PyTorch torch.nn library and the difference between nn and nn.functional

views

Tutorial explaining PyTorch’s torch.nn library and the differences between nn (module classes) and nn.functional (functional interfaces). Covers Parameters and ParameterList/ParameterDict, common layers (Conv2d, Linear, ReLU, MaxPool), containers (Sequential, ModuleList, ModuleDict), differences between parameters and buffers, dropout behavior in train/eval, and state_dict/load_state_dict usage.

Download / Open (k2s.cc)

Leave a Reply

Your email address will not be published. Required fields are marked *