model/transformer_flux.py

# Copyright 2025 Black Forest Labs, The HuggingFace Team and The InstantX Team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.


from typing import Any, Dict, Optional, Tuple, Union

import numpy as np
import torch
import torch.nn as nn

from diffusers.configuration_utils import ConfigMixin, register_to_config
from diffusers.loaders import FluxTransformer2DLoadersMixin, FromOriginalModelMixin, PeftAdapterMixin
from diffusers.utils import USE_PEFT_BACKEND, deprecate, logging, scale_lora_layers, unscale_lora_layers
from diffusers.utils.import_utils import is_torch_npu_available
from diffusers.utils.torch_utils import maybe_allow_in_graph
from diffusers.models.attention import FeedForward
from diffusers.models.attention_processor import (
    Attention,
    AttentionProcessor,
    FluxAttnProcessor2_0,
    FluxAttnProcessor2_0_NPU,
    FusedFluxAttnProcessor2_0,
)
from diffusers.models.cache_utils import CacheMixin
from diffusers.models.embeddings import CombinedTimestepGuidanceTextProjEmbeddings, CombinedTimestepTextProjEmbeddings, FluxPosEmbed
from diffusers.models.modeling_outputs import Transformer2DModelOutput
from diffusers.models.modeling_utils import ModelMixin
from diffusers.models.normalization import AdaLayerNormContinuous, AdaLayerNormZero, AdaLayerNormZeroSingle


logger = logging.get_logger(__name__)  # pylint: disable=invalid-name


@maybe_allow_in_graph
class FluxSingleTransformerBlock(nn.Module):
    def __init__(self, dim: int, num_attention_heads: int, attention_head_dim: int, mlp_ratio: float = 4.0):
        super().__init__()
        self.mlp_hidden_dim = int(dim * mlp_ratio)

        self.norm = AdaLayerNormZeroSingle(dim)
        self.proj_mlp = nn.Linear(dim, self.mlp_hidden_dim)
        self.act_mlp = nn.GELU(approximate="tanh")
        self.proj_out = nn.Linear(dim + self.mlp_hidden_dim, dim)

        if is_torch_npu_available():
            deprecation_message = (
                "Defaulting to FluxAttnProcessor2_0_NPU for NPU devices will be removed. Attention processors "
                "should be set explicitly using the `set_attn_processor` method."
            )
            deprecate("npu_processor", "0.34.0", deprecation_message)
            processor = FluxAttnProcessor2_0_NPU()
        else:
            processor = FluxAttnProcessor2_0()

        self.attn = Attention(
            query_dim=dim,
            cross_attention_dim=None,
            dim_head=attention_head_dim,
            heads=num_attention_heads,
            out_dim=dim,
            bias=True,
            processor=processor,
            qk_norm="rms_norm",
            eps=1e-6,
            pre_only=True,
        )

    def forward(
        self,
        hidden_states: torch.Tensor,
        temb: torch.Tensor,
        image_rotary_emb: Optional[Tuple[torch.Tensor, torch.Tensor]] = None,
        joint_attention_kwargs: Optional[Dict[str, Any]] = None,
    ) -> torch.Tensor:
        residual = hidden_states
        norm_hidden_states, gate = self.norm(hidden_states, emb=temb)
        mlp_hidden_states = self.act_mlp(self.proj_mlp(norm_hidden_states))
        joint_attention_kwargs = joint_attention_kwargs or {}
        attn_output = self.attn(
            hidden_states=norm_hidden_states,
            image_rotary_emb=image_rotary_emb,
            **joint_attention_kwargs,
        )

        hidden_states = torch.cat([attn_output, mlp_hidden_states], dim=2)
        gate = gate.unsqueeze(1)
        hidden_states = gate * self.proj_out(hidden_states)
        hidden_states = residual + hidden_states
        if hidden_states.dtype == torch.float16:
            hidden_states = hidden_states.clip(-65504, 65504)

        return hidden_states


@maybe_allow_in_graph
class FluxTransformerBlock(nn.Module):
    def __init__(
        self, dim: int, num_attention_heads: int, attention_head_dim: int, qk_norm: str = "rms_norm", eps: float = 1e-6
    ):
        super().__init__()

        self.norm1 = AdaLayerNormZero(dim)
        self.norm1_context = AdaLayerNormZero(dim)

        self.attn = Attention(
            query_dim=dim,
            cross_attention_dim=None,
            added_kv_proj_dim=dim,
            dim_head=attention_head_dim,
            heads=num_attention_heads,
            out_dim=dim,
            context_pre_only=False,
            bias=True,
            processor=FluxAttnProcessor2_0(),
            qk_norm=qk_norm,
            eps=eps,
        )

        self.norm2 = nn.LayerNorm(dim, elementwise_affine=False, eps=1e-6)
        self.ff = FeedForward(dim=dim, dim_out=dim, activation_fn="gelu-approximate")

        self.norm2_context = nn.LayerNorm(dim, elementwise_affine=False, eps=1e-6)
        self.ff_context = FeedForward(dim=dim, dim_out=dim, activation_fn="gelu-approximate")

    def forward(
        self,
        hidden_states: torch.Tensor,
        encoder_hidden_states: torch.Tensor,
        temb: torch.Tensor,
        image_rotary_emb: Optional[Tuple[torch.Tensor, torch.Tensor]] = None,
        joint_attention_kwargs: Optional[Dict[str, Any]] = None,
        ## ---- adding the modulation conditioning vector for controlling the strength ---- ## 
        modulation_condn: Optional[torch.Tensor] = None,  
    ) -> Tuple[torch.Tensor, torch.Tensor]:

        # add logic here for conditioning 
        norm_hidden_states, gate_msa, shift_mlp, scale_mlp, gate_mlp = self.norm1(hidden_states, emb=temb)
        
        norm_encoder_hidden_states, c_gate_msa, c_shift_mlp, c_scale_mlp, c_gate_mlp = self.norm1_context(
            encoder_hidden_states, emb=temb
        )
        
        # If modulation conditioning is passed, then we will use that to adjust the scale and shift parameters otherwise we will proceed with regular modulation function
        if modulation_condn is not None:
            modulation_condn = modulation_condn.squeeze(1)

            # chunking the modulation space here 
            modulation_scale, modulation_shift = modulation_condn.chunk(2, dim=1) # dividing the output into two parts, one for scale and another one for shift. 
            # print("modulation condn shape: {}".format(modulation_condn.shape)) # [1, out_dim]
            
            # adding a delta shift to the shift modulation vector 
            c_shift_mlp = c_shift_mlp + modulation_shift
            # adding a delta scale to the scale modulation vector 
            c_scale_mlp = c_scale_mlp + modulation_scale 

        joint_attention_kwargs = joint_attention_kwargs or {}
        # Attention.
        attention_outputs = self.attn(
            hidden_states=norm_hidden_states,
            encoder_hidden_states=norm_encoder_hidden_states,
            image_rotary_emb=image_rotary_emb,
            **joint_attention_kwargs,
        )

        if len(attention_outputs) == 2:
            attn_output, context_attn_output = attention_outputs
        elif len(attention_outputs) == 3:
            attn_output, context_attn_output, ip_attn_output = attention_outputs

        # Process attention outputs for the `hidden_states`.
        attn_output = gate_msa.unsqueeze(1) * attn_output
        hidden_states = hidden_states + attn_output

        norm_hidden_states = self.norm2(hidden_states)
        norm_hidden_states = norm_hidden_states * (1 + scale_mlp[:, None]) + shift_mlp[:, None]

        ff_output = self.ff(norm_hidden_states)
        ff_output = gate_mlp.unsqueeze(1) * ff_output

        hidden_states = hidden_states + ff_output
        if len(attention_outputs) == 3:
            hidden_states = hidden_states + ip_attn_output

        # Process attention outputs for the `encoder_hidden_states`.

        context_attn_output = c_gate_msa.unsqueeze(1) * 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[:, None]) + c_shift_mlp[:, None]

        context_ff_output = self.ff_context(norm_encoder_hidden_states)
        encoder_hidden_states = encoder_hidden_states + c_gate_mlp.unsqueeze(1) * 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 FluxTransformer2DModelwithSliderConditioning(
    ModelMixin, ConfigMixin, PeftAdapterMixin, FromOriginalModelMixin, FluxTransformer2DLoadersMixin, CacheMixin
):
    """
    The Transformer model introduced in Flux.

    Reference: https://blackforestlabs.ai/announcing-black-forest-labs/

    Args:
        patch_size (`int`, defaults to `1`):
            Patch size to turn the input data into small patches.
        in_channels (`int`, defaults to `64`):
            The number of channels in the input.
        out_channels (`int`, *optional*, defaults to `None`):
            The number of channels in the output. If not specified, it defaults to `in_channels`.
        num_layers (`int`, defaults to `19`):
            The number of layers of dual stream DiT blocks to use.
        num_single_layers (`int`, defaults to `38`):
            The number of layers of single stream DiT blocks to use.
        attention_head_dim (`int`, defaults to `128`):
            The number of dimensions to use for each attention head.
        num_attention_heads (`int`, defaults to `24`):
            The number of attention heads to use.
        joint_attention_dim (`int`, defaults to `4096`):
            The number of dimensions to use for the joint attention (embedding/channel dimension of
            `encoder_hidden_states`).
        pooled_projection_dim (`int`, defaults to `768`):
            The number of dimensions to use for the pooled projection.
        guidance_embeds (`bool`, defaults to `False`):
            Whether to use guidance embeddings for guidance-distilled variant of the model.
        axes_dims_rope (`Tuple[int]`, defaults to `(16, 56, 56)`):
            The dimensions to use for the rotary positional embeddings.
    """

    _supports_gradient_checkpointing = True
    _no_split_modules = ["FluxTransformerBlock", "FluxSingleTransformerBlock"]
    _skip_layerwise_casting_patterns = ["pos_embed", "norm"]
    _repeated_blocks = ["FluxTransformerBlock", "FluxSingleTransformerBlock"]

    @register_to_config
    def __init__(
        self,
        patch_size: int = 1,
        in_channels: int = 64,
        out_channels: Optional[int] = None,
        num_layers: int = 19,
        num_single_layers: int = 38,
        attention_head_dim: int = 128,
        num_attention_heads: int = 24,
        joint_attention_dim: int = 4096,
        pooled_projection_dim: int = 768,
        guidance_embeds: bool = False,
        axes_dims_rope: Tuple[int, int, int] = (16, 56, 56),
    ):
        super().__init__()
        self.out_channels = out_channels or in_channels
        self.inner_dim = num_attention_heads * attention_head_dim

        self.pos_embed = FluxPosEmbed(theta=10000, axes_dim=axes_dims_rope)

        text_time_guidance_cls = (
            CombinedTimestepGuidanceTextProjEmbeddings if guidance_embeds else CombinedTimestepTextProjEmbeddings
        )
        self.time_text_embed = text_time_guidance_cls(
            embedding_dim=self.inner_dim, pooled_projection_dim=pooled_projection_dim
        )

        self.context_embedder = nn.Linear(joint_attention_dim, self.inner_dim)
        self.x_embedder = nn.Linear(in_channels, self.inner_dim)

        self.transformer_blocks = nn.ModuleList(
            [   # we will add conditioning logic in this block for training with modulation space 
                FluxTransformerBlock(
                    dim=self.inner_dim,
                    num_attention_heads=num_attention_heads,
                    attention_head_dim=attention_head_dim,
                )
                for _ in range(num_layers)
            ]
        )

        self.single_transformer_blocks = nn.ModuleList(
            [
                FluxSingleTransformerBlock(
                    dim=self.inner_dim,
                    num_attention_heads=num_attention_heads,
                    attention_head_dim=attention_head_dim,
                )
                for _ in range(num_single_layers)
            ]
        )

        self.norm_out = AdaLayerNormContinuous(self.inner_dim, self.inner_dim, elementwise_affine=False, eps=1e-6)
        self.proj_out = nn.Linear(self.inner_dim, patch_size * patch_size * self.out_channels, bias=True)

        self.gradient_checkpointing = False

    @property
    # Copied from diffusers.models.unets.unet_2d_condition.UNet2DConditionModel.attn_processors
    def attn_processors(self) -> Dict[str, AttentionProcessor]:
        r"""
        Returns:
            `dict` of attention processors: A dictionary containing all attention processors used in the model with
            indexed by its weight name.
        """
        # set recursively
        processors = {}

        def fn_recursive_add_processors(name: str, module: torch.nn.Module, processors: Dict[str, AttentionProcessor]):
            if hasattr(module, "get_processor"):
                processors[f"{name}.processor"] = module.get_processor()

            for sub_name, child in module.named_children():
                fn_recursive_add_processors(f"{name}.{sub_name}", child, processors)

            return processors

        for name, module in self.named_children():
            fn_recursive_add_processors(name, module, processors)

        return processors

    # Copied from diffusers.models.unets.unet_2d_condition.UNet2DConditionModel.set_attn_processor
    def set_attn_processor(self, processor: Union[AttentionProcessor, Dict[str, AttentionProcessor]]):
        r"""
        Sets the attention processor to use to compute attention.

        Parameters:
            processor (`dict` of `AttentionProcessor` or only `AttentionProcessor`):
                The instantiated processor class or a dictionary of processor classes that will be set as the processor
                for **all** `Attention` layers.

                If `processor` is a dict, the key needs to define the path to the corresponding cross attention
                processor. This is strongly recommended when setting trainable attention processors.

        """
        count = len(self.attn_processors.keys())

        if isinstance(processor, dict) and len(processor) != count:
            raise ValueError(
                f"A dict of processors was passed, but the number of processors {len(processor)} does not match the"
                f" number of attention layers: {count}. Please make sure to pass {count} processor classes."
            )

        def fn_recursive_attn_processor(name: str, module: torch.nn.Module, processor):
            if hasattr(module, "set_processor"):
                if not isinstance(processor, dict):
                    module.set_processor(processor)
                else:
                    module.set_processor(processor.pop(f"{name}.processor"))

            for sub_name, child in module.named_children():
                fn_recursive_attn_processor(f"{name}.{sub_name}", child, processor)

        for name, module in self.named_children():
            fn_recursive_attn_processor(name, module, processor)

    # Copied from diffusers.models.unets.unet_2d_condition.UNet2DConditionModel.fuse_qkv_projections with FusedAttnProcessor2_0->FusedFluxAttnProcessor2_0
    def fuse_qkv_projections(self):
        """
        Enables fused QKV projections. For self-attention modules, all projection matrices (i.e., query, key, value)
        are fused. For cross-attention modules, key and value projection matrices are fused.

        <Tip warning={true}>

        This API is 🧪 experimental.

        </Tip>
        """
        self.original_attn_processors = None

        for _, attn_processor in self.attn_processors.items():
            if "Added" in str(attn_processor.__class__.__name__):
                raise ValueError("`fuse_qkv_projections()` is not supported for models having added KV projections.")

        self.original_attn_processors = self.attn_processors

        for module in self.modules():
            if isinstance(module, Attention):
                module.fuse_projections(fuse=True)

        self.set_attn_processor(FusedFluxAttnProcessor2_0())

    # Copied from diffusers.models.unets.unet_2d_condition.UNet2DConditionModel.unfuse_qkv_projections
    def unfuse_qkv_projections(self):
        """Disables the fused QKV projection if enabled.

        <Tip warning={true}>

        This API is 🧪 experimental.

        </Tip>

        """
        if self.original_attn_processors is not None:
            self.set_attn_processor(self.original_attn_processors)

    def forward(
        self,
        hidden_states: torch.Tensor,
        encoder_hidden_states: torch.Tensor = None,
        pooled_projections: 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,
        controlnet_block_samples=None,
        controlnet_single_block_samples=None,
        return_dict: bool = True,
        controlnet_blocks_repeat: bool = False,
        # adding a modulation conditioning, where an embedding is passed that can be used to modulate features of diffusion model 
        modulation_embeddings: Optional[torch.Tensor] = None, 
    ) -> Union[torch.Tensor, Transformer2DModelOutput]:
        """
        The [`FluxTransformer2DModelwithSliderConditioning`] forward method.

        Args:
            hidden_states (`torch.Tensor` of shape `(batch_size, image_sequence_length, in_channels)`):
                Input `hidden_states`.
            encoder_hidden_states (`torch.Tensor` of shape `(batch_size, text_sequence_length, joint_attention_dim)`):
                Conditional embeddings (embeddings computed from the input conditions such as prompts) to use.
            pooled_projections (`torch.Tensor` of shape `(batch_size, projection_dim)`): Embeddings projected
                from the embeddings of input conditions.
            timestep ( `torch.LongTensor`):
                Used to indicate denoising step.
            block_controlnet_hidden_states: (`list` of `torch.Tensor`):
                A list of tensors that if specified are added to the residuals of transformer blocks.
            joint_attention_kwargs (`dict`, *optional*):
                A kwargs dictionary that if specified is passed along to the `AttentionProcessor` as defined under
                `self.processor` in
                [diffusers.models.attention_processor](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/attention_processor.py).
            return_dict (`bool`, *optional*, defaults to `True`):
                Whether or not to return a [`~models.transformer_2d.Transformer2DModelOutput`] instead of a plain
                tuple.

        Returns:
            If `return_dict` is True, an [`~models.transformer_2d.Transformer2DModelOutput`] is returned, otherwise a
            `tuple` where the first element is the sample tensor.
        """
        
        # if modulation_embeddings is not None:
        #     print("working with modulation space conditioning ...") 
        #     print("modulation condn in main transformer call: {}".format(modulation_embeddings.shape)) 


        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

        if USE_PEFT_BACKEND:
            # weight the lora layers by setting `lora_scale` for each PEFT layer
            scale_lora_layers(self, lora_scale)
        else:
            if joint_attention_kwargs is not None and joint_attention_kwargs.get("scale", None) is not None:
                logger.warning(
                    "Passing `scale` via `joint_attention_kwargs` when not using the PEFT backend is ineffective."
                )

        hidden_states = self.x_embedder(hidden_states)

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

        temb = (
            self.time_text_embed(timestep, pooled_projections)
            if guidance is None
            else self.time_text_embed(timestep, guidance, pooled_projections)
        )

        # print("temb shape: {}".format(temb.shape)) # [1, 3072]

        # ------------------------ Logic to add the predicted embedding at the root of the modulation branch ------------------------- # 
        # modulation_embeddings = modulation_embeddings.squeeze(1)
        # scale_factor = 10
        # temb = temb + modulation_embeddings * scale_factor
        # ---------------------------------------------------------------------------------------------------------------------------- # 

        encoder_hidden_states = self.context_embedder(encoder_hidden_states)

        if txt_ids.ndim == 3:
            logger.warning(
                "Passing `txt_ids` 3d torch.Tensor is deprecated."
                "Please remove the batch dimension and pass it as a 2d torch Tensor"
            )
            txt_ids = txt_ids[0]
        if img_ids.ndim == 3:
            logger.warning(
                "Passing `img_ids` 3d torch.Tensor is deprecated."
                "Please remove the batch dimension and pass it as a 2d torch Tensor"
            )
            img_ids = img_ids[0]

        ids = torch.cat((txt_ids, img_ids), dim=0)
        image_rotary_emb = self.pos_embed(ids)

        if joint_attention_kwargs is not None and "ip_adapter_image_embeds" in joint_attention_kwargs:
            ip_adapter_image_embeds = joint_attention_kwargs.pop("ip_adapter_image_embeds")
            ip_hidden_states = self.encoder_hid_proj(ip_adapter_image_embeds)
            joint_attention_kwargs.update({"ip_hidden_states": ip_hidden_states})


        # Iterating over the transformer blocks that process the text and image conditionings separately and then will be combined later 
        for index_block, block in enumerate(self.transformer_blocks):
            if torch.is_grad_enabled() and self.gradient_checkpointing:

                # TODO: This we have to test and validate later, there is a possibility there is a bug in this and will not work for us.
                # # copied from syncd codebase for gradient checkpoint with new arugments 
                # def create_custom_forward(module, return_dict=None):
                #     def custom_forward(*inputs):
                #         if return_dict is not None:
                #             return module(*inputs, return_dict=return_dict)
                #         else:
                #             return module(*inputs)

                #     return custom_forward
                
                # new_kwargs = {
                #     "modulation_condn": modulation_condn, 
                # }
                
                # This line applies gradient checkpointing to the transformer block, allowing for reduced memory usage during training by recomputing intermediate activations in the backward pass.
                encoder_hidden_states, hidden_states = self._gradient_checkpointing_func(
                    block, 
                    hidden_states,
                    encoder_hidden_states,
                    temb,
                    image_rotary_emb,
                )

            else:
                # adding the modulation conditioning vector in the separate transformer blocks that will use it for adjusting the features 
                encoder_hidden_states, hidden_states = block(
                    hidden_states=hidden_states,
                    encoder_hidden_states=encoder_hidden_states,
                    temb=temb, # the temb vector is modified and a delta based on the new conditioning is added to it. 
                    image_rotary_emb=image_rotary_emb,
                    joint_attention_kwargs=joint_attention_kwargs,
                    # adding the modulation conditioning vector that can control the editing quality
                    modulation_condn = modulation_embeddings, # modulation_embeddings, | Not passing the modulation conditioning as it is already incroported in the temb vector 
                )

            # controlnet residual
            if controlnet_block_samples is not None:
                interval_control = len(self.transformer_blocks) / len(controlnet_block_samples)
                interval_control = int(np.ceil(interval_control))
                # For Xlabs ControlNet.
                if controlnet_blocks_repeat:
                    hidden_states = (
                        hidden_states + controlnet_block_samples[index_block % len(controlnet_block_samples)]
                    )
                else:
                    hidden_states = hidden_states + controlnet_block_samples[index_block // interval_control]
        hidden_states = torch.cat([encoder_hidden_states, hidden_states], dim=1)

        # single stream transformer blocks where the processing is happening in a single pass for the text and the image conditionings. 
        for index_block, block in enumerate(self.single_transformer_blocks):
            if torch.is_grad_enabled() and self.gradient_checkpointing:
                hidden_states = self._gradient_checkpointing_func(
                    block,
                    hidden_states,
                    temb,
                    image_rotary_emb,
                )

            else:
                hidden_states = block(
                    hidden_states=hidden_states,
                    temb=temb,
                    image_rotary_emb=image_rotary_emb,
                    joint_attention_kwargs=joint_attention_kwargs,
                )

            # controlnet residual
            if controlnet_single_block_samples is not None:
                interval_control = len(self.single_transformer_blocks) / len(controlnet_single_block_samples)
                interval_control = int(np.ceil(interval_control))
                hidden_states[:, encoder_hidden_states.shape[1] :, ...] = (
                    hidden_states[:, encoder_hidden_states.shape[1] :, ...]
                    + controlnet_single_block_samples[index_block // interval_control]
                )

        hidden_states = hidden_states[:, encoder_hidden_states.shape[1] :, ...]

        hidden_states = self.norm_out(hidden_states, temb)
        output = self.proj_out(hidden_states)

        if USE_PEFT_BACKEND:
            # remove `lora_scale` from each PEFT layer
            unscale_lora_layers(self, lora_scale)

        if not return_dict:
            return (output,)

        return Transformer2DModelOutput(sample=output)