MasteringPyTorchV2

MasteringPyTorchV2 copied to clipboard

The provided implementation of GoogLeNet contains several mistakes: - Wrong usage of undefined layers like a4, b4, etc. instead of the defined im3, im4, etc. - Incorrect data flow: output...

gangula-karthik

chapter 5 transformer.ipynb not able to reproduce

2

Was reproducing the code on Colab but got the following error : ->tr_iter = PennTreebank(split='train') tkzer = get_tokenizer('basic_english')  done pip install portalocker then got the following error:  Is...

ayusdixit

Numpy version conflict

1

We should update codebase wrt to numpy version 2.0 or limit the numpy version to 1.xx to avoid compatibility issues.

rnb007

Question about Skip connection block in ResNet

1

Hello, what is the `strd` and `multiplier` do here? what are the use for? ``` class BasicBlock(nn.Module): multiplier=1 def __init__(self, input_num_planes, num_planes, strd=1): super(BasicBlock, self).__init__() self.conv_layer1 = nn.Conv2d(in_channels=input_num_planes, out_channels=num_planes, kernel_size=3,...

m7mdhka