train_imagic.py

import argparse
import math
import os
from pathlib import Path
from typing import Optional

import torch
import torch.nn.functional as F
import torch.utils.checkpoint

from accelerate import Accelerator
from accelerate.logging import get_logger
from accelerate.utils import set_seed
from diffusers import AutoencoderKL, DDPMScheduler, StableDiffusionPipeline, UNet2DConditionModel
from huggingface_hub import HfFolder, Repository, whoami
from PIL import Image
import numpy as np
from torchvision import transforms
from tqdm.auto import tqdm
from transformers import CLIPTextModel, CLIPTokenizer


logger = get_logger(__name__)


def parse_args():
    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(
        "--tokenizer_name",
        type=str,
        default=None,
        help="Pretrained tokenizer name or path if not the same as model_name",
    )
    parser.add_argument(
        "--input_image",
        type=str,
        default=None,
        required=True,
        help="Path to input image to edit.",
    )
    parser.add_argument(
        "--target_text",
        type=str,
        default=None,
        help="The target text describing the output image.",
    )
    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(
        "--center_crop", action="store_true", help="Whether to center crop images before resizing to resolution"
    )
    parser.add_argument(
        "--train_batch_size", type=int, default=4, help="Batch size (per device) for the training dataloader."
    )
    parser.add_argument(
        "--emb_train_steps",
        type=int,
        default=500,
        help="Total number of training steps to perform.",
    )
    parser.add_argument(
        "--max_train_steps",
        type=int,
        default=1000,
        help="Total number of training steps to perform.",
    )
    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(
        "--emb_learning_rate",
        type=float,
        default=1e-3,
        help="Learning rate for optimizing the embeddings.",
    )
    parser.add_argument(
        "--learning_rate",
        type=float,
        default=1e-6,
        help="Learning rate for fine tuning the model.",
    )
    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(
        "--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_interval", type=int, default=10, help="Log every N steps.")
    parser.add_argument(
        "--mixed_precision",
        type=str,
        default="no",
        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."
        ),
    )
    parser.add_argument("--local_rank", type=int, default=-1, help="For distributed training: local_rank")

    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

    return args


class AverageMeter:
    def __init__(self, name=None):
        self.name = name
        self.reset()

    def reset(self):
        self.sum = self.count = self.avg = 0

    def update(self, val, n=1):
        self.sum += val * n
        self.count += n
        self.avg = self.sum / self.count


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 main():
    args = parse_args()
    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,
    )

    if args.seed is not None:
        set_seed(args.seed)

    # 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)
    elif args.pretrained_model_name_or_path:
        tokenizer = CLIPTokenizer.from_pretrained(args.pretrained_model_name_or_path, subfolder="tokenizer", use_auth_token=True)

    # Load models and create wrapper for stable diffusion
    text_encoder = CLIPTextModel.from_pretrained(args.pretrained_model_name_or_path, subfolder="text_encoder", use_auth_token=True)
    vae = AutoencoderKL.from_pretrained(args.pretrained_model_name_or_path, subfolder="vae", use_auth_token=True)
    unet = UNet2DConditionModel.from_pretrained(args.pretrained_model_name_or_path, subfolder="unet", use_auth_token=True)

    if args.gradient_checkpointing:
        unet.enable_gradient_checkpointing()

    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.Adam8bit
    else:
        optimizer_class = torch.optim.Adam

    noise_scheduler = DDPMScheduler(
        beta_start=0.00085, beta_end=0.012, beta_schedule="scaled_linear", num_train_timesteps=1000
    )

    weight_dtype = torch.float32
    if args.mixed_precision == "fp16":
        weight_dtype = torch.float16
    elif args.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.
    text_encoder.to(accelerator.device, dtype=weight_dtype)
    vae.to(accelerator.device, dtype=weight_dtype)

    # Encode the input image.
    input_image = Image.open(args.input_image).convert("RGB")

    image_transforms = transforms.Compose(
        [
            transforms.Resize(args.resolution, interpolation=transforms.InterpolationMode.BILINEAR),
            transforms.CenterCrop(args.resolution) if args.center_crop else transforms.RandomCrop(args.resolution),
            transforms.ToTensor(),
            transforms.Normalize([0.5], [0.5]),
        ]
    )

    init_image = image_transforms(input_image)
    init_image = init_image[None].to(device=accelerator.device, dtype=weight_dtype)
    with torch.inference_mode():
        init_latents = vae.encode(init_image).latent_dist.sample()
        init_latents = 0.18215 * init_latents

    # Encode the target text.
    text_ids = tokenizer(
        args.target_text,
        padding="max_length",
        truncation=True,
        max_length=tokenizer.model_max_length,
        return_tensors="pt",
    ).input_ids

    text_ids = text_ids.to(device=accelerator.device)
    with torch.inference_mode():
        target_embeddings = text_encoder(text_ids)[0]
    
    del vae, text_encoder
    if torch.cuda.is_available():
        torch.cuda.empty_cache()

    target_embeddings = target_embeddings.float()
    optimized_embeddings = target_embeddings.clone()

    # Optimize the text embeddings first.
    optimized_embeddings.requires_grad_(True)
    optimizer = optimizer_class(
        [optimized_embeddings],  # only optimize embeddings
        lr=args.emb_learning_rate,
        betas=(args.adam_beta1, args.adam_beta2),
        # weight_decay=args.adam_weight_decay,
        eps=args.adam_epsilon,
    )

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

    # 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("imagic", config=vars(args))

    def train_loop(pbar, optimizer, params):
        loss_avg = AverageMeter()
        for step in pbar:
            with accelerator.accumulate(unet):
                noise = torch.randn_like(init_latents)
                bsz = init_latents.shape[0]
                # Sample a random timestep for each image
                timesteps = torch.randint(0, noise_scheduler.config.num_train_timesteps, (bsz,), device=init_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(init_latents, noise, timesteps)

                noise_pred = unet(noisy_latents, timesteps, optimized_embeddings).sample
                
                loss = F.mse_loss(noise_pred.float(), noise.float(), reduction="mean")

                accelerator.backward(loss)
                # if accelerator.sync_gradients:     # results aren't good with it, may be will need more training with it.
                #     accelerator.clip_grad_norm_(params, args.max_grad_norm)
                optimizer.step()
                optimizer.zero_grad(set_to_none=True)
                loss_avg.update(loss.detach_(), bsz)

            if not step % args.log_interval:
                logs = {"loss": loss_avg.avg.item()}
                progress_bar.set_postfix(**logs)
                accelerator.log(logs, step=step)

        accelerator.wait_for_everyone()
    
    progress_bar = tqdm(range(args.emb_train_steps), disable=not accelerator.is_local_main_process)
    progress_bar.set_description("Optimizing embedding")
    
    train_loop(progress_bar, optimizer, optimized_embeddings)

    optimized_embeddings.requires_grad_(False)
    if accelerator.is_main_process:
        torch.save(target_embeddings.cpu(), os.path.join(args.output_dir, "target_embeddings.pt"))
        torch.save(optimized_embeddings.cpu(), os.path.join(args.output_dir, "optimized_embeddings.pt"))
        with open(os.path.join(args.output_dir, "target_text.txt"), "w") as f:
            f.write(args.target_text)

    # Fine tune the diffusion model.
    optimizer = optimizer_class(
        accelerator.unwrap_model(unet).parameters(),
        lr=args.learning_rate,
        betas=(args.adam_beta1, args.adam_beta2),
        # weight_decay=args.adam_weight_decay,
        eps=args.adam_epsilon,
    )
    optimizer = accelerator.prepare(optimizer)

    progress_bar = tqdm(range(args.max_train_steps), disable=not accelerator.is_local_main_process)
    progress_bar.set_description("Fine Tuning")
    unet.train()

    train_loop(progress_bar, optimizer, unet.parameters())

    # Create the pipeline using using the trained modules and save it.
    if accelerator.is_main_process:
        pipeline = StableDiffusionPipeline.from_pretrained(
            args.pretrained_model_name_or_path,
            unet=accelerator.unwrap_model(unet),
            use_auth_token=True
        )
        pipeline.save_pretrained(args.output_dir)

        if args.push_to_hub:
            repo.push_to_hub(commit_message="End of training", blocking=False, auto_lfs_prune=True)

    accelerator.end_training()


if __name__ == "__main__":
    main()