train.py

import argparse
import hashlib
import itertools
import math
import json
import random
import os
from pathlib import Path
from typing import Iterable, Optional
import subprocess
import sys
from typing import Optional
import inspect

import torch
import torch.nn.functional as F
import torch.utils.checkpoint
from torch.utils.data import Dataset
from torchinfo import summary

from accelerate import Accelerator
from accelerate.logging import get_logger
from accelerate.utils import set_seed
from diffusers import AutoencoderKL, DDPMScheduler, DDIMScheduler, StableDiffusionPipeline, UNet2DConditionModel
from diffusers.optimization import get_scheduler, get_cosine_with_hard_restarts_schedule_with_warmup
from diffusers.training_utils import EMAModel
from diffusers.utils.import_utils import is_xformers_available
from huggingface_hub import HfFolder, Repository, whoami
from PIL import Image
from torchvision import transforms
from tqdm.auto import tqdm
from transformers import CLIPTextModel, CLIPTokenizer

from lora_diffusion import (
    inject_trainable_lora,
    save_lora_weight,
    extract_lora_ups_down,
    monkeypatch_lora,
    tune_lora_scale,
)


logger = get_logger(__name__)


def parse_args(input_args=None):
    parser = argparse.ArgumentParser(description="Simple example of a training script.")
    parser.add_argument(
        "--pretrained_model_name_or_path",
        type=str,
        default=None,
        required=True,
        help="Path to pretrained model or model identifier from huggingface.co/models.",
    )
    parser.add_argument(
        "--pretrained_vae_name_or_path",
        type=str,
        default=None,
        help="Path to pretrained vae or vae identifier from huggingface.co/models.",
    )
    parser.add_argument(
        "--revision",
        type=str,
        default=None,
        required=False,
        help="Revision of pretrained model identifier from huggingface.co/models.",
    )
    parser.add_argument(
        "--tokenizer_name",
        type=str,
        default=None,
        help="Pretrained tokenizer name or path if not the same as model_name",
    )
    parser.add_argument(
        "--instance_data_dir",
        type=str,
        default=None,
        required=True,
        help="A folder containing the training data of instance images.",
    )
    parser.add_argument(
        "--class_data_dir",
        type=str,
        default=None,
        required=False,
        help="A folder containing the training data of class images.",
    )
    parser.add_argument(
        "--instance_prompt",
        type=str,
        default=None,
        required=True,
        help="The prompt with identifier specifying the instance",
    )
    parser.add_argument(
        "--class_prompt",
        type=str,
        default=None,
        help="The prompt to specify images in the same class as provided instance images.",
    )
    parser.add_argument(
        "--save_sample_prompt",
        type=str,
        default=None,
        help="The prompt used to generate sample outputs to save.",
    )
    parser.add_argument(
        "--save_sample_negative_prompt",
        type=str,
        default=None,
        help="The negative prompt used to generate sample outputs to save.",
    )
    parser.add_argument(
        "--n_save_sample",
        type=int,
        default=4,
        help="The number of samples to save.",
    )
    parser.add_argument("--save_interval", type=int, default=10_000, help="Save weights every N steps.")
    parser.add_argument("--save_min_steps", type=int, default=0, help="Start saving weights after N steps.")
    parser.add_argument(
        "--save_guidance_scale",
        type=float,
        default=7.5,
        help="CFG for save sample.",
    )
    parser.add_argument(
        "--save_infer_steps",
        type=int,
        default=50,
        help="The number of inference steps for save sample.",
    )
    parser.add_argument(
        "--with_prior_preservation",
        default=False,
        action="store_true",
        help="Flag to add prior preservation loss.",
    )
    parser.add_argument("--prior_loss_weight", type=float, default=1.0, help="The weight of prior preservation loss.")
    parser.add_argument(
        "--num_class_images",
        type=int,
        default=100,
        help=(
            "Minimal class images for prior preservation loss. If there are not enough images already present in"
            " class_data_dir, additional images will be sampled with class_prompt."
        ),
    )
    parser.add_argument(
        "--output_dir",
        type=str,
        default="text-inversion-model",
        help="The output directory where the model predictions and checkpoints will be written.",
    )
    parser.add_argument("--seed", type=int, default=None, help="A seed for reproducible training.")
    parser.add_argument(
        "--resolution",
        type=int,
        default=512,
        help=(
            "The resolution for input images, all the images in the train/validation dataset will be resized to this"
            " resolution"
        ),
    )
    parser.add_argument("--augment_min_resolution", type=int, default=None, help="Resize minimum image dimension before augmention pipeline.")
    parser.add_argument(
        "--augment_center_crop", action="store_true", help="Whether to center crop images before resizing to resolution"
    )
    parser.add_argument(
        "--augment_hflip", action="store_true", help="Whether to center crop images before resizing to resolution"
    )
    parser.add_argument("--train_text_encoder", action="store_true", help="Whether to train the text encoder")
    parser.add_argument(
        "--train_batch_size", type=int, default=4, help="Batch size (per device) for the training dataloader."
    )
    parser.add_argument(
        "--sample_batch_size", type=int, default=4, help="Batch size (per device) for sampling images."
    )
    parser.add_argument("--num_train_epochs", type=int, default=1)
    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(
        "--gradient_accumulation_steps",
        type=int,
        default=1,
        help="Number of updates steps to accumulate before performing a backward/update pass.",
    )
    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(
        "--learning_rate",
        type=float,
        default=5e-6,
        help="Initial learning rate (after the potential warmup period) to use.",
    )
    parser.add_argument(
        "--learning_rate_text",
        type=float,
        default=5e-6,
        help="Initial learning rate for text encoder (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(
        "--lr_cosine_num_cycles", type=float, default=1.0, help="Number of cycles when using cosine_with_restarts lr scheduler."
    )
    parser.add_argument("--use_ema", action="store_true", help="Whether to use EMA model.")
    parser.add_argument("--ema_inv_gamma", type=float, default=1.0, help="The inverse gamma parameter for the EMA model.")
    parser.add_argument("--ema_power", type=float, default=3 / 4, help="Exponential factor of EMA warmup.")
    parser.add_argument("--ema_min_value", type=float, default=0.0, help="The minimum EMA decay rate.")
    parser.add_argument("--ema_max_value", type=float, default=0.9999, help="The maximum EMA decay rate.")
    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("--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(
        "--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(
        "--log_gpu", action="store_true", help="Whether or not to log GPU memory usage."
    )
    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(
        "--use_image_captions",
        action="store_true",
        help="Get captions from textfile, otherwise filename",
    )
    parser.add_argument(
        "--conditioning_dropout_prob",
        type=float,
        default=0.0,
        help="Probability that conditioning is dropped.",
    )
    parser.add_argument(
        "--use_lora", action="store_true", help="Whether or not to use lora."
    )
    parser.add_argument(
        "--lora_rank",
        type=int,
        default=4,
        help="Rank reduction for LoRA.",
    )
    parser.add_argument(
        "--debug", action="store_true", help="Some exra verbosity."
    )
    parser.add_argument("--unconditional_prompt", type=str, default=" ", help="Prompt for conditioning dropout.")
    parser.add_argument("--local_rank", type=int, default=-1, help="For distributed training: local_rank")

    if input_args is not None:
        args = parser.parse_args(input_args)
    else:
        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.with_prior_preservation:
        if args.class_data_dir is None:
            raise ValueError("You must specify a data directory for class images.")
        if args.class_prompt is None:
            raise ValueError("You must specify prompt for class images.")

    return args


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

    def __init__(
        self,
        instance_data_root,
        instance_prompt,
        tokenizer,
        class_data_root=None,
        class_prompt=None,
        use_image_captions=False,
        unconditional_prompt=" ",
        size=512,
        augment_min_resolution=None,
        augment_center_crop=False,
        augment_hflip=False,
        debug=False,
    ):
        self.tokenizer = tokenizer
        self.use_image_captions = use_image_captions
        self.size = size
        self.augment_center_crop = augment_center_crop
        self.augment_hflip = augment_hflip

        self.instance_data_root = Path(instance_data_root)
        if not self.instance_data_root.exists():
            raise ValueError("Instance images root doesn't exists.")

        self.instance_images_path = [path for path in self.instance_data_root.glob('*') if '.txt' not in path.suffix]
        self.num_instance_images = len(self.instance_images_path)
        self.instance_prompt = instance_prompt
        self.debug = debug
        self._length = self.num_instance_images

        if class_data_root is not None:
            self.class_data_root = Path(class_data_root)
            self.class_data_root.mkdir(parents=True, exist_ok=True)

            self.class_images_path = [path for path in self.class_data_root.glob('*') if '.txt' not in path.suffix]
            random.shuffle(self.class_images_path)
            self.num_class_images = len(self.class_images_path)
            self._length = max(self.num_class_images, self.num_instance_images)
            self.class_prompt = class_prompt
        else:
            self.class_data_root = None

        self.unconditional_prompt = unconditional_prompt
        
        # Data augmentation pipeline
        augment_list = []
        if augment_min_resolution is not None:
            augment_list.append(transforms.Resize(augment_min_resolution))
        if augment_center_crop:
            augment_list.append(transforms.CenterCrop(size))
        else:
            augment_list.append(transforms.RandomCrop(size))
        if augment_hflip:
            augment_list.append(transforms.RandomHorizontalFlip(0.5))

        # Convert to format usable by model. 
        # Keep separate in case dumping augmentations to disk
        transform_list = []
        transform_list.append(transforms.ToTensor())
        transform_list.append(transforms.Normalize([0.5], [0.5]))
        
        if len(augment_list)>0:
            self.augment_transforms = transforms.Compose(augment_list)
        else:
            self.augment_transforms = None
            
        self.image_transforms = transforms.Compose(transform_list)

    def __len__(self):
        return self._length

    def __getitem__(self, index):
        example = {}
        image_path = self.instance_images_path[index % self.num_instance_images]
        instance_image = Image.open(image_path)
        if not instance_image.mode == "RGB":
            instance_image = instance_image.convert("RGB")
        if self.augment_transforms is not None:
            instance_image = self.augment_transforms(instance_image)
            if self.debug:
                hash_image = hashlib.sha1(instance_image.tobytes()).hexdigest()
                image_filename = image_path.stem + f"-{hash_image}.jpg"
                instance_image.save(os.path.join("/content/augment", image_filename))
        example["instance_images"] = self.image_transforms(instance_image)

        if self.use_image_captions:
            caption_path = image_path.with_suffix(".txt")
            if caption_path.exists():
                with open(caption_path) as f:
                    caption = f.read()
            else:
                caption = caption_path.stem
                
            caption = ''.join([i for i in caption if not i.isdigit()]) # not sure necessary
            caption = caption.replace("_"," ")
            self.instance_prompt = caption
            
        example["instance_prompt_ids"] = self.tokenizer(
            self.instance_prompt,
            padding="do_not_pad",
            truncation=True,
            max_length=self.tokenizer.model_max_length,
        ).input_ids
        
        if self.debug:
            print("\nInstance: " + str(image_path))
            print(self.instance_prompt)

        if self.class_data_root:
            image_path = self.class_images_path[index % self.num_class_images]
            class_image = Image.open(image_path)
            if not class_image.mode == "RGB":
                class_image = class_image.convert("RGB")
            if self.augment_transforms is not None:
                class_image = self.augment_transforms(class_image)
                if self.debug:
                    hash_image = hashlib.sha1(class_image.tobytes()).hexdigest()
                    image_filename = image_path.stem + f"-{hash_image}.jpg"
                    class_image.save(os.path.join("/content/augment", image_filename))
            example["class_images"] = self.image_transforms(class_image)
            
            if self.use_image_captions:
                caption_path = image_path.with_suffix(".txt")
                if caption_path.exists():
                    with open(caption_path) as f:
                        caption = f.read()
                else:
                    caption = caption_path.stem

                caption = ''.join([i for i in caption if not i.isdigit()]) # not sure necessary
                caption = caption.replace("_"," ")
                self.class_prompt = caption
            
            example["class_prompt_ids"] = self.tokenizer(
                self.class_prompt,
                padding="do_not_pad",
                truncation=True,
                max_length=self.tokenizer.model_max_length,
            ).input_ids
            
            if self.debug:
                print("\nClass: " + str(image_path))
                print(self.class_prompt)

        example["unconditional_prompt_ids"] = self.tokenizer(
                self.unconditional_prompt,
                padding="do_not_pad",
                truncation=True,
                max_length=self.tokenizer.model_max_length,
            ).input_ids

        return example


class PromptDataset(Dataset):
    "A simple dataset to prepare the prompts to generate class images on multiple GPUs."

    def __init__(self, prompt, num_samples):
        self.prompt = prompt
        self.num_samples = num_samples

    def __len__(self):
        return self.num_samples

    def __getitem__(self, index):
        example = {}
        example["prompt"] = self.prompt
        example["index"] = index
        return example


def get_full_repo_name(model_id: str, organization: Optional[str] = None, token: Optional[str] = None):
    if token is None:
        token = HfFolder.get_token()
    if organization is None:
        username = whoami(token)["name"]
        return f"{username}/{model_id}"
    else:
        return f"{organization}/{model_id}"

def get_gpu_memory_map():
    result = subprocess.check_output(
        [
            'nvidia-smi', '--query-gpu=memory.used',
            '--format=csv,nounits,noheader'
        ], encoding='utf-8')
    gpu_memory = [int(x) for x in result.strip().split('\n')]
    gpu_memory_map = dict(zip(range(len(gpu_memory)), gpu_memory))
    return gpu_memory_map
        
    
def image_grid(imgs, rows, cols):
    assert len(imgs) == rows*cols
    w, h = imgs[0].size
    grid = Image.new('RGB', size=(cols*w, rows*h))
    grid_w, grid_h = grid.size
    for i, img in enumerate(imgs):
        grid.paste(img, box=(i%cols*w, i//cols*h))
    return grid


def main(args):
    torch.set_printoptions(precision=10)
    logging_dir = Path(args.output_dir, args.logging_dir)

    accelerator = Accelerator(
        gradient_accumulation_steps=args.gradient_accumulation_steps,
        mixed_precision=args.mixed_precision,
        log_with="tensorboard",
        logging_dir=logging_dir,
    )

    # Currently, it's not possible to do gradient accumulation when training two models with accelerate.accumulate
    # This will be enabled soon in accelerate. For now, we don't allow gradient accumulation when training two models.
    # TODO (patil-suraj): Remove this check when gradient accumulation with two models is enabled in accelerate.
    if args.train_text_encoder and args.gradient_accumulation_steps > 1 and accelerator.num_processes > 1:
        raise ValueError(
            "Gradient accumulation is not supported when training the text encoder in distributed training. "
            "Please set gradient_accumulation_steps to 1. This feature will be supported in the future."
        )

    if args.seed is not None:
        #cudnn.benchmark = False
        #cudnn.deterministic = True
        set_seed(args.seed)

    if args.with_prior_preservation:
        pipeline = None
        class_images_dir = Path(args.class_data_dir)
        if not class_images_dir.exists():
            class_images_dir.mkdir(parents=True)
        cur_class_images = len(list(class_images_dir.iterdir()))

        if cur_class_images < args.num_class_images:
            torch_dtype = torch.float16 if accelerator.device.type == "cuda" else torch.float32
            if pipeline is None:
                pipeline = StableDiffusionPipeline.from_pretrained(
                    args.pretrained_model_name_or_path,
                    vae=AutoencoderKL.from_pretrained(
                        args.pretrained_vae_name_or_path or args.pretrained_model_name_or_path,
                        subfolder=None if args.pretrained_vae_name_or_path else "vae",
                        revision=None if args.pretrained_vae_name_or_path else args.revision,
                        torch_dtype=torch_dtype,
                    ),
                    torch_dtype=torch_dtype,
                    safety_checker=None,
                    revision=args.revision
                )
                pipeline.set_progress_bar_config(disable=True)
                pipeline.to(accelerator.device)

            num_new_images = args.num_class_images - cur_class_images
            logger.info(f"Number of class images to sample: {num_new_images}.")

            sample_dataset = PromptDataset(args.class_prompt, num_new_images)
            sample_dataloader = torch.utils.data.DataLoader(sample_dataset, batch_size=args.sample_batch_size)

            sample_dataloader = accelerator.prepare(sample_dataloader)

            for example in tqdm(
                sample_dataloader, desc="Generating class images", disable=not accelerator.is_local_main_process
            ):
                images = pipeline(example["prompt"]).images

                for i, image in enumerate(images):
                    hash_image = hashlib.sha1(image.tobytes()).hexdigest()
                    image_filename = class_images_dir / f"{example['index'][i] + cur_class_images}-{hash_image}.jpg"
                    image.save(image_filename)

            del pipeline
            if torch.cuda.is_available():
                torch.cuda.empty_cache()

    # Handle the repository creation
    if accelerator.is_main_process:
        if args.push_to_hub:
            if args.hub_model_id is None:
                repo_name = get_full_repo_name(Path(args.output_dir).name, token=args.hub_token)
            else:
                repo_name = args.hub_model_id
            repo = Repository(args.output_dir, clone_from=repo_name)

            with open(os.path.join(args.output_dir, ".gitignore"), "w+") as gitignore:
                if "step_*" not in gitignore:
                    gitignore.write("step_*\n")
                if "epoch_*" not in gitignore:
                    gitignore.write("epoch_*\n")
        elif args.output_dir is not None:
            os.makedirs(args.output_dir, exist_ok=True)

    # Load the tokenizer
    if args.tokenizer_name:
        tokenizer = CLIPTokenizer.from_pretrained(
            args.tokenizer_name,
            revision=args.revision,
        )
    elif args.pretrained_model_name_or_path:
        tokenizer = CLIPTokenizer.from_pretrained(
            args.pretrained_model_name_or_path,
            subfolder="tokenizer",
            revision=args.revision,
        )

    # Load models and create wrapper for stable diffusion
    text_encoder = CLIPTextModel.from_pretrained(
        args.pretrained_model_name_or_path,
        subfolder="text_encoder",
        revision=args.revision,
    )
    vae = AutoencoderKL.from_pretrained(        
        args.pretrained_vae_name_or_path or args.pretrained_model_name_or_path,
        subfolder=None if args.pretrained_vae_name_or_path else "vae",
        revision=None if args.pretrained_vae_name_or_path else args.revision,
    )
    unet = UNet2DConditionModel.from_pretrained(
        args.pretrained_model_name_or_path,
        subfolder="unet",
        revision=args.revision,
    )

    if is_xformers_available():
        try:
            unet.enable_xformers_memory_efficient_attention()
        except Exception as e:
            logger.warning(
                "Could not enable memory efficient attention. Make sure xformers is installed"
                f" correctly and a GPU is available: {e}"
            )

    if args.use_lora:
        unet.requires_grad_(False)
        unet_lora_params, unet_names = inject_trainable_lora(unet, r=args.lora_rank)
        if args.debug:
            for _up, _down in extract_lora_ups_down(unet):
                print("Before training: Unet First Layer lora up", _up.weight.data)
                print("Before training: Unet First Layer lora down", _down.weight.data)
                break
    
    vae.requires_grad_(False)
    
    if args.train_text_encoder and args.use_lora:
        text_encoder.requires_grad_(False)
        text_encoder_lora_params, text_encoder_names = inject_trainable_lora(
            text_encoder, target_replace_module=["CLIPAttention"],
            r=args.lora_rank,
        )
        if args.debug:
            for _up, _down in extract_lora_ups_down(
                text_encoder, target_replace_module=["CLIPAttention"]
            ):
                print("Before training: text encoder First Layer lora up", _up.weight.data)
                print("Before training: text encoder First Layer lora down", _down.weight.data)
                break
    elif not args.train_text_encoder:
        text_encoder.requires_grad_(False)
            
    if args.gradient_checkpointing:
        unet.enable_gradient_checkpointing()
        if args.train_text_encoder:
            text_encoder.gradient_checkpointing_enable()

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

    # 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

    text_lr = (
        args.learning_rate
        if args.learning_rate_text is None
        else args.learning_rate_text
    )
    
    if args.use_lora:
        params_to_optimize = (
            [
                {
                    "params": itertools.chain(*unet_lora_params), "lr": args.learning_rate
                },
                {
                    "params": itertools.chain(*text_encoder_lora_params),
                    "lr": text_lr,
                },
            ]
            if args.train_text_encoder
            else itertools.chain(*unet_lora_params)
        )   
    else: 
         params_to_optimize = (
            [
                {
                    "params": itertools.chain(unet.parameters()), "lr": args.learning_rate
                },
                {
                    "params": itertools.chain(text_encoder.parameters()),
                    "lr": text_lr,
                },
            ]
            if args.train_text_encoder
            else unet.parameters()
        )

    if args.debug:
        print(summary(vae, col_names=["num_params", "trainable"], verbose=1))
        print(summary(unet, col_names=["num_params", "trainable"], verbose=1))
        print(summary(text_encoder, col_names=["num_params", "trainable"], verbose=1))

        with open(os.path.join(args.output_dir, "vae.txt"), "w") as f:
            f.write(str(summary(vae, col_names=["num_params", "trainable"], verbose=2)))
            f.close()
        with open(os.path.join(args.output_dir, "unet.txt"), "w") as f:
            f.write(str(summary(unet, col_names=["num_params", "trainable"], verbose=2)))
            f.close()
        with open(os.path.join(args.output_dir, "text_encoder.txt"), "w") as f:
            f.write(str(summary(text_encoder, col_names=["num_params", "trainable"], verbose=2)))
            f.close()
        
    optimizer = optimizer_class(
        params_to_optimize,
        lr=args.learning_rate,
        betas=(args.adam_beta1, args.adam_beta2),
        weight_decay=args.adam_weight_decay,
        eps=args.adam_epsilon,
    )

    noise_scheduler = DDPMScheduler.from_pretrained(args.pretrained_model_name_or_path, subfolder="scheduler")

    train_dataset = DreamBoothDataset(
        instance_data_root=args.instance_data_dir,
        instance_prompt=args.instance_prompt,
        class_data_root=args.class_data_dir if args.with_prior_preservation else None,
        class_prompt=args.class_prompt,
        use_image_captions=args.use_image_captions,
        unconditional_prompt=args.unconditional_prompt,
        tokenizer=tokenizer,
        size=args.resolution,
        augment_min_resolution=args.augment_min_resolution,
        augment_center_crop=args.augment_center_crop,
        augment_hflip=args.augment_hflip,
        debug=args.debug,
    )

    def collate_fn(examples):
        input_ids = [example["instance_prompt_ids"] for example in examples]
        pixel_values = [example["instance_images"] for example in examples]

        # Concat class and instance examples for prior preservation.
        # We do this to avoid doing two forward passes.
        if args.with_prior_preservation:
            input_ids += [example["class_prompt_ids"] for example in examples]
            pixel_values += [example["class_images"] for example in examples]
        
        # Apply text-conditioning dropout by inserting uninformative prompt
        if args.conditioning_dropout_prob > 0:
            unconditional_ids = [example["unconditional_prompt_ids"] for example in examples]*2
            for i, input_id in enumerate(input_ids):
                if random.uniform(0.0, 1.0) <= args.conditioning_dropout_prob:
                    input_ids[i] = unconditional_ids[i]

        pixel_values = torch.stack(pixel_values)
        pixel_values = pixel_values.to(memory_format=torch.contiguous_format).float()

        input_ids = tokenizer.pad(
            {"input_ids": input_ids},
            padding="max_length",
            max_length=tokenizer.model_max_length,
            return_tensors="pt",
        ).input_ids

        batch = {
            "input_ids": input_ids,
            "pixel_values": pixel_values,
        }
        return batch

    train_dataloader = torch.utils.data.DataLoader(
        train_dataset, batch_size=args.train_batch_size, shuffle=True, collate_fn=collate_fn, num_workers=1
    )

    # 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.lr_scheduler=="cosine_with_restarts":
        lr_scheduler = get_cosine_with_hard_restarts_schedule_with_warmup(
            optimizer=optimizer,
            num_warmup_steps=args.lr_warmup_steps * args.gradient_accumulation_steps,
            num_training_steps=args.max_train_steps * args.gradient_accumulation_steps,
            num_cycles=args.lr_cosine_num_cycles,
        )        
    else:
        lr_scheduler = get_scheduler(
            args.lr_scheduler,
            optimizer=optimizer,
            num_warmup_steps=args.lr_warmup_steps * args.gradient_accumulation_steps,
            num_training_steps=args.max_train_steps * args.gradient_accumulation_steps,
        )

    if args.train_text_encoder:
        unet, text_encoder, optimizer, train_dataloader, lr_scheduler = accelerator.prepare(
            unet, text_encoder, optimizer, train_dataloader, lr_scheduler
        )
    else:
        unet, optimizer, train_dataloader, lr_scheduler = accelerator.prepare(
            unet, optimizer, train_dataloader, lr_scheduler
        )

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

    # Move text_encode and vae to gpu.
    # For mixed precision training we cast the text_encoder and vae weights to half-precision
    # as these models are only used for inference, keeping weights in full precision is not required.
    vae.to(accelerator.device, dtype=weight_dtype)
    if not args.train_text_encoder:
        text_encoder.to(accelerator.device, dtype=weight_dtype)

    # Create EMA for the unet.
    if args.use_ema:
        ema_unet = EMAModel(
            accelerator.unwrap_model(unet), 
            inv_gamma=args.ema_inv_gamma, 
            power=args.ema_power, 
            min_value=args.ema_min_value,
            max_value=args.ema_max_value
        )

    # 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:
        accelerator.init_trackers("dreambooth", config=vars(args))

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

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

    def save_weights(step):
        # Create the pipeline using using the trained modules and save it.
        if accelerator.is_main_process:
            save_dir = os.path.join(args.output_dir, f"{step}")
            if not os.path.exists(save_dir):
                os.makedirs(save_dir)

            with open(os.path.join(save_dir, "args.json"), "w") as f:
                json.dump(args.__dict__, f, indent=2)

            # https://github.com/huggingface/diffusers/issues/1566
            accepts_keep_fp32_wrapper = "keep_fp32_wrapper" in set(
                inspect.signature(accelerator.unwrap_model).parameters.keys()
            )
            extra_args = (
                {"keep_fp32_wrapper": True} if accepts_keep_fp32_wrapper else {}
            )
                    
            if args.train_text_encoder:
                text_enc_model = accelerator.unwrap_model(text_encoder, **extra_args)
            else:
                text_enc_model = CLIPTextModel.from_pretrained(args.pretrained_model_name_or_path, subfolder="text_encoder", revision=args.revision)

            pipeline = StableDiffusionPipeline.from_pretrained(
                args.pretrained_model_name_or_path,
                unet=accelerator.unwrap_model(
                        ema_unet.averaged_model if args.use_ema else unet,
                        **extra_args,
                    ),
                text_encoder=text_enc_model,
                vae=AutoencoderKL.from_pretrained(
                    args.pretrained_vae_name_or_path or args.pretrained_model_name_or_path,
                    subfolder=None if args.pretrained_vae_name_or_path else "vae",
                    revision=None if args.pretrained_vae_name_or_path else args.revision,
                ),
                safety_checker=None,
                torch_dtype=torch.float16,
                revision=args.revision,
            )
            
            if args.use_lora:
                save_lora_weight(pipeline.unet, os.path.join(save_dir, "lora_unet.pt"))
                if args.debug:
                    for _up, _down in extract_lora_ups_down(pipeline.unet):
                        print("First Unet Layer's Up Weight is now : ", _up.weight.data)
                        print("First Unet Layer's Down Weight is now : ", _down.weight.data)
                        break
                    
                if args.train_text_encoder:
                    save_lora_weight(
                        pipeline.text_encoder,
                        os.path.join(save_dir, "lora_text_encoder.pt"),
                        target_replace_module=["CLIPAttention"],
                    )
                    if args.debug:
                        for _up, _down in extract_lora_ups_down(
                            pipeline.text_encoder,
                            target_replace_module=["CLIPAttention"],
                        ):
                            print("First Text Encoder Layer's Up Weight is now : ", _up.weight.data)
                            print("First Text Encoder Layer's Down Weight is now : ", _down.weight.data)
                            break

                # No arguments yet, leave at defaults for now
                #tune_lora_scale(pipeline.text_encoder, 1.00)
                #tune_lora_scale(pipeline.unet, 1.00)
            else:
                pipeline.save_pretrained(save_dir)

            if args.save_sample_prompt is not None:
                pipeline = pipeline.to(accelerator.device)
                if is_xformers_available():
                    pipeline.enable_xformers_memory_efficient_attention()
                    
                save_sample_prompt = args.save_sample_prompt
                save_sample_prompt = list(map(str.strip, save_sample_prompt.split('//')))
                pipeline = pipeline.to(accelerator.device)
                g_cuda = torch.Generator(device=accelerator.device).manual_seed(args.seed)
                pipeline.set_progress_bar_config(disable=True)
                sample_dir = os.path.join(save_dir, "samples")
                os.makedirs(sample_dir, exist_ok=True)
                
                with torch.autocast("cuda"), torch.inference_mode():
                    all_images = []
                    for i in tqdm(range(args.n_save_sample), desc="Generating samples"):
                        images = pipeline(
                            save_sample_prompt,
                            negative_prompt=[args.save_sample_negative_prompt]*len(save_sample_prompt),
                            guidance_scale=args.save_guidance_scale,
                            num_inference_steps=args.save_infer_steps,
                            generator=g_cuda
                        ).images
                        all_images.extend(images)
                        
                    grid = image_grid(all_images, rows=args.n_save_sample, cols=len(save_sample_prompt))
                    grid.save(os.path.join(sample_dir, f"{step}.jpg"), quality=90, optimize=True)
                    
                del pipeline
                if torch.cuda.is_available():
                    torch.cuda.empty_cache()
            print(f"[*] Weights saved at {save_dir}")

    # Only show the progress bar once on each machine.
    progress_bar = tqdm(range(args.max_train_steps), disable=not accelerator.is_local_main_process)
    progress_bar.set_description("Steps")
    global_step = 0

    for epoch in range(args.num_train_epochs):
        unet.train()
        if args.train_text_encoder:
            text_encoder.train()
        for step, batch in enumerate(train_dataloader):
            with accelerator.accumulate(unet):
                # Convert images to latent space
                latents = vae.encode(batch["pixel_values"].to(dtype=weight_dtype)).latent_dist.sample()
                latents = latents * 0.18215

                # 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
                timesteps = torch.randint(0, noise_scheduler.config.num_train_timesteps, (bsz,), device=latents.device)
                timesteps = timesteps.long()

                # Add noise to the latents according to the noise magnitude at each timestep
                # (this is the forward diffusion process)
                noisy_latents = noise_scheduler.add_noise(latents, noise, timesteps)

                # Get the text embedding for conditioning
                encoder_hidden_states = text_encoder(batch["input_ids"])[0]

                # Predict the noise residual
                model_pred = unet(noisy_latents, timesteps, encoder_hidden_states).sample

                # Get the target for loss depending on the prediction type
                if noise_scheduler.config.prediction_type == "epsilon":
                    target = noise
                elif noise_scheduler.config.prediction_type == "v_prediction":
                    target = noise_scheduler.get_velocity(latents, noise, timesteps)
                else:
                    raise ValueError(f"Unknown prediction type {noise_scheduler.config.prediction_type}")

                if args.with_prior_preservation:
                    # Chunk the noise and model_pred into two parts and compute the loss on each part separately.
                    model_pred, model_pred_prior = torch.chunk(model_pred, 2, dim=0)
                    target, target_prior = torch.chunk(target, 2, dim=0)

                    # Compute instance loss
                    pred_loss = F.mse_loss(model_pred.float(), target.float(), reduction="none").mean([1, 2, 3]).mean()

                    # Compute prior loss
                    prior_loss = F.mse_loss(model_pred_prior.float(), target_prior.float(), reduction="mean")

                    # Add the prior loss to the instance loss.
                    loss = pred_loss + args.prior_loss_weight * prior_loss
                else:
                    loss = F.mse_loss(model_pred.float(), target.float(), reduction="mean")

                accelerator.backward(loss)
                if accelerator.sync_gradients:
                    params_to_clip = (
                        itertools.chain(unet.parameters(), text_encoder.parameters())
                        if args.train_text_encoder
                        else unet.parameters()
                    )
                    accelerator.clip_grad_norm_(params_to_clip, args.max_grad_norm)
                optimizer.step()
                lr_scheduler.step()
                if args.use_ema:
                    ema_unet.step(unet)
                optimizer.zero_grad()

            # Checks if the accelerator has performed an optimization step behind the scenes
            if accelerator.sync_gradients:
                progress_bar.update(1)
                global_step += 1
                
                if global_step > 0 and not global_step % args.save_interval and global_step >= args.save_min_steps:
                    save_weights(global_step)
                            
            if args.with_prior_preservation:
                logs = {"Loss/pred": pred_loss.detach().item(),
                        "Loss/prior": prior_loss.detach().item(),
                        "Loss/total": loss.detach().item(),
                       }
            else:
                logs = {"Loss/pred": loss.detach().item()}

            if args.train_text_encoder:
                logs["lr/unet"] = lr_scheduler.get_last_lr()[0]
                logs["lr/text"] = lr_scheduler.get_last_lr()[1]
            else:
                logs["lr"] = lr_scheduler.get_last_lr()[0]

                
            if args.log_gpu:
                logs["GPU"] = get_gpu_memory_map()[0]
                                
            if args.use_ema:
                logs["ema_decay"] = ema_unet.decay
                    
            progress_bar.set_postfix(**logs)
            accelerator.log(logs, step=global_step)


            if global_step >= args.max_train_steps:
                break
            
        accelerator.wait_for_everyone()

    if accelerator.is_main_process:
        save_weights(global_step)
    
    accelerator.end_training()


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