In this repository, I am gathering codes and toturials that use Pytorch to implement different neual networks. In addition to learning Torch, I emphasize on learning the Python implementation and using the library itself + mathematics of ML.
Most of the codes in this Git repository are from the YouTube channel by Aladdin Persson. Make sure to visit his channle and see his PyTorch playlist. It is simply wonderful.
Here is the link to Aladdin's YT channel: Link
Also codes are descriptive by their names, but here is a brief explaination of each code:
01 tensor initialization: Initializing different tensors and converting them to other data types.02 tensor math comparision: Addition, subtraction, division, inplace operations, matrix multiplication, broadcasting, clamping, etc.03 tensor indexing: Advanced indexing04 tensor reshaping: View, reshape, permute, transpose05 pytorch simple fullynet: Simple implementation of a fully connected NN: creating a network, seting device, hyperpeter setting, loading data, initializing network, loss and optimizer, train network, getting accuracy and testing model06 pytorch simple CNN: Simple implementation of a CNN.07 NN LSTM GRU: Simple implementation of a LSTM and a GRU.08 Bidrection LSTM: Simple implementation of a BiLSTM.09 Save Load: Saving and loading a CNN model.10 Transfer Learning: