Value Error: Optimizing a parameter that doesn't require gradients

[chengdazhi]

Value Error: Optimizing a parameter that doesn't require gradients

[hufengshuo07]

Ｉsolve this problem by adding a filter:
optimizer = torch.optim.SGD( filter(lambda p: p.requires_grad, model.parameters())
but a new problem appeared. Could you give some help?

Create new model
############# Starting Epoch 0 #############
loss: 0.01542340 | acc: 1.54: 0%| | 1/17291 [00:01<9:12:33, 1.92s/it]Traceback (most recent call last):
File "train.py", line 215, in
main()
File "train.py", line 173, in main
loss, acc = train(train_loader, m, criterion, optimizer, writer)
File "train.py", line 31, in train
out = m(inps)
File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/module.py", line 491, in call
result = self.forward(*input, **kwargs)
File "/usr/local/lib/python3.5/dist-packages/torch/nn/parallel/data_parallel.py", line 112, in forward
return self.module(*inputs[0], **kwargs[0])
File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/module.py", line 491, in call
result = self.forward(*input, **kwargs)
File "/media/sure/H/train_sppe/src/models/DCN_FastPose.py", line 41, in forward
out = self.preact(x)
File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/module.py", line 491, in call
result = self.forward(*input, **kwargs)
File "/media/sure/H/train_sppe/src/models/layers/DCN_SE_Resnet.py", line 84, in forward
x = self.layer3(x) # 1024 * h/16 * w/16
File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/module.py", line 491, in call
result = self.forward(*input, **kwargs)
File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/container.py", line 91, in forward
input = module(input)
File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/module.py", line 491, in call
result = self.forward(*input, **kwargs)
File "/media/sure/H/train_sppe/src/models/layers/DCN_SE_Resnet.py", line 39, in forward
out = self.conv3(out)
File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/module.py", line 491, in call
result = self.forward(*input, **kwargs)
File "/media/sure/H/train_sppe/src/DCNv2/modules/modulated_dcn.py", line 73, in forward
return func(input, offset, mask, self.weight, self.bias)
File "/media/sure/H/train_sppe/src/DCNv2/functions/modulated_dcn_func.py", line 36, in forward
self.deformable_groups)
File "/usr/local/lib/python3.5/dist-packages/torch/utils/ffi/init.py", line 197, in safe_call
result = torch._C._safe_call(*args, **kwargs)
torch.FatalError: cublas runtime error : an internal operation failed at /pytorch/aten/src/THC/THCBlas.cu:249