model/pipeline.py

import inspect
import os
from typing import Any, Callable, Dict, List, Optional, Union
import json
import sys
sys.path.append((os.path.dirname(__file__)))

import PIL
import PIL.Image
import numpy as np
import torch
from accelerate import load_checkpoint_in_model
from diffusers.utils.torch_utils import randn_tensor
from diffusers import FluxKontextPipeline
from diffusers.image_processor import PipelineImageInput
from diffusers.pipelines.flux.pipeline_output import FluxPipelineOutput
from diffusers.pipelines.flux.pipeline_flux_kontext import calculate_shift, retrieve_timesteps

from model.utils import compute_vae_encodings
from utils import prepare_image


class FluxKontextImg2ImgLoRAPipeline(FluxKontextPipeline):

    def get_base_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
        # 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)

        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."
            )

        if latents is None:
            latents = randn_tensor(shape, generator=generator, device=device, dtype=dtype)
        else:
            latents = latents.to(device)

        return latents
    
    @torch.no_grad()
    def __call__(self,
        image: Union[PIL.Image.Image, torch.Tensor],
        condition_images: List[Union[PIL.Image.Image, torch.Tensor]],
        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,
        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 used as the starting point. 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.
            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.
            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):
                Embedded guidance scale is enabled by setting `guidance_scale` > 1. Higher `guidance_scale` encourages
                a model to generate images more aligned with prompt at the expense of lower image quality.

                Guidance-distilled models approximates true classifier-free guidance for `guidance_scale` > 1. Refer to
                the [paper](https://huggingface.co/papers/2210.03142) to learn more.
            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

        # 1. Check inputs. Raise error if not correct
        self.check_inputs(
            prompt,
            prompt_2,
            height,
            width,
            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,
            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. 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,
            )

        # image, condition_image, mask = self.check_inputs(image, condition_image, mask, width, height)
        image = prepare_image(image).to(self.transformer.device, dtype=self.transformer.dtype)
        # condition_image_1 = prepare_image(condition_image_1).to(self.transformer.device, dtype=self.transformer.dtype)
        # condition_image_2 = prepare_image(condition_image_2).to(self.transformer.device, dtype=self.transformer.dtype)
        condition_images = [prepare_image(ci).to(self.transformer.device, dtype=self.transformer.dtype) for ci in condition_images]
        # VAE encoding
        # condition_1_latent = compute_vae_encodings(condition_image_1, self.vae)
        # condition_2_latent = compute_vae_encodings(condition_image_2, self.vae)
        # condition_latent = torch.cat([condition_1_latent, condition_2_latent], dim=2)
        condition_latents = [compute_vae_encodings(ci, self.vae, sample_mode="argmax") for ci in condition_images]
        # image_latent = compute_vae_encodings(image, self.vae)
        del condition_images
        # # Concatenate latents
        # cond_latents = torch.cat([
        #     image_latent,
        #     dp_latent,
        #     # condition_latent
        # ], dim=2)

        # 4. Prepare latent variables
        num_channels_latents = self.transformer.config.in_channels // 4
        latents, image_latents, latent_ids, image_ids = self.prepare_latents(
            image,
            batch_size * num_images_per_prompt,
            num_channels_latents,
            height,
            width,
            prompt_embeds.dtype,
            device,
            generator,
            latents,
        )

        # packed_dp_latent = self._pack_latents(
        #     dp_latent,
        #     dp_latent.shape[0],
        #     dp_latent.shape[1],
        #     dp_latent.shape[2],
        #     dp_latent.shape[3]
        # )
        # dp_latents, _, dp_latent_ids, _ = self.prepare_latents(
        #     None,
        #     dp_latent.shape[0],
        #     dp_latent.shape[1],
        #     dp_latent.shape[2] * self.vae_scale_factor,
        #     dp_latent.shape[3] * self.vae_scale_factor,
        #     prompt_embeds.dtype,
        #     device,
        #     generator,
        #     packed_dp_latent,
        # )
        # dp_latent_ids[:, 1] = dp_latent_ids[:, 1] + (int(height*2) // (self.vae_scale_factor * 2))
        # dp_ids = dp_latent_ids.clone()
        # dp_ids[..., 0] = 1

        packed_conds_latents = []
        cond_ids = []
        cond_latents = []
        cond_latents_ids = []
        for idx, condition_latent in enumerate(condition_latents):
            packed_conds_latent = self._pack_latents(
                condition_latent,
                condition_latent.shape[0],
                condition_latent.shape[1],
                condition_latent.shape[2],
                condition_latent.shape[3]
            )
            packed_conds_latents.append(packed_conds_latent)
            cond_latent, _, cond_latent_ids, _ = self.prepare_latents(
                None,
                condition_latent.shape[0],
                condition_latent.shape[1],
                condition_latent.shape[2] * self.vae_scale_factor,
                condition_latent.shape[3] * self.vae_scale_factor,
                prompt_embeds.dtype,
                device,
                generator,
                packed_conds_latent,
            )
            cond_latents.append(cond_latent)
            cond_id = cond_latent_ids.clone()
            # shift cond ids by condition image size
            cond_id[:, 1] = cond_id[:, 1] * (idx + 1) + (int(height) // (self.vae_scale_factor * 2))
            cond_id[:, 2] = cond_id[:, 2] * (idx + 1) + (int(width) // (self.vae_scale_factor * 2))
            cond_id[..., 0] = 1 # noise_latent is 0, image_latent is 1, cond_latents start from 2
            cond_ids.append(cond_id)
            cond_latent_ids[:, 1] = cond_latent_ids[:, 1] * (idx + 1) + (int(height) // (self.vae_scale_factor * 2))
            cond_latents_ids.append(cond_latent_ids)

        # concat all latent ids & image ids
        # latent_ids = torch.cat([
        #     latent_ids,
        #     dp_latent_ids,
        #     cond_latent_ids,
        # ], dim=0
        # )
        image_ids = torch.cat([
            image_ids,
            # dp_ids,
            *cond_ids
        ], dim=0
        )
        # latent_ids = torch.cat([latent_ids, *cond_latents_ids], dim=0)
        latent_ids = torch.cat([latent_ids, image_ids], dim=0)  # dim 0 is sequence dimension

        n_out_tokens = latents.shape[1]
        # latents = torch.cat([latents, *cond_latents], dim=1)
        # latent_id_len = latents.shape[1]//4
        # for i in range(1,4):
        #     w_offset = (int(width) // (self.vae_scale_factor * 2)) * i
        #     latent_image_ids[:,(latent_id_len*i) : int(latent_id_len*(i+1)), 2] = \
        #         latent_image_ids[:,(latent_id_len*i) : int(latent_id_len*(i+1)), 2] + w_offset
        #     h_offset = (int(height) // (self.vae_scale_factor * 2)) * i
        #     latent_image_ids[:,(latent_id_len*i) : int(latent_id_len*(i+1)), 1] = \
        #         latent_image_ids[:,(latent_id_len*i) : int(latent_id_len*(i+1)), 1] - h_offset

        # 5. Prepare timesteps
        sigmas = np.linspace(1.0, 1 / num_inference_steps, num_inference_steps) if sigmas is None else sigmas
        image_seq_len = latents.shape[1]
        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,
        )
        num_warmup_steps = max(len(timesteps) - num_inference_steps * self.scheduler.order, 0)
        self._num_timesteps = len(timesteps)

        # 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
        # We set the index here to remove DtoH sync, helpful especially during compilation.
        # Check out more details here: https://github.com/huggingface/diffusers/pull/11696
        self.scheduler.set_begin_index(0)        
        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_latents is not None:
                latent_model_input = torch.cat(
                    [latents, image_latents, 
                    #  dp_latents, 
                     *cond_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)
                # else:
                #     if i < 1:
                #         noise_pred = noise_pred * 0

                # 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]

                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[:,:n_out_tokens], 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)