flux/model.py

from dataclasses import dataclass
from typing import List

import torch
from torch import Tensor, nn

from flux.modules.layers import (
    DoubleStreamBlock,
    EmbedND,
    LastLayer,
    MLPEmbedder,
    SingleStreamBlock,
    timestep_embedding,
)
from flux.modules.lora import LinearLora, replace_linear_with_lora


@dataclass
class FluxParams:
    in_channels: int
    out_channels: int
    vec_in_dim: int
    context_in_dim: int
    hidden_size: int
    mlp_ratio: float
    num_heads: int
    depth: int
    depth_single_blocks: int
    axes_dim: list[int]
    theta: int
    qkv_bias: bool
    guidance_embed: bool


class Flux(nn.Module):
    """
    Transformer model for flow matching on sequences.
    """

    def __init__(self, params: FluxParams):
        super().__init__()

        self.params = params
        self.in_channels = params.in_channels
        self.out_channels = params.out_channels
        if params.hidden_size % params.num_heads != 0:
            raise ValueError(
                f"Hidden size {params.hidden_size} must be divisible by num_heads {params.num_heads}"
            )
        pe_dim = params.hidden_size // params.num_heads
        if sum(params.axes_dim) != pe_dim:
            raise ValueError(f"Got {params.axes_dim} but expected positional dim {pe_dim}")
        self.hidden_size = params.hidden_size
        self.num_heads = params.num_heads
        self.pe_embedder = EmbedND(dim=pe_dim, theta=params.theta, axes_dim=params.axes_dim)
        self.img_in = nn.Linear(self.in_channels, self.hidden_size, bias=True)
        self.time_in = MLPEmbedder(in_dim=256, hidden_dim=self.hidden_size)
        self.vector_in = MLPEmbedder(params.vec_in_dim, self.hidden_size)
        self.guidance_in = (
            MLPEmbedder(in_dim=256, hidden_dim=self.hidden_size) if params.guidance_embed else nn.Identity()
        )
        self.txt_in = nn.Linear(params.context_in_dim, self.hidden_size)

        self.double_blocks = nn.ModuleList(
            [
                DoubleStreamBlock(
                    self.hidden_size,
                    self.num_heads,
                    mlp_ratio=params.mlp_ratio,
                    qkv_bias=params.qkv_bias,
                )
                for _ in range(params.depth)
            ]
        )

        self.single_blocks = nn.ModuleList(
            [
                SingleStreamBlock(self.hidden_size, self.num_heads, mlp_ratio=params.mlp_ratio)
                for _ in range(params.depth_single_blocks)
            ]
        )

        self.final_layer = LastLayer(self.hidden_size, 1, self.out_channels)

    def forward(
        self,
        img: Tensor,
        img_ids: Tensor,
        txt: Tensor,
        txt_ids: Tensor,
        timesteps: Tensor,
        y: Tensor,
        txt_mask: Tensor = None,
        img_mask: Tensor = None,
        guidance: Tensor | None = None,
    ) -> Tensor:
        if img.ndim != 3 or txt.ndim != 3:
            raise ValueError("Input img and txt tensors must have 3 dimensions.")

        # running on sequences img
        img = self.img_in(img)
        vec = self.time_in(timestep_embedding(timesteps, 256))
        if self.params.guidance_embed:
            if guidance is None:
                raise ValueError("Didn't get guidance strength for guidance distilled model.")
            vec = vec + self.guidance_in(timestep_embedding(guidance, 256))
        vec = vec + self.vector_in(y)
        txt = self.txt_in(txt)

        ids = torch.cat((txt_ids, img_ids), dim=1)
        pe = self.pe_embedder(ids)

        for block in self.double_blocks:
            img, txt = block(img=img, txt=txt, vec=vec, pe=pe, img_mask=img_mask, txt_mask=txt_mask)

        img = torch.cat((txt, img), 1)
        attn_mask = torch.cat((txt_mask, img_mask), 1)
        for block in self.single_blocks:
            img = block(img, vec=vec, pe=pe, attn_mask=attn_mask)
        img = img[:, txt.shape[1] :, ...]

        img = self.final_layer(img, vec)  # (N, T, patch_size ** 2 * out_channels)
        # print(f'flux out {img.shape} {img.mean()}')
        return img

    def forward_with_cfg(
        self,
        img: Tensor,
        img_ids: Tensor,
        txt: Tensor,
        txt_ids: Tensor,
        timesteps: Tensor,
        y: Tensor,
        txt_mask: Tensor = None,
        img_mask: Tensor = None,
        guidance: Tensor | None = None,
        cfg_scale: float = 1.0,
    ) -> Tensor:
        half = img[: len(img) // 2]
        combined = torch.cat([half, half], dim=0)
        model_out = self.forward(img, img_ids, txt, txt_ids, timesteps, y, txt_mask, img_mask, guidance)
        cond_v, uncond_v = torch.split(model_out, len(model_out) // 2, dim=0)
        cond_v = uncond_v + cfg_scale * (cond_v - uncond_v)
        img = torch.cat([cond_v, uncond_v], dim=0)
        return img

    def get_fsdp_wrap_module_list(self) -> List[nn.Module]:
        return list(self.double_blocks) + list(self.single_blocks) + [self.final_layer] + [self.img_in, self.vector_in, self.guidance_in, self.txt_in, self.time_in]

    def get_checkpointing_wrap_module_list(self) -> List[nn.Module]:
        return list(self.double_blocks) + list(self.single_blocks) + [self.final_layer] + [self.img_in, self.vector_in, self.guidance_in, self.txt_in, self.time_in]


class FluxLoraWrapper(Flux):
    def __init__(
        self,
        lora_rank: int = 128,
        lora_scale: float = 1.0,
        *args,
        **kwargs,
    ) -> None:
        super().__init__(*args, **kwargs)

        self.lora_rank = lora_rank

        replace_linear_with_lora(
            self,
            max_rank=lora_rank,
            scale=lora_scale,
        )

    def set_lora_scale(self, scale: float) -> None:
        for module in self.modules():
            if isinstance(module, LinearLora):
                module.set_scale(scale=scale)