Excpetions when trying to use quartznet10x5 and quartznet15x5
farisalasmary opened this issue · comments
Dear @ivankunyankin ,
I've run the code on quartznet5x5 and it worked! when I switched to use quartznet10x5 or quartznet15x5, I got the following exception:
Traceback (most recent call last):
File "train.py", line 392, in <module>
trainer.train()
File "train.py", line 113, in train
self.train_step(epoch)
File "train.py", line 163, in train_step
output = self.model(specs)
File "/opt/conda/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1056, in _call_impl
return forward_call(*input, **kwargs)
File "/workspace/quartznet-asr/models.py", line 42, in forward
x = block(x)
File "/opt/conda/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1056, in _call_impl
return forward_call(*input, **kwargs)
File "/workspace/quartznet-asr/models.py", line 77, in forward
y = self.res(x)
File "/opt/conda/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1056, in _call_impl
return forward_call(*input, **kwargs)
File "/opt/conda/lib/python3.8/site-packages/torch/nn/modules/container.py", line 139, in forward
input = module(input)
File "/opt/conda/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1056, in _call_impl
return forward_call(*input, **kwargs)
File "/opt/conda/lib/python3.8/site-packages/torch/nn/modules/conv.py", line 301, in forward
return self._conv_forward(input, self.weight, self.bias)
File "/opt/conda/lib/python3.8/site-packages/torch/nn/modules/conv.py", line 297, in _conv_forward
return F.conv1d(input, weight, bias, self.stride,
RuntimeError: Given groups=1, weight of size [512, 256, 1], expected input[64, 512, 101] to have 256 channels, but got 512 channels instead
Any idea?
@farisalasmary Hi! I'm truly sorry for the late response. I created another branch with the fix of the problem. Could you verify that it works for you now?
Hi @ivankunyankin,
Thank you for your reply!! I've tried it and it worked!!
Thank you again for your time