log_rank0.txt

[2025-12-08 16:16:27 swin_tiny_patch4_window7_224] (main.py 348): INFO Full config saved to output/swin_tiny_patch4_window7_224/default/config.json
[2025-12-08 16:16:27 swin_tiny_patch4_window7_224] (main.py 351): INFO AMP_ENABLE: true
AMP_OPT_LEVEL: ''
AUG:
  AUTO_AUGMENT: rand-m9-mstd0.5-inc1
  COLOR_JITTER: 0.4
  CUTMIX: 1.0
  CUTMIX_MINMAX: null
  MIXUP: 0.8
  MIXUP_MODE: batch
  MIXUP_PROB: 1.0
  MIXUP_SWITCH_PROB: 0.5
  RECOUNT: 1
  REMODE: pixel
  REPROB: 0.25
BASE:
- ''
DATA:
  BATCH_SIZE: 16
  CACHE_MODE: part
  DATASET: imagenet
  DATA_PATH: /content/drive/MyDrive/Tulsi_train-70_val-30
  IMG_SIZE: 224
  INTERPOLATION: bicubic
  MASK_PATCH_SIZE: 32
  MASK_RATIO: 0.6
  NUM_WORKERS: 8
  PIN_MEMORY: true
  ZIP_MODE: false
ENABLE_AMP: false
EVAL_MODE: false
FUSED_LAYERNORM: false
FUSED_WINDOW_PROCESS: false
LOCAL_RANK: 0
MODEL:
  DROP_PATH_RATE: 0.2
  DROP_RATE: 0.0
  LABEL_SMOOTHING: 0.1
  NAME: swin_tiny_patch4_window7_224
  NUM_CLASSES: 1000
  PRETRAINED: /content/Swin-Transformer/pretrained_weights/swin_tiny_patch4_window7_224.pth
  RESUME: ''
  SIMMIM:
    NORM_TARGET:
      ENABLE: false
      PATCH_SIZE: 47
  SWIN:
    APE: false
    DEPTHS:
    - 2
    - 2
    - 6
    - 2
    EMBED_DIM: 96
    IN_CHANS: 3
    MLP_RATIO: 4.0
    NUM_HEADS:
    - 3
    - 6
    - 12
    - 24
    PATCH_NORM: true
    PATCH_SIZE: 4
    QKV_BIAS: true
    QK_SCALE: null
    WINDOW_SIZE: 7
  SWINV2:
    APE: false
    DEPTHS:
    - 2
    - 2
    - 6
    - 2
    EMBED_DIM: 96
    IN_CHANS: 3
    MLP_RATIO: 4.0
    NUM_HEADS:
    - 3
    - 6
    - 12
    - 24
    PATCH_NORM: true
    PATCH_SIZE: 4
    PRETRAINED_WINDOW_SIZES:
    - 0
    - 0
    - 0
    - 0
    QKV_BIAS: true
    WINDOW_SIZE: 7
  SWIN_MLP:
    APE: false
    DEPTHS:
    - 2
    - 2
    - 6
    - 2
    EMBED_DIM: 96
    IN_CHANS: 3
    MLP_RATIO: 4.0
    NUM_HEADS:
    - 3
    - 6
    - 12
    - 24
    PATCH_NORM: true
    PATCH_SIZE: 4
    WINDOW_SIZE: 7
  SWIN_MOE:
    APE: false
    AUX_LOSS_WEIGHT: 0.01
    CAPACITY_FACTOR: 1.25
    COSINE_ROUTER: false
    COSINE_ROUTER_DIM: 256
    COSINE_ROUTER_INIT_T: 0.5
    DEPTHS:
    - 2
    - 2
    - 6
    - 2
    EMBED_DIM: 96
    GATE_NOISE: 1.0
    INIT_STD: 0.02
    IN_CHANS: 3
    IS_GSHARD_LOSS: false
    MLP_FC2_BIAS: true
    MLP_RATIO: 4.0
    MOE_BLOCKS:
    - - -1
    - - -1
    - - -1
    - - -1
    MOE_DROP: 0.0
    NORMALIZE_GATE: false
    NUM_HEADS:
    - 3
    - 6
    - 12
    - 24
    NUM_LOCAL_EXPERTS: 1
    PATCH_NORM: true
    PATCH_SIZE: 4
    PRETRAINED_WINDOW_SIZES:
    - 0
    - 0
    - 0
    - 0
    QKV_BIAS: true
    QK_SCALE: null
    TOP_VALUE: 1
    USE_BPR: true
    WINDOW_SIZE: 7
  TYPE: swin
OUTPUT: output/swin_tiny_patch4_window7_224/default
PRINT_FREQ: 10
SAVE_FREQ: 1
SEED: 0
TAG: default
TEST:
  CROP: true
  SEQUENTIAL: false
  SHUFFLE: false
THROUGHPUT_MODE: false
TRAIN:
  ACCUMULATION_STEPS: 2
  AUTO_RESUME: true
  BASE_LR: 3.125e-05
  CLIP_GRAD: 5.0
  EPOCHS: 15
  LAYER_DECAY: 1.0
  LR_SCHEDULER:
    DECAY_EPOCHS: 30
    DECAY_RATE: 0.1
    GAMMA: 0.1
    MULTISTEPS: []
    NAME: cosine
    WARMUP_PREFIX: true
  MIN_LR: 3.125e-07
  MOE:
    SAVE_MASTER: false
  OPTIMIZER:
    BETAS:
    - 0.9
    - 0.999
    EPS: 1.0e-08
    MOMENTUM: 0.9
    NAME: adamw
  START_EPOCH: 0
  USE_CHECKPOINT: true
  WARMUP_EPOCHS: 3
  WARMUP_LR: 3.125e-08
  WEIGHT_DECAY: 0.05

[2025-12-08 16:16:27 swin_tiny_patch4_window7_224] (main.py 352): INFO {"cfg": "configs/swin/swin_tiny_patch4_window7_224.yaml", "opts": ["TRAIN.EPOCHS", "15", "TRAIN.WARMUP_EPOCHS", "3"], "batch_size": 16, "data_path": "/content/drive/MyDrive/Tulsi_train-70_val-30", "zip": false, "cache_mode": "part", "pretrained": "/content/Swin-Transformer/pretrained_weights/swin_tiny_patch4_window7_224.pth", "resume": null, "accumulation_steps": 2, "use_checkpoint": true, "disable_amp": false, "amp_opt_level": null, "output": "output", "tag": null, "eval": false, "throughput": false, "fused_window_process": false, "fused_layernorm": false, "optim": null}
[2025-12-08 16:16:35 swin_tiny_patch4_window7_224] (main.py 93): INFO Creating model:swin/swin_tiny_patch4_window7_224
[2025-12-08 16:16:36 swin_tiny_patch4_window7_224] (main.py 95): INFO SwinTransformer(
  (patch_embed): PatchEmbed(
    (proj): Conv2d(3, 96, kernel_size=(4, 4), stride=(4, 4))
    (norm): LayerNorm((96,), eps=1e-05, elementwise_affine=True)
  )
  (pos_drop): Dropout(p=0.0, inplace=False)
  (layers): ModuleList(
    (0): BasicLayer(
      dim=96, input_resolution=(56, 56), depth=2
      (blocks): ModuleList(
        (0): SwinTransformerBlock(
          dim=96, input_resolution=(56, 56), num_heads=3, window_size=7, shift_size=0, mlp_ratio=4.0
          (norm1): LayerNorm((96,), eps=1e-05, elementwise_affine=True)
          (attn): WindowAttention(
            dim=96, window_size=(7, 7), num_heads=3
            (qkv): Linear(in_features=96, out_features=288, bias=True)
            (attn_drop): Dropout(p=0.0, inplace=False)
            (proj): Linear(in_features=96, out_features=96, bias=True)
            (proj_drop): Dropout(p=0.0, inplace=False)
            (softmax): Softmax(dim=-1)
          )
          (drop_path): Identity()
          (norm2): LayerNorm((96,), eps=1e-05, elementwise_affine=True)
          (mlp): Mlp(
            (fc1): Linear(in_features=96, out_features=384, bias=True)
            (act): GELU(approximate='none')
            (fc2): Linear(in_features=384, out_features=96, bias=True)
            (drop): Dropout(p=0.0, inplace=False)
          )
        )
        (1): SwinTransformerBlock(
          dim=96, input_resolution=(56, 56), num_heads=3, window_size=7, shift_size=3, mlp_ratio=4.0
          (norm1): LayerNorm((96,), eps=1e-05, elementwise_affine=True)
          (attn): WindowAttention(
            dim=96, window_size=(7, 7), num_heads=3
            (qkv): Linear(in_features=96, out_features=288, bias=True)
            (attn_drop): Dropout(p=0.0, inplace=False)
            (proj): Linear(in_features=96, out_features=96, bias=True)
            (proj_drop): Dropout(p=0.0, inplace=False)
            (softmax): Softmax(dim=-1)
          )
          (drop_path): DropPath()
          (norm2): LayerNorm((96,), eps=1e-05, elementwise_affine=True)
          (mlp): Mlp(
            (fc1): Linear(in_features=96, out_features=384, bias=True)
            (act): GELU(approximate='none')
            (fc2): Linear(in_features=384, out_features=96, bias=True)
            (drop): Dropout(p=0.0, inplace=False)
          )
        )
      )
      (downsample): PatchMerging(
        input_resolution=(56, 56), dim=96
        (reduction): Linear(in_features=384, out_features=192, bias=False)
        (norm): LayerNorm((384,), eps=1e-05, elementwise_affine=True)
      )
    )
    (1): BasicLayer(
      dim=192, input_resolution=(28, 28), depth=2
      (blocks): ModuleList(
        (0): SwinTransformerBlock(
          dim=192, input_resolution=(28, 28), num_heads=6, window_size=7, shift_size=0, mlp_ratio=4.0
          (norm1): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
          (attn): WindowAttention(
            dim=192, window_size=(7, 7), num_heads=6
            (qkv): Linear(in_features=192, out_features=576, bias=True)
            (attn_drop): Dropout(p=0.0, inplace=False)
            (proj): Linear(in_features=192, out_features=192, bias=True)
            (proj_drop): Dropout(p=0.0, inplace=False)
            (softmax): Softmax(dim=-1)
          )
          (drop_path): DropPath()
          (norm2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
          (mlp): Mlp(
            (fc1): Linear(in_features=192, out_features=768, bias=True)
            (act): GELU(approximate='none')
            (fc2): Linear(in_features=768, out_features=192, bias=True)
            (drop): Dropout(p=0.0, inplace=False)
          )
        )
        (1): SwinTransformerBlock(
          dim=192, input_resolution=(28, 28), num_heads=6, window_size=7, shift_size=3, mlp_ratio=4.0
          (norm1): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
          (attn): WindowAttention(
            dim=192, window_size=(7, 7), num_heads=6
            (qkv): Linear(in_features=192, out_features=576, bias=True)
            (attn_drop): Dropout(p=0.0, inplace=False)
            (proj): Linear(in_features=192, out_features=192, bias=True)
            (proj_drop): Dropout(p=0.0, inplace=False)
            (softmax): Softmax(dim=-1)
          )
          (drop_path): DropPath()
          (norm2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
          (mlp): Mlp(
            (fc1): Linear(in_features=192, out_features=768, bias=True)
            (act): GELU(approximate='none')
            (fc2): Linear(in_features=768, out_features=192, bias=True)
            (drop): Dropout(p=0.0, inplace=False)
          )
        )
      )
      (downsample): PatchMerging(
        input_resolution=(28, 28), dim=192
        (reduction): Linear(in_features=768, out_features=384, bias=False)
        (norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
      )
    )
    (2): BasicLayer(
      dim=384, input_resolution=(14, 14), depth=6
      (blocks): ModuleList(
        (0): SwinTransformerBlock(
          dim=384, input_resolution=(14, 14), num_heads=12, window_size=7, shift_size=0, mlp_ratio=4.0
          (norm1): LayerNorm((384,), eps=1e-05, elementwise_affine=True)
          (attn): WindowAttention(
            dim=384, window_size=(7, 7), num_heads=12
            (qkv): Linear(in_features=384, out_features=1152, bias=True)
            (attn_drop): Dropout(p=0.0, inplace=False)
            (proj): Linear(in_features=384, out_features=384, bias=True)
            (proj_drop): Dropout(p=0.0, inplace=False)
            (softmax): Softmax(dim=-1)
          )
          (drop_path): DropPath()
          (norm2): LayerNorm((384,), eps=1e-05, elementwise_affine=True)
          (mlp): Mlp(
            (fc1): Linear(in_features=384, out_features=1536, bias=True)
            (act): GELU(approximate='none')
            (fc2): Linear(in_features=1536, out_features=384, bias=True)
            (drop): Dropout(p=0.0, inplace=False)
          )
        )
        (1): SwinTransformerBlock(
          dim=384, input_resolution=(14, 14), num_heads=12, window_size=7, shift_size=3, mlp_ratio=4.0
          (norm1): LayerNorm((384,), eps=1e-05, elementwise_affine=True)
          (attn): WindowAttention(
            dim=384, window_size=(7, 7), num_heads=12
            (qkv): Linear(in_features=384, out_features=1152, bias=True)
            (attn_drop): Dropout(p=0.0, inplace=False)
            (proj): Linear(in_features=384, out_features=384, bias=True)
            (proj_drop): Dropout(p=0.0, inplace=False)
            (softmax): Softmax(dim=-1)
          )
          (drop_path): DropPath()
          (norm2): LayerNorm((384,), eps=1e-05, elementwise_affine=True)
          (mlp): Mlp(
            (fc1): Linear(in_features=384, out_features=1536, bias=True)
            (act): GELU(approximate='none')
            (fc2): Linear(in_features=1536, out_features=384, bias=True)
            (drop): Dropout(p=0.0, inplace=False)
          )
        )
        (2): SwinTransformerBlock(
          dim=384, input_resolution=(14, 14), num_heads=12, window_size=7, shift_size=0, mlp_ratio=4.0
          (norm1): LayerNorm((384,), eps=1e-05, elementwise_affine=True)
          (attn): WindowAttention(
            dim=384, window_size=(7, 7), num_heads=12
            (qkv): Linear(in_features=384, out_features=1152, bias=True)
            (attn_drop): Dropout(p=0.0, inplace=False)
            (proj): Linear(in_features=384, out_features=384, bias=True)
            (proj_drop): Dropout(p=0.0, inplace=False)
            (softmax): Softmax(dim=-1)
          )
          (drop_path): DropPath()
          (norm2): LayerNorm((384,), eps=1e-05, elementwise_affine=True)
          (mlp): Mlp(
            (fc1): Linear(in_features=384, out_features=1536, bias=True)
            (act): GELU(approximate='none')
            (fc2): Linear(in_features=1536, out_features=384, bias=True)
            (drop): Dropout(p=0.0, inplace=False)
          )
        )
        (3): SwinTransformerBlock(
          dim=384, input_resolution=(14, 14), num_heads=12, window_size=7, shift_size=3, mlp_ratio=4.0
          (norm1): LayerNorm((384,), eps=1e-05, elementwise_affine=True)
          (attn): WindowAttention(
            dim=384, window_size=(7, 7), num_heads=12
            (qkv): Linear(in_features=384, out_features=1152, bias=True)
            (attn_drop): Dropout(p=0.0, inplace=False)
            (proj): Linear(in_features=384, out_features=384, bias=True)
            (proj_drop): Dropout(p=0.0, inplace=False)
            (softmax): Softmax(dim=-1)
          )
          (drop_path): DropPath()
          (norm2): LayerNorm((384,), eps=1e-05, elementwise_affine=True)
          (mlp): Mlp(
            (fc1): Linear(in_features=384, out_features=1536, bias=True)
            (act): GELU(approximate='none')
            (fc2): Linear(in_features=1536, out_features=384, bias=True)
            (drop): Dropout(p=0.0, inplace=False)
          )
        )
        (4): SwinTransformerBlock(
          dim=384, input_resolution=(14, 14), num_heads=12, window_size=7, shift_size=0, mlp_ratio=4.0
          (norm1): LayerNorm((384,), eps=1e-05, elementwise_affine=True)
          (attn): WindowAttention(
            dim=384, window_size=(7, 7), num_heads=12
            (qkv): Linear(in_features=384, out_features=1152, bias=True)
            (attn_drop): Dropout(p=0.0, inplace=False)
            (proj): Linear(in_features=384, out_features=384, bias=True)
            (proj_drop): Dropout(p=0.0, inplace=False)
            (softmax): Softmax(dim=-1)
          )
          (drop_path): DropPath()
          (norm2): LayerNorm((384,), eps=1e-05, elementwise_affine=True)
          (mlp): Mlp(
            (fc1): Linear(in_features=384, out_features=1536, bias=True)
            (act): GELU(approximate='none')
            (fc2): Linear(in_features=1536, out_features=384, bias=True)
            (drop): Dropout(p=0.0, inplace=False)
          )
        )
        (5): SwinTransformerBlock(
          dim=384, input_resolution=(14, 14), num_heads=12, window_size=7, shift_size=3, mlp_ratio=4.0
          (norm1): LayerNorm((384,), eps=1e-05, elementwise_affine=True)
          (attn): WindowAttention(
            dim=384, window_size=(7, 7), num_heads=12
            (qkv): Linear(in_features=384, out_features=1152, bias=True)
            (attn_drop): Dropout(p=0.0, inplace=False)
            (proj): Linear(in_features=384, out_features=384, bias=True)
            (proj_drop): Dropout(p=0.0, inplace=False)
            (softmax): Softmax(dim=-1)
          )
          (drop_path): DropPath()
          (norm2): LayerNorm((384,), eps=1e-05, elementwise_affine=True)
          (mlp): Mlp(
            (fc1): Linear(in_features=384, out_features=1536, bias=True)
            (act): GELU(approximate='none')
            (fc2): Linear(in_features=1536, out_features=384, bias=True)
            (drop): Dropout(p=0.0, inplace=False)
          )
        )
      )
      (downsample): PatchMerging(
        input_resolution=(14, 14), dim=384
        (reduction): Linear(in_features=1536, out_features=768, bias=False)
        (norm): LayerNorm((1536,), eps=1e-05, elementwise_affine=True)
      )
    )
    (3): BasicLayer(
      dim=768, input_resolution=(7, 7), depth=2
      (blocks): ModuleList(
        (0-1): 2 x SwinTransformerBlock(
          dim=768, input_resolution=(7, 7), num_heads=24, window_size=7, shift_size=0, mlp_ratio=4.0
          (norm1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (attn): WindowAttention(
            dim=768, window_size=(7, 7), num_heads=24
            (qkv): Linear(in_features=768, out_features=2304, bias=True)
            (attn_drop): Dropout(p=0.0, inplace=False)
            (proj): Linear(in_features=768, out_features=768, bias=True)
            (proj_drop): Dropout(p=0.0, inplace=False)
            (softmax): Softmax(dim=-1)
          )
          (drop_path): DropPath()
          (norm2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
          (mlp): Mlp(
            (fc1): Linear(in_features=768, out_features=3072, bias=True)
            (act): GELU(approximate='none')
            (fc2): Linear(in_features=3072, out_features=768, bias=True)
            (drop): Dropout(p=0.0, inplace=False)
          )
        )
      )
    )
  )
  (norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
  (avgpool): AdaptiveAvgPool1d(output_size=1)
  (head): Linear(in_features=768, out_features=1000, bias=True)
)
[2025-12-08 16:16:36 swin_tiny_patch4_window7_224] (main.py 98): INFO number of params: 28288354
[2025-12-08 16:16:36 swin_tiny_patch4_window7_224] (main.py 101): INFO number of GFLOPs: 4.49440512
[2025-12-08 16:16:36 swin_tiny_patch4_window7_224] (main.py 135): INFO no checkpoint found in output/swin_tiny_patch4_window7_224/default, ignoring auto resume
[2025-12-08 16:16:36 swin_tiny_patch4_window7_224] (utils.py 46): INFO ==============> Loading weight /content/Swin-Transformer/pretrained_weights/swin_tiny_patch4_window7_224.pth for fine-tuning......
[2025-12-08 16:16:36 swin_tiny_patch4_window7_224] (utils.py 127): WARNING _IncompatibleKeys(missing_keys=['layers.0.blocks.0.attn.relative_position_index', 'layers.0.blocks.1.attn_mask', 'layers.0.blocks.1.attn.relative_position_index', 'layers.1.blocks.0.attn.relative_position_index', 'layers.1.blocks.1.attn_mask', 'layers.1.blocks.1.attn.relative_position_index', 'layers.2.blocks.0.attn.relative_position_index', 'layers.2.blocks.1.attn_mask', 'layers.2.blocks.1.attn.relative_position_index', 'layers.2.blocks.2.attn.relative_position_index', 'layers.2.blocks.3.attn_mask', 'layers.2.blocks.3.attn.relative_position_index', 'layers.2.blocks.4.attn.relative_position_index', 'layers.2.blocks.5.attn_mask', 'layers.2.blocks.5.attn.relative_position_index', 'layers.3.blocks.0.attn.relative_position_index', 'layers.3.blocks.1.attn.relative_position_index'], unexpected_keys=[])
[2025-12-08 16:16:36 swin_tiny_patch4_window7_224] (utils.py 129): INFO => loaded successfully '/content/Swin-Transformer/pretrained_weights/swin_tiny_patch4_window7_224.pth'
[2025-12-08 16:16:44 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [0/110]	Time 8.141 (8.141)	Loss 9.8047 (9.8047)	Acc@1 0.000 (0.000)	Acc@5 0.000 (0.000)	Mem 426MB
[2025-12-08 16:16:50 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [10/110]	Time 2.495 (1.213)	Loss 9.0547 (10.1612)	Acc@1 0.000 (0.000)	Acc@5 0.000 (0.000)	Mem 426MB
[2025-12-08 16:16:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [20/110]	Time 0.511 (0.968)	Loss 9.3438 (9.7891)	Acc@1 0.000 (0.000)	Acc@5 0.000 (0.000)	Mem 426MB
[2025-12-08 16:17:04 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [30/110]	Time 0.038 (0.880)	Loss 8.9219 (9.5595)	Acc@1 0.000 (0.000)	Acc@5 0.000 (0.000)	Mem 426MB
[2025-12-08 16:17:19 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [40/110]	Time 0.050 (1.049)	Loss 9.8047 (9.5596)	Acc@1 0.000 (0.000)	Acc@5 0.000 (0.000)	Mem 426MB
[2025-12-08 16:17:26 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [50/110]	Time 0.032 (0.979)	Loss 9.3594 (9.5455)	Acc@1 0.000 (0.000)	Acc@5 0.000 (0.000)	Mem 426MB
[2025-12-08 16:17:38 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [60/110]	Time 6.048 (1.017)	Loss 9.9922 (9.5479)	Acc@1 0.000 (0.000)	Acc@5 0.000 (0.000)	Mem 426MB
[2025-12-08 16:17:46 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [70/110]	Time 0.050 (0.976)	Loss 9.6641 (9.5157)	Acc@1 0.000 (0.000)	Acc@5 0.000 (0.000)	Mem 426MB
[2025-12-08 16:17:53 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [80/110]	Time 0.048 (0.947)	Loss 9.5625 (9.5231)	Acc@1 0.000 (0.000)	Acc@5 0.000 (0.000)	Mem 426MB
[2025-12-08 16:17:59 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [90/110]	Time 0.033 (0.908)	Loss 9.0859 (9.5055)	Acc@1 0.000 (0.000)	Acc@5 0.000 (0.000)	Mem 426MB
[2025-12-08 16:18:11 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [100/110]	Time 5.920 (0.936)	Loss 8.4844 (9.4666)	Acc@1 0.000 (0.000)	Acc@5 0.000 (0.000)	Mem 426MB
[2025-12-08 16:18:17 swin_tiny_patch4_window7_224] (main.py 278): INFO  * Acc@1 0.000 Acc@5 0.000
[2025-12-08 16:18:17 swin_tiny_patch4_window7_224] (main.py 147): INFO Accuracy of the network on the 1760 test images: 0.0%
[2025-12-08 16:18:17 swin_tiny_patch4_window7_224] (main.py 153): INFO Start training
[2025-12-08 16:18:24 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][0/256]	eta 0:31:23 lr 0.000000	 wd 0.0500	time 7.3586 (7.3586)	loss 4.3156 (4.3156)	grad_norm 0.0000 (0.0000)	loss_scale 65536.0000 (65536.0000)	mem 695MB
[2025-12-08 16:18:31 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][10/256]	eta 0:05:07 lr 0.000000	 wd 0.0500	time 0.1247 (1.2482)	loss 4.4504 (4.2248)	grad_norm 26.0064 (nan)	loss_scale 16384.0000 (26810.1818)	mem 921MB
[2025-12-08 16:18:37 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][20/256]	eta 0:03:44 lr 0.000001	 wd 0.0500	time 0.1233 (0.9518)	loss 4.3685 (4.2801)	grad_norm 33.3361 (nan)	loss_scale 16384.0000 (21845.3333)	mem 921MB
[2025-12-08 16:18:44 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][30/256]	eta 0:03:15 lr 0.000001	 wd 0.0500	time 0.1286 (0.8653)	loss 4.3437 (4.2972)	grad_norm 31.1871 (nan)	loss_scale 16384.0000 (20083.6129)	mem 921MB
[2025-12-08 16:18:55 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][40/256]	eta 0:03:18 lr 0.000002	 wd 0.0500	time 4.0587 (0.9194)	loss 4.2975 (4.2785)	grad_norm 30.4545 (nan)	loss_scale 16384.0000 (19181.2683)	mem 921MB
[2025-12-08 16:19:02 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][50/256]	eta 0:03:03 lr 0.000002	 wd 0.0500	time 0.1274 (0.8894)	loss 3.9294 (4.2533)	grad_norm 26.2191 (nan)	loss_scale 16384.0000 (18632.7843)	mem 921MB
[2025-12-08 16:19:09 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][60/256]	eta 0:02:46 lr 0.000002	 wd 0.0500	time 0.1385 (0.8473)	loss 3.9494 (4.2275)	grad_norm 28.6338 (nan)	loss_scale 16384.0000 (18264.1311)	mem 921MB
[2025-12-08 16:19:15 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][70/256]	eta 0:02:30 lr 0.000003	 wd 0.0500	time 0.1253 (0.8114)	loss 4.0818 (4.1921)	grad_norm 29.5244 (nan)	loss_scale 16384.0000 (17999.3239)	mem 921MB
[2025-12-08 16:19:36 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][80/256]	eta 0:02:51 lr 0.000003	 wd 0.0500	time 14.7946 (0.9743)	loss 3.7849 (4.1402)	grad_norm 26.0718 (nan)	loss_scale 16384.0000 (17799.9012)	mem 921MB
[2025-12-08 16:19:44 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][90/256]	eta 0:02:39 lr 0.000004	 wd 0.0500	time 0.1256 (0.9598)	loss 3.5666 (4.0836)	grad_norm inf (nan)	loss_scale 8192.0000 (17464.2637)	mem 921MB
[2025-12-08 16:19:51 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][100/256]	eta 0:02:25 lr 0.000004	 wd 0.0500	time 0.1238 (0.9307)	loss 2.9667 (4.0115)	grad_norm 33.9108 (nan)	loss_scale 8192.0000 (16546.2178)	mem 921MB
[2025-12-08 16:19:58 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][110/256]	eta 0:02:12 lr 0.000004	 wd 0.0500	time 0.1292 (0.9093)	loss 2.9654 (3.9390)	grad_norm 40.0141 (nan)	loss_scale 8192.0000 (15793.5856)	mem 921MB
[2025-12-08 16:20:10 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][120/256]	eta 0:02:07 lr 0.000005	 wd 0.0500	time 5.6335 (0.9374)	loss 3.0443 (3.8617)	grad_norm 38.7037 (nan)	loss_scale 8192.0000 (15165.3554)	mem 921MB
[2025-12-08 16:20:16 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][130/256]	eta 0:01:54 lr 0.000005	 wd 0.0500	time 0.1265 (0.9104)	loss 2.4747 (3.7722)	grad_norm 39.0961 (nan)	loss_scale 8192.0000 (14633.0382)	mem 921MB
[2025-12-08 16:20:23 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][140/256]	eta 0:01:43 lr 0.000006	 wd 0.0500	time 0.1270 (0.8916)	loss 2.0418 (3.6738)	grad_norm inf (nan)	loss_scale 4096.0000 (14118.1277)	mem 921MB
[2025-12-08 16:20:29 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][150/256]	eta 0:01:32 lr 0.000006	 wd 0.0500	time 0.1263 (0.8744)	loss 2.0707 (3.5728)	grad_norm 31.9776 (nan)	loss_scale 4096.0000 (13454.4106)	mem 921MB
[2025-12-08 16:20:39 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][160/256]	eta 0:01:24 lr 0.000006	 wd 0.0500	time 4.0250 (0.8852)	loss 2.1233 (3.4721)	grad_norm 23.6168 (nan)	loss_scale 4096.0000 (12873.1429)	mem 921MB
[2025-12-08 16:20:46 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][170/256]	eta 0:01:14 lr 0.000007	 wd 0.0500	time 0.1288 (0.8716)	loss 1.8874 (3.3704)	grad_norm 26.8260 (nan)	loss_scale 4096.0000 (12359.8596)	mem 921MB
[2025-12-08 16:20:53 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][180/256]	eta 0:01:05 lr 0.000007	 wd 0.0500	time 0.1279 (0.8616)	loss 1.5643 (3.2755)	grad_norm 21.5001 (nan)	loss_scale 4096.0000 (11903.2928)	mem 921MB
[2025-12-08 16:21:01 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][190/256]	eta 0:00:56 lr 0.000008	 wd 0.0500	time 0.1351 (0.8577)	loss 1.3834 (3.1812)	grad_norm 26.4004 (nan)	loss_scale 4096.0000 (11494.5340)	mem 921MB
[2025-12-08 16:21:10 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][200/256]	eta 0:00:48 lr 0.000008	 wd 0.0500	time 2.6927 (0.8613)	loss 1.3477 (3.0889)	grad_norm 37.5644 (nan)	loss_scale 4096.0000 (11126.4478)	mem 921MB
[2025-12-08 16:21:16 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][210/256]	eta 0:00:39 lr 0.000008	 wd 0.0500	time 0.1283 (0.8504)	loss 1.5003 (3.0035)	grad_norm 23.1965 (nan)	loss_scale 4096.0000 (10793.2512)	mem 921MB
[2025-12-08 16:21:27 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][220/256]	eta 0:00:30 lr 0.000009	 wd 0.0500	time 0.1510 (0.8597)	loss 1.4713 (2.9287)	grad_norm 28.6159 (nan)	loss_scale 4096.0000 (10490.2081)	mem 921MB
[2025-12-08 16:21:33 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][230/256]	eta 0:00:22 lr 0.000009	 wd 0.0500	time 0.1479 (0.8502)	loss 1.1351 (2.8586)	grad_norm 20.6478 (nan)	loss_scale 4096.0000 (10213.4026)	mem 921MB
[2025-12-08 16:21:40 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][240/256]	eta 0:00:13 lr 0.000010	 wd 0.0500	time 1.3613 (0.8430)	loss 1.1721 (2.7905)	grad_norm 23.8414 (nan)	loss_scale 4096.0000 (9959.5685)	mem 921MB
[2025-12-08 16:21:46 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [0/15][250/256]	eta 0:00:05 lr 0.000010	 wd 0.0500	time 0.1304 (0.8338)	loss 1.0151 (2.7239)	grad_norm 14.7991 (nan)	loss_scale 4096.0000 (9725.9602)	mem 921MB
[2025-12-08 16:21:52 swin_tiny_patch4_window7_224] (main.py 231): INFO EPOCH 0 training takes 0:03:34
[2025-12-08 16:21:52 swin_tiny_patch4_window7_224] (utils.py 145): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_0.pth saving......
[2025-12-08 16:21:52 swin_tiny_patch4_window7_224] (utils.py 147): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_0.pth saved !!!
[2025-12-08 16:21:55 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [0/110]	Time 2.865 (2.865)	Loss 1.0010 (1.0010)	Acc@1 87.500 (87.500)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:21:56 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [10/110]	Time 0.109 (0.323)	Loss 0.6465 (0.9789)	Acc@1 87.500 (76.705)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:21:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [20/110]	Time 0.035 (0.203)	Loss 1.0342 (1.2305)	Acc@1 62.500 (66.071)	Acc@5 100.000 (97.619)	Mem 921MB
[2025-12-08 16:21:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [30/110]	Time 0.038 (0.163)	Loss 2.8555 (1.2320)	Acc@1 12.500 (66.734)	Acc@5 93.750 (98.185)	Mem 921MB
[2025-12-08 16:21:58 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [40/110]	Time 0.127 (0.146)	Loss 0.5859 (1.1318)	Acc@1 81.250 (68.902)	Acc@5 100.000 (98.628)	Mem 921MB
[2025-12-08 16:22:00 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [50/110]	Time 0.099 (0.143)	Loss 1.1348 (1.1021)	Acc@1 50.000 (69.363)	Acc@5 100.000 (98.897)	Mem 921MB
[2025-12-08 16:22:01 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [60/110]	Time 0.055 (0.137)	Loss 0.6143 (1.0382)	Acc@1 93.750 (72.746)	Acc@5 100.000 (99.078)	Mem 921MB
[2025-12-08 16:22:02 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [70/110]	Time 0.110 (0.134)	Loss 0.6538 (0.9915)	Acc@1 93.750 (74.208)	Acc@5 100.000 (99.208)	Mem 921MB
[2025-12-08 16:22:03 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [80/110]	Time 0.047 (0.132)	Loss 1.2695 (0.9400)	Acc@1 81.250 (76.620)	Acc@5 100.000 (99.306)	Mem 921MB
[2025-12-08 16:22:04 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [90/110]	Time 0.079 (0.127)	Loss 0.2128 (0.9305)	Acc@1 100.000 (76.992)	Acc@5 100.000 (99.382)	Mem 921MB
[2025-12-08 16:22:05 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [100/110]	Time 0.048 (0.122)	Loss 0.5830 (0.9425)	Acc@1 81.250 (74.381)	Acc@5 100.000 (99.443)	Mem 921MB
[2025-12-08 16:22:05 swin_tiny_patch4_window7_224] (main.py 278): INFO  * Acc@1 75.625 Acc@5 99.489
[2025-12-08 16:22:05 swin_tiny_patch4_window7_224] (main.py 165): INFO Accuracy of the network on the 1760 test images: 75.6%
[2025-12-08 16:22:05 swin_tiny_patch4_window7_224] (main.py 167): INFO Max accuracy: 75.62%
[2025-12-08 16:22:07 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][0/256]	eta 0:05:29 lr 0.000010	 wd 0.0500	time 1.2878 (1.2878)	loss 1.0676 (1.0676)	grad_norm 0.0000 (0.0000)	loss_scale 4096.0000 (4096.0000)	mem 921MB
[2025-12-08 16:22:09 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][10/256]	eta 0:01:18 lr 0.000011	 wd 0.0500	time 0.1810 (0.3198)	loss 1.1136 (1.1374)	grad_norm 22.2228 (22.9132)	loss_scale 4096.0000 (4096.0000)	mem 921MB
[2025-12-08 16:22:11 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][20/256]	eta 0:01:01 lr 0.000011	 wd 0.0500	time 0.2653 (0.2614)	loss 0.9679 (1.1324)	grad_norm 19.9265 (21.3481)	loss_scale 4096.0000 (4096.0000)	mem 921MB
[2025-12-08 16:22:12 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][30/256]	eta 0:00:51 lr 0.000012	 wd 0.0500	time 0.1346 (0.2271)	loss 1.0758 (1.1046)	grad_norm 15.7693 (20.4390)	loss_scale 4096.0000 (4096.0000)	mem 921MB
[2025-12-08 16:22:14 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][40/256]	eta 0:00:47 lr 0.000012	 wd 0.0500	time 0.2473 (0.2217)	loss 1.0190 (1.0791)	grad_norm 16.5814 (inf)	loss_scale 2048.0000 (3596.4878)	mem 921MB
[2025-12-08 16:22:17 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][50/256]	eta 0:00:46 lr 0.000012	 wd 0.0500	time 0.2555 (0.2253)	loss 1.2187 (1.0659)	grad_norm 29.2766 (inf)	loss_scale 2048.0000 (3292.8627)	mem 921MB
[2025-12-08 16:22:19 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][60/256]	eta 0:00:42 lr 0.000013	 wd 0.0500	time 0.1482 (0.2177)	loss 1.1547 (1.0616)	grad_norm 14.4174 (inf)	loss_scale 2048.0000 (3088.7869)	mem 921MB
[2025-12-08 16:22:20 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][70/256]	eta 0:00:38 lr 0.000013	 wd 0.0500	time 0.1405 (0.2096)	loss 0.9871 (1.0649)	grad_norm 22.9458 (inf)	loss_scale 2048.0000 (2942.1972)	mem 921MB
[2025-12-08 16:22:22 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][80/256]	eta 0:00:35 lr 0.000014	 wd 0.0500	time 0.1443 (0.2029)	loss 0.9865 (1.0548)	grad_norm 15.6917 (inf)	loss_scale 2048.0000 (2831.8025)	mem 921MB
[2025-12-08 16:22:23 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][90/256]	eta 0:00:32 lr 0.000014	 wd 0.0500	time 0.1412 (0.1974)	loss 0.7782 (1.0471)	grad_norm 22.5886 (inf)	loss_scale 2048.0000 (2745.6703)	mem 921MB
[2025-12-08 16:22:25 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][100/256]	eta 0:00:30 lr 0.000014	 wd 0.0500	time 0.1422 (0.1939)	loss 0.8515 (1.0393)	grad_norm 11.4153 (inf)	loss_scale 2048.0000 (2676.5941)	mem 921MB
[2025-12-08 16:22:26 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][110/256]	eta 0:00:27 lr 0.000015	 wd 0.0500	time 0.1385 (0.1907)	loss 0.9883 (1.0325)	grad_norm 11.6474 (inf)	loss_scale 2048.0000 (2619.9640)	mem 921MB
[2025-12-08 16:22:28 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][120/256]	eta 0:00:25 lr 0.000015	 wd 0.0500	time 0.2262 (0.1908)	loss 1.1074 (1.0301)	grad_norm 13.8704 (inf)	loss_scale 2048.0000 (2572.6942)	mem 921MB
[2025-12-08 16:22:31 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][130/256]	eta 0:00:24 lr 0.000016	 wd 0.0500	time 0.2580 (0.1955)	loss 1.1132 (1.0293)	grad_norm 21.1344 (inf)	loss_scale 2048.0000 (2532.6412)	mem 921MB
[2025-12-08 16:22:33 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][140/256]	eta 0:00:22 lr 0.000016	 wd 0.0500	time 0.1395 (0.1952)	loss 0.9727 (1.0231)	grad_norm 8.9138 (inf)	loss_scale 2048.0000 (2498.2695)	mem 921MB
[2025-12-08 16:22:34 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][150/256]	eta 0:00:20 lr 0.000016	 wd 0.0500	time 0.1394 (0.1928)	loss 0.9671 (1.0189)	grad_norm 24.9671 (inf)	loss_scale 2048.0000 (2468.4503)	mem 921MB
[2025-12-08 16:22:36 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][160/256]	eta 0:00:18 lr 0.000017	 wd 0.0500	time 0.1376 (0.1909)	loss 0.7812 (1.0136)	grad_norm 12.4752 (inf)	loss_scale 2048.0000 (2442.3354)	mem 921MB
[2025-12-08 16:22:38 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][170/256]	eta 0:00:16 lr 0.000017	 wd 0.0500	time 0.1420 (0.1892)	loss 0.9778 (1.0144)	grad_norm 23.9211 (inf)	loss_scale 2048.0000 (2419.2749)	mem 921MB
[2025-12-08 16:22:39 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][180/256]	eta 0:00:14 lr 0.000018	 wd 0.0500	time 0.1419 (0.1875)	loss 1.0092 (1.0096)	grad_norm 14.2538 (inf)	loss_scale 2048.0000 (2398.7624)	mem 921MB
[2025-12-08 16:22:41 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][190/256]	eta 0:00:12 lr 0.000018	 wd 0.0500	time 0.1650 (0.1862)	loss 0.7262 (1.0039)	grad_norm 10.2509 (inf)	loss_scale 2048.0000 (2380.3979)	mem 921MB
[2025-12-08 16:22:43 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][200/256]	eta 0:00:10 lr 0.000018	 wd 0.0500	time 0.1625 (0.1856)	loss 0.7198 (0.9956)	grad_norm 9.9701 (inf)	loss_scale 2048.0000 (2363.8607)	mem 921MB
[2025-12-08 16:22:45 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][210/256]	eta 0:00:08 lr 0.000019	 wd 0.0500	time 0.2463 (0.1886)	loss 0.7895 (0.9900)	grad_norm 8.7118 (inf)	loss_scale 2048.0000 (2348.8910)	mem 921MB
[2025-12-08 16:22:47 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][220/256]	eta 0:00:06 lr 0.000019	 wd 0.0500	time 0.1506 (0.1887)	loss 1.0708 (0.9847)	grad_norm 6.9734 (inf)	loss_scale 2048.0000 (2335.2760)	mem 921MB
[2025-12-08 16:22:48 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][230/256]	eta 0:00:04 lr 0.000020	 wd 0.0500	time 0.1468 (0.1871)	loss 0.9686 (0.9816)	grad_norm 10.6456 (inf)	loss_scale 2048.0000 (2322.8398)	mem 921MB
[2025-12-08 16:22:50 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][240/256]	eta 0:00:02 lr 0.000020	 wd 0.0500	time 0.1351 (0.1856)	loss 0.7823 (0.9772)	grad_norm 20.1823 (inf)	loss_scale 2048.0000 (2311.4357)	mem 921MB
[2025-12-08 16:22:51 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [1/15][250/256]	eta 0:00:01 lr 0.000021	 wd 0.0500	time 0.1308 (0.1837)	loss 1.0312 (0.9734)	grad_norm 9.2289 (inf)	loss_scale 2048.0000 (2300.9402)	mem 921MB
[2025-12-08 16:22:52 swin_tiny_patch4_window7_224] (main.py 231): INFO EPOCH 1 training takes 0:00:46
[2025-12-08 16:22:52 swin_tiny_patch4_window7_224] (utils.py 145): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_1.pth saving......
[2025-12-08 16:22:53 swin_tiny_patch4_window7_224] (utils.py 147): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_1.pth saved !!!
[2025-12-08 16:22:54 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [0/110]	Time 1.534 (1.534)	Loss 0.0524 (0.0524)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:22:55 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [10/110]	Time 0.066 (0.204)	Loss 0.2081 (0.0713)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:22:56 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [20/110]	Time 0.048 (0.139)	Loss 0.2686 (0.2864)	Acc@1 100.000 (95.536)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:22:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [30/110]	Time 0.037 (0.121)	Loss 0.1744 (0.2930)	Acc@1 100.000 (93.750)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:22:58 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [40/110]	Time 0.184 (0.133)	Loss 0.1865 (0.3332)	Acc@1 100.000 (91.768)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:22:59 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [50/110]	Time 0.133 (0.129)	Loss 0.4202 (0.3335)	Acc@1 100.000 (93.382)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:23:01 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [60/110]	Time 0.045 (0.126)	Loss 0.2434 (0.3278)	Acc@1 100.000 (94.467)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:23:01 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [70/110]	Time 0.049 (0.118)	Loss 0.2007 (0.3102)	Acc@1 100.000 (95.070)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:23:02 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [80/110]	Time 0.200 (0.115)	Loss 0.1945 (0.2872)	Acc@1 100.000 (95.679)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:23:03 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [90/110]	Time 0.088 (0.111)	Loss 0.0866 (0.2673)	Acc@1 100.000 (96.154)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:23:04 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [100/110]	Time 0.067 (0.107)	Loss 0.1549 (0.2692)	Acc@1 100.000 (96.411)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:23:04 swin_tiny_patch4_window7_224] (main.py 278): INFO  * Acc@1 96.705 Acc@5 100.000
[2025-12-08 16:23:04 swin_tiny_patch4_window7_224] (main.py 165): INFO Accuracy of the network on the 1760 test images: 96.7%
[2025-12-08 16:23:04 swin_tiny_patch4_window7_224] (main.py 167): INFO Max accuracy: 96.70%
[2025-12-08 16:23:05 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][0/256]	eta 0:05:19 lr 0.000021	 wd 0.0500	time 1.2498 (1.2498)	loss 0.9957 (0.9957)	grad_norm 0.0000 (0.0000)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:08 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][10/256]	eta 0:01:18 lr 0.000021	 wd 0.0500	time 0.1782 (0.3200)	loss 0.9628 (0.8816)	grad_norm 7.1425 (10.9805)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:09 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][20/256]	eta 0:00:58 lr 0.000022	 wd 0.0500	time 0.1333 (0.2475)	loss 0.9385 (0.8766)	grad_norm 9.6037 (10.9551)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:11 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][30/256]	eta 0:00:50 lr 0.000022	 wd 0.0500	time 0.3078 (0.2244)	loss 0.9667 (0.8817)	grad_norm 6.8127 (10.9777)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:14 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][40/256]	eta 0:00:50 lr 0.000022	 wd 0.0500	time 0.2312 (0.2319)	loss 1.1136 (0.8907)	grad_norm 9.5558 (10.8312)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:16 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][50/256]	eta 0:00:46 lr 0.000023	 wd 0.0500	time 0.1554 (0.2262)	loss 0.7757 (0.8893)	grad_norm 7.7459 (10.6998)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:17 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][60/256]	eta 0:00:42 lr 0.000023	 wd 0.0500	time 0.1547 (0.2152)	loss 0.8418 (0.8884)	grad_norm 9.2021 (11.0715)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:19 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][70/256]	eta 0:00:38 lr 0.000024	 wd 0.0500	time 0.1591 (0.2067)	loss 0.9457 (0.8882)	grad_norm 12.2236 (10.9613)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:20 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][80/256]	eta 0:00:35 lr 0.000024	 wd 0.0500	time 0.1319 (0.2008)	loss 0.7299 (0.8854)	grad_norm 16.1698 (10.9293)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:22 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][90/256]	eta 0:00:32 lr 0.000024	 wd 0.0500	time 0.1429 (0.1953)	loss 0.6606 (0.8787)	grad_norm 8.3588 (10.6541)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:24 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][100/256]	eta 0:00:29 lr 0.000025	 wd 0.0500	time 0.1332 (0.1914)	loss 0.9613 (0.8798)	grad_norm 10.5426 (10.5905)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:25 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][110/256]	eta 0:00:27 lr 0.000025	 wd 0.0500	time 0.1451 (0.1876)	loss 0.9109 (0.8818)	grad_norm 6.4798 (10.3049)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:27 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][120/256]	eta 0:00:26 lr 0.000026	 wd 0.0500	time 0.3378 (0.1925)	loss 0.8275 (0.8778)	grad_norm 6.1542 (10.1197)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:30 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][130/256]	eta 0:00:24 lr 0.000026	 wd 0.0500	time 0.1710 (0.1957)	loss 0.9310 (0.8721)	grad_norm 8.8601 (9.9562)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:31 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][140/256]	eta 0:00:22 lr 0.000026	 wd 0.0500	time 0.1567 (0.1926)	loss 0.8625 (0.8699)	grad_norm 8.7674 (10.0826)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:33 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][150/256]	eta 0:00:20 lr 0.000027	 wd 0.0500	time 0.1467 (0.1902)	loss 0.7005 (0.8635)	grad_norm 22.5902 (10.1523)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:34 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][160/256]	eta 0:00:18 lr 0.000027	 wd 0.0500	time 0.1327 (0.1882)	loss 0.9053 (0.8632)	grad_norm 8.0826 (9.9724)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:36 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][170/256]	eta 0:00:16 lr 0.000028	 wd 0.0500	time 0.1640 (0.1865)	loss 0.7058 (0.8642)	grad_norm 11.2763 (10.0654)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:38 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][180/256]	eta 0:00:14 lr 0.000028	 wd 0.0500	time 0.1391 (0.1853)	loss 1.0046 (0.8625)	grad_norm 8.2009 (10.0139)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:39 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][190/256]	eta 0:00:12 lr 0.000028	 wd 0.0500	time 0.1497 (0.1838)	loss 0.8600 (0.8575)	grad_norm 14.5035 (10.0326)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:42 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][200/256]	eta 0:00:10 lr 0.000029	 wd 0.0500	time 0.2177 (0.1857)	loss 0.7974 (0.8543)	grad_norm 37.2289 (10.2045)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:44 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][210/256]	eta 0:00:08 lr 0.000029	 wd 0.0500	time 0.1482 (0.1895)	loss 0.7676 (0.8512)	grad_norm 6.4352 (10.1807)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:46 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][220/256]	eta 0:00:06 lr 0.000030	 wd 0.0500	time 0.1435 (0.1879)	loss 0.8288 (0.8510)	grad_norm 6.9031 (10.0615)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:47 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][230/256]	eta 0:00:04 lr 0.000030	 wd 0.0500	time 0.1479 (0.1865)	loss 0.8576 (0.8529)	grad_norm 7.5688 (9.9845)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:49 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][240/256]	eta 0:00:02 lr 0.000031	 wd 0.0500	time 0.1311 (0.1853)	loss 0.6294 (0.8507)	grad_norm 7.8154 (9.9238)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:50 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [2/15][250/256]	eta 0:00:01 lr 0.000031	 wd 0.0500	time 0.1348 (0.1835)	loss 0.8328 (0.8515)	grad_norm 11.5288 (9.9968)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:23:51 swin_tiny_patch4_window7_224] (main.py 231): INFO EPOCH 2 training takes 0:00:46
[2025-12-08 16:23:51 swin_tiny_patch4_window7_224] (utils.py 145): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_2.pth saving......
[2025-12-08 16:23:52 swin_tiny_patch4_window7_224] (utils.py 147): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_2.pth saved !!!
[2025-12-08 16:23:53 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [0/110]	Time 1.213 (1.213)	Loss 0.1100 (0.1100)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:23:54 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [10/110]	Time 0.053 (0.177)	Loss 0.1223 (0.1153)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:23:55 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [20/110]	Time 0.042 (0.137)	Loss 0.1641 (0.1384)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:23:56 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [30/110]	Time 0.052 (0.132)	Loss 0.0850 (0.1348)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:23:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [40/110]	Time 0.112 (0.133)	Loss 0.1409 (0.1407)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:23:59 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [50/110]	Time 0.090 (0.133)	Loss 0.1776 (0.1428)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:23:59 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [60/110]	Time 0.077 (0.124)	Loss 0.1588 (0.1464)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:24:00 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [70/110]	Time 0.040 (0.116)	Loss 0.1434 (0.1419)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:24:01 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [80/110]	Time 0.167 (0.115)	Loss 0.1665 (0.1429)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:24:02 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [90/110]	Time 0.036 (0.109)	Loss 0.1190 (0.1429)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:24:02 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [100/110]	Time 0.073 (0.106)	Loss 0.1306 (0.1442)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:24:03 swin_tiny_patch4_window7_224] (main.py 278): INFO  * Acc@1 100.000 Acc@5 100.000
[2025-12-08 16:24:03 swin_tiny_patch4_window7_224] (main.py 165): INFO Accuracy of the network on the 1760 test images: 100.0%
[2025-12-08 16:24:03 swin_tiny_patch4_window7_224] (main.py 167): INFO Max accuracy: 100.00%
[2025-12-08 16:24:04 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][0/256]	eta 0:06:05 lr 0.000031	 wd 0.0500	time 1.4265 (1.4265)	loss 0.7532 (0.7532)	grad_norm 0.0000 (0.0000)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:07 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][10/256]	eta 0:01:21 lr 0.000031	 wd 0.0500	time 0.1838 (0.3330)	loss 0.6693 (0.8249)	grad_norm 10.6308 (12.0621)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:08 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][20/256]	eta 0:01:00 lr 0.000031	 wd 0.0500	time 0.1455 (0.2560)	loss 0.8135 (0.8230)	grad_norm 11.6380 (10.4814)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:11 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][30/256]	eta 0:00:55 lr 0.000031	 wd 0.0500	time 0.3185 (0.2476)	loss 0.8258 (0.8224)	grad_norm 5.8515 (9.8449)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:13 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][40/256]	eta 0:00:54 lr 0.000031	 wd 0.0500	time 0.2266 (0.2518)	loss 0.5776 (0.7955)	grad_norm 7.3597 (9.3166)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:15 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][50/256]	eta 0:00:48 lr 0.000031	 wd 0.0500	time 0.1375 (0.2345)	loss 0.9185 (0.8033)	grad_norm 5.1148 (9.1216)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:16 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][60/256]	eta 0:00:43 lr 0.000031	 wd 0.0500	time 0.1422 (0.2208)	loss 0.7834 (0.7982)	grad_norm 9.2445 (9.0716)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:18 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][70/256]	eta 0:00:39 lr 0.000031	 wd 0.0500	time 0.1630 (0.2128)	loss 0.9262 (0.8014)	grad_norm 9.0620 (9.0799)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:20 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][80/256]	eta 0:00:36 lr 0.000031	 wd 0.0500	time 0.1451 (0.2067)	loss 0.8117 (0.7975)	grad_norm 5.6181 (8.9801)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:21 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][90/256]	eta 0:00:33 lr 0.000031	 wd 0.0500	time 0.1510 (0.2011)	loss 0.9820 (0.8061)	grad_norm 6.6876 (8.8203)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:23 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][100/256]	eta 0:00:30 lr 0.000031	 wd 0.0500	time 0.1440 (0.1964)	loss 0.7384 (0.8120)	grad_norm 10.1008 (8.9033)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:25 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][110/256]	eta 0:00:28 lr 0.000031	 wd 0.0500	time 0.1363 (0.1973)	loss 0.7072 (0.8115)	grad_norm 7.5310 (8.9451)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:27 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][120/256]	eta 0:00:27 lr 0.000031	 wd 0.0500	time 0.2565 (0.2010)	loss 0.9690 (0.8107)	grad_norm 5.6539 (8.8396)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:29 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][130/256]	eta 0:00:25 lr 0.000031	 wd 0.0500	time 0.1458 (0.1992)	loss 0.6411 (0.8059)	grad_norm 10.2322 (8.7335)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:31 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][140/256]	eta 0:00:22 lr 0.000031	 wd 0.0500	time 0.1397 (0.1960)	loss 0.9288 (0.8087)	grad_norm 4.9849 (8.5720)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:32 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][150/256]	eta 0:00:20 lr 0.000031	 wd 0.0500	time 0.1443 (0.1932)	loss 0.8232 (0.8071)	grad_norm 4.2739 (8.4073)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:34 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][160/256]	eta 0:00:18 lr 0.000031	 wd 0.0500	time 0.1516 (0.1911)	loss 0.9264 (0.8103)	grad_norm 6.4459 (8.2648)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:35 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][170/256]	eta 0:00:16 lr 0.000031	 wd 0.0500	time 0.1461 (0.1890)	loss 0.9381 (0.8090)	grad_norm 5.5298 (8.3078)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:37 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][180/256]	eta 0:00:14 lr 0.000031	 wd 0.0500	time 0.1385 (0.1872)	loss 0.5937 (0.8056)	grad_norm 5.6355 (8.2062)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:39 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][190/256]	eta 0:00:12 lr 0.000031	 wd 0.0500	time 0.2391 (0.1876)	loss 0.6094 (0.8025)	grad_norm 7.8328 (8.1088)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:41 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][200/256]	eta 0:00:10 lr 0.000031	 wd 0.0500	time 0.1888 (0.1912)	loss 0.7243 (0.8045)	grad_norm 8.9012 (8.1219)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:43 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][210/256]	eta 0:00:08 lr 0.000031	 wd 0.0500	time 0.1678 (0.1905)	loss 0.7407 (0.8025)	grad_norm 4.3693 (8.0176)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:45 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][220/256]	eta 0:00:06 lr 0.000031	 wd 0.0500	time 0.1377 (0.1891)	loss 0.9220 (0.8026)	grad_norm 10.4994 (8.0207)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:46 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][230/256]	eta 0:00:04 lr 0.000031	 wd 0.0500	time 0.1470 (0.1875)	loss 0.9671 (0.8030)	grad_norm 6.6340 (8.0927)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:48 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][240/256]	eta 0:00:02 lr 0.000031	 wd 0.0500	time 0.1316 (0.1861)	loss 0.8317 (0.8021)	grad_norm 9.8353 (8.0835)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:49 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [3/15][250/256]	eta 0:00:01 lr 0.000031	 wd 0.0500	time 0.1354 (0.1843)	loss 0.7658 (0.8019)	grad_norm 3.6165 (8.0154)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:24:50 swin_tiny_patch4_window7_224] (main.py 231): INFO EPOCH 3 training takes 0:00:47
[2025-12-08 16:24:50 swin_tiny_patch4_window7_224] (utils.py 145): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_3.pth saving......
[2025-12-08 16:24:53 swin_tiny_patch4_window7_224] (utils.py 147): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_3.pth saved !!!
[2025-12-08 16:24:54 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [0/110]	Time 1.318 (1.318)	Loss 0.0782 (0.0782)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:24:55 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [10/110]	Time 0.095 (0.233)	Loss 0.0575 (0.0677)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:24:56 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [20/110]	Time 0.057 (0.165)	Loss 0.0933 (0.0878)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:24:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [30/110]	Time 0.035 (0.129)	Loss 0.1735 (0.0912)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:24:58 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [40/110]	Time 0.302 (0.120)	Loss 0.1156 (0.0975)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:24:58 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [50/110]	Time 0.032 (0.112)	Loss 0.1313 (0.1019)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:24:59 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [60/110]	Time 0.033 (0.105)	Loss 0.0926 (0.1046)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:25:00 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [70/110]	Time 0.040 (0.099)	Loss 0.1270 (0.1080)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:25:01 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [80/110]	Time 0.126 (0.100)	Loss 0.1504 (0.1118)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:25:02 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [90/110]	Time 0.046 (0.098)	Loss 0.1383 (0.1148)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:25:02 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [100/110]	Time 0.035 (0.094)	Loss 0.1327 (0.1165)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:25:03 swin_tiny_patch4_window7_224] (main.py 278): INFO  * Acc@1 100.000 Acc@5 100.000
[2025-12-08 16:25:03 swin_tiny_patch4_window7_224] (main.py 165): INFO Accuracy of the network on the 1760 test images: 100.0%
[2025-12-08 16:25:03 swin_tiny_patch4_window7_224] (main.py 167): INFO Max accuracy: 100.00%
[2025-12-08 16:25:04 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][0/256]	eta 0:06:00 lr 0.000031	 wd 0.0500	time 1.4090 (1.4090)	loss 0.5955 (0.5955)	grad_norm 0.0000 (0.0000)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:07 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][10/256]	eta 0:01:28 lr 0.000031	 wd 0.0500	time 0.1711 (0.3609)	loss 0.8487 (0.7392)	grad_norm 7.3779 (4.8038)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:09 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][20/256]	eta 0:01:12 lr 0.000031	 wd 0.0500	time 0.2612 (0.3088)	loss 0.7741 (0.7548)	grad_norm 7.8066 (5.6577)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:11 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][30/256]	eta 0:00:59 lr 0.000031	 wd 0.0500	time 0.1398 (0.2625)	loss 0.7522 (0.7751)	grad_norm 5.9684 (6.2799)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:12 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][40/256]	eta 0:00:50 lr 0.000031	 wd 0.0500	time 0.1453 (0.2357)	loss 0.9930 (0.7854)	grad_norm 7.5976 (6.0100)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:14 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][50/256]	eta 0:00:46 lr 0.000031	 wd 0.0500	time 0.1375 (0.2234)	loss 0.7545 (0.7754)	grad_norm 4.3221 (5.9255)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:16 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][60/256]	eta 0:00:41 lr 0.000030	 wd 0.0500	time 0.1373 (0.2113)	loss 0.6599 (0.7792)	grad_norm 6.5189 (5.8544)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:17 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][70/256]	eta 0:00:37 lr 0.000030	 wd 0.0500	time 0.1346 (0.2028)	loss 0.6953 (0.7779)	grad_norm 13.6016 (6.2729)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:19 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][80/256]	eta 0:00:34 lr 0.000030	 wd 0.0500	time 0.1431 (0.1970)	loss 0.8063 (0.7843)	grad_norm 5.3645 (6.4490)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:21 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][90/256]	eta 0:00:32 lr 0.000030	 wd 0.0500	time 0.2651 (0.1958)	loss 1.0345 (0.7861)	grad_norm 6.3408 (6.3224)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:23 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][100/256]	eta 0:00:31 lr 0.000030	 wd 0.0500	time 0.2216 (0.2011)	loss 0.9922 (0.7895)	grad_norm 6.1603 (6.4480)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:25 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][110/256]	eta 0:00:29 lr 0.000030	 wd 0.0500	time 0.1391 (0.1997)	loss 0.7551 (0.7883)	grad_norm 4.4496 (6.3885)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:27 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][120/256]	eta 0:00:26 lr 0.000030	 wd 0.0500	time 0.1409 (0.1957)	loss 0.8082 (0.7895)	grad_norm 12.0345 (6.5633)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:28 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][130/256]	eta 0:00:24 lr 0.000030	 wd 0.0500	time 0.1464 (0.1924)	loss 0.7863 (0.7882)	grad_norm 8.4305 (6.5830)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:30 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][140/256]	eta 0:00:22 lr 0.000030	 wd 0.0500	time 0.1547 (0.1903)	loss 0.8517 (0.7852)	grad_norm 22.7844 (6.7981)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:31 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][150/256]	eta 0:00:19 lr 0.000030	 wd 0.0500	time 0.1531 (0.1885)	loss 0.8475 (0.7907)	grad_norm 7.7142 (6.8177)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:33 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][160/256]	eta 0:00:17 lr 0.000030	 wd 0.0500	time 0.1502 (0.1864)	loss 0.7707 (0.7908)	grad_norm 8.1299 (6.7645)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:35 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][170/256]	eta 0:00:15 lr 0.000030	 wd 0.0500	time 0.1970 (0.1856)	loss 0.8253 (0.7890)	grad_norm 7.3139 (6.7053)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:37 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][180/256]	eta 0:00:14 lr 0.000030	 wd 0.0500	time 0.1984 (0.1893)	loss 0.6710 (0.7895)	grad_norm 8.5386 (6.8088)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:39 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][190/256]	eta 0:00:12 lr 0.000030	 wd 0.0500	time 0.1498 (0.1896)	loss 0.9398 (0.7892)	grad_norm 4.4760 (6.7191)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:41 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][200/256]	eta 0:00:10 lr 0.000030	 wd 0.0500	time 0.1375 (0.1880)	loss 0.8977 (0.7881)	grad_norm 4.9378 (6.8141)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:42 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][210/256]	eta 0:00:08 lr 0.000030	 wd 0.0500	time 0.1554 (0.1864)	loss 0.8788 (0.7869)	grad_norm 9.1093 (6.8153)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:44 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][220/256]	eta 0:00:06 lr 0.000029	 wd 0.0500	time 0.1665 (0.1850)	loss 0.7587 (0.7867)	grad_norm 5.5663 (6.7944)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:45 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][230/256]	eta 0:00:04 lr 0.000029	 wd 0.0500	time 0.1350 (0.1838)	loss 0.7371 (0.7840)	grad_norm 5.5462 (6.7886)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:47 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][240/256]	eta 0:00:02 lr 0.000029	 wd 0.0500	time 0.1326 (0.1826)	loss 1.0451 (0.7864)	grad_norm 9.6837 (6.8639)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:48 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [4/15][250/256]	eta 0:00:01 lr 0.000029	 wd 0.0500	time 0.1389 (0.1809)	loss 0.9244 (0.7867)	grad_norm 6.6220 (6.9064)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:25:49 swin_tiny_patch4_window7_224] (main.py 231): INFO EPOCH 4 training takes 0:00:46
[2025-12-08 16:25:49 swin_tiny_patch4_window7_224] (utils.py 145): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_4.pth saving......
[2025-12-08 16:25:50 swin_tiny_patch4_window7_224] (utils.py 147): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_4.pth saved !!!
[2025-12-08 16:25:52 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [0/110]	Time 1.341 (1.341)	Loss 0.1205 (0.1205)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:25:53 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [10/110]	Time 0.041 (0.221)	Loss 0.1138 (0.0981)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:25:53 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [20/110]	Time 0.074 (0.152)	Loss 0.1219 (0.1117)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:25:54 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [30/110]	Time 0.044 (0.126)	Loss 0.0678 (0.1068)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:25:55 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [40/110]	Time 0.059 (0.118)	Loss 0.1343 (0.1125)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:25:56 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [50/110]	Time 0.072 (0.110)	Loss 0.1265 (0.1140)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:25:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [60/110]	Time 0.039 (0.105)	Loss 0.1176 (0.1157)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:25:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [70/110]	Time 0.029 (0.101)	Loss 0.1387 (0.1197)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:25:58 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [80/110]	Time 0.069 (0.097)	Loss 0.1270 (0.1204)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:25:59 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [90/110]	Time 0.039 (0.098)	Loss 0.1120 (0.1212)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:26:00 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [100/110]	Time 0.037 (0.095)	Loss 0.1093 (0.1201)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:26:00 swin_tiny_patch4_window7_224] (main.py 278): INFO  * Acc@1 100.000 Acc@5 100.000
[2025-12-08 16:26:00 swin_tiny_patch4_window7_224] (main.py 165): INFO Accuracy of the network on the 1760 test images: 100.0%
[2025-12-08 16:26:00 swin_tiny_patch4_window7_224] (main.py 167): INFO Max accuracy: 100.00%
[2025-12-08 16:26:02 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][0/256]	eta 0:06:22 lr 0.000029	 wd 0.0500	time 1.4943 (1.4943)	loss 0.8072 (0.8072)	grad_norm 0.0000 (0.0000)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:05 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][10/256]	eta 0:01:44 lr 0.000029	 wd 0.0500	time 0.2486 (0.4228)	loss 0.8558 (0.8391)	grad_norm 5.4104 (6.8546)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:07 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][20/256]	eta 0:01:19 lr 0.000029	 wd 0.0500	time 0.1492 (0.3361)	loss 0.8321 (0.8115)	grad_norm 4.8415 (6.8705)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:09 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][30/256]	eta 0:01:02 lr 0.000029	 wd 0.0500	time 0.1414 (0.2787)	loss 0.5603 (0.7862)	grad_norm 6.3527 (6.9408)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:11 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][40/256]	eta 0:00:54 lr 0.000029	 wd 0.0500	time 0.1487 (0.2503)	loss 0.5664 (0.7734)	grad_norm 6.2474 (6.5782)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:12 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][50/256]	eta 0:00:47 lr 0.000029	 wd 0.0500	time 0.1378 (0.2318)	loss 0.8546 (0.7836)	grad_norm 5.4276 (6.6570)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:14 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][60/256]	eta 0:00:43 lr 0.000029	 wd 0.0500	time 0.1494 (0.2197)	loss 0.9095 (0.7866)	grad_norm 7.4483 (6.5612)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:16 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][70/256]	eta 0:00:39 lr 0.000029	 wd 0.0500	time 0.1624 (0.2144)	loss 0.7399 (0.7824)	grad_norm 6.1682 (6.4894)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:17 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][80/256]	eta 0:00:36 lr 0.000029	 wd 0.0500	time 0.2189 (0.2099)	loss 0.7997 (0.7862)	grad_norm 4.8407 (6.3563)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:20 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][90/256]	eta 0:00:35 lr 0.000028	 wd 0.0500	time 0.3070 (0.2149)	loss 0.8759 (0.7874)	grad_norm 6.6494 (6.2023)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:22 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][100/256]	eta 0:00:33 lr 0.000028	 wd 0.0500	time 0.1510 (0.2140)	loss 0.8385 (0.7892)	grad_norm 5.8821 (6.2349)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:23 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][110/256]	eta 0:00:30 lr 0.000028	 wd 0.0500	time 0.1621 (0.2090)	loss 0.8430 (0.7888)	grad_norm 6.3362 (6.1277)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:25 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][120/256]	eta 0:00:27 lr 0.000028	 wd 0.0500	time 0.1708 (0.2049)	loss 0.7827 (0.7871)	grad_norm 9.0008 (6.1901)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:27 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][130/256]	eta 0:00:25 lr 0.000028	 wd 0.0500	time 0.1476 (0.2017)	loss 0.7688 (0.7898)	grad_norm 9.4992 (6.3413)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:28 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][140/256]	eta 0:00:23 lr 0.000028	 wd 0.0500	time 0.1451 (0.1988)	loss 0.7386 (0.7891)	grad_norm 5.6367 (6.4103)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:30 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][150/256]	eta 0:00:20 lr 0.000028	 wd 0.0500	time 0.1377 (0.1962)	loss 0.9369 (0.7898)	grad_norm 5.3443 (6.3990)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:32 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][160/256]	eta 0:00:18 lr 0.000028	 wd 0.0500	time 0.2692 (0.1951)	loss 0.6632 (0.7874)	grad_norm 6.1040 (6.4356)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:34 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][170/256]	eta 0:00:17 lr 0.000028	 wd 0.0500	time 0.2423 (0.1988)	loss 0.8765 (0.7866)	grad_norm 2.6118 (6.3690)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:36 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][180/256]	eta 0:00:15 lr 0.000028	 wd 0.0500	time 0.1749 (0.1990)	loss 0.9281 (0.7866)	grad_norm 4.5129 (6.3787)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:38 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][190/256]	eta 0:00:12 lr 0.000027	 wd 0.0500	time 0.1459 (0.1966)	loss 0.6021 (0.7854)	grad_norm 4.7096 (6.4860)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:39 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][200/256]	eta 0:00:10 lr 0.000027	 wd 0.0500	time 0.1376 (0.1948)	loss 0.7932 (0.7829)	grad_norm 10.9294 (6.6325)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:41 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][210/256]	eta 0:00:08 lr 0.000027	 wd 0.0500	time 0.1521 (0.1929)	loss 0.7218 (0.7782)	grad_norm 5.7302 (6.5888)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:43 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][220/256]	eta 0:00:06 lr 0.000027	 wd 0.0500	time 0.1449 (0.1916)	loss 0.6969 (0.7782)	grad_norm 7.5702 (6.6128)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:44 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][230/256]	eta 0:00:04 lr 0.000027	 wd 0.0500	time 0.1584 (0.1902)	loss 0.8657 (0.7788)	grad_norm 19.1401 (6.6512)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:46 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][240/256]	eta 0:00:03 lr 0.000027	 wd 0.0500	time 0.1309 (0.1898)	loss 0.8596 (0.7817)	grad_norm 7.1591 (6.6963)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:47 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [5/15][250/256]	eta 0:00:01 lr 0.000027	 wd 0.0500	time 0.1313 (0.1879)	loss 0.5840 (0.7792)	grad_norm 15.8307 (6.7193)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO EPOCH 5 training takes 0:00:48
[2025-12-08 16:26:48 swin_tiny_patch4_window7_224] (utils.py 145): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_5.pth saving......
[2025-12-08 16:26:49 swin_tiny_patch4_window7_224] (utils.py 147): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_5.pth saved !!!
[2025-12-08 16:26:51 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [0/110]	Time 1.364 (1.364)	Loss 0.0694 (0.0694)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:26:51 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [10/110]	Time 0.098 (0.191)	Loss 0.1600 (0.0750)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:26:52 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [20/110]	Time 0.049 (0.141)	Loss 0.2118 (0.1241)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:26:53 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [30/110]	Time 0.041 (0.120)	Loss 0.1975 (0.1329)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:26:54 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [40/110]	Time 0.114 (0.113)	Loss 0.1370 (0.1338)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:26:55 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [50/110]	Time 0.244 (0.110)	Loss 0.1149 (0.1338)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:26:56 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [60/110]	Time 0.043 (0.104)	Loss 0.1085 (0.1332)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:26:56 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [70/110]	Time 0.067 (0.100)	Loss 0.1190 (0.1333)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:26:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [80/110]	Time 0.136 (0.097)	Loss 0.1208 (0.1317)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:26:58 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [90/110]	Time 0.208 (0.096)	Loss 0.1220 (0.1310)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:26:59 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [100/110]	Time 0.055 (0.095)	Loss 0.1289 (0.1297)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:27:00 swin_tiny_patch4_window7_224] (main.py 278): INFO  * Acc@1 100.000 Acc@5 100.000
[2025-12-08 16:27:00 swin_tiny_patch4_window7_224] (main.py 165): INFO Accuracy of the network on the 1760 test images: 100.0%
[2025-12-08 16:27:00 swin_tiny_patch4_window7_224] (main.py 167): INFO Max accuracy: 100.00%
[2025-12-08 16:27:02 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][0/256]	eta 0:09:37 lr 0.000027	 wd 0.0500	time 2.2564 (2.2564)	loss 0.7617 (0.7617)	grad_norm 0.0000 (0.0000)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:05 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][10/256]	eta 0:01:53 lr 0.000027	 wd 0.0500	time 0.1901 (0.4600)	loss 0.9595 (0.7408)	grad_norm 10.3595 (7.0659)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:07 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][20/256]	eta 0:01:16 lr 0.000027	 wd 0.0500	time 0.1580 (0.3230)	loss 0.7662 (0.7560)	grad_norm 6.1357 (7.5485)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:08 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][30/256]	eta 0:01:01 lr 0.000026	 wd 0.0500	time 0.1580 (0.2710)	loss 0.6708 (0.7343)	grad_norm 3.4190 (6.6012)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:10 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][40/256]	eta 0:00:52 lr 0.000026	 wd 0.0500	time 0.1371 (0.2444)	loss 0.6318 (0.7343)	grad_norm 5.9765 (6.2279)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:11 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][50/256]	eta 0:00:46 lr 0.000026	 wd 0.0500	time 0.1415 (0.2281)	loss 0.7354 (0.7416)	grad_norm 6.8356 (6.1764)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:13 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][60/256]	eta 0:00:42 lr 0.000026	 wd 0.0500	time 0.1750 (0.2171)	loss 0.8069 (0.7546)	grad_norm 3.0412 (6.1787)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:15 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][70/256]	eta 0:00:40 lr 0.000026	 wd 0.0500	time 0.1974 (0.2154)	loss 0.8400 (0.7637)	grad_norm 4.4955 (6.0757)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:18 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][80/256]	eta 0:00:39 lr 0.000026	 wd 0.0500	time 0.2289 (0.2216)	loss 0.7935 (0.7629)	grad_norm 6.2086 (6.4730)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:20 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][90/256]	eta 0:00:36 lr 0.000026	 wd 0.0500	time 0.1444 (0.2188)	loss 0.8321 (0.7665)	grad_norm 8.3337 (6.3899)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:21 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][100/256]	eta 0:00:33 lr 0.000026	 wd 0.0500	time 0.1489 (0.2132)	loss 0.7440 (0.7680)	grad_norm 4.0640 (6.2293)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:23 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][110/256]	eta 0:00:30 lr 0.000025	 wd 0.0500	time 0.1442 (0.2083)	loss 0.6930 (0.7599)	grad_norm 3.4570 (6.1902)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:25 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][120/256]	eta 0:00:27 lr 0.000025	 wd 0.0500	time 0.1420 (0.2052)	loss 0.8780 (0.7560)	grad_norm 2.7747 (6.1225)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:26 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][130/256]	eta 0:00:25 lr 0.000025	 wd 0.0500	time 0.1669 (0.2018)	loss 0.8440 (0.7527)	grad_norm 4.4960 (6.0134)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:28 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][140/256]	eta 0:00:23 lr 0.000025	 wd 0.0500	time 0.1471 (0.1987)	loss 0.6195 (0.7523)	grad_norm 4.8492 (6.0533)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:30 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][150/256]	eta 0:00:21 lr 0.000025	 wd 0.0500	time 0.1980 (0.1992)	loss 0.7937 (0.7543)	grad_norm 5.6069 (5.9888)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:32 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][160/256]	eta 0:00:19 lr 0.000025	 wd 0.0500	time 0.2716 (0.2026)	loss 0.7249 (0.7545)	grad_norm 9.5109 (5.9885)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:34 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][170/256]	eta 0:00:17 lr 0.000025	 wd 0.0500	time 0.1470 (0.2019)	loss 0.6618 (0.7543)	grad_norm 4.2011 (5.9627)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:36 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][180/256]	eta 0:00:15 lr 0.000025	 wd 0.0500	time 0.1486 (0.2000)	loss 0.7647 (0.7532)	grad_norm 9.7489 (5.9024)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:38 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][190/256]	eta 0:00:13 lr 0.000024	 wd 0.0500	time 0.1620 (0.1980)	loss 0.6412 (0.7526)	grad_norm 7.6488 (5.9950)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:39 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][200/256]	eta 0:00:10 lr 0.000024	 wd 0.0500	time 0.1375 (0.1961)	loss 0.7329 (0.7554)	grad_norm 6.9311 (6.0053)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:41 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][210/256]	eta 0:00:08 lr 0.000024	 wd 0.0500	time 0.1498 (0.1945)	loss 0.6696 (0.7562)	grad_norm 10.7045 (6.0953)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:42 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][220/256]	eta 0:00:06 lr 0.000024	 wd 0.0500	time 0.1382 (0.1929)	loss 0.7635 (0.7540)	grad_norm 14.3932 (6.1541)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:44 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][230/256]	eta 0:00:05 lr 0.000024	 wd 0.0500	time 0.1903 (0.1935)	loss 0.6858 (0.7563)	grad_norm 4.0602 (6.1950)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:47 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][240/256]	eta 0:00:03 lr 0.000024	 wd 0.0500	time 0.1931 (0.1957)	loss 0.8232 (0.7571)	grad_norm 8.7195 (6.1682)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:48 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [6/15][250/256]	eta 0:00:01 lr 0.000024	 wd 0.0500	time 0.1318 (0.1937)	loss 0.6421 (0.7558)	grad_norm 4.9146 (6.2626)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:27:49 swin_tiny_patch4_window7_224] (main.py 231): INFO EPOCH 6 training takes 0:00:49
[2025-12-08 16:27:49 swin_tiny_patch4_window7_224] (utils.py 145): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_6.pth saving......
[2025-12-08 16:27:50 swin_tiny_patch4_window7_224] (utils.py 147): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_6.pth saved !!!
[2025-12-08 16:27:51 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [0/110]	Time 1.194 (1.194)	Loss 0.0848 (0.0848)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:27:52 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [10/110]	Time 0.055 (0.188)	Loss 0.1383 (0.0883)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:27:53 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [20/110]	Time 0.081 (0.132)	Loss 0.1487 (0.1129)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:27:53 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [30/110]	Time 0.049 (0.114)	Loss 0.1193 (0.1191)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:27:54 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [40/110]	Time 0.327 (0.110)	Loss 0.1383 (0.1263)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:27:55 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [50/110]	Time 0.030 (0.105)	Loss 0.1300 (0.1275)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:27:56 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [60/110]	Time 0.048 (0.100)	Loss 0.1407 (0.1304)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:27:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [70/110]	Time 0.062 (0.098)	Loss 0.1290 (0.1340)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:27:58 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [80/110]	Time 0.397 (0.099)	Loss 0.1493 (0.1354)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:27:59 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [90/110]	Time 0.082 (0.103)	Loss 0.1517 (0.1369)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:28:00 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [100/110]	Time 0.040 (0.103)	Loss 0.1306 (0.1365)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:28:01 swin_tiny_patch4_window7_224] (main.py 278): INFO  * Acc@1 100.000 Acc@5 100.000
[2025-12-08 16:28:01 swin_tiny_patch4_window7_224] (main.py 165): INFO Accuracy of the network on the 1760 test images: 100.0%
[2025-12-08 16:28:01 swin_tiny_patch4_window7_224] (main.py 167): INFO Max accuracy: 100.00%
[2025-12-08 16:28:03 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][0/256]	eta 0:07:35 lr 0.000024	 wd 0.0500	time 1.7791 (1.7791)	loss 0.8487 (0.8487)	grad_norm 0.0000 (0.0000)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:05 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][10/256]	eta 0:01:33 lr 0.000023	 wd 0.0500	time 0.1403 (0.3811)	loss 0.7636 (0.7600)	grad_norm 3.5785 (4.5434)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:07 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][20/256]	eta 0:01:05 lr 0.000023	 wd 0.0500	time 0.1455 (0.2777)	loss 1.0069 (0.7652)	grad_norm 12.1065 (6.1191)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:09 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][30/256]	eta 0:00:54 lr 0.000023	 wd 0.0500	time 0.1473 (0.2404)	loss 0.8331 (0.7701)	grad_norm 4.6656 (6.2568)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:10 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][40/256]	eta 0:00:48 lr 0.000023	 wd 0.0500	time 0.2202 (0.2228)	loss 0.6657 (0.7712)	grad_norm 4.1285 (6.2178)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:12 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][50/256]	eta 0:00:43 lr 0.000023	 wd 0.0500	time 0.1355 (0.2127)	loss 0.7679 (0.7650)	grad_norm 2.9573 (5.8276)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:15 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][60/256]	eta 0:00:42 lr 0.000023	 wd 0.0500	time 0.3509 (0.2185)	loss 1.0055 (0.7651)	grad_norm 6.3671 (5.9073)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:17 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][70/256]	eta 0:00:41 lr 0.000023	 wd 0.0500	time 0.1583 (0.2219)	loss 0.6806 (0.7633)	grad_norm 5.6505 (6.0005)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:19 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][80/256]	eta 0:00:38 lr 0.000022	 wd 0.0500	time 0.1450 (0.2163)	loss 0.7875 (0.7563)	grad_norm 3.3324 (6.1957)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:20 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][90/256]	eta 0:00:34 lr 0.000022	 wd 0.0500	time 0.1326 (0.2098)	loss 0.5403 (0.7531)	grad_norm 5.0446 (6.1717)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:22 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][100/256]	eta 0:00:31 lr 0.000022	 wd 0.0500	time 0.1454 (0.2046)	loss 0.7948 (0.7510)	grad_norm 2.3271 (6.1004)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:23 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][110/256]	eta 0:00:29 lr 0.000022	 wd 0.0500	time 0.1352 (0.1997)	loss 0.7818 (0.7533)	grad_norm 6.6636 (6.1465)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:25 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][120/256]	eta 0:00:26 lr 0.000022	 wd 0.0500	time 0.1860 (0.1965)	loss 0.7125 (0.7541)	grad_norm 17.9424 (6.1643)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:27 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][130/256]	eta 0:00:24 lr 0.000022	 wd 0.0500	time 0.1508 (0.1937)	loss 0.7365 (0.7493)	grad_norm 4.8000 (6.0603)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:29 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][140/256]	eta 0:00:22 lr 0.000022	 wd 0.0500	time 0.2379 (0.1959)	loss 0.7900 (0.7478)	grad_norm 6.2770 (6.0462)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:31 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][150/256]	eta 0:00:21 lr 0.000021	 wd 0.0500	time 0.1643 (0.1995)	loss 0.6759 (0.7484)	grad_norm 6.2915 (6.1322)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:33 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][160/256]	eta 0:00:18 lr 0.000021	 wd 0.0500	time 0.1455 (0.1971)	loss 0.5766 (0.7449)	grad_norm 6.2666 (6.1488)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:35 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][170/256]	eta 0:00:16 lr 0.000021	 wd 0.0500	time 0.1645 (0.1952)	loss 0.6352 (0.7435)	grad_norm 7.9227 (6.1207)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:36 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][180/256]	eta 0:00:14 lr 0.000021	 wd 0.0500	time 0.1358 (0.1934)	loss 0.8084 (0.7415)	grad_norm 4.6597 (6.0776)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:38 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][190/256]	eta 0:00:12 lr 0.000021	 wd 0.0500	time 0.1564 (0.1917)	loss 0.8753 (0.7397)	grad_norm 4.6567 (6.0243)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:39 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][200/256]	eta 0:00:10 lr 0.000021	 wd 0.0500	time 0.1423 (0.1900)	loss 0.6809 (0.7393)	grad_norm 4.0550 (6.0192)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:41 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][210/256]	eta 0:00:08 lr 0.000021	 wd 0.0500	time 0.1516 (0.1886)	loss 0.7825 (0.7401)	grad_norm 3.7097 (6.0376)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:43 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][220/256]	eta 0:00:06 lr 0.000020	 wd 0.0500	time 0.2735 (0.1906)	loss 0.5838 (0.7436)	grad_norm 15.0082 (6.1353)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:46 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][230/256]	eta 0:00:05 lr 0.000020	 wd 0.0500	time 0.1502 (0.1924)	loss 0.8519 (0.7436)	grad_norm 4.2982 (6.0751)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:47 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][240/256]	eta 0:00:03 lr 0.000020	 wd 0.0500	time 0.1324 (0.1908)	loss 0.8367 (0.7465)	grad_norm 7.2134 (6.0367)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:49 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [7/15][250/256]	eta 0:00:01 lr 0.000020	 wd 0.0500	time 0.1322 (0.1889)	loss 0.6462 (0.7477)	grad_norm 6.6368 (6.0635)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:28:49 swin_tiny_patch4_window7_224] (main.py 231): INFO EPOCH 7 training takes 0:00:48
[2025-12-08 16:28:49 swin_tiny_patch4_window7_224] (utils.py 145): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_7.pth saving......
[2025-12-08 16:28:50 swin_tiny_patch4_window7_224] (utils.py 147): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_7.pth saved !!!
[2025-12-08 16:28:53 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [0/110]	Time 2.673 (2.673)	Loss 0.0989 (0.0989)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:28:54 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [10/110]	Time 0.057 (0.315)	Loss 0.1309 (0.1126)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:28:54 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [20/110]	Time 0.041 (0.201)	Loss 0.1454 (0.1238)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:28:55 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [30/110]	Time 0.035 (0.159)	Loss 0.1251 (0.1248)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:28:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [40/110]	Time 0.417 (0.155)	Loss 0.1141 (0.1230)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:28:58 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [50/110]	Time 0.061 (0.149)	Loss 0.0999 (0.1189)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:28:59 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [60/110]	Time 0.116 (0.143)	Loss 0.1050 (0.1169)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:29:00 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [70/110]	Time 0.056 (0.141)	Loss 0.0947 (0.1164)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:29:01 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [80/110]	Time 0.174 (0.132)	Loss 0.1099 (0.1157)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:29:02 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [90/110]	Time 0.049 (0.126)	Loss 0.1181 (0.1164)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:29:02 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [100/110]	Time 0.038 (0.121)	Loss 0.1033 (0.1155)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:29:03 swin_tiny_patch4_window7_224] (main.py 278): INFO  * Acc@1 100.000 Acc@5 100.000
[2025-12-08 16:29:03 swin_tiny_patch4_window7_224] (main.py 165): INFO Accuracy of the network on the 1760 test images: 100.0%
[2025-12-08 16:29:03 swin_tiny_patch4_window7_224] (main.py 167): INFO Max accuracy: 100.00%
[2025-12-08 16:29:04 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][0/256]	eta 0:05:52 lr 0.000020	 wd 0.0500	time 1.3787 (1.3787)	loss 0.7764 (0.7764)	grad_norm 0.0000 (0.0000)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:07 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][10/256]	eta 0:01:21 lr 0.000020	 wd 0.0500	time 0.1564 (0.3299)	loss 0.7203 (0.7456)	grad_norm 1.4736 (6.0720)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:08 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][20/256]	eta 0:01:00 lr 0.000020	 wd 0.0500	time 0.1573 (0.2578)	loss 0.7631 (0.7301)	grad_norm 1.7623 (5.1005)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:10 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][30/256]	eta 0:00:50 lr 0.000019	 wd 0.0500	time 0.1439 (0.2253)	loss 0.7234 (0.7472)	grad_norm 4.2765 (5.4088)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:12 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][40/256]	eta 0:00:48 lr 0.000019	 wd 0.0500	time 0.1891 (0.2253)	loss 0.7376 (0.7517)	grad_norm 5.3264 (5.3383)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:15 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][50/256]	eta 0:00:47 lr 0.000019	 wd 0.0500	time 0.1880 (0.2294)	loss 0.8232 (0.7533)	grad_norm 4.7656 (5.3576)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:16 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][60/256]	eta 0:00:42 lr 0.000019	 wd 0.0500	time 0.1459 (0.2182)	loss 0.8595 (0.7517)	grad_norm 4.1448 (5.1367)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:18 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][70/256]	eta 0:00:39 lr 0.000019	 wd 0.0500	time 0.2470 (0.2124)	loss 0.6871 (0.7496)	grad_norm 5.0007 (4.9659)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:20 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][80/256]	eta 0:00:36 lr 0.000019	 wd 0.0500	time 0.1528 (0.2068)	loss 0.6579 (0.7462)	grad_norm 2.8252 (5.1664)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:21 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][90/256]	eta 0:00:33 lr 0.000018	 wd 0.0500	time 0.1472 (0.2010)	loss 0.6171 (0.7499)	grad_norm 3.4453 (5.4208)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:23 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][100/256]	eta 0:00:30 lr 0.000018	 wd 0.0500	time 0.1501 (0.1969)	loss 0.5928 (0.7505)	grad_norm 6.2406 (5.4524)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:24 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][110/256]	eta 0:00:28 lr 0.000018	 wd 0.0500	time 0.1384 (0.1931)	loss 0.6098 (0.7532)	grad_norm 8.4194 (5.5832)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:27 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][120/256]	eta 0:00:26 lr 0.000018	 wd 0.0500	time 0.2976 (0.1969)	loss 0.7472 (0.7563)	grad_norm 5.0978 (5.5414)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:29 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][130/256]	eta 0:00:25 lr 0.000018	 wd 0.0500	time 0.1408 (0.2004)	loss 0.7559 (0.7535)	grad_norm 3.3100 (5.4450)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:31 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][140/256]	eta 0:00:22 lr 0.000018	 wd 0.0500	time 0.1426 (0.1973)	loss 0.8077 (0.7515)	grad_norm 3.9229 (5.3025)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:32 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][150/256]	eta 0:00:20 lr 0.000017	 wd 0.0500	time 0.1332 (0.1944)	loss 0.5435 (0.7494)	grad_norm 8.2021 (5.8890)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:34 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][160/256]	eta 0:00:18 lr 0.000017	 wd 0.0500	time 0.1755 (0.1924)	loss 0.8077 (0.7465)	grad_norm 7.0768 (5.7540)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:36 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][170/256]	eta 0:00:16 lr 0.000017	 wd 0.0500	time 0.1554 (0.1900)	loss 0.5834 (0.7456)	grad_norm 2.2758 (5.7020)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:37 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][180/256]	eta 0:00:14 lr 0.000017	 wd 0.0500	time 0.1416 (0.1881)	loss 0.8164 (0.7456)	grad_norm 5.6356 (5.7504)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:39 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][190/256]	eta 0:00:12 lr 0.000017	 wd 0.0500	time 0.1454 (0.1864)	loss 0.7470 (0.7434)	grad_norm 3.2374 (5.7052)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:41 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][200/256]	eta 0:00:10 lr 0.000017	 wd 0.0500	time 0.2699 (0.1891)	loss 0.6261 (0.7431)	grad_norm 5.0912 (5.7388)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:43 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][210/256]	eta 0:00:08 lr 0.000017	 wd 0.0500	time 0.1575 (0.1917)	loss 0.6507 (0.7449)	grad_norm 5.3345 (5.7716)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:45 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][220/256]	eta 0:00:06 lr 0.000016	 wd 0.0500	time 0.1429 (0.1901)	loss 0.6923 (0.7471)	grad_norm 8.6436 (5.7617)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:47 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][230/256]	eta 0:00:04 lr 0.000016	 wd 0.0500	time 0.1497 (0.1888)	loss 0.5748 (0.7466)	grad_norm 6.7108 (5.7195)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:48 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][240/256]	eta 0:00:02 lr 0.000016	 wd 0.0500	time 0.1321 (0.1874)	loss 0.6527 (0.7468)	grad_norm 2.5097 (5.7208)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:50 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [8/15][250/256]	eta 0:00:01 lr 0.000016	 wd 0.0500	time 0.1342 (0.1855)	loss 0.8257 (0.7470)	grad_norm 8.6320 (5.6927)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:29:50 swin_tiny_patch4_window7_224] (main.py 231): INFO EPOCH 8 training takes 0:00:47
[2025-12-08 16:29:50 swin_tiny_patch4_window7_224] (utils.py 145): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_8.pth saving......
[2025-12-08 16:29:56 swin_tiny_patch4_window7_224] (utils.py 147): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_8.pth saved !!!
[2025-12-08 16:29:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [0/110]	Time 0.937 (0.937)	Loss 0.1162 (0.1162)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:29:58 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [10/110]	Time 0.038 (0.151)	Loss 0.1422 (0.1201)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:29:58 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [20/110]	Time 0.050 (0.110)	Loss 0.1617 (0.1319)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:29:59 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [30/110]	Time 0.046 (0.100)	Loss 0.1250 (0.1320)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:30:00 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [40/110]	Time 0.325 (0.101)	Loss 0.1315 (0.1325)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:30:01 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [50/110]	Time 0.069 (0.096)	Loss 0.1117 (0.1300)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:30:02 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [60/110]	Time 0.031 (0.094)	Loss 0.1323 (0.1292)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:30:02 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [70/110]	Time 0.051 (0.089)	Loss 0.1165 (0.1286)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:30:03 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [80/110]	Time 0.201 (0.091)	Loss 0.1168 (0.1271)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:30:04 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [90/110]	Time 0.046 (0.089)	Loss 0.1090 (0.1266)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:30:05 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [100/110]	Time 0.080 (0.088)	Loss 0.1178 (0.1251)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:30:05 swin_tiny_patch4_window7_224] (main.py 278): INFO  * Acc@1 100.000 Acc@5 100.000
[2025-12-08 16:30:05 swin_tiny_patch4_window7_224] (main.py 165): INFO Accuracy of the network on the 1760 test images: 100.0%
[2025-12-08 16:30:05 swin_tiny_patch4_window7_224] (main.py 167): INFO Max accuracy: 100.00%
[2025-12-08 16:30:08 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][0/256]	eta 0:10:14 lr 0.000016	 wd 0.0500	time 2.4023 (2.4023)	loss 0.5762 (0.5762)	grad_norm 0.0000 (0.0000)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:11 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][10/256]	eta 0:01:57 lr 0.000016	 wd 0.0500	time 0.1919 (0.4777)	loss 0.6745 (0.7440)	grad_norm 5.1082 (4.9627)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:12 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][20/256]	eta 0:01:17 lr 0.000015	 wd 0.0500	time 0.1442 (0.3274)	loss 0.8101 (0.7562)	grad_norm 4.0289 (4.6094)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:14 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][30/256]	eta 0:01:01 lr 0.000015	 wd 0.0500	time 0.1348 (0.2714)	loss 0.8535 (0.7732)	grad_norm 6.1541 (5.0966)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:15 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][40/256]	eta 0:00:52 lr 0.000015	 wd 0.0500	time 0.1395 (0.2442)	loss 0.7309 (0.7652)	grad_norm 8.4809 (5.5882)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:17 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][50/256]	eta 0:00:47 lr 0.000015	 wd 0.0500	time 0.1416 (0.2302)	loss 0.6555 (0.7627)	grad_norm 4.3665 (5.3418)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:19 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][60/256]	eta 0:00:43 lr 0.000015	 wd 0.0500	time 0.2740 (0.2209)	loss 0.8080 (0.7546)	grad_norm 3.9977 (5.0427)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:21 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][70/256]	eta 0:00:40 lr 0.000015	 wd 0.0500	time 0.2246 (0.2161)	loss 0.5889 (0.7514)	grad_norm 5.2151 (4.9931)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:23 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][80/256]	eta 0:00:38 lr 0.000015	 wd 0.0500	time 0.2332 (0.2202)	loss 0.9296 (0.7563)	grad_norm 3.2309 (5.2909)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:25 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][90/256]	eta 0:00:35 lr 0.000014	 wd 0.0500	time 0.1443 (0.2167)	loss 0.7133 (0.7497)	grad_norm 5.1198 (5.1529)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:27 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][100/256]	eta 0:00:32 lr 0.000014	 wd 0.0500	time 0.1437 (0.2103)	loss 0.8170 (0.7507)	grad_norm 6.3726 (5.0559)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:28 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][110/256]	eta 0:00:29 lr 0.000014	 wd 0.0500	time 0.1422 (0.2050)	loss 0.8004 (0.7516)	grad_norm 6.7385 (5.2350)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:30 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][120/256]	eta 0:00:27 lr 0.000014	 wd 0.0500	time 0.1459 (0.2014)	loss 0.7156 (0.7490)	grad_norm 9.0274 (5.2500)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:31 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][130/256]	eta 0:00:24 lr 0.000014	 wd 0.0500	time 0.1402 (0.1979)	loss 0.6974 (0.7490)	grad_norm 7.2673 (5.2544)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:33 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][140/256]	eta 0:00:22 lr 0.000014	 wd 0.0500	time 0.1600 (0.1952)	loss 0.8056 (0.7499)	grad_norm 5.7423 (5.1948)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:35 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][150/256]	eta 0:00:20 lr 0.000013	 wd 0.0500	time 0.2565 (0.1939)	loss 0.8201 (0.7544)	grad_norm 8.1730 (5.2680)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:37 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][160/256]	eta 0:00:18 lr 0.000013	 wd 0.0500	time 0.1917 (0.1974)	loss 0.8524 (0.7521)	grad_norm 7.2876 (5.2171)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:39 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][170/256]	eta 0:00:17 lr 0.000013	 wd 0.0500	time 0.1566 (0.1982)	loss 0.8926 (0.7520)	grad_norm 9.5591 (5.2512)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:41 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][180/256]	eta 0:00:14 lr 0.000013	 wd 0.0500	time 0.1352 (0.1956)	loss 0.7035 (0.7521)	grad_norm 7.3813 (5.2225)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:42 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][190/256]	eta 0:00:12 lr 0.000013	 wd 0.0500	time 0.1589 (0.1937)	loss 0.7189 (0.7529)	grad_norm 2.8012 (5.1688)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:44 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][200/256]	eta 0:00:10 lr 0.000013	 wd 0.0500	time 0.1413 (0.1918)	loss 0.9204 (0.7518)	grad_norm 5.0932 (5.1456)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:45 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][210/256]	eta 0:00:08 lr 0.000013	 wd 0.0500	time 0.1375 (0.1902)	loss 0.7927 (0.7516)	grad_norm 4.6737 (5.1696)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:47 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][220/256]	eta 0:00:06 lr 0.000012	 wd 0.0500	time 0.1395 (0.1887)	loss 0.7662 (0.7527)	grad_norm 7.6711 (5.1997)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:49 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][230/256]	eta 0:00:04 lr 0.000012	 wd 0.0500	time 0.1933 (0.1879)	loss 0.8525 (0.7535)	grad_norm 6.3251 (5.2109)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:51 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][240/256]	eta 0:00:03 lr 0.000012	 wd 0.0500	time 0.1703 (0.1901)	loss 0.8049 (0.7537)	grad_norm 6.0166 (5.2624)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:53 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [9/15][250/256]	eta 0:00:01 lr 0.000012	 wd 0.0500	time 0.1313 (0.1882)	loss 0.7020 (0.7535)	grad_norm 7.3012 (5.2568)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO EPOCH 9 training takes 0:00:48
[2025-12-08 16:30:53 swin_tiny_patch4_window7_224] (utils.py 145): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_9.pth saving......
[2025-12-08 16:30:55 swin_tiny_patch4_window7_224] (utils.py 147): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_9.pth saved !!!
[2025-12-08 16:30:56 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [0/110]	Time 1.301 (1.301)	Loss 0.0773 (0.0773)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:30:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [10/110]	Time 0.043 (0.183)	Loss 0.1390 (0.0874)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:30:58 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [20/110]	Time 0.061 (0.132)	Loss 0.1602 (0.1131)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:30:59 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [30/110]	Time 0.037 (0.110)	Loss 0.1323 (0.1191)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:30:59 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [40/110]	Time 0.197 (0.105)	Loss 0.1399 (0.1248)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:31:00 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [50/110]	Time 0.092 (0.100)	Loss 0.1267 (0.1248)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:31:01 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [60/110]	Time 0.068 (0.096)	Loss 0.1373 (0.1270)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:31:02 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [70/110]	Time 0.051 (0.092)	Loss 0.1377 (0.1288)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:31:03 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [80/110]	Time 0.317 (0.095)	Loss 0.1273 (0.1302)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:31:04 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [90/110]	Time 0.126 (0.099)	Loss 0.1263 (0.1309)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:31:05 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [100/110]	Time 0.067 (0.100)	Loss 0.1387 (0.1307)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:31:06 swin_tiny_patch4_window7_224] (main.py 278): INFO  * Acc@1 100.000 Acc@5 100.000
[2025-12-08 16:31:06 swin_tiny_patch4_window7_224] (main.py 165): INFO Accuracy of the network on the 1760 test images: 100.0%
[2025-12-08 16:31:06 swin_tiny_patch4_window7_224] (main.py 167): INFO Max accuracy: 100.00%
[2025-12-08 16:31:08 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][0/256]	eta 0:06:23 lr 0.000012	 wd 0.0500	time 1.4990 (1.4990)	loss 0.7515 (0.7515)	grad_norm 0.0000 (0.0000)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:10 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][10/256]	eta 0:01:21 lr 0.000012	 wd 0.0500	time 0.1897 (0.3327)	loss 0.8015 (0.7828)	grad_norm 5.4163 (5.8290)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:11 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][20/256]	eta 0:01:01 lr 0.000012	 wd 0.0500	time 0.1334 (0.2596)	loss 0.7228 (0.7671)	grad_norm 6.5182 (5.0844)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:13 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][30/256]	eta 0:00:51 lr 0.000011	 wd 0.0500	time 0.1483 (0.2265)	loss 0.9259 (0.7499)	grad_norm 3.3979 (4.5890)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:15 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][40/256]	eta 0:00:45 lr 0.000011	 wd 0.0500	time 0.1513 (0.2099)	loss 0.8740 (0.7581)	grad_norm 6.7045 (4.6706)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:16 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][50/256]	eta 0:00:41 lr 0.000011	 wd 0.0500	time 0.1673 (0.1990)	loss 0.7783 (0.7511)	grad_norm 5.3590 (4.9582)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:18 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][60/256]	eta 0:00:38 lr 0.000011	 wd 0.0500	time 0.2381 (0.1984)	loss 0.7507 (0.7565)	grad_norm 5.6801 (4.9056)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:21 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][70/256]	eta 0:00:38 lr 0.000011	 wd 0.0500	time 0.2151 (0.2061)	loss 0.6891 (0.7534)	grad_norm 9.3247 (5.0042)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:23 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][80/256]	eta 0:00:35 lr 0.000011	 wd 0.0500	time 0.1417 (0.2039)	loss 0.8199 (0.7575)	grad_norm 3.2965 (4.9895)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:24 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][90/256]	eta 0:00:33 lr 0.000010	 wd 0.0500	time 0.1366 (0.2000)	loss 0.5660 (0.7542)	grad_norm 3.2476 (5.0307)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:26 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][100/256]	eta 0:00:30 lr 0.000010	 wd 0.0500	time 0.1414 (0.1963)	loss 0.7403 (0.7514)	grad_norm 4.0274 (4.9224)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:27 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][110/256]	eta 0:00:28 lr 0.000010	 wd 0.0500	time 0.1414 (0.1925)	loss 0.8217 (0.7524)	grad_norm 5.7161 (4.8845)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:29 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][120/256]	eta 0:00:25 lr 0.000010	 wd 0.0500	time 0.1602 (0.1894)	loss 0.8750 (0.7562)	grad_norm 5.8746 (4.9520)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:31 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][130/256]	eta 0:00:23 lr 0.000010	 wd 0.0500	time 0.1612 (0.1871)	loss 0.7177 (0.7538)	grad_norm 6.7667 (4.9799)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:32 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][140/256]	eta 0:00:21 lr 0.000010	 wd 0.0500	time 0.1607 (0.1865)	loss 0.7879 (0.7535)	grad_norm 7.8584 (4.9992)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:35 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][150/256]	eta 0:00:20 lr 0.000010	 wd 0.0500	time 0.2069 (0.1909)	loss 0.6887 (0.7567)	grad_norm 7.2049 (5.2332)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:37 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][160/256]	eta 0:00:18 lr 0.000009	 wd 0.0500	time 0.1625 (0.1915)	loss 0.8826 (0.7574)	grad_norm 11.1722 (5.2889)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:38 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][170/256]	eta 0:00:16 lr 0.000009	 wd 0.0500	time 0.1399 (0.1894)	loss 0.7468 (0.7584)	grad_norm 6.0606 (5.4079)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:40 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][180/256]	eta 0:00:14 lr 0.000009	 wd 0.0500	time 0.1376 (0.1876)	loss 0.6544 (0.7548)	grad_norm 3.3310 (5.3261)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:42 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][190/256]	eta 0:00:12 lr 0.000009	 wd 0.0500	time 0.1447 (0.1858)	loss 0.7922 (0.7545)	grad_norm 3.5781 (5.2750)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:43 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][200/256]	eta 0:00:10 lr 0.000009	 wd 0.0500	time 0.1538 (0.1842)	loss 0.6019 (0.7519)	grad_norm 2.5973 (5.3463)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:45 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][210/256]	eta 0:00:08 lr 0.000009	 wd 0.0500	time 0.1357 (0.1828)	loss 0.8165 (0.7534)	grad_norm 5.4969 (5.3262)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:46 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][220/256]	eta 0:00:06 lr 0.000009	 wd 0.0500	time 0.1609 (0.1821)	loss 0.6423 (0.7522)	grad_norm 3.1063 (5.3213)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:49 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][230/256]	eta 0:00:04 lr 0.000008	 wd 0.0500	time 0.2881 (0.1849)	loss 0.7387 (0.7498)	grad_norm 2.6302 (5.2878)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:51 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][240/256]	eta 0:00:02 lr 0.000008	 wd 0.0500	time 0.1309 (0.1862)	loss 0.7115 (0.7487)	grad_norm 6.2615 (5.2806)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:52 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [10/15][250/256]	eta 0:00:01 lr 0.000008	 wd 0.0500	time 0.1342 (0.1844)	loss 0.8173 (0.7482)	grad_norm 3.2573 (5.2074)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:31:53 swin_tiny_patch4_window7_224] (main.py 231): INFO EPOCH 10 training takes 0:00:47
[2025-12-08 16:31:53 swin_tiny_patch4_window7_224] (utils.py 145): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_10.pth saving......
[2025-12-08 16:31:54 swin_tiny_patch4_window7_224] (utils.py 147): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_10.pth saved !!!
[2025-12-08 16:31:55 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [0/110]	Time 1.053 (1.053)	Loss 0.0751 (0.0751)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:31:56 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [10/110]	Time 0.063 (0.165)	Loss 0.1539 (0.0890)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:31:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [20/110]	Time 0.049 (0.116)	Loss 0.1691 (0.1170)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:31:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [30/110]	Time 0.081 (0.106)	Loss 0.1329 (0.1242)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:31:58 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [40/110]	Time 0.076 (0.098)	Loss 0.1498 (0.1297)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:31:59 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [50/110]	Time 0.096 (0.097)	Loss 0.1160 (0.1287)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:32:00 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [60/110]	Time 0.051 (0.092)	Loss 0.1315 (0.1293)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:32:01 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [70/110]	Time 0.047 (0.090)	Loss 0.1214 (0.1310)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:32:02 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [80/110]	Time 0.163 (0.099)	Loss 0.1135 (0.1288)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:32:03 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [90/110]	Time 0.070 (0.100)	Loss 0.1107 (0.1275)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:32:04 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [100/110]	Time 0.050 (0.101)	Loss 0.1134 (0.1258)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:32:05 swin_tiny_patch4_window7_224] (main.py 278): INFO  * Acc@1 100.000 Acc@5 100.000
[2025-12-08 16:32:05 swin_tiny_patch4_window7_224] (main.py 165): INFO Accuracy of the network on the 1760 test images: 100.0%
[2025-12-08 16:32:05 swin_tiny_patch4_window7_224] (main.py 167): INFO Max accuracy: 100.00%
[2025-12-08 16:32:06 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][0/256]	eta 0:06:06 lr 0.000008	 wd 0.0500	time 1.4308 (1.4308)	loss 0.7186 (0.7186)	grad_norm 0.0000 (0.0000)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:09 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][10/256]	eta 0:01:21 lr 0.000008	 wd 0.0500	time 0.1729 (0.3319)	loss 0.6123 (0.6948)	grad_norm 7.0914 (4.7683)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:10 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][20/256]	eta 0:00:58 lr 0.000008	 wd 0.0500	time 0.1446 (0.2497)	loss 0.6391 (0.7162)	grad_norm 2.5713 (4.5157)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:12 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][30/256]	eta 0:00:49 lr 0.000008	 wd 0.0500	time 0.1465 (0.2192)	loss 0.6712 (0.7146)	grad_norm 5.3826 (4.6127)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:13 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][40/256]	eta 0:00:44 lr 0.000008	 wd 0.0500	time 0.1353 (0.2037)	loss 0.6711 (0.7145)	grad_norm 7.3599 (4.8555)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:15 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][50/256]	eta 0:00:40 lr 0.000007	 wd 0.0500	time 0.2395 (0.1957)	loss 0.6367 (0.7114)	grad_norm 5.4342 (4.7579)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:18 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][60/256]	eta 0:00:41 lr 0.000007	 wd 0.0500	time 0.2404 (0.2097)	loss 0.6705 (0.7162)	grad_norm 2.9193 (4.9543)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:20 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][70/256]	eta 0:00:38 lr 0.000007	 wd 0.0500	time 0.1363 (0.2078)	loss 0.5846 (0.7108)	grad_norm 2.0840 (4.7191)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:21 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][80/256]	eta 0:00:35 lr 0.000007	 wd 0.0500	time 0.1361 (0.2038)	loss 0.6939 (0.7152)	grad_norm 5.6949 (4.7979)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:23 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][90/256]	eta 0:00:32 lr 0.000007	 wd 0.0500	time 0.1346 (0.1987)	loss 0.6893 (0.7155)	grad_norm 1.4981 (4.7033)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:25 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][100/256]	eta 0:00:30 lr 0.000007	 wd 0.0500	time 0.1454 (0.1946)	loss 0.7500 (0.7209)	grad_norm 7.6919 (4.7504)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:26 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][110/256]	eta 0:00:27 lr 0.000007	 wd 0.0500	time 0.1478 (0.1912)	loss 0.7878 (0.7224)	grad_norm 4.2049 (4.9870)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:28 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][120/256]	eta 0:00:25 lr 0.000006	 wd 0.0500	time 0.1464 (0.1884)	loss 0.8855 (0.7310)	grad_norm 5.4299 (5.1126)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:29 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][130/256]	eta 0:00:23 lr 0.000006	 wd 0.0500	time 0.2706 (0.1870)	loss 0.5665 (0.7319)	grad_norm 4.7050 (5.1411)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:32 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][140/256]	eta 0:00:22 lr 0.000006	 wd 0.0500	time 0.2264 (0.1914)	loss 0.7141 (0.7315)	grad_norm 6.4771 (5.5384)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:34 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][150/256]	eta 0:00:20 lr 0.000006	 wd 0.0500	time 0.1375 (0.1918)	loss 0.6936 (0.7321)	grad_norm 3.6917 (5.4418)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:36 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][160/256]	eta 0:00:18 lr 0.000006	 wd 0.0500	time 0.1524 (0.1900)	loss 0.8502 (0.7348)	grad_norm 3.0263 (5.3190)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:37 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][170/256]	eta 0:00:16 lr 0.000006	 wd 0.0500	time 0.1454 (0.1878)	loss 0.8031 (0.7334)	grad_norm 5.7508 (5.2909)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:39 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][180/256]	eta 0:00:14 lr 0.000006	 wd 0.0500	time 0.1383 (0.1861)	loss 0.7955 (0.7354)	grad_norm 6.6560 (5.4133)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:40 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][190/256]	eta 0:00:12 lr 0.000006	 wd 0.0500	time 0.1427 (0.1843)	loss 0.7862 (0.7346)	grad_norm 14.3333 (5.4908)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:42 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][200/256]	eta 0:00:10 lr 0.000006	 wd 0.0500	time 0.1531 (0.1830)	loss 0.6122 (0.7344)	grad_norm 6.6772 (5.4472)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:43 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][210/256]	eta 0:00:08 lr 0.000005	 wd 0.0500	time 0.1978 (0.1819)	loss 0.6687 (0.7333)	grad_norm 3.2771 (5.3968)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:46 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][220/256]	eta 0:00:06 lr 0.000005	 wd 0.0500	time 0.2078 (0.1852)	loss 0.8776 (0.7348)	grad_norm 3.7404 (5.3705)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:48 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][230/256]	eta 0:00:04 lr 0.000005	 wd 0.0500	time 0.1627 (0.1862)	loss 0.6357 (0.7330)	grad_norm 5.7611 (5.3440)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:49 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][240/256]	eta 0:00:02 lr 0.000005	 wd 0.0500	time 0.1323 (0.1848)	loss 0.8577 (0.7349)	grad_norm 4.3855 (5.3000)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:51 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [11/15][250/256]	eta 0:00:01 lr 0.000005	 wd 0.0500	time 0.1329 (0.1830)	loss 0.6906 (0.7332)	grad_norm 2.9793 (5.4286)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:32:52 swin_tiny_patch4_window7_224] (main.py 231): INFO EPOCH 11 training takes 0:00:46
[2025-12-08 16:32:52 swin_tiny_patch4_window7_224] (utils.py 145): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_11.pth saving......
[2025-12-08 16:32:53 swin_tiny_patch4_window7_224] (utils.py 147): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_11.pth saved !!!
[2025-12-08 16:32:54 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [0/110]	Time 0.938 (0.938)	Loss 0.0862 (0.0862)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:32:55 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [10/110]	Time 0.064 (0.166)	Loss 0.1396 (0.1037)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:32:55 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [20/110]	Time 0.093 (0.121)	Loss 0.1486 (0.1165)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:32:56 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [30/110]	Time 0.033 (0.107)	Loss 0.1307 (0.1186)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:32:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [40/110]	Time 0.289 (0.102)	Loss 0.1591 (0.1271)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:32:58 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [50/110]	Time 0.090 (0.099)	Loss 0.1251 (0.1285)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:32:59 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [60/110]	Time 0.041 (0.105)	Loss 0.1453 (0.1310)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:33:00 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [70/110]	Time 0.034 (0.104)	Loss 0.1533 (0.1345)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:33:02 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [80/110]	Time 0.409 (0.111)	Loss 0.1267 (0.1352)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:33:02 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [90/110]	Time 0.048 (0.107)	Loss 0.1128 (0.1345)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:33:03 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [100/110]	Time 0.039 (0.103)	Loss 0.1340 (0.1333)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:33:04 swin_tiny_patch4_window7_224] (main.py 278): INFO  * Acc@1 100.000 Acc@5 100.000
[2025-12-08 16:33:04 swin_tiny_patch4_window7_224] (main.py 165): INFO Accuracy of the network on the 1760 test images: 100.0%
[2025-12-08 16:33:04 swin_tiny_patch4_window7_224] (main.py 167): INFO Max accuracy: 100.00%
[2025-12-08 16:33:05 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][0/256]	eta 0:05:50 lr 0.000005	 wd 0.0500	time 1.3678 (1.3678)	loss 0.7355 (0.7355)	grad_norm 0.0000 (0.0000)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:07 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][10/256]	eta 0:01:21 lr 0.000005	 wd 0.0500	time 0.1785 (0.3308)	loss 0.5577 (0.6993)	grad_norm 4.2685 (5.7733)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:09 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][20/256]	eta 0:00:59 lr 0.000005	 wd 0.0500	time 0.1401 (0.2515)	loss 0.9186 (0.7240)	grad_norm 5.3209 (6.2370)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:11 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][30/256]	eta 0:00:50 lr 0.000005	 wd 0.0500	time 0.1449 (0.2214)	loss 0.7291 (0.7337)	grad_norm 4.7453 (6.0492)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:12 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][40/256]	eta 0:00:45 lr 0.000004	 wd 0.0500	time 0.1752 (0.2102)	loss 0.6252 (0.7296)	grad_norm 4.7850 (5.9225)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:15 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][50/256]	eta 0:00:45 lr 0.000004	 wd 0.0500	time 0.2496 (0.2189)	loss 0.7672 (0.7321)	grad_norm 2.0590 (5.5376)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:17 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][60/256]	eta 0:00:41 lr 0.000004	 wd 0.0500	time 0.1713 (0.2139)	loss 0.7977 (0.7346)	grad_norm 6.3206 (5.6253)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:18 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][70/256]	eta 0:00:38 lr 0.000004	 wd 0.0500	time 0.1387 (0.2051)	loss 0.7355 (0.7247)	grad_norm 7.8269 (5.5101)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:20 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][80/256]	eta 0:00:34 lr 0.000004	 wd 0.0500	time 0.1397 (0.1988)	loss 0.8021 (0.7264)	grad_norm 8.3249 (5.4829)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:22 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][90/256]	eta 0:00:32 lr 0.000004	 wd 0.0500	time 0.2670 (0.1958)	loss 0.8114 (0.7261)	grad_norm 1.7377 (5.5060)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:23 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][100/256]	eta 0:00:30 lr 0.000004	 wd 0.0500	time 0.1563 (0.1928)	loss 0.7847 (0.7268)	grad_norm 5.2415 (5.4845)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:25 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][110/256]	eta 0:00:27 lr 0.000004	 wd 0.0500	time 0.1621 (0.1897)	loss 0.6136 (0.7257)	grad_norm 5.2840 (5.3923)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:27 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][120/256]	eta 0:00:25 lr 0.000004	 wd 0.0500	time 0.1888 (0.1892)	loss 0.9077 (0.7244)	grad_norm 1.2489 (5.2208)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:29 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][130/256]	eta 0:00:24 lr 0.000004	 wd 0.0500	time 0.2212 (0.1937)	loss 0.8115 (0.7281)	grad_norm 5.4075 (5.1794)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:31 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][140/256]	eta 0:00:22 lr 0.000003	 wd 0.0500	time 0.1445 (0.1932)	loss 0.7686 (0.7301)	grad_norm 5.0516 (5.1561)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:33 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][150/256]	eta 0:00:20 lr 0.000003	 wd 0.0500	time 0.1399 (0.1905)	loss 0.7386 (0.7287)	grad_norm 5.6940 (5.1138)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:34 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][160/256]	eta 0:00:18 lr 0.000003	 wd 0.0500	time 0.1400 (0.1885)	loss 0.7027 (0.7310)	grad_norm 4.0545 (5.1433)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:36 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][170/256]	eta 0:00:16 lr 0.000003	 wd 0.0500	time 0.1476 (0.1869)	loss 0.6928 (0.7355)	grad_norm 6.2465 (5.1840)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:37 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][180/256]	eta 0:00:14 lr 0.000003	 wd 0.0500	time 0.1357 (0.1854)	loss 0.5668 (0.7359)	grad_norm 5.6695 (5.2288)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:39 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][190/256]	eta 0:00:12 lr 0.000003	 wd 0.0500	time 0.1440 (0.1838)	loss 0.7258 (0.7354)	grad_norm 8.0476 (5.2093)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:41 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][200/256]	eta 0:00:10 lr 0.000003	 wd 0.0500	time 0.1726 (0.1833)	loss 0.7196 (0.7341)	grad_norm 1.5806 (5.1268)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:43 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][210/256]	eta 0:00:08 lr 0.000003	 wd 0.0500	time 0.2284 (0.1869)	loss 0.8503 (0.7345)	grad_norm 2.5741 (5.0945)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:45 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][220/256]	eta 0:00:06 lr 0.000003	 wd 0.0500	time 0.1428 (0.1869)	loss 0.8694 (0.7352)	grad_norm 10.3727 (5.1626)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:47 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][230/256]	eta 0:00:04 lr 0.000003	 wd 0.0500	time 0.1430 (0.1856)	loss 0.7880 (0.7368)	grad_norm 4.6327 (5.1813)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:48 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][240/256]	eta 0:00:02 lr 0.000003	 wd 0.0500	time 0.1320 (0.1843)	loss 0.7672 (0.7388)	grad_norm 8.7716 (5.3602)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:50 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [12/15][250/256]	eta 0:00:01 lr 0.000002	 wd 0.0500	time 0.1322 (0.1825)	loss 0.5516 (0.7382)	grad_norm 4.2036 (5.3405)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:33:50 swin_tiny_patch4_window7_224] (main.py 231): INFO EPOCH 12 training takes 0:00:46
[2025-12-08 16:33:50 swin_tiny_patch4_window7_224] (utils.py 145): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_12.pth saving......
[2025-12-08 16:33:51 swin_tiny_patch4_window7_224] (utils.py 147): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_12.pth saved !!!
[2025-12-08 16:33:53 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [0/110]	Time 1.199 (1.199)	Loss 0.0799 (0.0799)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:33:53 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [10/110]	Time 0.045 (0.172)	Loss 0.1526 (0.0964)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:33:54 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [20/110]	Time 0.066 (0.128)	Loss 0.1646 (0.1188)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:33:55 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [30/110]	Time 0.129 (0.131)	Loss 0.1429 (0.1241)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:33:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [40/110]	Time 0.260 (0.130)	Loss 0.1664 (0.1328)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:33:58 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [50/110]	Time 0.057 (0.127)	Loss 0.1270 (0.1335)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:33:59 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [60/110]	Time 0.089 (0.123)	Loss 0.1451 (0.1354)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:34:00 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [70/110]	Time 0.062 (0.117)	Loss 0.1473 (0.1383)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:34:00 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [80/110]	Time 0.138 (0.111)	Loss 0.1229 (0.1375)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:34:01 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [90/110]	Time 0.040 (0.107)	Loss 0.1133 (0.1363)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:34:02 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [100/110]	Time 0.047 (0.104)	Loss 0.1260 (0.1345)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:34:02 swin_tiny_patch4_window7_224] (main.py 278): INFO  * Acc@1 100.000 Acc@5 100.000
[2025-12-08 16:34:02 swin_tiny_patch4_window7_224] (main.py 165): INFO Accuracy of the network on the 1760 test images: 100.0%
[2025-12-08 16:34:02 swin_tiny_patch4_window7_224] (main.py 167): INFO Max accuracy: 100.00%
[2025-12-08 16:34:04 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][0/256]	eta 0:06:39 lr 0.000002	 wd 0.0500	time 1.5599 (1.5599)	loss 0.8039 (0.8039)	grad_norm 0.0000 (0.0000)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:06 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][10/256]	eta 0:01:21 lr 0.000002	 wd 0.0500	time 0.1569 (0.3322)	loss 0.7408 (0.7138)	grad_norm 3.7318 (3.2004)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:08 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][20/256]	eta 0:00:59 lr 0.000002	 wd 0.0500	time 0.1825 (0.2530)	loss 0.7315 (0.7109)	grad_norm 5.1478 (3.6737)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:10 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][30/256]	eta 0:00:52 lr 0.000002	 wd 0.0500	time 0.2578 (0.2339)	loss 0.6988 (0.7217)	grad_norm 7.6147 (4.1499)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:12 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][40/256]	eta 0:00:50 lr 0.000002	 wd 0.0500	time 0.1624 (0.2341)	loss 0.8467 (0.7223)	grad_norm 3.7151 (4.2126)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:14 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][50/256]	eta 0:00:46 lr 0.000002	 wd 0.0500	time 0.1639 (0.2244)	loss 0.9351 (0.7367)	grad_norm 6.3748 (4.4933)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:15 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][60/256]	eta 0:00:41 lr 0.000002	 wd 0.0500	time 0.1418 (0.2131)	loss 0.7332 (0.7389)	grad_norm 5.1177 (4.9165)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:17 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][70/256]	eta 0:00:38 lr 0.000002	 wd 0.0500	time 0.1710 (0.2076)	loss 0.5534 (0.7356)	grad_norm 3.2396 (5.1356)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:19 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][80/256]	eta 0:00:35 lr 0.000002	 wd 0.0500	time 0.1648 (0.2017)	loss 0.5700 (0.7376)	grad_norm 3.2301 (5.0283)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:20 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][90/256]	eta 0:00:32 lr 0.000002	 wd 0.0500	time 0.1379 (0.1964)	loss 0.7758 (0.7379)	grad_norm 5.2993 (5.0564)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:22 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][100/256]	eta 0:00:30 lr 0.000002	 wd 0.0500	time 0.1511 (0.1926)	loss 0.6753 (0.7352)	grad_norm 4.5957 (4.8898)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:24 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][110/256]	eta 0:00:28 lr 0.000002	 wd 0.0500	time 0.2352 (0.1945)	loss 0.8728 (0.7343)	grad_norm 1.9761 (4.7965)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:26 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][120/256]	eta 0:00:26 lr 0.000002	 wd 0.0500	time 0.1739 (0.1974)	loss 0.6021 (0.7332)	grad_norm 3.8834 (4.7364)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:28 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][130/256]	eta 0:00:24 lr 0.000001	 wd 0.0500	time 0.1805 (0.1972)	loss 0.5550 (0.7339)	grad_norm 6.1955 (4.8678)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:30 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][140/256]	eta 0:00:22 lr 0.000001	 wd 0.0500	time 0.1355 (0.1940)	loss 0.6844 (0.7308)	grad_norm 8.8701 (4.8550)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:31 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][150/256]	eta 0:00:20 lr 0.000001	 wd 0.0500	time 0.1375 (0.1914)	loss 0.6455 (0.7305)	grad_norm 3.9174 (4.9251)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:33 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][160/256]	eta 0:00:18 lr 0.000001	 wd 0.0500	time 0.1416 (0.1893)	loss 0.7068 (0.7294)	grad_norm 3.7725 (4.8382)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:34 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][170/256]	eta 0:00:16 lr 0.000001	 wd 0.0500	time 0.1538 (0.1875)	loss 0.6239 (0.7293)	grad_norm 3.2325 (4.9598)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:36 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][180/256]	eta 0:00:14 lr 0.000001	 wd 0.0500	time 0.1341 (0.1860)	loss 0.9226 (0.7327)	grad_norm 5.6996 (5.0206)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:38 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][190/256]	eta 0:00:12 lr 0.000001	 wd 0.0500	time 0.2301 (0.1865)	loss 0.7233 (0.7304)	grad_norm 9.4022 (5.0282)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:40 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][200/256]	eta 0:00:10 lr 0.000001	 wd 0.0500	time 0.1518 (0.1891)	loss 0.7418 (0.7308)	grad_norm 3.4207 (5.0596)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:42 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][210/256]	eta 0:00:08 lr 0.000001	 wd 0.0500	time 0.1451 (0.1892)	loss 0.7098 (0.7306)	grad_norm 4.8834 (5.1023)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:44 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][220/256]	eta 0:00:06 lr 0.000001	 wd 0.0500	time 0.1432 (0.1876)	loss 0.6552 (0.7295)	grad_norm 3.2501 (5.1727)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:45 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][230/256]	eta 0:00:04 lr 0.000001	 wd 0.0500	time 0.1478 (0.1863)	loss 0.7899 (0.7301)	grad_norm 3.9542 (5.1077)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:47 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][240/256]	eta 0:00:02 lr 0.000001	 wd 0.0500	time 0.1323 (0.1850)	loss 0.8665 (0.7339)	grad_norm 4.6473 (5.1453)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:48 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [13/15][250/256]	eta 0:00:01 lr 0.000001	 wd 0.0500	time 0.1316 (0.1832)	loss 0.7854 (0.7336)	grad_norm 4.3011 (5.0708)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:34:49 swin_tiny_patch4_window7_224] (main.py 231): INFO EPOCH 13 training takes 0:00:46
[2025-12-08 16:34:49 swin_tiny_patch4_window7_224] (utils.py 145): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_13.pth saving......
[2025-12-08 16:34:50 swin_tiny_patch4_window7_224] (utils.py 147): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_13.pth saved !!!
[2025-12-08 16:34:52 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [0/110]	Time 1.497 (1.497)	Loss 0.0798 (0.0798)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:34:53 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [10/110]	Time 0.066 (0.235)	Loss 0.1544 (0.0951)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:34:54 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [20/110]	Time 0.139 (0.180)	Loss 0.1675 (0.1190)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:34:55 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [30/110]	Time 0.066 (0.159)	Loss 0.1464 (0.1247)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:34:56 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [40/110]	Time 0.250 (0.152)	Loss 0.1602 (0.1325)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:34:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [50/110]	Time 0.034 (0.138)	Loss 0.1195 (0.1316)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:34:58 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [60/110]	Time 0.118 (0.128)	Loss 0.1382 (0.1327)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:34:59 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [70/110]	Time 0.063 (0.121)	Loss 0.1495 (0.1350)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:35:00 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [80/110]	Time 0.254 (0.118)	Loss 0.1245 (0.1348)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:35:00 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [90/110]	Time 0.040 (0.113)	Loss 0.1108 (0.1338)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:35:01 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [100/110]	Time 0.049 (0.110)	Loss 0.1273 (0.1322)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:35:02 swin_tiny_patch4_window7_224] (main.py 278): INFO  * Acc@1 100.000 Acc@5 100.000
[2025-12-08 16:35:02 swin_tiny_patch4_window7_224] (main.py 165): INFO Accuracy of the network on the 1760 test images: 100.0%
[2025-12-08 16:35:02 swin_tiny_patch4_window7_224] (main.py 167): INFO Max accuracy: 100.00%
[2025-12-08 16:35:03 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][0/256]	eta 0:05:57 lr 0.000001	 wd 0.0500	time 1.3958 (1.3958)	loss 0.8243 (0.8243)	grad_norm 0.0000 (0.0000)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:05 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][10/256]	eta 0:01:18 lr 0.000001	 wd 0.0500	time 0.1716 (0.3198)	loss 0.8483 (0.7425)	grad_norm 8.0017 (4.1114)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:07 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][20/256]	eta 0:01:02 lr 0.000001	 wd 0.0500	time 0.2345 (0.2660)	loss 0.9123 (0.7282)	grad_norm 3.5981 (4.1868)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:10 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][30/256]	eta 0:00:59 lr 0.000001	 wd 0.0500	time 0.2753 (0.2621)	loss 0.7928 (0.7333)	grad_norm 4.1669 (4.7620)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:11 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][40/256]	eta 0:00:51 lr 0.000001	 wd 0.0500	time 0.1567 (0.2395)	loss 0.8183 (0.7410)	grad_norm 4.5792 (4.6181)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:13 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][50/256]	eta 0:00:46 lr 0.000001	 wd 0.0500	time 0.1406 (0.2233)	loss 0.7116 (0.7418)	grad_norm 9.8330 (4.8923)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:15 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][60/256]	eta 0:00:41 lr 0.000001	 wd 0.0500	time 0.1353 (0.2121)	loss 0.9191 (0.7467)	grad_norm 5.9061 (5.1477)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:16 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][70/256]	eta 0:00:37 lr 0.000001	 wd 0.0500	time 0.1396 (0.2042)	loss 0.6681 (0.7462)	grad_norm 2.5210 (5.0721)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:18 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][80/256]	eta 0:00:34 lr 0.000001	 wd 0.0500	time 0.1438 (0.1980)	loss 0.7111 (0.7477)	grad_norm 7.4803 (5.2216)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:19 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][90/256]	eta 0:00:32 lr 0.000001	 wd 0.0500	time 0.1587 (0.1934)	loss 0.7028 (0.7498)	grad_norm 4.7857 (5.2217)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:21 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][100/256]	eta 0:00:30 lr 0.000001	 wd 0.0500	time 0.2480 (0.1943)	loss 0.7581 (0.7473)	grad_norm 2.8483 (5.1351)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:24 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][110/256]	eta 0:00:28 lr 0.000000	 wd 0.0500	time 0.2164 (0.1977)	loss 0.8027 (0.7442)	grad_norm 1.7251 (5.0321)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:26 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][120/256]	eta 0:00:27 lr 0.000000	 wd 0.0500	time 0.1909 (0.1988)	loss 0.6143 (0.7408)	grad_norm 4.3255 (4.9681)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:27 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][130/256]	eta 0:00:24 lr 0.000000	 wd 0.0500	time 0.1887 (0.1959)	loss 0.6982 (0.7425)	grad_norm 8.1386 (4.9412)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:29 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][140/256]	eta 0:00:22 lr 0.000000	 wd 0.0500	time 0.1593 (0.1933)	loss 0.7113 (0.7424)	grad_norm 3.1372 (4.8228)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:30 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][150/256]	eta 0:00:20 lr 0.000000	 wd 0.0500	time 0.1617 (0.1907)	loss 0.5528 (0.7427)	grad_norm 5.7226 (4.8027)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:32 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][160/256]	eta 0:00:18 lr 0.000000	 wd 0.0500	time 0.1713 (0.1888)	loss 0.7100 (0.7433)	grad_norm 3.5808 (4.7441)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:34 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][170/256]	eta 0:00:16 lr 0.000000	 wd 0.0500	time 0.1433 (0.1871)	loss 0.7381 (0.7393)	grad_norm 3.9870 (4.7184)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:35 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][180/256]	eta 0:00:14 lr 0.000000	 wd 0.0500	time 0.1926 (0.1864)	loss 0.5971 (0.7404)	grad_norm 6.6491 (4.7961)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:38 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][190/256]	eta 0:00:12 lr 0.000000	 wd 0.0500	time 0.2636 (0.1900)	loss 0.8499 (0.7406)	grad_norm 2.3209 (4.7740)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:40 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][200/256]	eta 0:00:10 lr 0.000000	 wd 0.0500	time 0.1522 (0.1898)	loss 0.7352 (0.7410)	grad_norm 9.6814 (4.8243)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:41 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][210/256]	eta 0:00:08 lr 0.000000	 wd 0.0500	time 0.1423 (0.1886)	loss 0.7086 (0.7417)	grad_norm 5.3860 (4.8673)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:43 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][220/256]	eta 0:00:06 lr 0.000000	 wd 0.0500	time 0.1413 (0.1872)	loss 0.5682 (0.7406)	grad_norm 3.6890 (4.8364)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:45 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][230/256]	eta 0:00:04 lr 0.000000	 wd 0.0500	time 0.1392 (0.1857)	loss 0.6115 (0.7379)	grad_norm 1.3407 (4.7977)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:46 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][240/256]	eta 0:00:02 lr 0.000000	 wd 0.0500	time 0.1322 (0.1843)	loss 0.8041 (0.7393)	grad_norm 6.6840 (4.8148)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:47 swin_tiny_patch4_window7_224] (main.py 222): INFO Train: [14/15][250/256]	eta 0:00:01 lr 0.000000	 wd 0.0500	time 0.1319 (0.1826)	loss 0.8263 (0.7393)	grad_norm 4.3044 (4.8162)	loss_scale 2048.0000 (2048.0000)	mem 921MB
[2025-12-08 16:35:48 swin_tiny_patch4_window7_224] (main.py 231): INFO EPOCH 14 training takes 0:00:46
[2025-12-08 16:35:48 swin_tiny_patch4_window7_224] (utils.py 145): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_14.pth saving......
[2025-12-08 16:35:49 swin_tiny_patch4_window7_224] (utils.py 147): INFO output/swin_tiny_patch4_window7_224/default/ckpt_epoch_14.pth saved !!!
[2025-12-08 16:35:51 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [0/110]	Time 1.809 (1.809)	Loss 0.0818 (0.0818)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:35:52 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [10/110]	Time 0.092 (0.261)	Loss 0.1542 (0.0968)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:35:53 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [20/110]	Time 0.035 (0.188)	Loss 0.1664 (0.1195)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:35:54 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [30/110]	Time 0.055 (0.150)	Loss 0.1472 (0.1249)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:35:55 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [40/110]	Time 0.148 (0.137)	Loss 0.1586 (0.1322)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:35:56 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [50/110]	Time 0.064 (0.132)	Loss 0.1194 (0.1311)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:35:56 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [60/110]	Time 0.034 (0.119)	Loss 0.1368 (0.1322)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:35:57 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [70/110]	Time 0.115 (0.117)	Loss 0.1461 (0.1344)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:35:58 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [80/110]	Time 0.248 (0.112)	Loss 0.1218 (0.1338)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:35:59 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [90/110]	Time 0.079 (0.108)	Loss 0.1109 (0.1328)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:36:00 swin_tiny_patch4_window7_224] (main.py 271): INFO Test: [100/110]	Time 0.037 (0.104)	Loss 0.1252 (0.1311)	Acc@1 100.000 (100.000)	Acc@5 100.000 (100.000)	Mem 921MB
[2025-12-08 16:36:00 swin_tiny_patch4_window7_224] (main.py 278): INFO  * Acc@1 100.000 Acc@5 100.000
[2025-12-08 16:36:00 swin_tiny_patch4_window7_224] (main.py 165): INFO Accuracy of the network on the 1760 test images: 100.0%
[2025-12-08 16:36:00 swin_tiny_patch4_window7_224] (main.py 167): INFO Max accuracy: 100.00%
[2025-12-08 16:36:00 swin_tiny_patch4_window7_224] (main.py 171): INFO Training time 0:17:43