2025-11-11 20:00:12.595 | INFO     | yolox_microbt.core.trainer:before_train:88 - args: Namespace(config='configs.sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb', experiment_name='sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb', name=None, dist_backend='nccl', dist_url=None, batch_size=64, devices=4, exp_file=None, resume=False, ckpt=None, start_epoch=None, num_machines=1, machine_rank=0, fp16=False, cache=None, occupy=False, logger='tensorboard', opts=[])
2025-11-11 20:00:12.600 | INFO     | yolox_microbt.core.trainer:before_train:89 - exp value:
╒═══════════════════╤═══════════════════════════════════════════════════════════════════════════════════╕
│ keys              │ values                                                                            │
╞═══════════════════╪═══════════════════════════════════════════════════════════════════════════════════╡
│ seed              │ None                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ output_dir        │ './YOLOX_outputs'                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ print_interval    │ 20                                                                                │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ eval_interval     │ 1                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ dataset           │ None                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ num_classes       │ 3                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ depth             │ 1.0                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ width             │ 1.0                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ act               │ 'silu'                                                                            │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ data_num_workers  │ 2                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ input_size        │ (416, 416)                                                                        │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ multiscale_range  │ 5                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ data_dir          │ None                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ train_ann         │ 'instances_train2017.json'                                                        │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ val_ann           │ 'instances_val2017.json'                                                          │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ test_ann          │ 'instances_test2017.json'                                                         │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ mosaic_prob       │ 1.0                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ mixup_prob        │ 0.5                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ hsv_prob          │ 0.5                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ flip_prob         │ 0.5                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ degrees           │ 10.0                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ translate         │ 0.1                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ mosaic_scale      │ (0.1, 2)                                                                          │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ enable_mixup      │ True                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ mixup_scale       │ (0.5, 1.5)                                                                        │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ shear             │ 2.0                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ warmup_epochs     │ 0                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ max_epoch         │ 120                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ warmup_lr         │ 0                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ min_lr_ratio      │ 0.05                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ basic_lr_per_img  │ 3.125e-05                                                                         │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ scheduler         │ 'warmcos'                                                                         │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ no_aug_epochs     │ 80                                                                                │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ ema               │ False                                                                             │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ weight_decay      │ 0.0005                                                                            │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ momentum          │ 0.9                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ save_history_ckpt │ True                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ exp_name          │ 'sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb' │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ test_size         │ (416, 416)                                                                        │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ test_conf         │ 0.01                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ nmsthre           │ 0.65                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ qat_warmup_epoch  │ 0                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ qat_clib_epoch    │ 2                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ all_reduce_norm   │ False                                                                             │
╘═══════════════════╧═══════════════════════════════════════════════════════════════════════════════════╛
2025-11-11 20:00:13.589 | INFO     | yolox_microbt.core.trainer:before_train:129 - init prefetcher, this might take one minute or less...
2025-11-11 20:00:16.543 | INFO     | yolox_microbt.core.trainer:before_train:168 - Training start...
2025-11-11 20:00:16.708 | INFO     | yolox_microbt.core.trainer:before_train:169 - 
DistributedDataParallel(
  (module): YOLOXTrainer(
    (yolox): GraphModule(
      (backbone0): Module(
        (backbone): Module(
          (0): Module(
            (0): Module(
              (conv): ConvReLU2d(
                3, 8, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (1): Module(
            (0): Module(
              (conv_dw): ConvReLU2d(
                8, 8, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=8
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pw): Conv2d(
                8, 8, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (2): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                8, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                32, 32, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), groups=32
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                32, 8, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
            (1): Module(
              (conv_pw): ConvReLU2d(
                8, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=32
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                32, 8, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (3): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                8, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                32, 32, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), groups=32
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                32, 10, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
            (1): Module(
              (conv_pw): ConvReLU2d(
                10, 40, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                40, 40, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=40
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                40, 10, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (4): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                10, 40, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                40, 40, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), groups=40
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                40, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
            (1): Module(
              (conv_pw): ConvReLU2d(
                32, 128, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=128
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                128, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
            (2): Module(
              (conv_pw): ConvReLU2d(
                32, 128, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=128
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                128, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (5): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                32, 128, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=128
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                128, 38, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (6): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                38, 152, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                152, 152, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=152
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                152, 42, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (7): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                42, 168, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                168, 168, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), groups=168
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                168, 80, kernel_size=(1, 1), stride=(1, 1), groups=2
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
            (1): Module(
              (conv_pw): ConvReLU2d(
                80, 320, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                320, 320, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=320
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                320, 80, kernel_size=(1, 1), stride=(1, 1), groups=2
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (8): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                80, 320, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                320, 320, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=320
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                320, 128, kernel_size=(1, 1), stride=(1, 1), groups=2
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
        )
      )
      (head0): Module(
        (shared_layer_8): Module(
          (conv0): Module(
            (conv0): ConvReLU2d(
              10, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv1): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=64
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv2): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_8_obj): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 1, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_8_cls): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 3, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_8_box): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 4, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (shared_layer_16): Module(
          (conv0): Module(
            (conv0): ConvReLU2d(
              42, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv1): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=64
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv2): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_16_obj): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 1, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_16_cls): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 3, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_16_box): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 4, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (shared_layer_32): Module(
          (conv0): Module(
            (conv0): ConvReLU2d(
              128, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv1): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=64
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv2): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_32_obj): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 1, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_32_cls): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 3, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_32_box): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 4, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
      )
      (x_post_act_fake_quantizer): FixedFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=0, quant_max=255, dtype=torch.quint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, scale=tensor([1.], device='cuda:0'), zero_point=tensor([0], device='cuda:0', dtype=torch.int32)
        (activation_post_process): PseudoObserver(min_val=0.0, max_val=1.0, pot=False)
      )
      (backbone0_backbone_0_0_conv_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_1_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_1_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_1_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_1_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_1_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_1_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_1_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_1_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_1_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_2_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_1_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_1_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_1_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_3_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_2_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_2_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_2_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_4_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_5_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_5_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_5_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_6_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_6_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_6_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_1_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_1_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_1_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_5_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_8_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_8_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_8_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_8_conv1_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_8_conv2_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_16_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_16_conv1_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_16_conv2_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_8_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_32_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_32_conv1_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_32_conv2_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_8_obj_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_8_cls_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_8_box_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_16_obj_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_16_cls_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_16_box_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_32_obj_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_32_cls_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_32_box_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
    )
    (loss): YOLOXLoss(
      (l1_loss): L1Loss()
      (bcewithlog_loss): BCEWithLogitsLoss()
      (iou_loss): IOUloss()
    )
  )
)
2025-11-11 20:00:16.709 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch1
2025-11-11 20:00:16.727 | INFO     | yolox_microbt.core.trainer:before_epoch:200 - --->No mosaic aug for calibration model!
2025-11-11 20:00:18.100 | ERROR    | yolox.core.trainer:train:79 - Exception in training: 
2025-11-11 20:00:18.101 | INFO     | yolox_microbt.core.trainer:after_train:172 - Training of experiment is done and the best AP is 0.00
2025-11-11 20:00:18.101 | ERROR    | yolox.core.launch:_distributed_worker:147 - An error has been caught in function '_distributed_worker', process 'SpawnProcess-1' (34530), thread 'MainThread' (140443641504704):
Traceback (most recent call last):

  File "<string>", line 1, in <module>
  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/multiprocessing/spawn.py", line 116, in spawn_main
    exitcode = _main(fd, parent_sentinel)
               │     │   └ 5
               │     └ 8
               └ <function _main at 0x7fbb94281e50>
  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/multiprocessing/spawn.py", line 129, in _main
    return self._bootstrap(parent_sentinel)
           │    │          └ 5
           │    └ <function BaseProcess._bootstrap at 0x7fbb943edd30>
           └ <SpawnProcess name='SpawnProcess-1' parent=34462 started>
  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/multiprocessing/process.py", line 315, in _bootstrap
    self.run()
    │    └ <function BaseProcess.run at 0x7fbb943ed3a0>
    └ <SpawnProcess name='SpawnProcess-1' parent=34462 started>
  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/multiprocessing/process.py", line 108, in run
    self._target(*self._args, **self._kwargs)
    │    │        │    │        │    └ {}
    │    │        │    │        └ <SpawnProcess name='SpawnProcess-1' parent=34462 started>
    │    │        │    └ <unprintable tuple object>
    │    │        └ <SpawnProcess name='SpawnProcess-1' parent=34462 started>
    │    └ <function _wrap at 0x7fb9ff5ea0d0>
    └ <SpawnProcess name='SpawnProcess-1' parent=34462 started>
  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/site-packages/torch/multiprocessing/spawn.py", line 59, in _wrap
    fn(i, *args)
    │  │   └ <unprintable tuple object>
    │  └ 0
    └ <function _distributed_worker at 0x7fb9ff1d4f70>

> File "/data1/guofeng/code/vnne_dev_tools/3rdparty/YOLOX/yolox/core/launch.py", line 147, in _distributed_worker
    main_func(*args)
    │          └ <unprintable tuple object>
    └ <function main at 0x7fb9f0b22790>

  File "/data1/guofeng/code/vnne_dev_tools/scripts/model_train.py", line 129, in main
    trainer.train()
    │       └ <function Trainer.train at 0x7fb9f02bab80>
    └ <yolox_microbt.core.trainer.Trainer_custom object at 0x7fb9f01016d0>

  File "/data1/guofeng/code/vnne_dev_tools/3rdparty/YOLOX/yolox/core/trainer.py", line 77, in train
    self.train_in_epoch()
    │    └ <function Trainer.train_in_epoch at 0x7fb9f02c4430>
    └ <yolox_microbt.core.trainer.Trainer_custom object at 0x7fb9f01016d0>

  File "/data1/guofeng/code/vnne_dev_tools/3rdparty/YOLOX/yolox/core/trainer.py", line 87, in train_in_epoch
    self.train_in_iter()
    │    └ <function Trainer.train_in_iter at 0x7fb9f02c44c0>
    └ <yolox_microbt.core.trainer.Trainer_custom object at 0x7fb9f01016d0>

  File "/data1/guofeng/code/vnne_dev_tools/3rdparty/YOLOX/yolox/core/trainer.py", line 93, in train_in_iter
    self.train_one_iter()
    │    └ <function Trainer_custom.train_one_iter at 0x7fba9da58280>
    └ <yolox_microbt.core.trainer.Trainer_custom object at 0x7fb9f01016d0>

  File "/data1/guofeng/code/vnne_dev_tools/yolox_microbt/core/trainer.py", line 58, in train_one_iter
    outputs = self.model(inps, targets)
              │    │     │     └ <unprintable Tensor object>
              │    │     └ <unprintable Tensor object>
              │    └ <unprintable DistributedDataParallel object>
              └ <yolox_microbt.core.trainer.Trainer_custom object at 0x7fb9f01016d0>

  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
           │             │        └ {}
           │             └ <unprintable tuple object>
           └ <unprintable method object>
  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/site-packages/torch/nn/parallel/distributed.py", line 799, in forward
    output = self.module(*inputs[0], **kwargs[0])
             │            │            └ ({},)
             │            └ <unprintable tuple object>
             └ <unprintable DistributedDataParallel object>
  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
           │             │        └ {}
           │             └ <unprintable tuple object>
           └ <unprintable method object>

  File "/data1/guofeng/code/vnne_dev_tools/yolox_microbt/model/yolox_detector.py", line 38, in forward
    loss, iou_loss, conf_loss, cls_loss, l1_loss, num_fg = self.loss(
                                                           └ <unprintable YOLOXTrainer object>

  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
           │             │        └ {}
           │             └ <unprintable tuple object>
           └ <bound method YOLOXLoss.forward of YOLOXLoss(
               (l1_loss): L1Loss()
               (bcewithlog_loss): BCEWithLogitsLoss()
               (iou_loss): IO...

  File "/data1/guofeng/code/vnne_dev_tools/yolox_microbt/model/loss/yolox_loss.py", line 75, in forward
    return self.get_losses(
           │    └ <function YOLOXLoss.get_losses at 0x7fb9f027fdc0>
           └ YOLOXLoss(
               (l1_loss): L1Loss()
               (bcewithlog_loss): BCEWithLogitsLoss()
               (iou_loss): IOUloss()
             )

  File "/data1/guofeng/code/vnne_dev_tools/yolox_microbt/model/loss/yolox_loss.py", line 181, in get_losses
    ) = self.get_assignments(  # noqa
        │    └ <function YOLOXLoss.get_assignments at 0x7fb9f027ff70>
        └ YOLOXLoss(
            (l1_loss): L1Loss()
            (bcewithlog_loss): BCEWithLogitsLoss()
            (iou_loss): IOUloss()
          )

  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/site-packages/torch/autograd/grad_mode.py", line 28, in decorate_context
    return func(*args, **kwargs)
           │     │       └ {}
           │     └ <unprintable tuple object>
           └ <function YOLOXLoss.get_assignments at 0x7fb9f027fee0>

  File "/data1/guofeng/code/vnne_dev_tools/yolox_microbt/model/loss/yolox_loss.py", line 367, in get_assignments
    ) = self.simota_matching(cost, pair_wise_ious, gt_classes, num_gt, fg_mask)
        │    │               │     │               │           │       └ <unprintable Tensor object>
        │    │               │     │               │           └ 1
        │    │               │     │               └ <unprintable Tensor object>
        │    │               │     └ <unprintable Tensor object>
        │    │               └ <unprintable Tensor object>
        │    └ <function YOLOXLoss.simota_matching at 0x7fb9f02800d0>
        └ YOLOXLoss(
            (l1_loss): L1Loss()
            (bcewithlog_loss): BCEWithLogitsLoss()
            (iou_loss): IOUloss()
          )

  File "/data1/guofeng/code/vnne_dev_tools/yolox_microbt/model/loss/yolox_loss.py", line 421, in simota_matching
    _, pos_idx = torch.topk(
    │            │     └ <built-in method topk of type object at 0x7fbb922c6ee0>
    │            └ <module 'torch' from '/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/site-packages/torch/__init__.py'>
    └ <unprintable Tensor object>

RuntimeError: CUDA error: device-side assert triggered
CUDA kernel errors might be asynchronously reported at some other API call,so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
2025-11-11 20:03:46.943 | INFO     | yolox_microbt.core.trainer:before_train:88 - args: Namespace(config='configs.sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb', experiment_name='sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb', name=None, dist_backend='nccl', dist_url=None, batch_size=64, devices=4, exp_file=None, resume=False, ckpt=None, start_epoch=None, num_machines=1, machine_rank=0, fp16=False, cache=None, occupy=False, logger='tensorboard', opts=[])
2025-11-11 20:03:46.948 | INFO     | yolox_microbt.core.trainer:before_train:89 - exp value:
╒═══════════════════╤═══════════════════════════════════════════════════════════════════════════════════╕
│ keys              │ values                                                                            │
╞═══════════════════╪═══════════════════════════════════════════════════════════════════════════════════╡
│ seed              │ None                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ output_dir        │ './YOLOX_outputs'                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ print_interval    │ 20                                                                                │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ eval_interval     │ 1                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ dataset           │ None                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ num_classes       │ 3                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ depth             │ 1.0                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ width             │ 1.0                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ act               │ 'silu'                                                                            │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ data_num_workers  │ 2                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ input_size        │ (416, 416)                                                                        │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ multiscale_range  │ 5                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ data_dir          │ None                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ train_ann         │ 'instances_train2017.json'                                                        │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ val_ann           │ 'instances_val2017.json'                                                          │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ test_ann          │ 'instances_test2017.json'                                                         │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ mosaic_prob       │ 1.0                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ mixup_prob        │ 0.5                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ hsv_prob          │ 0.5                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ flip_prob         │ 0.5                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ degrees           │ 10.0                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ translate         │ 0.1                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ mosaic_scale      │ (0.1, 2)                                                                          │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ enable_mixup      │ True                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ mixup_scale       │ (0.5, 1.5)                                                                        │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ shear             │ 2.0                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ warmup_epochs     │ 0                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ max_epoch         │ 120                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ warmup_lr         │ 0                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ min_lr_ratio      │ 0.05                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ basic_lr_per_img  │ 3.125e-05                                                                         │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ scheduler         │ 'warmcos'                                                                         │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ no_aug_epochs     │ 80                                                                                │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ ema               │ False                                                                             │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ weight_decay      │ 0.0005                                                                            │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ momentum          │ 0.9                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ save_history_ckpt │ True                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ exp_name          │ 'sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb' │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ test_size         │ (416, 416)                                                                        │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ test_conf         │ 0.01                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ nmsthre           │ 0.65                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ qat_warmup_epoch  │ 0                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ qat_clib_epoch    │ 2                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ all_reduce_norm   │ False                                                                             │
╘═══════════════════╧═══════════════════════════════════════════════════════════════════════════════════╛
2025-11-11 20:03:47.942 | INFO     | yolox_microbt.core.trainer:before_train:129 - init prefetcher, this might take one minute or less...
2025-11-11 20:03:51.054 | INFO     | yolox_microbt.core.trainer:before_train:168 - Training start...
2025-11-11 20:03:51.233 | INFO     | yolox_microbt.core.trainer:before_train:169 - 
DistributedDataParallel(
  (module): YOLOXTrainer(
    (yolox): GraphModule(
      (backbone0): Module(
        (backbone): Module(
          (0): Module(
            (0): Module(
              (conv): ConvReLU2d(
                3, 8, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (1): Module(
            (0): Module(
              (conv_dw): ConvReLU2d(
                8, 8, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=8
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pw): Conv2d(
                8, 8, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (2): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                8, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                32, 32, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), groups=32
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                32, 8, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
            (1): Module(
              (conv_pw): ConvReLU2d(
                8, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=32
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                32, 8, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (3): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                8, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                32, 32, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), groups=32
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                32, 10, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
            (1): Module(
              (conv_pw): ConvReLU2d(
                10, 40, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                40, 40, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=40
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                40, 10, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (4): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                10, 40, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                40, 40, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), groups=40
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                40, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
            (1): Module(
              (conv_pw): ConvReLU2d(
                32, 128, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=128
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                128, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
            (2): Module(
              (conv_pw): ConvReLU2d(
                32, 128, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=128
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                128, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (5): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                32, 128, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=128
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                128, 38, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (6): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                38, 152, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                152, 152, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=152
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                152, 42, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (7): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                42, 168, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                168, 168, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), groups=168
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                168, 80, kernel_size=(1, 1), stride=(1, 1), groups=2
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
            (1): Module(
              (conv_pw): ConvReLU2d(
                80, 320, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                320, 320, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=320
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                320, 80, kernel_size=(1, 1), stride=(1, 1), groups=2
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (8): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                80, 320, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                320, 320, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=320
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                320, 128, kernel_size=(1, 1), stride=(1, 1), groups=2
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
        )
      )
      (head0): Module(
        (shared_layer_8): Module(
          (conv0): Module(
            (conv0): ConvReLU2d(
              10, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv1): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=64
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv2): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_8_obj): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 1, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_8_cls): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 3, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_8_box): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 4, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (shared_layer_16): Module(
          (conv0): Module(
            (conv0): ConvReLU2d(
              42, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv1): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=64
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv2): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_16_obj): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 1, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_16_cls): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 3, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_16_box): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 4, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (shared_layer_32): Module(
          (conv0): Module(
            (conv0): ConvReLU2d(
              128, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv1): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=64
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv2): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_32_obj): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 1, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_32_cls): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 3, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_32_box): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 4, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
      )
      (x_post_act_fake_quantizer): FixedFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=0, quant_max=255, dtype=torch.quint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, scale=tensor([1.], device='cuda:0'), zero_point=tensor([0], device='cuda:0', dtype=torch.int32)
        (activation_post_process): PseudoObserver(min_val=0.0, max_val=1.0, pot=False)
      )
      (backbone0_backbone_0_0_conv_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_1_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_1_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_1_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_1_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_1_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_1_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_1_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_1_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_1_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_2_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_1_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_1_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_1_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_3_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_2_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_2_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_2_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_4_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_5_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_5_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_5_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_6_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_6_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_6_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_1_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_1_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_1_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_5_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_8_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_8_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_8_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_8_conv1_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_8_conv2_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_16_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_16_conv1_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_16_conv2_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_8_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_32_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_32_conv1_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_32_conv2_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_8_obj_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_8_cls_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_8_box_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_16_obj_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_16_cls_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_16_box_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_32_obj_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_32_cls_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_32_box_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
    )
    (loss): YOLOXLoss(
      (l1_loss): L1Loss()
      (bcewithlog_loss): BCEWithLogitsLoss()
      (iou_loss): IOUloss()
    )
  )
)
2025-11-11 20:03:51.234 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch1
2025-11-11 20:03:51.260 | INFO     | yolox_microbt.core.trainer:before_epoch:200 - --->No mosaic aug for calibration model!
2025-11-11 20:03:52.530 | ERROR    | yolox.core.trainer:train:79 - Exception in training: 
2025-11-11 20:03:52.530 | INFO     | yolox_microbt.core.trainer:after_train:172 - Training of experiment is done and the best AP is 0.00
2025-11-11 20:03:52.531 | ERROR    | yolox.core.launch:_distributed_worker:147 - An error has been caught in function '_distributed_worker', process 'SpawnProcess-1' (37666), thread 'MainThread' (140044107203520):
Traceback (most recent call last):

  File "<string>", line 1, in <module>
  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/multiprocessing/spawn.py", line 116, in spawn_main
    exitcode = _main(fd, parent_sentinel)
               │     │   └ 5
               │     └ 8
               └ <function _main at 0x7f5e8e0e7e50>
  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/multiprocessing/spawn.py", line 129, in _main
    return self._bootstrap(parent_sentinel)
           │    │          └ 5
           │    └ <function BaseProcess._bootstrap at 0x7f5e8e253d30>
           └ <SpawnProcess name='SpawnProcess-1' parent=37598 started>
  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/multiprocessing/process.py", line 315, in _bootstrap
    self.run()
    │    └ <function BaseProcess.run at 0x7f5e8e2533a0>
    └ <SpawnProcess name='SpawnProcess-1' parent=37598 started>
  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/multiprocessing/process.py", line 108, in run
    self._target(*self._args, **self._kwargs)
    │    │        │    │        │    └ {}
    │    │        │    │        └ <SpawnProcess name='SpawnProcess-1' parent=37598 started>
    │    │        │    └ <unprintable tuple object>
    │    │        └ <SpawnProcess name='SpawnProcess-1' parent=37598 started>
    │    └ <function _wrap at 0x7f5cf94510d0>
    └ <SpawnProcess name='SpawnProcess-1' parent=37598 started>
  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/site-packages/torch/multiprocessing/spawn.py", line 59, in _wrap
    fn(i, *args)
    │  │   └ <unprintable tuple object>
    │  └ 0
    └ <function _distributed_worker at 0x7f5cf9039f70>

> File "/data1/guofeng/code/vnne_dev_tools/3rdparty/YOLOX/yolox/core/launch.py", line 147, in _distributed_worker
    main_func(*args)
    │          └ <unprintable tuple object>
    └ <function main at 0x7f5cea987790>

  File "/data1/guofeng/code/vnne_dev_tools/scripts/model_train.py", line 129, in main
    trainer.train()
    │       └ <function Trainer.train at 0x7f5cea120b80>
    └ <yolox_microbt.core.trainer.Trainer_custom object at 0x7f5ce9f686d0>

  File "/data1/guofeng/code/vnne_dev_tools/3rdparty/YOLOX/yolox/core/trainer.py", line 77, in train
    self.train_in_epoch()
    │    └ <function Trainer.train_in_epoch at 0x7f5cea12a430>
    └ <yolox_microbt.core.trainer.Trainer_custom object at 0x7f5ce9f686d0>

  File "/data1/guofeng/code/vnne_dev_tools/3rdparty/YOLOX/yolox/core/trainer.py", line 87, in train_in_epoch
    self.train_in_iter()
    │    └ <function Trainer.train_in_iter at 0x7f5cea12a4c0>
    └ <yolox_microbt.core.trainer.Trainer_custom object at 0x7f5ce9f686d0>

  File "/data1/guofeng/code/vnne_dev_tools/3rdparty/YOLOX/yolox/core/trainer.py", line 93, in train_in_iter
    self.train_one_iter()
    │    └ <function Trainer_custom.train_one_iter at 0x7f5d978fc280>
    └ <yolox_microbt.core.trainer.Trainer_custom object at 0x7f5ce9f686d0>

  File "/data1/guofeng/code/vnne_dev_tools/yolox_microbt/core/trainer.py", line 58, in train_one_iter
    outputs = self.model(inps, targets)
              │    │     │     └ <unprintable Tensor object>
              │    │     └ <unprintable Tensor object>
              │    └ <unprintable DistributedDataParallel object>
              └ <yolox_microbt.core.trainer.Trainer_custom object at 0x7f5ce9f686d0>

  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
           │             │        └ {}
           │             └ <unprintable tuple object>
           └ <unprintable method object>
  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/site-packages/torch/nn/parallel/distributed.py", line 799, in forward
    output = self.module(*inputs[0], **kwargs[0])
             │            │            └ ({},)
             │            └ <unprintable tuple object>
             └ <unprintable DistributedDataParallel object>
  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
           │             │        └ {}
           │             └ <unprintable tuple object>
           └ <unprintable method object>

  File "/data1/guofeng/code/vnne_dev_tools/yolox_microbt/model/yolox_detector.py", line 38, in forward
    loss, iou_loss, conf_loss, cls_loss, l1_loss, num_fg = self.loss(
                                                           └ <unprintable YOLOXTrainer object>

  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
           │             │        └ {}
           │             └ <unprintable tuple object>
           └ <bound method YOLOXLoss.forward of YOLOXLoss(
               (l1_loss): L1Loss()
               (bcewithlog_loss): BCEWithLogitsLoss()
               (iou_loss): IO...

  File "/data1/guofeng/code/vnne_dev_tools/yolox_microbt/model/loss/yolox_loss.py", line 75, in forward
    return self.get_losses(
           │    └ <function YOLOXLoss.get_losses at 0x7f5cea0e5dc0>
           └ YOLOXLoss(
               (l1_loss): L1Loss()
               (bcewithlog_loss): BCEWithLogitsLoss()
               (iou_loss): IOUloss()
             )

  File "/data1/guofeng/code/vnne_dev_tools/yolox_microbt/model/loss/yolox_loss.py", line 181, in get_losses
    ) = self.get_assignments(  # noqa
        │    └ <function YOLOXLoss.get_assignments at 0x7f5cea0e5f70>
        └ YOLOXLoss(
            (l1_loss): L1Loss()
            (bcewithlog_loss): BCEWithLogitsLoss()
            (iou_loss): IOUloss()
          )

  File "/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/site-packages/torch/autograd/grad_mode.py", line 28, in decorate_context
    return func(*args, **kwargs)
           │     │       └ {}
           │     └ <unprintable tuple object>
           └ <function YOLOXLoss.get_assignments at 0x7f5cea0e5ee0>

  File "/data1/guofeng/code/vnne_dev_tools/yolox_microbt/model/loss/yolox_loss.py", line 367, in get_assignments
    ) = self.simota_matching(cost, pair_wise_ious, gt_classes, num_gt, fg_mask)
        │    │               │     │               │           │       └ <unprintable Tensor object>
        │    │               │     │               │           └ 7
        │    │               │     │               └ <unprintable Tensor object>
        │    │               │     └ <unprintable Tensor object>
        │    │               └ <unprintable Tensor object>
        │    └ <function YOLOXLoss.simota_matching at 0x7f5cea0e60d0>
        └ YOLOXLoss(
            (l1_loss): L1Loss()
            (bcewithlog_loss): BCEWithLogitsLoss()
            (iou_loss): IOUloss()
          )

  File "/data1/guofeng/code/vnne_dev_tools/yolox_microbt/model/loss/yolox_loss.py", line 421, in simota_matching
    _, pos_idx = torch.topk(
    │            │     └ <built-in method topk of type object at 0x7f5e8c12cee0>
    │            └ <module 'torch' from '/home/guofeng/miniconda3/envs/vnne_tools_new/lib/python3.9/site-packages/torch/__init__.py'>
    └ <unprintable Tensor object>

RuntimeError: CUDA error: device-side assert triggered
CUDA kernel errors might be asynchronously reported at some other API call,so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
2025-11-12 14:09:26.596 | INFO     | yolox_microbt.core.trainer:before_train:88 - args: Namespace(config='configs.sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb', experiment_name='sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb', name=None, dist_backend='nccl', dist_url=None, batch_size=64, devices=4, exp_file=None, resume=False, ckpt=None, start_epoch=None, num_machines=1, machine_rank=0, fp16=False, cache=None, occupy=False, logger='tensorboard', opts=[])
2025-11-12 14:09:26.602 | INFO     | yolox_microbt.core.trainer:before_train:89 - exp value:
╒═══════════════════╤═══════════════════════════════════════════════════════════════════════════════════╕
│ keys              │ values                                                                            │
╞═══════════════════╪═══════════════════════════════════════════════════════════════════════════════════╡
│ seed              │ None                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ output_dir        │ './YOLOX_outputs'                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ print_interval    │ 20                                                                                │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ eval_interval     │ 1                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ dataset           │ None                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ num_classes       │ 3                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ depth             │ 1.0                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ width             │ 1.0                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ act               │ 'silu'                                                                            │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ data_num_workers  │ 2                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ input_size        │ (416, 416)                                                                        │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ multiscale_range  │ 5                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ data_dir          │ None                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ train_ann         │ 'instances_train2017.json'                                                        │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ val_ann           │ 'instances_val2017.json'                                                          │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ test_ann          │ 'instances_test2017.json'                                                         │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ mosaic_prob       │ 1.0                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ mixup_prob        │ 0.5                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ hsv_prob          │ 0.5                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ flip_prob         │ 0.5                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ degrees           │ 10.0                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ translate         │ 0.1                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ mosaic_scale      │ (0.1, 2)                                                                          │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ enable_mixup      │ True                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ mixup_scale       │ (0.5, 1.5)                                                                        │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ shear             │ 2.0                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ warmup_epochs     │ 0                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ max_epoch         │ 120                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ warmup_lr         │ 0                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ min_lr_ratio      │ 0.05                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ basic_lr_per_img  │ 3.125e-05                                                                         │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ scheduler         │ 'warmcos'                                                                         │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ no_aug_epochs     │ 80                                                                                │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ ema               │ False                                                                             │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ weight_decay      │ 0.0005                                                                            │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ momentum          │ 0.9                                                                               │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ save_history_ckpt │ True                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ exp_name          │ 'sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb' │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ test_size         │ (416, 416)                                                                        │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ test_conf         │ 0.01                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ nmsthre           │ 0.65                                                                              │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ qat_warmup_epoch  │ 0                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ qat_clib_epoch    │ 2                                                                                 │
├───────────────────┼───────────────────────────────────────────────────────────────────────────────────┤
│ all_reduce_norm   │ False                                                                             │
╘═══════════════════╧═══════════════════════════════════════════════════════════════════════════════════╛
2025-11-12 14:09:27.629 | INFO     | yolox_microbt.core.trainer:before_train:129 - init prefetcher, this might take one minute or less...
2025-11-12 14:09:30.730 | INFO     | yolox_microbt.core.trainer:before_train:168 - Training start...
2025-11-12 14:09:30.867 | INFO     | yolox_microbt.core.trainer:before_train:169 - 
DistributedDataParallel(
  (module): YOLOXTrainer(
    (yolox): GraphModule(
      (backbone0): Module(
        (backbone): Module(
          (0): Module(
            (0): Module(
              (conv): ConvReLU2d(
                3, 8, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (1): Module(
            (0): Module(
              (conv_dw): ConvReLU2d(
                8, 8, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=8
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pw): Conv2d(
                8, 8, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (2): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                8, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                32, 32, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), groups=32
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                32, 8, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
            (1): Module(
              (conv_pw): ConvReLU2d(
                8, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=32
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                32, 8, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (3): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                8, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                32, 32, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), groups=32
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                32, 10, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
            (1): Module(
              (conv_pw): ConvReLU2d(
                10, 40, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                40, 40, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=40
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                40, 10, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (4): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                10, 40, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                40, 40, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), groups=40
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                40, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
            (1): Module(
              (conv_pw): ConvReLU2d(
                32, 128, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=128
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                128, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
            (2): Module(
              (conv_pw): ConvReLU2d(
                32, 128, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=128
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                128, 32, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (5): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                32, 128, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=128
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                128, 38, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (6): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                38, 152, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                152, 152, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=152
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                152, 42, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (7): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                42, 168, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                168, 168, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), groups=168
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                168, 80, kernel_size=(1, 1), stride=(1, 1), groups=2
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
            (1): Module(
              (conv_pw): ConvReLU2d(
                80, 320, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                320, 320, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=320
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                320, 80, kernel_size=(1, 1), stride=(1, 1), groups=2
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
          (8): Module(
            (0): Module(
              (conv_pw): ConvReLU2d(
                80, 320, kernel_size=(1, 1), stride=(1, 1)
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_dw): ConvReLU2d(
                320, 320, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=320
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
              (conv_pwl): Conv2d(
                320, 128, kernel_size=(1, 1), stride=(1, 1), groups=2
                (weight_fake_quant): LearnableFakeQuantize(
                  fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                  tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                  tensor([0.], device='cuda:0')
                  (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
                )
              )
            )
          )
        )
      )
      (head0): Module(
        (shared_layer_8): Module(
          (conv0): Module(
            (conv0): ConvReLU2d(
              10, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv1): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=64
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv2): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_8_obj): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 1, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_8_cls): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 3, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_8_box): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 4, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (shared_layer_16): Module(
          (conv0): Module(
            (conv0): ConvReLU2d(
              42, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv1): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=64
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv2): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_16_obj): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 1, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_16_cls): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 3, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_16_box): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 4, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (shared_layer_32): Module(
          (conv0): Module(
            (conv0): ConvReLU2d(
              128, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv1): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=64
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
          (conv2): Module(
            (conv0): ConvReLU2d(
              64, 64, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_32_obj): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 1, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_32_cls): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 3, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
        (layer_32_box): Module(
          (conv0): Module(
            (conv0): Conv2d(
              64, 4, kernel_size=(1, 1), stride=(1, 1)
              (weight_fake_quant): LearnableFakeQuantize(
                fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
                tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
                tensor([0.], device='cuda:0')
                (activation_post_process): MinMaxObserver(min_val=inf, max_val=-inf, pot=False)
              )
            )
          )
        )
      )
      (x_post_act_fake_quantizer): FixedFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=0, quant_max=255, dtype=torch.quint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, scale=tensor([1.], device='cuda:0'), zero_point=tensor([0], device='cuda:0', dtype=torch.int32)
        (activation_post_process): PseudoObserver(min_val=0.0, max_val=1.0, pot=False)
      )
      (backbone0_backbone_0_0_conv_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_1_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_1_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_1_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_1_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_2_1_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_1_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_1_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_1_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_3_1_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_2_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_1_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_1_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_1_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_3_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_2_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_2_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_4_2_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_4_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_5_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_5_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_5_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_6_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_6_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_6_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_1_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_1_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_7_1_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (add_5_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_8_0_conv_pw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_8_0_conv_dw_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_8_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_8_conv1_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_8_conv2_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_16_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_16_conv1_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_16_conv2_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (backbone0_backbone_8_0_conv_pwl_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_32_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_32_conv1_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_shared_layer_32_conv2_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_8_obj_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_8_cls_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_8_box_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_16_obj_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_16_cls_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_16_box_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_32_obj_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_32_cls_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
      (head0_layer_32_box_conv0_conv0_post_act_fake_quantizer): LearnableFakeQuantize(
        fake_quant_enabled=tensor([0], device='cuda:0', dtype=torch.uint8), observer_enabled=tensor([1], device='cuda:0', dtype=torch.uint8), is_initialized_params=tensor([0], device='cuda:0', dtype=torch.uint8), quant_min=-128, quant_max=127, dtype=torch.qint8, qscheme=torch.per_tensor_symmetric, ch_axis=-1, is_per_channel=False, is_symmetric_quant=True, scale=Parameter containing:
        tensor([1.], device='cuda:0', requires_grad=True), zero_point=Parameter containing:
        tensor([0.], device='cuda:0')
        (activation_post_process): EMAMinMaxObserver(min_val=inf, max_val=-inf, pot=False)
      )
    )
    (loss): YOLOXLoss(
      (l1_loss): L1Loss()
      (bcewithlog_loss): BCEWithLogitsLoss()
      (iou_loss): IOUloss()
    )
  )
)
2025-11-12 14:09:30.868 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch1
2025-11-12 14:09:30.891 | INFO     | yolox_microbt.core.trainer:before_epoch:200 - --->No mosaic aug for calibration model!
2025-11-12 14:09:34.660 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 1/120, iter: 20/129, gpu mem: 1075Mb, mem: 95.9Gb, iter_time: 0.188s, data_time: 0.002s, total_loss: 5.3, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.2, cls_loss: 0.6, lr: 2.000e-03, size: 448, ETA: 0:48:22
2025-11-12 14:09:37.847 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 1/120, iter: 40/129, gpu mem: 2127Mb, mem: 95.9Gb, iter_time: 0.158s, data_time: 0.003s, total_loss: 4.2, iou_loss: 2.2, l1_loss: 0.0, conf_loss: 1.4, cls_loss: 0.6, lr: 2.000e-03, size: 480, ETA: 0:44:30
2025-11-12 14:09:40.460 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 1/120, iter: 60/129, gpu mem: 2127Mb, mem: 96.0Gb, iter_time: 0.130s, data_time: 0.002s, total_loss: 5.0, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 1.8, cls_loss: 0.7, lr: 2.000e-03, size: 256, ETA: 0:40:45
2025-11-12 14:09:42.940 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 1/120, iter: 80/129, gpu mem: 2127Mb, mem: 96.0Gb, iter_time: 0.123s, data_time: 0.002s, total_loss: 4.2, iou_loss: 2.3, l1_loss: 0.0, conf_loss: 1.3, cls_loss: 0.6, lr: 2.000e-03, size: 448, ETA: 0:38:26
2025-11-12 14:09:45.532 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 1/120, iter: 100/129, gpu mem: 2127Mb, mem: 96.0Gb, iter_time: 0.129s, data_time: 0.002s, total_loss: 5.1, iou_loss: 2.1, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 0.7, lr: 2.000e-03, size: 416, ETA: 0:37:19
2025-11-12 14:09:48.427 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 1/120, iter: 120/129, gpu mem: 2127Mb, mem: 95.9Gb, iter_time: 0.144s, data_time: 0.002s, total_loss: 4.4, iou_loss: 2.1, l1_loss: 0.0, conf_loss: 1.6, cls_loss: 0.7, lr: 2.000e-03, size: 544, ETA: 0:37:11
2025-11-12 14:09:49.683 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:09:54.969 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:09:55.925 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:09:56.520 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6500
2025-11-12 14:09:56.649 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5795
2025-11-12 14:09:56.733 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4451
2025-11-12 14:09:56.733 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5582
2025-11-12 14:09:56.734 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:09:56.734 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:09:56.734 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.650
2025-11-12 14:09:56.734 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.580
2025-11-12 14:09:56.734 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.445
2025-11-12 14:09:56.734 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.558
2025-11-12 14:09:56.734 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:09:56.734 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:09:56.734 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:09:56.735 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:09:56.735 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:09:56.735 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:09:56.735 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:09:56.735 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:09:56.735 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:09:57.502 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:09:58.310 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:09:59.073 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:09:59.822 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:10:00.598 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:10:01.336 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:10:02.117 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:10:02.889 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:10:03.636 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:10:03.636 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 14:10:03.636 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 14:10:03.637 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:10:03.644 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.41 ms, Average NMS time: 0.59 ms, Average inference time: 3.01 ms

2025-11-12 14:10:03.645 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:10:03.721 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:10:03.838 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch2
2025-11-12 14:10:03.845 | INFO     | yolox_microbt.core.trainer:before_epoch:200 - --->No mosaic aug for calibration model!
2025-11-12 14:10:06.366 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 2/120, iter: 20/129, gpu mem: 2127Mb, mem: 96.0Gb, iter_time: 0.125s, data_time: 0.003s, total_loss: 3.4, iou_loss: 1.5, l1_loss: 0.0, conf_loss: 1.4, cls_loss: 0.5, lr: 2.000e-03, size: 544, ETA: 0:36:19
2025-11-12 14:10:09.089 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 2/120, iter: 40/129, gpu mem: 2127Mb, mem: 96.0Gb, iter_time: 0.135s, data_time: 0.003s, total_loss: 4.8, iou_loss: 2.3, l1_loss: 0.0, conf_loss: 1.8, cls_loss: 0.6, lr: 1.999e-03, size: 576, ETA: 0:36:03
2025-11-12 14:10:11.643 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 2/120, iter: 60/129, gpu mem: 2127Mb, mem: 96.0Gb, iter_time: 0.127s, data_time: 0.002s, total_loss: 3.5, iou_loss: 1.8, l1_loss: 0.0, conf_loss: 1.0, cls_loss: 0.7, lr: 1.999e-03, size: 352, ETA: 0:35:36
2025-11-12 14:10:14.179 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 2/120, iter: 80/129, gpu mem: 2127Mb, mem: 96.0Gb, iter_time: 0.126s, data_time: 0.002s, total_loss: 3.4, iou_loss: 1.6, l1_loss: 0.0, conf_loss: 1.3, cls_loss: 0.5, lr: 1.999e-03, size: 576, ETA: 0:35:14
2025-11-12 14:10:17.036 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 2/120, iter: 100/129, gpu mem: 2127Mb, mem: 96.0Gb, iter_time: 0.141s, data_time: 0.003s, total_loss: 3.4, iou_loss: 1.8, l1_loss: 0.0, conf_loss: 1.1, cls_loss: 0.5, lr: 1.999e-03, size: 480, ETA: 0:35:15
2025-11-12 14:10:19.506 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 2/120, iter: 120/129, gpu mem: 2127Mb, mem: 96.1Gb, iter_time: 0.122s, data_time: 0.002s, total_loss: 3.0, iou_loss: 1.3, l1_loss: 0.0, conf_loss: 1.2, cls_loss: 0.4, lr: 1.999e-03, size: 448, ETA: 0:34:52
2025-11-12 14:10:20.611 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:10:25.719 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:10:26.579 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:10:27.129 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6503
2025-11-12 14:10:27.258 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5867
2025-11-12 14:10:27.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4463
2025-11-12 14:10:27.333 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5611
2025-11-12 14:10:27.333 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:10:27.333 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:10:27.334 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.650
2025-11-12 14:10:27.334 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.587
2025-11-12 14:10:27.334 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.446
2025-11-12 14:10:27.334 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.561
2025-11-12 14:10:27.334 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:10:27.334 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:10:27.334 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:10:27.334 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:10:27.335 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:10:27.335 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:10:27.335 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:10:27.335 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:10:27.335 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:10:28.033 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:10:28.755 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:10:29.511 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:10:30.221 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:10:30.939 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:10:31.674 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:10:32.375 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:10:33.070 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:10:33.818 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:10:33.818 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 14:10:33.818 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 14:10:33.818 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:10:33.825 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.32 ms, Average NMS time: 0.58 ms, Average inference time: 2.90 ms

2025-11-12 14:10:33.826 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:10:33.904 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:10:33.983 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch3
2025-11-12 14:10:34.034 | INFO     | yolox_microbt.core.trainer:before_epoch:204 - --->enable mosaic aug for quantization training!
2025-11-12 14:10:37.389 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 3/120, iter: 20/129, gpu mem: 2127Mb, mem: 96.1Gb, iter_time: 0.167s, data_time: 0.002s, total_loss: 6.6, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.9, cls_loss: 0.9, lr: 1.998e-03, size: 384, ETA: 0:35:13
2025-11-12 14:10:41.113 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 3/120, iter: 40/129, gpu mem: 2159Mb, mem: 96.1Gb, iter_time: 0.184s, data_time: 0.027s, total_loss: 6.9, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 3.2, cls_loss: 0.9, lr: 1.998e-03, size: 544, ETA: 0:35:56
2025-11-12 14:10:45.008 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 3/120, iter: 60/129, gpu mem: 2160Mb, mem: 96.1Gb, iter_time: 0.194s, data_time: 0.041s, total_loss: 5.7, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.7, lr: 1.998e-03, size: 480, ETA: 0:36:43
2025-11-12 14:10:48.808 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 3/120, iter: 80/129, gpu mem: 2160Mb, mem: 96.1Gb, iter_time: 0.188s, data_time: 0.045s, total_loss: 6.6, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.9, cls_loss: 0.8, lr: 1.998e-03, size: 288, ETA: 0:37:18
2025-11-12 14:10:52.602 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 3/120, iter: 100/129, gpu mem: 2409Mb, mem: 96.1Gb, iter_time: 0.189s, data_time: 0.039s, total_loss: 7.4, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 3.5, cls_loss: 1.0, lr: 1.997e-03, size: 576, ETA: 0:37:50
2025-11-12 14:10:56.138 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 3/120, iter: 120/129, gpu mem: 2409Mb, mem: 96.1Gb, iter_time: 0.174s, data_time: 0.032s, total_loss: 6.0, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.7, lr: 1.997e-03, size: 288, ETA: 0:38:06
2025-11-12 14:10:57.840 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:11:02.816 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:11:04.918 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:11:06.198 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.5890
2025-11-12 14:11:06.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.4987
2025-11-12 14:11:06.535 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3334
2025-11-12 14:11:06.536 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.4737
2025-11-12 14:11:06.536 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:11:06.536 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:11:06.536 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.589
2025-11-12 14:11:06.536 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.499
2025-11-12 14:11:06.536 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.333
2025-11-12 14:11:06.536 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.474
2025-11-12 14:11:06.536 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:11:06.537 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:11:06.537 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:11:06.537 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:11:06.537 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:11:06.537 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:11:06.537 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:11:06.537 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:11:06.537 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:11:08.184 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:11:09.804 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:11:11.416 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:11:13.053 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:11:14.674 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:11:16.315 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:11:17.926 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:11:19.566 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:11:21.144 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:11:21.145 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.26
2025-11-12 14:11:21.145 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.47
2025-11-12 14:11:21.145 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:11:21.169 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.12 ms, Average NMS time: 0.57 ms, Average inference time: 2.69 ms

2025-11-12 14:11:21.171 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:11:21.283 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:11:21.364 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch4
2025-11-12 14:11:24.633 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 4/120, iter: 20/129, gpu mem: 2412Mb, mem: 96.1Gb, iter_time: 0.162s, data_time: 0.002s, total_loss: 6.6, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 3.1, cls_loss: 0.7, lr: 1.997e-03, size: 416, ETA: 0:38:22
2025-11-12 14:11:28.351 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 4/120, iter: 40/129, gpu mem: 2412Mb, mem: 94.9Gb, iter_time: 0.184s, data_time: 0.029s, total_loss: 6.7, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.9, cls_loss: 0.8, lr: 1.996e-03, size: 416, ETA: 0:38:41
2025-11-12 14:11:31.977 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 4/120, iter: 60/129, gpu mem: 2412Mb, mem: 94.9Gb, iter_time: 0.179s, data_time: 0.029s, total_loss: 7.0, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 3.1, cls_loss: 1.0, lr: 1.996e-03, size: 288, ETA: 0:38:55
2025-11-12 14:11:35.630 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 4/120, iter: 80/129, gpu mem: 2412Mb, mem: 94.9Gb, iter_time: 0.180s, data_time: 0.012s, total_loss: 5.8, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.2, cls_loss: 0.8, lr: 1.996e-03, size: 288, ETA: 0:39:08
2025-11-12 14:11:39.336 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 4/120, iter: 100/129, gpu mem: 2412Mb, mem: 94.9Gb, iter_time: 0.184s, data_time: 0.025s, total_loss: 7.8, iou_loss: 3.2, l1_loss: 0.0, conf_loss: 3.5, cls_loss: 1.1, lr: 1.995e-03, size: 288, ETA: 0:39:22
2025-11-12 14:11:42.973 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 4/120, iter: 120/129, gpu mem: 2412Mb, mem: 94.9Gb, iter_time: 0.180s, data_time: 0.020s, total_loss: 6.6, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 3.0, cls_loss: 0.8, lr: 1.995e-03, size: 480, ETA: 0:39:32
2025-11-12 14:11:44.683 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:11:49.739 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:11:52.525 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:11:54.284 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6222
2025-11-12 14:11:54.667 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5591
2025-11-12 14:11:54.777 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3741
2025-11-12 14:11:54.778 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5185
2025-11-12 14:11:54.778 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:11:54.778 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:11:54.778 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.622
2025-11-12 14:11:54.779 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.559
2025-11-12 14:11:54.779 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.374
2025-11-12 14:11:54.779 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.518
2025-11-12 14:11:54.779 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:11:54.779 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:11:54.779 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:11:54.779 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:11:54.779 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:11:54.780 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:11:54.780 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:11:54.780 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:11:54.780 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:11:57.021 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:11:59.232 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:12:01.408 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:12:03.589 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:12:05.742 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:12:07.910 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:12:10.088 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:12:12.252 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:12:14.415 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:12:14.415 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.28
2025-11-12 14:12:14.416 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.52
2025-11-12 14:12:14.416 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:12:14.440 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.21 ms, Average NMS time: 0.61 ms, Average inference time: 2.82 ms

2025-11-12 14:12:14.442 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:12:14.517 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:12:14.598 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch5
2025-11-12 14:12:17.946 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 5/120, iter: 20/129, gpu mem: 2412Mb, mem: 95.0Gb, iter_time: 0.166s, data_time: 0.006s, total_loss: 6.3, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.8, lr: 1.994e-03, size: 416, ETA: 0:39:39
2025-11-12 14:12:21.516 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 5/120, iter: 40/129, gpu mem: 2412Mb, mem: 94.9Gb, iter_time: 0.176s, data_time: 0.002s, total_loss: 7.9, iou_loss: 3.5, l1_loss: 0.0, conf_loss: 3.6, cls_loss: 0.9, lr: 1.994e-03, size: 352, ETA: 0:39:45
2025-11-12 14:12:25.136 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 5/120, iter: 60/129, gpu mem: 2412Mb, mem: 95.0Gb, iter_time: 0.178s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 0.8, lr: 1.993e-03, size: 352, ETA: 0:39:52
2025-11-12 14:12:28.724 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 5/120, iter: 80/129, gpu mem: 2412Mb, mem: 95.0Gb, iter_time: 0.177s, data_time: 0.002s, total_loss: 6.2, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.8, lr: 1.993e-03, size: 448, ETA: 0:39:57
2025-11-12 14:12:32.263 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 5/120, iter: 100/129, gpu mem: 2412Mb, mem: 94.9Gb, iter_time: 0.174s, data_time: 0.002s, total_loss: 6.4, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.8, lr: 1.992e-03, size: 448, ETA: 0:40:00
2025-11-12 14:12:35.865 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 5/120, iter: 120/129, gpu mem: 2412Mb, mem: 96.1Gb, iter_time: 0.177s, data_time: 0.004s, total_loss: 5.9, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.8, lr: 1.992e-03, size: 448, ETA: 0:40:04
2025-11-12 14:12:37.520 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:12:42.633 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:12:44.984 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:12:46.536 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6103
2025-11-12 14:12:46.820 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5628
2025-11-12 14:12:46.934 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3647
2025-11-12 14:12:46.935 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5126
2025-11-12 14:12:46.935 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:12:46.935 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:12:46.935 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.610
2025-11-12 14:12:46.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.563
2025-11-12 14:12:46.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.365
2025-11-12 14:12:46.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.513
2025-11-12 14:12:46.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:12:46.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:12:46.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:12:46.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:12:46.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:12:46.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:12:46.937 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:12:46.937 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:12:46.937 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:12:48.771 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:12:50.711 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:12:52.650 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:12:54.537 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:12:56.444 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:12:58.276 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:13:00.153 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:13:02.056 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:13:03.986 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:13:03.987 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.28
2025-11-12 14:13:03.987 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.51
2025-11-12 14:13:03.987 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:13:04.012 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.18 ms, Average NMS time: 0.57 ms, Average inference time: 2.76 ms

2025-11-12 14:13:04.014 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:13:04.089 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:13:04.171 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch6
2025-11-12 14:13:07.471 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 6/120, iter: 20/129, gpu mem: 2412Mb, mem: 95.6Gb, iter_time: 0.164s, data_time: 0.022s, total_loss: 6.0, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.9, lr: 1.991e-03, size: 416, ETA: 0:40:04
2025-11-12 14:13:11.248 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 6/120, iter: 40/129, gpu mem: 2416Mb, mem: 95.8Gb, iter_time: 0.186s, data_time: 0.018s, total_loss: 6.6, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 3.1, cls_loss: 0.8, lr: 1.990e-03, size: 576, ETA: 0:40:11
2025-11-12 14:13:14.868 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 6/120, iter: 60/129, gpu mem: 2416Mb, mem: 96.7Gb, iter_time: 0.180s, data_time: 0.025s, total_loss: 5.6, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 0.7, lr: 1.990e-03, size: 320, ETA: 0:40:15
2025-11-12 14:13:18.555 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 6/120, iter: 80/129, gpu mem: 2416Mb, mem: 94.9Gb, iter_time: 0.184s, data_time: 0.042s, total_loss: 5.4, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.2, cls_loss: 0.8, lr: 1.989e-03, size: 512, ETA: 0:40:20
2025-11-12 14:13:22.380 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 6/120, iter: 100/129, gpu mem: 2416Mb, mem: 94.9Gb, iter_time: 0.190s, data_time: 0.027s, total_loss: 5.4, iou_loss: 2.3, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 0.8, lr: 1.989e-03, size: 384, ETA: 0:40:27
2025-11-12 14:13:26.837 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 6/120, iter: 120/129, gpu mem: 2416Mb, mem: 94.9Gb, iter_time: 0.222s, data_time: 0.059s, total_loss: 5.8, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.4, cls_loss: 0.9, lr: 1.988e-03, size: 448, ETA: 0:40:46
2025-11-12 14:13:28.369 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:13:33.448 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:13:36.085 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:13:37.935 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6167
2025-11-12 14:13:38.176 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5404
2025-11-12 14:13:38.298 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3707
2025-11-12 14:13:38.299 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5093
2025-11-12 14:13:38.299 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:13:38.299 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:13:38.299 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.617
2025-11-12 14:13:38.299 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.540
2025-11-12 14:13:38.299 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.371
2025-11-12 14:13:38.300 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.509
2025-11-12 14:13:38.300 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:13:38.300 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:13:38.300 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:13:38.300 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:13:38.300 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:13:38.300 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:13:38.300 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:13:38.300 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:13:38.301 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:13:40.486 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:13:42.682 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:13:44.834 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:13:46.993 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:13:49.143 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:13:51.314 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:13:53.464 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:13:55.603 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:13:57.759 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:13:57.760 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.26
2025-11-12 14:13:57.760 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.51
2025-11-12 14:13:57.760 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:13:57.785 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.15 ms, Average NMS time: 0.58 ms, Average inference time: 2.73 ms

2025-11-12 14:13:57.786 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:13:57.861 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:13:57.943 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch7
2025-11-12 14:14:01.354 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 7/120, iter: 20/129, gpu mem: 2416Mb, mem: 95.0Gb, iter_time: 0.167s, data_time: 0.020s, total_loss: 6.5, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 1.0, lr: 1.987e-03, size: 544, ETA: 0:40:42
2025-11-12 14:14:05.026 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 7/120, iter: 40/129, gpu mem: 2416Mb, mem: 95.0Gb, iter_time: 0.182s, data_time: 0.011s, total_loss: 5.8, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.4, cls_loss: 0.8, lr: 1.986e-03, size: 544, ETA: 0:40:44
2025-11-12 14:14:08.489 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 7/120, iter: 60/129, gpu mem: 2416Mb, mem: 95.0Gb, iter_time: 0.171s, data_time: 0.016s, total_loss: 6.4, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.9, cls_loss: 0.9, lr: 1.986e-03, size: 576, ETA: 0:40:43
2025-11-12 14:14:12.142 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 7/120, iter: 80/129, gpu mem: 2416Mb, mem: 95.0Gb, iter_time: 0.180s, data_time: 0.017s, total_loss: 6.0, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.8, lr: 1.985e-03, size: 384, ETA: 0:40:44
2025-11-12 14:14:15.767 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 7/120, iter: 100/129, gpu mem: 2416Mb, mem: 95.0Gb, iter_time: 0.179s, data_time: 0.029s, total_loss: 6.5, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 1.0, lr: 1.984e-03, size: 288, ETA: 0:40:44
2025-11-12 14:14:19.538 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 7/120, iter: 120/129, gpu mem: 2416Mb, mem: 95.0Gb, iter_time: 0.186s, data_time: 0.036s, total_loss: 6.9, iou_loss: 3.0, l1_loss: 0.0, conf_loss: 3.0, cls_loss: 0.8, lr: 1.984e-03, size: 576, ETA: 0:40:47
2025-11-12 14:14:21.145 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:14:26.421 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:14:28.667 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:14:30.123 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.5985
2025-11-12 14:14:30.384 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5333
2025-11-12 14:14:30.497 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3887
2025-11-12 14:14:30.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5068
2025-11-12 14:14:30.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:14:30.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:14:30.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.599
2025-11-12 14:14:30.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.533
2025-11-12 14:14:30.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.389
2025-11-12 14:14:30.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.507
2025-11-12 14:14:30.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:14:30.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:14:30.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:14:30.500 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:14:30.500 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:14:30.500 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:14:30.500 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:14:30.500 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:14:30.500 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:14:32.254 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:14:34.076 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:14:35.867 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:14:37.610 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:14:39.436 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:14:41.224 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:14:42.966 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:14:44.764 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:14:47.036 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:14:47.036 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.27
2025-11-12 14:14:47.036 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.51
2025-11-12 14:14:47.036 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:14:47.061 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.24 ms, Average NMS time: 0.61 ms, Average inference time: 2.85 ms

2025-11-12 14:14:47.062 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:14:47.141 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:14:47.228 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch8
2025-11-12 14:14:50.758 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 8/120, iter: 20/129, gpu mem: 2416Mb, mem: 95.0Gb, iter_time: 0.173s, data_time: 0.001s, total_loss: 6.6, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.9, cls_loss: 0.9, lr: 1.983e-03, size: 448, ETA: 0:40:45
2025-11-12 14:14:54.248 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 8/120, iter: 40/129, gpu mem: 2418Mb, mem: 95.0Gb, iter_time: 0.172s, data_time: 0.003s, total_loss: 6.3, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.8, lr: 1.982e-03, size: 480, ETA: 0:40:43
2025-11-12 14:14:57.754 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 8/120, iter: 60/129, gpu mem: 2418Mb, mem: 95.0Gb, iter_time: 0.172s, data_time: 0.040s, total_loss: 6.1, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.8, lr: 1.981e-03, size: 384, ETA: 0:40:41
2025-11-12 14:15:01.373 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 8/120, iter: 80/129, gpu mem: 2418Mb, mem: 94.9Gb, iter_time: 0.178s, data_time: 0.033s, total_loss: 6.5, iou_loss: 3.1, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.7, lr: 1.980e-03, size: 256, ETA: 0:40:40
2025-11-12 14:15:05.270 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 8/120, iter: 100/129, gpu mem: 2418Mb, mem: 95.0Gb, iter_time: 0.193s, data_time: 0.051s, total_loss: 5.0, iou_loss: 2.1, l1_loss: 0.0, conf_loss: 2.2, cls_loss: 0.7, lr: 1.979e-03, size: 544, ETA: 0:40:44
2025-11-12 14:15:09.124 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 8/120, iter: 120/129, gpu mem: 2418Mb, mem: 95.0Gb, iter_time: 0.192s, data_time: 0.040s, total_loss: 5.6, iou_loss: 2.4, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 0.8, lr: 1.979e-03, size: 416, ETA: 0:40:47
2025-11-12 14:15:10.685 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:15:16.119 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:15:18.492 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:15:19.997 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6153
2025-11-12 14:15:20.276 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5777
2025-11-12 14:15:20.341 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3870
2025-11-12 14:15:20.342 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5267
2025-11-12 14:15:20.342 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:15:20.343 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:15:20.343 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.615
2025-11-12 14:15:20.343 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.578
2025-11-12 14:15:20.343 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.387
2025-11-12 14:15:20.343 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.527
2025-11-12 14:15:20.343 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:15:20.343 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:15:20.343 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:15:20.344 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:15:20.344 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:15:20.344 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:15:20.344 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:15:20.344 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:15:20.344 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:15:22.278 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:15:24.152 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:15:25.980 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:15:27.921 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:15:29.821 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:15:31.701 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:15:33.518 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:15:35.388 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:15:37.280 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:15:37.281 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.28
2025-11-12 14:15:37.281 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.53
2025-11-12 14:15:37.281 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:15:37.306 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.27 ms, Average NMS time: 0.64 ms, Average inference time: 2.91 ms

2025-11-12 14:15:37.307 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:15:37.383 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:15:37.465 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch9
2025-11-12 14:15:40.724 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 9/120, iter: 20/129, gpu mem: 2418Mb, mem: 95.0Gb, iter_time: 0.161s, data_time: 0.011s, total_loss: 6.2, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.9, lr: 1.977e-03, size: 256, ETA: 0:40:40
2025-11-12 14:15:44.511 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 9/120, iter: 40/129, gpu mem: 2418Mb, mem: 95.0Gb, iter_time: 0.188s, data_time: 0.034s, total_loss: 6.7, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 1.0, lr: 1.976e-03, size: 480, ETA: 0:40:42
2025-11-12 14:15:48.095 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 9/120, iter: 60/129, gpu mem: 2418Mb, mem: 95.0Gb, iter_time: 0.176s, data_time: 0.014s, total_loss: 6.3, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.9, cls_loss: 0.7, lr: 1.976e-03, size: 416, ETA: 0:40:40
2025-11-12 14:15:51.763 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 9/120, iter: 80/129, gpu mem: 2418Mb, mem: 95.0Gb, iter_time: 0.179s, data_time: 0.024s, total_loss: 5.4, iou_loss: 2.4, l1_loss: 0.0, conf_loss: 2.2, cls_loss: 0.7, lr: 1.975e-03, size: 256, ETA: 0:40:39
2025-11-12 14:15:55.407 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 9/120, iter: 100/129, gpu mem: 2418Mb, mem: 95.0Gb, iter_time: 0.181s, data_time: 0.028s, total_loss: 6.8, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 3.3, cls_loss: 0.8, lr: 1.974e-03, size: 416, ETA: 0:40:39
2025-11-12 14:15:59.184 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 9/120, iter: 120/129, gpu mem: 2418Mb, mem: 95.0Gb, iter_time: 0.187s, data_time: 0.031s, total_loss: 6.4, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.9, lr: 1.973e-03, size: 512, ETA: 0:40:40
2025-11-12 14:16:00.969 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:16:06.030 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:16:08.726 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:16:10.633 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6207
2025-11-12 14:16:10.999 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5529
2025-11-12 14:16:11.052 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3693
2025-11-12 14:16:11.053 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5143
2025-11-12 14:16:11.053 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:16:11.053 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:16:11.053 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.621
2025-11-12 14:16:11.053 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.553
2025-11-12 14:16:11.053 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.369
2025-11-12 14:16:11.053 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.514
2025-11-12 14:16:11.054 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:16:11.054 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:16:11.054 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:16:11.054 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:16:11.054 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:16:11.054 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:16:11.054 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:16:11.054 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:16:11.054 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:16:13.340 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:16:15.635 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:16:17.874 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:16:20.124 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:16:22.371 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:16:24.686 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:16:26.950 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:16:29.206 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:16:31.473 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:16:31.473 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.28
2025-11-12 14:16:31.473 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.51
2025-11-12 14:16:31.473 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:16:31.498 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.18 ms, Average NMS time: 0.59 ms, Average inference time: 2.76 ms

2025-11-12 14:16:31.500 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:16:31.576 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:16:31.657 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch10
2025-11-12 14:16:34.898 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 10/120, iter: 20/129, gpu mem: 2418Mb, mem: 95.0Gb, iter_time: 0.159s, data_time: 0.015s, total_loss: 5.8, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.2, cls_loss: 0.9, lr: 1.971e-03, size: 320, ETA: 0:40:35
2025-11-12 14:16:38.621 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 10/120, iter: 40/129, gpu mem: 2418Mb, mem: 95.0Gb, iter_time: 0.184s, data_time: 0.028s, total_loss: 5.8, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.4, cls_loss: 0.9, lr: 1.970e-03, size: 320, ETA: 0:40:35
2025-11-12 14:16:42.271 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 10/120, iter: 60/129, gpu mem: 2418Mb, mem: 94.9Gb, iter_time: 0.181s, data_time: 0.039s, total_loss: 7.1, iou_loss: 3.1, l1_loss: 0.0, conf_loss: 3.2, cls_loss: 0.8, lr: 1.969e-03, size: 384, ETA: 0:40:34
2025-11-12 14:16:46.195 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 10/120, iter: 80/129, gpu mem: 2418Mb, mem: 95.0Gb, iter_time: 0.195s, data_time: 0.054s, total_loss: 6.2, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.9, cls_loss: 0.7, lr: 1.968e-03, size: 512, ETA: 0:40:36
2025-11-12 14:16:50.130 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 10/120, iter: 100/129, gpu mem: 2418Mb, mem: 95.0Gb, iter_time: 0.196s, data_time: 0.037s, total_loss: 6.1, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.8, lr: 1.967e-03, size: 256, ETA: 0:40:38
2025-11-12 14:16:53.908 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 10/120, iter: 120/129, gpu mem: 2418Mb, mem: 95.0Gb, iter_time: 0.188s, data_time: 0.039s, total_loss: 5.7, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.4, cls_loss: 0.7, lr: 1.966e-03, size: 288, ETA: 0:40:39
2025-11-12 14:16:55.355 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:17:00.690 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:17:03.933 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:17:06.036 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6145
2025-11-12 14:17:06.458 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5574
2025-11-12 14:17:06.587 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3997
2025-11-12 14:17:06.588 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5238
2025-11-12 14:17:06.588 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:17:06.588 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:17:06.588 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.615
2025-11-12 14:17:06.588 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.557
2025-11-12 14:17:06.588 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.400
2025-11-12 14:17:06.589 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.524
2025-11-12 14:17:06.589 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:17:06.589 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:17:06.589 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:17:06.589 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:17:06.589 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:17:06.589 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:17:06.589 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:17:06.589 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:17:06.590 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:17:09.236 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:17:11.909 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:17:14.574 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:17:17.260 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:17:19.918 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:17:22.591 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:17:25.239 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:17:27.840 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:17:30.420 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:17:30.421 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.28
2025-11-12 14:17:30.421 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.52
2025-11-12 14:17:30.421 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:17:30.447 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.22 ms, Average NMS time: 0.63 ms, Average inference time: 2.85 ms

2025-11-12 14:17:30.448 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:17:30.524 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:17:30.644 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch11
2025-11-12 14:17:33.954 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 11/120, iter: 20/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.165s, data_time: 0.010s, total_loss: 6.6, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.9, cls_loss: 1.0, lr: 1.965e-03, size: 576, ETA: 0:40:31
2025-11-12 14:17:37.701 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 11/120, iter: 40/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.185s, data_time: 0.046s, total_loss: 6.5, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 3.0, cls_loss: 0.7, lr: 1.964e-03, size: 544, ETA: 0:40:30
2025-11-12 14:17:41.291 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 11/120, iter: 60/129, gpu mem: 2419Mb, mem: 95.1Gb, iter_time: 0.177s, data_time: 0.038s, total_loss: 6.7, iou_loss: 3.1, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.9, lr: 1.963e-03, size: 320, ETA: 0:40:28
2025-11-12 14:17:45.141 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 11/120, iter: 80/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.190s, data_time: 0.046s, total_loss: 6.6, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 3.0, cls_loss: 0.9, lr: 1.962e-03, size: 512, ETA: 0:40:28
2025-11-12 14:17:48.777 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 11/120, iter: 100/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.181s, data_time: 0.019s, total_loss: 6.4, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.8, lr: 1.960e-03, size: 576, ETA: 0:40:26
2025-11-12 14:17:52.269 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 11/120, iter: 120/129, gpu mem: 2419Mb, mem: 95.1Gb, iter_time: 0.172s, data_time: 0.009s, total_loss: 6.6, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.9, cls_loss: 0.9, lr: 1.959e-03, size: 288, ETA: 0:40:23
2025-11-12 14:17:54.040 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:17:59.120 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:18:02.339 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:18:04.552 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6001
2025-11-12 14:18:04.945 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5524
2025-11-12 14:18:05.026 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3897
2025-11-12 14:18:05.027 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5141
2025-11-12 14:18:05.027 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:18:05.027 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:18:05.027 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.600
2025-11-12 14:18:05.027 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.552
2025-11-12 14:18:05.027 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.390
2025-11-12 14:18:05.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.514
2025-11-12 14:18:05.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:18:05.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:18:05.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:18:05.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:18:05.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:18:05.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:18:05.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:18:05.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:18:05.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:18:07.692 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:18:10.309 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:18:12.966 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:18:15.554 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:18:18.128 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:18:20.778 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:18:23.408 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:18:25.990 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:18:28.613 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:18:28.613 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.26
2025-11-12 14:18:28.613 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.51
2025-11-12 14:18:28.614 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:18:28.639 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.17 ms, Average NMS time: 0.60 ms, Average inference time: 2.76 ms

2025-11-12 14:18:28.640 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:18:28.715 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:18:28.797 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch12
2025-11-12 14:18:32.125 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 12/120, iter: 20/129, gpu mem: 2419Mb, mem: 95.1Gb, iter_time: 0.165s, data_time: 0.018s, total_loss: 6.0, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.7, lr: 1.958e-03, size: 416, ETA: 0:40:18
2025-11-12 14:18:35.915 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 12/120, iter: 40/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.189s, data_time: 0.040s, total_loss: 6.8, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 3.0, cls_loss: 1.0, lr: 1.956e-03, size: 448, ETA: 0:40:18
2025-11-12 14:18:39.656 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 12/120, iter: 60/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.185s, data_time: 0.042s, total_loss: 6.0, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.8, lr: 1.955e-03, size: 448, ETA: 0:40:17
2025-11-12 14:18:43.132 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 12/120, iter: 80/129, gpu mem: 2419Mb, mem: 95.1Gb, iter_time: 0.173s, data_time: 0.028s, total_loss: 6.3, iou_loss: 3.0, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.8, lr: 1.954e-03, size: 416, ETA: 0:40:14
2025-11-12 14:18:46.973 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 12/120, iter: 100/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.191s, data_time: 0.044s, total_loss: 5.7, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.4, cls_loss: 0.8, lr: 1.953e-03, size: 512, ETA: 0:40:13
2025-11-12 14:18:50.759 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 12/120, iter: 120/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.188s, data_time: 0.046s, total_loss: 6.2, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.9, lr: 1.952e-03, size: 320, ETA: 0:40:13
2025-11-12 14:18:52.369 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:18:57.494 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:19:00.045 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:19:01.690 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6147
2025-11-12 14:19:02.041 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5616
2025-11-12 14:19:02.102 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3796
2025-11-12 14:19:02.103 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5186
2025-11-12 14:19:02.103 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:19:02.103 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:19:02.103 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.615
2025-11-12 14:19:02.104 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.562
2025-11-12 14:19:02.104 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.380
2025-11-12 14:19:02.104 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.519
2025-11-12 14:19:02.104 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:19:02.104 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:19:02.104 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:19:02.104 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:19:02.104 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:19:02.104 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:19:02.105 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:19:02.105 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:19:02.105 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:19:04.157 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:19:06.219 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:19:08.246 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:19:10.280 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:19:12.352 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:19:14.301 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:19:16.285 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:19:18.288 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:19:20.306 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:19:20.307 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.27
2025-11-12 14:19:20.307 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.52
2025-11-12 14:19:20.307 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:19:20.331 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.15 ms, Average NMS time: 0.58 ms, Average inference time: 2.73 ms

2025-11-12 14:19:20.333 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:19:20.408 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:19:20.526 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch13
2025-11-12 14:19:23.948 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 13/120, iter: 20/129, gpu mem: 2419Mb, mem: 95.1Gb, iter_time: 0.170s, data_time: 0.019s, total_loss: 7.1, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 3.3, cls_loss: 0.9, lr: 1.950e-03, size: 512, ETA: 0:40:08
2025-11-12 14:19:27.467 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 13/120, iter: 40/129, gpu mem: 2419Mb, mem: 95.1Gb, iter_time: 0.174s, data_time: 0.022s, total_loss: 6.4, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.9, lr: 1.949e-03, size: 320, ETA: 0:40:04
2025-11-12 14:19:31.026 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 13/120, iter: 60/129, gpu mem: 2419Mb, mem: 95.1Gb, iter_time: 0.177s, data_time: 0.027s, total_loss: 6.9, iou_loss: 3.3, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.7, lr: 1.947e-03, size: 320, ETA: 0:40:02
2025-11-12 14:19:34.745 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 13/120, iter: 80/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.185s, data_time: 0.043s, total_loss: 6.3, iou_loss: 3.0, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.7, lr: 1.946e-03, size: 288, ETA: 0:40:00
2025-11-12 14:19:38.456 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 13/120, iter: 100/129, gpu mem: 2419Mb, mem: 95.1Gb, iter_time: 0.185s, data_time: 0.034s, total_loss: 5.6, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.2, cls_loss: 0.8, lr: 1.945e-03, size: 384, ETA: 0:39:59
2025-11-12 14:19:42.243 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 13/120, iter: 120/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.188s, data_time: 0.034s, total_loss: 7.0, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 3.3, cls_loss: 0.9, lr: 1.943e-03, size: 480, ETA: 0:39:58
2025-11-12 14:19:43.906 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:19:49.052 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:19:52.033 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:19:53.866 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6077
2025-11-12 14:19:54.301 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5511
2025-11-12 14:19:54.433 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3791
2025-11-12 14:19:54.434 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5126
2025-11-12 14:19:54.434 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:19:54.434 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:19:54.434 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.608
2025-11-12 14:19:54.434 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.551
2025-11-12 14:19:54.434 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.379
2025-11-12 14:19:54.434 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.513
2025-11-12 14:19:54.434 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:19:54.435 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:19:54.435 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:19:54.435 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:19:54.435 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:19:54.435 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:19:54.435 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:19:54.435 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:19:54.435 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:19:56.781 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:19:59.187 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:20:01.521 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:20:03.873 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:20:06.262 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:20:08.588 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:20:10.903 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:20:13.268 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:20:15.602 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:20:15.602 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.27
2025-11-12 14:20:15.602 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.51
2025-11-12 14:20:15.603 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:20:15.628 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.19 ms, Average NMS time: 0.62 ms, Average inference time: 2.81 ms

2025-11-12 14:20:15.629 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:20:15.704 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:20:15.786 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch14
2025-11-12 14:20:19.254 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 14/120, iter: 20/129, gpu mem: 2419Mb, mem: 95.1Gb, iter_time: 0.172s, data_time: 0.003s, total_loss: 6.7, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 3.1, cls_loss: 0.9, lr: 1.941e-03, size: 352, ETA: 0:39:53
2025-11-12 14:20:23.016 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 14/120, iter: 40/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.186s, data_time: 0.043s, total_loss: 6.1, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.8, lr: 1.940e-03, size: 512, ETA: 0:39:52
2025-11-12 14:20:26.701 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 14/120, iter: 60/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.183s, data_time: 0.020s, total_loss: 6.3, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 1.0, lr: 1.939e-03, size: 544, ETA: 0:39:50
2025-11-12 14:20:30.377 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 14/120, iter: 80/129, gpu mem: 2419Mb, mem: 95.1Gb, iter_time: 0.183s, data_time: 0.011s, total_loss: 5.9, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.9, lr: 1.937e-03, size: 576, ETA: 0:39:48
2025-11-12 14:20:34.176 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 14/120, iter: 100/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.188s, data_time: 0.020s, total_loss: 5.8, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.7, lr: 1.936e-03, size: 480, ETA: 0:39:46
2025-11-12 14:20:38.010 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 14/120, iter: 120/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.190s, data_time: 0.039s, total_loss: 6.2, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.8, lr: 1.934e-03, size: 256, ETA: 0:39:45
2025-11-12 14:20:39.670 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:20:44.684 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:20:47.286 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:20:48.974 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6103
2025-11-12 14:20:49.320 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5661
2025-11-12 14:20:49.392 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3851
2025-11-12 14:20:49.393 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5205
2025-11-12 14:20:49.393 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:20:49.393 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:20:49.394 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.610
2025-11-12 14:20:49.394 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.566
2025-11-12 14:20:49.394 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.385
2025-11-12 14:20:49.394 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.521
2025-11-12 14:20:49.394 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:20:49.394 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:20:49.394 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:20:49.394 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:20:49.394 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:20:49.395 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:20:49.395 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:20:49.395 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:20:49.395 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:20:51.517 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:20:53.587 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:20:56.147 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:20:58.249 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:21:00.329 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:21:02.409 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:21:04.504 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:21:06.599 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:21:08.659 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:21:08.660 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.28
2025-11-12 14:21:08.660 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.52
2025-11-12 14:21:08.660 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:21:08.685 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.16 ms, Average NMS time: 0.58 ms, Average inference time: 2.74 ms

2025-11-12 14:21:08.686 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:21:08.762 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:21:08.844 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch15
2025-11-12 14:21:12.377 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 15/120, iter: 20/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.175s, data_time: 0.028s, total_loss: 5.5, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.2, cls_loss: 0.7, lr: 1.932e-03, size: 416, ETA: 0:39:41
2025-11-12 14:21:16.247 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 15/120, iter: 40/129, gpu mem: 2419Mb, mem: 95.1Gb, iter_time: 0.193s, data_time: 0.040s, total_loss: 5.8, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.4, cls_loss: 0.8, lr: 1.931e-03, size: 544, ETA: 0:39:40
2025-11-12 14:21:20.107 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 15/120, iter: 60/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.192s, data_time: 0.043s, total_loss: 5.8, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.7, lr: 1.929e-03, size: 544, ETA: 0:39:39
2025-11-12 14:21:23.896 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 15/120, iter: 80/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.189s, data_time: 0.043s, total_loss: 5.6, iou_loss: 2.4, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.7, lr: 1.928e-03, size: 448, ETA: 0:39:38
2025-11-12 14:21:27.693 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 15/120, iter: 100/129, gpu mem: 2419Mb, mem: 95.0Gb, iter_time: 0.189s, data_time: 0.034s, total_loss: 6.6, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 1.0, lr: 1.926e-03, size: 384, ETA: 0:39:36
2025-11-12 14:21:31.552 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 15/120, iter: 120/129, gpu mem: 2419Mb, mem: 95.1Gb, iter_time: 0.192s, data_time: 0.043s, total_loss: 6.9, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 3.3, cls_loss: 0.9, lr: 1.925e-03, size: 576, ETA: 0:39:35
2025-11-12 14:21:33.182 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:21:38.257 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:21:41.298 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:21:43.421 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6162
2025-11-12 14:21:43.737 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5554
2025-11-12 14:21:43.871 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3662
2025-11-12 14:21:43.872 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5126
2025-11-12 14:21:43.872 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:21:43.872 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:21:43.873 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.616
2025-11-12 14:21:43.873 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.555
2025-11-12 14:21:43.873 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.366
2025-11-12 14:21:43.873 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.513
2025-11-12 14:21:43.873 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:21:43.873 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:21:43.873 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:21:43.873 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:21:43.873 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:21:43.874 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:21:43.874 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:21:43.874 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:21:43.874 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:21:46.301 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:21:48.890 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:21:51.336 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:21:53.873 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:21:56.355 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:21:58.781 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:22:01.274 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:22:03.714 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:22:06.323 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:22:06.323 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.27
2025-11-12 14:22:06.323 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.51
2025-11-12 14:22:06.324 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:22:06.348 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.16 ms, Average NMS time: 0.59 ms, Average inference time: 2.76 ms

2025-11-12 14:22:06.350 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:22:06.425 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:22:06.506 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch16
2025-11-12 14:22:09.869 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 16/120, iter: 20/129, gpu mem: 2419Mb, mem: 95.1Gb, iter_time: 0.166s, data_time: 0.002s, total_loss: 6.2, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.9, lr: 1.922e-03, size: 544, ETA: 0:39:29
2025-11-12 14:22:13.411 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 16/120, iter: 40/129, gpu mem: 2419Mb, mem: 95.1Gb, iter_time: 0.175s, data_time: 0.003s, total_loss: 6.3, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.8, lr: 1.921e-03, size: 576, ETA: 0:39:25
2025-11-12 14:22:16.866 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 16/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.172s, data_time: 0.006s, total_loss: 6.4, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.9, lr: 1.919e-03, size: 512, ETA: 0:39:21
2025-11-12 14:22:20.392 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 16/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.173s, data_time: 0.015s, total_loss: 5.8, iou_loss: 2.4, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.8, lr: 1.918e-03, size: 480, ETA: 0:39:18
2025-11-12 14:22:23.970 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 16/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.178s, data_time: 0.018s, total_loss: 6.4, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.8, lr: 1.916e-03, size: 512, ETA: 0:39:15
2025-11-12 14:22:27.544 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 16/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.177s, data_time: 0.004s, total_loss: 5.8, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.2, cls_loss: 0.8, lr: 1.914e-03, size: 480, ETA: 0:39:11
2025-11-12 14:22:29.113 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:22:34.313 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:22:36.674 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:22:38.219 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6067
2025-11-12 14:22:38.496 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5664
2025-11-12 14:22:38.605 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3755
2025-11-12 14:22:38.606 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5162
2025-11-12 14:22:38.606 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:22:38.606 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:22:38.606 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.607
2025-11-12 14:22:38.607 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.566
2025-11-12 14:22:38.607 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.376
2025-11-12 14:22:38.607 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.516
2025-11-12 14:22:38.607 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:22:38.607 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:22:38.607 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:22:38.607 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:22:38.607 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:22:38.608 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:22:38.608 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:22:38.608 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:22:38.608 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:22:40.487 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:22:42.371 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:22:44.284 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:22:46.169 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:22:48.154 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:22:49.983 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:22:51.854 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:22:53.755 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:22:55.648 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:22:55.649 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.27
2025-11-12 14:22:55.649 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.52
2025-11-12 14:22:55.649 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:22:55.674 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.16 ms, Average NMS time: 0.58 ms, Average inference time: 2.74 ms

2025-11-12 14:22:55.675 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:22:55.750 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:22:55.831 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch17
2025-11-12 14:22:59.099 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 17/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.161s, data_time: 0.002s, total_loss: 6.4, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.9, lr: 1.912e-03, size: 448, ETA: 0:39:04
2025-11-12 14:23:02.552 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 17/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.169s, data_time: 0.020s, total_loss: 6.0, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.8, lr: 1.910e-03, size: 384, ETA: 0:39:00
2025-11-12 14:23:06.159 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 17/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.177s, data_time: 0.005s, total_loss: 6.9, iou_loss: 3.0, l1_loss: 0.0, conf_loss: 3.1, cls_loss: 0.8, lr: 1.909e-03, size: 288, ETA: 0:38:57
2025-11-12 14:23:09.637 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 17/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.0Gb, iter_time: 0.172s, data_time: 0.003s, total_loss: 6.3, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.9, cls_loss: 0.8, lr: 1.907e-03, size: 416, ETA: 0:38:53
2025-11-12 14:23:13.209 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 17/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.178s, data_time: 0.012s, total_loss: 5.8, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 0.7, lr: 1.905e-03, size: 384, ETA: 0:38:50
2025-11-12 14:23:16.820 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 17/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.178s, data_time: 0.012s, total_loss: 6.3, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.9, lr: 1.903e-03, size: 256, ETA: 0:38:47
2025-11-12 14:23:18.473 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:23:23.597 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:23:25.669 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:23:26.974 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6009
2025-11-12 14:23:27.296 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5371
2025-11-12 14:23:27.339 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3402
2025-11-12 14:23:27.339 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.4927
2025-11-12 14:23:27.340 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:23:27.340 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:23:27.340 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.601
2025-11-12 14:23:27.340 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.537
2025-11-12 14:23:27.340 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.340
2025-11-12 14:23:27.340 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.493
2025-11-12 14:23:27.340 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:23:27.340 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:23:27.340 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:23:27.341 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:23:27.341 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:23:27.341 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:23:27.341 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:23:27.341 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:23:27.341 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:23:29.070 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:23:30.707 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:23:32.396 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:23:34.074 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:23:35.744 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:23:37.384 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:23:39.296 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:23:40.946 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:23:42.636 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:23:42.636 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.26
2025-11-12 14:23:42.636 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.49
2025-11-12 14:23:42.636 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:23:42.661 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.17 ms, Average NMS time: 0.57 ms, Average inference time: 2.74 ms

2025-11-12 14:23:42.663 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:23:42.739 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:23:42.824 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch18
2025-11-12 14:23:46.273 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 18/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.170s, data_time: 0.003s, total_loss: 6.4, iou_loss: 3.0, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.9, lr: 1.901e-03, size: 256, ETA: 0:38:41
2025-11-12 14:23:49.769 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 18/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.172s, data_time: 0.024s, total_loss: 5.9, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 1.0, lr: 1.899e-03, size: 320, ETA: 0:38:38
2025-11-12 14:23:53.341 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 18/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.176s, data_time: 0.006s, total_loss: 6.4, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.8, lr: 1.897e-03, size: 288, ETA: 0:38:34
2025-11-12 14:23:57.304 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 18/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.0Gb, iter_time: 0.196s, data_time: 0.019s, total_loss: 6.6, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 3.1, cls_loss: 0.8, lr: 1.895e-03, size: 512, ETA: 0:38:33
2025-11-12 14:24:01.241 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 18/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.193s, data_time: 0.032s, total_loss: 6.2, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.8, lr: 1.894e-03, size: 448, ETA: 0:38:32
2025-11-12 14:24:04.773 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 18/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.176s, data_time: 0.015s, total_loss: 6.1, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.8, lr: 1.892e-03, size: 448, ETA: 0:38:28
2025-11-12 14:24:06.304 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:24:11.566 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:24:13.977 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:24:15.601 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6207
2025-11-12 14:24:15.885 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5715
2025-11-12 14:24:15.994 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4151
2025-11-12 14:24:15.996 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5358
2025-11-12 14:24:15.996 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:24:15.996 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:24:15.996 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.621
2025-11-12 14:24:15.996 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.572
2025-11-12 14:24:15.996 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.415
2025-11-12 14:24:15.996 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.536
2025-11-12 14:24:15.996 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:24:15.997 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:24:15.997 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:24:15.997 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:24:15.997 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:24:15.997 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:24:15.997 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:24:15.997 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:24:15.997 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:24:17.979 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:24:19.890 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:24:21.896 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:24:23.860 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:24:25.835 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:24:27.826 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:24:29.789 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:24:31.687 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:24:33.619 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:24:33.620 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.27
2025-11-12 14:24:33.620 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.54
2025-11-12 14:24:33.620 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:24:33.644 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.23 ms, Average NMS time: 0.61 ms, Average inference time: 2.84 ms

2025-11-12 14:24:33.646 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:24:33.720 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:24:33.801 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch19
2025-11-12 14:24:37.039 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 19/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.160s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.2, cls_loss: 0.8, lr: 1.889e-03, size: 480, ETA: 0:38:21
2025-11-12 14:24:40.659 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 19/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.178s, data_time: 0.003s, total_loss: 6.5, iou_loss: 3.0, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.8, lr: 1.887e-03, size: 480, ETA: 0:38:18
2025-11-12 14:24:44.130 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 19/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.170s, data_time: 0.002s, total_loss: 6.5, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.8, lr: 1.885e-03, size: 256, ETA: 0:38:14
2025-11-12 14:24:47.681 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 19/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.175s, data_time: 0.001s, total_loss: 6.1, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.8, lr: 1.884e-03, size: 416, ETA: 0:38:10
2025-11-12 14:24:51.242 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 19/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.175s, data_time: 0.002s, total_loss: 7.1, iou_loss: 3.3, l1_loss: 0.0, conf_loss: 2.9, cls_loss: 0.8, lr: 1.882e-03, size: 320, ETA: 0:38:07
2025-11-12 14:24:54.639 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 19/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.168s, data_time: 0.003s, total_loss: 6.4, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.8, lr: 1.880e-03, size: 320, ETA: 0:38:03
2025-11-12 14:24:56.297 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:25:01.485 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:25:04.193 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:25:05.862 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.5910
2025-11-12 14:25:06.269 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5514
2025-11-12 14:25:06.330 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3500
2025-11-12 14:25:06.331 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.4974
2025-11-12 14:25:06.331 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:25:06.331 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:25:06.331 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.591
2025-11-12 14:25:06.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.551
2025-11-12 14:25:06.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.350
2025-11-12 14:25:06.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.497
2025-11-12 14:25:06.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:25:06.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:25:06.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:25:06.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:25:06.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:25:06.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:25:06.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:25:06.333 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:25:06.333 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:25:08.515 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:25:10.643 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:25:12.800 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:25:14.931 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:25:17.067 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:25:19.163 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:25:21.258 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:25:23.407 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:25:25.521 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:25:25.522 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.25
2025-11-12 14:25:25.522 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.50
2025-11-12 14:25:25.522 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:25:25.546 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.21 ms, Average NMS time: 0.63 ms, Average inference time: 2.84 ms

2025-11-12 14:25:25.548 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:25:25.622 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:25:25.704 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch20
2025-11-12 14:25:28.943 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 20/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.159s, data_time: 0.011s, total_loss: 6.8, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 3.1, cls_loss: 0.8, lr: 1.877e-03, size: 512, ETA: 0:37:56
2025-11-12 14:25:32.505 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 20/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.177s, data_time: 0.028s, total_loss: 6.2, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.9, lr: 1.875e-03, size: 544, ETA: 0:37:53
2025-11-12 14:25:35.958 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 20/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.172s, data_time: 0.009s, total_loss: 5.9, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.8, lr: 1.873e-03, size: 512, ETA: 0:37:49
2025-11-12 14:25:39.556 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 20/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.0Gb, iter_time: 0.178s, data_time: 0.014s, total_loss: 6.3, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.9, lr: 1.871e-03, size: 416, ETA: 0:37:46
2025-11-12 14:25:42.950 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 20/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.169s, data_time: 0.013s, total_loss: 6.6, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 3.1, cls_loss: 0.8, lr: 1.869e-03, size: 512, ETA: 0:37:42
2025-11-12 14:25:46.377 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 20/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.0Gb, iter_time: 0.171s, data_time: 0.019s, total_loss: 6.4, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 1.0, lr: 1.867e-03, size: 576, ETA: 0:37:38
2025-11-12 14:25:47.985 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:25:53.221 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:25:55.586 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:25:57.212 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6191
2025-11-12 14:25:57.496 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5779
2025-11-12 14:25:57.603 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3855
2025-11-12 14:25:57.604 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5275
2025-11-12 14:25:57.604 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:25:57.604 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:25:57.605 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.619
2025-11-12 14:25:57.605 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.578
2025-11-12 14:25:57.605 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.385
2025-11-12 14:25:57.605 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.528
2025-11-12 14:25:57.605 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:25:57.605 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:25:57.605 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:25:57.605 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:25:57.605 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:25:57.605 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:25:57.606 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:25:57.606 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:25:57.606 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:25:59.502 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:26:01.418 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:26:03.400 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:26:05.321 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:26:07.265 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:26:09.177 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:26:11.087 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:26:13.016 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:26:14.949 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:26:14.949 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.27
2025-11-12 14:26:14.949 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.53
2025-11-12 14:26:14.949 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:26:14.973 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.21 ms, Average NMS time: 0.62 ms, Average inference time: 2.83 ms

2025-11-12 14:26:14.976 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:26:15.050 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:26:15.131 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch21
2025-11-12 14:26:18.643 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 21/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.174s, data_time: 0.002s, total_loss: 5.3, iou_loss: 2.3, l1_loss: 0.0, conf_loss: 2.2, cls_loss: 0.8, lr: 1.864e-03, size: 352, ETA: 0:37:33
2025-11-12 14:26:22.041 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 21/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.168s, data_time: 0.002s, total_loss: 6.3, iou_loss: 2.4, l1_loss: 0.0, conf_loss: 3.0, cls_loss: 0.8, lr: 1.862e-03, size: 576, ETA: 0:37:29
2025-11-12 14:26:25.644 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 21/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.178s, data_time: 0.002s, total_loss: 6.8, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 3.1, cls_loss: 1.0, lr: 1.860e-03, size: 448, ETA: 0:37:26
2025-11-12 14:26:29.201 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 21/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.0Gb, iter_time: 0.176s, data_time: 0.002s, total_loss: 6.0, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.8, lr: 1.858e-03, size: 544, ETA: 0:37:22
2025-11-12 14:26:32.798 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 21/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.178s, data_time: 0.003s, total_loss: 7.5, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 3.7, cls_loss: 1.0, lr: 1.856e-03, size: 512, ETA: 0:37:19
2025-11-12 14:26:36.289 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 21/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.173s, data_time: 0.003s, total_loss: 6.5, iou_loss: 3.0, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.8, lr: 1.854e-03, size: 512, ETA: 0:37:15
2025-11-12 14:26:37.786 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:26:42.907 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:26:46.114 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:26:48.345 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6047
2025-11-12 14:26:48.719 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5427
2025-11-12 14:26:48.812 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3778
2025-11-12 14:26:48.813 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5084
2025-11-12 14:26:48.813 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:26:48.814 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:26:48.814 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.605
2025-11-12 14:26:48.814 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.543
2025-11-12 14:26:48.814 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.378
2025-11-12 14:26:48.814 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.508
2025-11-12 14:26:48.814 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:26:48.814 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:26:48.814 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:26:48.814 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:26:48.815 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:26:48.815 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:26:48.815 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:26:48.815 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:26:48.815 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:26:51.450 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:26:54.082 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:26:56.661 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:26:59.330 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:27:01.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:27:04.514 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:27:07.084 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:27:09.694 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:27:12.311 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:27:12.312 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.25
2025-11-12 14:27:12.312 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.51
2025-11-12 14:27:12.312 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:27:12.338 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.15 ms, Average NMS time: 0.60 ms, Average inference time: 2.75 ms

2025-11-12 14:27:12.339 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:27:12.417 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:27:12.499 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch22
2025-11-12 14:27:15.894 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 22/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.168s, data_time: 0.002s, total_loss: 6.1, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.9, lr: 1.851e-03, size: 256, ETA: 0:37:09
2025-11-12 14:27:19.371 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 22/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.0Gb, iter_time: 0.172s, data_time: 0.008s, total_loss: 6.0, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.8, lr: 1.848e-03, size: 384, ETA: 0:37:05
2025-11-12 14:27:22.936 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 22/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.176s, data_time: 0.010s, total_loss: 6.0, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 0.7, lr: 1.846e-03, size: 480, ETA: 0:37:02
2025-11-12 14:27:26.346 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 22/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.168s, data_time: 0.015s, total_loss: 5.8, iou_loss: 2.2, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 1.1, lr: 1.844e-03, size: 256, ETA: 0:36:58
2025-11-12 14:27:29.971 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 22/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.0Gb, iter_time: 0.180s, data_time: 0.020s, total_loss: 5.2, iou_loss: 2.4, l1_loss: 0.0, conf_loss: 2.1, cls_loss: 0.7, lr: 1.842e-03, size: 544, ETA: 0:36:55
2025-11-12 14:27:33.412 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 22/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.170s, data_time: 0.010s, total_loss: 5.9, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.8, lr: 1.840e-03, size: 416, ETA: 0:36:51
2025-11-12 14:27:34.946 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:27:40.035 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:27:42.819 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:27:44.703 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6184
2025-11-12 14:27:45.054 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5599
2025-11-12 14:27:45.124 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3835
2025-11-12 14:27:45.124 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5206
2025-11-12 14:27:45.125 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:27:45.125 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:27:45.125 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.618
2025-11-12 14:27:45.125 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.560
2025-11-12 14:27:45.125 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.383
2025-11-12 14:27:45.125 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.521
2025-11-12 14:27:45.125 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:27:45.125 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:27:45.126 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:27:45.126 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:27:45.126 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:27:45.126 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:27:45.126 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:27:45.126 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:27:45.126 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:27:47.437 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:27:49.723 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:27:52.013 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:27:54.312 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:27:56.574 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:27:58.825 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:28:01.094 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:28:03.393 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:28:05.661 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:28:05.662 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.28
2025-11-12 14:28:05.662 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.52
2025-11-12 14:28:05.663 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:28:05.693 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.12 ms, Average NMS time: 0.59 ms, Average inference time: 2.72 ms

2025-11-12 14:28:05.694 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:28:05.769 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:28:05.850 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch23
2025-11-12 14:28:09.027 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 23/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.157s, data_time: 0.017s, total_loss: 6.1, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.9, lr: 1.836e-03, size: 416, ETA: 0:36:44
2025-11-12 14:28:12.542 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 23/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.174s, data_time: 0.020s, total_loss: 6.6, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.9, lr: 1.834e-03, size: 320, ETA: 0:36:40
2025-11-12 14:28:16.047 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 23/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.173s, data_time: 0.017s, total_loss: 6.6, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 3.0, cls_loss: 0.8, lr: 1.832e-03, size: 384, ETA: 0:36:37
2025-11-12 14:28:19.648 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 23/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.178s, data_time: 0.018s, total_loss: 5.8, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.9, lr: 1.830e-03, size: 576, ETA: 0:36:33
2025-11-12 14:28:23.217 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 23/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.176s, data_time: 0.005s, total_loss: 6.4, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 3.0, cls_loss: 0.8, lr: 1.827e-03, size: 480, ETA: 0:36:30
2025-11-12 14:28:26.728 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 23/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.172s, data_time: 0.002s, total_loss: 6.2, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.8, lr: 1.825e-03, size: 544, ETA: 0:36:26
2025-11-12 14:28:28.396 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:28:33.497 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:28:36.193 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:28:37.930 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.5953
2025-11-12 14:28:38.420 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5391
2025-11-12 14:28:38.497 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3487
2025-11-12 14:28:38.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.4944
2025-11-12 14:28:38.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:28:38.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:28:38.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.595
2025-11-12 14:28:38.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.539
2025-11-12 14:28:38.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.349
2025-11-12 14:28:38.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.494
2025-11-12 14:28:38.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:28:38.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:28:38.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:28:38.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:28:38.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:28:38.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:28:38.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:28:38.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:28:38.500 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:28:40.683 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:28:42.911 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:28:45.096 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:28:47.274 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:28:49.516 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:28:51.693 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:28:53.881 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:28:56.046 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:28:58.247 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:28:58.247 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.24
2025-11-12 14:28:58.248 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.49
2025-11-12 14:28:58.248 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:28:58.274 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.15 ms, Average NMS time: 0.57 ms, Average inference time: 2.72 ms

2025-11-12 14:28:58.275 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:28:58.355 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:28:58.442 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch24
2025-11-12 14:29:01.779 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 24/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.165s, data_time: 0.002s, total_loss: 6.0, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.9, lr: 1.822e-03, size: 544, ETA: 0:36:21
2025-11-12 14:29:05.468 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 24/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.181s, data_time: 0.002s, total_loss: 5.3, iou_loss: 2.4, l1_loss: 0.0, conf_loss: 2.2, cls_loss: 0.7, lr: 1.820e-03, size: 544, ETA: 0:36:18
2025-11-12 14:29:08.877 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 24/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.168s, data_time: 0.002s, total_loss: 6.7, iou_loss: 3.1, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.9, lr: 1.817e-03, size: 320, ETA: 0:36:14
2025-11-12 14:29:12.350 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 24/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.171s, data_time: 0.002s, total_loss: 6.5, iou_loss: 3.0, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 1.0, lr: 1.815e-03, size: 288, ETA: 0:36:10
2025-11-12 14:29:15.881 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 24/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.175s, data_time: 0.015s, total_loss: 5.6, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 0.7, lr: 1.812e-03, size: 352, ETA: 0:36:06
2025-11-12 14:29:19.594 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 24/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.185s, data_time: 0.017s, total_loss: 5.7, iou_loss: 2.4, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.7, lr: 1.810e-03, size: 544, ETA: 0:36:04
2025-11-12 14:29:21.208 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:29:26.333 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:29:29.962 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:29:32.406 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6055
2025-11-12 14:29:32.847 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5632
2025-11-12 14:29:32.923 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3658
2025-11-12 14:29:32.924 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5115
2025-11-12 14:29:32.924 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:29:32.924 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:29:32.924 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.606
2025-11-12 14:29:32.925 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.563
2025-11-12 14:29:32.925 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.366
2025-11-12 14:29:32.925 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.511
2025-11-12 14:29:32.925 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:29:32.925 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:29:32.925 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:29:32.925 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:29:32.925 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:29:32.925 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:29:32.925 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:29:32.926 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:29:32.926 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:29:35.896 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:29:38.839 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:29:41.805 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:29:44.935 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:29:47.844 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:29:50.754 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:29:53.652 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:29:56.505 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:29:59.355 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:29:59.355 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.26
2025-11-12 14:29:59.355 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.51
2025-11-12 14:29:59.356 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:29:59.380 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.17 ms, Average NMS time: 0.62 ms, Average inference time: 2.79 ms

2025-11-12 14:29:59.382 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:29:59.457 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:29:59.536 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch25
2025-11-12 14:30:02.741 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 25/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.158s, data_time: 0.013s, total_loss: 6.5, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.9, cls_loss: 0.9, lr: 1.807e-03, size: 384, ETA: 0:35:58
2025-11-12 14:30:06.181 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 25/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.171s, data_time: 0.021s, total_loss: 6.4, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.9, lr: 1.804e-03, size: 288, ETA: 0:35:54
2025-11-12 14:30:09.769 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 25/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.179s, data_time: 0.037s, total_loss: 6.5, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.9, cls_loss: 0.8, lr: 1.802e-03, size: 416, ETA: 0:35:51
2025-11-12 14:30:13.489 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 25/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.185s, data_time: 0.033s, total_loss: 6.8, iou_loss: 3.2, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.9, lr: 1.799e-03, size: 256, ETA: 0:35:48
2025-11-12 14:30:17.129 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 25/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.181s, data_time: 0.030s, total_loss: 6.4, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 1.0, lr: 1.797e-03, size: 384, ETA: 0:35:45
2025-11-12 14:30:20.611 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 25/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.173s, data_time: 0.042s, total_loss: 6.4, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.8, lr: 1.794e-03, size: 256, ETA: 0:35:41
2025-11-12 14:30:22.100 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:30:27.062 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:30:28.972 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:30:30.179 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.5804
2025-11-12 14:30:30.444 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5314
2025-11-12 14:30:30.484 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3333
2025-11-12 14:30:30.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.4817
2025-11-12 14:30:30.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:30:30.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:30:30.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.580
2025-11-12 14:30:30.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.531
2025-11-12 14:30:30.486 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.333
2025-11-12 14:30:30.486 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.482
2025-11-12 14:30:30.486 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:30:30.486 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:30:30.486 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:30:30.486 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:30:30.486 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:30:30.486 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:30:30.486 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:30:30.487 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:30:30.487 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:30:31.975 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:30:33.507 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:30:35.047 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:30:36.535 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:30:38.048 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:30:39.581 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:30:41.101 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:30:42.581 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:30:44.117 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:30:44.117 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.24
2025-11-12 14:30:44.117 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.48
2025-11-12 14:30:44.118 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:30:44.140 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.14 ms, Average NMS time: 0.57 ms, Average inference time: 2.70 ms

2025-11-12 14:30:44.142 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:30:44.218 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:30:44.299 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch26
2025-11-12 14:30:47.483 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 26/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.157s, data_time: 0.005s, total_loss: 6.2, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.8, lr: 1.791e-03, size: 480, ETA: 0:35:35
2025-11-12 14:30:50.999 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 26/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.174s, data_time: 0.028s, total_loss: 6.2, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.9, lr: 1.788e-03, size: 320, ETA: 0:35:31
2025-11-12 14:30:54.488 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 26/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.174s, data_time: 0.028s, total_loss: 6.1, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.8, lr: 1.786e-03, size: 480, ETA: 0:35:28
2025-11-12 14:30:58.101 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 26/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.177s, data_time: 0.014s, total_loss: 6.8, iou_loss: 3.0, l1_loss: 0.0, conf_loss: 3.0, cls_loss: 0.8, lr: 1.783e-03, size: 544, ETA: 0:35:24
2025-11-12 14:31:01.633 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 26/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.173s, data_time: 0.002s, total_loss: 6.6, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 3.0, cls_loss: 0.8, lr: 1.781e-03, size: 512, ETA: 0:35:21
2025-11-12 14:31:05.351 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 26/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.183s, data_time: 0.003s, total_loss: 5.7, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.4, cls_loss: 0.9, lr: 1.778e-03, size: 544, ETA: 0:35:18
2025-11-12 14:31:07.155 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:31:12.220 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:31:14.171 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:31:15.428 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6073
2025-11-12 14:31:15.692 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5576
2025-11-12 14:31:15.750 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3408
2025-11-12 14:31:15.750 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5019
2025-11-12 14:31:15.751 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:31:15.751 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:31:15.751 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.607
2025-11-12 14:31:15.751 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.558
2025-11-12 14:31:15.751 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.341
2025-11-12 14:31:15.751 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.502
2025-11-12 14:31:15.752 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:31:15.752 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:31:15.752 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:31:15.752 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:31:15.752 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:31:15.752 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:31:15.753 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:31:15.753 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:31:15.753 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:31:17.306 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:31:18.902 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:31:20.457 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:31:22.034 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:31:23.642 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:31:25.188 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:31:26.805 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:31:28.400 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:31:29.929 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:31:29.929 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.26
2025-11-12 14:31:29.929 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.50
2025-11-12 14:31:29.929 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:31:29.952 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.20 ms, Average NMS time: 0.58 ms, Average inference time: 2.79 ms

2025-11-12 14:31:29.953 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:31:30.028 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:31:30.147 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch27
2025-11-12 14:31:33.337 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 27/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.157s, data_time: 0.003s, total_loss: 6.7, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 3.0, cls_loss: 1.0, lr: 1.775e-03, size: 320, ETA: 0:35:12
2025-11-12 14:31:36.907 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 27/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.176s, data_time: 0.009s, total_loss: 6.1, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.8, lr: 1.772e-03, size: 384, ETA: 0:35:09
2025-11-12 14:31:40.392 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 27/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.173s, data_time: 0.016s, total_loss: 6.1, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.8, lr: 1.769e-03, size: 288, ETA: 0:35:05
2025-11-12 14:31:44.914 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 27/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.225s, data_time: 0.051s, total_loss: 6.3, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.9, lr: 1.767e-03, size: 352, ETA: 0:35:05
2025-11-12 14:31:48.603 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 27/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.181s, data_time: 0.027s, total_loss: 6.0, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.8, lr: 1.764e-03, size: 448, ETA: 0:35:02
2025-11-12 14:31:52.369 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 27/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.186s, data_time: 0.004s, total_loss: 6.7, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 3.0, cls_loss: 0.9, lr: 1.762e-03, size: 512, ETA: 0:34:59
2025-11-12 14:31:53.904 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:31:59.000 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:32:01.146 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:32:02.505 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6087
2025-11-12 14:32:02.751 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5761
2025-11-12 14:32:02.858 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3915
2025-11-12 14:32:02.859 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5254
2025-11-12 14:32:02.859 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:32:02.859 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:32:02.859 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.609
2025-11-12 14:32:02.859 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.576
2025-11-12 14:32:02.859 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.391
2025-11-12 14:32:02.859 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.525
2025-11-12 14:32:02.860 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:32:02.860 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:32:02.860 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:32:02.860 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:32:02.860 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:32:02.860 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:32:02.860 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:32:02.860 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:32:02.861 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:32:04.523 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:32:06.211 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:32:07.862 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:32:09.546 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:32:11.171 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:32:12.860 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:32:14.516 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:32:16.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:32:17.845 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:32:17.846 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.27
2025-11-12 14:32:17.846 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.53
2025-11-12 14:32:17.846 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:32:17.872 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.15 ms, Average NMS time: 0.57 ms, Average inference time: 2.72 ms

2025-11-12 14:32:17.874 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:32:17.949 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:32:18.034 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch28
2025-11-12 14:32:21.700 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 28/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.179s, data_time: 0.002s, total_loss: 7.6, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 3.7, cls_loss: 1.0, lr: 1.758e-03, size: 544, ETA: 0:34:54
2025-11-12 14:32:25.304 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 28/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.176s, data_time: 0.002s, total_loss: 6.3, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.9, cls_loss: 0.8, lr: 1.755e-03, size: 384, ETA: 0:34:51
2025-11-12 14:32:28.652 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 28/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.165s, data_time: 0.002s, total_loss: 6.5, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 3.0, cls_loss: 0.9, lr: 1.752e-03, size: 544, ETA: 0:34:47
2025-11-12 14:32:32.129 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 28/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.172s, data_time: 0.018s, total_loss: 6.6, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 3.1, cls_loss: 0.8, lr: 1.750e-03, size: 480, ETA: 0:34:43
2025-11-12 14:32:35.682 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 28/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.176s, data_time: 0.015s, total_loss: 6.2, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.7, lr: 1.747e-03, size: 320, ETA: 0:34:40
2025-11-12 14:32:39.147 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 28/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.171s, data_time: 0.012s, total_loss: 5.5, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.1, cls_loss: 0.8, lr: 1.744e-03, size: 320, ETA: 0:34:36
2025-11-12 14:32:40.751 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:32:45.823 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:32:48.487 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:32:50.262 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6136
2025-11-12 14:32:50.576 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5607
2025-11-12 14:32:50.634 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3630
2025-11-12 14:32:50.635 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5124
2025-11-12 14:32:50.635 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:32:50.635 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:32:50.635 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.614
2025-11-12 14:32:50.635 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.561
2025-11-12 14:32:50.636 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.363
2025-11-12 14:32:50.636 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.512
2025-11-12 14:32:50.636 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:32:50.636 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:32:50.636 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:32:50.636 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:32:50.636 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:32:50.636 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:32:50.636 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:32:50.637 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:32:50.637 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:32:52.840 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:32:55.026 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:32:57.173 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:32:59.300 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:33:01.469 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:33:03.622 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:33:05.750 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:33:07.870 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:33:09.988 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:33:09.989 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.27
2025-11-12 14:33:09.989 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.51
2025-11-12 14:33:09.989 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:33:10.014 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.18 ms, Average NMS time: 0.54 ms, Average inference time: 2.72 ms

2025-11-12 14:33:10.016 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:33:10.090 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:33:10.172 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch29
2025-11-12 14:33:13.433 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 29/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.160s, data_time: 0.006s, total_loss: 7.3, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 3.5, cls_loss: 1.0, lr: 1.740e-03, size: 384, ETA: 0:34:30
2025-11-12 14:33:17.090 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 29/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.180s, data_time: 0.008s, total_loss: 5.7, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.4, cls_loss: 0.8, lr: 1.738e-03, size: 576, ETA: 0:34:27
2025-11-12 14:33:20.571 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 29/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.171s, data_time: 0.002s, total_loss: 5.8, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 0.8, lr: 1.735e-03, size: 256, ETA: 0:34:23
2025-11-12 14:33:23.957 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 29/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.168s, data_time: 0.014s, total_loss: 6.0, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.9, lr: 1.732e-03, size: 320, ETA: 0:34:19
2025-11-12 14:33:27.474 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 29/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.174s, data_time: 0.020s, total_loss: 5.7, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.4, cls_loss: 0.8, lr: 1.729e-03, size: 384, ETA: 0:34:16
2025-11-12 14:33:31.033 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 29/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.177s, data_time: 0.015s, total_loss: 6.9, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 3.3, cls_loss: 0.9, lr: 1.727e-03, size: 576, ETA: 0:34:12
2025-11-12 14:33:32.644 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:33:37.698 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:33:40.155 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:33:41.738 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6182
2025-11-12 14:33:42.074 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5694
2025-11-12 14:33:42.144 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3579
2025-11-12 14:33:42.145 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5152
2025-11-12 14:33:42.145 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:33:42.145 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:33:42.145 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.618
2025-11-12 14:33:42.145 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.569
2025-11-12 14:33:42.145 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.358
2025-11-12 14:33:42.145 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.515
2025-11-12 14:33:42.145 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:33:42.145 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:33:42.146 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:33:42.146 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:33:42.146 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:33:42.146 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:33:42.146 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:33:42.146 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:33:42.146 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:33:44.114 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:33:46.093 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:33:48.056 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:33:50.007 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:33:51.981 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:33:53.917 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:33:55.845 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:33:57.784 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:33:59.723 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:33:59.724 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.27
2025-11-12 14:33:59.724 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.52
2025-11-12 14:33:59.724 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:33:59.749 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.16 ms, Average NMS time: 0.57 ms, Average inference time: 2.73 ms

2025-11-12 14:33:59.750 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:33:59.825 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:33:59.905 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch30
2025-11-12 14:34:03.353 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 30/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.170s, data_time: 0.001s, total_loss: 6.2, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.8, lr: 1.723e-03, size: 352, ETA: 0:34:07
2025-11-12 14:34:06.713 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 30/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.166s, data_time: 0.002s, total_loss: 5.8, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.8, lr: 1.720e-03, size: 384, ETA: 0:34:03
2025-11-12 14:34:10.176 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 30/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.170s, data_time: 0.003s, total_loss: 5.3, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.0, cls_loss: 0.7, lr: 1.717e-03, size: 320, ETA: 0:33:59
2025-11-12 14:34:13.664 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 30/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.171s, data_time: 0.002s, total_loss: 6.5, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.9, lr: 1.714e-03, size: 288, ETA: 0:33:55
2025-11-12 14:34:17.149 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 30/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.171s, data_time: 0.002s, total_loss: 6.5, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.9, cls_loss: 0.9, lr: 1.711e-03, size: 576, ETA: 0:33:52
2025-11-12 14:34:20.625 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 30/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.171s, data_time: 0.003s, total_loss: 6.3, iou_loss: 3.0, l1_loss: 0.0, conf_loss: 2.4, cls_loss: 0.9, lr: 1.708e-03, size: 320, ETA: 0:33:48
2025-11-12 14:34:22.140 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:34:27.304 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:34:30.319 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:34:32.283 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6263
2025-11-12 14:34:32.692 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5549
2025-11-12 14:34:32.806 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3323
2025-11-12 14:34:32.807 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5045
2025-11-12 14:34:32.807 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:34:32.807 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:34:32.807 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.626
2025-11-12 14:34:32.807 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.555
2025-11-12 14:34:32.807 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.332
2025-11-12 14:34:32.807 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.504
2025-11-12 14:34:32.807 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:34:32.808 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:34:32.808 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:34:32.808 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:34:32.808 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:34:32.808 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:34:32.808 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:34:32.808 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:34:32.808 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:34:35.262 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:34:37.748 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:34:40.237 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:34:42.695 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:34:45.179 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:34:47.636 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:34:50.116 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:34:52.566 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:34:55.108 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:34:55.109 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.25
2025-11-12 14:34:55.109 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.50
2025-11-12 14:34:55.110 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:34:55.140 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.19 ms, Average NMS time: 0.58 ms, Average inference time: 2.77 ms

2025-11-12 14:34:55.142 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:34:55.218 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:34:55.300 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch31
2025-11-12 14:34:58.520 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 31/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.160s, data_time: 0.016s, total_loss: 6.1, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.4, cls_loss: 1.1, lr: 1.704e-03, size: 320, ETA: 0:33:42
2025-11-12 14:35:02.018 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 31/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.174s, data_time: 0.033s, total_loss: 5.9, iou_loss: 2.3, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.9, lr: 1.701e-03, size: 352, ETA: 0:33:38
2025-11-12 14:35:05.465 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 31/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.171s, data_time: 0.012s, total_loss: 6.3, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.9, lr: 1.698e-03, size: 256, ETA: 0:33:35
2025-11-12 14:35:09.044 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 31/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.176s, data_time: 0.004s, total_loss: 5.9, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.8, lr: 1.696e-03, size: 512, ETA: 0:33:31
2025-11-12 14:35:12.507 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 31/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.170s, data_time: 0.012s, total_loss: 6.0, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.9, lr: 1.693e-03, size: 448, ETA: 0:33:28
2025-11-12 14:35:16.015 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 31/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.174s, data_time: 0.016s, total_loss: 5.3, iou_loss: 2.4, l1_loss: 0.0, conf_loss: 2.1, cls_loss: 0.8, lr: 1.690e-03, size: 384, ETA: 0:33:24
2025-11-12 14:35:17.616 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:35:22.899 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:35:25.060 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:35:26.445 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6119
2025-11-12 14:35:26.738 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5808
2025-11-12 14:35:26.795 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3572
2025-11-12 14:35:26.796 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5166
2025-11-12 14:35:26.796 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:35:26.796 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:35:26.797 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.612
2025-11-12 14:35:26.797 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.581
2025-11-12 14:35:26.797 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.357
2025-11-12 14:35:26.797 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.517
2025-11-12 14:35:26.797 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:35:26.797 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:35:26.798 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:35:26.798 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:35:26.798 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:35:26.798 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:35:26.798 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:35:26.798 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:35:26.799 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:35:28.579 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:35:30.278 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:35:32.023 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:35:33.788 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:35:35.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:35:37.262 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:35:38.947 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:35:40.692 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:35:42.447 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:35:42.447 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.25
2025-11-12 14:35:42.447 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.52
2025-11-12 14:35:42.448 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:35:42.472 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.20 ms, Average NMS time: 0.58 ms, Average inference time: 2.78 ms

2025-11-12 14:35:42.474 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:35:42.549 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:35:42.631 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch32
2025-11-12 14:35:46.069 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 32/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.170s, data_time: 0.002s, total_loss: 6.1, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 0.9, lr: 1.685e-03, size: 288, ETA: 0:33:19
2025-11-12 14:35:49.408 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 32/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.164s, data_time: 0.006s, total_loss: 6.3, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.9, lr: 1.682e-03, size: 512, ETA: 0:33:15
2025-11-12 14:35:52.952 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 32/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.177s, data_time: 0.020s, total_loss: 6.9, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 3.1, cls_loss: 0.9, lr: 1.679e-03, size: 352, ETA: 0:33:11
2025-11-12 14:35:56.481 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 32/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.176s, data_time: 0.026s, total_loss: 6.5, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 3.0, cls_loss: 0.8, lr: 1.676e-03, size: 256, ETA: 0:33:08
2025-11-12 14:35:59.877 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 32/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.169s, data_time: 0.014s, total_loss: 6.4, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.9, lr: 1.673e-03, size: 320, ETA: 0:33:04
2025-11-12 14:36:03.440 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 32/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.176s, data_time: 0.021s, total_loss: 6.2, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.9, lr: 1.670e-03, size: 320, ETA: 0:33:01
2025-11-12 14:36:05.118 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:36:10.218 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:36:13.152 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:36:15.191 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6121
2025-11-12 14:36:15.517 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5618
2025-11-12 14:36:15.584 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3844
2025-11-12 14:36:15.585 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5194
2025-11-12 14:36:15.585 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:36:15.585 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:36:15.585 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.612
2025-11-12 14:36:15.585 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.562
2025-11-12 14:36:15.585 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.384
2025-11-12 14:36:15.585 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.519
2025-11-12 14:36:15.586 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:36:15.586 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:36:15.586 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:36:15.586 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:36:15.586 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:36:15.586 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:36:15.586 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:36:15.586 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:36:15.587 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:36:17.950 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:36:20.428 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:36:22.807 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:36:25.175 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:36:27.663 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:36:30.117 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:36:32.704 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:36:35.197 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:36:37.611 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:36:37.612 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.28
2025-11-12 14:36:37.612 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.52
2025-11-12 14:36:37.612 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:36:37.643 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.13 ms, Average NMS time: 0.59 ms, Average inference time: 2.72 ms

2025-11-12 14:36:37.644 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:36:37.720 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:36:37.803 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch33
2025-11-12 14:36:41.020 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 33/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.159s, data_time: 0.005s, total_loss: 7.1, iou_loss: 3.1, l1_loss: 0.0, conf_loss: 3.1, cls_loss: 0.9, lr: 1.666e-03, size: 384, ETA: 0:32:55
2025-11-12 14:36:44.622 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 33/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.178s, data_time: 0.013s, total_loss: 6.4, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 3.0, cls_loss: 0.8, lr: 1.663e-03, size: 576, ETA: 0:32:52
2025-11-12 14:36:48.045 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 33/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.169s, data_time: 0.010s, total_loss: 6.6, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.9, lr: 1.660e-03, size: 256, ETA: 0:32:48
2025-11-12 14:36:51.652 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 33/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.177s, data_time: 0.010s, total_loss: 6.8, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 3.1, cls_loss: 0.8, lr: 1.657e-03, size: 544, ETA: 0:32:45
2025-11-12 14:36:55.227 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 33/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.176s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.4, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.8, lr: 1.654e-03, size: 512, ETA: 0:32:41
2025-11-12 14:36:58.628 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 33/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.167s, data_time: 0.005s, total_loss: 5.8, iou_loss: 2.4, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 1.1, lr: 1.651e-03, size: 288, ETA: 0:32:37
2025-11-12 14:37:00.143 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:37:05.293 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:37:07.910 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:37:09.694 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6121
2025-11-12 14:37:09.994 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5453
2025-11-12 14:37:10.054 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3506
2025-11-12 14:37:10.055 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5027
2025-11-12 14:37:10.055 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:37:10.055 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:37:10.055 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.612
2025-11-12 14:37:10.055 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.545
2025-11-12 14:37:10.055 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.351
2025-11-12 14:37:10.055 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.503
2025-11-12 14:37:10.055 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:37:10.055 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:37:10.056 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:37:10.056 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:37:10.056 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:37:10.056 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:37:10.056 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:37:10.056 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:37:10.056 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:37:12.213 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:37:14.408 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:37:16.579 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:37:18.734 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:37:20.903 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:37:23.037 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:37:25.169 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:37:27.308 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:37:29.471 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:37:29.471 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.26
2025-11-12 14:37:29.471 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.50
2025-11-12 14:37:29.472 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:37:29.497 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.24 ms, Average NMS time: 0.65 ms, Average inference time: 2.90 ms

2025-11-12 14:37:29.498 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:37:29.573 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:37:29.654 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch34
2025-11-12 14:37:32.982 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 34/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.163s, data_time: 0.002s, total_loss: 6.1, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.7, lr: 1.646e-03, size: 512, ETA: 0:32:32
2025-11-12 14:37:36.572 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 34/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.176s, data_time: 0.002s, total_loss: 5.8, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.7, lr: 1.643e-03, size: 288, ETA: 0:32:28
2025-11-12 14:37:39.942 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 34/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.165s, data_time: 0.002s, total_loss: 5.4, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.2, cls_loss: 0.8, lr: 1.640e-03, size: 320, ETA: 0:32:24
2025-11-12 14:37:43.507 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 34/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.176s, data_time: 0.002s, total_loss: 6.7, iou_loss: 3.0, l1_loss: 0.0, conf_loss: 2.9, cls_loss: 0.8, lr: 1.637e-03, size: 256, ETA: 0:32:21
2025-11-12 14:37:47.050 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 34/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.174s, data_time: 0.002s, total_loss: 6.5, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.9, lr: 1.634e-03, size: 320, ETA: 0:32:17
2025-11-12 14:37:50.448 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 34/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.167s, data_time: 0.002s, total_loss: 5.8, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 0.8, lr: 1.631e-03, size: 352, ETA: 0:32:14
2025-11-12 14:37:52.035 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:37:57.087 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:37:59.655 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:38:01.318 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6071
2025-11-12 14:38:01.723 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5644
2025-11-12 14:38:01.791 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3600
2025-11-12 14:38:01.792 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5105
2025-11-12 14:38:01.792 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:38:01.792 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:38:01.792 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.607
2025-11-12 14:38:01.792 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.564
2025-11-12 14:38:01.792 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.360
2025-11-12 14:38:01.792 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.510
2025-11-12 14:38:01.793 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:38:01.793 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:38:01.793 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:38:01.793 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:38:01.793 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:38:01.793 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:38:01.793 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:38:01.793 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:38:01.794 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:38:03.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:38:05.972 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:38:08.058 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:38:10.102 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:38:12.172 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:38:14.266 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:38:16.327 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:38:18.385 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:38:20.430 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:38:20.431 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.27
2025-11-12 14:38:20.431 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.51
2025-11-12 14:38:20.431 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:38:20.455 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.13 ms, Average NMS time: 0.57 ms, Average inference time: 2.70 ms

2025-11-12 14:38:20.457 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:38:20.531 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:38:20.613 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch35
2025-11-12 14:38:23.764 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 35/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.157s, data_time: 0.010s, total_loss: 5.6, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 0.8, lr: 1.626e-03, size: 416, ETA: 0:32:08
2025-11-12 14:38:27.261 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 35/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.172s, data_time: 0.017s, total_loss: 6.5, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.9, lr: 1.623e-03, size: 416, ETA: 0:32:04
2025-11-12 14:38:30.779 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 35/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.173s, data_time: 0.014s, total_loss: 6.6, iou_loss: 3.1, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.8, lr: 1.620e-03, size: 256, ETA: 0:32:00
2025-11-12 14:38:34.371 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 35/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.177s, data_time: 0.020s, total_loss: 6.4, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 1.0, lr: 1.617e-03, size: 288, ETA: 0:31:57
2025-11-12 14:38:37.817 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 35/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.170s, data_time: 0.018s, total_loss: 6.4, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.9, lr: 1.613e-03, size: 416, ETA: 0:31:53
2025-11-12 14:38:41.296 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 35/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.172s, data_time: 0.027s, total_loss: 5.7, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.4, cls_loss: 0.8, lr: 1.610e-03, size: 320, ETA: 0:31:50
2025-11-12 14:38:42.895 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:38:47.980 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:38:51.863 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:38:54.725 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6124
2025-11-12 14:38:55.097 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5550
2025-11-12 14:38:55.183 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3690
2025-11-12 14:38:55.184 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5121
2025-11-12 14:38:55.184 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:38:55.184 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:38:55.184 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.612
2025-11-12 14:38:55.184 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.555
2025-11-12 14:38:55.184 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.369
2025-11-12 14:38:55.185 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.512
2025-11-12 14:38:55.185 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:38:55.185 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:38:55.185 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:38:55.185 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:38:55.185 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:38:55.185 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:38:55.185 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:38:55.185 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:38:55.186 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:38:58.392 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:39:01.580 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:39:04.858 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:39:08.147 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:39:11.427 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:39:14.660 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:39:17.871 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:39:21.095 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:39:24.314 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:39:24.315 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.27
2025-11-12 14:39:24.315 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.51
2025-11-12 14:39:24.315 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:39:24.342 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.16 ms, Average NMS time: 0.64 ms, Average inference time: 2.80 ms

2025-11-12 14:39:24.343 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:39:24.418 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:39:24.500 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch36
2025-11-12 14:39:27.714 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 36/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.157s, data_time: 0.002s, total_loss: 6.7, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 3.1, cls_loss: 0.9, lr: 1.606e-03, size: 416, ETA: 0:31:44
2025-11-12 14:39:31.292 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 36/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.176s, data_time: 0.007s, total_loss: 6.7, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 1.1, lr: 1.602e-03, size: 288, ETA: 0:31:41
2025-11-12 14:39:34.839 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 36/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.3Gb, iter_time: 0.174s, data_time: 0.005s, total_loss: 5.3, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.2, cls_loss: 0.7, lr: 1.599e-03, size: 480, ETA: 0:31:37
2025-11-12 14:39:38.535 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 36/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.181s, data_time: 0.002s, total_loss: 6.1, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.8, lr: 1.596e-03, size: 512, ETA: 0:31:34
2025-11-12 14:39:42.345 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 36/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.3Gb, iter_time: 0.186s, data_time: 0.002s, total_loss: 6.1, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.9, lr: 1.593e-03, size: 352, ETA: 0:31:31
2025-11-12 14:39:46.058 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 36/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.182s, data_time: 0.003s, total_loss: 6.7, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 3.0, cls_loss: 0.8, lr: 1.589e-03, size: 416, ETA: 0:31:28
2025-11-12 14:39:47.737 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:39:52.721 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:39:54.953 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:39:56.475 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6169
2025-11-12 14:39:56.714 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5489
2025-11-12 14:39:56.818 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3865
2025-11-12 14:39:56.818 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5174
2025-11-12 14:39:56.819 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:39:56.819 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:39:56.819 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.617
2025-11-12 14:39:56.819 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.549
2025-11-12 14:39:56.819 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.386
2025-11-12 14:39:56.819 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.517
2025-11-12 14:39:56.819 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:39:56.819 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:39:56.820 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:39:56.820 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:39:56.820 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:39:56.820 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:39:56.820 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:39:56.820 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:39:56.820 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:39:58.586 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:40:00.469 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:40:02.291 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:40:04.070 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:40:05.897 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:40:07.768 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:40:09.573 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:40:11.392 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:40:13.248 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:40:13.248 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.27
2025-11-12 14:40:13.249 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.52
2025-11-12 14:40:13.249 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:40:13.273 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.14 ms, Average NMS time: 0.58 ms, Average inference time: 2.73 ms

2025-11-12 14:40:13.275 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:40:13.350 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:40:13.433 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch37
2025-11-12 14:40:16.621 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 37/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.158s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.8, lr: 1.584e-03, size: 448, ETA: 0:31:22
2025-11-12 14:40:20.070 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 37/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.171s, data_time: 0.022s, total_loss: 5.8, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 0.9, lr: 1.581e-03, size: 352, ETA: 0:31:19
2025-11-12 14:40:23.535 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 37/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.3Gb, iter_time: 0.170s, data_time: 0.020s, total_loss: 6.7, iou_loss: 3.1, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.8, lr: 1.578e-03, size: 320, ETA: 0:31:15
2025-11-12 14:40:27.212 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 37/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.180s, data_time: 0.006s, total_loss: 5.7, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.4, cls_loss: 0.8, lr: 1.575e-03, size: 512, ETA: 0:31:12
2025-11-12 14:40:30.956 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 37/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.184s, data_time: 0.002s, total_loss: 6.0, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.4, cls_loss: 0.8, lr: 1.571e-03, size: 320, ETA: 0:31:09
2025-11-12 14:40:34.549 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 37/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.177s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.6, cls_loss: 0.8, lr: 1.568e-03, size: 448, ETA: 0:31:05
2025-11-12 14:40:36.211 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:40:41.281 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:40:43.815 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:40:45.558 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6087
2025-11-12 14:40:45.823 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5587
2025-11-12 14:40:45.922 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3978
2025-11-12 14:40:45.923 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5217
2025-11-12 14:40:45.923 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:40:45.923 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:40:45.923 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.609
2025-11-12 14:40:45.923 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.559
2025-11-12 14:40:45.923 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.398
2025-11-12 14:40:45.924 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.522
2025-11-12 14:40:45.924 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:40:45.924 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:40:45.924 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:40:45.924 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:40:45.924 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:40:45.924 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:40:45.924 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:40:45.924 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:40:45.925 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:40:48.044 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:40:50.114 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:40:52.200 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:40:54.262 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:40:56.366 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:40:58.433 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:41:00.512 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:41:02.601 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:41:04.710 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:41:04.710 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.26
2025-11-12 14:41:04.710 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.52
2025-11-12 14:41:04.711 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:41:04.735 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.18 ms, Average NMS time: 0.60 ms, Average inference time: 2.78 ms

2025-11-12 14:41:04.736 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:41:04.811 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:41:04.892 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch38
2025-11-12 14:41:08.172 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 38/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.161s, data_time: 0.002s, total_loss: 6.2, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.9, lr: 1.563e-03, size: 544, ETA: 0:31:00
2025-11-12 14:41:11.796 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 38/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.178s, data_time: 0.002s, total_loss: 6.3, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.8, lr: 1.560e-03, size: 416, ETA: 0:30:57
2025-11-12 14:41:15.187 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 38/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.168s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.4, cls_loss: 0.8, lr: 1.556e-03, size: 320, ETA: 0:30:53
2025-11-12 14:41:18.574 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 38/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.169s, data_time: 0.003s, total_loss: 5.6, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.5, cls_loss: 0.7, lr: 1.553e-03, size: 352, ETA: 0:30:49
2025-11-12 14:41:22.148 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 38/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.178s, data_time: 0.023s, total_loss: 6.3, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.8, cls_loss: 0.8, lr: 1.550e-03, size: 512, ETA: 0:30:46
2025-11-12 14:41:25.677 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 38/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.176s, data_time: 0.023s, total_loss: 5.6, iou_loss: 2.5, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 0.8, lr: 1.546e-03, size: 512, ETA: 0:30:42
2025-11-12 14:41:27.220 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:41:32.220 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:41:34.557 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:41:35.967 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6063
2025-11-12 14:41:36.291 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5602
2025-11-12 14:41:36.345 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3399
2025-11-12 14:41:36.345 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5021
2025-11-12 14:41:36.345 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:41:36.345 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:41:36.346 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.606
2025-11-12 14:41:36.346 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.560
2025-11-12 14:41:36.346 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.340
2025-11-12 14:41:36.346 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.502
2025-11-12 14:41:36.346 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:41:36.346 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:41:36.346 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:41:36.346 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:41:36.346 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:41:36.346 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:41:36.347 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:41:36.347 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:41:36.347 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:41:38.195 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:41:40.023 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:41:41.780 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:41:43.610 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:41:45.440 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:41:47.299 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:41:49.062 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:41:50.871 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:41:52.736 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:41:52.736 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.27
2025-11-12 14:41:52.736 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.50
2025-11-12 14:41:52.736 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:41:52.761 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.14 ms, Average NMS time: 0.58 ms, Average inference time: 2.71 ms

2025-11-12 14:41:52.762 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:41:52.839 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:41:52.941 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch39
2025-11-12 14:41:56.301 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 39/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.3Gb, iter_time: 0.165s, data_time: 0.001s, total_loss: 6.1, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.8, lr: 1.541e-03, size: 416, ETA: 0:30:37
2025-11-12 14:41:59.643 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 39/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.165s, data_time: 0.013s, total_loss: 5.8, iou_loss: 2.7, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 0.8, lr: 1.538e-03, size: 352, ETA: 0:30:33
2025-11-12 14:42:03.315 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 39/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.3Gb, iter_time: 0.183s, data_time: 0.024s, total_loss: 6.1, iou_loss: 2.9, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 0.8, lr: 1.534e-03, size: 384, ETA: 0:30:30
2025-11-12 14:42:06.859 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 39/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.3Gb, iter_time: 0.176s, data_time: 0.021s, total_loss: 6.3, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.7, cls_loss: 0.8, lr: 1.531e-03, size: 288, ETA: 0:30:27
2025-11-12 14:42:10.277 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 39/120, iter: 100/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.170s, data_time: 0.027s, total_loss: 5.8, iou_loss: 2.6, l1_loss: 0.0, conf_loss: 2.4, cls_loss: 0.8, lr: 1.528e-03, size: 288, ETA: 0:30:23
2025-11-12 14:42:13.760 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 39/120, iter: 120/129, gpu mem: 2422Mb, mem: 95.2Gb, iter_time: 0.172s, data_time: 0.023s, total_loss: 6.1, iou_loss: 2.8, l1_loss: 0.0, conf_loss: 2.3, cls_loss: 0.9, lr: 1.524e-03, size: 544, ETA: 0:30:19
2025-11-12 14:42:15.326 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:42:20.389 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:42:22.054 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:42:23.032 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6135
2025-11-12 14:42:23.319 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5348
2025-11-12 14:42:23.367 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3480
2025-11-12 14:42:23.367 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.4988
2025-11-12 14:42:23.368 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:42:23.368 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:42:23.368 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.613
2025-11-12 14:42:23.368 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.535
2025-11-12 14:42:23.368 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.348
2025-11-12 14:42:23.368 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.499
2025-11-12 14:42:23.368 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:42:23.368 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:42:23.368 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:42:23.369 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:42:23.369 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:42:23.369 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:42:23.369 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:42:23.369 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:42:23.369 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:42:24.669 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:42:25.949 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:42:27.245 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:42:28.538 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:42:29.864 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:42:31.242 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:42:32.545 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:42:33.861 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:42:35.176 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:42:35.176 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.26
2025-11-12 14:42:35.176 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.50
2025-11-12 14:42:35.177 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:42:35.184 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.15 ms, Average NMS time: 0.58 ms, Average inference time: 2.73 ms

2025-11-12 14:42:35.186 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:42:35.262 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:42:35.345 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch40
2025-11-12 14:42:35.346 | INFO     | yolox_microbt.core.trainer:before_epoch:208 - --->No mosaic aug now!
2025-11-12 14:42:35.346 | INFO     | yolox_microbt.core.trainer:before_epoch:210 - --->Add additional L1 loss now!
2025-11-12 14:42:35.346 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:42:38.370 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 40/120, iter: 20/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.145s, data_time: 0.003s, total_loss: 5.0, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 1.4, cls_loss: 0.6, lr: 1.519e-03, size: 288, ETA: 0:30:13
2025-11-12 14:42:41.350 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 40/120, iter: 40/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 5.0, iou_loss: 2.0, l1_loss: 0.9, conf_loss: 1.6, cls_loss: 0.5, lr: 1.516e-03, size: 544, ETA: 0:30:08
2025-11-12 14:42:44.296 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 40/120, iter: 60/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 6.2, iou_loss: 1.9, l1_loss: 1.0, conf_loss: 2.8, cls_loss: 0.6, lr: 1.512e-03, size: 512, ETA: 0:30:04
2025-11-12 14:42:47.416 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 40/120, iter: 80/129, gpu mem: 2422Mb, mem: 95.1Gb, iter_time: 0.154s, data_time: 0.002s, total_loss: 7.3, iou_loss: 2.5, l1_loss: 1.0, conf_loss: 2.9, cls_loss: 0.8, lr: 1.509e-03, size: 576, ETA: 0:30:00
2025-11-12 14:42:50.346 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 40/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.144s, data_time: 0.002s, total_loss: 5.8, iou_loss: 2.3, l1_loss: 1.0, conf_loss: 1.9, cls_loss: 0.6, lr: 1.505e-03, size: 480, ETA: 0:29:55
2025-11-12 14:42:53.240 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 40/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.142s, data_time: 0.003s, total_loss: 5.3, iou_loss: 2.2, l1_loss: 0.6, conf_loss: 1.6, cls_loss: 0.9, lr: 1.502e-03, size: 256, ETA: 0:29:50
2025-11-12 14:42:54.470 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:42:59.645 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:43:01.096 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:43:02.040 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6268
2025-11-12 14:43:02.283 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5744
2025-11-12 14:43:02.336 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3955
2025-11-12 14:43:02.337 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5322
2025-11-12 14:43:02.337 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:43:02.337 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:43:02.337 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.627
2025-11-12 14:43:02.337 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.574
2025-11-12 14:43:02.337 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.396
2025-11-12 14:43:02.337 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.532
2025-11-12 14:43:02.338 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:43:02.338 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:43:02.338 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:43:02.338 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:43:02.338 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:43:02.338 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:43:02.338 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:43:02.338 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:43:02.338 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:43:03.570 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:43:04.814 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:43:06.035 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:43:07.228 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:43:08.443 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:43:09.668 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:43:10.887 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:43:12.102 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:43:13.325 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:43:13.326 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.29
2025-11-12 14:43:13.326 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.53
2025-11-12 14:43:13.326 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:43:13.334 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.22 ms, Average NMS time: 0.59 ms, Average inference time: 2.82 ms

2025-11-12 14:43:13.335 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:43:13.410 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:43:13.492 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch41
2025-11-12 14:43:16.521 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 41/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.149s, data_time: 0.001s, total_loss: 4.4, iou_loss: 1.5, l1_loss: 0.5, conf_loss: 2.0, cls_loss: 0.4, lr: 1.496e-03, size: 384, ETA: 0:29:43
2025-11-12 14:43:19.663 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 41/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.153s, data_time: 0.002s, total_loss: 5.0, iou_loss: 2.0, l1_loss: 0.8, conf_loss: 1.6, cls_loss: 0.7, lr: 1.493e-03, size: 448, ETA: 0:29:39
2025-11-12 14:43:22.926 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 41/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.160s, data_time: 0.002s, total_loss: 7.4, iou_loss: 3.0, l1_loss: 1.0, conf_loss: 2.7, cls_loss: 0.7, lr: 1.489e-03, size: 288, ETA: 0:29:35
2025-11-12 14:43:26.208 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 41/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.160s, data_time: 0.002s, total_loss: 5.1, iou_loss: 2.0, l1_loss: 0.8, conf_loss: 1.7, cls_loss: 0.7, lr: 1.486e-03, size: 480, ETA: 0:29:31
2025-11-12 14:43:29.440 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 41/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.157s, data_time: 0.003s, total_loss: 6.4, iou_loss: 2.5, l1_loss: 0.9, conf_loss: 2.4, cls_loss: 0.6, lr: 1.482e-03, size: 288, ETA: 0:29:27
2025-11-12 14:43:32.663 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 41/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.157s, data_time: 0.002s, total_loss: 6.5, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 2.7, cls_loss: 0.7, lr: 1.479e-03, size: 448, ETA: 0:29:23
2025-11-12 14:43:34.048 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:43:39.164 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:43:40.524 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:43:41.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6128
2025-11-12 14:43:41.616 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5588
2025-11-12 14:43:41.695 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3775
2025-11-12 14:43:41.696 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5164
2025-11-12 14:43:41.696 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:43:41.696 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:43:41.696 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.613
2025-11-12 14:43:41.697 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.559
2025-11-12 14:43:41.697 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.378
2025-11-12 14:43:41.697 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.516
2025-11-12 14:43:41.697 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:43:41.697 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:43:41.697 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:43:41.697 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:43:41.697 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:43:41.697 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:43:41.697 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:43:41.698 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:43:41.698 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:43:42.875 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:43:44.001 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:43:45.155 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:43:46.310 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:43:47.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:43:48.663 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:43:49.783 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:43:50.940 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:43:52.083 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:43:52.083 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.27
2025-11-12 14:43:52.083 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.52
2025-11-12 14:43:52.083 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:43:52.090 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.18 ms, Average NMS time: 0.56 ms, Average inference time: 2.75 ms

2025-11-12 14:43:52.092 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:43:52.167 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:43:52.248 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch42
2025-11-12 14:43:55.576 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 42/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.164s, data_time: 0.002s, total_loss: 5.2, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 1.7, cls_loss: 0.6, lr: 1.474e-03, size: 480, ETA: 0:29:17
2025-11-12 14:43:58.868 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 42/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.161s, data_time: 0.002s, total_loss: 7.0, iou_loss: 2.7, l1_loss: 0.9, conf_loss: 2.7, cls_loss: 0.7, lr: 1.470e-03, size: 256, ETA: 0:29:13
2025-11-12 14:44:02.234 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 42/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.164s, data_time: 0.002s, total_loss: 5.2, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.9, cls_loss: 0.7, lr: 1.466e-03, size: 416, ETA: 0:29:10
2025-11-12 14:44:05.624 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 42/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.166s, data_time: 0.002s, total_loss: 4.6, iou_loss: 1.7, l1_loss: 0.7, conf_loss: 1.6, cls_loss: 0.6, lr: 1.463e-03, size: 384, ETA: 0:29:06
2025-11-12 14:44:09.004 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 42/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.165s, data_time: 0.002s, total_loss: 5.8, iou_loss: 2.4, l1_loss: 0.9, conf_loss: 1.8, cls_loss: 0.7, lr: 1.459e-03, size: 544, ETA: 0:29:02
2025-11-12 14:44:12.310 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 42/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.161s, data_time: 0.001s, total_loss: 5.3, iou_loss: 2.0, l1_loss: 0.8, conf_loss: 1.8, cls_loss: 0.6, lr: 1.456e-03, size: 416, ETA: 0:28:58
2025-11-12 14:44:13.686 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:44:18.689 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:44:19.510 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:44:20.078 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6417
2025-11-12 14:44:20.178 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5698
2025-11-12 14:44:20.215 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4151
2025-11-12 14:44:20.216 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5422
2025-11-12 14:44:20.216 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:44:20.216 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:44:20.216 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.642
2025-11-12 14:44:20.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.570
2025-11-12 14:44:20.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.415
2025-11-12 14:44:20.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.542
2025-11-12 14:44:20.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:44:20.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:44:20.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:44:20.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:44:20.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:44:20.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:44:20.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:44:20.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:44:20.218 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:44:20.932 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:44:21.604 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:44:22.283 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:44:23.000 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:44:23.683 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:44:24.371 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:44:25.045 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:44:25.753 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:44:26.424 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:44:26.425 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.30
2025-11-12 14:44:26.425 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.54
2025-11-12 14:44:26.425 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:44:26.432 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.18 ms, Average NMS time: 0.55 ms, Average inference time: 2.73 ms

2025-11-12 14:44:26.433 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:44:26.511 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:44:26.591 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch43
2025-11-12 14:44:29.507 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 43/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.144s, data_time: 0.002s, total_loss: 5.4, iou_loss: 2.2, l1_loss: 0.7, conf_loss: 1.9, cls_loss: 0.7, lr: 1.450e-03, size: 288, ETA: 0:28:52
2025-11-12 14:44:32.335 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 43/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.138s, data_time: 0.004s, total_loss: 6.1, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 2.2, cls_loss: 0.7, lr: 1.447e-03, size: 352, ETA: 0:28:47
2025-11-12 14:44:35.324 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 43/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 7.5, iou_loss: 2.7, l1_loss: 1.0, conf_loss: 3.0, cls_loss: 0.8, lr: 1.443e-03, size: 288, ETA: 0:28:43
2025-11-12 14:44:38.334 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 43/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.148s, data_time: 0.003s, total_loss: 6.3, iou_loss: 2.6, l1_loss: 0.8, conf_loss: 2.2, cls_loss: 0.7, lr: 1.439e-03, size: 288, ETA: 0:28:38
2025-11-12 14:44:41.345 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 43/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.149s, data_time: 0.004s, total_loss: 6.3, iou_loss: 2.4, l1_loss: 0.9, conf_loss: 2.3, cls_loss: 0.6, lr: 1.436e-03, size: 512, ETA: 0:28:34
2025-11-12 14:44:44.377 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 43/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 5.8, iou_loss: 2.4, l1_loss: 0.8, conf_loss: 2.0, cls_loss: 0.6, lr: 1.432e-03, size: 512, ETA: 0:28:30
2025-11-12 14:44:45.646 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:44:50.789 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:44:52.558 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:44:53.721 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6299
2025-11-12 14:44:54.019 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5632
2025-11-12 14:44:54.065 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3867
2025-11-12 14:44:54.066 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5266
2025-11-12 14:44:54.066 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:44:54.066 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:44:54.066 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.630
2025-11-12 14:44:54.067 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.563
2025-11-12 14:44:54.067 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.387
2025-11-12 14:44:54.067 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.527
2025-11-12 14:44:54.067 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:44:54.067 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:44:54.067 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:44:54.067 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:44:54.067 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:44:54.067 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:44:54.067 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:44:54.068 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:44:54.068 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:44:55.615 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:44:57.125 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:44:58.608 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:45:00.077 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:45:01.535 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:45:03.046 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:45:04.536 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:45:06.041 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:45:07.502 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:45:07.502 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.27
2025-11-12 14:45:07.503 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.53
2025-11-12 14:45:07.503 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:45:07.526 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.28 ms, Average NMS time: 0.58 ms, Average inference time: 2.87 ms

2025-11-12 14:45:07.527 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:45:07.602 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:45:07.684 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch44
2025-11-12 14:45:10.461 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 44/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.136s, data_time: 0.003s, total_loss: 6.7, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 2.6, cls_loss: 0.8, lr: 1.427e-03, size: 352, ETA: 0:28:23
2025-11-12 14:45:13.360 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 44/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.142s, data_time: 0.003s, total_loss: 4.4, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.1, cls_loss: 0.5, lr: 1.423e-03, size: 320, ETA: 0:28:18
2025-11-12 14:45:16.244 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 44/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.143s, data_time: 0.004s, total_loss: 6.2, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 2.3, cls_loss: 0.7, lr: 1.419e-03, size: 480, ETA: 0:28:14
2025-11-12 14:45:19.074 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 44/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.138s, data_time: 0.003s, total_loss: 5.8, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 2.1, cls_loss: 0.6, lr: 1.416e-03, size: 352, ETA: 0:28:09
2025-11-12 14:45:22.051 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 44/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 5.3, iou_loss: 2.3, l1_loss: 0.8, conf_loss: 1.6, cls_loss: 0.6, lr: 1.412e-03, size: 480, ETA: 0:28:05
2025-11-12 14:45:24.953 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 44/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.143s, data_time: 0.003s, total_loss: 7.8, iou_loss: 2.7, l1_loss: 1.1, conf_loss: 3.1, cls_loss: 0.9, lr: 1.408e-03, size: 320, ETA: 0:28:01
2025-11-12 14:45:26.217 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:45:31.334 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:45:32.344 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:45:33.010 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6283
2025-11-12 14:45:33.137 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5877
2025-11-12 14:45:33.216 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4503
2025-11-12 14:45:33.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5554
2025-11-12 14:45:33.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:45:33.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:45:33.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.628
2025-11-12 14:45:33.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.588
2025-11-12 14:45:33.218 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.450
2025-11-12 14:45:33.218 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.555
2025-11-12 14:45:33.218 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:45:33.218 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:45:33.218 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:45:33.219 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:45:33.219 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:45:33.219 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:45:33.219 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:45:33.219 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:45:33.219 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:45:34.037 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:45:34.884 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:45:35.693 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:45:36.540 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:45:37.570 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:45:38.755 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:45:39.607 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:45:40.431 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:45:41.362 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:45:41.362 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 14:45:41.362 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 14:45:41.362 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:45:41.369 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.20 ms, Average NMS time: 0.58 ms, Average inference time: 2.78 ms

2025-11-12 14:45:41.372 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:45:41.450 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:45:41.532 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch45
2025-11-12 14:45:44.403 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 45/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.143s, data_time: 0.002s, total_loss: 6.0, iou_loss: 2.4, l1_loss: 0.9, conf_loss: 2.1, cls_loss: 0.6, lr: 1.403e-03, size: 448, ETA: 0:27:54
2025-11-12 14:45:47.388 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 45/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.148s, data_time: 0.002s, total_loss: 5.4, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 1.8, cls_loss: 0.6, lr: 1.399e-03, size: 416, ETA: 0:27:50
2025-11-12 14:45:50.387 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 45/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.146s, data_time: 0.003s, total_loss: 6.6, iou_loss: 2.2, l1_loss: 1.1, conf_loss: 2.6, cls_loss: 0.6, lr: 1.396e-03, size: 544, ETA: 0:27:46
2025-11-12 14:45:53.454 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 45/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.150s, data_time: 0.002s, total_loss: 6.3, iou_loss: 2.2, l1_loss: 0.9, conf_loss: 2.5, cls_loss: 0.7, lr: 1.392e-03, size: 576, ETA: 0:27:42
2025-11-12 14:45:56.512 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 45/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.151s, data_time: 0.004s, total_loss: 5.2, iou_loss: 2.4, l1_loss: 0.7, conf_loss: 1.5, cls_loss: 0.7, lr: 1.388e-03, size: 256, ETA: 0:27:38
2025-11-12 14:45:59.401 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 45/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.143s, data_time: 0.002s, total_loss: 5.1, iou_loss: 2.1, l1_loss: 0.9, conf_loss: 1.5, cls_loss: 0.6, lr: 1.384e-03, size: 512, ETA: 0:27:33
2025-11-12 14:46:00.867 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:46:05.980 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:46:06.971 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:46:07.607 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6337
2025-11-12 14:46:07.764 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5781
2025-11-12 14:46:07.805 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4341
2025-11-12 14:46:07.806 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5486
2025-11-12 14:46:07.806 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:46:07.806 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:46:07.806 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.634
2025-11-12 14:46:07.806 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.578
2025-11-12 14:46:07.806 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.434
2025-11-12 14:46:07.806 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.549
2025-11-12 14:46:07.806 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:46:07.807 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:46:07.807 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:46:07.807 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:46:07.807 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:46:07.807 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:46:07.807 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:46:07.807 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:46:07.807 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:46:08.674 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:46:09.506 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:46:10.355 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:46:11.152 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:46:11.997 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:46:12.796 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:46:13.597 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:46:14.432 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:46:15.239 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:46:15.239 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.30
2025-11-12 14:46:15.239 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.55
2025-11-12 14:46:15.240 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:46:15.247 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.19 ms, Average NMS time: 0.57 ms, Average inference time: 2.76 ms

2025-11-12 14:46:15.248 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:46:15.326 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:46:15.447 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch46
2025-11-12 14:46:18.446 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 46/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 6.8, iou_loss: 2.5, l1_loss: 0.9, conf_loss: 2.6, cls_loss: 0.8, lr: 1.379e-03, size: 256, ETA: 0:27:27
2025-11-12 14:46:21.353 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 46/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 6.5, iou_loss: 2.5, l1_loss: 1.0, conf_loss: 2.2, cls_loss: 0.8, lr: 1.375e-03, size: 352, ETA: 0:27:23
2025-11-12 14:46:24.225 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 46/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.141s, data_time: 0.002s, total_loss: 4.7, iou_loss: 1.7, l1_loss: 1.0, conf_loss: 1.5, cls_loss: 0.5, lr: 1.371e-03, size: 576, ETA: 0:27:18
2025-11-12 14:46:27.207 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 46/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 2.1, cls_loss: 0.9, lr: 1.368e-03, size: 416, ETA: 0:27:14
2025-11-12 14:46:30.055 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 46/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.140s, data_time: 0.003s, total_loss: 4.2, iou_loss: 1.8, l1_loss: 0.6, conf_loss: 1.3, cls_loss: 0.5, lr: 1.364e-03, size: 448, ETA: 0:27:10
2025-11-12 14:46:33.076 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 46/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.2, l1_loss: 1.1, conf_loss: 1.8, cls_loss: 0.6, lr: 1.360e-03, size: 576, ETA: 0:27:06
2025-11-12 14:46:34.446 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:46:39.583 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:46:40.479 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:46:41.029 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6317
2025-11-12 14:46:41.154 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5769
2025-11-12 14:46:41.235 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4346
2025-11-12 14:46:41.236 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5477
2025-11-12 14:46:41.236 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:46:41.236 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:46:41.236 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.632
2025-11-12 14:46:41.237 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.577
2025-11-12 14:46:41.237 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.435
2025-11-12 14:46:41.237 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.548
2025-11-12 14:46:41.237 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:46:41.237 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:46:41.237 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:46:41.237 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:46:41.237 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:46:41.238 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:46:41.238 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:46:41.238 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:46:41.238 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:46:41.943 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:46:42.635 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:46:43.366 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:46:44.057 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:46:44.766 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:46:45.508 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:46:46.202 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:46:46.908 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:46:47.655 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:46:47.655 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 14:46:47.656 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.55
2025-11-12 14:46:47.656 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:46:47.663 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.21 ms, Average NMS time: 0.61 ms, Average inference time: 2.82 ms

2025-11-12 14:46:47.664 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:46:47.740 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:46:47.820 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch47
2025-11-12 14:46:50.930 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 47/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.152s, data_time: 0.001s, total_loss: 4.1, iou_loss: 1.9, l1_loss: 0.5, conf_loss: 1.2, cls_loss: 0.5, lr: 1.355e-03, size: 416, ETA: 0:27:00
2025-11-12 14:46:54.092 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 47/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.155s, data_time: 0.002s, total_loss: 6.0, iou_loss: 2.3, l1_loss: 0.9, conf_loss: 2.2, cls_loss: 0.6, lr: 1.351e-03, size: 512, ETA: 0:26:56
2025-11-12 14:46:57.009 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 47/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.143s, data_time: 0.003s, total_loss: 6.5, iou_loss: 2.5, l1_loss: 0.9, conf_loss: 2.4, cls_loss: 0.7, lr: 1.347e-03, size: 448, ETA: 0:26:52
2025-11-12 14:47:00.000 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 47/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 6.3, iou_loss: 2.4, l1_loss: 0.9, conf_loss: 2.3, cls_loss: 0.7, lr: 1.343e-03, size: 448, ETA: 0:26:48
2025-11-12 14:47:03.038 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 47/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.4, l1_loss: 0.8, conf_loss: 2.0, cls_loss: 0.6, lr: 1.339e-03, size: 288, ETA: 0:26:44
2025-11-12 14:47:06.242 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 47/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.156s, data_time: 0.002s, total_loss: 6.6, iou_loss: 2.4, l1_loss: 0.8, conf_loss: 2.7, cls_loss: 0.7, lr: 1.336e-03, size: 288, ETA: 0:26:40
2025-11-12 14:47:07.686 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:47:12.712 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:47:13.724 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:47:14.372 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6328
2025-11-12 14:47:14.497 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5734
2025-11-12 14:47:14.541 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4235
2025-11-12 14:47:14.542 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5432
2025-11-12 14:47:14.542 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:47:14.542 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:47:14.543 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.633
2025-11-12 14:47:14.543 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.573
2025-11-12 14:47:14.543 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.423
2025-11-12 14:47:14.543 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.543
2025-11-12 14:47:14.543 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:47:14.543 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:47:14.543 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:47:14.543 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:47:14.543 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:47:14.543 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:47:14.543 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:47:14.544 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:47:14.544 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:47:15.401 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:47:16.208 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:47:17.035 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:47:17.829 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:47:18.661 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:47:19.451 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:47:20.294 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:47:21.100 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:47:21.894 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:47:21.894 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.30
2025-11-12 14:47:21.895 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.54
2025-11-12 14:47:21.895 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:47:21.902 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.15 ms, Average NMS time: 0.58 ms, Average inference time: 2.73 ms

2025-11-12 14:47:21.903 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:47:22.015 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:47:22.096 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch48
2025-11-12 14:47:24.950 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 48/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.141s, data_time: 0.002s, total_loss: 4.8, iou_loss: 1.9, l1_loss: 0.8, conf_loss: 1.5, cls_loss: 0.6, lr: 1.330e-03, size: 512, ETA: 0:26:34
2025-11-12 14:47:27.824 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 48/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.3Gb, iter_time: 0.140s, data_time: 0.002s, total_loss: 6.2, iou_loss: 2.5, l1_loss: 1.1, conf_loss: 2.0, cls_loss: 0.6, lr: 1.326e-03, size: 384, ETA: 0:26:29
2025-11-12 14:47:30.637 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 48/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.138s, data_time: 0.002s, total_loss: 6.0, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 2.2, cls_loss: 0.8, lr: 1.322e-03, size: 320, ETA: 0:26:25
2025-11-12 14:47:33.455 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 48/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.138s, data_time: 0.002s, total_loss: 4.6, iou_loss: 1.7, l1_loss: 0.8, conf_loss: 1.5, cls_loss: 0.6, lr: 1.318e-03, size: 448, ETA: 0:26:21
2025-11-12 14:47:36.365 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 48/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.143s, data_time: 0.003s, total_loss: 6.1, iou_loss: 2.2, l1_loss: 1.0, conf_loss: 2.3, cls_loss: 0.7, lr: 1.315e-03, size: 480, ETA: 0:26:17
2025-11-12 14:47:39.358 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 48/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.147s, data_time: 0.003s, total_loss: 6.0, iou_loss: 2.3, l1_loss: 1.1, conf_loss: 1.8, cls_loss: 0.7, lr: 1.311e-03, size: 512, ETA: 0:26:12
2025-11-12 14:47:40.762 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:47:45.974 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:47:46.921 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:47:47.510 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6421
2025-11-12 14:47:47.682 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5917
2025-11-12 14:47:47.726 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4603
2025-11-12 14:47:47.727 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5647
2025-11-12 14:47:47.727 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:47:47.727 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:47:47.728 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.642
2025-11-12 14:47:47.728 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.592
2025-11-12 14:47:47.728 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.460
2025-11-12 14:47:47.728 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.565
2025-11-12 14:47:47.728 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:47:47.728 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:47:47.728 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:47:47.728 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:47:47.728 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:47:47.729 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:47:47.729 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:47:47.729 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:47:47.729 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:47:48.486 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:47:49.272 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:47:50.021 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:47:50.805 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:47:51.554 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:47:52.310 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:47:53.098 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:47:53.844 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:47:54.649 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:47:54.649 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 14:47:54.649 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 14:47:54.649 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:47:54.657 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.27 ms, Average NMS time: 0.62 ms, Average inference time: 2.89 ms

2025-11-12 14:47:54.658 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:47:54.734 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:47:54.815 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch49
2025-11-12 14:47:57.712 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 49/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 5.1, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.8, cls_loss: 0.6, lr: 1.305e-03, size: 416, ETA: 0:26:07
2025-11-12 14:48:00.593 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 49/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 6.8, iou_loss: 2.4, l1_loss: 1.1, conf_loss: 2.4, cls_loss: 0.8, lr: 1.301e-03, size: 544, ETA: 0:26:02
2025-11-12 14:48:03.598 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 49/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.6, l1_loss: 0.9, conf_loss: 1.7, cls_loss: 0.6, lr: 1.297e-03, size: 480, ETA: 0:25:58
2025-11-12 14:48:06.956 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 49/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.164s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 2.0, cls_loss: 0.6, lr: 1.294e-03, size: 544, ETA: 0:25:55
2025-11-12 14:48:10.162 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 49/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.156s, data_time: 0.003s, total_loss: 5.1, iou_loss: 2.3, l1_loss: 0.6, conf_loss: 1.6, cls_loss: 0.6, lr: 1.290e-03, size: 256, ETA: 0:25:51
2025-11-12 14:48:13.256 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 49/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.150s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.4, l1_loss: 0.9, conf_loss: 1.8, cls_loss: 0.6, lr: 1.286e-03, size: 320, ETA: 0:25:47
2025-11-12 14:48:14.550 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:48:19.650 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:48:21.143 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:48:22.054 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6398
2025-11-12 14:48:22.284 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5898
2025-11-12 14:48:22.330 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4588
2025-11-12 14:48:22.331 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5628
2025-11-12 14:48:22.331 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:48:22.331 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:48:22.331 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.640
2025-11-12 14:48:22.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.590
2025-11-12 14:48:22.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.459
2025-11-12 14:48:22.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.563
2025-11-12 14:48:22.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:48:22.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:48:22.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:48:22.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:48:22.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:48:22.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:48:22.333 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:48:22.333 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:48:22.333 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:48:23.543 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:48:24.730 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:48:25.920 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:48:27.093 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:48:28.282 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:48:29.466 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:48:30.678 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:48:31.859 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:48:33.009 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:48:33.009 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 14:48:33.009 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 14:48:33.009 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:48:33.017 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.10 ms, Average NMS time: 0.57 ms, Average inference time: 2.67 ms

2025-11-12 14:48:33.019 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:48:33.131 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:48:33.214 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch50
2025-11-12 14:48:36.253 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 50/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.150s, data_time: 0.002s, total_loss: 5.5, iou_loss: 2.3, l1_loss: 0.9, conf_loss: 1.7, cls_loss: 0.6, lr: 1.280e-03, size: 512, ETA: 0:25:41
2025-11-12 14:48:39.210 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 50/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.145s, data_time: 0.002s, total_loss: 6.2, iou_loss: 2.2, l1_loss: 0.9, conf_loss: 2.4, cls_loss: 0.8, lr: 1.276e-03, size: 576, ETA: 0:25:37
2025-11-12 14:48:42.163 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 50/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.3Gb, iter_time: 0.144s, data_time: 0.002s, total_loss: 6.3, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 2.5, cls_loss: 0.8, lr: 1.272e-03, size: 384, ETA: 0:25:33
2025-11-12 14:48:44.933 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 50/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.135s, data_time: 0.002s, total_loss: 5.5, iou_loss: 2.3, l1_loss: 0.6, conf_loss: 1.9, cls_loss: 0.7, lr: 1.268e-03, size: 256, ETA: 0:25:29
2025-11-12 14:48:47.667 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 50/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.134s, data_time: 0.004s, total_loss: 7.3, iou_loss: 2.1, l1_loss: 1.1, conf_loss: 3.2, cls_loss: 0.8, lr: 1.264e-03, size: 544, ETA: 0:25:25
2025-11-12 14:48:50.713 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 50/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.150s, data_time: 0.002s, total_loss: 5.5, iou_loss: 2.2, l1_loss: 0.6, conf_loss: 2.0, cls_loss: 0.7, lr: 1.261e-03, size: 320, ETA: 0:25:21
2025-11-12 14:48:51.974 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:48:57.114 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:48:58.172 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:48:58.857 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6381
2025-11-12 14:48:59.042 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5995
2025-11-12 14:48:59.087 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4487
2025-11-12 14:48:59.088 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5621
2025-11-12 14:48:59.088 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:48:59.088 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:48:59.088 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.638
2025-11-12 14:48:59.088 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.599
2025-11-12 14:48:59.088 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.449
2025-11-12 14:48:59.088 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.562
2025-11-12 14:48:59.088 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:48:59.089 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:48:59.089 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:48:59.089 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:48:59.089 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:48:59.089 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:48:59.089 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:48:59.089 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:48:59.089 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:48:59.937 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:49:00.822 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:49:01.670 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:49:02.553 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:49:03.403 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:49:04.283 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:49:05.140 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:49:06.027 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:49:06.884 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:49:06.885 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 14:49:06.885 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 14:49:06.885 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:49:06.894 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.25 ms, Average NMS time: 0.62 ms, Average inference time: 2.87 ms

2025-11-12 14:49:06.895 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:49:07.008 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:49:07.094 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch51
2025-11-12 14:49:10.272 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 51/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.157s, data_time: 0.002s, total_loss: 5.3, iou_loss: 1.9, l1_loss: 0.8, conf_loss: 1.7, cls_loss: 0.8, lr: 1.255e-03, size: 480, ETA: 0:25:15
2025-11-12 14:49:13.500 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 51/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.157s, data_time: 0.002s, total_loss: 5.8, iou_loss: 2.4, l1_loss: 0.7, conf_loss: 2.0, cls_loss: 0.6, lr: 1.251e-03, size: 352, ETA: 0:25:11
2025-11-12 14:49:16.800 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 51/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.161s, data_time: 0.002s, total_loss: 5.8, iou_loss: 2.1, l1_loss: 0.8, conf_loss: 2.4, cls_loss: 0.6, lr: 1.247e-03, size: 448, ETA: 0:25:08
2025-11-12 14:49:19.917 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 51/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.151s, data_time: 0.002s, total_loss: 5.5, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.9, cls_loss: 0.8, lr: 1.243e-03, size: 352, ETA: 0:25:04
2025-11-12 14:49:23.014 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 51/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.152s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 1.8, cls_loss: 0.6, lr: 1.239e-03, size: 480, ETA: 0:25:00
2025-11-12 14:49:25.938 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 51/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.145s, data_time: 0.002s, total_loss: 5.3, iou_loss: 2.3, l1_loss: 0.8, conf_loss: 1.5, cls_loss: 0.6, lr: 1.235e-03, size: 448, ETA: 0:24:56
2025-11-12 14:49:27.310 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:49:32.203 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:49:32.927 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:49:33.388 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6233
2025-11-12 14:49:33.508 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5811
2025-11-12 14:49:33.575 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.3858
2025-11-12 14:49:33.576 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5301
2025-11-12 14:49:33.576 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:49:33.576 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:49:33.576 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.623
2025-11-12 14:49:33.576 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.581
2025-11-12 14:49:33.576 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.386
2025-11-12 14:49:33.576 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.530
2025-11-12 14:49:33.576 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:49:33.577 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:49:33.577 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:49:33.577 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:49:33.577 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:49:33.577 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:49:33.577 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:49:33.577 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:49:33.577 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:49:34.169 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:49:34.756 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:49:35.348 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:49:35.934 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:49:36.551 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:49:37.141 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:49:37.726 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:49:38.313 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:49:38.894 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:49:38.895 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.29
2025-11-12 14:49:38.895 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.53
2025-11-12 14:49:38.895 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:49:38.901 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.08 ms, Average NMS time: 0.53 ms, Average inference time: 2.61 ms

2025-11-12 14:49:38.904 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:49:39.013 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:49:39.095 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch52
2025-11-12 14:49:42.259 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 52/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.156s, data_time: 0.002s, total_loss: 6.9, iou_loss: 2.2, l1_loss: 1.0, conf_loss: 3.1, cls_loss: 0.7, lr: 1.229e-03, size: 448, ETA: 0:24:51
2025-11-12 14:49:45.507 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 52/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.158s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 2.0, cls_loss: 0.6, lr: 1.226e-03, size: 576, ETA: 0:24:47
2025-11-12 14:49:48.669 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 52/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.154s, data_time: 0.003s, total_loss: 4.6, iou_loss: 1.9, l1_loss: 0.7, conf_loss: 1.4, cls_loss: 0.6, lr: 1.222e-03, size: 416, ETA: 0:24:43
2025-11-12 14:49:51.948 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 52/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.160s, data_time: 0.002s, total_loss: 5.4, iou_loss: 2.3, l1_loss: 0.8, conf_loss: 1.6, cls_loss: 0.7, lr: 1.218e-03, size: 384, ETA: 0:24:40
2025-11-12 14:49:55.463 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 52/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.171s, data_time: 0.002s, total_loss: 6.5, iou_loss: 2.3, l1_loss: 1.1, conf_loss: 2.3, cls_loss: 0.7, lr: 1.214e-03, size: 576, ETA: 0:24:37
2025-11-12 14:49:58.825 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 52/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.164s, data_time: 0.002s, total_loss: 4.4, iou_loss: 1.9, l1_loss: 0.6, conf_loss: 1.5, cls_loss: 0.5, lr: 1.210e-03, size: 320, ETA: 0:24:33
2025-11-12 14:50:00.274 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:50:05.394 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:50:06.561 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:50:07.205 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6450
2025-11-12 14:50:07.404 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5865
2025-11-12 14:50:07.450 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4436
2025-11-12 14:50:07.451 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5584
2025-11-12 14:50:07.451 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:50:07.451 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:50:07.451 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.645
2025-11-12 14:50:07.452 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.586
2025-11-12 14:50:07.452 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.444
2025-11-12 14:50:07.452 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.558
2025-11-12 14:50:07.452 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:50:07.452 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:50:07.452 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:50:07.452 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:50:07.452 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:50:07.452 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:50:07.453 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:50:07.453 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:50:07.453 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:50:08.381 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:50:09.260 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:50:10.199 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:50:11.092 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:50:12.020 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:50:12.909 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:50:13.850 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:50:14.767 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:50:15.641 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:50:15.642 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 14:50:15.642 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 14:50:15.642 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:50:15.649 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.23 ms, Average NMS time: 0.61 ms, Average inference time: 2.84 ms

2025-11-12 14:50:15.651 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:50:15.726 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:50:15.806 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch53
2025-11-12 14:50:19.145 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 53/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.164s, data_time: 0.002s, total_loss: 6.4, iou_loss: 2.2, l1_loss: 1.2, conf_loss: 2.3, cls_loss: 0.6, lr: 1.204e-03, size: 576, ETA: 0:24:28
2025-11-12 14:50:22.649 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 53/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.172s, data_time: 0.002s, total_loss: 6.6, iou_loss: 2.2, l1_loss: 1.0, conf_loss: 2.7, cls_loss: 0.7, lr: 1.200e-03, size: 544, ETA: 0:24:25
2025-11-12 14:50:26.010 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 53/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.164s, data_time: 0.001s, total_loss: 4.4, iou_loss: 1.8, l1_loss: 0.6, conf_loss: 1.3, cls_loss: 0.5, lr: 1.196e-03, size: 416, ETA: 0:24:21
2025-11-12 14:50:29.255 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 53/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.158s, data_time: 0.001s, total_loss: 4.9, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.6, cls_loss: 0.6, lr: 1.192e-03, size: 416, ETA: 0:24:18
2025-11-12 14:50:32.586 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 53/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.163s, data_time: 0.003s, total_loss: 5.3, iou_loss: 2.0, l1_loss: 0.9, conf_loss: 1.8, cls_loss: 0.6, lr: 1.188e-03, size: 480, ETA: 0:24:14
2025-11-12 14:50:35.587 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 53/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 6.4, iou_loss: 2.5, l1_loss: 1.0, conf_loss: 2.2, cls_loss: 0.6, lr: 1.184e-03, size: 448, ETA: 0:24:10
2025-11-12 14:50:36.965 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:50:41.977 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:50:42.679 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:50:43.098 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6305
2025-11-12 14:50:43.204 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5974
2025-11-12 14:50:43.245 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4360
2025-11-12 14:50:43.246 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5547
2025-11-12 14:50:43.246 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:50:43.246 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:50:43.246 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.631
2025-11-12 14:50:43.247 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.597
2025-11-12 14:50:43.247 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.436
2025-11-12 14:50:43.247 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.555
2025-11-12 14:50:43.247 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:50:43.247 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:50:43.247 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:50:43.247 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:50:43.247 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:50:43.247 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:50:43.247 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:50:43.247 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:50:43.248 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:50:43.827 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:50:44.366 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:50:44.904 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:50:45.448 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:50:45.995 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:50:46.534 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:50:47.108 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:50:47.644 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:50:48.210 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:50:48.210 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 14:50:48.210 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.55
2025-11-12 14:50:48.211 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:50:48.217 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.12 ms, Average NMS time: 0.55 ms, Average inference time: 2.67 ms

2025-11-12 14:50:48.219 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:50:48.294 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:50:48.374 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch54
2025-11-12 14:50:51.235 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 54/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.141s, data_time: 0.001s, total_loss: 4.0, iou_loss: 1.9, l1_loss: 0.6, conf_loss: 1.0, cls_loss: 0.5, lr: 1.178e-03, size: 352, ETA: 0:24:05
2025-11-12 14:50:54.024 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 54/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.137s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.2, l1_loss: 0.6, conf_loss: 2.3, cls_loss: 0.6, lr: 1.174e-03, size: 256, ETA: 0:24:01
2025-11-12 14:50:57.139 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 54/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.153s, data_time: 0.002s, total_loss: 6.7, iou_loss: 2.5, l1_loss: 1.3, conf_loss: 2.1, cls_loss: 0.8, lr: 1.170e-03, size: 576, ETA: 0:23:57
2025-11-12 14:51:00.316 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 54/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.155s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.4, l1_loss: 0.8, conf_loss: 1.8, cls_loss: 0.7, lr: 1.166e-03, size: 512, ETA: 0:23:53
2025-11-12 14:51:03.304 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 54/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 5.4, iou_loss: 2.2, l1_loss: 1.1, conf_loss: 1.6, cls_loss: 0.6, lr: 1.162e-03, size: 448, ETA: 0:23:49
2025-11-12 14:51:06.273 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 54/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.145s, data_time: 0.002s, total_loss: 5.1, iou_loss: 2.1, l1_loss: 0.6, conf_loss: 1.9, cls_loss: 0.6, lr: 1.158e-03, size: 320, ETA: 0:23:45
2025-11-12 14:51:07.667 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:51:12.677 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:51:13.676 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:51:14.305 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6458
2025-11-12 14:51:14.446 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5708
2025-11-12 14:51:14.484 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4460
2025-11-12 14:51:14.484 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5542
2025-11-12 14:51:14.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:51:14.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:51:14.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.646
2025-11-12 14:51:14.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.571
2025-11-12 14:51:14.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.446
2025-11-12 14:51:14.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.554
2025-11-12 14:51:14.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:51:14.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:51:14.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:51:14.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:51:14.485 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:51:14.486 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:51:14.486 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:51:14.486 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:51:14.486 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:51:15.311 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:51:16.103 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:51:16.929 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:51:17.712 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:51:18.534 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:51:19.325 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:51:20.130 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:51:20.955 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:51:21.747 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:51:21.747 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 14:51:21.748 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.55
2025-11-12 14:51:21.748 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:51:21.755 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.12 ms, Average NMS time: 0.55 ms, Average inference time: 2.67 ms

2025-11-12 14:51:21.756 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:51:21.832 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:51:21.950 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch55
2025-11-12 14:51:25.143 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 55/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.3Gb, iter_time: 0.157s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.3, l1_loss: 0.7, conf_loss: 2.1, cls_loss: 0.6, lr: 1.152e-03, size: 320, ETA: 0:23:40
2025-11-12 14:51:28.213 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 55/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 4.2, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.1, cls_loss: 0.5, lr: 1.148e-03, size: 288, ETA: 0:23:36
2025-11-12 14:51:31.421 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 55/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.156s, data_time: 0.002s, total_loss: 5.3, iou_loss: 2.4, l1_loss: 0.7, conf_loss: 1.6, cls_loss: 0.6, lr: 1.144e-03, size: 256, ETA: 0:23:33
2025-11-12 14:51:34.490 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 55/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 4.4, iou_loss: 1.8, l1_loss: 0.7, conf_loss: 1.4, cls_loss: 0.5, lr: 1.140e-03, size: 384, ETA: 0:23:29
2025-11-12 14:51:37.700 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 55/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.156s, data_time: 0.002s, total_loss: 5.4, iou_loss: 2.1, l1_loss: 0.6, conf_loss: 2.0, cls_loss: 0.7, lr: 1.136e-03, size: 256, ETA: 0:23:25
2025-11-12 14:51:40.966 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 55/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.159s, data_time: 0.002s, total_loss: 6.7, iou_loss: 2.5, l1_loss: 1.4, conf_loss: 2.1, cls_loss: 0.7, lr: 1.132e-03, size: 576, ETA: 0:23:22
2025-11-12 14:51:42.315 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:51:47.568 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:51:48.501 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:51:49.035 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6273
2025-11-12 14:51:49.243 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.6025
2025-11-12 14:51:49.287 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4179
2025-11-12 14:51:49.288 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5492
2025-11-12 14:51:49.288 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:51:49.289 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:51:49.289 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.627
2025-11-12 14:51:49.289 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.603
2025-11-12 14:51:49.289 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.418
2025-11-12 14:51:49.289 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.549
2025-11-12 14:51:49.289 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:51:49.289 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:51:49.289 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:51:49.289 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:51:49.289 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:51:49.290 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:51:49.290 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:51:49.290 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:51:49.290 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:51:50.009 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:51:50.790 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:51:51.536 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:51:52.271 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:51:53.026 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:51:53.752 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:51:54.512 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:51:55.234 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:51:55.963 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:51:55.964 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.30
2025-11-12 14:51:55.964 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.55
2025-11-12 14:51:55.964 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:51:55.971 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.21 ms, Average NMS time: 0.63 ms, Average inference time: 2.84 ms

2025-11-12 14:51:55.972 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:51:56.048 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:51:56.169 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch56
2025-11-12 14:51:59.032 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 56/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.140s, data_time: 0.002s, total_loss: 6.5, iou_loss: 2.3, l1_loss: 0.9, conf_loss: 2.7, cls_loss: 0.7, lr: 1.127e-03, size: 352, ETA: 0:23:16
2025-11-12 14:52:01.801 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 56/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.137s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.3, l1_loss: 0.8, conf_loss: 1.9, cls_loss: 0.6, lr: 1.122e-03, size: 416, ETA: 0:23:12
2025-11-12 14:52:04.646 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 56/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.140s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.3, l1_loss: 0.8, conf_loss: 1.9, cls_loss: 0.7, lr: 1.118e-03, size: 384, ETA: 0:23:08
2025-11-12 14:52:07.447 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 56/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.138s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.4, l1_loss: 0.8, conf_loss: 1.9, cls_loss: 0.8, lr: 1.114e-03, size: 320, ETA: 0:23:04
2025-11-12 14:52:10.225 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 56/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.136s, data_time: 0.002s, total_loss: 4.0, iou_loss: 1.7, l1_loss: 0.7, conf_loss: 1.2, cls_loss: 0.5, lr: 1.110e-03, size: 352, ETA: 0:23:00
2025-11-12 14:52:12.977 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 56/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.136s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.0, l1_loss: 1.0, conf_loss: 2.0, cls_loss: 0.6, lr: 1.106e-03, size: 480, ETA: 0:22:56
2025-11-12 14:52:14.360 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:52:19.412 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:52:20.331 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:52:20.882 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6484
2025-11-12 14:52:21.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5877
2025-11-12 14:52:21.103 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4381
2025-11-12 14:52:21.103 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5581
2025-11-12 14:52:21.104 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:52:21.104 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:52:21.104 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.648
2025-11-12 14:52:21.104 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.588
2025-11-12 14:52:21.104 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.438
2025-11-12 14:52:21.104 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.558
2025-11-12 14:52:21.104 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:52:21.104 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:52:21.105 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:52:21.105 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:52:21.105 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:52:21.105 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:52:21.105 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:52:21.105 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:52:21.105 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:52:21.821 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:52:22.539 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:52:23.304 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:52:24.032 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:52:24.806 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:52:25.531 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:52:26.256 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:52:27.009 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:52:27.758 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:52:27.759 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 14:52:27.759 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 14:52:27.759 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:52:27.766 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.17 ms, Average NMS time: 0.55 ms, Average inference time: 2.73 ms

2025-11-12 14:52:27.767 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:52:27.844 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:52:27.926 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch57
2025-11-12 14:52:30.899 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 57/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.3Gb, iter_time: 0.144s, data_time: 0.003s, total_loss: 4.6, iou_loss: 1.7, l1_loss: 0.7, conf_loss: 1.6, cls_loss: 0.7, lr: 1.100e-03, size: 288, ETA: 0:22:50
2025-11-12 14:52:33.813 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 57/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.3Gb, iter_time: 0.143s, data_time: 0.002s, total_loss: 5.0, iou_loss: 2.2, l1_loss: 0.6, conf_loss: 1.5, cls_loss: 0.6, lr: 1.096e-03, size: 288, ETA: 0:22:47
2025-11-12 14:52:37.083 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 57/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.160s, data_time: 0.002s, total_loss: 6.7, iou_loss: 2.7, l1_loss: 1.0, conf_loss: 2.3, cls_loss: 0.7, lr: 1.092e-03, size: 576, ETA: 0:22:43
2025-11-12 14:52:40.393 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 57/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.161s, data_time: 0.002s, total_loss: 5.3, iou_loss: 2.1, l1_loss: 0.6, conf_loss: 1.9, cls_loss: 0.6, lr: 1.088e-03, size: 320, ETA: 0:22:40
2025-11-12 14:52:43.797 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 57/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.166s, data_time: 0.003s, total_loss: 5.4, iou_loss: 2.4, l1_loss: 0.8, conf_loss: 1.6, cls_loss: 0.6, lr: 1.084e-03, size: 352, ETA: 0:22:36
2025-11-12 14:52:47.015 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 57/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.3Gb, iter_time: 0.157s, data_time: 0.001s, total_loss: 7.1, iou_loss: 2.6, l1_loss: 0.8, conf_loss: 2.8, cls_loss: 0.9, lr: 1.080e-03, size: 256, ETA: 0:22:33
2025-11-12 14:52:48.320 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:52:53.410 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:52:54.775 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:52:55.715 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6296
2025-11-12 14:52:55.919 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5684
2025-11-12 14:52:55.962 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4478
2025-11-12 14:52:55.963 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5486
2025-11-12 14:52:55.963 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:52:55.963 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:52:55.963 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.630
2025-11-12 14:52:55.963 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.568
2025-11-12 14:52:55.963 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.448
2025-11-12 14:52:55.963 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.549
2025-11-12 14:52:55.963 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:52:55.963 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:52:55.963 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:52:55.964 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:52:55.964 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:52:55.964 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:52:55.964 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:52:55.964 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:52:55.964 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:52:57.138 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:52:58.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:52:59.457 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:53:00.605 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:53:01.763 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:53:02.954 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:53:04.116 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:53:05.244 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:53:06.403 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:53:06.404 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.29
2025-11-12 14:53:06.404 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.55
2025-11-12 14:53:06.404 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:53:06.412 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.17 ms, Average NMS time: 0.56 ms, Average inference time: 2.73 ms

2025-11-12 14:53:06.413 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:53:06.491 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:53:06.572 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch58
2025-11-12 14:53:09.640 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 58/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.150s, data_time: 0.002s, total_loss: 6.0, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 2.3, cls_loss: 0.6, lr: 1.074e-03, size: 480, ETA: 0:22:27
2025-11-12 14:53:12.861 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 58/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.157s, data_time: 0.001s, total_loss: 5.7, iou_loss: 2.4, l1_loss: 0.8, conf_loss: 1.9, cls_loss: 0.6, lr: 1.070e-03, size: 288, ETA: 0:22:24
2025-11-12 14:53:16.082 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 58/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.157s, data_time: 0.001s, total_loss: 5.2, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 1.6, cls_loss: 0.7, lr: 1.066e-03, size: 320, ETA: 0:22:20
2025-11-12 14:53:19.158 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 58/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.151s, data_time: 0.003s, total_loss: 4.5, iou_loss: 1.9, l1_loss: 0.6, conf_loss: 1.4, cls_loss: 0.6, lr: 1.062e-03, size: 416, ETA: 0:22:17
2025-11-12 14:53:22.349 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 58/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.156s, data_time: 0.004s, total_loss: 4.3, iou_loss: 1.9, l1_loss: 0.6, conf_loss: 1.3, cls_loss: 0.6, lr: 1.058e-03, size: 288, ETA: 0:22:13
2025-11-12 14:53:25.527 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 58/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.155s, data_time: 0.002s, total_loss: 4.6, iou_loss: 1.9, l1_loss: 0.7, conf_loss: 1.3, cls_loss: 0.6, lr: 1.054e-03, size: 416, ETA: 0:22:10
2025-11-12 14:53:26.959 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:53:32.154 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:53:33.054 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:53:33.659 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6506
2025-11-12 14:53:33.781 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5714
2025-11-12 14:53:33.867 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4511
2025-11-12 14:53:33.867 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5577
2025-11-12 14:53:33.868 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:53:33.868 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:53:33.868 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.651
2025-11-12 14:53:33.868 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.571
2025-11-12 14:53:33.868 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.451
2025-11-12 14:53:33.868 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.558
2025-11-12 14:53:33.868 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:53:33.868 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:53:33.868 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:53:33.869 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:53:33.869 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:53:33.869 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:53:33.869 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:53:33.869 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:53:33.869 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:53:34.597 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:53:35.333 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:53:36.108 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:53:36.838 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:53:37.574 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:53:38.345 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:53:39.084 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:53:39.856 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:53:40.595 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:53:40.596 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 14:53:40.596 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 14:53:40.596 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:53:40.603 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.21 ms, Average NMS time: 0.62 ms, Average inference time: 2.82 ms

2025-11-12 14:53:40.605 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:53:40.687 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:53:40.770 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch59
2025-11-12 14:53:43.814 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 59/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.3Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.4, l1_loss: 0.9, conf_loss: 1.7, cls_loss: 0.6, lr: 1.048e-03, size: 416, ETA: 0:22:04
2025-11-12 14:53:46.851 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 59/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.3Gb, iter_time: 0.149s, data_time: 0.001s, total_loss: 4.7, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.5, cls_loss: 0.5, lr: 1.044e-03, size: 352, ETA: 0:22:01
2025-11-12 14:53:49.729 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 59/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 1.9, cls_loss: 0.6, lr: 1.040e-03, size: 320, ETA: 0:21:57
2025-11-12 14:53:52.526 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 59/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.137s, data_time: 0.002s, total_loss: 8.0, iou_loss: 2.3, l1_loss: 1.1, conf_loss: 4.0, cls_loss: 0.7, lr: 1.036e-03, size: 512, ETA: 0:21:53
2025-11-12 14:53:55.559 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 59/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.150s, data_time: 0.003s, total_loss: 4.1, iou_loss: 1.6, l1_loss: 0.6, conf_loss: 1.4, cls_loss: 0.5, lr: 1.032e-03, size: 480, ETA: 0:21:49
2025-11-12 14:53:58.628 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 59/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.152s, data_time: 0.002s, total_loss: 5.4, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 1.7, cls_loss: 0.7, lr: 1.028e-03, size: 288, ETA: 0:21:46
2025-11-12 14:53:59.894 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:54:05.136 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:54:06.317 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:54:07.049 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6467
2025-11-12 14:54:07.221 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5715
2025-11-12 14:54:07.308 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4485
2025-11-12 14:54:07.309 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5556
2025-11-12 14:54:07.309 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:54:07.309 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:54:07.309 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.647
2025-11-12 14:54:07.309 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.571
2025-11-12 14:54:07.309 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.448
2025-11-12 14:54:07.309 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.556
2025-11-12 14:54:07.310 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:54:07.310 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:54:07.310 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:54:07.310 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:54:07.310 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:54:07.310 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:54:07.310 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:54:07.310 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:54:07.310 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:54:08.246 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:54:09.244 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:54:10.189 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:54:11.152 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:54:12.122 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:54:13.061 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:54:14.029 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:54:15.000 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:54:15.930 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:54:15.930 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 14:54:15.930 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 14:54:15.930 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:54:15.938 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.23 ms, Average NMS time: 0.61 ms, Average inference time: 2.84 ms

2025-11-12 14:54:15.939 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:54:16.018 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:54:16.099 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch60
2025-11-12 14:54:19.217 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 60/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.152s, data_time: 0.002s, total_loss: 4.8, iou_loss: 1.6, l1_loss: 0.6, conf_loss: 2.1, cls_loss: 0.5, lr: 1.022e-03, size: 448, ETA: 0:21:40
2025-11-12 14:54:22.606 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 60/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.165s, data_time: 0.002s, total_loss: 5.4, iou_loss: 2.3, l1_loss: 0.7, conf_loss: 1.9, cls_loss: 0.6, lr: 1.018e-03, size: 416, ETA: 0:21:37
2025-11-12 14:54:25.874 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 60/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.159s, data_time: 0.001s, total_loss: 5.8, iou_loss: 2.5, l1_loss: 0.7, conf_loss: 2.0, cls_loss: 0.6, lr: 1.014e-03, size: 288, ETA: 0:21:33
2025-11-12 14:54:29.162 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 60/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.160s, data_time: 0.003s, total_loss: 6.7, iou_loss: 2.5, l1_loss: 0.9, conf_loss: 2.6, cls_loss: 0.8, lr: 1.010e-03, size: 320, ETA: 0:21:30
2025-11-12 14:54:32.278 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 60/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.151s, data_time: 0.002s, total_loss: 7.3, iou_loss: 2.6, l1_loss: 0.8, conf_loss: 3.2, cls_loss: 0.7, lr: 1.006e-03, size: 288, ETA: 0:21:26
2025-11-12 14:54:35.258 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 60/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 4.9, iou_loss: 1.9, l1_loss: 0.8, conf_loss: 1.7, cls_loss: 0.6, lr: 1.002e-03, size: 576, ETA: 0:21:23
2025-11-12 14:54:36.674 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:54:41.629 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:54:42.466 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:54:42.978 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6436
2025-11-12 14:54:43.101 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5749
2025-11-12 14:54:43.138 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4495
2025-11-12 14:54:43.139 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5560
2025-11-12 14:54:43.139 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:54:43.139 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:54:43.139 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.644
2025-11-12 14:54:43.139 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.575
2025-11-12 14:54:43.139 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.449
2025-11-12 14:54:43.139 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.556
2025-11-12 14:54:43.139 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:54:43.140 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:54:43.140 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:54:43.140 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:54:43.140 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:54:43.140 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:54:43.140 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:54:43.140 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:54:43.140 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:54:43.823 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:54:44.470 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:54:45.142 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:54:45.861 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:54:46.510 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:54:47.166 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:54:47.811 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:54:48.490 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:54:49.139 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:54:49.139 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 14:54:49.139 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 14:54:49.139 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:54:49.146 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.14 ms, Average NMS time: 0.54 ms, Average inference time: 2.68 ms

2025-11-12 14:54:49.148 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:54:49.229 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:54:49.310 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch61
2025-11-12 14:54:52.450 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 61/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.155s, data_time: 0.002s, total_loss: 6.2, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 2.3, cls_loss: 0.6, lr: 9.959e-04, size: 512, ETA: 0:21:18
2025-11-12 14:54:55.624 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 61/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.154s, data_time: 0.002s, total_loss: 5.5, iou_loss: 2.3, l1_loss: 0.7, conf_loss: 1.9, cls_loss: 0.6, lr: 9.919e-04, size: 544, ETA: 0:21:14
2025-11-12 14:54:58.809 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 61/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.156s, data_time: 0.003s, total_loss: 5.8, iou_loss: 2.3, l1_loss: 0.9, conf_loss: 1.9, cls_loss: 0.7, lr: 9.878e-04, size: 480, ETA: 0:21:11
2025-11-12 14:55:01.869 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 61/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.148s, data_time: 0.002s, total_loss: 4.8, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.6, cls_loss: 0.6, lr: 9.838e-04, size: 288, ETA: 0:21:07
2025-11-12 14:55:05.087 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 61/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.157s, data_time: 0.003s, total_loss: 4.6, iou_loss: 1.9, l1_loss: 0.5, conf_loss: 1.6, cls_loss: 0.5, lr: 9.797e-04, size: 320, ETA: 0:21:03
2025-11-12 14:55:08.271 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 61/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.155s, data_time: 0.002s, total_loss: 4.7, iou_loss: 1.9, l1_loss: 0.6, conf_loss: 1.7, cls_loss: 0.6, lr: 9.756e-04, size: 352, ETA: 0:21:00
2025-11-12 14:55:09.551 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:55:14.539 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:55:15.854 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:55:16.680 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6450
2025-11-12 14:55:16.846 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5809
2025-11-12 14:55:16.934 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4446
2025-11-12 14:55:16.934 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5568
2025-11-12 14:55:16.935 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:55:16.935 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:55:16.935 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.645
2025-11-12 14:55:16.935 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.581
2025-11-12 14:55:16.935 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.445
2025-11-12 14:55:16.935 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.557
2025-11-12 14:55:16.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:55:16.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:55:16.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:55:16.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:55:16.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:55:16.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:55:16.937 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:55:16.937 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:55:16.937 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:55:18.019 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:55:19.034 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:55:20.087 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:55:21.160 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:55:22.202 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:55:23.281 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:55:24.339 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:55:25.407 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:55:26.437 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:55:26.438 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 14:55:26.438 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 14:55:26.438 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:55:26.445 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.10 ms, Average NMS time: 0.57 ms, Average inference time: 2.66 ms

2025-11-12 14:55:26.447 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:55:26.558 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:55:26.638 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch62
2025-11-12 14:55:29.366 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 62/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.134s, data_time: 0.002s, total_loss: 5.4, iou_loss: 2.1, l1_loss: 0.7, conf_loss: 2.0, cls_loss: 0.6, lr: 9.698e-04, size: 352, ETA: 0:20:54
2025-11-12 14:55:32.368 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 62/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.6, l1_loss: 0.8, conf_loss: 1.7, cls_loss: 0.7, lr: 9.657e-04, size: 384, ETA: 0:20:51
2025-11-12 14:55:35.272 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 62/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 4.9, iou_loss: 2.1, l1_loss: 0.7, conf_loss: 1.5, cls_loss: 0.6, lr: 9.617e-04, size: 384, ETA: 0:20:47
2025-11-12 14:55:38.101 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 62/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.138s, data_time: 0.002s, total_loss: 4.5, iou_loss: 1.9, l1_loss: 0.6, conf_loss: 1.4, cls_loss: 0.5, lr: 9.576e-04, size: 352, ETA: 0:20:43
2025-11-12 14:55:40.898 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 62/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.137s, data_time: 0.003s, total_loss: 6.5, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 2.5, cls_loss: 0.7, lr: 9.535e-04, size: 352, ETA: 0:20:39
2025-11-12 14:55:43.731 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 62/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.140s, data_time: 0.002s, total_loss: 5.2, iou_loss: 2.2, l1_loss: 0.7, conf_loss: 1.7, cls_loss: 0.6, lr: 9.495e-04, size: 256, ETA: 0:20:35
2025-11-12 14:55:45.082 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:55:50.012 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:55:51.021 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:55:51.685 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6520
2025-11-12 14:55:51.876 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5861
2025-11-12 14:55:51.914 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4710
2025-11-12 14:55:51.915 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5697
2025-11-12 14:55:51.915 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:55:51.915 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:55:51.915 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.652
2025-11-12 14:55:51.915 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.586
2025-11-12 14:55:51.915 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.471
2025-11-12 14:55:51.915 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.570
2025-11-12 14:55:51.915 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:55:51.915 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:55:51.916 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:55:51.916 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:55:51.916 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:55:51.916 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:55:51.916 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:55:51.916 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:55:51.916 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:55:52.733 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:55:53.583 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:55:54.396 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:55:55.243 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:55:56.060 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:55:56.904 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:55:57.702 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:55:58.561 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:55:59.389 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:55:59.389 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 14:55:59.389 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.57
2025-11-12 14:55:59.389 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:55:59.396 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.16 ms, Average NMS time: 0.57 ms, Average inference time: 2.73 ms

2025-11-12 14:55:59.398 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:55:59.476 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:55:59.557 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch63
2025-11-12 14:56:02.531 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 63/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 5.1, iou_loss: 1.9, l1_loss: 0.8, conf_loss: 1.6, cls_loss: 0.8, lr: 9.436e-04, size: 384, ETA: 0:20:30
2025-11-12 14:56:05.465 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 63/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.144s, data_time: 0.002s, total_loss: 5.3, iou_loss: 2.2, l1_loss: 0.7, conf_loss: 1.8, cls_loss: 0.6, lr: 9.396e-04, size: 352, ETA: 0:20:26
2025-11-12 14:56:08.285 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 63/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.138s, data_time: 0.002s, total_loss: 4.9, iou_loss: 1.9, l1_loss: 0.6, conf_loss: 1.7, cls_loss: 0.7, lr: 9.355e-04, size: 288, ETA: 0:20:23
2025-11-12 14:56:11.163 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 63/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 3.9, iou_loss: 1.6, l1_loss: 0.5, conf_loss: 1.2, cls_loss: 0.5, lr: 9.315e-04, size: 448, ETA: 0:20:19
2025-11-12 14:56:13.969 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 63/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.137s, data_time: 0.003s, total_loss: 5.5, iou_loss: 2.3, l1_loss: 0.7, conf_loss: 1.9, cls_loss: 0.6, lr: 9.274e-04, size: 448, ETA: 0:20:15
2025-11-12 14:56:16.833 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 63/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.140s, data_time: 0.002s, total_loss: 6.9, iou_loss: 2.7, l1_loss: 1.0, conf_loss: 2.5, cls_loss: 0.7, lr: 9.234e-04, size: 416, ETA: 0:20:11
2025-11-12 14:56:18.174 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:56:23.326 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:56:23.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:56:24.287 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6374
2025-11-12 14:56:24.408 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5725
2025-11-12 14:56:24.440 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4245
2025-11-12 14:56:24.441 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5448
2025-11-12 14:56:24.441 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:56:24.441 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:56:24.441 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.637
2025-11-12 14:56:24.441 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.573
2025-11-12 14:56:24.441 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.424
2025-11-12 14:56:24.442 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.545
2025-11-12 14:56:24.442 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:56:24.442 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:56:24.442 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:56:24.442 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:56:24.442 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:56:24.442 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:56:24.442 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:56:24.442 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:56:24.443 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:56:24.952 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:56:25.462 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:56:25.952 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:56:26.438 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:56:26.961 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:56:27.448 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:56:27.934 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:56:28.425 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:56:28.912 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:56:28.913 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.30
2025-11-12 14:56:28.913 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.54
2025-11-12 14:56:28.913 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:56:28.919 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.21 ms, Average NMS time: 0.56 ms, Average inference time: 2.77 ms

2025-11-12 14:56:28.920 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:56:28.996 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:56:29.076 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch64
2025-11-12 14:56:31.832 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 64/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.136s, data_time: 0.002s, total_loss: 4.4, iou_loss: 1.7, l1_loss: 0.5, conf_loss: 1.5, cls_loss: 0.6, lr: 9.175e-04, size: 384, ETA: 0:20:06
2025-11-12 14:56:34.920 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 64/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.002s, total_loss: 6.2, iou_loss: 2.7, l1_loss: 0.8, conf_loss: 1.9, cls_loss: 0.8, lr: 9.135e-04, size: 480, ETA: 0:20:03
2025-11-12 14:56:37.838 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 64/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.143s, data_time: 0.004s, total_loss: 5.3, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 1.8, cls_loss: 0.6, lr: 9.094e-04, size: 448, ETA: 0:19:59
2025-11-12 14:56:40.742 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 64/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.143s, data_time: 0.002s, total_loss: 6.8, iou_loss: 2.8, l1_loss: 0.9, conf_loss: 2.4, cls_loss: 0.7, lr: 9.054e-04, size: 320, ETA: 0:19:55
2025-11-12 14:56:43.600 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 64/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.141s, data_time: 0.003s, total_loss: 3.8, iou_loss: 1.6, l1_loss: 0.5, conf_loss: 1.1, cls_loss: 0.6, lr: 9.013e-04, size: 256, ETA: 0:19:51
2025-11-12 14:56:46.308 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 64/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.133s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.5, l1_loss: 0.7, conf_loss: 1.8, cls_loss: 0.6, lr: 8.973e-04, size: 384, ETA: 0:19:48
2025-11-12 14:56:47.658 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:56:52.697 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:56:53.669 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:56:54.315 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6511
2025-11-12 14:56:54.427 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5770
2025-11-12 14:56:54.466 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4368
2025-11-12 14:56:54.466 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5550
2025-11-12 14:56:54.467 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:56:54.467 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:56:54.467 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.651
2025-11-12 14:56:54.467 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.577
2025-11-12 14:56:54.467 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.437
2025-11-12 14:56:54.467 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.555
2025-11-12 14:56:54.467 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:56:54.467 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:56:54.467 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:56:54.468 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:56:54.468 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:56:54.468 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:56:54.468 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:56:54.468 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:56:54.468 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:56:55.279 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:56:56.055 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:56:56.883 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:56:57.670 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:56:58.483 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:56:59.255 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:57:00.033 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:57:00.851 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:57:01.636 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:57:01.636 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 14:57:01.636 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.55
2025-11-12 14:57:01.636 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:57:01.644 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.12 ms, Average NMS time: 0.57 ms, Average inference time: 2.68 ms

2025-11-12 14:57:01.645 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:57:01.721 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:57:01.803 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch65
2025-11-12 14:57:04.847 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 65/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.148s, data_time: 0.002s, total_loss: 5.5, iou_loss: 2.4, l1_loss: 0.8, conf_loss: 1.7, cls_loss: 0.6, lr: 8.914e-04, size: 480, ETA: 0:19:42
2025-11-12 14:57:07.934 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 65/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.002s, total_loss: 6.5, iou_loss: 2.5, l1_loss: 1.0, conf_loss: 2.3, cls_loss: 0.7, lr: 8.874e-04, size: 512, ETA: 0:19:39
2025-11-12 14:57:10.956 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 65/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 4.8, iou_loss: 2.0, l1_loss: 0.9, conf_loss: 1.3, cls_loss: 0.6, lr: 8.834e-04, size: 416, ETA: 0:19:35
2025-11-12 14:57:14.215 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 65/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.158s, data_time: 0.002s, total_loss: 7.3, iou_loss: 2.7, l1_loss: 0.9, conf_loss: 2.9, cls_loss: 0.8, lr: 8.793e-04, size: 480, ETA: 0:19:32
2025-11-12 14:57:17.350 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 65/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.152s, data_time: 0.003s, total_loss: 5.9, iou_loss: 2.3, l1_loss: 0.9, conf_loss: 2.1, cls_loss: 0.6, lr: 8.753e-04, size: 416, ETA: 0:19:28
2025-11-12 14:57:20.477 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 65/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.152s, data_time: 0.002s, total_loss: 5.8, iou_loss: 2.5, l1_loss: 0.9, conf_loss: 1.7, cls_loss: 0.6, lr: 8.713e-04, size: 512, ETA: 0:19:25
2025-11-12 14:57:21.994 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:57:27.015 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:57:27.846 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:57:28.396 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6392
2025-11-12 14:57:28.548 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5860
2025-11-12 14:57:28.616 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4185
2025-11-12 14:57:28.617 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5479
2025-11-12 14:57:28.617 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:57:28.618 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:57:28.618 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.639
2025-11-12 14:57:28.618 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.586
2025-11-12 14:57:28.618 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.418
2025-11-12 14:57:28.618 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.548
2025-11-12 14:57:28.618 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:57:28.618 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:57:28.618 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:57:28.618 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:57:28.618 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:57:28.619 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:57:28.619 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:57:28.619 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:57:28.619 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:57:29.299 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:57:29.996 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:57:30.676 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:57:31.387 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:57:32.059 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:57:32.737 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:57:33.458 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:57:34.151 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:57:34.846 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:57:34.847 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.30
2025-11-12 14:57:34.847 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.55
2025-11-12 14:57:34.847 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:57:34.854 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.18 ms, Average NMS time: 0.55 ms, Average inference time: 2.72 ms

2025-11-12 14:57:34.855 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:57:34.934 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:57:35.016 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch66
2025-11-12 14:57:37.963 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 66/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.144s, data_time: 0.002s, total_loss: 5.5, iou_loss: 2.4, l1_loss: 0.8, conf_loss: 1.7, cls_loss: 0.6, lr: 8.655e-04, size: 512, ETA: 0:19:20
2025-11-12 14:57:40.927 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 66/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.145s, data_time: 0.002s, total_loss: 6.3, iou_loss: 2.7, l1_loss: 0.8, conf_loss: 2.2, cls_loss: 0.7, lr: 8.614e-04, size: 480, ETA: 0:19:16
2025-11-12 14:57:43.805 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 66/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 4.0, iou_loss: 1.7, l1_loss: 0.8, conf_loss: 0.9, cls_loss: 0.5, lr: 8.574e-04, size: 512, ETA: 0:19:13
2025-11-12 14:57:46.788 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 66/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 6.6, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 2.6, cls_loss: 0.7, lr: 8.534e-04, size: 480, ETA: 0:19:09
2025-11-12 14:57:49.846 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 66/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.150s, data_time: 0.004s, total_loss: 4.9, iou_loss: 1.8, l1_loss: 0.8, conf_loss: 1.7, cls_loss: 0.6, lr: 8.494e-04, size: 512, ETA: 0:19:06
2025-11-12 14:57:52.936 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 66/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.150s, data_time: 0.002s, total_loss: 4.9, iou_loss: 1.6, l1_loss: 0.5, conf_loss: 2.3, cls_loss: 0.5, lr: 8.454e-04, size: 352, ETA: 0:19:02
2025-11-12 14:57:54.306 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:57:59.442 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:58:00.487 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:58:01.143 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6567
2025-11-12 14:58:01.328 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5738
2025-11-12 14:58:01.373 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4495
2025-11-12 14:58:01.374 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5600
2025-11-12 14:58:01.374 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:58:01.374 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:58:01.374 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.657
2025-11-12 14:58:01.374 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.574
2025-11-12 14:58:01.374 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.450
2025-11-12 14:58:01.375 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.560
2025-11-12 14:58:01.375 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:58:01.375 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:58:01.375 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:58:01.375 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:58:01.375 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:58:01.375 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:58:01.375 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:58:01.375 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:58:01.375 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:58:02.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:58:03.093 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:58:03.941 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:58:04.822 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:58:05.674 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:58:06.552 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:58:07.386 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:58:08.256 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:58:09.087 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:58:09.087 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 14:58:09.087 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 14:58:09.087 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:58:09.095 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.21 ms, Average NMS time: 0.59 ms, Average inference time: 2.79 ms

2025-11-12 14:58:09.096 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:58:09.171 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:58:09.252 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch67
2025-11-12 14:58:12.430 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 67/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.155s, data_time: 0.002s, total_loss: 5.2, iou_loss: 2.1, l1_loss: 0.8, conf_loss: 1.5, cls_loss: 0.7, lr: 8.396e-04, size: 320, ETA: 0:18:57
2025-11-12 14:58:15.463 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 67/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.148s, data_time: 0.002s, total_loss: 5.2, iou_loss: 2.1, l1_loss: 0.9, conf_loss: 1.7, cls_loss: 0.6, lr: 8.356e-04, size: 480, ETA: 0:18:54
2025-11-12 14:58:18.609 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 67/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.155s, data_time: 0.002s, total_loss: 6.3, iou_loss: 2.2, l1_loss: 1.1, conf_loss: 2.2, cls_loss: 0.7, lr: 8.316e-04, size: 576, ETA: 0:18:50
2025-11-12 14:58:21.763 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 67/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.155s, data_time: 0.002s, total_loss: 6.4, iou_loss: 2.7, l1_loss: 0.9, conf_loss: 2.2, cls_loss: 0.7, lr: 8.276e-04, size: 288, ETA: 0:18:47
2025-11-12 14:58:24.759 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 67/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.004s, total_loss: 3.9, iou_loss: 1.7, l1_loss: 0.8, conf_loss: 1.0, cls_loss: 0.5, lr: 8.236e-04, size: 544, ETA: 0:18:43
2025-11-12 14:58:27.820 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 67/120, iter: 120/129, gpu mem: 2423Mb, mem: 94.9Gb, iter_time: 0.150s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.3, l1_loss: 0.9, conf_loss: 1.8, cls_loss: 0.6, lr: 8.196e-04, size: 512, ETA: 0:18:40
2025-11-12 14:58:29.205 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:58:34.253 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:58:35.206 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:58:35.801 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6474
2025-11-12 14:58:35.987 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5807
2025-11-12 14:58:36.026 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4527
2025-11-12 14:58:36.026 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5603
2025-11-12 14:58:36.027 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:58:36.027 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:58:36.027 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.647
2025-11-12 14:58:36.027 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.581
2025-11-12 14:58:36.027 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.453
2025-11-12 14:58:36.027 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.560
2025-11-12 14:58:36.027 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:58:36.027 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:58:36.027 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:58:36.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:58:36.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:58:36.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:58:36.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:58:36.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:58:36.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:58:36.791 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:58:37.612 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:58:38.382 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:58:39.148 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:58:39.945 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:58:40.709 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:58:41.526 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:58:42.279 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:58:43.071 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:58:43.071 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 14:58:43.071 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 14:58:43.072 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:58:43.079 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.15 ms, Average NMS time: 0.57 ms, Average inference time: 2.72 ms

2025-11-12 14:58:43.080 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:58:43.156 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:58:43.237 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch68
2025-11-12 14:58:46.151 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 68/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 5.3, iou_loss: 1.8, l1_loss: 1.1, conf_loss: 1.8, cls_loss: 0.5, lr: 8.138e-04, size: 544, ETA: 0:18:34
2025-11-12 14:58:49.433 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 68/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.162s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.3, l1_loss: 0.9, conf_loss: 1.8, cls_loss: 0.6, lr: 8.098e-04, size: 544, ETA: 0:18:31
2025-11-12 14:58:52.892 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 68/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.168s, data_time: 0.002s, total_loss: 4.4, iou_loss: 1.8, l1_loss: 0.6, conf_loss: 1.3, cls_loss: 0.6, lr: 8.058e-04, size: 384, ETA: 0:18:28
2025-11-12 14:58:56.049 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 68/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.154s, data_time: 0.002s, total_loss: 4.6, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.5, cls_loss: 0.6, lr: 8.018e-04, size: 256, ETA: 0:18:25
2025-11-12 14:58:59.329 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 68/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.160s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.2, l1_loss: 1.0, conf_loss: 1.8, cls_loss: 0.6, lr: 7.978e-04, size: 576, ETA: 0:18:21
2025-11-12 14:59:02.569 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 68/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.157s, data_time: 0.002s, total_loss: 5.4, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 1.4, cls_loss: 0.6, lr: 7.939e-04, size: 480, ETA: 0:18:18
2025-11-12 14:59:04.098 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:59:09.159 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:59:10.130 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:59:10.687 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6438
2025-11-12 14:59:10.865 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5785
2025-11-12 14:59:10.900 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4112
2025-11-12 14:59:10.900 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5445
2025-11-12 14:59:10.901 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:59:10.901 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:59:10.901 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.644
2025-11-12 14:59:10.901 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.579
2025-11-12 14:59:10.901 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.411
2025-11-12 14:59:10.901 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.545
2025-11-12 14:59:10.901 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:59:10.901 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:59:10.901 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:59:10.901 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:59:10.902 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:59:10.902 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:59:10.902 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:59:10.902 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:59:10.902 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:59:11.669 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:59:12.464 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:59:13.222 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:59:13.976 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:59:14.776 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:59:15.542 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:59:16.342 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:59:17.102 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:59:17.872 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:59:17.873 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 14:59:17.873 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.54
2025-11-12 14:59:17.873 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:59:17.881 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.11 ms, Average NMS time: 0.56 ms, Average inference time: 2.68 ms

2025-11-12 14:59:17.882 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:59:18.002 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:59:18.082 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch69
2025-11-12 14:59:21.112 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 69/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 4.9, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.7, cls_loss: 0.6, lr: 7.881e-04, size: 288, ETA: 0:18:13
2025-11-12 14:59:23.898 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 69/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.137s, data_time: 0.002s, total_loss: 6.0, iou_loss: 2.1, l1_loss: 1.0, conf_loss: 2.3, cls_loss: 0.6, lr: 7.842e-04, size: 576, ETA: 0:18:09
2025-11-12 14:59:26.900 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 69/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.148s, data_time: 0.002s, total_loss: 6.3, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 2.2, cls_loss: 0.7, lr: 7.802e-04, size: 384, ETA: 0:18:06
2025-11-12 14:59:29.815 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 69/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 4.7, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.4, cls_loss: 0.6, lr: 7.762e-04, size: 256, ETA: 0:18:02
2025-11-12 14:59:32.719 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 69/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.143s, data_time: 0.003s, total_loss: 5.3, iou_loss: 2.2, l1_loss: 0.6, conf_loss: 1.7, cls_loss: 0.8, lr: 7.723e-04, size: 320, ETA: 0:17:59
2025-11-12 14:59:35.433 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 69/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.132s, data_time: 0.002s, total_loss: 6.1, iou_loss: 2.6, l1_loss: 0.8, conf_loss: 2.1, cls_loss: 0.6, lr: 7.683e-04, size: 288, ETA: 0:17:55
2025-11-12 14:59:36.640 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:59:41.646 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 14:59:42.718 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 14:59:43.406 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6430
2025-11-12 14:59:43.619 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5882
2025-11-12 14:59:43.658 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4412
2025-11-12 14:59:43.658 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5574
2025-11-12 14:59:43.658 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 14:59:43.658 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 14:59:43.659 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.643
2025-11-12 14:59:43.659 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.588
2025-11-12 14:59:43.659 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.441
2025-11-12 14:59:43.659 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.557
2025-11-12 14:59:43.659 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 14:59:43.659 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 14:59:43.659 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 14:59:43.659 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 14:59:43.659 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 14:59:43.660 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 14:59:43.660 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 14:59:43.660 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 14:59:43.660 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 14:59:44.518 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 14:59:45.413 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 14:59:46.271 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 14:59:47.158 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 14:59:48.045 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 14:59:48.890 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 14:59:49.790 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 14:59:50.667 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 14:59:51.554 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 14:59:51.554 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 14:59:51.554 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 14:59:51.554 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 14:59:51.561 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.15 ms, Average NMS time: 0.58 ms, Average inference time: 2.73 ms

2025-11-12 14:59:51.563 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:59:51.637 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 14:59:51.717 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch70
2025-11-12 14:59:54.416 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 70/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.134s, data_time: 0.002s, total_loss: 4.3, iou_loss: 2.1, l1_loss: 0.6, conf_loss: 1.1, cls_loss: 0.6, lr: 7.626e-04, size: 384, ETA: 0:17:49
2025-11-12 14:59:57.620 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 70/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.156s, data_time: 0.002s, total_loss: 6.3, iou_loss: 2.1, l1_loss: 1.2, conf_loss: 2.4, cls_loss: 0.6, lr: 7.587e-04, size: 480, ETA: 0:17:46
2025-11-12 15:00:00.812 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 70/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.156s, data_time: 0.002s, total_loss: 7.0, iou_loss: 2.2, l1_loss: 0.9, conf_loss: 3.2, cls_loss: 0.6, lr: 7.547e-04, size: 576, ETA: 0:17:43
2025-11-12 15:00:04.000 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 70/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.157s, data_time: 0.002s, total_loss: 5.4, iou_loss: 2.2, l1_loss: 0.7, conf_loss: 1.8, cls_loss: 0.7, lr: 7.508e-04, size: 352, ETA: 0:17:39
2025-11-12 15:00:06.764 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 70/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.135s, data_time: 0.004s, total_loss: 5.0, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.8, cls_loss: 0.5, lr: 7.469e-04, size: 512, ETA: 0:17:36
2025-11-12 15:00:09.863 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 70/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.152s, data_time: 0.002s, total_loss: 8.6, iou_loss: 2.7, l1_loss: 1.2, conf_loss: 4.0, cls_loss: 0.7, lr: 7.429e-04, size: 544, ETA: 0:17:32
2025-11-12 15:00:11.289 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:00:16.360 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:00:17.144 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:00:17.567 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6486
2025-11-12 15:00:17.676 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5832
2025-11-12 15:00:17.747 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4292
2025-11-12 15:00:17.748 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5537
2025-11-12 15:00:17.748 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:00:17.748 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:00:17.748 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.649
2025-11-12 15:00:17.748 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.583
2025-11-12 15:00:17.748 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.429
2025-11-12 15:00:17.748 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.554
2025-11-12 15:00:17.748 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:00:17.748 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:00:17.749 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:00:17.749 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:00:17.749 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:00:17.749 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:00:17.749 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:00:17.749 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:00:17.749 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:00:18.349 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:00:18.951 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:00:19.547 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:00:20.175 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:00:20.776 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:00:21.375 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:00:21.971 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:00:22.565 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:00:23.194 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:00:23.194 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 15:00:23.194 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.55
2025-11-12 15:00:23.195 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:00:23.201 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.23 ms, Average NMS time: 0.57 ms, Average inference time: 2.80 ms

2025-11-12 15:00:23.202 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:00:23.278 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:00:23.358 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch71
2025-11-12 15:00:26.494 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 71/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.154s, data_time: 0.002s, total_loss: 5.3, iou_loss: 2.0, l1_loss: 0.8, conf_loss: 1.9, cls_loss: 0.6, lr: 7.373e-04, size: 480, ETA: 0:17:27
2025-11-12 15:00:29.742 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 71/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.158s, data_time: 0.002s, total_loss: 6.5, iou_loss: 2.5, l1_loss: 0.9, conf_loss: 2.5, cls_loss: 0.6, lr: 7.333e-04, size: 320, ETA: 0:17:24
2025-11-12 15:00:32.904 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 71/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.154s, data_time: 0.002s, total_loss: 4.5, iou_loss: 2.1, l1_loss: 0.6, conf_loss: 1.2, cls_loss: 0.5, lr: 7.294e-04, size: 352, ETA: 0:17:21
2025-11-12 15:00:36.209 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 71/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.161s, data_time: 0.002s, total_loss: 5.2, iou_loss: 2.1, l1_loss: 0.8, conf_loss: 1.7, cls_loss: 0.6, lr: 7.255e-04, size: 384, ETA: 0:17:17
2025-11-12 15:00:39.262 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 71/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.3, l1_loss: 0.7, conf_loss: 2.3, cls_loss: 0.6, lr: 7.216e-04, size: 288, ETA: 0:17:14
2025-11-12 15:00:42.296 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 71/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.148s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.2, l1_loss: 0.6, conf_loss: 2.1, cls_loss: 0.6, lr: 7.177e-04, size: 288, ETA: 0:17:10
2025-11-12 15:00:43.610 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:00:48.652 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:00:49.594 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:00:50.143 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6461
2025-11-12 15:00:50.299 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5827
2025-11-12 15:00:50.348 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4601
2025-11-12 15:00:50.348 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5630
2025-11-12 15:00:50.348 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:00:50.348 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:00:50.349 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.646
2025-11-12 15:00:50.349 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.583
2025-11-12 15:00:50.349 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.460
2025-11-12 15:00:50.349 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.563
2025-11-12 15:00:50.349 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:00:50.349 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:00:50.349 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:00:50.349 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:00:50.350 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:00:50.350 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:00:50.350 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:00:50.350 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:00:50.350 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:00:51.134 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:00:51.885 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:00:52.636 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:00:53.416 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:00:54.158 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:00:54.948 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:00:55.696 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:00:56.452 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:00:57.257 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:00:57.258 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:00:57.258 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:00:57.258 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:00:57.265 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.07 ms, Average NMS time: 0.56 ms, Average inference time: 2.62 ms

2025-11-12 15:00:57.267 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:00:57.343 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:00:57.424 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch72
2025-11-12 15:01:00.584 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 72/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.156s, data_time: 0.003s, total_loss: 5.3, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 1.8, cls_loss: 0.6, lr: 7.121e-04, size: 480, ETA: 0:17:06
2025-11-12 15:01:03.494 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 72/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.143s, data_time: 0.002s, total_loss: 4.6, iou_loss: 2.1, l1_loss: 0.7, conf_loss: 1.3, cls_loss: 0.5, lr: 7.082e-04, size: 320, ETA: 0:17:02
2025-11-12 15:01:06.524 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 72/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.2, l1_loss: 0.9, conf_loss: 1.9, cls_loss: 0.6, lr: 7.043e-04, size: 448, ETA: 0:16:59
2025-11-12 15:01:09.647 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 72/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.154s, data_time: 0.002s, total_loss: 6.9, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 2.9, cls_loss: 0.7, lr: 7.005e-04, size: 448, ETA: 0:16:55
2025-11-12 15:01:12.900 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 72/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.159s, data_time: 0.002s, total_loss: 4.5, iou_loss: 1.9, l1_loss: 0.6, conf_loss: 1.4, cls_loss: 0.6, lr: 6.966e-04, size: 256, ETA: 0:16:52
2025-11-12 15:01:16.059 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 72/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.153s, data_time: 0.002s, total_loss: 5.2, iou_loss: 2.1, l1_loss: 0.6, conf_loss: 1.9, cls_loss: 0.6, lr: 6.927e-04, size: 384, ETA: 0:16:48
2025-11-12 15:01:17.531 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:01:22.814 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:01:23.846 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:01:24.535 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6508
2025-11-12 15:01:24.671 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5812
2025-11-12 15:01:24.710 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4582
2025-11-12 15:01:24.710 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5634
2025-11-12 15:01:24.710 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:01:24.710 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:01:24.711 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.651
2025-11-12 15:01:24.711 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.581
2025-11-12 15:01:24.711 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.458
2025-11-12 15:01:24.711 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.563
2025-11-12 15:01:24.711 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:01:24.711 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:01:24.711 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:01:24.711 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:01:24.712 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:01:24.712 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:01:24.712 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:01:24.712 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:01:24.712 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:01:25.609 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:01:26.456 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:01:27.337 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:01:28.188 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:01:29.068 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:01:29.920 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:01:30.802 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:01:31.658 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:01:32.564 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:01:32.565 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:01:32.565 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:01:32.565 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:01:32.572 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.21 ms, Average NMS time: 0.62 ms, Average inference time: 2.83 ms

2025-11-12 15:01:32.573 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:01:32.649 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:01:32.730 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch73
2025-11-12 15:01:35.782 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 73/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.002s, total_loss: 4.6, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.3, cls_loss: 0.6, lr: 6.871e-04, size: 512, ETA: 0:16:44
2025-11-12 15:01:38.978 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 73/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.156s, data_time: 0.002s, total_loss: 4.9, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.8, cls_loss: 0.5, lr: 6.833e-04, size: 320, ETA: 0:16:40
2025-11-12 15:01:42.072 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 73/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.150s, data_time: 0.002s, total_loss: 4.2, iou_loss: 1.9, l1_loss: 0.5, conf_loss: 1.2, cls_loss: 0.5, lr: 6.794e-04, size: 256, ETA: 0:16:37
2025-11-12 15:01:45.393 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 73/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.162s, data_time: 0.002s, total_loss: 5.1, iou_loss: 2.0, l1_loss: 0.9, conf_loss: 1.6, cls_loss: 0.6, lr: 6.756e-04, size: 416, ETA: 0:16:34
2025-11-12 15:01:48.591 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 73/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.156s, data_time: 0.002s, total_loss: 6.3, iou_loss: 2.2, l1_loss: 1.1, conf_loss: 2.4, cls_loss: 0.6, lr: 6.717e-04, size: 512, ETA: 0:16:30
2025-11-12 15:01:51.922 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 73/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.163s, data_time: 0.002s, total_loss: 5.1, iou_loss: 2.2, l1_loss: 0.7, conf_loss: 1.6, cls_loss: 0.6, lr: 6.679e-04, size: 448, ETA: 0:16:27
2025-11-12 15:01:53.425 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:01:58.536 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:01:59.560 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:02:00.182 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6456
2025-11-12 15:02:00.356 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5756
2025-11-12 15:02:00.387 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4111
2025-11-12 15:02:00.388 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5441
2025-11-12 15:02:00.388 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:02:00.388 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:02:00.388 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.646
2025-11-12 15:02:00.388 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.576
2025-11-12 15:02:00.388 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.411
2025-11-12 15:02:00.388 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.544
2025-11-12 15:02:00.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:02:00.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:02:00.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:02:00.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:02:00.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:02:00.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:02:00.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:02:00.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:02:00.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:02:01.215 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:02:02.062 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:02:02.867 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:02:03.700 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:02:04.510 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:02:05.350 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:02:06.142 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:02:06.954 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:02:07.808 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:02:07.808 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 15:02:07.808 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.54
2025-11-12 15:02:07.809 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:02:07.815 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.24 ms, Average NMS time: 0.59 ms, Average inference time: 2.83 ms

2025-11-12 15:02:07.817 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:02:07.896 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:02:07.978 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch74
2025-11-12 15:02:11.110 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 74/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.154s, data_time: 0.002s, total_loss: 5.1, iou_loss: 1.9, l1_loss: 0.8, conf_loss: 1.8, cls_loss: 0.5, lr: 6.624e-04, size: 512, ETA: 0:16:22
2025-11-12 15:02:14.258 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 74/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.153s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.2, l1_loss: 0.6, conf_loss: 2.1, cls_loss: 0.7, lr: 6.586e-04, size: 256, ETA: 0:16:19
2025-11-12 15:02:17.458 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 74/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.156s, data_time: 0.002s, total_loss: 5.5, iou_loss: 2.2, l1_loss: 0.9, conf_loss: 1.8, cls_loss: 0.6, lr: 6.547e-04, size: 544, ETA: 0:16:15
2025-11-12 15:02:20.656 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 74/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.157s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.3, l1_loss: 0.9, conf_loss: 1.9, cls_loss: 0.6, lr: 6.509e-04, size: 544, ETA: 0:16:12
2025-11-12 15:02:23.532 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 74/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.141s, data_time: 0.003s, total_loss: 5.5, iou_loss: 2.1, l1_loss: 0.8, conf_loss: 2.0, cls_loss: 0.6, lr: 6.471e-04, size: 256, ETA: 0:16:09
2025-11-12 15:02:26.549 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 74/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 5.3, iou_loss: 2.3, l1_loss: 0.7, conf_loss: 1.6, cls_loss: 0.6, lr: 6.433e-04, size: 288, ETA: 0:16:05
2025-11-12 15:02:27.781 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:02:32.800 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:02:33.729 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:02:34.320 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6477
2025-11-12 15:02:34.444 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5744
2025-11-12 15:02:34.497 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4536
2025-11-12 15:02:34.497 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5586
2025-11-12 15:02:34.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:02:34.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:02:34.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.648
2025-11-12 15:02:34.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.574
2025-11-12 15:02:34.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.454
2025-11-12 15:02:34.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.559
2025-11-12 15:02:34.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:02:34.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:02:34.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:02:34.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:02:34.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:02:34.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:02:34.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:02:34.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:02:34.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:02:35.285 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:02:36.014 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:02:36.743 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:02:37.511 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:02:38.251 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:02:39.022 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:02:39.745 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:02:40.473 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:02:41.250 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:02:41.250 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 15:02:41.251 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:02:41.251 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:02:41.258 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.20 ms, Average NMS time: 0.56 ms, Average inference time: 2.76 ms

2025-11-12 15:02:41.259 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:02:41.334 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:02:41.415 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch75
2025-11-12 15:02:44.337 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 75/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 6.3, iou_loss: 2.6, l1_loss: 0.8, conf_loss: 2.3, cls_loss: 0.6, lr: 6.378e-04, size: 256, ETA: 0:16:00
2025-11-12 15:02:47.203 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 75/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.141s, data_time: 0.002s, total_loss: 6.2, iou_loss: 2.6, l1_loss: 0.9, conf_loss: 2.0, cls_loss: 0.7, lr: 6.341e-04, size: 576, ETA: 0:15:57
2025-11-12 15:02:50.095 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 75/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 6.1, iou_loss: 2.4, l1_loss: 0.8, conf_loss: 2.2, cls_loss: 0.7, lr: 6.303e-04, size: 416, ETA: 0:15:53
2025-11-12 15:02:52.982 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 75/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.142s, data_time: 0.003s, total_loss: 5.6, iou_loss: 2.3, l1_loss: 0.9, conf_loss: 1.7, cls_loss: 0.6, lr: 6.265e-04, size: 512, ETA: 0:15:50
2025-11-12 15:02:55.768 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 75/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.137s, data_time: 0.002s, total_loss: 3.9, iou_loss: 1.8, l1_loss: 0.5, conf_loss: 1.0, cls_loss: 0.5, lr: 6.228e-04, size: 320, ETA: 0:15:46
2025-11-12 15:02:58.638 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 75/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.140s, data_time: 0.002s, total_loss: 3.8, iou_loss: 1.8, l1_loss: 0.5, conf_loss: 0.9, cls_loss: 0.5, lr: 6.190e-04, size: 384, ETA: 0:15:42
2025-11-12 15:02:59.935 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:03:05.014 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:03:06.361 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:03:07.245 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6517
2025-11-12 15:03:07.426 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.6056
2025-11-12 15:03:07.524 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4755
2025-11-12 15:03:07.525 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5776
2025-11-12 15:03:07.525 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:03:07.525 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:03:07.525 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.652
2025-11-12 15:03:07.526 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.606
2025-11-12 15:03:07.526 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.476
2025-11-12 15:03:07.526 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.578
2025-11-12 15:03:07.526 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:03:07.526 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:03:07.526 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:03:07.526 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:03:07.526 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:03:07.526 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:03:07.527 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:03:07.527 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:03:07.527 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:03:08.603 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:03:09.702 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:03:10.800 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:03:11.902 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:03:13.016 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:03:14.121 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:03:15.225 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:03:16.321 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:03:17.422 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:03:17.423 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:03:17.423 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.58
2025-11-12 15:03:17.423 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:03:17.430 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.10 ms, Average NMS time: 0.57 ms, Average inference time: 2.68 ms

2025-11-12 15:03:17.432 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:03:17.511 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:03:17.594 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch76
2025-11-12 15:03:20.632 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 76/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.148s, data_time: 0.002s, total_loss: 3.3, iou_loss: 1.5, l1_loss: 0.4, conf_loss: 0.9, cls_loss: 0.5, lr: 6.136e-04, size: 384, ETA: 0:15:37
2025-11-12 15:03:23.977 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 76/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.163s, data_time: 0.002s, total_loss: 5.3, iou_loss: 2.3, l1_loss: 0.7, conf_loss: 1.6, cls_loss: 0.7, lr: 6.098e-04, size: 256, ETA: 0:15:34
2025-11-12 15:03:27.090 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 76/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.3, l1_loss: 0.7, conf_loss: 2.2, cls_loss: 0.7, lr: 6.061e-04, size: 384, ETA: 0:15:31
2025-11-12 15:03:30.390 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 76/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.161s, data_time: 0.002s, total_loss: 5.8, iou_loss: 2.2, l1_loss: 0.9, conf_loss: 2.0, cls_loss: 0.6, lr: 6.024e-04, size: 544, ETA: 0:15:28
2025-11-12 15:03:33.705 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 76/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.162s, data_time: 0.002s, total_loss: 4.7, iou_loss: 2.1, l1_loss: 0.7, conf_loss: 1.3, cls_loss: 0.6, lr: 5.986e-04, size: 480, ETA: 0:15:24
2025-11-12 15:03:36.696 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 76/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.145s, data_time: 0.002s, total_loss: 5.2, iou_loss: 2.0, l1_loss: 0.9, conf_loss: 1.8, cls_loss: 0.6, lr: 5.949e-04, size: 544, ETA: 0:15:21
2025-11-12 15:03:38.025 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:03:43.148 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:03:44.176 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:03:44.794 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6477
2025-11-12 15:03:44.989 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5964
2025-11-12 15:03:45.028 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4491
2025-11-12 15:03:45.029 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5644
2025-11-12 15:03:45.029 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:03:45.029 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:03:45.029 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.648
2025-11-12 15:03:45.029 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.596
2025-11-12 15:03:45.029 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.449
2025-11-12 15:03:45.029 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.564
2025-11-12 15:03:45.029 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:03:45.030 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:03:45.030 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:03:45.030 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:03:45.030 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:03:45.030 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:03:45.030 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:03:45.030 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:03:45.030 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:03:45.920 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:03:46.758 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:03:47.634 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:03:48.476 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:03:49.352 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:03:50.202 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:03:51.097 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:03:51.959 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:03:52.837 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:03:52.837 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:03:52.838 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:03:52.838 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:03:52.845 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.15 ms, Average NMS time: 0.59 ms, Average inference time: 2.75 ms

2025-11-12 15:03:52.846 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:03:52.921 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:03:53.002 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch77
2025-11-12 15:03:55.951 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 77/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.145s, data_time: 0.002s, total_loss: 4.5, iou_loss: 2.0, l1_loss: 0.8, conf_loss: 1.2, cls_loss: 0.6, lr: 5.896e-04, size: 384, ETA: 0:15:16
2025-11-12 15:03:59.000 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 77/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.4, l1_loss: 0.7, conf_loss: 2.1, cls_loss: 0.7, lr: 5.859e-04, size: 576, ETA: 0:15:13
2025-11-12 15:04:02.000 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 77/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 2.7, iou_loss: 1.2, l1_loss: 0.4, conf_loss: 0.7, cls_loss: 0.4, lr: 5.822e-04, size: 288, ETA: 0:15:09
2025-11-12 15:04:05.069 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 77/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.150s, data_time: 0.003s, total_loss: 5.3, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 2.2, cls_loss: 0.5, lr: 5.785e-04, size: 512, ETA: 0:15:06
2025-11-12 15:04:08.203 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 77/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.154s, data_time: 0.002s, total_loss: 5.5, iou_loss: 2.1, l1_loss: 1.1, conf_loss: 1.7, cls_loss: 0.6, lr: 5.748e-04, size: 544, ETA: 0:15:02
2025-11-12 15:04:11.058 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 77/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.140s, data_time: 0.003s, total_loss: 4.0, iou_loss: 1.8, l1_loss: 0.6, conf_loss: 1.1, cls_loss: 0.5, lr: 5.711e-04, size: 416, ETA: 0:14:59
2025-11-12 15:04:12.318 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:04:17.283 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:04:18.215 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:04:18.737 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6451
2025-11-12 15:04:18.896 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5831
2025-11-12 15:04:18.935 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4513
2025-11-12 15:04:18.935 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5598
2025-11-12 15:04:18.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:04:18.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:04:18.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.645
2025-11-12 15:04:18.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.583
2025-11-12 15:04:18.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.451
2025-11-12 15:04:18.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.560
2025-11-12 15:04:18.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:04:18.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:04:18.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:04:18.937 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:04:18.937 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:04:18.937 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:04:18.937 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:04:18.937 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:04:18.937 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:04:19.677 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:04:20.440 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:04:21.158 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:04:21.877 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:04:22.644 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:04:23.360 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:04:24.084 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:04:24.833 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:04:25.603 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:04:25.603 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 15:04:25.603 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:04:25.603 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:04:25.611 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.17 ms, Average NMS time: 0.54 ms, Average inference time: 2.70 ms

2025-11-12 15:04:25.612 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:04:25.689 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:04:25.773 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch78
2025-11-12 15:04:28.939 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 78/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.154s, data_time: 0.002s, total_loss: 4.7, iou_loss: 1.9, l1_loss: 0.9, conf_loss: 1.5, cls_loss: 0.5, lr: 5.658e-04, size: 480, ETA: 0:14:54
2025-11-12 15:04:32.011 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 78/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 4.1, iou_loss: 1.9, l1_loss: 0.6, conf_loss: 1.1, cls_loss: 0.5, lr: 5.622e-04, size: 352, ETA: 0:14:51
2025-11-12 15:04:35.062 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 78/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 5.8, iou_loss: 2.4, l1_loss: 0.9, conf_loss: 1.7, cls_loss: 0.7, lr: 5.585e-04, size: 576, ETA: 0:14:47
2025-11-12 15:04:38.223 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 78/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.155s, data_time: 0.002s, total_loss: 6.0, iou_loss: 2.3, l1_loss: 1.1, conf_loss: 1.8, cls_loss: 0.7, lr: 5.549e-04, size: 576, ETA: 0:14:44
2025-11-12 15:04:41.264 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 78/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.148s, data_time: 0.002s, total_loss: 6.9, iou_loss: 2.5, l1_loss: 1.0, conf_loss: 2.7, cls_loss: 0.6, lr: 5.513e-04, size: 416, ETA: 0:14:41
2025-11-12 15:04:44.448 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 78/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.156s, data_time: 0.002s, total_loss: 5.2, iou_loss: 2.2, l1_loss: 0.7, conf_loss: 1.6, cls_loss: 0.7, lr: 5.476e-04, size: 256, ETA: 0:14:37
2025-11-12 15:04:45.705 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:04:50.758 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:04:51.908 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:04:52.648 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6528
2025-11-12 15:04:52.800 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.6005
2025-11-12 15:04:52.885 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4629
2025-11-12 15:04:52.886 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5721
2025-11-12 15:04:52.886 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:04:52.886 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:04:52.886 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.653
2025-11-12 15:04:52.887 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.600
2025-11-12 15:04:52.887 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.463
2025-11-12 15:04:52.887 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.572
2025-11-12 15:04:52.887 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:04:52.887 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:04:52.887 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:04:52.887 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:04:52.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:04:52.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:04:52.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:04:52.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:04:52.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:04:53.806 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:04:54.773 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:04:55.713 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:04:56.687 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:04:57.675 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:04:58.591 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:04:59.543 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:05:00.474 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:05:01.438 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:05:01.438 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:05:01.439 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.57
2025-11-12 15:05:01.439 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:05:01.446 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.20 ms, Average NMS time: 0.57 ms, Average inference time: 2.76 ms

2025-11-12 15:05:01.447 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:05:01.525 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:05:01.608 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch79
2025-11-12 15:05:04.766 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 79/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.154s, data_time: 0.002s, total_loss: 5.2, iou_loss: 2.2, l1_loss: 0.7, conf_loss: 1.7, cls_loss: 0.7, lr: 5.424e-04, size: 352, ETA: 0:14:32
2025-11-12 15:05:07.830 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 79/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 4.9, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.7, cls_loss: 0.5, lr: 5.388e-04, size: 512, ETA: 0:14:29
2025-11-12 15:05:11.049 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 79/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.157s, data_time: 0.002s, total_loss: 5.2, iou_loss: 2.1, l1_loss: 0.8, conf_loss: 1.8, cls_loss: 0.5, lr: 5.352e-04, size: 448, ETA: 0:14:26
2025-11-12 15:05:14.449 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 79/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.166s, data_time: 0.002s, total_loss: 5.2, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 2.0, cls_loss: 0.5, lr: 5.316e-04, size: 480, ETA: 0:14:23
2025-11-12 15:05:17.611 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 79/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.154s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.2, l1_loss: 0.7, conf_loss: 2.1, cls_loss: 0.6, lr: 5.280e-04, size: 384, ETA: 0:14:19
2025-11-12 15:05:20.576 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 79/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.145s, data_time: 0.002s, total_loss: 4.8, iou_loss: 2.1, l1_loss: 0.8, conf_loss: 1.4, cls_loss: 0.6, lr: 5.244e-04, size: 512, ETA: 0:14:16
2025-11-12 15:05:22.019 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:05:27.069 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:05:27.885 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:05:28.401 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6396
2025-11-12 15:05:28.526 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5893
2025-11-12 15:05:28.560 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4340
2025-11-12 15:05:28.560 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5543
2025-11-12 15:05:28.561 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:05:28.561 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:05:28.561 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.640
2025-11-12 15:05:28.561 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.589
2025-11-12 15:05:28.561 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.434
2025-11-12 15:05:28.561 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.554
2025-11-12 15:05:28.561 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:05:28.561 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:05:28.561 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:05:28.562 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:05:28.562 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:05:28.562 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:05:28.562 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:05:28.562 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:05:28.562 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:05:29.261 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:05:29.930 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:05:30.596 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:05:31.257 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:05:31.955 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:05:32.620 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:05:33.289 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:05:33.993 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:05:34.667 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:05:34.667 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.31
2025-11-12 15:05:34.667 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.55
2025-11-12 15:05:34.667 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:05:34.675 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.26 ms, Average NMS time: 0.57 ms, Average inference time: 2.82 ms

2025-11-12 15:05:34.677 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:05:34.752 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:05:34.833 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch80
2025-11-12 15:05:37.893 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 80/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 3.8, iou_loss: 1.5, l1_loss: 0.4, conf_loss: 1.4, cls_loss: 0.6, lr: 5.193e-04, size: 320, ETA: 0:14:11
2025-11-12 15:05:40.994 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 80/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.002s, total_loss: 3.9, iou_loss: 1.7, l1_loss: 0.7, conf_loss: 1.0, cls_loss: 0.5, lr: 5.157e-04, size: 512, ETA: 0:14:08
2025-11-12 15:05:44.419 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 80/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.167s, data_time: 0.002s, total_loss: 4.5, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.3, cls_loss: 0.5, lr: 5.122e-04, size: 352, ETA: 0:14:04
2025-11-12 15:05:47.753 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 80/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.162s, data_time: 0.002s, total_loss: 5.0, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.8, cls_loss: 0.6, lr: 5.086e-04, size: 256, ETA: 0:14:01
2025-11-12 15:05:50.850 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 80/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.150s, data_time: 0.002s, total_loss: 4.7, iou_loss: 2.1, l1_loss: 0.7, conf_loss: 1.3, cls_loss: 0.5, lr: 5.051e-04, size: 320, ETA: 0:13:58
2025-11-12 15:05:54.075 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 80/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.157s, data_time: 0.002s, total_loss: 3.2, iou_loss: 1.3, l1_loss: 0.5, conf_loss: 1.0, cls_loss: 0.4, lr: 5.016e-04, size: 416, ETA: 0:13:55
2025-11-12 15:05:55.535 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:06:00.636 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:06:01.698 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:06:02.406 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6545
2025-11-12 15:06:02.544 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.6016
2025-11-12 15:06:02.587 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4523
2025-11-12 15:06:02.587 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5695
2025-11-12 15:06:02.587 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:06:02.588 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:06:02.588 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.655
2025-11-12 15:06:02.588 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.602
2025-11-12 15:06:02.588 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.452
2025-11-12 15:06:02.588 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.569
2025-11-12 15:06:02.588 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:06:02.588 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:06:02.588 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:06:02.588 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:06:02.588 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:06:02.589 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:06:02.589 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:06:02.589 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:06:02.589 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:06:03.496 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:06:04.364 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:06:05.270 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:06:06.146 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:06:07.053 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:06:07.914 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:06:09.082 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:06:09.993 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:06:10.897 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:06:10.897 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:06:10.897 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.57
2025-11-12 15:06:10.897 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:06:10.906 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.20 ms, Average NMS time: 0.58 ms, Average inference time: 2.78 ms

2025-11-12 15:06:10.907 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:06:10.987 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:06:11.068 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch81
2025-11-12 15:06:13.967 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 81/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.144s, data_time: 0.003s, total_loss: 6.1, iou_loss: 2.3, l1_loss: 0.7, conf_loss: 2.3, cls_loss: 0.8, lr: 4.965e-04, size: 288, ETA: 0:13:50
2025-11-12 15:06:16.910 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 81/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 4.7, iou_loss: 1.8, l1_loss: 0.6, conf_loss: 1.7, cls_loss: 0.5, lr: 4.930e-04, size: 512, ETA: 0:13:46
2025-11-12 15:06:19.704 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 81/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.138s, data_time: 0.002s, total_loss: 5.2, iou_loss: 2.3, l1_loss: 0.7, conf_loss: 1.6, cls_loss: 0.7, lr: 4.895e-04, size: 320, ETA: 0:13:43
2025-11-12 15:06:22.415 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 81/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.135s, data_time: 0.004s, total_loss: 6.9, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 2.9, cls_loss: 0.7, lr: 4.860e-04, size: 384, ETA: 0:13:39
2025-11-12 15:06:25.372 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 81/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 4.7, iou_loss: 2.0, l1_loss: 0.8, conf_loss: 1.3, cls_loss: 0.5, lr: 4.825e-04, size: 448, ETA: 0:13:36
2025-11-12 15:06:28.186 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 81/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.139s, data_time: 0.002s, total_loss: 4.4, iou_loss: 1.9, l1_loss: 0.6, conf_loss: 1.4, cls_loss: 0.5, lr: 4.791e-04, size: 448, ETA: 0:13:33
2025-11-12 15:06:29.659 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:06:34.786 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:06:36.263 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:06:37.241 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6503
2025-11-12 15:06:37.482 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5911
2025-11-12 15:06:37.526 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4488
2025-11-12 15:06:37.527 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5634
2025-11-12 15:06:37.527 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:06:37.527 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:06:37.527 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.650
2025-11-12 15:06:37.527 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.591
2025-11-12 15:06:37.527 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.449
2025-11-12 15:06:37.527 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.563
2025-11-12 15:06:37.528 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:06:37.528 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:06:37.528 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:06:37.528 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:06:37.528 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:06:37.528 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:06:37.528 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:06:37.528 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:06:37.528 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:06:38.777 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:06:40.029 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:06:41.303 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:06:42.547 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:06:43.749 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:06:44.990 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:06:46.253 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:06:47.496 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:06:48.713 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:06:48.713 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:06:48.713 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:06:48.714 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:06:48.721 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.19 ms, Average NMS time: 0.61 ms, Average inference time: 2.80 ms

2025-11-12 15:06:48.723 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:06:48.798 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:06:48.878 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch82
2025-11-12 15:06:51.922 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 82/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.150s, data_time: 0.001s, total_loss: 4.9, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.6, cls_loss: 0.6, lr: 4.740e-04, size: 416, ETA: 0:13:28
2025-11-12 15:06:55.013 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 82/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.001s, total_loss: 5.2, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 2.0, cls_loss: 0.6, lr: 4.706e-04, size: 320, ETA: 0:13:24
2025-11-12 15:06:58.261 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 82/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.158s, data_time: 0.001s, total_loss: 4.5, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.3, cls_loss: 0.5, lr: 4.672e-04, size: 320, ETA: 0:13:21
2025-11-12 15:07:01.681 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 82/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.167s, data_time: 0.002s, total_loss: 5.2, iou_loss: 1.9, l1_loss: 0.7, conf_loss: 2.0, cls_loss: 0.6, lr: 4.637e-04, size: 512, ETA: 0:13:18
2025-11-12 15:07:04.851 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 82/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.155s, data_time: 0.001s, total_loss: 5.2, iou_loss: 2.1, l1_loss: 0.8, conf_loss: 1.8, cls_loss: 0.6, lr: 4.603e-04, size: 288, ETA: 0:13:15
2025-11-12 15:07:07.994 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 82/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.153s, data_time: 0.001s, total_loss: 2.8, iou_loss: 1.3, l1_loss: 0.4, conf_loss: 0.7, cls_loss: 0.4, lr: 4.569e-04, size: 256, ETA: 0:13:11
2025-11-12 15:07:09.502 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:07:14.483 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:07:15.521 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:07:16.208 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6483
2025-11-12 15:07:16.361 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5779
2025-11-12 15:07:16.435 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4630
2025-11-12 15:07:16.436 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5631
2025-11-12 15:07:16.436 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:07:16.436 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:07:16.436 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.648
2025-11-12 15:07:16.437 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.578
2025-11-12 15:07:16.437 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.463
2025-11-12 15:07:16.437 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.563
2025-11-12 15:07:16.437 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:07:16.437 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:07:16.437 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:07:16.437 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:07:16.437 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:07:16.437 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:07:16.437 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:07:16.438 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:07:16.438 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:07:17.306 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:07:18.202 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:07:19.058 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:07:19.943 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:07:20.805 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:07:21.684 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:07:22.534 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:07:23.417 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:07:24.270 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:07:24.271 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:07:24.271 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:07:24.271 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:07:24.278 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.15 ms, Average NMS time: 0.56 ms, Average inference time: 2.71 ms

2025-11-12 15:07:24.280 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:07:24.355 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:07:24.436 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch83
2025-11-12 15:07:27.188 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 83/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.136s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.3, l1_loss: 0.7, conf_loss: 2.0, cls_loss: 0.6, lr: 4.520e-04, size: 352, ETA: 0:13:07
2025-11-12 15:07:30.236 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 83/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.150s, data_time: 0.002s, total_loss: 5.3, iou_loss: 2.4, l1_loss: 0.8, conf_loss: 1.5, cls_loss: 0.6, lr: 4.486e-04, size: 384, ETA: 0:13:03
2025-11-12 15:07:33.207 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 83/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.145s, data_time: 0.002s, total_loss: 4.9, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.7, cls_loss: 0.6, lr: 4.452e-04, size: 384, ETA: 0:13:00
2025-11-12 15:07:36.139 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 83/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.144s, data_time: 0.006s, total_loss: 6.4, iou_loss: 2.4, l1_loss: 1.1, conf_loss: 2.2, cls_loss: 0.7, lr: 4.418e-04, size: 576, ETA: 0:12:56
2025-11-12 15:07:39.135 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 83/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 5.4, iou_loss: 2.3, l1_loss: 0.8, conf_loss: 1.8, cls_loss: 0.6, lr: 4.385e-04, size: 416, ETA: 0:12:53
2025-11-12 15:07:42.307 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 83/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.155s, data_time: 0.002s, total_loss: 4.9, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.7, cls_loss: 0.6, lr: 4.351e-04, size: 448, ETA: 0:12:50
2025-11-12 15:07:43.608 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:07:48.574 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:07:49.714 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:07:50.435 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6446
2025-11-12 15:07:50.643 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.6006
2025-11-12 15:07:50.687 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4480
2025-11-12 15:07:50.688 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5644
2025-11-12 15:07:50.688 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:07:50.688 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:07:50.688 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.645
2025-11-12 15:07:50.688 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.601
2025-11-12 15:07:50.688 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.448
2025-11-12 15:07:50.688 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.564
2025-11-12 15:07:50.689 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:07:50.689 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:07:50.689 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:07:50.689 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:07:50.689 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:07:50.689 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:07:50.689 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:07:50.689 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:07:50.690 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:07:51.606 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:07:52.547 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:07:53.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:07:54.402 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:07:55.338 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:07:56.257 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:07:57.208 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:07:58.128 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:07:59.080 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:07:59.080 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:07:59.081 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:07:59.081 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:07:59.088 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.13 ms, Average NMS time: 0.56 ms, Average inference time: 2.69 ms

2025-11-12 15:07:59.089 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:07:59.165 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:07:59.246 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch84
2025-11-12 15:08:02.217 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 84/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 4.6, iou_loss: 1.7, l1_loss: 0.7, conf_loss: 1.8, cls_loss: 0.5, lr: 4.303e-04, size: 480, ETA: 0:12:45
2025-11-12 15:08:05.264 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 84/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.1, l1_loss: 0.9, conf_loss: 2.0, cls_loss: 0.7, lr: 4.269e-04, size: 480, ETA: 0:12:42
2025-11-12 15:08:08.395 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 84/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.155s, data_time: 0.003s, total_loss: 5.0, iou_loss: 1.9, l1_loss: 0.7, conf_loss: 1.7, cls_loss: 0.6, lr: 4.236e-04, size: 384, ETA: 0:12:38
2025-11-12 15:08:11.382 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 84/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.004s, total_loss: 5.3, iou_loss: 2.1, l1_loss: 0.7, conf_loss: 1.9, cls_loss: 0.7, lr: 4.203e-04, size: 480, ETA: 0:12:35
2025-11-12 15:08:14.266 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 84/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.141s, data_time: 0.002s, total_loss: 4.5, iou_loss: 2.2, l1_loss: 0.7, conf_loss: 1.1, cls_loss: 0.6, lr: 4.170e-04, size: 256, ETA: 0:12:32
2025-11-12 15:08:17.175 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 84/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 4.6, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.3, cls_loss: 0.7, lr: 4.137e-04, size: 416, ETA: 0:12:28
2025-11-12 15:08:18.602 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:08:23.526 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:08:24.374 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:08:24.920 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6522
2025-11-12 15:08:25.040 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5775
2025-11-12 15:08:25.078 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4528
2025-11-12 15:08:25.079 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5608
2025-11-12 15:08:25.079 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:08:25.079 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:08:25.079 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.652
2025-11-12 15:08:25.079 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.577
2025-11-12 15:08:25.080 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.453
2025-11-12 15:08:25.080 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.561
2025-11-12 15:08:25.080 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:08:25.080 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:08:25.080 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:08:25.080 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:08:25.080 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:08:25.080 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:08:25.080 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:08:25.080 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:08:25.081 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:08:25.810 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:08:26.498 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:08:27.187 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:08:27.910 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:08:28.594 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:08:29.290 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:08:30.007 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:08:30.689 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:08:31.375 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:08:31.375 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:08:31.375 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:08:31.375 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:08:31.382 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.18 ms, Average NMS time: 0.52 ms, Average inference time: 2.70 ms

2025-11-12 15:08:31.384 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:08:31.463 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:08:31.544 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch85
2025-11-12 15:08:34.461 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 85/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.142s, data_time: 0.003s, total_loss: 5.8, iou_loss: 2.3, l1_loss: 1.0, conf_loss: 1.9, cls_loss: 0.6, lr: 4.089e-04, size: 448, ETA: 0:12:23
2025-11-12 15:08:37.318 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 85/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.140s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.0, l1_loss: 0.8, conf_loss: 2.2, cls_loss: 0.5, lr: 4.057e-04, size: 480, ETA: 0:12:20
2025-11-12 15:08:40.408 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 85/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.002s, total_loss: 4.7, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.4, cls_loss: 0.5, lr: 4.024e-04, size: 480, ETA: 0:12:17
2025-11-12 15:08:43.343 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 85/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.145s, data_time: 0.005s, total_loss: 5.0, iou_loss: 2.1, l1_loss: 0.8, conf_loss: 1.6, cls_loss: 0.6, lr: 3.992e-04, size: 512, ETA: 0:12:13
2025-11-12 15:08:46.126 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 85/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.137s, data_time: 0.002s, total_loss: 4.3, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.1, cls_loss: 0.6, lr: 3.959e-04, size: 352, ETA: 0:12:10
2025-11-12 15:08:49.147 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 85/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.148s, data_time: 0.002s, total_loss: 6.8, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 2.8, cls_loss: 0.6, lr: 3.927e-04, size: 320, ETA: 0:12:07
2025-11-12 15:08:50.450 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:08:55.465 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:08:56.445 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:08:57.080 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6463
2025-11-12 15:08:57.287 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5934
2025-11-12 15:08:57.357 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4363
2025-11-12 15:08:57.358 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5587
2025-11-12 15:08:57.358 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:08:57.358 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:08:57.358 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.646
2025-11-12 15:08:57.359 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.593
2025-11-12 15:08:57.359 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.436
2025-11-12 15:08:57.359 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.559
2025-11-12 15:08:57.359 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:08:57.359 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:08:57.359 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:08:57.360 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:08:57.360 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:08:57.360 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:08:57.360 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:08:57.360 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:08:57.361 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:08:58.186 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:08:59.024 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:08:59.817 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:09:00.656 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:09:01.450 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:09:02.262 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:09:03.122 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:09:03.923 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:09:04.756 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:09:04.756 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:09:04.756 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:09:04.756 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:09:04.763 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.13 ms, Average NMS time: 0.59 ms, Average inference time: 2.73 ms

2025-11-12 15:09:04.765 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:09:04.841 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:09:04.923 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch86
2025-11-12 15:09:08.030 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 86/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.154s, data_time: 0.002s, total_loss: 4.8, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 1.2, cls_loss: 0.6, lr: 3.880e-04, size: 448, ETA: 0:12:02
2025-11-12 15:09:10.994 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 86/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.004s, total_loss: 5.6, iou_loss: 2.6, l1_loss: 0.7, conf_loss: 1.6, cls_loss: 0.7, lr: 3.848e-04, size: 256, ETA: 0:11:58
2025-11-12 15:09:13.840 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 86/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.140s, data_time: 0.002s, total_loss: 7.1, iou_loss: 2.4, l1_loss: 1.1, conf_loss: 2.9, cls_loss: 0.7, lr: 3.816e-04, size: 544, ETA: 0:11:55
2025-11-12 15:09:16.829 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 86/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 4.3, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.2, cls_loss: 0.6, lr: 3.784e-04, size: 384, ETA: 0:11:52
2025-11-12 15:09:19.834 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 86/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 5.3, iou_loss: 2.2, l1_loss: 0.7, conf_loss: 1.8, cls_loss: 0.6, lr: 3.753e-04, size: 512, ETA: 0:11:48
2025-11-12 15:09:22.774 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 86/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.144s, data_time: 0.002s, total_loss: 4.5, iou_loss: 1.8, l1_loss: 0.8, conf_loss: 1.3, cls_loss: 0.6, lr: 3.721e-04, size: 512, ETA: 0:11:45
2025-11-12 15:09:24.196 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:09:29.195 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:09:30.072 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:09:30.568 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6501
2025-11-12 15:09:30.749 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5885
2025-11-12 15:09:30.789 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4514
2025-11-12 15:09:30.790 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5633
2025-11-12 15:09:30.790 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:09:30.790 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:09:30.790 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.650
2025-11-12 15:09:30.790 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.589
2025-11-12 15:09:30.790 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.451
2025-11-12 15:09:30.790 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.563
2025-11-12 15:09:30.790 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:09:30.791 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:09:30.791 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:09:30.791 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:09:30.791 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:09:30.791 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:09:30.791 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:09:30.791 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:09:30.791 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:09:31.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:09:32.266 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:09:32.972 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:09:33.679 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:09:34.424 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:09:35.128 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:09:35.831 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:09:36.573 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:09:37.280 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:09:37.281 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:09:37.281 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:09:37.281 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:09:37.288 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.17 ms, Average NMS time: 0.56 ms, Average inference time: 2.73 ms

2025-11-12 15:09:37.289 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:09:37.368 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:09:37.449 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch87
2025-11-12 15:09:40.631 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 87/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.155s, data_time: 0.002s, total_loss: 3.6, iou_loss: 1.6, l1_loss: 0.5, conf_loss: 1.1, cls_loss: 0.5, lr: 3.675e-04, size: 416, ETA: 0:11:40
2025-11-12 15:09:43.405 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 87/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.137s, data_time: 0.002s, total_loss: 4.8, iou_loss: 1.9, l1_loss: 0.9, conf_loss: 1.5, cls_loss: 0.5, lr: 3.644e-04, size: 512, ETA: 0:11:37
2025-11-12 15:09:46.446 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 87/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.148s, data_time: 0.003s, total_loss: 3.8, iou_loss: 1.7, l1_loss: 0.5, conf_loss: 1.0, cls_loss: 0.5, lr: 3.613e-04, size: 288, ETA: 0:11:34
2025-11-12 15:09:49.575 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 87/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.153s, data_time: 0.003s, total_loss: 5.3, iou_loss: 2.3, l1_loss: 0.7, conf_loss: 1.6, cls_loss: 0.6, lr: 3.581e-04, size: 320, ETA: 0:11:30
2025-11-12 15:09:52.519 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 87/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.145s, data_time: 0.002s, total_loss: 3.7, iou_loss: 1.8, l1_loss: 0.6, conf_loss: 0.9, cls_loss: 0.5, lr: 3.550e-04, size: 512, ETA: 0:11:27
2025-11-12 15:09:55.608 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 87/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.002s, total_loss: 4.7, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.4, cls_loss: 0.6, lr: 3.519e-04, size: 288, ETA: 0:11:24
2025-11-12 15:09:56.902 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:10:01.974 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:10:03.155 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:10:03.937 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6586
2025-11-12 15:10:04.084 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5711
2025-11-12 15:10:04.170 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4665
2025-11-12 15:10:04.171 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5654
2025-11-12 15:10:04.171 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:10:04.171 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:10:04.171 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.659
2025-11-12 15:10:04.171 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.571
2025-11-12 15:10:04.172 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.466
2025-11-12 15:10:04.172 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.565
2025-11-12 15:10:04.172 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:10:04.172 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:10:04.172 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:10:04.172 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:10:04.172 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:10:04.172 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:10:04.173 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:10:04.173 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:10:04.173 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:10:05.161 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:10:06.185 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:10:07.180 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:10:08.162 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:10:09.151 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:10:10.152 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:10:11.118 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:10:12.107 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:10:13.109 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:10:13.109 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:10:13.110 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.57
2025-11-12 15:10:13.110 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:10:13.117 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.16 ms, Average NMS time: 0.57 ms, Average inference time: 2.72 ms

2025-11-12 15:10:13.119 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:10:13.194 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:10:13.275 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch88
2025-11-12 15:10:16.100 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 88/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.138s, data_time: 0.002s, total_loss: 4.5, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.4, cls_loss: 0.5, lr: 3.475e-04, size: 288, ETA: 0:11:19
2025-11-12 15:10:19.228 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 88/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.153s, data_time: 0.002s, total_loss: 3.3, iou_loss: 1.4, l1_loss: 0.6, conf_loss: 1.0, cls_loss: 0.5, lr: 3.444e-04, size: 544, ETA: 0:11:16
2025-11-12 15:10:22.325 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 88/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.002s, total_loss: 5.0, iou_loss: 1.8, l1_loss: 0.6, conf_loss: 1.9, cls_loss: 0.6, lr: 3.413e-04, size: 256, ETA: 0:11:12
2025-11-12 15:10:25.326 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 88/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.148s, data_time: 0.003s, total_loss: 6.3, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 2.3, cls_loss: 0.7, lr: 3.383e-04, size: 256, ETA: 0:11:09
2025-11-12 15:10:28.284 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 88/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.144s, data_time: 0.001s, total_loss: 5.3, iou_loss: 2.2, l1_loss: 0.7, conf_loss: 1.8, cls_loss: 0.6, lr: 3.353e-04, size: 320, ETA: 0:11:06
2025-11-12 15:10:31.455 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 88/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.154s, data_time: 0.002s, total_loss: 4.5, iou_loss: 1.9, l1_loss: 0.6, conf_loss: 1.4, cls_loss: 0.6, lr: 3.322e-04, size: 256, ETA: 0:11:03
2025-11-12 15:10:32.709 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:10:37.770 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:10:38.929 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:10:39.620 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6515
2025-11-12 15:10:39.798 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5815
2025-11-12 15:10:39.843 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4563
2025-11-12 15:10:39.844 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5631
2025-11-12 15:10:39.844 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:10:39.844 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:10:39.844 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.651
2025-11-12 15:10:39.844 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.581
2025-11-12 15:10:39.844 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.456
2025-11-12 15:10:39.845 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.563
2025-11-12 15:10:39.845 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:10:39.845 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:10:39.845 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:10:39.845 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:10:39.845 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:10:39.846 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:10:39.846 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:10:39.846 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:10:39.846 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:10:40.789 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:10:41.699 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:10:42.639 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:10:43.545 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:10:44.488 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:10:45.410 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:10:46.357 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:10:47.308 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:10:48.202 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:10:48.203 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:10:48.203 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:10:48.203 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:10:48.210 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.20 ms, Average NMS time: 0.57 ms, Average inference time: 2.77 ms

2025-11-12 15:10:48.211 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:10:48.287 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:10:48.368 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch89
2025-11-12 15:10:51.404 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 89/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 5.0, iou_loss: 1.9, l1_loss: 0.7, conf_loss: 1.8, cls_loss: 0.5, lr: 3.279e-04, size: 576, ETA: 0:10:58
2025-11-12 15:10:54.810 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 89/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.167s, data_time: 0.002s, total_loss: 5.1, iou_loss: 2.4, l1_loss: 0.7, conf_loss: 1.5, cls_loss: 0.6, lr: 3.249e-04, size: 512, ETA: 0:10:55
2025-11-12 15:10:58.142 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 89/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.163s, data_time: 0.002s, total_loss: 4.4, iou_loss: 1.5, l1_loss: 0.7, conf_loss: 1.5, cls_loss: 0.6, lr: 3.219e-04, size: 512, ETA: 0:10:51
2025-11-12 15:11:01.280 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 89/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.154s, data_time: 0.003s, total_loss: 4.5, iou_loss: 1.8, l1_loss: 0.6, conf_loss: 1.7, cls_loss: 0.5, lr: 3.189e-04, size: 416, ETA: 0:10:48
2025-11-12 15:11:04.476 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 89/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.156s, data_time: 0.002s, total_loss: 5.0, iou_loss: 1.8, l1_loss: 0.8, conf_loss: 1.9, cls_loss: 0.5, lr: 3.159e-04, size: 576, ETA: 0:10:45
2025-11-12 15:11:07.906 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 89/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.168s, data_time: 0.002s, total_loss: 4.6, iou_loss: 1.9, l1_loss: 0.8, conf_loss: 1.3, cls_loss: 0.5, lr: 3.130e-04, size: 512, ETA: 0:10:42
2025-11-12 15:11:09.395 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:11:14.375 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:11:15.202 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:11:15.755 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6503
2025-11-12 15:11:15.873 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5986
2025-11-12 15:11:15.910 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4385
2025-11-12 15:11:15.911 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5625
2025-11-12 15:11:15.911 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:11:15.911 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:11:15.911 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.650
2025-11-12 15:11:15.911 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.599
2025-11-12 15:11:15.911 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.438
2025-11-12 15:11:15.911 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.562
2025-11-12 15:11:15.911 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:11:15.912 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:11:15.912 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:11:15.912 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:11:15.912 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:11:15.912 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:11:15.912 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:11:15.912 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:11:15.912 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:11:16.643 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:11:17.325 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:11:18.022 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:11:18.730 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:11:19.410 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:11:20.090 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:11:20.819 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:11:21.513 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:11:22.205 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:11:22.205 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:11:22.206 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:11:22.206 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:11:22.213 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.17 ms, Average NMS time: 0.57 ms, Average inference time: 2.74 ms

2025-11-12 15:11:22.215 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:11:22.290 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:11:22.371 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch90
2025-11-12 15:11:25.210 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 90/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.140s, data_time: 0.002s, total_loss: 5.0, iou_loss: 2.1, l1_loss: 0.8, conf_loss: 1.4, cls_loss: 0.7, lr: 3.087e-04, size: 256, ETA: 0:10:37
2025-11-12 15:11:28.094 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 90/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.141s, data_time: 0.003s, total_loss: 6.6, iou_loss: 2.4, l1_loss: 1.1, conf_loss: 2.5, cls_loss: 0.6, lr: 3.058e-04, size: 576, ETA: 0:10:34
2025-11-12 15:11:31.208 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 90/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.153s, data_time: 0.002s, total_loss: 4.3, iou_loss: 2.1, l1_loss: 0.7, conf_loss: 1.0, cls_loss: 0.5, lr: 3.029e-04, size: 352, ETA: 0:10:30
2025-11-12 15:11:34.138 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 90/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.003s, total_loss: 5.5, iou_loss: 2.2, l1_loss: 0.7, conf_loss: 1.9, cls_loss: 0.6, lr: 3.000e-04, size: 320, ETA: 0:10:27
2025-11-12 15:11:37.182 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 90/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.150s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.5, l1_loss: 0.9, conf_loss: 1.7, cls_loss: 0.6, lr: 2.971e-04, size: 416, ETA: 0:10:24
2025-11-12 15:11:40.166 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 90/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 4.0, iou_loss: 1.7, l1_loss: 0.5, conf_loss: 1.3, cls_loss: 0.5, lr: 2.942e-04, size: 320, ETA: 0:10:20
2025-11-12 15:11:41.405 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:11:46.402 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:11:47.455 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:11:48.097 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6601
2025-11-12 15:11:48.237 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.6045
2025-11-12 15:11:48.313 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4563
2025-11-12 15:11:48.314 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5736
2025-11-12 15:11:48.314 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:11:48.314 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:11:48.314 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.660
2025-11-12 15:11:48.314 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.604
2025-11-12 15:11:48.314 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.456
2025-11-12 15:11:48.314 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.574
2025-11-12 15:11:48.315 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:11:48.315 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:11:48.315 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:11:48.315 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:11:48.315 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:11:48.315 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:11:48.315 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:11:48.315 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:11:48.315 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:11:49.116 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:11:49.965 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:11:50.796 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:11:51.604 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:11:52.443 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:11:53.247 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:11:54.074 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:11:54.881 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:11:55.714 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:11:55.714 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.33
2025-11-12 15:11:55.714 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.57
2025-11-12 15:11:55.714 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:11:55.722 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.13 ms, Average NMS time: 0.55 ms, Average inference time: 2.68 ms

2025-11-12 15:11:55.723 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:11:55.803 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:11:55.923 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch91
2025-11-12 15:11:58.840 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 91/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.144s, data_time: 0.002s, total_loss: 4.0, iou_loss: 1.6, l1_loss: 0.6, conf_loss: 1.3, cls_loss: 0.5, lr: 2.900e-04, size: 512, ETA: 0:10:16
2025-11-12 15:12:01.891 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 91/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.003s, total_loss: 4.9, iou_loss: 1.8, l1_loss: 0.8, conf_loss: 1.7, cls_loss: 0.6, lr: 2.872e-04, size: 576, ETA: 0:10:12
2025-11-12 15:12:04.786 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 91/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 5.2, iou_loss: 1.9, l1_loss: 0.6, conf_loss: 2.1, cls_loss: 0.6, lr: 2.843e-04, size: 384, ETA: 0:10:09
2025-11-12 15:12:07.700 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 91/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 4.9, iou_loss: 1.9, l1_loss: 0.9, conf_loss: 1.4, cls_loss: 0.6, lr: 2.815e-04, size: 544, ETA: 0:10:06
2025-11-12 15:12:10.522 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 91/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.138s, data_time: 0.002s, total_loss: 5.0, iou_loss: 2.1, l1_loss: 0.6, conf_loss: 1.7, cls_loss: 0.6, lr: 2.787e-04, size: 288, ETA: 0:10:02
2025-11-12 15:12:13.364 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 91/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.140s, data_time: 0.002s, total_loss: 5.0, iou_loss: 2.0, l1_loss: 0.9, conf_loss: 1.6, cls_loss: 0.6, lr: 2.759e-04, size: 544, ETA: 0:09:59
2025-11-12 15:12:14.724 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:12:19.921 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:12:20.967 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:12:21.551 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6499
2025-11-12 15:12:21.738 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5840
2025-11-12 15:12:21.777 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4630
2025-11-12 15:12:21.778 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5657
2025-11-12 15:12:21.778 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:12:21.778 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:12:21.778 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.650
2025-11-12 15:12:21.778 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.584
2025-11-12 15:12:21.778 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.463
2025-11-12 15:12:21.778 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.566
2025-11-12 15:12:21.779 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:12:21.779 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:12:21.779 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:12:21.779 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:12:21.779 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:12:21.779 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:12:21.779 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:12:21.779 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:12:21.779 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:12:22.587 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:12:23.443 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:12:24.273 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:12:25.143 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:12:25.959 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:12:26.793 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:12:27.591 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:12:28.450 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:12:29.269 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:12:29.269 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:12:29.270 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.57
2025-11-12 15:12:29.270 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:12:29.277 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.24 ms, Average NMS time: 0.61 ms, Average inference time: 2.85 ms

2025-11-12 15:12:29.279 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:12:29.355 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:12:29.438 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch92
2025-11-12 15:12:32.377 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 92/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.145s, data_time: 0.002s, total_loss: 4.5, iou_loss: 1.8, l1_loss: 0.8, conf_loss: 1.4, cls_loss: 0.5, lr: 2.718e-04, size: 576, ETA: 0:09:54
2025-11-12 15:12:35.486 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 92/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.154s, data_time: 0.002s, total_loss: 3.7, iou_loss: 1.5, l1_loss: 0.7, conf_loss: 1.1, cls_loss: 0.5, lr: 2.691e-04, size: 512, ETA: 0:09:51
2025-11-12 15:12:38.375 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 92/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.143s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.2, l1_loss: 0.7, conf_loss: 2.3, cls_loss: 0.6, lr: 2.663e-04, size: 352, ETA: 0:09:48
2025-11-12 15:12:41.290 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 92/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.144s, data_time: 0.003s, total_loss: 3.8, iou_loss: 1.6, l1_loss: 0.5, conf_loss: 1.2, cls_loss: 0.5, lr: 2.635e-04, size: 256, ETA: 0:09:44
2025-11-12 15:12:44.212 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 92/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.143s, data_time: 0.002s, total_loss: 6.1, iou_loss: 2.4, l1_loss: 0.9, conf_loss: 2.2, cls_loss: 0.6, lr: 2.608e-04, size: 576, ETA: 0:09:41
2025-11-12 15:12:47.254 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 92/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 4.4, iou_loss: 1.8, l1_loss: 0.9, conf_loss: 1.2, cls_loss: 0.5, lr: 2.581e-04, size: 576, ETA: 0:09:38
2025-11-12 15:12:48.613 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:12:53.656 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:12:54.469 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:12:55.010 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6515
2025-11-12 15:12:55.126 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5847
2025-11-12 15:12:55.161 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4361
2025-11-12 15:12:55.161 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5574
2025-11-12 15:12:55.161 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:12:55.162 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:12:55.162 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.651
2025-11-12 15:12:55.162 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.585
2025-11-12 15:12:55.162 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.436
2025-11-12 15:12:55.162 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.557
2025-11-12 15:12:55.162 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:12:55.162 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:12:55.162 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:12:55.162 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:12:55.162 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:12:55.163 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:12:55.163 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:12:55.163 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:12:55.163 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:12:55.864 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:12:56.549 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:12:57.226 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:12:57.902 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:12:58.606 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:12:59.273 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:12:59.945 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:13:00.647 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:13:01.315 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:13:01.315 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:13:01.315 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:13:01.315 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:13:01.322 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.16 ms, Average NMS time: 0.56 ms, Average inference time: 2.72 ms

2025-11-12 15:13:01.335 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:13:01.410 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:13:01.491 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch93
2025-11-12 15:13:04.453 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 93/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.145s, data_time: 0.002s, total_loss: 4.9, iou_loss: 2.1, l1_loss: 0.6, conf_loss: 1.6, cls_loss: 0.6, lr: 2.541e-04, size: 288, ETA: 0:09:33
2025-11-12 15:13:07.555 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 93/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.002s, total_loss: 5.5, iou_loss: 2.1, l1_loss: 0.7, conf_loss: 2.0, cls_loss: 0.6, lr: 2.514e-04, size: 320, ETA: 0:09:30
2025-11-12 15:13:10.337 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 93/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.136s, data_time: 0.001s, total_loss: 3.4, iou_loss: 1.6, l1_loss: 0.4, conf_loss: 0.9, cls_loss: 0.5, lr: 2.488e-04, size: 448, ETA: 0:09:27
2025-11-12 15:13:13.629 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 93/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.161s, data_time: 0.002s, total_loss: 5.2, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 1.6, cls_loss: 0.6, lr: 2.461e-04, size: 448, ETA: 0:09:23
2025-11-12 15:13:16.534 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 93/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.143s, data_time: 0.003s, total_loss: 5.0, iou_loss: 2.0, l1_loss: 0.8, conf_loss: 1.6, cls_loss: 0.6, lr: 2.434e-04, size: 544, ETA: 0:09:20
2025-11-12 15:13:19.552 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 93/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.149s, data_time: 0.003s, total_loss: 5.3, iou_loss: 2.1, l1_loss: 0.7, conf_loss: 1.9, cls_loss: 0.6, lr: 2.408e-04, size: 256, ETA: 0:09:17
2025-11-12 15:13:20.789 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:13:25.868 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:13:27.205 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:13:28.112 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6538
2025-11-12 15:13:28.287 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5999
2025-11-12 15:13:28.366 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4450
2025-11-12 15:13:28.366 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5662
2025-11-12 15:13:28.367 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:13:28.367 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:13:28.367 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.654
2025-11-12 15:13:28.367 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.600
2025-11-12 15:13:28.367 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.445
2025-11-12 15:13:28.367 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.566
2025-11-12 15:13:28.367 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:13:28.367 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:13:28.367 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:13:28.368 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:13:28.368 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:13:28.368 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:13:28.368 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:13:28.368 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:13:28.368 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:13:29.461 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:13:30.595 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:13:31.713 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:13:32.838 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:13:33.980 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:13:35.085 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:13:36.220 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:13:37.329 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:13:38.446 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:13:38.446 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.33
2025-11-12 15:13:38.446 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.57
2025-11-12 15:13:38.446 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:13:38.454 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.17 ms, Average NMS time: 0.56 ms, Average inference time: 2.72 ms

2025-11-12 15:13:38.455 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:13:38.533 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:13:38.615 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch94
2025-11-12 15:13:41.343 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 94/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.135s, data_time: 0.002s, total_loss: 4.4, iou_loss: 1.7, l1_loss: 0.5, conf_loss: 1.6, cls_loss: 0.6, lr: 2.370e-04, size: 416, ETA: 0:09:12
2025-11-12 15:13:44.232 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 94/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 5.0, iou_loss: 2.1, l1_loss: 0.7, conf_loss: 1.6, cls_loss: 0.6, lr: 2.343e-04, size: 384, ETA: 0:09:09
2025-11-12 15:13:47.013 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 94/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.136s, data_time: 0.002s, total_loss: 3.9, iou_loss: 1.7, l1_loss: 0.6, conf_loss: 1.1, cls_loss: 0.5, lr: 2.317e-04, size: 448, ETA: 0:09:05
2025-11-12 15:13:50.029 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 94/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.148s, data_time: 0.002s, total_loss: 4.9, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.6, cls_loss: 0.6, lr: 2.292e-04, size: 448, ETA: 0:09:02
2025-11-12 15:13:52.981 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 94/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.145s, data_time: 0.002s, total_loss: 5.0, iou_loss: 2.4, l1_loss: 0.7, conf_loss: 1.2, cls_loss: 0.6, lr: 2.266e-04, size: 352, ETA: 0:08:59
2025-11-12 15:13:55.962 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 94/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 5.2, iou_loss: 2.0, l1_loss: 0.8, conf_loss: 1.8, cls_loss: 0.6, lr: 2.240e-04, size: 512, ETA: 0:08:56
2025-11-12 15:13:57.394 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:14:02.704 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:14:03.680 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:14:04.338 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6522
2025-11-12 15:14:04.463 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5825
2025-11-12 15:14:04.503 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4467
2025-11-12 15:14:04.503 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5605
2025-11-12 15:14:04.503 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:14:04.504 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:14:04.504 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.652
2025-11-12 15:14:04.504 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.583
2025-11-12 15:14:04.504 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.447
2025-11-12 15:14:04.504 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.560
2025-11-12 15:14:04.504 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:14:04.504 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:14:04.504 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:14:04.504 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:14:04.504 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:14:04.505 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:14:04.505 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:14:04.505 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:14:04.505 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:14:05.359 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:14:06.172 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:14:06.977 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:14:07.821 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:14:08.636 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:14:09.473 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:14:10.282 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:14:11.118 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:14:11.934 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:14:11.934 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.33
2025-11-12 15:14:11.934 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:14:11.934 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:14:11.941 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.20 ms, Average NMS time: 0.64 ms, Average inference time: 2.84 ms

2025-11-12 15:14:11.943 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:14:12.018 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:14:12.099 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch95
2025-11-12 15:14:14.847 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 95/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.135s, data_time: 0.001s, total_loss: 4.2, iou_loss: 1.7, l1_loss: 0.5, conf_loss: 1.5, cls_loss: 0.5, lr: 2.203e-04, size: 416, ETA: 0:08:51
2025-11-12 15:14:17.793 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 95/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.144s, data_time: 0.001s, total_loss: 5.7, iou_loss: 2.0, l1_loss: 0.8, conf_loss: 2.3, cls_loss: 0.6, lr: 2.178e-04, size: 448, ETA: 0:08:48
2025-11-12 15:14:20.565 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 95/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.136s, data_time: 0.002s, total_loss: 6.3, iou_loss: 2.6, l1_loss: 0.8, conf_loss: 2.3, cls_loss: 0.6, lr: 2.152e-04, size: 288, ETA: 0:08:44
2025-11-12 15:14:23.603 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 95/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 4.0, iou_loss: 1.6, l1_loss: 0.7, conf_loss: 1.3, cls_loss: 0.5, lr: 2.127e-04, size: 512, ETA: 0:08:41
2025-11-12 15:14:26.531 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 95/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.144s, data_time: 0.003s, total_loss: 6.3, iou_loss: 2.0, l1_loss: 1.1, conf_loss: 2.6, cls_loss: 0.6, lr: 2.102e-04, size: 576, ETA: 0:08:38
2025-11-12 15:14:29.524 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 95/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 4.5, iou_loss: 1.8, l1_loss: 0.5, conf_loss: 1.5, cls_loss: 0.6, lr: 2.078e-04, size: 256, ETA: 0:08:35
2025-11-12 15:14:30.754 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:14:35.707 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:14:36.857 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:14:37.624 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6528
2025-11-12 15:14:37.821 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5902
2025-11-12 15:14:37.864 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4449
2025-11-12 15:14:37.864 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5626
2025-11-12 15:14:37.864 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:14:37.864 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:14:37.865 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.653
2025-11-12 15:14:37.865 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.590
2025-11-12 15:14:37.865 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.445
2025-11-12 15:14:37.865 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.563
2025-11-12 15:14:37.865 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:14:37.865 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:14:37.865 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:14:37.865 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:14:37.866 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:14:37.866 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:14:37.866 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:14:37.866 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:14:37.866 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:14:38.817 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:14:39.799 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:14:40.783 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:14:41.739 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:14:42.719 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:14:43.662 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:14:44.656 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:14:45.639 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:14:46.591 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:14:46.592 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.33
2025-11-12 15:14:46.592 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:14:46.592 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:14:46.599 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.11 ms, Average NMS time: 0.54 ms, Average inference time: 2.65 ms

2025-11-12 15:14:46.600 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:14:46.676 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:14:46.758 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch96
2025-11-12 15:14:49.655 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 96/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.143s, data_time: 0.002s, total_loss: 4.8, iou_loss: 2.0, l1_loss: 0.8, conf_loss: 1.4, cls_loss: 0.5, lr: 2.042e-04, size: 512, ETA: 0:08:30
2025-11-12 15:14:52.623 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 96/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 4.5, iou_loss: 1.8, l1_loss: 0.5, conf_loss: 1.7, cls_loss: 0.5, lr: 2.017e-04, size: 384, ETA: 0:08:27
2025-11-12 15:14:55.570 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 96/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 4.8, iou_loss: 2.1, l1_loss: 0.7, conf_loss: 1.5, cls_loss: 0.6, lr: 1.993e-04, size: 480, ETA: 0:08:23
2025-11-12 15:14:58.436 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 96/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.140s, data_time: 0.005s, total_loss: 4.2, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.1, cls_loss: 0.5, lr: 1.969e-04, size: 320, ETA: 0:08:20
2025-11-12 15:15:01.453 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 96/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.3, l1_loss: 0.9, conf_loss: 1.9, cls_loss: 0.6, lr: 1.945e-04, size: 512, ETA: 0:08:17
2025-11-12 15:15:04.334 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 96/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.141s, data_time: 0.002s, total_loss: 3.9, iou_loss: 1.7, l1_loss: 0.6, conf_loss: 1.0, cls_loss: 0.5, lr: 1.921e-04, size: 416, ETA: 0:08:13
2025-11-12 15:15:05.686 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:15:10.707 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:15:11.530 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:15:12.170 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6560
2025-11-12 15:15:12.398 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5833
2025-11-12 15:15:12.490 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4599
2025-11-12 15:15:12.490 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5664
2025-11-12 15:15:12.491 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:15:12.491 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:15:12.491 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.656
2025-11-12 15:15:12.491 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.583
2025-11-12 15:15:12.491 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.460
2025-11-12 15:15:12.491 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.566
2025-11-12 15:15:12.492 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:15:12.492 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:15:12.492 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:15:12.492 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:15:12.492 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:15:12.492 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:15:12.493 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:15:12.493 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:15:12.493 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:15:13.149 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:15:13.816 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:15:14.471 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:15:15.159 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:15:15.823 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:15:16.482 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:15:17.170 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:15:17.819 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:15:18.478 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:15:18.478 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:15:18.479 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.57
2025-11-12 15:15:18.479 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:15:18.486 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.19 ms, Average NMS time: 0.55 ms, Average inference time: 2.74 ms

2025-11-12 15:15:18.486 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:15:18.562 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:15:18.643 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch97
2025-11-12 15:15:21.684 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 97/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.003s, total_loss: 4.7, iou_loss: 1.9, l1_loss: 0.8, conf_loss: 1.5, cls_loss: 0.6, lr: 1.886e-04, size: 544, ETA: 0:08:09
2025-11-12 15:15:24.671 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 97/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 2.2, cls_loss: 0.7, lr: 1.862e-04, size: 544, ETA: 0:08:06
2025-11-12 15:15:27.648 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 97/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 4.8, iou_loss: 1.9, l1_loss: 0.7, conf_loss: 1.7, cls_loss: 0.5, lr: 1.839e-04, size: 416, ETA: 0:08:02
2025-11-12 15:15:30.795 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 97/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.156s, data_time: 0.004s, total_loss: 4.2, iou_loss: 1.9, l1_loss: 0.6, conf_loss: 1.1, cls_loss: 0.5, lr: 1.815e-04, size: 416, ETA: 0:07:59
2025-11-12 15:15:33.599 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 97/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.137s, data_time: 0.002s, total_loss: 5.4, iou_loss: 2.3, l1_loss: 0.7, conf_loss: 1.8, cls_loss: 0.6, lr: 1.792e-04, size: 384, ETA: 0:07:56
2025-11-12 15:15:36.500 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 97/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 3.6, iou_loss: 1.5, l1_loss: 0.5, conf_loss: 1.1, cls_loss: 0.5, lr: 1.769e-04, size: 448, ETA: 0:07:53
2025-11-12 15:15:37.818 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:15:43.007 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:15:43.972 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:15:44.600 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6596
2025-11-12 15:15:44.742 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5897
2025-11-12 15:15:44.816 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4515
2025-11-12 15:15:44.817 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5670
2025-11-12 15:15:44.817 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:15:44.817 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:15:44.817 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.660
2025-11-12 15:15:44.817 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.590
2025-11-12 15:15:44.817 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.452
2025-11-12 15:15:44.818 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.567
2025-11-12 15:15:44.818 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:15:44.818 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:15:44.818 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:15:44.818 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:15:44.818 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:15:44.818 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:15:44.818 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:15:44.818 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:15:44.819 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:15:45.609 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:15:46.438 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:15:47.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:15:48.005 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:15:48.831 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:15:49.620 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:15:50.436 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:15:51.216 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:15:52.040 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:15:52.040 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:15:52.040 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.57
2025-11-12 15:15:52.040 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:15:52.047 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.24 ms, Average NMS time: 0.59 ms, Average inference time: 2.84 ms

2025-11-12 15:15:52.049 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:15:52.124 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:15:52.207 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch98
2025-11-12 15:15:55.209 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 98/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 5.0, iou_loss: 2.1, l1_loss: 0.7, conf_loss: 1.6, cls_loss: 0.6, lr: 1.736e-04, size: 448, ETA: 0:07:48
2025-11-12 15:15:58.188 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 98/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 4.9, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.6, cls_loss: 0.5, lr: 1.713e-04, size: 288, ETA: 0:07:45
2025-11-12 15:16:01.081 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 98/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.141s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.4, l1_loss: 0.8, conf_loss: 2.1, cls_loss: 0.6, lr: 1.690e-04, size: 416, ETA: 0:07:41
2025-11-12 15:16:04.099 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 98/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.148s, data_time: 0.003s, total_loss: 5.3, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 1.7, cls_loss: 0.6, lr: 1.668e-04, size: 416, ETA: 0:07:38
2025-11-12 15:16:07.145 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 98/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.149s, data_time: 0.003s, total_loss: 4.0, iou_loss: 1.8, l1_loss: 0.5, conf_loss: 1.2, cls_loss: 0.5, lr: 1.645e-04, size: 288, ETA: 0:07:35
2025-11-12 15:16:10.189 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 98/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.150s, data_time: 0.003s, total_loss: 4.2, iou_loss: 2.2, l1_loss: 0.6, conf_loss: 0.9, cls_loss: 0.6, lr: 1.623e-04, size: 544, ETA: 0:07:32
2025-11-12 15:16:11.595 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:16:16.517 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:16:17.398 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:16:17.942 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6540
2025-11-12 15:16:18.101 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5856
2025-11-12 15:16:18.138 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4362
2025-11-12 15:16:18.138 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5586
2025-11-12 15:16:18.138 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:16:18.139 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:16:18.139 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.654
2025-11-12 15:16:18.139 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.586
2025-11-12 15:16:18.139 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.436
2025-11-12 15:16:18.139 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.559
2025-11-12 15:16:18.139 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:16:18.140 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:16:18.140 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:16:18.140 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:16:18.140 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:16:18.140 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:16:18.140 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:16:18.141 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:16:18.141 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:16:18.821 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:16:19.493 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:16:20.200 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:16:20.873 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:16:21.548 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:16:22.277 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:16:22.936 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:16:23.608 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:16:24.275 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:16:24.276 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:16:24.276 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:16:24.276 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:16:24.283 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.14 ms, Average NMS time: 0.53 ms, Average inference time: 2.67 ms

2025-11-12 15:16:24.284 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:16:24.398 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:16:24.480 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch99
2025-11-12 15:16:27.401 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 99/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.145s, data_time: 0.002s, total_loss: 3.9, iou_loss: 1.8, l1_loss: 0.5, conf_loss: 1.0, cls_loss: 0.5, lr: 1.591e-04, size: 256, ETA: 0:07:27
2025-11-12 15:16:30.209 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 99/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.140s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 1.9, cls_loss: 0.7, lr: 1.569e-04, size: 576, ETA: 0:07:24
2025-11-12 15:16:33.270 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 99/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.002s, total_loss: 5.1, iou_loss: 2.4, l1_loss: 0.6, conf_loss: 1.5, cls_loss: 0.6, lr: 1.548e-04, size: 352, ETA: 0:07:21
2025-11-12 15:16:36.247 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 99/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.145s, data_time: 0.002s, total_loss: 4.8, iou_loss: 2.1, l1_loss: 0.6, conf_loss: 1.5, cls_loss: 0.6, lr: 1.526e-04, size: 352, ETA: 0:07:17
2025-11-12 15:16:39.003 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 99/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.136s, data_time: 0.002s, total_loss: 4.3, iou_loss: 1.9, l1_loss: 0.8, conf_loss: 1.1, cls_loss: 0.5, lr: 1.504e-04, size: 352, ETA: 0:07:14
2025-11-12 15:16:41.725 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 99/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.134s, data_time: 0.003s, total_loss: 5.6, iou_loss: 2.1, l1_loss: 0.8, conf_loss: 2.1, cls_loss: 0.6, lr: 1.483e-04, size: 480, ETA: 0:07:11
2025-11-12 15:16:43.172 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:16:48.183 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:16:49.090 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:16:49.678 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6524
2025-11-12 15:16:49.850 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5816
2025-11-12 15:16:49.886 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4573
2025-11-12 15:16:49.887 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5638
2025-11-12 15:16:49.887 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:16:49.887 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:16:49.887 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.652
2025-11-12 15:16:49.887 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.582
2025-11-12 15:16:49.887 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.457
2025-11-12 15:16:49.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.564
2025-11-12 15:16:49.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:16:49.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:16:49.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:16:49.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:16:49.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:16:49.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:16:49.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:16:49.889 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:16:49.889 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:16:50.634 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:16:51.410 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:16:52.144 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:16:52.881 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:16:53.667 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:16:54.408 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:16:55.157 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:16:55.947 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:16:56.688 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:16:56.689 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:16:56.689 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:16:56.689 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:16:56.696 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.13 ms, Average NMS time: 0.57 ms, Average inference time: 2.70 ms

2025-11-12 15:16:56.698 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:16:56.773 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:16:56.853 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch100
2025-11-12 15:16:59.833 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 100/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 5.3, iou_loss: 2.3, l1_loss: 0.7, conf_loss: 1.7, cls_loss: 0.6, lr: 1.452e-04, size: 352, ETA: 0:07:06
2025-11-12 15:17:02.639 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 100/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.138s, data_time: 0.002s, total_loss: 5.1, iou_loss: 1.8, l1_loss: 0.9, conf_loss: 1.9, cls_loss: 0.5, lr: 1.431e-04, size: 576, ETA: 0:07:03
2025-11-12 15:17:05.731 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 100/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.152s, data_time: 0.002s, total_loss: 7.0, iou_loss: 2.3, l1_loss: 1.1, conf_loss: 2.9, cls_loss: 0.7, lr: 1.411e-04, size: 544, ETA: 0:07:00
2025-11-12 15:17:08.974 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 100/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.159s, data_time: 0.003s, total_loss: 4.9, iou_loss: 2.1, l1_loss: 0.7, conf_loss: 1.4, cls_loss: 0.6, lr: 1.390e-04, size: 256, ETA: 0:06:56
2025-11-12 15:17:11.887 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 100/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.144s, data_time: 0.003s, total_loss: 3.1, iou_loss: 1.3, l1_loss: 0.6, conf_loss: 0.8, cls_loss: 0.4, lr: 1.369e-04, size: 384, ETA: 0:06:53
2025-11-12 15:17:14.745 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 100/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.140s, data_time: 0.003s, total_loss: 4.7, iou_loss: 1.8, l1_loss: 0.8, conf_loss: 1.5, cls_loss: 0.5, lr: 1.349e-04, size: 544, ETA: 0:06:50
2025-11-12 15:17:16.147 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:17:21.275 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:17:22.162 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:17:22.740 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6584
2025-11-12 15:17:22.864 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5836
2025-11-12 15:17:22.940 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4534
2025-11-12 15:17:22.941 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5651
2025-11-12 15:17:22.941 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:17:22.941 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:17:22.942 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.658
2025-11-12 15:17:22.942 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.584
2025-11-12 15:17:22.942 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.453
2025-11-12 15:17:22.942 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.565
2025-11-12 15:17:22.942 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:17:22.942 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:17:22.942 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:17:22.942 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:17:22.942 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:17:22.943 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:17:22.943 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:17:22.943 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:17:22.943 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:17:23.668 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:17:24.411 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:17:25.169 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:17:25.889 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:17:26.619 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:17:27.378 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:17:28.105 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:17:28.836 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:17:29.588 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:17:29.589 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.33
2025-11-12 15:17:29.589 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.57
2025-11-12 15:17:29.589 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:17:29.596 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.26 ms, Average NMS time: 0.58 ms, Average inference time: 2.84 ms

2025-11-12 15:17:29.597 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:17:29.676 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:17:29.756 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch101
2025-11-12 15:17:32.974 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 101/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.158s, data_time: 0.002s, total_loss: 6.8, iou_loss: 2.7, l1_loss: 1.0, conf_loss: 2.4, cls_loss: 0.7, lr: 1.320e-04, size: 576, ETA: 0:06:45
2025-11-12 15:17:36.339 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 101/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.164s, data_time: 0.003s, total_loss: 5.4, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 1.9, cls_loss: 0.6, lr: 1.299e-04, size: 512, ETA: 0:06:42
2025-11-12 15:17:39.601 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 101/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.160s, data_time: 0.002s, total_loss: 4.0, iou_loss: 1.7, l1_loss: 0.6, conf_loss: 1.2, cls_loss: 0.5, lr: 1.280e-04, size: 448, ETA: 0:06:39
2025-11-12 15:17:42.536 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 101/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.143s, data_time: 0.003s, total_loss: 3.7, iou_loss: 1.8, l1_loss: 0.6, conf_loss: 0.8, cls_loss: 0.5, lr: 1.260e-04, size: 288, ETA: 0:06:36
2025-11-12 15:17:45.394 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 101/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.140s, data_time: 0.002s, total_loss: 4.6, iou_loss: 2.1, l1_loss: 0.8, conf_loss: 1.2, cls_loss: 0.6, lr: 1.240e-04, size: 288, ETA: 0:06:33
2025-11-12 15:17:48.423 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 101/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.150s, data_time: 0.002s, total_loss: 3.9, iou_loss: 1.8, l1_loss: 0.6, conf_loss: 1.1, cls_loss: 0.5, lr: 1.221e-04, size: 256, ETA: 0:06:29
2025-11-12 15:17:49.831 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:17:54.900 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:17:55.979 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:17:56.711 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6552
2025-11-12 15:17:56.845 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5877
2025-11-12 15:17:56.887 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4283
2025-11-12 15:17:56.887 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5571
2025-11-12 15:17:56.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:17:56.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:17:56.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.655
2025-11-12 15:17:56.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.588
2025-11-12 15:17:56.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.428
2025-11-12 15:17:56.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.557
2025-11-12 15:17:56.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:17:56.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:17:56.888 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:17:56.889 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:17:56.889 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:17:56.889 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:17:56.889 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:17:56.889 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:17:56.889 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:17:57.822 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:17:58.736 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:17:59.690 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:18:00.594 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:18:01.531 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:18:02.442 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:18:03.383 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:18:04.320 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:18:05.229 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:18:05.230 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:18:05.230 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:18:05.230 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:18:05.237 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.21 ms, Average NMS time: 0.61 ms, Average inference time: 2.83 ms

2025-11-12 15:18:05.239 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:18:05.315 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:18:05.398 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch102
2025-11-12 15:18:08.661 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 102/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.161s, data_time: 0.002s, total_loss: 5.0, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.8, cls_loss: 0.6, lr: 1.193e-04, size: 352, ETA: 0:06:25
2025-11-12 15:18:11.617 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 102/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.144s, data_time: 0.002s, total_loss: 4.6, iou_loss: 2.2, l1_loss: 0.7, conf_loss: 1.2, cls_loss: 0.6, lr: 1.173e-04, size: 384, ETA: 0:06:22
2025-11-12 15:18:14.551 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 102/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.144s, data_time: 0.002s, total_loss: 4.6, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.4, cls_loss: 0.5, lr: 1.154e-04, size: 256, ETA: 0:06:18
2025-11-12 15:18:17.647 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 102/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.152s, data_time: 0.003s, total_loss: 4.5, iou_loss: 1.9, l1_loss: 0.7, conf_loss: 1.4, cls_loss: 0.6, lr: 1.136e-04, size: 448, ETA: 0:06:15
2025-11-12 15:18:20.647 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 102/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 5.1, iou_loss: 2.0, l1_loss: 1.0, conf_loss: 1.5, cls_loss: 0.5, lr: 1.117e-04, size: 480, ETA: 0:06:12
2025-11-12 15:18:23.663 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 102/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.148s, data_time: 0.002s, total_loss: 6.1, iou_loss: 2.4, l1_loss: 1.0, conf_loss: 2.1, cls_loss: 0.6, lr: 1.098e-04, size: 512, ETA: 0:06:09
2025-11-12 15:18:25.052 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:18:30.130 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:18:31.029 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:18:31.606 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6533
2025-11-12 15:18:31.738 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.6029
2025-11-12 15:18:31.773 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4358
2025-11-12 15:18:31.773 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5640
2025-11-12 15:18:31.773 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:18:31.774 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:18:31.774 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.653
2025-11-12 15:18:31.774 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.603
2025-11-12 15:18:31.774 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.436
2025-11-12 15:18:31.774 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.564
2025-11-12 15:18:31.774 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:18:31.774 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:18:31.774 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:18:31.774 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:18:31.775 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:18:31.775 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:18:31.775 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:18:31.775 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:18:31.775 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:18:32.554 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:18:33.293 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:18:34.066 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:18:34.799 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:18:35.533 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:18:36.303 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:18:37.043 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:18:37.790 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:18:38.570 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:18:38.570 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.33
2025-11-12 15:18:38.570 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:18:38.571 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:18:38.577 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.16 ms, Average NMS time: 0.61 ms, Average inference time: 2.77 ms

2025-11-12 15:18:38.579 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:18:38.654 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:18:38.734 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch103
2025-11-12 15:18:41.844 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 103/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.153s, data_time: 0.002s, total_loss: 6.1, iou_loss: 2.2, l1_loss: 0.9, conf_loss: 2.3, cls_loss: 0.6, lr: 1.072e-04, size: 480, ETA: 0:06:04
2025-11-12 15:18:45.181 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 103/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.163s, data_time: 0.001s, total_loss: 5.5, iou_loss: 2.1, l1_loss: 1.0, conf_loss: 1.9, cls_loss: 0.6, lr: 1.053e-04, size: 544, ETA: 0:06:01
2025-11-12 15:18:48.135 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 103/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.145s, data_time: 0.002s, total_loss: 6.4, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 2.6, cls_loss: 0.6, lr: 1.035e-04, size: 320, ETA: 0:05:58
2025-11-12 15:18:51.179 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 103/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.150s, data_time: 0.003s, total_loss: 6.0, iou_loss: 2.0, l1_loss: 1.1, conf_loss: 2.3, cls_loss: 0.6, lr: 1.017e-04, size: 576, ETA: 0:05:55
2025-11-12 15:18:54.294 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 103/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.154s, data_time: 0.002s, total_loss: 4.6, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.4, cls_loss: 0.5, lr: 9.996e-05, size: 448, ETA: 0:05:51
2025-11-12 15:18:57.372 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 103/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.003s, total_loss: 5.4, iou_loss: 1.9, l1_loss: 0.7, conf_loss: 2.2, cls_loss: 0.5, lr: 9.820e-05, size: 448, ETA: 0:05:48
2025-11-12 15:18:58.724 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:19:03.733 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:19:04.582 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:19:05.081 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6561
2025-11-12 15:19:05.237 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5880
2025-11-12 15:19:05.274 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4594
2025-11-12 15:19:05.275 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5678
2025-11-12 15:19:05.275 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:19:05.275 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:19:05.275 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.656
2025-11-12 15:19:05.275 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.588
2025-11-12 15:19:05.275 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.459
2025-11-12 15:19:05.276 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.568
2025-11-12 15:19:05.276 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:19:05.276 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:19:05.276 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:19:05.276 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:19:05.276 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:19:05.276 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:19:05.276 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:19:05.276 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:19:05.277 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:19:05.956 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:19:06.658 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:19:07.344 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:19:08.037 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:19:08.756 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:19:09.425 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:19:10.091 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:19:10.756 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:19:11.454 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:19:11.454 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:19:11.454 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.57
2025-11-12 15:19:11.455 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:19:11.461 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.17 ms, Average NMS time: 0.55 ms, Average inference time: 2.73 ms

2025-11-12 15:19:11.463 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:19:11.538 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:19:11.618 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch104
2025-11-12 15:19:14.444 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 104/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.139s, data_time: 0.002s, total_loss: 3.7, iou_loss: 1.5, l1_loss: 0.5, conf_loss: 1.1, cls_loss: 0.6, lr: 9.567e-05, size: 256, ETA: 0:05:44
2025-11-12 15:19:17.276 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 104/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.138s, data_time: 0.002s, total_loss: 6.1, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 2.2, cls_loss: 0.7, lr: 9.395e-05, size: 256, ETA: 0:05:40
2025-11-12 15:19:19.974 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 104/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.134s, data_time: 0.002s, total_loss: 5.1, iou_loss: 2.1, l1_loss: 0.8, conf_loss: 1.6, cls_loss: 0.6, lr: 9.224e-05, size: 384, ETA: 0:05:37
2025-11-12 15:19:23.005 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 104/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 5.0, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.8, cls_loss: 0.6, lr: 9.054e-05, size: 416, ETA: 0:05:34
2025-11-12 15:19:25.886 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 104/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 4.6, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 1.3, cls_loss: 0.6, lr: 8.886e-05, size: 416, ETA: 0:05:31
2025-11-12 15:19:28.874 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 104/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.3, l1_loss: 0.8, conf_loss: 2.1, cls_loss: 0.6, lr: 8.720e-05, size: 512, ETA: 0:05:27
2025-11-12 15:19:30.224 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:19:35.376 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:19:36.232 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:19:36.747 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6537
2025-11-12 15:19:36.907 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5875
2025-11-12 15:19:36.943 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4479
2025-11-12 15:19:36.944 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5630
2025-11-12 15:19:36.944 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:19:36.944 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:19:36.944 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.654
2025-11-12 15:19:36.944 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.588
2025-11-12 15:19:36.945 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.448
2025-11-12 15:19:36.945 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.563
2025-11-12 15:19:36.945 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:19:36.945 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:19:36.945 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:19:36.945 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:19:36.945 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:19:36.945 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:19:36.945 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:19:36.946 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:19:36.946 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:19:37.651 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:19:38.385 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:19:39.085 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:19:39.788 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:19:40.530 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:19:41.243 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:19:41.944 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:19:42.671 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:19:43.366 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:19:43.367 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:19:43.367 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:19:43.367 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:19:43.374 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.22 ms, Average NMS time: 0.56 ms, Average inference time: 2.78 ms

2025-11-12 15:19:43.375 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:19:43.451 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:19:43.530 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch105
2025-11-12 15:19:46.458 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 105/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.145s, data_time: 0.002s, total_loss: 4.5, iou_loss: 1.9, l1_loss: 0.8, conf_loss: 1.3, cls_loss: 0.5, lr: 8.481e-05, size: 480, ETA: 0:05:23
2025-11-12 15:19:49.860 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 105/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.166s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.0, l1_loss: 0.9, conf_loss: 2.0, cls_loss: 0.7, lr: 8.318e-05, size: 544, ETA: 0:05:20
2025-11-12 15:19:53.170 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 105/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.161s, data_time: 0.002s, total_loss: 4.4, iou_loss: 1.7, l1_loss: 0.6, conf_loss: 1.6, cls_loss: 0.5, lr: 8.157e-05, size: 480, ETA: 0:05:16
2025-11-12 15:19:56.644 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 105/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.170s, data_time: 0.003s, total_loss: 5.0, iou_loss: 2.0, l1_loss: 0.8, conf_loss: 1.6, cls_loss: 0.6, lr: 7.997e-05, size: 544, ETA: 0:05:13
2025-11-12 15:19:59.988 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 105/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.163s, data_time: 0.002s, total_loss: 4.6, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.4, cls_loss: 0.5, lr: 7.839e-05, size: 448, ETA: 0:05:10
2025-11-12 15:20:03.275 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 105/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.160s, data_time: 0.002s, total_loss: 4.1, iou_loss: 1.7, l1_loss: 0.7, conf_loss: 1.2, cls_loss: 0.5, lr: 7.682e-05, size: 544, ETA: 0:05:07
2025-11-12 15:20:04.786 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:20:09.839 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:20:10.747 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:20:11.349 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6528
2025-11-12 15:20:11.476 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5855
2025-11-12 15:20:11.511 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4493
2025-11-12 15:20:11.512 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5626
2025-11-12 15:20:11.512 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:20:11.512 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:20:11.512 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.653
2025-11-12 15:20:11.512 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.586
2025-11-12 15:20:11.512 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.449
2025-11-12 15:20:11.512 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.563
2025-11-12 15:20:11.513 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:20:11.513 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:20:11.513 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:20:11.513 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:20:11.513 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:20:11.513 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:20:11.513 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:20:11.513 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:20:11.513 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:20:12.293 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:20:13.035 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:20:13.812 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:20:14.548 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:20:15.296 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:20:16.074 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:20:16.819 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:20:17.569 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:20:18.374 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:20:18.375 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:20:18.375 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:20:18.375 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:20:18.382 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.20 ms, Average NMS time: 0.55 ms, Average inference time: 2.75 ms

2025-11-12 15:20:18.384 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:20:18.461 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:20:18.542 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch106
2025-11-12 15:20:21.440 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 106/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.141s, data_time: 0.003s, total_loss: 5.4, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 2.1, cls_loss: 0.7, lr: 7.457e-05, size: 256, ETA: 0:05:02
2025-11-12 15:20:24.567 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 106/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.152s, data_time: 0.002s, total_loss: 3.9, iou_loss: 1.8, l1_loss: 0.6, conf_loss: 1.0, cls_loss: 0.5, lr: 7.304e-05, size: 256, ETA: 0:04:59
2025-11-12 15:20:27.706 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 106/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.153s, data_time: 0.002s, total_loss: 4.5, iou_loss: 1.8, l1_loss: 0.7, conf_loss: 1.5, cls_loss: 0.5, lr: 7.153e-05, size: 544, ETA: 0:04:56
2025-11-12 15:20:30.722 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 106/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.147s, data_time: 0.003s, total_loss: 2.8, iou_loss: 1.3, l1_loss: 0.4, conf_loss: 0.7, cls_loss: 0.4, lr: 7.003e-05, size: 384, ETA: 0:04:53
2025-11-12 15:20:33.620 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 106/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.142s, data_time: 0.003s, total_loss: 4.6, iou_loss: 1.9, l1_loss: 0.7, conf_loss: 1.4, cls_loss: 0.5, lr: 6.854e-05, size: 576, ETA: 0:04:50
2025-11-12 15:20:36.719 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 106/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.152s, data_time: 0.002s, total_loss: 5.3, iou_loss: 2.3, l1_loss: 0.8, conf_loss: 1.6, cls_loss: 0.6, lr: 6.708e-05, size: 576, ETA: 0:04:46
2025-11-12 15:20:38.148 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:20:43.241 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:20:44.231 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:20:44.905 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6567
2025-11-12 15:20:45.044 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5823
2025-11-12 15:20:45.081 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4564
2025-11-12 15:20:45.081 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5651
2025-11-12 15:20:45.081 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:20:45.082 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:20:45.082 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.657
2025-11-12 15:20:45.082 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.582
2025-11-12 15:20:45.082 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.456
2025-11-12 15:20:45.082 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.565
2025-11-12 15:20:45.082 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:20:45.082 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:20:45.082 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:20:45.082 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:20:45.082 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:20:45.083 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:20:45.083 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:20:45.083 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:20:45.083 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:20:45.928 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:20:46.722 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:20:47.560 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:20:48.370 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:20:49.167 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:20:49.999 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:20:50.798 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:20:51.648 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:20:52.502 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:20:52.503 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:20:52.503 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.57
2025-11-12 15:20:52.503 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:20:52.512 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.22 ms, Average NMS time: 0.58 ms, Average inference time: 2.80 ms

2025-11-12 15:20:52.513 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:20:52.590 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:20:52.672 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch107
2025-11-12 15:20:55.584 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 107/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.143s, data_time: 0.002s, total_loss: 4.4, iou_loss: 1.8, l1_loss: 0.5, conf_loss: 1.5, cls_loss: 0.5, lr: 6.497e-05, size: 384, ETA: 0:04:42
2025-11-12 15:20:58.437 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 107/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 4.2, iou_loss: 1.6, l1_loss: 0.4, conf_loss: 1.5, cls_loss: 0.6, lr: 6.354e-05, size: 256, ETA: 0:04:39
2025-11-12 15:21:01.691 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 107/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.159s, data_time: 0.002s, total_loss: 5.3, iou_loss: 2.0, l1_loss: 0.7, conf_loss: 2.0, cls_loss: 0.6, lr: 6.213e-05, size: 384, ETA: 0:04:35
2025-11-12 15:21:04.781 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 107/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.150s, data_time: 0.005s, total_loss: 4.6, iou_loss: 1.9, l1_loss: 0.7, conf_loss: 1.4, cls_loss: 0.5, lr: 6.072e-05, size: 480, ETA: 0:04:32
2025-11-12 15:21:07.976 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 107/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.157s, data_time: 0.003s, total_loss: 4.2, iou_loss: 1.8, l1_loss: 0.6, conf_loss: 1.3, cls_loss: 0.5, lr: 5.934e-05, size: 480, ETA: 0:04:29
2025-11-12 15:21:10.998 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 107/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.148s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.4, l1_loss: 0.7, conf_loss: 2.2, cls_loss: 0.6, lr: 5.797e-05, size: 288, ETA: 0:04:26
2025-11-12 15:21:12.270 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:21:18.423 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:21:19.418 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:21:20.078 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6585
2025-11-12 15:21:20.250 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5858
2025-11-12 15:21:20.290 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4418
2025-11-12 15:21:20.291 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5620
2025-11-12 15:21:20.291 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:21:20.291 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:21:20.291 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.658
2025-11-12 15:21:20.291 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.586
2025-11-12 15:21:20.291 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.442
2025-11-12 15:21:20.291 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.562
2025-11-12 15:21:20.291 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:21:20.291 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:21:20.292 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:21:20.292 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:21:20.292 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:21:20.292 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:21:20.292 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:21:20.292 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:21:20.292 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:21:21.125 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:21:22.000 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:21:22.828 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:21:23.682 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:21:24.494 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:21:25.348 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:21:26.160 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:21:26.984 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:21:27.832 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:21:27.832 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:21:27.833 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:21:27.833 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:21:27.840 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.43 ms, Average NMS time: 0.68 ms, Average inference time: 3.11 ms

2025-11-12 15:21:27.841 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:21:27.919 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:21:28.000 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch108
2025-11-12 15:21:31.033 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 108/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 4.9, iou_loss: 2.1, l1_loss: 0.8, conf_loss: 1.4, cls_loss: 0.6, lr: 5.601e-05, size: 512, ETA: 0:04:21
2025-11-12 15:21:34.032 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 108/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.2, l1_loss: 0.7, conf_loss: 2.3, cls_loss: 0.6, lr: 5.468e-05, size: 320, ETA: 0:04:18
2025-11-12 15:21:37.084 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 108/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 6.6, iou_loss: 2.2, l1_loss: 1.0, conf_loss: 2.7, cls_loss: 0.6, lr: 5.336e-05, size: 576, ETA: 0:04:15
2025-11-12 15:21:40.092 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 108/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.147s, data_time: 0.003s, total_loss: 6.9, iou_loss: 2.4, l1_loss: 1.0, conf_loss: 2.5, cls_loss: 0.9, lr: 5.206e-05, size: 544, ETA: 0:04:12
2025-11-12 15:21:43.089 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 108/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.147s, data_time: 0.003s, total_loss: 5.3, iou_loss: 2.2, l1_loss: 0.7, conf_loss: 1.7, cls_loss: 0.6, lr: 5.078e-05, size: 512, ETA: 0:04:09
2025-11-12 15:21:46.252 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 108/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.155s, data_time: 0.003s, total_loss: 5.7, iou_loss: 2.2, l1_loss: 0.9, conf_loss: 2.0, cls_loss: 0.6, lr: 4.951e-05, size: 480, ETA: 0:04:05
2025-11-12 15:21:47.603 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:21:52.736 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:21:53.641 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:21:54.216 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6526
2025-11-12 15:21:54.352 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5946
2025-11-12 15:21:54.388 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4364
2025-11-12 15:21:54.388 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5612
2025-11-12 15:21:54.388 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:21:54.388 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:21:54.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.653
2025-11-12 15:21:54.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.595
2025-11-12 15:21:54.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.436
2025-11-12 15:21:54.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.561
2025-11-12 15:21:54.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:21:54.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:21:54.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:21:54.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:21:54.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:21:54.389 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:21:54.390 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:21:54.390 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:21:54.390 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:21:55.161 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:21:55.901 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:21:56.652 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:21:57.428 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:21:58.151 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:21:58.879 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:21:59.643 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:22:00.374 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:22:01.147 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:22:01.147 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:22:01.148 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:22:01.148 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:22:01.155 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.23 ms, Average NMS time: 0.57 ms, Average inference time: 2.80 ms

2025-11-12 15:22:01.157 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:22:01.234 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:22:01.319 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch109
2025-11-12 15:22:04.277 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 109/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.144s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 1.7, cls_loss: 0.6, lr: 4.770e-05, size: 352, ETA: 0:04:01
2025-11-12 15:22:07.297 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 109/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 4.1, iou_loss: 1.6, l1_loss: 0.6, conf_loss: 1.4, cls_loss: 0.5, lr: 4.647e-05, size: 416, ETA: 0:03:58
2025-11-12 15:22:10.545 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 109/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.158s, data_time: 0.002s, total_loss: 6.2, iou_loss: 2.4, l1_loss: 0.9, conf_loss: 2.3, cls_loss: 0.6, lr: 4.525e-05, size: 544, ETA: 0:03:54
2025-11-12 15:22:13.862 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 109/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.161s, data_time: 0.002s, total_loss: 5.2, iou_loss: 1.9, l1_loss: 0.8, conf_loss: 1.9, cls_loss: 0.6, lr: 4.405e-05, size: 576, ETA: 0:03:51
2025-11-12 15:22:17.101 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 109/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.158s, data_time: 0.002s, total_loss: 4.8, iou_loss: 1.8, l1_loss: 0.7, conf_loss: 1.9, cls_loss: 0.5, lr: 4.287e-05, size: 480, ETA: 0:03:48
2025-11-12 15:22:20.093 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 109/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.148s, data_time: 0.002s, total_loss: 4.2, iou_loss: 1.9, l1_loss: 0.6, conf_loss: 1.1, cls_loss: 0.6, lr: 4.170e-05, size: 320, ETA: 0:03:45
2025-11-12 15:22:21.411 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:22:26.367 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:22:27.390 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:22:27.995 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6553
2025-11-12 15:22:28.160 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5864
2025-11-12 15:22:28.201 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4385
2025-11-12 15:22:28.201 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5601
2025-11-12 15:22:28.202 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:22:28.202 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:22:28.202 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.655
2025-11-12 15:22:28.202 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.586
2025-11-12 15:22:28.202 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.438
2025-11-12 15:22:28.202 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.560
2025-11-12 15:22:28.202 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:22:28.202 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:22:28.202 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:22:28.203 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:22:28.203 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:22:28.203 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:22:28.203 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:22:28.203 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:22:28.203 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:22:29.011 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:22:29.842 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:22:30.671 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:22:31.516 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:22:32.303 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:22:33.082 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:22:33.910 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:22:34.699 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:22:35.523 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:22:35.523 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:22:35.523 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:22:35.523 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:22:35.530 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.11 ms, Average NMS time: 0.53 ms, Average inference time: 2.65 ms

2025-11-12 15:22:35.532 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:22:35.607 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:22:35.688 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch110
2025-11-12 15:22:38.558 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 110/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.141s, data_time: 0.002s, total_loss: 4.2, iou_loss: 1.7, l1_loss: 0.8, conf_loss: 1.2, cls_loss: 0.5, lr: 4.004e-05, size: 512, ETA: 0:03:40
2025-11-12 15:22:41.739 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 110/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.156s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.2, l1_loss: 0.9, conf_loss: 2.2, cls_loss: 0.6, lr: 3.891e-05, size: 576, ETA: 0:03:37
2025-11-12 15:22:44.694 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 110/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.144s, data_time: 0.003s, total_loss: 3.7, iou_loss: 1.6, l1_loss: 0.5, conf_loss: 1.1, cls_loss: 0.5, lr: 3.779e-05, size: 448, ETA: 0:03:34
2025-11-12 15:22:47.617 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 110/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 4.7, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.6, cls_loss: 0.5, lr: 3.670e-05, size: 256, ETA: 0:03:31
2025-11-12 15:22:50.525 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 110/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.144s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.1, l1_loss: 0.8, conf_loss: 2.1, cls_loss: 0.6, lr: 3.561e-05, size: 480, ETA: 0:03:28
2025-11-12 15:22:53.508 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 110/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 5.2, iou_loss: 2.2, l1_loss: 0.6, conf_loss: 1.7, cls_loss: 0.6, lr: 3.455e-05, size: 256, ETA: 0:03:24
2025-11-12 15:22:54.692 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:22:59.708 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:23:00.719 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:23:01.325 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6499
2025-11-12 15:23:01.506 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5887
2025-11-12 15:23:01.546 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4402
2025-11-12 15:23:01.547 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5596
2025-11-12 15:23:01.547 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:23:01.547 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:23:01.547 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.650
2025-11-12 15:23:01.547 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.589
2025-11-12 15:23:01.548 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.440
2025-11-12 15:23:01.548 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.560
2025-11-12 15:23:01.548 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:23:01.548 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:23:01.548 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:23:01.548 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:23:01.548 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:23:01.548 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:23:01.548 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:23:01.548 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:23:01.549 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:23:02.462 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:23:03.281 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:23:04.150 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:23:04.970 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:23:05.800 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:23:06.663 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:23:07.499 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:23:08.362 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:23:09.283 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:23:09.284 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:23:09.284 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:23:09.284 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:23:09.293 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.14 ms, Average NMS time: 0.55 ms, Average inference time: 2.69 ms

2025-11-12 15:23:09.294 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:23:09.371 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:23:09.453 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch111
2025-11-12 15:23:12.305 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 111/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.141s, data_time: 0.002s, total_loss: 4.9, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 1.3, cls_loss: 0.6, lr: 3.303e-05, size: 512, ETA: 0:03:20
2025-11-12 15:23:15.257 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 111/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 3.9, iou_loss: 1.8, l1_loss: 0.6, conf_loss: 1.0, cls_loss: 0.5, lr: 3.200e-05, size: 448, ETA: 0:03:17
2025-11-12 15:23:18.159 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 111/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.143s, data_time: 0.002s, total_loss: 4.4, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.3, cls_loss: 0.6, lr: 3.099e-05, size: 448, ETA: 0:03:13
2025-11-12 15:23:21.061 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 111/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 4.9, iou_loss: 1.9, l1_loss: 0.7, conf_loss: 1.5, cls_loss: 0.8, lr: 3.000e-05, size: 256, ETA: 0:03:10
2025-11-12 15:23:24.116 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 111/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.003s, total_loss: 5.6, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 2.0, cls_loss: 0.6, lr: 2.902e-05, size: 576, ETA: 0:03:07
2025-11-12 15:23:26.960 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 111/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.141s, data_time: 0.002s, total_loss: 5.2, iou_loss: 2.4, l1_loss: 0.8, conf_loss: 1.3, cls_loss: 0.6, lr: 2.806e-05, size: 416, ETA: 0:03:04
2025-11-12 15:23:28.318 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:23:33.469 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:23:34.440 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:23:35.075 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6611
2025-11-12 15:23:35.205 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5879
2025-11-12 15:23:35.284 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4493
2025-11-12 15:23:35.285 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5661
2025-11-12 15:23:35.285 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:23:35.285 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:23:35.285 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.661
2025-11-12 15:23:35.285 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.588
2025-11-12 15:23:35.285 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.449
2025-11-12 15:23:35.285 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.566
2025-11-12 15:23:35.286 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:23:35.286 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:23:35.286 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:23:35.286 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:23:35.286 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:23:35.286 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:23:35.286 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:23:35.286 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:23:35.287 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:23:36.047 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:23:36.858 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:23:37.631 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:23:38.412 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:23:39.216 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:23:39.983 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:23:40.749 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:23:41.560 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:23:42.407 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:23:42.407 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.33
2025-11-12 15:23:42.407 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.57
2025-11-12 15:23:42.407 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:23:42.416 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.17 ms, Average NMS time: 0.58 ms, Average inference time: 2.75 ms

2025-11-12 15:23:42.417 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:23:42.495 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:23:42.577 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch112
2025-11-12 15:23:45.504 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 112/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 5.5, iou_loss: 2.3, l1_loss: 0.7, conf_loss: 1.9, cls_loss: 0.6, lr: 2.669e-05, size: 448, ETA: 0:02:59
2025-11-12 15:23:48.384 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 112/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.143s, data_time: 0.002s, total_loss: 4.4, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.2, cls_loss: 0.6, lr: 2.577e-05, size: 416, ETA: 0:02:56
2025-11-12 15:23:51.438 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 112/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.004s, total_loss: 5.5, iou_loss: 2.1, l1_loss: 0.8, conf_loss: 2.0, cls_loss: 0.5, lr: 2.486e-05, size: 416, ETA: 0:02:53
2025-11-12 15:23:54.431 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 112/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.148s, data_time: 0.002s, total_loss: 4.2, iou_loss: 1.6, l1_loss: 0.6, conf_loss: 1.4, cls_loss: 0.5, lr: 2.397e-05, size: 544, ETA: 0:02:50
2025-11-12 15:23:57.490 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 112/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.150s, data_time: 0.002s, total_loss: 5.1, iou_loss: 2.0, l1_loss: 0.8, conf_loss: 1.6, cls_loss: 0.6, lr: 2.309e-05, size: 512, ETA: 0:02:47
2025-11-12 15:24:00.535 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 112/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.002s, total_loss: 4.5, iou_loss: 1.8, l1_loss: 0.7, conf_loss: 1.5, cls_loss: 0.5, lr: 2.223e-05, size: 512, ETA: 0:02:44
2025-11-12 15:24:01.836 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:24:06.810 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:24:07.606 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:24:08.111 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6539
2025-11-12 15:24:08.229 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5877
2025-11-12 15:24:08.303 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4477
2025-11-12 15:24:08.304 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5631
2025-11-12 15:24:08.304 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:24:08.304 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:24:08.304 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.654
2025-11-12 15:24:08.304 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.588
2025-11-12 15:24:08.305 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.448
2025-11-12 15:24:08.305 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.563
2025-11-12 15:24:08.305 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:24:08.305 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:24:08.305 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:24:08.305 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:24:08.305 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:24:08.305 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:24:08.305 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:24:08.305 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:24:08.306 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:24:08.951 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:24:09.599 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:24:10.243 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:24:10.918 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:24:11.559 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:24:12.220 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:24:12.858 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:24:13.542 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:24:14.211 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:24:14.211 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.33
2025-11-12 15:24:14.211 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:24:14.212 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:24:14.218 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.15 ms, Average NMS time: 0.57 ms, Average inference time: 2.72 ms

2025-11-12 15:24:14.220 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:24:14.296 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:24:14.377 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch113
2025-11-12 15:24:17.574 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 113/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.157s, data_time: 0.002s, total_loss: 4.2, iou_loss: 1.7, l1_loss: 0.6, conf_loss: 1.3, cls_loss: 0.5, lr: 2.102e-05, size: 448, ETA: 0:02:39
2025-11-12 15:24:20.855 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 113/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.160s, data_time: 0.002s, total_loss: 4.0, iou_loss: 1.8, l1_loss: 0.7, conf_loss: 1.0, cls_loss: 0.5, lr: 2.020e-05, size: 448, ETA: 0:02:36
2025-11-12 15:24:24.249 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 113/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.166s, data_time: 0.003s, total_loss: 5.6, iou_loss: 2.3, l1_loss: 1.0, conf_loss: 1.7, cls_loss: 0.6, lr: 1.939e-05, size: 576, ETA: 0:02:33
2025-11-12 15:24:27.369 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 113/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.152s, data_time: 0.001s, total_loss: 5.1, iou_loss: 2.3, l1_loss: 0.6, conf_loss: 1.5, cls_loss: 0.6, lr: 1.861e-05, size: 256, ETA: 0:02:30
2025-11-12 15:24:30.574 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 113/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.156s, data_time: 0.001s, total_loss: 3.9, iou_loss: 1.6, l1_loss: 0.5, conf_loss: 1.2, cls_loss: 0.5, lr: 1.783e-05, size: 352, ETA: 0:02:26
2025-11-12 15:24:33.572 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 113/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.146s, data_time: 0.002s, total_loss: 4.3, iou_loss: 1.7, l1_loss: 0.7, conf_loss: 1.4, cls_loss: 0.5, lr: 1.708e-05, size: 480, ETA: 0:02:23
2025-11-12 15:24:34.985 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:24:40.084 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:24:40.999 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:24:41.587 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6534
2025-11-12 15:24:41.719 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5960
2025-11-12 15:24:41.757 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4413
2025-11-12 15:24:41.758 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5636
2025-11-12 15:24:41.758 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:24:41.758 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:24:41.758 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.653
2025-11-12 15:24:41.758 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.596
2025-11-12 15:24:41.758 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.441
2025-11-12 15:24:41.758 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.564
2025-11-12 15:24:41.758 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:24:41.759 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:24:41.759 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:24:41.759 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:24:41.759 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:24:41.759 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:24:41.759 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:24:41.759 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:24:41.759 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:24:42.541 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:24:43.289 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:24:44.070 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:24:44.810 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:24:45.589 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:24:46.367 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:24:47.113 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:24:47.859 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:24:48.661 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:24:48.662 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.33
2025-11-12 15:24:48.662 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:24:48.662 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:24:48.669 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.19 ms, Average NMS time: 0.56 ms, Average inference time: 2.75 ms

2025-11-12 15:24:48.670 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:24:48.746 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:24:48.827 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch114
2025-11-12 15:24:51.881 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 114/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.150s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.3, l1_loss: 0.7, conf_loss: 2.0, cls_loss: 0.7, lr: 1.601e-05, size: 352, ETA: 0:02:19
2025-11-12 15:24:55.148 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 114/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.159s, data_time: 0.002s, total_loss: 5.8, iou_loss: 2.3, l1_loss: 0.8, conf_loss: 2.0, cls_loss: 0.6, lr: 1.530e-05, size: 576, ETA: 0:02:15
2025-11-12 15:24:58.519 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 114/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.165s, data_time: 0.003s, total_loss: 3.5, iou_loss: 1.6, l1_loss: 0.7, conf_loss: 0.8, cls_loss: 0.5, lr: 1.460e-05, size: 448, ETA: 0:02:12
2025-11-12 15:25:01.545 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 114/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.147s, data_time: 0.003s, total_loss: 7.0, iou_loss: 2.6, l1_loss: 0.9, conf_loss: 2.9, cls_loss: 0.6, lr: 1.392e-05, size: 352, ETA: 0:02:09
2025-11-12 15:25:04.742 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 114/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.156s, data_time: 0.002s, total_loss: 6.2, iou_loss: 2.5, l1_loss: 1.0, conf_loss: 1.9, cls_loss: 0.8, lr: 1.325e-05, size: 544, ETA: 0:02:06
2025-11-12 15:25:08.062 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 114/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.162s, data_time: 0.002s, total_loss: 6.8, iou_loss: 2.8, l1_loss: 1.1, conf_loss: 2.2, cls_loss: 0.7, lr: 1.260e-05, size: 384, ETA: 0:02:03
2025-11-12 15:25:09.564 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:25:14.693 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:25:15.581 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:25:16.159 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6550
2025-11-12 15:25:16.286 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5899
2025-11-12 15:25:16.324 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4436
2025-11-12 15:25:16.325 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5628
2025-11-12 15:25:16.325 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:25:16.325 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:25:16.325 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.655
2025-11-12 15:25:16.326 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.590
2025-11-12 15:25:16.326 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.444
2025-11-12 15:25:16.326 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.563
2025-11-12 15:25:16.326 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:25:16.326 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:25:16.326 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:25:16.327 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:25:16.327 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:25:16.327 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:25:16.327 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:25:16.327 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:25:16.327 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:25:17.097 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:25:17.839 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:25:18.596 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:25:19.381 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:25:20.111 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:25:20.849 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:25:21.612 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:25:22.332 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:25:23.071 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:25:23.071 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.33
2025-11-12 15:25:23.071 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:25:23.072 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:25:23.079 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.23 ms, Average NMS time: 0.57 ms, Average inference time: 2.80 ms

2025-11-12 15:25:23.080 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:25:23.157 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:25:23.280 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch115
2025-11-12 15:25:26.079 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 115/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.138s, data_time: 0.002s, total_loss: 6.0, iou_loss: 2.4, l1_loss: 0.7, conf_loss: 2.1, cls_loss: 0.7, lr: 1.168e-05, size: 256, ETA: 0:01:58
2025-11-12 15:25:28.886 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 115/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.3Gb, iter_time: 0.138s, data_time: 0.002s, total_loss: 4.0, iou_loss: 1.7, l1_loss: 0.8, conf_loss: 1.0, cls_loss: 0.5, lr: 1.107e-05, size: 480, ETA: 0:01:55
2025-11-12 15:25:31.829 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 115/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.3Gb, iter_time: 0.144s, data_time: 0.003s, total_loss: 5.3, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 1.3, cls_loss: 0.7, lr: 1.048e-05, size: 256, ETA: 0:01:52
2025-11-12 15:25:34.836 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 115/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.3Gb, iter_time: 0.147s, data_time: 0.001s, total_loss: 4.9, iou_loss: 2.1, l1_loss: 0.6, conf_loss: 1.6, cls_loss: 0.6, lr: 9.902e-06, size: 352, ETA: 0:01:49
2025-11-12 15:25:37.990 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 115/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.3Gb, iter_time: 0.154s, data_time: 0.002s, total_loss: 4.8, iou_loss: 1.8, l1_loss: 0.7, conf_loss: 1.7, cls_loss: 0.7, lr: 9.341e-06, size: 384, ETA: 0:01:46
2025-11-12 15:25:41.109 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 115/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.3Gb, iter_time: 0.152s, data_time: 0.002s, total_loss: 3.9, iou_loss: 1.6, l1_loss: 0.7, conf_loss: 1.2, cls_loss: 0.5, lr: 8.795e-06, size: 576, ETA: 0:01:42
2025-11-12 15:25:42.589 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:25:47.574 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:25:48.520 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:25:49.116 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6518
2025-11-12 15:25:49.245 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5843
2025-11-12 15:25:49.321 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4424
2025-11-12 15:25:49.322 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5595
2025-11-12 15:25:49.322 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:25:49.322 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:25:49.322 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.652
2025-11-12 15:25:49.322 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.584
2025-11-12 15:25:49.322 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.442
2025-11-12 15:25:49.323 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.560
2025-11-12 15:25:49.323 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:25:49.323 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:25:49.323 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:25:49.323 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:25:49.323 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:25:49.323 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:25:49.323 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:25:49.323 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:25:49.324 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:25:50.071 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:25:50.810 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:25:51.592 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:25:52.360 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:25:53.176 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:25:53.929 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:25:54.727 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:25:55.546 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:25:56.287 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:25:56.287 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.33
2025-11-12 15:25:56.288 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:25:56.288 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:25:56.295 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.20 ms, Average NMS time: 0.55 ms, Average inference time: 2.75 ms

2025-11-12 15:25:56.296 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:25:56.376 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:25:56.459 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch116
2025-11-12 15:25:59.458 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 116/120, iter: 20/129, gpu mem: 2423Mb, mem: 94.3Gb, iter_time: 0.148s, data_time: 0.002s, total_loss: 4.8, iou_loss: 2.1, l1_loss: 0.8, conf_loss: 1.3, cls_loss: 0.6, lr: 8.034e-06, size: 448, ETA: 0:01:38
2025-11-12 15:26:02.587 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 116/120, iter: 40/129, gpu mem: 2423Mb, mem: 94.4Gb, iter_time: 0.153s, data_time: 0.003s, total_loss: 5.7, iou_loss: 2.1, l1_loss: 0.9, conf_loss: 2.1, cls_loss: 0.6, lr: 7.528e-06, size: 512, ETA: 0:01:35
2025-11-12 15:26:05.513 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 116/120, iter: 60/129, gpu mem: 2423Mb, mem: 94.5Gb, iter_time: 0.143s, data_time: 0.003s, total_loss: 5.1, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 1.6, cls_loss: 0.6, lr: 7.039e-06, size: 320, ETA: 0:01:32
2025-11-12 15:26:08.300 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 116/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.138s, data_time: 0.002s, total_loss: 4.3, iou_loss: 2.1, l1_loss: 0.7, conf_loss: 1.0, cls_loss: 0.6, lr: 6.567e-06, size: 384, ETA: 0:01:28
2025-11-12 15:26:11.309 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 116/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.1Gb, iter_time: 0.147s, data_time: 0.002s, total_loss: 5.3, iou_loss: 2.2, l1_loss: 0.7, conf_loss: 1.7, cls_loss: 0.6, lr: 6.111e-06, size: 576, ETA: 0:01:25
2025-11-12 15:26:14.283 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 116/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.146s, data_time: 0.003s, total_loss: 4.7, iou_loss: 1.9, l1_loss: 0.5, conf_loss: 1.8, cls_loss: 0.5, lr: 5.671e-06, size: 320, ETA: 0:01:22
2025-11-12 15:26:15.545 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:26:20.723 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:26:21.649 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:26:22.254 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6569
2025-11-12 15:26:22.386 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5867
2025-11-12 15:26:22.422 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4405
2025-11-12 15:26:22.423 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5614
2025-11-12 15:26:22.423 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:26:22.423 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:26:22.423 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.657
2025-11-12 15:26:22.423 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.587
2025-11-12 15:26:22.424 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.441
2025-11-12 15:26:22.424 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.561
2025-11-12 15:26:22.424 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:26:22.424 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:26:22.424 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:26:22.424 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:26:22.424 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:26:22.424 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:26:22.424 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:26:22.424 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:26:22.425 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:26:23.217 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:26:23.980 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:26:24.770 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:26:25.527 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:26:26.287 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:26:27.088 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:26:27.847 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:26:28.608 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:26:29.427 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:26:29.427 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.33
2025-11-12 15:26:29.427 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:26:29.427 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:26:29.434 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.20 ms, Average NMS time: 0.59 ms, Average inference time: 2.79 ms

2025-11-12 15:26:29.436 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:26:29.511 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:26:29.592 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch117
2025-11-12 15:26:32.753 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 117/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.155s, data_time: 0.002s, total_loss: 5.0, iou_loss: 2.0, l1_loss: 0.6, conf_loss: 1.8, cls_loss: 0.5, lr: 5.062e-06, size: 384, ETA: 0:01:18
2025-11-12 15:26:35.868 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 117/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.152s, data_time: 0.002s, total_loss: 4.3, iou_loss: 1.7, l1_loss: 0.8, conf_loss: 1.3, cls_loss: 0.5, lr: 4.662e-06, size: 544, ETA: 0:01:14
2025-11-12 15:26:38.768 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 117/120, iter: 60/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.142s, data_time: 0.002s, total_loss: 5.0, iou_loss: 2.1, l1_loss: 0.8, conf_loss: 1.5, cls_loss: 0.6, lr: 4.279e-06, size: 288, ETA: 0:01:11
2025-11-12 15:26:41.637 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 117/120, iter: 80/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.140s, data_time: 0.002s, total_loss: 5.9, iou_loss: 2.2, l1_loss: 0.8, conf_loss: 2.3, cls_loss: 0.6, lr: 3.912e-06, size: 448, ETA: 0:01:08
2025-11-12 15:26:44.681 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 117/120, iter: 100/129, gpu mem: 2423Mb, mem: 95.2Gb, iter_time: 0.149s, data_time: 0.001s, total_loss: 3.8, iou_loss: 1.6, l1_loss: 0.5, conf_loss: 1.2, cls_loss: 0.5, lr: 3.562e-06, size: 416, ETA: 0:01:05
2025-11-12 15:26:47.750 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 117/120, iter: 120/129, gpu mem: 2423Mb, mem: 95.3Gb, iter_time: 0.150s, data_time: 0.001s, total_loss: 5.6, iou_loss: 2.5, l1_loss: 0.7, conf_loss: 1.7, cls_loss: 0.6, lr: 3.228e-06, size: 288, ETA: 0:01:02
2025-11-12 15:26:48.972 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:26:53.977 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:26:54.884 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:26:55.479 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6542
2025-11-12 15:26:55.609 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5871
2025-11-12 15:26:55.647 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4391
2025-11-12 15:26:55.648 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5601
2025-11-12 15:26:55.648 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:26:55.648 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:26:55.648 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.654
2025-11-12 15:26:55.648 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.587
2025-11-12 15:26:55.648 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.439
2025-11-12 15:26:55.648 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.560
2025-11-12 15:26:55.648 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:26:55.649 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:26:55.649 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:26:55.649 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:26:55.649 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:26:55.649 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:26:55.649 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:26:55.649 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:26:55.649 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:26:56.427 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:26:57.164 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:26:57.903 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:26:58.697 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:26:59.448 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:27:00.222 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:27:00.959 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:27:01.698 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:27:02.472 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:27:02.472 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.33
2025-11-12 15:27:02.472 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:27:02.472 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:27:02.479 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.18 ms, Average NMS time: 0.54 ms, Average inference time: 2.71 ms

2025-11-12 15:27:02.481 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:27:02.556 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:27:02.637 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch118
2025-11-12 15:27:05.655 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 118/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.148s, data_time: 0.002s, total_loss: 6.8, iou_loss: 2.6, l1_loss: 0.9, conf_loss: 2.7, cls_loss: 0.6, lr: 2.772e-06, size: 320, ETA: 0:00:57
2025-11-12 15:27:08.501 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 118/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.139s, data_time: 0.002s, total_loss: 6.4, iou_loss: 2.4, l1_loss: 1.0, conf_loss: 2.4, cls_loss: 0.6, lr: 2.479e-06, size: 576, ETA: 0:00:54
2025-11-12 15:27:11.573 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 118/120, iter: 60/129, gpu mem: 2423Mb, mem: 94.9Gb, iter_time: 0.152s, data_time: 0.004s, total_loss: 5.1, iou_loss: 2.2, l1_loss: 0.9, conf_loss: 1.4, cls_loss: 0.6, lr: 2.201e-06, size: 320, ETA: 0:00:51
2025-11-12 15:27:14.571 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 118/120, iter: 80/129, gpu mem: 2423Mb, mem: 94.9Gb, iter_time: 0.148s, data_time: 0.002s, total_loss: 5.6, iou_loss: 2.5, l1_loss: 0.8, conf_loss: 1.6, cls_loss: 0.6, lr: 1.940e-06, size: 416, ETA: 0:00:48
2025-11-12 15:27:17.559 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 118/120, iter: 100/129, gpu mem: 2423Mb, mem: 94.9Gb, iter_time: 0.146s, data_time: 0.003s, total_loss: 4.9, iou_loss: 2.1, l1_loss: 0.7, conf_loss: 1.5, cls_loss: 0.6, lr: 1.696e-06, size: 352, ETA: 0:00:45
2025-11-12 15:27:20.370 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 118/120, iter: 120/129, gpu mem: 2423Mb, mem: 94.9Gb, iter_time: 0.138s, data_time: 0.002s, total_loss: 4.9, iou_loss: 2.2, l1_loss: 0.9, conf_loss: 1.3, cls_loss: 0.5, lr: 1.468e-06, size: 480, ETA: 0:00:41
2025-11-12 15:27:21.808 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:27:26.862 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:27:27.789 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:27:28.354 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6543
2025-11-12 15:27:28.526 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5931
2025-11-12 15:27:28.565 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4425
2025-11-12 15:27:28.565 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5633
2025-11-12 15:27:28.565 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:27:28.565 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:27:28.566 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.654
2025-11-12 15:27:28.566 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.593
2025-11-12 15:27:28.566 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.442
2025-11-12 15:27:28.566 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.563
2025-11-12 15:27:28.566 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:27:28.566 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:27:28.566 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:27:28.566 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:27:28.566 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:27:28.567 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:27:28.567 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:27:28.567 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:27:28.567 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:27:29.356 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:27:30.106 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:27:30.862 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:27:31.649 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:27:32.402 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:27:33.151 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:27:33.937 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:27:34.701 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:27:35.497 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:27:35.498 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:27:35.498 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:27:35.498 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:27:35.504 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.13 ms, Average NMS time: 0.57 ms, Average inference time: 2.70 ms

2025-11-12 15:27:35.506 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:27:35.581 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:27:35.661 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch119
2025-11-12 15:27:38.731 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 119/120, iter: 20/129, gpu mem: 2423Mb, mem: 94.9Gb, iter_time: 0.150s, data_time: 0.002s, total_loss: 5.8, iou_loss: 2.4, l1_loss: 0.8, conf_loss: 1.9, cls_loss: 0.7, lr: 1.166e-06, size: 288, ETA: 0:00:37
2025-11-12 15:27:41.825 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 119/120, iter: 40/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.151s, data_time: 0.003s, total_loss: 4.5, iou_loss: 1.9, l1_loss: 0.6, conf_loss: 1.5, cls_loss: 0.5, lr: 9.785e-07, size: 320, ETA: 0:00:34
2025-11-12 15:27:44.889 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 119/120, iter: 60/129, gpu mem: 2423Mb, mem: 94.9Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 3.7, iou_loss: 1.7, l1_loss: 0.6, conf_loss: 0.9, cls_loss: 0.5, lr: 8.072e-07, size: 480, ETA: 0:00:31
2025-11-12 15:27:47.956 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 119/120, iter: 80/129, gpu mem: 2423Mb, mem: 94.9Gb, iter_time: 0.149s, data_time: 0.002s, total_loss: 5.4, iou_loss: 2.3, l1_loss: 0.7, conf_loss: 1.8, cls_loss: 0.6, lr: 6.524e-07, size: 288, ETA: 0:00:27
2025-11-12 15:27:51.174 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 119/120, iter: 100/129, gpu mem: 2423Mb, mem: 94.9Gb, iter_time: 0.157s, data_time: 0.002s, total_loss: 5.4, iou_loss: 2.3, l1_loss: 0.9, conf_loss: 1.5, cls_loss: 0.7, lr: 5.140e-07, size: 512, ETA: 0:00:24
2025-11-12 15:27:54.453 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 119/120, iter: 120/129, gpu mem: 2423Mb, mem: 94.9Gb, iter_time: 0.159s, data_time: 0.002s, total_loss: 4.4, iou_loss: 1.8, l1_loss: 0.5, conf_loss: 1.5, cls_loss: 0.5, lr: 3.922e-07, size: 352, ETA: 0:00:21
2025-11-12 15:27:55.830 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:28:01.085 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:28:02.030 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:28:02.590 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6534
2025-11-12 15:28:02.763 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5912
2025-11-12 15:28:02.803 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4464
2025-11-12 15:28:02.803 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5637
2025-11-12 15:28:02.803 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:28:02.804 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:28:02.804 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.653
2025-11-12 15:28:02.804 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.591
2025-11-12 15:28:02.804 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.446
2025-11-12 15:28:02.804 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.564
2025-11-12 15:28:02.804 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:28:02.804 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:28:02.804 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:28:02.804 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:28:02.805 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:28:02.805 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:28:02.805 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:28:02.805 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:28:02.805 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:28:03.574 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:28:04.407 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:28:05.202 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:28:06.008 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:28:06.771 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:28:07.585 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:28:08.347 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:28:09.118 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:28:09.929 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:28:09.929 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.33
2025-11-12 15:28:09.930 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:28:09.930 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:28:09.937 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.25 ms, Average NMS time: 0.59 ms, Average inference time: 2.84 ms

2025-11-12 15:28:09.938 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:28:10.020 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:28:10.103 | INFO     | yolox_microbt.core.trainer:before_epoch:190 - ---> start train epoch120
2025-11-12 15:28:13.228 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 120/120, iter: 20/129, gpu mem: 2423Mb, mem: 95.0Gb, iter_time: 0.152s, data_time: 0.002s, total_loss: 3.9, iou_loss: 1.8, l1_loss: 0.5, conf_loss: 1.1, cls_loss: 0.6, lr: 2.447e-07, size: 352, ETA: 0:00:17
2025-11-12 15:28:16.546 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 120/120, iter: 40/129, gpu mem: 2423Mb, mem: 94.9Gb, iter_time: 0.162s, data_time: 0.002s, total_loss: 4.1, iou_loss: 1.8, l1_loss: 0.5, conf_loss: 1.2, cls_loss: 0.5, lr: 1.631e-07, size: 416, ETA: 0:00:13
2025-11-12 15:28:19.658 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 120/120, iter: 60/129, gpu mem: 2423Mb, mem: 94.9Gb, iter_time: 0.152s, data_time: 0.003s, total_loss: 4.9, iou_loss: 2.1, l1_loss: 0.7, conf_loss: 1.5, cls_loss: 0.6, lr: 9.804e-08, size: 512, ETA: 0:00:10
2025-11-12 15:28:22.889 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 120/120, iter: 80/129, gpu mem: 2423Mb, mem: 94.9Gb, iter_time: 0.157s, data_time: 0.002s, total_loss: 5.3, iou_loss: 1.9, l1_loss: 0.7, conf_loss: 2.1, cls_loss: 0.6, lr: 4.944e-08, size: 544, ETA: 0:00:07
2025-11-12 15:28:26.207 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 120/120, iter: 100/129, gpu mem: 2423Mb, mem: 94.9Gb, iter_time: 0.163s, data_time: 0.002s, total_loss: 5.7, iou_loss: 2.6, l1_loss: 0.8, conf_loss: 1.7, cls_loss: 0.7, lr: 1.732e-08, size: 352, ETA: 0:00:04
2025-11-12 15:28:29.002 | INFO     | yolox_microbt.core.trainer:after_iter:258 - epoch: 120/120, iter: 120/129, gpu mem: 2423Mb, mem: 94.9Gb, iter_time: 0.137s, data_time: 0.002s, total_loss: 4.6, iou_loss: 2.1, l1_loss: 0.6, conf_loss: 1.2, cls_loss: 0.6, lr: 1.668e-09, size: 320, ETA: 0:00:01
2025-11-12 15:28:30.272 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:28:35.289 | INFO     | yolox.evaluators.voc_evaluator:evaluate_prediction:144 - Evaluate in main process...
2025-11-12 15:28:36.227 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.50
2025-11-12 15:28:36.841 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for person = 0.6542
2025-11-12 15:28:36.979 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for car = 0.5875
2025-11-12 15:28:37.017 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:324 - AP for dog = 0.4446
2025-11-12 15:28:37.017 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:329 - Mean AP = 0.5621
2025-11-12 15:28:37.017 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:330 - ~~~~~~~~
2025-11-12 15:28:37.018 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:331 - Results:
2025-11-12 15:28:37.018 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.654
2025-11-12 15:28:37.018 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.587
2025-11-12 15:28:37.018 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:333 - 0.445
2025-11-12 15:28:37.018 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:334 - 0.562
2025-11-12 15:28:37.018 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:335 - ~~~~~~~~
2025-11-12 15:28:37.018 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:336 - 
2025-11-12 15:28:37.018 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:337 - --------------------------------------------------------------
2025-11-12 15:28:37.018 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:338 - Results computed with the **unofficial** Python eval code.
2025-11-12 15:28:37.018 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:339 - Results should be very close to the official MATLAB eval code.
2025-11-12 15:28:37.019 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:340 - Recompute with `./tools/reval.py --matlab ...` for your paper.
2025-11-12 15:28:37.019 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:341 - -- Thanks, The Management
2025-11-12 15:28:37.019 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:342 - --------------------------------------------------------------
2025-11-12 15:28:37.019 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.55
2025-11-12 15:28:37.828 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.60
2025-11-12 15:28:38.577 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.65
2025-11-12 15:28:39.333 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.70
2025-11-12 15:28:40.122 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.75
2025-11-12 15:28:40.886 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.80
2025-11-12 15:28:41.670 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.85
2025-11-12 15:28:42.417 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.90
2025-11-12 15:28:43.185 | INFO     | yolox_microbt.data.datasets.voc:_do_python_eval:304 - Eval IoU : 0.95
2025-11-12 15:28:43.988 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:252 - --------------------------------------------------------------
2025-11-12 15:28:43.988 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:253 - map_5095: 0.32
2025-11-12 15:28:43.988 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:254 - map_50: 0.56
2025-11-12 15:28:43.989 | INFO     | yolox_microbt.data.datasets.voc:evaluate_detections:255 - --------------------------------------------------------------
2025-11-12 15:28:43.995 | INFO     | yolox_microbt.core.trainer:evaluate_and_save_model:368 - 
Average forward time: 2.14 ms, Average NMS time: 0.55 ms, Average inference time: 2.68 ms

2025-11-12 15:28:43.997 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:28:44.096 | INFO     | yolox_microbt.core.trainer:save_ckpt:389 - Save weights to ./YOLOX_outputs/sa6921_vnne_3classes_voc_quant_a8w8_120e_185k_trainset_fusebn_ranger_simple_rgb
2025-11-12 15:28:44.181 | INFO     | yolox_microbt.core.trainer:after_train:172 - Training of experiment is done and the best AP is 32.89
