The code for this repo is private.
Problem statement: Generate translation from one language to another.
The code include:
- Implementing the encoder part for the transformer based on the paper "Attenion is all you need".
- Encoder consists of an embedding layer, self attention layer, linear layer and the softmax layer to compute the output probabilities. The implementation is done from scratch.
- Next is implementing the the full tranformer encoder-decoder layer using Pytorch's buil-in Transformer function.