ycszen / pytorch-segmentation

Pytorch for Segmentation

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

weight init

skrish13 opened this issue · comments

For resnet and other architectures in general, the weight init has to be done separately right? The default init for nn modules of pytorch is uniform? From my understanding, there are better inits like glorot uniform, he init etc which we explicitly need to do. Because other libraries like Keras, Lasagne has the best init as the default, unlike PyTorch.

[1] - ResNet official: https://github.com/pytorch/vision/blob/master/torchvision/models/resnet.py#L113-L119
[2] - VGG official: https://github.com/pytorch/vision/blob/master/torchvision/models/vgg.py#L46-L59

[3] - Conv2D nn module: https://github.com/pytorch/pytorch/blob/master/torch/nn/modules/conv.py#L40-L47

There is special weight init in pytorch. You can find it in the document. In this repo, I don't care the init because I load the official weight.

Could you explain what you mean by 'official weight'. (Since there are other modules than just pre-trained layers in your code)

Just the model weights Pytorch offer.