Module transformer 

Transformer module

Structs

PositionWiseFeedForward

Applies the position-wise feed-forward network to the input tensor from the paper Attention Is All You Need.

PositionWiseFeedForwardConfig

Configuration to create a position-wise feed-forward layer using the init function.

TransformerDecoder

The transformer decoder module as describe in the paper Attention Is All You Need.

TransformerDecoderAutoregressiveCache

Autoregressive cache for the Transformer Decoder layer.

TransformerDecoderConfig

Configuration to create a Transformer Decoder layer using the init function.

TransformerDecoderInput

Transformer Decoder forward pass input argument.

TransformerDecoderLayer

Transformer Decoder layer module.

TransformerDecoderLayerAutoregressiveCache

Autoregressive cache for a single Transformer Decoder Layer.

TransformerEncoder

The transformer encoder module as describe in the paper Attention Is All You Need.

TransformerEncoderAutoregressiveCache

Autoregressive cache for the Transformer Encoder layer.

TransformerEncoderConfig

Configuration to create a Transformer Encoder layer using the init function.

TransformerEncoderInput

Transformer Encoder forward pass input argument.

TransformerEncoderLayer

Transformer encoder layer module.

TransformerEncoderLayerAutoregressiveCache

Autoregressive cache for a single Transformer Encoder Layer.