test_train_lcm_lora_bucket.py

#!/usr/bin/env python
# coding=utf-8
# Copyright 2023 The HuggingFace Inc. team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and

import argparse
import copy
import functools
import gc
import itertools
import json
import logging
import math
import time
import os
import random
import shutil
from pathlib import Path
from PIL import Image
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
import webdataset as wds
from functools import partial
from accelerate import Accelerator
from accelerate.logging import get_logger
from accelerate.utils import ProjectConfiguration, set_seed
from braceexpand import braceexpand
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 torchvision.transforms.functional import crop
from torch.utils.data import Dataset
from tqdm.auto import tqdm
from safetensors import safe_open
from safetensors.torch import save_file
from transformers import AutoTokenizer, PretrainedConfig
from webdataset.tariterators import (
    base_plus_ext,
    tar_file_expander,
    url_opener,
    valid_sample,
)

import diffusers
from diffusers import (
    AutoencoderKL,
    DDPMScheduler,
    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


MAX_SEQ_LENGTH = 77

if is_wandb_available():
    import wandb

# Will error if the minimal version of diffusers is not installed. Remove at your own risks.
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)

        # Set alpha parameter
        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


def save_lora_model(unet, output_dir, weight_dtype):
    lora_state_dict = get_module_kohya_state_dict(unet, "lora_unet", weight_dtype)
    output_path = os.path.join(output_dir, "unet_lora")
    StableDiffusionXLPipeline.save_lora_weights(output_path, lora_state_dict)

    # delete the "unet." from model state dict keys
    tensors = {}
    model_path = os.path.join(output_path, "pytorch_lora_weights.safetensors")
    with safe_open(model_path, framework="pt", device=0) as f:
        for k in f.keys():
            new_key = k.replace("unet.", "", 1)
            tensors[new_key] = f.get_tensor(k)
    save_file(tensors, model_path)


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

    unet = accelerator.unwrap_model(unet)
    pipeline = StableDiffusionXLPipeline.from_pretrained(
        args.pretrained_teacher_model,
        vae=vae,
        scheduler=LCMScheduler.from_pretrained(args.pretrained_teacher_model, subfolder="scheduler"),
        revision=args.revision,
        torch_dtype=weight_dtype,
    )
    pipeline = pipeline.to(accelerator.device)
    pipeline.set_progress_bar_config(disable=True)

    lora_state_dict = get_module_kohya_state_dict(unet, "lora_unet", weight_dtype)
    pipeline.load_lora_weights(lora_state_dict)
    pipeline.fuse_lora()

    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=4,
                num_images_per_prompt=4,
                generator=generator,
                guidance_scale=0.0,
            ).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({"validation": 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]


# From LCMScheduler.get_scalings_for_boundary_condition_discrete
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


# Compare LCMScheduler.step, Step 4
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":
        pred_x_0 = alphas[timesteps] * sample - sigmas[timesteps] * 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)))


class DDIMSolver:
    def __init__(self, alpha_cumprods, timesteps=1000, ddim_timesteps=50):
        # DDIM sampling parameters
        step_ratio = timesteps // ddim_timesteps

        self.ddim_timesteps = (np.arange(1, ddim_timesteps + 1) * step_ratio).round().astype(np.int64) - 1
        self.ddim_alpha_cumprods = alpha_cumprods[self.ddim_timesteps]
        self.ddim_alpha_cumprods_prev = np.asarray(
            [alpha_cumprods[0]] + alpha_cumprods[self.ddim_timesteps[:-1]].tolist()
        )
        # convert to torch tensors
        self.ddim_timesteps = torch.from_numpy(self.ddim_timesteps).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_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.")
    # ----------Model Checkpoint Loading Arguments----------
    parser.add_argument(
        "--pretrained_teacher_model",
        type=str,
        default="/home/model/SSD-1B",
        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.",
    )
    # ----------Training Arguments----------
    # ----General Training Arguments----
    parser.add_argument(
        "--output_dir",
        type=str,
        default="/home/fintune/lcm-xl-distilled_Songyun",
        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.")
    # ----Logging----
    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.'
        ),
    )
    # ----Checkpointing----
    parser.add_argument(
        "--checkpointing_steps",
        type=int,
        default=5000,
        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.'
        ),
    )
    # ----Image Processing----
    parser.add_argument(
        "--use_aspect_ratio_bucket",
        default=False,
        action="store_true",
        help="Use aspect ratio bucketing as image processing strategy, which may improve the quality of outputs."
    )
    parser.add_argument(
        "--train_data_dir",
        type=str,
        default=None,
        help=(
            "A folder containing the training data of images and caption."
        ),
    )
    parser.add_argument(
        "--pad_tokens", default=False, action="store_true", help="Whether to padding token."
    )
    parser.add_argument(
        "--ucg",
        type=float,
        default=0.0,
        help="Percentage chance of dropping out the text condition per batch. \
            Ranges from 0.0 to 1.0 where 1.0 means 100% text condition dropout."
    )
    parser.add_argument(
        "--debug_prompt",
        default=False,
        action="store_true",
        help="Print input prompt when training."
    )
    parser.add_argument(
        "--debug_arb",
        default=False,
        action="store_true",                          
        help="Enable debug logging on aspect ratio bucket."
    )
    parser.add_argument(
        "--arb_dim_limit",
        type=int,
        default=1024,
        help="Aspect ratio bucketing arguments: dim_limit."
    )
    parser.add_argument(
        "--arb_divisible",
        type=int,
        default=64,
        help="Aspect ratio bucketing arguments: divisbile."
    )
    parser.add_argument(
        "--arb_max_ar_error",
        type=int,
        default=4,
        help="Aspect ratio bucketing arguments: max_ar_error."
    )
    parser.add_argument(
        "--arb_max_size",
        type=int,
        nargs="+",
        default=(1024, 1024),
        help="Aspect ratio bucketing arguments: max_size. example: --arb_max_size 768 512"
    )
    parser.add_argument(
        "--arb_min_dim",
        type=int,
        default=256,
        help="Aspect ratio bucketing arguments: min_dim."
    )
    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(
        "--center_crop",
        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",
    )
    # ----Dataloader----
    parser.add_argument(
        "--dataloader_num_workers",
        type=int,
        default=1,
        help=(
            "Number of subprocesses to use for data loading. 0 means that the data will be loaded in the main process."
        ),
    )
    # ----Batch Size and Training Steps----
    parser.add_argument(
        "--train_batch_size", type=int, default=4, help="Batch size (per device) for the training dataloader."
    )
    parser.add_argument("--num_train_epochs", type=int, default=30)
    parser.add_argument(
        "--max_train_steps",
        type=int,
        default=None,
        help="Total number of training steps to perform.  If provided, overrides num_train_epochs.",
    )
    # ----Learning Rate----
    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=0, help="Number of steps for the warmup in the lr scheduler."
    )
    parser.add_argument(
        "--gradient_accumulation_steps",
        type=int,
        default=4,
        help="Number of updates steps to accumulate before performing a backward/update pass.",
    )
    # ----Optimizer (Adam)----
    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.")
    # ----Diffusion Training Arguments----
    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).",
    )
    # ----Latent Consistency Distillation (LCD) Specific Arguments----
    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=50,
        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.",
    )
    # ----Mixed Precision----
    parser.add_argument(
        "--mixed_precision",
        type=str,
        default="fp16",
        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`.",
    )
    # ----Training Optimizations----
    parser.add_argument(
        "--enable_xformers_memory_efficient_attention", action="store_true", default=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.",
    )
    # ----Distributed Training----
    parser.add_argument("--local_rank", type=int, default=-1, help="For distributed training: local_rank")
    # ----------Validation Arguments----------
    parser.add_argument(
        "--validation_steps",
        type=int,
        default=200,
        help="Run validation every X steps.",
    )
    # ----------Huggingface Hub Arguments-----------
    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`.",
    )
    # ----------Accelerate Arguments----------
    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

class AspectRatioBucket:
    '''
    Code from https://github.com/NovelAI/novelai-aspect-ratio-bucketing/blob/main/bucketmanager.py

    BucketManager impls NovelAI Aspect Ratio Bucketing, which may greatly improve the quality of outputs according to Novelai's blog (https://blog.novelai.net/novelai-improvements-on-stable-diffusion-e10d38db82ac)
    Requires a pickle with mapping of dataset IDs to resolutions called resolutions.pkl to use this.
    '''

    def __init__(self,
        id_size_map,
        train_data_dir,
        max_size=(768, 512),
        divisible=64,
        step_size=8,
        min_dim=256,
        base_res=(512, 512),
        bsz=1,
        world_size=1,
        global_rank=0,
        max_ar_error=4,
        seed=42,
        dim_limit=1024,
        debug=True,
    ):
        if global_rank == -1:
            global_rank = 0
            
        self.res_map = id_size_map
        self.max_size = max_size
        self.f = 8
        self.max_tokens = (max_size[0]/self.f) * (max_size[1]/self.f)
        self.div = divisible
        self.min_dim = min_dim
        self.dim_limit = dim_limit
        self.base_res = base_res
        self.bsz = bsz
        self.world_size = world_size
        self.global_rank = global_rank
        self.max_ar_error = max_ar_error
        self.prng = self.get_prng(seed)
        epoch_seed = self.prng.tomaxint() % (2**32-1)

        # separate prng for sharding use for increased thread resilience
        self.epoch_prng = self.get_prng(epoch_seed)
        self.epoch = None
        self.left_over = None
        self.batch_total = None
        self.batch_delivered = None

        self.debug = debug

        self.gen_buckets()
        self.assign_buckets()
        self.start_epoch()

    @staticmethod
    def get_prng(seed):
        return np.random.RandomState(seed)
    
    def __len__(self):
        return len(self.res_map) // self.bsz

    def gen_buckets(self):
        if self.debug:
            timer = time.perf_counter()
        resolutions = []
        aspects = []
        w = self.min_dim
        while (w/self.f) * (self.min_dim/self.f) <= self.max_tokens and w <= self.dim_limit:
            h = self.min_dim
            got_base = False
            while (w/self.f) * ((h+self.div)/self.f) <= self.max_tokens and (h+self.div) <= self.dim_limit:
                if w == self.base_res[0] and h == self.base_res[1]:
                    got_base = True
                h += self.div
            if (w != self.base_res[0] or h != self.base_res[1]) and got_base:
                resolutions.append(self.base_res)
                aspects.append(1)
            resolutions.append((w, h))
            aspects.append(float(w)/float(h))
            w += self.div
        h = self.min_dim
        while (h/self.f) * (self.min_dim/self.f) <= self.max_tokens and h <= self.dim_limit:
            w = self.min_dim
            got_base = False
            while (h/self.f) * ((w+self.div)/self.f) <= self.max_tokens and (w+self.div) <= self.dim_limit:
                if w == self.base_res[0] and h == self.base_res[1]:
                    got_base = True
                w += self.div
            resolutions.append((w, h))
            aspects.append(float(w)/float(h))
            h += self.div
        res_map = {}
        for i, res in enumerate(resolutions):
            res_map[res] = aspects[i]
        self.resolutions = sorted(
            res_map.keys(), key=lambda x: x[0] * 4096 - x[1])
        self.aspects = np.array(
            list(map(lambda x: res_map[x], self.resolutions)))
        self.resolutions = np.array(self.resolutions)
        if self.debug:
            timer = time.perf_counter() - timer
            print(f"resolutions:\n{self.resolutions}")
            print(f"aspects:\n{self.aspects}")
            print(f"gen_buckets: {timer:.5f}s")

    def assign_buckets(self):
        if self.debug:
            timer = time.perf_counter()
        self.buckets = {}
        self.aspect_errors = []
        skipped = 0
        skip_list = []
        for post_id in self.res_map.keys():
            w, h = self.res_map[post_id]
            aspect = float(w)/float(h)
            bucket_id = np.abs(self.aspects - aspect).argmin()
            if bucket_id not in self.buckets:
                self.buckets[bucket_id] = []
            error = abs(self.aspects[bucket_id] - aspect)
            if error < self.max_ar_error:
                self.buckets[bucket_id].append(post_id)
                if self.debug:
                    self.aspect_errors.append(error)
            else:
                skipped += 1
                skip_list.append(post_id)
        for post_id in skip_list:
            del self.res_map[post_id]
        if self.debug:
            timer = time.perf_counter() - timer
            self.aspect_errors = np.array(self.aspect_errors)
            try:
                print(f"skipped images: {skipped}")
                print(f"aspect error: mean {self.aspect_errors.mean()}, median {np.median(self.aspect_errors)}, max {self.aspect_errors.max()}")
                for bucket_id in reversed(sorted(self.buckets.keys(), key=lambda b: len(self.buckets[b]))):
                    print(
                        f"bucket {bucket_id}: {self.resolutions[bucket_id]}, aspect {self.aspects[bucket_id]:.5f}, entries {len(self.buckets[bucket_id])}")
                print(f"assign_buckets: {timer:.5f}s")
            except Exception as e:
                pass

    def start_epoch(self, world_size=None, global_rank=None):
        if self.debug:
            timer = time.perf_counter()
        if world_size is not None:
            self.world_size = world_size
        if global_rank is not None:
            self.global_rank = global_rank

        # select ids for this epoch/rank
        index = sorted(list(self.res_map.keys()))
        index_len = len(index)
        
        index = self.epoch_prng.permutation(index)
        index = index[:index_len - (index_len % (self.bsz * self.world_size))]
        # if self.debug:
            # print("perm", self.global_rank, index[0:16])
        
        index = index[self.global_rank::self.world_size]
        self.batch_total = len(index) // self.bsz
        assert (len(index) % self.bsz == 0)
        index = set(index)

        self.epoch = {}
        self.left_over = []
        self.batch_delivered = 0
        for bucket_id in sorted(self.buckets.keys()):
            if len(self.buckets[bucket_id]) > 0:
                self.epoch[bucket_id] = [post_id for post_id in self.buckets[bucket_id] if post_id in index]
                self.prng.shuffle(self.epoch[bucket_id])
                self.epoch[bucket_id] = list(self.epoch[bucket_id])
                overhang = len(self.epoch[bucket_id]) % self.bsz
                if overhang != 0:
                    self.left_over.extend(self.epoch[bucket_id][:overhang])
                    self.epoch[bucket_id] = self.epoch[bucket_id][overhang:]
                if len(self.epoch[bucket_id]) == 0:
                    del self.epoch[bucket_id]

        if self.debug:
            timer = time.perf_counter() - timer
            count = 0
            for bucket_id in self.epoch.keys():
                count += len(self.epoch[bucket_id])
            print(
                f"correct item count: {count == len(index)} ({count} of {len(index)})")
            print(f"start_epoch: {timer:.5f}s")

    def get_batch(self):
        if self.debug:
            timer = time.perf_counter()
        # check if no data left or no epoch initialized
        if self.epoch is None or self.left_over is None or (len(self.left_over) == 0 and not bool(self.epoch)) or self.batch_total == self.batch_delivered:
            self.start_epoch()

        found_batch = False
        batch_data = None
        resolution = self.base_res
        while not found_batch:
            bucket_ids = list(self.epoch.keys())
            if len(self.left_over) >= self.bsz:
                bucket_probs = [
                    len(self.left_over)] + [len(self.epoch[bucket_id]) for bucket_id in bucket_ids]
                bucket_ids = [-1] + bucket_ids
            else:
                bucket_probs = [len(self.epoch[bucket_id])
                                for bucket_id in bucket_ids]
            bucket_probs = np.array(bucket_probs, dtype=np.float32)
            bucket_lens = bucket_probs
            bucket_probs = bucket_probs / bucket_probs.sum()
            if bool(self.epoch):
                chosen_id = int(self.prng.choice(
                    bucket_ids, 1, p=bucket_probs)[0])
            else:
                chosen_id = -1

            if chosen_id == -1:
                # using leftover images that couldn't make it into a bucketed batch and returning them for use with basic square image
                self.prng.shuffle(self.left_over)
                batch_data = self.left_over[:self.bsz]
                self.left_over = self.left_over[self.bsz:]
                found_batch = True
            else:
                if len(self.epoch[chosen_id]) >= self.bsz:
                    # return bucket batch and resolution
                    batch_data = self.epoch[chosen_id][:self.bsz]
                    self.epoch[chosen_id] = self.epoch[chosen_id][self.bsz:]
                    resolution = tuple(self.resolutions[chosen_id])
                    found_batch = True
                    if len(self.epoch[chosen_id]) == 0:
                        del self.epoch[chosen_id]
                else:
                    # can't make a batch from this, not enough images. move them to leftovers and try again
                    self.left_over.extend(self.epoch[chosen_id])
                    del self.epoch[chosen_id]

            assert (found_batch or len(self.left_over)
                    >= self.bsz or bool(self.epoch))

        if self.debug:
            timer = time.perf_counter() - timer
            print(f"bucket probs: " +
                  ", ".join(map(lambda x: f"{x:.2f}", list(bucket_probs*100))))
            print(f"chosen id: {chosen_id}")
            print(f"batch data: {batch_data}")
            print(f"resolution: {resolution}")
            print(f"get_batch: {timer:.5f}s")

        self.batch_delivered += 1
        return (batch_data, resolution)

    def generator(self):
        if self.batch_delivered >= self.batch_total:
            self.start_epoch()
        while self.batch_delivered < self.batch_total:
            yield self.get_batch()


def init_arb_buckets(args, accelerator):
    arg_config = {
        "train_data_dir": args.train_data_dir,
        "bsz": args.train_batch_size,
        "seed": args.seed,
        "debug": args.debug_arb,
        "base_res": (args.resolution, args.resolution),
        "max_size": args.arb_max_size,
        "divisible": args.arb_divisible,
        "max_ar_error": args.arb_max_ar_error,
        "min_dim": args.arb_min_dim,
        "dim_limit": args.arb_dim_limit,
        "world_size": accelerator.num_processes,
        "global_rank": args.local_rank,
    }

    if args.debug_arb:
        print("BucketManager initialized using config:")
        print(json.dumps(arg_config, sort_keys=True, indent=4))  
    else:
        print(f"BucketManager initialized with base_res = {arg_config['base_res']}, max_size = {arg_config['max_size']}")
               
    def get_id_size_dict(entries):
        id_size_map = {}

        for entry in tqdm(entries, desc=f"Loading resolution from images", disable=args.local_rank not in [0, -1]):
            with Image.open(entry) as img:
                size = img.size
            id_size_map[entry] = size

        return id_size_map
    
    entries  = []
    inst_img_path = [x for x in Path(arg_config["train_data_dir"]).iterdir() if x.is_file() \
                    and x.suffix != ".txt" and x.suffix != ".caption" and x.suffix != ".csv"]
    entries.extend(inst_img_path)
            
    id_size_map = get_id_size_dict(entries)

    bucket_manager = AspectRatioBucket(id_size_map, **arg_config)
    
    return bucket_manager


class DistillationDataset(Dataset):
    """
    A dataset to prepare the images with the prompts for fine-tuning the model.
    It pre-processes the images and the tokenizes prompts.
    """

    def __init__(
        self,
        train_data_dir,
        resolution=1024,
        center_crop=False,
        random_flip=False,
        pad_tokens=False,
        ucg=0,
        debug_prompt=False,
        **kwargs
    ):
        self.train_data_dir = train_data_dir
        self.resolution = resolution
        self.center_crop = center_crop
        self.random_flip = random_flip
        self.pad_tokens = pad_tokens
        self.ucg = ucg
        self.debug_prompt = debug_prompt

        self.entries = []

        def prompt_resolver(x):
            img_item = (x, "")

            caption_path = os.path.splitext(x)[0] + ".caption"
            if os.path.exists(caption_path):
                with open(caption_path) as f:
                    content = f.read().strip()
                    img_item = (x, content)

            elif os.path.exists(os.path.splitext(x)[0] + ".txt"):
                with open(os.path.splitext(x)[0] + ".txt") as f:
                    content = f.read().strip()
                    img_item = (x, content)

            return img_item

        img_path = [prompt_resolver(x) for x in Path(self.train_data_dir).iterdir() if x.is_file() and x.suffix in [".jpg", ".png"]]
        self.entries.extend(img_path)

        random.shuffle(self.entries)
        self.num_images = len(self.entries)
        self._length = self.num_images

        # Preprocessing the datasets.
        self.train_resize = transforms.Resize(self.resolution, interpolation=transforms.InterpolationMode.BICUBIC)
        self.train_crop = transforms.CenterCrop(self.resolution) if self.center_crop else transforms.RandomCrop(self.resolution)
        self.train_flip = transforms.RandomHorizontalFlip(p=1.0)
        self.train_transforms = transforms.Compose([transforms.ToTensor(), transforms.Normalize([0.5], [0.5])])


    @staticmethod
    def read_img(filepath) -> Image:
        img = Image.open(filepath)

        if not img.mode == "RGB":
            img = img.convert("RGB")
        return img

    def __len__(self):
        return self._length

    def __getitem__(self, index):
        examples = {}
        image_path, prompt = self.entries[index % self.num_images]
        if random.random() <= self.ucg:
            prompt = ''

        image = self.read_img(image_path)
        if self.debug_prompt:
            print(f"prompt: {prompt}")

        # original_size = (image.height, image.width)
        original_size = (image.width, image.height)
        image = self.train_resize(image)

        if self.center_crop:
            y1 = max(0, int(round((image.height - self.resolution) / 2.0)))
            x1 = max(0, int(round((image.width - self.resolution) / 2.0)))
            image = self.train_crop(image)
        else:
            y1, x1, h, w = self.train_crop.get_params(image, (self.resolution, self.resolution))
            image = crop(image, y1, x1, h, w)

        if self.random_flip and random.random() < 0.5:
            x1 = image.width - x1
            image = self.train_flip(image)

        crop_top_left = (y1, x1)
        image = self.train_transforms(image)

        examples["original_size"] = original_size
        examples["crop_top_left"] = crop_top_left
        examples["pixel_values"] = image
        examples["caption"] = prompt

        return examples
    

class AspectRatioDataset(DistillationDataset):
    def __init__(self, debug_arb=False, enable_rotate=False, **kwargs):
        super().__init__(**kwargs)
        self.debug = debug_arb
        self.enable_rotate = enable_rotate
        self.prompt_cache = {}

        # cache prompts for reading
        for path, prompt in self.entries:
            self.prompt_cache[path] = prompt

    def denormalize(self, img, mean=0.5, std=0.5):
        res = transforms.Normalize((-1*mean/std), (1.0/std))(img.squeeze(0))
        res = torch.clamp(res, 0, 1)
        return res

    def transformer(self, image, size):
        x, y = image.size
        original_size = (x, y)
        short, long = (x, y) if x <= y else (y, x)

        w, h = size
        min_crop, max_crop = (w, h) if w <= h else (h, w)
        ratio_src, ratio_dst = float(long / short), float(max_crop / min_crop)
        
        if (x>y and w<h) or (x<y and w>h) and self.enable_rotate:
            # handle i/c mixed input
            image = image.rotate(90, expand=True)
            x, y = image.size
            
        if ratio_src > ratio_dst:
            new_w, new_h = (min_crop, int(min_crop * ratio_src)) if x<y else (int(min_crop * ratio_src), min_crop)
        elif ratio_src < ratio_dst:
            new_w, new_h = (max_crop, int(max_crop / ratio_src)) if x>y else (int(max_crop / ratio_src), max_crop)
        else:
            new_w, new_h = w, h

        # Preprocessing the datasets.
        self.train_resize = transforms.Resize((new_h, new_w), interpolation=transforms.InterpolationMode.BICUBIC)
        self.train_crop = transforms.CenterCrop((h, w)) if self.center_crop else transforms.RandomCrop((h, w))
        self.train_flip = transforms.RandomHorizontalFlip(p=1.0)
        self.train_transforms = transforms.Compose([transforms.ToTensor(), transforms.Normalize([0.5], [0.5])])

        new_image = self.train_resize(image)

        if self.center_crop:
            # print(new_image.height, new_image.width, h, w)
            y1 = max(0, int(round((new_image.height - h) / 2.0)))
            x1 = max(0, int(round((new_image.width - w) / 2.0)))
            new_image = self.train_crop(new_image)
        else:
            y1, x1, h, w = self.train_crop.get_params(new_image, (h, w))
            new_image = crop(new_image, y1, x1, h, w)

        if self.random_flip and random.random() < 0.5:
            x1 = new_image.width - x1
            new_image = self.train_flip(new_image)

        crop_top_left = (y1, x1)
        new_image = self.train_transforms(new_image)

        if self.debug:
            import uuid, torchvision
            print(x, y, "->", new_w, new_h, "->", new_image.shape)
            filename = str(uuid.uuid4())
            torchvision.utils.save_image(self.denormalize(new_image), f"/home/test/{filename}_1.jpg")
            torchvision.utils.save_image(torchvision.transforms.ToTensor()(image), f"/home/test/{filename}_2.jpg")
            print(f"saved: /tmp/{filename}")

        return new_image, crop_top_left, original_size

    def build_dict(self, item_id, size) -> dict:
        if item_id == "":
            return {}
        examples = {}
        prompt = self.prompt_cache[item_id]
        image = self.read_img(item_id)
        
        if random.random() < self.ucg:
            prompt = ''
            
        if self.debug_prompt:
            print("------------------------------------")
            print(item_id)
            print(f"prompt: {prompt}")
            print("------------------------------------")
        
        new_image, crop_top_left, original_size = self.transformer(image, size)

        examples["original_size"] = original_size
        examples["crop_top_left"] = crop_top_left
        examples["pixel_values"] = new_image
        examples["caption"] = prompt
        
        return examples

    def __getitem__(self, index):
        result = []
        for item in index:
            data_dict = self.build_dict(item["image"], item["size"])
            result.append({**data_dict})

        return result

class AspectRatioSampler(torch.utils.data.Sampler):
    def __init__(
        self, 
        buckets: AspectRatioBucket, 
        num_replicas: int = 1,
        debug: bool = False,
    ):
        super().__init__(None)
        self.bucket_manager = buckets
        self.num_replicas = num_replicas
        self.debug = debug
        self.iterator = buckets
            
    def build_res_id_dict(self, iter):
        base = {}
        for item, res in iter.generator():
            base.setdefault(res,[]).extend([item[0]])
        return base

    def find_closest(self, size, size_id_dict, typ):
        new_size = size
        if size not in size_id_dict or not any(size_id_dict[size]):
            kv = [(abs(s[0] / s[1] - size[0] / size[1]), s) for s in size_id_dict.keys() if any(size_id_dict[s])]            
            kv.sort(key=lambda e: e[0])
                
            new_size = kv[0][1]
            print(f"Warning: no {typ} image with {size} exists. Will use the closest ratio {new_size}.")

        return random.choice(size_id_dict[new_size])
    
    def __iter__(self):
        for batch, size in self.iterator.generator():
            result = []
            
            for item in batch:
                result.append({"image": item, "size": size})

            yield result

    def __len__(self):
        return len(self.iterator) // self.num_replicas


# Adapted from pipelines.StableDiffusionXLPipeline.encode_prompt
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)):
            # take a random caption if there are multiple
            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,
            )

            # We are only ALWAYS interested in the pooled output of the final text encoder
            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,
        split_batches=True,  # It's important to set this to True when using webdataset to get the right number of steps for lr scheduling. If set to False, the number of steps will be devide by the number of processes assuming batches are multiplied by the number of processes
    )

    # Make one log on every process with the configuration for debugging.
    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 passed along, set the training seed now.
    if args.seed is not None:
        set_seed(args.seed)

    # Handle the repository creation
    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

    # 1. Create the noise scheduler and the desired noise schedule.
    noise_scheduler = DDPMScheduler.from_pretrained(
        args.pretrained_teacher_model, subfolder="scheduler", revision=args.teacher_revision
    )

    # The scheduler calculates the alpha and sigma schedule for us
    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,
    )

    # 2. Load tokenizers from SD-XL checkpoint.
    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
    )

    # 3. Load text encoders from SD-XL checkpoint.
    # import correct text encoder classes
    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
    )

    # 4. Load VAE from SD-XL checkpoint (or more stable VAE)
    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,
    )

    # 5. Load teacher U-Net from SD-XL checkpoint
    teacher_unet = UNet2DConditionModel.from_pretrained(
        args.pretrained_teacher_model, subfolder="unet", revision=args.teacher_revision
    )

    # 6. Freeze teacher vae, text_encoders, and teacher_unet
    vae.requires_grad_(False)
    text_encoder_one.requires_grad_(False)
    text_encoder_two.requires_grad_(False)
    teacher_unet.requires_grad_(False)

    # 7. Create online (`unet`) student U-Nets.
    unet = UNet2DConditionModel.from_pretrained(
        args.pretrained_teacher_model, subfolder="unet", revision=args.teacher_revision
    )
    unet.train()

    # Check that all trainable models are in full precision
    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}"
        )

    # 8. Add LoRA to the student U-Net, only the LoRA projection matrix will be updated by the optimizer.
    lora_config = LoraConfig(
        r=args.lora_rank,
        target_modules=[
            "to_q",
            "to_k",
            "to_v",
            "to_out.0",
            "proj_in",
            "proj_out",
            "ff.net.0.proj",
            "ff.net.2",
            "conv1",
            "conv2",
            "conv_shortcut",
            "downsamplers.0.conv",
            "upsamplers.0.conv",
            "time_emb_proj",
        ],
    )
    unet = get_peft_model(unet, lora_config)

    # 9. Handle mixed precision and device placement
    # For mixed precision training we cast all non-trainable weigths to half-precision
    # as these weights are only used for inference, keeping weights in full precision is not required.
    weight_dtype = torch.float32
    if accelerator.mixed_precision == "fp16":
        weight_dtype = torch.float16
    elif accelerator.mixed_precision == "bf16":
        weight_dtype = torch.bfloat16

    # Move unet, vae and text_encoder to device and cast to weight_dtype
    # The VAE is in float32 to avoid NaN losses.
    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)

    # Move teacher_unet to device, optionally cast to weight_dtype
    teacher_unet.to(accelerator.device)
    if args.cast_teacher_unet:
        teacher_unet.to(dtype=weight_dtype)

    # Also move the alpha and sigma noise schedules to accelerator.device.
    alpha_schedule = alpha_schedule.to(accelerator.device)
    sigma_schedule = sigma_schedule.to(accelerator.device)
    solver = solver.to(accelerator.device)

    # 10. Handle saving and loading of checkpoints
    # `accelerate` 0.16.0 will have better support for customized saving
    if version.parse(accelerate.__version__) >= version.parse("0.16.0"):
        # create custom saving & loading hooks so that `accelerator.save_state(...)` serializes in a nice format
        def save_model_hook(models, weights, output_dir):
            if accelerator.is_main_process:
                unet_ = accelerator.unwrap_model(unet)
                lora_state_dict = get_peft_model_state_dict(unet_, adapter_name="default")
                StableDiffusionXLPipeline.save_lora_weights(os.path.join(output_dir, "unet_lora"), lora_state_dict)
                # save weights in peft format to be able to load them back
                unet_.save_pretrained(output_dir)

                for _, model in enumerate(models):
                    # make sure to pop weight so that corresponding model is not saved again
                    weights.pop()

        def load_model_hook(models, input_dir):
            # load the LoRA into the model
            unet_ = accelerator.unwrap_model(unet)
            unet_.load_adapter(input_dir, "default", is_trainable=True)

            for _ in range(len(models)):
                # pop models so that they are not loaded again
                models.pop()

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

    # 11. Enable optimizations
    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")

    # Enable TF32 for faster training on Ampere GPUs,
    # cf https://pytorch.org/docs/stable/notes/cuda.html#tensorfloat-32-tf32-on-ampere-devices
    if args.allow_tf32:
        torch.backends.cuda.matmul.allow_tf32 = True

    if args.gradient_checkpointing:
        unet.enable_gradient_checkpointing()

    # Use 8-bit Adam for lower memory usage or to fine-tune the model in 16GB GPUs
    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

    # 12. Optimizer creation
    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,
    )

    # 13. Dataset creation and data processing
    # Here, we compute not just the text embeddings but also the additional embeddings
    # needed for the SD XL UNet to operate.
    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(crop_coords, 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

        # Adapted from pipeline.StableDiffusionXLPipeline._get_add_time_ids
        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}

    # Create the dataset and the dataloader.
    dataset_class = AspectRatioDataset if args.use_aspect_ratio_bucket else DistillationDataset
    train_dataset = dataset_class(
        train_data_dir=args.train_data_dir,
        resolution=args.resolution,
        center_crop=args.center_crop,
        random_flip=args.random_flip,
        pad_tokens=args.pad_tokens,
        ucg=args.ucg,
        debug_prompt=args.debug_prompt,
        debug_arb=args.debug_arb,
    )

    def collate_fn_wrap(examples):
        # workround for variable list
        if len(examples) == 1:
            examples = examples[0]
        return collate_fn(examples)

    def collate_fn(examples):
        pixel_values = torch.stack([example["pixel_values"] for example in examples])
        pixel_values = pixel_values.to(memory_format=torch.contiguous_format).float()
        original_sizes = [example["original_size"] for example in examples]
        crop_top_lefts = [example["crop_top_left"] for example in examples]
        captions = [example["caption"] for example in examples]

        return {
            "pixel_values": pixel_values,
            "captions": captions,
            "original_sizes": original_sizes,
            "crop_top_lefts": crop_top_lefts,
        }
    
    if args.use_aspect_ratio_bucket:
        bucket_manager = init_arb_buckets(args, accelerator)
        sampler = AspectRatioSampler(bucket_manager, accelerator.num_processes)
        train_dataloader = torch.utils.data.DataLoader(
            train_dataset, 
            sampler=sampler, 
            collate_fn=collate_fn_wrap, 
            num_workers=args.dataloader_num_workers,
        )
    else:
        train_dataloader = torch.utils.data.DataLoader(
            train_dataset,
            batch_size=args.train_batch_size,
            shuffle=True,
            collate_fn=collate_fn, 
            num_workers=args.dataloader_num_workers,
        )

    # Let's first compute all the embeddings so that we can free up the text encoders
    # from memory.
    text_encoders = [text_encoder_one, text_encoder_two]
    tokenizers = [tokenizer_one, tokenizer_two]

    compute_embeddings_fn = functools.partial(
        compute_embeddings,
        proportion_empty_prompts=0,
        text_encoders=text_encoders,
        tokenizers=tokenizers,
    )

    # 14. LR Scheduler creation
    # Scheduler and math around the number of training steps.
    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,
    )

    # 15. Prepare for training
    # Prepare everything with our `accelerator`.
    unet, optimizer, lr_scheduler = accelerator.prepare(unet, optimizer, lr_scheduler)

    # We need to recalculate our total training steps as the size of the training dataloader may have changed.
    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
    # Afterwards we recalculate our number of training epochs
    args.num_train_epochs = math.ceil(args.max_train_steps / num_update_steps_per_epoch)

    # We need to initialize the trackers we use, and also store our configuration.
    # The trackers initializes automatically on the main process.
    if accelerator.is_main_process:
        tracker_config = dict(vars(args))
        accelerator.init_trackers(args.tracker_project_name, config=tracker_config)

    # Create uncond embeds for classifier free guidance
    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_dataset)}")
    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

    # Potentially load in the weights and states from a previous save
    if args.resume_from_checkpoint:
        if args.resume_from_checkpoint != "latest":
            path = os.path.basename(args.resume_from_checkpoint)
        else:
            # Get the most recent checkpoint
            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",
        # Only show the progress bar once on each machine.
        disable=not accelerator.is_local_main_process,
    )

    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["pixel_values"],
                    batch["captions"],
                    batch["original_sizes"],
                    batch["crop_top_lefts"],
                )
            

                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

                # encode pixel values with batch size of at most 8
                latents = []
                for i in range(0, pixel_values.shape[0], 8):
                    latents.append(vae.encode(pixel_values[i : i + 8]).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)

                # Sample noise that we'll add to the latents
                noise = torch.randn_like(latents)
                bsz = latents.shape[0]

                # Sample a random timestep for each image t_n ~ U[0, N - k - 1] without bias.
                topk = noise_scheduler.config.num_train_timesteps // args.num_ddim_timesteps
                index = torch.randint(0, args.num_ddim_timesteps, (bsz,), device=latents.device).long()
                start_timesteps = solver.ddim_timesteps[index]
                timesteps = start_timesteps - topk
                timesteps = torch.where(timesteps < 0, torch.zeros_like(timesteps), timesteps)

                # 20.4.4. Get boundary scalings for start_timesteps and (end) timesteps.
                c_skip_start, c_out_start = scalings_for_boundary_conditions(start_timesteps)
                c_skip_start, c_out_start = [append_dims(x, latents.ndim) for x in [c_skip_start, c_out_start]]
                c_skip, c_out = scalings_for_boundary_conditions(timesteps)
                c_skip, c_out = [append_dims(x, latents.ndim) for x in [c_skip, c_out]]

                # 20.4.5. Add noise to the latents according to the noise magnitude at each timestep
                # (this is the forward diffusion process) [z_{t_{n + k}} in Algorithm 1]
                noisy_model_input = noise_scheduler.add_noise(latents, noise, start_timesteps)

                # 20.4.6. Sample a random guidance scale w from U[w_min, w_max] and embed it
                w = (args.w_max - args.w_min) * torch.rand((bsz,)) + args.w_min
                w = w.reshape(bsz, 1, 1, 1)
                w = w.to(device=latents.device, dtype=latents.dtype)

                # 20.4.8. Prepare prompt embeds and unet_added_conditions
                prompt_embeds = encoded_text.pop("prompt_embeds")

                # 20.4.9. Get online LCM prediction on z_{t_{n + k}}, w, c, t_{n + k}
                noise_pred = unet(
                    noisy_model_input,
                    start_timesteps,
                    timestep_cond=None,
                    encoder_hidden_states=prompt_embeds.float(),
                    added_cond_kwargs=encoded_text,
                ).sample

                pred_x_0 = predicted_origin(
                    noise_pred,
                    start_timesteps,
                    noisy_model_input,
                    noise_scheduler.config.prediction_type,
                    alpha_schedule,
                    sigma_schedule,
                )

                model_pred = c_skip_start * noisy_model_input + c_out_start * pred_x_0

                # 20.4.10. Use the ODE solver to predict the kth step in the augmented PF-ODE trajectory after
                # noisy_latents with both the conditioning embedding c and unconditional embedding 0
                # Get teacher model prediction on noisy_latents and conditional embedding
                with torch.no_grad():
                    with torch.autocast("cuda"):
                        cond_teacher_output = teacher_unet(
                            noisy_model_input.to(weight_dtype),
                            start_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,
                            start_timesteps,
                            noisy_model_input,
                            noise_scheduler.config.prediction_type,
                            alpha_schedule,
                            sigma_schedule,
                        )

                        # Get teacher model prediction on noisy_latents and unconditional embedding
                        uncond_added_conditions = copy.deepcopy(encoded_text)
                        uncond_added_conditions["text_embeds"] = uncond_pooled_prompt_embeds
                        uncond_teacher_output = teacher_unet(
                            noisy_model_input.to(weight_dtype),
                            start_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,
                            start_timesteps,
                            noisy_model_input,
                            noise_scheduler.config.prediction_type,
                            alpha_schedule,
                            sigma_schedule,
                        )

                        # 20.4.11. Perform "CFG" to get x_prev estimate (using the LCM paper's CFG formulation)
                        pred_x0 = cond_pred_x0 + w * (cond_pred_x0 - uncond_pred_x0)
                        pred_noise = cond_teacher_output + w * (cond_teacher_output - uncond_teacher_output)
                        x_prev = solver.ddim_step(pred_x0, pred_noise, index)

                # 20.4.12. Get target LCM prediction on x_prev, w, c, t_n
                with torch.no_grad():
                    with torch.autocast("cuda", enabled=True, dtype=weight_dtype):
                        target_noise_pred = unet(
                            x_prev.float(),
                            timesteps,
                            timestep_cond=None,
                            encoder_hidden_states=prompt_embeds.float(),
                            added_cond_kwargs=encoded_text,
                        ).sample
                    pred_x_0 = predicted_origin(
                        target_noise_pred,
                        timesteps,
                        x_prev,
                        noise_scheduler.config.prediction_type,
                        alpha_schedule,
                        sigma_schedule,
                    )
                    target = c_skip * x_prev + c_out * pred_x_0

                # 20.4.13. Calculate loss
                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
                    )

                # 20.4.14. Backpropagate on the online student model (`unet`)
                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)

            # Checks if the accelerator has performed an optimization step behind the scenes
            if accelerator.sync_gradients:
                progress_bar.update(1)
                global_step += 1

                if accelerator.is_main_process:
                    if global_step % args.checkpointing_steps == 0:
                        # _before_ saving state, check if this save would set us over the `checkpoints_total_limit`
                        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]))

                            # before we save the new checkpoint, we need to have at _most_ `checkpoints_total_limit - 1` checkpoints
                            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)
                        unet = accelerator.unwrap_model(unet)
                        save_lora_model(unet, save_path, weight_dtype)
                        logger.info(f"Saved state to {save_path}")

                    # if global_step % args.validation_steps == 0:
                    #     log_validation(vae, unet, args, accelerator, weight_dtype, global_step)

            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()
    if accelerator.is_main_process:
        unet = accelerator.unwrap_model(unet)
        save_lora_model(unet, args.output_dir, weight_dtype)
        
    accelerator.end_training()


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