src/diffusers/pipelines/deepfloyd_if/pipeline_if_inpainting_superresolution.py

import html
import inspect
import re
import urllib.parse as ul
from typing import Any, Callable, Dict, List, Optional, Union

import numpy as np
import PIL.Image
import torch
import torch.nn.functional as F
from transformers import CLIPImageProcessor, T5EncoderModel, T5Tokenizer

from ...loaders import LoraLoaderMixin
from ...models import UNet2DConditionModel
from ...schedulers import DDPMScheduler
from ...utils import (
    BACKENDS_MAPPING,
    PIL_INTERPOLATION,
    is_bs4_available,
    is_ftfy_available,
    logging,
    replace_example_docstring,
)
from ...utils.torch_utils import randn_tensor
from ..pipeline_utils import DiffusionPipeline
from .pipeline_output import IFPipelineOutput
from .safety_checker import IFSafetyChecker
from .watermark import IFWatermarker


if is_bs4_available():
    from bs4 import BeautifulSoup

if is_ftfy_available():
    import ftfy


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


# Copied from diffusers.pipelines.deepfloyd_if.pipeline_if_img2img.resize
def resize(images: PIL.Image.Image, img_size: int) -> PIL.Image.Image:
    w, h = images.size

    coef = w / h

    w, h = img_size, img_size

    if coef >= 1:
        w = int(round(img_size / 8 * coef) * 8)
    else:
        h = int(round(img_size / 8 / coef) * 8)

    images = images.resize((w, h), resample=PIL_INTERPOLATION["bicubic"], reducing_gap=None)

    return images


EXAMPLE_DOC_STRING = """
    Examples:
        ```py
        >>> from diffusers import IFInpaintingPipeline, IFInpaintingSuperResolutionPipeline, DiffusionPipeline
        >>> from diffusers.utils import pt_to_pil
        >>> import torch
        >>> from PIL import Image
        >>> import requests
        >>> from io import BytesIO

        >>> url = "https://huggingface.co/datasets/diffusers/docs-images/resolve/main/if/person.png"
        >>> response = requests.get(url)
        >>> original_image = Image.open(BytesIO(response.content)).convert("RGB")
        >>> original_image = original_image

        >>> url = "https://huggingface.co/datasets/diffusers/docs-images/resolve/main/if/glasses_mask.png"
        >>> response = requests.get(url)
        >>> mask_image = Image.open(BytesIO(response.content))
        >>> mask_image = mask_image

        >>> pipe = IFInpaintingPipeline.from_pretrained(
        ...     "DeepFloyd/IF-I-XL-v1.0", variant="fp16", torch_dtype=torch.float16
        ... )
        >>> pipe.enable_model_cpu_offload()

        >>> prompt = "blue sunglasses"

        >>> prompt_embeds, negative_embeds = pipe.encode_prompt(prompt)
        >>> image = pipe(
        ...     image=original_image,
        ...     mask_image=mask_image,
        ...     prompt_embeds=prompt_embeds,
        ...     negative_prompt_embeds=negative_embeds,
        ...     output_type="pt",
        ... ).images

        >>> # save intermediate image
        >>> pil_image = pt_to_pil(image)
        >>> pil_image[0].save("./if_stage_I.png")

        >>> super_res_1_pipe = IFInpaintingSuperResolutionPipeline.from_pretrained(
        ...     "DeepFloyd/IF-II-L-v1.0", text_encoder=None, variant="fp16", torch_dtype=torch.float16
        ... )
        >>> super_res_1_pipe.enable_model_cpu_offload()

        >>> image = super_res_1_pipe(
        ...     image=image,
        ...     mask_image=mask_image,
        ...     original_image=original_image,
        ...     prompt_embeds=prompt_embeds,
        ...     negative_prompt_embeds=negative_embeds,
        ... ).images
        >>> image[0].save("./if_stage_II.png")
        ```
    """


class IFInpaintingSuperResolutionPipeline(DiffusionPipeline, LoraLoaderMixin):
    tokenizer: T5Tokenizer
    text_encoder: T5EncoderModel

    unet: UNet2DConditionModel
    scheduler: DDPMScheduler
    image_noising_scheduler: DDPMScheduler

    feature_extractor: Optional[CLIPImageProcessor]
    safety_checker: Optional[IFSafetyChecker]

    watermarker: Optional[IFWatermarker]

    bad_punct_regex = re.compile(
        r"["
        + "#®•©™&@·º½¾¿¡§~"
        + r"\)"
        + r"\("
        + r"\]"
        + r"\["
        + r"\}"
        + r"\{"
        + r"\|"
        + "\\"
        + r"\/"
        + r"\*"
        + r"]{1,}"
    )  # noqa

    model_cpu_offload_seq = "text_encoder->unet"
    _optional_components = ["tokenizer", "text_encoder", "safety_checker", "feature_extractor", "watermarker"]
    _exclude_from_cpu_offload = ["watermarker"]

    def __init__(
        self,
        tokenizer: T5Tokenizer,
        text_encoder: T5EncoderModel,
        unet: UNet2DConditionModel,
        scheduler: DDPMScheduler,
        image_noising_scheduler: DDPMScheduler,
        safety_checker: Optional[IFSafetyChecker],
        feature_extractor: Optional[CLIPImageProcessor],
        watermarker: Optional[IFWatermarker],
        requires_safety_checker: bool = True,
    ):
        super().__init__()

        if safety_checker is None and requires_safety_checker:
            logger.warning(
                f"You have disabled the safety checker for {self.__class__} by passing `safety_checker=None`. Ensure"
                " that you abide to the conditions of the IF license and do not expose unfiltered"
                " results in services or applications open to the public. Both the diffusers team and Hugging Face"
                " strongly recommend to keep the safety filter enabled in all public facing circumstances, disabling"
                " it only for use-cases that involve analyzing network behavior or auditing its results. For more"
                " information, please have a look at https://github.com/huggingface/diffusers/pull/254 ."
            )

        if safety_checker is not None and feature_extractor is None:
            raise ValueError(
                "Make sure to define a feature extractor when loading {self.__class__} if you want to use the safety"
                " checker. If you do not want to use the safety checker, you can pass `'safety_checker=None'` instead."
            )

        if unet.config.in_channels != 6:
            logger.warning(
                "It seems like you have loaded a checkpoint that shall not be used for super resolution from {unet.config._name_or_path} as it accepts {unet.config.in_channels} input channels instead of 6. Please make sure to pass a super resolution checkpoint as the `'unet'`: IFSuperResolutionPipeline.from_pretrained(unet=super_resolution_unet, ...)`."
            )

        self.register_modules(
            tokenizer=tokenizer,
            text_encoder=text_encoder,
            unet=unet,
            scheduler=scheduler,
            image_noising_scheduler=image_noising_scheduler,
            safety_checker=safety_checker,
            feature_extractor=feature_extractor,
            watermarker=watermarker,
        )
        self.register_to_config(requires_safety_checker=requires_safety_checker)

    # Copied from diffusers.pipelines.deepfloyd_if.pipeline_if.IFPipeline._text_preprocessing
    def _text_preprocessing(self, text, clean_caption=False):
        if clean_caption and not is_bs4_available():
            logger.warning(BACKENDS_MAPPING["bs4"][-1].format("Setting `clean_caption=True`"))
            logger.warning("Setting `clean_caption` to False...")
            clean_caption = False

        if clean_caption and not is_ftfy_available():
            logger.warning(BACKENDS_MAPPING["ftfy"][-1].format("Setting `clean_caption=True`"))
            logger.warning("Setting `clean_caption` to False...")
            clean_caption = False

        if not isinstance(text, (tuple, list)):
            text = [text]

        def process(text: str):
            if clean_caption:
                text = self._clean_caption(text)
                text = self._clean_caption(text)
            else:
                text = text.lower().strip()
            return text

        return [process(t) for t in text]

    # Copied from diffusers.pipelines.deepfloyd_if.pipeline_if.IFPipeline._clean_caption
    def _clean_caption(self, caption):
        caption = str(caption)
        caption = ul.unquote_plus(caption)
        caption = caption.strip().lower()
        caption = re.sub("<person>", "person", caption)
        # urls:
        caption = re.sub(
            r"\b((?:https?:(?:\/{1,3}|[a-zA-Z0-9%])|[a-zA-Z0-9.\-]+[.](?:com|co|ru|net|org|edu|gov|it)[\w/-]*\b\/?(?!@)))",  # noqa
            "",
            caption,
        )  # regex for urls
        caption = re.sub(
            r"\b((?:www:(?:\/{1,3}|[a-zA-Z0-9%])|[a-zA-Z0-9.\-]+[.](?:com|co|ru|net|org|edu|gov|it)[\w/-]*\b\/?(?!@)))",  # noqa
            "",
            caption,
        )  # regex for urls
        # html:
        caption = BeautifulSoup(caption, features="html.parser").text

        # @<nickname>
        caption = re.sub(r"@[\w\d]+\b", "", caption)

        # 31C0—31EF CJK Strokes
        # 31F0—31FF Katakana Phonetic Extensions
        # 3200—32FF Enclosed CJK Letters and Months
        # 3300—33FF CJK Compatibility
        # 3400—4DBF CJK Unified Ideographs Extension A
        # 4DC0—4DFF Yijing Hexagram Symbols
        # 4E00—9FFF CJK Unified Ideographs
        caption = re.sub(r"[\u31c0-\u31ef]+", "", caption)
        caption = re.sub(r"[\u31f0-\u31ff]+", "", caption)
        caption = re.sub(r"[\u3200-\u32ff]+", "", caption)
        caption = re.sub(r"[\u3300-\u33ff]+", "", caption)
        caption = re.sub(r"[\u3400-\u4dbf]+", "", caption)
        caption = re.sub(r"[\u4dc0-\u4dff]+", "", caption)
        caption = re.sub(r"[\u4e00-\u9fff]+", "", caption)
        #######################################################

        # все виды тире / all types of dash --> "-"
        caption = re.sub(
            r"[\u002D\u058A\u05BE\u1400\u1806\u2010-\u2015\u2E17\u2E1A\u2E3A\u2E3B\u2E40\u301C\u3030\u30A0\uFE31\uFE32\uFE58\uFE63\uFF0D]+",  # noqa
            "-",
            caption,
        )

        # кавычки к одному стандарту
        caption = re.sub(r"[`´«»“”¨]", '"', caption)
        caption = re.sub(r"[‘’]", "'", caption)

        # &quot;
        caption = re.sub(r"&quot;?", "", caption)
        # &amp
        caption = re.sub(r"&amp", "", caption)

        # ip adresses:
        caption = re.sub(r"\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}", " ", caption)

        # article ids:
        caption = re.sub(r"\d:\d\d\s+$", "", caption)

        # \n
        caption = re.sub(r"\\n", " ", caption)

        # "#123"
        caption = re.sub(r"#\d{1,3}\b", "", caption)
        # "#12345.."
        caption = re.sub(r"#\d{5,}\b", "", caption)
        # "123456.."
        caption = re.sub(r"\b\d{6,}\b", "", caption)
        # filenames:
        caption = re.sub(r"[\S]+\.(?:png|jpg|jpeg|bmp|webp|eps|pdf|apk|mp4)", "", caption)

        #
        caption = re.sub(r"[\"\']{2,}", r'"', caption)  # """AUSVERKAUFT"""
        caption = re.sub(r"[\.]{2,}", r" ", caption)  # """AUSVERKAUFT"""

        caption = re.sub(self.bad_punct_regex, r" ", caption)  # ***AUSVERKAUFT***, #AUSVERKAUFT
        caption = re.sub(r"\s+\.\s+", r" ", caption)  # " . "

        # this-is-my-cute-cat / this_is_my_cute_cat
        regex2 = re.compile(r"(?:\-|\_)")
        if len(re.findall(regex2, caption)) > 3:
            caption = re.sub(regex2, " ", caption)

        caption = ftfy.fix_text(caption)
        caption = html.unescape(html.unescape(caption))

        caption = re.sub(r"\b[a-zA-Z]{1,3}\d{3,15}\b", "", caption)  # jc6640
        caption = re.sub(r"\b[a-zA-Z]+\d+[a-zA-Z]+\b", "", caption)  # jc6640vc
        caption = re.sub(r"\b\d+[a-zA-Z]+\d+\b", "", caption)  # 6640vc231

        caption = re.sub(r"(worldwide\s+)?(free\s+)?shipping", "", caption)
        caption = re.sub(r"(free\s)?download(\sfree)?", "", caption)
        caption = re.sub(r"\bclick\b\s(?:for|on)\s\w+", "", caption)
        caption = re.sub(r"\b(?:png|jpg|jpeg|bmp|webp|eps|pdf|apk|mp4)(\simage[s]?)?", "", caption)
        caption = re.sub(r"\bpage\s+\d+\b", "", caption)

        caption = re.sub(r"\b\d*[a-zA-Z]+\d+[a-zA-Z]+\d+[a-zA-Z\d]*\b", r" ", caption)  # j2d1a2a...

        caption = re.sub(r"\b\d+\.?\d*[xх×]\d+\.?\d*\b", "", caption)

        caption = re.sub(r"\b\s+\:\s+", r": ", caption)
        caption = re.sub(r"(\D[,\./])\b", r"\1 ", caption)
        caption = re.sub(r"\s+", " ", caption)

        caption.strip()

        caption = re.sub(r"^[\"\']([\w\W]+)[\"\']$", r"\1", caption)
        caption = re.sub(r"^[\'\_,\-\:;]", r"", caption)
        caption = re.sub(r"[\'\_,\-\:\-\+]$", r"", caption)
        caption = re.sub(r"^\.\S+$", "", caption)

        return caption.strip()

    @torch.no_grad()
    # Copied from diffusers.pipelines.deepfloyd_if.pipeline_if.IFPipeline.encode_prompt
    def encode_prompt(
        self,
        prompt: Union[str, List[str]],
        do_classifier_free_guidance: bool = True,
        num_images_per_prompt: int = 1,
        device: Optional[torch.device] = None,
        negative_prompt: Optional[Union[str, List[str]]] = None,
        prompt_embeds: Optional[torch.Tensor] = None,
        negative_prompt_embeds: Optional[torch.Tensor] = None,
        clean_caption: bool = False,
    ):
        r"""
        Encodes the prompt into text encoder hidden states.

        Args:
            prompt (`str` or `List[str]`, *optional*):
                prompt to be encoded
            do_classifier_free_guidance (`bool`, *optional*, defaults to `True`):
                whether to use classifier free guidance or not
            num_images_per_prompt (`int`, *optional*, defaults to 1):
                number of images that should be generated per prompt
            device: (`torch.device`, *optional*):
                torch device to place the resulting embeddings on
            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. If not defined, one has to pass `negative_prompt_embeds`. instead.
                Ignored when not using guidance (i.e., ignored if `guidance_scale` is less than `1`).
            prompt_embeds (`torch.Tensor`, *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.
            negative_prompt_embeds (`torch.Tensor`, *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.
            clean_caption (bool, defaults to `False`):
                If `True`, the function will preprocess and clean the provided caption before encoding.
        """
        if prompt is not None and negative_prompt is not None:
            if type(prompt) is not type(negative_prompt):
                raise TypeError(
                    f"`negative_prompt` should be the same type to `prompt`, but got {type(negative_prompt)} !="
                    f" {type(prompt)}."
                )

        if device is None:
            device = self._execution_device

        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]

        # while T5 can handle much longer input sequences than 77, the text encoder was trained with a max length of 77 for IF
        max_length = 77

        if prompt_embeds is None:
            prompt = self._text_preprocessing(prompt, clean_caption=clean_caption)
            text_inputs = self.tokenizer(
                prompt,
                padding="max_length",
                max_length=max_length,
                truncation=True,
                add_special_tokens=True,
                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[:, max_length - 1 : -1])
                logger.warning(
                    "The following part of your input was truncated because CLIP can only handle sequences up to"
                    f" {max_length} tokens: {removed_text}"
                )

            attention_mask = text_inputs.attention_mask.to(device)

            prompt_embeds = self.text_encoder(
                text_input_ids.to(device),
                attention_mask=attention_mask,
            )
            prompt_embeds = prompt_embeds[0]

        if self.text_encoder is not None:
            dtype = self.text_encoder.dtype
        elif self.unet is not None:
            dtype = self.unet.dtype
        else:
            dtype = None

        prompt_embeds = prompt_embeds.to(dtype=dtype, device=device)

        bs_embed, seq_len, _ = prompt_embeds.shape
        # duplicate text embeddings 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(bs_embed * num_images_per_prompt, seq_len, -1)

        # get unconditional embeddings for classifier free guidance
        if do_classifier_free_guidance and negative_prompt_embeds is None:
            uncond_tokens: List[str]
            if negative_prompt is None:
                uncond_tokens = [""] * batch_size
            elif isinstance(negative_prompt, str):
                uncond_tokens = [negative_prompt]
            elif batch_size != len(negative_prompt):
                raise ValueError(
                    f"`negative_prompt`: {negative_prompt} has batch size {len(negative_prompt)}, but `prompt`:"
                    f" {prompt} has batch size {batch_size}. Please make sure that passed `negative_prompt` matches"
                    " the batch size of `prompt`."
                )
            else:
                uncond_tokens = negative_prompt

            uncond_tokens = self._text_preprocessing(uncond_tokens, clean_caption=clean_caption)
            max_length = prompt_embeds.shape[1]
            uncond_input = self.tokenizer(
                uncond_tokens,
                padding="max_length",
                max_length=max_length,
                truncation=True,
                return_attention_mask=True,
                add_special_tokens=True,
                return_tensors="pt",
            )
            attention_mask = uncond_input.attention_mask.to(device)

            negative_prompt_embeds = self.text_encoder(
                uncond_input.input_ids.to(device),
                attention_mask=attention_mask,
            )
            negative_prompt_embeds = negative_prompt_embeds[0]

        if do_classifier_free_guidance:
            # duplicate unconditional embeddings for each generation per prompt, using mps friendly method
            seq_len = negative_prompt_embeds.shape[1]

            negative_prompt_embeds = negative_prompt_embeds.to(dtype=dtype, device=device)

            negative_prompt_embeds = negative_prompt_embeds.repeat(1, num_images_per_prompt, 1)
            negative_prompt_embeds = negative_prompt_embeds.view(batch_size * num_images_per_prompt, seq_len, -1)

            # For classifier free guidance, we need to do two forward passes.
            # Here we concatenate the unconditional and text embeddings into a single batch
            # to avoid doing two forward passes
        else:
            negative_prompt_embeds = None

        return prompt_embeds, negative_prompt_embeds

    # Copied from diffusers.pipelines.deepfloyd_if.pipeline_if.IFPipeline.run_safety_checker
    def run_safety_checker(self, image, device, dtype):
        if self.safety_checker is not None:
            safety_checker_input = self.feature_extractor(self.numpy_to_pil(image), return_tensors="pt").to(device)
            image, nsfw_detected, watermark_detected = self.safety_checker(
                images=image,
                clip_input=safety_checker_input.pixel_values.to(dtype=dtype),
            )
        else:
            nsfw_detected = None
            watermark_detected = None

        return image, nsfw_detected, watermark_detected

    # Copied from diffusers.pipelines.deepfloyd_if.pipeline_if.IFPipeline.prepare_extra_step_kwargs
    def prepare_extra_step_kwargs(self, generator, eta):
        # prepare extra kwargs for the scheduler step, since not all schedulers have the same signature
        # eta (η) is only used with the DDIMScheduler, it will be ignored for other schedulers.
        # eta corresponds to η in DDIM paper: https://arxiv.org/abs/2010.02502
        # and should be between [0, 1]

        accepts_eta = "eta" in set(inspect.signature(self.scheduler.step).parameters.keys())
        extra_step_kwargs = {}
        if accepts_eta:
            extra_step_kwargs["eta"] = eta

        # check if the scheduler accepts generator
        accepts_generator = "generator" in set(inspect.signature(self.scheduler.step).parameters.keys())
        if accepts_generator:
            extra_step_kwargs["generator"] = generator
        return extra_step_kwargs

    def check_inputs(
        self,
        prompt,
        image,
        original_image,
        mask_image,
        batch_size,
        callback_steps,
        negative_prompt=None,
        prompt_embeds=None,
        negative_prompt_embeds=None,
    ):
        if (callback_steps is None) or (
            callback_steps is not None and (not isinstance(callback_steps, int) or callback_steps <= 0)
        ):
            raise ValueError(
                f"`callback_steps` has to be a positive integer but is {callback_steps} of type"
                f" {type(callback_steps)}."
            )

        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 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)}")

        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."
            )

        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}."
                )

        # image

        if isinstance(image, list):
            check_image_type = image[0]
        else:
            check_image_type = image

        if (
            not isinstance(check_image_type, torch.Tensor)
            and not isinstance(check_image_type, PIL.Image.Image)
            and not isinstance(check_image_type, np.ndarray)
        ):
            raise ValueError(
                "`image` has to be of type `torch.Tensor`, `PIL.Image.Image`, `np.ndarray`, or List[...] but is"
                f" {type(check_image_type)}"
            )

        if isinstance(image, list):
            image_batch_size = len(image)
        elif isinstance(image, torch.Tensor):
            image_batch_size = image.shape[0]
        elif isinstance(image, PIL.Image.Image):
            image_batch_size = 1
        elif isinstance(image, np.ndarray):
            image_batch_size = image.shape[0]
        else:
            assert False

        if batch_size != image_batch_size:
            raise ValueError(f"image batch size: {image_batch_size} must be same as prompt batch size {batch_size}")

        # original_image

        if isinstance(original_image, list):
            check_image_type = original_image[0]
        else:
            check_image_type = original_image

        if (
            not isinstance(check_image_type, torch.Tensor)
            and not isinstance(check_image_type, PIL.Image.Image)
            and not isinstance(check_image_type, np.ndarray)
        ):
            raise ValueError(
                "`original_image` has to be of type `torch.Tensor`, `PIL.Image.Image`, `np.ndarray`, or List[...] but is"
                f" {type(check_image_type)}"
            )

        if isinstance(original_image, list):
            image_batch_size = len(original_image)
        elif isinstance(original_image, torch.Tensor):
            image_batch_size = original_image.shape[0]
        elif isinstance(original_image, PIL.Image.Image):
            image_batch_size = 1
        elif isinstance(original_image, np.ndarray):
            image_batch_size = original_image.shape[0]
        else:
            assert False

        if batch_size != image_batch_size:
            raise ValueError(
                f"original_image batch size: {image_batch_size} must be same as prompt batch size {batch_size}"
            )

        # mask_image

        if isinstance(mask_image, list):
            check_image_type = mask_image[0]
        else:
            check_image_type = mask_image

        if (
            not isinstance(check_image_type, torch.Tensor)
            and not isinstance(check_image_type, PIL.Image.Image)
            and not isinstance(check_image_type, np.ndarray)
        ):
            raise ValueError(
                "`mask_image` has to be of type `torch.Tensor`, `PIL.Image.Image`, `np.ndarray`, or List[...] but is"
                f" {type(check_image_type)}"
            )

        if isinstance(mask_image, list):
            image_batch_size = len(mask_image)
        elif isinstance(mask_image, torch.Tensor):
            image_batch_size = mask_image.shape[0]
        elif isinstance(mask_image, PIL.Image.Image):
            image_batch_size = 1
        elif isinstance(mask_image, np.ndarray):
            image_batch_size = mask_image.shape[0]
        else:
            assert False

        if image_batch_size != 1 and batch_size != image_batch_size:
            raise ValueError(
                f"mask_image batch size: {image_batch_size} must be `1` or the same as prompt batch size {batch_size}"
            )

    # Copied from diffusers.pipelines.deepfloyd_if.pipeline_if_img2img.IFImg2ImgPipeline.preprocess_image with preprocess_image -> preprocess_original_image
    def preprocess_original_image(self, image: PIL.Image.Image) -> torch.Tensor:
        if not isinstance(image, list):
            image = [image]

        def numpy_to_pt(images):
            if images.ndim == 3:
                images = images[..., None]

            images = torch.from_numpy(images.transpose(0, 3, 1, 2))
            return images

        if isinstance(image[0], PIL.Image.Image):
            new_image = []

            for image_ in image:
                image_ = image_.convert("RGB")
                image_ = resize(image_, self.unet.config.sample_size)
                image_ = np.array(image_)
                image_ = image_.astype(np.float32)
                image_ = image_ / 127.5 - 1
                new_image.append(image_)

            image = new_image

            image = np.stack(image, axis=0)  # to np
            image = numpy_to_pt(image)  # to pt

        elif isinstance(image[0], np.ndarray):
            image = np.concatenate(image, axis=0) if image[0].ndim == 4 else np.stack(image, axis=0)
            image = numpy_to_pt(image)

        elif isinstance(image[0], torch.Tensor):
            image = torch.cat(image, axis=0) if image[0].ndim == 4 else torch.stack(image, axis=0)

        return image

    # Copied from diffusers.pipelines.deepfloyd_if.pipeline_if_superresolution.IFSuperResolutionPipeline.preprocess_image
    def preprocess_image(self, image: PIL.Image.Image, num_images_per_prompt, device) -> torch.Tensor:
        if not isinstance(image, torch.Tensor) and not isinstance(image, list):
            image = [image]

        if isinstance(image[0], PIL.Image.Image):
            image = [np.array(i).astype(np.float32) / 127.5 - 1.0 for i in image]

            image = np.stack(image, axis=0)  # to np
            image = torch.from_numpy(image.transpose(0, 3, 1, 2))
        elif isinstance(image[0], np.ndarray):
            image = np.stack(image, axis=0)  # to np
            if image.ndim == 5:
                image = image[0]

            image = torch.from_numpy(image.transpose(0, 3, 1, 2))
        elif isinstance(image, list) and isinstance(image[0], torch.Tensor):
            dims = image[0].ndim

            if dims == 3:
                image = torch.stack(image, dim=0)
            elif dims == 4:
                image = torch.concat(image, dim=0)
            else:
                raise ValueError(f"Image must have 3 or 4 dimensions, instead got {dims}")

        image = image.to(device=device, dtype=self.unet.dtype)

        image = image.repeat_interleave(num_images_per_prompt, dim=0)

        return image

    # Copied from diffusers.pipelines.deepfloyd_if.pipeline_if_inpainting.IFInpaintingPipeline.preprocess_mask_image
    def preprocess_mask_image(self, mask_image) -> torch.Tensor:
        if not isinstance(mask_image, list):
            mask_image = [mask_image]

        if isinstance(mask_image[0], torch.Tensor):
            mask_image = torch.cat(mask_image, axis=0) if mask_image[0].ndim == 4 else torch.stack(mask_image, axis=0)

            if mask_image.ndim == 2:
                # Batch and add channel dim for single mask
                mask_image = mask_image.unsqueeze(0).unsqueeze(0)
            elif mask_image.ndim == 3 and mask_image.shape[0] == 1:
                # Single mask, the 0'th dimension is considered to be
                # the existing batch size of 1
                mask_image = mask_image.unsqueeze(0)
            elif mask_image.ndim == 3 and mask_image.shape[0] != 1:
                # Batch of mask, the 0'th dimension is considered to be
                # the batching dimension
                mask_image = mask_image.unsqueeze(1)

            mask_image[mask_image < 0.5] = 0
            mask_image[mask_image >= 0.5] = 1

        elif isinstance(mask_image[0], PIL.Image.Image):
            new_mask_image = []

            for mask_image_ in mask_image:
                mask_image_ = mask_image_.convert("L")
                mask_image_ = resize(mask_image_, self.unet.config.sample_size)
                mask_image_ = np.array(mask_image_)
                mask_image_ = mask_image_[None, None, :]
                new_mask_image.append(mask_image_)

            mask_image = new_mask_image

            mask_image = np.concatenate(mask_image, axis=0)
            mask_image = mask_image.astype(np.float32) / 255.0
            mask_image[mask_image < 0.5] = 0
            mask_image[mask_image >= 0.5] = 1
            mask_image = torch.from_numpy(mask_image)

        elif isinstance(mask_image[0], np.ndarray):
            mask_image = np.concatenate([m[None, None, :] for m in mask_image], axis=0)

            mask_image[mask_image < 0.5] = 0
            mask_image[mask_image >= 0.5] = 1
            mask_image = torch.from_numpy(mask_image)

        return mask_image

    # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion_img2img.StableDiffusionImg2ImgPipeline.get_timesteps
    def get_timesteps(self, num_inference_steps, strength):
        # get the original timestep using init_timestep
        init_timestep = min(int(num_inference_steps * strength), num_inference_steps)

        t_start = 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.deepfloyd_if.pipeline_if_inpainting.IFInpaintingPipeline.prepare_intermediate_images
    def prepare_intermediate_images(
        self, image, timestep, batch_size, num_images_per_prompt, dtype, device, mask_image, generator=None
    ):
        image_batch_size, channels, height, width = image.shape

        batch_size = batch_size * num_images_per_prompt

        shape = (batch_size, channels, height, width)

        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."
            )

        noise = randn_tensor(shape, generator=generator, device=device, dtype=dtype)

        image = image.repeat_interleave(num_images_per_prompt, dim=0)
        noised_image = self.scheduler.add_noise(image, noise, timestep)

        image = (1 - mask_image) * image + mask_image * noised_image

        return image

    @torch.no_grad()
    @replace_example_docstring(EXAMPLE_DOC_STRING)
    def __call__(
        self,
        image: Union[PIL.Image.Image, np.ndarray, torch.Tensor],
        original_image: Union[
            PIL.Image.Image, torch.Tensor, np.ndarray, List[PIL.Image.Image], List[torch.Tensor], List[np.ndarray]
        ] = None,
        mask_image: Union[
            PIL.Image.Image, torch.Tensor, np.ndarray, List[PIL.Image.Image], List[torch.Tensor], List[np.ndarray]
        ] = None,
        strength: float = 0.8,
        prompt: Union[str, List[str]] = None,
        num_inference_steps: int = 100,
        timesteps: List[int] = None,
        guidance_scale: float = 4.0,
        negative_prompt: Optional[Union[str, List[str]]] = None,
        num_images_per_prompt: Optional[int] = 1,
        eta: float = 0.0,
        generator: Optional[Union[torch.Generator, List[torch.Generator]]] = None,
        prompt_embeds: Optional[torch.Tensor] = None,
        negative_prompt_embeds: Optional[torch.Tensor] = None,
        output_type: Optional[str] = "pil",
        return_dict: bool = True,
        callback: Optional[Callable[[int, int, torch.Tensor], None]] = None,
        callback_steps: int = 1,
        cross_attention_kwargs: Optional[Dict[str, Any]] = None,
        noise_level: int = 0,
        clean_caption: bool = True,
    ):
        """
        Function invoked when calling the pipeline for generation.

        Args:
            image (`torch.Tensor` or `PIL.Image.Image`):
                `Image`, or tensor representing an image batch, that will be used as the starting point for the
                process.
            original_image (`torch.Tensor` or `PIL.Image.Image`):
                The original image that `image` was varied from.
            mask_image (`PIL.Image.Image`):
                `Image`, or tensor representing an image batch, to mask `image`. White pixels in the mask will be
                repainted, while black pixels will be preserved. If `mask_image` is a PIL image, it will be converted
                to a single channel (luminance) before use. If it's a tensor, it should contain one color channel (L)
                instead of 3, so the expected shape would be `(B, H, W, 1)`.
            strength (`float`, *optional*, defaults to 0.8):
                Conceptually, indicates how much to transform the reference `image`. Must be between 0 and 1. `image`
                will be used as a starting point, adding more noise to it the larger the `strength`. The number of
                denoising steps depends on the amount of noise initially added. When `strength` is 1, added noise will
                be maximum and the denoising process will run for the full number of iterations specified in
                `num_inference_steps`. A value of 1, therefore, essentially ignores `image`.
            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.
            num_inference_steps (`int`, *optional*, defaults to 100):
                The number of denoising steps. More denoising steps usually lead to a higher quality image at the
                expense of slower inference.
            timesteps (`List[int]`, *optional*):
                Custom timesteps to use for the denoising process. If not defined, equal spaced `num_inference_steps`
                timesteps are used. Must be in descending order.
            guidance_scale (`float`, *optional*, defaults to 4.0):
                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
                `guidance_scale` is defined as `w` of equation 2. of [Imagen
                Paper](https://arxiv.org/pdf/2205.11487.pdf). 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.
            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 `guidance_scale` is
                less than `1`).
            num_images_per_prompt (`int`, *optional*, defaults to 1):
                The number of images to generate per prompt.
            eta (`float`, *optional*, defaults to 0.0):
                Corresponds to parameter eta (η) in the DDIM paper: https://arxiv.org/abs/2010.02502. Only applies to
                [`schedulers.DDIMScheduler`], will be ignored for others.
            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.
            prompt_embeds (`torch.Tensor`, *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.
            negative_prompt_embeds (`torch.Tensor`, *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.
            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.stable_diffusion.IFPipelineOutput`] instead of a plain tuple.
            callback (`Callable`, *optional*):
                A function that will be called every `callback_steps` steps during inference. The function will be
                called with the following arguments: `callback(step: int, timestep: int, latents: torch.Tensor)`.
            callback_steps (`int`, *optional*, defaults to 1):
                The frequency at which the `callback` function will be called. If not specified, the callback will be
                called at every step.
            cross_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).
            noise_level (`int`, *optional*, defaults to 0):
                The amount of noise to add to the upscaled image. Must be in the range `[0, 1000)`
            clean_caption (`bool`, *optional*, defaults to `True`):
                Whether or not to clean the caption before creating embeddings. Requires `beautifulsoup4` and `ftfy` to
                be installed. If the dependencies are not installed, the embeddings will be created from the raw
                prompt.

        Examples:

        Returns:
            [`~pipelines.stable_diffusion.IFPipelineOutput`] or `tuple`:
            [`~pipelines.stable_diffusion.IFPipelineOutput`] if `return_dict` is True, otherwise a `tuple. When
            returning a tuple, the first element is a list with the generated images, and the second element is a list
            of `bool`s denoting whether the corresponding generated image likely represents "not-safe-for-work" (nsfw)
            or watermarked content, according to the `safety_checker`.
        """
        # 1. Check inputs. Raise error if not correct
        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]

        self.check_inputs(
            prompt,
            image,
            original_image,
            mask_image,
            batch_size,
            callback_steps,
            negative_prompt,
            prompt_embeds,
            negative_prompt_embeds,
        )

        # 2. Define call parameters

        # here `guidance_scale` is defined analog to the guidance weight `w` of equation (2)
        # of the Imagen paper: https://arxiv.org/pdf/2205.11487.pdf . `guidance_scale = 1`
        # corresponds to doing no classifier free guidance.
        do_classifier_free_guidance = guidance_scale > 1.0

        device = self._execution_device

        # 3. Encode input prompt
        prompt_embeds, negative_prompt_embeds = self.encode_prompt(
            prompt,
            do_classifier_free_guidance,
            num_images_per_prompt=num_images_per_prompt,
            device=device,
            negative_prompt=negative_prompt,
            prompt_embeds=prompt_embeds,
            negative_prompt_embeds=negative_prompt_embeds,
            clean_caption=clean_caption,
        )

        if do_classifier_free_guidance:
            prompt_embeds = torch.cat([negative_prompt_embeds, prompt_embeds])

        dtype = prompt_embeds.dtype

        # 4. Prepare timesteps
        if timesteps is not None:
            self.scheduler.set_timesteps(timesteps=timesteps, device=device)
            timesteps = self.scheduler.timesteps
            num_inference_steps = len(timesteps)
        else:
            self.scheduler.set_timesteps(num_inference_steps, device=device)
            timesteps = self.scheduler.timesteps

        timesteps, num_inference_steps = self.get_timesteps(num_inference_steps, strength)

        # 5. prepare original image
        original_image = self.preprocess_original_image(original_image)
        original_image = original_image.to(device=device, dtype=dtype)

        # 6. prepare mask image
        mask_image = self.preprocess_mask_image(mask_image)
        mask_image = mask_image.to(device=device, dtype=dtype)

        if mask_image.shape[0] == 1:
            mask_image = mask_image.repeat_interleave(batch_size * num_images_per_prompt, dim=0)
        else:
            mask_image = mask_image.repeat_interleave(num_images_per_prompt, dim=0)

        # 6. Prepare intermediate images
        noise_timestep = timesteps[0:1]
        noise_timestep = noise_timestep.repeat(batch_size * num_images_per_prompt)

        intermediate_images = self.prepare_intermediate_images(
            original_image,
            noise_timestep,
            batch_size,
            num_images_per_prompt,
            dtype,
            device,
            mask_image,
            generator,
        )

        # 7. Prepare upscaled image and noise level
        _, _, height, width = original_image.shape

        image = self.preprocess_image(image, num_images_per_prompt, device)

        upscaled = F.interpolate(image, (height, width), mode="bilinear", align_corners=True)

        noise_level = torch.tensor([noise_level] * upscaled.shape[0], device=upscaled.device)
        noise = randn_tensor(upscaled.shape, generator=generator, device=upscaled.device, dtype=upscaled.dtype)
        upscaled = self.image_noising_scheduler.add_noise(upscaled, noise, timesteps=noise_level)

        if do_classifier_free_guidance:
            noise_level = torch.cat([noise_level] * 2)

        # 8. Prepare extra step kwargs. TODO: Logic should ideally just be moved out of the pipeline
        extra_step_kwargs = self.prepare_extra_step_kwargs(generator, eta)

        # HACK: see comment in `enable_model_cpu_offload`
        if hasattr(self, "text_encoder_offload_hook") and self.text_encoder_offload_hook is not None:
            self.text_encoder_offload_hook.offload()

        # 9. Denoising loop
        num_warmup_steps = len(timesteps) - num_inference_steps * self.scheduler.order
        with self.progress_bar(total=num_inference_steps) as progress_bar:
            for i, t in enumerate(timesteps):
                model_input = torch.cat([intermediate_images, upscaled], dim=1)

                model_input = torch.cat([model_input] * 2) if do_classifier_free_guidance else model_input
                model_input = self.scheduler.scale_model_input(model_input, t)

                # predict the noise residual
                noise_pred = self.unet(
                    model_input,
                    t,
                    encoder_hidden_states=prompt_embeds,
                    class_labels=noise_level,
                    cross_attention_kwargs=cross_attention_kwargs,
                    return_dict=False,
                )[0]

                # perform guidance
                if do_classifier_free_guidance:
                    noise_pred_uncond, noise_pred_text = noise_pred.chunk(2)
                    noise_pred_uncond, _ = noise_pred_uncond.split(model_input.shape[1] // 2, dim=1)
                    noise_pred_text, predicted_variance = noise_pred_text.split(model_input.shape[1] // 2, dim=1)
                    noise_pred = noise_pred_uncond + guidance_scale * (noise_pred_text - noise_pred_uncond)
                    noise_pred = torch.cat([noise_pred, predicted_variance], dim=1)

                if self.scheduler.config.variance_type not in ["learned", "learned_range"]:
                    noise_pred, _ = noise_pred.split(intermediate_images.shape[1], dim=1)

                # compute the previous noisy sample x_t -> x_t-1
                prev_intermediate_images = intermediate_images

                intermediate_images = self.scheduler.step(
                    noise_pred, t, intermediate_images, **extra_step_kwargs, return_dict=False
                )[0]

                intermediate_images = (1 - mask_image) * prev_intermediate_images + mask_image * intermediate_images

                # 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 callback is not None and i % callback_steps == 0:
                        callback(i, t, intermediate_images)

        image = intermediate_images

        if output_type == "pil":
            # 10. Post-processing
            image = (image / 2 + 0.5).clamp(0, 1)
            image = image.cpu().permute(0, 2, 3, 1).float().numpy()

            # 11. Run safety checker
            image, nsfw_detected, watermark_detected = self.run_safety_checker(image, device, prompt_embeds.dtype)

            # 12. Convert to PIL
            image = self.numpy_to_pil(image)

            # 13. Apply watermark
            if self.watermarker is not None:
                self.watermarker.apply_watermark(image, self.unet.config.sample_size)
        elif output_type == "pt":
            nsfw_detected = None
            watermark_detected = None

        else:
            # 10. Post-processing
            image = (image / 2 + 0.5).clamp(0, 1)
            image = image.cpu().permute(0, 2, 3, 1).float().numpy()

            # 11. Run safety checker
            image, nsfw_detected, watermark_detected = self.run_safety_checker(image, device, prompt_embeds.dtype)

        self.maybe_free_model_hooks()

        if not return_dict:
            return (image, nsfw_detected, watermark_detected)

        return IFPipelineOutput(images=image, nsfw_detected=nsfw_detected, watermark_detected=watermark_detected)