Yolact(
  (backbone): ResNetBackbone(
    (layers): ModuleList(
      (0): Sequential(
        (0): Bottleneck(
          (conv1): MLUConv2d(
            64, 64, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            64, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
          (downsample): Sequential(
            (0): MLUConv2d(
              64, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
              (observer): MLUMinMaxObserver(min_val=None, max_val=None)
            )
            (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): Bottleneck(
          (conv1): MLUConv2d(
            256, 64, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            64, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (2): Bottleneck(
          (conv1): MLUConv2d(
            256, 64, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            64, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
      )
      (1): Sequential(
        (0): Bottleneck(
          (conv1): MLUConv2d(
            256, 128, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            128, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            128, 512, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
          (downsample): Sequential(
            (0): MLUConv2d(
              256, 512, kernel_size=(1, 1), stride=(2, 2), scale=None, quantized_mode=None, input_mean=None, input_std=None
              (observer): MLUMinMaxObserver(min_val=None, max_val=None)
            )
            (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): Bottleneck(
          (conv1): MLUConv2d(
            512, 128, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            128, 512, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (2): Bottleneck(
          (conv1): MLUConv2d(
            512, 128, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            128, 512, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (3): Bottleneck(
          (conv1): MLUConv2d(
            512, 128, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            128, 512, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
      )
      (2): Sequential(
        (0): Bottleneck(
          (conv1): MLUConv2d(
            512, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
          (downsample): Sequential(
            (0): MLUConv2d(
              512, 1024, kernel_size=(1, 1), stride=(2, 2), scale=None, quantized_mode=None, input_mean=None, input_std=None
              (observer): MLUMinMaxObserver(min_val=None, max_val=None)
            )
            (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (2): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (3): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (4): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (5): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (6): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (7): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (8): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (9): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (10): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (11): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (12): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (13): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (14): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (15): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (16): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (17): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (18): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (19): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (20): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (21): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (22): Bottleneck(
          (conv1): MLUConv2d(
            1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            256, 1024, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
      )
      (3): Sequential(
        (0): Bottleneck(
          (conv1): MLUConv2d(
            1024, 512, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            512, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            512, 2048, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
          (downsample): Sequential(
            (0): MLUConv2d(
              1024, 2048, kernel_size=(1, 1), stride=(2, 2), scale=None, quantized_mode=None, input_mean=None, input_std=None
              (observer): MLUMinMaxObserver(min_val=None, max_val=None)
            )
            (1): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): Bottleneck(
          (conv1): MLUConv2d(
            2048, 512, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            512, 2048, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
        (2): Bottleneck(
          (conv1): MLUConv2d(
            2048, 512, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv2): MLUConv2d(
            512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (conv3): MLUConv2d(
            512, 2048, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
            (observer): MLUMinMaxObserver(min_val=None, max_val=None)
          )
          (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (relu): ReLU(inplace=True)
        )
      )
    )
    (conv1): MLUConv2d(
      3, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), scale=None, quantized_mode=None, input_mean=None, input_std=None
      (observer): MLUMinMaxObserver(min_val=None, max_val=None)
    )
    (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
    (relu): ReLU(inplace=True)
    (maxpool): MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False)
  )
  (proto_net): Sequential(
    (0): MLUConv2d(
      256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
      (observer): MLUMinMaxObserver(min_val=None, max_val=None)
    )
    (1): ReLU(inplace=True)
    (2): MLUConv2d(
      256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
      (observer): MLUMinMaxObserver(min_val=None, max_val=None)
    )
    (3): ReLU(inplace=True)
    (4): MLUConv2d(
      256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
      (observer): MLUMinMaxObserver(min_val=None, max_val=None)
    )
    (5): ReLU(inplace=True)
    (6): InterpolateModule()
    (7): ReLU(inplace=True)
    (8): MLUConv2d(
      256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
      (observer): MLUMinMaxObserver(min_val=None, max_val=None)
    )
    (9): ReLU(inplace=True)
    (10): MLUConv2d(
      256, 32, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
      (observer): MLUMinMaxObserver(min_val=None, max_val=None)
    )
  )
  (fpn): FPN(
    (lat_layers): ModuleList(
      (0): MLUConv2d(
        2048, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
        (observer): MLUMinMaxObserver(min_val=None, max_val=None)
      )
      (1): MLUConv2d(
        1024, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
        (observer): MLUMinMaxObserver(min_val=None, max_val=None)
      )
      (2): MLUConv2d(
        512, 256, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
        (observer): MLUMinMaxObserver(min_val=None, max_val=None)
      )
    )
    (pred_layers): ModuleList(
      (0): MLUConv2d(
        256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
        (observer): MLUMinMaxObserver(min_val=None, max_val=None)
      )
      (1): MLUConv2d(
        256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
        (observer): MLUMinMaxObserver(min_val=None, max_val=None)
      )
      (2): MLUConv2d(
        256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
        (observer): MLUMinMaxObserver(min_val=None, max_val=None)
      )
    )
    (downsample_layers): ModuleList(
      (0): MLUConv2d(
        256, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
        (observer): MLUMinMaxObserver(min_val=None, max_val=None)
      )
      (1): MLUConv2d(
        256, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
        (observer): MLUMinMaxObserver(min_val=None, max_val=None)
      )
    )
  )
  (prediction_layers): ModuleList(
    (0): PredictionModule(
      (upfeature): Sequential(
        (0): MLUConv2d(
          256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
          (observer): MLUMinMaxObserver(min_val=None, max_val=None)
        )
        (1): ReLU(inplace=True)
      )
      (bbox_layer): MLUConv2d(
        256, 12, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
        (observer): MLUMinMaxObserver(min_val=None, max_val=None)
      )
      (conf_layer): MLUConv2d(
        256, 15, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
        (observer): MLUMinMaxObserver(min_val=None, max_val=None)
      )
      (mask_layer): MLUConv2d(
        256, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
        (observer): MLUMinMaxObserver(min_val=None, max_val=None)
      )
    )
    (1): PredictionModule()
    (2): PredictionModule()
    (3): PredictionModule()
    (4): PredictionModule()
  )
  (semantic_seg_conv): MLUConv2d(
    256, 4, kernel_size=(1, 1), stride=(1, 1), scale=None, quantized_mode=None, input_mean=None, input_std=None
    (observer): MLUMinMaxObserver(min_val=None, max_val=None)
  )
)