Attention-Based-Models
Notes about some attention-based deep learning models
seq2seq_translation_tutorial_from_pytorch.ipynb consists of concrete notes of the shape of the tensor which in the Encoder, the Decoder and the training function, which helps comprehend the data flow of the model.
transformer_tutorial_from_pytorch.ipynb consists of concrete notes of the shape of the tensor which in the Transformer, which helps comprehend the data flow of the model. However, the model doesn't use TransformerDecoder.