Transformer

完成进度

• Embeddings and PositionalEncoding with example.

• MultiHeadAttention with Add & Norm.

• PositionwiseFeedForward with Add & Norm.

• EncoderLayer and DecoderLayer.

• Encoder and Decoder. TODO: vocab_size is undefined.

• NEXT: Generator.

• NEXT: EncoderDecoder.

• NEXT: Data.

GitHub

View Github