0500)  time: 0.7320  data: 0.0004  max mem: 57114
Epoch: [213]  [ 30/156]  eta: 0:01:41  lr: 0.001753  min_lr: 0.001753  loss: 2.9407 (2.8709)  weight_decay: 0.0500 (0.0500)  time: 0.7351  data: 0.0004  max mem: 57114
Epoch: [213]  [ 40/156]  eta: 0:01:31  lr: 0.001750  min_lr: 0.001750  loss: 2.8129 (2.8800)  weight_decay: 0.0500 (0.0500)  time: 0.7239  data: 0.0003  max mem: 57114
Epoch: [213]  [ 50/156]  eta: 0:01:22  lr: 0.001748  min_lr: 0.001748  loss: 2.8129 (2.8463)  weight_decay: 0.0500 (0.0500)  time: 0.7226  data: 0.0004  max mem: 57114
Epoch: [213]  [ 60/156]  eta: 0:01:13  lr: 0.001746  min_lr: 0.001746  loss: 2.8774 (2.8615)  weight_decay: 0.0500 (0.0500)  time: 0.7245  data: 0.0004  max mem: 57114
Epoch: [213]  [ 70/156]  eta: 0:01:05  lr: 0.001743  min_lr: 0.001743  loss: 2.9847 (2.8758)  weight_decay: 0.0500 (0.0500)  time: 0.7127  data: 0.0004  max mem: 57114
Epoch: [213]  [ 80/156]  eta: 0:00:57  lr: 0.001741  min_lr: 0.001741  loss: 2.8885 (2.8756)  weight_decay: 0.0500 (0.0500)  time: 0.7147  data: 0.0004  max mem: 57114
Epoch: [213]  [ 90/156]  eta: 0:00:49  lr: 0.001738  min_lr: 0.001738  loss: 2.9550 (2.8852)  weight_decay: 0.0500 (0.0500)  time: 0.7184  data: 0.0004  max mem: 57114
Epoch: [213]  [100/156]  eta: 0:00:41  lr: 0.001736  min_lr: 0.001736  loss: 2.9926 (2.8886)  weight_decay: 0.0500 (0.0500)  time: 0.7088  data: 0.0004  max mem: 57114
Epoch: [213]  [110/156]  eta: 0:00:34  lr: 0.001734  min_lr: 0.001734  loss: 2.9075 (2.8819)  weight_decay: 0.0500 (0.0500)  time: 0.6985  data: 0.0003  max mem: 57114
Epoch: [213]  [120/156]  eta: 0:00:26  lr: 0.001731  min_lr: 0.001731  loss: 2.7935 (2.8783)  weight_decay: 0.0500 (0.0500)  time: 0.6997  data: 0.0004  max mem: 57114
Epoch: [213]  [130/156]  eta: 0:00:19  lr: 0.001729  min_lr: 0.001729  loss: 2.8508 (2.8825)  weight_decay: 0.0500 (0.0500)  time: 0.7009  data: 0.0009  max mem: 57114
Epoch: [213]  [140/156]  eta: 0:00:11  lr: 0.001727  min_lr: 0.001727  loss: 2.8508 (2.8666)  weight_decay: 0.0500 (0.0500)  time: 0.6892  data: 0.0007  max mem: 57114
Epoch: [213]  [150/156]  eta: 0:00:04  lr: 0.001724  min_lr: 0.001724  loss: 2.8140 (2.8604)  weight_decay: 0.0500 (0.0500)  time: 0.6835  data: 0.0001  max mem: 57114
Epoch: [213]  [155/156]  eta: 0:00:00  lr: 0.001723  min_lr: 0.001723  loss: 2.8597 (2.8644)  weight_decay: 0.0500 (0.0500)  time: 0.6809  data: 0.0001  max mem: 57114
Epoch: [213] Total time: 0:01:53 (0.7277 s / it)
Averaged stats: lr: 0.001723  min_lr: 0.001723  loss: 2.8597 (2.8138)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6350 (0.6350)  acc1: 88.5417 (88.5417)  acc5: 98.9583 (98.9583)  time: 2.0278  data: 1.7716  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8248 (0.7654)  acc1: 84.8958 (83.1202)  acc5: 97.3958 (97.3146)  time: 0.5563  data: 0.3544  max mem: 57114
Test: Total time: 0:00:02 (0.5779 s / it)
* Acc@1 82.587 Acc@5 96.266 loss 0.864
Accuracy of the model on the 50000 test images: 82.6%
Max accuracy: 82.59%
Test:  [0/5]  eta: 0:00:09  loss: 0.5895 (0.5895)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 1.9865  data: 1.7430  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7082 (0.6806)  acc1: 80.7292 (78.9003)  acc5: 96.3542 (94.1177)  time: 0.5479  data: 0.3487  max mem: 57114
Test: Total time: 0:00:02 (0.5583 s / it)
* Acc@1 77.280 Acc@5 93.858 loss 0.823
Accuracy of the model EMA on 50000 test images: 77.3%
Max EMA accuracy: 77.28%
Epoch: [214]  [  0/156]  eta: 0:07:03  lr: 0.001723  min_lr: 0.001723  loss: 3.1968 (3.1968)  weight_decay: 0.0500 (0.0500)  time: 2.7135  data: 2.0637  max mem: 57114
Epoch: [214]  [ 10/156]  eta: 0:02:11  lr: 0.001720  min_lr: 0.001720  loss: 2.8807 (2.7945)  weight_decay: 0.0500 (0.0500)  time: 0.8992  data: 0.1879  max mem: 57114
Epoch: [214]  [ 20/156]  eta: 0:01:51  lr: 0.001718  min_lr: 0.001718  loss: 2.8179 (2.7614)  weight_decay: 0.0500 (0.0500)  time: 0.7240  data: 0.0003  max mem: 57114
Epoch: [214]  [ 30/156]  eta: 0:01:38  lr: 0.001716  min_lr: 0.001716  loss: 2.5409 (2.6516)  weight_decay: 0.0500 (0.0500)  time: 0.7149  data: 0.0004  max mem: 57114
Epoch: [214]  [ 40/156]  eta: 0:01:29  lr: 0.001713  min_lr: 0.001713  loss: 2.7182 (2.6841)  weight_decay: 0.0500 (0.0500)  time: 0.7169  data: 0.0004  max mem: 57114
Epoch: [214]  [ 50/156]  eta: 0:01:21  lr: 0.001711  min_lr: 0.001711  loss: 2.7617 (2.6942)  weight_decay: 0.0500 (0.0500)  time: 0.7407  data: 0.0004  max mem: 57114
Epoch: [214]  [ 60/156]  eta: 0:01:12  lr: 0.001709  min_lr: 0.001709  loss: 2.7617 (2.7030)  weight_decay: 0.0500 (0.0500)  time: 0.7360  data: 0.0004  max mem: 57114
Epoch: [214]  [ 70/156]  eta: 0:01:04  lr: 0.001706  min_lr: 0.001706  loss: 2.8029 (2.7121)  weight_decay: 0.0500 (0.0500)  time: 0.7240  data: 0.0003  max mem: 57114
Epoch: [214]  [ 80/156]  eta: 0:00:57  lr: 0.001704  min_lr: 0.001704  loss: 2.8139 (2.7303)  weight_decay: 0.0500 (0.0500)  time: 0.7302  data: 0.0003  max mem: 57114
Epoch: [214]  [ 90/156]  eta: 0:00:49  lr: 0.001702  min_lr: 0.001702  loss: 2.9338 (2.7680)  weight_decay: 0.0500 (0.0500)  time: 0.7218  data: 0.0004  max mem: 57114
Epoch: [214]  [100/156]  eta: 0:00:41  lr: 0.001699  min_lr: 0.001699  loss: 3.0048 (2.7920)  weight_decay: 0.0500 (0.0500)  time: 0.7064  data: 0.0004  max mem: 57114
Epoch: [214]  [110/156]  eta: 0:00:34  lr: 0.001697  min_lr: 0.001697  loss: 2.9215 (2.7781)  weight_decay: 0.0500 (0.0500)  time: 0.7066  data: 0.0004  max mem: 57114
Epoch: [214]  [120/156]  eta: 0:00:26  lr: 0.001695  min_lr: 0.001695  loss: 2.9665 (2.7995)  weight_decay: 0.0500 (0.0500)  time: 0.7046  data: 0.0004  max mem: 57114
Epoch: [214]  [130/156]  eta: 0:00:19  lr: 0.001692  min_lr: 0.001692  loss: 3.0128 (2.8098)  weight_decay: 0.0500 (0.0500)  time: 0.7066  data: 0.0008  max mem: 57114
Epoch: [214]  [140/156]  eta: 0:00:11  lr: 0.001690  min_lr: 0.001690  loss: 3.0215 (2.8177)  weight_decay: 0.0500 (0.0500)  time: 0.6969  data: 0.0007  max mem: 57114
Epoch: [214]  [150/156]  eta: 0:00:04  lr: 0.001688  min_lr: 0.001688  loss: 3.0253 (2.8284)  weight_decay: 0.0500 (0.0500)  time: 0.6816  data: 0.0001  max mem: 57114
Epoch: [214]  [155/156]  eta: 0:00:00  lr: 0.001686  min_lr: 0.001686  loss: 2.9849 (2.8256)  weight_decay: 0.0500 (0.0500)  time: 0.6769  data: 0.0001  max mem: 57114
Epoch: [214] Total time: 0:01:53 (0.7276 s / it)
Averaged stats: lr: 0.001686  min_lr: 0.001686  loss: 2.9849 (2.8047)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6697 (0.6697)  acc1: 89.0625 (89.0625)  acc5: 98.4375 (98.4375)  time: 2.0978  data: 1.8419  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8195 (0.8152)  acc1: 82.8125 (81.4578)  acc5: 97.9167 (97.0588)  time: 0.5703  data: 0.3684  max mem: 57114
Test: Total time: 0:00:02 (0.5916 s / it)
* Acc@1 82.121 Acc@5 96.076 loss 0.889
Accuracy of the model on the 50000 test images: 82.1%
Max accuracy: 82.59%
Test:  [0/5]  eta: 0:00:11  loss: 0.5888 (0.5888)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.3700  data: 2.1264  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7068 (0.6796)  acc1: 80.7292 (78.9003)  acc5: 96.3542 (94.1177)  time: 0.6247  data: 0.4254  max mem: 57114
Test: Total time: 0:00:03 (0.6361 s / it)
* Acc@1 77.292 Acc@5 93.874 loss 0.822
Accuracy of the model EMA on 50000 test images: 77.3%
Max EMA accuracy: 77.29%
Epoch: [215]  [  0/156]  eta: 0:06:55  lr: 0.001686  min_lr: 0.001686  loss: 2.9261 (2.9261)  weight_decay: 0.0500 (0.0500)  time: 2.6639  data: 2.0164  max mem: 57114
Epoch: [215]  [ 10/156]  eta: 0:02:12  lr: 0.001684  min_lr: 0.001684  loss: 2.9261 (2.8429)  weight_decay: 0.0500 (0.0500)  time: 0.9053  data: 0.1835  max mem: 57114
Epoch: [215]  [ 20/156]  eta: 0:01:51  lr: 0.001681  min_lr: 0.001681  loss: 3.0051 (2.9248)  weight_decay: 0.0500 (0.0500)  time: 0.7306  data: 0.0003  max mem: 57114
Epoch: [215]  [ 30/156]  eta: 0:01:38  lr: 0.001679  min_lr: 0.001679  loss: 2.9810 (2.8710)  weight_decay: 0.0500 (0.0500)  time: 0.7172  data: 0.0004  max mem: 57114
Epoch: [215]  [ 40/156]  eta: 0:01:29  lr: 0.001677  min_lr: 0.001677  loss: 2.9570 (2.8840)  weight_decay: 0.0500 (0.0500)  time: 0.7125  data: 0.0004  max mem: 57114
Epoch: [215]  [ 50/156]  eta: 0:01:20  lr: 0.001674  min_lr: 0.001674  loss: 3.0666 (2.8646)  weight_decay: 0.0500 (0.0500)  time: 0.7241  data: 0.0004  max mem: 57114
Epoch: [215]  [ 60/156]  eta: 0:01:12  lr: 0.001672  min_lr: 0.001672  loss: 2.9298 (2.8645)  weight_decay: 0.0500 (0.0500)  time: 0.7162  data: 0.0004  max mem: 57114
Epoch: [215]  [ 70/156]  eta: 0:01:04  lr: 0.001670  min_lr: 0.001670  loss: 2.9803 (2.8748)  weight_decay: 0.0500 (0.0500)  time: 0.7120  data: 0.0004  max mem: 57114
Epoch: [215]  [ 80/156]  eta: 0:00:56  lr: 0.001667  min_lr: 0.001667  loss: 2.9275 (2.8565)  weight_decay: 0.0500 (0.0500)  time: 0.7131  data: 0.0004  max mem: 57114
Epoch: [215]  [ 90/156]  eta: 0:00:48  lr: 0.001665  min_lr: 0.001665  loss: 3.0788 (2.8874)  weight_decay: 0.0500 (0.0500)  time: 0.7045  data: 0.0004  max mem: 57114
Epoch: [215]  [100/156]  eta: 0:00:41  lr: 0.001663  min_lr: 0.001663  loss: 3.0542 (2.8607)  weight_decay: 0.0500 (0.0500)  time: 0.6975  data: 0.0004  max mem: 57114
Epoch: [215]  [110/156]  eta: 0:00:33  lr: 0.001660  min_lr: 0.001660  loss: 2.6133 (2.8330)  weight_decay: 0.0500 (0.0500)  time: 0.6937  data: 0.0004  max mem: 57114
Epoch: [215]  [120/156]  eta: 0:00:26  lr: 0.001658  min_lr: 0.001658  loss: 2.7857 (2.8333)  weight_decay: 0.0500 (0.0500)  time: 0.7066  data: 0.0004  max mem: 57114
Epoch: [215]  [130/156]  eta: 0:00:18  lr: 0.001656  min_lr: 0.001656  loss: 2.8538 (2.8228)  weight_decay: 0.0500 (0.0500)  time: 0.7070  data: 0.0008  max mem: 57114
Epoch: [215]  [140/156]  eta: 0:00:11  lr: 0.001653  min_lr: 0.001653  loss: 2.8923 (2.8307)  weight_decay: 0.0500 (0.0500)  time: 0.6913  data: 0.0007  max mem: 57114
Epoch: [215]  [150/156]  eta: 0:00:04  lr: 0.001651  min_lr: 0.001651  loss: 2.7870 (2.8100)  weight_decay: 0.0500 (0.0500)  time: 0.6938  data: 0.0001  max mem: 57114
Epoch: [215]  [155/156]  eta: 0:00:00  lr: 0.001650  min_lr: 0.001650  loss: 2.7864 (2.8135)  weight_decay: 0.0500 (0.0500)  time: 0.6919  data: 0.0001  max mem: 57114
Epoch: [215] Total time: 0:01:52 (0.7227 s / it)
Averaged stats: lr: 0.001650  min_lr: 0.001650  loss: 2.7864 (2.8011)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6498 (0.6498)  acc1: 89.0625 (89.0625)  acc5: 98.9583 (98.9583)  time: 2.0225  data: 1.7670  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9024 (0.8064)  acc1: 82.2917 (82.6087)  acc5: 96.3542 (96.5473)  time: 0.5551  data: 0.3535  max mem: 57114
Test: Total time: 0:00:02 (0.5749 s / it)
* Acc@1 82.379 Acc@5 96.282 loss 0.880
Accuracy of the model on the 50000 test images: 82.4%
Max accuracy: 82.59%
Test:  [0/5]  eta: 0:00:11  loss: 0.5880 (0.5880)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.2700  data: 2.0264  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7054 (0.6786)  acc1: 80.7292 (78.9003)  acc5: 96.8750 (94.2455)  time: 0.6046  data: 0.4053  max mem: 57114
Test: Total time: 0:00:03 (0.6188 s / it)
* Acc@1 77.326 Acc@5 93.892 loss 0.820
Accuracy of the model EMA on 50000 test images: 77.3%
Max EMA accuracy: 77.33%
Epoch: [216]  [  0/156]  eta: 0:06:52  lr: 0.001650  min_lr: 0.001650  loss: 3.2552 (3.2552)  weight_decay: 0.0500 (0.0500)  time: 2.6449  data: 1.9959  max mem: 57114
Epoch: [216]  [ 10/156]  eta: 0:02:07  lr: 0.001647  min_lr: 0.001647  loss: 2.9157 (2.8355)  weight_decay: 0.0500 (0.0500)  time: 0.8757  data: 0.1817  max mem: 57114
Epoch: [216]  [ 20/156]  eta: 0:01:48  lr: 0.001645  min_lr: 0.001645  loss: 2.8921 (2.8048)  weight_decay: 0.0500 (0.0500)  time: 0.7089  data: 0.0003  max mem: 57114
Epoch: [216]  [ 30/156]  eta: 0:01:37  lr: 0.001643  min_lr: 0.001643  loss: 2.7639 (2.6903)  weight_decay: 0.0500 (0.0500)  time: 0.7204  data: 0.0003  max mem: 57114
Epoch: [216]  [ 40/156]  eta: 0:01:28  lr: 0.001640  min_lr: 0.001640  loss: 2.6247 (2.6805)  weight_decay: 0.0500 (0.0500)  time: 0.7263  data: 0.0004  max mem: 57114
Epoch: [216]  [ 50/156]  eta: 0:01:20  lr: 0.001638  min_lr: 0.001638  loss: 2.8831 (2.7425)  weight_decay: 0.0500 (0.0500)  time: 0.7258  data: 0.0005  max mem: 57114
Epoch: [216]  [ 60/156]  eta: 0:01:12  lr: 0.001636  min_lr: 0.001636  loss: 2.8512 (2.7135)  weight_decay: 0.0500 (0.0500)  time: 0.7344  data: 0.0004  max mem: 57114
Epoch: [216]  [ 70/156]  eta: 0:01:04  lr: 0.001633  min_lr: 0.001633  loss: 2.5431 (2.7032)  weight_decay: 0.0500 (0.0500)  time: 0.7321  data: 0.0004  max mem: 57114
Epoch: [216]  [ 80/156]  eta: 0:00:56  lr: 0.001631  min_lr: 0.001631  loss: 2.8804 (2.7082)  weight_decay: 0.0500 (0.0500)  time: 0.7125  data: 0.0004  max mem: 57114
Epoch: [216]  [ 90/156]  eta: 0:00:48  lr: 0.001629  min_lr: 0.001629  loss: 2.7075 (2.6930)  weight_decay: 0.0500 (0.0500)  time: 0.7089  data: 0.0004  max mem: 57114
Epoch: [216]  [100/156]  eta: 0:00:41  lr: 0.001626  min_lr: 0.001626  loss: 2.7075 (2.6913)  weight_decay: 0.0500 (0.0500)  time: 0.7049  data: 0.0004  max mem: 57114
Epoch: [216]  [110/156]  eta: 0:00:33  lr: 0.001624  min_lr: 0.001624  loss: 3.0020 (2.7203)  weight_decay: 0.0500 (0.0500)  time: 0.7109  data: 0.0004  max mem: 57114
Epoch: [216]  [120/156]  eta: 0:00:26  lr: 0.001622  min_lr: 0.001622  loss: 2.9910 (2.7291)  weight_decay: 0.0500 (0.0500)  time: 0.7159  data: 0.0004  max mem: 57114
Epoch: [216]  [130/156]  eta: 0:00:18  lr: 0.001620  min_lr: 0.001620  loss: 2.8891 (2.7343)  weight_decay: 0.0500 (0.0500)  time: 0.6997  data: 0.0005  max mem: 57114
Epoch: [216]  [140/156]  eta: 0:00:11  lr: 0.001617  min_lr: 0.001617  loss: 2.9321 (2.7400)  weight_decay: 0.0500 (0.0500)  time: 0.6834  data: 0.0004  max mem: 57114
Epoch: [216]  [150/156]  eta: 0:00:04  lr: 0.001615  min_lr: 0.001615  loss: 2.9792 (2.7486)  weight_decay: 0.0500 (0.0500)  time: 0.6830  data: 0.0001  max mem: 57114
Epoch: [216]  [155/156]  eta: 0:00:00  lr: 0.001614  min_lr: 0.001614  loss: 2.9907 (2.7574)  weight_decay: 0.0500 (0.0500)  time: 0.6879  data: 0.0001  max mem: 57114
Epoch: [216] Total time: 0:01:53 (0.7247 s / it)
Averaged stats: lr: 0.001614  min_lr: 0.001614  loss: 2.9907 (2.7977)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7400 (0.7400)  acc1: 90.1042 (90.1042)  acc5: 98.9583 (98.9583)  time: 2.0434  data: 1.7863  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9928 (0.9050)  acc1: 81.7708 (82.8645)  acc5: 97.3958 (96.2916)  time: 0.5596  data: 0.3574  max mem: 57114
Test: Total time: 0:00:02 (0.5831 s / it)
* Acc@1 82.709 Acc@5 96.282 loss 0.974
Accuracy of the model on the 50000 test images: 82.7%
Max accuracy: 82.71%
Test:  [0/5]  eta: 0:00:10  loss: 0.5876 (0.5876)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.0626  data: 1.8190  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7040 (0.6778)  acc1: 81.2500 (79.0281)  acc5: 96.8750 (94.2455)  time: 0.5631  data: 0.3639  max mem: 57114
Test: Total time: 0:00:02 (0.5732 s / it)
* Acc@1 77.316 Acc@5 93.908 loss 0.819
Accuracy of the model EMA on 50000 test images: 77.3%
Epoch: [217]  [  0/156]  eta: 0:13:12  lr: 0.001613  min_lr: 0.001613  loss: 2.4681 (2.4681)  weight_decay: 0.0500 (0.0500)  time: 5.0817  data: 3.5839  max mem: 57114
Epoch: [217]  [ 10/156]  eta: 0:02:45  lr: 0.001611  min_lr: 0.001611  loss: 3.0939 (2.9551)  weight_decay: 0.0500 (0.0500)  time: 1.1352  data: 0.3262  max mem: 57114
Epoch: [217]  [ 20/156]  eta: 0:02:07  lr: 0.001609  min_lr: 0.001609  loss: 3.0579 (2.9152)  weight_decay: 0.0500 (0.0500)  time: 0.7319  data: 0.0004  max mem: 57114
Epoch: [217]  [ 30/156]  eta: 0:01:50  lr: 0.001607  min_lr: 0.001607  loss: 2.9400 (2.8891)  weight_decay: 0.0500 (0.0500)  time: 0.7390  data: 0.0003  max mem: 57114
Epoch: [217]  [ 40/156]  eta: 0:01:37  lr: 0.001604  min_lr: 0.001604  loss: 2.7994 (2.8673)  weight_decay: 0.0500 (0.0500)  time: 0.7414  data: 0.0003  max mem: 57114
Epoch: [217]  [ 50/156]  eta: 0:01:26  lr: 0.001602  min_lr: 0.001602  loss: 2.9340 (2.8602)  weight_decay: 0.0500 (0.0500)  time: 0.7191  data: 0.0003  max mem: 57114
Epoch: [217]  [ 60/156]  eta: 0:01:17  lr: 0.001600  min_lr: 0.001600  loss: 2.8926 (2.7992)  weight_decay: 0.0500 (0.0500)  time: 0.7384  data: 0.0004  max mem: 57114
Epoch: [217]  [ 70/156]  eta: 0:01:08  lr: 0.001597  min_lr: 0.001597  loss: 2.4654 (2.7570)  weight_decay: 0.0500 (0.0500)  time: 0.7373  data: 0.0004  max mem: 57114
Epoch: [217]  [ 80/156]  eta: 0:00:59  lr: 0.001595  min_lr: 0.001595  loss: 2.7259 (2.7903)  weight_decay: 0.0500 (0.0500)  time: 0.7106  data: 0.0004  max mem: 57114
Epoch: [217]  [ 90/156]  eta: 0:00:51  lr: 0.001593  min_lr: 0.001593  loss: 2.9721 (2.7828)  weight_decay: 0.0500 (0.0500)  time: 0.7098  data: 0.0004  max mem: 57114
Epoch: [217]  [100/156]  eta: 0:00:43  lr: 0.001590  min_lr: 0.001590  loss: 2.5400 (2.7516)  weight_decay: 0.0500 (0.0500)  time: 0.7084  data: 0.0004  max mem: 57114
Epoch: [217]  [110/156]  eta: 0:00:35  lr: 0.001588  min_lr: 0.001588  loss: 2.6261 (2.7546)  weight_decay: 0.0500 (0.0500)  time: 0.7061  data: 0.0004  max mem: 57114
Epoch: [217]  [120/156]  eta: 0:00:27  lr: 0.001586  min_lr: 0.001586  loss: 2.7959 (2.7504)  weight_decay: 0.0500 (0.0500)  time: 0.7007  data: 0.0004  max mem: 57114
Epoch: [217]  [130/156]  eta: 0:00:19  lr: 0.001584  min_lr: 0.001584  loss: 2.8840 (2.7618)  weight_decay: 0.0500 (0.0500)  time: 0.6945  data: 0.0008  max mem: 57114
Epoch: [217]  [140/156]  eta: 0:00:11  lr: 0.001581  min_lr: 0.001581  loss: 2.9407 (2.7556)  weight_decay: 0.0500 (0.0500)  time: 0.6869  data: 0.0007  max mem: 57114
Epoch: [217]  [150/156]  eta: 0:00:04  lr: 0.001579  min_lr: 0.001579  loss: 2.9231 (2.7580)  weight_decay: 0.0500 (0.0500)  time: 0.6863  data: 0.0001  max mem: 57114
Epoch: [217]  [155/156]  eta: 0:00:00  lr: 0.001578  min_lr: 0.001578  loss: 2.9781 (2.7624)  weight_decay: 0.0500 (0.0500)  time: 0.6868  data: 0.0001  max mem: 57114
Epoch: [217] Total time: 0:01:56 (0.7444 s / it)
Averaged stats: lr: 0.001578  min_lr: 0.001578  loss: 2.9781 (2.7921)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5860 (0.5860)  acc1: 87.5000 (87.5000)  acc5: 98.9583 (98.9583)  time: 2.0266  data: 1.7709  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8304 (0.7369)  acc1: 81.7708 (82.7366)  acc5: 97.9167 (97.0588)  time: 0.5559  data: 0.3542  max mem: 57114
Test: Total time: 0:00:02 (0.5748 s / it)
* Acc@1 82.996 Acc@5 96.365 loss 0.828
Accuracy of the model on the 50000 test images: 83.0%
Max accuracy: 83.00%
Test:  [0/5]  eta: 0:00:10  loss: 0.5873 (0.5873)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.0302  data: 1.7866  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7028 (0.6770)  acc1: 81.2500 (79.2839)  acc5: 96.8750 (94.2455)  time: 0.5567  data: 0.3574  max mem: 57114
Test: Total time: 0:00:02 (0.5682 s / it)
* Acc@1 77.344 Acc@5 93.912 loss 0.818
Accuracy of the model EMA on 50000 test images: 77.3%
Max EMA accuracy: 77.34%
Epoch: [218]  [  0/156]  eta: 0:07:29  lr: 0.001578  min_lr: 0.001578  loss: 2.3260 (2.3260)  weight_decay: 0.0500 (0.0500)  time: 2.8837  data: 2.2381  max mem: 57114
Epoch: [218]  [ 10/156]  eta: 0:02:13  lr: 0.001575  min_lr: 0.001575  loss: 2.8035 (2.7279)  weight_decay: 0.0500 (0.0500)  time: 0.9132  data: 0.2037  max mem: 57114
Epoch: [218]  [ 20/156]  eta: 0:01:51  lr: 0.001573  min_lr: 0.001573  loss: 2.8035 (2.7381)  weight_decay: 0.0500 (0.0500)  time: 0.7195  data: 0.0003  max mem: 57114
Epoch: [218]  [ 30/156]  eta: 0:01:39  lr: 0.001571  min_lr: 0.001571  loss: 2.7775 (2.7589)  weight_decay: 0.0500 (0.0500)  time: 0.7236  data: 0.0004  max mem: 57114
Epoch: [218]  [ 40/156]  eta: 0:01:30  lr: 0.001568  min_lr: 0.001568  loss: 2.8025 (2.7586)  weight_decay: 0.0500 (0.0500)  time: 0.7385  data: 0.0004  max mem: 57114
Epoch: [218]  [ 50/156]  eta: 0:01:22  lr: 0.001566  min_lr: 0.001566  loss: 2.8025 (2.7556)  weight_decay: 0.0500 (0.0500)  time: 0.7607  data: 0.0004  max mem: 57114
Epoch: [218]  [ 60/156]  eta: 0:01:13  lr: 0.001564  min_lr: 0.001564  loss: 2.8507 (2.7838)  weight_decay: 0.0500 (0.0500)  time: 0.7446  data: 0.0004  max mem: 57114
Epoch: [218]  [ 70/156]  eta: 0:01:05  lr: 0.001562  min_lr: 0.001562  loss: 2.8507 (2.7864)  weight_decay: 0.0500 (0.0500)  time: 0.7329  data: 0.0004  max mem: 57114
Epoch: [218]  [ 80/156]  eta: 0:00:57  lr: 0.001559  min_lr: 0.001559  loss: 2.6080 (2.7627)  weight_decay: 0.0500 (0.0500)  time: 0.7427  data: 0.0004  max mem: 57114
Epoch: [218]  [ 90/156]  eta: 0:00:49  lr: 0.001557  min_lr: 0.001557  loss: 2.8761 (2.7800)  weight_decay: 0.0500 (0.0500)  time: 0.7211  data: 0.0004  max mem: 57114
Epoch: [218]  [100/156]  eta: 0:00:42  lr: 0.001555  min_lr: 0.001555  loss: 2.8273 (2.7560)  weight_decay: 0.0500 (0.0500)  time: 0.7068  data: 0.0004  max mem: 57114
Epoch: [218]  [110/156]  eta: 0:00:34  lr: 0.001553  min_lr: 0.001553  loss: 2.7250 (2.7534)  weight_decay: 0.0500 (0.0500)  time: 0.7080  data: 0.0004  max mem: 57114
Epoch: [218]  [120/156]  eta: 0:00:26  lr: 0.001550  min_lr: 0.001550  loss: 2.7892 (2.7541)  weight_decay: 0.0500 (0.0500)  time: 0.7018  data: 0.0004  max mem: 57114
Epoch: [218]  [130/156]  eta: 0:00:19  lr: 0.001548  min_lr: 0.001548  loss: 2.9273 (2.7759)  weight_decay: 0.0500 (0.0500)  time: 0.6973  data: 0.0004  max mem: 57114
Epoch: [218]  [140/156]  eta: 0:00:11  lr: 0.001546  min_lr: 0.001546  loss: 2.8515 (2.7598)  weight_decay: 0.0500 (0.0500)  time: 0.6945  data: 0.0003  max mem: 57114
Epoch: [218]  [150/156]  eta: 0:00:04  lr: 0.001543  min_lr: 0.001543  loss: 2.6936 (2.7698)  weight_decay: 0.0500 (0.0500)  time: 0.6896  data: 0.0001  max mem: 57114
Epoch: [218]  [155/156]  eta: 0:00:00  lr: 0.001542  min_lr: 0.001542  loss: 2.6598 (2.7703)  weight_decay: 0.0500 (0.0500)  time: 0.6833  data: 0.0001  max mem: 57114
Epoch: [218] Total time: 0:01:54 (0.7333 s / it)
Averaged stats: lr: 0.001542  min_lr: 0.001542  loss: 2.6598 (2.7905)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6929 (0.6929)  acc1: 88.0208 (88.0208)  acc5: 98.9583 (98.9583)  time: 2.0599  data: 1.8011  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8601 (0.8188)  acc1: 84.8958 (82.7366)  acc5: 97.9167 (97.0588)  time: 0.5628  data: 0.3603  max mem: 57114
Test: Total time: 0:00:02 (0.5846 s / it)
* Acc@1 82.585 Acc@5 96.264 loss 0.910
Accuracy of the model on the 50000 test images: 82.6%
Max accuracy: 83.00%
Test:  [0/5]  eta: 0:00:11  loss: 0.5871 (0.5871)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.3067  data: 2.0632  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7015 (0.6763)  acc1: 81.2500 (79.2839)  acc5: 96.8750 (94.2455)  time: 0.6126  data: 0.4127  max mem: 57114
Test: Total time: 0:00:03 (0.6258 s / it)
* Acc@1 77.378 Acc@5 93.912 loss 0.817
Accuracy of the model EMA on 50000 test images: 77.4%
Max EMA accuracy: 77.38%
Epoch: [219]  [  0/156]  eta: 0:06:36  lr: 0.001542  min_lr: 0.001542  loss: 3.2494 (3.2494)  weight_decay: 0.0500 (0.0500)  time: 2.5414  data: 1.8894  max mem: 57114
Epoch: [219]  [ 10/156]  eta: 0:02:09  lr: 0.001540  min_lr: 0.001540  loss: 2.6458 (2.6560)  weight_decay: 0.0500 (0.0500)  time: 0.8838  data: 0.1720  max mem: 57114
Epoch: [219]  [ 20/156]  eta: 0:01:49  lr: 0.001538  min_lr: 0.001538  loss: 2.8356 (2.6962)  weight_decay: 0.0500 (0.0500)  time: 0.7202  data: 0.0003  max mem: 57114
Epoch: [219]  [ 30/156]  eta: 0:01:37  lr: 0.001535  min_lr: 0.001535  loss: 2.8966 (2.7169)  weight_decay: 0.0500 (0.0500)  time: 0.7179  data: 0.0004  max mem: 57114
Epoch: [219]  [ 40/156]  eta: 0:01:28  lr: 0.001533  min_lr: 0.001533  loss: 2.7510 (2.6533)  weight_decay: 0.0500 (0.0500)  time: 0.7192  data: 0.0004  max mem: 57114
Epoch: [219]  [ 50/156]  eta: 0:01:20  lr: 0.001531  min_lr: 0.001531  loss: 2.5255 (2.6581)  weight_decay: 0.0500 (0.0500)  time: 0.7347  data: 0.0004  max mem: 57114
Epoch: [219]  [ 60/156]  eta: 0:01:12  lr: 0.001529  min_lr: 0.001529  loss: 2.5255 (2.6435)  weight_decay: 0.0500 (0.0500)  time: 0.7296  data: 0.0004  max mem: 57114
Epoch: [219]  [ 70/156]  eta: 0:01:04  lr: 0.001526  min_lr: 0.001526  loss: 2.7146 (2.6346)  weight_decay: 0.0500 (0.0500)  time: 0.7108  data: 0.0004  max mem: 57114
Epoch: [219]  [ 80/156]  eta: 0:00:56  lr: 0.001524  min_lr: 0.001524  loss: 2.8541 (2.6482)  weight_decay: 0.0500 (0.0500)  time: 0.7053  data: 0.0004  max mem: 57114
Epoch: [219]  [ 90/156]  eta: 0:00:48  lr: 0.001522  min_lr: 0.001522  loss: 2.9468 (2.6537)  weight_decay: 0.0500 (0.0500)  time: 0.7038  data: 0.0004  max mem: 57114
Epoch: [219]  [100/156]  eta: 0:00:41  lr: 0.001519  min_lr: 0.001519  loss: 2.9801 (2.6669)  weight_decay: 0.0500 (0.0500)  time: 0.7084  data: 0.0004  max mem: 57114
Epoch: [219]  [110/156]  eta: 0:00:33  lr: 0.001517  min_lr: 0.001517  loss: 2.9191 (2.6858)  weight_decay: 0.0500 (0.0500)  time: 0.7070  data: 0.0004  max mem: 57114
Epoch: [219]  [120/156]  eta: 0:00:26  lr: 0.001515  min_lr: 0.001515  loss: 2.9191 (2.6949)  weight_decay: 0.0500 (0.0500)  time: 0.6959  data: 0.0004  max mem: 57114
Epoch: [219]  [130/156]  eta: 0:00:18  lr: 0.001513  min_lr: 0.001513  loss: 2.9063 (2.7044)  weight_decay: 0.0500 (0.0500)  time: 0.6916  data: 0.0009  max mem: 57114
Epoch: [219]  [140/156]  eta: 0:00:11  lr: 0.001510  min_lr: 0.001510  loss: 2.9063 (2.7229)  weight_decay: 0.0500 (0.0500)  time: 0.6943  data: 0.0007  max mem: 57114
Epoch: [219]  [150/156]  eta: 0:00:04  lr: 0.001508  min_lr: 0.001508  loss: 2.8681 (2.7246)  weight_decay: 0.0500 (0.0500)  time: 0.6886  data: 0.0001  max mem: 57114
Epoch: [219]  [155/156]  eta: 0:00:00  lr: 0.001507  min_lr: 0.001507  loss: 2.9777 (2.7345)  weight_decay: 0.0500 (0.0500)  time: 0.6845  data: 0.0001  max mem: 57114
Epoch: [219] Total time: 0:01:52 (0.7217 s / it)
Averaged stats: lr: 0.001507  min_lr: 0.001507  loss: 2.9777 (2.7903)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7473 (0.7473)  acc1: 88.5417 (88.5417)  acc5: 98.9583 (98.9583)  time: 2.0559  data: 1.7994  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.9031 (0.8556)  acc1: 83.3333 (83.3760)  acc5: 96.3542 (96.5473)  time: 0.5619  data: 0.3600  max mem: 57114
Test: Total time: 0:00:02 (0.5832 s / it)
* Acc@1 82.846 Acc@5 96.367 loss 0.945
Accuracy of the model on the 50000 test images: 82.8%
Max accuracy: 83.00%
Test:  [0/5]  eta: 0:00:11  loss: 0.5866 (0.5866)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.3336  data: 2.0900  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7002 (0.6756)  acc1: 81.2500 (79.4118)  acc5: 96.8750 (94.2455)  time: 0.6174  data: 0.4181  max mem: 57114
Test: Total time: 0:00:03 (0.6284 s / it)
* Acc@1 77.364 Acc@5 93.926 loss 0.816
Accuracy of the model EMA on 50000 test images: 77.4%
Epoch: [220]  [  0/156]  eta: 0:13:06  lr: 0.001507  min_lr: 0.001507  loss: 3.0378 (3.0378)  weight_decay: 0.0500 (0.0500)  time: 5.0448  data: 4.1106  max mem: 57114
Epoch: [220]  [ 10/156]  eta: 0:02:44  lr: 0.001505  min_lr: 0.001505  loss: 3.0378 (2.8864)  weight_decay: 0.0500 (0.0500)  time: 1.1265  data: 0.3739  max mem: 57114
Epoch: [220]  [ 20/156]  eta: 0:02:06  lr: 0.001502  min_lr: 0.001502  loss: 2.8286 (2.7010)  weight_decay: 0.0500 (0.0500)  time: 0.7227  data: 0.0003  max mem: 57114
Epoch: [220]  [ 30/156]  eta: 0:01:48  lr: 0.001500  min_lr: 0.001500  loss: 2.6623 (2.7017)  weight_decay: 0.0500 (0.0500)  time: 0.7141  data: 0.0003  max mem: 57114
Epoch: [220]  [ 40/156]  eta: 0:01:35  lr: 0.001498  min_lr: 0.001498  loss: 2.8710 (2.7522)  weight_decay: 0.0500 (0.0500)  time: 0.7191  data: 0.0003  max mem: 57114
Epoch: [220]  [ 50/156]  eta: 0:01:25  lr: 0.001496  min_lr: 0.001496  loss: 2.9978 (2.7934)  weight_decay: 0.0500 (0.0500)  time: 0.7350  data: 0.0003  max mem: 57114
Epoch: [220]  [ 60/156]  eta: 0:01:16  lr: 0.001493  min_lr: 0.001493  loss: 3.0795 (2.8251)  weight_decay: 0.0500 (0.0500)  time: 0.7390  data: 0.0003  max mem: 57114
Epoch: [220]  [ 70/156]  eta: 0:01:07  lr: 0.001491  min_lr: 0.001491  loss: 2.8684 (2.8055)  weight_decay: 0.0500 (0.0500)  time: 0.7198  data: 0.0004  max mem: 57114
Epoch: [220]  [ 80/156]  eta: 0:00:58  lr: 0.001489  min_lr: 0.001489  loss: 2.7513 (2.8092)  weight_decay: 0.0500 (0.0500)  time: 0.7080  data: 0.0004  max mem: 57114
Epoch: [220]  [ 90/156]  eta: 0:00:50  lr: 0.001487  min_lr: 0.001487  loss: 2.8227 (2.8119)  weight_decay: 0.0500 (0.0500)  time: 0.7058  data: 0.0003  max mem: 57114
Epoch: [220]  [100/156]  eta: 0:00:42  lr: 0.001484  min_lr: 0.001484  loss: 2.8186 (2.7859)  weight_decay: 0.0500 (0.0500)  time: 0.7043  data: 0.0003  max mem: 57114
Epoch: [220]  [110/156]  eta: 0:00:34  lr: 0.001482  min_lr: 0.001482  loss: 2.6561 (2.7753)  weight_decay: 0.0500 (0.0500)  time: 0.7107  data: 0.0004  max mem: 57114
Epoch: [220]  [120/156]  eta: 0:00:27  lr: 0.001480  min_lr: 0.001480  loss: 2.7482 (2.7781)  weight_decay: 0.0500 (0.0500)  time: 0.7111  data: 0.0004  max mem: 57114
Epoch: [220]  [130/156]  eta: 0:00:19  lr: 0.001478  min_lr: 0.001478  loss: 2.9941 (2.7894)  weight_decay: 0.0500 (0.0500)  time: 0.6991  data: 0.0005  max mem: 57114
Epoch: [220]  [140/156]  eta: 0:00:11  lr: 0.001475  min_lr: 0.001475  loss: 2.9238 (2.7807)  weight_decay: 0.0500 (0.0500)  time: 0.6931  data: 0.0003  max mem: 57114
Epoch: [220]  [150/156]  eta: 0:00:04  lr: 0.001473  min_lr: 0.001473  loss: 2.5924 (2.7722)  weight_decay: 0.0500 (0.0500)  time: 0.6866  data: 0.0001  max mem: 57114
Epoch: [220]  [155/156]  eta: 0:00:00  lr: 0.001472  min_lr: 0.001472  loss: 2.8782 (2.7798)  weight_decay: 0.0500 (0.0500)  time: 0.6887  data: 0.0001  max mem: 57114
Epoch: [220] Total time: 0:01:55 (0.7407 s / it)
Averaged stats: lr: 0.001472  min_lr: 0.001472  loss: 2.8782 (2.7795)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6861 (0.6861)  acc1: 89.5833 (89.5833)  acc5: 98.4375 (98.4375)  time: 2.0828  data: 1.8220  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8673 (0.8323)  acc1: 84.8958 (83.5038)  acc5: 98.4375 (97.3146)  time: 0.5675  data: 0.3645  max mem: 57114
Test: Total time: 0:00:02 (0.5856 s / it)
* Acc@1 82.751 Acc@5 96.407 loss 0.910
Accuracy of the model on the 50000 test images: 82.8%
Max accuracy: 83.00%
Test:  [0/5]  eta: 0:00:11  loss: 0.5864 (0.5864)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.2531  data: 2.0096  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6988 (0.6749)  acc1: 81.2500 (79.4118)  acc5: 96.8750 (94.2455)  time: 0.6013  data: 0.4020  max mem: 57114
Test: Total time: 0:00:03 (0.6166 s / it)
* Acc@1 77.378 Acc@5 93.942 loss 0.815
Accuracy of the model EMA on 50000 test images: 77.4%
Max EMA accuracy: 77.38%
Epoch: [221]  [  0/156]  eta: 0:06:21  lr: 0.001472  min_lr: 0.001472  loss: 2.7650 (2.7650)  weight_decay: 0.0500 (0.0500)  time: 2.4487  data: 1.7726  max mem: 57114
Epoch: [221]  [ 10/156]  eta: 0:02:08  lr: 0.001470  min_lr: 0.001470  loss: 3.0584 (2.9570)  weight_decay: 0.0500 (0.0500)  time: 0.8813  data: 0.1614  max mem: 57114
Epoch: [221]  [ 20/156]  eta: 0:01:50  lr: 0.001467  min_lr: 0.001467  loss: 2.9664 (2.8711)  weight_decay: 0.0500 (0.0500)  time: 0.7281  data: 0.0003  max mem: 57114
Epoch: [221]  [ 30/156]  eta: 0:01:37  lr: 0.001465  min_lr: 0.001465  loss: 2.7610 (2.8373)  weight_decay: 0.0500 (0.0500)  time: 0.7146  data: 0.0004  max mem: 57114
Epoch: [221]  [ 40/156]  eta: 0:01:28  lr: 0.001463  min_lr: 0.001463  loss: 2.7642 (2.8025)  weight_decay: 0.0500 (0.0500)  time: 0.7160  data: 0.0004  max mem: 57114
Epoch: [221]  [ 50/156]  eta: 0:01:20  lr: 0.001461  min_lr: 0.001461  loss: 2.7642 (2.7875)  weight_decay: 0.0500 (0.0500)  time: 0.7386  data: 0.0004  max mem: 57114
Epoch: [221]  [ 60/156]  eta: 0:01:12  lr: 0.001459  min_lr: 0.001459  loss: 2.7157 (2.7681)  weight_decay: 0.0500 (0.0500)  time: 0.7444  data: 0.0004  max mem: 57114
Epoch: [221]  [ 70/156]  eta: 0:01:04  lr: 0.001456  min_lr: 0.001456  loss: 2.8333 (2.7549)  weight_decay: 0.0500 (0.0500)  time: 0.7349  data: 0.0004  max mem: 57114
Epoch: [221]  [ 80/156]  eta: 0:00:56  lr: 0.001454  min_lr: 0.001454  loss: 2.8593 (2.7543)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0003  max mem: 57114
Epoch: [221]  [ 90/156]  eta: 0:00:48  lr: 0.001452  min_lr: 0.001452  loss: 2.9528 (2.7625)  weight_decay: 0.0500 (0.0500)  time: 0.6996  data: 0.0004  max mem: 57114
Epoch: [221]  [100/156]  eta: 0:00:41  lr: 0.001450  min_lr: 0.001450  loss: 2.9030 (2.7653)  weight_decay: 0.0500 (0.0500)  time: 0.7108  data: 0.0004  max mem: 57114
Epoch: [221]  [110/156]  eta: 0:00:33  lr: 0.001447  min_lr: 0.001447  loss: 2.8823 (2.7729)  weight_decay: 0.0500 (0.0500)  time: 0.7151  data: 0.0003  max mem: 57114
Epoch: [221]  [120/156]  eta: 0:00:26  lr: 0.001445  min_lr: 0.001445  loss: 2.7856 (2.7574)  weight_decay: 0.0500 (0.0500)  time: 0.7198  data: 0.0003  max mem: 57114
Epoch: [221]  [130/156]  eta: 0:00:19  lr: 0.001443  min_lr: 0.001443  loss: 2.7856 (2.7585)  weight_decay: 0.0500 (0.0500)  time: 0.7096  data: 0.0010  max mem: 57114
Epoch: [221]  [140/156]  eta: 0:00:11  lr: 0.001441  min_lr: 0.001441  loss: 2.8090 (2.7424)  weight_decay: 0.0500 (0.0500)  time: 0.6880  data: 0.0009  max mem: 57114
Epoch: [221]  [150/156]  eta: 0:00:04  lr: 0.001439  min_lr: 0.001439  loss: 2.7462 (2.7345)  weight_decay: 0.0500 (0.0500)  time: 0.6795  data: 0.0001  max mem: 57114
Epoch: [221]  [155/156]  eta: 0:00:00  lr: 0.001438  min_lr: 0.001438  loss: 2.8642 (2.7378)  weight_decay: 0.0500 (0.0500)  time: 0.6829  data: 0.0001  max mem: 57114
Epoch: [221] Total time: 0:01:53 (0.7264 s / it)
Averaged stats: lr: 0.001438  min_lr: 0.001438  loss: 2.8642 (2.7692)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5499 (0.5499)  acc1: 90.1042 (90.1042)  acc5: 99.4792 (99.4792)  time: 2.0633  data: 1.8072  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7835 (0.6931)  acc1: 83.8542 (83.1202)  acc5: 97.9167 (97.1867)  time: 0.5633  data: 0.3615  max mem: 57114
Test: Total time: 0:00:02 (0.5865 s / it)
* Acc@1 83.232 Acc@5 96.551 loss 0.784
Accuracy of the model on the 50000 test images: 83.2%
Max accuracy: 83.23%
Test:  [0/5]  eta: 0:00:10  loss: 0.5863 (0.5863)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.0206  data: 1.7769  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6975 (0.6744)  acc1: 81.2500 (79.5396)  acc5: 96.8750 (94.2455)  time: 0.5548  data: 0.3555  max mem: 57114
Test: Total time: 0:00:02 (0.5688 s / it)
* Acc@1 77.402 Acc@5 93.960 loss 0.814
Accuracy of the model EMA on 50000 test images: 77.4%
Max EMA accuracy: 77.40%
Epoch: [222]  [  0/156]  eta: 0:08:35  lr: 0.001437  min_lr: 0.001437  loss: 3.1455 (3.1455)  weight_decay: 0.0500 (0.0500)  time: 3.3061  data: 2.6575  max mem: 57114
Epoch: [222]  [ 10/156]  eta: 0:02:20  lr: 0.001435  min_lr: 0.001435  loss: 2.5650 (2.5430)  weight_decay: 0.0500 (0.0500)  time: 0.9608  data: 0.2419  max mem: 57114
Epoch: [222]  [ 20/156]  eta: 0:01:58  lr: 0.001433  min_lr: 0.001433  loss: 2.5650 (2.6212)  weight_decay: 0.0500 (0.0500)  time: 0.7497  data: 0.0003  max mem: 57114
Epoch: [222]  [ 30/156]  eta: 0:01:44  lr: 0.001431  min_lr: 0.001431  loss: 2.7590 (2.6352)  weight_decay: 0.0500 (0.0500)  time: 0.7548  data: 0.0004  max mem: 57114
Epoch: [222]  [ 40/156]  eta: 0:01:32  lr: 0.001428  min_lr: 0.001428  loss: 2.7977 (2.7026)  weight_decay: 0.0500 (0.0500)  time: 0.7232  data: 0.0005  max mem: 57114
Epoch: [222]  [ 50/156]  eta: 0:01:23  lr: 0.001426  min_lr: 0.001426  loss: 2.8509 (2.6766)  weight_decay: 0.0500 (0.0500)  time: 0.7241  data: 0.0004  max mem: 57114
Epoch: [222]  [ 60/156]  eta: 0:01:14  lr: 0.001424  min_lr: 0.001424  loss: 2.8922 (2.7416)  weight_decay: 0.0500 (0.0500)  time: 0.7219  data: 0.0004  max mem: 57114
Epoch: [222]  [ 70/156]  eta: 0:01:06  lr: 0.001422  min_lr: 0.001422  loss: 3.0092 (2.7731)  weight_decay: 0.0500 (0.0500)  time: 0.7220  data: 0.0004  max mem: 57114
Epoch: [222]  [ 80/156]  eta: 0:00:57  lr: 0.001420  min_lr: 0.001420  loss: 3.0143 (2.8083)  weight_decay: 0.0500 (0.0500)  time: 0.7267  data: 0.0004  max mem: 57114
Epoch: [222]  [ 90/156]  eta: 0:00:49  lr: 0.001417  min_lr: 0.001417  loss: 3.0054 (2.8065)  weight_decay: 0.0500 (0.0500)  time: 0.7069  data: 0.0004  max mem: 57114
Epoch: [222]  [100/156]  eta: 0:00:42  lr: 0.001415  min_lr: 0.001415  loss: 2.7921 (2.8092)  weight_decay: 0.0500 (0.0500)  time: 0.7087  data: 0.0004  max mem: 57114
Epoch: [222]  [110/156]  eta: 0:00:34  lr: 0.001413  min_lr: 0.001413  loss: 2.8220 (2.7950)  weight_decay: 0.0500 (0.0500)  time: 0.7056  data: 0.0004  max mem: 57114
Epoch: [222]  [120/156]  eta: 0:00:26  lr: 0.001411  min_lr: 0.001411  loss: 2.8531 (2.7997)  weight_decay: 0.0500 (0.0500)  time: 0.7029  data: 0.0004  max mem: 57114
Epoch: [222]  [130/156]  eta: 0:00:19  lr: 0.001409  min_lr: 0.001409  loss: 2.9417 (2.8027)  weight_decay: 0.0500 (0.0500)  time: 0.7044  data: 0.0008  max mem: 57114
Epoch: [222]  [140/156]  eta: 0:00:11  lr: 0.001407  min_lr: 0.001407  loss: 2.9897 (2.8109)  weight_decay: 0.0500 (0.0500)  time: 0.6911  data: 0.0007  max mem: 57114
Epoch: [222]  [150/156]  eta: 0:00:04  lr: 0.001404  min_lr: 0.001404  loss: 2.9524 (2.8182)  weight_decay: 0.0500 (0.0500)  time: 0.6855  data: 0.0001  max mem: 57114
Epoch: [222]  [155/156]  eta: 0:00:00  lr: 0.001403  min_lr: 0.001403  loss: 2.9445 (2.8048)  weight_decay: 0.0500 (0.0500)  time: 0.6816  data: 0.0001  max mem: 57114
Epoch: [222] Total time: 0:01:54 (0.7334 s / it)
Averaged stats: lr: 0.001403  min_lr: 0.001403  loss: 2.9445 (2.7710)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:09  loss: 0.6109 (0.6109)  acc1: 90.1042 (90.1042)  acc5: 98.9583 (98.9583)  time: 1.9969  data: 1.7410  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7764 (0.7485)  acc1: 82.8125 (82.9923)  acc5: 97.9167 (96.8031)  time: 0.5500  data: 0.3483  max mem: 57114
Test: Total time: 0:00:02 (0.5700 s / it)
* Acc@1 82.890 Acc@5 96.447 loss 0.847
Accuracy of the model on the 50000 test images: 82.9%
Max accuracy: 83.23%
Test:  [0/5]  eta: 0:00:11  loss: 0.5861 (0.5861)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.3804  data: 2.1370  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6962 (0.6738)  acc1: 81.2500 (79.4118)  acc5: 96.8750 (94.3734)  time: 0.6268  data: 0.4275  max mem: 57114
Test: Total time: 0:00:03 (0.6391 s / it)
* Acc@1 77.398 Acc@5 93.962 loss 0.813
Accuracy of the model EMA on 50000 test images: 77.4%
Epoch: [223]  [  0/156]  eta: 0:13:27  lr: 0.001403  min_lr: 0.001403  loss: 2.9177 (2.9177)  weight_decay: 0.0500 (0.0500)  time: 5.1792  data: 3.0889  max mem: 57114
Epoch: [223]  [ 10/156]  eta: 0:02:46  lr: 0.001401  min_lr: 0.001401  loss: 2.8550 (2.6600)  weight_decay: 0.0500 (0.0500)  time: 1.1419  data: 0.2812  max mem: 57114
Epoch: [223]  [ 20/156]  eta: 0:02:07  lr: 0.001399  min_lr: 0.001399  loss: 2.8550 (2.6889)  weight_decay: 0.0500 (0.0500)  time: 0.7246  data: 0.0003  max mem: 57114
Epoch: [223]  [ 30/156]  eta: 0:01:49  lr: 0.001396  min_lr: 0.001396  loss: 2.9430 (2.7384)  weight_decay: 0.0500 (0.0500)  time: 0.7219  data: 0.0003  max mem: 57114
Epoch: [223]  [ 40/156]  eta: 0:01:36  lr: 0.001394  min_lr: 0.001394  loss: 2.7898 (2.7126)  weight_decay: 0.0500 (0.0500)  time: 0.7289  data: 0.0004  max mem: 57114
Epoch: [223]  [ 50/156]  eta: 0:01:26  lr: 0.001392  min_lr: 0.001392  loss: 2.8652 (2.7390)  weight_decay: 0.0500 (0.0500)  time: 0.7271  data: 0.0004  max mem: 57114
Epoch: [223]  [ 60/156]  eta: 0:01:16  lr: 0.001390  min_lr: 0.001390  loss: 2.8343 (2.7210)  weight_decay: 0.0500 (0.0500)  time: 0.7253  data: 0.0004  max mem: 57114
Epoch: [223]  [ 70/156]  eta: 0:01:07  lr: 0.001388  min_lr: 0.001388  loss: 2.7987 (2.7349)  weight_decay: 0.0500 (0.0500)  time: 0.7182  data: 0.0004  max mem: 57114
Epoch: [223]  [ 80/156]  eta: 0:00:59  lr: 0.001386  min_lr: 0.001386  loss: 2.8518 (2.7512)  weight_decay: 0.0500 (0.0500)  time: 0.7191  data: 0.0004  max mem: 57114
Epoch: [223]  [ 90/156]  eta: 0:00:50  lr: 0.001383  min_lr: 0.001383  loss: 3.0145 (2.7788)  weight_decay: 0.0500 (0.0500)  time: 0.7078  data: 0.0004  max mem: 57114
Epoch: [223]  [100/156]  eta: 0:00:42  lr: 0.001381  min_lr: 0.001381  loss: 2.8864 (2.7438)  weight_decay: 0.0500 (0.0500)  time: 0.6977  data: 0.0004  max mem: 57114
Epoch: [223]  [110/156]  eta: 0:00:34  lr: 0.001379  min_lr: 0.001379  loss: 2.5627 (2.7399)  weight_decay: 0.0500 (0.0500)  time: 0.7072  data: 0.0004  max mem: 57114
Epoch: [223]  [120/156]  eta: 0:00:27  lr: 0.001377  min_lr: 0.001377  loss: 2.9093 (2.7577)  weight_decay: 0.0500 (0.0500)  time: 0.7093  data: 0.0004  max mem: 57114
Epoch: [223]  [130/156]  eta: 0:00:19  lr: 0.001375  min_lr: 0.001375  loss: 2.9557 (2.7636)  weight_decay: 0.0500 (0.0500)  time: 0.7026  data: 0.0008  max mem: 57114
Epoch: [223]  [140/156]  eta: 0:00:11  lr: 0.001373  min_lr: 0.001373  loss: 2.8656 (2.7692)  weight_decay: 0.0500 (0.0500)  time: 0.6945  data: 0.0007  max mem: 57114
Epoch: [223]  [150/156]  eta: 0:00:04  lr: 0.001370  min_lr: 0.001370  loss: 2.8656 (2.7651)  weight_decay: 0.0500 (0.0500)  time: 0.6843  data: 0.0001  max mem: 57114
Epoch: [223]  [155/156]  eta: 0:00:00  lr: 0.001369  min_lr: 0.001369  loss: 2.6775 (2.7667)  weight_decay: 0.0500 (0.0500)  time: 0.6845  data: 0.0001  max mem: 57114
Epoch: [223] Total time: 0:01:55 (0.7417 s / it)
Averaged stats: lr: 0.001369  min_lr: 0.001369  loss: 2.6775 (2.7576)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6218 (0.6218)  acc1: 89.5833 (89.5833)  acc5: 99.4792 (99.4792)  time: 2.0444  data: 1.7884  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8097 (0.7626)  acc1: 83.3333 (83.3760)  acc5: 97.9167 (97.5703)  time: 0.5595  data: 0.3577  max mem: 57114
Test: Total time: 0:00:02 (0.5820 s / it)
* Acc@1 83.200 Acc@5 96.375 loss 0.849
Accuracy of the model on the 50000 test images: 83.2%
Max accuracy: 83.23%
Test:  [0/5]  eta: 0:00:11  loss: 0.5862 (0.5862)  acc1: 83.8542 (83.8542)  acc5: 97.3958 (97.3958)  time: 2.3517  data: 2.1081  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6950 (0.6734)  acc1: 81.2500 (79.2839)  acc5: 96.8750 (94.3734)  time: 0.6211  data: 0.4217  max mem: 57114
Test: Total time: 0:00:03 (0.6338 s / it)
* Acc@1 77.438 Acc@5 93.958 loss 0.813
Accuracy of the model EMA on 50000 test images: 77.4%
Max EMA accuracy: 77.44%
Epoch: [224]  [  0/156]  eta: 0:06:04  lr: 0.001369  min_lr: 0.001369  loss: 3.0654 (3.0654)  weight_decay: 0.0500 (0.0500)  time: 2.3344  data: 1.6857  max mem: 57114
Epoch: [224]  [ 10/156]  eta: 0:02:07  lr: 0.001367  min_lr: 0.001367  loss: 3.0654 (3.0235)  weight_decay: 0.0500 (0.0500)  time: 0.8701  data: 0.1535  max mem: 57114
Epoch: [224]  [ 20/156]  eta: 0:01:48  lr: 0.001365  min_lr: 0.001365  loss: 3.0065 (2.9860)  weight_decay: 0.0500 (0.0500)  time: 0.7219  data: 0.0003  max mem: 57114
Epoch: [224]  [ 30/156]  eta: 0:01:37  lr: 0.001363  min_lr: 0.001363  loss: 3.0065 (2.9359)  weight_decay: 0.0500 (0.0500)  time: 0.7163  data: 0.0004  max mem: 57114
Epoch: [224]  [ 40/156]  eta: 0:01:28  lr: 0.001360  min_lr: 0.001360  loss: 2.6563 (2.8035)  weight_decay: 0.0500 (0.0500)  time: 0.7327  data: 0.0004  max mem: 57114
Epoch: [224]  [ 50/156]  eta: 0:01:20  lr: 0.001358  min_lr: 0.001358  loss: 2.5123 (2.7748)  weight_decay: 0.0500 (0.0500)  time: 0.7382  data: 0.0004  max mem: 57114
Epoch: [224]  [ 60/156]  eta: 0:01:12  lr: 0.001356  min_lr: 0.001356  loss: 2.9556 (2.7767)  weight_decay: 0.0500 (0.0500)  time: 0.7184  data: 0.0004  max mem: 57114
Epoch: [224]  [ 70/156]  eta: 0:01:04  lr: 0.001354  min_lr: 0.001354  loss: 2.8081 (2.7555)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0004  max mem: 57114
Epoch: [224]  [ 80/156]  eta: 0:00:56  lr: 0.001352  min_lr: 0.001352  loss: 2.5825 (2.7514)  weight_decay: 0.0500 (0.0500)  time: 0.7021  data: 0.0004  max mem: 57114
Epoch: [224]  [ 90/156]  eta: 0:00:48  lr: 0.001350  min_lr: 0.001350  loss: 2.8448 (2.7533)  weight_decay: 0.0500 (0.0500)  time: 0.7094  data: 0.0003  max mem: 57114
Epoch: [224]  [100/156]  eta: 0:00:41  lr: 0.001347  min_lr: 0.001347  loss: 2.9100 (2.7427)  weight_decay: 0.0500 (0.0500)  time: 0.7235  data: 0.0003  max mem: 57114
Epoch: [224]  [110/156]  eta: 0:00:33  lr: 0.001345  min_lr: 0.001345  loss: 2.8074 (2.7418)  weight_decay: 0.0500 (0.0500)  time: 0.7181  data: 0.0004  max mem: 57114
Epoch: [224]  [120/156]  eta: 0:00:26  lr: 0.001343  min_lr: 0.001343  loss: 2.8074 (2.7323)  weight_decay: 0.0500 (0.0500)  time: 0.7044  data: 0.0004  max mem: 57114
Epoch: [224]  [130/156]  eta: 0:00:18  lr: 0.001341  min_lr: 0.001341  loss: 2.8850 (2.7437)  weight_decay: 0.0500 (0.0500)  time: 0.6980  data: 0.0010  max mem: 57114
Epoch: [224]  [140/156]  eta: 0:00:11  lr: 0.001339  min_lr: 0.001339  loss: 2.8806 (2.7348)  weight_decay: 0.0500 (0.0500)  time: 0.6871  data: 0.0009  max mem: 57114
Epoch: [224]  [150/156]  eta: 0:00:04  lr: 0.001337  min_lr: 0.001337  loss: 2.8376 (2.7498)  weight_decay: 0.0500 (0.0500)  time: 0.6817  data: 0.0001  max mem: 57114
Epoch: [224]  [155/156]  eta: 0:00:00  lr: 0.001336  min_lr: 0.001336  loss: 2.8806 (2.7433)  weight_decay: 0.0500 (0.0500)  time: 0.6860  data: 0.0001  max mem: 57114
Epoch: [224] Total time: 0:01:52 (0.7230 s / it)
Averaged stats: lr: 0.001336  min_lr: 0.001336  loss: 2.8806 (2.7593)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5708 (0.5708)  acc1: 88.0208 (88.0208)  acc5: 99.4792 (99.4792)  time: 2.0372  data: 1.7815  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7840 (0.7277)  acc1: 84.8958 (83.5038)  acc5: 97.9167 (97.4425)  time: 0.5581  data: 0.3564  max mem: 57114
Test: Total time: 0:00:02 (0.5776 s / it)
* Acc@1 82.828 Acc@5 96.379 loss 0.809
Accuracy of the model on the 50000 test images: 82.8%
Max accuracy: 83.23%
Test:  [0/5]  eta: 0:00:11  loss: 0.5863 (0.5863)  acc1: 83.8542 (83.8542)  acc5: 97.3958 (97.3958)  time: 2.3182  data: 2.0748  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6940 (0.6731)  acc1: 81.2500 (79.2839)  acc5: 96.8750 (94.3734)  time: 0.6143  data: 0.4150  max mem: 57114
Test: Total time: 0:00:03 (0.6250 s / it)
* Acc@1 77.458 Acc@5 93.964 loss 0.812
Accuracy of the model EMA on 50000 test images: 77.5%
Max EMA accuracy: 77.46%
Epoch: [225]  [  0/156]  eta: 0:07:22  lr: 0.001335  min_lr: 0.001335  loss: 2.8954 (2.8954)  weight_decay: 0.0500 (0.0500)  time: 2.8342  data: 2.1751  max mem: 57114
Epoch: [225]  [ 10/156]  eta: 0:02:14  lr: 0.001333  min_lr: 0.001333  loss: 2.8113 (2.7125)  weight_decay: 0.0500 (0.0500)  time: 0.9186  data: 0.1980  max mem: 57114
Epoch: [225]  [ 20/156]  eta: 0:01:53  lr: 0.001331  min_lr: 0.001331  loss: 2.9085 (2.8530)  weight_decay: 0.0500 (0.0500)  time: 0.7329  data: 0.0004  max mem: 57114
Epoch: [225]  [ 30/156]  eta: 0:01:40  lr: 0.001329  min_lr: 0.001329  loss: 3.0886 (2.9043)  weight_decay: 0.0500 (0.0500)  time: 0.7302  data: 0.0004  max mem: 57114
Epoch: [225]  [ 40/156]  eta: 0:01:29  lr: 0.001327  min_lr: 0.001327  loss: 3.0302 (2.9106)  weight_decay: 0.0500 (0.0500)  time: 0.7126  data: 0.0004  max mem: 57114
Epoch: [225]  [ 50/156]  eta: 0:01:21  lr: 0.001325  min_lr: 0.001325  loss: 2.7650 (2.8657)  weight_decay: 0.0500 (0.0500)  time: 0.7135  data: 0.0004  max mem: 57114
Epoch: [225]  [ 60/156]  eta: 0:01:12  lr: 0.001323  min_lr: 0.001323  loss: 2.8073 (2.8524)  weight_decay: 0.0500 (0.0500)  time: 0.7210  data: 0.0003  max mem: 57114
Epoch: [225]  [ 70/156]  eta: 0:01:04  lr: 0.001320  min_lr: 0.001320  loss: 2.9514 (2.8577)  weight_decay: 0.0500 (0.0500)  time: 0.7180  data: 0.0004  max mem: 57114
Epoch: [225]  [ 80/156]  eta: 0:00:56  lr: 0.001318  min_lr: 0.001318  loss: 2.9780 (2.8660)  weight_decay: 0.0500 (0.0500)  time: 0.7191  data: 0.0004  max mem: 57114
Epoch: [225]  [ 90/156]  eta: 0:00:49  lr: 0.001316  min_lr: 0.001316  loss: 2.8846 (2.8768)  weight_decay: 0.0500 (0.0500)  time: 0.7133  data: 0.0004  max mem: 57114
Epoch: [225]  [100/156]  eta: 0:00:41  lr: 0.001314  min_lr: 0.001314  loss: 2.9014 (2.8779)  weight_decay: 0.0500 (0.0500)  time: 0.7101  data: 0.0004  max mem: 57114
Epoch: [225]  [110/156]  eta: 0:00:33  lr: 0.001312  min_lr: 0.001312  loss: 2.9366 (2.8766)  weight_decay: 0.0500 (0.0500)  time: 0.7086  data: 0.0004  max mem: 57114
Epoch: [225]  [120/156]  eta: 0:00:26  lr: 0.001310  min_lr: 0.001310  loss: 2.9515 (2.8648)  weight_decay: 0.0500 (0.0500)  time: 0.6967  data: 0.0004  max mem: 57114
Epoch: [225]  [130/156]  eta: 0:00:18  lr: 0.001308  min_lr: 0.001308  loss: 2.7908 (2.8645)  weight_decay: 0.0500 (0.0500)  time: 0.6924  data: 0.0010  max mem: 57114
Epoch: [225]  [140/156]  eta: 0:00:11  lr: 0.001306  min_lr: 0.001306  loss: 2.7908 (2.8537)  weight_decay: 0.0500 (0.0500)  time: 0.6918  data: 0.0008  max mem: 57114
Epoch: [225]  [150/156]  eta: 0:00:04  lr: 0.001303  min_lr: 0.001303  loss: 2.9919 (2.8686)  weight_decay: 0.0500 (0.0500)  time: 0.6843  data: 0.0001  max mem: 57114
Epoch: [225]  [155/156]  eta: 0:00:00  lr: 0.001302  min_lr: 0.001302  loss: 2.9838 (2.8677)  weight_decay: 0.0500 (0.0500)  time: 0.6820  data: 0.0001  max mem: 57114
Epoch: [225] Total time: 0:01:53 (0.7245 s / it)
Averaged stats: lr: 0.001302  min_lr: 0.001302  loss: 2.9838 (2.7523)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7345 (0.7345)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.0555  data: 1.7937  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8947 (0.8415)  acc1: 81.7708 (83.2481)  acc5: 97.9167 (97.3146)  time: 0.5618  data: 0.3588  max mem: 57114
Test: Total time: 0:00:02 (0.5854 s / it)
* Acc@1 83.134 Acc@5 96.537 loss 0.929
Accuracy of the model on the 50000 test images: 83.1%
Max accuracy: 83.23%
Test:  [0/5]  eta: 0:00:12  loss: 0.5866 (0.5866)  acc1: 83.8542 (83.8542)  acc5: 97.3958 (97.3958)  time: 2.4083  data: 2.1646  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6933 (0.6729)  acc1: 81.7708 (79.4118)  acc5: 96.8750 (94.3734)  time: 0.6324  data: 0.4331  max mem: 57114
Test: Total time: 0:00:03 (0.6448 s / it)
* Acc@1 77.470 Acc@5 93.958 loss 0.812
Accuracy of the model EMA on 50000 test images: 77.5%
Max EMA accuracy: 77.47%
Epoch: [226]  [  0/156]  eta: 0:06:29  lr: 0.001302  min_lr: 0.001302  loss: 3.0514 (3.0514)  weight_decay: 0.0500 (0.0500)  time: 2.4954  data: 1.8428  max mem: 57114
Epoch: [226]  [ 10/156]  eta: 0:02:09  lr: 0.001300  min_lr: 0.001300  loss: 2.8489 (2.7058)  weight_decay: 0.0500 (0.0500)  time: 0.8901  data: 0.1679  max mem: 57114
Epoch: [226]  [ 20/156]  eta: 0:01:49  lr: 0.001298  min_lr: 0.001298  loss: 2.7731 (2.6387)  weight_decay: 0.0500 (0.0500)  time: 0.7210  data: 0.0004  max mem: 57114
Epoch: [226]  [ 30/156]  eta: 0:01:38  lr: 0.001296  min_lr: 0.001296  loss: 2.7007 (2.6587)  weight_decay: 0.0500 (0.0500)  time: 0.7183  data: 0.0005  max mem: 57114
Epoch: [226]  [ 40/156]  eta: 0:01:29  lr: 0.001294  min_lr: 0.001294  loss: 2.6164 (2.6408)  weight_decay: 0.0500 (0.0500)  time: 0.7314  data: 0.0005  max mem: 57114
Epoch: [226]  [ 50/156]  eta: 0:01:20  lr: 0.001292  min_lr: 0.001292  loss: 2.6164 (2.6455)  weight_decay: 0.0500 (0.0500)  time: 0.7268  data: 0.0004  max mem: 57114
Epoch: [226]  [ 60/156]  eta: 0:01:12  lr: 0.001289  min_lr: 0.001289  loss: 2.7531 (2.6716)  weight_decay: 0.0500 (0.0500)  time: 0.7216  data: 0.0004  max mem: 57114
Epoch: [226]  [ 70/156]  eta: 0:01:04  lr: 0.001287  min_lr: 0.001287  loss: 2.7531 (2.6712)  weight_decay: 0.0500 (0.0500)  time: 0.7248  data: 0.0004  max mem: 57114
Epoch: [226]  [ 80/156]  eta: 0:00:56  lr: 0.001285  min_lr: 0.001285  loss: 2.8228 (2.6750)  weight_decay: 0.0500 (0.0500)  time: 0.7146  data: 0.0003  max mem: 57114
Epoch: [226]  [ 90/156]  eta: 0:00:48  lr: 0.001283  min_lr: 0.001283  loss: 2.7391 (2.6736)  weight_decay: 0.0500 (0.0500)  time: 0.7032  data: 0.0004  max mem: 57114
Epoch: [226]  [100/156]  eta: 0:00:41  lr: 0.001281  min_lr: 0.001281  loss: 2.6262 (2.6656)  weight_decay: 0.0500 (0.0500)  time: 0.6990  data: 0.0003  max mem: 57114
Epoch: [226]  [110/156]  eta: 0:00:33  lr: 0.001279  min_lr: 0.001279  loss: 2.7728 (2.6670)  weight_decay: 0.0500 (0.0500)  time: 0.7051  data: 0.0003  max mem: 57114
Epoch: [226]  [120/156]  eta: 0:00:26  lr: 0.001277  min_lr: 0.001277  loss: 2.7430 (2.6669)  weight_decay: 0.0500 (0.0500)  time: 0.7074  data: 0.0003  max mem: 57114
Epoch: [226]  [130/156]  eta: 0:00:18  lr: 0.001275  min_lr: 0.001275  loss: 2.7430 (2.6693)  weight_decay: 0.0500 (0.0500)  time: 0.6989  data: 0.0007  max mem: 57114
Epoch: [226]  [140/156]  eta: 0:00:11  lr: 0.001273  min_lr: 0.001273  loss: 2.4783 (2.6474)  weight_decay: 0.0500 (0.0500)  time: 0.6935  data: 0.0006  max mem: 57114
Epoch: [226]  [150/156]  eta: 0:00:04  lr: 0.001270  min_lr: 0.001270  loss: 2.4783 (2.6509)  weight_decay: 0.0500 (0.0500)  time: 0.6883  data: 0.0001  max mem: 57114
Epoch: [226]  [155/156]  eta: 0:00:00  lr: 0.001269  min_lr: 0.001269  loss: 2.7925 (2.6527)  weight_decay: 0.0500 (0.0500)  time: 0.6851  data: 0.0001  max mem: 57114
Epoch: [226] Total time: 0:01:52 (0.7236 s / it)
Averaged stats: lr: 0.001269  min_lr: 0.001269  loss: 2.7925 (2.7450)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6153 (0.6153)  acc1: 89.0625 (89.0625)  acc5: 99.4792 (99.4792)  time: 2.0558  data: 1.7994  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8323 (0.7271)  acc1: 82.2917 (82.6087)  acc5: 96.8750 (96.5473)  time: 0.5619  data: 0.3600  max mem: 57114
Test: Total time: 0:00:02 (0.5830 s / it)
* Acc@1 83.236 Acc@5 96.483 loss 0.789
Accuracy of the model on the 50000 test images: 83.2%
Max accuracy: 83.24%
Test:  [0/5]  eta: 0:00:10  loss: 0.5868 (0.5868)  acc1: 83.8542 (83.8542)  acc5: 97.3958 (97.3958)  time: 2.0700  data: 1.8263  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6929 (0.6728)  acc1: 81.7708 (79.4118)  acc5: 96.8750 (94.3734)  time: 0.5647  data: 0.3654  max mem: 57114
Test: Total time: 0:00:02 (0.5755 s / it)
* Acc@1 77.472 Acc@5 93.954 loss 0.812
Accuracy of the model EMA on 50000 test images: 77.5%
Max EMA accuracy: 77.47%
Epoch: [227]  [  0/156]  eta: 0:07:26  lr: 0.001269  min_lr: 0.001269  loss: 2.5881 (2.5881)  weight_decay: 0.0500 (0.0500)  time: 2.8606  data: 2.2140  max mem: 57114
Epoch: [227]  [ 10/156]  eta: 0:02:14  lr: 0.001267  min_lr: 0.001267  loss: 2.7772 (2.7492)  weight_decay: 0.0500 (0.0500)  time: 0.9199  data: 0.2015  max mem: 57114
Epoch: [227]  [ 20/156]  eta: 0:01:53  lr: 0.001265  min_lr: 0.001265  loss: 2.8620 (2.8197)  weight_decay: 0.0500 (0.0500)  time: 0.7299  data: 0.0003  max mem: 57114
Epoch: [227]  [ 30/156]  eta: 0:01:39  lr: 0.001263  min_lr: 0.001263  loss: 2.8128 (2.7727)  weight_decay: 0.0500 (0.0500)  time: 0.7226  data: 0.0003  max mem: 57114
Epoch: [227]  [ 40/156]  eta: 0:01:30  lr: 0.001261  min_lr: 0.001261  loss: 2.6927 (2.7295)  weight_decay: 0.0500 (0.0500)  time: 0.7287  data: 0.0003  max mem: 57114
Epoch: [227]  [ 50/156]  eta: 0:01:21  lr: 0.001259  min_lr: 0.001259  loss: 2.7533 (2.7220)  weight_decay: 0.0500 (0.0500)  time: 0.7219  data: 0.0004  max mem: 57114
Epoch: [227]  [ 60/156]  eta: 0:01:13  lr: 0.001257  min_lr: 0.001257  loss: 2.8414 (2.7330)  weight_decay: 0.0500 (0.0500)  time: 0.7239  data: 0.0004  max mem: 57114
Epoch: [227]  [ 70/156]  eta: 0:01:05  lr: 0.001255  min_lr: 0.001255  loss: 2.9050 (2.7672)  weight_decay: 0.0500 (0.0500)  time: 0.7432  data: 0.0004  max mem: 57114
Epoch: [227]  [ 80/156]  eta: 0:00:57  lr: 0.001252  min_lr: 0.001252  loss: 2.9535 (2.7538)  weight_decay: 0.0500 (0.0500)  time: 0.7359  data: 0.0004  max mem: 57114
Epoch: [227]  [ 90/156]  eta: 0:00:49  lr: 0.001250  min_lr: 0.001250  loss: 2.9103 (2.7530)  weight_decay: 0.0500 (0.0500)  time: 0.7224  data: 0.0004  max mem: 57114
Epoch: [227]  [100/156]  eta: 0:00:41  lr: 0.001248  min_lr: 0.001248  loss: 2.9317 (2.7512)  weight_decay: 0.0500 (0.0500)  time: 0.7087  data: 0.0004  max mem: 57114
Epoch: [227]  [110/156]  eta: 0:00:34  lr: 0.001246  min_lr: 0.001246  loss: 2.8120 (2.7461)  weight_decay: 0.0500 (0.0500)  time: 0.7031  data: 0.0004  max mem: 57114
Epoch: [227]  [120/156]  eta: 0:00:26  lr: 0.001244  min_lr: 0.001244  loss: 2.7173 (2.7274)  weight_decay: 0.0500 (0.0500)  time: 0.7026  data: 0.0004  max mem: 57114
Epoch: [227]  [130/156]  eta: 0:00:19  lr: 0.001242  min_lr: 0.001242  loss: 2.4229 (2.7171)  weight_decay: 0.0500 (0.0500)  time: 0.6980  data: 0.0008  max mem: 57114
Epoch: [227]  [140/156]  eta: 0:00:11  lr: 0.001240  min_lr: 0.001240  loss: 2.8172 (2.7160)  weight_decay: 0.0500 (0.0500)  time: 0.6911  data: 0.0007  max mem: 57114
Epoch: [227]  [150/156]  eta: 0:00:04  lr: 0.001238  min_lr: 0.001238  loss: 2.8604 (2.7243)  weight_decay: 0.0500 (0.0500)  time: 0.6870  data: 0.0001  max mem: 57114
Epoch: [227]  [155/156]  eta: 0:00:00  lr: 0.001237  min_lr: 0.001237  loss: 2.8295 (2.7315)  weight_decay: 0.0500 (0.0500)  time: 0.6870  data: 0.0001  max mem: 57114
Epoch: [227] Total time: 0:01:53 (0.7299 s / it)
Averaged stats: lr: 0.001237  min_lr: 0.001237  loss: 2.8295 (2.7354)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6415 (0.6415)  acc1: 90.1042 (90.1042)  acc5: 98.9583 (98.9583)  time: 2.0464  data: 1.7903  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7984 (0.7832)  acc1: 84.8958 (83.6317)  acc5: 97.3958 (96.9310)  time: 0.5600  data: 0.3581  max mem: 57114
Test: Total time: 0:00:02 (0.5802 s / it)
* Acc@1 83.168 Acc@5 96.533 loss 0.873
Accuracy of the model on the 50000 test images: 83.2%
Max accuracy: 83.24%
Test:  [0/5]  eta: 0:00:11  loss: 0.5871 (0.5871)  acc1: 83.8542 (83.8542)  acc5: 97.3958 (97.3958)  time: 2.2728  data: 2.0294  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6923 (0.6725)  acc1: 81.7708 (79.5396)  acc5: 96.8750 (94.3734)  time: 0.6052  data: 0.4060  max mem: 57114
Test: Total time: 0:00:03 (0.6161 s / it)
* Acc@1 77.478 Acc@5 93.946 loss 0.811
Accuracy of the model EMA on 50000 test images: 77.5%
Max EMA accuracy: 77.48%
Epoch: [228]  [  0/156]  eta: 0:08:21  lr: 0.001237  min_lr: 0.001237  loss: 2.4061 (2.4061)  weight_decay: 0.0500 (0.0500)  time: 3.2153  data: 2.5622  max mem: 57114
Epoch: [228]  [ 10/156]  eta: 0:02:20  lr: 0.001235  min_lr: 0.001235  loss: 2.9539 (2.7955)  weight_decay: 0.0500 (0.0500)  time: 0.9606  data: 0.2332  max mem: 57114
Epoch: [228]  [ 20/156]  eta: 0:01:52  lr: 0.001232  min_lr: 0.001232  loss: 3.0168 (2.7747)  weight_decay: 0.0500 (0.0500)  time: 0.7100  data: 0.0003  max mem: 57114
Epoch: [228]  [ 30/156]  eta: 0:01:40  lr: 0.001230  min_lr: 0.001230  loss: 2.9454 (2.7993)  weight_decay: 0.0500 (0.0500)  time: 0.7065  data: 0.0004  max mem: 57114
Epoch: [228]  [ 40/156]  eta: 0:01:30  lr: 0.001228  min_lr: 0.001228  loss: 3.0014 (2.8667)  weight_decay: 0.0500 (0.0500)  time: 0.7324  data: 0.0005  max mem: 57114
Epoch: [228]  [ 50/156]  eta: 0:01:21  lr: 0.001226  min_lr: 0.001226  loss: 2.9682 (2.8324)  weight_decay: 0.0500 (0.0500)  time: 0.7376  data: 0.0003  max mem: 57114
Epoch: [228]  [ 60/156]  eta: 0:01:13  lr: 0.001224  min_lr: 0.001224  loss: 2.9210 (2.8520)  weight_decay: 0.0500 (0.0500)  time: 0.7263  data: 0.0004  max mem: 57114
Epoch: [228]  [ 70/156]  eta: 0:01:05  lr: 0.001222  min_lr: 0.001222  loss: 2.8650 (2.8279)  weight_decay: 0.0500 (0.0500)  time: 0.7187  data: 0.0004  max mem: 57114
Epoch: [228]  [ 80/156]  eta: 0:00:57  lr: 0.001220  min_lr: 0.001220  loss: 2.4848 (2.7655)  weight_decay: 0.0500 (0.0500)  time: 0.7136  data: 0.0004  max mem: 57114
Epoch: [228]  [ 90/156]  eta: 0:00:49  lr: 0.001218  min_lr: 0.001218  loss: 2.7752 (2.7877)  weight_decay: 0.0500 (0.0500)  time: 0.7004  data: 0.0004  max mem: 57114
Epoch: [228]  [100/156]  eta: 0:00:41  lr: 0.001216  min_lr: 0.001216  loss: 2.7752 (2.7619)  weight_decay: 0.0500 (0.0500)  time: 0.7052  data: 0.0004  max mem: 57114
Epoch: [228]  [110/156]  eta: 0:00:33  lr: 0.001214  min_lr: 0.001214  loss: 2.5209 (2.7535)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0003  max mem: 57114
Epoch: [228]  [120/156]  eta: 0:00:26  lr: 0.001212  min_lr: 0.001212  loss: 2.8213 (2.7431)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0004  max mem: 57114
Epoch: [228]  [130/156]  eta: 0:00:19  lr: 0.001210  min_lr: 0.001210  loss: 2.8547 (2.7538)  weight_decay: 0.0500 (0.0500)  time: 0.6994  data: 0.0009  max mem: 57114
Epoch: [228]  [140/156]  eta: 0:00:11  lr: 0.001208  min_lr: 0.001208  loss: 2.8882 (2.7527)  weight_decay: 0.0500 (0.0500)  time: 0.6899  data: 0.0008  max mem: 57114
Epoch: [228]  [150/156]  eta: 0:00:04  lr: 0.001206  min_lr: 0.001206  loss: 2.5569 (2.7290)  weight_decay: 0.0500 (0.0500)  time: 0.6844  data: 0.0001  max mem: 57114
Epoch: [228]  [155/156]  eta: 0:00:00  lr: 0.001205  min_lr: 0.001205  loss: 2.7336 (2.7368)  weight_decay: 0.0500 (0.0500)  time: 0.6782  data: 0.0001  max mem: 57114
Epoch: [228] Total time: 0:01:53 (0.7270 s / it)
Averaged stats: lr: 0.001205  min_lr: 0.001205  loss: 2.7336 (2.7309)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5877 (0.5877)  acc1: 90.6250 (90.6250)  acc5: 98.9583 (98.9583)  time: 2.1104  data: 1.8522  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7884 (0.7434)  acc1: 83.3333 (83.3760)  acc5: 97.9167 (97.4425)  time: 0.5728  data: 0.3705  max mem: 57114
Test: Total time: 0:00:02 (0.5953 s / it)
* Acc@1 83.434 Acc@5 96.565 loss 0.824
Accuracy of the model on the 50000 test images: 83.4%
Max accuracy: 83.43%
Test:  [0/5]  eta: 0:00:09  loss: 0.5874 (0.5874)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 1.9865  data: 1.7430  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6918 (0.6725)  acc1: 81.7708 (79.6675)  acc5: 96.8750 (94.3734)  time: 0.5479  data: 0.3487  max mem: 57114
Test: Total time: 0:00:02 (0.5598 s / it)
* Acc@1 77.490 Acc@5 93.938 loss 0.811
Accuracy of the model EMA on 50000 test images: 77.5%
Max EMA accuracy: 77.49%
Epoch: [229]  [  0/156]  eta: 0:07:19  lr: 0.001204  min_lr: 0.001204  loss: 2.3187 (2.3187)  weight_decay: 0.0500 (0.0500)  time: 2.8182  data: 2.0471  max mem: 57114
Epoch: [229]  [ 10/156]  eta: 0:02:16  lr: 0.001202  min_lr: 0.001202  loss: 2.8693 (2.6634)  weight_decay: 0.0500 (0.0500)  time: 0.9367  data: 0.1864  max mem: 57114
Epoch: [229]  [ 20/156]  eta: 0:01:51  lr: 0.001200  min_lr: 0.001200  loss: 2.8612 (2.6681)  weight_decay: 0.0500 (0.0500)  time: 0.7203  data: 0.0004  max mem: 57114
Epoch: [229]  [ 30/156]  eta: 0:01:39  lr: 0.001198  min_lr: 0.001198  loss: 2.8835 (2.7570)  weight_decay: 0.0500 (0.0500)  time: 0.7070  data: 0.0003  max mem: 57114
Epoch: [229]  [ 40/156]  eta: 0:01:30  lr: 0.001196  min_lr: 0.001196  loss: 2.8746 (2.6925)  weight_decay: 0.0500 (0.0500)  time: 0.7446  data: 0.0004  max mem: 57114
Epoch: [229]  [ 50/156]  eta: 0:01:21  lr: 0.001194  min_lr: 0.001194  loss: 2.7178 (2.6852)  weight_decay: 0.0500 (0.0500)  time: 0.7496  data: 0.0004  max mem: 57114
Epoch: [229]  [ 60/156]  eta: 0:01:13  lr: 0.001192  min_lr: 0.001192  loss: 2.6739 (2.6834)  weight_decay: 0.0500 (0.0500)  time: 0.7269  data: 0.0004  max mem: 57114
Epoch: [229]  [ 70/156]  eta: 0:01:05  lr: 0.001190  min_lr: 0.001190  loss: 2.6663 (2.6825)  weight_decay: 0.0500 (0.0500)  time: 0.7238  data: 0.0004  max mem: 57114
Epoch: [229]  [ 80/156]  eta: 0:00:57  lr: 0.001188  min_lr: 0.001188  loss: 2.8275 (2.7060)  weight_decay: 0.0500 (0.0500)  time: 0.7163  data: 0.0004  max mem: 57114
Epoch: [229]  [ 90/156]  eta: 0:00:49  lr: 0.001186  min_lr: 0.001186  loss: 2.7159 (2.6816)  weight_decay: 0.0500 (0.0500)  time: 0.7091  data: 0.0004  max mem: 57114
Epoch: [229]  [100/156]  eta: 0:00:41  lr: 0.001184  min_lr: 0.001184  loss: 2.6210 (2.6976)  weight_decay: 0.0500 (0.0500)  time: 0.7116  data: 0.0005  max mem: 57114
Epoch: [229]  [110/156]  eta: 0:00:34  lr: 0.001182  min_lr: 0.001182  loss: 2.8553 (2.7011)  weight_decay: 0.0500 (0.0500)  time: 0.7103  data: 0.0004  max mem: 57114
Epoch: [229]  [120/156]  eta: 0:00:26  lr: 0.001180  min_lr: 0.001180  loss: 2.8205 (2.7146)  weight_decay: 0.0500 (0.0500)  time: 0.7121  data: 0.0004  max mem: 57114
Epoch: [229]  [130/156]  eta: 0:00:19  lr: 0.001178  min_lr: 0.001178  loss: 2.7388 (2.7008)  weight_decay: 0.0500 (0.0500)  time: 0.7025  data: 0.0008  max mem: 57114
Epoch: [229]  [140/156]  eta: 0:00:11  lr: 0.001176  min_lr: 0.001176  loss: 2.7500 (2.7001)  weight_decay: 0.0500 (0.0500)  time: 0.6802  data: 0.0007  max mem: 57114
Epoch: [229]  [150/156]  eta: 0:00:04  lr: 0.001174  min_lr: 0.001174  loss: 2.7500 (2.6953)  weight_decay: 0.0500 (0.0500)  time: 0.6849  data: 0.0001  max mem: 57114
Epoch: [229]  [155/156]  eta: 0:00:00  lr: 0.001173  min_lr: 0.001173  loss: 2.7640 (2.6939)  weight_decay: 0.0500 (0.0500)  time: 0.6866  data: 0.0001  max mem: 57114
Epoch: [229] Total time: 0:01:53 (0.7296 s / it)
Averaged stats: lr: 0.001173  min_lr: 0.001173  loss: 2.7640 (2.7270)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6648 (0.6648)  acc1: 89.0625 (89.0625)  acc5: 98.9583 (98.9583)  time: 2.0688  data: 1.8128  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8269 (0.7945)  acc1: 83.3333 (83.6317)  acc5: 97.9167 (97.0588)  time: 0.5644  data: 0.3626  max mem: 57114
Test: Total time: 0:00:02 (0.5887 s / it)
* Acc@1 83.398 Acc@5 96.535 loss 0.871
Accuracy of the model on the 50000 test images: 83.4%
Max accuracy: 83.43%
Test:  [0/5]  eta: 0:00:12  loss: 0.5878 (0.5878)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.4001  data: 2.1566  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6915 (0.6726)  acc1: 81.7708 (79.6675)  acc5: 96.8750 (94.2455)  time: 0.6307  data: 0.4314  max mem: 57114
Test: Total time: 0:00:03 (0.6419 s / it)
* Acc@1 77.504 Acc@5 93.944 loss 0.811
Accuracy of the model EMA on 50000 test images: 77.5%
Max EMA accuracy: 77.50%
Epoch: [230]  [  0/156]  eta: 0:07:55  lr: 0.001172  min_lr: 0.001172  loss: 2.0727 (2.0727)  weight_decay: 0.0500 (0.0500)  time: 3.0483  data: 2.3952  max mem: 57114
Epoch: [230]  [ 10/156]  eta: 0:02:19  lr: 0.001170  min_lr: 0.001170  loss: 2.3149 (2.4736)  weight_decay: 0.0500 (0.0500)  time: 0.9589  data: 0.2181  max mem: 57114
Epoch: [230]  [ 20/156]  eta: 0:01:54  lr: 0.001168  min_lr: 0.001168  loss: 2.7766 (2.5862)  weight_decay: 0.0500 (0.0500)  time: 0.7341  data: 0.0004  max mem: 57114
Epoch: [230]  [ 30/156]  eta: 0:01:42  lr: 0.001166  min_lr: 0.001166  loss: 2.7795 (2.6346)  weight_decay: 0.0500 (0.0500)  time: 0.7290  data: 0.0004  max mem: 57114
Epoch: [230]  [ 40/156]  eta: 0:01:31  lr: 0.001164  min_lr: 0.001164  loss: 2.7795 (2.6464)  weight_decay: 0.0500 (0.0500)  time: 0.7295  data: 0.0004  max mem: 57114
Epoch: [230]  [ 50/156]  eta: 0:01:22  lr: 0.001162  min_lr: 0.001162  loss: 2.5921 (2.6178)  weight_decay: 0.0500 (0.0500)  time: 0.7177  data: 0.0004  max mem: 57114
Epoch: [230]  [ 60/156]  eta: 0:01:13  lr: 0.001160  min_lr: 0.001160  loss: 2.7835 (2.6629)  weight_decay: 0.0500 (0.0500)  time: 0.7149  data: 0.0003  max mem: 57114
Epoch: [230]  [ 70/156]  eta: 0:01:05  lr: 0.001158  min_lr: 0.001158  loss: 2.8188 (2.6664)  weight_decay: 0.0500 (0.0500)  time: 0.7147  data: 0.0004  max mem: 57114
Epoch: [230]  [ 80/156]  eta: 0:00:57  lr: 0.001156  min_lr: 0.001156  loss: 2.8600 (2.6948)  weight_decay: 0.0500 (0.0500)  time: 0.7184  data: 0.0004  max mem: 57114
Epoch: [230]  [ 90/156]  eta: 0:00:49  lr: 0.001154  min_lr: 0.001154  loss: 2.8600 (2.6768)  weight_decay: 0.0500 (0.0500)  time: 0.7155  data: 0.0004  max mem: 57114
Epoch: [230]  [100/156]  eta: 0:00:41  lr: 0.001152  min_lr: 0.001152  loss: 2.2587 (2.6483)  weight_decay: 0.0500 (0.0500)  time: 0.7168  data: 0.0004  max mem: 57114
Epoch: [230]  [110/156]  eta: 0:00:34  lr: 0.001150  min_lr: 0.001150  loss: 2.5822 (2.6714)  weight_decay: 0.0500 (0.0500)  time: 0.7163  data: 0.0004  max mem: 57114
Epoch: [230]  [120/156]  eta: 0:00:26  lr: 0.001148  min_lr: 0.001148  loss: 2.8271 (2.6564)  weight_decay: 0.0500 (0.0500)  time: 0.7089  data: 0.0003  max mem: 57114
Epoch: [230]  [130/156]  eta: 0:00:19  lr: 0.001146  min_lr: 0.001146  loss: 2.5110 (2.6569)  weight_decay: 0.0500 (0.0500)  time: 0.7009  data: 0.0004  max mem: 57114
Epoch: [230]  [140/156]  eta: 0:00:11  lr: 0.001144  min_lr: 0.001144  loss: 2.7200 (2.6523)  weight_decay: 0.0500 (0.0500)  time: 0.6856  data: 0.0003  max mem: 57114
Epoch: [230]  [150/156]  eta: 0:00:04  lr: 0.001142  min_lr: 0.001142  loss: 2.7369 (2.6686)  weight_decay: 0.0500 (0.0500)  time: 0.6782  data: 0.0001  max mem: 57114
Epoch: [230]  [155/156]  eta: 0:00:00  lr: 0.001141  min_lr: 0.001141  loss: 2.8375 (2.6777)  weight_decay: 0.0500 (0.0500)  time: 0.6841  data: 0.0001  max mem: 57114
Epoch: [230] Total time: 0:01:53 (0.7295 s / it)
Averaged stats: lr: 0.001141  min_lr: 0.001141  loss: 2.8375 (2.7189)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6953 (0.6953)  acc1: 92.1875 (92.1875)  acc5: 98.9583 (98.9583)  time: 2.0208  data: 1.7647  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8849 (0.8505)  acc1: 82.2917 (84.0153)  acc5: 97.9167 (97.3146)  time: 0.5549  data: 0.3530  max mem: 57114
Test: Total time: 0:00:02 (0.5770 s / it)
* Acc@1 83.542 Acc@5 96.535 loss 0.931
Accuracy of the model on the 50000 test images: 83.5%
Max accuracy: 83.54%
Test:  [0/5]  eta: 0:00:10  loss: 0.5882 (0.5882)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.0368  data: 1.7931  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6911 (0.6728)  acc1: 81.7708 (79.6675)  acc5: 96.8750 (94.2455)  time: 0.5581  data: 0.3587  max mem: 57114
Test: Total time: 0:00:02 (0.5679 s / it)
* Acc@1 77.512 Acc@5 93.962 loss 0.811
Accuracy of the model EMA on 50000 test images: 77.5%
Max EMA accuracy: 77.51%
Epoch: [231]  [  0/156]  eta: 0:07:10  lr: 0.001141  min_lr: 0.001141  loss: 1.9720 (1.9720)  weight_decay: 0.0500 (0.0500)  time: 2.7624  data: 1.9905  max mem: 57114
Epoch: [231]  [ 10/156]  eta: 0:02:13  lr: 0.001139  min_lr: 0.001139  loss: 2.6790 (2.7157)  weight_decay: 0.0500 (0.0500)  time: 0.9149  data: 0.1813  max mem: 57114
Epoch: [231]  [ 20/156]  eta: 0:01:54  lr: 0.001137  min_lr: 0.001137  loss: 2.6790 (2.6868)  weight_decay: 0.0500 (0.0500)  time: 0.7423  data: 0.0004  max mem: 57114
Epoch: [231]  [ 30/156]  eta: 0:01:42  lr: 0.001135  min_lr: 0.001135  loss: 2.7280 (2.7249)  weight_decay: 0.0500 (0.0500)  time: 0.7533  data: 0.0004  max mem: 57114
Epoch: [231]  [ 40/156]  eta: 0:01:32  lr: 0.001133  min_lr: 0.001133  loss: 2.7280 (2.7016)  weight_decay: 0.0500 (0.0500)  time: 0.7471  data: 0.0004  max mem: 57114
Epoch: [231]  [ 50/156]  eta: 0:01:22  lr: 0.001131  min_lr: 0.001131  loss: 2.6007 (2.6756)  weight_decay: 0.0500 (0.0500)  time: 0.7282  data: 0.0004  max mem: 57114
Epoch: [231]  [ 60/156]  eta: 0:01:13  lr: 0.001129  min_lr: 0.001129  loss: 2.6512 (2.6906)  weight_decay: 0.0500 (0.0500)  time: 0.7175  data: 0.0004  max mem: 57114
Epoch: [231]  [ 70/156]  eta: 0:01:05  lr: 0.001127  min_lr: 0.001127  loss: 2.6402 (2.6623)  weight_decay: 0.0500 (0.0500)  time: 0.7168  data: 0.0004  max mem: 57114
Epoch: [231]  [ 80/156]  eta: 0:00:57  lr: 0.001125  min_lr: 0.001125  loss: 2.5838 (2.6602)  weight_decay: 0.0500 (0.0500)  time: 0.7122  data: 0.0003  max mem: 57114
Epoch: [231]  [ 90/156]  eta: 0:00:49  lr: 0.001123  min_lr: 0.001123  loss: 2.7115 (2.6649)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0004  max mem: 57114
Epoch: [231]  [100/156]  eta: 0:00:41  lr: 0.001121  min_lr: 0.001121  loss: 2.7751 (2.6562)  weight_decay: 0.0500 (0.0500)  time: 0.7118  data: 0.0004  max mem: 57114
Epoch: [231]  [110/156]  eta: 0:00:34  lr: 0.001119  min_lr: 0.001119  loss: 2.7658 (2.6609)  weight_decay: 0.0500 (0.0500)  time: 0.7108  data: 0.0004  max mem: 57114
Epoch: [231]  [120/156]  eta: 0:00:26  lr: 0.001117  min_lr: 0.001117  loss: 2.8429 (2.6775)  weight_decay: 0.0500 (0.0500)  time: 0.6991  data: 0.0004  max mem: 57114
Epoch: [231]  [130/156]  eta: 0:00:19  lr: 0.001115  min_lr: 0.001115  loss: 2.8429 (2.6892)  weight_decay: 0.0500 (0.0500)  time: 0.6898  data: 0.0008  max mem: 57114
Epoch: [231]  [140/156]  eta: 0:00:11  lr: 0.001113  min_lr: 0.001113  loss: 2.8105 (2.6900)  weight_decay: 0.0500 (0.0500)  time: 0.6798  data: 0.0007  max mem: 57114
Epoch: [231]  [150/156]  eta: 0:00:04  lr: 0.001111  min_lr: 0.001111  loss: 2.6972 (2.6862)  weight_decay: 0.0500 (0.0500)  time: 0.6794  data: 0.0001  max mem: 57114
Epoch: [231]  [155/156]  eta: 0:00:00  lr: 0.001110  min_lr: 0.001110  loss: 2.7322 (2.6935)  weight_decay: 0.0500 (0.0500)  time: 0.6804  data: 0.0001  max mem: 57114
Epoch: [231] Total time: 0:01:53 (0.7278 s / it)
Averaged stats: lr: 0.001110  min_lr: 0.001110  loss: 2.7322 (2.7261)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6132 (0.6132)  acc1: 89.5833 (89.5833)  acc5: 98.9583 (98.9583)  time: 2.0770  data: 1.8208  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7791 (0.7553)  acc1: 84.8958 (84.2711)  acc5: 97.9167 (97.6982)  time: 0.5661  data: 0.3642  max mem: 57114
Test: Total time: 0:00:02 (0.5896 s / it)
* Acc@1 83.410 Acc@5 96.559 loss 0.841
Accuracy of the model on the 50000 test images: 83.4%
Max accuracy: 83.54%
Test:  [0/5]  eta: 0:00:11  loss: 0.5885 (0.5885)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.2478  data: 2.0042  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6907 (0.6729)  acc1: 81.7708 (79.6675)  acc5: 96.8750 (94.2455)  time: 0.6002  data: 0.4009  max mem: 57114
Test: Total time: 0:00:03 (0.6121 s / it)
* Acc@1 77.516 Acc@5 93.954 loss 0.812
Accuracy of the model EMA on 50000 test images: 77.5%
Max EMA accuracy: 77.52%
Epoch: [232]  [  0/156]  eta: 0:07:07  lr: 0.001110  min_lr: 0.001110  loss: 2.7848 (2.7848)  weight_decay: 0.0500 (0.0500)  time: 2.7427  data: 2.0883  max mem: 57114
Epoch: [232]  [ 10/156]  eta: 0:02:15  lr: 0.001108  min_lr: 0.001108  loss: 2.6979 (2.5724)  weight_decay: 0.0500 (0.0500)  time: 0.9280  data: 0.1901  max mem: 57114
Epoch: [232]  [ 20/156]  eta: 0:01:51  lr: 0.001106  min_lr: 0.001106  loss: 2.6979 (2.6688)  weight_decay: 0.0500 (0.0500)  time: 0.7259  data: 0.0002  max mem: 57114
Epoch: [232]  [ 30/156]  eta: 0:01:39  lr: 0.001104  min_lr: 0.001104  loss: 2.7527 (2.6980)  weight_decay: 0.0500 (0.0500)  time: 0.7197  data: 0.0003  max mem: 57114
Epoch: [232]  [ 40/156]  eta: 0:01:29  lr: 0.001102  min_lr: 0.001102  loss: 2.7527 (2.6874)  weight_decay: 0.0500 (0.0500)  time: 0.7262  data: 0.0003  max mem: 57114
Epoch: [232]  [ 50/156]  eta: 0:01:20  lr: 0.001100  min_lr: 0.001100  loss: 2.8026 (2.7023)  weight_decay: 0.0500 (0.0500)  time: 0.7179  data: 0.0004  max mem: 57114
Epoch: [232]  [ 60/156]  eta: 0:01:12  lr: 0.001098  min_lr: 0.001098  loss: 2.7688 (2.7096)  weight_decay: 0.0500 (0.0500)  time: 0.7273  data: 0.0004  max mem: 57114
Epoch: [232]  [ 70/156]  eta: 0:01:04  lr: 0.001096  min_lr: 0.001096  loss: 2.8857 (2.7321)  weight_decay: 0.0500 (0.0500)  time: 0.7322  data: 0.0003  max mem: 57114
Epoch: [232]  [ 80/156]  eta: 0:00:57  lr: 0.001094  min_lr: 0.001094  loss: 2.9388 (2.7560)  weight_decay: 0.0500 (0.0500)  time: 0.7230  data: 0.0003  max mem: 57114
Epoch: [232]  [ 90/156]  eta: 0:00:49  lr: 0.001092  min_lr: 0.001092  loss: 3.0140 (2.7661)  weight_decay: 0.0500 (0.0500)  time: 0.7129  data: 0.0004  max mem: 57114
Epoch: [232]  [100/156]  eta: 0:00:41  lr: 0.001090  min_lr: 0.001090  loss: 2.7291 (2.7606)  weight_decay: 0.0500 (0.0500)  time: 0.7018  data: 0.0004  max mem: 57114
Epoch: [232]  [110/156]  eta: 0:00:33  lr: 0.001088  min_lr: 0.001088  loss: 2.8275 (2.7569)  weight_decay: 0.0500 (0.0500)  time: 0.6990  data: 0.0004  max mem: 57114
Epoch: [232]  [120/156]  eta: 0:00:26  lr: 0.001086  min_lr: 0.001086  loss: 2.5624 (2.7381)  weight_decay: 0.0500 (0.0500)  time: 0.7064  data: 0.0003  max mem: 57114
Epoch: [232]  [130/156]  eta: 0:00:19  lr: 0.001084  min_lr: 0.001084  loss: 2.3740 (2.7228)  weight_decay: 0.0500 (0.0500)  time: 0.7055  data: 0.0004  max mem: 57114
Epoch: [232]  [140/156]  eta: 0:00:11  lr: 0.001082  min_lr: 0.001082  loss: 2.5278 (2.7116)  weight_decay: 0.0500 (0.0500)  time: 0.6920  data: 0.0003  max mem: 57114
Epoch: [232]  [150/156]  eta: 0:00:04  lr: 0.001080  min_lr: 0.001080  loss: 2.4782 (2.6949)  weight_decay: 0.0500 (0.0500)  time: 0.6841  data: 0.0001  max mem: 57114
Epoch: [232]  [155/156]  eta: 0:00:00  lr: 0.001079  min_lr: 0.001079  loss: 2.4766 (2.6924)  weight_decay: 0.0500 (0.0500)  time: 0.6837  data: 0.0001  max mem: 57114
Epoch: [232] Total time: 0:01:53 (0.7268 s / it)
Averaged stats: lr: 0.001079  min_lr: 0.001079  loss: 2.4766 (2.7124)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5006 (0.5006)  acc1: 91.1458 (91.1458)  acc5: 98.9583 (98.9583)  time: 2.0837  data: 1.8278  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6938 (0.6524)  acc1: 84.3750 (84.1432)  acc5: 97.9167 (97.5703)  time: 0.5675  data: 0.3656  max mem: 57114
Test: Total time: 0:00:02 (0.5901 s / it)
* Acc@1 83.630 Acc@5 96.701 loss 0.744
Accuracy of the model on the 50000 test images: 83.6%
Max accuracy: 83.63%
Test:  [0/5]  eta: 0:00:10  loss: 0.5888 (0.5888)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.0852  data: 1.8416  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6901 (0.6730)  acc1: 81.7708 (79.6675)  acc5: 96.8750 (94.2455)  time: 0.5677  data: 0.3684  max mem: 57114
Test: Total time: 0:00:02 (0.5815 s / it)
* Acc@1 77.524 Acc@5 93.944 loss 0.812
Accuracy of the model EMA on 50000 test images: 77.5%
Max EMA accuracy: 77.52%
Epoch: [233]  [  0/156]  eta: 0:07:10  lr: 0.001079  min_lr: 0.001079  loss: 3.1283 (3.1283)  weight_decay: 0.0500 (0.0500)  time: 2.7607  data: 2.1086  max mem: 57114
Epoch: [233]  [ 10/156]  eta: 0:02:11  lr: 0.001077  min_lr: 0.001077  loss: 2.9530 (2.7918)  weight_decay: 0.0500 (0.0500)  time: 0.9037  data: 0.1920  max mem: 57114
Epoch: [233]  [ 20/156]  eta: 0:01:50  lr: 0.001075  min_lr: 0.001075  loss: 2.8865 (2.7796)  weight_decay: 0.0500 (0.0500)  time: 0.7171  data: 0.0004  max mem: 57114
Epoch: [233]  [ 30/156]  eta: 0:01:38  lr: 0.001073  min_lr: 0.001073  loss: 2.8865 (2.7604)  weight_decay: 0.0500 (0.0500)  time: 0.7179  data: 0.0003  max mem: 57114
Epoch: [233]  [ 40/156]  eta: 0:01:29  lr: 0.001071  min_lr: 0.001071  loss: 2.7755 (2.7089)  weight_decay: 0.0500 (0.0500)  time: 0.7277  data: 0.0004  max mem: 57114
Epoch: [233]  [ 50/156]  eta: 0:01:21  lr: 0.001069  min_lr: 0.001069  loss: 2.7313 (2.7290)  weight_decay: 0.0500 (0.0500)  time: 0.7340  data: 0.0004  max mem: 57114
Epoch: [233]  [ 60/156]  eta: 0:01:13  lr: 0.001067  min_lr: 0.001067  loss: 2.9206 (2.7380)  weight_decay: 0.0500 (0.0500)  time: 0.7478  data: 0.0004  max mem: 57114
Epoch: [233]  [ 70/156]  eta: 0:01:05  lr: 0.001065  min_lr: 0.001065  loss: 2.9206 (2.7008)  weight_decay: 0.0500 (0.0500)  time: 0.7570  data: 0.0004  max mem: 57114
Epoch: [233]  [ 80/156]  eta: 0:00:57  lr: 0.001063  min_lr: 0.001063  loss: 2.6315 (2.6958)  weight_decay: 0.0500 (0.0500)  time: 0.7239  data: 0.0004  max mem: 57114
Epoch: [233]  [ 90/156]  eta: 0:00:49  lr: 0.001061  min_lr: 0.001061  loss: 2.9097 (2.7264)  weight_decay: 0.0500 (0.0500)  time: 0.7012  data: 0.0004  max mem: 57114
Epoch: [233]  [100/156]  eta: 0:00:41  lr: 0.001059  min_lr: 0.001059  loss: 2.9194 (2.7327)  weight_decay: 0.0500 (0.0500)  time: 0.7081  data: 0.0004  max mem: 57114
Epoch: [233]  [110/156]  eta: 0:00:34  lr: 0.001057  min_lr: 0.001057  loss: 2.8128 (2.7249)  weight_decay: 0.0500 (0.0500)  time: 0.7107  data: 0.0004  max mem: 57114
Epoch: [233]  [120/156]  eta: 0:00:26  lr: 0.001055  min_lr: 0.001055  loss: 2.5488 (2.7035)  weight_decay: 0.0500 (0.0500)  time: 0.7124  data: 0.0004  max mem: 57114
Epoch: [233]  [130/156]  eta: 0:00:19  lr: 0.001053  min_lr: 0.001053  loss: 2.4893 (2.6886)  weight_decay: 0.0500 (0.0500)  time: 0.7017  data: 0.0009  max mem: 57114
Epoch: [233]  [140/156]  eta: 0:00:11  lr: 0.001052  min_lr: 0.001052  loss: 2.7769 (2.6969)  weight_decay: 0.0500 (0.0500)  time: 0.6883  data: 0.0007  max mem: 57114
Epoch: [233]  [150/156]  eta: 0:00:04  lr: 0.001050  min_lr: 0.001050  loss: 2.8902 (2.7018)  weight_decay: 0.0500 (0.0500)  time: 0.6870  data: 0.0001  max mem: 57114
Epoch: [233]  [155/156]  eta: 0:00:00  lr: 0.001049  min_lr: 0.001049  loss: 2.7144 (2.6955)  weight_decay: 0.0500 (0.0500)  time: 0.6828  data: 0.0001  max mem: 57114
Epoch: [233] Total time: 0:01:53 (0.7297 s / it)
Averaged stats: lr: 0.001049  min_lr: 0.001049  loss: 2.7144 (2.7089)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5757 (0.5757)  acc1: 90.1042 (90.1042)  acc5: 98.9583 (98.9583)  time: 2.0601  data: 1.8044  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7382 (0.7088)  acc1: 83.8542 (84.3990)  acc5: 98.4375 (97.6982)  time: 0.5628  data: 0.3609  max mem: 57114
Test: Total time: 0:00:02 (0.5835 s / it)
* Acc@1 83.558 Acc@5 96.659 loss 0.800
Accuracy of the model on the 50000 test images: 83.6%
Max accuracy: 83.63%
Test:  [0/5]  eta: 0:00:11  loss: 0.5893 (0.5893)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.2964  data: 2.0530  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6898 (0.6733)  acc1: 81.2500 (79.5396)  acc5: 96.3542 (94.2455)  time: 0.6099  data: 0.4107  max mem: 57114
Test: Total time: 0:00:03 (0.6195 s / it)
* Acc@1 77.520 Acc@5 93.934 loss 0.812
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [234]  [  0/156]  eta: 0:12:47  lr: 0.001048  min_lr: 0.001048  loss: 2.9810 (2.9810)  weight_decay: 0.0500 (0.0500)  time: 4.9178  data: 3.8859  max mem: 57114
Epoch: [234]  [ 10/156]  eta: 0:02:42  lr: 0.001046  min_lr: 0.001046  loss: 2.8095 (2.6829)  weight_decay: 0.0500 (0.0500)  time: 1.1156  data: 0.3535  max mem: 57114
Epoch: [234]  [ 20/156]  eta: 0:02:07  lr: 0.001045  min_lr: 0.001045  loss: 2.8095 (2.7556)  weight_decay: 0.0500 (0.0500)  time: 0.7406  data: 0.0003  max mem: 57114
Epoch: [234]  [ 30/156]  eta: 0:01:49  lr: 0.001043  min_lr: 0.001043  loss: 2.7805 (2.6971)  weight_decay: 0.0500 (0.0500)  time: 0.7328  data: 0.0003  max mem: 57114
Epoch: [234]  [ 40/156]  eta: 0:01:36  lr: 0.001041  min_lr: 0.001041  loss: 2.7622 (2.7169)  weight_decay: 0.0500 (0.0500)  time: 0.7213  data: 0.0003  max mem: 57114
Epoch: [234]  [ 50/156]  eta: 0:01:26  lr: 0.001039  min_lr: 0.001039  loss: 2.6819 (2.7063)  weight_decay: 0.0500 (0.0500)  time: 0.7333  data: 0.0004  max mem: 57114
Epoch: [234]  [ 60/156]  eta: 0:01:16  lr: 0.001037  min_lr: 0.001037  loss: 2.6419 (2.6786)  weight_decay: 0.0500 (0.0500)  time: 0.7335  data: 0.0004  max mem: 57114
Epoch: [234]  [ 70/156]  eta: 0:01:07  lr: 0.001035  min_lr: 0.001035  loss: 2.7928 (2.7033)  weight_decay: 0.0500 (0.0500)  time: 0.7168  data: 0.0003  max mem: 57114
Epoch: [234]  [ 80/156]  eta: 0:00:58  lr: 0.001033  min_lr: 0.001033  loss: 2.8357 (2.6891)  weight_decay: 0.0500 (0.0500)  time: 0.6996  data: 0.0003  max mem: 57114
Epoch: [234]  [ 90/156]  eta: 0:00:50  lr: 0.001031  min_lr: 0.001031  loss: 2.9105 (2.7078)  weight_decay: 0.0500 (0.0500)  time: 0.6999  data: 0.0004  max mem: 57114
Epoch: [234]  [100/156]  eta: 0:00:42  lr: 0.001029  min_lr: 0.001029  loss: 2.9750 (2.7264)  weight_decay: 0.0500 (0.0500)  time: 0.7137  data: 0.0004  max mem: 57114
Epoch: [234]  [110/156]  eta: 0:00:34  lr: 0.001027  min_lr: 0.001027  loss: 2.8983 (2.7267)  weight_decay: 0.0500 (0.0500)  time: 0.7082  data: 0.0004  max mem: 57114
Epoch: [234]  [120/156]  eta: 0:00:27  lr: 0.001025  min_lr: 0.001025  loss: 2.7532 (2.7157)  weight_decay: 0.0500 (0.0500)  time: 0.7070  data: 0.0003  max mem: 57114
Epoch: [234]  [130/156]  eta: 0:00:19  lr: 0.001023  min_lr: 0.001023  loss: 2.5581 (2.7047)  weight_decay: 0.0500 (0.0500)  time: 0.7046  data: 0.0008  max mem: 57114
Epoch: [234]  [140/156]  eta: 0:00:11  lr: 0.001021  min_lr: 0.001021  loss: 2.7478 (2.7040)  weight_decay: 0.0500 (0.0500)  time: 0.6898  data: 0.0007  max mem: 57114
Epoch: [234]  [150/156]  eta: 0:00:04  lr: 0.001019  min_lr: 0.001019  loss: 2.8099 (2.7073)  weight_decay: 0.0500 (0.0500)  time: 0.6795  data: 0.0001  max mem: 57114
Epoch: [234]  [155/156]  eta: 0:00:00  lr: 0.001018  min_lr: 0.001018  loss: 2.8814 (2.7183)  weight_decay: 0.0500 (0.0500)  time: 0.6820  data: 0.0001  max mem: 57114
Epoch: [234] Total time: 0:01:55 (0.7408 s / it)
Averaged stats: lr: 0.001018  min_lr: 0.001018  loss: 2.8814 (2.7028)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.7171 (0.7171)  acc1: 91.1458 (91.1458)  acc5: 99.4792 (99.4792)  time: 2.0317  data: 1.7757  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8750 (0.8755)  acc1: 84.3750 (84.1432)  acc5: 97.9167 (97.0588)  time: 0.5570  data: 0.3552  max mem: 57114
Test: Total time: 0:00:02 (0.5770 s / it)
* Acc@1 83.642 Acc@5 96.727 loss 0.953
Accuracy of the model on the 50000 test images: 83.6%
Max accuracy: 83.64%
Test:  [0/5]  eta: 0:00:10  loss: 0.5899 (0.5899)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.0158  data: 1.7722  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6897 (0.6738)  acc1: 81.2500 (79.5396)  acc5: 96.3542 (94.2455)  time: 0.5538  data: 0.3545  max mem: 57114
Test: Total time: 0:00:02 (0.5646 s / it)
* Acc@1 77.508 Acc@5 93.926 loss 0.813
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [235]  [  0/156]  eta: 0:13:53  lr: 0.001018  min_lr: 0.001018  loss: 3.0371 (3.0371)  weight_decay: 0.0500 (0.0500)  time: 5.3441  data: 4.0995  max mem: 57114
Epoch: [235]  [ 10/156]  eta: 0:02:48  lr: 0.001016  min_lr: 0.001016  loss: 2.8588 (2.7936)  weight_decay: 0.0500 (0.0500)  time: 1.1520  data: 0.3730  max mem: 57114
Epoch: [235]  [ 20/156]  eta: 0:02:10  lr: 0.001014  min_lr: 0.001014  loss: 2.7924 (2.6915)  weight_decay: 0.0500 (0.0500)  time: 0.7423  data: 0.0003  max mem: 57114
Epoch: [235]  [ 30/156]  eta: 0:01:52  lr: 0.001013  min_lr: 0.001013  loss: 2.7536 (2.6816)  weight_decay: 0.0500 (0.0500)  time: 0.7533  data: 0.0004  max mem: 57114
Epoch: [235]  [ 40/156]  eta: 0:01:38  lr: 0.001011  min_lr: 0.001011  loss: 2.6150 (2.6426)  weight_decay: 0.0500 (0.0500)  time: 0.7398  data: 0.0004  max mem: 57114
Epoch: [235]  [ 50/156]  eta: 0:01:28  lr: 0.001009  min_lr: 0.001009  loss: 2.6190 (2.6509)  weight_decay: 0.0500 (0.0500)  time: 0.7332  data: 0.0004  max mem: 57114
Epoch: [235]  [ 60/156]  eta: 0:01:18  lr: 0.001007  min_lr: 0.001007  loss: 2.7548 (2.6325)  weight_decay: 0.0500 (0.0500)  time: 0.7334  data: 0.0004  max mem: 57114
Epoch: [235]  [ 70/156]  eta: 0:01:08  lr: 0.001005  min_lr: 0.001005  loss: 2.7395 (2.6261)  weight_decay: 0.0500 (0.0500)  time: 0.7201  data: 0.0004  max mem: 57114
Epoch: [235]  [ 80/156]  eta: 0:00:59  lr: 0.001003  min_lr: 0.001003  loss: 2.7395 (2.6220)  weight_decay: 0.0500 (0.0500)  time: 0.7142  data: 0.0004  max mem: 57114
Epoch: [235]  [ 90/156]  eta: 0:00:51  lr: 0.001001  min_lr: 0.001001  loss: 2.7703 (2.6388)  weight_decay: 0.0500 (0.0500)  time: 0.7167  data: 0.0004  max mem: 57114
Epoch: [235]  [100/156]  eta: 0:00:43  lr: 0.000999  min_lr: 0.000999  loss: 2.8958 (2.6718)  weight_decay: 0.0500 (0.0500)  time: 0.7110  data: 0.0004  max mem: 57114
Epoch: [235]  [110/156]  eta: 0:00:35  lr: 0.000997  min_lr: 0.000997  loss: 2.8958 (2.6765)  weight_decay: 0.0500 (0.0500)  time: 0.7081  data: 0.0004  max mem: 57114
Epoch: [235]  [120/156]  eta: 0:00:27  lr: 0.000995  min_lr: 0.000995  loss: 2.7869 (2.6778)  weight_decay: 0.0500 (0.0500)  time: 0.7129  data: 0.0003  max mem: 57114
Epoch: [235]  [130/156]  eta: 0:00:19  lr: 0.000994  min_lr: 0.000994  loss: 2.7869 (2.6787)  weight_decay: 0.0500 (0.0500)  time: 0.6984  data: 0.0009  max mem: 57114
Epoch: [235]  [140/156]  eta: 0:00:12  lr: 0.000992  min_lr: 0.000992  loss: 2.6535 (2.6669)  weight_decay: 0.0500 (0.0500)  time: 0.6820  data: 0.0008  max mem: 57114
Epoch: [235]  [150/156]  eta: 0:00:04  lr: 0.000990  min_lr: 0.000990  loss: 2.6120 (2.6634)  weight_decay: 0.0500 (0.0500)  time: 0.6865  data: 0.0001  max mem: 57114
Epoch: [235]  [155/156]  eta: 0:00:00  lr: 0.000989  min_lr: 0.000989  loss: 2.6541 (2.6524)  weight_decay: 0.0500 (0.0500)  time: 0.6893  data: 0.0001  max mem: 57114
Epoch: [235] Total time: 0:01:56 (0.7488 s / it)
Averaged stats: lr: 0.000989  min_lr: 0.000989  loss: 2.6541 (2.6983)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5209 (0.5209)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.0411  data: 1.7846  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6838 (0.6481)  acc1: 84.8958 (84.3990)  acc5: 98.4375 (97.1867)  time: 0.5589  data: 0.3570  max mem: 57114
Test: Total time: 0:00:02 (0.5801 s / it)
* Acc@1 83.560 Acc@5 96.635 loss 0.750
Accuracy of the model on the 50000 test images: 83.6%
Max accuracy: 83.64%
Test:  [0/5]  eta: 0:00:11  loss: 0.5903 (0.5903)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.2640  data: 2.0205  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6896 (0.6742)  acc1: 81.2500 (79.4118)  acc5: 96.3542 (94.2455)  time: 0.6035  data: 0.4042  max mem: 57114
Test: Total time: 0:00:03 (0.6163 s / it)
* Acc@1 77.518 Acc@5 93.942 loss 0.813
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [236]  [  0/156]  eta: 0:12:29  lr: 0.000989  min_lr: 0.000989  loss: 2.9109 (2.9109)  weight_decay: 0.0500 (0.0500)  time: 4.8028  data: 3.6036  max mem: 57114
Epoch: [236]  [ 10/156]  eta: 0:02:40  lr: 0.000987  min_lr: 0.000987  loss: 2.9536 (2.7662)  weight_decay: 0.0500 (0.0500)  time: 1.0999  data: 0.3279  max mem: 57114
Epoch: [236]  [ 20/156]  eta: 0:02:06  lr: 0.000985  min_lr: 0.000985  loss: 2.6551 (2.6318)  weight_decay: 0.0500 (0.0500)  time: 0.7388  data: 0.0004  max mem: 57114
Epoch: [236]  [ 30/156]  eta: 0:01:49  lr: 0.000983  min_lr: 0.000983  loss: 2.6551 (2.6821)  weight_decay: 0.0500 (0.0500)  time: 0.7430  data: 0.0004  max mem: 57114
Epoch: [236]  [ 40/156]  eta: 0:01:36  lr: 0.000981  min_lr: 0.000981  loss: 2.7871 (2.6441)  weight_decay: 0.0500 (0.0500)  time: 0.7272  data: 0.0004  max mem: 57114
Epoch: [236]  [ 50/156]  eta: 0:01:25  lr: 0.000979  min_lr: 0.000979  loss: 2.7619 (2.6791)  weight_decay: 0.0500 (0.0500)  time: 0.7085  data: 0.0004  max mem: 57114
Epoch: [236]  [ 60/156]  eta: 0:01:16  lr: 0.000977  min_lr: 0.000977  loss: 2.7658 (2.6806)  weight_decay: 0.0500 (0.0500)  time: 0.7147  data: 0.0004  max mem: 57114
Epoch: [236]  [ 70/156]  eta: 0:01:07  lr: 0.000975  min_lr: 0.000975  loss: 2.7658 (2.6914)  weight_decay: 0.0500 (0.0500)  time: 0.7178  data: 0.0004  max mem: 57114
Epoch: [236]  [ 80/156]  eta: 0:00:58  lr: 0.000973  min_lr: 0.000973  loss: 2.8656 (2.6987)  weight_decay: 0.0500 (0.0500)  time: 0.7142  data: 0.0004  max mem: 57114
Epoch: [236]  [ 90/156]  eta: 0:00:50  lr: 0.000972  min_lr: 0.000972  loss: 2.7085 (2.6887)  weight_decay: 0.0500 (0.0500)  time: 0.7118  data: 0.0004  max mem: 57114
Epoch: [236]  [100/156]  eta: 0:00:42  lr: 0.000970  min_lr: 0.000970  loss: 2.8138 (2.6971)  weight_decay: 0.0500 (0.0500)  time: 0.7027  data: 0.0004  max mem: 57114
Epoch: [236]  [110/156]  eta: 0:00:34  lr: 0.000968  min_lr: 0.000968  loss: 2.7719 (2.6797)  weight_decay: 0.0500 (0.0500)  time: 0.7042  data: 0.0004  max mem: 57114
Epoch: [236]  [120/156]  eta: 0:00:27  lr: 0.000966  min_lr: 0.000966  loss: 2.8754 (2.7094)  weight_decay: 0.0500 (0.0500)  time: 0.7022  data: 0.0004  max mem: 57114
Epoch: [236]  [130/156]  eta: 0:00:19  lr: 0.000964  min_lr: 0.000964  loss: 2.8420 (2.6946)  weight_decay: 0.0500 (0.0500)  time: 0.6991  data: 0.0009  max mem: 57114
Epoch: [236]  [140/156]  eta: 0:00:11  lr: 0.000962  min_lr: 0.000962  loss: 2.5744 (2.6932)  weight_decay: 0.0500 (0.0500)  time: 0.6900  data: 0.0007  max mem: 57114
Epoch: [236]  [150/156]  eta: 0:00:04  lr: 0.000960  min_lr: 0.000960  loss: 2.7102 (2.7066)  weight_decay: 0.0500 (0.0500)  time: 0.6838  data: 0.0001  max mem: 57114
Epoch: [236]  [155/156]  eta: 0:00:00  lr: 0.000959  min_lr: 0.000959  loss: 2.7984 (2.7030)  weight_decay: 0.0500 (0.0500)  time: 0.6846  data: 0.0001  max mem: 57114
Epoch: [236] Total time: 0:01:55 (0.7383 s / it)
Averaged stats: lr: 0.000959  min_lr: 0.000959  loss: 2.7984 (2.6875)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6525 (0.6525)  acc1: 90.6250 (90.6250)  acc5: 98.9583 (98.9583)  time: 2.0993  data: 1.8431  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8468 (0.7648)  acc1: 82.8125 (84.6547)  acc5: 98.4375 (97.3146)  time: 0.5705  data: 0.3687  max mem: 57114
Test: Total time: 0:00:02 (0.5962 s / it)
* Acc@1 83.422 Acc@5 96.531 loss 0.862
Accuracy of the model on the 50000 test images: 83.4%
Max accuracy: 83.64%
Test:  [0/5]  eta: 0:00:11  loss: 0.5910 (0.5910)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.3595  data: 2.1160  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6895 (0.6746)  acc1: 81.2500 (79.2839)  acc5: 96.3542 (94.2455)  time: 0.6226  data: 0.4233  max mem: 57114
Test: Total time: 0:00:03 (0.6347 s / it)
* Acc@1 77.506 Acc@5 93.946 loss 0.814
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [237]  [  0/156]  eta: 0:11:57  lr: 0.000959  min_lr: 0.000959  loss: 2.1521 (2.1521)  weight_decay: 0.0500 (0.0500)  time: 4.5995  data: 3.5763  max mem: 57114
Epoch: [237]  [ 10/156]  eta: 0:02:41  lr: 0.000957  min_lr: 0.000957  loss: 2.8411 (2.7515)  weight_decay: 0.0500 (0.0500)  time: 1.1041  data: 0.3254  max mem: 57114
Epoch: [237]  [ 20/156]  eta: 0:02:03  lr: 0.000956  min_lr: 0.000956  loss: 2.8443 (2.8129)  weight_decay: 0.0500 (0.0500)  time: 0.7262  data: 0.0003  max mem: 57114
Epoch: [237]  [ 30/156]  eta: 0:01:48  lr: 0.000954  min_lr: 0.000954  loss: 2.8378 (2.8264)  weight_decay: 0.0500 (0.0500)  time: 0.7250  data: 0.0003  max mem: 57114
Epoch: [237]  [ 40/156]  eta: 0:01:35  lr: 0.000952  min_lr: 0.000952  loss: 2.8003 (2.7736)  weight_decay: 0.0500 (0.0500)  time: 0.7362  data: 0.0004  max mem: 57114
Epoch: [237]  [ 50/156]  eta: 0:01:25  lr: 0.000950  min_lr: 0.000950  loss: 2.6161 (2.7377)  weight_decay: 0.0500 (0.0500)  time: 0.7300  data: 0.0004  max mem: 57114
Epoch: [237]  [ 60/156]  eta: 0:01:16  lr: 0.000948  min_lr: 0.000948  loss: 2.6554 (2.7201)  weight_decay: 0.0500 (0.0500)  time: 0.7258  data: 0.0004  max mem: 57114
Epoch: [237]  [ 70/156]  eta: 0:01:07  lr: 0.000946  min_lr: 0.000946  loss: 2.6192 (2.7035)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0003  max mem: 57114
Epoch: [237]  [ 80/156]  eta: 0:00:58  lr: 0.000944  min_lr: 0.000944  loss: 2.4737 (2.6761)  weight_decay: 0.0500 (0.0500)  time: 0.7014  data: 0.0004  max mem: 57114
Epoch: [237]  [ 90/156]  eta: 0:00:50  lr: 0.000943  min_lr: 0.000943  loss: 2.7668 (2.6908)  weight_decay: 0.0500 (0.0500)  time: 0.7055  data: 0.0004  max mem: 57114
Epoch: [237]  [100/156]  eta: 0:00:42  lr: 0.000941  min_lr: 0.000941  loss: 2.8812 (2.6912)  weight_decay: 0.0500 (0.0500)  time: 0.7160  data: 0.0004  max mem: 57114
Epoch: [237]  [110/156]  eta: 0:00:34  lr: 0.000939  min_lr: 0.000939  loss: 2.7770 (2.6893)  weight_decay: 0.0500 (0.0500)  time: 0.7099  data: 0.0004  max mem: 57114
Epoch: [237]  [120/156]  eta: 0:00:27  lr: 0.000937  min_lr: 0.000937  loss: 2.7173 (2.6735)  weight_decay: 0.0500 (0.0500)  time: 0.7041  data: 0.0004  max mem: 57114
Epoch: [237]  [130/156]  eta: 0:00:19  lr: 0.000935  min_lr: 0.000935  loss: 2.5927 (2.6787)  weight_decay: 0.0500 (0.0500)  time: 0.6950  data: 0.0009  max mem: 57114
Epoch: [237]  [140/156]  eta: 0:00:11  lr: 0.000933  min_lr: 0.000933  loss: 2.7390 (2.6829)  weight_decay: 0.0500 (0.0500)  time: 0.6860  data: 0.0007  max mem: 57114
Epoch: [237]  [150/156]  eta: 0:00:04  lr: 0.000931  min_lr: 0.000931  loss: 2.7286 (2.6772)  weight_decay: 0.0500 (0.0500)  time: 0.6811  data: 0.0001  max mem: 57114
Epoch: [237]  [155/156]  eta: 0:00:00  lr: 0.000930  min_lr: 0.000930  loss: 2.7286 (2.6728)  weight_decay: 0.0500 (0.0500)  time: 0.6834  data: 0.0001  max mem: 57114
Epoch: [237] Total time: 0:01:55 (0.7380 s / it)
Averaged stats: lr: 0.000930  min_lr: 0.000930  loss: 2.7286 (2.6932)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5504 (0.5504)  acc1: 91.6667 (91.6667)  acc5: 98.4375 (98.4375)  time: 2.0819  data: 1.8260  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7495 (0.7045)  acc1: 83.3333 (84.1432)  acc5: 97.9167 (96.6752)  time: 0.5670  data: 0.3653  max mem: 57114
Test: Total time: 0:00:02 (0.5878 s / it)
* Acc@1 83.670 Acc@5 96.643 loss 0.789
Accuracy of the model on the 50000 test images: 83.7%
Max accuracy: 83.67%
Test:  [0/5]  eta: 0:00:10  loss: 0.5917 (0.5917)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.0174  data: 1.7739  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6897 (0.6753)  acc1: 81.2500 (79.2839)  acc5: 95.8333 (94.1177)  time: 0.5541  data: 0.3548  max mem: 57114
Test: Total time: 0:00:02 (0.5648 s / it)
* Acc@1 77.504 Acc@5 93.944 loss 0.814
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [238]  [  0/156]  eta: 0:12:55  lr: 0.000930  min_lr: 0.000930  loss: 2.8718 (2.8718)  weight_decay: 0.0500 (0.0500)  time: 4.9737  data: 3.8017  max mem: 57114
Epoch: [238]  [ 10/156]  eta: 0:02:44  lr: 0.000928  min_lr: 0.000928  loss: 3.0933 (2.8711)  weight_decay: 0.0500 (0.0500)  time: 1.1266  data: 0.3459  max mem: 57114
Epoch: [238]  [ 20/156]  eta: 0:02:10  lr: 0.000927  min_lr: 0.000927  loss: 2.7523 (2.7640)  weight_decay: 0.0500 (0.0500)  time: 0.7578  data: 0.0003  max mem: 57114
Epoch: [238]  [ 30/156]  eta: 0:01:51  lr: 0.000925  min_lr: 0.000925  loss: 2.7523 (2.7909)  weight_decay: 0.0500 (0.0500)  time: 0.7552  data: 0.0003  max mem: 57114
Epoch: [238]  [ 40/156]  eta: 0:01:38  lr: 0.000923  min_lr: 0.000923  loss: 2.7491 (2.7246)  weight_decay: 0.0500 (0.0500)  time: 0.7263  data: 0.0004  max mem: 57114
Epoch: [238]  [ 50/156]  eta: 0:01:27  lr: 0.000921  min_lr: 0.000921  loss: 2.6577 (2.7565)  weight_decay: 0.0500 (0.0500)  time: 0.7188  data: 0.0004  max mem: 57114
Epoch: [238]  [ 60/156]  eta: 0:01:17  lr: 0.000919  min_lr: 0.000919  loss: 2.8916 (2.7426)  weight_decay: 0.0500 (0.0500)  time: 0.7314  data: 0.0004  max mem: 57114
Epoch: [238]  [ 70/156]  eta: 0:01:08  lr: 0.000917  min_lr: 0.000917  loss: 2.6711 (2.7029)  weight_decay: 0.0500 (0.0500)  time: 0.7378  data: 0.0004  max mem: 57114
Epoch: [238]  [ 80/156]  eta: 0:01:00  lr: 0.000916  min_lr: 0.000916  loss: 2.5190 (2.6737)  weight_decay: 0.0500 (0.0500)  time: 0.7380  data: 0.0004  max mem: 57114
Epoch: [238]  [ 90/156]  eta: 0:00:51  lr: 0.000914  min_lr: 0.000914  loss: 2.6794 (2.6905)  weight_decay: 0.0500 (0.0500)  time: 0.7257  data: 0.0003  max mem: 57114
Epoch: [238]  [100/156]  eta: 0:00:43  lr: 0.000912  min_lr: 0.000912  loss: 2.8271 (2.6906)  weight_decay: 0.0500 (0.0500)  time: 0.7063  data: 0.0003  max mem: 57114
Epoch: [238]  [110/156]  eta: 0:00:35  lr: 0.000910  min_lr: 0.000910  loss: 2.7779 (2.6929)  weight_decay: 0.0500 (0.0500)  time: 0.6984  data: 0.0004  max mem: 57114
Epoch: [238]  [120/156]  eta: 0:00:27  lr: 0.000908  min_lr: 0.000908  loss: 2.6617 (2.6812)  weight_decay: 0.0500 (0.0500)  time: 0.7005  data: 0.0004  max mem: 57114
Epoch: [238]  [130/156]  eta: 0:00:19  lr: 0.000906  min_lr: 0.000906  loss: 2.4958 (2.6770)  weight_decay: 0.0500 (0.0500)  time: 0.6996  data: 0.0008  max mem: 57114
Epoch: [238]  [140/156]  eta: 0:00:12  lr: 0.000905  min_lr: 0.000905  loss: 2.5970 (2.6778)  weight_decay: 0.0500 (0.0500)  time: 0.6917  data: 0.0007  max mem: 57114
Epoch: [238]  [150/156]  eta: 0:00:04  lr: 0.000903  min_lr: 0.000903  loss: 2.8178 (2.6765)  weight_decay: 0.0500 (0.0500)  time: 0.6879  data: 0.0001  max mem: 57114
Epoch: [238]  [155/156]  eta: 0:00:00  lr: 0.000902  min_lr: 0.000902  loss: 2.8296 (2.6867)  weight_decay: 0.0500 (0.0500)  time: 0.6895  data: 0.0001  max mem: 57114
Epoch: [238] Total time: 0:01:56 (0.7479 s / it)
Averaged stats: lr: 0.000902  min_lr: 0.000902  loss: 2.8296 (2.6815)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:12  loss: 0.6663 (0.6663)  acc1: 89.0625 (89.0625)  acc5: 98.9583 (98.9583)  time: 2.4615  data: 2.2056  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8463 (0.7948)  acc1: 83.3333 (83.7596)  acc5: 98.9583 (97.4425)  time: 0.6430  data: 0.4412  max mem: 57114
Test: Total time: 0:00:03 (0.6640 s / it)
* Acc@1 83.492 Acc@5 96.743 loss 0.880
Accuracy of the model on the 50000 test images: 83.5%
Max accuracy: 83.67%
Test:  [0/5]  eta: 0:00:12  loss: 0.5924 (0.5924)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.4486  data: 2.2050  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6898 (0.6758)  acc1: 81.2500 (79.2839)  acc5: 95.8333 (94.1177)  time: 0.6404  data: 0.4411  max mem: 57114
Test: Total time: 0:00:03 (0.6530 s / it)
* Acc@1 77.516 Acc@5 93.934 loss 0.815
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [239]  [  0/156]  eta: 0:16:03  lr: 0.000902  min_lr: 0.000902  loss: 2.9742 (2.9742)  weight_decay: 0.0500 (0.0500)  time: 6.1786  data: 4.7116  max mem: 57114
Epoch: [239]  [ 10/156]  eta: 0:02:59  lr: 0.000900  min_lr: 0.000900  loss: 2.8584 (2.6847)  weight_decay: 0.0500 (0.0500)  time: 1.2262  data: 0.4286  max mem: 57114
Epoch: [239]  [ 20/156]  eta: 0:02:13  lr: 0.000898  min_lr: 0.000898  loss: 2.6876 (2.6873)  weight_decay: 0.0500 (0.0500)  time: 0.7229  data: 0.0003  max mem: 57114
Epoch: [239]  [ 30/156]  eta: 0:01:53  lr: 0.000896  min_lr: 0.000896  loss: 2.6901 (2.7242)  weight_decay: 0.0500 (0.0500)  time: 0.7257  data: 0.0003  max mem: 57114
Epoch: [239]  [ 40/156]  eta: 0:01:39  lr: 0.000894  min_lr: 0.000894  loss: 2.8200 (2.7339)  weight_decay: 0.0500 (0.0500)  time: 0.7264  data: 0.0003  max mem: 57114
Epoch: [239]  [ 50/156]  eta: 0:01:28  lr: 0.000893  min_lr: 0.000893  loss: 2.8261 (2.7429)  weight_decay: 0.0500 (0.0500)  time: 0.7367  data: 0.0003  max mem: 57114
Epoch: [239]  [ 60/156]  eta: 0:01:18  lr: 0.000891  min_lr: 0.000891  loss: 2.6290 (2.7117)  weight_decay: 0.0500 (0.0500)  time: 0.7312  data: 0.0004  max mem: 57114
Epoch: [239]  [ 70/156]  eta: 0:01:09  lr: 0.000889  min_lr: 0.000889  loss: 2.7255 (2.7087)  weight_decay: 0.0500 (0.0500)  time: 0.7175  data: 0.0004  max mem: 57114
Epoch: [239]  [ 80/156]  eta: 0:01:00  lr: 0.000887  min_lr: 0.000887  loss: 2.8262 (2.7257)  weight_decay: 0.0500 (0.0500)  time: 0.7143  data: 0.0004  max mem: 57114
Epoch: [239]  [ 90/156]  eta: 0:00:51  lr: 0.000885  min_lr: 0.000885  loss: 2.6860 (2.7076)  weight_decay: 0.0500 (0.0500)  time: 0.7011  data: 0.0004  max mem: 57114
Epoch: [239]  [100/156]  eta: 0:00:43  lr: 0.000884  min_lr: 0.000884  loss: 2.6088 (2.7023)  weight_decay: 0.0500 (0.0500)  time: 0.7056  data: 0.0003  max mem: 57114
Epoch: [239]  [110/156]  eta: 0:00:35  lr: 0.000882  min_lr: 0.000882  loss: 2.8541 (2.7119)  weight_decay: 0.0500 (0.0500)  time: 0.6983  data: 0.0003  max mem: 57114
Epoch: [239]  [120/156]  eta: 0:00:27  lr: 0.000880  min_lr: 0.000880  loss: 2.8541 (2.6964)  weight_decay: 0.0500 (0.0500)  time: 0.7066  data: 0.0004  max mem: 57114
Epoch: [239]  [130/156]  eta: 0:00:19  lr: 0.000878  min_lr: 0.000878  loss: 2.5710 (2.6989)  weight_decay: 0.0500 (0.0500)  time: 0.7045  data: 0.0008  max mem: 57114
Epoch: [239]  [140/156]  eta: 0:00:12  lr: 0.000876  min_lr: 0.000876  loss: 2.8241 (2.7010)  weight_decay: 0.0500 (0.0500)  time: 0.6852  data: 0.0007  max mem: 57114
Epoch: [239]  [150/156]  eta: 0:00:04  lr: 0.000875  min_lr: 0.000875  loss: 2.8241 (2.6962)  weight_decay: 0.0500 (0.0500)  time: 0.6884  data: 0.0001  max mem: 57114
Epoch: [239]  [155/156]  eta: 0:00:00  lr: 0.000874  min_lr: 0.000874  loss: 2.7771 (2.6949)  weight_decay: 0.0500 (0.0500)  time: 0.6870  data: 0.0001  max mem: 57114
Epoch: [239] Total time: 0:01:56 (0.7481 s / it)
Averaged stats: lr: 0.000874  min_lr: 0.000874  loss: 2.7771 (2.6750)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6058 (0.6058)  acc1: 89.5833 (89.5833)  acc5: 98.4375 (98.4375)  time: 2.0994  data: 1.8383  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7907 (0.7284)  acc1: 82.2917 (83.6317)  acc5: 97.9167 (97.0588)  time: 0.5706  data: 0.3677  max mem: 57114
Test: Total time: 0:00:02 (0.5941 s / it)
* Acc@1 83.624 Acc@5 96.649 loss 0.820
Accuracy of the model on the 50000 test images: 83.6%
Max accuracy: 83.67%
Test:  [0/5]  eta: 0:00:12  loss: 0.5931 (0.5931)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.4062  data: 2.1624  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6898 (0.6765)  acc1: 81.2500 (79.2839)  acc5: 95.8333 (94.2455)  time: 0.6320  data: 0.4326  max mem: 57114
Test: Total time: 0:00:03 (0.6435 s / it)
* Acc@1 77.528 Acc@5 93.938 loss 0.815
Accuracy of the model EMA on 50000 test images: 77.5%
Max EMA accuracy: 77.53%
Epoch: [240]  [  0/156]  eta: 0:08:40  lr: 0.000874  min_lr: 0.000874  loss: 2.9621 (2.9621)  weight_decay: 0.0500 (0.0500)  time: 3.3350  data: 2.6808  max mem: 57114
Epoch: [240]  [ 10/156]  eta: 0:02:19  lr: 0.000872  min_lr: 0.000872  loss: 2.5732 (2.5811)  weight_decay: 0.0500 (0.0500)  time: 0.9564  data: 0.2440  max mem: 57114
Epoch: [240]  [ 20/156]  eta: 0:01:55  lr: 0.000870  min_lr: 0.000870  loss: 2.7109 (2.6075)  weight_decay: 0.0500 (0.0500)  time: 0.7244  data: 0.0003  max mem: 57114
Epoch: [240]  [ 30/156]  eta: 0:01:42  lr: 0.000868  min_lr: 0.000868  loss: 2.7219 (2.5809)  weight_decay: 0.0500 (0.0500)  time: 0.7355  data: 0.0004  max mem: 57114
Epoch: [240]  [ 40/156]  eta: 0:01:31  lr: 0.000866  min_lr: 0.000866  loss: 2.6848 (2.6164)  weight_decay: 0.0500 (0.0500)  time: 0.7331  data: 0.0004  max mem: 57114
Epoch: [240]  [ 50/156]  eta: 0:01:22  lr: 0.000865  min_lr: 0.000865  loss: 2.6848 (2.6271)  weight_decay: 0.0500 (0.0500)  time: 0.7253  data: 0.0004  max mem: 57114
Epoch: [240]  [ 60/156]  eta: 0:01:13  lr: 0.000863  min_lr: 0.000863  loss: 2.6349 (2.6040)  weight_decay: 0.0500 (0.0500)  time: 0.7137  data: 0.0004  max mem: 57114
Epoch: [240]  [ 70/156]  eta: 0:01:05  lr: 0.000861  min_lr: 0.000861  loss: 2.7275 (2.6113)  weight_decay: 0.0500 (0.0500)  time: 0.7093  data: 0.0004  max mem: 57114
Epoch: [240]  [ 80/156]  eta: 0:00:57  lr: 0.000859  min_lr: 0.000859  loss: 2.5979 (2.6016)  weight_decay: 0.0500 (0.0500)  time: 0.7177  data: 0.0004  max mem: 57114
Epoch: [240]  [ 90/156]  eta: 0:00:49  lr: 0.000857  min_lr: 0.000857  loss: 2.5394 (2.6114)  weight_decay: 0.0500 (0.0500)  time: 0.7138  data: 0.0003  max mem: 57114
Epoch: [240]  [100/156]  eta: 0:00:41  lr: 0.000856  min_lr: 0.000856  loss: 2.7878 (2.6176)  weight_decay: 0.0500 (0.0500)  time: 0.7044  data: 0.0003  max mem: 57114
Epoch: [240]  [110/156]  eta: 0:00:34  lr: 0.000854  min_lr: 0.000854  loss: 2.7733 (2.6195)  weight_decay: 0.0500 (0.0500)  time: 0.7042  data: 0.0003  max mem: 57114
Epoch: [240]  [120/156]  eta: 0:00:26  lr: 0.000852  min_lr: 0.000852  loss: 2.7176 (2.6247)  weight_decay: 0.0500 (0.0500)  time: 0.7178  data: 0.0003  max mem: 57114
Epoch: [240]  [130/156]  eta: 0:00:19  lr: 0.000850  min_lr: 0.000850  loss: 2.7176 (2.6231)  weight_decay: 0.0500 (0.0500)  time: 0.7122  data: 0.0008  max mem: 57114
Epoch: [240]  [140/156]  eta: 0:00:11  lr: 0.000849  min_lr: 0.000849  loss: 2.8247 (2.6348)  weight_decay: 0.0500 (0.0500)  time: 0.6912  data: 0.0007  max mem: 57114
Epoch: [240]  [150/156]  eta: 0:00:04  lr: 0.000847  min_lr: 0.000847  loss: 2.8384 (2.6451)  weight_decay: 0.0500 (0.0500)  time: 0.6895  data: 0.0001  max mem: 57114
Epoch: [240]  [155/156]  eta: 0:00:00  lr: 0.000846  min_lr: 0.000846  loss: 2.8529 (2.6375)  weight_decay: 0.0500 (0.0500)  time: 0.6894  data: 0.0001  max mem: 57114
Epoch: [240] Total time: 0:01:54 (0.7315 s / it)
Averaged stats: lr: 0.000846  min_lr: 0.000846  loss: 2.8529 (2.6700)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5677 (0.5677)  acc1: 91.6667 (91.6667)  acc5: 98.9583 (98.9583)  time: 2.0775  data: 1.8215  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7812 (0.7199)  acc1: 82.8125 (83.3760)  acc5: 98.4375 (97.3146)  time: 0.5662  data: 0.3644  max mem: 57114
Test: Total time: 0:00:02 (0.5853 s / it)
* Acc@1 83.698 Acc@5 96.729 loss 0.793
Accuracy of the model on the 50000 test images: 83.7%
Max accuracy: 83.70%
Test:  [0/5]  eta: 0:00:10  loss: 0.5938 (0.5938)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.0339  data: 1.7904  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6900 (0.6772)  acc1: 81.2500 (79.2839)  acc5: 95.8333 (94.2455)  time: 0.5574  data: 0.3582  max mem: 57114
Test: Total time: 0:00:02 (0.5708 s / it)
* Acc@1 77.508 Acc@5 93.928 loss 0.816
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [241]  [  0/156]  eta: 0:12:27  lr: 0.000846  min_lr: 0.000846  loss: 2.8060 (2.8060)  weight_decay: 0.0500 (0.0500)  time: 4.7924  data: 3.8471  max mem: 57114
Epoch: [241]  [ 10/156]  eta: 0:02:40  lr: 0.000844  min_lr: 0.000844  loss: 2.6657 (2.5579)  weight_decay: 0.0500 (0.0500)  time: 1.1007  data: 0.3501  max mem: 57114
Epoch: [241]  [ 20/156]  eta: 0:02:06  lr: 0.000842  min_lr: 0.000842  loss: 2.6460 (2.5315)  weight_decay: 0.0500 (0.0500)  time: 0.7339  data: 0.0004  max mem: 57114
Epoch: [241]  [ 30/156]  eta: 0:01:49  lr: 0.000840  min_lr: 0.000840  loss: 2.6737 (2.5384)  weight_decay: 0.0500 (0.0500)  time: 0.7434  data: 0.0004  max mem: 57114
Epoch: [241]  [ 40/156]  eta: 0:01:37  lr: 0.000839  min_lr: 0.000839  loss: 2.8474 (2.6238)  weight_decay: 0.0500 (0.0500)  time: 0.7537  data: 0.0004  max mem: 57114
Epoch: [241]  [ 50/156]  eta: 0:01:26  lr: 0.000837  min_lr: 0.000837  loss: 2.9437 (2.6691)  weight_decay: 0.0500 (0.0500)  time: 0.7427  data: 0.0004  max mem: 57114
Epoch: [241]  [ 60/156]  eta: 0:01:17  lr: 0.000835  min_lr: 0.000835  loss: 2.9515 (2.6886)  weight_decay: 0.0500 (0.0500)  time: 0.7240  data: 0.0003  max mem: 57114
Epoch: [241]  [ 70/156]  eta: 0:01:08  lr: 0.000833  min_lr: 0.000833  loss: 2.9310 (2.6967)  weight_decay: 0.0500 (0.0500)  time: 0.7334  data: 0.0003  max mem: 57114
Epoch: [241]  [ 80/156]  eta: 0:00:59  lr: 0.000832  min_lr: 0.000832  loss: 2.6387 (2.6795)  weight_decay: 0.0500 (0.0500)  time: 0.7263  data: 0.0003  max mem: 57114
Epoch: [241]  [ 90/156]  eta: 0:00:51  lr: 0.000830  min_lr: 0.000830  loss: 2.6281 (2.6807)  weight_decay: 0.0500 (0.0500)  time: 0.7098  data: 0.0004  max mem: 57114
Epoch: [241]  [100/156]  eta: 0:00:43  lr: 0.000828  min_lr: 0.000828  loss: 2.8104 (2.6856)  weight_decay: 0.0500 (0.0500)  time: 0.7131  data: 0.0004  max mem: 57114
Epoch: [241]  [110/156]  eta: 0:00:35  lr: 0.000826  min_lr: 0.000826  loss: 2.8928 (2.6955)  weight_decay: 0.0500 (0.0500)  time: 0.7103  data: 0.0003  max mem: 57114
Epoch: [241]  [120/156]  eta: 0:00:27  lr: 0.000825  min_lr: 0.000825  loss: 2.8187 (2.6998)  weight_decay: 0.0500 (0.0500)  time: 0.7009  data: 0.0004  max mem: 57114
Epoch: [241]  [130/156]  eta: 0:00:19  lr: 0.000823  min_lr: 0.000823  loss: 2.7643 (2.6955)  weight_decay: 0.0500 (0.0500)  time: 0.7056  data: 0.0005  max mem: 57114
Epoch: [241]  [140/156]  eta: 0:00:12  lr: 0.000821  min_lr: 0.000821  loss: 2.8489 (2.7003)  weight_decay: 0.0500 (0.0500)  time: 0.7217  data: 0.0003  max mem: 57114
Epoch: [241]  [150/156]  eta: 0:00:04  lr: 0.000819  min_lr: 0.000819  loss: 2.6469 (2.6679)  weight_decay: 0.0500 (0.0500)  time: 0.7067  data: 0.0001  max mem: 57114
Epoch: [241]  [155/156]  eta: 0:00:00  lr: 0.000819  min_lr: 0.000819  loss: 2.2003 (2.6544)  weight_decay: 0.0500 (0.0500)  time: 0.6946  data: 0.0001  max mem: 57114
Epoch: [241] Total time: 0:01:56 (0.7492 s / it)
Averaged stats: lr: 0.000819  min_lr: 0.000819  loss: 2.2003 (2.6624)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.4867 (0.4867)  acc1: 90.6250 (90.6250)  acc5: 99.4792 (99.4792)  time: 2.0909  data: 1.8352  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6725 (0.6267)  acc1: 84.3750 (84.2711)  acc5: 98.4375 (97.6982)  time: 0.5688  data: 0.3671  max mem: 57114
Test: Total time: 0:00:02 (0.5926 s / it)
* Acc@1 83.951 Acc@5 96.839 loss 0.720
Accuracy of the model on the 50000 test images: 84.0%
Max accuracy: 83.95%
Test:  [0/5]  eta: 0:00:10  loss: 0.5946 (0.5946)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.0281  data: 1.7846  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6901 (0.6779)  acc1: 81.2500 (79.2839)  acc5: 95.8333 (94.2455)  time: 0.5562  data: 0.3570  max mem: 57114
Test: Total time: 0:00:02 (0.5714 s / it)
* Acc@1 77.512 Acc@5 93.926 loss 0.817
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [242]  [  0/156]  eta: 0:13:05  lr: 0.000818  min_lr: 0.000818  loss: 2.3768 (2.3768)  weight_decay: 0.0500 (0.0500)  time: 5.0335  data: 3.5966  max mem: 57114
Epoch: [242]  [ 10/156]  eta: 0:02:46  lr: 0.000817  min_lr: 0.000817  loss: 2.7988 (2.7441)  weight_decay: 0.0500 (0.0500)  time: 1.1400  data: 0.3273  max mem: 57114
Epoch: [242]  [ 20/156]  eta: 0:02:09  lr: 0.000815  min_lr: 0.000815  loss: 2.7988 (2.7200)  weight_decay: 0.0500 (0.0500)  time: 0.7499  data: 0.0003  max mem: 57114
Epoch: [242]  [ 30/156]  eta: 0:01:51  lr: 0.000813  min_lr: 0.000813  loss: 2.5689 (2.7071)  weight_decay: 0.0500 (0.0500)  time: 0.7484  data: 0.0003  max mem: 57114
Epoch: [242]  [ 40/156]  eta: 0:01:39  lr: 0.000811  min_lr: 0.000811  loss: 2.5988 (2.6884)  weight_decay: 0.0500 (0.0500)  time: 0.7593  data: 0.0004  max mem: 57114
Epoch: [242]  [ 50/156]  eta: 0:01:28  lr: 0.000810  min_lr: 0.000810  loss: 2.6698 (2.7037)  weight_decay: 0.0500 (0.0500)  time: 0.7569  data: 0.0004  max mem: 57114
Epoch: [242]  [ 60/156]  eta: 0:01:18  lr: 0.000808  min_lr: 0.000808  loss: 2.5607 (2.6634)  weight_decay: 0.0500 (0.0500)  time: 0.7441  data: 0.0004  max mem: 57114
Epoch: [242]  [ 70/156]  eta: 0:01:09  lr: 0.000806  min_lr: 0.000806  loss: 2.5607 (2.6645)  weight_decay: 0.0500 (0.0500)  time: 0.7390  data: 0.0004  max mem: 57114
Epoch: [242]  [ 80/156]  eta: 0:01:00  lr: 0.000805  min_lr: 0.000805  loss: 2.6089 (2.6434)  weight_decay: 0.0500 (0.0500)  time: 0.7294  data: 0.0004  max mem: 57114
Epoch: [242]  [ 90/156]  eta: 0:00:52  lr: 0.000803  min_lr: 0.000803  loss: 2.5645 (2.6337)  weight_decay: 0.0500 (0.0500)  time: 0.7209  data: 0.0004  max mem: 57114
Epoch: [242]  [100/156]  eta: 0:00:43  lr: 0.000801  min_lr: 0.000801  loss: 2.7806 (2.6592)  weight_decay: 0.0500 (0.0500)  time: 0.7177  data: 0.0004  max mem: 57114
Epoch: [242]  [110/156]  eta: 0:00:35  lr: 0.000799  min_lr: 0.000799  loss: 2.8728 (2.6600)  weight_decay: 0.0500 (0.0500)  time: 0.7360  data: 0.0004  max mem: 57114
Epoch: [242]  [120/156]  eta: 0:00:28  lr: 0.000798  min_lr: 0.000798  loss: 2.7782 (2.6619)  weight_decay: 0.0500 (0.0500)  time: 0.7551  data: 0.0004  max mem: 57114
Epoch: [242]  [130/156]  eta: 0:00:20  lr: 0.000796  min_lr: 0.000796  loss: 2.8597 (2.6769)  weight_decay: 0.0500 (0.0500)  time: 0.7429  data: 0.0009  max mem: 57114
Epoch: [242]  [140/156]  eta: 0:00:12  lr: 0.000794  min_lr: 0.000794  loss: 2.8860 (2.6749)  weight_decay: 0.0500 (0.0500)  time: 0.7255  data: 0.0008  max mem: 57114
Epoch: [242]  [150/156]  eta: 0:00:04  lr: 0.000792  min_lr: 0.000792  loss: 2.7185 (2.6648)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0001  max mem: 57114
Epoch: [242]  [155/156]  eta: 0:00:00  lr: 0.000792  min_lr: 0.000792  loss: 2.7185 (2.6673)  weight_decay: 0.0500 (0.0500)  time: 0.6949  data: 0.0001  max mem: 57114
Epoch: [242] Total time: 0:01:59 (0.7654 s / it)
Averaged stats: lr: 0.000792  min_lr: 0.000792  loss: 2.7185 (2.6574)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6182 (0.6182)  acc1: 91.1458 (91.1458)  acc5: 98.9583 (98.9583)  time: 2.0254  data: 1.7699  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7742 (0.7532)  acc1: 84.8958 (84.3990)  acc5: 98.9583 (97.5703)  time: 0.5558  data: 0.3541  max mem: 57114
Test: Total time: 0:00:02 (0.5756 s / it)
* Acc@1 83.766 Acc@5 96.805 loss 0.835
Accuracy of the model on the 50000 test images: 83.8%
Max accuracy: 83.95%
Test:  [0/5]  eta: 0:00:11  loss: 0.5952 (0.5952)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.2551  data: 2.0116  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6903 (0.6785)  acc1: 81.2500 (79.2839)  acc5: 95.8333 (94.1177)  time: 0.6017  data: 0.4024  max mem: 57114
Test: Total time: 0:00:03 (0.6126 s / it)
* Acc@1 77.498 Acc@5 93.916 loss 0.818
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [243]  [  0/156]  eta: 0:13:00  lr: 0.000791  min_lr: 0.000791  loss: 2.5057 (2.5057)  weight_decay: 0.0500 (0.0500)  time: 5.0028  data: 3.5833  max mem: 57114
Epoch: [243]  [ 10/156]  eta: 0:02:44  lr: 0.000790  min_lr: 0.000790  loss: 2.8692 (2.6019)  weight_decay: 0.0500 (0.0500)  time: 1.1271  data: 0.3260  max mem: 57114
Epoch: [243]  [ 20/156]  eta: 0:02:08  lr: 0.000788  min_lr: 0.000788  loss: 2.8816 (2.6357)  weight_decay: 0.0500 (0.0500)  time: 0.7396  data: 0.0003  max mem: 57114
Epoch: [243]  [ 30/156]  eta: 0:01:50  lr: 0.000786  min_lr: 0.000786  loss: 2.7979 (2.6052)  weight_decay: 0.0500 (0.0500)  time: 0.7368  data: 0.0003  max mem: 57114
Epoch: [243]  [ 40/156]  eta: 0:01:37  lr: 0.000785  min_lr: 0.000785  loss: 2.5276 (2.5870)  weight_decay: 0.0500 (0.0500)  time: 0.7367  data: 0.0003  max mem: 57114
Epoch: [243]  [ 50/156]  eta: 0:01:26  lr: 0.000783  min_lr: 0.000783  loss: 2.6567 (2.5875)  weight_decay: 0.0500 (0.0500)  time: 0.7325  data: 0.0004  max mem: 57114
Epoch: [243]  [ 60/156]  eta: 0:01:17  lr: 0.000781  min_lr: 0.000781  loss: 2.6791 (2.5993)  weight_decay: 0.0500 (0.0500)  time: 0.7308  data: 0.0004  max mem: 57114
Epoch: [243]  [ 70/156]  eta: 0:01:08  lr: 0.000779  min_lr: 0.000779  loss: 2.6791 (2.6142)  weight_decay: 0.0500 (0.0500)  time: 0.7197  data: 0.0004  max mem: 57114
Epoch: [243]  [ 80/156]  eta: 0:00:59  lr: 0.000778  min_lr: 0.000778  loss: 2.6636 (2.6028)  weight_decay: 0.0500 (0.0500)  time: 0.6968  data: 0.0004  max mem: 57114
Epoch: [243]  [ 90/156]  eta: 0:00:50  lr: 0.000776  min_lr: 0.000776  loss: 2.7099 (2.6269)  weight_decay: 0.0500 (0.0500)  time: 0.6943  data: 0.0004  max mem: 57114
Epoch: [243]  [100/156]  eta: 0:00:43  lr: 0.000774  min_lr: 0.000774  loss: 2.8396 (2.6112)  weight_decay: 0.0500 (0.0500)  time: 0.7264  data: 0.0003  max mem: 57114
Epoch: [243]  [110/156]  eta: 0:00:35  lr: 0.000773  min_lr: 0.000773  loss: 2.7292 (2.6140)  weight_decay: 0.0500 (0.0500)  time: 0.7338  data: 0.0003  max mem: 57114
Epoch: [243]  [120/156]  eta: 0:00:27  lr: 0.000771  min_lr: 0.000771  loss: 2.7292 (2.6199)  weight_decay: 0.0500 (0.0500)  time: 0.7168  data: 0.0004  max mem: 57114
Epoch: [243]  [130/156]  eta: 0:00:19  lr: 0.000769  min_lr: 0.000769  loss: 2.6704 (2.6120)  weight_decay: 0.0500 (0.0500)  time: 0.7294  data: 0.0009  max mem: 57114
Epoch: [243]  [140/156]  eta: 0:00:12  lr: 0.000768  min_lr: 0.000768  loss: 2.7085 (2.6142)  weight_decay: 0.0500 (0.0500)  time: 0.7293  data: 0.0007  max mem: 57114
Epoch: [243]  [150/156]  eta: 0:00:04  lr: 0.000766  min_lr: 0.000766  loss: 2.7268 (2.6171)  weight_decay: 0.0500 (0.0500)  time: 0.7205  data: 0.0001  max mem: 57114
Epoch: [243]  [155/156]  eta: 0:00:00  lr: 0.000765  min_lr: 0.000765  loss: 2.7140 (2.6163)  weight_decay: 0.0500 (0.0500)  time: 0.7253  data: 0.0001  max mem: 57114
Epoch: [243] Total time: 0:01:57 (0.7551 s / it)
Averaged stats: lr: 0.000765  min_lr: 0.000765  loss: 2.7140 (2.6527)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:17  loss: 0.6121 (0.6121)  acc1: 89.0625 (89.0625)  acc5: 98.9583 (98.9583)  time: 3.4770  data: 3.1987  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7548 (0.7176)  acc1: 84.8958 (84.0153)  acc5: 98.9583 (97.9540)  time: 0.8509  data: 0.6398  max mem: 57114
Test: Total time: 0:00:04 (0.8880 s / it)
* Acc@1 83.875 Acc@5 96.847 loss 0.803
Accuracy of the model on the 50000 test images: 83.9%
Max accuracy: 83.95%
Test:  [0/5]  eta: 0:00:18  loss: 0.5958 (0.5958)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 3.7638  data: 3.5197  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6905 (0.6791)  acc1: 81.2500 (79.2839)  acc5: 95.8333 (94.1177)  time: 0.9060  data: 0.7040  max mem: 57114
Test: Total time: 0:00:04 (0.9214 s / it)
* Acc@1 77.490 Acc@5 93.916 loss 0.818
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [244]  [  0/156]  eta: 0:13:26  lr: 0.000765  min_lr: 0.000765  loss: 3.0913 (3.0913)  weight_decay: 0.0500 (0.0500)  time: 5.1727  data: 4.0515  max mem: 57114
Epoch: [244]  [ 10/156]  eta: 0:02:45  lr: 0.000763  min_lr: 0.000763  loss: 2.7114 (2.5481)  weight_decay: 0.0500 (0.0500)  time: 1.1342  data: 0.3686  max mem: 57114
Epoch: [244]  [ 20/156]  eta: 0:02:08  lr: 0.000761  min_lr: 0.000761  loss: 2.5909 (2.5043)  weight_decay: 0.0500 (0.0500)  time: 0.7342  data: 0.0003  max mem: 57114
Epoch: [244]  [ 30/156]  eta: 0:01:50  lr: 0.000760  min_lr: 0.000760  loss: 2.6295 (2.5620)  weight_decay: 0.0500 (0.0500)  time: 0.7349  data: 0.0004  max mem: 57114
Epoch: [244]  [ 40/156]  eta: 0:01:37  lr: 0.000758  min_lr: 0.000758  loss: 2.9121 (2.6571)  weight_decay: 0.0500 (0.0500)  time: 0.7339  data: 0.0004  max mem: 57114
Epoch: [244]  [ 50/156]  eta: 0:01:26  lr: 0.000756  min_lr: 0.000756  loss: 2.9121 (2.6724)  weight_decay: 0.0500 (0.0500)  time: 0.7250  data: 0.0004  max mem: 57114
Epoch: [244]  [ 60/156]  eta: 0:01:17  lr: 0.000755  min_lr: 0.000755  loss: 2.8832 (2.6736)  weight_decay: 0.0500 (0.0500)  time: 0.7249  data: 0.0004  max mem: 57114
Epoch: [244]  [ 70/156]  eta: 0:01:07  lr: 0.000753  min_lr: 0.000753  loss: 2.8578 (2.6698)  weight_decay: 0.0500 (0.0500)  time: 0.7191  data: 0.0004  max mem: 57114
Epoch: [244]  [ 80/156]  eta: 0:00:59  lr: 0.000751  min_lr: 0.000751  loss: 2.7743 (2.6751)  weight_decay: 0.0500 (0.0500)  time: 0.7140  data: 0.0004  max mem: 57114
Epoch: [244]  [ 90/156]  eta: 0:00:51  lr: 0.000750  min_lr: 0.000750  loss: 2.8261 (2.6991)  weight_decay: 0.0500 (0.0500)  time: 0.7166  data: 0.0004  max mem: 57114
Epoch: [244]  [100/156]  eta: 0:00:43  lr: 0.000748  min_lr: 0.000748  loss: 2.8484 (2.6978)  weight_decay: 0.0500 (0.0500)  time: 0.7137  data: 0.0004  max mem: 57114
Epoch: [244]  [110/156]  eta: 0:00:35  lr: 0.000746  min_lr: 0.000746  loss: 2.6802 (2.6951)  weight_decay: 0.0500 (0.0500)  time: 0.7161  data: 0.0004  max mem: 57114
Epoch: [244]  [120/156]  eta: 0:00:27  lr: 0.000745  min_lr: 0.000745  loss: 2.6092 (2.6797)  weight_decay: 0.0500 (0.0500)  time: 0.7081  data: 0.0004  max mem: 57114
Epoch: [244]  [130/156]  eta: 0:00:19  lr: 0.000743  min_lr: 0.000743  loss: 2.6092 (2.6811)  weight_decay: 0.0500 (0.0500)  time: 0.7052  data: 0.0009  max mem: 57114
Epoch: [244]  [140/156]  eta: 0:00:11  lr: 0.000741  min_lr: 0.000741  loss: 2.8109 (2.6842)  weight_decay: 0.0500 (0.0500)  time: 0.6976  data: 0.0007  max mem: 57114
Epoch: [244]  [150/156]  eta: 0:00:04  lr: 0.000740  min_lr: 0.000740  loss: 2.8982 (2.6973)  weight_decay: 0.0500 (0.0500)  time: 0.6848  data: 0.0001  max mem: 57114
Epoch: [244]  [155/156]  eta: 0:00:00  lr: 0.000739  min_lr: 0.000739  loss: 2.8019 (2.6859)  weight_decay: 0.0500 (0.0500)  time: 0.6849  data: 0.0001  max mem: 57114
Epoch: [244] Total time: 0:01:56 (0.7456 s / it)
Averaged stats: lr: 0.000739  min_lr: 0.000739  loss: 2.8019 (2.6483)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5932 (0.5932)  acc1: 92.1875 (92.1875)  acc5: 98.9583 (98.9583)  time: 2.1024  data: 1.8466  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7528 (0.7026)  acc1: 86.4583 (85.4220)  acc5: 98.9583 (97.4425)  time: 0.5712  data: 0.3694  max mem: 57114
Test: Total time: 0:00:02 (0.5938 s / it)
* Acc@1 83.871 Acc@5 96.831 loss 0.795
Accuracy of the model on the 50000 test images: 83.9%
Max accuracy: 83.95%
Test:  [0/5]  eta: 0:00:11  loss: 0.5966 (0.5966)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.2536  data: 2.0101  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6908 (0.6799)  acc1: 81.2500 (79.4118)  acc5: 95.8333 (94.1177)  time: 0.6013  data: 0.4021  max mem: 57114
Test: Total time: 0:00:03 (0.6138 s / it)
* Acc@1 77.492 Acc@5 93.906 loss 0.819
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [245]  [  0/156]  eta: 0:12:52  lr: 0.000739  min_lr: 0.000739  loss: 2.8720 (2.8720)  weight_decay: 0.0500 (0.0500)  time: 4.9510  data: 3.6757  max mem: 57114
Epoch: [245]  [ 10/156]  eta: 0:02:40  lr: 0.000737  min_lr: 0.000737  loss: 2.6145 (2.6116)  weight_decay: 0.0500 (0.0500)  time: 1.1021  data: 0.3344  max mem: 57114
Epoch: [245]  [ 20/156]  eta: 0:02:04  lr: 0.000735  min_lr: 0.000735  loss: 2.6808 (2.6834)  weight_decay: 0.0500 (0.0500)  time: 0.7158  data: 0.0003  max mem: 57114
Epoch: [245]  [ 30/156]  eta: 0:01:47  lr: 0.000734  min_lr: 0.000734  loss: 2.7031 (2.6464)  weight_decay: 0.0500 (0.0500)  time: 0.7178  data: 0.0002  max mem: 57114
Epoch: [245]  [ 40/156]  eta: 0:01:35  lr: 0.000732  min_lr: 0.000732  loss: 2.7331 (2.6911)  weight_decay: 0.0500 (0.0500)  time: 0.7202  data: 0.0003  max mem: 57114
Epoch: [245]  [ 50/156]  eta: 0:01:25  lr: 0.000730  min_lr: 0.000730  loss: 2.8630 (2.7042)  weight_decay: 0.0500 (0.0500)  time: 0.7330  data: 0.0004  max mem: 57114
Epoch: [245]  [ 60/156]  eta: 0:01:16  lr: 0.000729  min_lr: 0.000729  loss: 2.8623 (2.7095)  weight_decay: 0.0500 (0.0500)  time: 0.7333  data: 0.0003  max mem: 57114
Epoch: [245]  [ 70/156]  eta: 0:01:07  lr: 0.000727  min_lr: 0.000727  loss: 2.7005 (2.7033)  weight_decay: 0.0500 (0.0500)  time: 0.7189  data: 0.0003  max mem: 57114
Epoch: [245]  [ 80/156]  eta: 0:00:58  lr: 0.000725  min_lr: 0.000725  loss: 2.5755 (2.6697)  weight_decay: 0.0500 (0.0500)  time: 0.7109  data: 0.0003  max mem: 57114
Epoch: [245]  [ 90/156]  eta: 0:00:50  lr: 0.000724  min_lr: 0.000724  loss: 2.5310 (2.6335)  weight_decay: 0.0500 (0.0500)  time: 0.7131  data: 0.0003  max mem: 57114
Epoch: [245]  [100/156]  eta: 0:00:42  lr: 0.000722  min_lr: 0.000722  loss: 2.5122 (2.6275)  weight_decay: 0.0500 (0.0500)  time: 0.7097  data: 0.0004  max mem: 57114
Epoch: [245]  [110/156]  eta: 0:00:34  lr: 0.000720  min_lr: 0.000720  loss: 2.5436 (2.6225)  weight_decay: 0.0500 (0.0500)  time: 0.7053  data: 0.0004  max mem: 57114
Epoch: [245]  [120/156]  eta: 0:00:27  lr: 0.000719  min_lr: 0.000719  loss: 2.6193 (2.6157)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0003  max mem: 57114
Epoch: [245]  [130/156]  eta: 0:00:19  lr: 0.000717  min_lr: 0.000717  loss: 2.6752 (2.6067)  weight_decay: 0.0500 (0.0500)  time: 0.7008  data: 0.0007  max mem: 57114
Epoch: [245]  [140/156]  eta: 0:00:11  lr: 0.000716  min_lr: 0.000716  loss: 2.6286 (2.6124)  weight_decay: 0.0500 (0.0500)  time: 0.6930  data: 0.0006  max mem: 57114
Epoch: [245]  [150/156]  eta: 0:00:04  lr: 0.000714  min_lr: 0.000714  loss: 2.6179 (2.6138)  weight_decay: 0.0500 (0.0500)  time: 0.6867  data: 0.0001  max mem: 57114
Epoch: [245]  [155/156]  eta: 0:00:00  lr: 0.000713  min_lr: 0.000713  loss: 2.6179 (2.6105)  weight_decay: 0.0500 (0.0500)  time: 0.6893  data: 0.0001  max mem: 57114
Epoch: [245] Total time: 0:01:55 (0.7402 s / it)
Averaged stats: lr: 0.000713  min_lr: 0.000713  loss: 2.6179 (2.6433)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5833 (0.5833)  acc1: 90.1042 (90.1042)  acc5: 98.9583 (98.9583)  time: 2.0839  data: 1.8285  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7926 (0.6977)  acc1: 82.8125 (84.6547)  acc5: 97.9167 (97.3146)  time: 0.5676  data: 0.3658  max mem: 57114
Test: Total time: 0:00:02 (0.5865 s / it)
* Acc@1 83.905 Acc@5 96.751 loss 0.784
Accuracy of the model on the 50000 test images: 83.9%
Max accuracy: 83.95%
Test:  [0/5]  eta: 0:00:11  loss: 0.5973 (0.5973)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.3988  data: 2.1554  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6912 (0.6808)  acc1: 80.7292 (79.2839)  acc5: 95.8333 (94.1177)  time: 0.6305  data: 0.4312  max mem: 57114
Test: Total time: 0:00:03 (0.6445 s / it)
* Acc@1 77.510 Acc@5 93.906 loss 0.820
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [246]  [  0/156]  eta: 0:13:00  lr: 0.000713  min_lr: 0.000713  loss: 2.3376 (2.3376)  weight_decay: 0.0500 (0.0500)  time: 5.0018  data: 4.0003  max mem: 57114
Epoch: [246]  [ 10/156]  eta: 0:02:46  lr: 0.000711  min_lr: 0.000711  loss: 2.7793 (2.5905)  weight_decay: 0.0500 (0.0500)  time: 1.1380  data: 0.3640  max mem: 57114
Epoch: [246]  [ 20/156]  eta: 0:02:07  lr: 0.000710  min_lr: 0.000710  loss: 2.8654 (2.6670)  weight_decay: 0.0500 (0.0500)  time: 0.7318  data: 0.0004  max mem: 57114
Epoch: [246]  [ 30/156]  eta: 0:01:49  lr: 0.000708  min_lr: 0.000708  loss: 2.8654 (2.5866)  weight_decay: 0.0500 (0.0500)  time: 0.7220  data: 0.0003  max mem: 57114
Epoch: [246]  [ 40/156]  eta: 0:01:36  lr: 0.000706  min_lr: 0.000706  loss: 2.5279 (2.5766)  weight_decay: 0.0500 (0.0500)  time: 0.7270  data: 0.0003  max mem: 57114
Epoch: [246]  [ 50/156]  eta: 0:01:26  lr: 0.000705  min_lr: 0.000705  loss: 2.6798 (2.6198)  weight_decay: 0.0500 (0.0500)  time: 0.7291  data: 0.0004  max mem: 57114
Epoch: [246]  [ 60/156]  eta: 0:01:16  lr: 0.000703  min_lr: 0.000703  loss: 2.8773 (2.6314)  weight_decay: 0.0500 (0.0500)  time: 0.7348  data: 0.0004  max mem: 57114
Epoch: [246]  [ 70/156]  eta: 0:01:07  lr: 0.000701  min_lr: 0.000701  loss: 2.7511 (2.6085)  weight_decay: 0.0500 (0.0500)  time: 0.7243  data: 0.0004  max mem: 57114
Epoch: [246]  [ 80/156]  eta: 0:00:59  lr: 0.000700  min_lr: 0.000700  loss: 2.7114 (2.6127)  weight_decay: 0.0500 (0.0500)  time: 0.7156  data: 0.0004  max mem: 57114
Epoch: [246]  [ 90/156]  eta: 0:00:51  lr: 0.000698  min_lr: 0.000698  loss: 2.7114 (2.6131)  weight_decay: 0.0500 (0.0500)  time: 0.7174  data: 0.0004  max mem: 57114
Epoch: [246]  [100/156]  eta: 0:00:42  lr: 0.000697  min_lr: 0.000697  loss: 2.6416 (2.6156)  weight_decay: 0.0500 (0.0500)  time: 0.7023  data: 0.0004  max mem: 57114
Epoch: [246]  [110/156]  eta: 0:00:34  lr: 0.000695  min_lr: 0.000695  loss: 2.6091 (2.6055)  weight_decay: 0.0500 (0.0500)  time: 0.6927  data: 0.0004  max mem: 57114
Epoch: [246]  [120/156]  eta: 0:00:27  lr: 0.000693  min_lr: 0.000693  loss: 2.5283 (2.5992)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0004  max mem: 57114
Epoch: [246]  [130/156]  eta: 0:00:19  lr: 0.000692  min_lr: 0.000692  loss: 2.7953 (2.6131)  weight_decay: 0.0500 (0.0500)  time: 0.7088  data: 0.0008  max mem: 57114
Epoch: [246]  [140/156]  eta: 0:00:11  lr: 0.000690  min_lr: 0.000690  loss: 2.8219 (2.6206)  weight_decay: 0.0500 (0.0500)  time: 0.6912  data: 0.0007  max mem: 57114
Epoch: [246]  [150/156]  eta: 0:00:04  lr: 0.000689  min_lr: 0.000689  loss: 2.7377 (2.6212)  weight_decay: 0.0500 (0.0500)  time: 0.6830  data: 0.0001  max mem: 57114
Epoch: [246]  [155/156]  eta: 0:00:00  lr: 0.000688  min_lr: 0.000688  loss: 2.7377 (2.6188)  weight_decay: 0.0500 (0.0500)  time: 0.6830  data: 0.0001  max mem: 57114
Epoch: [246] Total time: 0:01:55 (0.7422 s / it)
Averaged stats: lr: 0.000688  min_lr: 0.000688  loss: 2.7377 (2.6424)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6205 (0.6205)  acc1: 89.5833 (89.5833)  acc5: 98.9583 (98.9583)  time: 2.1437  data: 1.8880  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8362 (0.7453)  acc1: 82.8125 (84.3990)  acc5: 98.4375 (97.1867)  time: 0.5794  data: 0.3777  max mem: 57114
Test: Total time: 0:00:03 (0.6014 s / it)
* Acc@1 83.863 Acc@5 96.815 loss 0.825
Accuracy of the model on the 50000 test images: 83.9%
Max accuracy: 83.95%
Test:  [0/5]  eta: 0:00:11  loss: 0.5983 (0.5983)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.2811  data: 2.0376  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6917 (0.6817)  acc1: 80.7292 (79.2839)  acc5: 95.8333 (94.1177)  time: 0.6072  data: 0.4076  max mem: 57114
Test: Total time: 0:00:03 (0.6215 s / it)
* Acc@1 77.512 Acc@5 93.898 loss 0.821
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [247]  [  0/156]  eta: 0:13:33  lr: 0.000688  min_lr: 0.000688  loss: 2.5759 (2.5759)  weight_decay: 0.0500 (0.0500)  time: 5.2172  data: 4.2051  max mem: 57114
Epoch: [247]  [ 10/156]  eta: 0:02:47  lr: 0.000686  min_lr: 0.000686  loss: 2.7932 (2.6614)  weight_decay: 0.0500 (0.0500)  time: 1.1457  data: 0.3826  max mem: 57114
Epoch: [247]  [ 20/156]  eta: 0:02:08  lr: 0.000684  min_lr: 0.000684  loss: 2.5118 (2.5469)  weight_decay: 0.0500 (0.0500)  time: 0.7317  data: 0.0003  max mem: 57114
Epoch: [247]  [ 30/156]  eta: 0:01:49  lr: 0.000683  min_lr: 0.000683  loss: 2.5304 (2.5473)  weight_decay: 0.0500 (0.0500)  time: 0.7204  data: 0.0003  max mem: 57114
Epoch: [247]  [ 40/156]  eta: 0:01:37  lr: 0.000681  min_lr: 0.000681  loss: 2.5749 (2.5897)  weight_decay: 0.0500 (0.0500)  time: 0.7310  data: 0.0004  max mem: 57114
Epoch: [247]  [ 50/156]  eta: 0:01:26  lr: 0.000680  min_lr: 0.000680  loss: 2.7890 (2.6151)  weight_decay: 0.0500 (0.0500)  time: 0.7339  data: 0.0004  max mem: 57114
Epoch: [247]  [ 60/156]  eta: 0:01:16  lr: 0.000678  min_lr: 0.000678  loss: 2.6851 (2.5663)  weight_decay: 0.0500 (0.0500)  time: 0.7125  data: 0.0004  max mem: 57114
Epoch: [247]  [ 70/156]  eta: 0:01:07  lr: 0.000676  min_lr: 0.000676  loss: 2.6200 (2.5777)  weight_decay: 0.0500 (0.0500)  time: 0.7110  data: 0.0004  max mem: 57114
Epoch: [247]  [ 80/156]  eta: 0:00:59  lr: 0.000675  min_lr: 0.000675  loss: 2.6481 (2.5791)  weight_decay: 0.0500 (0.0500)  time: 0.7099  data: 0.0004  max mem: 57114
Epoch: [247]  [ 90/156]  eta: 0:00:50  lr: 0.000673  min_lr: 0.000673  loss: 2.6339 (2.5838)  weight_decay: 0.0500 (0.0500)  time: 0.7032  data: 0.0003  max mem: 57114
Epoch: [247]  [100/156]  eta: 0:00:42  lr: 0.000672  min_lr: 0.000672  loss: 2.6339 (2.5773)  weight_decay: 0.0500 (0.0500)  time: 0.7112  data: 0.0004  max mem: 57114
Epoch: [247]  [110/156]  eta: 0:00:34  lr: 0.000670  min_lr: 0.000670  loss: 2.6916 (2.5783)  weight_decay: 0.0500 (0.0500)  time: 0.7191  data: 0.0004  max mem: 57114
Epoch: [247]  [120/156]  eta: 0:00:27  lr: 0.000668  min_lr: 0.000668  loss: 2.6182 (2.5797)  weight_decay: 0.0500 (0.0500)  time: 0.7153  data: 0.0004  max mem: 57114
Epoch: [247]  [130/156]  eta: 0:00:19  lr: 0.000667  min_lr: 0.000667  loss: 2.6182 (2.5807)  weight_decay: 0.0500 (0.0500)  time: 0.6966  data: 0.0009  max mem: 57114
Epoch: [247]  [140/156]  eta: 0:00:11  lr: 0.000665  min_lr: 0.000665  loss: 2.7976 (2.5839)  weight_decay: 0.0500 (0.0500)  time: 0.6822  data: 0.0008  max mem: 57114
Epoch: [247]  [150/156]  eta: 0:00:04  lr: 0.000664  min_lr: 0.000664  loss: 2.7740 (2.5863)  weight_decay: 0.0500 (0.0500)  time: 0.6806  data: 0.0001  max mem: 57114
Epoch: [247]  [155/156]  eta: 0:00:00  lr: 0.000663  min_lr: 0.000663  loss: 2.7876 (2.5898)  weight_decay: 0.0500 (0.0500)  time: 0.6806  data: 0.0001  max mem: 57114
Epoch: [247] Total time: 0:01:55 (0.7413 s / it)
Averaged stats: lr: 0.000663  min_lr: 0.000663  loss: 2.7876 (2.6359)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5737 (0.5737)  acc1: 90.6250 (90.6250)  acc5: 98.4375 (98.4375)  time: 2.0881  data: 1.8259  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7504 (0.7021)  acc1: 82.8125 (83.6317)  acc5: 97.9167 (97.3146)  time: 0.5683  data: 0.3653  max mem: 57114
Test: Total time: 0:00:02 (0.5901 s / it)
* Acc@1 84.085 Acc@5 96.855 loss 0.789
Accuracy of the model on the 50000 test images: 84.1%
Max accuracy: 84.09%
Test:  [0/5]  eta: 0:00:10  loss: 0.5992 (0.5992)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.0147  data: 1.7714  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6922 (0.6826)  acc1: 80.7292 (79.2839)  acc5: 95.8333 (94.1177)  time: 0.5535  data: 0.3544  max mem: 57114
Test: Total time: 0:00:02 (0.5652 s / it)
* Acc@1 77.492 Acc@5 93.892 loss 0.822
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [248]  [  0/156]  eta: 0:14:26  lr: 0.000663  min_lr: 0.000663  loss: 1.8118 (1.8118)  weight_decay: 0.0500 (0.0500)  time: 5.5553  data: 3.8552  max mem: 57114
Epoch: [248]  [ 10/156]  eta: 0:02:53  lr: 0.000661  min_lr: 0.000661  loss: 2.6814 (2.5731)  weight_decay: 0.0500 (0.0500)  time: 1.1914  data: 0.3507  max mem: 57114
Epoch: [248]  [ 20/156]  eta: 0:02:12  lr: 0.000659  min_lr: 0.000659  loss: 2.6628 (2.6067)  weight_decay: 0.0500 (0.0500)  time: 0.7485  data: 0.0002  max mem: 57114
Epoch: [248]  [ 30/156]  eta: 0:01:53  lr: 0.000658  min_lr: 0.000658  loss: 2.6628 (2.5990)  weight_decay: 0.0500 (0.0500)  time: 0.7355  data: 0.0003  max mem: 57114
Epoch: [248]  [ 40/156]  eta: 0:01:39  lr: 0.000656  min_lr: 0.000656  loss: 2.6644 (2.5867)  weight_decay: 0.0500 (0.0500)  time: 0.7289  data: 0.0003  max mem: 57114
Epoch: [248]  [ 50/156]  eta: 0:01:28  lr: 0.000655  min_lr: 0.000655  loss: 2.7507 (2.6448)  weight_decay: 0.0500 (0.0500)  time: 0.7346  data: 0.0004  max mem: 57114
Epoch: [248]  [ 60/156]  eta: 0:01:18  lr: 0.000653  min_lr: 0.000653  loss: 2.8245 (2.6416)  weight_decay: 0.0500 (0.0500)  time: 0.7348  data: 0.0004  max mem: 57114
Epoch: [248]  [ 70/156]  eta: 0:01:08  lr: 0.000652  min_lr: 0.000652  loss: 2.7360 (2.6415)  weight_decay: 0.0500 (0.0500)  time: 0.7169  data: 0.0004  max mem: 57114
Epoch: [248]  [ 80/156]  eta: 0:01:00  lr: 0.000650  min_lr: 0.000650  loss: 2.5921 (2.6063)  weight_decay: 0.0500 (0.0500)  time: 0.7276  data: 0.0004  max mem: 57114
Epoch: [248]  [ 90/156]  eta: 0:00:51  lr: 0.000648  min_lr: 0.000648  loss: 2.5932 (2.6159)  weight_decay: 0.0500 (0.0500)  time: 0.7327  data: 0.0004  max mem: 57114
Epoch: [248]  [100/156]  eta: 0:00:43  lr: 0.000647  min_lr: 0.000647  loss: 2.6907 (2.6046)  weight_decay: 0.0500 (0.0500)  time: 0.7079  data: 0.0004  max mem: 57114
Epoch: [248]  [110/156]  eta: 0:00:35  lr: 0.000645  min_lr: 0.000645  loss: 2.7962 (2.6126)  weight_decay: 0.0500 (0.0500)  time: 0.6992  data: 0.0004  max mem: 57114
Epoch: [248]  [120/156]  eta: 0:00:27  lr: 0.000644  min_lr: 0.000644  loss: 2.7962 (2.6146)  weight_decay: 0.0500 (0.0500)  time: 0.6989  data: 0.0004  max mem: 57114
Epoch: [248]  [130/156]  eta: 0:00:19  lr: 0.000642  min_lr: 0.000642  loss: 2.5656 (2.5925)  weight_decay: 0.0500 (0.0500)  time: 0.6983  data: 0.0009  max mem: 57114
Epoch: [248]  [140/156]  eta: 0:00:12  lr: 0.000641  min_lr: 0.000641  loss: 2.2189 (2.5720)  weight_decay: 0.0500 (0.0500)  time: 0.6899  data: 0.0008  max mem: 57114
Epoch: [248]  [150/156]  eta: 0:00:04  lr: 0.000639  min_lr: 0.000639  loss: 2.3940 (2.5696)  weight_decay: 0.0500 (0.0500)  time: 0.6843  data: 0.0001  max mem: 57114
Epoch: [248]  [155/156]  eta: 0:00:00  lr: 0.000638  min_lr: 0.000638  loss: 2.4921 (2.5722)  weight_decay: 0.0500 (0.0500)  time: 0.6873  data: 0.0001  max mem: 57114
Epoch: [248] Total time: 0:01:56 (0.7494 s / it)
Averaged stats: lr: 0.000638  min_lr: 0.000638  loss: 2.4921 (2.6246)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5484 (0.5484)  acc1: 89.0625 (89.0625)  acc5: 98.9583 (98.9583)  time: 2.0591  data: 1.8034  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7541 (0.6870)  acc1: 83.8542 (83.5038)  acc5: 98.9583 (97.8261)  time: 0.5624  data: 0.3607  max mem: 57114
Test: Total time: 0:00:02 (0.5849 s / it)
* Acc@1 83.901 Acc@5 96.847 loss 0.766
Accuracy of the model on the 50000 test images: 83.9%
Max accuracy: 84.09%
Test:  [0/5]  eta: 0:00:11  loss: 0.6002 (0.6002)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.2413  data: 1.9978  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6929 (0.6837)  acc1: 80.7292 (79.2839)  acc5: 96.3542 (94.2455)  time: 0.5989  data: 0.3996  max mem: 57114
Test: Total time: 0:00:03 (0.6107 s / it)
* Acc@1 77.512 Acc@5 93.892 loss 0.823
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [249]  [  0/156]  eta: 0:13:23  lr: 0.000638  min_lr: 0.000638  loss: 2.7060 (2.7060)  weight_decay: 0.0500 (0.0500)  time: 5.1492  data: 4.3648  max mem: 57114
Epoch: [249]  [ 10/156]  eta: 0:02:45  lr: 0.000637  min_lr: 0.000637  loss: 2.6864 (2.4748)  weight_decay: 0.0500 (0.0500)  time: 1.1347  data: 0.3972  max mem: 57114
Epoch: [249]  [ 20/156]  eta: 0:02:07  lr: 0.000635  min_lr: 0.000635  loss: 2.6864 (2.5448)  weight_decay: 0.0500 (0.0500)  time: 0.7235  data: 0.0004  max mem: 57114
Epoch: [249]  [ 30/156]  eta: 0:01:49  lr: 0.000633  min_lr: 0.000633  loss: 2.7257 (2.5872)  weight_decay: 0.0500 (0.0500)  time: 0.7233  data: 0.0004  max mem: 57114
Epoch: [249]  [ 40/156]  eta: 0:01:36  lr: 0.000632  min_lr: 0.000632  loss: 2.7257 (2.6142)  weight_decay: 0.0500 (0.0500)  time: 0.7214  data: 0.0004  max mem: 57114
Epoch: [249]  [ 50/156]  eta: 0:01:25  lr: 0.000630  min_lr: 0.000630  loss: 2.8093 (2.6605)  weight_decay: 0.0500 (0.0500)  time: 0.7170  data: 0.0004  max mem: 57114
Epoch: [249]  [ 60/156]  eta: 0:01:16  lr: 0.000629  min_lr: 0.000629  loss: 2.8093 (2.6791)  weight_decay: 0.0500 (0.0500)  time: 0.7279  data: 0.0004  max mem: 57114
Epoch: [249]  [ 70/156]  eta: 0:01:07  lr: 0.000627  min_lr: 0.000627  loss: 2.7719 (2.6812)  weight_decay: 0.0500 (0.0500)  time: 0.7211  data: 0.0004  max mem: 57114
Epoch: [249]  [ 80/156]  eta: 0:00:58  lr: 0.000626  min_lr: 0.000626  loss: 2.6628 (2.6526)  weight_decay: 0.0500 (0.0500)  time: 0.7093  data: 0.0004  max mem: 57114
Epoch: [249]  [ 90/156]  eta: 0:00:50  lr: 0.000624  min_lr: 0.000624  loss: 2.6628 (2.6703)  weight_decay: 0.0500 (0.0500)  time: 0.7079  data: 0.0004  max mem: 57114
Epoch: [249]  [100/156]  eta: 0:00:42  lr: 0.000623  min_lr: 0.000623  loss: 2.7872 (2.6548)  weight_decay: 0.0500 (0.0500)  time: 0.7061  data: 0.0004  max mem: 57114
Epoch: [249]  [110/156]  eta: 0:00:34  lr: 0.000621  min_lr: 0.000621  loss: 2.7385 (2.6629)  weight_decay: 0.0500 (0.0500)  time: 0.7085  data: 0.0004  max mem: 57114
Epoch: [249]  [120/156]  eta: 0:00:27  lr: 0.000620  min_lr: 0.000620  loss: 2.7913 (2.6594)  weight_decay: 0.0500 (0.0500)  time: 0.7122  data: 0.0004  max mem: 57114
Epoch: [249]  [130/156]  eta: 0:00:19  lr: 0.000618  min_lr: 0.000618  loss: 2.7562 (2.6524)  weight_decay: 0.0500 (0.0500)  time: 0.7043  data: 0.0009  max mem: 57114
Epoch: [249]  [140/156]  eta: 0:00:11  lr: 0.000616  min_lr: 0.000616  loss: 2.7235 (2.6610)  weight_decay: 0.0500 (0.0500)  time: 0.6898  data: 0.0007  max mem: 57114
Epoch: [249]  [150/156]  eta: 0:00:04  lr: 0.000615  min_lr: 0.000615  loss: 2.7437 (2.6661)  weight_decay: 0.0500 (0.0500)  time: 0.6870  data: 0.0001  max mem: 57114
Epoch: [249]  [155/156]  eta: 0:00:00  lr: 0.000614  min_lr: 0.000614  loss: 2.7235 (2.6585)  weight_decay: 0.0500 (0.0500)  time: 0.6882  data: 0.0001  max mem: 57114
Epoch: [249] Total time: 0:01:55 (0.7411 s / it)
Averaged stats: lr: 0.000614  min_lr: 0.000614  loss: 2.7235 (2.6252)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5820 (0.5820)  acc1: 89.5833 (89.5833)  acc5: 98.9583 (98.9583)  time: 2.0508  data: 1.7949  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7595 (0.6953)  acc1: 83.8542 (84.0153)  acc5: 98.9583 (97.6982)  time: 0.5608  data: 0.3591  max mem: 57114
Test: Total time: 0:00:02 (0.5843 s / it)
* Acc@1 84.011 Acc@5 96.963 loss 0.782
Accuracy of the model on the 50000 test images: 84.0%
Max accuracy: 84.09%
Test:  [0/5]  eta: 0:00:11  loss: 0.6012 (0.6012)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.2537  data: 2.0099  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6935 (0.6847)  acc1: 80.7292 (79.2839)  acc5: 96.3542 (94.2455)  time: 0.6014  data: 0.4021  max mem: 57114
Test: Total time: 0:00:03 (0.6123 s / it)
* Acc@1 77.504 Acc@5 93.888 loss 0.824
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [250]  [  0/156]  eta: 0:12:46  lr: 0.000614  min_lr: 0.000614  loss: 2.9340 (2.9340)  weight_decay: 0.0500 (0.0500)  time: 4.9152  data: 4.2479  max mem: 57114
Epoch: [250]  [ 10/156]  eta: 0:02:45  lr: 0.000612  min_lr: 0.000612  loss: 2.4727 (2.5303)  weight_decay: 0.0500 (0.0500)  time: 1.1337  data: 0.3865  max mem: 57114
Epoch: [250]  [ 20/156]  eta: 0:02:06  lr: 0.000611  min_lr: 0.000611  loss: 2.3573 (2.4863)  weight_decay: 0.0500 (0.0500)  time: 0.7326  data: 0.0003  max mem: 57114
Epoch: [250]  [ 30/156]  eta: 0:01:48  lr: 0.000609  min_lr: 0.000609  loss: 2.5561 (2.5418)  weight_decay: 0.0500 (0.0500)  time: 0.7136  data: 0.0004  max mem: 57114
Epoch: [250]  [ 40/156]  eta: 0:01:36  lr: 0.000608  min_lr: 0.000608  loss: 2.6881 (2.5638)  weight_decay: 0.0500 (0.0500)  time: 0.7217  data: 0.0004  max mem: 57114
Epoch: [250]  [ 50/156]  eta: 0:01:25  lr: 0.000606  min_lr: 0.000606  loss: 2.8775 (2.6293)  weight_decay: 0.0500 (0.0500)  time: 0.7186  data: 0.0003  max mem: 57114
Epoch: [250]  [ 60/156]  eta: 0:01:16  lr: 0.000605  min_lr: 0.000605  loss: 2.8564 (2.6126)  weight_decay: 0.0500 (0.0500)  time: 0.7191  data: 0.0004  max mem: 57114
Epoch: [250]  [ 70/156]  eta: 0:01:07  lr: 0.000603  min_lr: 0.000603  loss: 2.6989 (2.6259)  weight_decay: 0.0500 (0.0500)  time: 0.7153  data: 0.0004  max mem: 57114
Epoch: [250]  [ 80/156]  eta: 0:00:58  lr: 0.000602  min_lr: 0.000602  loss: 2.7123 (2.6394)  weight_decay: 0.0500 (0.0500)  time: 0.7138  data: 0.0004  max mem: 57114
Epoch: [250]  [ 90/156]  eta: 0:00:50  lr: 0.000600  min_lr: 0.000600  loss: 2.7493 (2.6568)  weight_decay: 0.0500 (0.0500)  time: 0.7188  data: 0.0004  max mem: 57114
Epoch: [250]  [100/156]  eta: 0:00:42  lr: 0.000599  min_lr: 0.000599  loss: 2.7398 (2.6381)  weight_decay: 0.0500 (0.0500)  time: 0.7047  data: 0.0004  max mem: 57114
Epoch: [250]  [110/156]  eta: 0:00:34  lr: 0.000597  min_lr: 0.000597  loss: 2.7846 (2.6593)  weight_decay: 0.0500 (0.0500)  time: 0.7023  data: 0.0004  max mem: 57114
Epoch: [250]  [120/156]  eta: 0:00:27  lr: 0.000596  min_lr: 0.000596  loss: 2.7676 (2.6512)  weight_decay: 0.0500 (0.0500)  time: 0.7178  data: 0.0004  max mem: 57114
Epoch: [250]  [130/156]  eta: 0:00:19  lr: 0.000594  min_lr: 0.000594  loss: 2.7364 (2.6441)  weight_decay: 0.0500 (0.0500)  time: 0.7330  data: 0.0009  max mem: 57114
Epoch: [250]  [140/156]  eta: 0:00:11  lr: 0.000593  min_lr: 0.000593  loss: 2.6282 (2.6338)  weight_decay: 0.0500 (0.0500)  time: 0.7236  data: 0.0008  max mem: 57114
Epoch: [250]  [150/156]  eta: 0:00:04  lr: 0.000591  min_lr: 0.000591  loss: 2.5085 (2.6269)  weight_decay: 0.0500 (0.0500)  time: 0.7252  data: 0.0001  max mem: 57114
Epoch: [250]  [155/156]  eta: 0:00:00  lr: 0.000591  min_lr: 0.000591  loss: 2.8433 (2.6256)  weight_decay: 0.0500 (0.0500)  time: 0.7270  data: 0.0001  max mem: 57114
Epoch: [250] Total time: 0:01:56 (0.7489 s / it)
Averaged stats: lr: 0.000591  min_lr: 0.000591  loss: 2.8433 (2.6255)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6032 (0.6032)  acc1: 89.0625 (89.0625)  acc5: 98.9583 (98.9583)  time: 2.0538  data: 1.7981  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7971 (0.7386)  acc1: 83.3333 (83.6317)  acc5: 97.9167 (97.3146)  time: 0.5614  data: 0.3597  max mem: 57114
Test: Total time: 0:00:02 (0.5823 s / it)
* Acc@1 83.991 Acc@5 96.937 loss 0.813
Accuracy of the model on the 50000 test images: 84.0%
Max accuracy: 84.09%
Test:  [0/5]  eta: 0:00:11  loss: 0.6020 (0.6020)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.2794  data: 2.0358  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6939 (0.6856)  acc1: 80.7292 (79.2839)  acc5: 96.3542 (94.2455)  time: 0.6066  data: 0.4072  max mem: 57114
Test: Total time: 0:00:03 (0.6195 s / it)
* Acc@1 77.506 Acc@5 93.880 loss 0.825
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [251]  [  0/156]  eta: 0:14:14  lr: 0.000590  min_lr: 0.000590  loss: 2.0616 (2.0616)  weight_decay: 0.0500 (0.0500)  time: 5.4805  data: 4.6445  max mem: 57114
Epoch: [251]  [ 10/156]  eta: 0:02:50  lr: 0.000589  min_lr: 0.000589  loss: 2.5969 (2.5012)  weight_decay: 0.0500 (0.0500)  time: 1.1681  data: 0.4226  max mem: 57114
Epoch: [251]  [ 20/156]  eta: 0:02:09  lr: 0.000587  min_lr: 0.000587  loss: 2.7396 (2.6225)  weight_decay: 0.0500 (0.0500)  time: 0.7274  data: 0.0003  max mem: 57114
Epoch: [251]  [ 30/156]  eta: 0:01:50  lr: 0.000586  min_lr: 0.000586  loss: 2.7590 (2.6398)  weight_decay: 0.0500 (0.0500)  time: 0.7179  data: 0.0003  max mem: 57114
Epoch: [251]  [ 40/156]  eta: 0:01:37  lr: 0.000584  min_lr: 0.000584  loss: 2.6591 (2.6412)  weight_decay: 0.0500 (0.0500)  time: 0.7190  data: 0.0004  max mem: 57114
Epoch: [251]  [ 50/156]  eta: 0:01:26  lr: 0.000583  min_lr: 0.000583  loss: 2.6369 (2.6260)  weight_decay: 0.0500 (0.0500)  time: 0.7166  data: 0.0004  max mem: 57114
Epoch: [251]  [ 60/156]  eta: 0:01:16  lr: 0.000581  min_lr: 0.000581  loss: 2.6369 (2.6356)  weight_decay: 0.0500 (0.0500)  time: 0.7144  data: 0.0004  max mem: 57114
Epoch: [251]  [ 70/156]  eta: 0:01:07  lr: 0.000580  min_lr: 0.000580  loss: 2.6655 (2.6213)  weight_decay: 0.0500 (0.0500)  time: 0.7216  data: 0.0004  max mem: 57114
Epoch: [251]  [ 80/156]  eta: 0:00:59  lr: 0.000578  min_lr: 0.000578  loss: 2.7111 (2.6320)  weight_decay: 0.0500 (0.0500)  time: 0.7263  data: 0.0004  max mem: 57114
Epoch: [251]  [ 90/156]  eta: 0:00:50  lr: 0.000577  min_lr: 0.000577  loss: 2.6536 (2.6128)  weight_decay: 0.0500 (0.0500)  time: 0.7136  data: 0.0003  max mem: 57114
Epoch: [251]  [100/156]  eta: 0:00:42  lr: 0.000575  min_lr: 0.000575  loss: 2.6841 (2.6220)  weight_decay: 0.0500 (0.0500)  time: 0.7031  data: 0.0003  max mem: 57114
Epoch: [251]  [110/156]  eta: 0:00:34  lr: 0.000574  min_lr: 0.000574  loss: 2.7757 (2.6319)  weight_decay: 0.0500 (0.0500)  time: 0.6996  data: 0.0004  max mem: 57114
Epoch: [251]  [120/156]  eta: 0:00:27  lr: 0.000572  min_lr: 0.000572  loss: 2.6875 (2.6202)  weight_decay: 0.0500 (0.0500)  time: 0.6982  data: 0.0004  max mem: 57114
Epoch: [251]  [130/156]  eta: 0:00:19  lr: 0.000571  min_lr: 0.000571  loss: 2.5121 (2.6105)  weight_decay: 0.0500 (0.0500)  time: 0.7018  data: 0.0008  max mem: 57114
Epoch: [251]  [140/156]  eta: 0:00:11  lr: 0.000569  min_lr: 0.000569  loss: 2.4401 (2.5938)  weight_decay: 0.0500 (0.0500)  time: 0.6950  data: 0.0007  max mem: 57114
Epoch: [251]  [150/156]  eta: 0:00:04  lr: 0.000568  min_lr: 0.000568  loss: 2.4905 (2.6064)  weight_decay: 0.0500 (0.0500)  time: 0.6835  data: 0.0001  max mem: 57114
Epoch: [251]  [155/156]  eta: 0:00:00  lr: 0.000567  min_lr: 0.000567  loss: 2.6901 (2.6095)  weight_decay: 0.0500 (0.0500)  time: 0.7070  data: 0.0001  max mem: 57114
Epoch: [251] Total time: 0:01:56 (0.7452 s / it)
Averaged stats: lr: 0.000567  min_lr: 0.000567  loss: 2.6901 (2.6151)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:16  loss: 0.5465 (0.5465)  acc1: 91.1458 (91.1458)  acc5: 98.9583 (98.9583)  time: 3.2219  data: 2.9491  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7764 (0.7103)  acc1: 84.8958 (84.7826)  acc5: 97.9167 (97.1867)  time: 0.7976  data: 0.5899  max mem: 57114
Test: Total time: 0:00:04 (0.8446 s / it)
* Acc@1 84.147 Acc@5 96.947 loss 0.792
Accuracy of the model on the 50000 test images: 84.1%
Max accuracy: 84.15%
Test:  [0/5]  eta: 0:00:24  loss: 0.6030 (0.6030)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 4.8594  data: 4.6159  max mem: 57114
Test:  [4/5]  eta: 0:00:01  loss: 0.6943 (0.6865)  acc1: 80.7292 (79.2839)  acc5: 96.3542 (94.2455)  time: 1.1272  data: 0.9233  max mem: 57114
Test: Total time: 0:00:05 (1.1694 s / it)
* Acc@1 77.506 Acc@5 93.858 loss 0.826
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [252]  [  0/156]  eta: 0:21:09  lr: 0.000567  min_lr: 0.000567  loss: 2.5979 (2.5979)  weight_decay: 0.0500 (0.0500)  time: 8.1388  data: 6.1163  max mem: 57114
Epoch: [252]  [ 10/156]  eta: 0:03:34  lr: 0.000566  min_lr: 0.000566  loss: 2.5979 (2.6648)  weight_decay: 0.0500 (0.0500)  time: 1.4679  data: 0.5564  max mem: 57114
Epoch: [252]  [ 20/156]  eta: 0:02:35  lr: 0.000564  min_lr: 0.000564  loss: 2.5624 (2.5127)  weight_decay: 0.0500 (0.0500)  time: 0.7949  data: 0.0004  max mem: 57114
Epoch: [252]  [ 30/156]  eta: 0:02:09  lr: 0.000563  min_lr: 0.000563  loss: 2.3942 (2.5271)  weight_decay: 0.0500 (0.0500)  time: 0.7895  data: 0.0004  max mem: 57114
Epoch: [252]  [ 40/156]  eta: 0:01:55  lr: 0.000561  min_lr: 0.000561  loss: 2.6721 (2.5833)  weight_decay: 0.0500 (0.0500)  time: 0.8340  data: 0.0006  max mem: 57114
Epoch: [252]  [ 50/156]  eta: 0:01:41  lr: 0.000560  min_lr: 0.000560  loss: 2.7276 (2.6033)  weight_decay: 0.0500 (0.0500)  time: 0.8403  data: 0.0006  max mem: 57114
Epoch: [252]  [ 60/156]  eta: 0:01:28  lr: 0.000558  min_lr: 0.000558  loss: 2.6623 (2.5979)  weight_decay: 0.0500 (0.0500)  time: 0.7887  data: 0.0017  max mem: 57114
Epoch: [252]  [ 70/156]  eta: 0:01:17  lr: 0.000557  min_lr: 0.000557  loss: 2.8437 (2.6075)  weight_decay: 0.0500 (0.0500)  time: 0.7613  data: 0.0017  max mem: 57114
Epoch: [252]  [ 80/156]  eta: 0:01:06  lr: 0.000555  min_lr: 0.000555  loss: 2.8022 (2.6082)  weight_decay: 0.0500 (0.0500)  time: 0.7335  data: 0.0004  max mem: 57114
Epoch: [252]  [ 90/156]  eta: 0:00:56  lr: 0.000554  min_lr: 0.000554  loss: 2.8022 (2.6409)  weight_decay: 0.0500 (0.0500)  time: 0.7086  data: 0.0004  max mem: 57114
Epoch: [252]  [100/156]  eta: 0:00:47  lr: 0.000552  min_lr: 0.000552  loss: 2.8128 (2.6322)  weight_decay: 0.0500 (0.0500)  time: 0.7063  data: 0.0004  max mem: 57114
Epoch: [252]  [110/156]  eta: 0:00:38  lr: 0.000551  min_lr: 0.000551  loss: 2.8067 (2.6372)  weight_decay: 0.0500 (0.0500)  time: 0.7033  data: 0.0004  max mem: 57114
Epoch: [252]  [120/156]  eta: 0:00:29  lr: 0.000550  min_lr: 0.000550  loss: 2.7859 (2.6332)  weight_decay: 0.0500 (0.0500)  time: 0.7054  data: 0.0004  max mem: 57114
Epoch: [252]  [130/156]  eta: 0:00:21  lr: 0.000548  min_lr: 0.000548  loss: 2.6981 (2.6288)  weight_decay: 0.0500 (0.0500)  time: 0.7082  data: 0.0008  max mem: 57114
Epoch: [252]  [140/156]  eta: 0:00:12  lr: 0.000547  min_lr: 0.000547  loss: 2.7129 (2.6308)  weight_decay: 0.0500 (0.0500)  time: 0.6913  data: 0.0007  max mem: 57114
Epoch: [252]  [150/156]  eta: 0:00:04  lr: 0.000545  min_lr: 0.000545  loss: 2.7468 (2.6303)  weight_decay: 0.0500 (0.0500)  time: 0.6861  data: 0.0001  max mem: 57114
Epoch: [252]  [155/156]  eta: 0:00:00  lr: 0.000544  min_lr: 0.000544  loss: 2.7266 (2.6289)  weight_decay: 0.0500 (0.0500)  time: 0.6826  data: 0.0001  max mem: 57114
Epoch: [252] Total time: 0:02:03 (0.7934 s / it)
Averaged stats: lr: 0.000544  min_lr: 0.000544  loss: 2.7266 (2.6123)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6096 (0.6096)  acc1: 90.6250 (90.6250)  acc5: 98.9583 (98.9583)  time: 2.1314  data: 1.8756  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7776 (0.7038)  acc1: 84.3750 (85.0384)  acc5: 98.9583 (97.8261)  time: 0.5775  data: 0.3752  max mem: 57114
Test: Total time: 0:00:02 (0.5988 s / it)
* Acc@1 84.143 Acc@5 96.897 loss 0.816
Accuracy of the model on the 50000 test images: 84.1%
Max accuracy: 84.15%
Test:  [0/5]  eta: 0:00:11  loss: 0.6039 (0.6039)  acc1: 83.8542 (83.8542)  acc5: 97.3958 (97.3958)  time: 2.3950  data: 2.1514  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6946 (0.6873)  acc1: 80.7292 (79.1560)  acc5: 96.3542 (94.2455)  time: 0.6297  data: 0.4304  max mem: 57114
Test: Total time: 0:00:03 (0.6415 s / it)
* Acc@1 77.494 Acc@5 93.860 loss 0.827
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [253]  [  0/156]  eta: 0:13:22  lr: 0.000544  min_lr: 0.000544  loss: 2.7542 (2.7542)  weight_decay: 0.0500 (0.0500)  time: 5.1430  data: 3.8511  max mem: 57114
Epoch: [253]  [ 10/156]  eta: 0:02:44  lr: 0.000543  min_lr: 0.000543  loss: 2.7542 (2.6189)  weight_decay: 0.0500 (0.0500)  time: 1.1247  data: 0.3504  max mem: 57114
Epoch: [253]  [ 20/156]  eta: 0:02:06  lr: 0.000541  min_lr: 0.000541  loss: 2.6728 (2.6386)  weight_decay: 0.0500 (0.0500)  time: 0.7232  data: 0.0003  max mem: 57114
Epoch: [253]  [ 30/156]  eta: 0:01:49  lr: 0.000540  min_lr: 0.000540  loss: 2.6728 (2.6021)  weight_decay: 0.0500 (0.0500)  time: 0.7226  data: 0.0003  max mem: 57114
Epoch: [253]  [ 40/156]  eta: 0:01:36  lr: 0.000539  min_lr: 0.000539  loss: 2.7314 (2.6556)  weight_decay: 0.0500 (0.0500)  time: 0.7334  data: 0.0003  max mem: 57114
Epoch: [253]  [ 50/156]  eta: 0:01:26  lr: 0.000537  min_lr: 0.000537  loss: 2.7932 (2.6755)  weight_decay: 0.0500 (0.0500)  time: 0.7337  data: 0.0004  max mem: 57114
Epoch: [253]  [ 60/156]  eta: 0:01:16  lr: 0.000536  min_lr: 0.000536  loss: 2.8332 (2.6838)  weight_decay: 0.0500 (0.0500)  time: 0.7262  data: 0.0004  max mem: 57114
Epoch: [253]  [ 70/156]  eta: 0:01:07  lr: 0.000534  min_lr: 0.000534  loss: 2.8234 (2.6521)  weight_decay: 0.0500 (0.0500)  time: 0.7226  data: 0.0003  max mem: 57114
Epoch: [253]  [ 80/156]  eta: 0:00:58  lr: 0.000533  min_lr: 0.000533  loss: 2.5578 (2.6393)  weight_decay: 0.0500 (0.0500)  time: 0.7041  data: 0.0004  max mem: 57114
Epoch: [253]  [ 90/156]  eta: 0:00:50  lr: 0.000531  min_lr: 0.000531  loss: 2.5578 (2.6134)  weight_decay: 0.0500 (0.0500)  time: 0.6959  data: 0.0003  max mem: 57114
Epoch: [253]  [100/156]  eta: 0:00:42  lr: 0.000530  min_lr: 0.000530  loss: 2.6849 (2.6022)  weight_decay: 0.0500 (0.0500)  time: 0.7045  data: 0.0003  max mem: 57114
Epoch: [253]  [110/156]  eta: 0:00:34  lr: 0.000529  min_lr: 0.000529  loss: 2.7135 (2.6246)  weight_decay: 0.0500 (0.0500)  time: 0.7120  data: 0.0003  max mem: 57114
Epoch: [253]  [120/156]  eta: 0:00:27  lr: 0.000527  min_lr: 0.000527  loss: 2.7666 (2.6370)  weight_decay: 0.0500 (0.0500)  time: 0.7091  data: 0.0003  max mem: 57114
Epoch: [253]  [130/156]  eta: 0:00:19  lr: 0.000526  min_lr: 0.000526  loss: 2.8190 (2.6407)  weight_decay: 0.0500 (0.0500)  time: 0.7005  data: 0.0008  max mem: 57114
Epoch: [253]  [140/156]  eta: 0:00:11  lr: 0.000524  min_lr: 0.000524  loss: 2.7481 (2.6378)  weight_decay: 0.0500 (0.0500)  time: 0.6914  data: 0.0007  max mem: 57114
Epoch: [253]  [150/156]  eta: 0:00:04  lr: 0.000523  min_lr: 0.000523  loss: 2.6424 (2.6350)  weight_decay: 0.0500 (0.0500)  time: 0.6816  data: 0.0001  max mem: 57114
Epoch: [253]  [155/156]  eta: 0:00:00  lr: 0.000522  min_lr: 0.000522  loss: 2.6844 (2.6400)  weight_decay: 0.0500 (0.0500)  time: 0.6818  data: 0.0001  max mem: 57114
Epoch: [253] Total time: 0:01:55 (0.7406 s / it)
Averaged stats: lr: 0.000522  min_lr: 0.000522  loss: 2.6844 (2.6044)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 0.6225 (0.6225)  acc1: 90.6250 (90.6250)  acc5: 98.9583 (98.9583)  time: 2.2443  data: 1.9881  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8000 (0.7481)  acc1: 84.3750 (84.6547)  acc5: 98.9583 (97.9540)  time: 0.5995  data: 0.3977  max mem: 57114
Test: Total time: 0:00:03 (0.6244 s / it)
* Acc@1 84.195 Acc@5 96.859 loss 0.844
Accuracy of the model on the 50000 test images: 84.2%
Max accuracy: 84.20%
Test:  [0/5]  eta: 0:00:09  loss: 0.6049 (0.6049)  acc1: 83.8542 (83.8542)  acc5: 97.3958 (97.3958)  time: 1.9989  data: 1.7554  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6951 (0.6882)  acc1: 80.7292 (79.1560)  acc5: 96.3542 (94.2455)  time: 0.5504  data: 0.3511  max mem: 57114
Test: Total time: 0:00:02 (0.5614 s / it)
* Acc@1 77.498 Acc@5 93.852 loss 0.828
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [254]  [  0/156]  eta: 0:14:13  lr: 0.000522  min_lr: 0.000522  loss: 2.7188 (2.7188)  weight_decay: 0.0500 (0.0500)  time: 5.4689  data: 4.3263  max mem: 57114
Epoch: [254]  [ 10/156]  eta: 0:02:52  lr: 0.000521  min_lr: 0.000521  loss: 2.5659 (2.5981)  weight_decay: 0.0500 (0.0500)  time: 1.1818  data: 0.3935  max mem: 57114
Epoch: [254]  [ 20/156]  eta: 0:02:11  lr: 0.000519  min_lr: 0.000519  loss: 2.6097 (2.6361)  weight_decay: 0.0500 (0.0500)  time: 0.7431  data: 0.0003  max mem: 57114
Epoch: [254]  [ 30/156]  eta: 0:01:52  lr: 0.000518  min_lr: 0.000518  loss: 2.8088 (2.6316)  weight_decay: 0.0500 (0.0500)  time: 0.7349  data: 0.0003  max mem: 57114
Epoch: [254]  [ 40/156]  eta: 0:01:38  lr: 0.000516  min_lr: 0.000516  loss: 2.6722 (2.6035)  weight_decay: 0.0500 (0.0500)  time: 0.7312  data: 0.0004  max mem: 57114
Epoch: [254]  [ 50/156]  eta: 0:01:28  lr: 0.000515  min_lr: 0.000515  loss: 2.6782 (2.6177)  weight_decay: 0.0500 (0.0500)  time: 0.7342  data: 0.0004  max mem: 57114
Epoch: [254]  [ 60/156]  eta: 0:01:18  lr: 0.000513  min_lr: 0.000513  loss: 2.7681 (2.6256)  weight_decay: 0.0500 (0.0500)  time: 0.7405  data: 0.0004  max mem: 57114
Epoch: [254]  [ 70/156]  eta: 0:01:08  lr: 0.000512  min_lr: 0.000512  loss: 2.7681 (2.6251)  weight_decay: 0.0500 (0.0500)  time: 0.7277  data: 0.0004  max mem: 57114
Epoch: [254]  [ 80/156]  eta: 0:01:00  lr: 0.000511  min_lr: 0.000511  loss: 2.6802 (2.6245)  weight_decay: 0.0500 (0.0500)  time: 0.7115  data: 0.0004  max mem: 57114
Epoch: [254]  [ 90/156]  eta: 0:00:51  lr: 0.000509  min_lr: 0.000509  loss: 2.7436 (2.6271)  weight_decay: 0.0500 (0.0500)  time: 0.7113  data: 0.0004  max mem: 57114
Epoch: [254]  [100/156]  eta: 0:00:43  lr: 0.000508  min_lr: 0.000508  loss: 2.7793 (2.6467)  weight_decay: 0.0500 (0.0500)  time: 0.7139  data: 0.0004  max mem: 57114
Epoch: [254]  [110/156]  eta: 0:00:35  lr: 0.000506  min_lr: 0.000506  loss: 2.7840 (2.6467)  weight_decay: 0.0500 (0.0500)  time: 0.7065  data: 0.0004  max mem: 57114
Epoch: [254]  [120/156]  eta: 0:00:27  lr: 0.000505  min_lr: 0.000505  loss: 2.7784 (2.6429)  weight_decay: 0.0500 (0.0500)  time: 0.7003  data: 0.0004  max mem: 57114
Epoch: [254]  [130/156]  eta: 0:00:19  lr: 0.000504  min_lr: 0.000504  loss: 2.6897 (2.6392)  weight_decay: 0.0500 (0.0500)  time: 0.6973  data: 0.0008  max mem: 57114
Epoch: [254]  [140/156]  eta: 0:00:12  lr: 0.000502  min_lr: 0.000502  loss: 2.6861 (2.6277)  weight_decay: 0.0500 (0.0500)  time: 0.6923  data: 0.0007  max mem: 57114
Epoch: [254]  [150/156]  eta: 0:00:04  lr: 0.000501  min_lr: 0.000501  loss: 2.5829 (2.6215)  weight_decay: 0.0500 (0.0500)  time: 0.6873  data: 0.0001  max mem: 57114
Epoch: [254]  [155/156]  eta: 0:00:00  lr: 0.000500  min_lr: 0.000500  loss: 2.5829 (2.6123)  weight_decay: 0.0500 (0.0500)  time: 0.6943  data: 0.0001  max mem: 57114
Epoch: [254] Total time: 0:01:56 (0.7490 s / it)
Averaged stats: lr: 0.000500  min_lr: 0.000500  loss: 2.5829 (2.6016)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5571 (0.5571)  acc1: 89.5833 (89.5833)  acc5: 98.9583 (98.9583)  time: 2.1233  data: 1.8588  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7267 (0.6565)  acc1: 85.9375 (84.7826)  acc5: 98.9583 (97.9540)  time: 0.5753  data: 0.3718  max mem: 57114
Test: Total time: 0:00:02 (0.5999 s / it)
* Acc@1 84.275 Acc@5 96.881 loss 0.768
Accuracy of the model on the 50000 test images: 84.3%
Max accuracy: 84.28%
Test:  [0/5]  eta: 0:00:10  loss: 0.6058 (0.6058)  acc1: 83.8542 (83.8542)  acc5: 97.3958 (97.3958)  time: 2.1178  data: 1.8743  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6955 (0.6891)  acc1: 80.7292 (79.1560)  acc5: 96.3542 (94.2455)  time: 0.5742  data: 0.3750  max mem: 57114
Test: Total time: 0:00:02 (0.5879 s / it)
* Acc@1 77.494 Acc@5 93.828 loss 0.829
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [255]  [  0/156]  eta: 0:13:39  lr: 0.000500  min_lr: 0.000500  loss: 2.4848 (2.4848)  weight_decay: 0.0500 (0.0500)  time: 5.2532  data: 4.3519  max mem: 57114
Epoch: [255]  [ 10/156]  eta: 0:02:50  lr: 0.000499  min_lr: 0.000499  loss: 2.5768 (2.6002)  weight_decay: 0.0500 (0.0500)  time: 1.1712  data: 0.3959  max mem: 57114
Epoch: [255]  [ 20/156]  eta: 0:02:09  lr: 0.000497  min_lr: 0.000497  loss: 2.6992 (2.5569)  weight_decay: 0.0500 (0.0500)  time: 0.7367  data: 0.0004  max mem: 57114
Epoch: [255]  [ 30/156]  eta: 0:01:51  lr: 0.000496  min_lr: 0.000496  loss: 2.7822 (2.6106)  weight_decay: 0.0500 (0.0500)  time: 0.7319  data: 0.0004  max mem: 57114
Epoch: [255]  [ 40/156]  eta: 0:01:38  lr: 0.000494  min_lr: 0.000494  loss: 2.8568 (2.6298)  weight_decay: 0.0500 (0.0500)  time: 0.7463  data: 0.0004  max mem: 57114
Epoch: [255]  [ 50/156]  eta: 0:01:27  lr: 0.000493  min_lr: 0.000493  loss: 2.7546 (2.6265)  weight_decay: 0.0500 (0.0500)  time: 0.7311  data: 0.0004  max mem: 57114
Epoch: [255]  [ 60/156]  eta: 0:01:17  lr: 0.000492  min_lr: 0.000492  loss: 2.6516 (2.6150)  weight_decay: 0.0500 (0.0500)  time: 0.7310  data: 0.0004  max mem: 57114
Epoch: [255]  [ 70/156]  eta: 0:01:08  lr: 0.000490  min_lr: 0.000490  loss: 2.6455 (2.5943)  weight_decay: 0.0500 (0.0500)  time: 0.7377  data: 0.0004  max mem: 57114
Epoch: [255]  [ 80/156]  eta: 0:00:59  lr: 0.000489  min_lr: 0.000489  loss: 2.6455 (2.6009)  weight_decay: 0.0500 (0.0500)  time: 0.7200  data: 0.0004  max mem: 57114
Epoch: [255]  [ 90/156]  eta: 0:00:51  lr: 0.000488  min_lr: 0.000488  loss: 2.7531 (2.6190)  weight_decay: 0.0500 (0.0500)  time: 0.7047  data: 0.0004  max mem: 57114
Epoch: [255]  [100/156]  eta: 0:00:43  lr: 0.000486  min_lr: 0.000486  loss: 2.6991 (2.6152)  weight_decay: 0.0500 (0.0500)  time: 0.7051  data: 0.0004  max mem: 57114
Epoch: [255]  [110/156]  eta: 0:00:35  lr: 0.000485  min_lr: 0.000485  loss: 2.3781 (2.5875)  weight_decay: 0.0500 (0.0500)  time: 0.7086  data: 0.0004  max mem: 57114
Epoch: [255]  [120/156]  eta: 0:00:27  lr: 0.000483  min_lr: 0.000483  loss: 2.4055 (2.5987)  weight_decay: 0.0500 (0.0500)  time: 0.7132  data: 0.0004  max mem: 57114
Epoch: [255]  [130/156]  eta: 0:00:19  lr: 0.000482  min_lr: 0.000482  loss: 2.6495 (2.5852)  weight_decay: 0.0500 (0.0500)  time: 0.7027  data: 0.0008  max mem: 57114
Epoch: [255]  [140/156]  eta: 0:00:12  lr: 0.000481  min_lr: 0.000481  loss: 2.7044 (2.5959)  weight_decay: 0.0500 (0.0500)  time: 0.6917  data: 0.0007  max mem: 57114
Epoch: [255]  [150/156]  eta: 0:00:04  lr: 0.000479  min_lr: 0.000479  loss: 2.6431 (2.5901)  weight_decay: 0.0500 (0.0500)  time: 0.6867  data: 0.0001  max mem: 57114
Epoch: [255]  [155/156]  eta: 0:00:00  lr: 0.000479  min_lr: 0.000479  loss: 2.6431 (2.5883)  weight_decay: 0.0500 (0.0500)  time: 0.6860  data: 0.0001  max mem: 57114
Epoch: [255] Total time: 0:01:56 (0.7483 s / it)
Averaged stats: lr: 0.000479  min_lr: 0.000479  loss: 2.6431 (2.5968)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5618 (0.5618)  acc1: 90.1042 (90.1042)  acc5: 98.9583 (98.9583)  time: 2.1034  data: 1.8474  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6854 (0.6592)  acc1: 85.4167 (84.3990)  acc5: 98.4375 (97.8261)  time: 0.5713  data: 0.3695  max mem: 57114
Test: Total time: 0:00:02 (0.5953 s / it)
* Acc@1 84.305 Acc@5 96.979 loss 0.754
Accuracy of the model on the 50000 test images: 84.3%
Max accuracy: 84.31%
Test:  [0/5]  eta: 0:00:09  loss: 0.6068 (0.6068)  acc1: 83.8542 (83.8542)  acc5: 97.3958 (97.3958)  time: 1.9859  data: 1.7424  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6960 (0.6901)  acc1: 80.7292 (79.1560)  acc5: 96.3542 (94.2455)  time: 0.5478  data: 0.3486  max mem: 57114
Test: Total time: 0:00:02 (0.5576 s / it)
* Acc@1 77.486 Acc@5 93.814 loss 0.830
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [256]  [  0/156]  eta: 0:13:32  lr: 0.000479  min_lr: 0.000479  loss: 2.8983 (2.8983)  weight_decay: 0.0500 (0.0500)  time: 5.2058  data: 3.4702  max mem: 57114
Epoch: [256]  [ 10/156]  eta: 0:02:47  lr: 0.000477  min_lr: 0.000477  loss: 2.7825 (2.7121)  weight_decay: 0.0500 (0.0500)  time: 1.1485  data: 0.3157  max mem: 57114
Epoch: [256]  [ 20/156]  eta: 0:02:08  lr: 0.000476  min_lr: 0.000476  loss: 2.7553 (2.6581)  weight_decay: 0.0500 (0.0500)  time: 0.7356  data: 0.0003  max mem: 57114
Epoch: [256]  [ 30/156]  eta: 0:01:51  lr: 0.000474  min_lr: 0.000474  loss: 2.5730 (2.5884)  weight_decay: 0.0500 (0.0500)  time: 0.7452  data: 0.0003  max mem: 57114
Epoch: [256]  [ 40/156]  eta: 0:01:38  lr: 0.000473  min_lr: 0.000473  loss: 2.5730 (2.5827)  weight_decay: 0.0500 (0.0500)  time: 0.7516  data: 0.0003  max mem: 57114
Epoch: [256]  [ 50/156]  eta: 0:01:27  lr: 0.000472  min_lr: 0.000472  loss: 2.7847 (2.6370)  weight_decay: 0.0500 (0.0500)  time: 0.7376  data: 0.0004  max mem: 57114
Epoch: [256]  [ 60/156]  eta: 0:01:18  lr: 0.000470  min_lr: 0.000470  loss: 2.8034 (2.6570)  weight_decay: 0.0500 (0.0500)  time: 0.7376  data: 0.0004  max mem: 57114
Epoch: [256]  [ 70/156]  eta: 0:01:08  lr: 0.000469  min_lr: 0.000469  loss: 2.8914 (2.6527)  weight_decay: 0.0500 (0.0500)  time: 0.7323  data: 0.0004  max mem: 57114
Epoch: [256]  [ 80/156]  eta: 0:01:00  lr: 0.000468  min_lr: 0.000468  loss: 2.4917 (2.6330)  weight_decay: 0.0500 (0.0500)  time: 0.7166  data: 0.0004  max mem: 57114
Epoch: [256]  [ 90/156]  eta: 0:00:51  lr: 0.000466  min_lr: 0.000466  loss: 2.4012 (2.6056)  weight_decay: 0.0500 (0.0500)  time: 0.7070  data: 0.0004  max mem: 57114
Epoch: [256]  [100/156]  eta: 0:00:43  lr: 0.000465  min_lr: 0.000465  loss: 2.4302 (2.5959)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0003  max mem: 57114
Epoch: [256]  [110/156]  eta: 0:00:35  lr: 0.000464  min_lr: 0.000464  loss: 2.5487 (2.5869)  weight_decay: 0.0500 (0.0500)  time: 0.7082  data: 0.0004  max mem: 57114
Epoch: [256]  [120/156]  eta: 0:00:27  lr: 0.000462  min_lr: 0.000462  loss: 2.5146 (2.5828)  weight_decay: 0.0500 (0.0500)  time: 0.7057  data: 0.0004  max mem: 57114
Epoch: [256]  [130/156]  eta: 0:00:19  lr: 0.000461  min_lr: 0.000461  loss: 2.7132 (2.5917)  weight_decay: 0.0500 (0.0500)  time: 0.6989  data: 0.0008  max mem: 57114
Epoch: [256]  [140/156]  eta: 0:00:12  lr: 0.000460  min_lr: 0.000460  loss: 2.8263 (2.5961)  weight_decay: 0.0500 (0.0500)  time: 0.6940  data: 0.0007  max mem: 57114
Epoch: [256]  [150/156]  eta: 0:00:04  lr: 0.000458  min_lr: 0.000458  loss: 2.7189 (2.5916)  weight_decay: 0.0500 (0.0500)  time: 0.6929  data: 0.0001  max mem: 57114
Epoch: [256]  [155/156]  eta: 0:00:00  lr: 0.000458  min_lr: 0.000458  loss: 2.7914 (2.6014)  weight_decay: 0.0500 (0.0500)  time: 0.6878  data: 0.0001  max mem: 57114
Epoch: [256] Total time: 0:01:56 (0.7493 s / it)
Averaged stats: lr: 0.000458  min_lr: 0.000458  loss: 2.7914 (2.5953)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6207 (0.6207)  acc1: 90.6250 (90.6250)  acc5: 98.9583 (98.9583)  time: 2.1262  data: 1.8705  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7989 (0.7471)  acc1: 84.3750 (85.0384)  acc5: 98.4375 (97.5703)  time: 0.5759  data: 0.3742  max mem: 57114
Test: Total time: 0:00:02 (0.5969 s / it)
* Acc@1 84.247 Acc@5 96.877 loss 0.835
Accuracy of the model on the 50000 test images: 84.2%
Max accuracy: 84.31%
Test:  [0/5]  eta: 0:00:11  loss: 0.6075 (0.6075)  acc1: 83.8542 (83.8542)  acc5: 97.3958 (97.3958)  time: 2.3359  data: 2.0925  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6965 (0.6910)  acc1: 80.7292 (79.1560)  acc5: 96.3542 (94.2455)  time: 0.6178  data: 0.4186  max mem: 57114
Test: Total time: 0:00:03 (0.6292 s / it)
* Acc@1 77.510 Acc@5 93.804 loss 0.831
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [257]  [  0/156]  eta: 0:12:48  lr: 0.000458  min_lr: 0.000458  loss: 3.0291 (3.0291)  weight_decay: 0.0500 (0.0500)  time: 4.9233  data: 3.9156  max mem: 57114
Epoch: [257]  [ 10/156]  eta: 0:02:43  lr: 0.000456  min_lr: 0.000456  loss: 2.6930 (2.4872)  weight_decay: 0.0500 (0.0500)  time: 1.1215  data: 0.3562  max mem: 57114
Epoch: [257]  [ 20/156]  eta: 0:02:06  lr: 0.000455  min_lr: 0.000455  loss: 2.7837 (2.5718)  weight_decay: 0.0500 (0.0500)  time: 0.7321  data: 0.0003  max mem: 57114
Epoch: [257]  [ 30/156]  eta: 0:01:49  lr: 0.000454  min_lr: 0.000454  loss: 2.7837 (2.5466)  weight_decay: 0.0500 (0.0500)  time: 0.7252  data: 0.0003  max mem: 57114
Epoch: [257]  [ 40/156]  eta: 0:01:36  lr: 0.000452  min_lr: 0.000452  loss: 2.5079 (2.5455)  weight_decay: 0.0500 (0.0500)  time: 0.7347  data: 0.0003  max mem: 57114
Epoch: [257]  [ 50/156]  eta: 0:01:26  lr: 0.000451  min_lr: 0.000451  loss: 2.5580 (2.5830)  weight_decay: 0.0500 (0.0500)  time: 0.7350  data: 0.0003  max mem: 57114
Epoch: [257]  [ 60/156]  eta: 0:01:17  lr: 0.000450  min_lr: 0.000450  loss: 2.7322 (2.6007)  weight_decay: 0.0500 (0.0500)  time: 0.7397  data: 0.0004  max mem: 57114
Epoch: [257]  [ 70/156]  eta: 0:01:07  lr: 0.000448  min_lr: 0.000448  loss: 2.6685 (2.5896)  weight_decay: 0.0500 (0.0500)  time: 0.7245  data: 0.0004  max mem: 57114
Epoch: [257]  [ 80/156]  eta: 0:00:59  lr: 0.000447  min_lr: 0.000447  loss: 2.5544 (2.5971)  weight_decay: 0.0500 (0.0500)  time: 0.6955  data: 0.0003  max mem: 57114
Epoch: [257]  [ 90/156]  eta: 0:00:50  lr: 0.000446  min_lr: 0.000446  loss: 2.5277 (2.5931)  weight_decay: 0.0500 (0.0500)  time: 0.7091  data: 0.0004  max mem: 57114
Epoch: [257]  [100/156]  eta: 0:00:42  lr: 0.000444  min_lr: 0.000444  loss: 2.4014 (2.5776)  weight_decay: 0.0500 (0.0500)  time: 0.7155  data: 0.0004  max mem: 57114
Epoch: [257]  [110/156]  eta: 0:00:34  lr: 0.000443  min_lr: 0.000443  loss: 2.5911 (2.5857)  weight_decay: 0.0500 (0.0500)  time: 0.7078  data: 0.0004  max mem: 57114
Epoch: [257]  [120/156]  eta: 0:00:27  lr: 0.000442  min_lr: 0.000442  loss: 2.6836 (2.5770)  weight_decay: 0.0500 (0.0500)  time: 0.7244  data: 0.0004  max mem: 57114
Epoch: [257]  [130/156]  eta: 0:00:19  lr: 0.000440  min_lr: 0.000440  loss: 2.5418 (2.5803)  weight_decay: 0.0500 (0.0500)  time: 0.7389  data: 0.0010  max mem: 57114
Epoch: [257]  [140/156]  eta: 0:00:12  lr: 0.000439  min_lr: 0.000439  loss: 2.6598 (2.5872)  weight_decay: 0.0500 (0.0500)  time: 0.7109  data: 0.0008  max mem: 57114
Epoch: [257]  [150/156]  eta: 0:00:04  lr: 0.000438  min_lr: 0.000438  loss: 2.5339 (2.5735)  weight_decay: 0.0500 (0.0500)  time: 0.6807  data: 0.0001  max mem: 57114
Epoch: [257]  [155/156]  eta: 0:00:00  lr: 0.000437  min_lr: 0.000437  loss: 2.5339 (2.5717)  weight_decay: 0.0500 (0.0500)  time: 0.6852  data: 0.0001  max mem: 57114
Epoch: [257] Total time: 0:01:56 (0.7477 s / it)
Averaged stats: lr: 0.000437  min_lr: 0.000437  loss: 2.5339 (2.5868)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5512 (0.5512)  acc1: 90.6250 (90.6250)  acc5: 98.9583 (98.9583)  time: 2.0411  data: 1.7853  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7548 (0.6864)  acc1: 84.3750 (84.5269)  acc5: 98.4375 (97.6982)  time: 0.5588  data: 0.3571  max mem: 57114
Test: Total time: 0:00:02 (0.5822 s / it)
* Acc@1 84.337 Acc@5 96.999 loss 0.773
Accuracy of the model on the 50000 test images: 84.3%
Max accuracy: 84.34%
Test:  [0/5]  eta: 0:00:10  loss: 0.6084 (0.6084)  acc1: 83.8542 (83.8542)  acc5: 97.3958 (97.3958)  time: 2.0051  data: 1.7618  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6969 (0.6919)  acc1: 80.7292 (79.1560)  acc5: 96.3542 (94.2455)  time: 0.5517  data: 0.3525  max mem: 57114
Test: Total time: 0:00:02 (0.5650 s / it)
* Acc@1 77.496 Acc@5 93.810 loss 0.832
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [258]  [  0/156]  eta: 0:13:50  lr: 0.000437  min_lr: 0.000437  loss: 2.9091 (2.9091)  weight_decay: 0.0500 (0.0500)  time: 5.3248  data: 4.1412  max mem: 57114
Epoch: [258]  [ 10/156]  eta: 0:02:53  lr: 0.000436  min_lr: 0.000436  loss: 2.7499 (2.6950)  weight_decay: 0.0500 (0.0500)  time: 1.1906  data: 0.3767  max mem: 57114
Epoch: [258]  [ 20/156]  eta: 0:02:12  lr: 0.000434  min_lr: 0.000434  loss: 2.7499 (2.7730)  weight_decay: 0.0500 (0.0500)  time: 0.7554  data: 0.0003  max mem: 57114
Epoch: [258]  [ 30/156]  eta: 0:01:54  lr: 0.000433  min_lr: 0.000433  loss: 2.7351 (2.7780)  weight_decay: 0.0500 (0.0500)  time: 0.7504  data: 0.0003  max mem: 57114
Epoch: [258]  [ 40/156]  eta: 0:01:40  lr: 0.000432  min_lr: 0.000432  loss: 2.7716 (2.7901)  weight_decay: 0.0500 (0.0500)  time: 0.7628  data: 0.0004  max mem: 57114
Epoch: [258]  [ 50/156]  eta: 0:01:29  lr: 0.000430  min_lr: 0.000430  loss: 2.8460 (2.7857)  weight_decay: 0.0500 (0.0500)  time: 0.7513  data: 0.0004  max mem: 57114
Epoch: [258]  [ 60/156]  eta: 0:01:20  lr: 0.000429  min_lr: 0.000429  loss: 2.8259 (2.7806)  weight_decay: 0.0500 (0.0500)  time: 0.7670  data: 0.0004  max mem: 57114
Epoch: [258]  [ 70/156]  eta: 0:01:11  lr: 0.000428  min_lr: 0.000428  loss: 2.7404 (2.7484)  weight_decay: 0.0500 (0.0500)  time: 0.7890  data: 0.0004  max mem: 57114
Epoch: [258]  [ 80/156]  eta: 0:01:02  lr: 0.000427  min_lr: 0.000427  loss: 2.5072 (2.7163)  weight_decay: 0.0500 (0.0500)  time: 0.7993  data: 0.0004  max mem: 57114
Epoch: [258]  [ 90/156]  eta: 0:00:54  lr: 0.000425  min_lr: 0.000425  loss: 2.5304 (2.6833)  weight_decay: 0.0500 (0.0500)  time: 0.7993  data: 0.0004  max mem: 57114
Epoch: [258]  [100/156]  eta: 0:00:45  lr: 0.000424  min_lr: 0.000424  loss: 2.5304 (2.6680)  weight_decay: 0.0500 (0.0500)  time: 0.7757  data: 0.0004  max mem: 57114
Epoch: [258]  [110/156]  eta: 0:00:37  lr: 0.000423  min_lr: 0.000423  loss: 2.6236 (2.6604)  weight_decay: 0.0500 (0.0500)  time: 0.7621  data: 0.0004  max mem: 57114
Epoch: [258]  [120/156]  eta: 0:00:29  lr: 0.000421  min_lr: 0.000421  loss: 2.6801 (2.6574)  weight_decay: 0.0500 (0.0500)  time: 0.7597  data: 0.0004  max mem: 57114
Epoch: [258]  [130/156]  eta: 0:00:20  lr: 0.000420  min_lr: 0.000420  loss: 2.6864 (2.6583)  weight_decay: 0.0500 (0.0500)  time: 0.7600  data: 0.0009  max mem: 57114
Epoch: [258]  [140/156]  eta: 0:00:12  lr: 0.000419  min_lr: 0.000419  loss: 2.7849 (2.6644)  weight_decay: 0.0500 (0.0500)  time: 0.7498  data: 0.0008  max mem: 57114
Epoch: [258]  [150/156]  eta: 0:00:04  lr: 0.000418  min_lr: 0.000418  loss: 2.7281 (2.6533)  weight_decay: 0.0500 (0.0500)  time: 0.7220  data: 0.0001  max mem: 57114
Epoch: [258]  [155/156]  eta: 0:00:00  lr: 0.000417  min_lr: 0.000417  loss: 2.6447 (2.6438)  weight_decay: 0.0500 (0.0500)  time: 0.7079  data: 0.0001  max mem: 57114
Epoch: [258] Total time: 0:02:03 (0.7920 s / it)
Averaged stats: lr: 0.000417  min_lr: 0.000417  loss: 2.6447 (2.5861)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5550 (0.5550)  acc1: 91.1458 (91.1458)  acc5: 98.4375 (98.4375)  time: 2.1352  data: 1.8793  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7019 (0.6704)  acc1: 85.4167 (84.9105)  acc5: 98.4375 (97.5703)  time: 0.5777  data: 0.3759  max mem: 57114
Test: Total time: 0:00:02 (0.5987 s / it)
* Acc@1 84.457 Acc@5 96.935 loss 0.761
Accuracy of the model on the 50000 test images: 84.5%
Max accuracy: 84.46%
Test:  [0/5]  eta: 0:00:10  loss: 0.6093 (0.6093)  acc1: 83.8542 (83.8542)  acc5: 97.3958 (97.3958)  time: 2.0808  data: 1.8374  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6976 (0.6928)  acc1: 80.7292 (79.1560)  acc5: 96.3542 (94.1177)  time: 0.5668  data: 0.3676  max mem: 57114
Test: Total time: 0:00:02 (0.5789 s / it)
* Acc@1 77.492 Acc@5 93.814 loss 0.833
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [259]  [  0/156]  eta: 0:15:49  lr: 0.000417  min_lr: 0.000417  loss: 2.4172 (2.4172)  weight_decay: 0.0500 (0.0500)  time: 6.0859  data: 5.2331  max mem: 57114
Epoch: [259]  [ 10/156]  eta: 0:03:00  lr: 0.000415  min_lr: 0.000415  loss: 2.5586 (2.5532)  weight_decay: 0.0500 (0.0500)  time: 1.2333  data: 0.4760  max mem: 57114
Epoch: [259]  [ 20/156]  eta: 0:02:15  lr: 0.000414  min_lr: 0.000414  loss: 2.6571 (2.6128)  weight_decay: 0.0500 (0.0500)  time: 0.7437  data: 0.0003  max mem: 57114
Epoch: [259]  [ 30/156]  eta: 0:01:54  lr: 0.000413  min_lr: 0.000413  loss: 2.6482 (2.5308)  weight_decay: 0.0500 (0.0500)  time: 0.7360  data: 0.0004  max mem: 57114
Epoch: [259]  [ 40/156]  eta: 0:01:40  lr: 0.000412  min_lr: 0.000412  loss: 2.5767 (2.5624)  weight_decay: 0.0500 (0.0500)  time: 0.7295  data: 0.0004  max mem: 57114
Epoch: [259]  [ 50/156]  eta: 0:01:28  lr: 0.000410  min_lr: 0.000410  loss: 2.7950 (2.6071)  weight_decay: 0.0500 (0.0500)  time: 0.7212  data: 0.0004  max mem: 57114
Epoch: [259]  [ 60/156]  eta: 0:01:18  lr: 0.000409  min_lr: 0.000409  loss: 2.7977 (2.6315)  weight_decay: 0.0500 (0.0500)  time: 0.7269  data: 0.0004  max mem: 57114
Epoch: [259]  [ 70/156]  eta: 0:01:09  lr: 0.000408  min_lr: 0.000408  loss: 2.7403 (2.6276)  weight_decay: 0.0500 (0.0500)  time: 0.7418  data: 0.0004  max mem: 57114
Epoch: [259]  [ 80/156]  eta: 0:01:00  lr: 0.000407  min_lr: 0.000407  loss: 2.8201 (2.6570)  weight_decay: 0.0500 (0.0500)  time: 0.7408  data: 0.0004  max mem: 57114
Epoch: [259]  [ 90/156]  eta: 0:00:52  lr: 0.000405  min_lr: 0.000405  loss: 2.8128 (2.6475)  weight_decay: 0.0500 (0.0500)  time: 0.7234  data: 0.0004  max mem: 57114
Epoch: [259]  [100/156]  eta: 0:00:43  lr: 0.000404  min_lr: 0.000404  loss: 2.6223 (2.6420)  weight_decay: 0.0500 (0.0500)  time: 0.7104  data: 0.0004  max mem: 57114
Epoch: [259]  [110/156]  eta: 0:00:35  lr: 0.000403  min_lr: 0.000403  loss: 2.7185 (2.6556)  weight_decay: 0.0500 (0.0500)  time: 0.7005  data: 0.0003  max mem: 57114
Epoch: [259]  [120/156]  eta: 0:00:27  lr: 0.000402  min_lr: 0.000402  loss: 2.8441 (2.6672)  weight_decay: 0.0500 (0.0500)  time: 0.6928  data: 0.0003  max mem: 57114
Epoch: [259]  [130/156]  eta: 0:00:19  lr: 0.000400  min_lr: 0.000400  loss: 2.5830 (2.6493)  weight_decay: 0.0500 (0.0500)  time: 0.6934  data: 0.0007  max mem: 57114
Epoch: [259]  [140/156]  eta: 0:00:12  lr: 0.000399  min_lr: 0.000399  loss: 2.4169 (2.6304)  weight_decay: 0.0500 (0.0500)  time: 0.6897  data: 0.0006  max mem: 57114
Epoch: [259]  [150/156]  eta: 0:00:04  lr: 0.000398  min_lr: 0.000398  loss: 2.4140 (2.6157)  weight_decay: 0.0500 (0.0500)  time: 0.6925  data: 0.0001  max mem: 57114
Epoch: [259]  [155/156]  eta: 0:00:00  lr: 0.000397  min_lr: 0.000397  loss: 2.3520 (2.6009)  weight_decay: 0.0500 (0.0500)  time: 0.6918  data: 0.0001  max mem: 57114
Epoch: [259] Total time: 0:01:57 (0.7532 s / it)
Averaged stats: lr: 0.000397  min_lr: 0.000397  loss: 2.3520 (2.5761)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 0.5068 (0.5068)  acc1: 91.6667 (91.6667)  acc5: 98.4375 (98.4375)  time: 2.2027  data: 1.9471  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6846 (0.6386)  acc1: 84.8958 (84.7826)  acc5: 97.9167 (97.1867)  time: 0.5911  data: 0.3895  max mem: 57114
Test: Total time: 0:00:03 (0.6132 s / it)
* Acc@1 84.355 Acc@5 96.963 loss 0.732
Accuracy of the model on the 50000 test images: 84.4%
Max accuracy: 84.46%
Test:  [0/5]  eta: 0:00:11  loss: 0.6102 (0.6102)  acc1: 83.8542 (83.8542)  acc5: 97.3958 (97.3958)  time: 2.3941  data: 2.1506  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6983 (0.6938)  acc1: 80.7292 (79.1560)  acc5: 96.3542 (94.1177)  time: 0.6295  data: 0.4302  max mem: 57114
Test: Total time: 0:00:03 (0.6399 s / it)
* Acc@1 77.504 Acc@5 93.816 loss 0.834
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [260]  [  0/156]  eta: 0:15:14  lr: 0.000397  min_lr: 0.000397  loss: 2.4455 (2.4455)  weight_decay: 0.0500 (0.0500)  time: 5.8598  data: 4.9529  max mem: 57114
Epoch: [260]  [ 10/156]  eta: 0:02:53  lr: 0.000396  min_lr: 0.000396  loss: 2.6752 (2.7494)  weight_decay: 0.0500 (0.0500)  time: 1.1915  data: 0.4506  max mem: 57114
Epoch: [260]  [ 20/156]  eta: 0:02:10  lr: 0.000395  min_lr: 0.000395  loss: 2.6752 (2.6734)  weight_decay: 0.0500 (0.0500)  time: 0.7138  data: 0.0005  max mem: 57114
Epoch: [260]  [ 30/156]  eta: 0:01:50  lr: 0.000393  min_lr: 0.000393  loss: 2.5365 (2.6287)  weight_decay: 0.0500 (0.0500)  time: 0.7025  data: 0.0006  max mem: 57114
Epoch: [260]  [ 40/156]  eta: 0:01:37  lr: 0.000392  min_lr: 0.000392  loss: 2.7129 (2.6026)  weight_decay: 0.0500 (0.0500)  time: 0.7170  data: 0.0005  max mem: 57114
Epoch: [260]  [ 50/156]  eta: 0:01:26  lr: 0.000391  min_lr: 0.000391  loss: 2.5612 (2.5656)  weight_decay: 0.0500 (0.0500)  time: 0.7181  data: 0.0004  max mem: 57114
Epoch: [260]  [ 60/156]  eta: 0:01:17  lr: 0.000390  min_lr: 0.000390  loss: 2.6271 (2.5676)  weight_decay: 0.0500 (0.0500)  time: 0.7262  data: 0.0004  max mem: 57114
Epoch: [260]  [ 70/156]  eta: 0:01:08  lr: 0.000388  min_lr: 0.000388  loss: 2.6507 (2.5632)  weight_decay: 0.0500 (0.0500)  time: 0.7338  data: 0.0004  max mem: 57114
Epoch: [260]  [ 80/156]  eta: 0:00:59  lr: 0.000387  min_lr: 0.000387  loss: 2.7169 (2.5582)  weight_decay: 0.0500 (0.0500)  time: 0.7168  data: 0.0004  max mem: 57114
Epoch: [260]  [ 90/156]  eta: 0:00:50  lr: 0.000386  min_lr: 0.000386  loss: 2.7719 (2.5612)  weight_decay: 0.0500 (0.0500)  time: 0.7008  data: 0.0004  max mem: 57114
Epoch: [260]  [100/156]  eta: 0:00:42  lr: 0.000385  min_lr: 0.000385  loss: 2.6595 (2.5584)  weight_decay: 0.0500 (0.0500)  time: 0.6925  data: 0.0004  max mem: 57114
Epoch: [260]  [110/156]  eta: 0:00:34  lr: 0.000383  min_lr: 0.000383  loss: 2.6168 (2.5740)  weight_decay: 0.0500 (0.0500)  time: 0.6972  data: 0.0004  max mem: 57114
Epoch: [260]  [120/156]  eta: 0:00:27  lr: 0.000382  min_lr: 0.000382  loss: 2.6096 (2.5693)  weight_decay: 0.0500 (0.0500)  time: 0.6982  data: 0.0004  max mem: 57114
Epoch: [260]  [130/156]  eta: 0:00:19  lr: 0.000381  min_lr: 0.000381  loss: 2.4402 (2.5531)  weight_decay: 0.0500 (0.0500)  time: 0.7248  data: 0.0011  max mem: 57114
Epoch: [260]  [140/156]  eta: 0:00:11  lr: 0.000380  min_lr: 0.000380  loss: 2.4985 (2.5487)  weight_decay: 0.0500 (0.0500)  time: 0.7239  data: 0.0010  max mem: 57114
Epoch: [260]  [150/156]  eta: 0:00:04  lr: 0.000379  min_lr: 0.000379  loss: 2.6165 (2.5482)  weight_decay: 0.0500 (0.0500)  time: 0.6895  data: 0.0001  max mem: 57114
Epoch: [260]  [155/156]  eta: 0:00:00  lr: 0.000378  min_lr: 0.000378  loss: 2.6599 (2.5540)  weight_decay: 0.0500 (0.0500)  time: 0.6831  data: 0.0001  max mem: 57114
Epoch: [260] Total time: 0:01:56 (0.7452 s / it)
Averaged stats: lr: 0.000378  min_lr: 0.000378  loss: 2.6599 (2.5774)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5918 (0.5918)  acc1: 90.6250 (90.6250)  acc5: 98.4375 (98.4375)  time: 2.1144  data: 1.8581  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7799 (0.7209)  acc1: 85.4167 (84.1432)  acc5: 98.4375 (97.4425)  time: 0.5736  data: 0.3717  max mem: 57114
Test: Total time: 0:00:02 (0.5965 s / it)
* Acc@1 84.309 Acc@5 96.949 loss 0.805
Accuracy of the model on the 50000 test images: 84.3%
Max accuracy: 84.46%
Test:  [0/5]  eta: 0:00:11  loss: 0.6110 (0.6110)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.3155  data: 2.0719  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6988 (0.6945)  acc1: 80.7292 (79.2839)  acc5: 96.3542 (94.1177)  time: 0.6138  data: 0.4145  max mem: 57114
Test: Total time: 0:00:03 (0.6267 s / it)
* Acc@1 77.492 Acc@5 93.802 loss 0.835
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [261]  [  0/156]  eta: 0:12:59  lr: 0.000378  min_lr: 0.000378  loss: 2.9627 (2.9627)  weight_decay: 0.0500 (0.0500)  time: 4.9952  data: 3.5486  max mem: 57114
Epoch: [261]  [ 10/156]  eta: 0:02:43  lr: 0.000377  min_lr: 0.000377  loss: 2.8482 (2.7978)  weight_decay: 0.0500 (0.0500)  time: 1.1231  data: 0.3229  max mem: 57114
Epoch: [261]  [ 20/156]  eta: 0:02:05  lr: 0.000375  min_lr: 0.000375  loss: 2.5630 (2.6267)  weight_decay: 0.0500 (0.0500)  time: 0.7194  data: 0.0003  max mem: 57114
Epoch: [261]  [ 30/156]  eta: 0:01:47  lr: 0.000374  min_lr: 0.000374  loss: 2.5630 (2.6030)  weight_decay: 0.0500 (0.0500)  time: 0.7054  data: 0.0003  max mem: 57114
Epoch: [261]  [ 40/156]  eta: 0:01:35  lr: 0.000373  min_lr: 0.000373  loss: 2.6856 (2.5716)  weight_decay: 0.0500 (0.0500)  time: 0.7156  data: 0.0003  max mem: 57114
Epoch: [261]  [ 50/156]  eta: 0:01:26  lr: 0.000372  min_lr: 0.000372  loss: 2.6392 (2.5770)  weight_decay: 0.0500 (0.0500)  time: 0.7468  data: 0.0004  max mem: 57114
Epoch: [261]  [ 60/156]  eta: 0:01:17  lr: 0.000371  min_lr: 0.000371  loss: 2.5051 (2.5496)  weight_decay: 0.0500 (0.0500)  time: 0.7803  data: 0.0004  max mem: 57114
Epoch: [261]  [ 70/156]  eta: 0:01:08  lr: 0.000369  min_lr: 0.000369  loss: 2.3328 (2.5428)  weight_decay: 0.0500 (0.0500)  time: 0.7654  data: 0.0004  max mem: 57114
Epoch: [261]  [ 80/156]  eta: 0:01:00  lr: 0.000368  min_lr: 0.000368  loss: 2.4943 (2.5409)  weight_decay: 0.0500 (0.0500)  time: 0.7366  data: 0.0004  max mem: 57114
Epoch: [261]  [ 90/156]  eta: 0:00:51  lr: 0.000367  min_lr: 0.000367  loss: 2.5539 (2.5350)  weight_decay: 0.0500 (0.0500)  time: 0.7462  data: 0.0004  max mem: 57114
Epoch: [261]  [100/156]  eta: 0:00:44  lr: 0.000366  min_lr: 0.000366  loss: 2.6163 (2.5371)  weight_decay: 0.0500 (0.0500)  time: 0.7796  data: 0.0005  max mem: 57114
Epoch: [261]  [110/156]  eta: 0:00:36  lr: 0.000365  min_lr: 0.000365  loss: 2.5189 (2.5219)  weight_decay: 0.0500 (0.0500)  time: 0.7848  data: 0.0010  max mem: 57114
Epoch: [261]  [120/156]  eta: 0:00:28  lr: 0.000363  min_lr: 0.000363  loss: 2.3393 (2.5118)  weight_decay: 0.0500 (0.0500)  time: 0.7638  data: 0.0011  max mem: 57114
Epoch: [261]  [130/156]  eta: 0:00:20  lr: 0.000362  min_lr: 0.000362  loss: 2.4034 (2.5067)  weight_decay: 0.0500 (0.0500)  time: 0.7554  data: 0.0013  max mem: 57114
Epoch: [261]  [140/156]  eta: 0:00:12  lr: 0.000361  min_lr: 0.000361  loss: 2.5631 (2.5116)  weight_decay: 0.0500 (0.0500)  time: 0.7515  data: 0.0009  max mem: 57114
Epoch: [261]  [150/156]  eta: 0:00:04  lr: 0.000360  min_lr: 0.000360  loss: 2.7038 (2.5184)  weight_decay: 0.0500 (0.0500)  time: 0.7384  data: 0.0001  max mem: 57114
Epoch: [261]  [155/156]  eta: 0:00:00  lr: 0.000359  min_lr: 0.000359  loss: 2.8579 (2.5304)  weight_decay: 0.0500 (0.0500)  time: 0.7244  data: 0.0001  max mem: 57114
Epoch: [261] Total time: 0:02:01 (0.7770 s / it)
Averaged stats: lr: 0.000359  min_lr: 0.000359  loss: 2.8579 (2.5716)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:16  loss: 0.6648 (0.6648)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 3.2814  data: 3.0190  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8232 (0.7794)  acc1: 83.8542 (84.3990)  acc5: 98.4375 (97.6982)  time: 0.8072  data: 0.6039  max mem: 57114
Test: Total time: 0:00:04 (0.8284 s / it)
* Acc@1 84.293 Acc@5 96.963 loss 0.865
Accuracy of the model on the 50000 test images: 84.3%
Max accuracy: 84.46%
Test:  [0/5]  eta: 0:00:11  loss: 0.6118 (0.6118)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.2768  data: 2.0331  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6996 (0.6954)  acc1: 80.7292 (79.1560)  acc5: 96.3542 (94.1177)  time: 0.6061  data: 0.4067  max mem: 57114
Test: Total time: 0:00:03 (0.6179 s / it)
* Acc@1 77.488 Acc@5 93.798 loss 0.836
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [262]  [  0/156]  eta: 0:15:57  lr: 0.000359  min_lr: 0.000359  loss: 2.0623 (2.0623)  weight_decay: 0.0500 (0.0500)  time: 6.1383  data: 4.7778  max mem: 57114
Epoch: [262]  [ 10/156]  eta: 0:03:01  lr: 0.000358  min_lr: 0.000358  loss: 2.5234 (2.4944)  weight_decay: 0.0500 (0.0500)  time: 1.2427  data: 0.4347  max mem: 57114
Epoch: [262]  [ 20/156]  eta: 0:02:14  lr: 0.000357  min_lr: 0.000357  loss: 2.4687 (2.4161)  weight_decay: 0.0500 (0.0500)  time: 0.7343  data: 0.0004  max mem: 57114
Epoch: [262]  [ 30/156]  eta: 0:01:54  lr: 0.000355  min_lr: 0.000355  loss: 2.6503 (2.5195)  weight_decay: 0.0500 (0.0500)  time: 0.7271  data: 0.0004  max mem: 57114
Epoch: [262]  [ 40/156]  eta: 0:01:40  lr: 0.000354  min_lr: 0.000354  loss: 2.7258 (2.5015)  weight_decay: 0.0500 (0.0500)  time: 0.7421  data: 0.0004  max mem: 57114
Epoch: [262]  [ 50/156]  eta: 0:01:29  lr: 0.000353  min_lr: 0.000353  loss: 2.5803 (2.5415)  weight_decay: 0.0500 (0.0500)  time: 0.7322  data: 0.0004  max mem: 57114
Epoch: [262]  [ 60/156]  eta: 0:01:18  lr: 0.000352  min_lr: 0.000352  loss: 2.6079 (2.5396)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0004  max mem: 57114
Epoch: [262]  [ 70/156]  eta: 0:01:09  lr: 0.000351  min_lr: 0.000351  loss: 2.4161 (2.5184)  weight_decay: 0.0500 (0.0500)  time: 0.7117  data: 0.0004  max mem: 57114
Epoch: [262]  [ 80/156]  eta: 0:01:00  lr: 0.000350  min_lr: 0.000350  loss: 2.5355 (2.5166)  weight_decay: 0.0500 (0.0500)  time: 0.7182  data: 0.0004  max mem: 57114
Epoch: [262]  [ 90/156]  eta: 0:00:51  lr: 0.000348  min_lr: 0.000348  loss: 2.5942 (2.5156)  weight_decay: 0.0500 (0.0500)  time: 0.7097  data: 0.0003  max mem: 57114
Epoch: [262]  [100/156]  eta: 0:00:43  lr: 0.000347  min_lr: 0.000347  loss: 2.5942 (2.5244)  weight_decay: 0.0500 (0.0500)  time: 0.7048  data: 0.0004  max mem: 57114
Epoch: [262]  [110/156]  eta: 0:00:35  lr: 0.000346  min_lr: 0.000346  loss: 2.6452 (2.5385)  weight_decay: 0.0500 (0.0500)  time: 0.7046  data: 0.0004  max mem: 57114
Epoch: [262]  [120/156]  eta: 0:00:27  lr: 0.000345  min_lr: 0.000345  loss: 2.6452 (2.5405)  weight_decay: 0.0500 (0.0500)  time: 0.7082  data: 0.0004  max mem: 57114
Epoch: [262]  [130/156]  eta: 0:00:19  lr: 0.000344  min_lr: 0.000344  loss: 2.6340 (2.5419)  weight_decay: 0.0500 (0.0500)  time: 0.7000  data: 0.0008  max mem: 57114
Epoch: [262]  [140/156]  eta: 0:00:12  lr: 0.000343  min_lr: 0.000343  loss: 2.5951 (2.5403)  weight_decay: 0.0500 (0.0500)  time: 0.6931  data: 0.0007  max mem: 57114
Epoch: [262]  [150/156]  eta: 0:00:04  lr: 0.000341  min_lr: 0.000341  loss: 2.5923 (2.5476)  weight_decay: 0.0500 (0.0500)  time: 0.6922  data: 0.0001  max mem: 57114
Epoch: [262]  [155/156]  eta: 0:00:00  lr: 0.000341  min_lr: 0.000341  loss: 2.6234 (2.5534)  weight_decay: 0.0500 (0.0500)  time: 0.6908  data: 0.0001  max mem: 57114
Epoch: [262] Total time: 0:01:57 (0.7502 s / it)
Averaged stats: lr: 0.000341  min_lr: 0.000341  loss: 2.6234 (2.5675)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6187 (0.6187)  acc1: 91.6667 (91.6667)  acc5: 98.4375 (98.4375)  time: 2.0410  data: 1.7853  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7903 (0.7366)  acc1: 83.8542 (84.6547)  acc5: 98.4375 (97.1867)  time: 0.5589  data: 0.3571  max mem: 57114
Test: Total time: 0:00:02 (0.5792 s / it)
* Acc@1 84.437 Acc@5 96.975 loss 0.828
Accuracy of the model on the 50000 test images: 84.4%
Max accuracy: 84.46%
Test:  [0/5]  eta: 0:00:11  loss: 0.6126 (0.6126)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.2865  data: 2.0429  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7001 (0.6962)  acc1: 80.7292 (79.1560)  acc5: 96.3542 (93.9898)  time: 0.6080  data: 0.4087  max mem: 57114
Test: Total time: 0:00:03 (0.6191 s / it)
* Acc@1 77.478 Acc@5 93.788 loss 0.837
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [263]  [  0/156]  eta: 0:15:22  lr: 0.000341  min_lr: 0.000341  loss: 2.7037 (2.7037)  weight_decay: 0.0500 (0.0500)  time: 5.9120  data: 4.7539  max mem: 57114
Epoch: [263]  [ 10/156]  eta: 0:02:57  lr: 0.000340  min_lr: 0.000340  loss: 2.6838 (2.5865)  weight_decay: 0.0500 (0.0500)  time: 1.2169  data: 0.4324  max mem: 57114
Epoch: [263]  [ 20/156]  eta: 0:02:13  lr: 0.000338  min_lr: 0.000338  loss: 2.6838 (2.6382)  weight_decay: 0.0500 (0.0500)  time: 0.7327  data: 0.0003  max mem: 57114
Epoch: [263]  [ 30/156]  eta: 0:01:54  lr: 0.000337  min_lr: 0.000337  loss: 2.6412 (2.5630)  weight_decay: 0.0500 (0.0500)  time: 0.7336  data: 0.0003  max mem: 57114
Epoch: [263]  [ 40/156]  eta: 0:01:39  lr: 0.000336  min_lr: 0.000336  loss: 2.5524 (2.5922)  weight_decay: 0.0500 (0.0500)  time: 0.7307  data: 0.0003  max mem: 57114
Epoch: [263]  [ 50/156]  eta: 0:01:27  lr: 0.000335  min_lr: 0.000335  loss: 2.7656 (2.6153)  weight_decay: 0.0500 (0.0500)  time: 0.7085  data: 0.0004  max mem: 57114
Epoch: [263]  [ 60/156]  eta: 0:01:17  lr: 0.000334  min_lr: 0.000334  loss: 2.7514 (2.5989)  weight_decay: 0.0500 (0.0500)  time: 0.7131  data: 0.0004  max mem: 57114
Epoch: [263]  [ 70/156]  eta: 0:01:08  lr: 0.000333  min_lr: 0.000333  loss: 2.7514 (2.6181)  weight_decay: 0.0500 (0.0500)  time: 0.7251  data: 0.0004  max mem: 57114
Epoch: [263]  [ 80/156]  eta: 0:00:59  lr: 0.000331  min_lr: 0.000331  loss: 2.6816 (2.5917)  weight_decay: 0.0500 (0.0500)  time: 0.7159  data: 0.0004  max mem: 57114
Epoch: [263]  [ 90/156]  eta: 0:00:51  lr: 0.000330  min_lr: 0.000330  loss: 2.7875 (2.6101)  weight_decay: 0.0500 (0.0500)  time: 0.7086  data: 0.0004  max mem: 57114
Epoch: [263]  [100/156]  eta: 0:00:43  lr: 0.000329  min_lr: 0.000329  loss: 2.8539 (2.6165)  weight_decay: 0.0500 (0.0500)  time: 0.7109  data: 0.0003  max mem: 57114
Epoch: [263]  [110/156]  eta: 0:00:35  lr: 0.000328  min_lr: 0.000328  loss: 2.7711 (2.6318)  weight_decay: 0.0500 (0.0500)  time: 0.7027  data: 0.0003  max mem: 57114
Epoch: [263]  [120/156]  eta: 0:00:27  lr: 0.000327  min_lr: 0.000327  loss: 2.7443 (2.6390)  weight_decay: 0.0500 (0.0500)  time: 0.6998  data: 0.0004  max mem: 57114
Epoch: [263]  [130/156]  eta: 0:00:19  lr: 0.000326  min_lr: 0.000326  loss: 2.7440 (2.6380)  weight_decay: 0.0500 (0.0500)  time: 0.7003  data: 0.0009  max mem: 57114
Epoch: [263]  [140/156]  eta: 0:00:12  lr: 0.000325  min_lr: 0.000325  loss: 2.6894 (2.6319)  weight_decay: 0.0500 (0.0500)  time: 0.6898  data: 0.0007  max mem: 57114
Epoch: [263]  [150/156]  eta: 0:00:04  lr: 0.000324  min_lr: 0.000324  loss: 2.5892 (2.6252)  weight_decay: 0.0500 (0.0500)  time: 0.6823  data: 0.0001  max mem: 57114
Epoch: [263]  [155/156]  eta: 0:00:00  lr: 0.000323  min_lr: 0.000323  loss: 2.5383 (2.6269)  weight_decay: 0.0500 (0.0500)  time: 0.6873  data: 0.0001  max mem: 57114
Epoch: [263] Total time: 0:01:56 (0.7467 s / it)
Averaged stats: lr: 0.000323  min_lr: 0.000323  loss: 2.5383 (2.5639)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5826 (0.5826)  acc1: 91.1458 (91.1458)  acc5: 98.4375 (98.4375)  time: 2.0218  data: 1.7661  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7475 (0.7091)  acc1: 85.4167 (84.5269)  acc5: 97.9167 (97.8261)  time: 0.5551  data: 0.3533  max mem: 57114
Test: Total time: 0:00:02 (0.5771 s / it)
* Acc@1 84.471 Acc@5 96.999 loss 0.804
Accuracy of the model on the 50000 test images: 84.5%
Max accuracy: 84.47%
Test:  [0/5]  eta: 0:00:09  loss: 0.6134 (0.6134)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 1.9982  data: 1.7540  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7005 (0.6970)  acc1: 80.7292 (79.1560)  acc5: 96.3542 (94.1177)  time: 0.5503  data: 0.3509  max mem: 57114
Test: Total time: 0:00:02 (0.5617 s / it)
* Acc@1 77.482 Acc@5 93.784 loss 0.838
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [264]  [  0/156]  eta: 0:14:08  lr: 0.000323  min_lr: 0.000323  loss: 1.9886 (1.9886)  weight_decay: 0.0500 (0.0500)  time: 5.4400  data: 4.6566  max mem: 57114
Epoch: [264]  [ 10/156]  eta: 0:02:50  lr: 0.000322  min_lr: 0.000322  loss: 2.5858 (2.4804)  weight_decay: 0.0500 (0.0500)  time: 1.1687  data: 0.4237  max mem: 57114
Epoch: [264]  [ 20/156]  eta: 0:02:09  lr: 0.000321  min_lr: 0.000321  loss: 2.6470 (2.5429)  weight_decay: 0.0500 (0.0500)  time: 0.7314  data: 0.0003  max mem: 57114
Epoch: [264]  [ 30/156]  eta: 0:01:50  lr: 0.000319  min_lr: 0.000319  loss: 2.6891 (2.5782)  weight_decay: 0.0500 (0.0500)  time: 0.7174  data: 0.0003  max mem: 57114
Epoch: [264]  [ 40/156]  eta: 0:01:37  lr: 0.000318  min_lr: 0.000318  loss: 2.5269 (2.5363)  weight_decay: 0.0500 (0.0500)  time: 0.7283  data: 0.0004  max mem: 57114
Epoch: [264]  [ 50/156]  eta: 0:01:27  lr: 0.000317  min_lr: 0.000317  loss: 2.3410 (2.5071)  weight_decay: 0.0500 (0.0500)  time: 0.7336  data: 0.0004  max mem: 57114
Epoch: [264]  [ 60/156]  eta: 0:01:17  lr: 0.000316  min_lr: 0.000316  loss: 2.4997 (2.5008)  weight_decay: 0.0500 (0.0500)  time: 0.7283  data: 0.0004  max mem: 57114
Epoch: [264]  [ 70/156]  eta: 0:01:08  lr: 0.000315  min_lr: 0.000315  loss: 2.5345 (2.4886)  weight_decay: 0.0500 (0.0500)  time: 0.7230  data: 0.0004  max mem: 57114
Epoch: [264]  [ 80/156]  eta: 0:00:59  lr: 0.000314  min_lr: 0.000314  loss: 2.5669 (2.4852)  weight_decay: 0.0500 (0.0500)  time: 0.7174  data: 0.0004  max mem: 57114
Epoch: [264]  [ 90/156]  eta: 0:00:51  lr: 0.000313  min_lr: 0.000313  loss: 2.6153 (2.5085)  weight_decay: 0.0500 (0.0500)  time: 0.7122  data: 0.0004  max mem: 57114
Epoch: [264]  [100/156]  eta: 0:00:43  lr: 0.000312  min_lr: 0.000312  loss: 2.7477 (2.5350)  weight_decay: 0.0500 (0.0500)  time: 0.7012  data: 0.0004  max mem: 57114
Epoch: [264]  [110/156]  eta: 0:00:35  lr: 0.000311  min_lr: 0.000311  loss: 2.7912 (2.5385)  weight_decay: 0.0500 (0.0500)  time: 0.7051  data: 0.0003  max mem: 57114
Epoch: [264]  [120/156]  eta: 0:00:27  lr: 0.000309  min_lr: 0.000309  loss: 2.6958 (2.5354)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0003  max mem: 57114
Epoch: [264]  [130/156]  eta: 0:00:19  lr: 0.000308  min_lr: 0.000308  loss: 2.6424 (2.5371)  weight_decay: 0.0500 (0.0500)  time: 0.6980  data: 0.0008  max mem: 57114
Epoch: [264]  [140/156]  eta: 0:00:11  lr: 0.000307  min_lr: 0.000307  loss: 2.6258 (2.5347)  weight_decay: 0.0500 (0.0500)  time: 0.6833  data: 0.0007  max mem: 57114
Epoch: [264]  [150/156]  eta: 0:00:04  lr: 0.000306  min_lr: 0.000306  loss: 2.6390 (2.5329)  weight_decay: 0.0500 (0.0500)  time: 0.6813  data: 0.0001  max mem: 57114
Epoch: [264]  [155/156]  eta: 0:00:00  lr: 0.000306  min_lr: 0.000306  loss: 2.6634 (2.5233)  weight_decay: 0.0500 (0.0500)  time: 0.6807  data: 0.0001  max mem: 57114
Epoch: [264] Total time: 0:01:56 (0.7438 s / it)
Averaged stats: lr: 0.000306  min_lr: 0.000306  loss: 2.6634 (2.5600)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 0.5162 (0.5162)  acc1: 90.6250 (90.6250)  acc5: 98.9583 (98.9583)  time: 2.2847  data: 2.0285  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7062 (0.6333)  acc1: 84.8958 (85.2941)  acc5: 98.4375 (97.4425)  time: 0.6076  data: 0.4058  max mem: 57114
Test: Total time: 0:00:03 (0.6290 s / it)
* Acc@1 84.489 Acc@5 97.025 loss 0.729
Accuracy of the model on the 50000 test images: 84.5%
Max accuracy: 84.49%
Test:  [0/5]  eta: 0:00:10  loss: 0.6141 (0.6141)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.0086  data: 1.7651  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7009 (0.6976)  acc1: 80.7292 (79.1560)  acc5: 96.3542 (93.9898)  time: 0.5523  data: 0.3531  max mem: 57114
Test: Total time: 0:00:02 (0.5652 s / it)
* Acc@1 77.490 Acc@5 93.784 loss 0.839
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [265]  [  0/156]  eta: 0:15:49  lr: 0.000305  min_lr: 0.000305  loss: 1.7179 (1.7179)  weight_decay: 0.0500 (0.0500)  time: 6.0865  data: 4.0951  max mem: 57114
Epoch: [265]  [ 10/156]  eta: 0:02:57  lr: 0.000304  min_lr: 0.000304  loss: 2.5420 (2.4362)  weight_decay: 0.0500 (0.0500)  time: 1.2184  data: 0.3726  max mem: 57114
Epoch: [265]  [ 20/156]  eta: 0:02:13  lr: 0.000303  min_lr: 0.000303  loss: 2.6301 (2.4677)  weight_decay: 0.0500 (0.0500)  time: 0.7296  data: 0.0003  max mem: 57114
Epoch: [265]  [ 30/156]  eta: 0:01:54  lr: 0.000302  min_lr: 0.000302  loss: 2.6496 (2.4989)  weight_decay: 0.0500 (0.0500)  time: 0.7410  data: 0.0003  max mem: 57114
Epoch: [265]  [ 40/156]  eta: 0:01:41  lr: 0.000301  min_lr: 0.000301  loss: 2.7701 (2.5295)  weight_decay: 0.0500 (0.0500)  time: 0.7553  data: 0.0004  max mem: 57114
Epoch: [265]  [ 50/156]  eta: 0:01:29  lr: 0.000300  min_lr: 0.000300  loss: 2.6469 (2.5393)  weight_decay: 0.0500 (0.0500)  time: 0.7404  data: 0.0004  max mem: 57114
Epoch: [265]  [ 60/156]  eta: 0:01:19  lr: 0.000299  min_lr: 0.000299  loss: 2.5542 (2.5449)  weight_decay: 0.0500 (0.0500)  time: 0.7293  data: 0.0004  max mem: 57114
Epoch: [265]  [ 70/156]  eta: 0:01:09  lr: 0.000298  min_lr: 0.000298  loss: 2.5542 (2.5393)  weight_decay: 0.0500 (0.0500)  time: 0.7233  data: 0.0004  max mem: 57114
Epoch: [265]  [ 80/156]  eta: 0:01:00  lr: 0.000297  min_lr: 0.000297  loss: 2.5031 (2.5237)  weight_decay: 0.0500 (0.0500)  time: 0.7123  data: 0.0004  max mem: 57114
Epoch: [265]  [ 90/156]  eta: 0:00:51  lr: 0.000296  min_lr: 0.000296  loss: 2.5418 (2.5377)  weight_decay: 0.0500 (0.0500)  time: 0.7065  data: 0.0004  max mem: 57114
Epoch: [265]  [100/156]  eta: 0:00:43  lr: 0.000295  min_lr: 0.000295  loss: 2.4837 (2.5255)  weight_decay: 0.0500 (0.0500)  time: 0.7079  data: 0.0004  max mem: 57114
Epoch: [265]  [110/156]  eta: 0:00:35  lr: 0.000293  min_lr: 0.000293  loss: 2.5465 (2.5263)  weight_decay: 0.0500 (0.0500)  time: 0.7123  data: 0.0004  max mem: 57114
Epoch: [265]  [120/156]  eta: 0:00:27  lr: 0.000292  min_lr: 0.000292  loss: 2.6863 (2.5320)  weight_decay: 0.0500 (0.0500)  time: 0.7047  data: 0.0004  max mem: 57114
Epoch: [265]  [130/156]  eta: 0:00:19  lr: 0.000291  min_lr: 0.000291  loss: 2.6635 (2.5229)  weight_decay: 0.0500 (0.0500)  time: 0.6922  data: 0.0009  max mem: 57114
Epoch: [265]  [140/156]  eta: 0:00:12  lr: 0.000290  min_lr: 0.000290  loss: 2.6054 (2.5224)  weight_decay: 0.0500 (0.0500)  time: 0.6863  data: 0.0007  max mem: 57114
Epoch: [265]  [150/156]  eta: 0:00:04  lr: 0.000289  min_lr: 0.000289  loss: 2.6054 (2.5248)  weight_decay: 0.0500 (0.0500)  time: 0.6840  data: 0.0001  max mem: 57114
Epoch: [265]  [155/156]  eta: 0:00:00  lr: 0.000289  min_lr: 0.000289  loss: 2.5729 (2.5265)  weight_decay: 0.0500 (0.0500)  time: 0.6842  data: 0.0001  max mem: 57114
Epoch: [265] Total time: 0:01:57 (0.7508 s / it)
Averaged stats: lr: 0.000289  min_lr: 0.000289  loss: 2.5729 (2.5621)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5742 (0.5742)  acc1: 91.6667 (91.6667)  acc5: 98.9583 (98.9583)  time: 2.0726  data: 1.8168  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7442 (0.6870)  acc1: 83.8542 (84.3990)  acc5: 98.4375 (97.5703)  time: 0.5653  data: 0.3634  max mem: 57114
Test: Total time: 0:00:02 (0.5845 s / it)
* Acc@1 84.395 Acc@5 96.993 loss 0.776
Accuracy of the model on the 50000 test images: 84.4%
Max accuracy: 84.49%
Test:  [0/5]  eta: 0:00:11  loss: 0.6147 (0.6147)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.3226  data: 2.0791  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7013 (0.6982)  acc1: 80.7292 (79.1560)  acc5: 96.3542 (93.9898)  time: 0.6151  data: 0.4159  max mem: 57114
Test: Total time: 0:00:03 (0.6276 s / it)
* Acc@1 77.492 Acc@5 93.782 loss 0.839
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [266]  [  0/156]  eta: 0:14:10  lr: 0.000289  min_lr: 0.000289  loss: 1.6959 (1.6959)  weight_decay: 0.0500 (0.0500)  time: 5.4547  data: 4.4488  max mem: 57114
Epoch: [266]  [ 10/156]  eta: 0:02:50  lr: 0.000287  min_lr: 0.000287  loss: 2.4515 (2.3374)  weight_decay: 0.0500 (0.0500)  time: 1.1674  data: 0.4048  max mem: 57114
Epoch: [266]  [ 20/156]  eta: 0:02:11  lr: 0.000286  min_lr: 0.000286  loss: 2.6115 (2.4954)  weight_decay: 0.0500 (0.0500)  time: 0.7429  data: 0.0004  max mem: 57114
Epoch: [266]  [ 30/156]  eta: 0:01:51  lr: 0.000285  min_lr: 0.000285  loss: 2.6115 (2.4464)  weight_decay: 0.0500 (0.0500)  time: 0.7356  data: 0.0003  max mem: 57114
Epoch: [266]  [ 40/156]  eta: 0:01:38  lr: 0.000284  min_lr: 0.000284  loss: 2.6038 (2.4662)  weight_decay: 0.0500 (0.0500)  time: 0.7170  data: 0.0003  max mem: 57114
Epoch: [266]  [ 50/156]  eta: 0:01:26  lr: 0.000283  min_lr: 0.000283  loss: 2.6140 (2.4679)  weight_decay: 0.0500 (0.0500)  time: 0.7090  data: 0.0004  max mem: 57114
Epoch: [266]  [ 60/156]  eta: 0:01:17  lr: 0.000282  min_lr: 0.000282  loss: 2.5531 (2.4658)  weight_decay: 0.0500 (0.0500)  time: 0.7209  data: 0.0004  max mem: 57114
Epoch: [266]  [ 70/156]  eta: 0:01:08  lr: 0.000281  min_lr: 0.000281  loss: 2.6665 (2.5133)  weight_decay: 0.0500 (0.0500)  time: 0.7239  data: 0.0004  max mem: 57114
Epoch: [266]  [ 80/156]  eta: 0:00:59  lr: 0.000280  min_lr: 0.000280  loss: 2.7105 (2.5264)  weight_decay: 0.0500 (0.0500)  time: 0.7110  data: 0.0004  max mem: 57114
Epoch: [266]  [ 90/156]  eta: 0:00:51  lr: 0.000279  min_lr: 0.000279  loss: 2.7105 (2.5532)  weight_decay: 0.0500 (0.0500)  time: 0.7190  data: 0.0004  max mem: 57114
Epoch: [266]  [100/156]  eta: 0:00:43  lr: 0.000278  min_lr: 0.000278  loss: 2.7296 (2.5593)  weight_decay: 0.0500 (0.0500)  time: 0.7188  data: 0.0004  max mem: 57114
Epoch: [266]  [110/156]  eta: 0:00:35  lr: 0.000277  min_lr: 0.000277  loss: 2.4902 (2.5419)  weight_decay: 0.0500 (0.0500)  time: 0.6982  data: 0.0004  max mem: 57114
Epoch: [266]  [120/156]  eta: 0:00:27  lr: 0.000276  min_lr: 0.000276  loss: 2.4368 (2.5467)  weight_decay: 0.0500 (0.0500)  time: 0.6948  data: 0.0004  max mem: 57114
Epoch: [266]  [130/156]  eta: 0:00:19  lr: 0.000275  min_lr: 0.000275  loss: 2.6208 (2.5521)  weight_decay: 0.0500 (0.0500)  time: 0.6945  data: 0.0009  max mem: 57114
Epoch: [266]  [140/156]  eta: 0:00:11  lr: 0.000274  min_lr: 0.000274  loss: 2.5477 (2.5380)  weight_decay: 0.0500 (0.0500)  time: 0.6815  data: 0.0007  max mem: 57114
Epoch: [266]  [150/156]  eta: 0:00:04  lr: 0.000273  min_lr: 0.000273  loss: 2.3826 (2.5234)  weight_decay: 0.0500 (0.0500)  time: 0.6832  data: 0.0001  max mem: 57114
Epoch: [266]  [155/156]  eta: 0:00:00  lr: 0.000272  min_lr: 0.000272  loss: 2.4860 (2.5281)  weight_decay: 0.0500 (0.0500)  time: 0.6861  data: 0.0001  max mem: 57114
Epoch: [266] Total time: 0:01:55 (0.7425 s / it)
Averaged stats: lr: 0.000272  min_lr: 0.000272  loss: 2.4860 (2.5544)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5489 (0.5489)  acc1: 90.6250 (90.6250)  acc5: 98.4375 (98.4375)  time: 2.0815  data: 1.8242  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7506 (0.6820)  acc1: 84.8958 (84.7826)  acc5: 98.4375 (97.3146)  time: 0.5669  data: 0.3649  max mem: 57114
Test: Total time: 0:00:02 (0.5868 s / it)
* Acc@1 84.451 Acc@5 96.981 loss 0.768
Accuracy of the model on the 50000 test images: 84.5%
Max accuracy: 84.49%
Test:  [0/5]  eta: 0:00:11  loss: 0.6154 (0.6154)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.3162  data: 2.0727  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7017 (0.6988)  acc1: 80.7292 (79.1560)  acc5: 96.3542 (93.9898)  time: 0.6139  data: 0.4146  max mem: 57114
Test: Total time: 0:00:03 (0.6277 s / it)
* Acc@1 77.524 Acc@5 93.776 loss 0.840
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [267]  [  0/156]  eta: 0:14:14  lr: 0.000272  min_lr: 0.000272  loss: 2.5445 (2.5445)  weight_decay: 0.0500 (0.0500)  time: 5.4786  data: 3.7973  max mem: 57114
Epoch: [267]  [ 10/156]  eta: 0:02:50  lr: 0.000271  min_lr: 0.000271  loss: 2.5445 (2.4688)  weight_decay: 0.0500 (0.0500)  time: 1.1655  data: 0.3455  max mem: 57114
Epoch: [267]  [ 20/156]  eta: 0:02:10  lr: 0.000270  min_lr: 0.000270  loss: 2.6471 (2.4878)  weight_decay: 0.0500 (0.0500)  time: 0.7347  data: 0.0004  max mem: 57114
Epoch: [267]  [ 30/156]  eta: 0:01:51  lr: 0.000269  min_lr: 0.000269  loss: 2.7584 (2.4949)  weight_decay: 0.0500 (0.0500)  time: 0.7285  data: 0.0004  max mem: 57114
Epoch: [267]  [ 40/156]  eta: 0:01:37  lr: 0.000268  min_lr: 0.000268  loss: 2.5784 (2.5036)  weight_decay: 0.0500 (0.0500)  time: 0.7152  data: 0.0003  max mem: 57114
Epoch: [267]  [ 50/156]  eta: 0:01:26  lr: 0.000267  min_lr: 0.000267  loss: 2.5328 (2.4796)  weight_decay: 0.0500 (0.0500)  time: 0.7149  data: 0.0003  max mem: 57114
Epoch: [267]  [ 60/156]  eta: 0:01:16  lr: 0.000266  min_lr: 0.000266  loss: 2.6151 (2.5005)  weight_decay: 0.0500 (0.0500)  time: 0.7216  data: 0.0004  max mem: 57114
Epoch: [267]  [ 70/156]  eta: 0:01:07  lr: 0.000265  min_lr: 0.000265  loss: 2.6151 (2.4976)  weight_decay: 0.0500 (0.0500)  time: 0.7166  data: 0.0004  max mem: 57114
Epoch: [267]  [ 80/156]  eta: 0:00:59  lr: 0.000264  min_lr: 0.000264  loss: 2.6178 (2.4936)  weight_decay: 0.0500 (0.0500)  time: 0.7161  data: 0.0004  max mem: 57114
Epoch: [267]  [ 90/156]  eta: 0:00:50  lr: 0.000263  min_lr: 0.000263  loss: 2.6230 (2.5062)  weight_decay: 0.0500 (0.0500)  time: 0.7129  data: 0.0004  max mem: 57114
Epoch: [267]  [100/156]  eta: 0:00:42  lr: 0.000262  min_lr: 0.000262  loss: 2.6104 (2.5015)  weight_decay: 0.0500 (0.0500)  time: 0.7033  data: 0.0004  max mem: 57114
Epoch: [267]  [110/156]  eta: 0:00:34  lr: 0.000261  min_lr: 0.000261  loss: 2.6104 (2.5101)  weight_decay: 0.0500 (0.0500)  time: 0.6991  data: 0.0004  max mem: 57114
Epoch: [267]  [120/156]  eta: 0:00:27  lr: 0.000260  min_lr: 0.000260  loss: 2.5249 (2.4986)  weight_decay: 0.0500 (0.0500)  time: 0.7066  data: 0.0004  max mem: 57114
Epoch: [267]  [130/156]  eta: 0:00:19  lr: 0.000259  min_lr: 0.000259  loss: 2.5424 (2.5018)  weight_decay: 0.0500 (0.0500)  time: 0.7051  data: 0.0008  max mem: 57114
Epoch: [267]  [140/156]  eta: 0:00:11  lr: 0.000258  min_lr: 0.000258  loss: 2.6770 (2.5083)  weight_decay: 0.0500 (0.0500)  time: 0.6938  data: 0.0007  max mem: 57114
Epoch: [267]  [150/156]  eta: 0:00:04  lr: 0.000257  min_lr: 0.000257  loss: 2.6493 (2.5093)  weight_decay: 0.0500 (0.0500)  time: 0.6874  data: 0.0001  max mem: 57114
Epoch: [267]  [155/156]  eta: 0:00:00  lr: 0.000256  min_lr: 0.000256  loss: 2.5402 (2.5006)  weight_decay: 0.0500 (0.0500)  time: 0.6867  data: 0.0001  max mem: 57114
Epoch: [267] Total time: 0:01:55 (0.7426 s / it)
Averaged stats: lr: 0.000256  min_lr: 0.000256  loss: 2.5402 (2.5527)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5416 (0.5416)  acc1: 90.6250 (90.6250)  acc5: 98.9583 (98.9583)  time: 2.1738  data: 1.9179  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7355 (0.6741)  acc1: 85.4167 (84.6547)  acc5: 98.4375 (97.8261)  time: 0.5854  data: 0.3836  max mem: 57114
Test: Total time: 0:00:03 (0.6105 s / it)
* Acc@1 84.335 Acc@5 97.041 loss 0.762
Accuracy of the model on the 50000 test images: 84.3%
Max accuracy: 84.49%
Test:  [0/5]  eta: 0:00:11  loss: 0.6162 (0.6162)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.2851  data: 2.0416  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7021 (0.6995)  acc1: 81.2500 (79.1560)  acc5: 96.3542 (94.1177)  time: 0.6077  data: 0.4084  max mem: 57114
Test: Total time: 0:00:03 (0.6202 s / it)
* Acc@1 77.518 Acc@5 93.776 loss 0.841
Accuracy of the model EMA on 50000 test images: 77.5%
Epoch: [268]  [  0/156]  eta: 0:15:53  lr: 0.000256  min_lr: 0.000256  loss: 2.9535 (2.9535)  weight_decay: 0.0500 (0.0500)  time: 6.1126  data: 5.1222  max mem: 57114
Epoch: [268]  [ 10/156]  eta: 0:02:57  lr: 0.000255  min_lr: 0.000255  loss: 2.5231 (2.4558)  weight_decay: 0.0500 (0.0500)  time: 1.2170  data: 0.4660  max mem: 57114
Epoch: [268]  [ 20/156]  eta: 0:02:13  lr: 0.000254  min_lr: 0.000254  loss: 2.4281 (2.3977)  weight_decay: 0.0500 (0.0500)  time: 0.7227  data: 0.0004  max mem: 57114
Epoch: [268]  [ 30/156]  eta: 0:01:52  lr: 0.000253  min_lr: 0.000253  loss: 2.1784 (2.3573)  weight_decay: 0.0500 (0.0500)  time: 0.7184  data: 0.0003  max mem: 57114
Epoch: [268]  [ 40/156]  eta: 0:01:39  lr: 0.000252  min_lr: 0.000252  loss: 2.5282 (2.4053)  weight_decay: 0.0500 (0.0500)  time: 0.7216  data: 0.0004  max mem: 57114
Epoch: [268]  [ 50/156]  eta: 0:01:27  lr: 0.000251  min_lr: 0.000251  loss: 2.7161 (2.4475)  weight_decay: 0.0500 (0.0500)  time: 0.7252  data: 0.0004  max mem: 57114
Epoch: [268]  [ 60/156]  eta: 0:01:17  lr: 0.000250  min_lr: 0.000250  loss: 2.7136 (2.4500)  weight_decay: 0.0500 (0.0500)  time: 0.7198  data: 0.0004  max mem: 57114
Epoch: [268]  [ 70/156]  eta: 0:01:08  lr: 0.000249  min_lr: 0.000249  loss: 2.6740 (2.4865)  weight_decay: 0.0500 (0.0500)  time: 0.7155  data: 0.0004  max mem: 57114
Epoch: [268]  [ 80/156]  eta: 0:00:59  lr: 0.000248  min_lr: 0.000248  loss: 2.6981 (2.4944)  weight_decay: 0.0500 (0.0500)  time: 0.7153  data: 0.0004  max mem: 57114
Epoch: [268]  [ 90/156]  eta: 0:00:51  lr: 0.000247  min_lr: 0.000247  loss: 2.6981 (2.5062)  weight_decay: 0.0500 (0.0500)  time: 0.7104  data: 0.0004  max mem: 57114
Epoch: [268]  [100/156]  eta: 0:00:43  lr: 0.000246  min_lr: 0.000246  loss: 2.7829 (2.5344)  weight_decay: 0.0500 (0.0500)  time: 0.7103  data: 0.0004  max mem: 57114
Epoch: [268]  [110/156]  eta: 0:00:35  lr: 0.000245  min_lr: 0.000245  loss: 2.7285 (2.5242)  weight_decay: 0.0500 (0.0500)  time: 0.7142  data: 0.0004  max mem: 57114
Epoch: [268]  [120/156]  eta: 0:00:27  lr: 0.000244  min_lr: 0.000244  loss: 2.5016 (2.5128)  weight_decay: 0.0500 (0.0500)  time: 0.7123  data: 0.0004  max mem: 57114
Epoch: [268]  [130/156]  eta: 0:00:19  lr: 0.000243  min_lr: 0.000243  loss: 2.5016 (2.5051)  weight_decay: 0.0500 (0.0500)  time: 0.7015  data: 0.0008  max mem: 57114
Epoch: [268]  [140/156]  eta: 0:00:12  lr: 0.000242  min_lr: 0.000242  loss: 2.5331 (2.5099)  weight_decay: 0.0500 (0.0500)  time: 0.6934  data: 0.0007  max mem: 57114
Epoch: [268]  [150/156]  eta: 0:00:04  lr: 0.000241  min_lr: 0.000241  loss: 2.5961 (2.5253)  weight_decay: 0.0500 (0.0500)  time: 0.6859  data: 0.0001  max mem: 57114
Epoch: [268]  [155/156]  eta: 0:00:00  lr: 0.000241  min_lr: 0.000241  loss: 2.7041 (2.5353)  weight_decay: 0.0500 (0.0500)  time: 0.6843  data: 0.0001  max mem: 57114
Epoch: [268] Total time: 0:01:56 (0.7470 s / it)
Averaged stats: lr: 0.000241  min_lr: 0.000241  loss: 2.7041 (2.5517)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6582 (0.6582)  acc1: 90.6250 (90.6250)  acc5: 98.9583 (98.9583)  time: 2.0312  data: 1.7747  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8220 (0.7710)  acc1: 83.8542 (84.5269)  acc5: 97.9167 (97.1867)  time: 0.5569  data: 0.3550  max mem: 57114
Test: Total time: 0:00:02 (0.5770 s / it)
* Acc@1 84.413 Acc@5 96.999 loss 0.850
Accuracy of the model on the 50000 test images: 84.4%
Max accuracy: 84.49%
Test:  [0/5]  eta: 0:00:11  loss: 0.6167 (0.6167)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.2867  data: 2.0432  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7024 (0.6999)  acc1: 81.2500 (79.1560)  acc5: 96.3542 (94.1177)  time: 0.6081  data: 0.4087  max mem: 57114
Test: Total time: 0:00:03 (0.6597 s / it)
* Acc@1 77.530 Acc@5 93.774 loss 0.841
Accuracy of the model EMA on 50000 test images: 77.5%
Max EMA accuracy: 77.53%
Epoch: [269]  [  0/156]  eta: 0:12:25  lr: 0.000240  min_lr: 0.000240  loss: 2.8958 (2.8958)  weight_decay: 0.0500 (0.0500)  time: 4.7788  data: 4.1289  max mem: 57114
Epoch: [269]  [ 10/156]  eta: 0:02:40  lr: 0.000240  min_lr: 0.000240  loss: 2.5879 (2.3688)  weight_decay: 0.0500 (0.0500)  time: 1.1008  data: 0.3756  max mem: 57114
Epoch: [269]  [ 20/156]  eta: 0:02:05  lr: 0.000239  min_lr: 0.000239  loss: 2.6957 (2.5579)  weight_decay: 0.0500 (0.0500)  time: 0.7317  data: 0.0003  max mem: 57114
Epoch: [269]  [ 30/156]  eta: 0:01:48  lr: 0.000238  min_lr: 0.000238  loss: 2.7764 (2.5716)  weight_decay: 0.0500 (0.0500)  time: 0.7250  data: 0.0003  max mem: 57114
Epoch: [269]  [ 40/156]  eta: 0:01:36  lr: 0.000237  min_lr: 0.000237  loss: 2.7060 (2.5826)  weight_decay: 0.0500 (0.0500)  time: 0.7283  data: 0.0004  max mem: 57114
Epoch: [269]  [ 50/156]  eta: 0:01:25  lr: 0.000236  min_lr: 0.000236  loss: 2.6324 (2.5705)  weight_decay: 0.0500 (0.0500)  time: 0.7237  data: 0.0004  max mem: 57114
Epoch: [269]  [ 60/156]  eta: 0:01:16  lr: 0.000235  min_lr: 0.000235  loss: 2.5319 (2.5490)  weight_decay: 0.0500 (0.0500)  time: 0.7225  data: 0.0004  max mem: 57114
Epoch: [269]  [ 70/156]  eta: 0:01:07  lr: 0.000234  min_lr: 0.000234  loss: 2.6588 (2.5508)  weight_decay: 0.0500 (0.0500)  time: 0.7291  data: 0.0004  max mem: 57114
Epoch: [269]  [ 80/156]  eta: 0:00:59  lr: 0.000233  min_lr: 0.000233  loss: 2.6588 (2.5599)  weight_decay: 0.0500 (0.0500)  time: 0.7245  data: 0.0004  max mem: 57114
Epoch: [269]  [ 90/156]  eta: 0:00:50  lr: 0.000232  min_lr: 0.000232  loss: 2.5973 (2.5638)  weight_decay: 0.0500 (0.0500)  time: 0.7158  data: 0.0004  max mem: 57114
Epoch: [269]  [100/156]  eta: 0:00:42  lr: 0.000231  min_lr: 0.000231  loss: 2.5973 (2.5540)  weight_decay: 0.0500 (0.0500)  time: 0.7159  data: 0.0004  max mem: 57114
Epoch: [269]  [110/156]  eta: 0:00:34  lr: 0.000230  min_lr: 0.000230  loss: 2.5489 (2.5535)  weight_decay: 0.0500 (0.0500)  time: 0.7148  data: 0.0004  max mem: 57114
Epoch: [269]  [120/156]  eta: 0:00:27  lr: 0.000229  min_lr: 0.000229  loss: 2.6330 (2.5479)  weight_decay: 0.0500 (0.0500)  time: 0.7000  data: 0.0004  max mem: 57114
Epoch: [269]  [130/156]  eta: 0:00:19  lr: 0.000228  min_lr: 0.000228  loss: 2.6330 (2.5282)  weight_decay: 0.0500 (0.0500)  time: 0.6931  data: 0.0008  max mem: 57114
Epoch: [269]  [140/156]  eta: 0:00:11  lr: 0.000227  min_lr: 0.000227  loss: 2.5789 (2.5348)  weight_decay: 0.0500 (0.0500)  time: 0.6880  data: 0.0007  max mem: 57114
Epoch: [269]  [150/156]  eta: 0:00:04  lr: 0.000226  min_lr: 0.000226  loss: 2.5555 (2.5404)  weight_decay: 0.0500 (0.0500)  time: 0.6829  data: 0.0001  max mem: 57114
Epoch: [269]  [155/156]  eta: 0:00:00  lr: 0.000226  min_lr: 0.000226  loss: 2.5553 (2.5414)  weight_decay: 0.0500 (0.0500)  time: 0.6848  data: 0.0001  max mem: 57114
Epoch: [269] Total time: 0:01:55 (0.7409 s / it)
Averaged stats: lr: 0.000226  min_lr: 0.000226  loss: 2.5553 (2.5483)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6151 (0.6151)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.0906  data: 1.8306  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8042 (0.7405)  acc1: 82.8125 (83.8875)  acc5: 98.4375 (97.1867)  time: 0.5689  data: 0.3662  max mem: 57114
Test: Total time: 0:00:02 (0.5893 s / it)
* Acc@1 84.429 Acc@5 96.947 loss 0.824
Accuracy of the model on the 50000 test images: 84.4%
Max accuracy: 84.49%
Test:  [0/5]  eta: 0:00:11  loss: 0.6171 (0.6171)  acc1: 84.3750 (84.3750)  acc5: 97.3958 (97.3958)  time: 2.3655  data: 2.1221  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7027 (0.7004)  acc1: 81.2500 (79.2839)  acc5: 96.3542 (94.1177)  time: 0.6238  data: 0.4245  max mem: 57114
Test: Total time: 0:00:03 (0.6342 s / it)
* Acc@1 77.534 Acc@5 93.770 loss 0.842
Accuracy of the model EMA on 50000 test images: 77.5%
Max EMA accuracy: 77.53%
Epoch: [270]  [  0/156]  eta: 0:10:36  lr: 0.000225  min_lr: 0.000225  loss: 2.8529 (2.8529)  weight_decay: 0.0500 (0.0500)  time: 4.0828  data: 3.4315  max mem: 57114
Epoch: [270]  [ 10/156]  eta: 0:02:31  lr: 0.000224  min_lr: 0.000224  loss: 2.4895 (2.4408)  weight_decay: 0.0500 (0.0500)  time: 1.0374  data: 0.3122  max mem: 57114
Epoch: [270]  [ 20/156]  eta: 0:02:01  lr: 0.000224  min_lr: 0.000224  loss: 2.4908 (2.5522)  weight_decay: 0.0500 (0.0500)  time: 0.7336  data: 0.0003  max mem: 57114
Epoch: [270]  [ 30/156]  eta: 0:01:45  lr: 0.000223  min_lr: 0.000223  loss: 2.6121 (2.5352)  weight_decay: 0.0500 (0.0500)  time: 0.7259  data: 0.0003  max mem: 57114
Epoch: [270]  [ 40/156]  eta: 0:01:33  lr: 0.000222  min_lr: 0.000222  loss: 2.5526 (2.5108)  weight_decay: 0.0500 (0.0500)  time: 0.7175  data: 0.0003  max mem: 57114
Epoch: [270]  [ 50/156]  eta: 0:01:23  lr: 0.000221  min_lr: 0.000221  loss: 2.5526 (2.5116)  weight_decay: 0.0500 (0.0500)  time: 0.7209  data: 0.0003  max mem: 57114
Epoch: [270]  [ 60/156]  eta: 0:01:14  lr: 0.000220  min_lr: 0.000220  loss: 2.6718 (2.5352)  weight_decay: 0.0500 (0.0500)  time: 0.7209  data: 0.0003  max mem: 57114
Epoch: [270]  [ 70/156]  eta: 0:01:06  lr: 0.000219  min_lr: 0.000219  loss: 2.6718 (2.5359)  weight_decay: 0.0500 (0.0500)  time: 0.7158  data: 0.0003  max mem: 57114
Epoch: [270]  [ 80/156]  eta: 0:00:58  lr: 0.000218  min_lr: 0.000218  loss: 2.6827 (2.5555)  weight_decay: 0.0500 (0.0500)  time: 0.7163  data: 0.0004  max mem: 57114
Epoch: [270]  [ 90/156]  eta: 0:00:50  lr: 0.000217  min_lr: 0.000217  loss: 2.5931 (2.5359)  weight_decay: 0.0500 (0.0500)  time: 0.7182  data: 0.0004  max mem: 57114
Epoch: [270]  [100/156]  eta: 0:00:42  lr: 0.000216  min_lr: 0.000216  loss: 2.4775 (2.5383)  weight_decay: 0.0500 (0.0500)  time: 0.7145  data: 0.0004  max mem: 57114
Epoch: [270]  [110/156]  eta: 0:00:34  lr: 0.000215  min_lr: 0.000215  loss: 2.5762 (2.5413)  weight_decay: 0.0500 (0.0500)  time: 0.7151  data: 0.0004  max mem: 57114
Epoch: [270]  [120/156]  eta: 0:00:26  lr: 0.000214  min_lr: 0.000214  loss: 2.5884 (2.5522)  weight_decay: 0.0500 (0.0500)  time: 0.7133  data: 0.0004  max mem: 57114
Epoch: [270]  [130/156]  eta: 0:00:19  lr: 0.000213  min_lr: 0.000213  loss: 2.5787 (2.5460)  weight_decay: 0.0500 (0.0500)  time: 0.7001  data: 0.0008  max mem: 57114
Epoch: [270]  [140/156]  eta: 0:00:11  lr: 0.000212  min_lr: 0.000212  loss: 2.6945 (2.5636)  weight_decay: 0.0500 (0.0500)  time: 0.6856  data: 0.0007  max mem: 57114
Epoch: [270]  [150/156]  eta: 0:00:04  lr: 0.000211  min_lr: 0.000211  loss: 2.6766 (2.5411)  weight_decay: 0.0500 (0.0500)  time: 0.6805  data: 0.0001  max mem: 57114
Epoch: [270]  [155/156]  eta: 0:00:00  lr: 0.000211  min_lr: 0.000211  loss: 2.4261 (2.5354)  weight_decay: 0.0500 (0.0500)  time: 0.6823  data: 0.0001  max mem: 57114
Epoch: [270] Total time: 0:01:54 (0.7353 s / it)
Averaged stats: lr: 0.000211  min_lr: 0.000211  loss: 2.4261 (2.5402)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5246 (0.5246)  acc1: 90.6250 (90.6250)  acc5: 98.9583 (98.9583)  time: 2.0758  data: 1.8201  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6984 (0.6575)  acc1: 84.3750 (84.6547)  acc5: 98.9583 (97.3146)  time: 0.5658  data: 0.3641  max mem: 57114
Test: Total time: 0:00:02 (0.5919 s / it)
* Acc@1 84.513 Acc@5 96.997 loss 0.741
Accuracy of the model on the 50000 test images: 84.5%
Max accuracy: 84.51%
Test:  [0/5]  eta: 0:00:10  loss: 0.6175 (0.6175)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.0514  data: 1.8078  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7030 (0.7008)  acc1: 81.2500 (79.4118)  acc5: 96.3542 (94.1177)  time: 0.5610  data: 0.3617  max mem: 57114
Test: Total time: 0:00:02 (0.5741 s / it)
* Acc@1 77.540 Acc@5 93.792 loss 0.842
Accuracy of the model EMA on 50000 test images: 77.5%
Max EMA accuracy: 77.54%
Epoch: [271]  [  0/156]  eta: 0:11:44  lr: 0.000211  min_lr: 0.000211  loss: 2.7026 (2.7026)  weight_decay: 0.0500 (0.0500)  time: 4.5175  data: 3.8375  max mem: 57114
Epoch: [271]  [ 10/156]  eta: 0:02:40  lr: 0.000210  min_lr: 0.000210  loss: 2.7026 (2.5202)  weight_decay: 0.0500 (0.0500)  time: 1.0989  data: 0.3630  max mem: 57114
Epoch: [271]  [ 20/156]  eta: 0:02:05  lr: 0.000209  min_lr: 0.000209  loss: 2.6149 (2.4751)  weight_decay: 0.0500 (0.0500)  time: 0.7457  data: 0.0080  max mem: 57114
Epoch: [271]  [ 30/156]  eta: 0:01:48  lr: 0.000208  min_lr: 0.000208  loss: 2.6149 (2.5288)  weight_decay: 0.0500 (0.0500)  time: 0.7283  data: 0.0003  max mem: 57114
Epoch: [271]  [ 40/156]  eta: 0:01:36  lr: 0.000207  min_lr: 0.000207  loss: 2.5938 (2.5358)  weight_decay: 0.0500 (0.0500)  time: 0.7265  data: 0.0003  max mem: 57114
Epoch: [271]  [ 50/156]  eta: 0:01:25  lr: 0.000206  min_lr: 0.000206  loss: 2.6873 (2.5843)  weight_decay: 0.0500 (0.0500)  time: 0.7273  data: 0.0004  max mem: 57114
Epoch: [271]  [ 60/156]  eta: 0:01:16  lr: 0.000205  min_lr: 0.000205  loss: 2.7115 (2.5955)  weight_decay: 0.0500 (0.0500)  time: 0.7357  data: 0.0004  max mem: 57114
Epoch: [271]  [ 70/156]  eta: 0:01:07  lr: 0.000204  min_lr: 0.000204  loss: 2.7031 (2.5930)  weight_decay: 0.0500 (0.0500)  time: 0.7354  data: 0.0003  max mem: 57114
Epoch: [271]  [ 80/156]  eta: 0:00:59  lr: 0.000204  min_lr: 0.000204  loss: 2.5770 (2.5798)  weight_decay: 0.0500 (0.0500)  time: 0.7238  data: 0.0004  max mem: 57114
Epoch: [271]  [ 90/156]  eta: 0:00:50  lr: 0.000203  min_lr: 0.000203  loss: 2.5257 (2.5652)  weight_decay: 0.0500 (0.0500)  time: 0.7108  data: 0.0005  max mem: 57114
Epoch: [271]  [100/156]  eta: 0:00:42  lr: 0.000202  min_lr: 0.000202  loss: 2.6019 (2.5694)  weight_decay: 0.0500 (0.0500)  time: 0.6990  data: 0.0005  max mem: 57114
Epoch: [271]  [110/156]  eta: 0:00:34  lr: 0.000201  min_lr: 0.000201  loss: 2.5869 (2.5473)  weight_decay: 0.0500 (0.0500)  time: 0.6991  data: 0.0003  max mem: 57114
Epoch: [271]  [120/156]  eta: 0:00:27  lr: 0.000200  min_lr: 0.000200  loss: 2.7086 (2.5668)  weight_decay: 0.0500 (0.0500)  time: 0.7041  data: 0.0003  max mem: 57114
Epoch: [271]  [130/156]  eta: 0:00:19  lr: 0.000199  min_lr: 0.000199  loss: 2.7420 (2.5671)  weight_decay: 0.0500 (0.0500)  time: 0.7003  data: 0.0008  max mem: 57114
Epoch: [271]  [140/156]  eta: 0:00:11  lr: 0.000198  min_lr: 0.000198  loss: 2.4818 (2.5545)  weight_decay: 0.0500 (0.0500)  time: 0.6874  data: 0.0007  max mem: 57114
Epoch: [271]  [150/156]  eta: 0:00:04  lr: 0.000197  min_lr: 0.000197  loss: 2.5461 (2.5615)  weight_decay: 0.0500 (0.0500)  time: 0.6819  data: 0.0001  max mem: 57114
Epoch: [271]  [155/156]  eta: 0:00:00  lr: 0.000197  min_lr: 0.000197  loss: 2.5843 (2.5621)  weight_decay: 0.0500 (0.0500)  time: 0.6788  data: 0.0001  max mem: 57114
Epoch: [271] Total time: 0:01:55 (0.7400 s / it)
Averaged stats: lr: 0.000197  min_lr: 0.000197  loss: 2.5843 (2.5376)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5870 (0.5870)  acc1: 91.1458 (91.1458)  acc5: 98.9583 (98.9583)  time: 2.0634  data: 1.8077  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7596 (0.7156)  acc1: 85.4167 (85.0384)  acc5: 98.4375 (97.4425)  time: 0.5633  data: 0.3616  max mem: 57114
Test: Total time: 0:00:02 (0.5890 s / it)
* Acc@1 84.567 Acc@5 96.987 loss 0.800
Accuracy of the model on the 50000 test images: 84.6%
Max accuracy: 84.57%
Test:  [0/5]  eta: 0:00:10  loss: 0.6179 (0.6179)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.1079  data: 1.8643  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7032 (0.7013)  acc1: 81.2500 (79.2839)  acc5: 96.3542 (94.1177)  time: 0.5723  data: 0.3730  max mem: 57114
Test: Total time: 0:00:02 (0.5830 s / it)
* Acc@1 77.550 Acc@5 93.796 loss 0.842
Accuracy of the model EMA on 50000 test images: 77.5%
Max EMA accuracy: 77.55%
Epoch: [272]  [  0/156]  eta: 0:09:08  lr: 0.000197  min_lr: 0.000197  loss: 2.7494 (2.7494)  weight_decay: 0.0500 (0.0500)  time: 3.5144  data: 2.8588  max mem: 57114
Epoch: [272]  [ 10/156]  eta: 0:02:24  lr: 0.000196  min_lr: 0.000196  loss: 2.6175 (2.4093)  weight_decay: 0.0500 (0.0500)  time: 0.9911  data: 0.2601  max mem: 57114
Epoch: [272]  [ 20/156]  eta: 0:02:00  lr: 0.000195  min_lr: 0.000195  loss: 2.5696 (2.4989)  weight_decay: 0.0500 (0.0500)  time: 0.7524  data: 0.0003  max mem: 57114
Epoch: [272]  [ 30/156]  eta: 0:01:44  lr: 0.000194  min_lr: 0.000194  loss: 2.5696 (2.4863)  weight_decay: 0.0500 (0.0500)  time: 0.7405  data: 0.0004  max mem: 57114
Epoch: [272]  [ 40/156]  eta: 0:01:33  lr: 0.000193  min_lr: 0.000193  loss: 2.6080 (2.4721)  weight_decay: 0.0500 (0.0500)  time: 0.7301  data: 0.0004  max mem: 57114
Epoch: [272]  [ 50/156]  eta: 0:01:24  lr: 0.000192  min_lr: 0.000192  loss: 2.5856 (2.4898)  weight_decay: 0.0500 (0.0500)  time: 0.7359  data: 0.0004  max mem: 57114
Epoch: [272]  [ 60/156]  eta: 0:01:14  lr: 0.000191  min_lr: 0.000191  loss: 2.5631 (2.5033)  weight_decay: 0.0500 (0.0500)  time: 0.7146  data: 0.0004  max mem: 57114
Epoch: [272]  [ 70/156]  eta: 0:01:06  lr: 0.000191  min_lr: 0.000191  loss: 2.7561 (2.5318)  weight_decay: 0.0500 (0.0500)  time: 0.7139  data: 0.0004  max mem: 57114
Epoch: [272]  [ 80/156]  eta: 0:00:57  lr: 0.000190  min_lr: 0.000190  loss: 2.7142 (2.5347)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0004  max mem: 57114
Epoch: [272]  [ 90/156]  eta: 0:00:49  lr: 0.000189  min_lr: 0.000189  loss: 2.6476 (2.5400)  weight_decay: 0.0500 (0.0500)  time: 0.7080  data: 0.0004  max mem: 57114
Epoch: [272]  [100/156]  eta: 0:00:42  lr: 0.000188  min_lr: 0.000188  loss: 2.4965 (2.5283)  weight_decay: 0.0500 (0.0500)  time: 0.7067  data: 0.0004  max mem: 57114
Epoch: [272]  [110/156]  eta: 0:00:34  lr: 0.000187  min_lr: 0.000187  loss: 2.6037 (2.5466)  weight_decay: 0.0500 (0.0500)  time: 0.7088  data: 0.0004  max mem: 57114
Epoch: [272]  [120/156]  eta: 0:00:26  lr: 0.000186  min_lr: 0.000186  loss: 2.6725 (2.5450)  weight_decay: 0.0500 (0.0500)  time: 0.7114  data: 0.0004  max mem: 57114
Epoch: [272]  [130/156]  eta: 0:00:19  lr: 0.000185  min_lr: 0.000185  loss: 2.2468 (2.5098)  weight_decay: 0.0500 (0.0500)  time: 0.6941  data: 0.0005  max mem: 57114
Epoch: [272]  [140/156]  eta: 0:00:11  lr: 0.000184  min_lr: 0.000184  loss: 2.6032 (2.5241)  weight_decay: 0.0500 (0.0500)  time: 0.6920  data: 0.0004  max mem: 57114
Epoch: [272]  [150/156]  eta: 0:00:04  lr: 0.000184  min_lr: 0.000184  loss: 2.6547 (2.5209)  weight_decay: 0.0500 (0.0500)  time: 0.6897  data: 0.0001  max mem: 57114
Epoch: [272]  [155/156]  eta: 0:00:00  lr: 0.000183  min_lr: 0.000183  loss: 2.6575 (2.5239)  weight_decay: 0.0500 (0.0500)  time: 0.6878  data: 0.0001  max mem: 57114
Epoch: [272] Total time: 0:01:54 (0.7341 s / it)
Averaged stats: lr: 0.000183  min_lr: 0.000183  loss: 2.6575 (2.5398)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 0.5561 (0.5561)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.2320  data: 1.9714  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7319 (0.6844)  acc1: 83.8542 (84.6547)  acc5: 98.4375 (97.5703)  time: 0.5972  data: 0.3944  max mem: 57114
Test: Total time: 0:00:03 (0.6225 s / it)
* Acc@1 84.581 Acc@5 97.067 loss 0.770
Accuracy of the model on the 50000 test images: 84.6%
Max accuracy: 84.58%
Test:  [0/5]  eta: 0:00:09  loss: 0.6181 (0.6181)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 1.9948  data: 1.7514  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7034 (0.7016)  acc1: 81.2500 (79.2839)  acc5: 96.3542 (94.2455)  time: 0.5496  data: 0.3504  max mem: 57114
Test: Total time: 0:00:02 (0.5609 s / it)
* Acc@1 77.554 Acc@5 93.804 loss 0.843
Accuracy of the model EMA on 50000 test images: 77.6%
Max EMA accuracy: 77.55%
Epoch: [273]  [  0/156]  eta: 0:08:25  lr: 0.000183  min_lr: 0.000183  loss: 2.9475 (2.9475)  weight_decay: 0.0500 (0.0500)  time: 3.2431  data: 2.5925  max mem: 57114
Epoch: [273]  [ 10/156]  eta: 0:02:21  lr: 0.000182  min_lr: 0.000182  loss: 2.5582 (2.3839)  weight_decay: 0.0500 (0.0500)  time: 0.9699  data: 0.2359  max mem: 57114
Epoch: [273]  [ 20/156]  eta: 0:01:55  lr: 0.000181  min_lr: 0.000181  loss: 2.5025 (2.4199)  weight_decay: 0.0500 (0.0500)  time: 0.7332  data: 0.0004  max mem: 57114
Epoch: [273]  [ 30/156]  eta: 0:01:43  lr: 0.000181  min_lr: 0.000181  loss: 2.4808 (2.4127)  weight_decay: 0.0500 (0.0500)  time: 0.7447  data: 0.0004  max mem: 57114
Epoch: [273]  [ 40/156]  eta: 0:01:33  lr: 0.000180  min_lr: 0.000180  loss: 2.5352 (2.4185)  weight_decay: 0.0500 (0.0500)  time: 0.7502  data: 0.0003  max mem: 57114
Epoch: [273]  [ 50/156]  eta: 0:01:23  lr: 0.000179  min_lr: 0.000179  loss: 2.6604 (2.4609)  weight_decay: 0.0500 (0.0500)  time: 0.7373  data: 0.0004  max mem: 57114
Epoch: [273]  [ 60/156]  eta: 0:01:15  lr: 0.000178  min_lr: 0.000178  loss: 2.7362 (2.4880)  weight_decay: 0.0500 (0.0500)  time: 0.7378  data: 0.0004  max mem: 57114
Epoch: [273]  [ 70/156]  eta: 0:01:06  lr: 0.000177  min_lr: 0.000177  loss: 2.6977 (2.4733)  weight_decay: 0.0500 (0.0500)  time: 0.7217  data: 0.0004  max mem: 57114
Epoch: [273]  [ 80/156]  eta: 0:00:57  lr: 0.000176  min_lr: 0.000176  loss: 2.6380 (2.4789)  weight_decay: 0.0500 (0.0500)  time: 0.7042  data: 0.0004  max mem: 57114
Epoch: [273]  [ 90/156]  eta: 0:00:50  lr: 0.000175  min_lr: 0.000175  loss: 2.6580 (2.4890)  weight_decay: 0.0500 (0.0500)  time: 0.7120  data: 0.0003  max mem: 57114
Epoch: [273]  [100/156]  eta: 0:00:42  lr: 0.000175  min_lr: 0.000175  loss: 2.6140 (2.4768)  weight_decay: 0.0500 (0.0500)  time: 0.7138  data: 0.0003  max mem: 57114
Epoch: [273]  [110/156]  eta: 0:00:34  lr: 0.000174  min_lr: 0.000174  loss: 2.3191 (2.4680)  weight_decay: 0.0500 (0.0500)  time: 0.7059  data: 0.0004  max mem: 57114
Epoch: [273]  [120/156]  eta: 0:00:26  lr: 0.000173  min_lr: 0.000173  loss: 2.2966 (2.4608)  weight_decay: 0.0500 (0.0500)  time: 0.6999  data: 0.0004  max mem: 57114
Epoch: [273]  [130/156]  eta: 0:00:19  lr: 0.000172  min_lr: 0.000172  loss: 2.4524 (2.4711)  weight_decay: 0.0500 (0.0500)  time: 0.6882  data: 0.0008  max mem: 57114
Epoch: [273]  [140/156]  eta: 0:00:11  lr: 0.000171  min_lr: 0.000171  loss: 2.5177 (2.4658)  weight_decay: 0.0500 (0.0500)  time: 0.6828  data: 0.0007  max mem: 57114
Epoch: [273]  [150/156]  eta: 0:00:04  lr: 0.000170  min_lr: 0.000170  loss: 2.6965 (2.4839)  weight_decay: 0.0500 (0.0500)  time: 0.6844  data: 0.0001  max mem: 57114
Epoch: [273]  [155/156]  eta: 0:00:00  lr: 0.000170  min_lr: 0.000170  loss: 2.6965 (2.4854)  weight_decay: 0.0500 (0.0500)  time: 0.6841  data: 0.0001  max mem: 57114
Epoch: [273] Total time: 0:01:54 (0.7328 s / it)
Averaged stats: lr: 0.000170  min_lr: 0.000170  loss: 2.6965 (2.5301)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6365 (0.6365)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.1875  data: 1.9314  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8155 (0.7596)  acc1: 84.8958 (84.9105)  acc5: 98.4375 (97.4425)  time: 0.5881  data: 0.3863  max mem: 57114
Test: Total time: 0:00:03 (0.6134 s / it)
* Acc@1 84.569 Acc@5 96.983 loss 0.838
Accuracy of the model on the 50000 test images: 84.6%
Max accuracy: 84.58%
Test:  [0/5]  eta: 0:00:11  loss: 0.6182 (0.6182)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.2924  data: 2.0489  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7035 (0.7018)  acc1: 81.2500 (79.2839)  acc5: 96.3542 (94.2455)  time: 0.6091  data: 0.4099  max mem: 57114
Test: Total time: 0:00:03 (0.6190 s / it)
* Acc@1 77.566 Acc@5 93.814 loss 0.843
Accuracy of the model EMA on 50000 test images: 77.6%
Max EMA accuracy: 77.57%
Epoch: [274]  [  0/156]  eta: 0:08:29  lr: 0.000170  min_lr: 0.000170  loss: 2.8898 (2.8898)  weight_decay: 0.0500 (0.0500)  time: 3.2648  data: 2.6097  max mem: 57114
Epoch: [274]  [ 10/156]  eta: 0:02:21  lr: 0.000169  min_lr: 0.000169  loss: 2.5481 (2.3952)  weight_decay: 0.0500 (0.0500)  time: 0.9680  data: 0.2375  max mem: 57114
Epoch: [274]  [ 20/156]  eta: 0:01:54  lr: 0.000168  min_lr: 0.000168  loss: 2.5481 (2.5074)  weight_decay: 0.0500 (0.0500)  time: 0.7209  data: 0.0003  max mem: 57114
Epoch: [274]  [ 30/156]  eta: 0:01:40  lr: 0.000168  min_lr: 0.000168  loss: 2.6699 (2.4988)  weight_decay: 0.0500 (0.0500)  time: 0.7071  data: 0.0003  max mem: 57114
Epoch: [274]  [ 40/156]  eta: 0:01:30  lr: 0.000167  min_lr: 0.000167  loss: 2.3832 (2.4834)  weight_decay: 0.0500 (0.0500)  time: 0.7087  data: 0.0003  max mem: 57114
Epoch: [274]  [ 50/156]  eta: 0:01:21  lr: 0.000166  min_lr: 0.000166  loss: 2.4530 (2.5025)  weight_decay: 0.0500 (0.0500)  time: 0.7243  data: 0.0003  max mem: 57114
Epoch: [274]  [ 60/156]  eta: 0:01:12  lr: 0.000165  min_lr: 0.000165  loss: 2.6667 (2.5302)  weight_decay: 0.0500 (0.0500)  time: 0.7217  data: 0.0003  max mem: 57114
Epoch: [274]  [ 70/156]  eta: 0:01:04  lr: 0.000164  min_lr: 0.000164  loss: 2.5449 (2.5016)  weight_decay: 0.0500 (0.0500)  time: 0.7177  data: 0.0004  max mem: 57114
Epoch: [274]  [ 80/156]  eta: 0:00:56  lr: 0.000163  min_lr: 0.000163  loss: 2.6305 (2.5248)  weight_decay: 0.0500 (0.0500)  time: 0.7224  data: 0.0004  max mem: 57114
Epoch: [274]  [ 90/156]  eta: 0:00:49  lr: 0.000163  min_lr: 0.000163  loss: 2.6803 (2.5226)  weight_decay: 0.0500 (0.0500)  time: 0.7008  data: 0.0003  max mem: 57114
Epoch: [274]  [100/156]  eta: 0:00:41  lr: 0.000162  min_lr: 0.000162  loss: 2.5795 (2.5240)  weight_decay: 0.0500 (0.0500)  time: 0.7016  data: 0.0004  max mem: 57114
Epoch: [274]  [110/156]  eta: 0:00:33  lr: 0.000161  min_lr: 0.000161  loss: 2.6473 (2.5205)  weight_decay: 0.0500 (0.0500)  time: 0.6998  data: 0.0004  max mem: 57114
Epoch: [274]  [120/156]  eta: 0:00:26  lr: 0.000160  min_lr: 0.000160  loss: 2.7350 (2.5282)  weight_decay: 0.0500 (0.0500)  time: 0.6923  data: 0.0004  max mem: 57114
Epoch: [274]  [130/156]  eta: 0:00:18  lr: 0.000159  min_lr: 0.000159  loss: 2.6973 (2.5326)  weight_decay: 0.0500 (0.0500)  time: 0.6996  data: 0.0009  max mem: 57114
Epoch: [274]  [140/156]  eta: 0:00:11  lr: 0.000159  min_lr: 0.000159  loss: 2.6101 (2.5244)  weight_decay: 0.0500 (0.0500)  time: 0.6958  data: 0.0007  max mem: 57114
Epoch: [274]  [150/156]  eta: 0:00:04  lr: 0.000158  min_lr: 0.000158  loss: 2.1762 (2.5086)  weight_decay: 0.0500 (0.0500)  time: 0.6886  data: 0.0001  max mem: 57114
Epoch: [274]  [155/156]  eta: 0:00:00  lr: 0.000157  min_lr: 0.000157  loss: 2.5338 (2.5110)  weight_decay: 0.0500 (0.0500)  time: 0.6914  data: 0.0001  max mem: 57114
Epoch: [274] Total time: 0:01:53 (0.7258 s / it)
Averaged stats: lr: 0.000157  min_lr: 0.000157  loss: 2.5338 (2.5283)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5361 (0.5361)  acc1: 91.1458 (91.1458)  acc5: 98.9583 (98.9583)  time: 2.0419  data: 1.7859  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7270 (0.6629)  acc1: 83.3333 (84.3990)  acc5: 98.4375 (97.4425)  time: 0.5590  data: 0.3572  max mem: 57114
Test: Total time: 0:00:02 (0.5836 s / it)
* Acc@1 84.511 Acc@5 97.027 loss 0.750
Accuracy of the model on the 50000 test images: 84.5%
Max accuracy: 84.58%
Test:  [0/5]  eta: 0:00:11  loss: 0.6185 (0.6185)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.3213  data: 2.0778  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7034 (0.7020)  acc1: 81.2500 (79.4118)  acc5: 97.3958 (94.5013)  time: 0.6149  data: 0.4156  max mem: 57114
Test: Total time: 0:00:03 (0.6273 s / it)
* Acc@1 77.584 Acc@5 93.822 loss 0.843
Accuracy of the model EMA on 50000 test images: 77.6%
Max EMA accuracy: 77.58%
Epoch: [275]  [  0/156]  eta: 0:08:39  lr: 0.000157  min_lr: 0.000157  loss: 2.0686 (2.0686)  weight_decay: 0.0500 (0.0500)  time: 3.3271  data: 2.6696  max mem: 57114
Epoch: [275]  [ 10/156]  eta: 0:02:20  lr: 0.000157  min_lr: 0.000157  loss: 2.5409 (2.4712)  weight_decay: 0.0500 (0.0500)  time: 0.9597  data: 0.2430  max mem: 57114
Epoch: [275]  [ 20/156]  eta: 0:01:54  lr: 0.000156  min_lr: 0.000156  loss: 2.5567 (2.4595)  weight_decay: 0.0500 (0.0500)  time: 0.7176  data: 0.0003  max mem: 57114
Epoch: [275]  [ 30/156]  eta: 0:01:40  lr: 0.000155  min_lr: 0.000155  loss: 2.5640 (2.4804)  weight_decay: 0.0500 (0.0500)  time: 0.7098  data: 0.0003  max mem: 57114
Epoch: [275]  [ 40/156]  eta: 0:01:30  lr: 0.000154  min_lr: 0.000154  loss: 2.3592 (2.4389)  weight_decay: 0.0500 (0.0500)  time: 0.7213  data: 0.0003  max mem: 57114
Epoch: [275]  [ 50/156]  eta: 0:01:21  lr: 0.000153  min_lr: 0.000153  loss: 2.3257 (2.4385)  weight_decay: 0.0500 (0.0500)  time: 0.7309  data: 0.0004  max mem: 57114
Epoch: [275]  [ 60/156]  eta: 0:01:13  lr: 0.000153  min_lr: 0.000153  loss: 2.4714 (2.4329)  weight_decay: 0.0500 (0.0500)  time: 0.7289  data: 0.0004  max mem: 57114
Epoch: [275]  [ 70/156]  eta: 0:01:05  lr: 0.000152  min_lr: 0.000152  loss: 2.5979 (2.4388)  weight_decay: 0.0500 (0.0500)  time: 0.7148  data: 0.0004  max mem: 57114
Epoch: [275]  [ 80/156]  eta: 0:00:57  lr: 0.000151  min_lr: 0.000151  loss: 2.6292 (2.4564)  weight_decay: 0.0500 (0.0500)  time: 0.7074  data: 0.0004  max mem: 57114
Epoch: [275]  [ 90/156]  eta: 0:00:49  lr: 0.000150  min_lr: 0.000150  loss: 2.6414 (2.4609)  weight_decay: 0.0500 (0.0500)  time: 0.7109  data: 0.0004  max mem: 57114
Epoch: [275]  [100/156]  eta: 0:00:41  lr: 0.000149  min_lr: 0.000149  loss: 2.6185 (2.4537)  weight_decay: 0.0500 (0.0500)  time: 0.7007  data: 0.0004  max mem: 57114
Epoch: [275]  [110/156]  eta: 0:00:33  lr: 0.000149  min_lr: 0.000149  loss: 2.4728 (2.4653)  weight_decay: 0.0500 (0.0500)  time: 0.7017  data: 0.0003  max mem: 57114
Epoch: [275]  [120/156]  eta: 0:00:26  lr: 0.000148  min_lr: 0.000148  loss: 2.6112 (2.4674)  weight_decay: 0.0500 (0.0500)  time: 0.7104  data: 0.0004  max mem: 57114
Epoch: [275]  [130/156]  eta: 0:00:19  lr: 0.000147  min_lr: 0.000147  loss: 2.6779 (2.4774)  weight_decay: 0.0500 (0.0500)  time: 0.7020  data: 0.0008  max mem: 57114
Epoch: [275]  [140/156]  eta: 0:00:11  lr: 0.000146  min_lr: 0.000146  loss: 2.7205 (2.4847)  weight_decay: 0.0500 (0.0500)  time: 0.6948  data: 0.0007  max mem: 57114
Epoch: [275]  [150/156]  eta: 0:00:04  lr: 0.000146  min_lr: 0.000146  loss: 2.6341 (2.4981)  weight_decay: 0.0500 (0.0500)  time: 0.6882  data: 0.0001  max mem: 57114
Epoch: [275]  [155/156]  eta: 0:00:00  lr: 0.000145  min_lr: 0.000145  loss: 2.6288 (2.4965)  weight_decay: 0.0500 (0.0500)  time: 0.6907  data: 0.0001  max mem: 57114
Epoch: [275] Total time: 0:01:53 (0.7277 s / it)
Averaged stats: lr: 0.000145  min_lr: 0.000145  loss: 2.6288 (2.5257)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5884 (0.5884)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.1439  data: 1.8876  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7677 (0.7135)  acc1: 84.8958 (84.5269)  acc5: 98.4375 (97.6982)  time: 0.5796  data: 0.3776  max mem: 57114
Test: Total time: 0:00:03 (0.6057 s / it)
* Acc@1 84.627 Acc@5 97.027 loss 0.797
Accuracy of the model on the 50000 test images: 84.6%
Max accuracy: 84.63%
Test:  [0/5]  eta: 0:00:10  loss: 0.6186 (0.6186)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.0292  data: 1.7858  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7033 (0.7022)  acc1: 81.2500 (79.2839)  acc5: 97.3958 (94.5013)  time: 0.5566  data: 0.3572  max mem: 57114
Test: Total time: 0:00:02 (0.5683 s / it)
* Acc@1 77.596 Acc@5 93.828 loss 0.843
Accuracy of the model EMA on 50000 test images: 77.6%
Max EMA accuracy: 77.60%
Epoch: [276]  [  0/156]  eta: 0:07:19  lr: 0.000145  min_lr: 0.000145  loss: 2.6308 (2.6308)  weight_decay: 0.0500 (0.0500)  time: 2.8204  data: 2.1719  max mem: 57114
Epoch: [276]  [ 10/156]  eta: 0:02:14  lr: 0.000144  min_lr: 0.000144  loss: 2.8230 (2.6378)  weight_decay: 0.0500 (0.0500)  time: 0.9205  data: 0.1977  max mem: 57114
Epoch: [276]  [ 20/156]  eta: 0:01:54  lr: 0.000144  min_lr: 0.000144  loss: 2.8230 (2.7037)  weight_decay: 0.0500 (0.0500)  time: 0.7466  data: 0.0003  max mem: 57114
Epoch: [276]  [ 30/156]  eta: 0:01:41  lr: 0.000143  min_lr: 0.000143  loss: 2.7381 (2.6881)  weight_decay: 0.0500 (0.0500)  time: 0.7366  data: 0.0003  max mem: 57114
Epoch: [276]  [ 40/156]  eta: 0:01:30  lr: 0.000142  min_lr: 0.000142  loss: 2.6959 (2.6769)  weight_decay: 0.0500 (0.0500)  time: 0.7190  data: 0.0003  max mem: 57114
Epoch: [276]  [ 50/156]  eta: 0:01:22  lr: 0.000141  min_lr: 0.000141  loss: 2.6880 (2.6328)  weight_decay: 0.0500 (0.0500)  time: 0.7304  data: 0.0003  max mem: 57114
Epoch: [276]  [ 60/156]  eta: 0:01:14  lr: 0.000141  min_lr: 0.000141  loss: 2.6771 (2.6348)  weight_decay: 0.0500 (0.0500)  time: 0.7471  data: 0.0004  max mem: 57114
Epoch: [276]  [ 70/156]  eta: 0:01:05  lr: 0.000140  min_lr: 0.000140  loss: 2.5726 (2.6053)  weight_decay: 0.0500 (0.0500)  time: 0.7435  data: 0.0004  max mem: 57114
Epoch: [276]  [ 80/156]  eta: 0:00:57  lr: 0.000139  min_lr: 0.000139  loss: 2.3472 (2.5646)  weight_decay: 0.0500 (0.0500)  time: 0.7109  data: 0.0004  max mem: 57114
Epoch: [276]  [ 90/156]  eta: 0:00:49  lr: 0.000138  min_lr: 0.000138  loss: 2.6817 (2.5802)  weight_decay: 0.0500 (0.0500)  time: 0.7013  data: 0.0004  max mem: 57114
Epoch: [276]  [100/156]  eta: 0:00:41  lr: 0.000138  min_lr: 0.000138  loss: 2.6854 (2.5681)  weight_decay: 0.0500 (0.0500)  time: 0.6971  data: 0.0004  max mem: 57114
Epoch: [276]  [110/156]  eta: 0:00:34  lr: 0.000137  min_lr: 0.000137  loss: 2.5407 (2.5535)  weight_decay: 0.0500 (0.0500)  time: 0.6983  data: 0.0004  max mem: 57114
Epoch: [276]  [120/156]  eta: 0:00:26  lr: 0.000136  min_lr: 0.000136  loss: 2.5728 (2.5638)  weight_decay: 0.0500 (0.0500)  time: 0.7062  data: 0.0004  max mem: 57114
Epoch: [276]  [130/156]  eta: 0:00:19  lr: 0.000135  min_lr: 0.000135  loss: 2.6851 (2.5613)  weight_decay: 0.0500 (0.0500)  time: 0.7016  data: 0.0006  max mem: 57114
Epoch: [276]  [140/156]  eta: 0:00:11  lr: 0.000135  min_lr: 0.000135  loss: 2.6543 (2.5601)  weight_decay: 0.0500 (0.0500)  time: 0.6951  data: 0.0005  max mem: 57114
Epoch: [276]  [150/156]  eta: 0:00:04  lr: 0.000134  min_lr: 0.000134  loss: 2.6413 (2.5609)  weight_decay: 0.0500 (0.0500)  time: 0.6902  data: 0.0001  max mem: 57114
Epoch: [276]  [155/156]  eta: 0:00:00  lr: 0.000134  min_lr: 0.000134  loss: 2.6073 (2.5634)  weight_decay: 0.0500 (0.0500)  time: 0.6868  data: 0.0001  max mem: 57114
Epoch: [276] Total time: 0:01:53 (0.7301 s / it)
Averaged stats: lr: 0.000134  min_lr: 0.000134  loss: 2.6073 (2.5280)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6183 (0.6183)  acc1: 89.5833 (89.5833)  acc5: 98.4375 (98.4375)  time: 2.1820  data: 1.9263  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8000 (0.7454)  acc1: 84.3750 (84.2711)  acc5: 98.4375 (97.4425)  time: 0.5872  data: 0.3853  max mem: 57114
Test: Total time: 0:00:03 (0.6094 s / it)
* Acc@1 84.591 Acc@5 97.011 loss 0.825
Accuracy of the model on the 50000 test images: 84.6%
Max accuracy: 84.63%
Test:  [0/5]  eta: 0:00:12  loss: 0.6186 (0.6186)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.4789  data: 2.2354  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7031 (0.7022)  acc1: 81.2500 (79.2839)  acc5: 97.3958 (94.6292)  time: 0.6465  data: 0.4472  max mem: 57114
Test: Total time: 0:00:03 (0.6578 s / it)
* Acc@1 77.621 Acc@5 93.838 loss 0.843
Accuracy of the model EMA on 50000 test images: 77.6%
Max EMA accuracy: 77.62%
Epoch: [277]  [  0/156]  eta: 0:08:31  lr: 0.000133  min_lr: 0.000133  loss: 2.9906 (2.9906)  weight_decay: 0.0500 (0.0500)  time: 3.2793  data: 2.6265  max mem: 57114
Epoch: [277]  [ 10/156]  eta: 0:02:20  lr: 0.000133  min_lr: 0.000133  loss: 2.6784 (2.5895)  weight_decay: 0.0500 (0.0500)  time: 0.9637  data: 0.2390  max mem: 57114
Epoch: [277]  [ 20/156]  eta: 0:01:57  lr: 0.000132  min_lr: 0.000132  loss: 2.6784 (2.5778)  weight_decay: 0.0500 (0.0500)  time: 0.7417  data: 0.0003  max mem: 57114
Epoch: [277]  [ 30/156]  eta: 0:01:44  lr: 0.000131  min_lr: 0.000131  loss: 2.7109 (2.6261)  weight_decay: 0.0500 (0.0500)  time: 0.7562  data: 0.0004  max mem: 57114
Epoch: [277]  [ 40/156]  eta: 0:01:34  lr: 0.000131  min_lr: 0.000131  loss: 2.6990 (2.6304)  weight_decay: 0.0500 (0.0500)  time: 0.7624  data: 0.0006  max mem: 57114
Epoch: [277]  [ 50/156]  eta: 0:01:25  lr: 0.000130  min_lr: 0.000130  loss: 2.6799 (2.6148)  weight_decay: 0.0500 (0.0500)  time: 0.7625  data: 0.0006  max mem: 57114
Epoch: [277]  [ 60/156]  eta: 0:01:16  lr: 0.000129  min_lr: 0.000129  loss: 2.6799 (2.6232)  weight_decay: 0.0500 (0.0500)  time: 0.7659  data: 0.0004  max mem: 57114
Epoch: [277]  [ 70/156]  eta: 0:01:07  lr: 0.000128  min_lr: 0.000128  loss: 2.6759 (2.6181)  weight_decay: 0.0500 (0.0500)  time: 0.7430  data: 0.0004  max mem: 57114
Epoch: [277]  [ 80/156]  eta: 0:00:58  lr: 0.000128  min_lr: 0.000128  loss: 2.7325 (2.6312)  weight_decay: 0.0500 (0.0500)  time: 0.7076  data: 0.0004  max mem: 57114
Epoch: [277]  [ 90/156]  eta: 0:00:50  lr: 0.000127  min_lr: 0.000127  loss: 2.6772 (2.6027)  weight_decay: 0.0500 (0.0500)  time: 0.6885  data: 0.0004  max mem: 57114
Epoch: [277]  [100/156]  eta: 0:00:42  lr: 0.000126  min_lr: 0.000126  loss: 2.5353 (2.5864)  weight_decay: 0.0500 (0.0500)  time: 0.6945  data: 0.0004  max mem: 57114
Epoch: [277]  [110/156]  eta: 0:00:34  lr: 0.000125  min_lr: 0.000125  loss: 2.5675 (2.5721)  weight_decay: 0.0500 (0.0500)  time: 0.7120  data: 0.0004  max mem: 57114
Epoch: [277]  [120/156]  eta: 0:00:27  lr: 0.000125  min_lr: 0.000125  loss: 2.6732 (2.5768)  weight_decay: 0.0500 (0.0500)  time: 0.7103  data: 0.0004  max mem: 57114
Epoch: [277]  [130/156]  eta: 0:00:19  lr: 0.000124  min_lr: 0.000124  loss: 2.6282 (2.5778)  weight_decay: 0.0500 (0.0500)  time: 0.7001  data: 0.0008  max mem: 57114
Epoch: [277]  [140/156]  eta: 0:00:11  lr: 0.000123  min_lr: 0.000123  loss: 2.6122 (2.5770)  weight_decay: 0.0500 (0.0500)  time: 0.6918  data: 0.0007  max mem: 57114
Epoch: [277]  [150/156]  eta: 0:00:04  lr: 0.000123  min_lr: 0.000123  loss: 2.6122 (2.5618)  weight_decay: 0.0500 (0.0500)  time: 0.6832  data: 0.0001  max mem: 57114
Epoch: [277]  [155/156]  eta: 0:00:00  lr: 0.000122  min_lr: 0.000122  loss: 2.4647 (2.5651)  weight_decay: 0.0500 (0.0500)  time: 0.6889  data: 0.0001  max mem: 57114
Epoch: [277] Total time: 0:01:55 (0.7392 s / it)
Averaged stats: lr: 0.000122  min_lr: 0.000122  loss: 2.4647 (2.5267)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5786 (0.5786)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.0925  data: 1.8365  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7586 (0.6979)  acc1: 83.8542 (84.2711)  acc5: 98.4375 (97.4425)  time: 0.5692  data: 0.3674  max mem: 57114
Test: Total time: 0:00:02 (0.5915 s / it)
* Acc@1 84.585 Acc@5 97.027 loss 0.780
Accuracy of the model on the 50000 test images: 84.6%
Max accuracy: 84.63%
Test:  [0/5]  eta: 0:00:11  loss: 0.6185 (0.6185)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.2367  data: 1.9931  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7029 (0.7022)  acc1: 81.2500 (79.2839)  acc5: 97.3958 (94.5013)  time: 0.5980  data: 0.3987  max mem: 57114
Test: Total time: 0:00:03 (0.6126 s / it)
* Acc@1 77.639 Acc@5 93.850 loss 0.843
Accuracy of the model EMA on 50000 test images: 77.6%
Max EMA accuracy: 77.64%
Epoch: [278]  [  0/156]  eta: 0:09:30  lr: 0.000122  min_lr: 0.000122  loss: 2.9940 (2.9940)  weight_decay: 0.0500 (0.0500)  time: 3.6559  data: 2.9993  max mem: 57114
Epoch: [278]  [ 10/156]  eta: 0:02:25  lr: 0.000122  min_lr: 0.000122  loss: 2.6514 (2.5315)  weight_decay: 0.0500 (0.0500)  time: 0.9982  data: 0.2730  max mem: 57114
Epoch: [278]  [ 20/156]  eta: 0:01:59  lr: 0.000121  min_lr: 0.000121  loss: 2.6763 (2.6114)  weight_decay: 0.0500 (0.0500)  time: 0.7400  data: 0.0004  max mem: 57114
Epoch: [278]  [ 30/156]  eta: 0:01:44  lr: 0.000120  min_lr: 0.000120  loss: 2.7458 (2.5705)  weight_decay: 0.0500 (0.0500)  time: 0.7389  data: 0.0004  max mem: 57114
Epoch: [278]  [ 40/156]  eta: 0:01:33  lr: 0.000119  min_lr: 0.000119  loss: 2.6677 (2.5719)  weight_decay: 0.0500 (0.0500)  time: 0.7284  data: 0.0004  max mem: 57114
Epoch: [278]  [ 50/156]  eta: 0:01:23  lr: 0.000119  min_lr: 0.000119  loss: 2.6598 (2.5660)  weight_decay: 0.0500 (0.0500)  time: 0.7290  data: 0.0004  max mem: 57114
Epoch: [278]  [ 60/156]  eta: 0:01:14  lr: 0.000118  min_lr: 0.000118  loss: 2.6598 (2.5835)  weight_decay: 0.0500 (0.0500)  time: 0.7190  data: 0.0004  max mem: 57114
Epoch: [278]  [ 70/156]  eta: 0:01:05  lr: 0.000117  min_lr: 0.000117  loss: 2.6063 (2.5924)  weight_decay: 0.0500 (0.0500)  time: 0.7039  data: 0.0004  max mem: 57114
Epoch: [278]  [ 80/156]  eta: 0:00:57  lr: 0.000117  min_lr: 0.000117  loss: 2.5584 (2.5824)  weight_decay: 0.0500 (0.0500)  time: 0.7064  data: 0.0004  max mem: 57114
Epoch: [278]  [ 90/156]  eta: 0:00:49  lr: 0.000116  min_lr: 0.000116  loss: 2.6633 (2.5874)  weight_decay: 0.0500 (0.0500)  time: 0.7149  data: 0.0004  max mem: 57114
Epoch: [278]  [100/156]  eta: 0:00:42  lr: 0.000115  min_lr: 0.000115  loss: 2.6428 (2.5818)  weight_decay: 0.0500 (0.0500)  time: 0.7162  data: 0.0004  max mem: 57114
Epoch: [278]  [110/156]  eta: 0:00:34  lr: 0.000115  min_lr: 0.000115  loss: 2.6150 (2.5763)  weight_decay: 0.0500 (0.0500)  time: 0.7069  data: 0.0004  max mem: 57114
Epoch: [278]  [120/156]  eta: 0:00:26  lr: 0.000114  min_lr: 0.000114  loss: 2.5643 (2.5615)  weight_decay: 0.0500 (0.0500)  time: 0.7037  data: 0.0004  max mem: 57114
Epoch: [278]  [130/156]  eta: 0:00:19  lr: 0.000113  min_lr: 0.000113  loss: 2.4336 (2.5476)  weight_decay: 0.0500 (0.0500)  time: 0.6945  data: 0.0009  max mem: 57114
Epoch: [278]  [140/156]  eta: 0:00:11  lr: 0.000113  min_lr: 0.000113  loss: 2.4336 (2.5496)  weight_decay: 0.0500 (0.0500)  time: 0.6830  data: 0.0008  max mem: 57114
Epoch: [278]  [150/156]  eta: 0:00:04  lr: 0.000112  min_lr: 0.000112  loss: 2.4944 (2.5352)  weight_decay: 0.0500 (0.0500)  time: 0.6845  data: 0.0002  max mem: 57114
Epoch: [278]  [155/156]  eta: 0:00:00  lr: 0.000112  min_lr: 0.000112  loss: 2.5171 (2.5440)  weight_decay: 0.0500 (0.0500)  time: 0.6850  data: 0.0001  max mem: 57114
Epoch: [278] Total time: 0:01:54 (0.7321 s / it)
Averaged stats: lr: 0.000112  min_lr: 0.000112  loss: 2.5171 (2.5199)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5961 (0.5961)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.0965  data: 1.8407  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7699 (0.7191)  acc1: 84.3750 (84.9105)  acc5: 98.4375 (97.3146)  time: 0.5700  data: 0.3682  max mem: 57114
Test: Total time: 0:00:02 (0.5922 s / it)
* Acc@1 84.545 Acc@5 96.991 loss 0.805
Accuracy of the model on the 50000 test images: 84.5%
Max accuracy: 84.63%
Test:  [0/5]  eta: 0:00:11  loss: 0.6184 (0.6184)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.3275  data: 2.0839  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7025 (0.7021)  acc1: 81.2500 (79.2839)  acc5: 97.3958 (94.5013)  time: 0.6162  data: 0.4169  max mem: 57114
Test: Total time: 0:00:03 (0.6298 s / it)
* Acc@1 77.663 Acc@5 93.870 loss 0.842
Accuracy of the model EMA on 50000 test images: 77.7%
Max EMA accuracy: 77.66%
Epoch: [279]  [  0/156]  eta: 0:08:13  lr: 0.000112  min_lr: 0.000112  loss: 2.7675 (2.7675)  weight_decay: 0.0500 (0.0500)  time: 3.1649  data: 2.5151  max mem: 57114
Epoch: [279]  [ 10/156]  eta: 0:02:17  lr: 0.000111  min_lr: 0.000111  loss: 2.4939 (2.4161)  weight_decay: 0.0500 (0.0500)  time: 0.9424  data: 0.2289  max mem: 57114
Epoch: [279]  [ 20/156]  eta: 0:01:54  lr: 0.000110  min_lr: 0.000110  loss: 2.4052 (2.4082)  weight_decay: 0.0500 (0.0500)  time: 0.7243  data: 0.0003  max mem: 57114
Epoch: [279]  [ 30/156]  eta: 0:01:41  lr: 0.000110  min_lr: 0.000110  loss: 2.5672 (2.4718)  weight_decay: 0.0500 (0.0500)  time: 0.7246  data: 0.0003  max mem: 57114
Epoch: [279]  [ 40/156]  eta: 0:01:30  lr: 0.000109  min_lr: 0.000109  loss: 2.5976 (2.4885)  weight_decay: 0.0500 (0.0500)  time: 0.7185  data: 0.0004  max mem: 57114
Epoch: [279]  [ 50/156]  eta: 0:01:21  lr: 0.000108  min_lr: 0.000108  loss: 2.5682 (2.4819)  weight_decay: 0.0500 (0.0500)  time: 0.7199  data: 0.0004  max mem: 57114
Epoch: [279]  [ 60/156]  eta: 0:01:13  lr: 0.000108  min_lr: 0.000108  loss: 2.5611 (2.4551)  weight_decay: 0.0500 (0.0500)  time: 0.7183  data: 0.0004  max mem: 57114
Epoch: [279]  [ 70/156]  eta: 0:01:05  lr: 0.000107  min_lr: 0.000107  loss: 2.5735 (2.4690)  weight_decay: 0.0500 (0.0500)  time: 0.7222  data: 0.0004  max mem: 57114
Epoch: [279]  [ 80/156]  eta: 0:00:57  lr: 0.000106  min_lr: 0.000106  loss: 2.6889 (2.4976)  weight_decay: 0.0500 (0.0500)  time: 0.7270  data: 0.0004  max mem: 57114
Epoch: [279]  [ 90/156]  eta: 0:00:49  lr: 0.000106  min_lr: 0.000106  loss: 2.5843 (2.4905)  weight_decay: 0.0500 (0.0500)  time: 0.7133  data: 0.0004  max mem: 57114
Epoch: [279]  [100/156]  eta: 0:00:41  lr: 0.000105  min_lr: 0.000105  loss: 2.3322 (2.4799)  weight_decay: 0.0500 (0.0500)  time: 0.7042  data: 0.0004  max mem: 57114
Epoch: [279]  [110/156]  eta: 0:00:33  lr: 0.000104  min_lr: 0.000104  loss: 2.5467 (2.4869)  weight_decay: 0.0500 (0.0500)  time: 0.6977  data: 0.0004  max mem: 57114
Epoch: [279]  [120/156]  eta: 0:00:26  lr: 0.000104  min_lr: 0.000104  loss: 2.6755 (2.4852)  weight_decay: 0.0500 (0.0500)  time: 0.7048  data: 0.0004  max mem: 57114
Epoch: [279]  [130/156]  eta: 0:00:19  lr: 0.000103  min_lr: 0.000103  loss: 2.6444 (2.5002)  weight_decay: 0.0500 (0.0500)  time: 0.7042  data: 0.0010  max mem: 57114
Epoch: [279]  [140/156]  eta: 0:00:11  lr: 0.000102  min_lr: 0.000102  loss: 2.6444 (2.5042)  weight_decay: 0.0500 (0.0500)  time: 0.6873  data: 0.0008  max mem: 57114
Epoch: [279]  [150/156]  eta: 0:00:04  lr: 0.000102  min_lr: 0.000102  loss: 2.6368 (2.5086)  weight_decay: 0.0500 (0.0500)  time: 0.6888  data: 0.0001  max mem: 57114
Epoch: [279]  [155/156]  eta: 0:00:00  lr: 0.000101  min_lr: 0.000101  loss: 2.5307 (2.5094)  weight_decay: 0.0500 (0.0500)  time: 0.6889  data: 0.0001  max mem: 57114
Epoch: [279] Total time: 0:01:53 (0.7278 s / it)
Averaged stats: lr: 0.000101  min_lr: 0.000101  loss: 2.5307 (2.5278)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5715 (0.5715)  acc1: 90.6250 (90.6250)  acc5: 98.4375 (98.4375)  time: 2.0948  data: 1.8364  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7594 (0.6950)  acc1: 84.3750 (85.0384)  acc5: 98.4375 (97.4425)  time: 0.5697  data: 0.3673  max mem: 57114
Test: Total time: 0:00:02 (0.5927 s / it)
* Acc@1 84.721 Acc@5 97.091 loss 0.780
Accuracy of the model on the 50000 test images: 84.7%
Max accuracy: 84.72%
Test:  [0/5]  eta: 0:00:09  loss: 0.6182 (0.6182)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 1.9789  data: 1.7354  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7021 (0.7020)  acc1: 81.2500 (79.4118)  acc5: 97.3958 (94.5013)  time: 0.5464  data: 0.3471  max mem: 57114
Test: Total time: 0:00:02 (0.5564 s / it)
* Acc@1 77.697 Acc@5 93.884 loss 0.842
Accuracy of the model EMA on 50000 test images: 77.7%
Max EMA accuracy: 77.70%
Epoch: [280]  [  0/156]  eta: 0:08:04  lr: 0.000101  min_lr: 0.000101  loss: 2.0101 (2.0101)  weight_decay: 0.0500 (0.0500)  time: 3.1062  data: 2.4455  max mem: 57114
Epoch: [280]  [ 10/156]  eta: 0:02:19  lr: 0.000101  min_lr: 0.000101  loss: 2.6123 (2.5334)  weight_decay: 0.0500 (0.0500)  time: 0.9576  data: 0.2227  max mem: 57114
Epoch: [280]  [ 20/156]  eta: 0:01:54  lr: 0.000100  min_lr: 0.000100  loss: 2.6640 (2.5595)  weight_decay: 0.0500 (0.0500)  time: 0.7288  data: 0.0004  max mem: 57114
Epoch: [280]  [ 30/156]  eta: 0:01:42  lr: 0.000099  min_lr: 0.000099  loss: 2.6287 (2.5270)  weight_decay: 0.0500 (0.0500)  time: 0.7307  data: 0.0004  max mem: 57114
Epoch: [280]  [ 40/156]  eta: 0:01:31  lr: 0.000099  min_lr: 0.000099  loss: 2.5919 (2.5726)  weight_decay: 0.0500 (0.0500)  time: 0.7398  data: 0.0004  max mem: 57114
Epoch: [280]  [ 50/156]  eta: 0:01:22  lr: 0.000098  min_lr: 0.000098  loss: 2.6243 (2.5325)  weight_decay: 0.0500 (0.0500)  time: 0.7185  data: 0.0004  max mem: 57114
Epoch: [280]  [ 60/156]  eta: 0:01:14  lr: 0.000097  min_lr: 0.000097  loss: 2.6243 (2.5189)  weight_decay: 0.0500 (0.0500)  time: 0.7315  data: 0.0004  max mem: 57114
Epoch: [280]  [ 70/156]  eta: 0:01:05  lr: 0.000097  min_lr: 0.000097  loss: 2.5152 (2.5166)  weight_decay: 0.0500 (0.0500)  time: 0.7337  data: 0.0004  max mem: 57114
Epoch: [280]  [ 80/156]  eta: 0:00:57  lr: 0.000096  min_lr: 0.000096  loss: 2.5083 (2.5138)  weight_decay: 0.0500 (0.0500)  time: 0.7135  data: 0.0004  max mem: 57114
Epoch: [280]  [ 90/156]  eta: 0:00:49  lr: 0.000096  min_lr: 0.000096  loss: 2.4619 (2.4801)  weight_decay: 0.0500 (0.0500)  time: 0.7170  data: 0.0004  max mem: 57114
Epoch: [280]  [100/156]  eta: 0:00:41  lr: 0.000095  min_lr: 0.000095  loss: 2.4251 (2.4770)  weight_decay: 0.0500 (0.0500)  time: 0.7038  data: 0.0004  max mem: 57114
Epoch: [280]  [110/156]  eta: 0:00:34  lr: 0.000094  min_lr: 0.000094  loss: 2.6519 (2.4847)  weight_decay: 0.0500 (0.0500)  time: 0.7015  data: 0.0004  max mem: 57114
Epoch: [280]  [120/156]  eta: 0:00:26  lr: 0.000094  min_lr: 0.000094  loss: 2.6109 (2.4612)  weight_decay: 0.0500 (0.0500)  time: 0.7052  data: 0.0003  max mem: 57114
Epoch: [280]  [130/156]  eta: 0:00:19  lr: 0.000093  min_lr: 0.000093  loss: 2.3724 (2.4595)  weight_decay: 0.0500 (0.0500)  time: 0.7012  data: 0.0008  max mem: 57114
Epoch: [280]  [140/156]  eta: 0:00:11  lr: 0.000093  min_lr: 0.000093  loss: 2.6619 (2.4804)  weight_decay: 0.0500 (0.0500)  time: 0.6949  data: 0.0007  max mem: 57114
Epoch: [280]  [150/156]  eta: 0:00:04  lr: 0.000092  min_lr: 0.000092  loss: 2.6904 (2.4923)  weight_decay: 0.0500 (0.0500)  time: 0.6868  data: 0.0001  max mem: 57114
Epoch: [280]  [155/156]  eta: 0:00:00  lr: 0.000092  min_lr: 0.000092  loss: 2.6255 (2.4873)  weight_decay: 0.0500 (0.0500)  time: 0.6856  data: 0.0001  max mem: 57114
Epoch: [280] Total time: 0:01:54 (0.7310 s / it)
Averaged stats: lr: 0.000092  min_lr: 0.000092  loss: 2.6255 (2.5200)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5654 (0.5654)  acc1: 90.6250 (90.6250)  acc5: 98.9583 (98.9583)  time: 2.1104  data: 1.8546  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7381 (0.6843)  acc1: 83.8542 (84.7826)  acc5: 98.9583 (97.6982)  time: 0.5728  data: 0.3710  max mem: 57114
Test: Total time: 0:00:02 (0.5964 s / it)
* Acc@1 84.651 Acc@5 97.033 loss 0.765
Accuracy of the model on the 50000 test images: 84.7%
Max accuracy: 84.72%
Test:  [0/5]  eta: 0:00:11  loss: 0.6181 (0.6181)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.3193  data: 2.0759  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7018 (0.7018)  acc1: 81.7708 (79.6675)  acc5: 97.3958 (94.5013)  time: 0.6145  data: 0.4153  max mem: 57114
Test: Total time: 0:00:03 (0.6267 s / it)
* Acc@1 77.741 Acc@5 93.892 loss 0.841
Accuracy of the model EMA on 50000 test images: 77.7%
Max EMA accuracy: 77.74%
Epoch: [281]  [  0/156]  eta: 0:06:42  lr: 0.000092  min_lr: 0.000092  loss: 2.0319 (2.0319)  weight_decay: 0.0500 (0.0500)  time: 2.5800  data: 1.9213  max mem: 57114
Epoch: [281]  [ 10/156]  eta: 0:02:09  lr: 0.000091  min_lr: 0.000091  loss: 2.5829 (2.5059)  weight_decay: 0.0500 (0.0500)  time: 0.8840  data: 0.1749  max mem: 57114
Epoch: [281]  [ 20/156]  eta: 0:01:48  lr: 0.000090  min_lr: 0.000090  loss: 2.7030 (2.5965)  weight_decay: 0.0500 (0.0500)  time: 0.7110  data: 0.0003  max mem: 57114
Epoch: [281]  [ 30/156]  eta: 0:01:37  lr: 0.000090  min_lr: 0.000090  loss: 2.5181 (2.5277)  weight_decay: 0.0500 (0.0500)  time: 0.7102  data: 0.0004  max mem: 57114
Epoch: [281]  [ 40/156]  eta: 0:01:28  lr: 0.000089  min_lr: 0.000089  loss: 2.4610 (2.5294)  weight_decay: 0.0500 (0.0500)  time: 0.7172  data: 0.0004  max mem: 57114
Epoch: [281]  [ 50/156]  eta: 0:01:20  lr: 0.000089  min_lr: 0.000089  loss: 2.5784 (2.5171)  weight_decay: 0.0500 (0.0500)  time: 0.7378  data: 0.0004  max mem: 57114
Epoch: [281]  [ 60/156]  eta: 0:01:12  lr: 0.000088  min_lr: 0.000088  loss: 2.3786 (2.4867)  weight_decay: 0.0500 (0.0500)  time: 0.7333  data: 0.0004  max mem: 57114
Epoch: [281]  [ 70/156]  eta: 0:01:04  lr: 0.000087  min_lr: 0.000087  loss: 2.6111 (2.5108)  weight_decay: 0.0500 (0.0500)  time: 0.7231  data: 0.0004  max mem: 57114
Epoch: [281]  [ 80/156]  eta: 0:00:56  lr: 0.000087  min_lr: 0.000087  loss: 2.6111 (2.5168)  weight_decay: 0.0500 (0.0500)  time: 0.7111  data: 0.0004  max mem: 57114
Epoch: [281]  [ 90/156]  eta: 0:00:48  lr: 0.000086  min_lr: 0.000086  loss: 2.6211 (2.5299)  weight_decay: 0.0500 (0.0500)  time: 0.6963  data: 0.0004  max mem: 57114
Epoch: [281]  [100/156]  eta: 0:00:41  lr: 0.000086  min_lr: 0.000086  loss: 2.7173 (2.5435)  weight_decay: 0.0500 (0.0500)  time: 0.7038  data: 0.0004  max mem: 57114
Epoch: [281]  [110/156]  eta: 0:00:33  lr: 0.000085  min_lr: 0.000085  loss: 2.6497 (2.5286)  weight_decay: 0.0500 (0.0500)  time: 0.7089  data: 0.0004  max mem: 57114
Epoch: [281]  [120/156]  eta: 0:00:26  lr: 0.000084  min_lr: 0.000084  loss: 2.4717 (2.5220)  weight_decay: 0.0500 (0.0500)  time: 0.7068  data: 0.0003  max mem: 57114
Epoch: [281]  [130/156]  eta: 0:00:18  lr: 0.000084  min_lr: 0.000084  loss: 2.5929 (2.5235)  weight_decay: 0.0500 (0.0500)  time: 0.7011  data: 0.0010  max mem: 57114
Epoch: [281]  [140/156]  eta: 0:00:11  lr: 0.000083  min_lr: 0.000083  loss: 2.6425 (2.5316)  weight_decay: 0.0500 (0.0500)  time: 0.6930  data: 0.0009  max mem: 57114
Epoch: [281]  [150/156]  eta: 0:00:04  lr: 0.000083  min_lr: 0.000083  loss: 2.5734 (2.5229)  weight_decay: 0.0500 (0.0500)  time: 0.6787  data: 0.0001  max mem: 57114
Epoch: [281]  [155/156]  eta: 0:00:00  lr: 0.000082  min_lr: 0.000082  loss: 2.6197 (2.5322)  weight_decay: 0.0500 (0.0500)  time: 0.6762  data: 0.0001  max mem: 57114
Epoch: [281] Total time: 0:01:52 (0.7218 s / it)
Averaged stats: lr: 0.000082  min_lr: 0.000082  loss: 2.6197 (2.5195)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 0.5799 (0.5799)  acc1: 91.6667 (91.6667)  acc5: 98.4375 (98.4375)  time: 2.2449  data: 1.9916  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7706 (0.7116)  acc1: 83.3333 (84.7826)  acc5: 98.4375 (97.6982)  time: 0.5996  data: 0.3984  max mem: 57114
Test: Total time: 0:00:03 (0.6244 s / it)
* Acc@1 84.709 Acc@5 97.061 loss 0.790
Accuracy of the model on the 50000 test images: 84.7%
Max accuracy: 84.72%
Test:  [0/5]  eta: 0:00:11  loss: 0.6175 (0.6175)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.3023  data: 2.0589  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7012 (0.7014)  acc1: 81.7708 (79.7954)  acc5: 97.3958 (94.5013)  time: 0.6111  data: 0.4119  max mem: 57114
Test: Total time: 0:00:03 (0.6268 s / it)
* Acc@1 77.757 Acc@5 93.908 loss 0.841
Accuracy of the model EMA on 50000 test images: 77.8%
Max EMA accuracy: 77.76%
Epoch: [282]  [  0/156]  eta: 0:07:52  lr: 0.000082  min_lr: 0.000082  loss: 2.7780 (2.7780)  weight_decay: 0.0500 (0.0500)  time: 3.0257  data: 2.3717  max mem: 57114
Epoch: [282]  [ 10/156]  eta: 0:02:12  lr: 0.000082  min_lr: 0.000082  loss: 2.7384 (2.6031)  weight_decay: 0.0500 (0.0500)  time: 0.9080  data: 0.2159  max mem: 57114
Epoch: [282]  [ 20/156]  eta: 0:01:52  lr: 0.000081  min_lr: 0.000081  loss: 2.5619 (2.5258)  weight_decay: 0.0500 (0.0500)  time: 0.7150  data: 0.0003  max mem: 57114
Epoch: [282]  [ 30/156]  eta: 0:01:39  lr: 0.000081  min_lr: 0.000081  loss: 2.4692 (2.4633)  weight_decay: 0.0500 (0.0500)  time: 0.7281  data: 0.0004  max mem: 57114
Epoch: [282]  [ 40/156]  eta: 0:01:29  lr: 0.000080  min_lr: 0.000080  loss: 2.5466 (2.5219)  weight_decay: 0.0500 (0.0500)  time: 0.7219  data: 0.0004  max mem: 57114
Epoch: [282]  [ 50/156]  eta: 0:01:21  lr: 0.000079  min_lr: 0.000079  loss: 2.6479 (2.5355)  weight_decay: 0.0500 (0.0500)  time: 0.7323  data: 0.0004  max mem: 57114
Epoch: [282]  [ 60/156]  eta: 0:01:12  lr: 0.000079  min_lr: 0.000079  loss: 2.5866 (2.5310)  weight_decay: 0.0500 (0.0500)  time: 0.7188  data: 0.0004  max mem: 57114
Epoch: [282]  [ 70/156]  eta: 0:01:04  lr: 0.000078  min_lr: 0.000078  loss: 2.5687 (2.5242)  weight_decay: 0.0500 (0.0500)  time: 0.6923  data: 0.0004  max mem: 57114
Epoch: [282]  [ 80/156]  eta: 0:00:56  lr: 0.000078  min_lr: 0.000078  loss: 2.5604 (2.5387)  weight_decay: 0.0500 (0.0500)  time: 0.6986  data: 0.0004  max mem: 57114
Epoch: [282]  [ 90/156]  eta: 0:00:48  lr: 0.000077  min_lr: 0.000077  loss: 2.6537 (2.5318)  weight_decay: 0.0500 (0.0500)  time: 0.7122  data: 0.0004  max mem: 57114
Epoch: [282]  [100/156]  eta: 0:00:41  lr: 0.000077  min_lr: 0.000077  loss: 2.6818 (2.5245)  weight_decay: 0.0500 (0.0500)  time: 0.7163  data: 0.0004  max mem: 57114
Epoch: [282]  [110/156]  eta: 0:00:33  lr: 0.000076  min_lr: 0.000076  loss: 2.5424 (2.5240)  weight_decay: 0.0500 (0.0500)  time: 0.7117  data: 0.0004  max mem: 57114
Epoch: [282]  [120/156]  eta: 0:00:26  lr: 0.000076  min_lr: 0.000076  loss: 2.6644 (2.5406)  weight_decay: 0.0500 (0.0500)  time: 0.7102  data: 0.0004  max mem: 57114
Epoch: [282]  [130/156]  eta: 0:00:18  lr: 0.000075  min_lr: 0.000075  loss: 2.6857 (2.5490)  weight_decay: 0.0500 (0.0500)  time: 0.7051  data: 0.0008  max mem: 57114
Epoch: [282]  [140/156]  eta: 0:00:11  lr: 0.000074  min_lr: 0.000074  loss: 2.6800 (2.5483)  weight_decay: 0.0500 (0.0500)  time: 0.6869  data: 0.0007  max mem: 57114
Epoch: [282]  [150/156]  eta: 0:00:04  lr: 0.000074  min_lr: 0.000074  loss: 2.6228 (2.5441)  weight_decay: 0.0500 (0.0500)  time: 0.6796  data: 0.0001  max mem: 57114
Epoch: [282]  [155/156]  eta: 0:00:00  lr: 0.000074  min_lr: 0.000074  loss: 2.6228 (2.5454)  weight_decay: 0.0500 (0.0500)  time: 0.6837  data: 0.0001  max mem: 57114
Epoch: [282] Total time: 0:01:53 (0.7244 s / it)
Averaged stats: lr: 0.000074  min_lr: 0.000074  loss: 2.6228 (2.5120)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5702 (0.5702)  acc1: 90.6250 (90.6250)  acc5: 98.4375 (98.4375)  time: 2.0697  data: 1.8142  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7490 (0.6934)  acc1: 83.8542 (84.3990)  acc5: 98.4375 (97.5703)  time: 0.5646  data: 0.3629  max mem: 57114
Test: Total time: 0:00:02 (0.5844 s / it)
* Acc@1 84.687 Acc@5 97.029 loss 0.777
Accuracy of the model on the 50000 test images: 84.7%
Max accuracy: 84.72%
Test:  [0/5]  eta: 0:00:11  loss: 0.6171 (0.6171)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.3623  data: 2.1187  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7006 (0.7011)  acc1: 81.7708 (79.7954)  acc5: 97.3958 (94.5013)  time: 0.6232  data: 0.4238  max mem: 57114
Test: Total time: 0:00:03 (0.6346 s / it)
* Acc@1 77.791 Acc@5 93.918 loss 0.840
Accuracy of the model EMA on 50000 test images: 77.8%
Max EMA accuracy: 77.79%
Epoch: [283]  [  0/156]  eta: 0:07:03  lr: 0.000074  min_lr: 0.000074  loss: 2.7463 (2.7463)  weight_decay: 0.0500 (0.0500)  time: 2.7157  data: 2.0655  max mem: 57114
Epoch: [283]  [ 10/156]  eta: 0:02:12  lr: 0.000073  min_lr: 0.000073  loss: 2.7463 (2.6880)  weight_decay: 0.0500 (0.0500)  time: 0.9048  data: 0.1881  max mem: 57114
Epoch: [283]  [ 20/156]  eta: 0:01:50  lr: 0.000072  min_lr: 0.000072  loss: 2.7157 (2.6856)  weight_decay: 0.0500 (0.0500)  time: 0.7208  data: 0.0003  max mem: 57114
Epoch: [283]  [ 30/156]  eta: 0:01:38  lr: 0.000072  min_lr: 0.000072  loss: 2.7080 (2.6388)  weight_decay: 0.0500 (0.0500)  time: 0.7172  data: 0.0004  max mem: 57114
Epoch: [283]  [ 40/156]  eta: 0:01:29  lr: 0.000071  min_lr: 0.000071  loss: 2.7357 (2.6669)  weight_decay: 0.0500 (0.0500)  time: 0.7189  data: 0.0003  max mem: 57114
Epoch: [283]  [ 50/156]  eta: 0:01:20  lr: 0.000071  min_lr: 0.000071  loss: 2.7862 (2.6503)  weight_decay: 0.0500 (0.0500)  time: 0.7138  data: 0.0004  max mem: 57114
Epoch: [283]  [ 60/156]  eta: 0:01:11  lr: 0.000070  min_lr: 0.000070  loss: 2.5500 (2.6036)  weight_decay: 0.0500 (0.0500)  time: 0.7095  data: 0.0004  max mem: 57114
Epoch: [283]  [ 70/156]  eta: 0:01:04  lr: 0.000070  min_lr: 0.000070  loss: 2.4660 (2.5957)  weight_decay: 0.0500 (0.0500)  time: 0.7147  data: 0.0004  max mem: 57114
Epoch: [283]  [ 80/156]  eta: 0:00:56  lr: 0.000069  min_lr: 0.000069  loss: 2.6602 (2.6022)  weight_decay: 0.0500 (0.0500)  time: 0.7238  data: 0.0003  max mem: 57114
Epoch: [283]  [ 90/156]  eta: 0:00:48  lr: 0.000069  min_lr: 0.000069  loss: 2.6602 (2.5967)  weight_decay: 0.0500 (0.0500)  time: 0.7248  data: 0.0003  max mem: 57114
Epoch: [283]  [100/156]  eta: 0:00:41  lr: 0.000068  min_lr: 0.000068  loss: 2.7206 (2.6046)  weight_decay: 0.0500 (0.0500)  time: 0.7089  data: 0.0004  max mem: 57114
Epoch: [283]  [110/156]  eta: 0:00:33  lr: 0.000068  min_lr: 0.000068  loss: 2.6347 (2.5866)  weight_decay: 0.0500 (0.0500)  time: 0.7021  data: 0.0004  max mem: 57114
Epoch: [283]  [120/156]  eta: 0:00:26  lr: 0.000067  min_lr: 0.000067  loss: 2.5466 (2.5815)  weight_decay: 0.0500 (0.0500)  time: 0.7073  data: 0.0004  max mem: 57114
Epoch: [283]  [130/156]  eta: 0:00:18  lr: 0.000067  min_lr: 0.000067  loss: 2.5845 (2.5796)  weight_decay: 0.0500 (0.0500)  time: 0.7010  data: 0.0008  max mem: 57114
Epoch: [283]  [140/156]  eta: 0:00:11  lr: 0.000066  min_lr: 0.000066  loss: 2.4853 (2.5553)  weight_decay: 0.0500 (0.0500)  time: 0.6873  data: 0.0007  max mem: 57114
Epoch: [283]  [150/156]  eta: 0:00:04  lr: 0.000066  min_lr: 0.000066  loss: 2.5443 (2.5631)  weight_decay: 0.0500 (0.0500)  time: 0.6851  data: 0.0001  max mem: 57114
Epoch: [283]  [155/156]  eta: 0:00:00  lr: 0.000065  min_lr: 0.000065  loss: 2.5885 (2.5652)  weight_decay: 0.0500 (0.0500)  time: 0.6883  data: 0.0001  max mem: 57114
Epoch: [283] Total time: 0:01:52 (0.7239 s / it)
Averaged stats: lr: 0.000065  min_lr: 0.000065  loss: 2.5885 (2.5152)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6114 (0.6114)  acc1: 90.6250 (90.6250)  acc5: 98.4375 (98.4375)  time: 2.1080  data: 1.8518  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7970 (0.7374)  acc1: 83.8542 (84.6547)  acc5: 98.4375 (97.5703)  time: 0.5724  data: 0.3704  max mem: 57114
Test: Total time: 0:00:02 (0.5942 s / it)
* Acc@1 84.657 Acc@5 97.033 loss 0.820
Accuracy of the model on the 50000 test images: 84.7%
Max accuracy: 84.72%
Test:  [0/5]  eta: 0:00:11  loss: 0.6163 (0.6163)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 2.3043  data: 2.0609  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7000 (0.7005)  acc1: 81.7708 (79.9233)  acc5: 97.3958 (94.5013)  time: 0.6120  data: 0.4123  max mem: 57114
Test: Total time: 0:00:03 (0.6250 s / it)
* Acc@1 77.831 Acc@5 93.940 loss 0.839
Accuracy of the model EMA on 50000 test images: 77.8%
Max EMA accuracy: 77.83%
Epoch: [284]  [  0/156]  eta: 0:06:13  lr: 0.000065  min_lr: 0.000065  loss: 1.8651 (1.8651)  weight_decay: 0.0500 (0.0500)  time: 2.3912  data: 1.7376  max mem: 57114
Epoch: [284]  [ 10/156]  eta: 0:02:11  lr: 0.000065  min_lr: 0.000065  loss: 2.5416 (2.4717)  weight_decay: 0.0500 (0.0500)  time: 0.9001  data: 0.1582  max mem: 57114
Epoch: [284]  [ 20/156]  eta: 0:01:51  lr: 0.000064  min_lr: 0.000064  loss: 2.4742 (2.4277)  weight_decay: 0.0500 (0.0500)  time: 0.7391  data: 0.0003  max mem: 57114
Epoch: [284]  [ 30/156]  eta: 0:01:38  lr: 0.000064  min_lr: 0.000064  loss: 2.4772 (2.4982)  weight_decay: 0.0500 (0.0500)  time: 0.7165  data: 0.0004  max mem: 57114
Epoch: [284]  [ 40/156]  eta: 0:01:28  lr: 0.000063  min_lr: 0.000063  loss: 2.7330 (2.4825)  weight_decay: 0.0500 (0.0500)  time: 0.7039  data: 0.0004  max mem: 57114
Epoch: [284]  [ 50/156]  eta: 0:01:19  lr: 0.000063  min_lr: 0.000063  loss: 2.7312 (2.5367)  weight_decay: 0.0500 (0.0500)  time: 0.7111  data: 0.0004  max mem: 57114
Epoch: [284]  [ 60/156]  eta: 0:01:12  lr: 0.000062  min_lr: 0.000062  loss: 2.8083 (2.5514)  weight_decay: 0.0500 (0.0500)  time: 0.7341  data: 0.0004  max mem: 57114
Epoch: [284]  [ 70/156]  eta: 0:01:04  lr: 0.000062  min_lr: 0.000062  loss: 2.6008 (2.5296)  weight_decay: 0.0500 (0.0500)  time: 0.7341  data: 0.0004  max mem: 57114
Epoch: [284]  [ 80/156]  eta: 0:00:56  lr: 0.000061  min_lr: 0.000061  loss: 2.6066 (2.5470)  weight_decay: 0.0500 (0.0500)  time: 0.7196  data: 0.0004  max mem: 57114
Epoch: [284]  [ 90/156]  eta: 0:00:48  lr: 0.000061  min_lr: 0.000061  loss: 2.6556 (2.5513)  weight_decay: 0.0500 (0.0500)  time: 0.7127  data: 0.0004  max mem: 57114
Epoch: [284]  [100/156]  eta: 0:00:41  lr: 0.000060  min_lr: 0.000060  loss: 2.5621 (2.5446)  weight_decay: 0.0500 (0.0500)  time: 0.7055  data: 0.0004  max mem: 57114
Epoch: [284]  [110/156]  eta: 0:00:33  lr: 0.000060  min_lr: 0.000060  loss: 2.4257 (2.5192)  weight_decay: 0.0500 (0.0500)  time: 0.6987  data: 0.0004  max mem: 57114
Epoch: [284]  [120/156]  eta: 0:00:26  lr: 0.000059  min_lr: 0.000059  loss: 2.3880 (2.5055)  weight_decay: 0.0500 (0.0500)  time: 0.7019  data: 0.0004  max mem: 57114
Epoch: [284]  [130/156]  eta: 0:00:18  lr: 0.000059  min_lr: 0.000059  loss: 2.5295 (2.5137)  weight_decay: 0.0500 (0.0500)  time: 0.7044  data: 0.0009  max mem: 57114
Epoch: [284]  [140/156]  eta: 0:00:11  lr: 0.000058  min_lr: 0.000058  loss: 2.5729 (2.5118)  weight_decay: 0.0500 (0.0500)  time: 0.6919  data: 0.0007  max mem: 57114
Epoch: [284]  [150/156]  eta: 0:00:04  lr: 0.000058  min_lr: 0.000058  loss: 2.5729 (2.5025)  weight_decay: 0.0500 (0.0500)  time: 0.6846  data: 0.0001  max mem: 57114
Epoch: [284]  [155/156]  eta: 0:00:00  lr: 0.000058  min_lr: 0.000058  loss: 2.5982 (2.5077)  weight_decay: 0.0500 (0.0500)  time: 0.6843  data: 0.0001  max mem: 57114
Epoch: [284] Total time: 0:01:52 (0.7239 s / it)
Averaged stats: lr: 0.000058  min_lr: 0.000058  loss: 2.5982 (2.5137)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5661 (0.5661)  acc1: 90.6250 (90.6250)  acc5: 98.4375 (98.4375)  time: 2.0450  data: 1.7891  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7463 (0.6932)  acc1: 85.4167 (85.2941)  acc5: 98.4375 (97.5703)  time: 0.5597  data: 0.3579  max mem: 57114
Test: Total time: 0:00:02 (0.5832 s / it)
* Acc@1 84.793 Acc@5 97.079 loss 0.773
Accuracy of the model on the 50000 test images: 84.8%
Max accuracy: 84.79%
Test:  [0/5]  eta: 0:00:09  loss: 0.6157 (0.6157)  acc1: 84.8958 (84.8958)  acc5: 97.3958 (97.3958)  time: 1.9786  data: 1.7351  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6992 (0.6999)  acc1: 81.7708 (79.9233)  acc5: 97.3958 (94.5013)  time: 0.5464  data: 0.3471  max mem: 57114
Test: Total time: 0:00:02 (0.5573 s / it)
* Acc@1 77.869 Acc@5 93.964 loss 0.838
Accuracy of the model EMA on 50000 test images: 77.9%
Max EMA accuracy: 77.87%
Epoch: [285]  [  0/156]  eta: 0:07:19  lr: 0.000058  min_lr: 0.000058  loss: 2.0487 (2.0487)  weight_decay: 0.0500 (0.0500)  time: 2.8171  data: 2.1672  max mem: 57114
Epoch: [285]  [ 10/156]  eta: 0:02:13  lr: 0.000057  min_lr: 0.000057  loss: 2.7155 (2.6322)  weight_decay: 0.0500 (0.0500)  time: 0.9176  data: 0.1973  max mem: 57114
Epoch: [285]  [ 20/156]  eta: 0:01:52  lr: 0.000057  min_lr: 0.000057  loss: 2.6917 (2.5097)  weight_decay: 0.0500 (0.0500)  time: 0.7270  data: 0.0003  max mem: 57114
Epoch: [285]  [ 30/156]  eta: 0:01:41  lr: 0.000056  min_lr: 0.000056  loss: 2.4987 (2.4941)  weight_decay: 0.0500 (0.0500)  time: 0.7384  data: 0.0003  max mem: 57114
Epoch: [285]  [ 40/156]  eta: 0:01:31  lr: 0.000056  min_lr: 0.000056  loss: 2.5763 (2.5146)  weight_decay: 0.0500 (0.0500)  time: 0.7508  data: 0.0003  max mem: 57114
Epoch: [285]  [ 50/156]  eta: 0:01:22  lr: 0.000055  min_lr: 0.000055  loss: 2.5237 (2.4638)  weight_decay: 0.0500 (0.0500)  time: 0.7394  data: 0.0003  max mem: 57114
Epoch: [285]  [ 60/156]  eta: 0:01:13  lr: 0.000055  min_lr: 0.000055  loss: 2.3858 (2.4815)  weight_decay: 0.0500 (0.0500)  time: 0.7316  data: 0.0004  max mem: 57114
Epoch: [285]  [ 70/156]  eta: 0:01:05  lr: 0.000054  min_lr: 0.000054  loss: 2.3858 (2.4527)  weight_decay: 0.0500 (0.0500)  time: 0.7178  data: 0.0004  max mem: 57114
Epoch: [285]  [ 80/156]  eta: 0:00:57  lr: 0.000054  min_lr: 0.000054  loss: 2.4953 (2.4712)  weight_decay: 0.0500 (0.0500)  time: 0.7166  data: 0.0003  max mem: 57114
Epoch: [285]  [ 90/156]  eta: 0:00:49  lr: 0.000053  min_lr: 0.000053  loss: 2.7730 (2.4935)  weight_decay: 0.0500 (0.0500)  time: 0.7219  data: 0.0004  max mem: 57114
Epoch: [285]  [100/156]  eta: 0:00:41  lr: 0.000053  min_lr: 0.000053  loss: 2.4529 (2.4725)  weight_decay: 0.0500 (0.0500)  time: 0.7072  data: 0.0005  max mem: 57114
Epoch: [285]  [110/156]  eta: 0:00:34  lr: 0.000052  min_lr: 0.000052  loss: 2.4529 (2.4687)  weight_decay: 0.0500 (0.0500)  time: 0.7111  data: 0.0004  max mem: 57114
Epoch: [285]  [120/156]  eta: 0:00:26  lr: 0.000052  min_lr: 0.000052  loss: 2.6559 (2.4792)  weight_decay: 0.0500 (0.0500)  time: 0.7111  data: 0.0004  max mem: 57114
Epoch: [285]  [130/156]  eta: 0:00:19  lr: 0.000051  min_lr: 0.000051  loss: 2.6877 (2.4853)  weight_decay: 0.0500 (0.0500)  time: 0.6942  data: 0.0008  max mem: 57114
Epoch: [285]  [140/156]  eta: 0:00:11  lr: 0.000051  min_lr: 0.000051  loss: 2.6935 (2.4910)  weight_decay: 0.0500 (0.0500)  time: 0.6821  data: 0.0007  max mem: 57114
Epoch: [285]  [150/156]  eta: 0:00:04  lr: 0.000051  min_lr: 0.000051  loss: 2.6941 (2.5037)  weight_decay: 0.0500 (0.0500)  time: 0.6812  data: 0.0001  max mem: 57114
Epoch: [285]  [155/156]  eta: 0:00:00  lr: 0.000050  min_lr: 0.000050  loss: 2.6977 (2.5044)  weight_decay: 0.0500 (0.0500)  time: 0.6835  data: 0.0001  max mem: 57114
Epoch: [285] Total time: 0:01:53 (0.7307 s / it)
Averaged stats: lr: 0.000050  min_lr: 0.000050  loss: 2.6977 (2.5112)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5980 (0.5980)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.0521  data: 1.7959  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7663 (0.7226)  acc1: 83.8542 (84.2711)  acc5: 98.4375 (97.4425)  time: 0.5611  data: 0.3592  max mem: 57114
Test: Total time: 0:00:02 (0.5836 s / it)
* Acc@1 84.709 Acc@5 97.055 loss 0.802
Accuracy of the model on the 50000 test images: 84.7%
Max accuracy: 84.79%
Test:  [0/5]  eta: 0:00:12  loss: 0.6150 (0.6150)  acc1: 84.8958 (84.8958)  acc5: 96.8750 (96.8750)  time: 2.4194  data: 2.1758  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6985 (0.6993)  acc1: 81.7708 (79.9233)  acc5: 96.8750 (94.5013)  time: 0.6345  data: 0.4352  max mem: 57114
Test: Total time: 0:00:03 (0.6463 s / it)
* Acc@1 77.925 Acc@5 93.988 loss 0.837
Accuracy of the model EMA on 50000 test images: 77.9%
Max EMA accuracy: 77.93%
Epoch: [286]  [  0/156]  eta: 0:06:55  lr: 0.000050  min_lr: 0.000050  loss: 2.3051 (2.3051)  weight_decay: 0.0500 (0.0500)  time: 2.6626  data: 2.0022  max mem: 57114
Epoch: [286]  [ 10/156]  eta: 0:02:10  lr: 0.000050  min_lr: 0.000050  loss: 2.6897 (2.4922)  weight_decay: 0.0500 (0.0500)  time: 0.8971  data: 0.1823  max mem: 57114
Epoch: [286]  [ 20/156]  eta: 0:01:51  lr: 0.000049  min_lr: 0.000049  loss: 2.5623 (2.4376)  weight_decay: 0.0500 (0.0500)  time: 0.7267  data: 0.0003  max mem: 57114
Epoch: [286]  [ 30/156]  eta: 0:01:39  lr: 0.000049  min_lr: 0.000049  loss: 2.5736 (2.5136)  weight_decay: 0.0500 (0.0500)  time: 0.7321  data: 0.0003  max mem: 57114
Epoch: [286]  [ 40/156]  eta: 0:01:29  lr: 0.000048  min_lr: 0.000048  loss: 2.6281 (2.4966)  weight_decay: 0.0500 (0.0500)  time: 0.7176  data: 0.0004  max mem: 57114
Epoch: [286]  [ 50/156]  eta: 0:01:20  lr: 0.000048  min_lr: 0.000048  loss: 2.6205 (2.5222)  weight_decay: 0.0500 (0.0500)  time: 0.7049  data: 0.0004  max mem: 57114
Epoch: [286]  [ 60/156]  eta: 0:01:12  lr: 0.000048  min_lr: 0.000048  loss: 2.5620 (2.4906)  weight_decay: 0.0500 (0.0500)  time: 0.7240  data: 0.0004  max mem: 57114
Epoch: [286]  [ 70/156]  eta: 0:01:04  lr: 0.000047  min_lr: 0.000047  loss: 2.4241 (2.4898)  weight_decay: 0.0500 (0.0500)  time: 0.7227  data: 0.0004  max mem: 57114
Epoch: [286]  [ 80/156]  eta: 0:00:56  lr: 0.000047  min_lr: 0.000047  loss: 2.4241 (2.4708)  weight_decay: 0.0500 (0.0500)  time: 0.7058  data: 0.0004  max mem: 57114
Epoch: [286]  [ 90/156]  eta: 0:00:48  lr: 0.000046  min_lr: 0.000046  loss: 2.4434 (2.4855)  weight_decay: 0.0500 (0.0500)  time: 0.7111  data: 0.0004  max mem: 57114
Epoch: [286]  [100/156]  eta: 0:00:41  lr: 0.000046  min_lr: 0.000046  loss: 2.7451 (2.5051)  weight_decay: 0.0500 (0.0500)  time: 0.7195  data: 0.0004  max mem: 57114
Epoch: [286]  [110/156]  eta: 0:00:33  lr: 0.000045  min_lr: 0.000045  loss: 2.7451 (2.5164)  weight_decay: 0.0500 (0.0500)  time: 0.7181  data: 0.0004  max mem: 57114
Epoch: [286]  [120/156]  eta: 0:00:26  lr: 0.000045  min_lr: 0.000045  loss: 2.6937 (2.5238)  weight_decay: 0.0500 (0.0500)  time: 0.7070  data: 0.0004  max mem: 57114
Epoch: [286]  [130/156]  eta: 0:00:18  lr: 0.000045  min_lr: 0.000045  loss: 2.6251 (2.5322)  weight_decay: 0.0500 (0.0500)  time: 0.6971  data: 0.0008  max mem: 57114
Epoch: [286]  [140/156]  eta: 0:00:11  lr: 0.000044  min_lr: 0.000044  loss: 2.5842 (2.5290)  weight_decay: 0.0500 (0.0500)  time: 0.6860  data: 0.0007  max mem: 57114
Epoch: [286]  [150/156]  eta: 0:00:04  lr: 0.000044  min_lr: 0.000044  loss: 2.5613 (2.5208)  weight_decay: 0.0500 (0.0500)  time: 0.6825  data: 0.0001  max mem: 57114
Epoch: [286]  [155/156]  eta: 0:00:00  lr: 0.000044  min_lr: 0.000044  loss: 2.5842 (2.5185)  weight_decay: 0.0500 (0.0500)  time: 0.6817  data: 0.0001  max mem: 57114
Epoch: [286] Total time: 0:01:52 (0.7240 s / it)
Averaged stats: lr: 0.000044  min_lr: 0.000044  loss: 2.5842 (2.5045)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5670 (0.5670)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.1480  data: 1.8927  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7437 (0.6922)  acc1: 83.8542 (84.3990)  acc5: 98.4375 (97.3146)  time: 0.5802  data: 0.3786  max mem: 57114
Test: Total time: 0:00:03 (0.6051 s / it)
* Acc@1 84.685 Acc@5 97.045 loss 0.774
Accuracy of the model on the 50000 test images: 84.7%
Max accuracy: 84.79%
Test:  [0/5]  eta: 0:00:11  loss: 0.6142 (0.6142)  acc1: 84.8958 (84.8958)  acc5: 96.8750 (96.8750)  time: 2.2818  data: 2.0383  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6976 (0.6986)  acc1: 81.7708 (79.9233)  acc5: 96.8750 (94.5013)  time: 0.6070  data: 0.4077  max mem: 57114
Test: Total time: 0:00:03 (0.6181 s / it)
* Acc@1 77.949 Acc@5 94.002 loss 0.836
Accuracy of the model EMA on 50000 test images: 77.9%
Max EMA accuracy: 77.95%
Epoch: [287]  [  0/156]  eta: 0:07:05  lr: 0.000043  min_lr: 0.000043  loss: 2.5977 (2.5977)  weight_decay: 0.0500 (0.0500)  time: 2.7259  data: 2.0650  max mem: 57114
Epoch: [287]  [ 10/156]  eta: 0:02:15  lr: 0.000043  min_lr: 0.000043  loss: 2.5977 (2.5203)  weight_decay: 0.0500 (0.0500)  time: 0.9263  data: 0.1880  max mem: 57114
Epoch: [287]  [ 20/156]  eta: 0:01:53  lr: 0.000043  min_lr: 0.000043  loss: 2.5763 (2.5384)  weight_decay: 0.0500 (0.0500)  time: 0.7401  data: 0.0004  max mem: 57114
Epoch: [287]  [ 30/156]  eta: 0:01:41  lr: 0.000042  min_lr: 0.000042  loss: 2.6164 (2.5391)  weight_decay: 0.0500 (0.0500)  time: 0.7369  data: 0.0004  max mem: 57114
Epoch: [287]  [ 40/156]  eta: 0:01:30  lr: 0.000042  min_lr: 0.000042  loss: 2.5698 (2.5130)  weight_decay: 0.0500 (0.0500)  time: 0.7281  data: 0.0004  max mem: 57114
Epoch: [287]  [ 50/156]  eta: 0:01:21  lr: 0.000041  min_lr: 0.000041  loss: 2.5698 (2.5020)  weight_decay: 0.0500 (0.0500)  time: 0.7186  data: 0.0004  max mem: 57114
Epoch: [287]  [ 60/156]  eta: 0:01:13  lr: 0.000041  min_lr: 0.000041  loss: 2.6130 (2.5120)  weight_decay: 0.0500 (0.0500)  time: 0.7167  data: 0.0003  max mem: 57114
Epoch: [287]  [ 70/156]  eta: 0:01:04  lr: 0.000041  min_lr: 0.000041  loss: 2.4508 (2.4926)  weight_decay: 0.0500 (0.0500)  time: 0.7153  data: 0.0004  max mem: 57114
Epoch: [287]  [ 80/156]  eta: 0:00:56  lr: 0.000040  min_lr: 0.000040  loss: 2.2803 (2.4795)  weight_decay: 0.0500 (0.0500)  time: 0.7078  data: 0.0004  max mem: 57114
Epoch: [287]  [ 90/156]  eta: 0:00:49  lr: 0.000040  min_lr: 0.000040  loss: 2.4824 (2.4829)  weight_decay: 0.0500 (0.0500)  time: 0.7054  data: 0.0004  max mem: 57114
Epoch: [287]  [100/156]  eta: 0:00:41  lr: 0.000039  min_lr: 0.000039  loss: 2.4824 (2.4636)  weight_decay: 0.0500 (0.0500)  time: 0.7123  data: 0.0004  max mem: 57114
Epoch: [287]  [110/156]  eta: 0:00:33  lr: 0.000039  min_lr: 0.000039  loss: 2.3190 (2.4653)  weight_decay: 0.0500 (0.0500)  time: 0.7121  data: 0.0004  max mem: 57114
Epoch: [287]  [120/156]  eta: 0:00:26  lr: 0.000039  min_lr: 0.000039  loss: 2.5500 (2.4781)  weight_decay: 0.0500 (0.0500)  time: 0.7116  data: 0.0004  max mem: 57114
Epoch: [287]  [130/156]  eta: 0:00:19  lr: 0.000038  min_lr: 0.000038  loss: 2.5588 (2.4704)  weight_decay: 0.0500 (0.0500)  time: 0.7050  data: 0.0004  max mem: 57114
Epoch: [287]  [140/156]  eta: 0:00:11  lr: 0.000038  min_lr: 0.000038  loss: 2.5314 (2.4693)  weight_decay: 0.0500 (0.0500)  time: 0.6916  data: 0.0003  max mem: 57114
Epoch: [287]  [150/156]  eta: 0:00:04  lr: 0.000037  min_lr: 0.000037  loss: 2.5206 (2.4703)  weight_decay: 0.0500 (0.0500)  time: 0.6860  data: 0.0001  max mem: 57114
Epoch: [287]  [155/156]  eta: 0:00:00  lr: 0.000037  min_lr: 0.000037  loss: 2.5743 (2.4714)  weight_decay: 0.0500 (0.0500)  time: 0.6867  data: 0.0001  max mem: 57114
Epoch: [287] Total time: 0:01:53 (0.7276 s / it)
Averaged stats: lr: 0.000037  min_lr: 0.000037  loss: 2.5743 (2.5020)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5495 (0.5495)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.0501  data: 1.7941  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7302 (0.6758)  acc1: 85.4167 (84.6547)  acc5: 98.4375 (97.4425)  time: 0.5607  data: 0.3589  max mem: 57114
Test: Total time: 0:00:02 (0.5837 s / it)
* Acc@1 84.685 Acc@5 97.015 loss 0.758
Accuracy of the model on the 50000 test images: 84.7%
Max accuracy: 84.79%
Test:  [0/5]  eta: 0:00:11  loss: 0.6133 (0.6133)  acc1: 85.9375 (85.9375)  acc5: 96.8750 (96.8750)  time: 2.3188  data: 2.0752  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6967 (0.6977)  acc1: 81.7708 (80.1790)  acc5: 96.8750 (94.5013)  time: 0.6144  data: 0.4151  max mem: 57114
Test: Total time: 0:00:03 (0.6297 s / it)
* Acc@1 77.985 Acc@5 94.022 loss 0.835
Accuracy of the model EMA on 50000 test images: 78.0%
Max EMA accuracy: 77.99%
Epoch: [288]  [  0/156]  eta: 0:08:29  lr: 0.000037  min_lr: 0.000037  loss: 2.9290 (2.9290)  weight_decay: 0.0500 (0.0500)  time: 3.2635  data: 2.6115  max mem: 57114
Epoch: [288]  [ 10/156]  eta: 0:02:18  lr: 0.000037  min_lr: 0.000037  loss: 2.7100 (2.4954)  weight_decay: 0.0500 (0.0500)  time: 0.9477  data: 0.2377  max mem: 57114
Epoch: [288]  [ 20/156]  eta: 0:01:54  lr: 0.000036  min_lr: 0.000036  loss: 2.6650 (2.5679)  weight_decay: 0.0500 (0.0500)  time: 0.7188  data: 0.0003  max mem: 57114
Epoch: [288]  [ 30/156]  eta: 0:01:41  lr: 0.000036  min_lr: 0.000036  loss: 2.6228 (2.5639)  weight_decay: 0.0500 (0.0500)  time: 0.7236  data: 0.0003  max mem: 57114
Epoch: [288]  [ 40/156]  eta: 0:01:31  lr: 0.000036  min_lr: 0.000036  loss: 2.6119 (2.5470)  weight_decay: 0.0500 (0.0500)  time: 0.7263  data: 0.0004  max mem: 57114
Epoch: [288]  [ 50/156]  eta: 0:01:21  lr: 0.000035  min_lr: 0.000035  loss: 2.6867 (2.5441)  weight_decay: 0.0500 (0.0500)  time: 0.7226  data: 0.0004  max mem: 57114
Epoch: [288]  [ 60/156]  eta: 0:01:13  lr: 0.000035  min_lr: 0.000035  loss: 2.5066 (2.5137)  weight_decay: 0.0500 (0.0500)  time: 0.7286  data: 0.0003  max mem: 57114
Epoch: [288]  [ 70/156]  eta: 0:01:05  lr: 0.000035  min_lr: 0.000035  loss: 2.4245 (2.5211)  weight_decay: 0.0500 (0.0500)  time: 0.7287  data: 0.0004  max mem: 57114
Epoch: [288]  [ 80/156]  eta: 0:00:57  lr: 0.000034  min_lr: 0.000034  loss: 2.6435 (2.5272)  weight_decay: 0.0500 (0.0500)  time: 0.7037  data: 0.0003  max mem: 57114
Epoch: [288]  [ 90/156]  eta: 0:00:49  lr: 0.000034  min_lr: 0.000034  loss: 2.5801 (2.5201)  weight_decay: 0.0500 (0.0500)  time: 0.6936  data: 0.0004  max mem: 57114
Epoch: [288]  [100/156]  eta: 0:00:41  lr: 0.000033  min_lr: 0.000033  loss: 2.5690 (2.5252)  weight_decay: 0.0500 (0.0500)  time: 0.7047  data: 0.0004  max mem: 57114
Epoch: [288]  [110/156]  eta: 0:00:33  lr: 0.000033  min_lr: 0.000033  loss: 2.6338 (2.5313)  weight_decay: 0.0500 (0.0500)  time: 0.7077  data: 0.0004  max mem: 57114
Epoch: [288]  [120/156]  eta: 0:00:26  lr: 0.000033  min_lr: 0.000033  loss: 2.6341 (2.5350)  weight_decay: 0.0500 (0.0500)  time: 0.7143  data: 0.0004  max mem: 57114
Epoch: [288]  [130/156]  eta: 0:00:19  lr: 0.000032  min_lr: 0.000032  loss: 2.4690 (2.5268)  weight_decay: 0.0500 (0.0500)  time: 0.7129  data: 0.0009  max mem: 57114
Epoch: [288]  [140/156]  eta: 0:00:11  lr: 0.000032  min_lr: 0.000032  loss: 2.4469 (2.5256)  weight_decay: 0.0500 (0.0500)  time: 0.6874  data: 0.0007  max mem: 57114
Epoch: [288]  [150/156]  eta: 0:00:04  lr: 0.000032  min_lr: 0.000032  loss: 2.6307 (2.5376)  weight_decay: 0.0500 (0.0500)  time: 0.6802  data: 0.0001  max mem: 57114
Epoch: [288]  [155/156]  eta: 0:00:00  lr: 0.000031  min_lr: 0.000031  loss: 2.7074 (2.5428)  weight_decay: 0.0500 (0.0500)  time: 0.6818  data: 0.0001  max mem: 57114
Epoch: [288] Total time: 0:01:53 (0.7278 s / it)
Averaged stats: lr: 0.000031  min_lr: 0.000031  loss: 2.7074 (2.5047)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 0.6495 (0.6495)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.2140  data: 1.9581  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8243 (0.7781)  acc1: 84.3750 (84.5269)  acc5: 98.4375 (97.4425)  time: 0.5935  data: 0.3917  max mem: 57114
Test: Total time: 0:00:03 (0.6203 s / it)
* Acc@1 84.669 Acc@5 97.029 loss 0.853
Accuracy of the model on the 50000 test images: 84.7%
Max accuracy: 84.79%
Test:  [0/5]  eta: 0:00:12  loss: 0.6122 (0.6122)  acc1: 85.9375 (85.9375)  acc5: 96.8750 (96.8750)  time: 2.4189  data: 2.1753  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6957 (0.6968)  acc1: 81.7708 (80.1790)  acc5: 96.8750 (94.5013)  time: 0.6345  data: 0.4351  max mem: 57114
Test: Total time: 0:00:03 (0.6460 s / it)
* Acc@1 78.035 Acc@5 94.050 loss 0.834
Accuracy of the model EMA on 50000 test images: 78.0%
Max EMA accuracy: 78.04%
Epoch: [289]  [  0/156]  eta: 0:07:30  lr: 0.000031  min_lr: 0.000031  loss: 2.9510 (2.9510)  weight_decay: 0.0500 (0.0500)  time: 2.8904  data: 2.2397  max mem: 57114
Epoch: [289]  [ 10/156]  eta: 0:02:13  lr: 0.000031  min_lr: 0.000031  loss: 2.5571 (2.6379)  weight_decay: 0.0500 (0.0500)  time: 0.9128  data: 0.2038  max mem: 57114
Epoch: [289]  [ 20/156]  eta: 0:01:51  lr: 0.000031  min_lr: 0.000031  loss: 2.5571 (2.6010)  weight_decay: 0.0500 (0.0500)  time: 0.7135  data: 0.0003  max mem: 57114
Epoch: [289]  [ 30/156]  eta: 0:01:39  lr: 0.000030  min_lr: 0.000030  loss: 2.6174 (2.5680)  weight_decay: 0.0500 (0.0500)  time: 0.7187  data: 0.0003  max mem: 57114
Epoch: [289]  [ 40/156]  eta: 0:01:29  lr: 0.000030  min_lr: 0.000030  loss: 2.6043 (2.5404)  weight_decay: 0.0500 (0.0500)  time: 0.7292  data: 0.0004  max mem: 57114
Epoch: [289]  [ 50/156]  eta: 0:01:20  lr: 0.000030  min_lr: 0.000030  loss: 2.6372 (2.5466)  weight_decay: 0.0500 (0.0500)  time: 0.7262  data: 0.0004  max mem: 57114
Epoch: [289]  [ 60/156]  eta: 0:01:12  lr: 0.000029  min_lr: 0.000029  loss: 2.6668 (2.5410)  weight_decay: 0.0500 (0.0500)  time: 0.7152  data: 0.0004  max mem: 57114
Epoch: [289]  [ 70/156]  eta: 0:01:04  lr: 0.000029  min_lr: 0.000029  loss: 2.6467 (2.5480)  weight_decay: 0.0500 (0.0500)  time: 0.7103  data: 0.0005  max mem: 57114
Epoch: [289]  [ 80/156]  eta: 0:00:56  lr: 0.000029  min_lr: 0.000029  loss: 2.6421 (2.5324)  weight_decay: 0.0500 (0.0500)  time: 0.7101  data: 0.0005  max mem: 57114
Epoch: [289]  [ 90/156]  eta: 0:00:48  lr: 0.000028  min_lr: 0.000028  loss: 2.5703 (2.5395)  weight_decay: 0.0500 (0.0500)  time: 0.7135  data: 0.0005  max mem: 57114
Epoch: [289]  [100/156]  eta: 0:00:41  lr: 0.000028  min_lr: 0.000028  loss: 2.5703 (2.5403)  weight_decay: 0.0500 (0.0500)  time: 0.7095  data: 0.0005  max mem: 57114
Epoch: [289]  [110/156]  eta: 0:00:33  lr: 0.000028  min_lr: 0.000028  loss: 2.4352 (2.5262)  weight_decay: 0.0500 (0.0500)  time: 0.6997  data: 0.0005  max mem: 57114
Epoch: [289]  [120/156]  eta: 0:00:26  lr: 0.000027  min_lr: 0.000027  loss: 2.5858 (2.5244)  weight_decay: 0.0500 (0.0500)  time: 0.7089  data: 0.0004  max mem: 57114
Epoch: [289]  [130/156]  eta: 0:00:18  lr: 0.000027  min_lr: 0.000027  loss: 2.5944 (2.5198)  weight_decay: 0.0500 (0.0500)  time: 0.7088  data: 0.0014  max mem: 57114
Epoch: [289]  [140/156]  eta: 0:00:11  lr: 0.000027  min_lr: 0.000027  loss: 2.5415 (2.5197)  weight_decay: 0.0500 (0.0500)  time: 0.6913  data: 0.0012  max mem: 57114
Epoch: [289]  [150/156]  eta: 0:00:04  lr: 0.000026  min_lr: 0.000026  loss: 2.5415 (2.5198)  weight_decay: 0.0500 (0.0500)  time: 0.6820  data: 0.0001  max mem: 57114
Epoch: [289]  [155/156]  eta: 0:00:00  lr: 0.000026  min_lr: 0.000026  loss: 2.6123 (2.5191)  weight_decay: 0.0500 (0.0500)  time: 0.6817  data: 0.0001  max mem: 57114
Epoch: [289] Total time: 0:01:52 (0.7243 s / it)
Averaged stats: lr: 0.000026  min_lr: 0.000026  loss: 2.6123 (2.5050)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5744 (0.5744)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.0417  data: 1.7860  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7468 (0.6986)  acc1: 85.4167 (84.7826)  acc5: 98.4375 (97.4425)  time: 0.5590  data: 0.3573  max mem: 57114
Test: Total time: 0:00:02 (0.5814 s / it)
* Acc@1 84.697 Acc@5 97.071 loss 0.778
Accuracy of the model on the 50000 test images: 84.7%
Max accuracy: 84.79%
Test:  [0/5]  eta: 0:00:11  loss: 0.6112 (0.6112)  acc1: 85.9375 (85.9375)  acc5: 96.8750 (96.8750)  time: 2.2393  data: 1.9957  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6947 (0.6959)  acc1: 81.7708 (80.1790)  acc5: 96.8750 (94.6292)  time: 0.5985  data: 0.3992  max mem: 57114
Test: Total time: 0:00:03 (0.6142 s / it)
* Acc@1 78.079 Acc@5 94.082 loss 0.832
Accuracy of the model EMA on 50000 test images: 78.1%
Max EMA accuracy: 78.08%
Epoch: [290]  [  0/156]  eta: 0:06:18  lr: 0.000026  min_lr: 0.000026  loss: 1.8939 (1.8939)  weight_decay: 0.0500 (0.0500)  time: 2.4266  data: 1.7791  max mem: 57114
Epoch: [290]  [ 10/156]  eta: 0:02:08  lr: 0.000026  min_lr: 0.000026  loss: 2.1026 (2.2630)  weight_decay: 0.0500 (0.0500)  time: 0.8773  data: 0.1621  max mem: 57114
Epoch: [290]  [ 20/156]  eta: 0:01:49  lr: 0.000026  min_lr: 0.000026  loss: 2.6515 (2.4758)  weight_decay: 0.0500 (0.0500)  time: 0.7255  data: 0.0004  max mem: 57114
Epoch: [290]  [ 30/156]  eta: 0:01:39  lr: 0.000025  min_lr: 0.000025  loss: 2.7228 (2.5068)  weight_decay: 0.0500 (0.0500)  time: 0.7439  data: 0.0005  max mem: 57114
Epoch: [290]  [ 40/156]  eta: 0:01:29  lr: 0.000025  min_lr: 0.000025  loss: 2.6402 (2.5195)  weight_decay: 0.0500 (0.0500)  time: 0.7362  data: 0.0005  max mem: 57114
Epoch: [290]  [ 50/156]  eta: 0:01:20  lr: 0.000025  min_lr: 0.000025  loss: 2.6174 (2.5321)  weight_decay: 0.0500 (0.0500)  time: 0.7114  data: 0.0005  max mem: 57114
Epoch: [290]  [ 60/156]  eta: 0:01:12  lr: 0.000024  min_lr: 0.000024  loss: 2.7401 (2.5570)  weight_decay: 0.0500 (0.0500)  time: 0.7141  data: 0.0005  max mem: 57114
Epoch: [290]  [ 70/156]  eta: 0:01:04  lr: 0.000024  min_lr: 0.000024  loss: 2.7443 (2.5822)  weight_decay: 0.0500 (0.0500)  time: 0.7060  data: 0.0005  max mem: 57114
Epoch: [290]  [ 80/156]  eta: 0:00:56  lr: 0.000024  min_lr: 0.000024  loss: 2.6812 (2.5756)  weight_decay: 0.0500 (0.0500)  time: 0.7170  data: 0.0005  max mem: 57114
Epoch: [290]  [ 90/156]  eta: 0:00:48  lr: 0.000023  min_lr: 0.000023  loss: 2.6033 (2.5458)  weight_decay: 0.0500 (0.0500)  time: 0.7201  data: 0.0006  max mem: 57114
Epoch: [290]  [100/156]  eta: 0:00:41  lr: 0.000023  min_lr: 0.000023  loss: 2.5716 (2.5450)  weight_decay: 0.0500 (0.0500)  time: 0.7039  data: 0.0007  max mem: 57114
Epoch: [290]  [110/156]  eta: 0:00:33  lr: 0.000023  min_lr: 0.000023  loss: 2.6301 (2.5443)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0007  max mem: 57114
Epoch: [290]  [120/156]  eta: 0:00:26  lr: 0.000022  min_lr: 0.000022  loss: 2.5429 (2.5286)  weight_decay: 0.0500 (0.0500)  time: 0.7042  data: 0.0007  max mem: 57114
Epoch: [290]  [130/156]  eta: 0:00:18  lr: 0.000022  min_lr: 0.000022  loss: 2.4638 (2.5219)  weight_decay: 0.0500 (0.0500)  time: 0.6906  data: 0.0012  max mem: 57114
Epoch: [290]  [140/156]  eta: 0:00:11  lr: 0.000022  min_lr: 0.000022  loss: 2.3159 (2.4977)  weight_decay: 0.0500 (0.0500)  time: 0.6848  data: 0.0009  max mem: 57114
Epoch: [290]  [150/156]  eta: 0:00:04  lr: 0.000022  min_lr: 0.000022  loss: 2.4545 (2.5127)  weight_decay: 0.0500 (0.0500)  time: 0.6852  data: 0.0001  max mem: 57114
Epoch: [290]  [155/156]  eta: 0:00:00  lr: 0.000021  min_lr: 0.000021  loss: 2.6882 (2.5168)  weight_decay: 0.0500 (0.0500)  time: 0.6906  data: 0.0001  max mem: 57114
Epoch: [290] Total time: 0:01:52 (0.7234 s / it)
Averaged stats: lr: 0.000021  min_lr: 0.000021  loss: 2.6882 (2.5092)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6281 (0.6281)  acc1: 90.6250 (90.6250)  acc5: 98.4375 (98.4375)  time: 2.0458  data: 1.7900  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8001 (0.7500)  acc1: 84.8958 (84.6547)  acc5: 98.4375 (97.6982)  time: 0.5599  data: 0.3581  max mem: 57114
Test: Total time: 0:00:02 (0.5830 s / it)
* Acc@1 84.727 Acc@5 97.045 loss 0.830
Accuracy of the model on the 50000 test images: 84.7%
Max accuracy: 84.79%
Test:  [0/5]  eta: 0:00:11  loss: 0.6101 (0.6101)  acc1: 85.9375 (85.9375)  acc5: 96.8750 (96.8750)  time: 2.3236  data: 2.0801  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6937 (0.6949)  acc1: 81.7708 (80.1790)  acc5: 96.8750 (94.7570)  time: 0.6154  data: 0.4161  max mem: 57114
Test: Total time: 0:00:03 (0.6271 s / it)
* Acc@1 78.153 Acc@5 94.126 loss 0.831
Accuracy of the model EMA on 50000 test images: 78.2%
Max EMA accuracy: 78.15%
Epoch: [291]  [  0/156]  eta: 0:07:57  lr: 0.000021  min_lr: 0.000021  loss: 2.4369 (2.4369)  weight_decay: 0.0500 (0.0500)  time: 3.0620  data: 2.2816  max mem: 57114
Epoch: [291]  [ 10/156]  eta: 0:02:17  lr: 0.000021  min_lr: 0.000021  loss: 2.6194 (2.5349)  weight_decay: 0.0500 (0.0500)  time: 0.9391  data: 0.2077  max mem: 57114
Epoch: [291]  [ 20/156]  eta: 0:01:52  lr: 0.000021  min_lr: 0.000021  loss: 2.5129 (2.5072)  weight_decay: 0.0500 (0.0500)  time: 0.7191  data: 0.0003  max mem: 57114
Epoch: [291]  [ 30/156]  eta: 0:01:40  lr: 0.000021  min_lr: 0.000021  loss: 2.6319 (2.5125)  weight_decay: 0.0500 (0.0500)  time: 0.7198  data: 0.0003  max mem: 57114
Epoch: [291]  [ 40/156]  eta: 0:01:29  lr: 0.000020  min_lr: 0.000020  loss: 2.6650 (2.5147)  weight_decay: 0.0500 (0.0500)  time: 0.7097  data: 0.0004  max mem: 57114
Epoch: [291]  [ 50/156]  eta: 0:01:20  lr: 0.000020  min_lr: 0.000020  loss: 2.5581 (2.5249)  weight_decay: 0.0500 (0.0500)  time: 0.7074  data: 0.0003  max mem: 57114
Epoch: [291]  [ 60/156]  eta: 0:01:12  lr: 0.000020  min_lr: 0.000020  loss: 2.5581 (2.5072)  weight_decay: 0.0500 (0.0500)  time: 0.7224  data: 0.0004  max mem: 57114
Epoch: [291]  [ 70/156]  eta: 0:01:04  lr: 0.000019  min_lr: 0.000019  loss: 2.5337 (2.5052)  weight_decay: 0.0500 (0.0500)  time: 0.7189  data: 0.0004  max mem: 57114
Epoch: [291]  [ 80/156]  eta: 0:00:56  lr: 0.000019  min_lr: 0.000019  loss: 2.5068 (2.4891)  weight_decay: 0.0500 (0.0500)  time: 0.7219  data: 0.0004  max mem: 57114
Epoch: [291]  [ 90/156]  eta: 0:00:49  lr: 0.000019  min_lr: 0.000019  loss: 2.6655 (2.5205)  weight_decay: 0.0500 (0.0500)  time: 0.7170  data: 0.0004  max mem: 57114
Epoch: [291]  [100/156]  eta: 0:00:41  lr: 0.000019  min_lr: 0.000019  loss: 2.7381 (2.5045)  weight_decay: 0.0500 (0.0500)  time: 0.7049  data: 0.0004  max mem: 57114
Epoch: [291]  [110/156]  eta: 0:00:33  lr: 0.000018  min_lr: 0.000018  loss: 2.6517 (2.5078)  weight_decay: 0.0500 (0.0500)  time: 0.6979  data: 0.0004  max mem: 57114
Epoch: [291]  [120/156]  eta: 0:00:26  lr: 0.000018  min_lr: 0.000018  loss: 2.4140 (2.4954)  weight_decay: 0.0500 (0.0500)  time: 0.7041  data: 0.0004  max mem: 57114
Epoch: [291]  [130/156]  eta: 0:00:18  lr: 0.000018  min_lr: 0.000018  loss: 2.1917 (2.4872)  weight_decay: 0.0500 (0.0500)  time: 0.7043  data: 0.0008  max mem: 57114
Epoch: [291]  [140/156]  eta: 0:00:11  lr: 0.000018  min_lr: 0.000018  loss: 2.0702 (2.4630)  weight_decay: 0.0500 (0.0500)  time: 0.6968  data: 0.0007  max mem: 57114
Epoch: [291]  [150/156]  eta: 0:00:04  lr: 0.000017  min_lr: 0.000017  loss: 2.4423 (2.4762)  weight_decay: 0.0500 (0.0500)  time: 0.6918  data: 0.0001  max mem: 57114
Epoch: [291]  [155/156]  eta: 0:00:00  lr: 0.000017  min_lr: 0.000017  loss: 2.5906 (2.4753)  weight_decay: 0.0500 (0.0500)  time: 0.6864  data: 0.0001  max mem: 57114
Epoch: [291] Total time: 0:01:53 (0.7257 s / it)
Averaged stats: lr: 0.000017  min_lr: 0.000017  loss: 2.5906 (2.5065)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5628 (0.5628)  acc1: 90.6250 (90.6250)  acc5: 98.4375 (98.4375)  time: 2.1020  data: 1.8454  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7401 (0.6913)  acc1: 83.8542 (84.6547)  acc5: 98.4375 (97.4425)  time: 0.5712  data: 0.3692  max mem: 57114
Test: Total time: 0:00:02 (0.5964 s / it)
* Acc@1 84.777 Acc@5 97.083 loss 0.771
Accuracy of the model on the 50000 test images: 84.8%
Max accuracy: 84.79%
Test:  [0/5]  eta: 0:00:11  loss: 0.6089 (0.6089)  acc1: 85.9375 (85.9375)  acc5: 96.8750 (96.8750)  time: 2.2280  data: 1.9842  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6926 (0.6939)  acc1: 81.7708 (80.1790)  acc5: 96.8750 (94.7570)  time: 0.5963  data: 0.3969  max mem: 57114
Test: Total time: 0:00:03 (0.6071 s / it)
* Acc@1 78.227 Acc@5 94.144 loss 0.829
Accuracy of the model EMA on 50000 test images: 78.2%
Max EMA accuracy: 78.23%
Epoch: [292]  [  0/156]  eta: 0:06:39  lr: 0.000017  min_lr: 0.000017  loss: 3.2064 (3.2064)  weight_decay: 0.0500 (0.0500)  time: 2.5608  data: 1.9092  max mem: 57114
Epoch: [292]  [ 10/156]  eta: 0:02:09  lr: 0.000017  min_lr: 0.000017  loss: 2.7371 (2.6324)  weight_decay: 0.0500 (0.0500)  time: 0.8857  data: 0.1739  max mem: 57114
Epoch: [292]  [ 20/156]  eta: 0:01:49  lr: 0.000017  min_lr: 0.000017  loss: 2.4316 (2.4661)  weight_decay: 0.0500 (0.0500)  time: 0.7145  data: 0.0003  max mem: 57114
Epoch: [292]  [ 30/156]  eta: 0:01:37  lr: 0.000016  min_lr: 0.000016  loss: 2.4029 (2.4729)  weight_decay: 0.0500 (0.0500)  time: 0.7088  data: 0.0003  max mem: 57114
Epoch: [292]  [ 40/156]  eta: 0:01:27  lr: 0.000016  min_lr: 0.000016  loss: 2.6779 (2.5035)  weight_decay: 0.0500 (0.0500)  time: 0.7111  data: 0.0004  max mem: 57114
Epoch: [292]  [ 50/156]  eta: 0:01:20  lr: 0.000016  min_lr: 0.000016  loss: 2.6779 (2.4896)  weight_decay: 0.0500 (0.0500)  time: 0.7369  data: 0.0004  max mem: 57114
Epoch: [292]  [ 60/156]  eta: 0:01:12  lr: 0.000016  min_lr: 0.000016  loss: 2.5801 (2.5044)  weight_decay: 0.0500 (0.0500)  time: 0.7404  data: 0.0004  max mem: 57114
Epoch: [292]  [ 70/156]  eta: 0:01:04  lr: 0.000015  min_lr: 0.000015  loss: 2.4614 (2.4813)  weight_decay: 0.0500 (0.0500)  time: 0.7201  data: 0.0004  max mem: 57114
Epoch: [292]  [ 80/156]  eta: 0:00:56  lr: 0.000015  min_lr: 0.000015  loss: 2.3316 (2.4786)  weight_decay: 0.0500 (0.0500)  time: 0.7152  data: 0.0003  max mem: 57114
Epoch: [292]  [ 90/156]  eta: 0:00:48  lr: 0.000015  min_lr: 0.000015  loss: 2.4559 (2.4754)  weight_decay: 0.0500 (0.0500)  time: 0.7045  data: 0.0003  max mem: 57114
Epoch: [292]  [100/156]  eta: 0:00:41  lr: 0.000015  min_lr: 0.000015  loss: 2.4261 (2.4548)  weight_decay: 0.0500 (0.0500)  time: 0.7012  data: 0.0003  max mem: 57114
Epoch: [292]  [110/156]  eta: 0:00:33  lr: 0.000014  min_lr: 0.000014  loss: 2.4106 (2.4661)  weight_decay: 0.0500 (0.0500)  time: 0.7001  data: 0.0003  max mem: 57114
Epoch: [292]  [120/156]  eta: 0:00:26  lr: 0.000014  min_lr: 0.000014  loss: 2.7057 (2.4797)  weight_decay: 0.0500 (0.0500)  time: 0.7026  data: 0.0004  max mem: 57114
Epoch: [292]  [130/156]  eta: 0:00:18  lr: 0.000014  min_lr: 0.000014  loss: 2.6765 (2.4927)  weight_decay: 0.0500 (0.0500)  time: 0.7081  data: 0.0009  max mem: 57114
Epoch: [292]  [140/156]  eta: 0:00:11  lr: 0.000014  min_lr: 0.000014  loss: 2.6741 (2.4971)  weight_decay: 0.0500 (0.0500)  time: 0.6971  data: 0.0007  max mem: 57114
Epoch: [292]  [150/156]  eta: 0:00:04  lr: 0.000013  min_lr: 0.000013  loss: 2.7464 (2.5149)  weight_decay: 0.0500 (0.0500)  time: 0.6836  data: 0.0001  max mem: 57114
Epoch: [292]  [155/156]  eta: 0:00:00  lr: 0.000013  min_lr: 0.000013  loss: 2.4193 (2.5025)  weight_decay: 0.0500 (0.0500)  time: 0.6780  data: 0.0001  max mem: 57114
Epoch: [292] Total time: 0:01:52 (0.7223 s / it)
Averaged stats: lr: 0.000013  min_lr: 0.000013  loss: 2.4193 (2.5086)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5520 (0.5520)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.0270  data: 1.7706  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7297 (0.6767)  acc1: 83.8542 (84.3990)  acc5: 98.4375 (97.4425)  time: 0.5561  data: 0.3542  max mem: 57114
Test: Total time: 0:00:02 (0.5787 s / it)
* Acc@1 84.765 Acc@5 97.053 loss 0.758
Accuracy of the model on the 50000 test images: 84.8%
Max accuracy: 84.79%
Test:  [0/5]  eta: 0:00:11  loss: 0.6077 (0.6077)  acc1: 85.9375 (85.9375)  acc5: 96.8750 (96.8750)  time: 2.3173  data: 2.0738  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6915 (0.6928)  acc1: 81.7708 (80.1790)  acc5: 96.8750 (94.7570)  time: 0.6142  data: 0.4149  max mem: 57114
Test: Total time: 0:00:03 (0.6274 s / it)
* Acc@1 78.257 Acc@5 94.160 loss 0.828
Accuracy of the model EMA on 50000 test images: 78.3%
Max EMA accuracy: 78.26%
Epoch: [293]  [  0/156]  eta: 0:07:08  lr: 0.000013  min_lr: 0.000013  loss: 2.7907 (2.7907)  weight_decay: 0.0500 (0.0500)  time: 2.7471  data: 2.0949  max mem: 57114
Epoch: [293]  [ 10/156]  eta: 0:02:11  lr: 0.000013  min_lr: 0.000013  loss: 2.6522 (2.6252)  weight_decay: 0.0500 (0.0500)  time: 0.9038  data: 0.1908  max mem: 57114
Epoch: [293]  [ 20/156]  eta: 0:01:50  lr: 0.000013  min_lr: 0.000013  loss: 2.6381 (2.6263)  weight_decay: 0.0500 (0.0500)  time: 0.7120  data: 0.0004  max mem: 57114
Epoch: [293]  [ 30/156]  eta: 0:01:38  lr: 0.000013  min_lr: 0.000013  loss: 2.4811 (2.5331)  weight_decay: 0.0500 (0.0500)  time: 0.7092  data: 0.0004  max mem: 57114
Epoch: [293]  [ 40/156]  eta: 0:01:29  lr: 0.000012  min_lr: 0.000012  loss: 2.4811 (2.5411)  weight_decay: 0.0500 (0.0500)  time: 0.7335  data: 0.0004  max mem: 57114
Epoch: [293]  [ 50/156]  eta: 0:01:20  lr: 0.000012  min_lr: 0.000012  loss: 2.5444 (2.5122)  weight_decay: 0.0500 (0.0500)  time: 0.7412  data: 0.0004  max mem: 57114
Epoch: [293]  [ 60/156]  eta: 0:01:12  lr: 0.000012  min_lr: 0.000012  loss: 2.5444 (2.5222)  weight_decay: 0.0500 (0.0500)  time: 0.7336  data: 0.0004  max mem: 57114
Epoch: [293]  [ 70/156]  eta: 0:01:04  lr: 0.000012  min_lr: 0.000012  loss: 2.6275 (2.5305)  weight_decay: 0.0500 (0.0500)  time: 0.7192  data: 0.0004  max mem: 57114
Epoch: [293]  [ 80/156]  eta: 0:00:56  lr: 0.000012  min_lr: 0.000012  loss: 2.5172 (2.5248)  weight_decay: 0.0500 (0.0500)  time: 0.7012  data: 0.0004  max mem: 57114
Epoch: [293]  [ 90/156]  eta: 0:00:48  lr: 0.000011  min_lr: 0.000011  loss: 2.4601 (2.4927)  weight_decay: 0.0500 (0.0500)  time: 0.7033  data: 0.0003  max mem: 57114
Epoch: [293]  [100/156]  eta: 0:00:41  lr: 0.000011  min_lr: 0.000011  loss: 2.2144 (2.4749)  weight_decay: 0.0500 (0.0500)  time: 0.7144  data: 0.0003  max mem: 57114
Epoch: [293]  [110/156]  eta: 0:00:33  lr: 0.000011  min_lr: 0.000011  loss: 2.5694 (2.4791)  weight_decay: 0.0500 (0.0500)  time: 0.7151  data: 0.0003  max mem: 57114
Epoch: [293]  [120/156]  eta: 0:00:26  lr: 0.000011  min_lr: 0.000011  loss: 2.6839 (2.4910)  weight_decay: 0.0500 (0.0500)  time: 0.7171  data: 0.0004  max mem: 57114
Epoch: [293]  [130/156]  eta: 0:00:19  lr: 0.000011  min_lr: 0.000011  loss: 2.6763 (2.4882)  weight_decay: 0.0500 (0.0500)  time: 0.7089  data: 0.0008  max mem: 57114
Epoch: [293]  [140/156]  eta: 0:00:11  lr: 0.000010  min_lr: 0.000010  loss: 2.4970 (2.4854)  weight_decay: 0.0500 (0.0500)  time: 0.6809  data: 0.0007  max mem: 57114
Epoch: [293]  [150/156]  eta: 0:00:04  lr: 0.000010  min_lr: 0.000010  loss: 2.3130 (2.4741)  weight_decay: 0.0500 (0.0500)  time: 0.6807  data: 0.0001  max mem: 57114
Epoch: [293]  [155/156]  eta: 0:00:00  lr: 0.000010  min_lr: 0.000010  loss: 2.4863 (2.4688)  weight_decay: 0.0500 (0.0500)  time: 0.6810  data: 0.0001  max mem: 57114
Epoch: [293] Total time: 0:01:53 (0.7258 s / it)
Averaged stats: lr: 0.000010  min_lr: 0.000010  loss: 2.4863 (2.5083)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5093 (0.5093)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.0590  data: 1.8032  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6869 (0.6353)  acc1: 84.3750 (84.7826)  acc5: 98.4375 (97.5703)  time: 0.5624  data: 0.3607  max mem: 57114
Test: Total time: 0:00:02 (0.5780 s / it)
* Acc@1 84.779 Acc@5 97.073 loss 0.719
Accuracy of the model on the 50000 test images: 84.8%
Max accuracy: 84.79%
Test:  [0/5]  eta: 0:00:12  loss: 0.6063 (0.6063)  acc1: 85.9375 (85.9375)  acc5: 96.8750 (96.8750)  time: 2.4231  data: 2.1795  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6903 (0.6915)  acc1: 81.7708 (80.1790)  acc5: 96.8750 (94.8849)  time: 0.6352  data: 0.4360  max mem: 57114
Test: Total time: 0:00:03 (0.6449 s / it)
* Acc@1 78.329 Acc@5 94.200 loss 0.826
Accuracy of the model EMA on 50000 test images: 78.3%
Max EMA accuracy: 78.33%
Epoch: [294]  [  0/156]  eta: 0:06:41  lr: 0.000010  min_lr: 0.000010  loss: 2.0365 (2.0365)  weight_decay: 0.0500 (0.0500)  time: 2.5766  data: 1.9111  max mem: 57114
Epoch: [294]  [ 10/156]  eta: 0:02:07  lr: 0.000010  min_lr: 0.000010  loss: 2.3410 (2.2773)  weight_decay: 0.0500 (0.0500)  time: 0.8764  data: 0.1741  max mem: 57114
Epoch: [294]  [ 20/156]  eta: 0:01:49  lr: 0.000010  min_lr: 0.000010  loss: 2.4458 (2.4052)  weight_decay: 0.0500 (0.0500)  time: 0.7152  data: 0.0004  max mem: 57114
Epoch: [294]  [ 30/156]  eta: 0:01:38  lr: 0.000009  min_lr: 0.000009  loss: 2.5658 (2.4263)  weight_decay: 0.0500 (0.0500)  time: 0.7310  data: 0.0004  max mem: 57114
Epoch: [294]  [ 40/156]  eta: 0:01:28  lr: 0.000009  min_lr: 0.000009  loss: 2.5658 (2.4307)  weight_decay: 0.0500 (0.0500)  time: 0.7269  data: 0.0004  max mem: 57114
Epoch: [294]  [ 50/156]  eta: 0:01:20  lr: 0.000009  min_lr: 0.000009  loss: 2.5516 (2.4317)  weight_decay: 0.0500 (0.0500)  time: 0.7161  data: 0.0004  max mem: 57114
Epoch: [294]  [ 60/156]  eta: 0:01:12  lr: 0.000009  min_lr: 0.000009  loss: 2.5999 (2.4596)  weight_decay: 0.0500 (0.0500)  time: 0.7222  data: 0.0005  max mem: 57114
Epoch: [294]  [ 70/156]  eta: 0:01:03  lr: 0.000009  min_lr: 0.000009  loss: 2.6115 (2.4771)  weight_decay: 0.0500 (0.0500)  time: 0.7104  data: 0.0005  max mem: 57114
Epoch: [294]  [ 80/156]  eta: 0:00:56  lr: 0.000009  min_lr: 0.000009  loss: 2.6246 (2.4919)  weight_decay: 0.0500 (0.0500)  time: 0.7044  data: 0.0004  max mem: 57114
Epoch: [294]  [ 90/156]  eta: 0:00:48  lr: 0.000008  min_lr: 0.000008  loss: 2.6375 (2.4988)  weight_decay: 0.0500 (0.0500)  time: 0.7121  data: 0.0004  max mem: 57114
Epoch: [294]  [100/156]  eta: 0:00:41  lr: 0.000008  min_lr: 0.000008  loss: 2.6510 (2.5073)  weight_decay: 0.0500 (0.0500)  time: 0.7042  data: 0.0004  max mem: 57114
Epoch: [294]  [110/156]  eta: 0:00:33  lr: 0.000008  min_lr: 0.000008  loss: 2.7255 (2.5110)  weight_decay: 0.0500 (0.0500)  time: 0.6999  data: 0.0004  max mem: 57114
Epoch: [294]  [120/156]  eta: 0:00:26  lr: 0.000008  min_lr: 0.000008  loss: 2.6268 (2.5038)  weight_decay: 0.0500 (0.0500)  time: 0.7075  data: 0.0004  max mem: 57114
Epoch: [294]  [130/156]  eta: 0:00:18  lr: 0.000008  min_lr: 0.000008  loss: 2.5542 (2.5035)  weight_decay: 0.0500 (0.0500)  time: 0.7034  data: 0.0010  max mem: 57114
Epoch: [294]  [140/156]  eta: 0:00:11  lr: 0.000008  min_lr: 0.000008  loss: 2.6635 (2.5159)  weight_decay: 0.0500 (0.0500)  time: 0.6935  data: 0.0008  max mem: 57114
Epoch: [294]  [150/156]  eta: 0:00:04  lr: 0.000007  min_lr: 0.000007  loss: 2.6635 (2.5126)  weight_decay: 0.0500 (0.0500)  time: 0.6879  data: 0.0001  max mem: 57114
Epoch: [294]  [155/156]  eta: 0:00:00  lr: 0.000007  min_lr: 0.000007  loss: 2.6635 (2.5195)  weight_decay: 0.0500 (0.0500)  time: 0.6839  data: 0.0001  max mem: 57114
Epoch: [294] Total time: 0:01:52 (0.7220 s / it)
Averaged stats: lr: 0.000007  min_lr: 0.000007  loss: 2.6635 (2.5020)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5924 (0.5924)  acc1: 90.6250 (90.6250)  acc5: 98.4375 (98.4375)  time: 2.0959  data: 1.8399  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7761 (0.7191)  acc1: 84.3750 (84.7826)  acc5: 98.4375 (97.4425)  time: 0.5699  data: 0.3680  max mem: 57114
Test: Total time: 0:00:02 (0.5927 s / it)
* Acc@1 84.833 Acc@5 97.059 loss 0.799
Accuracy of the model on the 50000 test images: 84.8%
Max accuracy: 84.83%
Test:  [0/5]  eta: 0:00:10  loss: 0.6050 (0.6050)  acc1: 85.9375 (85.9375)  acc5: 96.8750 (96.8750)  time: 2.0136  data: 1.7700  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6892 (0.6904)  acc1: 81.7708 (80.1790)  acc5: 96.8750 (94.8849)  time: 0.5534  data: 0.3541  max mem: 57114
Test: Total time: 0:00:02 (0.5639 s / it)
* Acc@1 78.399 Acc@5 94.252 loss 0.824
Accuracy of the model EMA on 50000 test images: 78.4%
Max EMA accuracy: 78.40%
Epoch: [295]  [  0/156]  eta: 0:06:52  lr: 0.000007  min_lr: 0.000007  loss: 2.3895 (2.3895)  weight_decay: 0.0500 (0.0500)  time: 2.6443  data: 1.8387  max mem: 57114
Epoch: [295]  [ 10/156]  eta: 0:02:10  lr: 0.000007  min_lr: 0.000007  loss: 2.6025 (2.6275)  weight_decay: 0.0500 (0.0500)  time: 0.8933  data: 0.1674  max mem: 57114
Epoch: [295]  [ 20/156]  eta: 0:01:51  lr: 0.000007  min_lr: 0.000007  loss: 2.6643 (2.5480)  weight_decay: 0.0500 (0.0500)  time: 0.7296  data: 0.0003  max mem: 57114
Epoch: [295]  [ 30/156]  eta: 0:01:39  lr: 0.000007  min_lr: 0.000007  loss: 2.5019 (2.5094)  weight_decay: 0.0500 (0.0500)  time: 0.7289  data: 0.0004  max mem: 57114
Epoch: [295]  [ 40/156]  eta: 0:01:30  lr: 0.000007  min_lr: 0.000007  loss: 2.5482 (2.5102)  weight_decay: 0.0500 (0.0500)  time: 0.7367  data: 0.0004  max mem: 57114
Epoch: [295]  [ 50/156]  eta: 0:01:21  lr: 0.000007  min_lr: 0.000007  loss: 2.5628 (2.4941)  weight_decay: 0.0500 (0.0500)  time: 0.7491  data: 0.0004  max mem: 57114
Epoch: [295]  [ 60/156]  eta: 0:01:13  lr: 0.000006  min_lr: 0.000006  loss: 2.6078 (2.5075)  weight_decay: 0.0500 (0.0500)  time: 0.7249  data: 0.0004  max mem: 57114
Epoch: [295]  [ 70/156]  eta: 0:01:05  lr: 0.000006  min_lr: 0.000006  loss: 2.6078 (2.4854)  weight_decay: 0.0500 (0.0500)  time: 0.7195  data: 0.0004  max mem: 57114
Epoch: [295]  [ 80/156]  eta: 0:00:57  lr: 0.000006  min_lr: 0.000006  loss: 2.6119 (2.4894)  weight_decay: 0.0500 (0.0500)  time: 0.7305  data: 0.0003  max mem: 57114
Epoch: [295]  [ 90/156]  eta: 0:00:49  lr: 0.000006  min_lr: 0.000006  loss: 2.6125 (2.4959)  weight_decay: 0.0500 (0.0500)  time: 0.7178  data: 0.0004  max mem: 57114
Epoch: [295]  [100/156]  eta: 0:00:41  lr: 0.000006  min_lr: 0.000006  loss: 2.6881 (2.5095)  weight_decay: 0.0500 (0.0500)  time: 0.7056  data: 0.0004  max mem: 57114
Epoch: [295]  [110/156]  eta: 0:00:34  lr: 0.000006  min_lr: 0.000006  loss: 2.6908 (2.4989)  weight_decay: 0.0500 (0.0500)  time: 0.7148  data: 0.0004  max mem: 57114
Epoch: [295]  [120/156]  eta: 0:00:26  lr: 0.000006  min_lr: 0.000006  loss: 2.5974 (2.4996)  weight_decay: 0.0500 (0.0500)  time: 0.7151  data: 0.0004  max mem: 57114
Epoch: [295]  [130/156]  eta: 0:00:19  lr: 0.000005  min_lr: 0.000005  loss: 2.6427 (2.5104)  weight_decay: 0.0500 (0.0500)  time: 0.7011  data: 0.0008  max mem: 57114
Epoch: [295]  [140/156]  eta: 0:00:11  lr: 0.000005  min_lr: 0.000005  loss: 2.6487 (2.5094)  weight_decay: 0.0500 (0.0500)  time: 0.6893  data: 0.0007  max mem: 57114
Epoch: [295]  [150/156]  eta: 0:00:04  lr: 0.000005  min_lr: 0.000005  loss: 2.5977 (2.5141)  weight_decay: 0.0500 (0.0500)  time: 0.6813  data: 0.0001  max mem: 57114
Epoch: [295]  [155/156]  eta: 0:00:00  lr: 0.000005  min_lr: 0.000005  loss: 2.7192 (2.5177)  weight_decay: 0.0500 (0.0500)  time: 0.6828  data: 0.0001  max mem: 57114
Epoch: [295] Total time: 0:01:53 (0.7300 s / it)
Averaged stats: lr: 0.000005  min_lr: 0.000005  loss: 2.7192 (2.5041)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:11  loss: 0.6084 (0.6084)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.2014  data: 1.9454  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7823 (0.7332)  acc1: 84.8958 (84.6547)  acc5: 98.4375 (97.3146)  time: 0.5910  data: 0.3892  max mem: 57114
Test: Total time: 0:00:03 (0.6130 s / it)
* Acc@1 84.675 Acc@5 97.027 loss 0.813
Accuracy of the model on the 50000 test images: 84.7%
Max accuracy: 84.83%
Test:  [0/5]  eta: 0:00:11  loss: 0.6035 (0.6035)  acc1: 85.9375 (85.9375)  acc5: 96.8750 (96.8750)  time: 2.3486  data: 2.1051  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6878 (0.6891)  acc1: 81.7708 (80.3069)  acc5: 96.8750 (94.7570)  time: 0.6204  data: 0.4211  max mem: 57114
Test: Total time: 0:00:03 (0.6354 s / it)
* Acc@1 78.487 Acc@5 94.279 loss 0.822
Accuracy of the model EMA on 50000 test images: 78.5%
Max EMA accuracy: 78.49%
Epoch: [296]  [  0/156]  eta: 0:07:07  lr: 0.000005  min_lr: 0.000005  loss: 2.4646 (2.4646)  weight_decay: 0.0500 (0.0500)  time: 2.7417  data: 2.0795  max mem: 57114
Epoch: [296]  [ 10/156]  eta: 0:02:11  lr: 0.000005  min_lr: 0.000005  loss: 2.4646 (2.5019)  weight_decay: 0.0500 (0.0500)  time: 0.8987  data: 0.1893  max mem: 57114
Epoch: [296]  [ 20/156]  eta: 0:01:51  lr: 0.000005  min_lr: 0.000005  loss: 2.6728 (2.6178)  weight_decay: 0.0500 (0.0500)  time: 0.7212  data: 0.0003  max mem: 57114
Epoch: [296]  [ 30/156]  eta: 0:01:39  lr: 0.000005  min_lr: 0.000005  loss: 2.6437 (2.5451)  weight_decay: 0.0500 (0.0500)  time: 0.7331  data: 0.0003  max mem: 57114
Epoch: [296]  [ 40/156]  eta: 0:01:29  lr: 0.000005  min_lr: 0.000005  loss: 2.4283 (2.5375)  weight_decay: 0.0500 (0.0500)  time: 0.7274  data: 0.0004  max mem: 57114
Epoch: [296]  [ 50/156]  eta: 0:01:21  lr: 0.000004  min_lr: 0.000004  loss: 2.6016 (2.5368)  weight_decay: 0.0500 (0.0500)  time: 0.7222  data: 0.0004  max mem: 57114
Epoch: [296]  [ 60/156]  eta: 0:01:12  lr: 0.000004  min_lr: 0.000004  loss: 2.5970 (2.5297)  weight_decay: 0.0500 (0.0500)  time: 0.7159  data: 0.0004  max mem: 57114
Epoch: [296]  [ 70/156]  eta: 0:01:04  lr: 0.000004  min_lr: 0.000004  loss: 2.5120 (2.5103)  weight_decay: 0.0500 (0.0500)  time: 0.7094  data: 0.0004  max mem: 57114
Epoch: [296]  [ 80/156]  eta: 0:00:56  lr: 0.000004  min_lr: 0.000004  loss: 2.4565 (2.5058)  weight_decay: 0.0500 (0.0500)  time: 0.7102  data: 0.0004  max mem: 57114
Epoch: [296]  [ 90/156]  eta: 0:00:48  lr: 0.000004  min_lr: 0.000004  loss: 2.5497 (2.5064)  weight_decay: 0.0500 (0.0500)  time: 0.7083  data: 0.0004  max mem: 57114
Epoch: [296]  [100/156]  eta: 0:00:41  lr: 0.000004  min_lr: 0.000004  loss: 2.4261 (2.5003)  weight_decay: 0.0500 (0.0500)  time: 0.7112  data: 0.0004  max mem: 57114
Epoch: [296]  [110/156]  eta: 0:00:33  lr: 0.000004  min_lr: 0.000004  loss: 2.6450 (2.5150)  weight_decay: 0.0500 (0.0500)  time: 0.7086  data: 0.0004  max mem: 57114
Epoch: [296]  [120/156]  eta: 0:00:26  lr: 0.000004  min_lr: 0.000004  loss: 2.6711 (2.5222)  weight_decay: 0.0500 (0.0500)  time: 0.7046  data: 0.0004  max mem: 57114
Epoch: [296]  [130/156]  eta: 0:00:18  lr: 0.000004  min_lr: 0.000004  loss: 2.4670 (2.5039)  weight_decay: 0.0500 (0.0500)  time: 0.6952  data: 0.0009  max mem: 57114
Epoch: [296]  [140/156]  eta: 0:00:11  lr: 0.000003  min_lr: 0.000003  loss: 2.5317 (2.5099)  weight_decay: 0.0500 (0.0500)  time: 0.6850  data: 0.0007  max mem: 57114
Epoch: [296]  [150/156]  eta: 0:00:04  lr: 0.000003  min_lr: 0.000003  loss: 2.5604 (2.5054)  weight_decay: 0.0500 (0.0500)  time: 0.6849  data: 0.0001  max mem: 57114
Epoch: [296]  [155/156]  eta: 0:00:00  lr: 0.000003  min_lr: 0.000003  loss: 2.5956 (2.5122)  weight_decay: 0.0500 (0.0500)  time: 0.6834  data: 0.0001  max mem: 57114
Epoch: [296] Total time: 0:01:52 (0.7234 s / it)
Averaged stats: lr: 0.000003  min_lr: 0.000003  loss: 2.5956 (2.5032)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6213 (0.6213)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.0709  data: 1.8149  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8005 (0.7461)  acc1: 84.3750 (84.5269)  acc5: 98.4375 (97.4425)  time: 0.5649  data: 0.3631  max mem: 57114
Test: Total time: 0:00:02 (0.5871 s / it)
* Acc@1 84.717 Acc@5 97.043 loss 0.824
Accuracy of the model on the 50000 test images: 84.7%
Max accuracy: 84.83%
Test:  [0/5]  eta: 0:00:11  loss: 0.6021 (0.6021)  acc1: 85.9375 (85.9375)  acc5: 96.8750 (96.8750)  time: 2.3107  data: 2.0672  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6866 (0.6878)  acc1: 81.7708 (80.4348)  acc5: 96.8750 (94.7570)  time: 0.6128  data: 0.4135  max mem: 57114
Test: Total time: 0:00:03 (0.6284 s / it)
* Acc@1 78.575 Acc@5 94.313 loss 0.820
Accuracy of the model EMA on 50000 test images: 78.6%
Max EMA accuracy: 78.57%
Epoch: [297]  [  0/156]  eta: 0:07:23  lr: 0.000003  min_lr: 0.000003  loss: 2.9631 (2.9631)  weight_decay: 0.0500 (0.0500)  time: 2.8439  data: 2.1893  max mem: 57114
Epoch: [297]  [ 10/156]  eta: 0:02:13  lr: 0.000003  min_lr: 0.000003  loss: 2.6833 (2.6589)  weight_decay: 0.0500 (0.0500)  time: 0.9155  data: 0.1993  max mem: 57114
Epoch: [297]  [ 20/156]  eta: 0:01:53  lr: 0.000003  min_lr: 0.000003  loss: 2.5248 (2.5060)  weight_decay: 0.0500 (0.0500)  time: 0.7354  data: 0.0003  max mem: 57114
Epoch: [297]  [ 30/156]  eta: 0:01:40  lr: 0.000003  min_lr: 0.000003  loss: 2.4723 (2.5361)  weight_decay: 0.0500 (0.0500)  time: 0.7321  data: 0.0003  max mem: 57114
Epoch: [297]  [ 40/156]  eta: 0:01:30  lr: 0.000003  min_lr: 0.000003  loss: 2.7450 (2.5773)  weight_decay: 0.0500 (0.0500)  time: 0.7173  data: 0.0003  max mem: 57114
Epoch: [297]  [ 50/156]  eta: 0:01:21  lr: 0.000003  min_lr: 0.000003  loss: 2.6115 (2.5466)  weight_decay: 0.0500 (0.0500)  time: 0.7134  data: 0.0004  max mem: 57114
Epoch: [297]  [ 60/156]  eta: 0:01:12  lr: 0.000003  min_lr: 0.000003  loss: 2.5492 (2.5281)  weight_decay: 0.0500 (0.0500)  time: 0.7117  data: 0.0004  max mem: 57114
Epoch: [297]  [ 70/156]  eta: 0:01:04  lr: 0.000003  min_lr: 0.000003  loss: 2.5492 (2.5340)  weight_decay: 0.0500 (0.0500)  time: 0.7212  data: 0.0004  max mem: 57114
Epoch: [297]  [ 80/156]  eta: 0:00:56  lr: 0.000003  min_lr: 0.000003  loss: 2.6727 (2.5355)  weight_decay: 0.0500 (0.0500)  time: 0.7248  data: 0.0004  max mem: 57114
Epoch: [297]  [ 90/156]  eta: 0:00:49  lr: 0.000002  min_lr: 0.000002  loss: 2.5943 (2.5264)  weight_decay: 0.0500 (0.0500)  time: 0.7122  data: 0.0004  max mem: 57114
Epoch: [297]  [100/156]  eta: 0:00:41  lr: 0.000002  min_lr: 0.000002  loss: 2.5025 (2.5207)  weight_decay: 0.0500 (0.0500)  time: 0.7048  data: 0.0003  max mem: 57114
Epoch: [297]  [110/156]  eta: 0:00:33  lr: 0.000002  min_lr: 0.000002  loss: 2.5011 (2.5104)  weight_decay: 0.0500 (0.0500)  time: 0.7133  data: 0.0003  max mem: 57114
Epoch: [297]  [120/156]  eta: 0:00:26  lr: 0.000002  min_lr: 0.000002  loss: 2.6513 (2.5223)  weight_decay: 0.0500 (0.0500)  time: 0.7060  data: 0.0004  max mem: 57114
Epoch: [297]  [130/156]  eta: 0:00:19  lr: 0.000002  min_lr: 0.000002  loss: 2.6333 (2.5245)  weight_decay: 0.0500 (0.0500)  time: 0.6983  data: 0.0008  max mem: 57114
Epoch: [297]  [140/156]  eta: 0:00:11  lr: 0.000002  min_lr: 0.000002  loss: 2.5670 (2.5290)  weight_decay: 0.0500 (0.0500)  time: 0.6947  data: 0.0007  max mem: 57114
Epoch: [297]  [150/156]  eta: 0:00:04  lr: 0.000002  min_lr: 0.000002  loss: 2.4846 (2.5101)  weight_decay: 0.0500 (0.0500)  time: 0.6852  data: 0.0001  max mem: 57114
Epoch: [297]  [155/156]  eta: 0:00:00  lr: 0.000002  min_lr: 0.000002  loss: 2.4130 (2.5068)  weight_decay: 0.0500 (0.0500)  time: 0.6832  data: 0.0001  max mem: 57114
Epoch: [297] Total time: 0:01:53 (0.7264 s / it)
Averaged stats: lr: 0.000002  min_lr: 0.000002  loss: 2.4130 (2.5037)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.5424 (0.5424)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.0932  data: 1.8369  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7229 (0.6704)  acc1: 83.8542 (84.5269)  acc5: 98.4375 (97.4425)  time: 0.5695  data: 0.3675  max mem: 57114
Test: Total time: 0:00:02 (0.5927 s / it)
* Acc@1 84.785 Acc@5 97.079 loss 0.750
Accuracy of the model on the 50000 test images: 84.8%
Max accuracy: 84.83%
Test:  [0/5]  eta: 0:00:11  loss: 0.6005 (0.6005)  acc1: 85.9375 (85.9375)  acc5: 96.8750 (96.8750)  time: 2.2689  data: 2.0252  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6853 (0.6864)  acc1: 81.7708 (80.5627)  acc5: 96.8750 (94.7570)  time: 0.6045  data: 0.4051  max mem: 57114
Test: Total time: 0:00:03 (0.6171 s / it)
* Acc@1 78.645 Acc@5 94.337 loss 0.818
Accuracy of the model EMA on 50000 test images: 78.6%
Max EMA accuracy: 78.64%
Epoch: [298]  [  0/156]  eta: 0:07:43  lr: 0.000002  min_lr: 0.000002  loss: 3.0060 (3.0060)  weight_decay: 0.0500 (0.0500)  time: 2.9698  data: 2.3232  max mem: 57114
Epoch: [298]  [ 10/156]  eta: 0:02:14  lr: 0.000002  min_lr: 0.000002  loss: 2.3719 (2.3607)  weight_decay: 0.0500 (0.0500)  time: 0.9244  data: 0.2115  max mem: 57114
Epoch: [298]  [ 20/156]  eta: 0:01:53  lr: 0.000002  min_lr: 0.000002  loss: 2.2991 (2.3634)  weight_decay: 0.0500 (0.0500)  time: 0.7256  data: 0.0003  max mem: 57114
Epoch: [298]  [ 30/156]  eta: 0:01:39  lr: 0.000002  min_lr: 0.000002  loss: 2.5927 (2.4091)  weight_decay: 0.0500 (0.0500)  time: 0.7170  data: 0.0003  max mem: 57114
Epoch: [298]  [ 40/156]  eta: 0:01:29  lr: 0.000002  min_lr: 0.000002  loss: 2.6478 (2.4383)  weight_decay: 0.0500 (0.0500)  time: 0.7096  data: 0.0004  max mem: 57114
Epoch: [298]  [ 50/156]  eta: 0:01:21  lr: 0.000002  min_lr: 0.000002  loss: 2.6122 (2.4473)  weight_decay: 0.0500 (0.0500)  time: 0.7281  data: 0.0004  max mem: 57114
Epoch: [298]  [ 60/156]  eta: 0:01:12  lr: 0.000002  min_lr: 0.000002  loss: 2.5659 (2.4547)  weight_decay: 0.0500 (0.0500)  time: 0.7326  data: 0.0004  max mem: 57114
Epoch: [298]  [ 70/156]  eta: 0:01:04  lr: 0.000002  min_lr: 0.000002  loss: 2.5010 (2.4386)  weight_decay: 0.0500 (0.0500)  time: 0.7198  data: 0.0004  max mem: 57114
Epoch: [298]  [ 80/156]  eta: 0:00:56  lr: 0.000002  min_lr: 0.000002  loss: 2.6040 (2.4529)  weight_decay: 0.0500 (0.0500)  time: 0.7083  data: 0.0004  max mem: 57114
Epoch: [298]  [ 90/156]  eta: 0:00:48  lr: 0.000002  min_lr: 0.000002  loss: 2.7497 (2.4752)  weight_decay: 0.0500 (0.0500)  time: 0.7043  data: 0.0004  max mem: 57114
Epoch: [298]  [100/156]  eta: 0:00:41  lr: 0.000001  min_lr: 0.000001  loss: 2.6637 (2.4762)  weight_decay: 0.0500 (0.0500)  time: 0.7046  data: 0.0004  max mem: 57114
Epoch: [298]  [110/156]  eta: 0:00:33  lr: 0.000001  min_lr: 0.000001  loss: 2.5131 (2.4739)  weight_decay: 0.0500 (0.0500)  time: 0.7037  data: 0.0004  max mem: 57114
Epoch: [298]  [120/156]  eta: 0:00:26  lr: 0.000001  min_lr: 0.000001  loss: 2.5660 (2.4840)  weight_decay: 0.0500 (0.0500)  time: 0.7011  data: 0.0004  max mem: 57114
Epoch: [298]  [130/156]  eta: 0:00:18  lr: 0.000001  min_lr: 0.000001  loss: 2.5683 (2.4732)  weight_decay: 0.0500 (0.0500)  time: 0.6988  data: 0.0009  max mem: 57114
Epoch: [298]  [140/156]  eta: 0:00:11  lr: 0.000001  min_lr: 0.000001  loss: 2.4904 (2.4752)  weight_decay: 0.0500 (0.0500)  time: 0.6933  data: 0.0007  max mem: 57114
Epoch: [298]  [150/156]  eta: 0:00:04  lr: 0.000001  min_lr: 0.000001  loss: 2.6625 (2.4878)  weight_decay: 0.0500 (0.0500)  time: 0.6858  data: 0.0001  max mem: 57114
Epoch: [298]  [155/156]  eta: 0:00:00  lr: 0.000001  min_lr: 0.000001  loss: 2.6963 (2.4966)  weight_decay: 0.0500 (0.0500)  time: 0.6850  data: 0.0001  max mem: 57114
Epoch: [298] Total time: 0:01:53 (0.7253 s / it)
Averaged stats: lr: 0.000001  min_lr: 0.000001  loss: 2.6963 (2.5021)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6530 (0.6530)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.0908  data: 1.8350  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.8257 (0.7780)  acc1: 84.3750 (84.5269)  acc5: 98.4375 (97.4425)  time: 0.5688  data: 0.3671  max mem: 57114
Test: Total time: 0:00:02 (0.5904 s / it)
* Acc@1 84.745 Acc@5 97.051 loss 0.854
Accuracy of the model on the 50000 test images: 84.7%
Max accuracy: 84.83%
Test:  [0/5]  eta: 0:00:11  loss: 0.5989 (0.5989)  acc1: 85.9375 (85.9375)  acc5: 96.8750 (96.8750)  time: 2.3525  data: 2.1089  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6839 (0.6850)  acc1: 81.7708 (80.4348)  acc5: 96.8750 (94.7570)  time: 0.6212  data: 0.4219  max mem: 57114
Test: Total time: 0:00:03 (0.6348 s / it)
* Acc@1 78.698 Acc@5 94.385 loss 0.816
Accuracy of the model EMA on 50000 test images: 78.7%
Max EMA accuracy: 78.70%
Epoch: [299]  [  0/156]  eta: 0:09:19  lr: 0.000001  min_lr: 0.000001  loss: 2.8604 (2.8604)  weight_decay: 0.0500 (0.0500)  time: 3.5871  data: 2.9147  max mem: 57114
Epoch: [299]  [ 10/156]  eta: 0:02:23  lr: 0.000001  min_lr: 0.000001  loss: 2.5645 (2.4544)  weight_decay: 0.0500 (0.0500)  time: 0.9834  data: 0.2653  max mem: 57114
Epoch: [299]  [ 20/156]  eta: 0:01:57  lr: 0.000001  min_lr: 0.000001  loss: 2.5580 (2.4386)  weight_decay: 0.0500 (0.0500)  time: 0.7282  data: 0.0003  max mem: 57114
Epoch: [299]  [ 30/156]  eta: 0:01:43  lr: 0.000001  min_lr: 0.000001  loss: 2.6423 (2.4805)  weight_decay: 0.0500 (0.0500)  time: 0.7307  data: 0.0004  max mem: 57114
Epoch: [299]  [ 40/156]  eta: 0:01:32  lr: 0.000001  min_lr: 0.000001  loss: 2.4540 (2.4560)  weight_decay: 0.0500 (0.0500)  time: 0.7249  data: 0.0004  max mem: 57114
Epoch: [299]  [ 50/156]  eta: 0:01:23  lr: 0.000001  min_lr: 0.000001  loss: 2.4540 (2.4715)  weight_decay: 0.0500 (0.0500)  time: 0.7302  data: 0.0005  max mem: 57114
Epoch: [299]  [ 60/156]  eta: 0:01:14  lr: 0.000001  min_lr: 0.000001  loss: 2.5515 (2.4447)  weight_decay: 0.0500 (0.0500)  time: 0.7334  data: 0.0004  max mem: 57114
Epoch: [299]  [ 70/156]  eta: 0:01:05  lr: 0.000001  min_lr: 0.000001  loss: 2.5515 (2.4604)  weight_decay: 0.0500 (0.0500)  time: 0.7214  data: 0.0004  max mem: 57114
Epoch: [299]  [ 80/156]  eta: 0:00:57  lr: 0.000001  min_lr: 0.000001  loss: 2.6335 (2.4607)  weight_decay: 0.0500 (0.0500)  time: 0.7101  data: 0.0004  max mem: 57114
Epoch: [299]  [ 90/156]  eta: 0:00:49  lr: 0.000001  min_lr: 0.000001  loss: 2.4350 (2.4426)  weight_decay: 0.0500 (0.0500)  time: 0.7093  data: 0.0004  max mem: 57114
Epoch: [299]  [100/156]  eta: 0:00:41  lr: 0.000001  min_lr: 0.000001  loss: 2.3645 (2.4381)  weight_decay: 0.0500 (0.0500)  time: 0.7074  data: 0.0004  max mem: 57114
Epoch: [299]  [110/156]  eta: 0:00:34  lr: 0.000001  min_lr: 0.000001  loss: 2.4013 (2.4403)  weight_decay: 0.0500 (0.0500)  time: 0.7046  data: 0.0004  max mem: 57114
Epoch: [299]  [120/156]  eta: 0:00:26  lr: 0.000001  min_lr: 0.000001  loss: 2.6522 (2.4513)  weight_decay: 0.0500 (0.0500)  time: 0.7140  data: 0.0004  max mem: 57114
Epoch: [299]  [130/156]  eta: 0:00:19  lr: 0.000001  min_lr: 0.000001  loss: 2.6739 (2.4640)  weight_decay: 0.0500 (0.0500)  time: 0.7120  data: 0.0009  max mem: 57114
Epoch: [299]  [140/156]  eta: 0:00:11  lr: 0.000001  min_lr: 0.000001  loss: 2.6540 (2.4751)  weight_decay: 0.0500 (0.0500)  time: 0.6906  data: 0.0007  max mem: 57114
Epoch: [299]  [150/156]  eta: 0:00:04  lr: 0.000001  min_lr: 0.000001  loss: 2.7069 (2.4901)  weight_decay: 0.0500 (0.0500)  time: 0.6819  data: 0.0001  max mem: 57114
Epoch: [299]  [155/156]  eta: 0:00:00  lr: 0.000001  min_lr: 0.000001  loss: 2.7069 (2.4889)  weight_decay: 0.0500 (0.0500)  time: 0.6845  data: 0.0001  max mem: 57114
Epoch: [299] Total time: 0:01:54 (0.7335 s / it)
Averaged stats: lr: 0.000001  min_lr: 0.000001  loss: 2.7069 (2.5071)  weight_decay: 0.0500 (0.0500)
Test:  [0/5]  eta: 0:00:10  loss: 0.6196 (0.6196)  acc1: 90.1042 (90.1042)  acc5: 98.4375 (98.4375)  time: 2.0909  data: 1.8353  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.7903 (0.7446)  acc1: 84.3750 (84.5269)  acc5: 98.4375 (97.3146)  time: 0.5688  data: 0.3671  max mem: 57114
Test: Total time: 0:00:02 (0.5928 s / it)
* Acc@1 84.637 Acc@5 97.023 loss 0.823
Accuracy of the model on the 50000 test images: 84.6%
Max accuracy: 84.83%
Test:  [0/5]  eta: 0:00:11  loss: 0.5974 (0.5974)  acc1: 85.9375 (85.9375)  acc5: 96.8750 (96.8750)  time: 2.3034  data: 2.0599  max mem: 57114
Test:  [4/5]  eta: 0:00:00  loss: 0.6825 (0.6835)  acc1: 81.7708 (80.4348)  acc5: 96.8750 (94.7570)  time: 0.6113  data: 0.4121  max mem: 57114
Test: Total time: 0:00:03 (0.6266 s / it)
* Acc@1 78.760 Acc@5 94.427 loss 0.814
Accuracy of the model EMA on 50000 test images: 78.8%
Max EMA accuracy: 78.76%
Training time 10:47:47