Skip to content

SUPRA-Tex: 3D Generation with Super-Aligned Texture

38.9 KB
Raw Blame
import inspect
from typing import Any, Dict, List, Optional, Tuple, Union

import torch, math
import torch.nn as nn
import torch.nn.functional as F
import numpy as np
from ..core.attention import attention_forward
from ..core.gradient import gradient_checkpoint_forward


def get_timestep_embedding(
    timesteps: torch.Tensor,
    embedding_dim: int,
    flip_sin_to_cos: bool = False,
    downscale_freq_shift: float = 1,
    scale: float = 1,
    max_period: int = 10000,
) -> torch.Tensor:
    """
    This matches the implementation in Denoising Diffusion Probabilistic Models: Create sinusoidal timestep embeddings.

    Args
        timesteps (torch.Tensor):
            a 1-D Tensor of N indices, one per batch element. These may be fractional.
        embedding_dim (int):
            the dimension of the output.
        flip_sin_to_cos (bool):
            Whether the embedding order should be `cos, sin` (if True) or `sin, cos` (if False)
        downscale_freq_shift (float):
            Controls the delta between frequencies between dimensions
        scale (float):
            Scaling factor applied to the embeddings.
        max_period (int):
            Controls the maximum frequency of the embeddings
    Returns
        torch.Tensor: an [N x dim] Tensor of positional embeddings.
    """
    assert len(timesteps.shape) == 1, "Timesteps should be a 1d-array"

    half_dim = embedding_dim // 2
    exponent = -math.log(max_period) * torch.arange(
        start=0, end=half_dim, dtype=torch.float32, device=timesteps.device
    )
    exponent = exponent / (half_dim - downscale_freq_shift)

    emb = torch.exp(exponent)
    emb = timesteps[:, None].float() * emb[None, :]

    # scale embeddings
    emb = scale * emb

    # concat sine and cosine embeddings
    emb = torch.cat([torch.sin(emb), torch.cos(emb)], dim=-1)

    # flip sine and cosine embeddings
    if flip_sin_to_cos:
        emb = torch.cat([emb[:, half_dim:], emb[:, :half_dim]], dim=-1)

    # zero pad
    if embedding_dim % 2 == 1:
        emb = torch.nn.functional.pad(emb, (0, 1, 0, 0))
    return emb


class TimestepEmbedding(nn.Module):
    def __init__(
        self,
        in_channels: int,
        time_embed_dim: int,
        act_fn: str = "silu",
        out_dim: int = None,
        post_act_fn: Optional[str] = None,
        cond_proj_dim=None,
        sample_proj_bias=True,
    ):
        super().__init__()

        self.linear_1 = nn.Linear(in_channels, time_embed_dim, sample_proj_bias)

        if cond_proj_dim is not None:
            self.cond_proj = nn.Linear(cond_proj_dim, in_channels, bias=False)
        else:
            self.cond_proj = None

        self.act = torch.nn.SiLU()

        if out_dim is not None:
            time_embed_dim_out = out_dim
        else:
            time_embed_dim_out = time_embed_dim
        self.linear_2 = nn.Linear(time_embed_dim, time_embed_dim_out, sample_proj_bias)

        if post_act_fn is None:
            self.post_act = None

    def forward(self, sample, condition=None):
        if condition is not None:
            sample = sample + self.cond_proj(condition)
        sample = self.linear_1(sample)

        if self.act is not None:
            sample = self.act(sample)

        sample = self.linear_2(sample)

        if self.post_act is not None:
            sample = self.post_act(sample)
        return sample


class Timesteps(nn.Module):
    def __init__(self, num_channels: int, flip_sin_to_cos: bool, downscale_freq_shift: float, scale: int = 1):
        super().__init__()
        self.num_channels = num_channels
        self.flip_sin_to_cos = flip_sin_to_cos
        self.downscale_freq_shift = downscale_freq_shift
        self.scale = scale

    def forward(self, timesteps: torch.Tensor) -> torch.Tensor:
        t_emb = get_timestep_embedding(
            timesteps,
            self.num_channels,
            flip_sin_to_cos=self.flip_sin_to_cos,
            downscale_freq_shift=self.downscale_freq_shift,
            scale=self.scale,
        )
        return t_emb


class AdaLayerNormContinuous(nn.Module):
    r"""
    Adaptive normalization layer with a norm layer (layer_norm or rms_norm).

    Args:
        embedding_dim (`int`): Embedding dimension to use during projection.
        conditioning_embedding_dim (`int`): Dimension of the input condition.
        elementwise_affine (`bool`, defaults to `True`):
            Boolean flag to denote if affine transformation should be applied.
        eps (`float`, defaults to 1e-5): Epsilon factor.
        bias (`bias`, defaults to `True`): Boolean flag to denote if bias should be use.
        norm_type (`str`, defaults to `"layer_norm"`):
            Normalization layer to use. Values supported: "layer_norm", "rms_norm".
    """

    def __init__(
        self,
        embedding_dim: int,
        conditioning_embedding_dim: int,
        # NOTE: It is a bit weird that the norm layer can be configured to have scale and shift parameters
        # because the output is immediately scaled and shifted by the projected conditioning embeddings.
        # Note that AdaLayerNorm does not let the norm layer have scale and shift parameters.
        # However, this is how it was implemented in the original code, and it's rather likely you should
        # set `elementwise_affine` to False.
        elementwise_affine=True,
        eps=1e-5,
        bias=True,
        norm_type="layer_norm",
    ):
        super().__init__()
        self.silu = nn.SiLU()
        self.linear = nn.Linear(conditioning_embedding_dim, embedding_dim * 2, bias=bias)
        if norm_type == "layer_norm":
            self.norm = nn.LayerNorm(embedding_dim, eps, elementwise_affine, bias)

    def forward(self, x: torch.Tensor, conditioning_embedding: torch.Tensor) -> torch.Tensor:
        # convert back to the original dtype in case `conditioning_embedding`` is upcasted to float32 (needed for hunyuanDiT)
        emb = self.linear(self.silu(conditioning_embedding).to(x.dtype))
        scale, shift = torch.chunk(emb, 2, dim=1)
        x = self.norm(x) * (1 + scale)[:, None, :] + shift[:, None, :]
        return x


def get_1d_rotary_pos_embed(
    dim: int,
    pos: Union[np.ndarray, int],
    theta: float = 10000.0,
    use_real=False,
    linear_factor=1.0,
    ntk_factor=1.0,
    repeat_interleave_real=True,
    freqs_dtype=torch.float32,  #  torch.float32, torch.float64 (flux)
):
    """
    Precompute the frequency tensor for complex exponentials (cis) with given dimensions.

    This function calculates a frequency tensor with complex exponentials using the given dimension 'dim' and the end
    index 'end'. The 'theta' parameter scales the frequencies. The returned tensor contains complex values in complex64
    data type.

    Args:
        dim (`int`): Dimension of the frequency tensor.
        pos (`np.ndarray` or `int`): Position indices for the frequency tensor. [S] or scalar
        theta (`float`, *optional*, defaults to 10000.0):
            Scaling factor for frequency computation. Defaults to 10000.0.
        use_real (`bool`, *optional*):
            If True, return real part and imaginary part separately. Otherwise, return complex numbers.
        linear_factor (`float`, *optional*, defaults to 1.0):
            Scaling factor for the context extrapolation. Defaults to 1.0.
        ntk_factor (`float`, *optional*, defaults to 1.0):
            Scaling factor for the NTK-Aware RoPE. Defaults to 1.0.
        repeat_interleave_real (`bool`, *optional*, defaults to `True`):
            If `True` and `use_real`, real part and imaginary part are each interleaved with themselves to reach `dim`.
            Otherwise, they are concateanted with themselves.
        freqs_dtype (`torch.float32` or `torch.float64`, *optional*, defaults to `torch.float32`):
            the dtype of the frequency tensor.
    Returns:
        `torch.Tensor`: Precomputed frequency tensor with complex exponentials. [S, D/2]
    """
    assert dim % 2 == 0

    if isinstance(pos, int):
        pos = torch.arange(pos)
    if isinstance(pos, np.ndarray):
        pos = torch.from_numpy(pos)  # type: ignore  # [S]

    theta = theta * ntk_factor
    freqs = (
        1.0 / (theta ** (torch.arange(0, dim, 2, dtype=freqs_dtype, device=pos.device) / dim)) / linear_factor
    )  # [D/2]
    freqs = torch.outer(pos, freqs)  # type: ignore   # [S, D/2]
    is_npu = freqs.device.type == "npu"
    if is_npu:
        freqs = freqs.float()
    if use_real and repeat_interleave_real:
        # flux, hunyuan-dit, cogvideox
        freqs_cos = freqs.cos().repeat_interleave(2, dim=1, output_size=freqs.shape[1] * 2).float()  # [S, D]
        freqs_sin = freqs.sin().repeat_interleave(2, dim=1, output_size=freqs.shape[1] * 2).float()  # [S, D]
        return freqs_cos, freqs_sin
    elif use_real:
        # stable audio, allegro
        freqs_cos = torch.cat([freqs.cos(), freqs.cos()], dim=-1).float()  # [S, D]
        freqs_sin = torch.cat([freqs.sin(), freqs.sin()], dim=-1).float()  # [S, D]
        return freqs_cos, freqs_sin
    else:
        # lumina
        freqs_cis = torch.polar(torch.ones_like(freqs), freqs)  # complex64     # [S, D/2]
        return freqs_cis


def apply_rotary_emb(
    x: torch.Tensor,
    freqs_cis: Union[torch.Tensor, Tuple[torch.Tensor]],
    use_real: bool = True,
    use_real_unbind_dim: int = -1,
    sequence_dim: int = 2,
) -> Tuple[torch.Tensor, torch.Tensor]:
    """
    Apply rotary embeddings to input tensors using the given frequency tensor. This function applies rotary embeddings
    to the given query or key 'x' tensors using the provided frequency tensor 'freqs_cis'. The input tensors are
    reshaped as complex numbers, and the frequency tensor is reshaped for broadcasting compatibility. The resulting
    tensors contain rotary embeddings and are returned as real tensors.

    Args:
        x (`torch.Tensor`):
            Query or key tensor to apply rotary embeddings. [B, H, S, D] xk (torch.Tensor): Key tensor to apply
        freqs_cis (`Tuple[torch.Tensor]`): Precomputed frequency tensor for complex exponentials. ([S, D], [S, D],)

    Returns:
        Tuple[torch.Tensor, torch.Tensor]: Tuple of modified query tensor and key tensor with rotary embeddings.
    """
    if use_real:
        cos, sin = freqs_cis  # [S, D]
        if sequence_dim == 2:
            cos = cos[None, None, :, :]
            sin = sin[None, None, :, :]
        elif sequence_dim == 1:
            cos = cos[None, :, None, :]
            sin = sin[None, :, None, :]
        else:
            raise ValueError(f"`sequence_dim={sequence_dim}` but should be 1 or 2.")

        cos, sin = cos.to(x.device), sin.to(x.device)

        if use_real_unbind_dim == -1:
            # Used for flux, cogvideox, hunyuan-dit
            x_real, x_imag = x.reshape(*x.shape[:-1], -1, 2).unbind(-1)  # [B, H, S, D//2]
            x_rotated = torch.stack([-x_imag, x_real], dim=-1).flatten(3)
        elif use_real_unbind_dim == -2:
            # Used for Stable Audio, OmniGen, CogView4 and Cosmos
            x_real, x_imag = x.reshape(*x.shape[:-1], 2, -1).unbind(-2)  # [B, H, S, D//2]
            x_rotated = torch.cat([-x_imag, x_real], dim=-1)
        else:
            raise ValueError(f"`use_real_unbind_dim={use_real_unbind_dim}` but should be -1 or -2.")

        out = (x.float() * cos + x_rotated.float() * sin).to(x.dtype)

        return out
    else:
        # used for lumina
        x_rotated = torch.view_as_complex(x.float().reshape(*x.shape[:-1], -1, 2))
        freqs_cis = freqs_cis.unsqueeze(2)
        x_out = torch.view_as_real(x_rotated * freqs_cis).flatten(3)

        return x_out.type_as(x)

def _get_projections(attn: "Flux2Attention", hidden_states, encoder_hidden_states=None):
    query = attn.to_q(hidden_states)
    key = attn.to_k(hidden_states)
    value = attn.to_v(hidden_states)

    encoder_query = encoder_key = encoder_value = None
    if encoder_hidden_states is not None and attn.added_kv_proj_dim is not None:
        encoder_query = attn.add_q_proj(encoder_hidden_states)
        encoder_key = attn.add_k_proj(encoder_hidden_states)
        encoder_value = attn.add_v_proj(encoder_hidden_states)

    return query, key, value, encoder_query, encoder_key, encoder_value


def _get_fused_projections(attn: "Flux2Attention", hidden_states, encoder_hidden_states=None):
    query, key, value = attn.to_qkv(hidden_states).chunk(3, dim=-1)

    encoder_query = encoder_key = encoder_value = (None,)
    if encoder_hidden_states is not None and hasattr(attn, "to_added_qkv"):
        encoder_query, encoder_key, encoder_value = attn.to_added_qkv(encoder_hidden_states).chunk(3, dim=-1)

    return query, key, value, encoder_query, encoder_key, encoder_value


def _get_qkv_projections(attn: "Flux2Attention", hidden_states, encoder_hidden_states=None):
    return _get_projections(attn, hidden_states, encoder_hidden_states)


class Flux2SwiGLU(nn.Module):
    """
    Flux 2 uses a SwiGLU-style activation in the transformer feedforward sub-blocks, but with the linear projection
    layer fused into the first linear layer of the FF sub-block. Thus, this module has no trainable parameters.
    """

    def __init__(self):
        super().__init__()
        self.gate_fn = nn.SiLU()

    def forward(self, x: torch.Tensor) -> torch.Tensor:
        x1, x2 = x.chunk(2, dim=-1)
        x = self.gate_fn(x1) * x2
        return x


class Flux2FeedForward(nn.Module):
    def __init__(
        self,
        dim: int,
        dim_out: Optional[int] = None,
        mult: float = 3.0,
        inner_dim: Optional[int] = None,
        bias: bool = False,
    ):
        super().__init__()
        if inner_dim is None:
            inner_dim = int(dim * mult)
        dim_out = dim_out or dim

        # Flux2SwiGLU will reduce the dimension by half
        self.linear_in = nn.Linear(dim, inner_dim * 2, bias=bias)
        self.act_fn = Flux2SwiGLU()
        self.linear_out = nn.Linear(inner_dim, dim_out, bias=bias)

    def forward(self, x: torch.Tensor) -> torch.Tensor:
        x = self.linear_in(x)
        x = self.act_fn(x)
        x = self.linear_out(x)
        return x


class Flux2Attention(torch.nn.Module):
    def __init__(
        self,
        query_dim: int,
        heads: int = 8,
        dim_head: int = 64,
        dropout: float = 0.0,
        bias: bool = False,
        added_kv_proj_dim: Optional[int] = None,
        added_proj_bias: Optional[bool] = True,
        out_bias: bool = True,
        eps: float = 1e-5,
        out_dim: int = None,
        elementwise_affine: bool = True,
    ):
        super().__init__()

        self.head_dim = dim_head
        self.inner_dim = out_dim if out_dim is not None else dim_head * heads
        self.query_dim = query_dim
        self.out_dim = out_dim if out_dim is not None else query_dim
        self.heads = out_dim // dim_head if out_dim is not None else heads

        self.use_bias = bias
        self.dropout = dropout

        self.added_kv_proj_dim = added_kv_proj_dim
        self.added_proj_bias = added_proj_bias

        self.to_q = torch.nn.Linear(query_dim, self.inner_dim, bias=bias)
        self.to_k = torch.nn.Linear(query_dim, self.inner_dim, bias=bias)
        self.to_v = torch.nn.Linear(query_dim, self.inner_dim, bias=bias)

        # QK Norm
        self.norm_q = torch.nn.RMSNorm(dim_head, eps=eps, elementwise_affine=elementwise_affine)
        self.norm_k = torch.nn.RMSNorm(dim_head, eps=eps, elementwise_affine=elementwise_affine)

        self.to_out = torch.nn.ModuleList([])
        self.to_out.append(torch.nn.Linear(self.inner_dim, self.out_dim, bias=out_bias))
        self.to_out.append(torch.nn.Dropout(dropout))

        if added_kv_proj_dim is not None:
            self.norm_added_q = torch.nn.RMSNorm(dim_head, eps=eps)
            self.norm_added_k = torch.nn.RMSNorm(dim_head, eps=eps)
            self.add_q_proj = torch.nn.Linear(added_kv_proj_dim, self.inner_dim, bias=added_proj_bias)
            self.add_k_proj = torch.nn.Linear(added_kv_proj_dim, self.inner_dim, bias=added_proj_bias)
            self.add_v_proj = torch.nn.Linear(added_kv_proj_dim, self.inner_dim, bias=added_proj_bias)
            self.to_add_out = torch.nn.Linear(self.inner_dim, query_dim, bias=out_bias)

    def forward(
        self,
        hidden_states: torch.Tensor,
        encoder_hidden_states: Optional[torch.Tensor] = None,
        attention_mask: Optional[torch.Tensor] = None,
        image_rotary_emb: Optional[torch.Tensor] = None,
        kv_cache = None,
        **kwargs,
    ) -> torch.Tensor:
        query, key, value, encoder_query, encoder_key, encoder_value = _get_qkv_projections(
            self, hidden_states, encoder_hidden_states
        )

        query = query.unflatten(-1, (self.heads, -1))
        key = key.unflatten(-1, (self.heads, -1))
        value = value.unflatten(-1, (self.heads, -1))

        query = self.norm_q(query)
        key = self.norm_k(key)

        if self.added_kv_proj_dim is not None:
            encoder_query = encoder_query.unflatten(-1, (self.heads, -1))
            encoder_key = encoder_key.unflatten(-1, (self.heads, -1))
            encoder_value = encoder_value.unflatten(-1, (self.heads, -1))

            encoder_query = self.norm_added_q(encoder_query)
            encoder_key = self.norm_added_k(encoder_key)

            query = torch.cat([encoder_query, query], dim=1)
            key = torch.cat([encoder_key, key], dim=1)
            value = torch.cat([encoder_value, value], dim=1)

        if image_rotary_emb is not None:
            query = apply_rotary_emb(query, image_rotary_emb, sequence_dim=1)
            key = apply_rotary_emb(key, image_rotary_emb, sequence_dim=1)

        if kv_cache is not None:
            key = torch.concat([key, kv_cache[0]], dim=1)
            value = torch.concat([value, kv_cache[1]], dim=1)
        hidden_states = attention_forward(
            query,
            key,
            value,
            q_pattern="b s n d", k_pattern="b s n d", v_pattern="b s n d", out_pattern="b s n d",
        )
        hidden_states = hidden_states.flatten(2, 3)
        hidden_states = hidden_states.to(query.dtype)

        if encoder_hidden_states is not None:
            encoder_hidden_states, hidden_states = hidden_states.split_with_sizes(
                [encoder_hidden_states.shape[1], hidden_states.shape[1] - encoder_hidden_states.shape[1]], dim=1
            )
            encoder_hidden_states = self.to_add_out(encoder_hidden_states)

        hidden_states = self.to_out[0](hidden_states)
        hidden_states = self.to_out[1](hidden_states)

        if encoder_hidden_states is not None:
            return hidden_states, encoder_hidden_states
        else:
            return hidden_states


class Flux2ParallelSelfAttention(torch.nn.Module):
    def __init__(
        self,
        query_dim: int,
        heads: int = 8,
        dim_head: int = 64,
        dropout: float = 0.0,
        bias: bool = False,
        out_bias: bool = True,
        eps: float = 1e-5,
        out_dim: int = None,
        elementwise_affine: bool = True,
        mlp_ratio: float = 4.0,
        mlp_mult_factor: int = 2,
        processor=None,
    ):
        super().__init__()

        self.head_dim = dim_head
        self.inner_dim = out_dim if out_dim is not None else dim_head * heads
        self.query_dim = query_dim
        self.out_dim = out_dim if out_dim is not None else query_dim
        self.heads = out_dim // dim_head if out_dim is not None else heads

        self.use_bias = bias
        self.dropout = dropout

        self.mlp_ratio = mlp_ratio
        self.mlp_hidden_dim = int(query_dim * self.mlp_ratio)
        self.mlp_mult_factor = mlp_mult_factor

        # Fused QKV projections + MLP input projection
        self.to_qkv_mlp_proj = torch.nn.Linear(
            self.query_dim, self.inner_dim * 3 + self.mlp_hidden_dim * self.mlp_mult_factor, bias=bias
        )
        self.mlp_act_fn = Flux2SwiGLU()

        # QK Norm
        self.norm_q = torch.nn.RMSNorm(dim_head, eps=eps, elementwise_affine=elementwise_affine)
        self.norm_k = torch.nn.RMSNorm(dim_head, eps=eps, elementwise_affine=elementwise_affine)

        # Fused attention output projection + MLP output projection
        self.to_out = torch.nn.Linear(self.inner_dim + self.mlp_hidden_dim, self.out_dim, bias=out_bias)

    def forward(
        self,
        hidden_states: torch.Tensor,
        attention_mask: Optional[torch.Tensor] = None,
        image_rotary_emb: Optional[torch.Tensor] = None,
        kv_cache = None,
        **kwargs,
    ) -> torch.Tensor:
        # Parallel in (QKV + MLP in) projection
        hidden_states = self.to_qkv_mlp_proj(hidden_states)
        qkv, mlp_hidden_states = torch.split(
            hidden_states, [3 * self.inner_dim, self.mlp_hidden_dim * self.mlp_mult_factor], dim=-1
        )

        # Handle the attention logic
        query, key, value = qkv.chunk(3, dim=-1)

        query = query.unflatten(-1, (self.heads, -1))
        key = key.unflatten(-1, (self.heads, -1))
        value = value.unflatten(-1, (self.heads, -1))

        query = self.norm_q(query)
        key = self.norm_k(key)

        if image_rotary_emb is not None:
            query = apply_rotary_emb(query, image_rotary_emb, sequence_dim=1)
            key = apply_rotary_emb(key, image_rotary_emb, sequence_dim=1)

        if kv_cache is not None:
            key = torch.concat([key, kv_cache[0]], dim=1)
            value = torch.concat([value, kv_cache[1]], dim=1)
        hidden_states = attention_forward(
            query,
            key,
            value,
            q_pattern="b s n d", k_pattern="b s n d", v_pattern="b s n d", out_pattern="b s n d",
        )
        hidden_states = hidden_states.flatten(2, 3)
        hidden_states = hidden_states.to(query.dtype)

        # Handle the feedforward (FF) logic
        mlp_hidden_states = self.mlp_act_fn(mlp_hidden_states)

        # Concatenate and parallel output projection
        hidden_states = torch.cat([hidden_states, mlp_hidden_states], dim=-1)
        hidden_states = self.to_out(hidden_states)

        return hidden_states


class Flux2SingleTransformerBlock(nn.Module):
    def __init__(
        self,
        dim: int,
        num_attention_heads: int,
        attention_head_dim: int,
        mlp_ratio: float = 3.0,
        eps: float = 1e-6,
        bias: bool = False,
    ):
        super().__init__()

        self.norm = nn.LayerNorm(dim, elementwise_affine=False, eps=eps)

        # Note that the MLP in/out linear layers are fused with the attention QKV/out projections, respectively; this
        # is often called a "parallel" transformer block. See the [ViT-22B paper](https://arxiv.org/abs/2302.05442)
        # for a visual depiction of this type of transformer block.
        self.attn = Flux2ParallelSelfAttention(
            query_dim=dim,
            dim_head=attention_head_dim,
            heads=num_attention_heads,
            out_dim=dim,
            bias=bias,
            out_bias=bias,
            eps=eps,
            mlp_ratio=mlp_ratio,
            mlp_mult_factor=2,
        )

    def forward(
        self,
        hidden_states: torch.Tensor,
        encoder_hidden_states: Optional[torch.Tensor],
        temb_mod_params: Tuple[torch.Tensor, torch.Tensor, torch.Tensor],
        image_rotary_emb: Optional[Tuple[torch.Tensor, torch.Tensor]] = None,
        joint_attention_kwargs: Optional[Dict[str, Any]] = None,
        split_hidden_states: bool = False,
        text_seq_len: Optional[int] = None,
        kv_cache = None,
    ) -> Tuple[torch.Tensor, torch.Tensor]:
        # If encoder_hidden_states is None, hidden_states is assumed to have encoder_hidden_states already
        # concatenated
        if encoder_hidden_states is not None:
            text_seq_len = encoder_hidden_states.shape[1]
            hidden_states = torch.cat([encoder_hidden_states, hidden_states], dim=1)

        mod_shift, mod_scale, mod_gate = temb_mod_params

        norm_hidden_states = self.norm(hidden_states)
        norm_hidden_states = (1 + mod_scale) * norm_hidden_states + mod_shift

        joint_attention_kwargs = joint_attention_kwargs or {}
        attn_output = self.attn(
            hidden_states=norm_hidden_states,
            image_rotary_emb=image_rotary_emb,
            kv_cache=kv_cache,
            **joint_attention_kwargs,
        )

        hidden_states = hidden_states + mod_gate * attn_output
        if hidden_states.dtype == torch.float16:
            hidden_states = hidden_states.clip(-65504, 65504)

        if split_hidden_states:
            encoder_hidden_states, hidden_states = hidden_states[:, :text_seq_len], hidden_states[:, text_seq_len:]
            return encoder_hidden_states, hidden_states
        else:
            return hidden_states


class Flux2TransformerBlock(nn.Module):
    def __init__(
        self,
        dim: int,
        num_attention_heads: int,
        attention_head_dim: int,
        mlp_ratio: float = 3.0,
        eps: float = 1e-6,
        bias: bool = False,
    ):
        super().__init__()
        self.mlp_hidden_dim = int(dim * mlp_ratio)

        self.norm1 = nn.LayerNorm(dim, elementwise_affine=False, eps=eps)
        self.norm1_context = nn.LayerNorm(dim, elementwise_affine=False, eps=eps)

        self.attn = Flux2Attention(
            query_dim=dim,
            added_kv_proj_dim=dim,
            dim_head=attention_head_dim,
            heads=num_attention_heads,
            out_dim=dim,
            bias=bias,
            added_proj_bias=bias,
            out_bias=bias,
            eps=eps,
        )

        self.norm2 = nn.LayerNorm(dim, elementwise_affine=False, eps=eps)
        self.ff = Flux2FeedForward(dim=dim, dim_out=dim, mult=mlp_ratio, bias=bias)

        self.norm2_context = nn.LayerNorm(dim, elementwise_affine=False, eps=eps)
        self.ff_context = Flux2FeedForward(dim=dim, dim_out=dim, mult=mlp_ratio, bias=bias)

    def forward(
        self,
        hidden_states: torch.Tensor,
        encoder_hidden_states: torch.Tensor,
        temb_mod_params_img: Tuple[Tuple[torch.Tensor, torch.Tensor, torch.Tensor], ...],
        temb_mod_params_txt: Tuple[Tuple[torch.Tensor, torch.Tensor, torch.Tensor], ...],
        image_rotary_emb: Optional[Tuple[torch.Tensor, torch.Tensor]] = None,
        joint_attention_kwargs: Optional[Dict[str, Any]] = None,
        kv_cache = None,
    ) -> Tuple[torch.Tensor, torch.Tensor]:
        joint_attention_kwargs = joint_attention_kwargs or {}

        # Modulation parameters shape: [1, 1, self.dim]
        (shift_msa, scale_msa, gate_msa), (shift_mlp, scale_mlp, gate_mlp) = temb_mod_params_img
        (c_shift_msa, c_scale_msa, c_gate_msa), (c_shift_mlp, c_scale_mlp, c_gate_mlp) = temb_mod_params_txt

        # Img stream
        norm_hidden_states = self.norm1(hidden_states)
        norm_hidden_states = (1 + scale_msa) * norm_hidden_states + shift_msa

        # Conditioning txt stream
        norm_encoder_hidden_states = self.norm1_context(encoder_hidden_states)
        norm_encoder_hidden_states = (1 + c_scale_msa) * norm_encoder_hidden_states + c_shift_msa

        # Attention on concatenated img + txt stream
        attention_outputs = self.attn(
            hidden_states=norm_hidden_states,
            encoder_hidden_states=norm_encoder_hidden_states,
            image_rotary_emb=image_rotary_emb,
            kv_cache=kv_cache,
            **joint_attention_kwargs,
        )

        attn_output, context_attn_output = attention_outputs

        # Process attention outputs for the image stream (`hidden_states`).
        attn_output = gate_msa * attn_output
        hidden_states = hidden_states + attn_output

        norm_hidden_states = self.norm2(hidden_states)
        norm_hidden_states = norm_hidden_states * (1 + scale_mlp) + shift_mlp

        ff_output = self.ff(norm_hidden_states)
        hidden_states = hidden_states + gate_mlp * ff_output

        # Process attention outputs for the text stream (`encoder_hidden_states`).
        context_attn_output = c_gate_msa * context_attn_output
        encoder_hidden_states = encoder_hidden_states + context_attn_output

        norm_encoder_hidden_states = self.norm2_context(encoder_hidden_states)
        norm_encoder_hidden_states = norm_encoder_hidden_states * (1 + c_scale_mlp) + c_shift_mlp

        context_ff_output = self.ff_context(norm_encoder_hidden_states)
        encoder_hidden_states = encoder_hidden_states + c_gate_mlp * context_ff_output
        if encoder_hidden_states.dtype == torch.float16:
            encoder_hidden_states = encoder_hidden_states.clip(-65504, 65504)

        return encoder_hidden_states, hidden_states


class Flux2PosEmbed(nn.Module):
    # modified from https://github.com/black-forest-labs/flux/blob/c00d7c60b085fce8058b9df845e036090873f2ce/src/flux/modules/layers.py#L11
    def __init__(self, theta: int, axes_dim: List[int]):
        super().__init__()
        self.theta = theta
        self.axes_dim = axes_dim

    def forward(self, ids: torch.Tensor) -> torch.Tensor:
        # Expected ids shape: [S, len(self.axes_dim)]
        cos_out = []
        sin_out = []
        pos = ids.float()
        is_mps = ids.device.type == "mps"
        is_npu = ids.device.type == "npu"
        freqs_dtype = torch.float32 if (is_mps or is_npu) else torch.float64
        # Unlike Flux 1, loop over len(self.axes_dim) rather than ids.shape[-1]
        for i in range(len(self.axes_dim)):
            cos, sin = get_1d_rotary_pos_embed(
                self.axes_dim[i],
                pos[..., i],
                theta=self.theta,
                repeat_interleave_real=True,
                use_real=True,
                freqs_dtype=freqs_dtype,
            )
            cos_out.append(cos)
            sin_out.append(sin)
        freqs_cos = torch.cat(cos_out, dim=-1).to(ids.device)
        freqs_sin = torch.cat(sin_out, dim=-1).to(ids.device)
        return freqs_cos, freqs_sin


class Flux2TimestepGuidanceEmbeddings(nn.Module):
    def __init__(
        self,
        in_channels: int = 256,
        embedding_dim: int = 6144,
        bias: bool = False,
        guidance_embeds: bool = True,
    ):
        super().__init__()

        self.time_proj = Timesteps(num_channels=in_channels, flip_sin_to_cos=True, downscale_freq_shift=0)
        self.timestep_embedder = TimestepEmbedding(
            in_channels=in_channels, time_embed_dim=embedding_dim, sample_proj_bias=bias
        )

        if guidance_embeds:
            self.guidance_embedder = TimestepEmbedding(
                in_channels=in_channels, time_embed_dim=embedding_dim, sample_proj_bias=bias
            )
        else:
            self.guidance_embedder = None

    def forward(self, timestep: torch.Tensor, guidance: torch.Tensor) -> torch.Tensor:
        timesteps_proj = self.time_proj(timestep)
        timesteps_emb = self.timestep_embedder(timesteps_proj.to(timestep.dtype))  # (N, D)

        if guidance is not None and self.guidance_embedder is not None:
            guidance_proj = self.time_proj(guidance)
            guidance_emb = self.guidance_embedder(guidance_proj.to(guidance.dtype))  # (N, D)
            time_guidance_emb = timesteps_emb + guidance_emb
            return time_guidance_emb
        else:
            return timesteps_emb


class Flux2Modulation(nn.Module):
    def __init__(self, dim: int, mod_param_sets: int = 2, bias: bool = False):
        super().__init__()
        self.mod_param_sets = mod_param_sets

        self.linear = nn.Linear(dim, dim * 3 * self.mod_param_sets, bias=bias)
        self.act_fn = nn.SiLU()

    def forward(self, temb: torch.Tensor) -> Tuple[Tuple[torch.Tensor, torch.Tensor, torch.Tensor], ...]:
        mod = self.act_fn(temb)
        mod = self.linear(mod)

        if mod.ndim == 2:
            mod = mod.unsqueeze(1)
        mod_params = torch.chunk(mod, 3 * self.mod_param_sets, dim=-1)
        # Return tuple of 3-tuples of modulation params shift/scale/gate
        return tuple(mod_params[3 * i : 3 * (i + 1)] for i in range(self.mod_param_sets))


class Flux2DiT(torch.nn.Module):

    _repeated_blocks = ["Flux2TransformerBlock", "Flux2SingleTransformerBlock"]

    def __init__(
        self,
        patch_size: int = 1,
        in_channels: int = 128,
        out_channels: Optional[int] = None,
        num_layers: int = 8,
        num_single_layers: int = 48,
        attention_head_dim: int = 128,
        num_attention_heads: int = 48,
        joint_attention_dim: int = 15360,
        timestep_guidance_channels: int = 256,
        mlp_ratio: float = 3.0,
        axes_dims_rope: Tuple[int, ...] = (32, 32, 32, 32),
        rope_theta: int = 2000,
        eps: float = 1e-6,
        guidance_embeds: bool = True,
    ):
        super().__init__()
        self.out_channels = out_channels or in_channels
        self.inner_dim = num_attention_heads * attention_head_dim

        # 1. Sinusoidal positional embedding for RoPE on image and text tokens
        self.pos_embed = Flux2PosEmbed(theta=rope_theta, axes_dim=axes_dims_rope)

        # 2. Combined timestep + guidance embedding
        self.time_guidance_embed = Flux2TimestepGuidanceEmbeddings(
            in_channels=timestep_guidance_channels,
            embedding_dim=self.inner_dim,
            bias=False,
            guidance_embeds=guidance_embeds,
        )

        # 3. Modulation (double stream and single stream blocks share modulation parameters, resp.)
        # Two sets of shift/scale/gate modulation parameters for the double stream attn and FF sub-blocks
        self.double_stream_modulation_img = Flux2Modulation(self.inner_dim, mod_param_sets=2, bias=False)
        self.double_stream_modulation_txt = Flux2Modulation(self.inner_dim, mod_param_sets=2, bias=False)
        # Only one set of modulation parameters as the attn and FF sub-blocks are run in parallel for single stream
        self.single_stream_modulation = Flux2Modulation(self.inner_dim, mod_param_sets=1, bias=False)

        # 4. Input projections
        self.x_embedder = nn.Linear(in_channels, self.inner_dim, bias=False)
        self.context_embedder = nn.Linear(joint_attention_dim, self.inner_dim, bias=False)

        # 5. Double Stream Transformer Blocks
        self.transformer_blocks = nn.ModuleList(
            [
                Flux2TransformerBlock(
                    dim=self.inner_dim,
                    num_attention_heads=num_attention_heads,
                    attention_head_dim=attention_head_dim,
                    mlp_ratio=mlp_ratio,
                    eps=eps,
                    bias=False,
                )
                for _ in range(num_layers)
            ]
        )

        # 6. Single Stream Transformer Blocks
        self.single_transformer_blocks = nn.ModuleList(
            [
                Flux2SingleTransformerBlock(
                    dim=self.inner_dim,
                    num_attention_heads=num_attention_heads,
                    attention_head_dim=attention_head_dim,
                    mlp_ratio=mlp_ratio,
                    eps=eps,
                    bias=False,
                )
                for _ in range(num_single_layers)
            ]
        )

        # 7. Output layers
        self.norm_out = AdaLayerNormContinuous(
            self.inner_dim, self.inner_dim, elementwise_affine=False, eps=eps, bias=False
        )
        self.proj_out = nn.Linear(self.inner_dim, patch_size * patch_size * self.out_channels, bias=False)

        self.gradient_checkpointing = False

    def forward(
        self,
        hidden_states: torch.Tensor,
        encoder_hidden_states: torch.Tensor = None,
        timestep: torch.LongTensor = None,
        img_ids: torch.Tensor = None,
        txt_ids: torch.Tensor = None,
        guidance: torch.Tensor = None,
        joint_attention_kwargs: Optional[Dict[str, Any]] = None,
        kv_cache = None,
        use_gradient_checkpointing=False,
        use_gradient_checkpointing_offload=False,
    ):
        # 0. Handle input arguments
        if joint_attention_kwargs is not None:
            joint_attention_kwargs = joint_attention_kwargs.copy()
            lora_scale = joint_attention_kwargs.pop("scale", 1.0)
        else:
            lora_scale = 1.0

        num_txt_tokens = encoder_hidden_states.shape[1]

        # 1. Calculate timestep embedding and modulation parameters
        timestep = timestep.to(hidden_states.dtype) * 1000

        if guidance is not None:
            guidance = guidance.to(hidden_states.dtype) * 1000

        temb = self.time_guidance_embed(timestep, guidance)

        double_stream_mod_img = self.double_stream_modulation_img(temb)
        double_stream_mod_txt = self.double_stream_modulation_txt(temb)
        single_stream_mod = self.single_stream_modulation(temb)[0]

        # 2. Input projection for image (hidden_states) and conditioning text (encoder_hidden_states)
        hidden_states = self.x_embedder(hidden_states)
        encoder_hidden_states = self.context_embedder(encoder_hidden_states)

        # 3. Calculate RoPE embeddings from image and text tokens
        # NOTE: the below logic means that we can't support batched inference with images of different resolutions or
        # text prompts of differents lengths. Is this a use case we want to support?
        if img_ids.ndim == 3:
            img_ids = img_ids[0]
        if txt_ids.ndim == 3:
            txt_ids = txt_ids[0]

        image_rotary_emb = self.pos_embed(img_ids)
        text_rotary_emb = self.pos_embed(txt_ids)
        concat_rotary_emb = (
            torch.cat([text_rotary_emb[0], image_rotary_emb[0]], dim=0),
            torch.cat([text_rotary_emb[1], image_rotary_emb[1]], dim=0),
        )

        # 4. Double Stream Transformer Blocks
        for block_id, block in enumerate(self.transformer_blocks):
            encoder_hidden_states, hidden_states = gradient_checkpoint_forward(
                block,
                use_gradient_checkpointing=use_gradient_checkpointing,
                use_gradient_checkpointing_offload=use_gradient_checkpointing_offload,
                hidden_states=hidden_states,
                encoder_hidden_states=encoder_hidden_states,
                temb_mod_params_img=double_stream_mod_img,
                temb_mod_params_txt=double_stream_mod_txt,
                image_rotary_emb=concat_rotary_emb,
                joint_attention_kwargs=joint_attention_kwargs,
                kv_cache=None if kv_cache is None else kv_cache.get(f"double_{block_id}"),
            )
        # Concatenate text and image streams for single-block inference
        hidden_states = torch.cat([encoder_hidden_states, hidden_states], dim=1)

        # 5. Single Stream Transformer Blocks
        for block_id, block in enumerate(self.single_transformer_blocks):
            hidden_states = gradient_checkpoint_forward(
                block,
                use_gradient_checkpointing=use_gradient_checkpointing,
                use_gradient_checkpointing_offload=use_gradient_checkpointing_offload,
                hidden_states=hidden_states,
                encoder_hidden_states=None,
                temb_mod_params=single_stream_mod,
                image_rotary_emb=concat_rotary_emb,
                joint_attention_kwargs=joint_attention_kwargs,
                kv_cache=None if kv_cache is None else kv_cache.get(f"single_{block_id}"),
            )
        # Remove text tokens from concatenated stream
        hidden_states = hidden_states[:, num_txt_tokens:, ...]

        # 6. Output layers
        hidden_states = self.norm_out(hidden_states, temb)
        output = self.proj_out(hidden_states)

        return output