ComfyUI-WanVideoWrapper/LongCat/layers.py

import torch.nn as nn
import torch.nn.functional as F
import torch
import math

class FeedForwardSwiGLU(nn.Module):
    def __init__(
        self,
        dim: int,
        hidden_dim: int,
        multiple_of: int = 256,
    ):
        super().__init__()
        hidden_dim = int(2 * hidden_dim / 3)
        hidden_dim = multiple_of * ((hidden_dim + multiple_of - 1) // multiple_of)

        self.dim = dim
        self.hidden_dim = hidden_dim
        self.w1 = nn.Linear(dim, hidden_dim, bias=False)
        self.w2 = nn.Linear(hidden_dim, dim, bias=False)
        self.w3 = nn.Linear(dim, hidden_dim, bias=False)

    def forward(self, x):
        return self.w2(F.silu(self.w1(x)) * self.w3(x))
    
class TimestepEmbedder(nn.Module):
    """
    Embeds scalar timesteps into vector representations.
    """

    def __init__(self, t_embed_dim, frequency_embedding_size=256):
        super().__init__()
        self.t_embed_dim = t_embed_dim
        self.frequency_embedding_size = frequency_embedding_size
        self.mlp = nn.Sequential(
            nn.Linear(frequency_embedding_size, t_embed_dim, bias=True),
            nn.SiLU(),
            nn.Linear(t_embed_dim, t_embed_dim, bias=True),
        )

    @staticmethod
    def timestep_embedding(t, dim, max_period=10000):
        """
        Create sinusoidal timestep embeddings.
        :param t: a 1-D Tensor of N indices, one per batch element.
                          These may be fractional.
        :param dim: the dimension of the output.
        :param max_period: controls the minimum frequency of the embeddings.
        :return: an (N, D) Tensor of positional embeddings.
        """
        half = dim // 2
        freqs = torch.exp(-math.log(max_period) * torch.arange(start=0, end=half, dtype=torch.float32) / half)
        freqs = freqs.to(device=t.device)
        args = t[:, None].float() * freqs[None]
        embedding = torch.cat([torch.cos(args), torch.sin(args)], dim=-1)
        if dim % 2:
            embedding = torch.cat([embedding, torch.zeros_like(embedding[:, :1])], dim=-1)
        return embedding

    def forward(self, t, dtype):
        t_freq = self.timestep_embedding(t, self.frequency_embedding_size)
        if t_freq.dtype != dtype:
            t_freq = t_freq.to(dtype)
        t_emb = self.mlp(t_freq)
        return t_emb