Method forward

torch/nn/modules/transformer.py:594–717

r"""Pass the input through the encoder layer. Args: src: the sequence to the encoder layer (required). src_mask: the mask for the src sequence (optional). src_key_padding_mask: the mask for the src keys per batch (optional). is_causal: If spec

(
            self,
            src: Tensor,
            src_mask: Optional[Tensor] = None,
            src_key_padding_mask: Optional[Tensor] = None,
            is_causal: bool = False)

    def forward(
            self,
            src: Tensor,
            src_mask: Optional[Tensor] = None,
            src_key_padding_mask: Optional[Tensor] = None,
            is_causal: bool = False) -> Tensor:
        r"""Pass the input through the encoder layer.

        Args:
            src: the sequence to the encoder layer (required).
            src_mask: the mask for the src sequence (optional).
            src_key_padding_mask: the mask for the src keys per batch (optional).
            is_causal: If specified, applies a causal mask as ``src mask``.
                Default: ``False``.
                Warning:
                ``is_causal`` provides a hint that ``src_mask`` is the
                causal mask. Providing incorrect hints can result in
                incorrect execution, including forward and backward
                compatibility.

        Shape:
            see the docs in Transformer class.
        """
        src_key_padding_mask = F._canonical_mask(
            mask=src_key_padding_mask,
            mask_name="src_key_padding_mask",
            other_type=F._none_or_dtype(src_mask),
            other_name="src_mask",
            target_type=src.dtype
        )

        src_mask = F._canonical_mask(
            mask=src_mask,
            mask_name="src_mask",
            other_type=None,
            other_name="",
            target_type=src.dtype,
            check_other=False,
        )

        # see Fig. 1 of https://arxiv.org/pdf/2002.04745v1.pdf
        why_not_sparsity_fast_path = ''
        if not src.dim() == 3:
            why_not_sparsity_fast_path = f"input not batched; expected src.dim() of 3 but got {src.dim()}"
        elif self.training:
            why_not_sparsity_fast_path = "training is enabled"
        elif not self.self_attn.batch_first :
            why_not_sparsity_fast_path = "self_attn.batch_first was not True"
        elif not self.self_attn._qkv_same_embed_dim :
            why_not_sparsity_fast_path = "self_attn._qkv_same_embed_dim was not True"
        elif not self.activation_relu_or_gelu:
            why_not_sparsity_fast_path = "activation_relu_or_gelu was not True"
        elif not (self.norm1.eps == self.norm2.eps):
            why_not_sparsity_fast_path = "norm1.eps is not equal to norm2.eps"
        elif src.is_nested and (src_key_padding_mask is not None or src_mask is not None):
            why_not_sparsity_fast_path = "neither src_key_padding_mask nor src_mask are not supported with NestedTensor input"
        elif self.self_attn.num_heads % 2 == 1:
            why_not_sparsity_fast_path = "num_head is odd"