update: FluxKontextInpaintPipeline support (#11820)

* update: FluxKontextInpaintPipeline support

* fix: Refactor code, remove mask_image_latents and ruff check

* feat: Add test case and fix with pytest

* Apply style fixes

* copies

---------
Co-authored-by: YiYi Xu <yixu310@gmail.com>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

src/diffusers/__init__.py

@@ -381,6 +381,7 @@ else:
             "FluxFillPipeline",
             "FluxImg2ImgPipeline",
             "FluxInpaintPipeline",
+            "FluxKontextInpaintPipeline",
             "FluxKontextPipeline",
             "FluxPipeline",
             "FluxPriorReduxPipeline",
@@ -975,6 +976,7 @@ if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
             FluxFillPipeline,
             FluxImg2ImgPipeline,
             FluxInpaintPipeline,
+            FluxKontextInpaintPipeline,
             FluxKontextPipeline,
             FluxPipeline,
             FluxPriorReduxPipeline,

src/diffusers/pipelines/__init__.py

@@ -141,6 +141,7 @@ else:
         "FluxPriorReduxPipeline",
         "ReduxImageEncoder",
         "FluxKontextPipeline",
+        "FluxKontextInpaintPipeline",
     ]
     _import_structure["audioldm"] = ["AudioLDMPipeline"]
     _import_structure["audioldm2"] = [
@@ -610,6 +611,7 @@ if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
             FluxFillPipeline,
             FluxImg2ImgPipeline,
             FluxInpaintPipeline,
+            FluxKontextInpaintPipeline,
             FluxKontextPipeline,
             FluxPipeline,
             FluxPriorReduxPipeline,

src/diffusers/pipelines/flux/__init__.py

@@ -34,6 +34,7 @@ else:
     _import_structure["pipeline_flux_img2img"] = ["FluxImg2ImgPipeline"]
     _import_structure["pipeline_flux_inpaint"] = ["FluxInpaintPipeline"]
     _import_structure["pipeline_flux_kontext"] = ["FluxKontextPipeline"]
+    _import_structure["pipeline_flux_kontext_inpaint"] = ["FluxKontextInpaintPipeline"]
     _import_structure["pipeline_flux_prior_redux"] = ["FluxPriorReduxPipeline"]
 if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
     try:
@@ -54,6 +55,7 @@ if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
         from .pipeline_flux_img2img import FluxImg2ImgPipeline
         from .pipeline_flux_inpaint import FluxInpaintPipeline
         from .pipeline_flux_kontext import FluxKontextPipeline
+        from .pipeline_flux_kontext_inpaint import FluxKontextInpaintPipeline
         from .pipeline_flux_prior_redux import FluxPriorReduxPipeline
 else:
     import sys

src/diffusers/pipelines/flux/pipeline_flux_kontext_inpaint.py

+# Copyright 2025 ZenAI. All rights reserved.
+# author: @vuongminh1907
+
+import inspect
+from typing import Any, Callable, Dict, List, Optional, Union
+
+import numpy as np
+import PIL.Image
+import torch
+from transformers import (
+    CLIPImageProcessor,
+    CLIPTextModel,
+    CLIPTokenizer,
+    CLIPVisionModelWithProjection,
+    T5EncoderModel,
+    T5TokenizerFast,
+)
+
+from ...image_processor import PipelineImageInput, VaeImageProcessor
+from ...loaders import FluxIPAdapterMixin, FluxLoraLoaderMixin, FromSingleFileMixin, TextualInversionLoaderMixin
+from ...models import AutoencoderKL, FluxTransformer2DModel
+from ...schedulers import FlowMatchEulerDiscreteScheduler
+from ...utils import (
+    USE_PEFT_BACKEND,
+    is_torch_xla_available,
+    logging,
+    replace_example_docstring,
+    scale_lora_layers,
+    unscale_lora_layers,
+)
+from ...utils.torch_utils import randn_tensor
+from ..pipeline_utils import DiffusionPipeline
+from .pipeline_output import FluxPipelineOutput
+
+
+if is_torch_xla_available():
+    import torch_xla.core.xla_model as xm
+
+    XLA_AVAILABLE = True
+else:
+    XLA_AVAILABLE = False
+
+
+logger = logging.get_logger(__name__)  # pylint: disable=invalid-name
+
+EXAMPLE_DOC_STRING = """
+    Examples:
+        # Inpainting with text only
+        ```py
+        >>> import torch
+        >>> from diffusers import FluxKontextInpaintPipeline
+        >>> from diffusers.utils import load_image
+
+        >>> prompt = "Change the yellow dinosaur to green one"
+        >>> img_url = (
+        ...     "https://github.com/ZenAI-Vietnam/Flux-Kontext-pipelines/blob/main/assets/dinosaur_input.jpeg?raw=true"
+        ... )
+        >>> mask_url = (
+        ...     "https://github.com/ZenAI-Vietnam/Flux-Kontext-pipelines/blob/main/assets/dinosaur_mask.png?raw=true"
+        ... )
+
+        >>> source = load_image(img_url)
+        >>> mask = load_image(mask_url)
+
+        >>> pipe = FluxKontextInpaintPipeline.from_pretrained(
+        ...     "black-forest-labs/FLUX.1-Kontext-dev", torch_dtype=torch.bfloat16
+        ... )
+        >>> pipe.to("cuda")
+
+        >>> image = pipe(prompt=prompt, image=source, mask_image=mask, strength=1.0).images[0]
+        >>> image.save("kontext_inpainting_normal.png")
+        ```
+
+        # Inpainting with image conditioning
+        ```py
+        >>> import torch
+        >>> from diffusers import FluxKontextInpaintPipeline
+        >>> from diffusers.utils import load_image
+
+        >>> pipe = FluxKontextInpaintPipeline.from_pretrained(
+        ...     "black-forest-labs/FLUX.1-Kontext-dev", torch_dtype=torch.bfloat16
+        ... )
+        >>> pipe.to("cuda")
+
+        >>> prompt = "Replace this ball"
+        >>> img_url = "https://images.pexels.com/photos/39362/the-ball-stadion-football-the-pitch-39362.jpeg?auto=compress&cs=tinysrgb&dpr=1&w=500"
+        >>> mask_url = (
+        ...     "https://github.com/ZenAI-Vietnam/Flux-Kontext-pipelines/blob/main/assets/ball_mask.png?raw=true"
+        ... )
+        >>> image_reference_url = (
+        ...     "https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcTah3x6OL_ECMBaZ5ZlJJhNsyC-OSMLWAI-xw&s"
+        ... )
+
+        >>> source = load_image(img_url)
+        >>> mask = load_image(mask_url)
+        >>> image_reference = load_image(image_reference_url)
+
+        >>> mask = pipe.mask_processor.blur(mask, blur_factor=12)
+        >>> image = pipe(
+        ...     prompt=prompt, image=source, mask_image=mask, image_reference=image_reference, strength=1.0
+        ... ).images[0]
+        >>> image.save("kontext_inpainting_ref.png")
+        ```
+"""
+
+PREFERRED_KONTEXT_RESOLUTIONS = [
+    (672, 1568),
+    (688, 1504),
+    (720, 1456),
+    (752, 1392),
+    (800, 1328),
+    (832, 1248),
+    (880, 1184),
+    (944, 1104),
+    (1024, 1024),
+    (1104, 944),
+    (1184, 880),
+    (1248, 832),
+    (1328, 800),
+    (1392, 752),
+    (1456, 720),
+    (1504, 688),
+    (1568, 672),
+]
+
+
+def calculate_shift(
+    image_seq_len,
+    base_seq_len: int = 256,
+    max_seq_len: int = 4096,
+    base_shift: float = 0.5,
+    max_shift: float = 1.15,
+):
+    m = (max_shift - base_shift) / (max_seq_len - base_seq_len)
+    b = base_shift - m * base_seq_len
+    mu = image_seq_len * m + b
+    return mu
+
+
+# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.retrieve_timesteps
+def retrieve_timesteps(
+    scheduler,
+    num_inference_steps: Optional[int] = None,
+    device: Optional[Union[str, torch.device]] = None,
+    timesteps: Optional[List[int]] = None,
+    sigmas: Optional[List[float]] = None,
+    **kwargs,
+):
+    r"""
+    Calls the scheduler's `set_timesteps` method and retrieves timesteps from the scheduler after the call. Handles
+    custom timesteps. Any kwargs will be supplied to `scheduler.set_timesteps`.
+
+    Args:
+        scheduler (`SchedulerMixin`):
+            The scheduler to get timesteps from.
+        num_inference_steps (`int`):
+            The number of diffusion steps used when generating samples with a pre-trained model. If used, `timesteps`
+            must be `None`.
+        device (`str` or `torch.device`, *optional*):
+            The device to which the timesteps should be moved to. If `None`, the timesteps are not moved.
+        timesteps (`List[int]`, *optional*):
+            Custom timesteps used to override the timestep spacing strategy of the scheduler. If `timesteps` is passed,
+            `num_inference_steps` and `sigmas` must be `None`.
+        sigmas (`List[float]`, *optional*):
+            Custom sigmas used to override the timestep spacing strategy of the scheduler. If `sigmas` is passed,
+            `num_inference_steps` and `timesteps` must be `None`.
+
+    Returns:
+        `Tuple[torch.Tensor, int]`: A tuple where the first element is the timestep schedule from the scheduler and the
+        second element is the number of inference steps.
+    """
+    if timesteps is not None and sigmas is not None:
+        raise ValueError("Only one of `timesteps` or `sigmas` can be passed. Please choose one to set custom values")
+    if timesteps is not None:
+        accepts_timesteps = "timesteps" in set(inspect.signature(scheduler.set_timesteps).parameters.keys())
+        if not accepts_timesteps:
+            raise ValueError(
+                f"The current scheduler class {scheduler.__class__}'s `set_timesteps` does not support custom"
+                f" timestep schedules. Please check whether you are using the correct scheduler."
+            )
+        scheduler.set_timesteps(timesteps=timesteps, device=device, **kwargs)
+        timesteps = scheduler.timesteps
+        num_inference_steps = len(timesteps)
+    elif sigmas is not None:
+        accept_sigmas = "sigmas" in set(inspect.signature(scheduler.set_timesteps).parameters.keys())
+        if not accept_sigmas:
+            raise ValueError(
+                f"The current scheduler class {scheduler.__class__}'s `set_timesteps` does not support custom"
+                f" sigmas schedules. Please check whether you are using the correct scheduler."
+            )
+        scheduler.set_timesteps(sigmas=sigmas, device=device, **kwargs)
+        timesteps = scheduler.timesteps
+        num_inference_steps = len(timesteps)
+    else:
+        scheduler.set_timesteps(num_inference_steps, device=device, **kwargs)
+        timesteps = scheduler.timesteps
+    return timesteps, num_inference_steps
+
+
+# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion_img2img.retrieve_latents
+def retrieve_latents(
+    encoder_output: torch.Tensor, generator: Optional[torch.Generator] = None, sample_mode: str = "sample"
+):
+    if hasattr(encoder_output, "latent_dist") and sample_mode == "sample":
+        return encoder_output.latent_dist.sample(generator)
+    elif hasattr(encoder_output, "latent_dist") and sample_mode == "argmax":
+        return encoder_output.latent_dist.mode()
+    elif hasattr(encoder_output, "latents"):
+        return encoder_output.latents
+    else:
+        raise AttributeError("Could not access latents of provided encoder_output")
+
+
+class FluxKontextInpaintPipeline(
+    DiffusionPipeline,
+    FluxLoraLoaderMixin,
+    FromSingleFileMixin,
+    TextualInversionLoaderMixin,
+    FluxIPAdapterMixin,
+):
+    r"""
+    The Flux Kontext pipeline for text-to-image generation.
+
+    Reference: https://blackforestlabs.ai/announcing-black-forest-labs/
+
+    Args:
+        transformer ([`FluxTransformer2DModel`]):
+            Conditional Transformer (MMDiT) architecture to denoise the encoded image latents.
+        scheduler ([`FlowMatchEulerDiscreteScheduler`]):
+            A scheduler to be used in combination with `transformer` to denoise the encoded image latents.
+        vae ([`AutoencoderKL`]):
+            Variational Auto-Encoder (VAE) Model to encode and decode images to and from latent representations.
+        text_encoder ([`CLIPTextModel`]):
+            [CLIP](https://huggingface.co/docs/transformers/model_doc/clip#transformers.CLIPTextModel), specifically
+            the [clip-vit-large-patch14](https://huggingface.co/openai/clip-vit-large-patch14) variant.
+        text_encoder_2 ([`T5EncoderModel`]):
+            [T5](https://huggingface.co/docs/transformers/en/model_doc/t5#transformers.T5EncoderModel), specifically
+            the [google/t5-v1_1-xxl](https://huggingface.co/google/t5-v1_1-xxl) variant.
+        tokenizer (`CLIPTokenizer`):
+            Tokenizer of class
+            [CLIPTokenizer](https://huggingface.co/docs/transformers/en/model_doc/clip#transformers.CLIPTokenizer).
+        tokenizer_2 (`T5TokenizerFast`):
+            Second Tokenizer of class
+            [T5TokenizerFast](https://huggingface.co/docs/transformers/en/model_doc/t5#transformers.T5TokenizerFast).
+    """
+
+    model_cpu_offload_seq = "text_encoder->text_encoder_2->image_encoder->transformer->vae"
+    _optional_components = ["image_encoder", "feature_extractor"]
+    _callback_tensor_inputs = ["latents", "prompt_embeds"]
+
+    def __init__(
+        self,
+        scheduler: FlowMatchEulerDiscreteScheduler,
+        vae: AutoencoderKL,
+        text_encoder: CLIPTextModel,
+        tokenizer: CLIPTokenizer,
+        text_encoder_2: T5EncoderModel,
+        tokenizer_2: T5TokenizerFast,
+        transformer: FluxTransformer2DModel,
+        image_encoder: CLIPVisionModelWithProjection = None,
+        feature_extractor: CLIPImageProcessor = None,
+    ):
+        super().__init__()
+
+        self.register_modules(
+            vae=vae,
+            text_encoder=text_encoder,
+            text_encoder_2=text_encoder_2,
+            tokenizer=tokenizer,
+            tokenizer_2=tokenizer_2,
+            transformer=transformer,
+            scheduler=scheduler,
+            image_encoder=image_encoder,
+            feature_extractor=feature_extractor,
+        )
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
+        # Flux latents are turned into 2x2 patches and packed. This means the latent width and height has to be divisible
+        # by the patch size. So the vae scale factor is multiplied by the patch size to account for this
+        self.latent_channels = self.vae.config.latent_channels if getattr(self, "vae", None) else 16
+        self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor * 2)
+
+        self.mask_processor = VaeImageProcessor(
+            vae_scale_factor=self.vae_scale_factor * 2,
+            vae_latent_channels=self.latent_channels,
+            do_normalize=False,
+            do_binarize=True,
+            do_convert_grayscale=True,
+        )
+
+        self.tokenizer_max_length = (
+            self.tokenizer.model_max_length if hasattr(self, "tokenizer") and self.tokenizer is not None else 77
+        )
+        self.default_sample_size = 128
+
+    # Copied from diffusers.pipelines.flux.pipeline_flux.FluxPipeline._get_t5_prompt_embeds
+    def _get_t5_prompt_embeds(
+        self,
+        prompt: Union[str, List[str]] = None,
+        num_images_per_prompt: int = 1,
+        max_sequence_length: int = 512,
+        device: Optional[torch.device] = None,
+        dtype: Optional[torch.dtype] = None,
+    ):
+        device = device or self._execution_device
+        dtype = dtype or self.text_encoder.dtype
+
+        prompt = [prompt] if isinstance(prompt, str) else prompt
+        batch_size = len(prompt)
+
+        if isinstance(self, TextualInversionLoaderMixin):
+            prompt = self.maybe_convert_prompt(prompt, self.tokenizer_2)
+
+        text_inputs = self.tokenizer_2(
+            prompt,
+            padding="max_length",
+            max_length=max_sequence_length,
+            truncation=True,
+            return_length=False,
+            return_overflowing_tokens=False,
+            return_tensors="pt",
+        )
+        text_input_ids = text_inputs.input_ids
+        untruncated_ids = self.tokenizer_2(prompt, padding="longest", return_tensors="pt").input_ids
+
+        if untruncated_ids.shape[-1] >= text_input_ids.shape[-1] and not torch.equal(text_input_ids, untruncated_ids):
+            removed_text = self.tokenizer_2.batch_decode(untruncated_ids[:, self.tokenizer_max_length - 1 : -1])
+            logger.warning(
+                "The following part of your input was truncated because `max_sequence_length` is set to "
+                f" {max_sequence_length} tokens: {removed_text}"
+            )
+
+        prompt_embeds = self.text_encoder_2(text_input_ids.to(device), output_hidden_states=False)[0]
+
+        dtype = self.text_encoder_2.dtype
+        prompt_embeds = prompt_embeds.to(dtype=dtype, device=device)
+
+        _, seq_len, _ = prompt_embeds.shape
+
+        # duplicate text embeddings and attention mask for each generation per prompt, using mps friendly method
+        prompt_embeds = prompt_embeds.repeat(1, num_images_per_prompt, 1)
+        prompt_embeds = prompt_embeds.view(batch_size * num_images_per_prompt, seq_len, -1)
+
+        return prompt_embeds
+
+    # Copied from diffusers.pipelines.flux.pipeline_flux.FluxPipeline._get_clip_prompt_embeds
+    def _get_clip_prompt_embeds(
+        self,
+        prompt: Union[str, List[str]],
+        num_images_per_prompt: int = 1,
+        device: Optional[torch.device] = None,
+    ):
+        device = device or self._execution_device
+
+        prompt = [prompt] if isinstance(prompt, str) else prompt
+        batch_size = len(prompt)
+
+        if isinstance(self, TextualInversionLoaderMixin):
+            prompt = self.maybe_convert_prompt(prompt, self.tokenizer)
+
+        text_inputs = self.tokenizer(
+            prompt,
+            padding="max_length",
+            max_length=self.tokenizer_max_length,
+            truncation=True,
+            return_overflowing_tokens=False,
+            return_length=False,
+            return_tensors="pt",
+        )
+
+        text_input_ids = text_inputs.input_ids
+        untruncated_ids = self.tokenizer(prompt, padding="longest", return_tensors="pt").input_ids
+        if untruncated_ids.shape[-1] >= text_input_ids.shape[-1] and not torch.equal(text_input_ids, untruncated_ids):
+            removed_text = self.tokenizer.batch_decode(untruncated_ids[:, self.tokenizer_max_length - 1 : -1])
+            logger.warning(
+                "The following part of your input was truncated because CLIP can only handle sequences up to"
+                f" {self.tokenizer_max_length} tokens: {removed_text}"
+            )
+        prompt_embeds = self.text_encoder(text_input_ids.to(device), output_hidden_states=False)
+
+        # Use pooled output of CLIPTextModel
+        prompt_embeds = prompt_embeds.pooler_output
+        prompt_embeds = prompt_embeds.to(dtype=self.text_encoder.dtype, device=device)
+
+        # duplicate text embeddings for each generation per prompt, using mps friendly method
+        prompt_embeds = prompt_embeds.repeat(1, num_images_per_prompt)
+        prompt_embeds = prompt_embeds.view(batch_size * num_images_per_prompt, -1)
+
+        return prompt_embeds
+
+    # Copied from diffusers.pipelines.flux.pipeline_flux.FluxPipeline.encode_prompt
+    def encode_prompt(
+        self,
+        prompt: Union[str, List[str]],
+        prompt_2: Union[str, List[str]],
+        device: Optional[torch.device] = None,
+        num_images_per_prompt: int = 1,
+        prompt_embeds: Optional[torch.FloatTensor] = None,
+        pooled_prompt_embeds: Optional[torch.FloatTensor] = None,
+        max_sequence_length: int = 512,
+        lora_scale: Optional[float] = None,
+    ):
+        r"""
+
+        Args:
+            prompt (`str` or `List[str]`, *optional*):
+                prompt to be encoded
+            prompt_2 (`str` or `List[str]`, *optional*):
+                The prompt or prompts to be sent to the `tokenizer_2` and `text_encoder_2`. If not defined, `prompt` is
+                used in all text-encoders
+            device: (`torch.device`):
+                torch device
+            num_images_per_prompt (`int`):
+                number of images that should be generated per prompt
+            prompt_embeds (`torch.FloatTensor`, *optional*):
+                Pre-generated text embeddings. Can be used to easily tweak text inputs, *e.g.* prompt weighting. If not
+                provided, text embeddings will be generated from `prompt` input argument.
+            pooled_prompt_embeds (`torch.FloatTensor`, *optional*):
+                Pre-generated pooled text embeddings. Can be used to easily tweak text inputs, *e.g.* prompt weighting.
+                If not provided, pooled text embeddings will be generated from `prompt` input argument.
+            lora_scale (`float`, *optional*):
+                A lora scale that will be applied to all LoRA layers of the text encoder if LoRA layers are loaded.
+        """
+        device = device or self._execution_device
+
+        # set lora scale so that monkey patched LoRA
+        # function of text encoder can correctly access it
+        if lora_scale is not None and isinstance(self, FluxLoraLoaderMixin):
+            self._lora_scale = lora_scale
+
+            # dynamically adjust the LoRA scale
+            if self.text_encoder is not None and USE_PEFT_BACKEND:
+                scale_lora_layers(self.text_encoder, lora_scale)
+            if self.text_encoder_2 is not None and USE_PEFT_BACKEND:
+                scale_lora_layers(self.text_encoder_2, lora_scale)
+
+        prompt = [prompt] if isinstance(prompt, str) else prompt
+
+        if prompt_embeds is None:
+            prompt_2 = prompt_2 or prompt
+            prompt_2 = [prompt_2] if isinstance(prompt_2, str) else prompt_2
+
+            # We only use the pooled prompt output from the CLIPTextModel
+            pooled_prompt_embeds = self._get_clip_prompt_embeds(
+                prompt=prompt,
+                device=device,
+                num_images_per_prompt=num_images_per_prompt,
+            )
+            prompt_embeds = self._get_t5_prompt_embeds(
+                prompt=prompt_2,
+                num_images_per_prompt=num_images_per_prompt,
+                max_sequence_length=max_sequence_length,
+                device=device,
+            )
+
+        if self.text_encoder is not None:
+            if isinstance(self, FluxLoraLoaderMixin) and USE_PEFT_BACKEND:
+                # Retrieve the original scale by scaling back the LoRA layers
+                unscale_lora_layers(self.text_encoder, lora_scale)
+
+        if self.text_encoder_2 is not None:
+            if isinstance(self, FluxLoraLoaderMixin) and USE_PEFT_BACKEND:
+                # Retrieve the original scale by scaling back the LoRA layers
+                unscale_lora_layers(self.text_encoder_2, lora_scale)
+
+        dtype = self.text_encoder.dtype if self.text_encoder is not None else self.transformer.dtype
+        text_ids = torch.zeros(prompt_embeds.shape[1], 3).to(device=device, dtype=dtype)
+
+        return prompt_embeds, pooled_prompt_embeds, text_ids
+
+    # Copied from diffusers.pipelines.flux.pipeline_flux.FluxPipeline.encode_image
+    def encode_image(self, image, device, num_images_per_prompt):
+        dtype = next(self.image_encoder.parameters()).dtype
+
+        if not isinstance(image, torch.Tensor):
+            image = self.feature_extractor(image, return_tensors="pt").pixel_values
+
+        image = image.to(device=device, dtype=dtype)
+        image_embeds = self.image_encoder(image).image_embeds
+        image_embeds = image_embeds.repeat_interleave(num_images_per_prompt, dim=0)
+        return image_embeds
+
+    # Copied from diffusers.pipelines.flux.pipeline_flux.FluxPipeline.prepare_ip_adapter_image_embeds
+    def prepare_ip_adapter_image_embeds(
+        self, ip_adapter_image, ip_adapter_image_embeds, device, num_images_per_prompt
+    ):
+        image_embeds = []
+        if ip_adapter_image_embeds is None:
+            if not isinstance(ip_adapter_image, list):
+                ip_adapter_image = [ip_adapter_image]
+
+            if len(ip_adapter_image) != self.transformer.encoder_hid_proj.num_ip_adapters:
+                raise ValueError(
+                    f"`ip_adapter_image` must have same length as the number of IP Adapters. Got {len(ip_adapter_image)} images and {self.transformer.encoder_hid_proj.num_ip_adapters} IP Adapters."
+                )
+
+            for single_ip_adapter_image in ip_adapter_image:
+                single_image_embeds = self.encode_image(single_ip_adapter_image, device, 1)
+                image_embeds.append(single_image_embeds[None, :])
+        else:
+            if not isinstance(ip_adapter_image_embeds, list):
+                ip_adapter_image_embeds = [ip_adapter_image_embeds]
+
+            if len(ip_adapter_image_embeds) != self.transformer.encoder_hid_proj.num_ip_adapters:
+                raise ValueError(
+                    f"`ip_adapter_image_embeds` must have same length as the number of IP Adapters. Got {len(ip_adapter_image_embeds)} image embeds and {self.transformer.encoder_hid_proj.num_ip_adapters} IP Adapters."
+                )
+
+            for single_image_embeds in ip_adapter_image_embeds:
+                image_embeds.append(single_image_embeds)
+
+        ip_adapter_image_embeds = []
+        for single_image_embeds in image_embeds:
+            single_image_embeds = torch.cat([single_image_embeds] * num_images_per_prompt, dim=0)
+            single_image_embeds = single_image_embeds.to(device=device)
+            ip_adapter_image_embeds.append(single_image_embeds)
+
+        return ip_adapter_image_embeds
+
+    # Copied from diffusers.pipelines.stable_diffusion_3.pipeline_stable_diffusion_3_img2img.StableDiffusion3Img2ImgPipeline.get_timesteps
+    def get_timesteps(self, num_inference_steps, strength, device):
+        # get the original timestep using init_timestep
+        init_timestep = min(num_inference_steps * strength, num_inference_steps)
+
+        t_start = int(max(num_inference_steps - init_timestep, 0))
+        timesteps = self.scheduler.timesteps[t_start * self.scheduler.order :]
+        if hasattr(self.scheduler, "set_begin_index"):
+            self.scheduler.set_begin_index(t_start * self.scheduler.order)
+
+        return timesteps, num_inference_steps - t_start
+
+    # Copied from diffusers.pipelines.flux.pipeline_flux_inpaint.FluxInpaintPipeline.check_inputs
+    def check_inputs(
+        self,
+        prompt,
+        prompt_2,
+        image,
+        mask_image,
+        strength,
+        height,
+        width,
+        output_type,
+        negative_prompt=None,
+        negative_prompt_2=None,
+        prompt_embeds=None,
+        negative_prompt_embeds=None,
+        pooled_prompt_embeds=None,
+        negative_pooled_prompt_embeds=None,
+        callback_on_step_end_tensor_inputs=None,
+        padding_mask_crop=None,
+        max_sequence_length=None,
+    ):
+        if strength < 0 or strength > 1:
+            raise ValueError(f"The value of strength should in [0.0, 1.0] but is {strength}")
+
+        if height % (self.vae_scale_factor * 2) != 0 or width % (self.vae_scale_factor * 2) != 0:
+            logger.warning(
+                f"`height` and `width` have to be divisible by {self.vae_scale_factor * 2} but are {height} and {width}. Dimensions will be resized accordingly"
+            )
+
+        if callback_on_step_end_tensor_inputs is not None and not all(
+            k in self._callback_tensor_inputs for k in callback_on_step_end_tensor_inputs
+        ):
+            raise ValueError(
+                f"`callback_on_step_end_tensor_inputs` has to be in {self._callback_tensor_inputs}, but found {[k for k in callback_on_step_end_tensor_inputs if k not in self._callback_tensor_inputs]}"
+            )
+
+        if prompt is not None and prompt_embeds is not None:
+            raise ValueError(
+                f"Cannot forward both `prompt`: {prompt} and `prompt_embeds`: {prompt_embeds}. Please make sure to"
+                " only forward one of the two."
+            )
+        elif prompt_2 is not None and prompt_embeds is not None:
+            raise ValueError(
+                f"Cannot forward both `prompt_2`: {prompt_2} and `prompt_embeds`: {prompt_embeds}. Please make sure to"
+                " only forward one of the two."
+            )
+        elif prompt is None and prompt_embeds is None:
+            raise ValueError(
+                "Provide either `prompt` or `prompt_embeds`. Cannot leave both `prompt` and `prompt_embeds` undefined."
+            )
+        elif prompt is not None and (not isinstance(prompt, str) and not isinstance(prompt, list)):
+            raise ValueError(f"`prompt` has to be of type `str` or `list` but is {type(prompt)}")
+        elif prompt_2 is not None and (not isinstance(prompt_2, str) and not isinstance(prompt_2, list)):
+            raise ValueError(f"`prompt_2` has to be of type `str` or `list` but is {type(prompt_2)}")
+
+        if negative_prompt is not None and negative_prompt_embeds is not None:
+            raise ValueError(
+                f"Cannot forward both `negative_prompt`: {negative_prompt} and `negative_prompt_embeds`:"
+                f" {negative_prompt_embeds}. Please make sure to only forward one of the two."
+            )
+        elif negative_prompt_2 is not None and negative_prompt_embeds is not None:
+            raise ValueError(
+                f"Cannot forward both `negative_prompt_2`: {negative_prompt_2} and `negative_prompt_embeds`:"
+                f" {negative_prompt_embeds}. Please make sure to only forward one of the two."
+            )
+
+        if prompt_embeds is not None and negative_prompt_embeds is not None:
+            if prompt_embeds.shape != negative_prompt_embeds.shape:
+                raise ValueError(
+                    "`prompt_embeds` and `negative_prompt_embeds` must have the same shape when passed directly, but"
+                    f" got: `prompt_embeds` {prompt_embeds.shape} != `negative_prompt_embeds`"
+                    f" {negative_prompt_embeds.shape}."
+                )
+
+        if prompt_embeds is not None and pooled_prompt_embeds is None:
+            raise ValueError(
+                "If `prompt_embeds` are provided, `pooled_prompt_embeds` also have to be passed. Make sure to generate `pooled_prompt_embeds` from the same text encoder that was used to generate `prompt_embeds`."
+            )
+        if negative_prompt_embeds is not None and negative_pooled_prompt_embeds is None:
+            raise ValueError(
+                "If `negative_prompt_embeds` are provided, `negative_pooled_prompt_embeds` also have to be passed. Make sure to generate `negative_pooled_prompt_embeds` from the same text encoder that was used to generate `negative_prompt_embeds`."
+            )
+
+        if padding_mask_crop is not None:
+            if not isinstance(image, PIL.Image.Image):
+                raise ValueError(
+                    f"The image should be a PIL image when inpainting mask crop, but is of type {type(image)}."
+                )
+            if not isinstance(mask_image, PIL.Image.Image):
+                raise ValueError(
+                    f"The mask image should be a PIL image when inpainting mask crop, but is of type"
+                    f" {type(mask_image)}."
+                )
+            if output_type != "pil":
+                raise ValueError(f"The output type should be PIL when inpainting mask crop, but is {output_type}.")
+
+        if max_sequence_length is not None and max_sequence_length > 512:
+            raise ValueError(f"`max_sequence_length` cannot be greater than 512 but is {max_sequence_length}")
+
+    @staticmethod
+    # Copied from diffusers.pipelines.flux.pipeline_flux.FluxPipeline._prepare_latent_image_ids
+    def _prepare_latent_image_ids(batch_size, height, width, device, dtype):
+        latent_image_ids = torch.zeros(height, width, 3)
+        latent_image_ids[..., 1] = latent_image_ids[..., 1] + torch.arange(height)[:, None]
+        latent_image_ids[..., 2] = latent_image_ids[..., 2] + torch.arange(width)[None, :]
+
+        latent_image_id_height, latent_image_id_width, latent_image_id_channels = latent_image_ids.shape
+
+        latent_image_ids = latent_image_ids.reshape(
+            latent_image_id_height * latent_image_id_width, latent_image_id_channels
+        )
+
+        return latent_image_ids.to(device=device, dtype=dtype)
+
+    @staticmethod
+    # Copied from diffusers.pipelines.flux.pipeline_flux.FluxPipeline._pack_latents
+    def _pack_latents(latents, batch_size, num_channels_latents, height, width):
+        latents = latents.view(batch_size, num_channels_latents, height // 2, 2, width // 2, 2)
+        latents = latents.permute(0, 2, 4, 1, 3, 5)
+        latents = latents.reshape(batch_size, (height // 2) * (width // 2), num_channels_latents * 4)
+
+        return latents
+
+    @staticmethod
+    # Copied from diffusers.pipelines.flux.pipeline_flux.FluxPipeline._unpack_latents
+    def _unpack_latents(latents, height, width, vae_scale_factor):
+        batch_size, num_patches, channels = latents.shape
+
+        # VAE applies 8x compression on images but we must also account for packing which requires
+        # latent height and width to be divisible by 2.
+        height = 2 * (int(height) // (vae_scale_factor * 2))
+        width = 2 * (int(width) // (vae_scale_factor * 2))
+
+        latents = latents.view(batch_size, height // 2, width // 2, channels // 4, 2, 2)
+        latents = latents.permute(0, 3, 1, 4, 2, 5)
+
+        latents = latents.reshape(batch_size, channels // (2 * 2), height, width)
+
+        return latents
+
+    def _encode_vae_image(self, image: torch.Tensor, generator: torch.Generator):
+        if isinstance(generator, list):
+            image_latents = [
+                retrieve_latents(self.vae.encode(image[i : i + 1]), generator=generator[i], sample_mode="argmax")
+                for i in range(image.shape[0])
+            ]
+            image_latents = torch.cat(image_latents, dim=0)
+        else:
+            image_latents = retrieve_latents(self.vae.encode(image), generator=generator, sample_mode="argmax")
+
+        image_latents = (image_latents - self.vae.config.shift_factor) * self.vae.config.scaling_factor
+
+        return image_latents
+
+    # Copied from diffusers.pipelines.flux.pipeline_flux.FluxPipeline.enable_vae_slicing
+    def enable_vae_slicing(self):
+        r"""
+        Enable sliced VAE decoding. When this option is enabled, the VAE will split the input tensor in slices to
+        compute decoding in several steps. This is useful to save some memory and allow larger batch sizes.
+        """
+        self.vae.enable_slicing()
+
+    # Copied from diffusers.pipelines.flux.pipeline_flux.FluxPipeline.disable_vae_slicing
+    def disable_vae_slicing(self):
+        r"""
+        Disable sliced VAE decoding. If `enable_vae_slicing` was previously enabled, this method will go back to
+        computing decoding in one step.
+        """
+        self.vae.disable_slicing()
+
+    # Copied from diffusers.pipelines.flux.pipeline_flux.FluxPipeline.enable_vae_tiling
+    def enable_vae_tiling(self):
+        r"""
+        Enable tiled VAE decoding. When this option is enabled, the VAE will split the input tensor into tiles to
+        compute decoding and encoding in several steps. This is useful for saving a large amount of memory and to allow
+        processing larger images.
+        """
+        self.vae.enable_tiling()
+
+    # Copied from diffusers.pipelines.flux.pipeline_flux.FluxPipeline.disable_vae_tiling
+    def disable_vae_tiling(self):
+        r"""
+        Disable tiled VAE decoding. If `enable_vae_tiling` was previously enabled, this method will go back to
+        computing decoding in one step.
+        """
+        self.vae.disable_tiling()
+
+    def prepare_latents(
+        self,
+        image: Optional[torch.Tensor],
+        timestep: int,
+        batch_size: int,
+        num_channels_latents: int,
+        height: int,
+        width: int,
+        dtype: torch.dtype,
+        device: torch.device,
+        generator: Optional[Union[torch.Generator, List[torch.Generator]]] = None,
+        latents: Optional[torch.Tensor] = None,
+        image_reference: Optional[torch.Tensor] = None,
+    ):
+        if isinstance(generator, list) and len(generator) != batch_size:
+            raise ValueError(
+                f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"
+                f" size of {batch_size}. Make sure the batch size matches the length of the generators."
+            )
+
+        # VAE applies 8x compression on images but we must also account for packing which requires
+        # latent height and width to be divisible by 2.
+        height = 2 * (int(height) // (self.vae_scale_factor * 2))
+        width = 2 * (int(width) // (self.vae_scale_factor * 2))
+        shape = (batch_size, num_channels_latents, height, width)
+
+        # Prepare image latents
+        image_latents = image_ids = None
+        if image is not None:
+            image = image.to(device=device, dtype=dtype)
+            if image.shape[1] != self.latent_channels:
+                image_latents = self._encode_vae_image(image=image, generator=generator)
+            else:
+                image_latents = image
+            if batch_size > image_latents.shape[0] and batch_size % image_latents.shape[0] == 0:
+                # expand init_latents for batch_size
+                additional_image_per_prompt = batch_size // image_latents.shape[0]
+                image_latents = torch.cat([image_latents] * additional_image_per_prompt, dim=0)
+            elif batch_size > image_latents.shape[0] and batch_size % image_latents.shape[0] != 0:
+                raise ValueError(
+                    f"Cannot duplicate `image` of batch size {image_latents.shape[0]} to {batch_size} text prompts."
+                )
+            else:
+                image_latents = torch.cat([image_latents], dim=0)
+
+        # Prepare image reference latents
+        image_reference_latents = image_reference_ids = None
+        if image_reference is not None:
+            image_reference = image_reference.to(device=device, dtype=dtype)
+            if image_reference.shape[1] != self.latent_channels:
+                image_reference_latents = self._encode_vae_image(image=image_reference, generator=generator)
+            else:
+                image_reference_latents = image_reference
+            if batch_size > image_reference_latents.shape[0] and batch_size % image_reference_latents.shape[0] == 0:
+                # expand init_latents for batch_size
+                additional_image_per_prompt = batch_size // image_reference_latents.shape[0]
+                image_reference_latents = torch.cat([image_reference_latents] * additional_image_per_prompt, dim=0)
+            elif batch_size > image_reference_latents.shape[0] and batch_size % image_reference_latents.shape[0] != 0:
+                raise ValueError(
+                    f"Cannot duplicate `image_reference` of batch size {image_reference_latents.shape[0]} to {batch_size} text prompts."
+                )
+            else:
+                image_reference_latents = torch.cat([image_reference_latents], dim=0)
+
+        latent_ids = self._prepare_latent_image_ids(batch_size, height // 2, width // 2, device, dtype)
+
+        if latents is None:
+            noise = randn_tensor(shape, generator=generator, device=device, dtype=dtype)
+            latents = self.scheduler.scale_noise(image_latents, timestep, noise)
+        else:
+            noise = latents.to(device=device, dtype=dtype)
+            latents = noise
+
+        image_latent_height, image_latent_width = image_latents.shape[2:]
+        image_latents = self._pack_latents(
+            image_latents, batch_size, num_channels_latents, image_latent_height, image_latent_width
+        )
+        image_ids = self._prepare_latent_image_ids(
+            batch_size, image_latent_height // 2, image_latent_width // 2, device, dtype
+        )
+        # image ids are the same as latent ids with the first dimension set to 1 instead of 0
+        image_ids[..., 0] = 1
+
+        if image_reference_latents is not None:
+            image_reference_latent_height, image_reference_latent_width = image_reference_latents.shape[2:]
+            image_reference_latents = self._pack_latents(
+                image_reference_latents,
+                batch_size,
+                num_channels_latents,
+                image_reference_latent_height,
+                image_reference_latent_width,
+            )
+            image_reference_ids = self._prepare_latent_image_ids(
+                batch_size, image_reference_latent_height // 2, image_reference_latent_width // 2, device, dtype
+            )
+            # image_reference_ids are the same as latent ids with the first dimension set to 1 instead of 0
+            image_reference_ids[..., 0] = 1
+
+        noise = self._pack_latents(noise, batch_size, num_channels_latents, height, width)
+        latents = self._pack_latents(latents, batch_size, num_channels_latents, height, width)
+
+        return latents, image_latents, image_reference_latents, latent_ids, image_ids, image_reference_ids, noise
+
+    # Copied from diffusers.pipelines.flux.pipeline_flux_inpaint.FluxInpaintPipeline.prepare_mask_latents
+    def prepare_mask_latents(
+        self,
+        mask,
+        masked_image,
+        batch_size,
+        num_channels_latents,
+        num_images_per_prompt,
+        height,
+        width,
+        dtype,
+        device,
+        generator,
+    ):
+        # VAE applies 8x compression on images but we must also account for packing which requires
+        # latent height and width to be divisible by 2.
+        height = 2 * (int(height) // (self.vae_scale_factor * 2))
+        width = 2 * (int(width) // (self.vae_scale_factor * 2))
+        # resize the mask to latents shape as we concatenate the mask to the latents
+        # we do that before converting to dtype to avoid breaking in case we're using cpu_offload
+        # and half precision
+        mask = torch.nn.functional.interpolate(mask, size=(height, width))
+        mask = mask.to(device=device, dtype=dtype)
+
+        batch_size = batch_size * num_images_per_prompt
+
+        masked_image = masked_image.to(device=device, dtype=dtype)
+
+        if masked_image.shape[1] == 16:
+            masked_image_latents = masked_image
+        else:
+            masked_image_latents = retrieve_latents(self.vae.encode(masked_image), generator=generator)
+
+            masked_image_latents = (
+                masked_image_latents - self.vae.config.shift_factor
+            ) * self.vae.config.scaling_factor
+
+        # duplicate mask and masked_image_latents for each generation per prompt, using mps friendly method
+        if mask.shape[0] < batch_size:
+            if not batch_size % mask.shape[0] == 0:
+                raise ValueError(
+                    "The passed mask and the required batch size don't match. Masks are supposed to be duplicated to"
+                    f" a total batch size of {batch_size}, but {mask.shape[0]} masks were passed. Make sure the number"
+                    " of masks that you pass is divisible by the total requested batch size."
+                )
+            mask = mask.repeat(batch_size // mask.shape[0], 1, 1, 1)
+        if masked_image_latents.shape[0] < batch_size:
+            if not batch_size % masked_image_latents.shape[0] == 0:
+                raise ValueError(
+                    "The passed images and the required batch size don't match. Images are supposed to be duplicated"
+                    f" to a total batch size of {batch_size}, but {masked_image_latents.shape[0]} images were passed."
+                    " Make sure the number of images that you pass is divisible by the total requested batch size."
+                )
+            masked_image_latents = masked_image_latents.repeat(batch_size // masked_image_latents.shape[0], 1, 1, 1)
+
+        # aligning device to prevent device errors when concating it with the latent model input
+        masked_image_latents = masked_image_latents.to(device=device, dtype=dtype)
+        masked_image_latents = self._pack_latents(
+            masked_image_latents,
+            batch_size,
+            num_channels_latents,
+            height,
+            width,
+        )
+        mask = self._pack_latents(
+            mask.repeat(1, num_channels_latents, 1, 1),
+            batch_size,
+            num_channels_latents,
+            height,
+            width,
+        )
+
+        return mask, masked_image_latents
+
+    @property
+    def guidance_scale(self):
+        return self._guidance_scale
+
+    @property
+    def joint_attention_kwargs(self):
+        return self._joint_attention_kwargs
+
+    @property
+    def num_timesteps(self):
+        return self._num_timesteps
+
+    @property
+    def current_timestep(self):
+        return self._current_timestep
+
+    @property
+    def interrupt(self):
+        return self._interrupt
+
+    @torch.no_grad()
+    @replace_example_docstring(EXAMPLE_DOC_STRING)
+    def __call__(
+        self,
+        image: Optional[PipelineImageInput] = None,
+        image_reference: Optional[PipelineImageInput] = None,
+        mask_image: PipelineImageInput = None,
+        prompt: Union[str, List[str]] = None,
+        prompt_2: Optional[Union[str, List[str]]] = None,
+        negative_prompt: Union[str, List[str]] = None,
+        negative_prompt_2: Optional[Union[str, List[str]]] = None,
+        true_cfg_scale: float = 1.0,
+        height: Optional[int] = None,
+        width: Optional[int] = None,
+        strength: float = 1.0,
+        padding_mask_crop: Optional[int] = None,
+        num_inference_steps: int = 28,
+        sigmas: Optional[List[float]] = None,
+        guidance_scale: float = 3.5,
+        num_images_per_prompt: Optional[int] = 1,
+        generator: Optional[Union[torch.Generator, List[torch.Generator]]] = None,
+        latents: Optional[torch.FloatTensor] = None,
+        prompt_embeds: Optional[torch.FloatTensor] = None,
+        pooled_prompt_embeds: Optional[torch.FloatTensor] = None,
+        ip_adapter_image: Optional[PipelineImageInput] = None,
+        ip_adapter_image_embeds: Optional[List[torch.Tensor]] = None,
+        negative_ip_adapter_image: Optional[PipelineImageInput] = None,
+        negative_ip_adapter_image_embeds: Optional[List[torch.Tensor]] = None,
+        negative_prompt_embeds: Optional[torch.FloatTensor] = None,
+        negative_pooled_prompt_embeds: Optional[torch.FloatTensor] = None,
+        output_type: Optional[str] = "pil",
+        return_dict: bool = True,
+        joint_attention_kwargs: Optional[Dict[str, Any]] = None,
+        callback_on_step_end: Optional[Callable[[int, int, Dict], None]] = None,
+        callback_on_step_end_tensor_inputs: List[str] = ["latents"],
+        max_sequence_length: int = 512,
+        max_area: int = 1024**2,
+        _auto_resize: bool = True,
+    ):
+        r"""
+        Function invoked when calling the pipeline for generation.
+
+        Args:
+            image (`torch.Tensor`, `PIL.Image.Image`, `np.ndarray`, `List[torch.Tensor]`, `List[PIL.Image.Image]`, or `List[np.ndarray]`):
+                `Image`, numpy array or tensor representing an image batch to be be inpainted (which parts of the image
+                to be masked out with `mask_image` and repainted according to `prompt` and `image_reference`). For both
+                numpy array and pytorch tensor, the expected value range is between `[0, 1]` If it's a tensor or a list
+                or tensors, the expected shape should be `(B, C, H, W)` or `(C, H, W)`. If it is a numpy array or a
+                list of arrays, the expected shape should be `(B, H, W, C)` or `(H, W, C)` It can also accept image
+                latents as `image`, but if passing latents directly it is not encoded again.
+            image_reference (`torch.Tensor`, `PIL.Image.Image`, `np.ndarray`, `List[torch.Tensor]`, `List[PIL.Image.Image]`, or `List[np.ndarray]`):
+                `Image`, numpy array or tensor representing an image batch to be used as the starting point for the
+                masked area. For both numpy array and pytorch tensor, the expected value range is between `[0, 1]` If
+                it's a tensor or a list or tensors, the expected shape should be `(B, C, H, W)` or `(C, H, W)` If it is
+                a numpy array or a list of arrays, the expected shape should be `(B, H, W, C)` or `(H, W, C)` It can
+                also accept image latents as `image`, but if passing latents directly it is not encoded again.
+            mask_image (`torch.Tensor`, `PIL.Image.Image`, `np.ndarray`, `List[torch.Tensor]`, `List[PIL.Image.Image]`, or `List[np.ndarray]`):
+                `Image`, numpy array or tensor representing an image batch to mask `image`. White pixels in the mask
+                are repainted while black pixels are preserved. If `mask_image` is a PIL image, it is converted to a
+                single channel (luminance) before use. If it's a numpy array or pytorch tensor, it should contain one
+                color channel (L) instead of 3, so the expected shape for pytorch tensor would be `(B, 1, H, W)`, `(B,
+                H, W)`, `(1, H, W)`, `(H, W)`. And for numpy array would be for `(B, H, W, 1)`, `(B, H, W)`, `(H, W,
+                1)`, or `(H, W)`.
+            prompt (`str` or `List[str]`, *optional*):
+                The prompt or prompts to guide the image generation. If not defined, one has to pass `prompt_embeds`.
+                instead.
+            prompt_2 (`str` or `List[str]`, *optional*):
+                The prompt or prompts to be sent to `tokenizer_2` and `text_encoder_2`. If not defined, `prompt` is
+                will be used instead.
+            negative_prompt (`str` or `List[str]`, *optional*):
+                The prompt or prompts not to guide the image generation. If not defined, one has to pass
+                `negative_prompt_embeds` instead. Ignored when not using guidance (i.e., ignored if `true_cfg_scale` is
+                not greater than `1`).
+            negative_prompt_2 (`str` or `List[str]`, *optional*):
+                The prompt or prompts not to guide the image generation to be sent to `tokenizer_2` and
+                `text_encoder_2`. If not defined, `negative_prompt` is used in all the text-encoders.
+            true_cfg_scale (`float`, *optional*, defaults to 1.0):
+                When > 1.0 and a provided `negative_prompt`, enables true classifier-free guidance.
+            height (`int`, *optional*, defaults to self.unet.config.sample_size * self.vae_scale_factor):
+                The height in pixels of the generated image. This is set to 1024 by default for the best results.
+            width (`int`, *optional*, defaults to self.unet.config.sample_size * self.vae_scale_factor):
+                The width in pixels of the generated image. This is set to 1024 by default for the best results.
+            strength (`float`, *optional*, defaults to 1.0):
+                Indicates extent to transform the reference `image`. Must be between 0 and 1. `image` is used as a
+                starting point and more noise is added the higher the `strength`. The number of denoising steps depends
+                on the amount of noise initially added. When `strength` is 1, added noise is maximum and the denoising
+                process runs for the full number of iterations specified in `num_inference_steps`. A value of 1
+                essentially ignores `image`.
+            padding_mask_crop (`int`, *optional*, defaults to `None`):
+                The size of margin in the crop to be applied to the image and masking. If `None`, no crop is applied to
+                image and mask_image. If `padding_mask_crop` is not `None`, it will first find a rectangular region
+                with the same aspect ration of the image and contains all masked area, and then expand that area based
+                on `padding_mask_crop`. The image and mask_image will then be cropped based on the expanded area before
+                resizing to the original image size for inpainting. This is useful when the masked area is small while
+                the image is large and contain information irrelevant for inpainting, such as background.
+            num_inference_steps (`int`, *optional*, defaults to 50):
+                The number of denoising steps. More denoising steps usually lead to a higher quality image at the
+                expense of slower inference.
+            sigmas (`List[float]`, *optional*):
+                Custom sigmas to use for the denoising process with schedulers which support a `sigmas` argument in
+                their `set_timesteps` method. If not defined, the default behavior when `num_inference_steps` is passed
+                will be used.
+            guidance_scale (`float`, *optional*, defaults to 3.5):
+                Guidance scale as defined in [Classifier-Free Diffusion
+                Guidance](https://huggingface.co/papers/2207.12598). `guidance_scale` is defined as `w` of equation 2.
+                of [Imagen Paper](https://huggingface.co/papers/2205.11487). Guidance scale is enabled by setting
+                `guidance_scale > 1`. Higher guidance scale encourages to generate images that are closely linked to
+                the text `prompt`, usually at the expense of lower image quality.
+            num_images_per_prompt (`int`, *optional*, defaults to 1):
+                The number of images to generate per prompt.
+            generator (`torch.Generator` or `List[torch.Generator]`, *optional*):
+                One or a list of [torch generator(s)](https://pytorch.org/docs/stable/generated/torch.Generator.html)
+                to make generation deterministic.
+            latents (`torch.FloatTensor`, *optional*):
+                Pre-generated noisy latents, sampled from a Gaussian distribution, to be used as inputs for image
+                generation. Can be used to tweak the same generation with different prompts. If not provided, a latents
+                tensor will ge generated by sampling using the supplied random `generator`.
+            prompt_embeds (`torch.FloatTensor`, *optional*):
+                Pre-generated text embeddings. Can be used to easily tweak text inputs, *e.g.* prompt weighting. If not
+                provided, text embeddings will be generated from `prompt` input argument.
+            pooled_prompt_embeds (`torch.FloatTensor`, *optional*):
+                Pre-generated pooled text embeddings. Can be used to easily tweak text inputs, *e.g.* prompt weighting.
+                If not provided, pooled text embeddings will be generated from `prompt` input argument.
+            ip_adapter_image: (`PipelineImageInput`, *optional*):
+                Optional image input to work with IP Adapters.
+            ip_adapter_image_embeds (`List[torch.Tensor]`, *optional*):
+                Pre-generated image embeddings for IP-Adapter. It should be a list of length same as number of
+                IP-adapters. Each element should be a tensor of shape `(batch_size, num_images, emb_dim)`. If not
+                provided, embeddings are computed from the `ip_adapter_image` input argument.
+            negative_ip_adapter_image:
+                (`PipelineImageInput`, *optional*): Optional image input to work with IP Adapters.
+            negative_ip_adapter_image_embeds (`List[torch.Tensor]`, *optional*):
+                Pre-generated image embeddings for IP-Adapter. It should be a list of length same as number of
+                IP-adapters. Each element should be a tensor of shape `(batch_size, num_images, emb_dim)`. If not
+                provided, embeddings are computed from the `ip_adapter_image` input argument.
+            negative_prompt_embeds (`torch.FloatTensor`, *optional*):
+                Pre-generated negative text embeddings. Can be used to easily tweak text inputs, *e.g.* prompt
+                weighting. If not provided, negative_prompt_embeds will be generated from `negative_prompt` input
+                argument.
+            negative_pooled_prompt_embeds (`torch.FloatTensor`, *optional*):
+                Pre-generated negative pooled text embeddings. Can be used to easily tweak text inputs, *e.g.* prompt
+                weighting. If not provided, pooled negative_prompt_embeds will be generated from `negative_prompt`
+                input argument.
+            output_type (`str`, *optional*, defaults to `"pil"`):
+                The output format of the generate image. Choose between
+                [PIL](https://pillow.readthedocs.io/en/stable/): `PIL.Image.Image` or `np.array`.
+            return_dict (`bool`, *optional*, defaults to `True`):
+                Whether or not to return a [`~pipelines.flux.FluxPipelineOutput`] instead of a plain tuple.
+            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).
+            callback_on_step_end (`Callable`, *optional*):
+                A function that calls at the end of each denoising steps during the inference. The function is called
+                with the following arguments: `callback_on_step_end(self: DiffusionPipeline, step: int, timestep: int,
+                callback_kwargs: Dict)`. `callback_kwargs` will include a list of all tensors as specified by
+                `callback_on_step_end_tensor_inputs`.
+            callback_on_step_end_tensor_inputs (`List`, *optional*):
+                The list of tensor inputs for the `callback_on_step_end` function. The tensors specified in the list
+                will be passed as `callback_kwargs` argument. You will only be able to include variables listed in the
+                `._callback_tensor_inputs` attribute of your pipeline class.
+            max_sequence_length (`int` defaults to 512):
+                Maximum sequence length to use with the `prompt`.
+            max_area (`int`, defaults to `1024 ** 2`):
+                The maximum area of the generated image in pixels. The height and width will be adjusted to fit this
+                area while maintaining the aspect ratio.
+
+        Examples:
+
+        Returns:
+            [`~pipelines.flux.FluxPipelineOutput`] or `tuple`: [`~pipelines.flux.FluxPipelineOutput`] if `return_dict`
+            is True, otherwise a `tuple`. When returning a tuple, the first element is a list with the generated
+            images.
+        """
+
+        height = height or self.default_sample_size * self.vae_scale_factor
+        width = width or self.default_sample_size * self.vae_scale_factor
+
+        original_height, original_width = height, width
+        aspect_ratio = width / height
+        width = round((max_area * aspect_ratio) ** 0.5)
+        height = round((max_area / aspect_ratio) ** 0.5)
+
+        multiple_of = self.vae_scale_factor * 2
+        width = width // multiple_of * multiple_of
+        height = height // multiple_of * multiple_of
+
+        if height != original_height or width != original_width:
+            logger.warning(
+                f"Generation `height` and `width` have been adjusted to {height} and {width} to fit the model requirements."
+            )
+
+        # 1. Check inputs. Raise error if not correct
+        self.check_inputs(
+            prompt,
+            prompt_2,
+            image,
+            mask_image,
+            strength,
+            height,
+            width,
+            output_type=output_type,
+            negative_prompt=negative_prompt,
+            negative_prompt_2=negative_prompt_2,
+            prompt_embeds=prompt_embeds,
+            negative_prompt_embeds=negative_prompt_embeds,
+            pooled_prompt_embeds=pooled_prompt_embeds,
+            negative_pooled_prompt_embeds=negative_pooled_prompt_embeds,
+            callback_on_step_end_tensor_inputs=callback_on_step_end_tensor_inputs,
+            padding_mask_crop=padding_mask_crop,
+            max_sequence_length=max_sequence_length,
+        )
+
+        self._guidance_scale = guidance_scale
+        self._joint_attention_kwargs = joint_attention_kwargs
+        self._current_timestep = None
+        self._interrupt = False
+
+        # 2. Preprocess image
+        if image is not None and not (isinstance(image, torch.Tensor) and image.size(1) == self.latent_channels):
+            if isinstance(image, list) and isinstance(image[0], torch.Tensor) and image[0].ndim == 4:
+                image = torch.cat(image, dim=0)
+            img = image[0] if isinstance(image, list) else image
+            image_height, image_width = self.image_processor.get_default_height_width(img)
+            aspect_ratio = image_width / image_height
+            if _auto_resize:
+                # Kontext is trained on specific resolutions, using one of them is recommended
+                _, image_width, image_height = min(
+                    (abs(aspect_ratio - w / h), w, h) for w, h in PREFERRED_KONTEXT_RESOLUTIONS
+                )
+            image_width = image_width // multiple_of * multiple_of
+            image_height = image_height // multiple_of * multiple_of
+            image = self.image_processor.resize(image, image_height, image_width)
+
+            # Choose the resolution of the image to be the same as the image
+            width = image_width
+            height = image_height
+
+            # 2.1 Preprocess mask
+            if padding_mask_crop is not None:
+                crops_coords = self.mask_processor.get_crop_region(mask_image, width, height, pad=padding_mask_crop)
+                resize_mode = "fill"
+            else:
+                crops_coords = None
+                resize_mode = "default"
+
+            image = self.image_processor.preprocess(
+                image, image_height, image_width, crops_coords=crops_coords, resize_mode=resize_mode
+            )
+        else:
+            raise ValueError("image must be provided correctly for inpainting")
+
+        init_image = image.to(dtype=torch.float32)
+
+        # 2.1 Preprocess image_reference
+        if image_reference is not None and not (
+            isinstance(image_reference, torch.Tensor) and image_reference.size(1) == self.latent_channels
+        ):
+            if (
+                isinstance(image_reference, list)
+                and isinstance(image_reference[0], torch.Tensor)
+                and image_reference[0].ndim == 4
+            ):
+                image_reference = torch.cat(image_reference, dim=0)
+            img_reference = image_reference[0] if isinstance(image_reference, list) else image_reference
+            image_reference_height, image_reference_width = self.image_processor.get_default_height_width(
+                img_reference
+            )
+            aspect_ratio = image_reference_width / image_reference_height
+            if _auto_resize:
+                # Kontext is trained on specific resolutions, using one of them is recommended
+                _, image_reference_width, image_reference_height = min(
+                    (abs(aspect_ratio - w / h), w, h) for w, h in PREFERRED_KONTEXT_RESOLUTIONS
+                )
+            image_reference_width = image_reference_width // multiple_of * multiple_of
+            image_reference_height = image_reference_height // multiple_of * multiple_of
+            image_reference = self.image_processor.resize(
+                image_reference, image_reference_height, image_reference_width
+            )
+            image_reference = self.image_processor.preprocess(
+                image_reference,
+                image_reference_height,
+                image_reference_width,
+                crops_coords=crops_coords,
+                resize_mode=resize_mode,
+            )
+        else:
+            image_reference = None
+
+        # 3. Define call parameters
+        if prompt is not None and isinstance(prompt, str):
+            batch_size = 1
+        elif prompt is not None and isinstance(prompt, list):
+            batch_size = len(prompt)
+        else:
+            batch_size = prompt_embeds.shape[0]
+
+        device = self._execution_device
+
+        lora_scale = (
+            self.joint_attention_kwargs.get("scale", None) if self.joint_attention_kwargs is not None else None
+        )
+        has_neg_prompt = negative_prompt is not None or (
+            negative_prompt_embeds is not None and negative_pooled_prompt_embeds is not None
+        )
+        do_true_cfg = true_cfg_scale > 1 and has_neg_prompt
+        (
+            prompt_embeds,
+            pooled_prompt_embeds,
+            text_ids,
+        ) = self.encode_prompt(
+            prompt=prompt,
+            prompt_2=prompt_2,
+            prompt_embeds=prompt_embeds,
+            pooled_prompt_embeds=pooled_prompt_embeds,
+            device=device,
+            num_images_per_prompt=num_images_per_prompt,
+            max_sequence_length=max_sequence_length,
+            lora_scale=lora_scale,
+        )
+        if do_true_cfg:
+            (
+                negative_prompt_embeds,
+                negative_pooled_prompt_embeds,
+                negative_text_ids,
+            ) = self.encode_prompt(
+                prompt=negative_prompt,
+                prompt_2=negative_prompt_2,
+                prompt_embeds=negative_prompt_embeds,
+                pooled_prompt_embeds=negative_pooled_prompt_embeds,
+                device=device,
+                num_images_per_prompt=num_images_per_prompt,
+                max_sequence_length=max_sequence_length,
+                lora_scale=lora_scale,
+            )
+
+        # 4. Prepare timesteps
+        sigmas = np.linspace(1.0, 1 / num_inference_steps, num_inference_steps) if sigmas is None else sigmas
+        image_seq_len = (int(height) // self.vae_scale_factor // 2) * (int(width) // self.vae_scale_factor // 2)
+        mu = calculate_shift(
+            image_seq_len,
+            self.scheduler.config.get("base_image_seq_len", 256),
+            self.scheduler.config.get("max_image_seq_len", 4096),
+            self.scheduler.config.get("base_shift", 0.5),
+            self.scheduler.config.get("max_shift", 1.15),
+        )
+        timesteps, num_inference_steps = retrieve_timesteps(
+            self.scheduler,
+            num_inference_steps,
+            device,
+            sigmas=sigmas,
+            mu=mu,
+        )
+        timesteps, num_inference_steps = self.get_timesteps(num_inference_steps, strength, device)
+        if num_inference_steps < 1:
+            raise ValueError(
+                f"After adjusting the num_inference_steps by strength parameter: {strength}, the number of pipeline"
+                f"steps is {num_inference_steps} which is < 1 and not appropriate for this pipeline."
+            )
+        latent_timestep = timesteps[:1].repeat(batch_size * num_images_per_prompt)
+
+        num_warmup_steps = max(len(timesteps) - num_inference_steps * self.scheduler.order, 0)
+        self._num_timesteps = len(timesteps)
+
+        # 5. Prepare latent variables
+        num_channels_latents = self.transformer.config.in_channels // 4
+        latents, image_latents, image_reference_latents, latent_ids, image_ids, image_reference_ids, noise = (
+            self.prepare_latents(
+                init_image,
+                latent_timestep,
+                batch_size * num_images_per_prompt,
+                num_channels_latents,
+                height,
+                width,
+                prompt_embeds.dtype,
+                device,
+                generator,
+                latents,
+                image_reference,
+            )
+        )
+
+        if image_reference_ids is not None:
+            latent_ids = torch.cat([latent_ids, image_reference_ids], dim=0)  # dim 0 is sequence dimension
+        elif image_ids is not None:
+            latent_ids = torch.cat([latent_ids, image_ids], dim=0)  # dim 0 is sequence dimension
+
+        mask_condition = self.mask_processor.preprocess(
+            mask_image, height=height, width=width, resize_mode=resize_mode, crops_coords=crops_coords
+        )
+
+        masked_image = init_image * (mask_condition < 0.5)
+
+        mask, _ = self.prepare_mask_latents(
+            mask_condition,
+            masked_image,
+            batch_size,
+            num_channels_latents,
+            num_images_per_prompt,
+            height,
+            width,
+            prompt_embeds.dtype,
+            device,
+            generator,
+        )
+
+        # handle guidance
+        if self.transformer.config.guidance_embeds:
+            guidance = torch.full([1], guidance_scale, device=device, dtype=torch.float32)
+            guidance = guidance.expand(latents.shape[0])
+        else:
+            guidance = None
+
+        if (ip_adapter_image is not None or ip_adapter_image_embeds is not None) and (
+            negative_ip_adapter_image is None and negative_ip_adapter_image_embeds is None
+        ):
+            negative_ip_adapter_image = np.zeros((width, height, 3), dtype=np.uint8)
+            negative_ip_adapter_image = [negative_ip_adapter_image] * self.transformer.encoder_hid_proj.num_ip_adapters
+
+        elif (ip_adapter_image is None and ip_adapter_image_embeds is None) and (
+            negative_ip_adapter_image is not None or negative_ip_adapter_image_embeds is not None
+        ):
+            ip_adapter_image = np.zeros((width, height, 3), dtype=np.uint8)
+            ip_adapter_image = [ip_adapter_image] * self.transformer.encoder_hid_proj.num_ip_adapters
+
+        if self.joint_attention_kwargs is None:
+            self._joint_attention_kwargs = {}
+
+        image_embeds = None
+        negative_image_embeds = None
+        if ip_adapter_image is not None or ip_adapter_image_embeds is not None:
+            image_embeds = self.prepare_ip_adapter_image_embeds(
+                ip_adapter_image,
+                ip_adapter_image_embeds,
+                device,
+                batch_size * num_images_per_prompt,
+            )
+        if negative_ip_adapter_image is not None or negative_ip_adapter_image_embeds is not None:
+            negative_image_embeds = self.prepare_ip_adapter_image_embeds(
+                negative_ip_adapter_image,
+                negative_ip_adapter_image_embeds,
+                device,
+                batch_size * num_images_per_prompt,
+            )
+
+        # 6. Denoising loop
+        with self.progress_bar(total=num_inference_steps) as progress_bar:
+            for i, t in enumerate(timesteps):
+                if self.interrupt:
+                    continue
+
+                self._current_timestep = t
+                if image_embeds is not None:
+                    self._joint_attention_kwargs["ip_adapter_image_embeds"] = image_embeds
+
+                latent_model_input = latents
+                if image_reference_latents is not None:
+                    latent_model_input = torch.cat([latents, image_reference_latents], dim=1)
+                elif image_latents is not None:
+                    latent_model_input = torch.cat([latents, image_latents], dim=1)
+                timestep = t.expand(latents.shape[0]).to(latents.dtype)
+
+                noise_pred = self.transformer(
+                    hidden_states=latent_model_input,
+                    timestep=timestep / 1000,
+                    guidance=guidance,
+                    pooled_projections=pooled_prompt_embeds,
+                    encoder_hidden_states=prompt_embeds,
+                    txt_ids=text_ids,
+                    img_ids=latent_ids,
+                    joint_attention_kwargs=self.joint_attention_kwargs,
+                    return_dict=False,
+                )[0]
+                noise_pred = noise_pred[:, : latents.size(1)]
+
+                if do_true_cfg:
+                    if negative_image_embeds is not None:
+                        self._joint_attention_kwargs["ip_adapter_image_embeds"] = negative_image_embeds
+                    neg_noise_pred = self.transformer(
+                        hidden_states=latent_model_input,
+                        timestep=timestep / 1000,
+                        guidance=guidance,
+                        pooled_projections=negative_pooled_prompt_embeds,
+                        encoder_hidden_states=negative_prompt_embeds,
+                        txt_ids=negative_text_ids,
+                        img_ids=latent_ids,
+                        joint_attention_kwargs=self.joint_attention_kwargs,
+                        return_dict=False,
+                    )[0]
+                    neg_noise_pred = neg_noise_pred[:, : latents.size(1)]
+                    noise_pred = neg_noise_pred + true_cfg_scale * (noise_pred - neg_noise_pred)
+
+                # compute the previous noisy sample x_t -> x_t-1
+                latents_dtype = latents.dtype
+                latents = self.scheduler.step(noise_pred, t, latents, return_dict=False)[0]
+
+                init_latents_proper = image_latents
+                init_mask = mask
+
+                if i < len(timesteps) - 1:
+                    noise_timestep = timesteps[i + 1]
+                    init_latents_proper = self.scheduler.scale_noise(
+                        init_latents_proper, torch.tensor([noise_timestep]), noise
+                    )
+
+                latents = (1 - init_mask) * init_latents_proper + init_mask * latents
+
+                if latents.dtype != latents_dtype:
+                    if torch.backends.mps.is_available():
+                        # some platforms (eg. apple mps) misbehave due to a pytorch bug: https://github.com/pytorch/pytorch/pull/99272
+                        latents = latents.to(latents_dtype)
+
+                if callback_on_step_end is not None:
+                    callback_kwargs = {}
+                    for k in callback_on_step_end_tensor_inputs:
+                        callback_kwargs[k] = locals()[k]
+                    callback_outputs = callback_on_step_end(self, i, t, callback_kwargs)
+
+                    latents = callback_outputs.pop("latents", latents)
+                    prompt_embeds = callback_outputs.pop("prompt_embeds", prompt_embeds)
+
+                # call the callback, if provided
+                if i == len(timesteps) - 1 or ((i + 1) > num_warmup_steps and (i + 1) % self.scheduler.order == 0):
+                    progress_bar.update()
+
+                if XLA_AVAILABLE:
+                    xm.mark_step()
+
+        self._current_timestep = None
+
+        if output_type == "latent":
+            image = latents
+        else:
+            latents = self._unpack_latents(latents, height, width, self.vae_scale_factor)
+            latents = (latents / self.vae.config.scaling_factor) + self.vae.config.shift_factor
+            image = self.vae.decode(latents, return_dict=False)[0]
+            image = self.image_processor.postprocess(image, output_type=output_type)
+
+        # Offload all models
+        self.maybe_free_model_hooks()
+
+        if not return_dict:
+            return (image,)
+
+        return FluxPipelineOutput(images=image)

src/diffusers/utils/dummy_torch_and_transformers_objects.py

@@ -692,6 +692,21 @@ class FluxInpaintPipeline(metaclass=DummyObject):
         requires_backends(cls, ["torch", "transformers"])
 
 
+class FluxKontextInpaintPipeline(metaclass=DummyObject):
+    _backends = ["torch", "transformers"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch", "transformers"])
+
+    @classmethod
+    def from_config(cls, *args, **kwargs):
+        requires_backends(cls, ["torch", "transformers"])
+
+    @classmethod
+    def from_pretrained(cls, *args, **kwargs):
+        requires_backends(cls, ["torch", "transformers"])
+
+
 class FluxKontextPipeline(metaclass=DummyObject):
     _backends = ["torch", "transformers"]
 

tests/pipelines/flux/test_pipeline_flux_kontext_inpaint.py

+import random
+import unittest
+
+import numpy as np
+import torch
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
+
+from diffusers import (
+    AutoencoderKL,
+    FasterCacheConfig,
+    FlowMatchEulerDiscreteScheduler,
+    FluxKontextInpaintPipeline,
+    FluxTransformer2DModel,
+)
+from diffusers.utils.testing_utils import floats_tensor, torch_device
+
+from ..test_pipelines_common import (
+    FasterCacheTesterMixin,
+    FluxIPAdapterTesterMixin,
+    PipelineTesterMixin,
+    PyramidAttentionBroadcastTesterMixin,
+)
+
+
+class FluxKontextInpaintPipelineFastTests(
+    unittest.TestCase,
+    PipelineTesterMixin,
+    FluxIPAdapterTesterMixin,
+    PyramidAttentionBroadcastTesterMixin,
+    FasterCacheTesterMixin,
+):
+    pipeline_class = FluxKontextInpaintPipeline
+    params = frozenset(
+        ["image", "prompt", "height", "width", "guidance_scale", "prompt_embeds", "pooled_prompt_embeds"]
+    )
+    batch_params = frozenset(["image", "prompt"])
+
+    # there is no xformers processor for Flux
+    test_xformers_attention = False
+    test_layerwise_casting = True
+    test_group_offloading = True
+
+    faster_cache_config = FasterCacheConfig(
+        spatial_attention_block_skip_range=2,
+        spatial_attention_timestep_skip_range=(-1, 901),
+        unconditional_batch_skip_range=2,
+        attention_weight_callback=lambda _: 0.5,
+        is_guidance_distilled=True,
+    )
+
+    def get_dummy_components(self, num_layers: int = 1, num_single_layers: int = 1):
+        torch.manual_seed(0)
+        transformer = FluxTransformer2DModel(
+            patch_size=1,
+            in_channels=4,
+            num_layers=num_layers,
+            num_single_layers=num_single_layers,
+            attention_head_dim=16,
+            num_attention_heads=2,
+            joint_attention_dim=32,
+            pooled_projection_dim=32,
+            axes_dims_rope=[4, 4, 8],
+        )
+        clip_text_encoder_config = CLIPTextConfig(
+            bos_token_id=0,
+            eos_token_id=2,
+            hidden_size=32,
+            intermediate_size=37,
+            layer_norm_eps=1e-05,
+            num_attention_heads=4,
+            num_hidden_layers=5,
+            pad_token_id=1,
+            vocab_size=1000,
+            hidden_act="gelu",
+            projection_dim=32,
+        )
+
+        torch.manual_seed(0)
+        text_encoder = CLIPTextModel(clip_text_encoder_config)
+
+        torch.manual_seed(0)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
+
+        tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
+        tokenizer_2 = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
+
+        torch.manual_seed(0)
+        vae = AutoencoderKL(
+            sample_size=32,
+            in_channels=3,
+            out_channels=3,
+            block_out_channels=(4,),
+            layers_per_block=1,
+            latent_channels=1,
+            norm_num_groups=1,
+            use_quant_conv=False,
+            use_post_quant_conv=False,
+            shift_factor=0.0609,
+            scaling_factor=1.5035,
+        )
+
+        scheduler = FlowMatchEulerDiscreteScheduler()
+
+        return {
+            "scheduler": scheduler,
+            "text_encoder": text_encoder,
+            "text_encoder_2": text_encoder_2,
+            "tokenizer": tokenizer,
+            "tokenizer_2": tokenizer_2,
+            "transformer": transformer,
+            "vae": vae,
+            "image_encoder": None,
+            "feature_extractor": None,
+        }
+
+    def get_dummy_inputs(self, device, seed=0):
+        image = floats_tensor((1, 3, 32, 32), rng=random.Random(seed)).to(device)
+        mask_image = torch.ones((1, 1, 32, 32)).to(device)
+        if str(device).startswith("mps"):
+            generator = torch.manual_seed(seed)
+        else:
+            generator = torch.Generator(device="cpu").manual_seed(seed)
+
+        inputs = {
+            "prompt": "A painting of a squirrel eating a burger",
+            "image": image,
+            "mask_image": mask_image,
+            "generator": generator,
+            "num_inference_steps": 2,
+            "guidance_scale": 5.0,
+            "height": 32,
+            "width": 32,
+            "max_sequence_length": 48,
+            "strength": 0.8,
+            "output_type": "np",
+            "_auto_resize": False,
+        }
+        return inputs
+
+    def test_flux_inpaint_different_prompts(self):
+        pipe = self.pipeline_class(**self.get_dummy_components()).to(torch_device)
+
+        inputs = self.get_dummy_inputs(torch_device)
+        output_same_prompt = pipe(**inputs).images[0]
+
+        inputs = self.get_dummy_inputs(torch_device)
+        inputs["prompt_2"] = "a different prompt"
+        output_different_prompts = pipe(**inputs).images[0]
+
+        max_diff = np.abs(output_same_prompt - output_different_prompts).max()
+
+        # Outputs should be different here
+        # For some reasons, they don't show large differences
+        assert max_diff > 1e-6
+
+    def test_flux_image_output_shape(self):
+        pipe = self.pipeline_class(**self.get_dummy_components()).to(torch_device)
+        inputs = self.get_dummy_inputs(torch_device)
+
+        height_width_pairs = [(32, 32), (72, 56)]
+        for height, width in height_width_pairs:
+            expected_height = height - height % (pipe.vae_scale_factor * 2)
+            expected_width = width - width % (pipe.vae_scale_factor * 2)
+            # Because output shape is the same as the input shape, we need to create a dummy image and mask image
+            image = floats_tensor((1, 3, height, width), rng=random.Random(0)).to(torch_device)
+            mask_image = torch.ones((1, 1, height, width)).to(torch_device)
+
+            inputs.update(
+                {
+                    "height": height,
+                    "width": width,
+                    "max_area": height * width,
+                    "image": image,
+                    "mask_image": mask_image,
+                }
+            )
+            image = pipe(**inputs).images[0]
+            output_height, output_width, _ = image.shape
+            assert (output_height, output_width) == (expected_height, expected_width)
+
+    def test_flux_true_cfg(self):
+        pipe = self.pipeline_class(**self.get_dummy_components()).to(torch_device)
+        inputs = self.get_dummy_inputs(torch_device)
+        inputs.pop("generator")
+
+        no_true_cfg_out = pipe(**inputs, generator=torch.manual_seed(0)).images[0]
+        inputs["negative_prompt"] = "bad quality"
+        inputs["true_cfg_scale"] = 2.0
+        true_cfg_out = pipe(**inputs, generator=torch.manual_seed(0)).images[0]
+        assert not np.allclose(no_true_cfg_out, true_cfg_out)