RuntimeError: Given groups=1, weight of size [128, 128, 3, 3], expected input[32, 256, 20, 20] to have 128 channels, but got 256 channels instead

[sunmingqi-1]

(YOLO_Mamba) [pc_stu2@localhost FER-YOLO-Mamba-main]$ python train.py No module named 'selective_scan_cuda_core' initialize network with normal type Load weights model_data/yolox_s.pth.

Successful Load Key: ['backbone.backbone.stem.conv.conv.weight', 'backbone.backbone.stem.conv.bn.weight', 'backbone.backbone.stem.conv.bn.bias', 'backbone.backbone.stem.conv.bn.running_mean', 'backbone.backbone.stem.conv.bn.running_var', 'backbone.backbone.stem.conv.bn.num_batches_tracked', 'backbone.backbone.dark2.0.conv.weight', 'backbone.backbone.dark2.0.bn.weight', 'backbone.backbone.dark2.0.bn.bias', 'backbone.backbone.dark2.0.bn.running_mean', 'backbone.backbone.dark2.0.bn.running_var', 'backbone.backbone.dark …… Successful Load Key Num: 334

Fail To Load Key: ['backbone.C3_p4.conv1.conv.weight', 'backbone.C3_p4.conv1.bn.weight', 'backbone.C3_p4.conv1.bn.bias', 'backbone.C3_p4.conv1.bn.running_mean', 'backbone.C3_p4.conv1.bn.running_var', 'backbone.C3_p4.conv1.bn.num_batches_tracked', 'backbone.C3_p4.conv2.conv.weight', 'backbone.C3_p4.conv2.bn.weight', 'backbone.C3_p4.conv2.bn.bias', 'backbone.C3_p4.conv2.bn.running_mean', 'backbone.C3_p4.conv2.bn.running_var', 'backbone.C3_p4.conv2.bn.num_batches_tracked', 'backbone.C3_p4.conv3.conv.weight', 'backbo …… Fail To Load Key num: 128 Expected u.is_cuda() to be true, but got false. (Could this error message be improved? If so, please report an enhancement request to PyTorch.) Error occurs, No graph saved Configurations:

| keys | values|

| classes_path | model_data/sfew_classes.txt| | model_path | model_data/yolox_s.pth| | input_shape | [320, 320]| | Init_Epoch | 0| | Freeze_Epoch | 100| | UnFreeze_Epoch | 300| | Freeze_batch_size | 32| | Unfreeze_batch_size | 16| | Freeze_Train | True| | Init_lr | 0.01| | Min_lr | 0.0001| | optimizer_type | sgd| | momentum | 0.937| | lr_decay_type | cos| | save_period | 10| | save_dir | logs| | num_workers | 0| | num_train | 1125| | num_val | 126|

Start Train Epoch 1/300: 0%| | 0/35 [00:00<?, ?it/s<class 'dict'>]Traceback (most recent call last): File "/home/pc_stu2/Projects/FER-YOLO-Mamba-main/train.py", line 236, in fit_one_epoch(model_train, model, ema, yolo_loss, loss_history, eval_callback, optimizer, epoch, epoch_step, epoch_step_val, gen, gen_val, UnFreeze_Epoch, Cuda, fp16, scaler, save_period, save_dir, local_rank) File "/home/pc_stu2/Projects/FER-YOLO-Mamba-main/utils/utils_fit.py", line 34, in fit_one_epoch outputs = model_train(images) ^^^^^^^^^^^^^^^^^^^ File "/home/pc_stu2/anaconda3/envs/YOLO_Mamba/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl return forward_call(*args, kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/pc_stu2/Projects/FER-YOLO-Mamba-main/nets/yolo.py", line 996, in forward outputs = self.head.forward(fpn_outs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/pc_stu2/Projects/FER-YOLO-Mamba-main/nets/yolo.py", line 892, in forward cls_feat = self.cls_convsk ^^^^^^^^^^^^^^^^^^^^ File "/home/pc_stu2/anaconda3/envs/YOLO_Mamba/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl return forward_call(*args, *kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/pc_stu2/anaconda3/envs/YOLO_Mamba/lib/python3.11/site-packages/torch/nn/modules/container.py", line 217, in forward input = module(input) ^^^^^^^^^^^^^ File "/home/pc_stu2/anaconda3/envs/YOLO_Mamba/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl return forward_call(args, kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/pc_stu2/Projects/FER-YOLO-Mamba-main/nets/darknet.py", line 46, in forward return self.act(self.bn(self.conv(x))) ^^^^^^^^^^^^ File "/home/pc_stu2/anaconda3/envs/YOLO_Mamba/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl return forward_call(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/pc_stu2/anaconda3/envs/YOLO_Mamba/lib/python3.11/site-packages/torch/nn/modules/conv.py", line 463, in forward return self._conv_forward(input, self.weight, self.bias) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/pc_stu2/anaconda3/envs/YOLO_Mamba/lib/python3.11/site-packages/torch/nn/modules/conv.py", line 459, in _conv_forward return F.conv2d(input, weight, bias, self.stride, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ RuntimeError: Given groups=1, weight of size [128, 128, 3, 3], expected input[32, 256, 20, 20] to have 128 channels, but got 256 channels instead Epoch 1/300: 0%|

[hafsa390]

@sunmingqi-1 , did you solve it?

[sunmingqi-1]

@hafsa390 是的，我已经解决了这个问题。你只需要修改一下YOLOHead 的网络结构。在其 forward中的for 循环中，添加 x=self.stemsk 即可。