train_phased_sdxl.py

import argparse
import copy
import functools
import gc
import itertools
import json
import logging
import math
import os
import random
import shutil
from pathlib import Path
from typing import List, Union

import accelerate
import numpy as np
import torch
import torch.nn.functional as F
import torch.utils.checkpoint
import torchvision.transforms.functional as TF
import transformers
from accelerate import Accelerator
from accelerate.logging import get_logger
from accelerate.utils import ProjectConfiguration, set_seed
from huggingface_hub import create_repo
from packaging import version
from peft import LoraConfig, get_peft_model, get_peft_model_state_dict
from torch.utils.data import default_collate
from torchvision import transforms
from tqdm.auto import tqdm
from transformers import AutoTokenizer, PretrainedConfig
import diffusers
from diffusers import (
    AutoencoderKL,
    DDPMScheduler,
    DDIMScheduler,
    LCMScheduler,
    StableDiffusionXLPipeline,
    UNet2DConditionModel,
)
from diffusers.optimization import get_scheduler
from diffusers.utils import check_min_version, is_wandb_available
from diffusers.utils.import_utils import is_xformers_available
from torch.utils.data import DataLoader, Dataset
from PIL import Image


MAX_SEQ_LENGTH = 77

if is_wandb_available():
    import wandb

check_min_version("0.18.0.dev0")

logger = get_logger(__name__)



def get_module_kohya_state_dict(
    module, prefix: str, dtype: torch.dtype, adapter_name: str = "default"
):
    kohya_ss_state_dict = {}
    for peft_key, weight in get_peft_model_state_dict(
        module, adapter_name=adapter_name
    ).items():
        kohya_key = peft_key.replace("base_model.model", prefix)
        kohya_key = kohya_key.replace("lora_A", "lora_down")
        kohya_key = kohya_key.replace("lora_B", "lora_up")
        kohya_key = kohya_key.replace(".", "_", kohya_key.count(".") - 2)
        kohya_ss_state_dict[kohya_key] = weight.to(dtype)

        if "lora_down" in kohya_key:
            alpha_key = f'{kohya_key.split(".")[0]}.alpha'
            kohya_ss_state_dict[alpha_key] = torch.tensor(
                module.peft_config[adapter_name].lora_alpha
            ).to(dtype)

    return kohya_ss_state_dict


class CustomImageDataset(Dataset):
    def __init__(self, img_dir, sample_size):
        """
        Args:
            img_dir (string): Directory with all the images and text files.
            sample_size (tuple): Desired sample size as (height, width).
        """
        self.img_dir = img_dir
        self.sample_size = sample_size

        self.img_names = [
            f for f in tqdm(os.listdir(img_dir)) if f.endswith((".png", ".jpg"))
        ]
        print("finish load meta file")
        self.transform = transforms.Compose(
            [
                transforms.Resize(
                    self.sample_size, interpolation=transforms.InterpolationMode.LANCZOS
                ),
                transforms.CenterCrop(self.sample_size),
                transforms.ToTensor(),
                transforms.Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5]),
            ]
        )

    def __len__(self):
        return len(self.img_names)

    def __getitem__(self, idx):

        img_name = self.img_names[idx]
        img_path = os.path.join(self.img_dir, img_name)
        image = Image.open(img_path).convert("RGB")
        image = TF.resize(
            image, self.sample_size, interpolation=transforms.InterpolationMode.LANCZOS
        )
        c_top, c_left, _, _ = transforms.RandomCrop.get_params(
            image, output_size=(self.sample_size, self.sample_size)
        )
        image = TF.crop(image, c_top, c_left, self.sample_size, self.sample_size)
        image = TF.to_tensor(image)
        image = TF.normalize(image, [0.5], [0.5])
        text_name = img_name.rsplit(".", 1)[0] + ".txt"
        text_path = os.path.join(self.img_dir, text_name)
        with open(text_path, "r") as f:
            text = f.read().strip()
        return image, text, (self.sample_size, self.sample_size), (c_top, c_left)


def log_validation(
    vae, unet, args, accelerator, weight_dtype, step, inference_steps, cfg
):
    logger.info("Running validation... ")

    unet = accelerator.unwrap_model(unet)
    pipeline = StableDiffusionXLPipeline.from_pretrained(
        args.pretrained_teacher_model,
        vae=vae,
        unet=unet,
        scheduler=DDIMScheduler(
            num_train_timesteps=1000,
            beta_start=0.00085,
            beta_end=0.012,
            beta_schedule="scaled_linear",
            timestep_spacing="trailing",
            clip_sample=False,
            set_alpha_to_one=False,
        ),
    )
    pipeline.set_progress_bar_config(disable=True)
    pipeline = pipeline.to(accelerator.device)
    pipeline.enable_vae_slicing()

    if args.enable_xformers_memory_efficient_attention:
        pipeline.enable_xformers_memory_efficient_attention()

    if args.seed is None:
        generator = None
    else:
        generator = torch.Generator(device=accelerator.device).manual_seed(args.seed)

    validation_prompts = [
        "portrait photo of a girl, photograph, highly detailed face, depth of field, moody light, golden hour, style by Dan Winters, Russell James, Steve McCurry, centered, extremely detailed, Nikon D850, award winning photography",
        "Self-portrait oil painting, a beautiful cyborg with golden hair, 8k",
        "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k",
        "A photo of beautiful mountain with realistic sunset and blue lake, highly detailed, masterpiece",
    ]

    image_logs = []

    for _, prompt in enumerate(validation_prompts):
        images = []
        with torch.autocast("cuda", dtype=weight_dtype):
            images = pipeline(
                prompt=prompt,
                num_inference_steps=inference_steps,
                num_images_per_prompt=4,
                generator=generator,
                guidance_scale=cfg,
            ).images
        image_logs.append({"validation_prompt": prompt, "images": images})

    for tracker in accelerator.trackers:
        if tracker.name == "tensorboard":
            for log in image_logs:
                images = log["images"]
                validation_prompt = log["validation_prompt"]
                formatted_images = []
                for image in images:
                    formatted_images.append(np.asarray(image))

                formatted_images = np.stack(formatted_images)

                tracker.writer.add_images(
                    validation_prompt, formatted_images, step, dataformats="NHWC"
                )
        elif tracker.name == "wandb":
            formatted_images = []

            for log in image_logs:
                images = log["images"]
                validation_prompt = log["validation_prompt"]
                for image in images:
                    image = wandb.Image(image, caption=validation_prompt)
                    formatted_images.append(image)

            tracker.log({f"validation-{cfg}-{inference_steps}": formatted_images})
        else:
            logger.warn(f"image logging not implemented for {tracker.name}")

        del pipeline
        gc.collect()
        torch.cuda.empty_cache()

        return image_logs


def append_dims(x, target_dims):
    """Appends dimensions to the end of a tensor until it has target_dims dimensions."""
    dims_to_append = target_dims - x.ndim
    if dims_to_append < 0:
        raise ValueError(
            f"input has {x.ndim} dims but target_dims is {target_dims}, which is less"
        )
    return x[(...,) + (None,) * dims_to_append]


def scalings_for_boundary_conditions_target(index, selected_indices):
    c_skip = torch.isin(index, selected_indices).float()
    c_out = 1.0 - c_skip
    return c_skip, c_out


def scalings_for_boundary_conditions_online(index, selected_indices):
    c_skip = torch.zeros_like(index).float()
    c_out = torch.ones_like(index).float()
    return c_skip, c_out


def scalings_for_boundary_conditions(timestep, sigma_data=0.5, timestep_scaling=10.0):
    c_skip = sigma_data**2 / ((timestep / 0.1) ** 2 + sigma_data**2)
    c_out = (timestep / 0.1) / ((timestep / 0.1) ** 2 + sigma_data**2) ** 0.5
    return c_skip, c_out


def predicted_origin(model_output, timesteps, sample, prediction_type, alphas, sigmas):
    if prediction_type == "epsilon":
        sigmas = extract_into_tensor(sigmas, timesteps, sample.shape)
        alphas = extract_into_tensor(alphas, timesteps, sample.shape)
        pred_x_0 = (sample - sigmas * model_output) / alphas
    elif prediction_type == "v_prediction":
        sigmas = extract_into_tensor(sigmas, timesteps, sample.shape)
        alphas = extract_into_tensor(alphas, timesteps, sample.shape)
        pred_x_0 = alphas * sample - sigmas * model_output
    else:
        raise ValueError(f"Prediction type {prediction_type} currently not supported.")

    return pred_x_0


def extract_into_tensor(a, t, x_shape):
    b, *_ = t.shape
    out = a.gather(-1, t)
    return out.reshape(b, *((1,) * (len(x_shape) - 1)))


@torch.no_grad()
def update_ema(target_params, source_params, rate=0.99):
    """
    Update target parameters to be closer to those of source parameters using
    an exponential moving average.

    :param target_params: the target parameter sequence.
    :param source_params: the source parameter sequence.
    :param rate: the EMA rate (closer to 1 means slower).
    """
    for targ, src in zip(target_params, source_params):
        targ.detach().mul_(rate).add_(src, alpha=1 - rate)


class DDIMSolver:
    def __init__(self, alpha_cumprods, timesteps=1000, ddim_timesteps=40):
        self.step_ratio = timesteps // ddim_timesteps
        self.ddim_timesteps = (
            np.arange(1, ddim_timesteps + 1) * self.step_ratio
        ).round().astype(np.int64) - 1
        self.ddim_alpha_cumprods = alpha_cumprods[self.ddim_timesteps]
        self.ddim_timesteps_prev = np.asarray([0] + self.ddim_timesteps[:-1].tolist())
        self.ddim_alpha_cumprods_prev = np.asarray(
            [alpha_cumprods[0]] + alpha_cumprods[self.ddim_timesteps[:-1]].tolist()
        )
        self.ddim_timesteps = torch.from_numpy(self.ddim_timesteps).long()
        self.ddim_timesteps_prev = torch.from_numpy(self.ddim_timesteps_prev).long()
        self.ddim_alpha_cumprods = torch.from_numpy(self.ddim_alpha_cumprods)
        self.ddim_alpha_cumprods_prev = torch.from_numpy(self.ddim_alpha_cumprods_prev)

    def to(self, device):
        self.ddim_timesteps = self.ddim_timesteps.to(device)
        self.ddim_timesteps_prev = self.ddim_timesteps_prev.to(device)

        self.ddim_alpha_cumprods = self.ddim_alpha_cumprods.to(device)
        self.ddim_alpha_cumprods_prev = self.ddim_alpha_cumprods_prev.to(device)
        return self

    def ddim_step(self, pred_x0, pred_noise, timestep_index):
        alpha_cumprod_prev = extract_into_tensor(
            self.ddim_alpha_cumprods_prev, timestep_index, pred_x0.shape
        )
        dir_xt = (1.0 - alpha_cumprod_prev).sqrt() * pred_noise
        x_prev = alpha_cumprod_prev.sqrt() * pred_x0 + dir_xt
        return x_prev


def import_model_class_from_model_name_or_path(
    pretrained_model_name_or_path: str, revision: str, subfolder: str = "text_encoder"
):
    text_encoder_config = PretrainedConfig.from_pretrained(
        pretrained_model_name_or_path,
        subfolder=subfolder,
        revision=revision,
        use_auth_token=True,
    )
    model_class = text_encoder_config.architectures[0]

    if model_class == "CLIPTextModel":
        from transformers import CLIPTextModel

        return CLIPTextModel
    elif model_class == "CLIPTextModelWithProjection":
        from transformers import CLIPTextModelWithProjection

        return CLIPTextModelWithProjection
    else:
        raise ValueError(f"{model_class} is not supported.")


def parse_args():
    parser = argparse.ArgumentParser(description="Simple example of a training script.")
    parser.add_argument(
        "--pretrained_teacher_model",
        type=str,
        default=None,
        required=True,
        help="Path to pretrained LDM teacher model or model identifier from huggingface.co/models.",
    )
    parser.add_argument(
        "--pretrained_vae_model_name_or_path",
        type=str,
        default=None,
        help="Path to pretrained VAE model with better numerical stability. More details: https://github.com/huggingface/diffusers/pull/4038.",
    )
    parser.add_argument(
        "--teacher_revision",
        type=str,
        default=None,
        required=False,
        help="Revision of pretrained LDM teacher model identifier from huggingface.co/models.",
    )
    parser.add_argument(
        "--revision",
        type=str,
        default=None,
        required=False,
        help="Revision of pretrained LDM model identifier from huggingface.co/models.",
    )
    parser.add_argument(
        "--output_dir",
        type=str,
        default="lcm-xl-distilled",
        help="The output directory where the model predictions and checkpoints will be written.",
    )
    parser.add_argument(
        "--cache_dir",
        type=str,
        default=None,
        help="The directory where the downloaded models and datasets will be stored.",
    )
    parser.add_argument(
        "--seed", type=int, default=None, help="A seed for reproducible training."
    )
    parser.add_argument(
        "--logging_dir",
        type=str,
        default="logs",
        help=(
            "[TensorBoard](https://www.tensorflow.org/tensorboard) log directory. Will default to"
            " *output_dir/runs/**CURRENT_DATETIME_HOSTNAME***."
        ),
    )
    parser.add_argument(
        "--report_to",
        type=str,
        default="tensorboard",
        help=(
            'The integration to report the results and logs to. Supported platforms are `"tensorboard"`'
            ' (default), `"wandb"` and `"comet_ml"`. Use `"all"` to report to all integrations.'
        ),
    )
    parser.add_argument(
        "--checkpointing_steps",
        type=int,
        default=500,
        help=(
            "Save a checkpoint of the training state every X updates. These checkpoints are only suitable for resuming"
            " training using `--resume_from_checkpoint`."
        ),
    )
    parser.add_argument(
        "--checkpoints_total_limit",
        type=int,
        default=None,
        help=("Max number of checkpoints to store."),
    )
    parser.add_argument(
        "--resume_from_checkpoint",
        type=str,
        default=None,
        help=(
            "Whether training should be resumed from a previous checkpoint. Use a path saved by"
            ' `--checkpointing_steps`, or `"latest"` to automatically select the last available checkpoint.'
        ),
    )
    parser.add_argument(
        "--train_shards_path_or_url",
        type=str,
        default=None,
        help=(
            "The name of the Dataset (from the HuggingFace hub) to train on (could be your own, possibly private,"
            " dataset). It can also be a path pointing to a local copy of a dataset in your filesystem,"
            " or to a folder containing files that 🤗 Datasets can understand."
        ),
    )
    parser.add_argument(
        "--resolution",
        type=int,
        default=1024,
        help=(
            "The resolution for input images, all the images in the train/validation dataset will be resized to this"
            " resolution"
        ),
    )
    parser.add_argument(
        "--use_fix_crop_and_size",
        action="store_true",
        help="Whether or not to use the fixed crop and size for the teacher model.",
        default=False,
    )
    parser.add_argument(
        "--center_crop",
        default=False,
        action="store_true",
        help=(
            "Whether to center crop the input images to the resolution. If not set, the images will be randomly"
            " cropped. The images will be resized to the resolution first before cropping."
        ),
    )
    parser.add_argument(
        "--random_flip",
        action="store_true",
        help="whether to randomly flip images horizontally",
    )
    parser.add_argument(
        "--dataloader_num_workers",
        type=int,
        default=0,
        help=(
            "Number of subprocesses to use for data loading. 0 means that the data will be loaded in the main process."
        ),
    )
    parser.add_argument(
        "--train_batch_size",
        type=int,
        default=16,
        help="Batch size (per device) for the training dataloader.",
    )
    parser.add_argument("--num_train_epochs", type=int, default=100)
    parser.add_argument(
        "--max_train_steps",
        type=int,
        default=None,
        help="Total number of training steps to perform.  If provided, overrides num_train_epochs.",
    )
    parser.add_argument(
        "--max_train_samples",
        type=int,
        default=None,
        help=(
            "For debugging purposes or quicker training, truncate the number of training examples to this "
            "value if set."
        ),
    )
    parser.add_argument(
        "--learning_rate",
        type=float,
        default=1e-4,
        help="Initial learning rate (after the potential warmup period) to use.",
    )
    parser.add_argument(
        "--scale_lr",
        action="store_true",
        default=False,
        help="Scale the learning rate by the number of GPUs, gradient accumulation steps, and batch size.",
    )
    parser.add_argument(
        "--lr_scheduler",
        type=str,
        default="constant",
        help=(
            'The scheduler type to use. Choose between ["linear", "cosine", "cosine_with_restarts", "polynomial",'
            ' "constant", "constant_with_warmup"]'
        ),
    )
    parser.add_argument(
        "--lr_warmup_steps",
        type=int,
        default=500,
        help="Number of steps for the warmup in the lr scheduler.",
    )
    parser.add_argument(
        "--gradient_accumulation_steps",
        type=int,
        default=1,
        help="Number of updates steps to accumulate before performing a backward/update pass.",
    )
    parser.add_argument(
        "--use_8bit_adam",
        action="store_true",
        help="Whether or not to use 8-bit Adam from bitsandbytes.",
    )
    parser.add_argument(
        "--adam_beta1",
        type=float,
        default=0.9,
        help="The beta1 parameter for the Adam optimizer.",
    )
    parser.add_argument(
        "--adam_beta2",
        type=float,
        default=0.999,
        help="The beta2 parameter for the Adam optimizer.",
    )
    parser.add_argument(
        "--adam_weight_decay", type=float, default=1e-2, help="Weight decay to use."
    )
    parser.add_argument(
        "--adam_epsilon",
        type=float,
        default=1e-08,
        help="Epsilon value for the Adam optimizer",
    )
    parser.add_argument(
        "--max_grad_norm", default=1.0, type=float, help="Max gradient norm."
    )
    parser.add_argument(
        "--proportion_empty_prompts",
        type=float,
        default=0,
        help="Proportion of image prompts to be replaced with empty strings. Defaults to 0 (no prompt replacement).",
    )
    parser.add_argument(
        "--w_min",
        type=float,
        default=3.0,
        required=False,
        help=(
            "The minimum guidance scale value for guidance scale sampling. Note that we are using the Imagen CFG"
            " formulation rather than the LCM formulation, which means all guidance scales have 1 added to them as"
            " compared to the original paper."
        ),
    )
    parser.add_argument(
        "--w_max",
        type=float,
        default=15.0,
        required=False,
        help=(
            "The maximum guidance scale value for guidance scale sampling. Note that we are using the Imagen CFG"
            " formulation rather than the LCM formulation, which means all guidance scales have 1 added to them as"
            " compared to the original paper."
        ),
    )
    parser.add_argument(
        "--num_ddim_timesteps",
        type=int,
        default=40,
        help="The number of timesteps to use for DDIM sampling.",
    )
    parser.add_argument(
        "--loss_type",
        type=str,
        default="l2",
        choices=["l2", "huber"],
        help="The type of loss to use for the LCD loss.",
    )
    parser.add_argument(
        "--huber_c",
        type=float,
        default=0.001,
        help="The huber loss parameter. Only used if `--loss_type=huber`.",
    )
    parser.add_argument(
        "--lora_rank",
        type=int,
        default=64,
        help="The rank of the LoRA projection matrix.",
    )
    parser.add_argument(
        "--mixed_precision",
        type=str,
        default=None,
        choices=["no", "fp16", "bf16"],
        help=(
            "Whether to use mixed precision. Choose between fp16 and bf16 (bfloat16). Bf16 requires PyTorch >="
            " 1.10.and an Nvidia Ampere GPU.  Default to the value of accelerate config of the current system or the"
            " flag passed with the `accelerate.launch` command. Use this argument to override the accelerate config."
        ),
    )
    parser.add_argument(
        "--allow_tf32",
        action="store_true",
        help=(
            "Whether or not to allow TF32 on Ampere GPUs. Can be used to speed up training. For more information, see"
            " https://pytorch.org/docs/stable/notes/cuda.html#tensorfloat-32-tf32-on-ampere-devices"
        ),
    )
    parser.add_argument(
        "--cast_teacher_unet",
        action="store_true",
        help="Whether to cast the teacher U-Net to the precision specified by `--mixed_precision`.",
    )
    parser.add_argument(
        "--enable_xformers_memory_efficient_attention",
        action="store_true",
        help="Whether or not to use xformers.",
    )
    parser.add_argument(
        "--gradient_checkpointing",
        action="store_true",
        help="Whether or not to use gradient checkpointing to save memory at the expense of slower backward pass.",
    )
    parser.add_argument(
        "--local_rank",
        type=int,
        default=-1,
        help="For distributed training: local_rank",
    )
    parser.add_argument(
        "--validation_steps",
        type=int,
        default=200,
        help="Run validation every X steps.",
    )
    parser.add_argument(
        "--push_to_hub",
        action="store_true",
        help="Whether or not to push the model to the Hub.",
    )
    parser.add_argument(
        "--hub_token",
        type=str,
        default=None,
        help="The token to use to push to the Model Hub.",
    )
    parser.add_argument(
        "--hub_model_id",
        type=str,
        default=None,
        help="The name of the repository to keep in sync with the local `output_dir`.",
    )
    parser.add_argument(
        "--tracker_project_name",
        type=str,
        default="text2image-fine-tune",
        help=(
            "The `project_name` argument passed to Accelerator.init_trackers for"
            " more information see https://huggingface.co/docs/accelerate/v0.17.0/en/package_reference/accelerator#accelerate.Accelerator"
        ),
    )

    args = parser.parse_args()
    env_local_rank = int(os.environ.get("LOCAL_RANK", -1))
    if env_local_rank != -1 and env_local_rank != args.local_rank:
        args.local_rank = env_local_rank

    if args.proportion_empty_prompts < 0 or args.proportion_empty_prompts > 1:
        raise ValueError("`--proportion_empty_prompts` must be in the range [0, 1].")

    return args


def encode_prompt(
    prompt_batch, text_encoders, tokenizers, proportion_empty_prompts, is_train=True
):
    prompt_embeds_list = []

    captions = []
    for caption in prompt_batch:
        if random.random() < proportion_empty_prompts:
            captions.append("")
        elif isinstance(caption, str):
            captions.append(caption)
        elif isinstance(caption, (list, np.ndarray)):
            captions.append(random.choice(caption) if is_train else caption[0])

    with torch.no_grad():
        for tokenizer, text_encoder in zip(tokenizers, text_encoders):
            text_inputs = tokenizer(
                captions,
                padding="max_length",
                max_length=tokenizer.model_max_length,
                truncation=True,
                return_tensors="pt",
            )
            text_input_ids = text_inputs.input_ids
            prompt_embeds = text_encoder(
                text_input_ids.to(text_encoder.device),
                output_hidden_states=True,
            )

            pooled_prompt_embeds = prompt_embeds[0]
            prompt_embeds = prompt_embeds.hidden_states[-2]
            bs_embed, seq_len, _ = prompt_embeds.shape
            prompt_embeds = prompt_embeds.view(bs_embed, seq_len, -1)
            prompt_embeds_list.append(prompt_embeds)

    prompt_embeds = torch.concat(prompt_embeds_list, dim=-1)
    pooled_prompt_embeds = pooled_prompt_embeds.view(bs_embed, -1)
    return prompt_embeds, pooled_prompt_embeds


def main(args):
    logging_dir = Path(args.output_dir, args.logging_dir)

    accelerator_project_config = ProjectConfiguration(
        project_dir=args.output_dir, logging_dir=logging_dir
    )

    accelerator = Accelerator(
        gradient_accumulation_steps=args.gradient_accumulation_steps,
        mixed_precision=args.mixed_precision,
        log_with=args.report_to,
        project_config=accelerator_project_config,
    )

    logging.basicConfig(
        format="%(asctime)s - %(levelname)s - %(name)s - %(message)s",
        datefmt="%m/%d/%Y %H:%M:%S",
        level=logging.INFO,
    )
    logger.info(accelerator.state, main_process_only=False)
    if accelerator.is_local_main_process:
        transformers.utils.logging.set_verbosity_warning()
        diffusers.utils.logging.set_verbosity_info()
    else:
        transformers.utils.logging.set_verbosity_error()
        diffusers.utils.logging.set_verbosity_error()

    if args.seed is not None:
        set_seed(args.seed + accelerator.process_index)

    if accelerator.is_main_process:
        if args.output_dir is not None:
            os.makedirs(args.output_dir, exist_ok=True)

        if args.push_to_hub:
            create_repo(
                repo_id=args.hub_model_id or Path(args.output_dir).name,
                exist_ok=True,
                token=args.hub_token,
                private=True,
            ).repo_id

    noise_scheduler = DDPMScheduler.from_pretrained(
        args.pretrained_teacher_model,
        subfolder="scheduler",
        revision=args.teacher_revision,
    )
    alpha_schedule = torch.sqrt(noise_scheduler.alphas_cumprod)
    sigma_schedule = torch.sqrt(1 - noise_scheduler.alphas_cumprod)
    solver = DDIMSolver(
        noise_scheduler.alphas_cumprod.numpy(),
        timesteps=noise_scheduler.config.num_train_timesteps,
        ddim_timesteps=args.num_ddim_timesteps,
    )

    tokenizer_one = AutoTokenizer.from_pretrained(
        args.pretrained_teacher_model,
        subfolder="tokenizer",
        revision=args.teacher_revision,
        use_fast=False,
    )
    tokenizer_two = AutoTokenizer.from_pretrained(
        args.pretrained_teacher_model,
        subfolder="tokenizer_2",
        revision=args.teacher_revision,
        use_fast=False,
    )

    text_encoder_cls_one = import_model_class_from_model_name_or_path(
        args.pretrained_teacher_model, args.teacher_revision
    )
    text_encoder_cls_two = import_model_class_from_model_name_or_path(
        args.pretrained_teacher_model, args.teacher_revision, subfolder="text_encoder_2"
    )

    text_encoder_one = text_encoder_cls_one.from_pretrained(
        args.pretrained_teacher_model,
        subfolder="text_encoder",
        revision=args.teacher_revision,
    )
    text_encoder_two = text_encoder_cls_two.from_pretrained(
        args.pretrained_teacher_model,
        subfolder="text_encoder_2",
        revision=args.teacher_revision,
    )

    vae_path = (
        args.pretrained_teacher_model
        if args.pretrained_vae_model_name_or_path is None
        else args.pretrained_vae_model_name_or_path
    )
    vae = AutoencoderKL.from_pretrained(
        vae_path,
        subfolder="vae" if args.pretrained_vae_model_name_or_path is None else None,
        revision=args.teacher_revision,
    )

    teacher_unet = UNet2DConditionModel.from_pretrained(
        args.pretrained_teacher_model, subfolder="unet", revision=args.teacher_revision
    )
    target_unet = UNet2DConditionModel.from_pretrained(
        args.pretrained_teacher_model,
        subfolder="unet",
        revision=args.teacher_revision,
    )
    unet = UNet2DConditionModel.from_pretrained(
        args.pretrained_teacher_model, subfolder="unet", revision=args.teacher_revision
    )
    unet.train()
    vae.requires_grad_(False)
    text_encoder_one.requires_grad_(False)
    text_encoder_two.requires_grad_(False)
    target_unet.requires_grad_(False)
    teacher_unet.requires_grad_(False)
    unet.requires_grad_(True)
    low_precision_error_string = (
        " Please make sure to always have all model weights in full float32 precision when starting training - even if"
        " doing mixed precision training, copy of the weights should still be float32."
    )

    if accelerator.unwrap_model(unet).dtype != torch.float32:
        raise ValueError(
            f"Controlnet loaded as datatype {accelerator.unwrap_model(unet).dtype}. {low_precision_error_string}"
        )

    weight_dtype = torch.float32
    if accelerator.mixed_precision == "fp16":
        weight_dtype = torch.float16
    elif accelerator.mixed_precision == "bf16":
        weight_dtype = torch.bfloat16

    vae.to(accelerator.device)
    if args.pretrained_vae_model_name_or_path is not None:
        vae.to(dtype=weight_dtype)
    text_encoder_one.to(accelerator.device, dtype=weight_dtype)
    text_encoder_two.to(accelerator.device, dtype=weight_dtype)

    teacher_unet.to(accelerator.device)
    target_unet.to(accelerator.device)
    if args.cast_teacher_unet:
        teacher_unet.to(dtype=weight_dtype)

    alpha_schedule = alpha_schedule.to(accelerator.device)
    sigma_schedule = sigma_schedule.to(accelerator.device)
    solver = solver.to(accelerator.device)

    if version.parse(accelerate.__version__) >= version.parse("0.16.0"):

        def save_model_hook(models, weights, output_dir):
            if accelerator.is_main_process:
                unet_ = accelerator.unwrap_model(unet)
                torch.save(unet_.state_dict(), os.path.join(output_dir, "unet.ckpt"))
                torch.save(
                    target_unet.state_dict(), os.path.join(output_dir, "ema_unet.ckpt")
                )

                for _, model in enumerate(models):
                    weights.pop()

        def load_model_hook(models, input_dir):
            unet_ = accelerator.unwrap_model(unet)
            unet_.load_state_dict(
                torch.load(os.path.join(input_dir, "unet.ckpt"), map_location="cpu")
            )
            target_unet.load_state_dict(
                torch.load(os.path.join(input_dir, "ema_unet.ckpt"), map_location="cpu")
            )
            for _ in range(len(models)):
                models.pop()

        accelerator.register_save_state_pre_hook(save_model_hook)
        accelerator.register_load_state_pre_hook(load_model_hook)

    if args.enable_xformers_memory_efficient_attention:
        if is_xformers_available():
            import xformers

            xformers_version = version.parse(xformers.__version__)
            if xformers_version == version.parse("0.0.16"):
                logger.warn(
                    "xFormers 0.0.16 cannot be used for training in some GPUs. If you observe problems during training, please update xFormers to at least 0.0.17. See https://huggingface.co/docs/diffusers/main/en/optimization/xformers for more details."
                )
            unet.enable_xformers_memory_efficient_attention()
            teacher_unet.enable_xformers_memory_efficient_attention()
            target_unet.enable_xformers_memory_efficient_attention()
        else:
            raise ValueError(
                "xformers is not available. Make sure it is installed correctly"
            )

    if args.allow_tf32:
        torch.backends.cuda.matmul.allow_tf32 = True

    if args.gradient_checkpointing:
        unet.enable_gradient_checkpointing()

    if args.use_8bit_adam:
        try:
            import bitsandbytes as bnb
        except ImportError:
            raise ImportError(
                "To use 8-bit Adam, please install the bitsandbytes library: `pip install bitsandbytes`."
            )

        optimizer_class = bnb.optim.AdamW8bit
    else:
        optimizer_class = torch.optim.AdamW

    optimizer = optimizer_class(
        unet.parameters(),
        lr=args.learning_rate,
        betas=(args.adam_beta1, args.adam_beta2),
        weight_decay=args.adam_weight_decay,
        eps=args.adam_epsilon,
    )

    def compute_embeddings(
        prompt_batch,
        original_sizes,
        crop_coords,
        proportion_empty_prompts,
        text_encoders,
        tokenizers,
        is_train=True,
    ):
        target_size = (args.resolution, args.resolution)
        original_sizes = list(map(list, zip(*original_sizes)))
        crops_coords_top_left = list(map(list, zip(*crop_coords)))

        original_sizes = torch.tensor(original_sizes, dtype=torch.long)
        crops_coords_top_left = torch.tensor(crops_coords_top_left, dtype=torch.long)

        prompt_embeds, pooled_prompt_embeds = encode_prompt(
            prompt_batch, text_encoders, tokenizers, proportion_empty_prompts, is_train
        )
        add_text_embeds = pooled_prompt_embeds

        add_time_ids = list(target_size)
        add_time_ids = torch.tensor([add_time_ids])
        add_time_ids = add_time_ids.repeat(len(prompt_batch), 1)
        add_time_ids = torch.cat(
            [original_sizes, crops_coords_top_left, add_time_ids], dim=-1
        )
        add_time_ids = add_time_ids.to(accelerator.device, dtype=prompt_embeds.dtype)

        prompt_embeds = prompt_embeds.to(accelerator.device)
        add_text_embeds = add_text_embeds.to(accelerator.device)
        unet_added_cond_kwargs = {
            "text_embeds": add_text_embeds,
            "time_ids": add_time_ids,
        }

        return {"prompt_embeds": prompt_embeds, **unet_added_cond_kwargs}

    train_dataset = CustomImageDataset(
        "/mnt/wangfuyun/laion50w/images", args.resolution
    )
    train_dataloader = DataLoader(
        train_dataset,
        shuffle=True,
        batch_size=args.train_batch_size,
        num_workers=args.dataloader_num_workers,
    )

    text_encoders = [text_encoder_one, text_encoder_two]
    tokenizers = [tokenizer_one, tokenizer_two]

    compute_embeddings_fn = functools.partial(
        compute_embeddings,
        proportion_empty_prompts=0.0,
        text_encoders=text_encoders,
        tokenizers=tokenizers,
    )

    overrode_max_train_steps = False
    num_update_steps_per_epoch = math.ceil(
        len(train_dataloader) / args.gradient_accumulation_steps
    )
    if args.max_train_steps is None:
        args.max_train_steps = args.num_train_epochs * num_update_steps_per_epoch
        overrode_max_train_steps = True

    if args.scale_lr:
        args.learning_rate = (
            args.learning_rate
            * args.gradient_accumulation_steps
            * args.train_batch_size
            * accelerator.num_processes
        )

    lr_scheduler = get_scheduler(
        args.lr_scheduler,
        optimizer=optimizer,
        num_warmup_steps=args.lr_warmup_steps,
        num_training_steps=args.max_train_steps,
    )

    unet, optimizer, lr_scheduler, train_dataloader = accelerator.prepare(
        unet, optimizer, lr_scheduler, train_dataloader
    )

    num_update_steps_per_epoch = math.ceil(
        len(train_dataloader) / args.gradient_accumulation_steps
    )
    if overrode_max_train_steps:
        args.max_train_steps = args.num_train_epochs * num_update_steps_per_epoch
    args.num_train_epochs = math.ceil(args.max_train_steps / num_update_steps_per_epoch)

    if accelerator.is_main_process:
        tracker_config = dict(vars(args))
        accelerator.init_trackers(args.tracker_project_name, config=tracker_config)

    uncond_prompt_embeds = torch.zeros(args.train_batch_size, 77, 2048).to(
        accelerator.device
    )
    uncond_pooled_prompt_embeds = torch.zeros(args.train_batch_size, 1280).to(
        accelerator.device
    )

    # 16. Train!
    total_batch_size = (
        args.train_batch_size
        * accelerator.num_processes
        * args.gradient_accumulation_steps
    )

    logger.info("***** Running training *****")
    logger.info(f"  Num batches each epoch = {len(train_dataloader)}")
    logger.info(f"  Num Epochs = {args.num_train_epochs}")
    logger.info(f"  Instantaneous batch size per device = {args.train_batch_size}")
    logger.info(
        f"  Total train batch size (w. parallel, distributed & accumulation) = {total_batch_size}"
    )
    logger.info(f"  Gradient Accumulation steps = {args.gradient_accumulation_steps}")
    logger.info(f"  Total optimization steps = {args.max_train_steps}")
    global_step = 0
    first_epoch = 0

    if args.resume_from_checkpoint:
        if args.resume_from_checkpoint != "latest":
            path = os.path.basename(args.resume_from_checkpoint)
        else:
            dirs = os.listdir(args.output_dir)
            dirs = [d for d in dirs if d.startswith("checkpoint")]
            dirs = sorted(dirs, key=lambda x: int(x.split("-")[1]))
            path = dirs[-1] if len(dirs) > 0 else None

        if path is None:
            accelerator.print(
                f"Checkpoint '{args.resume_from_checkpoint}' does not exist. Starting a new training run."
            )
            args.resume_from_checkpoint = None
            initial_global_step = 0
        else:
            accelerator.print(f"Resuming from checkpoint {path}")
            accelerator.load_state(os.path.join(args.output_dir, path))
            global_step = int(path.split("-")[1])

            initial_global_step = global_step
            first_epoch = global_step // num_update_steps_per_epoch
    else:
        initial_global_step = 0

    progress_bar = tqdm(
        range(0, args.max_train_steps),
        initial=initial_global_step,
        desc="Steps",
        disable=not accelerator.is_local_main_process,
    )

    print("hello, start training now!")
    for epoch in range(first_epoch, args.num_train_epochs):
        for step, batch in enumerate(train_dataloader):
            with accelerator.accumulate(unet):
                image, text, orig_size, crop_coords = batch

                image = image.to(accelerator.device, non_blocking=True)
                encoded_text = compute_embeddings_fn(text, orig_size, crop_coords)

                if args.pretrained_vae_model_name_or_path is not None:
                    pixel_values = image.to(dtype=weight_dtype)
                    if vae.dtype != weight_dtype:
                        vae.to(dtype=weight_dtype)
                else:
                    pixel_values = image

                with torch.no_grad():
                    latents = []
                    for i in range(0, pixel_values.shape[0], 1):
                        with torch.autocast("cuda"):
                            latents.append(
                                vae.encode(pixel_values[i : i + 1]).latent_dist.sample()
                            )
                    latents = torch.cat(latents, dim=0)

                latents = latents * vae.config.scaling_factor
                if args.pretrained_vae_model_name_or_path is None:
                    latents = latents.to(weight_dtype)

                noise = torch.randn_like(latents)
                bsz = latents.shape[0]

                topk = (
                    noise_scheduler.config.num_train_timesteps
                    // args.num_ddim_timesteps
                )
                index = torch.randint(
                    0, args.num_ddim_timesteps, (bsz,), device=latents.device
                ).long()
                index = (index // 10 * 10 + 10) - 1

                start_timesteps = solver.ddim_timesteps[index]

                noisy_model_input = noise_scheduler.add_noise(
                    latents, noise, start_timesteps
                )

                prompt_embeds = encoded_text.pop("prompt_embeds")

                tmp_timesteps = start_timesteps
                tmp_model_input = noisy_model_input
                with torch.no_grad():
                    with torch.autocast("cuda"):
                        w = 6.5
                        mask_solver = torch.ones_like(index).float()
                        for _ in range(10):
                            cond_teacher_output = teacher_unet(
                                tmp_model_input.float(),
                                tmp_timesteps,
                                encoder_hidden_states=prompt_embeds.to(weight_dtype),
                                added_cond_kwargs={
                                    k: v.to(weight_dtype)
                                    for k, v in encoded_text.items()
                                },
                            ).sample
                            cond_pred_x0 = predicted_origin(
                                cond_teacher_output,
                                tmp_timesteps,
                                tmp_model_input,
                                noise_scheduler.config.prediction_type,
                                alpha_schedule,
                                sigma_schedule,
                            )

                            if args.not_apply_cfg_solver:
                                uncond_teacher_output = cond_teacher_output
                                uncond_pred_x0 = cond_pred_x0
                            else:
                                uncond_added_conditions = copy.deepcopy(encoded_text)
                                uncond_added_conditions["text_embeds"] = (
                                    uncond_pooled_prompt_embeds
                                )
                                uncond_teacher_output = teacher_unet(
                                    tmp_model_input.float(),
                                    tmp_timesteps,
                                    encoder_hidden_states=uncond_prompt_embeds.to(
                                        weight_dtype
                                    ),
                                    added_cond_kwargs={
                                        k: v.to(weight_dtype)
                                        for k, v in uncond_added_conditions.items()
                                    },
                                ).sample
                                uncond_pred_x0 = predicted_origin(
                                    uncond_teacher_output,
                                    tmp_timesteps,
                                    tmp_model_input,
                                    noise_scheduler.config.prediction_type,
                                    alpha_schedule,
                                    sigma_schedule,
                                )

                            pred_x0 = cond_pred_x0 + w * (cond_pred_x0 - uncond_pred_x0)
                            pred_noise = cond_teacher_output + w * (
                                cond_teacher_output - uncond_teacher_output
                            )
                            tmp_model_input = solver.ddim_step(
                                pred_x0, pred_noise, index
                            ).float() * append_dims(
                                mask_solver, latents.ndim
                            ) + tmp_model_input * append_dims(
                                (1 - mask_solver), latents.ndim
                            )
                            index = index - 1
                            mask_solver = torch.where(
                                index < 0, 0.0, mask_solver
                            ).float()
                            index = torch.where(
                                index < 0, torch.zeros_like(index), index
                            )
                            tmp_timesteps = tmp_timesteps - topk
                            tmp_timesteps = torch.where(
                                tmp_timesteps < 0,
                                torch.zeros_like(tmp_timesteps),
                                tmp_timesteps,
                            )
                            if (mask_solver == 0.0).all():
                                break
                        x_prev = tmp_model_input
                        timesteps = tmp_timesteps

                start_sigmas = extract_into_tensor(
                    sigma_schedule, start_timesteps, latents.shape
                )
                start_alphas = extract_into_tensor(
                    alpha_schedule, start_timesteps, latents.shape
                )
                start_nsr = start_sigmas / start_alphas
                sigmas = extract_into_tensor(sigma_schedule, timesteps, latents.shape)
                alphas = extract_into_tensor(alpha_schedule, timesteps, latents.shape)
                nsr = sigmas / alphas

                b = start_timesteps.shape[0]

                sampled_timesteps = torch.empty(
                    b, dtype=torch.long, device=accelerator.device
                )

                for i in range(b):
                    sampled_timesteps[i] = timesteps[i] + torch.randint(
                        1,
                        (start_timesteps[i] - timesteps[i] + 1).item(),
                        (1,),
                        dtype=torch.long,
                        device=accelerator.device,
                    )
                accelerator.print(timesteps)
                accelerator.print(start_timesteps)
                accelerator.print(sampled_timesteps)
                sampled_sigmas = extract_into_tensor(
                    sigma_schedule, sampled_timesteps, latents.shape
                )
                sampled_alphas = extract_into_tensor(
                    alpha_schedule, sampled_timesteps, latents.shape
                )
                sampled_nsr = sampled_sigmas / sampled_alphas

                accelerator.print("x_prev", x_prev.mean())
                accelerator.print("start_nsr", start_nsr.mean())
                accelerator.print("nsr", nsr.mean())
                noise = ((x_prev / alphas) - (noisy_model_input / start_alphas)) / (
                    nsr - start_nsr
                )
                target = x_prev

                sampled_x = (
                    (noisy_model_input / start_alphas)
                    + (sampled_nsr - start_nsr) * noise
                ) * sampled_alphas

                model_pred = unet(
                    sampled_x.float(),
                    sampled_timesteps,
                    encoder_hidden_states=prompt_embeds.float(),
                    added_cond_kwargs={
                        k: v.to(weight_dtype) for k, v in encoded_text.items()
                    },
                ).sample

                model_pred = (
                    sampled_x / sampled_alphas + (nsr - sampled_nsr) * model_pred
                ) * alphas

                if args.loss_type == "l2":
                    loss = F.mse_loss(
                        model_pred.float(), target.float(), reduction="mean"
                    )
                elif args.loss_type == "huber":
                    loss = torch.mean(
                        torch.sqrt(
                            (model_pred.float() - target.float()) ** 2 + args.huber_c**2
                        )
                        - args.huber_c
                    )

                accelerator.backward(loss)
                if accelerator.sync_gradients:
                    accelerator.clip_grad_norm_(unet.parameters(), args.max_grad_norm)
                optimizer.step()
                lr_scheduler.step()
                optimizer.zero_grad(set_to_none=True)

            if accelerator.sync_gradients:
                progress_bar.update(1)
                global_step += 1
                update_ema(target_unet.parameters(), unet.parameters(), 0.9999)

                if accelerator.is_main_process:
                    if global_step % args.checkpointing_steps == 0:
                        if args.checkpoints_total_limit is not None:
                            checkpoints = os.listdir(args.output_dir)
                            checkpoints = [
                                d for d in checkpoints if d.startswith("checkpoint")
                            ]
                            checkpoints = sorted(
                                checkpoints, key=lambda x: int(x.split("-")[1])
                            )

                            if len(checkpoints) >= args.checkpoints_total_limit:
                                num_to_remove = (
                                    len(checkpoints) - args.checkpoints_total_limit + 1
                                )
                                removing_checkpoints = checkpoints[0:num_to_remove]

                                logger.info(
                                    f"{len(checkpoints)} checkpoints already exist, removing {len(removing_checkpoints)} checkpoints"
                                )
                                logger.info(
                                    f"removing checkpoints: {', '.join(removing_checkpoints)}"
                                )

                                for removing_checkpoint in removing_checkpoints:
                                    removing_checkpoint = os.path.join(
                                        args.output_dir, removing_checkpoint
                                    )
                                    shutil.rmtree(removing_checkpoint)

                        save_path = os.path.join(
                            args.output_dir, f"checkpoint-{global_step}"
                        )
                        accelerator.save_state(save_path)
                        logger.info(f"Saved state to {save_path}")

                    if global_step % args.validation_steps == 0:
                        with torch.no_grad():
                            log_validation(
                                vae,
                                unet,
                                args,
                                accelerator,
                                weight_dtype,
                                global_step,
                                4,
                                1.5,
                            )
                            log_validation(
                                vae,
                                unet,
                                args,
                                accelerator,
                                weight_dtype,
                                global_step,
                                24,
                                1.5,
                            )
            logs = {"loss": loss.detach().item(), "lr": lr_scheduler.get_last_lr()[0]}
            progress_bar.set_postfix(**logs)
            accelerator.log(logs, step=global_step)

            if global_step >= args.max_train_steps:
                break

    # Create the pipeline using using the trained modules and save it.
    accelerator.wait_for_everyone()
    accelerator.end_training()


if __name__ == "__main__":
    args = parse_args()
    main(args)