Class FeedForward

src/diffusers/models/attention.py:1682–1742

r""" A feed-forward layer. Parameters: dim (`int`): The number of channels in the input. dim_out (`int`, *optional*): The number of channels in the output. If not given, defaults to `dim`. mult (`int`, *optional*, defaults to 4): The multiplier to use for the hidden

class FeedForward(nn.Module):
    r"""
    A feed-forward layer.

    Parameters:
        dim (`int`): The number of channels in the input.
        dim_out (`int`, *optional*): The number of channels in the output. If not given, defaults to `dim`.
        mult (`int`, *optional*, defaults to 4): The multiplier to use for the hidden dimension.
        dropout (`float`, *optional*, defaults to 0.0): The dropout probability to use.
        activation_fn (`str`, *optional*, defaults to `"geglu"`): Activation function to be used in feed-forward.
        final_dropout (`bool` *optional*, defaults to False): Apply a final dropout.
        bias (`bool`, defaults to True): Whether to use a bias in the linear layer.
    """

    def __init__(
        self,
        dim: int,
        dim_out: int | None = None,
        mult: int = 4,
        dropout: float = 0.0,
        activation_fn: str = "geglu",
        final_dropout: bool = False,
        inner_dim=None,
        bias: bool = True,
    ):
        super().__init__()
        if inner_dim is None:
            inner_dim = int(dim * mult)
        dim_out = dim_out if dim_out is not None else dim

        if activation_fn == "gelu":
            act_fn = GELU(dim, inner_dim, bias=bias)
        if activation_fn == "gelu-approximate":
            act_fn = GELU(dim, inner_dim, approximate="tanh", bias=bias)
        elif activation_fn == "geglu":
            act_fn = GEGLU(dim, inner_dim, bias=bias)
        elif activation_fn == "geglu-approximate":
            act_fn = ApproximateGELU(dim, inner_dim, bias=bias)
        elif activation_fn == "swiglu":
            act_fn = SwiGLU(dim, inner_dim, bias=bias)
        elif activation_fn == "linear-silu":
            act_fn = LinearActivation(dim, inner_dim, bias=bias, activation="silu")

        self.net = nn.ModuleList([])
        # project in
        self.net.append(act_fn)
        # project dropout
        self.net.append(nn.Dropout(dropout))
        # project out
        self.net.append(nn.Linear(inner_dim, dim_out, bias=bias))
        # FF as used in Vision Transformer, MLP-Mixer, etc. have a final dropout
        if final_dropout:
            self.net.append(nn.Dropout(dropout))

    def forward(self, hidden_states: torch.Tensor, *args, **kwargs) -> torch.Tensor:
        if len(args) > 0 or kwargs.get("scale", None) is not None:
            deprecation_message = "The `scale` argument is deprecated and will be ignored. Please remove it, as passing it will raise an error in the future. `scale` should directly be passed while calling the underlying pipeline component i.e., via `cross_attention_kwargs`."
            deprecate("scale", "1.0.0", deprecation_message)